Draft Superannuation Alternative Default Models

Productivity Commission
Draft Report
Superannuation:
Alternative Default Models
March 2017
This is a draft report
prepared for further public
consultation and input.
The Commission will finalise
its report after these processes
have taken place.

 Commonwealth of Australia 2017
Except for the Commonwealth Coat of Arms and content supplied by third parties, this copyright work is
licensed under a Creative Commons Attribution 3.0 Australia licence. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/3.0/au. In essence, you are free to copy, communicate and adapt the
work, as long as you attribute the work to the Productivity Commission (but not in any way that suggests the
Commission endorses you or your use) and abide by the other licence terms.
Use of the Commonwealth Coat of Arms
For terms of use of the Coat of Arms visit the ‘It’s an Honour’ website: http://www.itsanhonour.gov.au
Third party copyright
Wherever a third party holds copyright in this material, the copyright remains with that party. Their
permission may be required to use the material, please contact them directly.
Attribution
This work should be attributed as follows, Source: Productivity Commission, Superannuation: Alternative
Default Models, Draft Report.
If you have adapted, modified or transformed this work in anyway, please use the following, Source: based
on Productivity Commission data, Superannuation: Alternative Default Models, Draft Report.
An appropriate reference for this publication is:
Productivity Commission 2017, Superannuation: Alternative Default Models, Draft Report, Canberra.
Publications enquiries
Media and Publications, phone: (03) 9653 2244 or email: maps@pc.gov.au
The Productivity Commission
The Productivity Commission is the Australian Government’s independent research and
advisory body on a range of economic, social and environmental issues affecting the welfare of
Australians. Its role, expressed most simply, is to help governments make better policies, in the
long term interest of the Australian community.
The Commission’s independence is underpinned by an Act of Parliament. Its processes and
outputs are open to public scrutiny and are driven by concern for the wellbeing of the
community as a whole.
Further information on the Productivity Commission can be obtained from the Commission’s
website (www.pc.gov.au).

OPPORTUNITY FOR FURTHER COMMENT
DRAFT REPORT
iii
Opportunity for further comment
You are invited to examine this draft report and comment on it by written submission to
the Productivity Commission by Friday 28 April 2017. Further information on how to
provide a submission is included on the inquiry website at www.pc.gov.au/inquiries/
current/superannuation/make-submission
The final report will be prepared after further submissions have been received and public
hearings have been held, and will be forwarded to the Australian Government by
17 August 2017.
Public hearing dates and venues
Location Date Venue
Melbourne 3 May 2017 Productivity Commission
Level 12, 530 Collins Street
Sydney 8 May 2017 The Grace Hotel
77 York Street
Commissioners
For the purposes of this inquiry and draft report, in accordance with section 40 of the
Productivity Commission Act 1998 the powers of the Productivity Commission have been
exercised by:
Peter Harris
Karen Chester
Chair
Deputy Chair

iv SUPERANNUATION: ALTERNATIVE DEFAULT MODELS
DRAFT REPORT
Terms of reference
EFFICIENCY AND COMPETITIVENESS OF THE SUPERANNUATION SYSTEM
I, Scott Morrison, Treasurer, pursuant to Parts 2, 3 and 4 of the Productivity Commission
Act 1998, hereby request that the Productivity Commission conduct: a study to develop
criteria to assess the efficiency and competitiveness of the superannuation system; and an
inquiry to develop alternative models for a formal competitive process for allocating
default fund members to products.
Background
An efficient superannuation system is critical to help Australia meet the economic and
fiscal challenges of an ageing population. The superannuation system has accumulated
over $2 trillion in assets. Given the system’s size and growth, the system is of central
importance to funding the economy and delivering retirement incomes.
MySuper has been a strong step in the right direction but more needs to be done to reduce
fees and improve after-fee returns for fund members. The Financial System Inquiry noted
that fees have not fallen by as much as would be expected given the substantial increase in
the scale of the superannuation system, a major reason for this being the absence of
consumer-driven competition, particularly in the default fund market.
These Terms of Reference follow from the Government’s response to Financial System
Inquiry Recommendation 10 on efficiency in superannuation. The Government committed
to tasking the Productivity Commission to develop and release criteria to assess the
efficiency and competitiveness of the superannuation system, including the choice and
default markets and to develop alternative models for allocating default fund members to
products.
This work will inform a review of the efficiency and competitiveness of the
superannuation system, which the Productivity Commission will be asked to undertake
following the full implementation of the MySuper reforms (after 1 July 2017).
Process
The Productivity Commission is to develop criteria to assess the efficiency and
competitiveness of the superannuation system and release the criteria within nine months
of receiving these Terms of Reference. The release of these criteria is intended to provide

TERMS OF REFERENCE
DRAFT REPORT
v
transparency and certainty to the superannuation industry about how it will be assessed
ahead of the full implementation of MySuper.
The Productivity Commission is to develop alternative models for a formal competitive
process for allocating default fund members to products. In developing alternative models,
the Productivity Commission should be informed by the criteria it develops to assess the
efficiency and competitiveness of the superannuation system. The Productivity
Commission should report on alternative models within 18 months of receiving these
Terms of Reference.
For both elements, the Productivity Commission should consult widely and undertake
appropriate public consultation processes, including inviting public submissions and
conducting industry roundtables. The Productivity Commission is to provide both draft and
final reports and the reports will be published.
Scope of study: development of criteria to assess efficiency of super system
The Productivity Commission should develop criteria to assess whether and the extent to
which the superannuation system is efficient and competitive and delivers the best
outcomes for members and retirees, including optimising risk-adjusted after fee returns. In
determining the criteria to assess the efficiency and competitiveness of the superannuation
system, the Productivity Commission may have regard to:
• operational efficiency, where products and services are delivered in a way that
minimises costs and maximises value, which can be enhanced by competition and
innovation from new entrants and incumbents;
• allocative efficiency, where the system allocates resources to the most productive use
and optimally allocates risks;
• dynamic efficiency, including services to members, where the system induces the
optimal balance between consumption and saving over time; and
• the extent to which the system encourages optimal behaviour on the part of consumers,
including consideration of the learnings from behavioural finance.
The Productivity Commission should consider the nature of competition in the
superannuation industry, the effect of government policy and regulation on the
competitiveness and efficiency of the system and relevant international experience.
Scope of inquiry: development of alternative models
The Productivity Commission is to examine alternative models for a formal competitive
process for allocating default fund members in the superannuation system to products and
to develop a workable model, or models, that could be implemented by Government if a
new model for allocating default fund members to products is desirable. These model(s)
would provide viable alternatives for the Government’s consideration, depending on the

vi SUPERANNUATION: ALTERNATIVE DEFAULT MODELS
DRAFT REPORT
outcomes of the review of the efficiency and competitiveness of the superannuation
system, which the Productivity Commission will be asked to undertake following the full
implementation of the MySuper reforms.
The developed model(s) should enhance efficiency in the superannuation system in order
to improve retirement incomes, including through optimising long-term net returns to
members, and build trust and confidence in funds regulated by the Australian Prudential
Regulation Authority (APRA). The models developed should consider default fund
selection across the superannuation system as a whole.
The Productivity Commission may consider auction, tender and other types of competitive
processes. The Productivity Commission should consider the merits of different
approaches, the metrics for conducting them and their frequency. This should include
consideration of:
• the strengths and weaknesses of competitive processes used internationally, such as
Chile, New Zealand and Sweden, as well as those used in large corporate tenders by the
Northern Territory Government and in other jurisdictions;
• the costs and benefits of different mechanisms, including:
– optimising long-term after fee returns;
– the administrative, fiscal, individual and complexity costs;
• and in examining different processes, consider:
– the robustness of the process, including against gaming and collusion;
– whether the structure achieves efficient outcomes and facilitates ongoing innovation
over the long run;
– the effect on system stability and market concentration;
– who should run the process; and
– the extent to which the process promotes the interests of consumers.
• regulatory impediments to optimal competition under the preferred model(s).
Principles for designing a model for a competitive process should include:
• Best interests: ensure incentive compatibility with meeting the best interests of
members, encourage long-term investing, and encourage a focus on expected after-fee
returns based on asset allocation and investment strategy.
• Competition: drive pressure on funds to be innovative and efficient, diversify asset
allocation and optimise long-term after-fee returns by rewarding best performers.
Facilitate new superannuation fund entrants to the market.
• Feasibility: ensure the process is low-cost and easy to administer and minimises
regulatory costs on industry, including business and employers.

TERMS OF REFERENCE
DRAFT REPORT
vii
• Credibility and transparency: make relevant information public; avoid room for
gaming the process; and ensure metrics are clear, simple, difficult to dispute and
difficult to manipulate.
• Regular assessment and accountability: regularly conduct a repeat process that
requires default funds to earn their right to receive new default members, and ensure
funds are accountable for the outcomes they deliver members.
• Fiscal implications: the extent to which the process can reduce reliance on the Age
Pension and/or give rise to other risks or costs to Government.
The Productivity Commission should draw on expertise in the field of competitive models.
S. MORRISON
Treasurer
[Received 17 February 2016]

Draft Superannuation Alternative Default Models

CONTENTS
DRAFT REPORT
ix
Contents
Opportunity for further comment iii
Terms of reference iv
Abbreviations xi
Overview 1
Draft findings and recommendations 19
Information requests 25
1 Choice and competition in superannuation 29
1.1 What is this inquiry about? 30
1.2 The role and evolution of defaults in Australia 32
1.3 How are competition and choice relevant to the default
segment of the system? 39
1.4 A guide to the rest of the report 47
2 The Commission’s assessment framework 49
2.1 The objective of this inquiry 50
2.2 The model assessment criteria 50
2.3 The no defaults baseline is not a model 59
3 Approach to developing alternative models 61
3.1 Specifying alternative models for evaluation 62
3.2 Which employees will the models cover? 67
3.3 Characteristics of the default product 76
3.4 Other common foundations for the models 82
Annex: Estimating the effects of alternative definitions of new
default members 91
4 The baseline 95
4.1 Establishing the baseline 96
4.2 Assessing the baseline 99

x SUPERANNUATION: ALTERNATIVE DEFAULT MODELS
DRAFT REPORT
5 Assisted employee choice 115
5.1 A shortlist of products 117
5.2 Other employee protections 127
5.3 Assessment against the baseline 132
6 Fee-based auction 139
6.1 Elements of auction design 142
6.2 The auction model considered for assessment 151
6.3 Assessment of the model 155
6.4 Overall assessment 161
7 Multi-criteria tender 165
7.1 Elements of tender design 167
7.2 The tender model considered for assessment 174
7.3 Assessment of the model 179
7.4 Overall assessment 186
8 Assisted employer choice (with employee protections) 189
8.1 Why the Commission chose this model for consideration 190
8.2 How would administrative filters operate? 193
8.3 Institutional arrangements 208
8.4 Evaluation of the model 212
A Public consultation 219
B Select default models overseas 223
B.1 Chile 223
B.2 New Zealand 228
B.3 United Kingdom 235
B.4 Sweden 239
References 243

ABBREVIATIONS
DRAFT REPORT
xi
Abbreviations
AFP Administradoras de Fondos de Pensiones (Chile)
APRA Australian Prudential Regulation Authority
ASIC Australian Securities and Investments Commission
ATO Australian Taxation Office
BOOT Better off overall test
CPI Consumer price index
FSI Financial System Inquiry
FWC Fair Work Commission
NEST National Employment Savings Trust (United Kingdom)
PC Productivity Commission
RSE Registrable superannuation entity
SG Superannuation Guarantee
SG Act Superannuation Guarantee (Administration) Act 1992 (Cwlth)
SIS Act Superannuation Industry (Supervision) Act 1993 (Cwlth)
SMSF Self-managed superannuation fund
SPA Swedish Pensions Authority
TPD Total and permanent disability

2 SUPERANNUATION: ALTERNATIVE DEFAULT MODELS
DRAFT REPORT
Key points
• Superannuation has evolved much since compulsory superannuation was first introduced a
quarter of a century ago. Today’s default arrangements evolved historically within the
workplace relations system, and provide a safety net for employees that don’t or cannot
make a decision in a world of compulsion.
• So after 25 years, this Inquiry is a timely opportunity to look at potential ways to introduce
more competition into a system that benefits from a large flow of mandated superannuation
contributions, and much of that from disengaged members.
• Complementary policy action (including to extend genuine member choice to all employees)
is needed to deliver the full potential benefits of member-driven competition under the
models we identify and even under current default arrangements. The freedom to make
choices is necessary to realise the benefits of competition.
• Two-thirds of members stick with their default fund. If the system is going to rely on defaults,
it needs to guide members to products that at a minimum seek to maximise long-term net
returns.
• In this draft report the Commission has developed four alternative models that are likely to
outperform a baseline of ‘no defaults’ on member benefits and competition.
− They try to address the core problem in superannuation — the sheer complexity of
decision making coupled with compulsion — by increasing the availability and quality of
information or nudging choice to a smaller set of high-quality products.
− But these interventions come at a cost, and each model has different relative strengths
and weaknesses.
• Members who do not exercise choice should be allocated to a default product only once.
The current system’s propensity to create multiple accounts is an egregious systemic failure.
It warrants more than the incremental remediation to date.
− This approach would result in a smaller pool of employees being defaulted each year,
but it should be sufficient to generate competitive dynamics.
• There should be a government-run centralised online information service, with universal
participation by employees and employers, to facilitate more efficient allocation of default
members to products.
• A centralised clearing house (akin to New Zealand’s), while a more ambitious undertaking,
would have wider and more enduring member benefits.
• Member outcomes would also benefit from more transparent disclosure by funds regarding
merger considerations, to hasten the exit of underperforming funds.
• Certain reforms indicated in this report will be examined further in the stage 3 review.

OVERVIEW
DRAFT REPORT
3
Overview
In Australia’s superannuation system, most risks — in relation to the level of contributions,
investment returns, fees and ultimately the benefit in retirement — lie with the individual.
And yet the superannuation system hasn’t always afforded or encouraged individual
decision making. Some is well-intentioned paternalism, reflecting the compulsory and
complex nature of superannuation, while other arrangements are merely a historical
overhang. Both elements coexist in Australia’s arrangements for allocating default
superannuation members to products (‘default arrangements’).
In many ways default arrangements can be beneficial to members and are common around
the world in retirement savings systems. But they can also stifle competition and
innovation that would otherwise occur when consumers make active decisions, and
discourage individuals learning about a sizeable asset held on their behalf. Striking the
right balance between these benefits is imperative for a system that collectively compels
large mandated savings and affords individual responsibility (and with it risk), and an
economy where job mobility is the norm.
The Commission has been asked to develop alternative workable models for a formal
competitive process to allocate default superannuation members to products. These new
alternative models could be implemented by the Australian Government if deemed
desirable following the Commission’s future review of the efficiency and competitiveness
of the superannuation system — to be undertaken sometime after 1 July 2017.
Figure 1 Stage two in a three-stage superannuation review

DRAFT REPORT
This Inquiry represents the second of three related pieces of work on superannuation to be
undertaken by the Productivity Commission (figure 1). These stem from the Australian
Government’s response to the recommendations of the 2014 Financial System Inquiry
(FSI). The FSI found that the superannuation system was not operationally efficient due to
a lack of price-based competition in the superannuation default market.
Importantly, the Commission’s task in this Inquiry (stage 2) is to develop new competitive
models for future consideration by the Australian Government. The Commission is not
tasked in this Inquiry to form a view on whether alternative models are better or worse than
the current default arrangements, nor on the merits of the current default arrangements.
This is core to the overall competitiveness and efficiency of the superannuation system and
will therefore be examined as part of the stage 3 review of the competitiveness and
efficiency of the superannuation system.
The role of competition in superannuation
Competition can promote better outcomes for consumers, such as lower prices, improved
service quality and product innovation. There are several reasons why these competitive
ideals may not be realised in the superannuation market.
First and foremost, superannuation in Australia is based on compulsion, which can limit
the demand-side pressure exerted by members. Member passivity and disengagement are
further compounded by the complexity of retirement saving decisions, long time horizons,
various behavioural and cognitive biases, and the costs of active involvement. At their
core, these are information problems — incomplete information or asymmetric
information — exacerbated by cognitive constraints.
The absence of strong member engagement can dull competitive pressure on
superannuation funds, regardless of the presence of many funds in the market. Moreover,
the information problems can create a risk of unhealthy and wasteful competition, such as
excessive advertising and product proliferation.
Policy intervention can potentially improve on these outcomes by helping members make
better choices, or via default selection processes reducing risk as members opt out of
choice. Policymakers can also intervene on the supply side to steer competition and trustee
behaviour to more beneficial aspects.
Default arrangements in Australia
Defaults have been integral to the development of Australia’s superannuation system,
largely stemming from the decision to make superannuation compulsory and the inherent
complexity for individuals in making decisions about retirement saving. Defaults also
potentially reduce the cost and complexity for employers of complying with their

OVERVIEW
DRAFT REPORT
5
Superannuation Guarantee (SG) obligations. Most private pension systems around the
world rely on defaults to some extent.
Box 1 Current default arrangements
Under the Superannuation Guarantee (Administration) Act 1992 (Cwlth), employers are
required to make contributions to their employee’s superannuation fund (with limited
exemptions). Most employees have the opportunity to choose which fund these contributions
are paid into. However, for those employees that do not make an active choice, employers are
required to nominate a fund — which is thus the default fund. The rules surrounding this
process vary significantly across industrial awards, agreements, contracts and other
employment arrangements. Once a default fund has been selected, fund trustees are required
to place default contributions into that fund’s default product, which in most cases must be a
MySuper product.
These arrangements are the result of historical workplace arrangements and the introduction of
MySuper products as part of the Stronger Super reforms in 2013. In summary, there are up to
four ‘third parties’ involved in the selection of default products for members that fail to make an
active choice — the regulator (APRA), an expert panel within the Fair Work Commission
(FWC), the full bench of the FWC and, finally, employers. These third parties have differing
objectives and accountabilities. Employers in particular report disparate appetite and capability
in selecting default funds for their employees.
The issue of default arrangements is highly contentious and politically sensitive. While
defaults (MySuper products) only make up about a quarter of total superannuation assets
(figure 2), they are more significant in terms of their share of members (putting a number
on this is difficult due to systemic data gaps). Survey evidence suggests about two thirds of
superannuation fund members stick with their default fund (and fewer than 5 per cent of
members switch once they are in a fund). Opening up access to these members and their
contributions is a lucrative proposition for new entrants and a competitive threat for
incumbents (especially the relative poor performers).
MySuper authorisation
• License RSEs
• Standard features
• Enhanced trustee
obligations
• Annual reporting on
fees, investment
strategy & scale
Product choice
• Prescribed or limited
choice in award or
agreement
• Full choice, select
from any MySuper
product
APRA EmployersFWC
(Process not yet finalised)
Expert Panel
• Select eligible
MySuper products
and compile the
Default List
• Assess applications
based on factors set
out in legislation
Full Bench
• Review default fund
terms of awards
• Specify (typically)
2-15 products in
awards
• 4-yearly review
In some cases employers can select from any MySuper product

DRAFT REPORT
Figure 2 Size of the superannuation market, June 2016
Other includes ‘balance of life office statutory funds’ and small APRA funds.
Some inquiry participants argued that the current default arrangements have served
members well by placing them in funds with a culture of serving others and operating in
the best interests of members, not shareholders or other related parties. In support of this
view, many of these same participants highlighted empirical evidence of superior returns in
not-for-profit funds relative to for-profit funds, and for default funds named in modern
awards in particular. And while this is historically true on average, it does not hold for all
industry funds.
Other participants, including retail funds, suggested that supply-side barriers (particularly
in industrial awards and agreements) ultimately constrain competition and innovation
which could benefit members. Moreover, they highlighted that even within the population
of current ‘no frills’ MySuper products there is significant diversity in investment
performance.
But it is unwise, as many have to date, to portray this Inquiry as just another ‘industry fund
versus retail fund’ debate. This Inquiry is much more than that — it is a wake-up call to the
entire industry, which some claim has become complacent with a steady flow of mandated
contributions from disengaged members, and as an industry has failed to improve its scale
and efficiency and deliver better outcomes for members (despite the MySuper reforms).
From this perspective, the Inquiry has managed to unite the superannuation industry
against the Inquiry’s potential contemplation of more-than-incremental reform. A healthy
For profit
$55b
SMSF
$622b
Not for profit
$419b
MySuper
$0.5t
Choice
$1.6t
For profit
$491b
Not for profit
(& other)
$513b

OVERVIEW
DRAFT REPORT
7
dose of scepticism would suggest that there must be rents to be recovered for the benefit of
members for such unanimity to be valid.
Most participants from the superannuation industry argued that some form of default
arrangements is needed for the foreseeable future (if not into perpetuity). Many
participants voiced a familiar objection to the view that competition could promote
member interests, instead placing faith in the goodwill and legal obligations of trustees.
This Inquiry will test that proposition more fully.
How we developed the alternative models
Competition that merely sought to increase the already daunting number of choices (more
than 110 MySuper products alone) would almost certainly do nothing to assist the
potentially disengaged member, default or otherwise. Such is not, however, the paradigm
for this Inquiry. The aim, as set by the Terms of Reference, is to determine if and how
competition might make choice simpler or to lower the risk of not exercising choice.
There are various ways that a competition lens can inform the design of alternative models
for allocating default members. Broadly, competition can be promoted in one of two ways:
• competition in the market — competition based on conventional demand- and
supply-side interactions between employees (members) and trustees (funds)
• competition for the market — competition between funds via some form of formal
process to earn the right to gain access to default members.
Promoting competition in the market needs consumers to make choices. While most
employees already have the opportunity to choose which superannuation product their
contributions are paid into, some employees are restricted from doing so (previously
estimated to be in the order of 20 per cent of members). Freedom to make choices —
which includes the freedom to make mistakes — is a necessary condition to realise the
benefits of competition in the superannuation market. It is therefore essential to extend
genuine member choice to all employees.
Promoting competition for the market may be desirable where it is not feasible to promote
effective competition in the market, such as where member engagement is low or there are
barriers to entry and exit. Typical ways to promote competition for a market include a
tender or auction process.
The Commission has taken a broad interpretation of what constitutes a formal competitive
process for the default market. It is any new alternative model that permits open
participation, encourages rivalry between funds to the benefit of members, and involves
products being selected for members based on merit.

DRAFT REPORT
There are three key steps to developing alternative competitive models:
1. specifying the market (members and products) to which the model applies
2. determining the type of selection mechanism for default products
3. determining the decision-making process for allocating members to those products.
The first step — what is the default market — is consistent across models and is discussed
further below. The four models presented in this draft report differ on the second and third
steps. The product selection mechanisms (step 2) are either administrative (such as a
product filter based on certain standards) or market based (such as an auction). The
distinction between administrative and market mechanisms is not necessarily stark: a
‘heavier’ filter, which includes relative performance standards, is very similar to a
multi-criteria tender. The choice of the selection mechanism ultimately depends on the
degree of filtering and who will be allocating members to products (step 3). The
Commission has chosen the combinations that work well together.
Some models have not been considered in this Inquiry because they are not consistent with
the competitive focus in the Terms of Reference — for example, a single sovereign fund
that receives default contributions by fiat.
What is the market at stake?
The default member
Currently, an employee typically acquires a new default superannuation account every time
they commence a new job and do not actively choose a fund themselves (the turnover
pool). This creates a significant proliferation of accounts, with over 40 per cent of
members holding more than one account.
This Inquiry is taking an alternative approach as suggested by the FSI in 2014: employees
who fail to exercise choice1 will be allocated to a default product only if they do not have
an existing superannuation account. These would typically be new entrants to the
workforce (the first-timer pool), and they will retain that account (including through a
change in employer) unless they actively switch. Other employees who commence a new
job would see their SG contributions flow to one of their existing superannuation accounts,
or a new account of their choice.
There would be about 400 000 new default members each year under the Commission’s
first-timer definition, with about $800 million in annual contributions. This is evidently
much lower than the contributions in the turnover pool. This pool, while being hard to
1 None of the models presented in this report mandate that employees join or remain with a particular
default product. Employees would have the right to choose their own fund and product, and to switch
funds or products after joining.

OVERVIEW
DRAFT REPORT
9
estimate because of systemic data gaps, would have more members (in the short term) and
higher average incomes.
Nonetheless, the first-timer pool is not insignificant, particularly in the medium term. A
key reason for its initial modest size is that the average income of first-timer members is
about $21 000 (which reflects the part-time or casual nature of many first jobs).
Contributions from each year’s cohort would grow at a faster rate than for people in the
turnover pool as they converge towards average incomes; applying average incomes to this
cohort yields a pool of just over $2 billion. These younger members would likely remain in
their original product for a longer period, and thereby offer successful funds the benefits of
improved inflow–outflow ratios and thereby readier liquidity management.
In this draft report, the Commission is keen for participant feedback on the proposal to
restrict the default models to people who do not already have a superannuation account,
and what this implies for the competitive dynamics of any alternative model. For example,
which funds would not compete for this default pool, and why not? The impact on funds
heavily dependent on constant inflows of default members might at first glance appear
self-evident but in the Commission’s thinking this might be wrong. The presence or
absence of complementary reforms to barriers affecting choice appears likely to make
some difference.
The key benefit of restricting the new default allocation models to the first-timer pool is
the immediate ‘circuit breaker’ of reducing wasteful account proliferation, which erodes
member balances via multiple sets of fees and insurance policies. The FSI found that
addressing account proliferation and lost accounts could increase superannuation balances
at retirement by around $25 000. Multiple accounts are especially problematic among
younger members, for whom it is less likely to be a deliberate outcome. This erosion is
further exacerbated where young members have inappropriate insurance policies with their
superannuation funds. Overall, account proliferation is one of the superannuation system’s
worst systemic failings and warrants more than the incremental remediation to date.
This approach is contingent on universal use of a centralised online information service to
enable employers to identify which employees do not already have an existing account and
to facilitate members to consolidate their accounts. At a system level, this service will
generate information about employee decisions that will fill important evidence and data
gaps for future policy decisions. The ATO is currently developing platforms that might be
adaptable to the task.
A centralised clearing house2 would bring additional member (and efficiency) benefits to
the superannuation system, as evident in New Zealand with the role played by the Inland
Revenue Department. These benefits include administrative simplicity for employers and
funds, which makes it more attractive to new entrants, and easier oversight of employer
2 Whereby employers would remit all superannuation contributions directly to the clearing house, with the
clearing house then transferring those contributions to the relevant superannuation funds.

DRAFT REPORT
compliance with SG contributions. A clearing house will be explored further in the stage 3
review.
Implications for existing default members
While this Inquiry’s focus is on how to allocate new default members, any new allocation
process poses the question of how to treat the large stock of existing default members. How
these existing members are treated has implications for competitive dynamics, as well as
broader system efficiency and fairness.
A fund whose product ‘wins’ default status for contributions of new default members
would have to extend the same fees and service terms to its existing members in the default
product. This would be simple and fair — it would avoid a situation where funds have
multiple tiers of default members, each in a different product or fee structure — and was
broadly supported by inquiry participants. It would also spread the benefits of competition
to a wider cohort of default members. But the models in this report do not envisage moving
the entire system’s pre-existing default members to the new default product. This avoids
material upheaval, with the potential for near-term instability, in the superannuation
system.
The default product
The bundle of superannuation services (the default product) to which default members
would be allocated is also common to all models, and pertains to the accumulation stage
only. The default product should be simple, providing the minimum services necessary to
meet the Government’s stated objective for superannuation — to provide income in
retirement to substitute or supplement the Age Pension.
These would include investment, administration of member accounts and intrafund advice.
The administrative services and intrafund advice should help members understand the
superannuation system and be informed about the performance of their investment product,
the fees they are paying, the intrafund switching options and the process for exiting the
fund. Importantly, the quality or range of ancillary services per se should not be the main
driving factor in selecting default products under any of the allocation models. Funds
should primarily compete on long-term net returns and fees for a threshold level of service.
The alternative models will be evaluated on the quality of their default superannuation
product only — and not on insurance, which is currently bundled with default
superannuation on an opt-out basis. In this environment, insurance is best dealt with as a
regulatory add-on. The Commission will assume that the trustee chooses to bundle the
default superannuation product with insurance, subject to meeting their fiduciary and
legislative obligations, but the models do not involve comparing the quality or price of
bundled insurance.

OVERVIEW
DRAFT REPORT
11
The Commission has noted the various tensions of bundling insurance with superannuation
in its stage 1 report, and it should also be self-evident to trustees that the Government’s
stated objective for the superannuation system does not envisage insurance as an essential
element of the system.
The assessment framework
Each model in this draft report has been assessed against five criteria and the assessment is
mostly qualitative, given the developmental nature of this Inquiry. The assessment
framework was informed by the Terms of Reference and the Commission’s overarching
objective to promote the wellbeing of the Australian community as a whole.
• Member benefits: Does the model create incentives for funds to maximise long-term
net returns and allocate members to products that meet their needs?
• Competition: Does the model encourage open participation (contestability) and rivalry
between funds (competition) to drive innovation, cost reductions and more efficient
long-term outcomes for members?
• Integrity: Does the model promote a high-degree of integrity in the selection and
delivery of default superannuation products, and the ongoing behaviour of
superannuation funds more broadly?
• Stability: Is the model likely to create instability in the superannuation system that
leads to significant systemic risks?
• System-wide costs: Does the model minimise overall system-wide costs, taking into
account costs on members, employers, funds and governments?
There was broad agreement among inquiry participants that members’ best interests should
be paramount in the assessment. In many (if not most) cases, a model that promotes
members’ best interests could also be expected to promote the wellbeing of the Australian
community.
The assessment of alternative models requires a common reference point or baseline for
comparison. The Commission is using the baseline of no default arrangements. This
proposal elicited a substantial amount of participant feedback, which fell into two broad
categories: participants who wanted the models compared to current default arrangements,
and participants who interpreted the no-defaults baseline as an alternative model. Neither
interpretation is correct.
The no-defaults baseline is not an alternative model in its own right. Rather, the
Commission considers this to be as close as is possible to an objective starting point from
which to identify the problems that occur in a no-defaults world (the raison d’être of a
default system is to address these).

DRAFT REPORT
It is also the baseline that can best expose the relative abilities of alternative models to
address the problems of the default cohort today, rather than past problems and perception,
some of which may no longer be relevant. The alternative models need to beat the baseline
against the assessment criteria to be considered viable, then prove workable.
What does the baseline look like?
The baseline is a world of unassisted active choice. In the baseline, employers must make
SG contributions on behalf of employees. Superannuation funds and products would be
subject to prudential, competition and consumer laws and regulations. Employees would
be required to choose a superannuation product to receive their contributions. There would
be unrestricted choice, but with regulations — employees can choose from any publicly
offered legally complying superannuation product available in the market (including a
self-managed fund).
Employees would be responsible for researching and evaluating available products
themselves, albeit with the potential to seek support from third parties of their own
initiative — such as employers, unions, financial advisers and ratings agencies. Employers,
unions and other intermediaries would also be free to bargain with superannuation funds
on behalf of employees to secure group discounts. This would mimic the current practice
offered by corporate funds or industry funds, the only difference being that employees
would need to actively choose that fund.
… And how does it perform against the assessment criteria?
Some employees would struggle to make good long-term decisions in this unassisted active
choice world. In particular, they may choose products with high fees or poor investment
performance, or pay for additional services (such as a plethora of investment options) they
may not need. Further, some employees may choose overly conservative products,
reducing their potential retirement income. Importantly, and unlike poor consumer choices
in other markets, these costs extend beyond the employee to also be borne by future
taxpayers, to the extent that they result in higher Age Pension reliance.
In contrast, a significant benefit of the baseline — which is less evident under default
arrangements — is that requiring employees to interact with their superannuation and make
an active choice can potentially drive member engagement and ultimately exert
competitive forces.
The baseline is considered the purest form of competition, albeit operating in a highly
regulated market. The supply side of the market is likely to be highly contestable, with low
barriers to entry and exit, and low levels of market concentration. Rivalry between funds is
expected to drive innovation over time. However, there is also scope for unhealthy and
wasteful competition in the form of superfluous product differentiation and inducements,
and excessive advertising.

OVERVIEW
DRAFT REPORT
13
An unassisted active choice world minimises the number of principal–agent relationships
that can lead members astray. That said, potential for conflicts of interest still exists in an
unassisted active choice world, primarily where members seek advice from other parties.
The baseline is unlikely to have a destabilising effect on the superannuation system. There
would be no reason to expect sudden large-scale changes in the flow of members or assets
between funds, or excessive market concentration.
In an unassisted active choice world, members bear most of the costs of allocating
themselves to a relevant superannuation product (largely search costs, though some are
likely to put in more time and effort than others). While this may not be significant at the
individual level, it may be cumulatively significant across the entire system. Costs for
other system participants (for example, employers) are expected to be low.
The Commission is undertaking an experimental survey of superannuation decision
making which will further add to this evidence base. The results and analysis will be made
publicly available (upon completion) and will also be incorporated in the Commission’s
final report.
The alternative models in brief
As noted earlier, information problems are at the core of the market failures in
superannuation. There are only two pathways to address information problems: increase
the availability and quality of information (at a cost) or limit choice to a smaller set of
better products. The Commission has developed four new alternative models that do this in
some way (figure 3). The key features of each model are noted briefly below, and
additional detail about each of the models can be found in their corresponding chapters.
• Assisted employee choice — this model leverages the competition benefits that arise
when members exercise choice, but with information and nudges to assist members to
make informed choices.
• Assisted employer choice (with employee protections) — this model injects
competition by giving employers choice in selecting a default product for their
employees, as long as the default product meets some minimum standards, while also
simplifying choices for small and medium sized businesses.
• A fee-based auction and a multi-criteria tender — these models incorporate a
market-based mechanism into the selection of default products, with sequential
allocation of members among winning products.

DRAFT REPORT
Figure 3 The building blocks of the allocation models
Model 1: Assisted employee choice
In this model, employees are required to choose a superannuation product themselves
(active choice), but are assisted in doing so by a set of policy interventions.
• The primary feature is a non-mandatory shortlist of carefully selected products that are
deemed to be ‘good’. It would nudge employees towards good products without going
as far as to impose one by default. The shortlist would consist of between 4 to 10
products, selected by a government body using a heavy administrative filter (see
model 2). The shortlist would be accompanied by simple information on key features of
each product in a consistent and comparable format.
• A system of product accreditation would encourage funds to provide products which
have a common, simple design that allows for easy comparisons across different
products, and thereby makes it easier for employees to choose a product that meets
their needs. Product accreditation would be voluntary for funds and be akin to a
strengthened MySuper authorisation process (with a stronger emphasis on minimum
performance standards). Products would have to be accredited to be considered for
shortlisting.
• A simple, low-cost last-resort fund (run by existing eligible rollover funds or possibly
the Future Fund) would hold contributions from employees who fail to exercise choice
and encourage them to choose their own product.
Model 2: Assisted employer choice (with employee protections)
In this model, employers choose a default product for their employees who do not exercise
choice. In doing so, employers must select a product from one of two lists. The model
recognises that some (mainly larger) employers are well placed to choose a default product
Degree of
filtering
Type of
filtering
None High Very high
Who
allocates?
Low
Baseline
1 Assisted
employee
choice
2 Assisted
employer choice
(with employee
protections)
4 Fee-based
auction
3 Multi-criteria
tender
None AuctionHeavy filter
Employer GovernmentEmployee
TenderMinimum standards

OVERVIEW
DRAFT REPORT
15
and negotiate favourable arrangements for their employees, while many (mainly small and
medium size) employers are not well placed to select a default product.
• Employers must choose a product that meets mandatory minimum standards: a light
filter. Mandatory minimum standards are important to protect member interests, given
the principal–agent issues inherent in the employer choice environment. The light filter
would be akin to a strengthened MySuper authorisation process (with a stronger
emphasis on minimum performance standards).
• A heavy filter would apply stricter criteria around investment performance and other
product features (similar to the tender criteria below) to identify a narrower ‘preferred
default’ list of the best performing products. The preferred default list would be
voluntary for employers, and is designed to assist employers who are not well placed to
select appropriate default products for their employees.
Model 3: Multi-criteria tender
In this model, participants compete for rights to a share of the default pool by making
proposals against a number of different assessment criteria, which are weighted by their
relative importance. A well designed multi-criteria tender could therefore engender a
product that performs well on multiple characteristics that meet members’ needs. The key
features of this model are:
• a pre-qualification stage to ensure participating funds meet minimum criteria relating
to governance and transparency
• the tendering stage, where participants submit tenders against the following criteria:
– past performance on net returns and member satisfaction
– investment strategy
– the quality of member services, engagement and intrafund advice
– fee levels and transparency
– innovation in unspecified areas
• a comparative evaluation stage — tenders are evaluated against the criteria and
assigned an index score, with up to 10 of the overall best performers being deemed as
winners of the tender. A best-and-final-offer stage could be used to encourage
convergence where proposals are close, in the interest of members having near
equivalence in standards despite being sequentially allocated to a product
• a performance monitoring and enforcement stage, with winning providers
self-reporting against the commitments made in their winning offers, and a credible
threat of losing default provider status in the absence of immediate remediation of a
contractual breach.

DRAFT REPORT
Box 2 Who would run the selection process?
All of the models require a selecting body. Because of the competitive assessment nature of the
exercise, this body should not be a regulator like APRA or ASIC. The selecting body (or panel)
would be established specifically for the selection task each round, and then cease operations
— it would have no ongoing role. The selecting body would be overseen by and accountable to
the Australian Government. While this places some of the responsibility for the outcomes on the
government and so may increase perceived moral responsibility by the government for later
poor performance, this is necessary to ensure the integrity and credibility of the process.
Government administration of the process should not preclude private-sector participation in an
advisory sense.
There are a number of key principles that are relevant to the composition and conduct of this
selection body:
• an obligation to act in the best interests of fund members
• an understanding of the needs of default members and how these are best met
• sufficient expertise and knowledge of superannuation and finance
• transparent processes
• procedural fairness (that is, enabling all relevant parties to participate equally)
• being dispassionate and free of conflicts of interest
• being accountable for its decisions and having credibility with members and funds.
The product selection process should be repeated every four years in the early stages of the
operation of the allocation model. The frequency can be reduced in the future.
Model 4: Fee-based auction
This model has products competing for default status by out-bidding each other on member
fees. A fee-based auction is worthy of consideration for two reasons. First, there is
evidence that variation in member fees explains a significant amount of variation in net
returns. Second, realising economies of scale would allow the superannuation system to
achieve lower member fees without diluting net returns. While both points may be
contested, they are not without merit.
• A pre-qualification stage is used to ensure participants meet certain minimum
standards. Fund-level minimum standards relate to the fund’s integrity and capacity to
handle expected member inflows. Product-level standards relate to a certain level of
member and investment services.
• The auction’s bidding metric is a composite fee that accounts for administration and
investment fees, as these make up the bulk of member charges. The bidding format is
a first-price sealed-bid.
• The auction would be able to produce multiple winners (but also a single winner). The
advantage of multiple winners is the wider spread of the lowered fee across the market
(noting the requirement cited earlier that winning funds must extend the new fee to

OVERVIEW
DRAFT REPORT
17
their existing members of the default product). This would be done through a carefully
managed best-and-final-offer mechanism, whereby a select number of losing funds may
have an opportunity to match the fee and service offering of the winner.
• Post-auction monitoring and enforcement mechanisms will be needed to keep the
winning fund(s) to the commitments made in the auction process, including a credible
threat of losing default provider status in the absence of immediate remediation of a
contractual breach.
Assessing and comparing the models
This section compares how the four alternative models compare to the baseline of
unassisted active choice. Unsurprisingly, all of the alternative models outperform the
baseline on member benefits, because they guide more members to better performing
products.
• Models that involve a multi-criteria tender or heavy filter perform relatively better, as
their multifaceted selection framework will more comprehensively reflect member
needs.
• There is a risk under the fee-based auction that winning funds will eschew higher-cost
illiquid assets, which have been credited with providing superior diversification and net
returns in the past to some funds. Nonetheless, putting young disengaged members into
a low-cost passive fund may be a reasonable choice, which many overseas default
models apply (albeit with many combining passive with conservative investment
strategies).
• The models with defaults could deter some employees (with moderate levels of
engagement and literacy) from actively researching products themselves. The higher
level of member engagement in the baseline is better maintained in the assisted
employee choice model.
All of the models also outperform the baseline on competition, because they focus
competition on aspects of member value.
• The models that centralise allocation decisions in the hands of government (auction,
tender and heavy filter) have high levels of filtering and therefore focus competition for
the market on the factors that matter for members, at least in the short to medium term.
• The models with employee or employer choice have high competition in the market
(due to a large number of both buyers and sellers). Any models that have high
competition in the market will invariably have high marketing costs and some scope for
unhealthy and wasteful competition (but less than the baseline). But this type of
competition is healthier in the long term, because it is better able to innovate and
anticipate shifts in consumer needs.

DRAFT REPORT
All of the models bring new parties into the decision-making process compared to the
baseline, some more so than others. This will unavoidably introduce conflicts of interests
undermining integrity.
• The auction, tender and heavy filter models put a significant onus on government to get
the selection right. This invariably heightens the risk of getting it wrong — a lot rests
on the integrity of the selection body and process. There is a risk that the selection
process will be too conservative, or try to be all things to all people. And this is
apparent from international experience.
– These risks are probably smallest for the fee-based auction because of its relative
simplicity. The complexity of weighing multiple criteria may create a perception (or
reality) that the preferred list panel moves to a lowest common denominator and
fails to objectively and robustly evaluate products.
• Additional integrity concerns arise in models with employer choice, particularly given
the implausibility of imposing a regulatory obligation upon employers to act in their
employees’ best interests.
• But the models can also promote integrity by reducing scope for exploitative behaviour
that can occur in the presence of compulsion and complexity. The additional consumer
protections and the selection processes can build trust and confidence in funds and the
system. Moreover, the accountability and transparency mechanisms embedded in each
model are likely to be more effective in holding funds to account than unassisted active
choice.
No model is expected to introduce excessive levels of market concentration or
unmanageable risks to the stability of the system.
System-wide costs under all of the models are either similar or slightly lower than the
baseline — a simple function of the smaller decision making cohort. All of the models
reduce search costs for members and, because the set of choices is narrowed under all
models, there are likely to be reductions in system-wide marketing costs. This will be
offset to some extent by additional costs associated with selection processes and consumer
protections.
• The models with mandatory minimum standards or product accreditation (assisted
employer choice (with employee protections) and assisted employee choice) have
relatively higher costs because these models require ongoing regulatory oversight of a
potentially large number of products.
• System-wide costs are likely to be lower for the auction and tender. The cost of a
quadrennial selection exercise is likely to be small, and would require self-reporting
and monitoring of only a small number of products.

DRAFT FINDINGS AND RECOMMENDATIONS
DRAFT REPORT
19
Draft findings and recommendations
DRAFT FINDING 1.1
Complementary policy action (including to extend genuine member choice to all
employees) will be needed to deliver the full potential of member-driven competition
under the alternative models developed in this Inquiry and also under current default
arrangements.
DRAFT FINDING 1.2
Where there is third-party involvement in the selection of a default product, there
needs to be effective regulation or arrangements in place to ensure these third parties
act in the best interests of members.
DRAFT FINDING 1.3
For the purposes of this Inquiry, a formal competitive process for allocating default
members constitutes any new alternative process that permits open participation
(contestability), encourages rivalry between funds (competition) to the benefit of
members, and involves products being selected for members based on merit.
DRAFT FINDING 3.1
The current arrangements where members can be defaulted to a new account on
every change of job increases the proliferation of accounts. This materially adds to
costs for members and reduces member balances at retirement.
DRAFT RECOMMENDATION 3.1
To avoid perpetuating the legacy problems of the current system, any future
alternative system for allocating members to default products should be premised on
employees being assigned a default product only once, when they join the workforce.

DRAFT REPORT
The Australian Government should establish a centralised online service for members,
employers and the Government that builds on existing functionality of myGov and
Single Touch Payroll. The service should:
• allow members to register online their choice to open, close or consolidate
accounts when they are submitting their Tax File Number on starting a new job
• facilitate the carryover of existing member accounts when members change jobs
• collect information about member choices (including on whether they are electing
to open a default account) for their employer and the Government.
There should be universal participation in this process by employees and employers.
DRAFT FINDING 3.2
There are strong grounds for requiring a fund that wins default status for contributions
of new default members to extend the same fees and service terms to all its existing
members of the default product.
DRAFT FINDING 3.3
The default product in all models will focus on the accumulation stage and include
investment, administration of member accounts and intrafund advice. It will be a
simple and low-cost (but not conservative) product aimed at the needs of those who
are disengaged from the system.
• A bundled insurance product will not be a factor in the selection of products and is
best addressed through regulation and regulator oversight.
• Funds will primarily compete on long-term net investment returns and costs,
subject to meeting a threshold quality of service, not on the quality or range of
ancillary services per se.
DRAFT FINDING 3.4
The desirable frequency for the selection process is between four and eight years,
with the greater frequency best used in the early period.

DRAFT REPORT
21
DRAFT FINDING 3.5
The selection of eligible default products should be administered by a government
body, and be subject to strong governance rules. The decision-making body must:
• have a strong focus on fund member interests
• have sufficient expertise to evaluate products
• be independent and free of real or perceived conflicts of interest
• have processes that are transparent and afford procedural fairness
• be accountable for its decisions.
The Australian Government should introduce a formal framework that specifies the
process and obligations of trustees when making or considering merger proposals. As
part of the framework, trustees would be required to disclose all merger attempts
involving their fund, as well as the reasons for any decisions.
DRAFT FINDING 5.1
Relative to the baseline, the assisted employee choice model would:
• significantly reduce the complexity employees face in choosing a product and lead
to more employees choosing high-performing products that meet their needs,
thereby increasing member benefits
• focus competition on product aspects of value to members, put downward
pressure on fees (through greater product comparability) and likely curtail wasteful
product proliferation
• better align funds’ interests with those of members
• support a stable superannuation system
• have lower search costs for many employees and very low costs for employers,
but with government and funds incurring additional costs associated with
regulatory structures.

DRAFT REPORT
DRAFT FINDING 6.1
Relative to the baseline, the fee-based auction model would:
• promote member benefits by exerting downward pressure on fees
• focus competition on, and elevate transparency of, member fees
• likely assist integrity due to its simplicity and accountability mechanisms
• be unlikely to compromise long-term stability because the model as designed
would not lead to excessive concentration of funds or volatile movement of
members and assets
• have slightly lower system-wide costs, mainly due to lower search costs for
members.
DRAFT FINDING 7.1
Relative to the baseline, the multi-criteria tender model would:
• promote member benefits by focusing competition on member satisfaction and
long-term net returns
• focus competition on aspects of performance that matter to members, and more
generally through the winning bid providing a market-wide performance benchmark
• create risks for integrity due to its vulnerability to subjective judgments, yet on the
other hand promote integrity through stronger accountability mechanisms
• not create any material risks to stability, since it is unlikely to lead to excessive
concentration or volatile movement of members and assets
members.

DRAFT REPORT
23
DRAFT FINDING 8.1
Relative to the baseline, assisted employer choice (with employee protections) —
employing both a light filter for mandatory minimum standards and a heavy filter for a
preferred default list — would:
• enhance member benefits by increasing the likelihood of members being placed in
higher quality products, and reducing the likelihood of them being in a poor product
• promote healthy competition by presenting product providers with incentives to
perform strongly against the preferred list selection criteria or compete for
corporate tenders and facilitating greater comparability, but there would still be
some scope for unhealthy and wasteful marketing to employers
• increase the potential for agency problems given the involvement of employers,
although the risk to the integrity of the system would lie primarily in the
appointment process for the preferred default list selection panel
• create few stability concerns
• reduce search costs for employees, while increasing search costs for employers
and regulatory costs for funds and government.

INFORMATION REQUESTS
DRAFT REPORT
25
Information requests
INFORMATION REQUEST 3.1
The Commission is seeking comment from the industry on the effect on fund
participation in the competitive process of moving to the first-timer pool of default
members.
• Are there any funds that might choose not to participate due to the lower value of
the pool, and what is the threshold minimum pool size (per fund) at which this might
happen?
• How would the move to a first-timer pool of default members affect the fees and
product features that funds would offer when competing for new members?
The Commission is seeking feedback on the merits and implementation issues of its
proposal to improve transparency around fund merger activity.
• What (if any) complementary regulatory action would be needed to ensure that the
framework is effective in promoting mergers and the exit of underperforming funds?
• Are there any likely unintended consequences from introducing the proposed
framework? To what extent and how could those unintended consequences be
addressed through policy design?
The Commission is seeking comment on its approach to, and alternative ways of
estimating the size and value of, the turnover and first-timer pools and the benefits
from reducing account proliferation due to moving to the first-timer approach.

DRAFT REPORT
In terms of a shortlist of superannuation products, what evidence is available on the
size of the list that would best facilitate the majority of employees to choose a fund that
meets their needs?
What specific information should be included alongside such a shortlist to help
employees to choose between products? In what format should the information be
presented? What evidence is there for how the metrics would assist employees to
make decisions?
What institutional arrangements would best suit a last-resort fund? Should it be
managed by existing eligible rollover funds or the Future Fund?
Under a system of active employee choice, what would be the costs and benefits of
prohibiting funds or related parties from offering prospective members a short-term
benefit that is unrelated to the superannuation product? What specific form should any
such prohibitions take?
Is the format of a first-priced sealed bid with a best-and-final-offer stage and potentially
multiple winners the best fit for a fee-based auction? Are there any risks associated
with these design elements that have not been identified?
Regarding transparency, what would fall under the exemption of information that, if it
were disclosed, could reasonably be expected to have commercially adverse
consequences for the bidder?
Could a fee-based auction be designed to facilitate second-order competition between
funds on non-fee aspects of performance, such as member services?
Which aspects of member services should be included in the tender criteria? Do
default members value the same services as choice members?
How should default members be allocated across the set of winning products? Are
there problems with sequentially allocating members into products with different
investment strategies, fees and services, as is implied by this model? What role could
a best-and-final-offer stage play in providing a select number of funds the opportunity
to improve their offer?
Regarding transparency, are there any problems with only withholding information that
can reasonably be shown to be commercially sensitive? What information would fall
under this exemption?

INFORMATION REQUESTS
DRAFT REPORT
27
What are the main drivers of costs to employers in selecting default products on behalf
of their employees? Would a shortlist of preferred default products make this task
easier for employers? Is there an ideal minimum number of products that should be
nominated on the preferred default list?
Are there other specific criteria in addition to those proposed under the minimum
standards criteria that default products should meet to protect members and help to
achieve better outcomes for them in the long term?
Would a dual-list approach, allowing employers to select a product from one of two
lists, provide them with sufficient flexibility to select tailored default products that best
meet the needs of their employees?
Which types of employers prefer to retain a role in default product selection? To what
extent are default products or corporate fund offerings considered important benefits
offered to prospective employees in competitive labour markets?

CHOICE AND COMPETITION IN SUPERANNUATION
DRAFT REPORT
29
1 Choice and competition in
superannuation
Key points
• The Commission has been asked to examine new alternative models for a formal
competitive process to allocate default superannuation members to products.
− This inquiry is stage 2 of a 3 stage process, and stems from the Australian
Government’s response to the 2014 Financial System Inquiry.
• Superannuation has evolved much since compulsory superannuation was first introduced a
quarter of a century ago. Today’s default arrangements evolved historically within the
workplace relations system, and provide a safety net for employees that fail to make a
decision given compulsion.
• But while the superannuation system uses a default model to address member
disengagement, this system may at the same time have acted to perpetuate that
disengagement.
• Competition is a means to an end in superannuation, as in other service arrangements.
This inquiry is specifically charged with considering how competition can be factored into
the design of arrangements to allocate default members in order to enhance efficiency, and
thereby member outcomes, in the superannuation system.
• The superannuation system — particularly the default segment — has unique features.
− On the demand side, it is characterised by the compulsory nature of contributions,
complex long-term decisions, and various cognitive and behavioural biases.
− On the supply side, there are unique features such as non-traditional barriers to entry
(industrial agreements), as well as various forms of regulation.
• The ability for employees to exercise choice is integral to the design of Australia’s
(predominantly) defined contribution system, and the option to exercise choice is integral to
competition. While employee choice in superannuation has evolved over time, and a wide
range of funds can compete for new members to join, a substantial share of employees
continue to have restricted or no choice of their superannuation fund.
• Complementary policy action (including to extend genuine member choice to all employees)
will be needed to deliver the full potential of member-driven competition under the
alternative models developed in this inquiry and under current default arrangements.
• Where employees are not able to make choices in their best interests and exert effective
competitive pressure (in the market), there may be a case for a default arrangement. But
even here it is still possible to inject competition for the default market as a whole.
• In this report, a formal competitive process for allocating default members constitutes any
new alternative process that permits open participation and encourages rivalry between
funds to the benefit of members, and involves products being selected based on merit.

DRAFT REPORT
1.1 What is this inquiry about?
The Commission has been asked to examine alternative models for a formal competitive
process to allocate default superannuation members to products. The essence of the task is
to develop a workable model, or models, that could be implemented by the Australian
Government if deemed desirable following the future review of the efficiency and
competitiveness of the superannuation system (which the Productivity Commission will be
requested to undertake after 1 July 2017).
This inquiry represents the second of three related pieces of work on superannuation to be
undertaken by the Productivity Commission (figure 1.1) stemming from the Australian
Government’s response to the recommendations of the 2014 Financial System Inquiry
(FSI) (Murray et al. 2014).3 The FSI found that the superannuation system was not
operationally efficient due to a lack of price-based competition in the sector. To improve
operational efficiency during the accumulation phase of superannuation, the FSI
recommended introduction of ‘a formal competitive process to allocate new default fund
members to MySuper products’, unless a review by 2020 finds such a move would be
unnecessary (Murray et al. 2014, p. 101).
Figure 1.1 Stage 2 in a three-stage superannuation review
What is the scope of this inquiry?
The focus of the Commission’s task in this inquiry (stage 2) is on developing alternative
models for future consideration by the Australian Government. The Commission is not
required in this inquiry to reach views on whether alternative models are better or worse
than the current default arrangements, nor the merits of the MySuper arrangements or the
3 The Commission released its stage 1 study on How to Assess the Competitiveness and Efficiency of the
Superannuation System on 25 November 2016.

DRAFT REPORT
31
competitiveness and efficiency of the system. These matters will be examined in detail as
part of the next phase (stage 3) review of the competitiveness and efficiency of the
superannuation system.
The scope of this inquiry is relatively broad. The terms of reference encourage the
Commission to consider various types of models that could be applied to allocate default
members, to draw on the strengths and weaknesses of approaches used internationally, and
to draw on expertise in the field of competitive models.
Notably, the focus of this inquiry is broader than the Commission’s 2012 inquiry
(PC 2012) into default superannuation funds, which focused on criteria for the selection
and ongoing assessment of superannuation funds eligible for nomination as default funds
in modern awards. In this inquiry, the Commission is considering alternative models for a
formal competitive process for allocating default members across the system as a whole.
The terms of reference also ask the Commission to develop a process for allocating default
fund members to products. This is a slight departure from one part of the current default
architecture, where modern awards are required under the Fair Work Act 2009 (Cwlth) to
specify default funds to which contributions can be made. The reference to funds rather
than products in legislation is a result of the history of the superannuation system, which
pre-dates the Stronger Super reforms introduced in 2013. From October 2013,
superannuation funds have been required to direct all default contributions to a MySuper
product authorised by the Australian Prudential Regulation Authority (APRA). All
members under previous default arrangements must be transferred to a suitable MySuper
product by 1 July 2017 (Australian Government 2011, p. 3).
In practice, under the current arrangements, the terms default fund and product can almost
be used interchangeably. With limited exceptions, only funds that have a MySuper product
can be included in the default fund term of an award, and typically a fund can only offer a
single MySuper product (Australian Government 2011, p. 4). Consistent with the language
of the terms of reference, all of the alternative models assessed in this draft report focus on
allocating default members to products (rather than funds). Default fund members consume
a certain bundle of services (a product) from a superannuation fund. Furthermore,
superannuation funds can offer many different products. For reasons explained further in
chapter 3, the alternative models in this report will focus on allocating members to
products for the accumulation phase only.
Evidence base and further consultation
In keeping with the Productivity Commission Act 1998 (Cwlth), the Commission has
conducted this inquiry using transparent and public processes, with an overarching concern
for the wellbeing of the Australian community as a whole. The Commission published an
issues paper in September 2016, and has met with a range of interested parties including
industry bodies, superannuation funds, specialists, academics, and government officials.
The full list of the study participants that the Commission consulted prior to the draft

DRAFT REPORT
report is detailed in appendix A. The Commission also received 52 submissions and had 43
meetings with stakeholders in the lead up to the release of this draft report.
In developing this draft report, the Commission has given consideration to all submissions
and considered the most relevant available evidence. This includes learnings from
international experiences, the fields of behavioural finance and competitive market models,
and survey-based evidence, which have added to the Commission’s evidence base.
The Commission invites additional comments and submissions on the draft report, and will
also be seeking further feedback via public hearings — further information can be found
on the Commission’s website.
1.2 The role and evolution of defaults in Australia
Default arrangements have been integral to the development of Australia’s superannuation
system. It is useful to explore the role and evolution of, and rationale for, defaults in
Australia as a precursor to considering alternative models for allocating default members.
The nature of the default market
In a standard competitive market, well informed and engaged consumers make decisions in
their own best interests, driving demand for goods or services they value. In turn, suppliers
respond to these signals and dynamic rivalry leads to innovations and cost reductions over
time. There are several reasons why the superannuation system, and in particular the
default segment, does not live up to this textbook ideal.
First and foremost, the system is based on compulsion.
Superannuation, as we know it today, originated from a number of different but related
traditions: employer paternalism, industrial agreements and government compulsion. Each of
these traditions has one thing in common: compulsion. (UniSuper, sub. 20, p. 4)
The compulsory nature of contributions can act to limit the demand-side pressure exerted
by members. Several mutually reinforcing aspects of the system can further accentuate
member passivity and disengagement, such as the complexity of retirement income
decisions, long time horizons, various behavioural and cognitive biases, and the costs of
active involvement (PC 2016; ISA, sub. 40, p. 56; REST, sub. 23, p. 3).
Many people have little sense of the risks they face; many do not understand probability well;
and many do not understand basic concepts in finance … Even if someone has the knowledge
to choose well, the gain from choosing more effectively in any particular month is small,
whereas the transactions costs in terms of time are significant. (Barr 2013, p. 73)
While most employees now have the opportunity to choose the superannuation fund and
product they wish to direct their compulsory contributions to, some remain unable to

DRAFT REPORT
33
choose (although estimating the precise number of people in this situation is not
straightforward (box 1.1)). In any case, most employees (even when they can) do not
appear to exercise choice for a superannuation product. The available survey evidence
suggests up to two-thirds of fund members stay with the default nominated for them by
their employer (PC 2016, p. 199) — hence they can be considered part of the default
segment of the system — and recent research indicates that about one-quarter of those in a
default plan are there by conscious choice (Butt et al. 2015, p. 14). Trust in system
participants has also been identified as a primary motivating factor for people defaulting
(AIST, sub. 28, p. 3; Butt et al. 2015; CIFR, sub. 7, p. 3). The evidence also suggests that
relatively few members switch once they are in a fund (ISA, sub. 40, pp. 42–43), and
CHOICE (sub. 31, p. 7) indicated that superannuation switching rates are below those for
other financial and utility products.
Box 1.1 The evolution of employee choice in superannuation
The ability for employees to choose their superannuation product has evolved over time. From
1 July 2005, most employees have had the right to choose the superannuation fund to which
their contributions are made, including those in modern awards. The main exceptions are those
employed under some workplace determinations and enterprise agreements (chapter 3; Fear
and Pace 2008, p. 12; O’Dwyer 2016b, p. 23; PJCCFS 2007, pp. 16–17). Previous estimates
have indicated that about 20 per cent of the Australian workforce still do not have choice of fund
(ASFA 2012, p. 7; Clare 2010, p. 9; Fear and Pace 2008, p. 12), although it is difficult to verify
this figure. Further, these estimates encapsulate people in different circumstances, and where
choice of fund might not be guaranteed by legislation, it might (in some instances) still be
offered by the employer voluntarily.
In March 2016, the Australian Government introduced legislation that would enable employees
covered under enterprise agreements and workplace determinations that are made after 1 July
2016 to choose their own superannuation fund (Australian Government 2016). The Explanatory
Memorandum to the Bill noted that while comprehensive data are not available, it cited the
Association of Superannuation Funds of Australia estimate in 2012 that ‘20 per cent of
employees continue to have restricted or no choice of superannuation fund’, which the
Australian Government equated to approximately two million individuals (O’Dwyer 2016b,
p. 15). The Australian Government assumed, based on Department of Employment figures, that
its proposed legislative changes would extend choice to up to 800 000 employees under
enterprise agreements and workplace determinations (O’Dwyer 2016a, 2016b, p. 23). This
legislation lapsed at the prorogation of the 44th
parliament and has not yet been re-introduced.
Unlike the accumulation phase, the need to make an active choice (in other words, no
default system) exists in the retirement phase of superannuation, though the Australian
Government recently announced it would seek policy changes in this area (chapter 3). In a
recent research project examining superannuation policy for post-retirement (PC 2015a,
p. 2), the Commission found evidence that despite people being afforded much flexibility
in the way they drawdown their superannuation once they have access, most retirees are
prudent in their drawdown behaviour (including those that take lump sums). The necessity
to make choices is evidently better established in members’ minds by the time of
retirement.

DRAFT REPORT
Current default arrangements
The rules and processes surrounding how default members are allocated to a default
product vary significantly across awards, agreements, contracts and other employment
arrangements (figure 1.2). These arrangements have largely evolved through the industrial
relations system (box 1.2).
Figure 1.2 Default fundsa in Australia’s workplaces
a All default funds must offer a MySuper product or be an Exempt Public Sector Superannuation Scheme.
b Numbers in brackets are the approximate proportion of total employed persons under the specified work
arrangements, sourced from (PC 2015b, p. 5). cThis includes state-based systems and common law
employment contracts. d Currently, the vast majority of the 122 listed current modern awards stipulate one
or more default funds from which an employer may select one. Where an award does not list any funds,
employers are free to choose any fund with a MySuper product as their default. Where there are
grandfathering provisions in a modern award, employers can generally continue to make contributions (on
behalf of their employees) to their previously selected default fund, even if it is no longer listed in the
award (PC 2012, p. 6).
Employment arrangements
(11.8 million employed persons)
Otherc
Awardd Agreement
More than
one default
fund listed
No default
fund listed
One default
fund listed
Refer to
award
List one or
more
default
funds
No default
fund
specified
Stipulated
fund
Exemption
Does employer
have to pay
super?
Employer
must
choose
from list
Employer
free to
choose any
fund
Employer
has no
choice
Employer
must
choose
from list
Employer
free to
choose any
fund
Employee
has no
choice of
fund
Fair Work Act 2009
Yes
No
(71%)b (29%)b

DRAFT REPORT
35
Box 1.2 The origins of default funds and products
The obligation on employers to make superannuation contributions (on behalf of their
employees) progressively became part of various industrial awards during the 1980s. In 1992,
compulsory employer contributions were extended to almost all employees with the introduction
of Commonwealth legislation to establish the Superannuation Guarantee.
Existing default arrangements evolved in the context of the workplace relations system. Specific
superannuation funds were named in awards when superannuation became an industrial matter
in national wage bargaining in the 1980s. Several industry-based funds were established as
not-for-profit entities to cater for employees in specific industries.
From 1 July 2005, most employees have been able to choose their superannuation fund and (in
effect) the product to which they want their contributions directed (box 1.1). Where employees
have not made a choice, the funds named in awards have become the default option in many
cases. Default arrangements provide a safety net for employees that fail to make a decision, in
view of the compulsory nature of superannuation.
Various rationales have been put forward as to why the majority of people fail to make a
decision, creating the need for defaults. These include the inherent complexity that individuals
face in making decisions about retirement incomes, leading to fears that employees who were
disengaged or ill-equipped to make decisions could make poor decisions, which in turn would
increase future reliance on the Age Pension. Default arrangements were also seen as reducing
the cost and complexity for employers of complying with their Superannuation Guarantee
obligations (given concerns, particularly from unions, about non compliance by some employers
in the absence of relatively straightforward default arrangements).
The superannuation system has evolved in a myriad of ways since compulsory superannuation
was first conceived. For example, several industry funds have merged to create larger funds
with more diverse memberships, and many of these industry funds have become public-offer
funds. More broadly, there has been significant consolidation in the number of institutional funds
from about 4700 to 250 funds over the 1997 to 2016 period. There have also been
accompanying improvements in transparency. MySuper is intended to be a simple and cost
effective superannuation product replacing previous default products. It is designed to ensure
members do not pay for any unnecessary features they do not need or use. MySuper products
are intended to have a relatively standard set of features to enable members, employers and
market analysts to compare funds more easily.
Sources: Australian Government (2011); Productivity Commission (2012, 2016); Shorten (2011); Vision
Super (sub. 4, pp. 4–5); Weaven (2016).
There are up to four ‘third parties’ involved in the selection of default products for
employees that fail to make an active choice (at least as currently envisaged under the
legislated arrangements) — the regulator (APRA), an expert panel within the Fair Work
Commission (FWC), the full bench of the FWC and, finally, employers (figure 1.3).
Employers’ level of discretion in the product selected for their employees depends on the
specific employment arrangements in place. For example, they could effectively have no
choice (where an employee’s award prescribes one fund) to wide-ranging choice (where an
employer can select any authorised MySuper product).

DRAFT REPORT
Figure 1.3 Decision steps in selecting default products
A stylised view
a The process set out in legislative changes introduced in 2013 — to establish a two-stage process within
the Fair Work Commission (FWC) to determine a default list of products and review those to be included in
each award — has not yet been finalised (FWC, sub. 51, p. 2). bThis includes that a MySuper product
must be based on a single diversified investment strategy (which may be a life-cycle strategy) (s. 29TC of
the SIS Act), adhere to rules and restrictions on product fees (Division 5 of Part 2C of the SIS Act), and a
requirement that the product must incorporate life and total and permanent disability insurance on an opt
out basis (s. 68AA of the SIS Act). cApplications must include a range of information about the applying
fund and its standard MySuper product performance. The relevant criteria are set out in s. 156F of the
FWA. dA standard MySuper product satisfies this second stage test if it is on the Default List and the FWC
is satisfied that specifying it would be in the best interests of the default fund employees to whom the
modern award applies (s. 156H of the FWA).
Sources: APRA (2012); Australian Government (2011, pp. 5–6); Fair Work Act 2009 (Cwlth) (FWA); FWC
(2016); Superannuation Industry (Supervision) Act 1993 (Cwlth) (SIS Act).
Fund trustees are subject to fiduciary and statutory obligations to operate in the best
interests of their members (CIFR, sub. 7, p. 2; Kinetic Superannuation, sub. 45, p. 2). In
practice, these obligations translate into a detailed set of prudential standards, reporting
requirements, supervisory frameworks and other regulatory guidance material provided to
trustees, which include additional duties for MySuper products (APRA, sub. 33, pp. 2–3).
Although employers are responsible for selecting a default product for their employees,
they do not face similar legal obligations to act in the best interests of employees (ASIC,
sub. 41, p. 2). They are, however, provided with regulatory guidance on a range of matters
such as on illegal inducements and factors to consider in default fund selection, and are
required to comply with their Superannuation Guarantee obligations.
Participant views on current default arrangements
Several inquiry participants expressed concern that the current supply-side arrangements in
the default segment (particularly their interaction with industrial awards and agreements),
MySuper authorisation
• License RSEs
• Standard featuresb
• Enhanced trustee
obligations
• Annual reporting on
fees, investment
strategy & scale
Product choice
• Prescribed or limited
choice in award or
agreement
• Full choice, select
from any MySuper
product
APRA EmployersFWCa
(Process not yet finalised)
Expert Panel
• Select eligible
MySuper products
and compile the
Default List
• Assess applications
based on factors set
out in legislationc
Full Bench
• Review default fund
terms of awards
• Specify (typically)
2-15 products in
awardsd
• 4-yearly review
In some cases employers can select from any MySuper product

DRAFT REPORT
37
restrict competition and innovation which could benefit members, and are duplicative and
inefficient (for example, Colonial First State, sub. 25, p. 1; FSC, sub. 38, p. 3, p. 31).
Other participants suggested that the current default arrangements have served an
important purpose by linking default members to funds with a culture of serving others and
operating in the best interests of members, not shareholders or other related parties (for
example, ISA, sub. 40, pp. 7, 25; Vision Super, sub. 4, pp. 6–9). In support of this view,
many of these same participants highlighted empirical evidence of superior returns
delivered by not-for-profit funds relative to for-profit funds, and for default funds
nominated in modern awards particular. Other participants noted that there is significant
diversity in investment performance even within the population of current MySuper
products (box 1.3).
Box 1.3 Selected participants’ views on default sector performance
The relative performance of superannuation funds (and products), and how these are best
measured and presented, is a source of ongoing debate in the superannuation system and
examination in the literature (PC 2016, pp. 112–130). This box focuses on views raised by
inquiry participants about the relative performance of the ‘default segment’.
Industry Super Australia (sub. 40, p. 10) noted (based on SuperRatings data) that over the past
10 years, funds named in modern awards as eligible defaults have outperformed non-default
options, with average annual net returns of about 5.9 per cent compared with 4.9 per cent. The
Australian Institute of Superannuation Trustees (sub. 28, pp. 15–16) noted that a list of top
performing funds provided by Chant West for the 10 years to 30 June 2016 included funds that
were all default funds in modern awards.
Some participants also noted the relative superior investment performance of industry-fund
default products within the default segment. For example, Industry Super Australia (sub. 40,
p. 46) noted that over the past 10 years, average net returns (for products listed within awards)
were 6.0 per cent for industry funds compared with about 4.0 per cent for retail funds, while the
Australian Institute of Superannuation Trustees (sub. 28, p. 15) cited broadly similar
comparative results.
The Financial Services Council (sub. 38, p. 32) noted evidence that — since MySuper started in
2013, and over the past five and 10 years — there has been significant variation in the
performance of different default products (now MySuper products) that are listed in modern
awards as default funds.
Based on the data it collects, the Australian Prudential Regulation Authority (sub. 33, p. 2) said:
… since 1 July 2013, there has been considerable variation in net returns and fees for different
MySuper products, leading to a wide range of outcomes for members across these different products.
While there has been some evidence of reductions in fees and costs since MySuper products were
introduced, particularly for products with previously very high fee levels, there is clearly room for further
improvement. Further, while many MySuper products have achieved their net return targets over the
past few years, some have fallen well short.

DRAFT REPORT
The Commission’s analysis of APRA (2017d) product-level net returns data4 showed that
while the top quartile MySuper products is dominated by industry funds, the bottom
quartile is populated by a roughly even number of industry, retail and corporate funds.
Admittedly this analysis is much caveated by data limitations, but it does suggest that poor
performance is not necessarily dictated by fund type. APRA recently noted that it is
looking closely at the bottom performing funds (across a range of factors in addition to net
returns), of whom a large proportion are not-for-profit and industry funds (Rowell 2017,
p. 28).
Most participants consider there is an ongoing need for defaults …
Inquiry participants from within the superannuation industry generally suggested that some
form of default arrangements will continue to be a necessary feature of the system for the
foreseeable future (if not into perpetuity) irrespective of the process for allocating default
members (for example, AIST, sub. 28, p. 4; ASFA, sub. 24, p. 3; ISA, sub. 40, p. 8; Vision
Super, sub. 4, p. 4). Four inter-related factors were generally cited as reasons for the
ongoing need for default arrangements.
• The compulsory nature of superannuation.
• The level of employee/member disengagement (especially for young members).
• The complexity and costs associated with members making decisions.
• The inability of some members to make good decisions due to a lack of financial
literacy or the influence of various behavioural or cognitive biases.
On this basis, most participants were highly sceptical of (or explicitly rejected) any model
framed around active member choice without a default safety net (for example, CIFR,
sub. 7, p. 5; ISA, sub. 40, p. 55), at least in the near term (for example, FSC, sub. 38, p. 3).
… but views differed on the allocation mechanism
Several inquiry participants favoured the continuation of an administrative allocation
model for allocating default members within the industrial relations system (for example,
AIST, sub. 28, pp. 4–5; ISA, sub. 40, p. 5). Participants noted that the current system
already has elements of an administrative and market-based model — such as the
legislated (but not yet finalised) FWC expert panel evaluation process for being listed in an
award — and that this system should be given an opportunity to operate (for example,
AustralianSuper, sub. 19, p. 1; REST, sub. 23, p. 2). Some participants also emphasised
that employers should still be considered central to the allocation process within this
framework (for example, AHA, sub. 6, pp. 3–4; ISA, sub. 40, pp. 30–31, 44; UniSuper,
4 Data availability and consistency limited the Commission’s analysis to 76 products over 11 quarters to
31 December 2016.

DRAFT REPORT
39
sub. 20, pp. 8–9) and that MySuper authorisation is an insufficient filter (ISA, sub. 40, p. 5;
REST, sub. 23, p. 5).
Other participants favoured opening up access to default members to a wider range of
potential suppliers using the MySuper authorisation as the starting point for a filter, with
scope to enhance it (for example, FSC, sub. 38, p. 21). There were different views among
this group on whether employers should be compelled to select a default for their
employees, the scope for treating large and small employers differently, or whether another
party should be designated the allocation task (for example, CIFR, sub. 7, pp. 6–7; FSC,
sub. 38, p. 21; WSSA, sub. 11, p. 3). However, the general view of this group (explicitly or
implicitly) was to consider models where defaults are administered outside the industrial
relations system.
Use of an auction or tender model as part of a default mechanism was rejected by most
industry participants (discussed further in section 1.3). More detailed participant views on
specific models are outlined in the relevant chapters.
1.3 How are competition and choice relevant to the
default segment of the system?
The Commission has been tasked to develop alternative models for a new formal
competitive process for allocating default members. Competition is a means to an end, with
the ultimate objective being arrangements that promote members’ best interests and
community wellbeing.
Several inquiry participants questioned the merits of focusing on a formal competitive
process (and competition in superannuation more generally) as a desirable objective for
superannuation, and in particular for default members (for example, ISA, sub. 40, pp. 7,
51). Others supported the introduction of greater competition in the processes used to
select eligible default funds and products, and see expanded active member choice (even if
unlikely in the near term) as an objective worth striving towards (for example, FSC,
sub. 38, p. 3).
This section explores the potential role of competition and choice in the superannuation
system (and the relevance for the design of alternative models), as a precursor to outlining
the Commission’s approach to assessing alternative models in chapter 2.
Competition and member choice are relevant to default members
Competition — meaning rivalry among providers — can be an important force for
promoting efficiency. It will generally be effective where the opportunity exists for
incentives and consumer pressure to address price and quality preferences of potential
members. Through this process, competition can lead to reduced fees, improved service
quality and product innovation. Competition can also ‘weed out’ inefficient or

DRAFT REPORT
unsustainable providers. However, competition in superannuation may not always promote
outcomes consistent with members’ best interests or community wellbeing. It will
generally be less effective where members are not well informed or engaged, preferences
are not identified, products are difficult to compare in a meaningful way or where
providers compete on features of little value to consumers.
It is also worth considering what is often observable in markets where competition is not
present: incentives are skewed towards interests other than those of consumers. While it is
always possible to point to examples of selfless service providers, with large mandated
cash flows in superannuation, risks remain that practices will not always coincide and
evolve with the best interests of disengaged individuals. The need to innovate or to
anticipate shifts in consumer needs is weaker without the threat of members moving their
contributions.
Several inquiry participants that were sceptical about the use of competition as a force for
member interests appeared to put most faith in the wisdom of trustees as capable of
addressing these needs without any greater incentive than their current fiduciary and
statutory obligations. This inquiry will provide an opportunity to test that proposition more
fully.
There are various ways that principles of competition and member choice can inform the
design of alternative arrangements for allocating default members. An important first step
is to consider the nature of competition and the functional level of the market at which it
can be targeted. Broadly, competition could be promoted in one of two ways:
• competition in the market — which refers to competition based on conventional
demand- and supply-side interactions between employees (members) and trustees
(funds)
• competition for the market — which refers to competition between funds via some
form of formal process to earn the right to gain access to default members.
Member choice can drive competition in the market
Promoting competition in the market relies on consumers being able to make choices. As
noted above, while most employees in the superannuation system already have the
opportunity to choose which superannuation product their contributions are paid into, some
employees remain unable to choose (box 1.1). Most participants supported members’
freedom to choose their fund and product (for example, AMP, sub. 42, p. 1; IPA,
sub. 17, p. 12; Mercer, sub. 15, p. 4), although some emphasised that choice can lead to
poorer performance outcomes (for example, ISA, sub. 40, p. 8).
Several participants noted that employee choice is already driving competition in the
market to benefit members. For example, AIST (sub. 28, p. 27) cited analysis from Chant
West that suggests there is considerable pressure on funds to attract and retain members
and the fees funds charge represent an important point of difference (especially under

DRAFT REPORT
41
MySuper arrangements). Competition outside the boundaries of the default segment —
such as from self-managed superannuation funds — further contributes to the competitive
dynamics in the system (Ai Group, sub. 21, p. 1; REST, sub. 23, p. 4).
Other participants were sceptical of the level of member-driven competition. For example,
Kinetic Superannuation noted that:
As identified by recent inquiries, the Australian superannuation industry lacks a strong
consumer-driven competition that is particularly pronounced in the default fund market. This
contributes to perceptions of inertia and complacency in the default fund system. (sub. 45, p. 1)
Downside risks of promoting greater competition and choice were also highlighted by
participants. For example, inquiry participants indicated that models that seek to promote
greater competition may lead to: increased funds’ marketing and distribution costs (AIST,
sub. 28, p. 50; REST, sub. 23, pp. 3–4), more sales driven switching that draws members
into underperforming products (ISA, sub. 40, p. 5, 41), higher member (or employer)
search costs (AIST, sub, 20, p. 30), and exposure of employees and (particularly small)
employers to choices they are ill-equipped to make (CSA, sub. 35, p. 2). These can
apparently occur despite obligations already imposed on trustees to act in members best
interests.
It can also be reasonably argued that there is a great deal of choice in the superannuation
system already — perhaps too much for many new members to cope with, contributing to
persistently high use of the default option. And there is little doubt that freedom of choice
includes the freedom to make mistakes.
Another issue raised by participants is that the present system may be unable to prevent an
employer coercing its employees into a particular fund, perhaps based on inducements the
business receives that do not benefit the member. While legislative provisions are designed
to prevent funds from offering employers inducements5, participants had differing views
on their efficacy (for example, FSC, sub. 38, p. 13; ISA, sub. 40, p. 41). And the regulator,
the Australian Securities and Investments Commission noted that:
… the real prospect remains of employers selecting a fund as a result of inducement, including
lawful inducements and on the basis of competition directed to the employer rather than to the
quality of the product for retirement income purposes of the employees. (sub. 41, p. 4)
Improvements to member choice in the future are thus likely to be a necessary but
insufficient condition to help realise the full potential of member-driven competition in the
superannuation system. Such improvements might include, for example, the removal of
restrictions in enterprise and workplace agreements that limit employees’ choice of fund,
or further action to discourage coercion. A maturing system may improve the scope for
member-driven competition. Thus, barriers to relying on competition include a potential
dependence on complementary policy actions, rather than the simple opening up of more
choice. The policy choice is more complicated than it might first seem.
5 Section 68A of the Superannuation Industry Supervision Act 1993 (Cwlth).

DRAFT REPORT
A maturing system may improve the scope for member-driven competition
The decision to choose a superannuation fund or product, or stay in the default, is not the
only time members can exercise choice in order to drive competition. The ongoing process
of monitoring their fund’s performance, comparing market offerings and (and at the
extreme) switching providers (or deciding stay with their provider) are all opportunities to
engage. Members are also more likely to become engaged with their superannuation as
they get older and approach retirement (PC 2016, p. 192). Therefore, as the system and
member population matures, a greater proportion of members will, in effect, act as agents
for competitive pressure intensifying on funds’ product and pricing decisions.
The system has matured in various other ways over the past quarter century. For example,
employees and employers are more familiar with the concept of superannuation and
employees switch jobs more often (on average) and therefore have more prompts to
potentially make decisions about their superannuation (ABS 2013, 2016). There has been
some consolidation in the funds market (box 1.2). Funds are also utilising new ways —
such as mobile phone applications and social media platforms — to interact with members
and to capture new members (Frost 2016; Willis Towers Watson 2016).
Superannuation policy settings have also evolved. As noted above, employee choice was
extended in 2005 (box 1.1) and regulations introduced as part of MySuper have also
enhanced the ability for members, employers and market analysts to compare products
more easily. Another system change has been the introduction of SuperStream and Single
Touch Payroll which aim to streamline and simplify the manner in which employers
interact with superannuation funds and comply with their Superannuation Guarantee
obligations.
Notwithstanding the system evolving in various ways, several participants suggested that
the factors which underpin the rationale for defaults have not changed — namely, the
inherent complexity of decisions about retirement income, employee disengagement and
employees not feeling qualified to make an informed decision (for example, AIST, sub. 28,
pp. 47–49; Vision Super, sub. 4, p. 4). Participants also identified areas where the system
could be further improved, including:
• centralised information to enable more meaningful MySuper product comparisons,
fund selection and consolidation (CHOICE, sub. 31, p. 9; WSSA, sub. 11, p. 2)
• member education on financial and superannuation issues (WSSA, sub. 11, p. 5)
• more effective design and enforcement of regulation in particular areas, such as
employer inducements and disclosure obligations (AIST, sub. 28, pp. 23–24; ISA,
sub. 40, p. 41).
The factors which signal a maturing system do not necessarily suggest there will be no role
for defaults in the near future, or that a particular type of alternative allocation model
should take precedence in this inquiry. However, the above factors provide important
context for the development of new alternative models which could be deemed workable.

DRAFT REPORT
43
They also suggest that some of the earlier concerns which formed the basis for the need for
defaults — such as a lack of transparency and limited member choice — simply do not
hold the same sway today. In other words, the Commission needs to develop alternative
models to meet the needs of superannuation members of the future.
Further, while the superannuation system uses a default model to address member
disengagement, this system may at the same time have acted to increase or perpetuate that
disengagement (PC 2016, p. 38). Some participants in this inquiry noted that member
disengagement can, in part, be attributed to the existing institutional settings, including the
operation of the current default arrangements. For example, the Financial Services Council
said:
… in evaluating consumer engagement and behavioural biases, it would be remiss of the PC to
conclude that consumers will never be engaged with the superannuation system. The current
system of having a default model by definition creates an incentive for consumers to be
disengaged. (sub. 38, pp. 7–8)
To better understand how employees would react in a world where there was a greater
reliance on them to choose their superannuation product, the Commission is undertaking
an experimental survey of superannuation decision making. The results and analysis will
be made publicly available (upon completion) and will also be incorporated in the
Commission’s final report.
Where used, defaults need to be supported by clear objectives and effective
regulation
There is a complex interplay between defaults and member engagement, which has
implications for the ultimate goal of a default system. Some default models may be
designed with reference to member engagement, namely to nudge default members
towards taking active decisions to shift out of their default fund or product — for example
the KiwiSaver default model in New Zealand (appendix B).
The current default arrangements in the Australian system are framed around the simpler
objective of meeting members’ best interests — including through the initial product they
are allocated to — rather than to nudge them towards active choice per se. In the presence
of a default system where other agents are making decisions on behalf of members, it is
imperative that those agents face clear objectives, and in the context of the Australian
system, are subject to explicit and enforceable obligations to act in members’ best interests.
As noted above, fund trustees have statutory and fiduciary obligations to operate in the best
interests of their members. Although employers are responsible for ultimately selecting a
default product for their employees, they do not face similar legal obligations to act in the
best interests of employees. Inquiry participants suggested that changes to current default
arrangements — such as requiring employers to choose from a long list of products —
would impose significant costs on employers (for example, REST, sub. 23, p. 14).

DRAFT REPORT
DRAFT FINDING 1.1
Complementary policy action (including to extend genuine member choice to all
employees) will be needed to deliver the full potential of member-driven competition
under the alternative models developed in this Inquiry and also under current default
arrangements.
DRAFT FINDING 1.2
Where there is third-party involvement in the selection of a default product, there
needs to be effective regulation or arrangements in place to ensure these third parties
act in the best interests of members.
A competitive process can stimulate competition for the market
Promoting competition for the market may be desirable where it is not feasible to promote
effective competition in the market, such as due to barriers to entry and exit, or monopoly
supply conditions.6 It may best be thought of as a way of encouraging the generation of
beneficial variations to products in return for a fixed percentage of new member inflows.
Typical ways to promote competition for a market include a tender or auction process.
These mechanisms can encourage participants to make a competitive (service and price)
offering for initial rights to the market, and incentives to continue to act in a competitive
and efficient manner over time where the process is repeated regularly (chapters 6 and 7).
In this way, regular competition for the market also means it becomes contestable.7
Tenders and auctions are widely used in the public and private sectors to procure a broad
range of products and services. Tenders, in particular, already feature in elements of the
Australian superannuation system, for example, the use of corporate tenders by employers
and some funds’ tendering of administration and insurance providers (REST, sub. 23, p. 4).
There was near universal agreement among participants across all parts of the industry —
industry and retail superannuation funds included — that it would not be desirable for the
Commission to contemplate a tender or auction model for the selection of default products
and the allocation of members to those products (box 1.4). Only a few participants saw
merit in the use of a tender-type mechanism for the allocation of default members under
6 Promoting competition for the market could also go hand-in-hand with efforts to promote competition in
the market, such as a tender process followed by a focus on member choice.
7 Contestability can be defined as the degree of ease with which firms can enter or leave a market reflecting
the level of potential competition. In a contestable market the threat of new entrants causes the incumbent
firms to operate at levels approaching that expected in a competitive market (Industry Commission 1996,
p. xix).

DRAFT REPORT
45
different models (CMD, sub. 18, p. 2; IPA, sub. 17, p. 7, Kinetic Superannuation,
sub. 45, p. 1).
Box 1.4 Participant concerns about auction and tender models
Inquiry participants raised a wide range of concerns about the potential use of an auction or
tender model to allocate default members to superannuation products.
• A narrow focus on fees would lead to incentives for funds to cut service quality and shift to
low-cost, passive investment strategies that may not maximise long-term net returns (for
example, IFAA, sub. 13, pp. 3, 8; ISA, sub. 40, p. 54).
• These mechanisms would invite commoditisation when tailoring of products to meet diverse
member needs and innovation is needed (for example, AMP, sub. 42, p. 4; ASFA, sub. 24,
p. 13; CIFR, sub. 7, p. 5).
• A tender for a single fund that is held infrequently would destroy competition because of a
lack of possible participants (ISA, sub. 40, p. 54), and a national tender or auction could lead
to a much more concentrated market in the long term with adverse implications for fund
viability and competition (for example, CSA, sub. 35, p. 7; UniSuper, sub. 20, p. 7). It would
be of concern if the process was ‘used as a de facto tool to achieve fund consolidation’
(IFAA, sub. 13, p. 5).
• The tender model introduced in Chile reduced administration fees for default members but
has not led to better member outcomes (including investment returns) and is unlikely to in
the long term (AIST, sub. 28, p. 3; FSC, sub. 49, p. 1; ISA, sub. 40, pp. 12–13). The
Australian default system substantially outperforms the Chilean system (ISA, sub. 40, p. 11).
• These mechanisms introduce the risk of gaming (for example, ASFA, sub. 24, p. 13; First
State Super, sub. 26, p. 1; FSC, sub. 38, p. 30) and, for an auction model in particular, tacit
collusion (for example, FPA, sub. 29, p. 11; FSC, sub. 38, p. 30).
• There are high costs associated with funds participating in these types of processes, and for
administering them (for example, ASFA, sub. 24, p. 14; AustralianSuper, sub. 19, p. 3).
The ability for this issue to unite an otherwise disparate set of industry voices against the
introduction of an auction or tender model may, in itself, be instructive. Notwithstanding
that there may be legitimate concerns about the suitability of various tender models
conducted overseas, it may also suggest that, given the massive flows of mandated
contributions at stake, there are rents to be protected by funds, which should be recovered
for the benefit of members. It certainly requires some sceptical thought from a
policy-maker’s perspective, and thus some analysis from the Commission on alternative
and workable models. And, in any event, this is what the terms of reference require.
At the outset, some concerns raised about the design of a tender or auction model appear
less than compelling. For example, concerns about gaming need to be tempered by the fact
that funds (and trustees more specifically) would still be subject to a range of statutory and
fiduciary duties which bind all fund trustees (Kinetic Superannuation, sub. 45, p. 2). And
collusion between proponents is a difficult proposition to argue other than in the abstract,
with a cohort of potential participants numbering in excess of 100.

DRAFT REPORT
Other participant concerns raise more substantive issues which will require careful thought,
evaluation and model design mitigation in this inquiry — for example, funds’ incentives to
shift to low-cost investments. Chapters 6 (fee-based auction) and 7 (multi-criteria tender)
provide more detail on participants’ views, the likely materiality of specific risks, and the
ways they could be mitigated (if at all).
Auctions and tenders have been used overseas (in Chile and New Zealand; appendix B),
and in the Northern Territory (chapter 7). These experiences provide useful evidence for
the Commission’s consideration and iterative development of alternative models afforded
by this inquiry.
A new formal competitive process does not necessarily mean an auction or tender
The terms of reference ask the Commission to consider alternative models for a formal
competitive process for allocating default members, and indicate it ‘may consider auction,
tender and other types of competitive processes’. The focus of this inquiry on a formal
competitive process received mixed reactions from inquiry participants.
Some expressed concern that it favoured market-based models, such as a tender or auction.
For example, Industry Super Australia said:
It is not appropriate for the Inquiry to insist that a model includes a “competitive process”
because, as the Commission has elsewhere acknowledged, competition is only a means to an
end. An Inquiry that required the default selection model to be a competitive process would
place form over substance. (sub. 40, p. 51)
The use of the term appropriate is odd in this context — it is strange in a policy review
context to imagine there is something improper about considering an idea. And indeed, the
terms of reference seek such consideration.
CIFR (sub. 7, p. 2) submitted that focusing on models which are based on a formal
competitive process narrowly presents the issue as a matter of the mechanism by which
default members are allocated, whereas who chooses is the more important issue given the
nature of the default superannuation segment and default members.
Other participants appeared to interpret models for a formal competitive process in a
broader sense. For example, REST (sub. 23, p. 4) submitted that the current administrative
model for allocating default members already has elements of a market-based model, such
as the proposed process for being listed as a default fund after consideration by an expert
panel. The Australian Institute of Superannuation Trustees submitted that:
The framework proposed by the Commission requires a competitive process to determine
which funds are eligible to be used as default funds. AIST has no difficulty with a competitive
process, provided it is correctly framed and will deliver a result in the best interests of fund
members. (sub. 28, p. 25)

DRAFT REPORT
47
Rather than adopt a narrow approach, in developing alternative models in this inquiry the
Commission has taken a broad interpretation of what constitutes a formal competitive
process. In essence, the Commission considered any new alternative process that is
relatively open, capable of encouraging rivalry between providers in order to generate
competitive tension which influences the products and services funds offer, and where
there is some basis for objective and independent assessment about products based on
merit. The processes designed would need to:
• be contestable for efficient and innovative providers to become eligible to supply
default products
• improve competitive tension to members’ advantage in the default market (and possibly
beyond it)
• encourage persistent under-performers to exit.
Based on this definition, the Commission has examined several types of models in this
draft report, including those based on an administrative filter with employer choice, a
fee-based auction, a multi-criteria tender, or assisted employee choice. An important
design consideration in each model is who has responsibility for making decisions
(chapter 3).
DRAFT FINDING 1.3
For the purposes of this Inquiry, a formal competitive process for allocating default
members constitutes any new alternative process that permits open participation
(contestability), encourages rivalry between funds (competition) to the benefit of
members, and involves products being selected for members based on merit.
1.4 A guide to the rest of the report
This report consists of an overview, eight chapters (including this one), and two
appendixes. Chapter 2 outlines the Commission’s assessment framework for assessing
alternative models. Chapter 3 outlines the Commission’s approach to the development of
alternative models for assessment, including the common foundations across all models.
Chapter 4 details the Commission’s no defaults baseline that alternative models will be
compared against. The next four chapters examine the alternative models against the
assessment framework outlined in chapter 2.
• Chapter 5 examines the assisted employee choice model.
• Chapter 6 examines a fee-based auction model.
• Chapter 7 examines a multi-criteria tender model.
• Chapter 8 examines an assisted employer choice (with employee protections) model.

DRAFT REPORT
Appendix A outlines the Commission’s public consultation process.
Appendix B provides background on retirement income systems and default arrangements
in Chile, New Zealand, the United Kingdom and Sweden.

THE COMMISSION’S ASSESSMENT FRAMEWORK
DRAFT REPORT
49
2 The Commission’s assessment
framework
Key points
• The default segment (as best measured) represents $474 billion as at June 2016,
23 per cent of the about $2 trillion in the superannuation system, and importantly, about
two-thirds of fund members.
• In assessing new alternative models, primary consideration will be given to whether each
model is capable of meeting members’ best interests and community wellbeing. The
assessment will be mostly qualitative given the developmental nature of this inquiry to
identify workable alternative models.
• More specifically, five model assessment criteria will be applied.
− Member benefits: whether the model creates incentives for funds to deliver default
products that maximise long-term net returns and allocate members to products that
meet their needs.
− Competition: the extent to which the model can harness the benefits of open
participation and rivalry between funds to drive innovation, cost reductions and
improved long-term outcomes for members.
− Integrity: whether the model promotes a high degree of integrity in the selection,
filtering, allocation, and in the behaviour of system participants.
− Stability: whether the model is likely to create problems or risks which have the potential
to develop into significant systemic problems across the system.
 Churn and the exit (or takeover) of underperforming funds do not necessarily
represent stability concerns, and could instead be a sign of healthy competition
delivering higher long-term net returns.
− System-wide costs: the impact the model would have on overall system-wide net costs
taking into account the impacts on members, funds, employers and government
(including the implications for future reliance on the Age Pension).
• A baseline of no default arrangements will be applied. It is the only objective baseline that
can feasibly expose the relative abilities of alternative models to address today’s problems
rather than past issues (some of which may no longer be relevant). It is not an alternative
model.
• In this inquiry, alternative models will need to beat the baseline against the assessment
criteria.
This chapter outlines the approach the Commission will take in assessing the merits of new
alternative models. This includes specifying the objective of this inquiry (section 2.1), the
model assessment criteria (section 2.2) and the baseline that will be applied as the

DRAFT REPORT
reference point in the analysis (section 2.3). Chapter 3 outlines the approach to developing
models.
2.1 The objective of this inquiry
The overarching objective of this inquiry is to identify new alternative workable models
for allocating default members to superannuation products. In assessing whether a model is
workable or not, the Commission will consider whether it is capable of delivering
outcomes which promote members’ best interests and (in keeping with the Productivity
Commission Act 1998 (Cwlth)) the wellbeing of the Australian community as a whole.
This will be informed primarily by the extent to which a model meets the relevant
assessment criteria (outlined in the next section).
The Commission’s assessment will necessarily be mostly qualitative given the focus of this
inquiry is on identifying workable models.
In many (if not most) cases, a model that promotes members’ best interests could also be
expected to promote the wellbeing of the Australian community. For example, better
investment outcomes for members will ultimately benefit the Australian community as a
whole where it improves wellbeing in retirement and reduces fiscal pressure on the
government. However, instances may arise where members’ interests and those of the
broader community do not align. For example, a process where employers are responsible
for selecting default products may lead to better outcomes for their employees, but it may
also impose costs on employers and regulators. Additional processes that involve
significant investments by taxpayers would also need to offer significant community
benefit, even if they were in members’ individual interests. Trade-offs between member
and community objectives will be identified and factored into the Commission’s analysis
of each new alternative model.
2.2 The model assessment criteria
The Commission’s issues paper set out five proposed model assessment criteria to evaluate
each new alternative model.8 The criteria were informed by the terms of reference and the
efficiency and competitiveness criteria developed for the stage 1 study (PC 2016).9
There was mixed feedback from inquiry participants on the criteria set out in the
Commission’s issues paper. Several participants expressed broad support (for example,
8 These are distinct from the criteria (or metrics) that would be used within each model to assess the merits
of individual superannuation products.
9 The focus of this inquiry is narrower. The criteria developed to examine new alternative models
(a component of the system) are designed for this specific purpose.

DRAFT REPORT
51
CSA, sub. 35, p. 4; FPA, sub. 29, p. 3; Kinetic Superannuation, sub. 45, p. 2), although
some suggested they were broad and would be difficult to operationalise in places (for
example, APRA, sub. 33, p. 1; PwC, sub. 12, p. 3), and that the criterion on members’ best
interests should supersede all others (for example, ACTU, sub. 34, p. 9; ISA, sub. 40,
p. 38). Others queried specific criteria and suggested areas for amendment (for example,
FSC, sub. 38, pp. 7–15; p. 17; ISA, sub. 40, pp. 38–39; REST, sub. 23, p. 7).
The following sections outline the Commission’s five model assessment criteria, which
have been refined based on inquiry participant feedback. The Commission has also sought
to provide more guidance on how these criteria will be applied, and potential trade-offs
between criteria that may need to be considered. These criteria will be considered and
interpreted collectively to inform the overall assessment of each model (figure 2.1).
Figure 2.1 The framework for assessing alternative models
Criterion 1: Member benefits
Does the model create incentives for funds to maximise long-term net returns and allocate
members to products that meet their needs?
As noted above, a primary factor in examining the merits of new alternative models is
whether, and to what extent, each is capable of meeting members’ best interests. The
Commission flagged in its issues paper that members’ best interests equates to
consideration of the extent to which an alternative model maximises long-term net returns
and allocates members to products that meet their needs.
Objective of
this inquiry
To develop a new alternative workable model, or models, that could
be considered following a future review of the system
Model
assessment
criteria
System-
wide costs
Members’ best interests
Member
benefits
Competition Integrity Stability
Objective of
alternative
model
Community-wide wellbeing

DRAFT REPORT
There was broad agreement among inquiry participants that members’ best interests should
be paramount in the assessment, and several argued this criterion should supersede all
other considerations (ACTU, sub. 34, p. 9; AIST, sub. 28, p. 3; AustralianSuper, sub. 19,
p. 1; REST, sub. 23, p. 7; Vision Super, sub. 4, p. 5). For example, ISA submitted that:
Members’ best interests could arguably be the sole criterion for evaluating selection processes,
since selection processes that are strong on the other criteria, but weak on members’ best
interests, would be difficult to justify. Moreover, systems that have high integrity, and low
system wide costs, for example, should be more efficient in securing members’ best interests.
(ISA, sub. 40, p. 38)
The Australian Council of Trade Unions submitted that it:
… continues to support the Members’ best interest (or the ultimate benefit to member) as the
principal criteria against which any assessment of alternate models (against the existing default
arrangement) should be made. The ACTU does not support ‘trade-offs’ as such in this
evaluation. In specific terms, maximising the net benefit to the member at their retirement is the
single, substantive goal which the system should be aiming to produce. (ACTU, sub. 34, p. 9)
There was also broad agreement among participants that serving members’ best interests
primarily translates into maximising long-term net returns (taking account of risk) as a key
input to delivering retirement income (Ai Group, sub. 21, p. 2; AIST, sub. 28, p. 20;
AustralianSuper’, sub. 19, p. 2; Kinetic Superannuation, sub. 45, p. 2; FPA, sub. 29, p. 3).
For example, AIST said:
Members’ best interests certainly includes a requirement that an allocated default fund can meet
the needs of the cohort of workers to be allocated to a fund. In determining the needs of the
employees, a clear focus should be on the expected long-term net returns to members. (AIST,
sub. 28, p. 20)
Additional indicators to flesh out an assessment of members’ best interests were also
suggested. For example, the range and quality of services provided (such as financial
advice and insurance cover), the level of transparency, and the value and culture of the
fund (AIST, sub. 28, p. 20; ISA, sub. 40, p. 38; REST, sub. 23, p. 8).
The Commission agrees that members’ best interests should sit above other model
assessment criteria. Hence, it will assess whether the model creates incentives for funds to
deliver default products that maximise member benefits — defined here as maximising
long-term net returns and effectively matching members to products that meet their needs.
In the Commission’s view, these outcomes are an input into (and consistent with) meeting
members’ best interests. In assessing this criterion, the Commission will consider the
following factors:
• whether the selection and allocation process provides sufficient incentives and
discipline on funds to focus on maximising long-term net returns to default members,
taking account of risks
• whether members are able to undertake meaningful comparisons across product and
service offerings in the market

DRAFT REPORT
53
• whether the selection and allocation process provides sufficient incentives and
discipline on funds to provide products which incorporate services (such as call centres,
online resources, and intrafund advice) that provide a minimum level of service quality,
but do not add unnecessary costs that may detract from net returns. Chapter 3 provides
further examples and guidance on what the Commission considers to be a minimum
level of service quality.
Criterion 2: Competition
Does the model encourage open participation (contestability) and rivalry between funds
(competition) that drives innovation, cost reductions and more efficient long-term
outcomes for members?
Competition may offer benefits beyond long-term net returns and members being allocated
to products which meet their needs. It is potentially a powerful tool to drive efficiency in
the superannuation system over time, including by providing incentives for funds to lower
their costs, improve their service quality and innovate. However, competition may not
always promote more efficient long-term outcomes for members, and can be promoted in
different ways (chapter 1).
This leaves the question of how competition is factored into an assessment of new
alternative models. This was a source of debate among inquiry participants. Several
participants expressed concern that including competition as a standalone criterion gave it
too much prominence, and confuses means with ends. The potential for competition to lead
to adverse outcomes for default members was raised, and if competition was retained as a
criterion, a common view was that it should be subservient to criteria focused on efficiency
or members’ best interests (for example, AIST, sub. 28, pp. 20–21; ISA, sub. 40).
Other participants supported competition as a standalone criterion. For example, FSC
(sub. 38, p. 7) suggested that competition should be one of the primary criterion used,
while Kinetic Superannuation (sub. 45, p. 2) submitted that competition drives cost
reductions and encourages innovation which in turn leads to long-run improvements in
efficiency. Some participants noted that barriers prevent certain funds from competing for
default members in the current system (for example, FSC, sub. 38, p. 31), and that a new
alternative model’s impact on competition should focus on the number and severity of
barriers to entry and exit in the market (for example, FPA, sub. 29, p. 3).
Probably the best way to consider whether competition is a relevant criterion is to consider
a system for default provision that offers none. In such a circumstance, reliance would
have to be made on actions from outside the super system to offer dynamic efficiency
improvements. It might be possible to envisage such a system at a point in time
maximising member benefits by a combination of well-informed benign intervention and
well-intentioned management. But over time and with many providers, variation in
performance by both regulator and manager is inevitable and experience suggests neither
are sustainable at such an exceptional level without an incentive to adapt to better ideas or

DRAFT REPORT
new needs. Competition provides an incentive to address such opportunities and needs; and
the degree to which it is present determines the timing and spread of such actions.
Innovation delayed is clearly a cost to members.
The Commission nevertheless agrees that competition in the superannuation system is an
intermediate objective insofar as it drives improved outcomes for members: lower prices,
better products, and improved choices. It will be important to identify the likelihood and
materiality of the potential risks and rewards of greater competition. That said, the need to
balance competition with other considerations is not unique to the superannuation sector —
for example, various provisions within industry-specific and general competition law
provide scope to balance and weigh the impacts of policy changes on competition,
efficiency and community-wide interests.10
In summary, the Commission considers that it is relevant to retain competition as one of
the five assessment criteria. In assessing the merits of each model, the Commission will
assess the extent to which it can harness the benefits of open participation (contestability)
and rivalry between funds (competition) to drive innovation, cost reductions, and improved
long-term outcomes for members — this will feed directly into an assessment of members’
best interests. More specifically, the Commission will consider the following factors:
• the type of competition that the model seeks to inject into the process for allocating
default members to products (that is, whether it creates ‘competition in the market’ or
‘competition for the market’, or both) (chapter 1)
• the extent to which the model enables the efficient entry and exit of funds, and provides
funds with an equal opportunity to be assessed as a default provider (provided they
meet the necessary minimum standards)
• whether the model provides incentives for funds to develop more innovative products
and services over time (which meet default members’ needs) and reduce their costs
with a strong focus on benefits to their members
• whether the model is likely to require or motivate funds to provide meaningful and
comparable information to enable members to make well-informed decisions and
undertake meaningful comparisons across product and service offerings in the market
• the extent to which there are likely to be spillover benefits or costs in terms of the
products and services provided to existing default and non-default members in the
system.
In its assessment, the Commission will remain mindful of potential trade-offs between
competition and other criteria, such as member benefits and system stability.
10 For example, Part VII and Part XIC of the Competition and Consumer Act 2010 (Cwlth).

DRAFT REPORT
55
Criterion 3: Integrity
Does the model promote a high-degree of integrity in the selection and delivery of default
superannuation products, and the ongoing behaviour of superannuation funds more
broadly?
Members’ best interests are likely to be promoted by a default system which has and
promotes a high degree of integrity among its participants. A new alternative model for
allocating default members will invariably involve some form of filtering or selection
process (even if it is minimal or involves ‘nudges’). On the supplier side, funds
participating in any process will have more information about their products, services and
future intentions than the agent responsible for filtering or selecting products and members.
This introduces the risk that commitments by funds during the selection or filtering process
do not materialise later. Given the significance of the default sector, the risk that funds
could seek to coordinate their behaviour to manage (or game) the process needs to be
considered. Some models may be more prone to these risks than others. These risks would
work against the best interests of members.
Defaults involve someone else (for example, a fund, employer or another agent) choosing a
superannuation product on behalf of an employee, which can introduce ‘principal–agent’
problems. The integrity of the decision maker involved in this process is therefore an
important consideration. Qualities such as relevant expertise and capabilities,
understanding members’ perspective and needs, freedom from conflicts of interest,
transparency and accountability would add to the integrity of the process, while issues such
as poorly managed conflicts of interest or a decision maker being captured by the supplier
industry would erode it (CIFR, sub. 7, p. 4; ISA, sub. 40, p. 39).
A criterion focused on integrity was broadly supported by inquiry participants (for example
Ai Group, sub. 21, p. 3; FPA, sub. 29, p. 3), although views differed on the specific matters
to be taken into account. At the selection stage, risks associated with: a lack of knowledge
and skill (CIFR, sub. 7, p. 4); regulatory capture (ISA, sub. 40, p. 39); conflicts of interest
(ISA, sub. 40, p. 39); political influence (FSC, sub. 38, p. 24); and a lack of transparency
and accountability (CIFR, sub. 7, p. 4) were highlighted. In the delivery stage, a consistent
theme was that any model would need to have appropriate transparency, governance and
accountability mechanisms in place to ensure default providers are held to account for their
commitments (AIST, sub. 28, p. 21; Kinetic Superannuation, sub. 45, p. 2) and compliance
with legal obligations, such as the Superannuation Guarantee (AustralianSuper, sub. 19,
p. 3).
However, participants also delivered words of caution. The FSC (sub. 38, p. 17) suggested
that the integrity criterion proposed in the Commission’s issues paper seemed
inappropriately designed to only consider auction metrics. Other participants warned
against over-engineering the regulatory arrangements that would underpin any new
alternative model given that strict governance arrangements are already in place to govern
fund activities in the superannuation system (for example, Kinetic Superannuation, sub. 45,

DRAFT REPORT
pp. 2–4; REST, sub. 23, p. 13); although such responsibilities self-evidently do not extend
beyond an individual fund.
In its assessment of new alternative models, the Commission will consider whether each
one promotes a high degree of integrity in the selection, filtering, allocation and delivery
process. A lack of integrity in these areas could affect confidence in the system more
generally. More specifically, the Commission will consider the following factors:
• whether agents responsible for making decisions about default products have
appropriate incentives, relevant expertise and capabilities, and whether conflicts can be
effectively disclosed and managed
• the extent to which the selection process is transparent and consistent with any
pre-process guidance that is in place
• whether and how funds are held to account for commitments they make during the
process
• the extent to which the model provides incentives for participating funds and other third
parties involved in the process (such as employers and advisers) to act in good faith and
in the best interests of all their members (current and prospective)
• the extent to which the model incorporates new regulations to address integrity risks
• the extent to which the model exacerbates peer risk in the system (PC 2016). (Peer risk
may also be a factor considered in the context of assessing member benefits and system
stability depending on the model being assessed.)
The Commission recognises there is a potential trade-off between introducing new
measures to promote integrity under a new alternative model and the additional costs of
regulation (which could flow back through to members as higher fees) and costs of
administration on government. The system-wide costs criterion will take into account the
costs of additional regulation.
Criterion 4: Stability
Is the model likely to create instability in the superannuation system that leads to
significant systemic risks?
The stability of the superannuation system was identified as a core objective in the
Commission’s study to develop criteria for assessing the efficiency and competitiveness of
the system (PC 2016). Given the system is growing in size, and is linked to the financial
system and the real economy through a multitude of channels, there is potential for
systemic risks to move across the system. The default segment represents a large
component of the system — for example, total assets in MySuper products (a proxy for the
size of the default segment) were $474 billion as at June 2016, 23 per cent of the about
$2 trillion in the superannuation system (APRA 2017c, pp. 11, 24), and importantly, about
two-thirds of fund members (chapter 1).

DRAFT REPORT
57
There was broad support for inclusion of a criterion focused on system stability (for
example, Ai Group, sub. 21, p. 3; APRA, sub. 33, p. 1; ISA, sub. 40, p. 39), and regard to
whether a model promotes or detracts from system stability (Kinetic Superannuation,
sub. 45, p. 6). The need to consider stability from the perspective of members, employers
and the broader community (including government) was raised (AIST, sub. 28, p. 22;
REST, sub. 23, p. 13). From members’ perspective, the stability of the system was seen as
a means of fostering confidence and trust in APRA-regulated funds. From funds’
perspective, greater certainty of net cash flows provides them with scope to take a
long-term approach to investment (AIST, sub. 28, p. 22; ISA, sub. 40, p. 39).
A stable superannuation system that supports efficient outcomes will not be one without
change. Some concerns raised by inquiry participants about the impact of new alternative
models on stability go to realities that are synonymous with the cut and thrust of any
genuinely competitive and dynamic market environment — such as movements in market
share for individual firms or inefficient firms exiting or being taken over. And the
continuing opportunities for further exploiting economies of scale apparent in
superannuation should not be impeded by confusing the number of funds with system
stability. These are not systemic risks to the system more broadly. Further, while additional
regulation is also often argued as a factor creating uncertainty and working against system
stability, well-designed changes to the regulatory environment can promote as much as
reduce system stability.
In summary, the Commission does not see churn per se, and the exit (or takeover) of funds,
as a sign of stability concerns. Indeed, churn and the exit of underperforming funds is often
a sign of healthy competition that should deliver better long-term net returns. Nonetheless,
chapter 3 will consider how the design of new allocation models could influence the
structure of the market within which funds operate and the incentives for mergers between
funds.
In assessing this criterion, the Commission will focus on problems or risks arising which
have the potential to develop into significant systemic problems. For example, it will be
relevant to consider whether any of the models are likely to cause (APRA, sub. 33, pp. 2–
3; ASFA, sub. 24, p. 3):
• excessive and volatile movement of members and assets between funds, thereby
exacerbating liquidity risk and peer risk in the system
• funds to adopt unsustainable loss leading strategies that could cause broader and
significant system disruption in a later period
• excessive market concentration and interconnectedness between upstream service
providers and funds in the system that significantly reduces rivalry in the market.
This criterion is framed as a ‘negative test’ — in other words, is a new alternative model
likely to compromise stability? The Commission will consider the extent to which a model
may create systemic instability based on the factors listed above.

DRAFT REPORT
Criterion 5: System-wide costs
Does the model minimise overall system-wide costs, taking into account costs on members,
employers, funds and governments?
A new alternative model has the potential to introduce costs into the system (such as new
administration, marketing, regulatory or compliance, transaction or search costs) and will
likely change the costs of participation and their incidence. However, it may also reduce
costs in other areas, such as lower search costs for members or a lower burden associated
with raising taxes to fund the Age Pension over the long term (as a result of members
being in higher performing products). In considering the merits of new alternative models,
it is relevant to consider who bears increased or reduced costs, and the overall impact on
system-wide costs (in other terms, the incremental economic cost impact on the
community compared to the baseline).
There is potential for an alternative model to result in transfers between parties. For
example, a model based on employer choice would elevate costs to employers while
reducing the costs to members who will expend less resources in searching for products
and could be guided to a better outcome. A new alternative model may also add to a
party’s costs in one area but reduce costs in another. For example, a model that involves a
selection process would add to funds’ costs (as they devote resources to participating) but
could save them money on future marketing expenses if they are successful in the process.
Inquiry participants recognised the relevance of examining the total system-wide costs of
introducing a default model with a view to comparing this feature across models (for
example, FPA, sub. 29, p. 3). There was a view that the Commission should take a broad
lens in applying this criterion. For example, ISA (sub. 40, p. 39) argued the cost criterion
should focus on efficiency, while AIST (sub. 28, p. 22) suggested that the focus should be
on net costs across superannuation and retirement income systems. The Ai Group (sub. 21,
p. 3) cautioned the Commission not to double-count costs already included in the
assessment of net returns, but recommended a focus on costs borne by employers and
governments in administering the models.
A model that minimises total system-wide net costs is clearly in the interests of the broader
community, and in most cases, members’ best interests, however, there may be instances
of trade-offs. For example, minimising the resources devoted to a filtering or selection
process may be consistent with minimising total system-wide net costs, but may come at
the expense of integrity in the selection and delivery phase.
In its assessment of each new alternative model, the Commission will examine the impact
that the introduction of the model would have on costs borne by parties within the system
overall. The balance of these impacts will be difficult to discern with precision and will
require judgment, which will be informed by the available evidence and feedback from
participants. The Commission will not be attempting to calculate the dollar cost of different
models. The focus of the qualitative assessment will be on the following factors:

DRAFT REPORT
59
• what is the likely costs impact on members, funds, employers and government, and the
materiality of any changes from the baseline?
• is the model likely to lead to overall higher or lower system-wide costs compared to the
baseline? What is the expected materiality of any changes from the baseline?
In considering this criterion, the Commission will consider the impact of cost transfers
between parties, the potential for costs to be passed through to other parties, and will avoid
double counting benefits or costs from a system perspective. For example, a model that
results in better member outcomes may reduce government spending on the Age Pension
(compared to the baseline), but some component of these cost savings will be reflected in
increased member benefits.
Some participants suggested the Commission should take into account the risks and costs
associated with transitioning to an alternative model from the current system (for example,
AIST, sub. 28, p. 5). As the Commission is not comparing these models to the current
system, these types of transition costs are not relevant to this study, although it is expected
that they will be an important consideration in stage 3.
2.3 The no defaults baseline is not a model
The assessment of new alternative models requires a common reference point for
comparison. In its issues paper, the Commission proposed a baseline of no default
arrangements. This proposal elicited a substantial amount of participant feedback.
Some participants indicated that the Commission’s ‘greenfields’ approach of no defaults is
a reasonable reference point of comparison for new alternative models (CSA, sub. 35, p. 4;
PwC, sub. 12, p. 3). However, most participants were critical of a no defaults baseline.
Broadly, criticisms fell into two categories:
• a no default baseline should not be the Commission’s preferred model as it would not
meet members’ best interests (for example AustralianSuper, sub. 19, p. 1; First State
Super, sub. 26, p. 2; IFAA, sub. 13, p. 6), at least in the short term (FSC, sub. 38, p. 3)
• a no default baseline lacks analytical rigour, it will be difficult (if not impossible) to
operationalise in a meaningful way given there is no real-world evidence to assess it,
and it could yield misleading results (for example, AIST, sub. 28, p. 13; ISA, sub. 40,
pp. 8–9).
Several participants argued that it would be more appropriate to use the current default
arrangements as the basis for comparison (for example, ACTU, sub. 34, p. 6; AIST,
sub. 28, p. 4; ASFA, sub. 24, p. 2; ISA, sub. 40, p. 9; UniSuper, sub. 20, p. 3).
The Commission’s intended application of the no defaults baseline in the context of this
inquiry appears to have been misunderstood by some participants. The Commission
adopted the baseline as the only objective and feasible reference point against which to

DRAFT REPORT
assess new alternative models. Use of this baseline will ensure that the models examined
do not merely offer benefits (or costs) above those of the present system but instead
actually address what a default system should be aimed at delivering today, after nearly 25
years’ experience with a major compulsory allocation system of increasing complexity and
significance. The absence of any system is a logical way of exposing what would be lost
without defaults; and allows consideration of why that is important and whether there are
other ways (for example, outside the system itself) to address that.
Moreover, use of a no defaults baseline exhibits a refreshing absence of ideological
preference and is about as objective a benchmark as possible. Finally, as a baseline aimed
at solving today’s problems, it is substantially more oriented towards the future than a
baseline that represents the combined best past efforts (wise or foolish) to address past
problems and perceptions. Thus the no defaults baseline is not a new alternative model,
although at times the Commission will test it in order to see if an apparent problem it
creates is substantive or not.
In stage 3 of the Commission’s work on superannuation, the present system can be
considered against potential new alternative models, and indeed against the Commission’s
objective baseline. But it will be its ability to deliver on today and future needs that will
matter.
Applying a no defaults baseline is not without challenges. Inquiry participants raised valid
criticisms and questions on how it would be meaningfully applied to inform the evaluation
of new alternative models, and what evidence would be drawn on given it is a hypothetical
scenario (for example, AIST, sub. 28, p. 13; ISA, sub. 40, pp. 8–9). Of course this problem
of hypotheticals applies to all options (including the untested expert panel process within
the FWC), other than advocacy for the incumbent system.
The Commission accepts that in order to apply the baseline of no defaults it needs to
provide further explanation on how it will be incorporated into the evaluation of new
alternative models. As such, chapter 4 provides a detailed explanation of the evidence and
assumptions the Commission has used to inform the baseline of no defaults.

APPROACH TO DEVELOPING ALTERNATIVE MODELS
DRAFT REPORT
61
3 Approach to developing alternative
models
Key points
• In developing its models, the Commission sought to design a workable system for the
future, and in doing so, overcome some of the legacy problems affecting the system today.
• This report presents four alternative models: assisted employee choice, assisted employer
choice (with employee protections), a multi-criteria tender, and a fee-based auction.
• The current system is plagued by the issue of members holding multiple accounts. This is
largely a product of current policy settings, whereby members are automatically allocated a
new default account every time they change jobs and do not exercise choice. Account
duplication erodes member balances and imposes costs on other system participants.
• The Commission in principle supports a recommendation by the Financial System Inquiry
that employees should be allocated to a default product only once — upon first entering the
workforce — and retain this account for future jobs unless they choose to switch.
− While this would be a smaller pool of defaulting members (some 400 000 members
each year) to be allocated under the Commission’s models, it may prove relatively
attractive to some funds and not others.
• A fund that wins default status for contributions of new members should be required to offer
the revised winning fees and service terms to all current members of its default product.
This would diffuse the benefits of competition to the broader system.
• The default product should focus on the accumulation stage and consist of services that are
the minimum necessary to meet the objective for the superannuation system. The bundle
should include investment, administration of member accounts and intrafund advice.
− The bundled insurance product would not be considered in the selection of products.
Funds would be assessed on their nominated accumulation product, subject to meeting
a threshold level of service for bundled insurance.
• The body responsible for selecting eligible default products must have an explicit focus on
member interests and be accountable for its decisions. The body must be free of conflicts of
interest, display high integrity and have sufficient financial expertise to evaluate products.
• There is strong merit in establishing a centralised online service to assist employees and
employers in carrying over existing accounts between jobs, and to inform the Government
about member decisions on opening, closing or consolidating their accounts.
− The ATO is well placed to undertake this task, as it has to issue a Tax File Number to
new entrants to the workforce, and operates the Single Touch Payroll system.
• There is merit in immediately establishing a clear, transparent and accountable framework
to guide trustees when proposing or considering mergers. A key element would be an
obligation to disclose all mergers considered involving their fund, as well as the reasons for
the ultimate outcome.

DRAFT REPORT
This chapter sets out the framework and some common foundations for developing the
alternative models presented in subsequent chapters. Section 3.1 outlines the
Commission’s approach to specifying those alternative models. Section 3.2 details which
employees would be covered by the allocation models. Section 3.3 draws on international
experience and past Commission work to identify broad features of the default product at
the centre of the allocation models. Finally, section 3.4 identifies other common features of
the allocation models, such as the frequency of the process and the characteristics of the
body responsible for making the final decisions.
3.1 Specifying alternative models for evaluation
The issues paper for this inquiry identified several steps for developing an allocation model
and set out the spectrum of options for most of those steps. At a high level, the models
differ on two aspects:
• the type of selection mechanism for default products
• the mechanism and decision maker for allocating members to those products.
However, there is a broad range of options within those variables and evaluating every
possible model permutation is neither tractable nor necessary. For ease of analysis, the
Commission has formulated a discrete set of model types for assessment.
How did the Commission select its alternative models?
In narrowing the focus to specific model types, the Commission was guided by several
principles. First, as required by its terms of reference, the models had to fit the description
of an alternative model — that is, they needed to be different from the current
arrangements and from each other. Second, the models had to be comprehensive in
covering the options across the spectrum of the Commission’s framework. When the
models are taken in combination, no design element from the framework is left
unexplored. Third, as discussed in chapter 2, the models needed to be workable and
consistent with the terms of reference. To determine whether a hypothetical model would
be workable, the Commission drew on international examples, in addition to analysing
Australian institutional settings and market characteristics. Several approaches have been
excluded from consideration due to being inconsistent with the terms of reference or for
being manifestly unworkable under the current institutional settings (discussed in the next
section). Consequently, this report presents and evaluates four types of allocation models.
• Assisted employee choice — this model leverages the competition benefits that arise
when members exercise choice, but with stronger and formal protections to assist
members make informed choices.
• Assisted employer choice (with employee protections) — this model injects
competition by giving employers choice in selecting a default product for their

DRAFT REPORT
63
employees, as long as the default products meet some minimum standards, while also
simplifying choices for small and medium-sized businesses.
• A fee-based auction and a multi-criteria tender — these models incorporate a
market-based mechanism into the selection of default products, with sequential
allocation of members among winning products (figure 3.1).
Figure 3.1 The building blocks for the four allocation models
Dashed lines indicate where use of filtering is voluntary.
For each of those models the objective was to specify a design that would perform best
against the assessment criteria. The Commission took an iterative approach to refining the
models based on its own research and analysis, as well as participant feedback. Finally,
having specified a ‘best-in-class’ design for each category, the Commission evaluated the
models against the assessment criteria proposed in chapter 2, relative to the baseline of a
world of no defaults.
System stability was a threshold consideration for model choice and design
As discussed in chapter 2, the Commission interpreted the stability criterion as a threshold
consideration for its models. That is, all of the models presented in this report have been
designed to avoid generating material risks to the superannuation system.
More specifically, the Commission targeted two types of stability concerns:
• not generating volatile and excessive movement of members and assets between funds
• avoiding unsustainable strategic behaviour by funds.
On the first issue, the key considerations are the size of the pool of default members (both
new and existing) to which the models apply, as well as the flow-on implications for other
members. A further consideration is how the models accommodate members who are in
products that lose default status. Those matters are discussed in this chapter.
Degree of
filtering
Type of
filtering
None High Very high
Who
allocates?
Low
Baseline
1 Assisted
employee
choice
2 Assisted
employer choice
(with employee
protections)
4 Fee-based
auction
3 Multi-criteria
tender
None AuctionHeavy filter
Employer GovernmentEmployee
TenderMinimum standards

DRAFT REPORT
The Commission has addressed the second issue through model design — in particular,
avoiding excessively frequent repetition of the competitive process (discussed in this
chapter) and various safeguards within the models themselves (discussed in chapters 6–8).
Several participants also raised the issue of stability risks to the system from creating an
excessively concentrated and interconnected market between funds and upstream suppliers
(chapter 2). The Commission was less concerned about its models creating those types of
risks. There are too many funds and default products in Australia, many of which are
subscale, and some consolidation would be beneficial (APRA, sub. 33). Moreover, other
features of the Commission’s approach, such as its proposed narrowing of the definition of
new default members and regular repetition of the competitive process, safeguard against
excessive market concentration.
Which models have been excluded from the outset?
A single sovereign fund model is not an option for this inquiry
The Commission will not be evaluating models that involve a single or dominant sovereign
fund that receives default contributions by fiat.11 This structure of the pension system is
quite common internationally (for example, in France, Sweden, Norway and Japan). Some
of these models imply or specify a government guarantee. A government guarantee is not
inherent in the Australian system and this is a desirable position to retain. It is also
inconsistent with the focus on a competitive process, sought by the terms of reference.
The models that focus on the wholesale level of the market will not be considered
In Australia, the superannuation system is characterised by a degree of competition at both
the retail and wholesale levels of the market (PC 2016). Funds either insource specific
investment and administration services or outsource them to wholesale providers, and then
deliver a service bundle (the accumulation product) to the member. In some international
systems the competitive process is targeted at the wholesale level. This is typically the case
in centralised systems (for example, in Sweden and the United Kingdom), where a single
public trustee procures services from a pool of investment managers (appendix B). The US
Thrift Savings Plan for federal government employees is another example (box 3.1). A
review of the Swedish pension system found that there are cost savings in decoupling
administration from investment and centralising the former in the hands of the government
(Barr 2013, p. 75). Some participants proposed that the Commission should consider
models that involve a government-run competitive process at the wholesale level. This
could be done either in combination with a process at the retail level or instead of it, as in
the centralised systems referred to above.
11 This is distinct from a model discussed in chapter 5 that involves a temporary ‘last-resort’ fund as a
protection against members forfeiting their contributions due to failing to nominate a product in the open
market.

DRAFT REPORT
65
Box 3.1 The US Thrift Savings Plan
As at December 2015, the US Thrift Savings Plan had over US$458 billion in funds under
management. It is available for all federal employees, who are automatically enrolled when they
begin work. The default employee contribution rate is 3 per cent of the salary, and employers
contribute a further 4 per cent. Employees can opt out at any time, and if they reduce their
contribution rate the employer contribution will also fall. Individuals can invest their money in
one of five life-cycle funds or manage their own account by choosing an investment mix from
five individual funds: a government securities fund, a fixed-income fund, a common stock fund,
a small capitalisation fund and an international stock fund.
Funds must be managed passively, and thus the individual funds aim to track specific
asset-class indices. Management of four individual funds is tendered for and managed
externally (the government securities fund is managed in house). The Thrift Savings Plan uses
an internal procurement team to operate this tender process.
Sources: FRTIB (2016); TSF(2015); White (2011).
The Commission has decided not to assess such a concept at this stage of the inquiry.12
The explicit focus of this inquiry is to inject competitive pressure at the trustee (retail)
level. It is at this level where, despite the existence of some member choice, there are
fundamental challenges for members to make an informed and rational decision
(chapter 1). Centralised tendering for wholesale services targets competition in the
wholesale market (investment and administration services), which does not have the same
issues of disengagement on the demand side. Researchers have argued that trustees already
place strong pressure on service providers. For example, Chant West (2015, p. 1) stated:
We contend that, while consumers (i.e. fund members) are not driving competition, the funds
themselves and their service providers are. Every day, in the course of our business, we observe
funds and providers competing fiercely, and we believe this ensures competitive pricing.
The remainder of the chapter specifies some common foundations for the models
considered in this report. In particular, it describes the market that is at the centre of all of
the models, from the perspective of members and providers. In looking at the relevant
variables the Commission has examined selected features of default private pension
systems in other countries (table 3.1).
12 That said, the Commission may still revisit particular aspects of the Swedish and US Thrift models in the
context of its stage 3 inquiry.

66
Table 3.1 Selected features of international default systemsa
Australia New Zealand United States United Kingdom Sweden Chile Canada
Compulsory? Yes No (opt out) No (opt out) No (opt out) Yes Yes Yes
Life stages Accumulation Accumulation Accumulationb Accumulation All Accumulation All
Scope of
coverage
Universal,
segmented by
modern award
Universal, subject to
some restrictions
Universal eligibility,
segmented by
employer
Universal, subject
to some restrictions
Universal Universal Universal, subject to
some restrictions
Definition of
new default
member
New job and not
exercising choice
First job and no active
employee or
employer choice
Automatic enrolment
by employer
New job and not
exercising choice
First job and not
exercising choice
First job First job; employees
have no choice
Market structure
and service
provision model
Employer chooses
fund as default
provider of the
bundle of services
for its employees;
funds insource or
outsource services
Employees
sequentially allocated
to an eligible fund that
provides a bundle of
services; funds
insource or outsource
services
Employer chooses
fund as default
provider of the
bundle of services for
its employees; funds
insource or
outsource services
Employers choose
public trustee as
default provider of
service bundle;
public trustee
procures services
from private sector
Single product
provided by public
agency; in-house
administration and
mix of insourcing
and outsourcing of
investment
Single default
provider (winner of a
tender) offering a
bundle of insourced
or outsourced
services to
employees
Single product
provided by public
agency; in-house
administration, and
investment by
independent statutory
board
Services
bundled in
default
Administration,
investment, intrafund
advice, life and
TPDc insurance
Administration,
investment, member
education
Administration,
investment; can
include employee
advice and education
Administration,
investment,
intrafund advice
Administration,
investment,
intrafund advice,
pension
Administration,
advice, life and
TPDc insurance
Administration,
advice, life and TPDc
insurance, pension
Centralised
administration?
Largely
decentralised
Centralised clearing
house
Largely decentralised Centralised Centralised Decentralised Centralised
Prescribed
investment
approach?
No prescription;
single option for
members (life-cycle
or diversified)
Mandatory low risk
portfolio; single option
for members
No prescription for
employers; single
option for members
Single target date
option for members
Passive or active
investment; single
option for members
(life-cycle)
Passive investment;
five risk-based
investment options
Passive and active
investment over long
term; largely defined
benefit for members
a The international systems covered in the table are the NZ KiwiSaver, the US 401(k) auto-enrolment, the UK NEST, the Swedish AP7, the Chilean default AFP, and the Canada
Pension Plan. b Retirement income products are not directly provided by 401(k) providers but can be purchased by them for their members. c Total and permanent disability.
Sources: Appendix B (for Chile, NZ, Sweden, UK systems); IRS (2017); US DOL and IRS (2013); CP Investment Board (2016); Government of Canada (2016).

DRAFT REPORT
67
3.2 Which employees will the models cover?
A fundamental design consideration, which is common to all models, is what employees
they will ultimately cover and how.
Employee choice would remain
The Commission has adopted the definition of ‘default products’ as products to which
employees are allocated if they fail to exercise their right to choose a product. None of the
models presented in this report mandate that employees join or remain with a particular
default product and it is envisaged that employees will have the right to choose their own
product, and switch products after joining.
This interpretation is largely consistent with the Choice of Fund requirements legislated by
the Australian Government, notwithstanding some constraints on the coverage of those
provisions (discussed below). It is also consistent with international practice, and generally
accepted by industry participants in Australia, despite some disagreement on the merits of
active member choice. Beyond that, there are three aspects to the employee coverage issue:
• the models’ coverage of the flow of employees — specifically, the definition of a new
default member and the trigger point for being allocated to the default product
• whether and how the models should apply to the stock of existing default members
• the institutional constraints on the application of the models to some members.
Who is a ‘new default member’?
Under the current arrangements, an employee typically acquires a default product when
they enter the workforce for the first time, and thereafter would get a new account every
time they change jobs and do not actively choose a product. The current defaulting
structure leads to unnecessary account proliferation and erosion of member balances.
This structure is a historical consequence of the origins of superannuation in industrial
relations traditional defined benefit schemes. It is essentially an outcome of not providing
adequate infrastructure to support individuals following the introduction of a compulsory
defined contribution system. The purpose of this inquiry is to design allocation models for
the future, and this includes revisiting legacy policies that might no longer be in the best
interests of members. In this context, the Financial System Inquiry (FSI) (Murray et
al. 2014) recommended that employees should be allocated to a default product only once
— typically when they first join the workforce — and should subsequently retain that
account unless they choose a different fund. The Commission has conducted a preliminary
assessment of this proposal.

DRAFT REPORT
Evaluating the merits of the proposed definition of a new default member
Account proliferation is a significant issue, notwithstanding some recent evidence of
consolidation activity.13 Available data for 2015 indicate that there were 2.2 accounts for
each member of the system. Well over 40 per cent of members retain multiple accounts
(ATO, pers. comm., 29 September 2016; 2016c) There is also evidence of members
holding duplicate insurance (PC 2016, p. 150).
Costs of account proliferation
Some members would hold multiple accounts deliberately for various reasons (for
example, retaining an account in a corporate fund or a defined benefit scheme). But the
majority likely have multiple accounts due to the operation of the current default system
that potentially triggers the creation of a new default account with every change of job.
Multiple accounts are a source of significant inefficiency in the system. Account
proliferation erodes economies of scale in member administration and reduces member
balances through duplicated provision of services. And duplicate insurance policies
(particularly where the member can only claim under one policy) are a source of
inefficiency in their own right. Account proliferation is also likely to be regressive and
have the greatest effect on those who are least well off. Employees that work multiple
part-time or casual jobs and in the course of their lives have acquired a large number of
accounts with low balances in each would be particularly affected by paying the fixed costs
of maintaining those accounts.
The Australian Chamber of Commerce and Industry (sub. 27, p. 11) argued:
If it is accepted that account proliferation across and within funds is systemically undesirable,
any replacement default allocation mechanism should not operate to allocate a default fund for
an eligible employee with each new job. This suggests that, in principle and subject to the
employee exercising choice at any time, an eligible employee should bring his or her allocated
fund into his or her second and subsequent jobs.
There are significant gaps in the data that preclude an accurate estimate of ‘unintended’
account proliferation (box 3.2). These data gaps are a policy problem in their own right,
and warrant attention by the Australian Government (discussed in section 3.4). However,
in all likelihood the benefit from reducing the proliferation of member accounts and
duplication of costs charged to members is significant. For example, the Grattan Institute
(Minifie, Cameron and Savage 2015) estimated that removing about 12 million excess
accounts from the system would save about $360 million in administration costs per year.
The FSI (Murray et al. 2014, p. 110) estimated that moving from an average of 2.5
accounts to a single account over a person’s working life could increase superannuation
balances by about $25 000.
13 For example, between 1 July 2013 and 1 January 2017 about 1.4 million accounts with an aggregate value
of about $6.8 billion were consolidated through the Australian Taxation Office’s online system (ATO,
pers. comm., 28 February 2017).

DRAFT REPORT
69
Box 3.2 Estimating account proliferation — the data gaps
Several agencies collect a large volume of superannuation-related data from funds, employers
and employees. However, there are material gaps in the data that relate to member decisions
with respect to their accounts as well as the types of products they are in. This precludes an
accurate estimate of how many accounts have been opened unintentionally and are surplus to
member requirements.
Specifically, the Australian Prudential Regulation Authority (APRA) collects data on the stock
and flow of accounts within APRA-regulated funds by product type (choice or MySuper), but
there is no ready way to link this information to particular members across funds. For example,
the data would not allow a precise estimate of the overall number of members with MySuper
accounts in any given year and, consequently, a precise estimate of account proliferation in
MySuper.
The Australian Taxation Office (ATO) collects data on the number of members that open new
accounts, and on the number of accounts they open and hold. However, it does not collect data
on the types of accounts held by members, nor on whether those accounts were selected by
active choice or through default. There is also no way to attribute the reasons for members
opening new accounts, nor a comprehensive record of whether existing balances were
consolidated into a newly opened account.
For its estimate of the number of accounts opened under the first-timer scenario, the
Commission has relied on ATO data on the number of members that open their first ever
account every year. Data limitations have prevented the Commission from estimating the
number of accounts added to the system unnecessarily under the current job turnover
defaulting arrangements. Thus, the estimate of the benefit from reducing account proliferation
due to moving to the first-timer approach is only intended to be indicative.
Based on the Commission’s indicative estimates, the annual system-wide savings in
administrative fees and insurance premiums from moving to a first-timer default could be
of the order of $150 million for every 500 000 to 600 000 of default accounts permanently
removed from the system (see the annex to this chapter). There would also be time saved
by members who would need to manage fewer accounts. Furthermore, this approach could
mitigate ‘system leakages’ in the form of lost member accounts — an ongoing issue in the
superannuation system, despite significant recent work by the Australian Taxation Office
(ATO) to reconnect members with lost accounts (PC 2016, pp. 296–297).
DRAFT FINDING 3.1
The current arrangements where members can be defaulted to a new account on
every change of jobs increase the proliferation of accounts. This materially adds to
costs for members and reduces member balances at retirement.
The effect of the FSI approach on the incentives of funds to participate
Notwithstanding those benefits, there are implications for a competitive model in moving
from the current ‘job turnover’ pool of new default members to the FSI’s approach of only

DRAFT REPORT
targeting those who are first entering the workforce. What will ultimately matter is whether
the pool size would be sufficiently attractive to induce funds to participate in a competitive
process and offer additional benefits for members.
A preliminary estimate of the initial value of the first-timer pool is about $800 million per
annum (annex). In the short run, this value is substantially lower than that of the current
turnover pool. This is because the turnover pool comprises more people who are on higher
incomes than first entrants into the workforce, and because some of those people
consolidate their existing balances into the new account. The value of the first-timer pool is
also relatively small in the context of the aggregate flow of default contributions, which
amounted to $74 billion in 2016 (APRA 2017b). Large funds that would have to offer all
their current default members the flow-on of a successful bid (discussed below) might not
find it attractive to offer any improvements over their current default products.
However, the value of the first-timer pool of contributions is certainly not immaterial,
particularly when the contributions of each cohort are examined over the medium term. A
key reason for the initial small size of the pool is that the average income of members is
about $21 000 (which reflects the part-time or casual nature of the first job). Contributions
from each year’s first-timer cohort would grow strongly in the short term as those members
catch up to average incomes, and those members would be likely to remain in their original
default product for extended periods because there would be no automatic switching of
default products on changing jobs. Applying the average income of the turnover group to
the first-timer cohort yields a pool of annual contributions of over $2 billion per year. This
is on par with, or greater than the gross inflow of contributions for all but the largest 10
MySuper products in 2016 (table 3.2). The pool of young long-term members (figure 3.4 in
annex) is even more attractive in the context of the ageing member bases of many funds
and the net flow of contributions into default products. In 2016, only five MySuper
products registered net flows above $2 billion. Nearly 80 APRA-regulated funds —
accounting for almost a quarter of the assets in all APRA funds — recorded a net outflow
in 2016 (APRA 2017a).
Furthermore, attaining default status would also generate some flow-on ‘signalling’
benefits for the fund from being able to market to (and attract) some members outside of
the narrower definition of new default members.
Ultimately, moving to a first-timer definition of new default members is unlikely to
adversely affect competition between funds for that pool of members. Competition is not
solely contingent on participation by the largest funds, nor does it require the involvement
of many funds.

DRAFT REPORT
71
.
Table 3.2 MySuper assets and flows of the largest MySuper funds,
2016a
Average MySuper
assets
Average MySuper
contributions
Average MySuper
net flows
$b $b $b
Top 5 funds 32.9 5.1 2.5
Funds ranked 6–10 21.0 2.9 1.9
Funds ranked 16–20 7.0 1.2 0.7
a Funds are ranked on the level of gross MySuper contributions, not net flows or assets.
Source: Productivity Commission estimates based on APRA (2017a).
The Commission is seeking comment from the industry on the effect on fund
participation in the competitive process of moving to the first-timer pool of default
members.
• Are there any funds that might choose not to participate due to the lower value of
the pool, and what is the threshold minimum pool size (per fund) at which this might
happen?
• How would the move to a first-timer pool of default members affect the fees and
product features that funds would offer when competing for new members?
Other implications for efficiency
Adopting a first-timer definition for new default members could have other implications
for efficiency, although none of these appear insurmountable and most would be transitory.
As the target group of employees who would be the direct beneficiaries of the competitive
process is smaller, there is a risk that the gains from the competitive process would not
diffuse as quickly or as broadly within the superannuation system. However, to some
extent this should be addressed by a requirement that successful funds extend the same fees
and product terms to their existing default members (discussed below). Also, as noted
above, there would be some people who switch into the successful products voluntarily, as
funds market those products more broadly.
There may also be potential adverse effects for employees who, in the absence of other
actions would remain in their default fund for many years. Under the current arrangements,
being asked to select a product on changing jobs is an opportunity for employees to review
the suitability and performance of their existing superannuation product. However,
retaining the existing turnover default arrangements is an indirect way of encouraging
members to engage with their superannuation, and current evidence indicates that members
rarely review their product when changing jobs, in any case.

DRAFT REPORT
To avoid perpetuating the legacy problems of the current system, any future
alternative system for allocating members to default products should be premised on
employees being assigned a default product only once, when they join the workforce.
The need for additional infrastructure to support employees and employers
Additional infrastructure is needed to immediately address account proliferation and
critical data gaps. It would also be needed to facilitate the effective operation of the
alternative default arrangements. From an employer’s perspective, having access to an
online service that identifies the fund and account details of employees that already have
an account would simplify compliance with Superannuation Guarantee obligations and
reduce the risk of opening additional accounts. For employees, a service that simplifies the
carryover of their existing accounts when changing jobs would be similarly helpful. As
discussed in section 3.4, current and forthcoming initiatives by the ATO could be readily
adapted to the task.
How should the models accommodate existing members?
A well-designed competitive process would be expected to lead to lower fees, higher net
returns and other benefits for members. A threshold issue is whether these benefits should
apply only to those employees that had been directly defaulted into the product following
the competitive process, or extend to other members. For example, in Chile, a fund that
wins the fee-based auction for new default members must offer the same fee to its existing
default members (appendix B). The Commission considered two options for extending the
mechanism to existing members:
• the fees, terms and conditions of the winning product are extended to current members
of the fund’s default product
• the fees, terms and conditions of the winning products are extended to all existing
default members in the system.
Accommodating other default members of the winning fund(s)
As discussed above, extending the fees and terms of the winning product to other default
members of the fund could discourage participation by funds with a large stock of default
members that hold higher-margin products. Nevertheless, there is a strong case to adopt
this approach on fairness, simplicity and efficiency grounds. First, as noted, it would
accelerate the diffusion of the benefits of the competitive process across the system.
Second, it would avoid the problem of creating two classes of default members in a fund,
with some being in a superior product through sheer luck of timing. It is also likely to be

DRAFT REPORT
73
more consistent with a trustee’s fiduciary duty to act in the best interests of all members.
Third, it reduces the risk of unsustainable loss-leading behaviour by funds competing for
new default members, simply to grow market share.
Fourth, it is likely to be simpler to administer and navigate for all parties, and would also
be consistent with the motivations behind the current MySuper architecture (Cooper et
al. 2010). In the absence of a process to consolidate default members within a fund, any
allocation model could, over time, result in layers of default products and add to
complexity in an already complex system.
To ensure that the flow-on benefits reach the other default members of the fund, the new
product terms and fees would need to be extended immediately to everyone in the generic
default product of the fund.
Should default members in other funds be automatically moved to winning
products?
There may appear on first glance to be an equity rationale in favour of extending any gains
for new default fund members to members in all other default products, regardless of
whether their fund won a contest or not. They are, after all, default product members too,
who have been guided into their current (now relatively worse) position by the
government.
However, on closer examination, there is little merit in automatically moving the entire
system’s pre-existing default product members to the new winning product/s at the end of
each iteration of the selection process. The implications of this are very large and mostly
negative. In effect, this would make the product selection process a trigger point for
reallocating all default members to new default product(s). The pool for which funds
would be competing would be the contributions of all members in default products — over
$74 billion in 2016, and growing (APRA 2017a).
This approach would likely significantly increase member churn with adverse
consequences not just for liquidity but for system stability. As noted earlier, the
Commission has eschewed options that could adversely affect system stability. The
benefits of any lowering of fees or improved service quality would be swept away should
such a scenario emerge. The Commission agrees with the Australian Institute of
Superannuation Trustees (sub. 28, p. 33), which argued:
Existing members should remain in their fund until they either exercise choice, provided the
relevant fund remains entitled to receive default superannuation payments.
Nonetheless, to facilitate broader diffusion of competitive pressure, all of the models need
to accommodate voluntary switching by members into the successful products. This is
consistent with current policy. As a general principle, successful products should be open
to members of other funds.

DRAFT REPORT
DRAFT FINDING 3.2
There are strong grounds for requiring a fund that wins default status for contributions
of new default members to extend the same fees and service terms to all its existing
members of the default product.
Dealing with members in products that lose default status
A fundamental long-term issue for all allocation models is how they deal with default
members of a fund if that fund’s product loses default status or is unsuccessful in a future
iteration of the competitive process. The question is particularly important where a default
member is only allocated to a product once. The issue has not arisen in Australia to date,
and no regulatory or policy guidance has been developed yet to clarify the implications for
the fund and its existing members if and when it would.14 There are two interrelated
aspects to the question:
1. what should happen to the accumulated balances and future contributions of existing
default members of the fund?
2. if the fund retains its existing default members, should there be any safeguards to
protect those members when the fund is no longer bound by its original default
contractual or regulatory obligations?
There is a spectrum of options for addressing the first question, with automatic transfer of
members to another product being the most far-reaching. In general, to foster system
stability, automatic transfer of members should apply only in exceptional circumstances.
This would involve situations where a fund is wound up or deregistered as a
superannuation entity, or fails to meet mandatory requirements for a default product (as
discussed in chapter 8). This would generally be expected to be a rare occurrence.
For situations where a fund is unsuccessful in a future round of the competitive process,
the best approach is making this information publicly available, and leaving the decision to
members.
With respect to the second question, the key risk is of strategic behaviour by funds seeking
to capture a pool of disengaged default members and subsequently raising fees or reducing
product quality. The magnitude of this risk with a repeated process over an extended
timeframe is unclear. As discussed in chapters 6–8, to some extent it could be addressed
through fund selection criteria (for example, focusing on past history of performance and
member satisfaction) and other features of the allocation model. However, it is challenging
to eliminate the risk entirely. It is neither practical nor desirable to require funds to commit
14 The Deputy Chairman of APRA has recently indicated that APRA is focusing on ensuring that trustees
are proactively considering their strategy, and where necessary, ‘planning for a (hopefully graceful) exit
from the industry’ (Yeates 2015).

DRAFT REPORT
75
to a particular level of fees or product quality in perpetuity. Thus, much of the task would
ultimately fall to adequate and well-enforced (ex-post) regulation of trustee duties to act in
members’ best interests.
In sum, the preferred approach to the above issues would generally avoid automatically
transferring members and would focus on transparency and ongoing prudential and
consumer protection regulation. The allocation models in subsequent chapters specify
particular safeguards in this context.
Institutional constraints on coverage
The scope of coverage of the current default system is governed by a complex interplay of
various Commonwealth and State legislation. At the Commonwealth level, the two key
pieces of legislation are the Superannuation Guarantee (Administration) Act 1992 (Cwlth)
(SG Act) and the Fair Work Act 2009 (Cwlth). The SG Act requires employers to offer
choice of fund and provide a default fund, and the Fair Work Act establishes the current
process for selecting default funds. The two statutes differ in their general coverage, with
the SG Act having broader application. Each of those Acts also provides a range of
exemptions.
Specifically, the SG Act is binding on all jurisdictions, but provides exemptions to the
choice of fund requirement for:
• situations where an enterprise agreement specifies a fund
• Commonwealth defined benefit schemes, such as PSS and CSS, as well as unfunded
public sector schemes
• a number of ‘prescribed’ State Government schemes (currently 18) for state and local
government employees (s. 32C of SG Act).
Not all of those exemptions are an immovable constraint on the Australian Government in
implementing a national model for allocating members to superannuation products. For
example, there are no apparent constitutional barriers to removing the ‘deemed choice’
exemption for enterprise agreements.15 There might be constitutional constraints on
extending the model to prescribed and defined benefit State schemes. However, even in
those cases, most jurisdictions have voluntarily offered choice to their public servants,16
and this segment of the system is also generally closed to new members.
On the other hand, the Commission estimated that in 2015, the Fair Work Act covered
about 71 per cent of workers (and 85 per cent of employees) (PC 2015b, p. 78). Most State
15 The Superannuation Legislation Amendment (Choice of Fund) Bill 2016 that sought to achieve this
lapsed at prorogation of the 44th Parliament.
16 For example, the largest schemes for public servants in all states, except Queensland and South Australia,
offer choice. QSuper — the largest public sector fund in Queensland — allows members to transfer their
benefits to another fund on an annual basis (QSuper 2017).

DRAFT REPORT
Governments have not referred the industrial relations powers that govern their public
services to the Commonwealth and the national system does not apply to most state and
local government employees (table 3.3). The Commission estimated that up to 63 per cent
of public sector workers were not covered by the Fair Work Act (PC 2015b, p. 779). The
Act also does not cover some private sector workers, such as independent contractors.
Table 3.3 The application of the federal workplace relations system
across jurisdictions
Jurisdiction Who is covered by the Fair Work Act? Who is covered by State laws?
NSW Private sector employees and employees of
constitutional corporations
State public sector workers and
local government employees
Vic All employees
Qld Private sector employees and employees of
SA Private sector employees and employees of
WA Employees of constitutional corporations with some
additional areas where employers and employees have
other connections with the federal system
All other employees
Tas Private sector employees, local government employees
and employees of constitutional corporations
State public sector workers
NT All employees
ACT All employees
Source: CPSU SPSF Group (sub. 90 to PC 2015b).
The Commission’s approach to the above institutional constraints would be to take them as
given where they are a hard Constitutional constraint and relevant to the specific model. At
this stage, the Commission does not consider that these constraints will have a material
impact on the potential coverage of the alternative models.
3.3 Characteristics of the default product
Coverage of life-cycle phases
Consistent with the terms of reference, the primary focus of this inquiry has been to design
a model for the accumulation phase of the member’s life-cycle. The Commission has also
considered the merits of incorporating the retirement phase into the models. At present,
there are no default arrangements for the retirement phase of superannuation. However, as
part of the Australian Government’s response to a recommendation by the FSI, the
Treasury (2016) is developing a policy framework for a comprehensive income product for
retirement. It is envisaged that trustees will develop a ‘MyRetirement product’ for their
members, which will operate as a ‘soft default’, that is, it will be offered to members on an
opt-in basis.

DRAFT REPORT
77
The case for integrating accumulation and retirement phases in a single selection process
rests on the objective of designing a model that will remain relevant over time in the face
of ongoing ageing of the Australian population. As the proportion of the population in the
retirement phases of their lives grows, so would the importance of reflecting the needs of
retirees in superannuation policy. There is a further issue of ensuring the ongoing viability
and scale of default providers in the context of demographic change. A default allocation
model that integrates the accumulation and retirement phases is one way of addressing
those issues.
However, there are also compelling arguments against this approach. Notwithstanding a
common goal for members across all life stages — the objective to maximise long-term net
returns on contributions — there are also material differences between the accumulation
and retirement phases (PC 2016). These stem from the distinct characteristics of members
(including the extent and quality of their engagement with the system), as well as the
nature of the decisions they make and the services they demand. Retirement often brings
additional complexity into superannuation decisions. It requires the person to manage their
longevity and short-term investment risks in addition to long-term investment risks that are
the only factor for most of the accumulation phase. The significant differences between the
two phases is an argument for specialisation by providers within each phase.
In Australia, funds have historically been slow to develop retirement income products for
their members (PC 2016, appendix D). While this might be partly attributable to
immaturity of the system, past regulatory barriers, and potentially a lack of competition, it
could also be due to limited scope for efficiency gains from bundling the two types of
service within a single provider.
Internationally, the private pension systems most similar to Australia’s have a sole focus
on accumulation (table 3.1). Those that incorporate retirement (for example, Sweden) are
typically centralised systems with a single government provider and mandatory pooling of
risk across the entire population. This is a market structure that might better lend itself to
‘hard defaults’ in the retirement phase of superannuation, both due to administrative
savings on the supply side and the pooling of risks on the demand side.
The Commission’s view is that the models should focus solely on the factors relevant to
the accumulation phase of members.17 Additional complexity is not likely to illuminate the
question of default models at the start of a working life. Stage 3 of this inquiry process may
allow for the wider issue of failure to develop pension products to be better exposed.
17 Some researchers (for example, Drew, Walk and West (2015)) also identify an additional ‘transition’
stage between accumulation and retirement. There may be a case for providing some generic advice on
managing transitional risks in the accumulation stage.

DRAFT REPORT
Specifying the services within the default product
At a high level, the bundle of superannuation services to which default members would be
assigned is common to all allocation models considered in this report. This section sets out
the broad characteristics of the default product at the core of each model. There are two
threshold questions:
1. What services should form part of the default product?
2. To what extent should the nature of the services in the default product (in particular,
investment) be prescriptively specified?
What services are part of the default product?
At a high level, the approach to specifying the default product is informed by the
Government’s stated objective for the superannuation system, which refers to ‘providing
income in retirement to substitute or supplement the Age Pension’. Beyond that, the
Cooper Review (Cooper et al. 2010, p. 11) identified several key features for the current
MySuper architecture.
• MySuper products are intended to be simple and cost-effective to service a core
demographic of members, the majority of whom have low balances and are disengaged
from the system.
• The products should also be suitable for those who actively elect to be in a simple
default product.
• The products would have an enhanced focus on maximising net returns and reducing
costs, and would have a degree of homogeneity to facilitate comparisons and improve
competition on price and performance.
The Commission agrees with this approach, but would emphasise that the needs of default
members who are disengaged and do not exercise choice should be the primary factor in
specifying the default product. Some engaged members might elect to be in the default
product, but this is not the ‘core demographic’ for it, and default services should not be
improved to reflect their needs at the expense of those who are disengaged and may not
value those enhancements.
In sum, the guiding principle is that the bundle of services in the default product should be
the minimum necessary to meet the Government’s stated objective for superannuation. The
key services that must be included in any default product are investment, administration of
member accounts and intrafund advice. The administrative services and intrafund advice
provided to default members should meet a standard that allows members to be reasonably
informed about the performance of their investment product, the fees that are associated
with it, the intrafund switching options and the process for exiting the fund. The advisory
services that would be recoverable through a general administrative fee would also
incorporate provision of information about the operation of the superannuation system and
general financial advice (for example, online calculators and guides on managing transition

DRAFT REPORT
79
risk prior to retirement). For the most part, those services can be delivered through digital
channels (website and online account access) and functional call centre support.
Superannuation funds provide a large range of services, some of which are incidental to the
main objectives of the system, including flexible investment options, individual financial
advice, and various banking services (figure 3.2). For the most part, those services are
offered in the choice segment of the system, and are not needed in a simple default
product.
Figure 3.2 Examples of services offered by funds that are not necessary
in a default product
Source: Productivity Commission analysis of Rainmaker (nd).
Internationally, some default systems mandate the provision of additional services to
support members. For example, the NZ KiwiSaver scheme requires default funds to
educate members on making appropriate investment choices, and funds have been
addressing this in a range of ways, including communication campaigns, seminars and
provision of financial advice (appendix B).
The Commission does not support this approach for Australia. While some funds can and
do offer investor education to their members (as distinct from educating members about
Account Flexibility
• EFT/BPAY
Contributions
• Spouse accounts
• Contribution
splitting
• Anti-detriment
payments
• Linked account fee
discounting
• Children’s
accounts
Investment Flexibility
• Investment choice
• Multi-manager
investment choice
• Online investment
switching
• Monthly or quarterly
investment reports
• Choice of asset
allocation
• Ethical investment
choice
• Term deposit
investment choice
• Choice of fund
managers
• Automatic portfolio
rebalancing
• Direct shares
investment facility
Ancillary Services
• Financial advice –
limited
• Non-super
investment
products
• Loans (home,
credit, personal
and business)
• Savings accounts
• Non-super
insurance
• Shopping
discounts

DRAFT REPORT
the system and intrafund financial advice), it is not apparent that they are universally best
placed to do so, nor is this service necessary to meet the Government’s stated objective for
superannuation.
Accounting for service quality
The Commission received a large volume of comment from participants in stage 1 of its
superannuation work (PC 2016), and for this inquiry, on the importance of service quality
and the shortcomings of focusing on fees and net returns at the expense of service quality.
However, as discussed above this is largely a debate for the choice segment of the system,
where engaged and (ideally) well-informed members can communicate their preferences
about the tailoring and flexibility of the accumulation product.
In effect, the preferred approach is that service quality should not be the primary factor in
selecting default products under any of the allocation models. Funds should primarily
compete on net returns and costs, subject to meeting a threshold quality of service, not on
the quality or range of ancillary services per se.
The general approach outlined above is consistent with the bulk of international default
systems (table 3.1), as well as the sole purpose test for superannuation fund trustees and
the structure of MySuper products (with the exception of bundled insurance, which is
discussed below).
Treatment of bundled insurance
The issues paper proposed that the allocation models will focus on accumulation products,
with life and total and permanent disability insurance being allocated through a separate
process. Some participants argued that if insurance were to remain bundled with the default
accumulation product, a separate allocation process for insurance would not be efficient or
practical. Mercer (sub. 15, p. 6) observed:
We are concerned that such an approach would represent a significant and unnecessary burden
for industry, have an adverse impact on acceptance conditions and may be unworkable (e.g. we
envisage each fund may have to deal with a multitude of insurers).
The Australian Institute of Superannuation Trustees (sub. 28, pp. 34–5) argued:
The allocation of default funds must incorporate a process which facilitates market or
occupation segmentation to ensure default funds in each industry or occupation incorporates
relevant and cost effective insurance coverage … It is unclear why it would be desirable, or
indeed possible to undertake a separate competitive tender process for insurance without
undermining the benefits of group insurance.
The Commission (PC 2016, pp. 75, 149) recently noted the conflicting objectives placed
on trustees, the difficulty in articulating a clear purpose for insurance in superannuation,
and the efficiency implications (both positive and negative) of bundling insurance with the

DRAFT REPORT
81
accumulation product. It should also be self-evident to funds and their advisers that the
Government’s stated objective for the superannuation system does not envisage insurance
as an essential element of the system. However, this inquiry will not be determining
whether insurance should be bundled with the default product. In this respect, the policy
settings will be taken as given.
The sole focus at this stage is on allocating members to default superannuation
accumulation products, not insurance products. The former requires the decision maker in
all models to make very challenging trade-offs, even before insurance is considered. Thus,
the allocation models will be evaluated on the quality of their default superannuation
products only.
The Commission will assume that trustees choose to bundle default superannuation
products with insurance, subject to meeting their fiduciary and legislative obligations
relating to insurance, but the models will not involve comparing the quality or price of
bundled insurance. Ultimately, protecting member interests with respect to bundled
insurance is a task for well-specified and adequately enforced regulation that sets out clear
rules for minimum product quality standards and level of cover. Kinetic Superannuation
(sub. 45, p. 7), for example, argued that there are several existing instruments that already
govern trustees, including trustees’ insurance covenants and insurance strategy and
framework, as well as regulated minimum standards of cover.
How prescriptively should the investment strategy of providers be specified?
Default systems around the world differ in the degree of flexibility they allow providers
and members in electing an investment strategy (table 3.1). There are differences in the
investment approach (mandatory passive or at the discretion of the provider), whether the
level of portfolio risk is prescribed for providers, and whether members have any choice in
assembling their portfolio.
The optimal approach to investment in the context of private pension systems is a matter of
great debate, which is far from settled (PC 2016). Some commentators (for example, CIFR,
sub. 7) cautioned against a ‘commoditised’ approach and argued that member diversity
plays a significant role in selecting a sound investment strategy. On the latter point,
however, the Commission considers that a commoditised approach is unavoidable in a
default product, and it would not necessarily be problematic given the cost savings from a
no-frills approach. Ultimately, product customisation (where it is warranted) is the
province of the choice segment.
At present, most funds adopt a balanced fund approach for default members, which has a
greater exposure to growth assets than many no frills products used overseas for default
options. Since the advent of compulsory superannuation 25 years ago, this has generally
been a wise choice. In the context of this inquiry, whether and how the investment
approach is incorporated into the criteria for selecting products is a matter for specific
models, and the issue is discussed in subsequent chapters. In most models, the focus of the

DRAFT REPORT
selection criteria is on relative performance, rather than a mandated investment strategy,
and this is generally a better approach for selecting well-performing products. However, in
the case of the auction model (chapter 6), some prescription of the investment strategy was
an unavoidable consequence of model design.
DRAFT FINDING 3.3
The default product in all models will focus on the accumulation stage and include
investment, administration of member accounts and intrafund advice. It will be a
simple and low-cost product aimed at the needs of those who are disengaged from the
system.
• A bundled insurance product will not be a factor in the selection of products and is
best addressed through regulation and regulator oversight.
• Funds will primarily compete on net returns and costs, subject to meeting a
threshold quality of service, not on the quality or range of ancillary services per se.
3.4 Other common foundations for the models
Several additional design characteristics are common to all models presented in this report.
This section discusses how frequently the competitive process would be repeated and the
principles for developing a body that would administer this process.
Frequency of the selection process
A key element common to all of the allocation models relates to the frequency of the
selection process. There are clear trade-offs in making this decision. Running the process
very frequently would:
• raise costs for all participants
• reduce the size of the prize for participating funds and potentially adversely affect
participation
• create risks of short-termism in the way funds set their investment strategies
• potentially undermine system stability due to both short-termism and higher churn of
default products.
However, running the process very infrequently also creates several risks, including:
• creating a market with high barriers to entry and exit
• lower dynamic efficiency, because the winning funds secure default status for
prolonged periods

DRAFT REPORT
83
• higher cost of any errors made by the body that selects the winning products, because
its decisions would be locked in for a longer period of time.
Proposals from participants on the optimal frequency of the process generally ranged
between two and five years (AIST, sub. 28, p. 39; IPA, sub. 17, p. 8; Mercer, sub 15, p. 3).
In its 2012 report, when looking at the issue in the context of default superannuation in
modern awards, the Commission concluded that the process should be run no more
frequently than once every four years, and no less frequently than once every eight years
(PC 2012, p. 26). The Commission considers that this range strikes the right balance for all
its models. Nonetheless, there would be merit in running the process more frequently for
the first iterations (that is, once every four years), to sort through any teething problems in
the selection process, and gradually move to a less frequent process in the future.
DRAFT FINDING 3.4
The desirable frequency for the selection process is between four and eight years,
with the greater frequency best used in the early period.
Who will be responsible for selecting products — key principles
The Commission received a large volume of comment on the suitability of particular
bodies to the task of selecting eligible default products, as well as on the outcomes under
the current arrangements. To be clear, this stage of the inquiry is not assessing the current
arrangements and will not draw conclusions on the merits of retaining responsibility for
selecting default products within the Fair Work Commission.
Beyond that, the characteristics of a body that would perform the task partly depend on the
features of the allocation model it would oversee. Nevertheless, several best practice
principles apply irrespective of the model. The Commission considers an organisation
suited to running the selection process would possess the following characteristics:
• be dispassionate and free of conflicts of interest
• be accountable for its decisions and have credibility with members and funds.
In the context of the above principles, the Commission does not support the proposal by
the Centre for International Finance and Regulation (sub. 7, pp. 6–7) to vest the function
with a special class of private-sector member agents. As argued by several participants (for

DRAFT REPORT
example, CMD, sub. 18; AIST, sub. 28; Mercer, sub. 15), there is a strong case for the
process being administered by the Australian Government. While this places some of the
responsibility for the outcomes on the government, this is necessary to ensure the integrity
and credibility of the process. It is also consistent with the fact that the superannuation
system, and in particular, defaults, is largely a construct of regulation that compels member
contributions. Nonetheless, government administration of the process should not preclude
some private-sector participation in an advisory sense. The body responsible for making
decisions should be able to draw on expert advice from all sources.
With respect to the suitability of some existing government bodies, the Commission does
not support direct involvement of APRA and the Australian Securities and Investments
Commission in the final step of selecting preferred default products for members. It agrees
with APRA (sub. 33, p. 5) and Industry Super Australia (sub. 40, p. 33) that involving the
key financial regulators in this decision would create material conflicts of interest between
their regulatory functions and the task of picking winners from the regulated entities.
A key focus would be ensuring that those appointed to the decision-making body have both
the expertise and independence to perform their functions. The two elements may
sometimes conflict — for example, those with the greatest expertise might not be fully
independent. To some extent this might be mitigated by not requiring that all members
have expertise in superannuation and finance. Rather, they could be accomplished
individuals that have expertise in collecting and evaluating evidence and advice, but also
being able to see beyond it. Academic and some (not all) judicial fields generate these
capabilities.
A significant risk in constituting the body is if the process for selecting its members
becomes politicised and genuine independence is replaced with an objective of ensuring
that all of the relevant interests — such as employee and employer groups — are
represented. As discussed in the Commission’s report on workplace relations (PC 2015b,
pp. 161–167), the latter is a recipe for partisanship and unresolvable conflicts of interest.
More broadly, the quality and integrity of the process would need to be facilitated through:
• a robust selection process of members of the responsible body, including a process for
termination and replacement of appointments
• adequate reporting and oversight of the process and decisions of the body.
Subsequent chapters build on these best-practice principles in the context of specific
allocation models. In particular, chapter 8 expands on the desirable features of the
appointment process for a body that could apply a heavy administrative filter.

DRAFT REPORT
85
DRAFT FINDING 3.5
The selection of eligible default products should be administered by a government
body, and be subject to strong governance rules. The decision-making body must:
• have a strong focus on fund member interests
• have sufficient expertise to evaluate products
• be independent and free of real or perceived conflicts of interest
• have processes that are transparent and afford procedural fairness
• be accountable for its decisions.
A centralised online service to assist members and employers and
inform the Government
There are currently material and critical gaps in the information available to the Australian
Government about members and their choices with respect to opening, closing or
consolidating superannuation accounts. There is also a need for additional infrastructure to
facilitate the flow of information between employees and employers, and in the process
make it easier for employees to carry-over their existing accounts between jobs.
The ATO is currently developing a new online system that will streamline superannuation
fund choice and could be customised to assist both employers and employees, as well as
fill an information gap for the Government. As part of the Single Touch Payroll reforms,
people starting a job with a participating employer would be able to fill out their ‘standard
choice’ form through the myGov website when declaring their Tax File Number (the
system is currently intended to be voluntary for employers). The website would show each
employee their existing superannuation accounts and give them the option to select one of
those to receive contributions from their new employer. Employees would still be able to
nominate a different fund or select their employer’s default, as at present (ATO 2016b).
There is merit in presenting the information on that website in a way that nudges
employees away from opening additional accounts. The employee could be presented with
three options as outlined in figure 3.3.
This electronic system would be the starting point for a more streamlined choice process.
Extending it to all employees would allow them to more easily pick one of their existing
products and make their decision in their own time (provided it is within the required
timeframe). In time, the system could also be extended to a wholesale integrated payroll
service for employers that would remove the need for employees to log into myGov.
Further customisation of the service for employees could include provision of key
information on an employee’s existing products to allow them to compare their current
product with other options. This was favoured by some inquiry participants
(CMD, sub. 18; CHOICE, sub. 31). In a similar vein, the Grattan Institute has proposed

DRAFT REPORT
that people also be given this information each time they fill out their tax return (Minifie,
Cameron and Savage 2014, p. 30).
Finally, collecting the information submitted by members in a form that enables analysis of
system-wide trends in member behaviour with respect to reliance on default allocation,
product switching and account consolidation would be useful in informing future
government policy.
The Commission will continue consulting with the ATO on the feasibility and design of
these online functions. However, its preliminary conclusion is that leveraging off the
existing myGov platform is a relatively low cost and credible way of addressing the above
objectives.
Figure 3.3 A desirable order of options on an online choice form
Choose an existing
account
(select from a list of existing accounts)
Choose a default
account
(you will be allocated a new account
where you do not have one)
Open a new account
(please specify the fund and product)
1
2
3

DRAFT REPORT
87
The Australian Government should establish a centralised online service for members,
employers and the Government that builds on existing functionality of myGov and
Single Touch Payroll. The service should:
• allow members to register online their choice to open, close or consolidate
accounts when they are submitting their Tax File Number on starting a new job
• facilitate the carryover of existing member accounts when members change jobs
• collect information about member choices (including on whether they are electing
to open a default account) for their employer and the Government.
There should be universal participation in this process by employees and employers.
The case for a central clearing house
A centralised online service described above could be a first step to government provision
of a more comprehensive service for employers and employees. The FSI (Murray et
al. 2014, p. 116) recommended that ‘Government should consider implementing a national
‘payment hub’ or ‘clearing house’ by which employers make superannuation contributions
to multiple funds’.
Government-run clearing houses that direct the flow of funds from employees, or their
employers, to relevant funds are a feature of many private pension systems around the
world, including those of New Zealand and Sweden (appendix B). Typically, they have
been established to reduce administration costs for employers and funds. In New Zealand,
the central clearing house run by Inland Revenue is broadly recognised as a successful
feature of the system that also facilitates the entry of new funds into the market (box 3.3).
In Australia, clearing house services for superannuation have traditionally been the
province of the private sector, including payroll operators and some funds. However, those
services are contingent on members or employers providing the fund and account details.
They would not address situations where the employee inadvertently fails to provide those
details. And a fragmented system is not conducive to capturing economies of scale that
would arise from a centralised approach.
The SuperStream reforms recently implemented by the ATO have already established
some foundations for a centralised clearing house. Employers must now pay their
superannuation contributions to superannuation funds electronically, and the payment is
linked to a unique payment reference number, which is automatically sent to the ATO for
compliance monitoring. The system allows employers to make the contributions for all
employees in a single transaction, even if the money is going to multiple funds. The ATO
has also developed a small business clearing house service, albeit with limited
functionality. Developing a centralised clearing house and housing it within the ATO
would be a natural extension of this recent work. However, the Commission acknowledges

DRAFT REPORT
that this would involve a significant change to the role traditionally undertaken by the ATO
in superannuation.
Box 3.3 Inland Revenue’s role as a clearing house
Inland Revenue uses New Zealand’s pre-existing income tax system for its clearing house role.
Employers deduct employee contributions, make compulsory employer contributions and remit
these directly to Inland Revenue. Inland Revenue transfers KiwiSaver contributions to each
employee’s relevant KiwiSaver scheme provider (Inland Revenue 2017).
The clearing house is widely considered to be a critical and successful part of KiwiSaver. The
Commission consulted with KiwiSaver funds, New Zealand government organisations and
academics, most of whom highlighted what they consider to be the clearing house’s benefits:
• ensuring each individual has only one KiwiSaver account
• low administration costs for employers and funds
• the ease with which Inland Revenue can monitor employers’ compliance with mandatory
contributions
• simplicity for funds facilitates market entry and competition between providers.
One drawback noted by some parties is that Inland Revenue does not have up-to-date contact
details for all employees, making it harder for providers to communicate with their members.
(However, this issue would likely still exist under a different system.)
Regulation should promote competition
Irrespective of the allocation model selected, the regulatory regime governing the system
(both the content and administration) should be conducive to broader competition in the
system and not create any impediments of its own. For example, mergers and acquisitions
can be a way to inject competition for the default market. Participants in the Commission’s
stage 1 study (PC 2016), as well as this inquiry, identified several areas where more
effective regulation of the system could potentially improve competition and outcomes for
members.
Improved disclosure of fund costs and fees
The current lack of transparency on fund fees and costs was a key theme for many
participants in the stage 1 study (PC 2016, pp. 193–194). Problems with fund disclosure
included misattribution of costs and fees (across investment and administration), missing
or inconsistent data for some types of costs (for example, indirect and related party costs)
and a likely misalignment between the structure of member fees and underlying costs.
This lack of transparency is likely to materially impede both member- and regulator-driven
competition between funds. On the regulator side, the Deputy Chairman of APRA
(Rowell 2016) observed that there needed to be far greater transparency on the underlying

DRAFT REPORT
89
costs of running a superannuation fund for APRA to be able to put pressure on the funds to
reduce costs or merge with a more efficient fund. Of even more direct relevance to this
inquiry, the lack of transparency will further complicate the task of the body responsible
for selecting default products.
Successor fund transfer rules
An important consequence of the competitive process is that it should enable efficient exit
of funds (chapter 1). Current successor fund transfer rules are a potential barrier to this
process. The regulations impose a requirement on trustees of funds wishing to exit and
transfer their membership via a ‘bulk transfer’ to meet an ‘equivalent rights’ test (r. 1.03 of
the Superannuation Industry (Supervision) Regulations 1994 (Cwlth)). This is combined
with the general requirement under section 52 of the Superannuation Industry
(Supervision) Act 1993 (Cwlth) that the trustees of both funds act in the best interests of
members.
Some participants in stage 1 argued that, in the absence of clear guidance by APRA,
trustees interpreted the test conservatively and would not consolidate unless they could
establish literal equivalence of all rights for every member in both funds. A frequently
cited example was the difficulty in securing equivalent terms and conditions for bundled
insurance. The Commission concluded that the rules could be a material barrier to efficient
consolidation if they prevented mergers that would benefit the member group as a whole,
while leaving some members worse off.
More recently, APRA (2016) has released new draft guidance for successor fund transfers,
which if implemented, could address the concerns about a literal interpretation of the rules.
Specifically, the guidelines indicate that equivalent rights only apply to legally enforceable
rights of members (such as the right to accrued balances). They would not apply to product
aspects, such as the insurance offering, and the level of fees, product features and
investment options for the accumulation product. The guidelines also propose that trustees
can assess groups of members with common characteristics rather than examining the right
of every member.
The Commission will monitor the development of these guidelines, but considers the draft
document to be a step in the right direction.
Transparency around merger proposals
The stage 1 report (PC 2016, p. 96) noted the debate in the industry about mergers not
going ahead due to principal–agent problems. Several stakeholders have contended that
some trustee board directors (and their sponsoring bodies) have an incentive to avoid
mergers that would force them to relinquish their position on the board.

DRAFT REPORT
The Commission will not take a view in this inquiry on the relative merits of different
board structures. However, it recognises that there is a general lack of transparency for
members and APRA in situations where a fund was approached for a merger, but the
merger did not go ahead. This is in stark contrast to what is required of listed companies
under Part 6.5 of the Corporations Act 2001 (Cwlth). The latter clearly specifies the
takeover procedure and requires anyone making a takeover bid to disclose this information
to the Australian Securities and Investments Commission (ASIC) and the shareholders of
the company targeted for the takeover. The Act also requires directors of the targeted
company to respond with statements to ASIC, the bidder and their own shareholders.
Rice Warner (sub. 43, p. 10) argued:
The issue appears to be with the processes by which those proposals are considered and how
decisions are reached in the best interests of members. The process would be enhanced by a
formal framework for approaches and the consideration of those approaches.
While trustees already have a general obligation to act in member interests, there is merit
in Rice Warner’s proposal to introduce a formal framework specifically for mergers to
provide confidence to members and APRA that the trustee is discharging this obligation.
The Commission agrees with Rice Warner that the framework should include:
• clear guidelines for funds on how to approach another fund with a merger proposal
• an obligation for trustees to give genuine consideration to merger proposals in the
context of member interests
• disclosure by trustees to their members and APRA of all merger approaches and the
reasons for any decisions.
The Australian Government should introduce a formal framework that specifies the
process and obligations of trustees when making or considering merger proposals. As
part of the framework, trustees would be required to disclose all merger attempts
involving their fund, as well as the reasons for any decisions.
The Commission is seeking feedback on the merits and implementation issues of its
proposal to improve transparency around fund merger activity.
• What (if any) complementary regulatory action would be needed to ensure that the
framework is effective in promoting mergers and the exit of underperforming funds?
• Are there any likely unintended consequences from introducing the proposed
framework? To what extent, and how could those unintended consequences be
addressed through policy design?

DRAFT REPORT
91
Annex Estimating the effects of alternative definitions
of new default members
The definition of a new default member affects the size and composition of the default
market contested by funds, as well as the proliferation of default superannuation accounts
within the system. The Commission has estimated some of those effects under the current
‘labour turnover’ definition and the ‘first-timer’ definition proposed by the Financial
System Inquiry. These estimates are not a precise assessment and are only intended to
provide an order of magnitude of the effects.
Size of the market
ATO data on the number and taxable income of members that open a new superannuation
account each year, and the number of accounts they opened, were used to provide an
indicative estimate of the size of the pool of default members for the turnover scenario
(table 3.4). To derive the number of members and accounts in the turnover pool, the
Commission applied a 70 per cent default rate to the aggregate member and account data
provided by the ATO. The default rate is based on a survey by Delpachitra and
Rafizadeh (2014) and is also consistent with a 2010 survey by Roy Morgan, cited by
Cooper et al. (2010).18
For the first timer pool, the Commission used ATO data on the number of members
opening an account for the first time each year and assumed that each member would only
hold one account. The Commission assumed a 90 per cent default rate, in line with broader
findings in the literature that young people are more likely to default (PC 2016). In
calculating the value of the contributions from the first-timer group, the Commission drew
on average income estimates for that group provided by the ATO. The estimated value of
contributions underestimates the longer-term attractiveness of each year’s cohort, because
it is effectively based on the average income of a person when entering the workforce.
Applying average incomes of the turnover pool to the first-timer cohort yields a pool of
over $2 billion per annum.
Member switching behaviour is not incorporated into the estimates
Many of the factors that would drive the relative value of the two pools cannot be
estimated. The Commission has not attempted to predict the switching behaviour under
either scenario. For example, the estimate of contributions in the turnover pool may
understate its short-term attractiveness to the (unknown) extent that existing members may
consolidate balances from an established account into the new default product. Many funds
actively market this option to new members. Conversely, the value of the turnover pool
would decline if members carry over their existing account into a new job. And all account
18 The latter found an 80 per cent default rate.

DRAFT REPORT
consolidation activity under the current arrangements would reduce the estimate of account
proliferation in the turnover scenario.
Table 3.4 Indicative size of the turnover and first-timer poolsa
Turnover poolb First-timer poolc
Number of accounts 1 070 000 410 000
Number of members 960 000 410 000
Average income $56 800 $21 400
Value of contributions $5.2 billion $0.8 billion
a Average per year, between 2013-14 and 2015-16. Income and value of contribution estimates expressed
in 2015 dollars. b Based on the number of people opening accounts, the number of accounts they opened
and a 70 per cent default rate. c Based on the number of people opening accounts for the first time and a
90 per cent default rate. Assumes one account per person.
Sources: ATO (pers. comm., 24 January 2017, 15 February 2017).
Consequently, the estimates for the value of the turnover pool and the number of accounts
that would be removed from the system due to moving to the first-timer approach are
presented only to indicate an order of magnitude of the impacts.
Fund membership demographics
To provide context for the estimate of the first-timer pool, the Commission examined the
demographic distribution of member benefits in APRA-regulated funds (table 3.5). Nearly
three-quarters of total member assets in APRA funds, and well over half of assets in
MySuper products, are held by those who are over 45. This contrasts with the demographic
distribution of new entrants into the system — of whom about two-thirds are younger
than 25 (figure 3.4). The flow of contributions from the first-timer pool would be material
in the context of the cohort of existing young members — those that are most relevant for
the longer-term sustainability of the funds.
Table 3.5 Member benefits in APRA funds by age bracket, 2016a
<25 25–44 45–54 55+ Total
All APRA funds Value of benefits ($b) 11.7 306.0 299.2 603.4 1 220
Share of total (%) 1 25 25 49 100
MySuper products Value of benefits ($b) 9.3 191.2 138.5 117.5 457
Share of total (%) 2 42 30 26 100
a Does not include the benefits of members whose age was not reported. Those members account for fewer
than 1% of balances, for both MySuper products and APRA funds overall.
Sources: APRA (2017a, 2017b).

DRAFT REPORT
93
Figure 3.4 Age distribution of new entrants into the super systema
a Average for the period 2013-14 to 2015-16.
Data source: ATO (pers. comm., 10 January 2017).
Savings in administrative fees and insurance premiums from reduced
account fragmentation
A reduction in the number of default accounts would lead to savings to members in
account administration fees, as well as in the price of duplicate bundled insurance policies.
These gains to members would be cumulative — every year there would be fewer accounts
opened relative to current arrangements, leading to an additional long-term stream of cost
savings.
Chant West (2016) estimated the average fixed component of the administrative fee on a
MySuper account at $75. Given the indicative nature of the estimates, the Commission
assumed that this value would remain constant under the turnover and first-timer scenarios.
This implies total annual fee savings of about $45 million for every permanent reduction of
600 000 accounts relative to the turnover scenario.
Estimating the savings in insurance premiums paid by members from reduced duplication
of bundled insurance is challenging, particularly due to the complexity and heterogeneity
of the policies. The Commission assumed an average premium of about $255 per annum
on the basis of aggregate APRA (2017c) data on premiums collected and the number of
accounts with bundled insurance. The Commission also assumed that 20 per cent of default
members in the turnover pool opt out of bundled insurance when they open a new default
account. The latter implies that for every reduction of 600 000 default accounts, there
0
5
10
15
20
25
30
35
40
45
<18 18–25 25–35 35–45 45–55 55–65 65+
Percent
Age of members (years)

DRAFT REPORT
would be a reduction of about 500 000 bundled insurance policies. Accordingly, every
permanent reduction of 500 000 bundled insurance policies relative to the turnover
scenario could lead to annual cost savings to members of over $100 million.
Overall, the savings to members in administrative fees and insurance premiums from
removing between 500 000 and 600 000 accounts from the system would be of the order of
$150 million per annum.
The Commission is seeking comment on its approach to, and alternative ways of
estimating the size and value of, the turnover and first-timer pools and the benefits
from reducing account proliferation due to moving to the first-timer approach.

THE BASELINE
DRAFT REPORT
95
4 The baseline
Key points
• The Commission has used a baseline of no default arrangements. This would require
employees to choose their own superannuation product — described in this report as
unassisted active choice.
• The baseline is not an alternative model being considered by the Commission — the
performance of the alternative models relative to the baseline is what matters.
− The baseline also focuses attention on what an absence of default arrangements would
imply for members in today’s superannuation environment.
• A world without default arrangements has a number of positives.
− The supply side of the market would be contestable and rivalrous, lending itself to
strong competition. Such organic competition is an important driver of innovation over
time.
− It minimises the number of formal principal–agent relationships and encourages more
members to engage with their superannuation.
− It imposes a minimal regulatory burden and has low risk of system instability.
• Despite these positives, such a baseline also exposes serious weaknesses inherent in
compelling people to make complex long-term financial decisions that some may find
difficult.
− Although employees are notionally free to act in their own best interests, cognitive
constraints, behavioural biases and inadequate information can limit their ability to do
so. Moreover, compulsory membership in the system inevitably exacerbates
disengagement.
− There is scope for unhealthy or wasteful competition (including marketing) in the
market.
− Employees are likely to rely on advice from third parties (such as employers and
financial advisers) when making their decision, but potential conflicts of interest can
lead some employees astray.
− Employees face the bulk of the costs of allocation, including search costs.
• In this inquiry, alternative models need to beat the baseline to be considered viable, and
then prove workable.
The Commission’s task for this inquiry is to develop and assess alternative models for
allocating default members in the superannuation system to products. The Commission’s
assessment of alternative models requires a common reference point for comparison: a
baseline. This chapter establishes the baseline for this inquiry.

DRAFT REPORT
It is important to note that the baseline is not an alternative model in itself — but is used as
a tool to assess alternative models proposed by the Commission. For this purpose, the no
defaults approach has been labelled as ‘unassisted active choice’ — a baseline that requires
employees to choose a superannuation product without assistance.
The baseline was primarily chosen to identify and understand the problems that would
arise under a completely decentralised allocation mechanism (chapter 2). This guides the
Commission in developing alternative models that directly address these problems — the
rationale for government intervention using defaults.
Characterising and assessing an unassisted active choice baseline is not without its
difficulties. There is no obvious precedent in Australia or internationally that can provide
real-world evidence for how a compulsory retirement savings market without defaults
would function; but neither is that essential unless the intent is to apply such an approach.
Further, many participants questioned the use of a no defaults baseline, suggesting that it
lacks analytical rigour, is difficult to operationalise and could yield misleading results
(chapter 2).
However, there are several ways to gather evidence to inform the baseline. These draw
upon established and tested economic principles and analysis, academic research, current
system settings, and lessons from similar markets. The Commission has gathered evidence
to provide a reasonable analysis of how employees would behave in an unassisted active
choice world to provide a qualitative assessment of the baseline and address participant
concerns.
This chapter begins by establishing the main characteristics of the baseline to give an idea
of how it would work in practice (section 4.1). Then, section 4.2 assesses the baseline
against the Commission’s five assessment criteria (described in chapter 2). This is useful
for two reasons. First, it identifies the potential problems of the baseline in order to guide
the development of alternative models. Second, the assessment provides a common point
of comparison for the alternative models.
4.1 Establishing the baseline
The Commission’s baseline for this inquiry is a world of unassisted active choice and thus
no defaults. The defining characteristic of the baseline is that it places the onus entirely on
the employee to choose a superannuation product.
The baseline reflects an objective ‘first principles’ approach to a standard competitive
market: the consumer (employee) chooses and directly purchases a service (superannuation
product) provided by a supplier (superannuation fund). By using a completely
decentralised approach to product allocation, an assessment of the baseline helps identify
possible market failures and indicates the problems to be addressed through government
intervention in alternative default models.

THE BASELINE
DRAFT REPORT
97
The baseline has two key characteristics. First, employees would be required to choose a
superannuation product to receive their employer contributions. There would be no default
products that employees are automatically allocated to if they do not make a choice.
Second, employees could choose from any publicly offered legally complying
superannuation product available in the market. All employees would have unrestricted
choice. This means there would be no role for workplace instruments to allocate employees
to funds or products.
Importantly, superannuation products would still be subject to today’s prudential,
competition and consumer laws and regulations. This is significant because unassisted
choice does not mean no regulatory standards at all. Further, regulatory advances since the
introduction of the compulsory superannuation system in the early 1990s may have
lessened some problems that the default system was originally designed to solve. For
example, changes in product disclosure standards have improved transparency and
comparability of products over time.
A legally complying product in unassisted active choice would be any product offered by a
superannuation fund regulated by the Australian Prudential Regulation Authority (APRA),
Australian Taxation Office (ATO) or other legislation (exempt public sector
superannuation schemes). In 2016, there were over 200 large superannuation funds
regulated by APRA, offering more than 40 000 investment products, many of which were
open to the public (APRA, sub. 32, pp. 7, 13 to PC 2016). In addition, employees can set
up a self-managed superannuation fund if they wish to have complete autonomy over their
investments. By June 2016, there were over 577 000 self-managed superannuation funds
(ATO 2016a).
Under the baseline, choice is ultimately driven by employees. The current role played by
employers in choosing a default fund would not apply. That being said, employers, unions
or other intermediaries would be free to bargain with superannuation funds on behalf of
employees to secure group discounts on fees. This would mimic the current practice
offered by corporate funds or industry funds, the only difference being that employees
would be under no obligation to choose that fund. Further, although the superannuation of
most members is in defined contribution products (PC 2016, p. 32), there would be no
restriction on employers and superannuation funds offering defined benefit products.
Similarly, product offerings would be decided by superannuation funds. This means that
funds would determine their own fees, investment strategies and ancillary services.
Importantly, current legislation requires superannuation funds to offer a minimum level of
insurance in their default products. However, given there are no default products under the
baseline, funds would not be subject to this requirement. Funds would be free to offer
products with or without insurance as part of their product offering, noting that some
legislative provisions and regulations may apply to how funds offer and provide insurance
to members.
In practice, employees would follow a simple process for choosing a superannuation
product (figure 4.1). Employees would ultimately be responsible for evaluating available

DRAFT REPORT
products in the market and choosing their preferred product or fund. Third parties can also
assist employees in making their decision, such as employers, unions, financial advisers
and private ratings entities. Once the employee has chosen a product, they would be
responsible for joining the fund and obtaining their product and account information. They
would then provide the account details to their employer in a similar process to filling out
the current superannuation ‘standard choice form’ (ATO 2015b), but without any default
option. In addition, and mimicking the current rules, the employee would be required to
obtain a letter from the fund stating that it can receive contributions from their employer
(ATO 2015b, p. 2). This process is available now on many larger super fund websites.
Figure 4.1 Illustrative process for employees to choose a super product
a An employee can choose just the fund (not the specific product) if the fund automatically allocates them
to a product. However, they would still be required to provide the account number to their employer.
This active choice nomination process is not novel. It mirrors the current process whereby
new employees are required to choose their own bank account and provide the details to
their employer in order to receive wages.
In an unassisted active choice world it is possible that some employees will either refuse or
fail to make a choice. It is difficult to predict how many employees will fail to nominate a
product. Nonetheless, the Commission considers that nonresponse is generally likely to be
low. In a related example, Carroll et al (2009, p. 1643) studied a natural experiment
whereby employees were required to actively choose whether to enrol in a retirement plan
within 30 days of being hired. Five per cent of new employees failed to return the form.
Under the baseline, there would be significant consequences for non-responding
employees and their employers. Non-responding employees would forfeit their employer
contributions, as the employer would have nowhere to make the payments. And given that
employers are under a legal obligation to make these contributions, they could face
substantial consequences if these obligations are not met (Ai Group, sub. 21, p. 4).
Under the current rules, employers are only required to make contributions quarterly
(ATO 2015a). This is likely to give new employees time to choose and nominate a
superannuation product. In addition, the threat of non-compliance action from the ATO
may act as an incentive for the employer to follow up with the employee on their choice of
fund. It is reasonable to assume that eventually a product would be nominated, but delay
and uncertainty ultimately impose a cost to non-responding employees.
Evaluate
products in the
market
Choose the
preferred
product
(or fund)a
Join
superannuation
fund and obtain
account and
product
information
Supply account
information to
employer
(using standard
choice form)

THE BASELINE
DRAFT REPORT
99
4.2 Assessing the baseline
This section assesses the benefits and costs of the baseline against the five criteria outlined
in chapter 2 — member benefits, competition, integrity, stability and system-wide costs.
After discussing the benefits and costs against each criterion, this section summarises the
overall assessment of the baseline.
Member benefits
The first assessment criterion focuses on whether the allocation process creates incentives
for funds to maximise long-term net returns and allocate members to products that meet
their needs.
Employees can often struggle to make good decisions
Employees ultimately represent the demand-side of the superannuation system. Under the
baseline, employees would be free to make decisions in their own interests. They could
bargain directly with suppliers (or use an appointed agent) and choose a superannuation
product according to their own risk tolerance and personal preferences. Or, they may make
a choice at random. Further, unions, employers and other agents would be free to negotiate
with suppliers to secure group discounts on fees for a particular cohort of employees, such
as those currently available to corporate funds. However, the incentive to do so may well
be reduced by the fact that in an unassisted active choice world, no default regulation
would exist to guide them to join such an option. There would also be no legal constraints
on employees that restrict their ability to choose a product or an agent.
However, employees do not always make decisions in their best interests. Freedom to
choose means they may choose products that are likely to result in worse retirement
outcomes relative to outcomes from other available products. For example, employees may
choose products with higher fees, poorer investment performance or pay for additional
services (such as a plethora of investment options) they do not need. Further, many
employees may choose overly conservative products, reducing their potential retirement
income. There is evidence that many people fail to properly consider risk when
diversifying assets (Benartzi and Thaler 2007, p. 86), are excessively sensitive to financial
losses relative to gains (Kahneman and Tversky 1979, p. 263), and can be slow to move
out of underperforming products (Bateman et al. 2016a, p. 20).
There are a number of reasons why employees may make poor decisions — especially
long-term decisions such as superannuation. People can struggle when making long-term
decisions due to various constraints, heuristics and behavioural biases (box 4.1). They are
also not always confident in making financial decisions or where to look for help, and they
can be easily misguided by irrelevant information or advice that is not in their best interest.

DRAFT REPORT
Moreover, compulsion in superannuation (in part created to address some of these
constraints and biases) inevitably exacerbates disengagement of members — a
characteristic present under the baseline as well as any alternative allocation model.
Given that people generally join the workforce at a young age, most people who are
required to make a decision under the baseline (and all the models) are young. In fact,
about one quarter of members are under 18 when they are first prompted to choose a
superannuation product (chapter 3). It is then particularly concerning that younger people
tend to show little interest in their superannuation and are particularly vulnerable to
making poor decisions. In a 2014 survey of 25–35 year olds, Ali et al found serious
deficiencies in the superannuation knowledge, behaviours and attitudes of young adults.
Knowledge of basic facts about superannuation is poor. Young adults are unengaged by and
uninterested in superannuation or retirement planning. Attitudes can be described as worried
and sceptical. (Ali et al. 2014, p. 1)
Rice Warner (sub. 43, p. 6) shared this concern, noting that many young people hold too
many accounts (an outcome encouraged by the current default system), pay high fees
relative to their balances, and pay for excessive insurance relative to their needs.
Box 4.1 Constraints, heuristics and behavioural biases
Constraints on decision making
When people face complex problems, such as retirement planning, behavioural economics
suggests they do not always make decisions in their best interests. There are several possible
reasons for this. For example, Mullainathan and Thaler (2000, p. 2) suggest people do not
always put their interests first (bounded self-interest) or may lack the willpower (bounded
self-control) to carry out their intentions. Further, in a well-known view, Simon (1955) proposed
that people face ‘bounded rationality’ in the presence of three main constraints:
• cognitive constraints — people find it difficult to compute complex problems
• time — there are limits to the time people have available to solve problems
• imperfect information — relevant information may not be accessible or understandable.
A prominent example of a constraint in the superannuation system is the large number of
products available for employees to choose from. The need to decide between an excessive
number of options can lead to ‘choice overload’ and lessen an employee’s motivation to
choose. Iyengar, Huberman and Jiang (2004) show that when people are presented with a
greater menu of superannuation investment options to select from, they are less likely to
participate in a retirement savings plan.
(continued next page)

THE BASELINE
DRAFT REPORT
101
Box 4.1 (continued)
Heuristics
Given these constraints, people often resort to using mental rules of thumb or shortcuts —
commonly termed ‘heuristics’ — to simplify problems and make decisions. There are many
different heuristics people use when making retirement decisions.
Benartzi and Thaler (2007, pp. 84–86) noted that people with US retirement savings accounts
choose contribution rates in a number of ways, for example: choosing multiples of 5 per cent;
matching the employer contribution; or simply choosing the maximum allowed. Further, people
tend to diversify when choosing their asset allocation, often dividing their assets evenly across a
subset of the available options — termed the 1/N heuristic. In addition, people may rely on
patterns in a small sample of information (representative heuristic), may only rely on readily
available information (availability heuristic) or rely on the default option (default heuristic)
(Tversky and Kahneman 1974).
Behavioural biases
While heuristics can be useful in making decisions under constraints, they can also introduce
systematic biases into decision making. These biases can prevent people from taking account
of all relevant and available information, making it difficult to act in their best long-term interests.
• Status quo bias — researchers have observed that many people display procrastination
(delaying decisions) and inertia (failing to revise decisions) when saving for retirement (Clark
et al. 2013).
• Framing bias — experiments have shown that the way information is presented, such as a
menu of investment options, has a strong effect on the portfolio people choose. For
example, anchoring can occur when people rely heavily on one trait or piece of information
when making a decision (Mitchell and Utkus 2004).
• Loss aversion — experiments show that people tend to be more sensitive to financial losses
than gains of the same amount (Mitchell and Utkus 2004).
• Familiarity bias — some investors may prefer to invest in shares of companies they are
familiar with because they view it as less risky than others (Clark et al. 2013).
• Present bias and myopia — research suggests that some people focus on short-term
consumption to the detriment of long-term consumption. In particular, some people may be
subject to ‘hyperbolic discounting’, whereby the time value of money is discounted more
heavily in the near term than the long term. Indeed, one of the reasons superannuation is
compulsory is to mandate savings because people may not do so on their own accord
(Goda et al. 2015; Mitchell and Utkus 2004).
• Compulsion — it is inevitable that some people who are compelled to make superannuation
contributions may not be interested in or capable of making any relevant decisions, leading
them to disengage from the decision-making process (Fear and Pace 2008, p. vii).
Beyond disengagement, simple lack of awareness can also compound the problem.
Between 2002 and 2014, about 2 per cent of people surveyed every three years were not
aware that employers are required to make superannuation payments on behalf of their
employees (ANZ 2015, p. 71). Employers of such employees may fall short of their

DRAFT REPORT
Superannuation Guarantee requirements, but the extent to which this occurs even today is
subject to debate and will likely be examined by the Commission in its stage 3 inquiry.
Collectively, these problems have the potential to lead some employees astray, choosing
products which either perform poorly or do not meet their needs. Several submissions
echoed these concerns about the quality of employee decision making under unassisted
active choice (for example, First State Super, sub. 26, p. 2; Kinetic Superannuation,
sub. 45, p. 6).
Alternative models should demonstrate how, and how effectively, they can address these
problems in employee decision making.
What does unassisted active choice mean for member engagement?
Much of the evidence on member disengagement is drawn from a world of defaults that
may perversely encourage it. A significant benefit present in the baseline, but less evident
under other default arrangements, is that requiring employees to interact with their
superannuation by making active choices can potentially drive member engagement
(ASFA, sub. 24, p. 21; Carroll et al. 2009, p. 1641; FSC, sub. 38, p. 16).
Under the baseline, there are significant penalties for indecision — if an employee fails to
choose a product, they may miss out on contributions. This penalty can be expected to act
as an incentive for employees to evaluate and engage with their superannuation in order to
make a choice. In fact, researchers have examined how an unattractive default option can
encourage people to make their own active choices (Carroll et al. 2009). For example, New
Zealand’s use of a very conservative investment default has led many people to change
providers: about 34 per cent of members who were initially in a default fund in 2008 had
transferred to a new provider by 2012 and another 5 per cent had switched investment
options within the same default provider (MBIE 2012, p. 34).
Employees have the most to benefit from active and meaningful engagement. If employees
feel encouraged to take an interest in — and educate themselves about — their
superannuation, they are better placed to manage their savings in line with their individual
financial situation and risk preferences. This is important because employees ‘ … would
have more information about their circumstances than any other decision maker’ (ASIC,
sub. 41, p. 3). Greater member engagement can also benefit employees by encouraging
them to consolidate multiple accounts in favour of a single account. This can save
employees from paying excess fees on multiple superannuation accounts.
Overall, the baseline is likely to encourage employees to engage with their superannuation.
This is important because member engagement can drive overall market (demand-side)
competition within the superannuation system, improving system-wide efficiency.
Alternative models will have to offer ways of improving on member engagement if they
are to improve upon the baseline.

THE BASELINE
DRAFT REPORT
103
Greater choice benefits from better information
Members may not engage with their superannuation, or struggle to make decisions,
because there are problems with the information they use. In the superannuation system,
useful information is not always available (PC 2016, p. 193). Information tends to be
complex, can be hard to access or is not always consistently disclosed. This makes it
difficult to compare options, especially given the plethora of superannuation products
available in the market. These information problems lead to high search costs for
employees, who often have to spend a lot of time and effort to find, understand and
evaluate information in order to make meaningful decisions.
Consumer protection regulations can help improve the usefulness and transparency of
information, but they are not a silver bullet. For example, the Australian Securities and
Investments Commission (ASIC, sub. 41, p. 2) aims to improve information transparency
and comparability in the superannuation system through its ongoing work with product
dashboards and product disclosure statements. Further, the private sector already provides
easily accessible information and comparison tools that employees can use to simplify their
decision making and reduce their search costs. But, as noted by ASIC the effectiveness of
all of this information is ultimately limited by the extent to which consumers are willing to
use it.
Disclosure initiatives … while useful for consumers, cannot deliver investor protection and
effective competition on their own. In part, this is driven by consumer disengagement and
behavioural factors constraining the effectiveness of these disclosure measures.
(ASIC, sub. 35, p. 2 to PC 2016)
Under the baseline, superannuation funds can be expected to lift their efforts to actively
inform potential members of their services and performance in order to attract and retain
members. Undoubtedly, there is a cost to providing this information. But, no matter who
provides it, there will always be a cost associated with improving information in the
superannuation system. Absent the option of a single sovereign fund (chapter 3), there are
only two pathways to address information asymmetry: increase the availability and quality
of information (at a cost) or limit choice to a smaller set of funds.
In an assessment of competitive models, which is the task of this inquiry, choice is a given
and thus improved consumer information is likely necessary. Therefore, all models will
face the cost of information provision to some extent. That said, the relative cost will be an
important consideration as to whether and how competitive allocation can work.
Other industries experience similar consumer decision-making problems
In order to understand the impacts of unassisted active choice on the superannuation
system, the Commission looked at the way consumers make decisions in similar industries.
Other markets, such as residential home loans, electricity and private health insurance, rely
on consumers actively making difficult decisions (box 4.2).

DRAFT REPORT
Box 4.2 Consumer decision making and competition in like markets
Superannuation has similar qualities to other markets in Australia that rely on active choice.
Consumers often face the same difficulties in decision making and suppliers are subject to a
similar landscape of competition.
Residential home loans
Residential home loans share many of the same characteristics as superannuation. Consumers
struggle with similar decision-making problems, such as cognitive constraints and behavioural
biases (Skelton 2015, pp. 65–94). However, they are still expected to make complex financial
decisions between products with varying features and long-term consequences. This can lead
them to taking on home loans that are unnecessarily costly or do not meet their needs. For
example, some market aggregators suggest that the highest and lowest home loan interest
rates can differ by about 2 percentage points (Canstar 2016; RateCity 2017).
Further, banks face similar barriers to entry and exit as those found in superannuation, including
scale, vertical integration, regulatory requirements and incumbency advantages (CIFR 2015,
p. 92; Deloitte Access Economics 2014, pp. 29–34). However, rivalry is more limited. While
products are somewhat homogenous and many substitutes are available, the market is
dominated by the big four banks, which consistently retain over 70 per cent of the market share
for household loans (CIFR 2015, p. 58).
Electricity
While not a completely unassisted choice market (State Governments provide comparison
websites), studies of consumer attitudes in Australia’s electricity market show consumers
generally have a low level of interest or understanding of electricity or its pricing, and are likely
to be more interested at specific times (such as signing up to electricity services) (AEMC 2012,
p. 43). In particular, vulnerable consumers experience some of the same behavioural biases
found in the superannuation system, such as inertia and loss aversion (AEMC 2016, p. 46).
Competition in retail electricity markets is also similar to that in superannuation. There are
several regulatory barriers to entry, such as price regulation and regulatory uncertainty
(AEMC 2016, p. 91). Overall, there is strong retail competition in most jurisdictions
(AEMC 2016, pp. i–iii). Consumers can receive significant discounts (in the order of hundreds of
dollars a year) by switching to a more competitive offer. But disengagement can leave
consumers paying more. For example, the AEMC (2016, p. iii) suggests that consumers who
have not switched electricity retailers in the past five years could probably find a better deal.
Private health insurance
About half of all Australians have private health insurance (ACCC 2015, p. 6). Like
superannuation, regulatory settings encourage consumers to obtain private health insurance —
through insurance rebates and the Medicare Levy Surcharge. However, imperfect information
and complexity in decision making can cause consumers to disengage from their private health
insurance and choose policies with little value. For example, research shows that while
consumers engage at the start of the purchasing process, many who think about changing their
arrangements fail to do so (ACCC 2015, p. 29). In addition, many consumers are thought to
hold ‘junk’ policies that provide them little value (CHOICE 2016).
Contestability issues in the private health insurance market closely resemble those in
superannuation. For example, barriers to entry include prudential capital requirements, product
regulation, incumbency advantages and economies of scale (PHIAC 2015, pp. 14–21).

THE BASELINE
DRAFT REPORT
105
Some inquiry participants suggested that superannuation is not necessarily comparable to
other industries or products, such as bank accounts, as the decisions are more complex and
made for the long term (ASFA, sub. 24, p. 22; IFAA et al, sub. 13, p. 6; Mercer, sub. 15,
p. 4). Further, involvement in the superannuation system is compulsory.
However, the Institute for Public Affairs likened superannuation to other industries,
arguing that:
In the same way that these entities are not involved in the selection of employee bank accounts,
electricity or telecommunications providers, the selection of a superannuation fund should
ultimately be a matter for an employee alone. (IPA, sub. 17, p. 8)
Overall, this does not mean that other markets offer no useful insights and lessons. The
industries in box 4.2 tend to cater to most Australians and some of the decisions, such as
choice of home loan, can be complex, made for the long term, and have very large
financial consequences. Moreover, a mortgage is harder to unwind (penalties, significant
documentation) than a shift across super funds; and comparisons of health insurance
products are particularly difficult to assess given an individual’s need (expectation of
making a claim) is far less certain than their eventual retirement.
Viewed in this way, the superannuation default system might be considered an outlier in
terms of intervention.
Competition
The second assessment criterion focuses on encouraging open participation (contestability)
and rivalry between funds (competition) that drives innovation, cost reductions and more
efficient long-term outcomes for members.
Competition on both the demand and supply sides of the superannuation system can be
useful in improving outcomes for members, by reducing prices or increasing service
quality (PC 2016, p. 65).
The baseline is the purest form of competition. There are few regulatory constraints on the
demand side — employees are free to choose any superannuation product that meets their
needs. However, some aspects of the system naturally distort the signals members send to
superannuation providers. As noted above, some employees are disengaged or fail to make
decisions in their own best interests. This softens the pressure on superannuation funds to
provide their services at or close to the underlying costs of provision.
That said, the threat of switching by engaged employees under the baseline may exert
enough competitive pressure on funds, driving better outcomes for all members. And as
noted earlier, engagement is encouraged under the baseline conditions.

DRAFT REPORT
The baseline is highly contestable …
On the supply side of the superannuation system, the baseline imposes few material
constraints on providers. The baseline is considered to be highly contestable given there
are low barriers to entry (ASFA, sub. 24, p. 21; FSC, sub. 38, p. 16). This contestability
allows superannuation funds to compete in the same way as suppliers in other competitive
markets.
The Australian Competition and Consumer Commission (2008, p. 40) classifies barriers to
entry and advantages of incumbency as regulatory, structural or strategic. Under the
baseline, some barriers to entry and incumbency advantages exist.
• Regulatory barriers — under the baseline, superannuation funds are subject to similar
regulatory barriers to entry as most other financial service providers, including
competition, prudential and consumer regulations. Regulatory barriers are lowest under
the baseline compared with alternative models.
• Structural barriers — because superannuation is an investment service, there may be
some structural barriers to new superannuation funds entering the market. These
include advantages from economies of scale, large initial sunk costs, and access to
distribution channels.
• Strategic barriers — incumbency may act as a strategic barrier. For example, an
incumbent fund can benefit from low product switching as a result of consumer inertia
and can also benefit from advertising historical investment performance.
While these barriers to entry persist under the baseline, they will generally also be present
under any alternative model. Further, these barriers tend to resemble those found in similar
competitive markets (box 4.2). And while they may present some challenges for new
entrants, they are not expected to be high enough to significantly diminish contestability.
The more important question may thus be the degree to which any alternative model adds
to the barriers to entry.
Along with barriers to entry, barriers to exit — the ability of providers to leave the market
or merge with more efficient competitors — is an important condition for competition that
drives efficiency.
Under the baseline (and all the models), the complex process for funds to wind up and exit
the market is likely to remain, and there are some regulatory barriers to mergers along
competition, prudential and consumer protection lines. For example, some have argued that
bulk transfer rules — which require the successor fund to give the new members
equivalent rights to those in the original fund — restrict potential mergers (chapter 3).
The case of fund exits relative to the baseline will be particularly interesting — models that
deprive most of the current 100 odd MySuper products of access to default inflows will
undoubtedly hasten exits. Indeed, APRA stated that consolidation of funds via mergers
may promote efficiency in the system (APRA, sub. 32, p. 6 to PC 2016). While it is

THE BASELINE
DRAFT REPORT
107
unclear whether the baseline would encourage mergers, some alternative models may
improve on this situation.
… and there is strong rivalry between superannuation funds
The baseline also features a high degree of substitutability between superannuation
products on the supply side. This is because many products are somewhat homogenous and
employees are free to switch to new products. Given the baseline allows any fund to
compete in the market, there would be many products available for employees to substitute
into. There are currently more than 40 000 investment options in the market (APRA,
sub. 32, p. 7 to PC 2016), and this high degree of substitutability could also be expected
under the baseline.
Substitutability of products also drives rivalrous behaviour between superannuation funds,
increasing the pressure to compete by lowering costs or increasing value to retain or attract
employees. A common indicator of rivalry in the superannuation system is market
concentration. Given there are more than 200 superannuation funds in the market,
concentration under the baseline is low. Further, market concentration is likely to be low as
there are low barriers to entry.
Strong competition can also be an important driver of innovation in the superannuation
system. In a highly competitive market, funds would face strong incentives to innovate and
become more productive in order to extract or retain profits and/or market share. Under the
baseline, strong competition between superannuation funds can be expected to lead to
more innovation over time (FSC, sub. 38, p. 17).
Overall, the factors that make up competition in the superannuation system —
contestability and rivalry — are strong, and tend to mirror those found in similar markets,
such as residential home loans, electricity and private health insurance (box 4.2).
There is scope for unhealthy or wasteful competition
Competition can improve outcomes for members, by either reducing prices (primarily fees)
or increasing service quality (such as investment performance), or both. However, as noted
by many inquiry participants, competition is a means to promoting efficiency, not an end in
itself (for example, AIST, sub. 28, p. 3). This sentiment was also emphasised by the
Commission in its stage 1 study.
There may be scope for unhealthy or wasteful competition under the baseline, whereby
funds compete on attributes that are not in members’ best interests and therefore
detrimental to the efficiency of the system. In particular, superannuation funds may have
an incentive to differentiate their product on non-price aspects and spend more on
advertising (PC 2016, p. 107).

DRAFT REPORT
Some funds may differentiate their product by offering highly customisable investment
options, perhaps in an attempt to attract or retain a small subset of members. Such products
may offer an immediate benefit to the member, but their higher cost base will significantly
erode that member’s ultimate retirement balance. These problems can be exacerbated if
funds do not adequately disclose the costs of these features.
Further, competition can be damaging to members’ interests if it leads to advertising
expenditure beyond what is socially optimal, although optimality may be particularly
difficult to determine in such an environment. Several participants noted the potential for
high marketing costs under an active choice process (Club Plus Super, sub. 32, p. 2; Vision
Super, sub. 4, p. 10).
In general, the Commission considers that increasing competition in the market will
invariably lead to increases in marketing and distribution costs.
There may also be situations where superannuation funds offer marketing incentives
directly to employees in order to attract new members to the fund. For example, a fund
may offer a discount on its administration fees for the first few years if an employee joins
the fund. Currently, it is not illegal for a superannuation fund to give benefits to an
employee as an incentive to choose their fund (ASIC, sub. 41, p. 4). Incentives start to
become problematic when they are unrelated to the product and take advantage of
behavioural biases by offering a short-term benefit at the expense of the member’s
long-term interests — for example, a free gift card or petrol voucher. However, drawing a
line between a legitimate member benefit and an unhealthy inducement is nigh on
impossible (chapter 5). While the extent to which this practice occurs in the current system
is unclear, the risk of such behaviour by providers would likely be greater under the
baseline.
These product differentiation and marketing strategies may be further amplified by the
decision making difficulties faced by employees. If some employees fail to switch to
superior products over time, this can reduce the pressure on rivalry between funds.
Overall, the baseline can be expected to encourage strong competition. However, there is
scope for unhealthy and wasteful competition to impose some costs under the baseline,
reducing its efficiency.
Integrity
The third assessment criterion focuses on whether the allocation process promotes a
high-degree of integrity in the selection and delivery of default superannuation products,
and the ongoing behaviour of superannuation funds more broadly.

THE BASELINE
DRAFT REPORT
109
The number of principal–agent relationships are minimised …
The nature of the superannuation system and the number and diversity of its participants
have led to many principal–agent relationships: between employees, employers, funds,
service providers, insurers and regulators. Problems can arise when there is a conflict of
interest between the parties and the agent has the opportunity to exploit asymmetric
information to act in their own interest to the detriment of the principal.
There are two determinants that impact the extent to which principal–agent relationships
create problems.
• The number of relationships that exist between the employee and intermediaries. The
greater it is, the greater the opportunity for intermediaries to exploit conflicts of
interest.
• The extent to which conflicts of interest in these relationships impact on the employee’s
best interests.
Across all of the allocation processes, active choice minimises the number of these
relationships. This is because the baseline requires all employees to have a direct
relationship with their superannuation fund, minimising the number of formal relationships
with intermediaries. This can reduce the opportunity for conflicting interests to lead
employees astray.
That said, some intermediary relationships can provide value to members. For example,
employers, unions and consumer groups can still seek to negotiate wholesale discounts on
fees with superannuation funds to provide a superior product for a group employees.
However, under unassisted active choice there are no regulations dictating this type of bulk
purchasing and the employee would still have to choose to be involved.
… but some scope for conflicts of interest still exists
The majority of Australians receive advice on financial matters, often from a variety of
sources, such as specialists, banks, family or friends, superannuation funds and employers
(ANZ 2015, p. 45). This can be very useful to members who want expert advice and help.
However, this can become a problem when they are influenced by advice from parties with
conflicting interests.
In principle, the more choice afforded by an allocation model, the greater the opportunity
for advising parties to influence an employee’s decision. This is particularly pertinent in
the open choice world of the baseline. In the absence of a default option, new entrants to
the workforce could be influenced to choose a particular fund by another party which has
an affiliation or interest with a particular superannuation fund. The Australian Institute of
Superannuation Trustees (sub. 28, p. 47) suggested that, under the baseline, new
employees may essentially default to bank-related superannuation funds simply because
this is the only financial relationship they have.

DRAFT REPORT
That said, while there may be scope for influencing members under the baseline, it would
be a mistake to assume that those risks would not exist in any world involving significant
rights to choose, including the current system. Moreover, since employers have no direct
control over their employees’ superannuation fund under the baseline, there is possibly less
scope for them to influence their employees’ decisions compared with an employer choice
allocation mechanism.
Another relationship with potential for conflicts of interest is that between members and
financial advisers. Under the baseline, the influence of financial advisers would most likely
be higher than today. Yet the quality of that advice may not improve. Outcomes under the
baseline benefit from the Future of Financial Advice reforms in 2013, whereby financial
advisers face a legal requirement to act in the best interests of their clients (ASIC 2016).
In assessing alternative models against the baseline, the potential for conflicts of interest
will need to be addressed. Similarly, any regulations designed to protect members can give
rise to new, potentially conflicted, relationships. But since a rollback of consumer choice is
not under consideration (chapter 2) it is not a fatal blow to either the baseline or any of the
alternative models that such behaviours may be present.
Stability
The fourth assessment criterion focuses on determining whether the allocation process
creates instability in the superannuation system that leads to significant systemic risks.
This can include excessive and volatile movement of members and assets between funds,
unsustainable loss leading strategies or excessive market concentration.
Under the baseline, flows of members and assets between superannuation funds would
likely rely on individual decisions and be driven by competition in the market in a similar
way to other financial services, such as banking and insurance.
[Active choice] could also lead to relatively stable outcomes as individual decisions would be
unlikely to lead to large shifts in market shares in the short term, with funds both gaining and
losing members over time. (ASFA, sub. 24, p. 21)
Therefore, there would be no reason to expect sudden large-scale changes in the flow of
members or assets between funds that impact on fund liquidity or peer risk.
Further, the baseline does not explicitly reward funds attempting to adopt an unsustainable
loss leading strategy, and competitive market forces will generally act to prevent this
behaviour or limit its effectiveness in the long term.
Finally, as noted above, there is no evident reason to believe the baseline would lead to
market concentration between funds or upstream providers. Importantly, the baseline is
still subject to current competition regulations. And like other financial institutions,
superannuation under the baseline would still face prudential supervision by APRA, which
monitors fund behaviour and industry trends in order to maintain a stable financial system.

THE BASELINE
DRAFT REPORT
111
Therefore, the baseline is unlikely to create instability in the superannuation system that
leads to significant systemic risks.
System-wide costs
The final assessment criterion focuses on minimising overall system-wide costs, taking
into account costs to members, employers, funds and government.
The cost of allocation ultimately falls on members
In an active choice world, members ultimately face the cost of choosing products. These
members may face high search costs, and it has been suggested that transaction costs might
also be significant.
Search costs mainly involve the time and effort expended by the employee to find,
compare and evaluate products to determine which one is in their best interest. While the
materiality of these search costs is unclear, a 2016 regulation impact statement estimated
them at about $87 per person based on an assumption that three hours would be required to
research and choose a product (O’Dwyer 2016b, p. 25). Search costs may also include the
upfront cost of purchasing professional advice to help make the decision. Therefore,
imposing these search costs on a large number of members within the system means
greater search costs overall at the system level. That said, disengaged members may forgo
thoroughly searching for a product and rely on heuristics. In this case, the search costs
would be lower, but so would the quality of the decision.
Members will also be faced with transaction costs: under the baseline, employees need to
join the superannuation fund first and then provide the product details to the employer.
Further, superannuation funds face the transaction costs associated with transferring
members between funds. These switching costs may actually be desirable if they are the
result of increased member engagement. But to the extent that suboptimal advertising
encourages churn, this would be undesirable. Overall, there is no way of accurately
attributing these effects.
There may be some merit in the argument that purchasing superannuation products as an
individual is also likely to lead to higher costs relative to a group of employees negotiating
on a wholesale basis (ASFA, sub. 24, p. 21). That said, the baseline has no restriction on
the ability for employees or an intermediary negotiating a group discount on fees.
Minimal burden for most employers and superannuation funds
Under the baseline, the employer’s role in the superannuation system is limited to making
Superannuation Guarantee contributions on behalf of employees. Therefore, employers
face potentially low regulatory and compliance burdens (ASFA, sub. 24, p. 21; FSC,

DRAFT REPORT
sub. 38, p. 16). Any employer wishing to be involved in their employee’s superannuation
decision can still offer a corporate fund for their employees to choose from or an education
service to help members make a choice.
While there is an argument that employers may be required to send contributions to many
different funds under the baseline, superannuation clearinghouses and the introduction of
the ATO’s Single Touch Payroll system are likely to alleviate this burden to a large extent
(chapter 3).
Despite significant member engagement under the baseline, employers may face the
burden of making sure all of their employees choose a product. This is because, if an
employee does not choose a product, there is a risk that the employer does not comply with
their Superannuation Guarantee obligations. It is difficult to estimate the extent of this
burden at the system level under the baseline, but as already discussed, it is likely to be
low. Further, it is worth noting that after 25 years of experience, most employers will be
aware that failure to make employer contributions inevitably invites penalties.
Superannuation funds also face low regulatory burden. For example, funds would not face
the costs associated with MySuper regulations. This minimises the regulatory costs that are
passed through to members.
On the other hand, as noted above, intense competition between superannuation funds can
be expected to spur significant marketing expenses aimed at retaining or attracting
members. Not only will superannuation funds initially face higher marketing expenses, it is
likely that at least a proportion of these costs will ultimately be passed through to members
as fees (Vision Super, sub. 4, p. 10).
The overall cost to government
Under the baseline, there is likely to be little or no ongoing administration costs to
government (ASFA, sub. 24, p. 21). The Government would have no additional costs
above the ordinary competition, prudential and consumer protection regulations it currently
provides to all financial markets. It is nevertheless possible to envisage that the ATO, with
its sophisticated electronic connections to payroll providers and other employers, might
face greater calls for proactive scrutiny that contributions are paid by employers. The
Single Touch Payroll system appears capable of being used in such a fashion, albeit at a
cost.
To the extent that there is an increase in members choosing poorly under the baseline, the
government is likely to face greater costs in the long term. These members are more likely
to rely on the Age Pension to provide or top up their retirement income (Mercer, sub. 15,
p. 4). This may ultimately lead to significant fiscal costs to the Australian Government,
borne by taxpayers. However, the extent to which this might occur is highly speculative.

THE BASELINE
DRAFT REPORT
113
Overall assessment
Overall, the baseline exposes some key problems with a completely unassisted approach to
active choice allocation, which alternative models should aim to better. It also
demonstrates several benefits that would be worth retaining. Table 4.1 presents a summary
assessment of the baseline.
Table 4.1 The baseline: assessment summary
Criterion Assessment
Member benefits Members are free to act in their own interests and requiring active choice can
trigger member engagement.
However, cognitive constraints, heuristics, behavioural biases and problems with
information can lead some members to make poor long-term financial decisions.
Many superannuation funds can be expected to aim to attract new members, and in
doing so improve the availability of information.
Competition Competition between superannuation funds is highly contestable and rivalrous.
There are many similarities with other active choice markets.
However, there is scope for superannuation funds to engage in unhealthy and
wasteful competition that is not in the interest of members.
Poor decision making and inertia can lead to low demand-side competition.
Integrity Members are required to engage directly with superannuation funds.
But there is potential for conflicts of interest, for example with financial advisers,
which can negatively influence member decision making.
Stability Individual decisions by members are likely to lead to small and predictable member
flows. This is supportive of system-wide stability.
System-wide costs The cost of allocation, particularly search costs, ultimately falls on members.
A minimal role for most employers leads to low administrative burden.
Superannuation funds face minimal regulatory burden, but are likely to incur
significant marketing costs which are likely to be passed on to members.
The overall cost to government is likely to be low.
Retirement planning is inherently complex and decisions need to be made over the long
term. Cognitive constraints, heuristics, behavioural biases and imperfect information can
lead some members astray, making poor decisions which are not in their best long-term
interests. Thus members can potentially select products that do not meet their needs, such
as products with high fees, poor investment performance, unnecessary services or overly
conservative investments. Alternative models should aim to improve employee decision
making or guide them to high quality products in other ways.
That said, the baseline benefits from giving all employees the freedom to act in their own
best interest and prompting more employees to engage meaningfully with their
superannuation. Alternative models should aim to retain these benefits.
Under the level playing field of the baseline, superannuation funds are expected to face
strong competition. However, the baseline reveals that there would be scope for
superannuation funds to engage in unhealthy or wasteful competition. Alternative models
should aim to reduce these undesirable features.

DRAFT REPORT
Importantly, the baseline acknowledges that members are likely to lean more heavily on
other parties for advice, carrying with it the potential for conflicts of interest that could
threaten the integrity of the superannuation system. Alternative models should adequately
manage the risks caused by conflicts of interest.
The baseline is unlikely to cause instability or lead to systemic risk. Alternative models
should do likewise.
Finally, although employer and superannuation costs are likely to be low under the
baseline, members ultimately face high costs of allocation. Alternative models should aim
to reduce search costs faced by members, such as effort and advice.

ASSISTED EMPLOYEE CHOICE
DRAFT REPORT
115
5 Assisted employee choice
Key points
• This chapter sets out the first of four alternative allocation models for a new default system.
• A model of assisted employee choice would consist of:
− a non-mandatory product shortlist and associated information to nudge people towards
a small set of good products
− a set of protections to help employees choose from a broader set of products and
reduce the risks of not making a choice.
• The product shortlist would reduce the complexity of choosing a superannuation product for
a majority of employees by nudging them towards good products without going so far as to
impose one by default. This would considerably reduce ‘choice overload’ and make
disengaged employees more likely to make good choices.
− The shortlist would comprise (indicatively) 4 to 10 products, selected through a rigorous
administrative filter process run by an independent government body.
− Simple information on the shortlisted products would be provided in a consistent format
to help people choose between options and consider which product best meets their
needs, with care taken in the presentation to avoid undesirable biases.
• Additional protections would help employees to compare products more broadly to identify
one that best meets their needs, and reduce some of the material risks under active choice.
− A system of voluntary product accreditation (a strengthened version of MySuper) would
encourage funds to provide a ‘no frills’ product for employees who are not well placed to
select specific investment options themselves.
− A last-resort fund would hold the contributions of employees who do not make prompt
choices until they have chosen a product.
• Relative to the baseline, the assisted employee choice model performs well against the
assessment criteria and would:
− significantly reduce the complexity people face and lead to more people choosing
high-performing products that meet their needs
− focus competition on product aspects of value to members, put downward pressure on
fees and potentially curtail wasteful product proliferation
− better align funds’ interests with those of members
− have lower search costs for many employees, partly offset by additional costs to funds
and government associated with new regulatory structures.

DRAFT REPORT
The previous chapter described what unassisted active choice of superannuation products
might look like: some employees may struggle to identify which product best meets their
needs, may procrastinate and put off making a choice, or could be swayed by marketing or
conflicted interests to choose a product that leaves them worse off in the long term.
This chapter presents a model of ‘assisted’ choice that combines active choice by
employees with guidance and a set of protections (table 5.1). First, a shortlist of products
would ‘nudge’ employees towards good superannuation products, supported by the
provision of simple and comparable information on key features of those products
(section 5.1). Second, a set of regulatory protections, comprising a voluntary system of
product accreditation and a ‘last-resort’ fund, would help employees to compare products
more broadly and reduce the risks of not making a choice (section 5.2). Compared to the
baseline, this model performs well against the Commission’s assessment criteria
(section 5.3).
Table 5.1 Summary of key features of assisted employee choice
Main design element Design features
Product shortlist Employees are presented with a shortlist of (indicatively) 4 to 10 products,
listed in random order.
These products are selected by an independent government body on the
basis of being good products for a majority of employees, following a
rigorous administrative filter process.
Employees remain free to choose any legally complying product that is not
on the shortlist.
Information on shortlisted
products
Information on the shortlisted products is provided to employees in a
consistent format to help employees choose between options and consider
which product best meets their needs.
The information is simple and covers investment performance, risks and
fees (with the specific metrics to be based on detailed consumer testing).
Additional information on the products remains available to employees
through other channels.
Product accreditation A government body administers a system of product accreditation to
encourage funds to provide a ‘no frills’ product for employees who are not
well placed to select specific investment options themselves.
Accreditation is similar to MySuper, but with stronger standards on fund
scale and performance, and clearer processes for removing accreditation
from non-complying products.
Accreditation is voluntary for funds. Employees remain free to choose any
legally complying product, regardless of whether it is accredited.
Last-resort fund A low-cost fund is established to hold the contributions of employees who
do not nominate their own superannuation product. The fund is tightly
regulated and could be based on existing funds for lost accounts, or
provided by the Future Fund.
This fund has a remit to encourage its members to select a product from
another provider in the market. Once they do, members’ balances are
transferred to their chosen provider.

DRAFT REPORT
117
5.1 A shortlist of products
The central feature of the assisted employee choice model is a shortlist of superannuation
products. In essence, employees would be given a list of high-quality products to pick
from, while retaining the option to nominate any other product in the market. The shortlist
would be supplemented by information on the key features of those products to help
employees make an informed choice about which product best meets their individual
needs.
There is a strong body of evidence that shows that the way choices are presented and the
way information is provided can have powerful impacts on people’s choices. This evidence
from behavioural finance has inspired the use of ‘nudges’. Thaler and Sunstein (2009)
defined a nudge as an aspect of the way that choices are presented that alters behaviour
towards options that make people better off, without forbidding any options and making it
easy and cheap for people to choose a different option if they wish.
A nudge is not the same as a default. While defaults can be a form of nudge, it is still
possible to steer people towards particular options without any form of automatic
assignment. Sometimes this will be preferable, especially where there are strong
advantages in encouraging each individual to consider the available options, or there are
concerns about potentially misleading or harming consumers through the imposition of
defaults (Clark, Strauss and Knox-Hayes 2012, p. 142).
A shortlist of superannuation products would fulfil this definition of a nudge. The basic
principle is that the choice ‘architecture’ is modified to simplify the set from which people
can choose, and to steer them towards selecting a higher-quality product than they might
otherwise — but without going so far as to impose a product chosen by somebody else as a
default. Employees would remain free to nominate a product not on the shortlist —
including a product sponsored by an employer or union — or to self-manage their
superannuation. In this way, they must still make an active choice, but are assisted in the
task.
The main benefit of the shortlist is that it would make it easier and simpler for employees
to choose a good product that meets their needs — especially for those who have limited
financial knowledge or would put little effort into researching products. Even employees
who select randomly from the shortlist would be better off than choosing a bad product or
making no choice at all, compared with the baseline of unassisted active choice (chapter 4).
The shortlist approach also differs from the present system in that there is no compulsion
to use the list, whereas today various instruments apply compulsion to many employees’
choice processes.
Yet this shortlist will have costs. Which products make it on the list — and who selects
them — would have a powerful influence on which products many employees choose.
Accurate and comparable information can be difficult to provide, and may not reflect all
product features that matter to everyone (though employees would remain free to seek

DRAFT REPORT
further information on the products from other sources). The costs and benefits are
explored in greater detail in section 5.3.
Which products should be shortlisted?
The shortlist would only contain products that have been assessed as being good options
for the majority of employees on the basis of a ‘heavy’ administrative filter, with the
selection criteria and process the same as detailed for the preferred default list in chapter 8
(each fund would also need to offer an accredited product, as defined in section 5.2). This
would mean that each product would need to meet criteria for long-term net returns,
investment approach, governance and other factors. Importantly, the stringency of the
criteria will need to be adjusted such that the number of successful funds that appear on the
shortlist is within an acceptable range (as described below).
Figure 5.1 provides an example of what the shortlist might look like on a standard choice
form. When an employee selects one of the products on the list, an account with the
relevant fund would be opened on their behalf to receive their superannuation contributions
(by transmitting relevant information from the choice form to the fund and employer). This
would result in lower transaction costs for employees compared with the current rules,
whereby employees wishing to nominate a specific fund must already have opened an
account with that fund, and must obtain a letter from the fund stating that it can receive
contributions from their employer (ATO 2015b, p. 2). The new process would be
facilitated by the centralised online service discussed in chapter 3, which would also allow
employees to choose an existing superannuation account they already have.
The effectiveness of the shortlist will depend to a large degree on how many products are
listed, the order they are listed in, and who compiles the list and on what basis. It will also
hinge on what supplementary information is provided about the products to help people
make a choice. This information should be unbiased and readily comparable across
products. The Commission does not envision a shortlist without any supplementary
information to assist employees.
To better understand how employees would use a shortlist and supplementary information
in practice, the Commission has commissioned an experimental survey of the working age
population (chapter 1). Among other things, this survey examined how people’s choices
were influenced by the length of the shortlist and the types of product information
provided. The work is underway but not completed. The results will be analysed following
publication of this draft report and incorporated in the Commission’s final report.

DRAFT REPORT
119
Figure 5.1 Shortlist of products — example only
Based on ATO (2015b, p. 5).
How many products should be shortlisted?
The length of the shortlist will bear on its effectiveness. Several studies have found that
presenting too many options can overwhelm people and lead them to make poor choices
(box 5.1). Such ‘choice overload’ may mean that a very long list of products is no more
helpful than no list. However, evidence varies for how long is too long.
Restrictions on the number of options presented have been put forward as one way to
reduce choice overload and help people to make better choices (Abaluck and Gruber 2011,
p. 22; Ketcham, Kuminoff and Powers 2016, p. 2). In a review of the literature, Johnson et
al. (2012, p. 490) suggested that four or five options is a reasonable initial value for a
choice set, potentially with an option for people to see more options if they wish. Other
reviews point to the importance of contextual and individual factors, noting that the
amount of choice overload people experience depends on the complexity of the decision
Employee to complete
Choice of superannuation product
I request that all my future super contributions be paid to (select one
option only):
 My existing superannuation account (Western Super)
OR
One of the following shortlisted products:
 Northern Super – Super Saver
 ABC Super – Balanced
 XYZ Fund – Invest for the Future
 Southern Super – Aspire to Retire
 Investment Expert Super – Balanced
 JKL Fund – Balanced
 Super Duper – Nest Egg
Information on these products can be found below
OR
 The APRA fund or retirement savings account I nominate (provide
details below)

DRAFT REPORT
and choice set, how certain people are in their preferences, and how much effort they are
willing to put into the decision (Chernev, Böckenholt and Goodman 2015). Financial
literacy is also likely to play a role.
Box 5.1 Evidence of choice overload in financial decisions
Behavioural researchers in the United States have found that asking people to select from long
lists of financial products can quickly lead to them feeling overwhelmed, especially when they
have lower levels of financial knowledge. In a well-known experiment, Iyengar, Huberman and
Jiang (2004, pp. 88–91) found that every additional 10 products on a list of retirement savings
product options was associated with a 1.5 to 2 per cent fall in scheme participation, with
participation rates falling fastest when more than 30 products were listed.
A separate study of superannuation product choices found that an increase in the number of
options from 10 to 19 led to a jump in the proportion of employees who ended up in the default
option from one fifth to one third (Morrin et al. 2012, p. 546). Other research has found that the
more investment options people are presented with, the more likely they are to opt for the status
quo (a product they already have) (Samuelson and Zeckhauser 1988, p. 8). In all these cases,
adding more options appears to discourage people from selecting any of them.
There is also evidence that people struggle to make good choices when the number of options
is large. In one study, researchers found that halving the list of investment options (from about
90) led to employees choosing options with significantly lower costs, which they estimated
would generate large financial savings over time (Keim and Mitchell 2016). Similarly, research
on health insurance has found that a high number of product options is associated with
poorer-quality choices: people are more likely to select expensive products and overlook
products that better meet their needs (Abaluck and Gruber 2011; Johnson et al. 2013).
However, shorter is not always better. Too few products could increase the risk of people
selecting a product that does not meet their needs, or leave people more susceptible to
framing effects (discussed below). It could also reduce welfare for people who would be
better off under options that are not included in the set (Ketcham, Kuminoff and
Powers 2016, p. 34).
The supply side of the market is relevant too. In the event that there is a moderate number
of ‘good’ products, such that it is difficult to definitively classify some as better than
others, artificially reducing the number could cut out products that are otherwise
worthwhile. There is also a risk that a very short list could restrict competition between
providers, depending on how frequently the list is changed (ASFA, sub. 24, p. 23).
Drawing together these considerations and the available evidence, it would appear that a
shortlist should ideally contain somewhere in the range from 4 to about 10 products. Fewer
than four would likely curtail choice too much; more than 10 would likely lead to choice
overload for some employees and reduce the value in having a shortlist at all. The
Commission expects to refine these bounds using the results from its survey, and is also
asking participants for additional evidence that is relevant to this matter. Ultimately, the
actual implementation of a shortlist would need to be preceded by consumer testing and/or
trials to determine what works best in this specific superannuation context.

DRAFT REPORT
121
How should the list be ordered?
The order in which products are listed also matters. Psychologists have long recognised
that the order in which people are asked questions has a detectable influence on their
responses (Krosnick and Alwin 1987). In a well-known experiment, Benartzi and Thaler
(2002, p. 1610) found that people were averse to picking extreme options when selecting
from a set of investment portfolios ordered from low to high risk, to the extent that their
preferred option among two alternatives changed when a third option was presented either
to the left or right side of the pair.
In a context where all the products on the superannuation shortlist are ‘good’, their order
should be randomised to remove any ordering biases. This approach is already used for
election ballots and in other contexts (such as surveys). Randomisation would be relatively
straightforward for an online choice form (as part of the centralised online service
described in chapter 3), and the order could be re-randomised at regular intervals.
What information should be provided?
Providing information on key features of the products on the shortlist would further help to
nudge people towards choosing good products that suit their needs.19 Some studies have
found that people often fail to seek out information on financial products, even when it is
easy to access, and thus providing it to them directly can lead to better choices (Kling et
al. 2012). More generally, a review of research on financial education found that it is
usually most effective when provided ‘just in time’ for people to make a decision
(Fernandes, Lynch Jr and Netemeyer 2014).
Providing relevant product information alongside the choice form would reduce the search
costs of employees finding this information themselves, and would make it easier for them
to compare the products. Several inquiry participants supported government provision of
basic information about the features and performance of selected products (CHOICE,
sub. 31, p. 14; CMD, sub. 18, p. 8; IPA, sub. 17, p. 8). The consumer group CHOICE
(sub. 31, p. 10) argued that its market research showed that ‘many consumers had lost faith
in the super funds or financial advisers to help them with decisions around
superannuation’.
The amount and presentation of information matter
To be effective, information should be relevant, easy to understand, comparable, and
sufficiently complete so that people can make the right choices — especially if they are
19 In principle, such information could be provided without a shortlist, though the products featured would
need to be determined through a similar process (such as an administrative filter) — and since only a
limited number of products could realistically be featured, such information would in any case be a de
facto shortlist.

DRAFT REPORT
unlikely to seek information from other sources. Helpful (versus defensive) disclaimers
should accompany the information to explain how it was sourced, where to find further
information or assistance (such as on fund or government websites), and the need for
employees to consider their individual circumstances.
Providing simple yet effective information is easier said than done. A large volume of
complex information exists for individual superannuation products, and distilling this to
the attributes that matter most will be difficult. Providing too much information risks
confusing people or deterring them from reading it. People can be overwhelmed when
presented with too many different attributes and tend to respond by using a single attribute
to simplify their decision (Johnson et al. 2012, p. 495). There is also evidence that
providing people with information on particular attributes of superannuation products
(namely asset allocations) can lead them to make worse choices than had these attributes
been omitted (Bateman et al. 2016b).
On the other hand, providing too little information may mean that employees are unaware
of important product features that they care about. The Australian Securities and
Investments Commission (sub. 41, p. 2) submitted that that ‘being able to provide the right
information which enables effective comparability of funds is of critical importance’.
The way that information is presented also matters. There is considerable evidence that
decision making is subject to biases that arise from the way choices are framed (box 5.2).
Some experts have even presented evidence that the way options are presented can have a
stronger influence on people’s investment decisions (in aggregate) than the underlying risk
and return characteristics of the investments (Mitchell and Utkus 2004, p. 16). The
literature offers evidence favouring a range of techniques to minimise biases, such as
including graphs to aid decisions (Bhandari, Hassanein and Deaves 2008, p. 399),
highlighting the consequences of each option (Keller et al. 2011, p. 382), and presenting a
small number of information items for multiple products simultaneously to make it easier
for people to compare specific attributes (Bateman et al. 2016a, p. 20).
Metrics for the shortlist
Taken collectively, the above evidence suggests that the information provided with the
shortlist (that is, alongside the superannuation choice form) should err on the side of being
simple rather than comprehensive. Additional information would remain available to
employees through other channels. Funds would still be required to issue product
disclosure statements and comply with other reporting requirements, and private-sector
information (including product ratings and comparison tools) would be available
(chapter 4).

DRAFT REPORT
123
Box 5.2 Framing effects on decision making
Psychologists have long recognised that people have limited capacity for mentally processing
information (Benartzi and Lehrer 2015; Mayer and Moreno 2003, p. 43). As a result, they may
resort to heuristics (mental shortcuts), such as focusing on more salient information or using
rules of thumb to simplify decision making (Kahneman 2011). This means that the way
investment information is presented can influence the choices people make (Benartzi and
Thaler 2007, pp. 92–94). However, these influences can be complex and difficult to predict.
Scale and time periods
Psychological experiments have shown that people are especially sensitive to how numbers are
presented, and in particular the scale and time period. A common finding is that people appear
more willing to take risk when returns are presented over a longer time horizon. For example,
Benartzi and Thaler (1999, p. 377) found that survey participants said they would invest more of
their retirement savings in shares — and thus were willing to accept more risk — when they
were shown 30-year rates of return rather than 1-year returns. Other studies have found that
people’s willingness to accept a given risk can change depending on whether it is framed as a
probability of success or of failure (Kahneman 2011), and that people are likely to take more
risk when investment choices are presented graphically (Kaufmann, Weber and Haisley 2013,
p. 325).
Another key research finding is that people tend to give more weight to fees when presented in
dollar terms rather than percentages (Bateman et al. 2016a, p. 13). For example, in consumer
testing for MySuper dashboards, many participants did not notice when the scale on an annual
return graph was varied (from a maximum of 10 per cent to 20 per cent) (ASIC 2013, p. 32).
Anchoring and availability biases
There is evidence that people’s responses to questions can be influenced by the presence of
other numbers they have recently seen, even when these numbers are clearly unrelated
(anchoring biases). There is also evidence for a tendency for people to use only the information
in front of them when making a decision, while neglecting other relevant information that is not
immediately to hand (availability biases) (Kahneman 2011).
Diversification
Framing effects have been shown to have a particularly strong impact on asset allocation
decisions. In a widely cited study, Benartzi and Thaler (2001, p. 79) found evidence that, when
presented with a number of investment options, people tend to allocate their assets evenly
among them regardless of what those options are — which they term the 1/N heuristic. Others
have found that when a large number of options are available, people tend to choose a subset
and divide assets equally among them (Gerrans and Yap 2013, p. 1; Huberman and
Jiang 2006, p. 763).
Bateman et al. (2016b, p. 60) presented experimental evidence that this phenomenon applies
for pre-mixed investments, with participants tending to favour portfolios with more evenly sized
shares of assets across categories, even when this comes at the expense of lower returns or
higher risk. As a consequence, presenting information on asset allocation can distract people
from focusing on the underlying risks and returns.

DRAFT REPORT
Designing simple metrics for superannuation products has been a long-running challenge,
especially where the objective is to inform people who are relatively disengaged with
superannuation or have low levels of financial literacy. Some stakeholders are of the view
that existing disclosures do not make it easy for people to compare products across funds
(for example, AIST, sub. 28, p. 4; ISA, sub. 40, p. 15). Bateman et al. (2016a, p. 7) argued
that Australian regulators have tended to focus on how well consumers understand the
information rather than whether it assists in decision making. They noted that research on
simplified financial disclosures in the United States and European Union has also found
that disclosures are not always easily understood by consumers and do not always lead to
better decision making (Bateman et al. 2016a, pp. 2, 13).
Significant work has been done in recent years to develop product dashboards for MySuper
(default) superannuation products, which were intended to make it easy for members to
compare products offered by different funds, and thereby improve competition and
transparency in the industry (Shorten 2013). However, these dashboards have been
criticised by a number of parties (box 5.3), and it is unclear how many people use them
when choosing between superannuation products.
Although universal agreement on the exact types and formats of information that should be
provided is lacking, at a minimum, the Commission considers that it would be most helpful
to provide indicators that cover longer-term investment performance (returns), risks and
fees, as well as an indication of how well a product has performed relative to the industry
average. The Commission’s experimental survey presented participants with some of these
types of information in the context of a shortlist (table 5.2). It is also requesting
participants’ views on relevant information to provide, and evidence to support this.
Ultimately, detailed consumer testing (including real-world trials) will be crucial prior to
final design and implementation to better understand how people use the information and
to refine the presentation.
Table 5.2 Product information used in the Commission’s survey
Hypothetical example only
Label Description
Return Gained X% a year (on average) over past 10 years
Risk Expect a loss in X out of 20 years
Target To increase balance by CPI + X% per year, over next 10 years
Fees Pay X% of account balance in fees per year

DRAFT REPORT
125
Box 5.3 MySuper product dashboards
MySuper product dashboards have been mandatory since December 2013. They are required
to contain, in a prescribed format, information on:
• returns for each of the 10 previous financial years (where available)
• the return target over 10 years
• a comparison between the return target and actual returns for previous years
• the level of investment risk
• a statement of fees and other costs, measured in dollar terms for a $50 000 account balance
(ASIC 2014).
MySuper was introduced after a period of consumer and industry testing by the Australian
Securities and Investments Commission. It found that consumers favour simple information,
with returns presented graphically and fees customised to individual balances (ASIC 2013).
Fee disclosure
The fee metric used on MySuper dashboards has been criticised. Fee structures can differ
significantly across funds, and so the reported dollar figure (for a representative $50 000
balance) may not be a useful guide for members with differently sized balances. The Actuaries
Institute (2016, pp. 1, 3) has recommended that dollar-value fees also be included for $10 000
and $250 000 balances, and that administrative and investment fees be separately reported.
Standard risk measure
The ‘standard risk measure’ used for MySuper dashboards — the likely number of negative
annual returns over a 20-year period — was criticised by participants in the Commission’s
stage 1 superannuation study for not conveying the likely magnitude, path or timing of negative
returns (PC 2016, p. 115). In a recent survey, participants displayed low levels of understanding
of this risk measure (Bateman et al. 2016a, p. 14). The Australian Prudential Regulation
Authority has noted that the standard risk measure was developed by the superannuation
industry, and that it has been encouraging the industry to develop other measures to
communicate risk (APRA, sub. DR111, p. 5 to PC 2016).
Some stakeholders have proposed alternative risk measures. For example, the Actuaries
Institute (2016, p. 14) suggested reporting the number of years in which returns are likely to
exceed a given benchmark (such as inflation plus 4 per cent). In experiments, participants have
been observed to make more rational investment choices when returns are presented as a
range or using probability-based descriptions, as opposed to describing risks as a frequency
above or below a given threshold (as per the standard risk measure) (Bateman et al. 2016c).
Return graphs and warnings
A recent study surveyed people to examine how they use the return and fee disclosures on the
MySuper dashboards (Bateman et al. 2016a). It found that while respondents switched between
products fairly quickly when the fee differential changed, they were more reluctant to switch
when short-term returns changed. This was the case regardless of the volatility of returns or
how they were presented (graphically or in table format), though simplifying the return
information led to prompter responses. The researchers suggested that their results may reflect
the relatively complex presentation of return information on the dashboards (as a graph) and
the prominence of warnings related to the uncertainty of future returns.

DRAFT REPORT
Who should compile the shortlist?
An independent body, established by government, is best placed to select which products
should be shortlisted (as described in chapter 3). This body would be an impartial arbiter. It
would compile the shortlist by deploying a ‘heavy’ administrative filter (chapter 8).
While it would be possible to task a private-sector provider (such as a ratings entity) or
individual employers with compiling shortlists, this could give rise to conflicts of interest.
Commercial relationships between ratings entities and particular funds — and advisory
activities undertaken by entities otherwise primarily involved in rating funds — are
common enough today to suggest that, at a minimum, perceptions of conflict would arise.
This does not, however, suggest that private ratings entities should have no role; simply
that there must be a public-interest decision maker, lest the credibility of the process be
damaged.
Inquiry participants generally supported a government-managed process. For example, the
Financial Planning Association of Australia (sub. 29, p. 13) submitted that the body
assessing funds should be independent of industry interests, and noted that there is a risk
that private ratings entities would be captured by special interests that differ from the
interests of employees. It argued that, instead, the government should provide fund
comparison data that are at employees’ fingertips when they make their choice (FPA,
sub. 29, p. 12). Other participants also pointed to the risks of relying on privately provided
information (for example, CHOICE, sub. 31, p. 10), or argued that ‘public confidence
would be stronger in a government-managed exercise’ (AIST, sub. 28, p. 51).
The main risk with a government body compiling the shortlist is excessive risk aversion,
resulting in only conservative products being listed. An additional risk is that of demands
for compensation if one of the products underperforms or leaves members worse off
(ASFA, sub. 24, p. 23; PwC, sub. 12, p. 6). It is likely that these risks could be
significantly mitigated through the selection of qualified investment advisers to the body,
through the structure of institutional or legislative arrangements, and by promoting clarity
in the voluntary nature of the information given. The current system in Australia has for
the most part successfully avoided the perception that the government guarantees positive
returns, even where criminal activity has seen significant defrauding of superannuation
investors.
A final consideration is how frequently the list is updated. Changing the list too frequently
would be administratively costly (as the filter process would need to be re-run). On the
other hand, changing it too infrequently could give some products an unfair competitive
advantage over others and risk employees being guided towards products that no longer
perform well. As discussed in chapter 3, the process should initially be run every four
years. In the event that a product loses its place on the shortlist or fails to reapply for
inclusion, and provided the fund remains legally compliant, the fund would not
automatically lose its existing members (who remain free to choose a product from a
different provider).

DRAFT REPORT
127
5.2 Other employee protections
While a shortlist should be of great assistance for many employees who would struggle to
make decisions under unassisted active choice, the benefits would mostly accrue to those
employees who act and select one of the products on the shortlist. The gap between
knowledge and action will still arise as funds that do not make the shortlist will respond
with more intense marketing efforts. And there is nothing in a non-mandatory approach
like this to ensure that the utterly indifferent employee makes any choice at all.
This section discusses a set of regulation-based protections that aim to help employees to
compare products more broadly — and ultimately select one that best meets their
individual needs — while reducing the risks of not making any choice at all. These
interventions are targeted at specific problems that an examination of the world of
unassisted active choice demonstrated (chapter 4), including the complexity of comparing
different products, asymmetries in information between funds and employees, and the risk
that employees who fail to make any choice will forfeit their superannuation contributions.
These protections form part of the assisted employee choice model, and could apply across
the superannuation market generally. They could be implemented separately or in
combination. The details set out here are illustrative; ultimately, any regulatory change
should only proceed if it offers a net benefit to the community. The Commission’s final
(stage 3) inquiry will consider that.
Product accreditation
The difficulties that most employees would face in selecting a superannuation product
under active choice would only be partly ameliorated by a shortlist. Those who wish to
compare other products available in the market may still find it difficult to compare
product features and performance, especially if they have limited knowledge, capacity or
confidence to shop around.
The Commission sees merit in a product accreditation regime as a way of engendering a
harmonised subset of simple products in the market. Accreditation would be voluntary for
funds, and funds could apply at any time (though accreditation would be mandatory for
funds seeking inclusion on the shortlist discussed earlier). Any product that meets a
minimum set of standards could apply to be assessed by a regulatory body — in this case,
the Australian Prudential Regulation Authority (APRA) — and given a stamp of
accreditation. The criteria used to assess products would be very similar to those used to
determine which products are on the ‘minimum standards’ list described in chapter 8.
The standards would specify a ‘no frills’ product that provides administration, investment
and intrafund advice services. Each fund’s product would contain a single diversified
investment strategy (with funds permitted to offer life-cycle variations) and single fee
structure for all members. But a specific asset allocation or investment strategy would not

DRAFT REPORT
be prescribed, leaving funds free to tailor their products to the needs of different types of
members. Funds could use the accreditation in their marketing and promotional activities.
This regime would have many similarities to the current MySuper regulations, which are
administered by APRA (table 5.3). The intention of these reforms was for a ‘no frills’
product with simple features that are easy to compare across funds (Australian
Government 2011, p. 3). It was not intended to preclude anyone from using a more
complex or tailored choice product, or from self-managing their own superannuation
(Cooper et al. 2010, p. 10).
Table 5.3 Comparison of MySuper and product accreditation
Characteristic Current MySuper
authorisation process
Product accreditation in the
Commission’s model
Similarities
Accumulation product
features
Administration, investment,
intrafund advice
Administration, investment, intrafund advice
Investment strategy Single diversified strategy or set
of life-cycle strategies
Single diversified strategy or set of life-cycle
strategies
Fee structure Single fee structure that applies
to all members
Single fee structure that applies to all
members
Exit fees, switching fees,
and buy–sell spreads
Levied on cost recovery basis Levied on cost recovery basis
Commissions Prohibited Prohibited
Number of accredited
products per fund
Onea (including any life-cycle
variations)
One (including any life-cycle variations)
Governance standards Funds must be likely to comply
with trustee obligations
Funds must be likely to comply with trustee
obligations
Differences
Need for accreditation Funds must obtain MySuper
authorisation to receive default
contributions
Voluntary for funds (but mandatory for funds
seeking shortlisting)
Accreditation process One-off licencing, with ongoing
monitoring
Formal ongoing monitoring of fund
performance following accreditation, with
funds required to self-report against criteriab
Investment performance Funds must report return and
risk targets, but no specific
performance standard
Products that persistently underperform
(over multiple years) relative to pre-stated
targets lose accreditation
Fund scale Funds must demonstrate that
the scale of the fund does not
disadvantage members relative
to members of other MySuper
products
Funds must regularly demonstrate, and
have independently verified, that fund scale
is not an impediment to performance
Insurance Life cover and total and
permanent disability cover must
be provided on an opt-out basis
No requirements, but any insurance must
meet minimum regulatory standards
(chapter 3), and details must be clearly and
accessibly disclosed
a In some circumstances, funds can offer different administrative fees to members who are employees of
large employers. b Self-reporting is to be backed up by independent auditing by, or on behalf of, APRA.
Sources: Australian Government (2011); PC (2016).

DRAFT REPORT
129
Inquiry participants generally supported some kind of accreditation process. Some
favoured the retention of MySuper under alternative models, with the process strengthened
to include greater protections for members. For example, the Financial Services Council
proposed an enhanced MySuper approval process to strengthen the safety net for members
who are relatively disengaged. It envisioned that this would involve funds obtaining
authorisation on an ongoing basis from APRA, based on a ‘multifactor “member
outcomes” assessment’ that considers scale, fees and longer-term net investment returns
(FSC, sub. 38, pp. 21–22). The consumer group CHOICE (sub. 31, p. 3) argued that ‘any
consumer choice needs to occur with an appropriate safety-net in place’.
APRA (sub. 33, p. 2) also supported a stronger MySuper process, submitting that there is
‘merit in legislating both stronger authorisation requirements and a broader member
outcomes assessment in lieu of the current scale test’. APRA noted that there has been
considerable variation in net returns and fees across MySuper products, and that some
products have fallen well short of meeting their net return targets.
While many elements would be similar between the existing MySuper process and the
product accreditation in the Commission’s assisted employee choice model, there would be
some key differences (table 5.3). Specifically:
• accreditation would not be a legal requirement for funds to accept default contributions
— since there are no defaults in this model — but a voluntary process
• while funds would still need to demonstrate (and independently and regularly verify)
that the fund’s scale does not impede investment performance, funds would also lose
accreditation if their product persistently underperforms (over multiple years) relative
to pre-stated targets
• there would be a more formal process for monitoring the performance of accredited
funds: funds must demonstrate on an ongoing basis (for example, annually) that they
continue to meet the minimum standards; those that do not would lose accreditation.
The common thread linking all the accreditation standards is that assessment would be
relatively objective. While attributes such as investment performance and fund scale can
invite a degree of subjectivity, the standards for these would be based on ‘negative tests’
— that is, funds must demonstrate (and independently and regularly verify) that they have
not adversely affected members through poor investment performance or scale. Framed in
this way, the standards would not risk punishing otherwise sound funds affected by
short-term market movements. Negative tests also avoid imposing barriers to new entrants
seeking accreditation, as funds lacking a performance history cannot have breached the
test.
This system of product accreditation would offer several key benefits (section 5.3). It
would make it easier for relatively more engaged employees to compare products, since the
product design would be harmonised across funds (which would lead to lower search
costs). Accreditation would offer protection to fund members who are not experienced or
confident investors and simply want a ‘no frills’ product without having to pick specific

DRAFT REPORT
investment options themselves. Importantly, funds would have the ability to tailor
accredited products to different types of members (for example, in terms of investment
strategy), thereby giving members scope to seek out a product that meets their individual
needs and personal circumstances.
As accreditation would be voluntary for funds (with the exception of funds seeking
inclusion on the shortlist), a balance needs to be struck between setting high standards and
making it an attractive proposition for funds to seek accreditation in the first place.
Complying with the standards would have a cost to funds, though in many instances this
cost would be offset by the additional members and reputational benefits that would flow
from offering an accredited product.
Any accreditation process introduces the complication of how to deal with members of a
fund that loses its accreditation. Given that employees are responsible for choosing their
own product under this model, any process that leads to them being automatically moved
to another fund risks leaving some worse off and would be difficult to implement. Instead,
funds that lose their accreditation should be required to communicate this fact clearly and
immediately to members of the affected product (who may then choose to switch to an
accredited product with a different fund).
A last-resort fund
Some employees may fail to nominate a product under active choice (chapter 4), and this
risk may remain despite the other protections discussed above. A ‘last-resort’ fund would
act to protect these employees so their superannuation contributions are not forfeited. This
fund would be established to hold the contributions of these employees until they select a
product themselves (in most cases they would be new entrants to the workforce who do not
already have a superannuation account).
The last-resort fund would effectively be a default, but not in the usual sense of the term: it
would simply provide a temporary holding account. It would not compete with other funds
for members, and members would not be able to switch into the last-resort fund once they
have already chosen a superannuation product. Importantly, the fund would have a remit to
encourage its members to select a product from another provider in the market. It would
safeguard a member’s superannuation contributions until they have exercised choice, and
then transfer their balance to the product chosen.
PricewaterhouseCoopers proposed such a fund. It submitted that:
If there was just one default fund which was considered the most basic, lowest cost, no frills
and possibly a government-run fund, designed to catch the least number of default employees
as possible, then it may just encourage more people to actively select a more suitable default
fund. (PwC, sub. 12, p. 2)
Similarly, the Association of Superannuation Funds of Australia (sub. 24, p. 22) argued
that the Australian Taxation Office (ATO) may need to operate a facility of last resort

DRAFT REPORT
131
where an employee fails to nominate a superannuation fund under an active choice model.
Indeed, researchers have argued that providing an unattractive default option can be a way
to encourage people to make their own active choices, and may be useful when people
have highly heterogeneous preferences (Carroll et al. 2009, p. 1671).
The last-resort fund could be based on existing eligible rollover funds (which hold member
balances rolled over from lost or inactive accounts). These typically offer a very basic
product, with low fees and returns, low-risk investments and no insurance. The regulator
(APRA) would stipulate what services must be provided and the fees that can be charged,
and require that the fund help its members make an active choice.
An alternative would be for such a fund to be operated by the ATO, potentially in
conjunction with the Future Fund. The Commission seeks participant views on suitable
institutional arrangements for a last-resort fund.
Should marketing practices be further regulated?
The baseline — unassisted active choice by employees — would be characterised by
significant marketing and promotional activity by superannuation funds and other parties,
such as related financial institutions and financial advisers (chapter 4). Some of these
marketing activities would benefit employees by helping them to identify products that suit
their needs. However, this marketing effort would come at significant system-wide cost,
and much will reflect efforts by funds to retain members or attract new members from
other funds (a zero-sum game). Employees may find it difficult to assess the accuracy of
the information or whether advice is subject to conflicts of interest.
Marketing and promotional practices are already subject to regulation. Consumer and
financial services laws make misleading and deceptive conduct unlawful, and a suite of
financial regulations require funds to measure and disclosure their fees and costs in a
prescribed manner. Further, financial advice is subject to regulation, and workplace laws
prohibit employers and unions from coercing employees into joining a specific
superannuation fund.
Given the extent of marketing activity likely to prevail under any system of active
employee choice, government resources will need to be directed towards monitoring
behaviour in the marketplace and enforcing existing rules. These rules are generally
broadly defined, with regulators and courts having discretion in how they are applied.
However, it is not currently illegal per se for a fund to give benefits to a prospective
member as an incentive to choose the fund (ASIC, sub. 41, p. 4). Under any alternative
model that encourages active marketing, stronger regulator powers may be necessary to
protect member balances in the event that funds (or related parties) offer short-term
benefits to prospective members that are unrelated to the superannuation product itself
(such as consumer goods or discounts on banking or loan products).

DRAFT REPORT
That said, limits on such behaviour can be difficult to specify and enforce, and would come
at the risk of prohibiting behaviour that may be beneficial to members (for example, poorly
crafted limits could rule out legitimate payments to members in the decumulation phase,
such as payments reflecting an accrued tax benefit). At this stage, the Commission is
requesting participants’ views on whether additional restrictions on marketing practices
would be warranted and, if so, what form such restrictions should take.
5.3 Assessment against the baseline
The Commission’s model of assisted employee choice contains several components. While
these could be implemented separately, this assessment is based on the joint
implementation of all components.
Member benefits
Some benefits will vary across members
The model would improve member benefits by making it easier and simpler for employees
to choose a good product that meets their needs. The benefits would be significant — and
apply broadly, since by definition there are no ‘default members’ in the model — though
their form and extent would likely vary.
• Low-engagement members, who lack the motivation, knowledge or confidence to
research and select products themselves, would be significantly less likely to choose a
bad product (based on limited or no research) if they opt to select a product from the
shortlist.
– The shortlist would make the process of choosing a product simpler, while
significantly reducing the likelihood of members being unduly influenced by
conflicting advice, confusing information, brand loyalty or marketing. As such, it is
expected to have a significant positive impact on these members’ balances over the
long term.
– While the shortlist would likely reduce the proportion of employees that fail to
make any choice at all, employees who do end up in this situation would have the
protection of the last-resort fund, and so would not be forgoing their employer
contributions.
• Moderate-engagement members, who would do some comparison of available products
under the baseline, would benefit from significantly lower search costs.
– Those who use the shortlist would benefit from the objective and straightforward
information provided, and would likely experience a significant reduction in the
amount of time and effort they need to dedicate to the choice process.

DRAFT REPORT
133
– Those who undertake broader research would particularly benefit from product
accreditation, which would lead to harmonised products and help members to
identify a product that best meets their individual needs — which could increase
engagement and consequently reduce the need to pay for financial advice.
– Both the shortlist and accreditation process would reduce scope for confusion and
cognitive biases, and would make it less likely that members choose products with
overly conservative investment strategies or that persistently underperform.
• High-engagement members, who possess sufficient expertise and confidence to choose
products from the wider market, could use the shortlisted and accredited products as a
benchmark to assist with evaluating other products.
As noted earlier, the Commission has initiated a survey to further explore some of the
impacts of a shortlist and information provision on members’ decision making. The results
of this survey will be incorporated into the assessment of this model in the final report.
Other features will apply more broadly
Other benefits will extend across the spectrum of member engagement. The shortlist could
make the inflow of new members to shortlisted funds more stable, relative to the baseline,
as these funds would not be as dependent on marketing efforts to persuade prospective
members to join. This could give the funds slightly greater capacity to invest in ways that
maximise long-term returns, given they would likely have lower liquidity needs.
However, there are some risks to the benefits of the model being realised. In particular,
there could be low rates of participation in the accreditation or shortlisting processes by
funds. When a product loses accreditation, members in that product could be adversely
affected if they do not re-evaluate their situation. And, even if there are high rates of
participation in the accreditation process by funds, the existence of an underperformance
standard based on meeting pre-stated targets could lead to some funds setting relatively
conservative targets.
There is a narrow area of risk that the shortlist reduces the incentive for some employees to
seek out a product better suited to their individual needs. This could be a particular concern
for people with atypical characteristics (such as irregular work patterns or insurance needs)
that might be better off seeking professional advice or undertaking their own research.
However, these risks would also arise from default products (and thus are common to all
the Commission’s models and the current arrangements), and employees would remain
free to select a non-accredited or non-shortlisted product, seek financial advice or
self-manage their superannuation.
Importantly, the protections set out in this chapter to assist employees to choose a product
would be supplemented by the broader architecture of prudential and consumer laws that
apply even in the baseline (chapter 4). Nevertheless, none of these protections will
guarantee that all employees make good choices. Under any model where employees are

DRAFT REPORT
free to choose a superannuation product, some will inevitably make bad decisions
(chapter 1). But the number who do is expected to be materially lower under this model of
assisted employee choice compared with the baseline.
Competition
Competition would be more focused on member outcomes
A shortlist could substantially change the nature of competition in a world of active choice.
Instead of competing directly in the market for members’ contributions, some of funds’
efforts would be directed to competing for the market — that is, to be on the shortlist. One
direct effect of this may be to somewhat reduce expenditure on marketing by funds that are
shortlisted, although these funds would still likely advertise to promote their status on the
shortlist, and non-listed funds would likely devote increased effort to marketing.
Product accreditation would also facilitate rivalry between funds by enabling members
(and intermediaries) to more easily compare product features and performance across
funds. Improved comparability can level the playing field between funds and reduce
information asymmetries between funds and their members. This would make it easier for
members to switch funds and thus keep competitive pressure on providers. It can also put
downward pressure on fees — as, of course, can prohibitions on certain fees within
accredited products. For example, there was general agreement among participants in the
Commission’s stage 1 study that the MySuper reforms have increased competitive pressure
on funds (PC 2016).
While both shortlisting and accreditation would lead to funds changing existing products or
introducing new products that meet the required standards, both interventions would also
likely divert the focus of competition away from product proliferation and providing
irrelevant product features. Relative to the baseline, this would reduce the number of
members who are in highly customised products with multiple investment or insurance
choices (which many such members may neither need nor value). Further benefits could
arise from the effects of shortlisting and accreditation on fund consolidation, in large part
due to scale-related criteria in the assessment of products.
Importantly, the benefits of accreditation will hinge on its uptake by funds, which is
uncertain. Any accreditation system carries the risk of stifling innovation or imposing
higher compliance costs that do not lead to commensurate benefits — either to funds or to
members (who could ultimately be left worse off in the long term). This can discourage
funds from seeking accreditation in the first place, thereby defeating the intent of the
reform.

DRAFT REPORT
135
The market would remain highly contestable
The fact that only a very limited number of products would be included on the shortlist
could pose a barrier to new funds entering the market (or existing funds developing new
products). Equally, it could provide a guide to new funds for how to gain early acceptance
by members.
Shortlisting could also create barriers to exit if it means that funds must continue to accept
new members while they are listed. These barriers have the potential to reduce competitive
pressure and innovation, especially if the list is not reviewed regularly.
However, this too would not necessarily lead to bad outcomes. Indeed, the shortlisting
process handled sensibly would strengthen competition between funds to maximise
long-term member benefits and reduce costs. All funds would be eligible to participate in
the process, and so the market would remain contestable, provided that the shortlisting
process is re-conducted at regular intervals. This would give non-listed funds an incentive
to improve their product offering to become listed (including through mergers to increase
fund scale), while the constant pressure of losing a place on the list would give funds
already on the shortlist an incentive to keep constantly improving in areas relevant to the
filter, in anticipation of the next contest.
Finally, the competition impacts of a last-resort fund (or funds) are expected to be small,
regardless of whether it is government owned or not. The fund would be designed to
capture only the small proportion of employees who fail to make any choice (even with the
assistance offered by the other interventions), and it would have a requirement to guide its
members towards the products of other providers. Any differences in regulatory treatment
for this fund, or an implicit government guarantee, are not expected to unduly favour it
over other providers in the market.
Integrity
The integrity of the interventions set out in this chapter will reflect both the efficacy of
their design and the performance of the regulator. Integrity will also depend on how well
rules are enforced — and, especially, whether funds that fail to meet the required standards
lose their accreditation or place on the shortlist. Institutional aspects are important too, as
discussed in chapter 3. Specifically, government involvement in shortlisting can bring with
it a risk of choosing products that are too conservative, and the quality of the decision
making will ultimately depend on the quality of the individuals comprising the responsible
body.
As noted above, the shortlist may reduce the incentive for funds to undertake expensive or
wasteful marketing (including on features unrelated to long-term member interests). Risks
of collusion and other gaming are probably minimal with a voluntary listing and a process
involving tens if not hundreds of funds, along with a well-developed private-sector ratings

DRAFT REPORT
system as exists today in Australia. However, to the extent that they might apply, chapter 8
considers these risks along with ways to mitigate them further.
More generally, both shortlisting and product accreditation could reduce scope for
employees to make bad decisions on the basis of conflicting advice or interests. This would
be due to a combination of direct prohibitions on commissions and by providing a form of
benchmark to help employees to evaluate other products or advice. However, employees
would still be required to actively choose a product, and thus some risk of being persuaded
or coerced into a particular product would largely remain under this model.
Stability
The assisted employee choice model is expected to have minimal if any adverse impacts on
the stability of the superannuation system. Sudden large-scale movements in balances or
members are highly unlikely, and the shortlisting process would be unlikely to result in
excessive levels of market concentration. Funds that lose their position on the shortlist
would not automatically lose their existing members. Those that lose accreditation would
not face enforced loss of members either, though would be required to inform members in
their default product that it is no longer accredited, and thus may experience some outflows
of members to other funds.
System-wide costs
The assisted employee choice model will likely see material reductions in search costs for
some employees, mostly arising from spending less time and effort on choosing a product,
and less money on financial advice, relative to the baseline. All employees will experience
lower transaction costs with the ease of using the centralised online service to select an
existing account (compared with the need to provide documentation to the employer, as in
the baseline) or to select a new product from the shortlist. In addition, employees who fail
to exercise choice would not be forfeiting their superannuation contributions as they would
be placed in the last-resort fund.
This model may also result in a reduction in some marketing activities by funds, with a
corresponding reduction in marketing costs. However, significant marketing effort will still
likely occur across the superannuation system.
At the same time, the model would impose some (but not large) system-wide costs that are
not present in the baseline. Funds would incur some additional costs, primarily in the form
of administrative costs in seeking accreditation or shortlisting and compliance costs in
designing products that meet the requirements of these systems. There may also be costs
associated with collecting and reporting the supplementary information for the shortlist.
While some of these costs may initially be significant, they would fall over time once
structures have been established and funds have adjusted their internal processes.

DRAFT REPORT
137
Government would also incur costs in administering a shortlist and accreditation system.
These include the cost of establishing a body to compile the shortlist, as well as the
ongoing regulatory costs associated with monitoring the behaviour of funds. And there
would be the possibility of calls for government compensation in the event that one of the
shortlisted funds fails — a very unlikely event given the current level of high prudential
scrutiny. All of these regulatory costs are likely to be similar in magnitude to those that
would be incurred under the assisted employer choice model set out in chapter 8.
Further costs to governments would be incurred in establishing a last-resort fund (though
some of this cost would be recovered from its members) and in establishing and running a
centralised online service and/or centralised clearing house (as discussed in chapter 3).
However, over the long term, the model is expected to have significantly lower costs
associated with raising revenue to fund Age Pension liabilities, relative to the baseline.
The impacts on employers are likely to be small relative to the baseline, where employer
costs are already very low. The risk of employers being found in breach of their
superannuation obligations because an employee has failed to nominate a product will be
reduced due to the last-resort fund.
Overall assessment
The assisted employee choice model described in this chapter reflects what the
Commission considers to be a workable alternative to a system of defaults — it is a
collection of information improvements and regulatory interventions that could be
realistically implemented. It is also expected to lead to significantly better outcomes
overall than the baseline (table 5.4).
DRAFT FINDING 5.1
Relative to the baseline, the assisted employee choice model would:
• significantly reduce the complexity employees face in choosing a product and lead
to more employees choosing high-performing products that meet their needs,
thereby increasing member benefits
• focus competition on product aspects of value to members, put downward
pressure on fees (through greater product comparability) and likely curtail wasteful
product proliferation
• better align funds’ interests with those of members
• support a stable superannuation system
• have lower search costs for many employees and very low costs for employers,
but with government and funds incurring additional costs associated with
regulatory structures.

DRAFT REPORT
Table 5.4 Assisted employee choice model: summary
Member
benefits
The shortlist would make it simpler and easier for employees to compare products and
select a suitable product. It would also reduce the probability of employees relying on
irrelevant or conflicted sources of information.
Product accreditation would make it easier to compare products from a wider set of
providers and make decisions simpler.
The shortlist and accreditation would make it more likely that employees exercise
choice and the last-resort fund would reduce the penalty associated with not exercising
choice. But some risk remains that employees fail to exercise choice or choose a good
product.
Employees would not be forced to choose accredited or shortlisted products and would
remain free to choose other products.
Competition Improved product comparability from accreditation would put downward pressure on
fees. Accreditation would also encourage funds to provide a ‘no frills’ product for
members who are not well placed to select specific investment options themselves.
Accreditation may curtail wasteful product proliferation in the market as a whole.
Shortlisting could raise barriers to entry, though would not be inefficient given the
process would be contestable and repeated.
Accreditation and shortlisting can potentially reduce innovation, and the associated
compliance costs can discourage funds from participating.
Integrity The integrity of the interventions would reflect both the efficacy of their design and the
performance of the regulator and shortlisting body.
Both shortlisting and product accreditation could reduce scope for employees to make
bad decisions on the basis of conflicting advice or interests.
Stability Sudden large-scale movements in balances or members are unlikely, and the model is
not expected to result in excessive levels of market concentration.
System-wide
costs
Search costs are likely to be materially lower for many employees.
There will be additional compliance costs to funds, which would be offset by a reduction
in marketing expenditure (though marketing activity would likely remain significant).
There are relatively small monitoring and enforcement costs to government.
In terms of a shortlist of superannuation products, what evidence is available on the
size of the list that would best facilitate the majority of employees to choose a fund that
meets their needs?
What specific information should be included alongside such a shortlist to help
employees to choose between products? In what format should the information be
presented? What evidence is there for how the metrics would assist employees to
make decisions?
What institutional arrangements would best suit a last resort fund? Should it be
managed by existing eligible rollover funds or the Future Fund?
Under a system of active employee choice, what would be the costs and benefits of
prohibiting funds or related parties from offering prospective members a short-term
benefit that is unrelated to the superannuation product? What specific form should any
such prohibitions take?

FEE-BASED AUCTION
DRAFT REPORT
139
6 Fee-based auction
Key points
• A fee-based auction would require funds to compete for default fund status by out-bidding
each other on member fees.
• A reasonable presumption underpinning this model is that, over time, variation in member
fees explains a significant amount of variation in net returns. This suggests that an auction
— which would drive fees down — has the capacity to lift long-term net returns.
• Using an auction to select providers of default superannuation products is not without
precedent: Chile uses this approach to select its default providers.
• But there are some potential risks.
− Funds may bid an unsustainable fee structure, in the hope that the greater scale
economies from new members eventually decrease costs. Funds may also seek to
upsell a higher fee service at a later date or raise fees not committed to under the
auction.
− Non-fee aspects of performance such as member services and investment
management may suffer. Investment performance may suffer due to funds avoiding or
decreasing exposure to higher-cost illiquid investment opportunities, like infrastructure.
• There are many well-tested levers in the design of auctions available to address these risks.
These levers are complemented by existing superannuation regulations that also work to
mitigate these risks.
• To test the scope for a fee-based auction to meet members’ needs, the Commission
designed a prototype auction for assessment against its five model assessment criteria.
• The auction assessed performs well against the baseline. Relative to the baseline:
− member benefits would be higher, due to the likelihood of reduced fees
− competition would be enhanced in the area of fees (and fee transparency)
− integrity would be assisted due to the auction’s simplicity and accountability
mechanisms
− stability is unlikely to be compromised because the model would not lead to excessive
concentration of funds or volatile movement of members and assets
− system-wide costs are expected to be slightly lower, mainly due to lower search costs
for members.
The fees charged by Australian superannuation funds have been the subject of much policy
debate. The Financial System inquiry (FSI) (Murray et al. 2014) compared the fees
charged in Australia with other countries, and considered whether superannuation fees are
higher due to a lack of competitive pressures.

DRAFT REPORT
Some participants to the FSI argued that simple fee comparisons with other countries are
misleading. These participants highlighted that:
• the high returns achieved in Australia are partly due to superannuation funds’ willingness
to invest in higher cost asset classes, such as direct property, infrastructure, hedge funds
and private equity (ASFA 2014; AustralianSuper 2014; Chant West 2014a;
Deloitte 2014). Costs are also driven by the provision of legislated and expected member
services (FSC 2014). ASFA (2014) argued that in international terms, Australian defined
contribution members are paying fees consistent with members of similar funds overseas
• some of the costs incurred by superannuation funds are unique to the Australian system
and outside their direct control, such as the costs from adjusting to recent regulatory
reforms (Deloitte 2014)
• other countries have different approaches to fee disclosure, making it difficult to
directly compare fees in Australia with those charged elsewhere (Chant West 2014a).
While the FSI agreed that features of the Australian superannuation system contribute to
higher costs, it nonetheless concluded that fees are still too high given the scale of the
system, and that Australian funds could lower fees without compromising net returns.
Since the Australian superannuation system is several times larger than [defined contribution]
systems overseas, Australian funds could be expected to have lower fees after accounting for
differences in features. A major concern of the Inquiry, shared by the Super System Review, is
that the Australian system as a whole has been unable to realise the full benefits of scale.
(Murray et al. 2014, p. 102)
Researchers at the Grattan Institute also argued that superannuation fees in Australia are
excessive, and reported evidence that suggests long-term net returns are primarily driven
by fees (Minifie, Cameron and Savage 2014). On this basis the Grattan Institute argued
that policy makers should consider selecting default superannuation providers using a
fee-based auction.20 They cited Chile’s experience with an auction for selecting the default
superannuation provider, which has reduced fees for default members.
A number of criticisms are levelled at fee-based auctions, including that an exclusive focus
on fees would drive funds to adopt a passive, low-cost investment strategy, reducing
member returns (FSC, sub. 38, p. 29; ISA, sub. 40, p. 54; AustralianSuper, sub. 19, p. 3).
The debate over the merits of a fee-based auction is thus linked to the broader debate over
whether gross investment returns are largely market driven, and the relative merits of
active and passive investment management (PC 2016, chapter 6).
20 In a subsequent paper Minifie, Cameron and Savage (2015) said that for some asset classes, a sole focus
on fees may not give investment managers sufficient incentive to perform.

FEE-BASED AUCTION
DRAFT REPORT
141
Box 6.1 The relationship between fees and returns
There is an extensive literature that investigates the relationship between fees and net returns.
Overall, there is evidence that fees are a significant predictor of net returns, because on
average it is not possible to outperform the market (Johnson et al. 2015; Jones and
Wermers 2011). Researchers at Morningstar compared the net returns of active and passive
investment managers in the United States, and concluded:
… actively managed funds have generally underperformed their passive counterparts, especially over
longer time horizons, and experienced high mortality rates … failure tended to be positively correlated
with fees (that is, higher cost funds were more likely to underperform or be shuttered or merged away
and lower-cost funds were likelier to survive and enjoyed greater odds of success). Fees matter. They
are one of the only reliable predictors of success. (Johnson et al. 2015, p. 1)
This of course does not preclude some individual fund managers from outperforming the market
average by pursuing active management strategies (associated with higher fees), including by
investing in alternative asset classes (PC 2016, pp. 126–128).
In the case of the Australian superannuation system, there is evidence that fees are a primary
determinant of net returns. As noted, the Grattan Institute (2014) found evidence that, on average,
Australian funds that charge higher total (investment and administration) fees deliver lower net
returns once fees are accounted for. Their analysis suggests that a fund with fees 1 percentage
point below the average fee could be expected to generate net returns about 1.4 percentage
points above average.
The Grattan Institute’s findings are consistent with some other studies. Basu and Andrew
(2014) drew on data from 2004 to 2012 and found a (statistically) significant and negative
relationship between gross returns from Australian superannuation funds and expense ratios
(incorporating both investment and operating expenses, which are the primary determinants of
investment and administration fees). They found that a 1 per cent decrease in fees is
associated with a 0.1 per cent increase in gross returns (and by extension a larger increase in
net returns). Drew, Stanford and Veeraraghavan (2002) examined the investment performance
of Australian superannuation funds between 1991 and 1999, and found that higher investment
fees were associated with lower net investment returns.
Chant West (2014a) responded to the Grattan Institute’s report by arguing that the higher
investment fees charged by the ten largest MySuper products are justified by higher returns. This
conclusion is based on evidence that these ten products, which had allocations in unlisted assets
between 10 to 36 per cent, had higher average annual investment returns over a 15-year period
to 2014 compared to a passive benchmark portfolio. The Chant West measure of returns is not
adjusted for administration costs (and therefore does not represent net returns as defined in this
report).
Other studies have found that while higher fees lead to higher returns, these higher returns may
simply be compensating investors for higher risk. Ainsworth et al. (2016) found that Australian
superannuation funds with higher investment fees usually have higher allocations to riskier
asset classes, which are more expensive to manage. The study found evidence that funds that
charged higher investment fees produced higher net returns than the cheapest funds; however,
the most expensive funds did not realise significantly higher net returns once risks were taken
into account. Similarly, Cummings and Ellis (2015) found that some Australian superannuation
funds realised returns from illiquid investments (net of investment costs) that compensated for
the non-diversifiable risk (such as liquidity risk) that the investments contributed to the
investment portfolio.

DRAFT REPORT
There are two reasons why it is worthwhile exploring a fee-based auction as a competitive
allocation model. First, there is evidence that variation in member fees explains a
significant amount of variation in net returns (box 6.1). Second, realising economies of
scale would allow the superannuation system to achieve lower fees without diluting net
returns. While both points may be contested, they are not without merit.
As such, the Commission has explored whether a fee-based auction could be designed to
meet members’ needs (summarised in table 6.1). The chapter first considers the key design
features of an auction (section 6.1), then proposes an auction model for assessment
(section 6.2), and finally assesses that model against the baseline using the criteria outlined
in chapter 2 (section 6.3).
Table 6.1 Key features of the auction model
Pre-qualification Products must meet a set of minimum standards relating to investment
strategy and member services.
Funds must prove that they can manage expected inflows of default fund
members.
Bidding format First-price sealed-bid, and potential best-and-final-offer stage.
Assessment criteria Administration and investment fees.
Number of winners One to five.
Allocation of members Sequential allocation.
Transparency Information to be withheld only if its publication can be reasonably expected to
have commercially adverse consequences for the bidder.
Monitoring and
enforcement
Terms of appointment (and termination) specified in a contract.
Account for winners’ performance in future auctions.
Option to shorten default period in case of one winner.
6.1 Elements of auction design
A typical auction requires bidders to compete with each other on a single, price-based
metric. The seller elicits higher and higher bids from participants until all but one
participant has withdrawn and the good is allocated to that last remaining bidder at a price
equal to their most recent bid. In a well-designed auction, the final price approximates the
highest bidder valuation.
In a fee-based default superannuation auction, funds would bid for the right to receive the
flow of new default superannuation members for a designated period of time, based on the
fees they would charge members. The process aims to minimise costs for new default
members by allocating the right to the fund or funds with the lowest fees.

FEE-BASED AUCTION
DRAFT REPORT
143
There are fundamental challenges in designing an auction
Selecting default superannuation products using a fee-based auction would present several
risks. Some of these risks are not unique to auctions, arising in other allocation processes.
Unsustainable loss leading and raising other fees
Funds could use a short term ‘loss leading’ strategy, whereby the winning fund bids a fee
below near term cost. A fund could pursue this strategy in the hope that the greater scale
economies from new members eventually decreases its costs. If this strategy proves
unsustainable, the fund’s members are likely to incur higher costs over time as funds try to
correct for losses by reducing services and/or increasing fees.
Another possibility is that a fund pursues this strategy with the intent to upsell a higher fee
service at a later date (ASFA, sub. 24, p. 13). The scope for such short-term behaviour
should be considered in an assessment of the auction’s performance in delivering member
benefits.
There is also a more general risk that the winners of an auction could later raise fees not
committed to under the auction. Providers of superannuation products levy a number of
fees, including administration, investment, insurance, switching, exit, activity and advice
fees. Member fees can also be built into asset prices and buy-sell spreads.21
AustralianSuper commented:
… if a tendering system considers only administration fees, then investment management fees
would increase and defeat the intentions of the tendering process to keep costs down. (sub. 19,
p. 3)
Another risk is that fees increase following a merger between a winning fund and a
non-participant.
Non-fee aspects of performance suffer
The auction’s primary focus on fees could reduce non-fee aspects of performance. In an
effort to provide the lowest fee possible, funds may fail to compete on areas of
performance such as innovation, the suitability of the investment approach and the quality
of member services (ASFA, sub. 24, p. 3; CIFR, sub. 7, p. 5; First State Super, sub. 26,
p. 6; FPA, sub. 29, p. 10; FSC, subs. 38, 49, pp. 27, 1; IFAA et al., sub. 13, p. 5; ISA,
sub. 40, p. 54; Vision Super, sub. 4, p. 9).
21 A buy-sell spread is the difference in the buy and sell unit price of an asset.

DRAFT REPORT
In commenting on tender processes generally, The Association of Superannuation Funds of
Australia (ASFA) argued:
To reduce costs to a minimum, all non-core services for members such as [income stream
offerings, advice, and website functionality] may be reduced or dropped altogether. In addition,
it is also very difficult to build concepts like innovation into a tender process. This combined
with a focus on price would restrict innovation in the sector. It would also drive homogeneity in
default funds. (sub. 24, p. 13)
The Retail Employees Superannuation Trust (REST) warned that an excessive focus on
fees would see the investment approach suffer.
Any over emphasis on lower investment management costs favours a short term investment
philosophy requiring greater liquidity requirements and the adoption of a cheaper passive
indexing strategy with the prospect of lower returns. (sub. 23, p. 12)
Bids may not fall as far as costs might allow
An auction would enhance the pressures on funds to reduce fees. In the current system,
for-profit funds have twin incentives, having to consider the interests of both shareholders
and members when setting fees and meeting costs. And while non-profit funds do not have
to provide a profit to shareholders, this means that they may have relatively weaker
incentives to reduce costs (and with them fees). Running a competitive auction in such an
environment helps to discover how far funds, otherwise not fully under pressure to reduce
fees and costs, can do so.
Achieving competitive outcomes however relies on funds being able to make reasonably
confident estimates of the value of winning the auction. Funds need to estimate the likely
number of new default members in the designated default period, and the value of their
contributions into the future. This valuation also depends on how many choice members
will choose the winning product as a result of the quality signal that winning sends, and
how many members will leave the product. To estimate the net present value of future
contributions, funds will have to forecast future costs and investment returns, among other
things.
Information used to produce these estimates can be sourced from funds’ own knowledge
— large, well-established funds will have good internal estimates of the value of default
members; and those working presently across multiple industries will be particularly
well-placed. Bids by competing funds and any information provided by the seller will also
be informative. On the latter source, offering estimates to the less well-informed bidders
may assist the auction to operate at its most competitive, but comes with the risk that the
estimates prove misleading and bids prove unsustainable.
The key takeaway from these observations is that while the auction should set out terms
that maximise the opportunity for price discovery, it cannot be expected to produce a fee
level fully reflective of costs, both present and future. This is because it cannot match
bidders’ own knowledge nor ameliorate the effect of valuation uncertainty. Nevertheless,

FEE-BASED AUCTION
DRAFT REPORT
145
auction design should take into account that more information means more confident
bidding and a better result for members.
Bids may be subject to collusion
A risk in an auction is that funds collude to coordinate behaviour such that the fees offered
do not reflect a competitive outcome. It can, however, be difficult to identify whether
funds are colluding or not. Funds have more information about their costs than the
auctioneer (an information asymmetry), and variation between funds may in any event not
be large. Moreover, some strategic behaviour can appear collusive but in practice is not.
In general, the key risk factors for collusion are multiple rounds, repetition over time,
relatively few competitors and multiple winners. Having multiple rounds provides scope
for signalling, and repetition over time allows more sophisticated signalling across
auctions. Further, coordination is easier the fewer the number of competitors. Agreements
are also easier to reach when units can be allocated across bidders, as is the case when
there are multiple winners (Chan, Laplagne and Appels 2003).
While some of these risk factors are present in a fee-based auction — for example, the
auction would be repeated over time — the Commission considers that the overall risk of
collusion is low. Numerous participants are likely given the large number of active funds
and the dependence of many on new member inflows. Further, the market would not be
captive to any collusive agreement, because all members can still exercise choice at any
time. More generally, reputation risk, as well as current fiduciary and statutory obligations
(supported by various regulations), are likely to discourage such behaviour. The
Commission is also not aware of any international evidence of collusion in similar
processes, like the Chilean auction and New Zealand tender.
Designing an auction: what are the risk mitigant levers?
Auction design is critical to achieving the auction’s objectives, and for helping to address
the above identified risks. Policy makers have a number of levers when it comes to auction
design, many of which are well-tested in other markets or auction processes. The levers cut
across the three main stages of an auction — pre-auction, conduct during the auction, and
post-auction — and include the:
• pre-qualification stage
• bidding metric
• bidding format
• governance arrangements for the auction process
• number of ‘winners’ identified
• frequency with which auctions are held

DRAFT REPORT
• monitoring and enforcement mechanisms and rules that apply after the auction.
Notably, these levers would not operate in a vacuum: much of the existing policy and
regulatory framework for products and funds — which also work to further mitigate these
risks — would apply (chapter 3).
Pre-qualification stage
Prior to bidding, most auctions require participants to meet certain conditions in a
pre-qualification stage. Typically bidders must provide basic information about themselves
and their services by lodging an expression of interest to compete in the auction. In New
Zealand’s 2014 tender process, participants had to, among other things, undergo stress
testing of their systems and processes to ensure they could deal with expected member
volumes and be operational when allocation began (appendix B). By excluding funds that
clearly fail to meet minimum standards, pre-qualification can reduce the time spent
assessing bids and enhance the integrity of the process.
In the case of superannuation, demonstrating the capacity to absorb and manage a large
influx of new members annually would be an essential pre-qualification requirement.
Governance standards of an order higher than the regulator’s requirement would also seem
essential given the need to possibly deal with a higher level of public exposure. A level of
reserves that avoids dependency on finance-as-you-go management of administrative costs,
including the costs incurred from receiving an influx of new members, is also important.
Pre-qualification can also commit bidders to minimum levels of service and other aspects
of performance not directly accounted for in the bidding stage. As such, it could potentially
help to address the risk under an auction that non-price aspects of performance will receive
little-to-no attention.
A number of inquiry participants supported using MySuper, or a strengthened version of it,
as a pre-requisite for default products (AIST, sub. 28, p. 33; CIFR, sub. 7, p. 7; IFAA et al.,
sub. 13, p. 8; Kinetic Superannuation, sub. 45, p. 4; Mercer, sub. 15, p. 3; PwC, sub. 12,
p. 3; WSSA, sub. 11, p. 2).
Other proposals include limiting the set of default providers to those funds that:
• successfully retrieve unpaid Superannuation Guarantee contributions (AustralianSuper,
sub. 19, p. 3)
• provide unbiased advice and transparent performance indicators (Centre for Market
Design, sub. 18, p. 5)
• satisfy regulatory audits (Kinetic Superannuation, sub. 45, p. 4)
• are not in the lowest 10 per cent of performers on net returns (AIST, sub. 28, p. 44).
A pre-qualification stage has its own risks — for example, unknowingly excluding new
participants by designing for the known participants — and so must be designed carefully.

FEE-BASED AUCTION
DRAFT REPORT
147
The bidding metric
Another choice in the design of an auction is which fees should be the basis for bidding.
Inquiry participants did not provide views on which fees are most suitable for an auction.
In Chile, the right to provide default superannuation products is auctioned on the basis of
administration fees. In Chile, these fees cover administration, distribution, and internal
investment management costs (Chant West 2014b). They do not include investment fees
charged by external fund managers, which (along with the investments themselves) are
separately regulated and scrutinised in Chile. In the Chilean system, the fees that are the
subject of bidding form a majority of total fees on average across all pension providers
(FSC, sub. 49, p. 1).
The Commission considers that fees forming the bidding metric should account for a large
proportion of the total charges levied on members. This better ensures that the auction
materially benefits members. Fees should also have scope to fall from their current levels,
either because they well exceed efficient costs, or because there is scope for funds to
reduce costs over time.
AustralianSuper (sub. 19, p. 3) highlighted the risk that targeting administration fees leads
to the winners of the auction raising investment fees (a ‘waterbed’ effect). Another risk is
that in seeking to cut costs, funds reduce aspects of performance valued by members. For
example, setting investment fees in an auction could result in funds adopting a lower cost
investment strategy, which may come at the expense of diversification and long-term net
returns if the fund develops an investment portfolio with less exposure to potentially
higher-yielding asset classes that are higher cost and/or require more active management.
There is an inevitable trade-off between including a given fee in an auction or not:
excluding it could lead to a waterbed effect, while including it could reduce aspects of
product quality. Both risks may be mitigated through a well-designed pre-qualification
stage, and contract obligations to lock in agreements.
The bidding format
The bidding format is arguably the most fundamental design element of an auction. A vital
consideration for the choice of format is the nature of the asset being priced. In the case of
superannuation, unlike many other assets, the asset being sold is impermanent (a number
of years access to default members); and more importantly the seller has an ongoing
interest in the welfare of the members who make up the asset. Unlike the sale of
telecommunications spectrum or asset privatisations, the asset stewardship is in effect
shared. Leaving aside the degree to which this might be a legal matter, it is clearly going to
be a matter of interest to policy-makers and the winning fund alike that the offers are
delivered.
There are two high-level bidding formats: open and sealed. Each have their own variations
(box 6.2). Broadly, an open format is one where competition is more direct — bidders

DRAFT REPORT
observe each other’s bidding in real time and the result is revealed to the bidders and the
auctioneer simultaneously. A sealed format involves bidders submitting a single bid to the
auctioneer, who then has control over how information is disseminated. It may also
provide the auctioneer more flexibility to negotiate multiple winners via a
best-and-final-offer stage.
Box 6.2 Bidding auction formats
Open formats: English and Dutch
An English auction is the well-known ascending open-cry format, common to real estate sales.
In the regular case the price is raised until one bidder remains and the good is sold to the last
remaining bidder at their last stated price. A fee-based default superannuation auction that uses
the English format would require bidders to bid so long as the current bid is above their
‘threshold’ fee (the lowest fee that would ensure their expected return is non-negative).
The Dutch format is the reverse of an English auction. In the regular case the price falls until a
bidder accepts an announced price and pays that price. There is no dynamic competitive
bidding; the format relies on the bidder with the highest valuation submitting the first and only
bid. For a default superannuation fee-based auction, the Dutch format would involve the fee
starting at some reserve level (potentially zero) and being raised until a bidder accepts a fee
level. Some initial public offerings (IPO) have used a variant of the Dutch format to determine
the market clearing price for newly-issued company stock (for example, Google’s initial public
offering in 2004 took such a form).
Sealed formats: First-price sealed-bid and Vickrey
A first-price sealed-bid format requires bidders to submit a single bid to the seller. The highest
bid wins and pays that bid. This would look much the same in a default superannuation
fee-based auction (aside from the winner being the lowest submitted bid). This format is used in
the United States Forest Service timber auctions, where large portions of publicly owned
timberland are auctioned to private companies for logging.
A Vickrey is a second-price sealed-bid format in which the bidder that makes the highest bid
wins but pays the next highest bid. This would also look much the same in a default
superannuation fee-based auction, with the exception that the winner would charge the second
lowest bid. Examples of second-price auctions appear to be rare, however, examples cited in
the literature include stamp auctions and more recently online eBay auctions where bidders are
able to utilise proxy bidding.
Sources: Athey, Levin and Seria (2011), Ausubel and Milgrom (2005), Chan et al. (2003), Luckling-Reiley
(2000), Weinberg (2004).
Responsible body and transparency
As discussed in chapter 3, the responsible body should have an understanding of members’
needs and an obligation to act in their best interests. Practically, this means the auctioneer
should be independent and free of any interests in the participating funds. It also means
they should be accountable, and have no ongoing regulatory role, lest there be an incentive
to favour a fund that offers some form of regulator advantage.

FEE-BASED AUCTION
DRAFT REPORT
149
However, it is worth noting that this design element is of less importance in a typical
fee-based auction than other models (such as a multi-criteria tender). This is because the
allocation process is a function of a single, objective metric, reducing the scope for
interference in an efficient allocation by the auctioneer.
Where not at risk of misleading bidders, more information flows — within and across
auctions — are desirable given they decrease uncertainty and increase bidding confidence.
Therefore, the process should be as transparent as possible to foster participant confidence
and thereby greater participation. A high degree of transparency will also promote public
trust in the process. However, there may be information that would reasonably be expected
to have commercially adverse consequences for the bidder. It is important to balance the
confidentiality needs of funds against the preference for transparency, as reflected in the
Commonwealth Procurement Rules:
The need to maintain the confidentiality of information should always be balanced against the
public accountability and transparency requirements of the Australian Government.
(Department of Finance 2014, p. 21)
It is also important to note that the superannuation industry is already highly regulated with
respect to transparency, with regulator datasets and product disclosure statements going
into significant detail, particularly on fees and expenses. Given that Chile’s auction process
publishes unsuccessful bids, it would equally be reasonable for this feature to be included
in Australia should we adopt such a process.
Number of winners
The auction could have a single winner or multiple winners. Restricting the auction to
having a single winner would ensure that default members are in products of equal quality.
On the other hand, there are potential benefits to having multiple winners. More of the
existing stock of default members will benefit as each winning fund extends their winning
product to their existing default members (chapter 3). Having multiple winners also
increases the probability for any one fund to win, which would encourage participation.
Further, multiple winners spreads the distribution of contributions which reduces the risk
of dominance by a few big players in the longer run but reduces the participant’s
confidence in forecasting the inflows secured from being one of potentially many winners.
Having a known band of potential winners (one to five) allows a reasonable confidence
interval for that assessment.
Frequency
As discussed in chapter 3, there are risks associated with the process being too frequent or
too infrequent. Frequent auctions may increase the risk of a short-term approach to
investment, and would increase costs for participating funds and the Government (Mercer,
sub. 15, p. 6; ACTU, sub. 34, p. 10). Further, as explained above, running the process more
frequently could increase the risk of collusive agreements being maintained if only a few

DRAFT REPORT
funds are regularly participating (seemingly an unlikely prospect with over 100 MySuper
products). On the other hand, if the process is held many years apart, excessive industry
concentration could harm competition in the long term. The auction process also reveals
useful information, which is another argument for greater frequency.
Monitoring and enforcement mechanisms
Another consideration in the design of an auction is how the winners will be kept to their
winning fees and any other commitments they make in the auction process. Also important
is guarding against the risks that the winners raise other fees, or reduce effort in non-price
aspects of performance. The relationship between a winner of an auction and its default
members — and therefore the risk that funds fail to meet their commitments — can extend
well beyond the relevant default period.
There is a range of measures already in place that would regulate the conduct of the
winners of an auction. These measures include trustees’ fiduciary duties, statutory
obligations under the Superannuation Industry (Supervision) Act 1993 (Cwlth) (SIS Act),
and various provisions in competition and corporations law such as the dishonest conduct,
inducing persons to deal, and misleading or deceptive conduct provisions. More generally,
self-enforcing mechanisms such as the incentives created by reputational risk would
motivate winning funds to meet their commitments.
However, while these measures play an important role, they would be insufficient to
ensure winning funds meet all of their commitments. The auction process would need to
include specific measures for this purpose. Options include the following.
• Accounting for performance in subsequent auctions, for example, precluding a previous
winner from competing through the pre-qualification stage in a subsequent auction if it
reneged on its past commitments.
• Regulatory mechanisms, such as a licensing regime for default fund providers with
ongoing monitoring and reporting requirements, with the threat of removal of default
status and potentially loss of default members if there is non-compliance. The Chilean
Superintendence of Pensions receives reports from funds that outline fees and charges,
enabling it to scrutinise the fees of default funds. Alternatively, the commitments made
by default providers could be confirmed in a voluntary code of conduct or
memorandum of understanding.
• Contracts that specify fees and service levels, and outline consequences for failing to
meet these commitments. A contract could require winners to self-report against their
commitments on an ongoing basis.
Unlike the supply of other products and services, the provision of a superannuation product
is a long-term proposition and subject to various market forces that are outside the control
of trustees. These factors may need to be taken into account in the design of mechanisms to
ensure performance following an auction, without undermining system credibility.

FEE-BASED AUCTION
DRAFT REPORT
151
6.2 The auction model considered for assessment
The Commission has designed a prototype auction for assessment. The auction would
provide the winners with access to the flow of new default member contributions over the
designated default period (and beyond if default members stay with the winning fund). As
outlined in chapter 3, any competitive allocation mechanism, including an auction, should
initially be held every four years.
The auction would be administered by an independent panel, constituted by government.
The panel would advise the relevant Minister on which fund or funds should be selected
for default status, with this advice made public. The Minister would be free to accept or
reject the advice, although the reasons for any departure from the panel’s recommendations
would also be made public.
The model for assessment consists of three main stages.
1. Pre-qualification stage — participants lodge an expression of interest to compete in the
auction. This stage limits participation in the auction to those products and funds that
meet certain criteria. Participants must commit to a minimum level of service and other
aspects of performance not directly accounted for in the bidding stage.
2. Bidding stage — participants bid against one another in a first-price sealed-bid auction.
The panel would have the right to conclude this stage by asking participants to submit a
‘best-and-final-offer’. The winning offers would be formalised in a contract, and (if
there are multiple winners) default employees would be sequentially allocated22 across
winning funds.
3. Performance monitoring stage — the winning providers would self-report against the
commitments made in their offers (formalised in a contract that applies for the duration
of the default period). Providers that materially breach their commitments during the
default period would lose default status.
The pre-qualification stage is intended to filter out funds and products that are clearly
unsuitable to winning default status. To pass the pre-qualification stage funds and products
must meet minimum conditions.
Funds must prove that they can safely handle expected inflows of default fund members.
As such, the assessment panel would assess funds on factors such as stability, liquidity,
regulatory compliance, transparency and governance arrangements. The assessment panel
22 Under a model of sequential allocation, each default fund takes a turn in being allocated a new entrant
default member, with the process (including the order of allocation across funds) being repeated
indefinitely. This model results in an even allocation of default members across winning funds.

DRAFT REPORT
should also be satisfied that winners have sufficient capital to pay for the resources needed
to accept large inflows of new members.
Products should be balanced across capital growth and income assets, and would be
required to limit fees not included in the bidding stage to cost recovery. The
pre-qualification process should also include a performance filter: funds that underperform
for five or more years across their investment products relative to a benchmark on
long-term net returns would not qualify. To facilitate market entry, performance on past
returns need not be confined to superannuation products: performance history could come
from a similar investment product. Bidders without a performance history in the Australian
superannuation system could be assessed based on their relevant performance in offshore
markets.
To reduce the risk that an auction achieves low fees at the expense of other aspects of
performance, participants would also be required to meet minimum member service
standards in the following areas (chapter 3):
• administration and member engagement tools
• communication materials and intrafund advice
• quality of record keeping system
• transparency of fees and performance.
The pre-qualification process for funds and products would be similar to a strengthened
MySuper authorisation process (discussed further in chapters 5 and 8). APRA (sub. 33,
p. 2) has argued there would be merit in lifting the bar for MySuper by introducing
‘stronger authorisation requirements and a broader member outcomes assessment’.
The Commission considered whether, instead of imposing minimum standards, funds
could offer whatever service levels they want, with the panel then scoring each fund’s
service bundle and assigning a basis points penalty in the bidding stage to funds with
relatively low service scores. This approach would facilitate competition between funds on
aspects of member services. On the other hand, it could also lead to gold plating of
member services. Regardless, it would be difficult to ensure objective and consistent
evaluation of bids, short of the auction becoming a multi-criteria tender, and a rule would
need to be developed for converting the service level score into a basis points penalty in
the bidding stage. The Commission is seeking feedback from participants on whether a
fee-based auction can be designed to facilitate second-order competition between funds on
non-fee aspects of performance, such as member services.

FEE-BASED AUCTION
DRAFT REPORT
153
Bidding stage
The bidding metric
The bidding metric would incorporate both administration and investment fees.23
Together, these fees account for more than half of the total charges levied on MySuper
members (APRA 2017c), and some research suggests there is scope for them to fall from
current levels (Minifie, Cameron and Savage 2015; Murray et al. 2014). As such,
accounting for both fees would be expected to deliver material benefits to members, and
help to drive down costs and fees over time. Accounting for a large part of the total fee
base also reduces the risk that the winners shift costs to other fees and charges (the
waterbed effect).
The process for submitting bids should be designed carefully to ensure that the process
cannot be gamed and achieves outcomes that meet default members’ needs. The particular
fee structure that is bid (such as the mix of fixed and variable fees) can result in a different
overall fee burden for different account sizes. The panel should bear in mind that funds are
bidding for the right to provide a default product to new workforce entrants, and therefore
reward funds that set low fees for low account sizes, at least in the near term. However, to
provide funds with flexibility, the panel should not rule any particular fee structure in or
out.
While including both administration and investment fees reduces the risk of the waterbed
effect, it increases the likelihood that winning funds adopt a lower cost investment
strategy. For example, funds may avoid or reduce exposure to illiquid investment
opportunities such as infrastructure, which may otherwise have increased long-term net
returns. The Commission considers this risk in its assessment of the auction model in
section 6.3.
The bidding format and transparency
A first-price sealed-bid auction format is likely to be the best fit. There are three benefits to
using this format. First, it better allows for multiple winners. An open format produces a
result that appears more ‘final’. Whereas, under a sealed format, a best-and-final-offer
stage could be used to produce multiple winners. Second, it may better encourage
participation relative to other formats. This is because funds would face less risk in
revealing sensitive information in the auction process to other funds (as the size of
incremental bids in an open format can be informative in and of itself), particularly
information that may put them at a disadvantage in the choice market if they are
unsuccessful. This increased participation should transfer to more competitive bidding.
23 Administration and investment fees are defined in s. 29V(3) of the SIS Act. Administration fees relate to
the administration or operation of the fund and investment fees relate to the investment of assets. These
fees include costs incurred that are not otherwise charged as a buy-sell spread or a switching, exit,
activity, advice or insurance (non premium) fee.

DRAFT REPORT
And last, a sealed-bid format also minimises the scope for collusion (though as explained
above the risks of collusion are low in any case). The Commission is seeking feedback on
the particular bidding format used.
There is also the question of whether bids should be released to the public. As explained in
section 6.1, there are benefits for public trust and bidding competiveness from more
transparency, but this needs to be balanced against the commercial sensitivity for funds.
The principle of only withholding information if it can be shown to be reasonably expected
to have commercially adverse consequences for the bidder should apply. The Commission
is seeking further feedback on the optimal level of transparency.
Number of winners and best-and-final-offer
The auction is designed such that it could produce a single winner or multiple winners. As
mentioned above, the key benefit of multiple winners is that more of the incumbent stock
of default members will benefit, as each winning fund extends their winning product to
their existing default members. Multiple winners also spreads the distribution of
contributions, reducing the risk of dominance by a few big players in the longer run,
although this is considered remote in any case. While having one winner reduces the risk
of collusion, this risk is considered low (as detailed above).
The mechanism for acquiring multiple winners will be through a carefully managed
best-and-final-offer stage. In particular, the panel would have the discretion to provide the
opportunity for up to four of the most competitive funds to match the fee of the lowest
bidder. The final decision on how many funds are able to match the lowest bidder would
depend on the number of participants; if participation is low then fewer funds would be
given the opportunity.
If there are multiple winners, the government would sequentially allocate employees across
winning products. This approach, used in New Zealand, would impose a lower burden on
employers and employees than if they were required to choose.
Performance monitoring stage
The following mechanisms would increase incentives for the winning bidder to keep their
commitments made during the auction process.
• Contract with ongoing communication and disclosure regime. The commitments made
by the winning bidder would be confirmed in a contract, which would apply in the
default period. Winners would need to transparently report against their core standard
requirements on an ongoing basis. The contract would provide a reference for
considering whether a past winner has kept their commitments, and would outline
processes for managing the exit of underperforming funds. Ultimately, the relevant
Minister would be responsible for enforcing the contractual obligations of successful
funds.

FEE-BASED AUCTION
DRAFT REPORT
155
– In the event that a fund has its default status revoked mid-term, its default members
would be notified, and unless they opted to remain with the fund, would be
sequentially allocated across the remaining default providers. If there is only one
default provider, the default period would be truncated and a new auction would be
held.
• Accounting for past performance in future auctions. As part of the pre-qualification
stage, any past auction winners would be excluded from progressing further if they
materially fell short of previous commitments. For example, a winning fund may fail to
provide the quality of member services promised in a past pre-qualification stage, or
may seek to charge much higher fees than committed to under the previous auction. As
already noted, products provided by funds that consistently underperform on long-term
net returns would be excluded in the pre-qualification stage.
These mechanisms would be reinforced by existing measures, including trustees’ fiduciary
duties, statutory obligations under the SIS Act, and various provisions in competition and
corporations law (see above).
The risk of fee increases following a merger between a winning fund and a non-participant
would be collectively addressed by the contract and current regulatory measures. Default
members who stand to lose from a merger would be protected by the SIS Act, which
requires that bulk transfers of members to successor funds be made in the best interests of
members. As discussed in chapter 3, a formal framework that requires fund trustees to give
genuine consideration to merger proposals in the context of member interests would also
need to be considered.
In the event that a default provider fails to win a future auction, it would be free of its
contractual obligations. The provider would then have more flexibility to change its fees
and conditions, which may not best meet the needs of existing default members. Existing
regulatory measures would provide some protection for default members. To help keep
members informed, the results of subsequent auctions would be made publicly available
(chapter 3).
6.3 Assessment of the model
The Commission has assessed the proposed auction model against its five assessment
criteria: member benefits, competition, integrity, stability and system-wide costs
(chapter 2). In doing so, it has drawn on a range of evidence sources, including
submissions from inquiry participants, conceptual analysis, literature on mechanism and
auction design, and outcomes from other sectors and countries.

DRAFT REPORT
Member benefits
Member benefits are best met when funds have incentives to maximise long-term net
returns and allocate members to products that meet their needs (chapter 2).
Outcomes under the auction
The auction would create strong incentives for funds to lower their fees, potentially
boosting net returns. Importantly, the inclusion of both administration and investment fees
reduces the risk that competition occurs on one type of fee, only for others to be raised.
At the same time, competition on fees could increase incentives for funds to adopt a
low-cost passive approach to investment. Funds would have incentives to passively
manage within an asset class (such as tracking an equity index instead of actively choosing
listed equities), and to avoid or limit exposure to higher cost asset classes (such as illiquid
investment opportunities like infrastructure and unlisted property). These incentives would
be strengthened by the inclusion of investment fees in the bidding metric.
The impact of passive investment management within asset classes on long-term net
returns is unclear. As noted in the introduction, there are disparate views on this issue, and
it remains a source of ongoing debate. It is also important to consider the role that current
regulatory arrangements would play in conditioning the behaviour of funds. In
commenting on concerns that funds would resort to low-risk, low-return investments,
Kinetic Superannuation argued:
If this is correct it means that the fund’s directors and management would have knowingly
adopted a strategy in contravention of its statutory covenants of best interests and balanced
investment as well as the fund’s investment governance and risk management frameworks.
This critique does not consider the regulatory environment within which the bidding fund
operates. The fund’s reckless abandonment of quality and compliance controls and service
standards, in addition to its investment and best interests covenants, or even its adoption of
pricing manipulation or bidder collusion, amounts to a repudiation of a fund’s governance
framework and invites exposure of its directors and managers to legal liability. (sub. 45, p. 3)
On the other hand, lower exposure to illiquid investments could reduce diversification in
default product investment portfolios, potentially reducing long-term, risk-adjusted net
returns. Industry Super Australia highlighted a working paper by Cummings and Ellis
(2011) that suggests greater exposure to alternative illiquid assets has contributed to
increased risk-adjusted net returns for not-for-profit funds in Australia:
The outcomes delivered by not-for-profit funds is due to their ‘members first’ culture, and the
fact they have invested a higher proportion of their members’ savings in illiquid assets (such as
nation building infrastructure, property and private equity) which deliver superior long-term
returns, and lead to greater capital formation. (sub. 40, p. 11)
However, Cummings and Ellis (2011, p. 24) noted that their finding ‘may be specific to
this time period [2004–2010] and does not guarantee that funds with illiquid portfolios will

FEE-BASED AUCTION
DRAFT REPORT
157
always outperform funds with more liquid portfolios’. Others have also argued that there
are difficulties in drawing conclusions about the long-term performance of alternative asset
classes:
For many alternative asset classes, the quality and duration of the empirical data on assets and
managers is not sufficient to draw conclusions about the potential for profitable active
management. Despite recent dramatic trends in institutional investing, little is known
empirically about the performance of alternative assets over long business cycles. (Ang,
Goetzmann and Schaefer 2010, pp. 69–70)
Further, in a later version of their paper (published in an academic journal), Cummings and
Ellis (2015, p. 473) concluded that ‘taking account of risk, there is no evidence of
outperformance for funds with higher allocation to these [illiquid] investments’. However,
the authors acknowledge that as long as the returns from illiquid asset classes continue to
at least reflect the non-diversifiable risks they introduce to portfolios, then there will be
diversification benefits to the portfolio as a whole and a better risk-return trade-off.
As for other aspects of member benefits, the pre-qualification stage reduces the risk of low
fees coming at the expense of member services. Another important aspect of member
benefits is that employees are allocated to products that meet their needs. With a single
winner, there is a risk under the auction that some members will not be suited to the
product offered — one size may not fit all. While this risk can be mitigated with multiple
winners, the winning products would be expected to be very similar, and there is still no
certainty that any one individual will be allocated to a fund that fully meets their needs.
Assessment of the auction relative to the baseline
The auction is expected to perform better than the baseline on member benefits. The
baseline carries risks to member benefits, due mainly to poor employee decision making
arising from cognitive constraints, behavioural biases and inadequate information
(chapter 4). While the auction could reduce diversification across asset classes through
lower exposure to illiquid investments, relative to the baseline, member benefits would be
promoted by reduced fees and thereby higher net returns. In the end, putting young
disengaged members in a low-cost, passive growth investment may not be an unreasonable
approach.
Competition
The auction should foster competition that drives innovation, cost reductions and more
efficient long-term outcomes for members (chapter 2).

DRAFT REPORT
Competition for access to default members in the auction would potentially be strong.
Notably, a well-designed auction that effectively addresses the risk of collusion does not
necessarily require a large number of competitors to generate sufficient competitive
tension. Rather, in many circumstances it should be sufficient for the number of
competitors that are capable of winning the auction to exceed the number of winners. It
would be important however to consider the impact of the number of winners on incentives
to compete — too many would dilute the pool of new entrant default members, reducing
the intensity of competition. And the competitive tension from having nine winners out of
ten participants is likely to be much lower than where there is four winners out of five
participants. While ongoing participation would depend on the integrity of the process and
the costs of bidding, these issues can be addressed through careful auction design.
The focus of competition under the auction is on member fees. Including administration
and investment fees in the bidding metric opens up competition to a broader set of funds
than if fees covered only one of these measures. This also creates strong incentives for
funds to innovate to find lower cost ways of providing both administration and investment
services.
The auction process could also affect competition in the broader market for superannuation
products. One way this could occur is through the winning fund and its fee structure
providing a performance benchmark for other funds and choice members and their
advisers. However, the lack of historical success in encouraging member engagement
makes this at best an uncertain proposition, as does the experience in Chile, where funds
that have not competed for the default market have maintained much higher administration
fees (FSC, sub. 49, p. 1).
The auction could also affect competition in the broader market by altering market
structure or creating barriers to entry. Some participants were concerned that a competitive
allocation process would lead to excessive market concentration (APRA, sub. 33, p. 3;
FSC, sub. 38, p. 15). APRA said:
… significant industry consolidation may result in much higher levels of industry
concentration, potentially removing the incentive for the remaining, much smaller number of,
large incumbent funds to vigorously compete under the default allocation model. (sub. 33, p. 3)
The auction would likely help facilitate some market consolidation. There are currently
115 MySuper products, and some MySuper providers may seek to merge with each other if
they lose access to new entrant default members for four years (albeit they would still
retain access to the legacy job churn component of the current default member market).
To the extent there is some market consolidation, this is unlikely to substantially affect
competition. As highlighted above, some stakeholders have even suggested there is room
for industry consolidation (APRA, sub. 33, p. 3; Chant West 2015; Minifie, Cameron and
Savage 2014; Murray et al. 2014). Fewer products and competitors could even provide

FEE-BASED AUCTION
DRAFT REPORT
159
benefits, including improved productive efficiency and greater competitive pressure on
upstream providers of superannuation services.
The auction would provide a marginal improvement to competition over the baseline.
Competition for choice members under both the auction and the baseline would be highly
contestable, with strong rivalry between funds. The nature of competition however,
particularly with respect to default employees, would differ. Under the baseline there is
scope for superannuation funds to engage in unhealthy competition, including wasteful
marketing expenditure (chapter 4). While this same risk exists outside the auction process,
the nature of competition within the process is mainly focused on fees.
Integrity
The auction should assist integrity in the selection and delivery of default superannuation
products, and the behaviour of superannuation funds more broadly (chapter 2). Three
factors warrant particular attention: the potential for the allocation process to be
manipulated, including by collusion among auction participants; the integrity of the agent
assessing bids; and the strength of accountability and transparency mechanisms to ensure
funds are held to account for the outcomes they deliver.
There is a relatively low risk of manipulation of outcomes under the auction. The panel is
unlikely to be able to distort the allocation given the objectivity of the process, and the
potential for collusion is low given the factors discussed in section 6.1.
Ensuring winning funds meet their commitments during the default period is another
important issue. This is enforced primarily via a contractual obligation, and the potential
truncation of the designated period or exclusion from future processes. While breaches
may be obvious in some cases, disputes may be more difficult to resolve in cases where the
fund has followed the ‘letter’ but not the ‘spirit’ of the commitment. The resolution of
disputes regarding the outcome of the process, and public trust more generally, are well
supported by maximising transparency subject to commercial considerations.
Overall, the auction could perform better than the baseline against the integrity criterion.
There is an onus on the panel to get its advice to the Minister right in order to build public
trust over time. The risk that the assessment panel fails to objectively and robustly assess
bids is countered by the relative simplicity of the auction. The auction also reduces the
scope for exploitative behaviour that can occur in the presence of compulsion and

DRAFT REPORT
complexity under unassisted active choice. Last, the accountability and transparency
mechanisms of the auction are likely to be more effective in holding funds to account than
unassisted active choice.
Stability
The auction should not create instability in the superannuation system that leads to
significant systemic risks. To this end, the auction should not lead to: excessive
concentration in the superannuation product market or upstream supplier markets;
excessive and volatile movement of members and assets between funds; or funds focusing
only on short-term outcomes (chapter 2).
The auction is unlikely to require an excessive movement of funds, or lead to excessive
concentration in the superannuation product market. To the extent there is some market
consolidation, this would be unlikely to compromise stability. As highlighted above, some
stakeholders have suggested there is room for industry consolidation (APRA, sub. 33, p. 3;
Chant West 2015). Some consolidation could even improve stability if it leads to a more
efficient allocation of resources.
While the auction may result in some funds adopting a short-term focus (such as through
an unsustainable loss leading strategy), which would have implications for member
benefits, it is unlikely to lead to funds ceasing operations abruptly, and as such is unlikely
to have wider implications for system stability.
Neither the auction nor the baseline pose a material risk to stability. Both scenarios are
unlikely to lead to excessive market concentration, given the Commission’s definition of a
new default member. Nor would either scenario promote a short-term focus that
compromises stability.
System-wide costs
Minimising system-wide costs involves minimising the total costs across members,
employers, funds, and government (chapter 2). An assessment of this criterion requires an
examination of whether the auction is likely to increase or reduce the costs on each of these
four parties, the likely materiality of the change, and a judgment on the overall cost impact.

FEE-BASED AUCTION
DRAFT REPORT
161
Total costs to members under the auction are likely to be low. The auction involves a
fundamental shift in responsibility from individuals to government that means only those
members who actively opt out of defaults face search costs.
Total costs to employers under the auction are likely to be very low. Employers will be
relatively detached from the process with no obligations other than to direct contributions
to allocated funds.
Total costs could be high for some funds. While participation costs would be low, there
would be material marketing expenses. Funds that did not participate or are unsuccessful
will face high marketing costs in trying to capture choice members. Ultimately, these costs
will impact on members and/or shareholders.
Total costs to government should be low. Establishing the panel and running the auction
could cost somewhere in the order of $2.5 million per process. Monitoring would also add
to regulatory costs given that funds are required to commit to minimum levels of member
services in the pre-qualification stage (although the number of funds to be monitored is
low). Establishing a system for sequentially allocating members to default products would
also constitute an additional cost (in the event there are multiple winners). This system
could be an extended function of the centralised online service (chapter 3).
However, if appropriate products are selected and more people are in products that
maximise long-term net returns, the auction may reduce reliance on the Aged Pension, and
with it the deadweight losses from raising tax revenue to fund it.
Overall, system-wide costs are likely to be slightly lower under the auction than the
baseline. Under the baseline, search costs may be small for many members if they
outsource the decision or make a relatively passive one. However, the decision and action
required of all new entrant members under the baseline, as well as higher rates of switching
among existing members, suggests that in aggregate these costs would likely outweigh the
higher costs for government and funds under the auction.
6.4 Overall assessment
Table 6.2 summarises the costs and benefits of the auction model.
The main risks of the model are unsustainable bidding strategies and that the primary focus
on fees comes at the expense of other aspects of performance, including innovation or
higher cost investment strategies focused on long-term net returns. Other risks include bids
well exceeding costs and winners raising other fees not committed to under the auction.

DRAFT REPORT
Some of the features of the prototype auction model proposed are intended to address these
risks. Implementing such design features however presents its own challenges.
• Accounting for past performance in the pre-qualification stages of future auctions is
intended to increase incentives for winners to keep their commitments. Yet to be
effective this must present a credible threat to the winning bidder.
• Including investment fees is intended to reduce the risk that funds raise fees not
committed to under the auction. However, this may come at the expense of exposure to
potentially higher-yielding, but costlier, asset classes that can benefit diversification
and long-term net returns.
The Commission welcomes participants’ views on the effectiveness and workability of
these and other aspects of the auction model assessed.
Table 6.2 Auction model: summary
Member
benefits
To the extent that investment returns are largely market driven, focusing the nature of
competition on fees would promote long-term net returns, despite the potential for it to
push funds away from particular higher cost asset classes.
The pre-qualification stage (if well designed), would reduce the risk of low fees coming
at the expense of member services.
Competition Funds mainly compete on member fees.
The inclusion of both administration and investment fees opens up competition to a
broader set of funds than if bidding were only on the basis of administration or
investment fees.
competition. Fewer competitors could even provide benefits, including improved
productive efficiency and greater competitive pressure on upstream providers.
Integrity The risk of the outcome being manipulated by funds or the panel is low, as the bidding
metric is objective and collusion is unlikely.
Funds are held accountable by many mechanisms, including mechanisms specific to
the auction like a contractual obligation and potential exclusion from future processes,
as well as existing regulations and reputation risk.
Stability Unlikely to create instability, because the model would not lead to excessive market
concentration or volatile movement of members and assets.
System-wide
costs
High costs for government and funds are likely to be offset by the lower costs for
members and employers.

FEE-BASED AUCTION
DRAFT REPORT
163
DRAFT FINDING 6.1
Relative to the baseline, a fee-based auction model would:
• promote member benefits by exerting downward pressure on fees
• focus competition on, and elevate transparency of, member fees
• likely assist integrity due to its simplicity and accountability mechanisms
• be unlikely to compromise long-term stability because the model as designed
would not lead to excessive concentration of funds or volatile movement of
members and assets
members.
Is the format of a first-price sealed-bid with a best-and-final-offer stage and potentially
multiple winners the best fit for a fee-based auction? Are there any risks associated
with these design elements that have not been identified?
Regarding transparency, what would fall under the exemption of information that, if it
were disclosed, could reasonably be expected to have commercially adverse
consequences for the bidder?
Could a fee-based auction be designed to facilitate second-order competition between
funds on non-fee aspects of performance, such as member services?

MULTI-CRITERIA TENDER
DRAFT REPORT
165
7 Multi-criteria tender
Key points
• A multi-criteria tender would require funds to compete for default fund status by making
proposals against multiple assessment criteria.
• If well-designed, a multi-criteria tender can help engender a product that performs well
against characteristics that matter to members. Using a multi-criteria tender for this purpose
is not without precedent: New Zealand uses this approach to select its default providers.
• There are some risks, however.
− The criteria and their weighting may not reflect the breadth of all default members’
needs.
− The complexity of weighing multiple criteria may create a perception (or reality) that the
assessment panel failed to objectively and consistently evaluate tenders.
• Tender design is critical to achieving the tender’s objectives, and for helping address the
above risks. The Commission designed a prototype tender for assessment against its five
model assessment criteria.
• The multi-criteria tender assessed matches or betters the performance of the baseline on
most model assessment criteria. Relative to the baseline:
− member benefits would be better served by the inclusion and relatively high weighting
of long-term net returns in the assessment criteria
− competition would be focused on aspects of performance that matter to members, and
would be further enhanced by the winning bid providing a market-wide performance
benchmark
− integrity may be promoted due to the multi-criteria tender’s strong accountability
mechanisms, but on the other hand there would be risks to integrity due to the tender’s
vulnerability to subjective judgments
− stability would not be materially threatened over the longer term since the tender is
unlikely to lead to excessive concentration of funds or volatile movement of members
and assets
− system-wide costs are expected to be slightly lower, with higher costs for government
and funds likely to be outweighed by lower costs for members and employers.
A broad range of products and services, from defence equipment to public housing
management, is procured using multi-criteria tenders. New Zealand uses a multi-criteria
tender to select its default superannuation providers, and the Commission’s terms of
reference require it to consider the strengths and weaknesses of this model. One of the
main benefits of multi-criteria tenders is that they allow the procuring agency to target the

DRAFT REPORT
most valuable aspects of the product or service. A tender could therefore help to address
the risk of lower investment diversification under a fee-based auction (chapter 6).
Few participants commented directly on multi-criteria tenders. Of those that did, the Centre
for Market Design (sub. 18, p. 7), Institute of Public Affairs (sub. 17, p. 8) and Kinetic
Superannuation (sub. 45, pp. 2, 7) provided some support for the model. Kinetic
Superannuation argued:
Kinetic Superannuation favours a market-based model of a competitive tender process … A
market-based tender is the most effective solution because of its ability to revitalise
benchmarking against the winning product and its international acceptance as promoting
competition. Because of its adaptability throughout commercial relationships and public sector
procurement a tender can accommodate itself to the regulatory regime with minimal disruption
and without causing undue cost and complexity. (sub. 45, pp. 2, 7)
A number of other participants were opposed to using a multi-criteria tender, in part due to
the difficulty involved in assessing subjective criteria.
While a system-wide auction or tender model has the potential to bring more market forces to
bear on default allocation, ASFA considers that the risks and potential costs of this model far
outweigh these potential benefits. Some of the most significant potential negative outcomes …
[include] … difficulties in establishing objective and quantifiable criteria to evaluate the bids
for factors like levels of service. (ASFA sub. 24, p. 3)
Under a market based model, there would need to be several criteria, including qualitative
components. This raises the question as to how to assess the tender responses of different
Funds, and how to trade off against competing results on different measures. Creating an
appropriately rigorous and transparent framework would be difficult and potentially complex.
This may be particularly difficult in the case of qualitative factors. (IFAA et al., sub. 13,
pp. 7–8)
Expanding an auction or tender to consider other criteria introduces additional complexity and
inefficiency. In particular, the government has to select a set of parameters upon which to
optimise. There is no set of criteria that could in the real world be dynamically efficient without
perfect information. (FSC sub. 38, p. 15)
Some of the concerns about the complexity of tenders are legitimate, but they are not
insurmountable. Tenders are well-established mechanisms for injecting competition into
allocation models. As such, the Commission has explored whether a multi-criteria tender
could be designed that adequately meets the five assessment criteria presented in chapter 2.
The chapter first considers the key design elements of a tender (section 7.1), then proposes
a tender model for assessment (section 7.2) (table 7.1), and finally assesses the model
against the assessment criteria (sections 7.3 and 7.4).

DRAFT REPORT
167
Table 7.1 Summary of key features of tender model
Pre-qualification Products must pass a pre-qualification process with respect to services and
fees.
Funds must prove they can manage expected inflows of default funds.
Assessment criteria Past performance on net returns and independent surveys of member
satisfaction.
Investment strategy and governance.
The quality of member services, engagement and advice.
Fee levels and transparency.
Innovation in other areas.
Number of winners Five to ten.
Allocation of members Sequential allocation.
Transparency Information to be disclosed unless its publication can reasonably be expected
to have commercially adverse consequences for the bidder.
Monitoring and
enforcement
Terms of appointment specified in a contract.
Account for the winners’ performance in considering eligibility for future
tenders.
7.1 Elements of tender design
Participants in a multi-criteria tender compete by making proposals against a number of
different assessment criteria. Criteria are typically weighted by their relative importance.
While some tender criteria may be relatively easy to measure (such as fees), some require
judgment about the quality of each participant’s proposal. For example, a tender for the
right to provide a service may require participants to explain the quality of proposed
member services. In this case, the quality of each participant’s proposed member services
would be assigned a score. The weighted scores for each criterion are then added together
to produce an overall score for each participant.24 The overall scores provide a basis for
ranking bids and identifying the winners of the tender.
Multi-criteria tenders offer a number of potential benefits. A well-designed tender can help
engender a product that performs well on characteristics that meet members’ needs.
Non-prescriptive criteria give funds an incentive to innovate, and the flexibility to offer
products that specialise on certain characteristics. To the extent that the criteria and their
weightings reflect the needs of members, the tender process can thus drive funds to
discover the unique combination of product characteristics that best meet members’ needs.
24 Scores may first be normalised, usually by indexing against the ‘best’ score for that criteria. In the case of
a fee-based criterion, the best score would be the lowest fee.

DRAFT REPORT
There are fundamental challenges in designing a tender
Selecting default superannuation products using a multi-criteria tender would entail several
risks. Some of these risks are not unique to tenders, arising in other allocation processes.
Inconsistent assessments and failing to keep the winners to their commitments
There is a risk that assessments of tender proposals will be inaccurate or inconsistent. This
risk arises from the subjective nature of some tender criteria, such as the quality of the
investment approach or member services. While the quality of member services can be
measured, as illustrated by ratings services such as SuperRatings and Chant West, doing so
nonetheless requires some judgment, which can be disputed.
The subjective nature of the criteria also increase the risk of not being able to keep funds to
their commitments. Again, this risk arises from the subjective nature of some criteria:
while assessing whether winners have kept their commitments on an objective measure
such as fees is relatively straightforward, determining whether they have met their
commitments on their investment approach or member services is more difficult.
The criteria and their weighting may fail to reflect members’ needs
The organisation procuring a service through a tender often does so on behalf of a separate
end user. For example, a government procuring infrastructure construction services is
doing so on behalf of the broader public, which will pay for and use that infrastructure. As
such, a fundamental challenge is ensuring that the criteria and their weightings reflect end
user needs. In this context, the criteria specified in the tender process can be seen as an
attempt to substitute for the signals that would arise in an open and competitive market.
Developing criteria that reflect default members’ needs is difficult in a tender for default
superannuation products, given that most default members are disengaged from
superannuation and thus do not send market signals.
Compounding this difficulty is that members’ needs can change over time. However, to the
extent that changes in members’ needs can be identified, criteria can be updated and re-
weighted as tenders are repeated, as occurred in New Zealand (Mercer, sub. 15, p. 9). As
noted in chapter 3, periods between tenders should be shorter in the early stages of a new
model.
The process may be non-competitive due to bidder collusion
The potential for collusion arises as the seller has better information than the procurer. As
discussed in chapter 6, the key risk factors for collusion are having multiple rounds,
repetition over time, a low number of bidders, and multiple winners. Multiple rounds
provides scope for signalling, and repetition over time allows more sophisticated signalling
across tenders. Tenders with relatively few bidders make organising and maintaining a

DRAFT REPORT
169
collusive agreement easier. And in multiple-winner tenders, an agreement is easier to form
as units can be more easily shared among bidders.
While some of these risk factors would apply to a multi-criteria tender (repetition over time
and multiple winners) the Commission considers the overall risk of collusion is low.
Collusion would be made more difficult by three factors. First, participation is likely to be
relatively high. For a large group of funds, the pool of new default members would provide
a material source of contributions that would assist those funds in securing long-term
sustainability (chapter 3). Second, the tender can be designed to limit signalling before and
during the process; for example, by using sealed bids and prohibiting any communication
before the process. Prohibiting pre-tender communication would be easier to enforce than
in traditional procurement processes given bidding parties will be stand-alone funds rather
than consortia. Further, signalling over time is made more difficult for funds given criteria
are subject to change between processes. And third, given the market is not captive
(members can exercise choice at any time) the incentives to collude are weakened.
Designing a tender: what are the levers for mitigating risk?
Tender design is critical to achieving the tender’s objectives, and for helping to address the
above risks. Policy makers have a number of choices when it comes to tender design.
These choices cut across the three main stages of a tender — pre-tender, conduct during
the tender, and post-tender — and include the:
• pre-qualification stage
• bidding criteria
• governance arrangements for the tender process
• number of ‘winners’ identified
• frequency with which tenders are held
• monitoring and enforcement mechanisms and rules that apply after the tender.
Notably, these levers would not operate in a vacuum: much of the existing policy and
regulatory framework for products and funds — which also work to further mitigate these
risks — would apply.
Most multi-criteria tenders require participants to meet certain conditions in a
pre-qualification stage. Bidders typically provide basic information about themselves and
their services through an expression of interest to compete in the tender. In New Zealand’s
2014 tender process, participants had to, among other things, undergo stress testing of their
systems and processes to ensure they could deal with expected member volumes and be
operational when allocation began (appendix B).

DRAFT REPORT
Pre-qualification can serve a number of purposes. By excluding firms that clearly fail to
meet minimum required standards, pre-qualification can reduce the total number of tenders
to assess. This is important given that assessing a tender is time-intensive and thus costly.
A pre-qualification stage has its own risks, and so must be designed carefully. The criteria
should not unnecessarily exclude participants and in so doing reduce competition and the
diversity of bids. Another challenge is ensuring the criteria provide incentives for
participants to innovate to reduce costs and better meet members’ needs, rather than just
meet minimum requirements.
A number of inquiry participants supported using MySuper (or a strengthened version of
it) as a pre-requisite for default products (AIST, sub. 28, p. 33; CIFR, sub. 7, p. 7; IFAA et
al., sub. 13, p. 8; Kinetic Superannuation, sub. 45, p. 4; Mercer, sub. 15, p. 3; PwC,
sub. 12, p. 3; WSSA, sub. 11, p. 2). Other proposals included limiting the set of default
providers to funds that successfully retrieve unpaid Superannuation Guarantee
contributions (AustralianSuper, sub. 19, p. 3), providing unbiased advice and transparent
performance indicators (Centre for Market Design, sub. 18, p. 5), and satisfying regulatory
audits (Kinetic Superannuation, sub. 45, p. 4). AIST (sub. 28, p. 44) argued that products
should be ineligible if they consistently fall below a minimum net return standard, such as
the lowest 10 per cent of performers.
Criteria used in the comparative evaluation stage
An important design element of a tender is the criteria that participants bid on. The criteria
should reflect members’ needs and in doing so promote their best interests.
Inquiry participants identified a number of features and performance metrics of funds and
products, many of which could be used as criteria in a tender. The criteria identified fall
into two main groups.
• Fund and product features, including the investment strategy across asset classes
(accounting for risk), quality of member services, insurance offering, governance
quality, fee levels and the ability to tailor to members’ needs (AIST, sub. 28, pp. 43–
44; CIFR, sub. 7, p. 7; First State Super, sub. 26, p. 2; IFAA et al., sub. 13, p. 7;
Mercer, sub. 15, p. 5; REST, sub. 23, p. 7).
• Past performance measures, namely historical net returns and the results from member
satisfaction surveys (AIST, sub. 28, p. 43; FPA, sub. 29, p. 8; Kinetic Superannuation,
sub. 45, p. 4; Mercer, sub. 15, p. 5; REST, sub. 23, p. 7).
These criteria broadly reflect current industry practice for comparing superannuation
funds. Industry consultants, for example, also account for current fund features and past
performance measures when assessing and rating funds (Chant West 2017;
SuperRatings 2017). And corporate tenders are typically based on similar criteria (ISA,
sub. 38 and Mercer, sub. 31 to PC 2016). The Northern Territory Government (2014) noted
that its default fund for government employees has a proven strong investment

DRAFT REPORT
171
performance and low capped administration fees. The Commission’s stage 1 study, which
drew on considerable input from industry stakeholders, identified various system-level
indicators that align with the fund- and product-level criteria above.
The two main groups of criteria — fund and product features, and past performance
measures — each serve a slightly different purpose. Fund and product features help to
identify products that are most suitable for default members. Past performance measures
on the other hand create ongoing incentives for good performance, since performance will
be accounted for in the next tender (a ‘repeat game’).25 As such, past performance
measures — regardless of the extent to which they are indicative of future performance —
help to reduce the risk that funds will provide sub-optimal effort or renege on their
commitments after winning a tender.
Each criterion must be assigned a weighting that reflects its relative importance. In the
New Zealand tender, criteria related to organisational capability, member education and
investment capability (including a proven track record in funds management performance)
received a 70 per cent weighting, with pricing levels accounting for the remaining 30 per
cent. The Tasmanian Government’s (2014) general procurement guidelines suggest that
non-price criteria be weighted between 10 to 40 per cent, and price criteria between 60 to
90 per cent.
A key challenge in developing criteria and their weights is ensuring they reflect the needs
of default members. Notably, these needs may differ from the needs of choice members.
For example, while many active members may place a high value on a mobile application
that tracks investments in real time, default members may not. As such, the criteria would
need to be developed carefully.
Selection body and transparency
As discussed in chapter 3, the selection body should have an obligation to act in members’
best interests and an understanding of their needs. Practically, this means they should be
independent and free of any interests and involvement in the participating funds. It also
suggests they should have no ongoing regulatory role, lest there be an incentive to favour a
fund that offers some form of advantage to the regulator.
Given the degree of judgment that may be needed in assessing and ranking bids, some
expertise would be required. However, good incentives are likely far more important —
panel members need not all be technical experts in finance and investment. The ability to
comprehend relevant information (itself requiring a high level of financial literacy) should
be sufficient. Collectively, the panel should offer a wide range of skills and expertise. The
potential for subjectivity in assessment also reinforces the case for independence and
accountability in the form of transparency and review processes. A panel that makes or is
25 Past performance indicators also have a role to play in identifying suitable products. However, the role of
net returns may be limited, since past returns are not necessarily indicative of future returns.

DRAFT REPORT
in the process of making a poor choice risks creating a loss of confidence in the system.
While design can mitigate some of this risk, the ultimate sanction is to cancel the process
and replace the panel. This power must be available to the government.
Broadly, the more transparent the process the better. A high degree of transparency will
promote integrity (and ultimately trust) in the process. However, there may be information
that could reasonably be expected to have commercially adverse consequences for the
bidder. As discussed in chapter 6, it is important to balance the confidentiality needs of
funds against the need for transparency. The Commonwealth Procurement Rules state:
The need to maintain the confidentiality of information should always be balanced against the
public accountability and transparency requirements of the Australian Government.
(Department of Finance 2014, p. 21)
The Commission is seeking further feedback on the optimal level of transparency.
Number of winners
A tender could identify a single winner or multiple winners. A number of participants
argued there are (in general) risks from having a single or limited number of winners,
including reduced system stability and increased market concentration (ASFA, sub. 24,
pp. 13–14; FSC, sub. 38, p. 27; IFAA et al., sub. 13, p. 4), albeit these comments were
framed in the context of the Commission’s issues paper, which countenanced an allocation
system that applies to all default members, rather than just new entrant default members.
Some participants contended that a single winner (or a small number of winners) could
come at the expense of product variety. The Association of Superannuation Funds of
Australia argued:
A limited number of generic default super funds would exclude funds from specifically creating
niche products to meet particular default fund markets. There may be some industries which
have specific characteristics that can be met by the existing default fund as that fund was
created to service that particular type of employee or demographic. Such specific characteristics
could not be provided in a generic default product designed for people from many different and
heterogeneous industries. (sub. 24, p. 13)
If the tender has multiple winners, there would need to be a rule for determining how
winners are selected. Two potential approaches are:
• selecting a number of the best overall performers. The New Zealand tender process
uses this approach, selecting nine winners in its 2014 tender
• selecting the best performers on particular criteria categories. For example, the funds
that perform best on performance history, member services and investment strategy
respectively.
Selecting multiple winners would also require consideration of how employees are
allocated to default products. In the New Zealand tender employees are allocated to default

DRAFT REPORT
173
products sequentially.26 Having multiple winners and sequential allocation would mean
that default members are allocated to products that, though all may be high quality, will
each provide a slightly different bundle of services. This is an inevitable outcome of such a
model. Alternatively, employees could be allocated by employer (chapter 8) or employee
(chapter 5) choice. In these cases, it would also be important to consider how information
on the winning products is presented.
Frequency
As discussed in chapter 3, there are risks associated with the process being too frequent or
too infrequent. Holding the process too frequently risks encouraging short-termism and
would exacerbate the costs to participating funds and government (Mercer, sub. 15, p. 6;
ACTU, sub. 34, p. 10). And although the risk is low in any case, high frequency could also
increase the risk of collusive agreements being maintained (see above). On the other hand,
there may be an industry concentration concern if the process is held too infrequently
(CSA, sub. 35, p. 9). The auction process also reveals useful information, which is another
argument against less frequency.
Monitoring and enforcement mechanisms
There is a risk that the winners of a tender fail to meet their commitments made against the
tender criteria. For example, trustees could fail to meet promised standards of service
quality or investment management. The relationship between the winners of a tender and
its default members — and therefore the risk that funds fail to meet their commitments —
can extend well beyond the relevant default period.
There is a range of measures already in place that would help to regulate the conduct of the
winners of a tender. These measures include trustees’ fiduciary duties, statutory obligations
(including those in the Superannuation Industry (Supervision) Act 1993 (Cwlth) (SIS
Act)), and various provisions in competition and corporations law such as the dishonest
conduct, inducing persons to deal, and misleading or deceptive conduct provisions.
However, while these measures play an important role, they would be insufficient to
ensure winning funds meet all of their specific commitments.
Further options for addressing the risk that funds fail to meet their commitments include
the following.
• Accounting for past performance in subsequent tenders, such as by accounting for
historical long-term net returns and the results of member satisfaction surveys, would
motivate previous winners to meet their commitments. The pre-qualification stage in a
26 Under a model of sequential allocation, each default fund takes a turn in being allocated a new entrant
default member, with the process (including the order of allocation across funds) indefinitely repeated.
This model results in an even allocation of default members across winning funds.

DRAFT REPORT
subsequent tender could further assist by precluding a previous winner from competing
if it reneged on its past commitments.
• Regulatory mechanisms, such as a licensing regime for default fund providers with
ongoing monitoring and reporting requirements, with the threat of removal of default
status and potentially loss of default members if there is non-compliance. Alternatively,
the commitments made by the winning bidders could be confirmed in a voluntary code
of conduct or memorandum of understanding.
• Contracts that specify fees and service levels, with ongoing self-reporting obligations.
Winners of the New Zealand default tender must sign a contract with the government
(referred to as a ‘terms of appointment’). Corporate tenders are also usually confirmed
in a contract.
Unlike the supply of many other products and services, the provision of a superannuation
product is a long-term proposition and subject to various market forces outside the control
of superannuation funds and trustees. As such, it would be misleading (and difficult) to
assess the performance of the supplier over short periods. These factors need to be taken
into account in the design of mechanisms to ensure performance following a tender.
7.2 The tender model considered for assessment
The Commission has designed a prototype multi-criteria tender for assessment. The tender
would provide the winners with access to the flow of new entrant default members over the
designated default period (and beyond if default members stay with the winning fund)
(chapter 3). As outlined in chapter 3, any competitive allocation mechanism, including a
tender, should initially be held every four years.
The tender would be administered by an independent panel, constituted by the Australian
Government. The panel would advise the relevant Minister on which fund or funds should
be selected for default status, with this advice made public. The Minister would be free to
accept or reject the advice, although the reasons for any departure from the panel’s
recommendations would have to be made public to ensure transparency.
The model for assessment consists of three main stages.
1. Pre-qualification stage — a tender evaluation plan is released and participants lodge an
expression of interest to compete in the tender. This stage limits participation in the
tender to those products and funds that meet certain criteria.
2. Comparative evaluation stage — participants submit a detailed proposal that specifies
pricing, services and other characteristics of their default product. Service proposals are
evaluated and assigned an index score. The panel would have the right to conclude this
stage by asking some participants to submit a ‘best-and-final-offer’, which would help
to encourage convergence where offers are close. The winning offer would be

DRAFT REPORT
175
formalised in a contract, and default employees would be sequentially allocated across
winning funds.
3. Performance monitoring stage — the winning providers would self-report against the
commitments in their contracts. Winning providers that materially breach their
commitments during the default period would lose default status.
The pre-qualification stage is intended to filter out funds that are clearly unsuitable to
winning default status. To pass the pre-qualification stage products and funds must each
meet minimum conditions.
• Funds must prove they can safely handle expected inflows of default fund members. As
such, the assessment panel would assess funds on factors such as stability, liquidity,
regulatory compliance and governance arrangements.
• Product features would be limited, corresponding to the criteria assessed in the
comparative evaluation stage. Any fees excluded from the comparative evaluation stage
would be limited to cost recovery. Requirements relating to the investment approach
would not be specified (as occurs under the pre-qualification stage for the auction),
since this is considered in the comparative evaluation stage (see below).
– The assessment panel would have the option to use pre-determined thresholds (for
example, a minimum level of improvement in fees or a minimum reduction in
investment delay before funds are placed) to help sieve bids. To inform funds of
future expectations, these could be published.
The pre-qualification process for funds and products would be similar to a strengthened
MySuper authorisation process (discussed further in chapters 5 and 8). The Australian
Prudential Regulation Authority (sub. 33, p. 2) has argued there would be merit in lifting
the bar for MySuper by introducing ‘stronger authorisation requirements and a broader
member outcomes assessment’.
Comparative evaluation stage
Bidding criteria and transparency
The current legislated criteria in the Fair Work Act 2009 (Cwlth) (for determining whether
a product can be included as a default in modern awards) are too vague to be used in a
multi-criteria tender. Specific criteria would need to be developed for this purpose.
The default product should focus on the accumulation stage and consist of a bundle of
services that is the minimum necessary to meet the Government’s stated objective of the
superannuation system (chapter 3).

DRAFT REPORT
For the purposes of assessing a multi-criteria tender, and reflecting submissions from
inquiry participants, it is proposed that the comparative evaluation stage would require
participants to submit tenders against the following criteria.
• Past performance on net returns and independent surveys of member satisfaction. The
net returns measure — which should have a relatively high weighting — would be the
average of annual past net returns across the fund’s products over the past five years
(which may all be non-default products). To facilitate market entry, the assessment of
past returns would not be confined to superannuation products: performance history
could come from a similar investment product. Bidders without a performance history in
the Australian superannuation system would be assessed based on their performance in
offshore markets. The scores for each bidder would be indexed against one another (that
is, the best performing product would receive a score of, for example, 100, with other
products’ relative performance determining their score relative to this benchmark).
• Investment strategy. This would include an assessment of the planned asset allocation,
investment governance, quality of the investment team and management, and the
processes for selecting external fund managers. The assessment would centre on
whether the investment approach is consistent with investment performance targets and
maximising long-term net returns (including through effective diversification). These
latter measures do not require a performance record, and as such put potential entrants
on an equal footing with incumbents, helping to reduce entry barriers.
• The quality of member services, engagement and advice. This would include an
assessment of the proposed administration services for members, and tools that help
members to understand and plan for their retirement, and to resolve any issues or
questions that may arise. Participants would not be rewarded for proposing services
considered superfluous to default members’ needs (chapter 3).
• Fee levels and transparency. Participants would tender the administration and
investment fees they intend to charge. Participants would also explain how they intend
to make all of their fees transparent to members.
• Innovation in other areas. This criterion allows for the likelihood that there is scope for
superannuation providers to innovate and improve their services and products in areas
not explicitly included in the above criteria.
The quantifiable criteria — past net returns, results from member satisfaction surveys, and
fee levels — would be indexed against the best performing tender on each measure. For the
other criteria, the assessment panel (chapter 3) would issue each participant a score out of
ten, with this score indexed against the highest score across all participants for that criteria.
The tender evaluation plan would outline what is required to receive a score of 10, 9, 8 and
so on.
There is also the question of whether bids should be released to the public. As explained in
section 7.1, there are benefits for public trust and bidding competiveness from more
transparency, but this needs to be balanced against commercial sensitivity for funds.
Information should only be withheld if it can reasonably be expected to have commercially

DRAFT REPORT
177
adverse consequences for the bidder. An example of information that may be reasonably
withheld would be a novel product feature included in a bid. Such information has the
potential to create value for members, and mandating its disclosure could blunt the
incentive to produce it in the first place. However, at the very least, high-level information
such as the number of participants, and the number that met pre-qualification and
minimum improvement expectations should be published to give funds a feel for the
competitive landscape.
Number of winners
The panel would identify up to 10 of the overall best performers as the recommended
winners of the tender. An indicative range may be between 5 to 10 winners. The number of
recommended winning funds should be decided by the assessment panel on a case-by-case
basis. The panel should account for the following factors when deciding how many
winners to recommend.
• The number of participants. The number of winners should be fewer than the number
of participants. Otherwise, the tender process would be non-competitive. For example,
if the panel announces there will be ten products recommended for default status, but
market knowledge suggests only ten or so have a reasonable chance of winning, funds
could submit relatively non-competitive offerings and still expect to be selected.
• The distribution of scores. The number of winners should not arbitrarily exclude a
relatively well-performing bidder. For example, if it is announced there will be seven
winners, but there are eight stand-out bidders with very little separating the seventh and
eight best scores, the number of winners should be revised to eight.
Allocation of default members and the best-and-final-offer
Employees are allocated to products sequentially.27 This would occur by the government
sequentially allocating new default members across products. This approach, used in New
Zealand, would impose a lower burden on employers and employees than if they were
required to choose.
Sequential allocation would generally result in members being allocated to products with
slightly different bundles of services. Allowing service bundles to differ would help to
facilitate a competitive process whereby funds discover the set of product features that best
meet members’ needs. At the same time, there is a small risk that a fund could get on the
winners list despite an aspect of its service offering being clearly inferior to other winners.
A best-and-final-offer stage would provide the assessment panel with the opportunity to
reduce the extent of this disparity, and enable funds to rectify an element of its service in
27 Chapter 5 considers a scenario where employees instead choose their product. Chapter 8 considers a
scenario where employers choose (with an administrative filter).

DRAFT REPORT
what is an otherwise competitive offer. The best-and-final-offer stage would only be open
to a limited number of participants, as judged by the assessment panel on the basis of the
overall number of participants, and the distribution of tender scores.
Performance monitoring stage
The following mechanisms would increase incentives for winning bidders to meet their
commitments.
• Contract with ongoing communication and disclosure. Commitments made by winning
bidders would be confirmed in a contract with the responsible Minister. The contract
would apply in the default period and provide a reference point for assessing whether
past winners have kept their commitments. Winners would need to transparently
self-report against their commitments on an ongoing basis. Ultimately, the relevant
Minister would be responsible for enforcing the contractual obligations of successful
funds.
– The contract would outline processes for managing the exit of underperforming
funds within the 4-year default period. In the event that a fund has its default status
revoked, its default members would be notified, and unless they opted to remain
with the fund, would be sequentially allocated across the remaining default
providers.
• Accounting for past performance in future tenders. Past performance would be
accounted for in two ways.
– As part of the pre-qualification stage, any past tender winners would be excluded
from progressing further if they fell materially short of previous commitments. A
breach of previous commitments would arise, for example, if a fund did not provide
the level of member services promised in a past comparative evaluation stage.
– The criteria would explicitly account for past long-term net returns and the results
of member satisfaction surveys. If these are given sufficient weighting in the
assessment of tenders, they would motivate funds to continue to act in members’
best interests, in order to win future tenders.
The risk of a fund reducing the quality of its default product following a merger with a
non-participant would be addressed by the contract and current regulatory measures.
Default members who stood to lose from a merger would be protected by the SIS Act28.
As discussed in chapter 3, a formal framework that requires fund trustees to give genuine
consideration to merger proposals in the context of member interests would also need to be
considered.
28 The SIS Act imposes a general obligation on trustees to act in members’ best interests, as well as a
specific ‘equivalent rights’ test for successor fund transfers (chapter 3).

DRAFT REPORT
179
In the event that a default provider fails to win a future tender, it would be free of any
contractual obligations. The provider would then have more flexibility to change its fees
and conditions, which may not best meet the needs of existing default members. As noted
by Mercer (sub. 15, p. 10), this scenario must be carefully considered in any tender system.
Existing regulatory measures would provide some protection for default members. Funds
should not, however, be forced to notify members themselves if they fail to win default
status in a future tender, lest members be given an indication that performance is poor
when it is in fact just outside the top 4 or 5. It is quite plausible that a default product may
not win a subsequent tender but still be a good option. It is in practice the difference
between being in the top 4 or 5 at a point in time or being in the top 10. Instead, the
Government can help to keep the system participants informed by transparently publishing
and announcing the results of future tenders.29
7.3 Assessment of the model
The Commission has assessed the prototype tender against its five assessment criteria:
member benefits, competition, integrity, stability and system-wide costs (chapter 2). In
doing so, it has drawn on a range of evidence sources, including submissions from inquiry
participants, conceptual analysis, the literature on mechanism design, and outcomes from
other sectors and countries.
Member benefits
Member benefits are met where funds have incentives to maximise long-term net returns
and allocate members to products that meet their needs (chapter 2).
Outcomes under the tender
The selection criteria in the multi-criteria tender help to directly promote member benefits.
The criteria include the results of member satisfaction surveys and long-term net returns,
with the latter complemented by an assessment of the current investment approach and
investment return targets. The results of member satisfaction surveys and long-term net
returns would be weighted higher than the other criteria. As such, a fund that performs
relatively poorly on these measures would be at a significant disadvantage in the tender.
The benefits extended to members would be magnified with multiple winners, since more
incumbent default members would receive the conditions secured in the auction (although
scale economies and incentives to compete would be lower with more winners).
29 However, as noted above, in the event that a product has its default status revoked within the four year
default period for failure to meet contracted standards, its default members should be notified.

DRAFT REPORT
Because the tender process would initially be repeated every four years, the inclusion of
past performance measures in the selection criteria provides ongoing incentives for good
performance. These incentives would apply to any fund that wants to win a future tender,
regardless of whether they were successful or unsuccessful in a past tender. Benchmarking
participants’ long-term net returns against each other would also mean that the more a
participant out-performs its competitors, the higher its score in a tender, reducing the
likelihood that funds cluster around middling levels of performance, to the long-term
detriment of their members.
The benefits from selection criteria that target member benefits are likely to extend into the
choice segment of the superannuation product market. This is because winning bids, which
would be made public, would provide a market-wide performance benchmark for members
that choose their superannuation provider. Kinetic Superannuation argued:
… market models are generally more efficient than administrative allocation models and have
the advantage of revealing the relative performance of products. Not only does a tender
promote visible competitive bidding but its regularity of recurrence can facilitate cost reduction
and innovation as funds benchmark themselves against successful bidders. (sub. 45, p. 2)
The Centre for Market Design similarly argued that well-managed default funds are
valuable insofar as they:
… create clear reference points, against which comparison may be made using credible
independent metrics. (sub. 18, p. 4)
Further effects on competition are considered in the assessment of competition below.
Assessment of the tender relative to the baseline
The tender is expected to better promote member benefits than the baseline. While
employees are more likely to be engaged under the baseline, they might be constrained by
information problems and the complexity involved in choosing a fund. This may in turn
reduce incentives for funds to maximise long-term net returns (chapter 4). In contrast,
under the multi-criteria tender the focus on long-term net returns would increase incentives
for funds to compete on this measure, including through the winning bids providing a
market wide performance benchmark.
Competition
The tender should foster competition that drives innovation, cost reductions and more
efficient long-term outcomes for members (chapter 2).

DRAFT REPORT
181
Competition in the tender is expected to be sufficiently strong to engender good outcomes.
While not all funds will choose to participate — indeed the very largest may have good
reasons not to — the bulk of mid-sized funds currently dependent on default inflows or
with deteriorating net outflow ratios will find strong reason to do so.
Notably, a well-designed tender does not necessarily require a large number of competitors
to generate sufficient competitive tension. Rather, in many circumstances it should be
sufficient for the number of competitors that are capable of winning the tender to exceed
the number of winners. It would be important however to consider the impact of the
number of winners on incentives to compete — too many would dilute the pool of new
entrant default members, reducing the intensity of competition. And the competitive
tension from having nine winners out of ten participants is likely to be much lower than
where there are four winners out of five participants. While ongoing participation would
depend on the integrity of the process and the costs of bidding, these issues can be
addressed through careful design of the tender.
In terms of the nature of competition, to the extent that the criteria and their weightings
reflect members’ needs, the tender goes some way to achieving the competitive tension
that would exist in a market with active and informed users. Moreover, multiple criteria
provide funds with flexibility to discover the best combination of fees, investment
approach and member services. The use of non-prescriptive criteria with no pre-defined
solutions can provide incentives for innovation. Where a beneficial development in
member services is observable in other markets, a minimum improvement criterion could
also encourage change. And the criteria should be varied over iterations of the tender, in
conjunction with expert advice.
At the same time, the inclusion of past performance measures — and in particular
requiring a performance record of at least five years — would create a barrier to entry. The
height of this barrier would be limited however by allowing funds that have not provided
default superannuation products in Australia to use their performance record in similar
investment products locally or offshore. More generally, entry could still occur in the
choice segment of the market. Once funds have survived for five years in an investment
market (which in itself provides a sort of ‘competency filter’), they could seek to win
access to default members though the tender. More generally, by excluding new suppliers
with no performance history anywhere, the inclusion of past performance measures
improves the integrity of the process (see below).
As highlighted in the above section on member benefits, the tender process would also
affect competition in the broader market for superannuation products. Winning bids would
attract attention from other funds and choice members and their advisers, providing a
market-wide performance benchmark. Other funds may seek to improve their products to
prevent their members from switching to the winning fund. Moreover, those funds that
miss out on default status should have greater incentives to improve their products in order

DRAFT REPORT
to attract new members through active choice. While Chile’s experience with a fee-based
auction suggests that these incentives might be limited in their effect, the likelihood of
consolidation in Australia (see below) is a more obvious driver of product improvement
amongst unsuccessful bidders than it may be in concentrated overseas markets.
The tender could also affect competition in the broader market by altering market structure.
Some participants were concerned that a competitive allocation process would create
excessive market concentration (APRA, sub. 33, p. 3; FSC, sub. 38, p. 15). APRA said:
… significant industry consolidation may result in much higher levels of industry
concentration, potentially removing the incentive for the remaining, much smaller number of,
large incumbent funds to vigorously compete under the default allocation model. (sub. 33, p. 3)
A tender would likely help to facilitate some market consolidation. There are currently
115 MySuper products (APRA 2017c) and some MySuper providers may seek to merge
with another fund if they lose access to default members for four years.
competition given the very deep market in Australia. Indeed, as highlighted above some
stakeholders have suggested there is room for industry consolidation (for example, APRA,
sub. 33, p. 3; Chant West 2015). Fewer products and competitors could provide benefits,
including improved productive efficiency through scale economies and greater competitive
pressure on upstream providers of superannuation services.
The tender is expected to improve competition over the baseline. Under the baseline, the
market would be highly contestable and there would be strong rivalry between funds. As
noted in chapter 4 however, there is still scope for funds to compete on attributes that are
not in members’ best interests. In contrast, under the multi-criteria tender the selection
criteria and likelihood of consolidation would focus the nature of competition on
maximising long-term net returns and member satisfaction.
Integrity
The multi-criteria tender could promote a high-degree of integrity in the selection and
delivery of default superannuation products, and the behaviour of superannuation funds
more broadly (chapter 2). Two factors warrant particular attention in the context of a
multi-criteria tender: the potential for the allocation process to be manipulated, including
by a misalignment of incentives between the selector and member, or of collusion among
tender participants; and the strength of accountability and transparency mechanisms in
place to ensure funds are held to account for the outcomes they deliver.

DRAFT REPORT
183
Given the degree of subjectivity in bid assessment, there is a risk that the selector can be
captured by the industry (or particular segments of the industry) or is too risk averse and is
captured by political concerns (ISA, sub. 40, p. 32). There may be difficulties in
maintaining a truly independent panel, although accountability and transparency measures
help counteract this.
Collusion poses a relatively small risk to the integrity of the multi-criteria tender for three
reasons. First, participation is likely to be relatively high given the existing depth of the
market and the reliance on default flows for many funds. Second, funds would submit
sealed bids. This would effectively eliminate scope for signalling during the process.
Further to this, dynamic criteria make signalling over time more difficult, and
communication between funds before the process would be explicitly disallowed and
easier to enforce than in traditional procurement processes given bidding parties will be
stand-alone funds rather than consortia. Third, given the market is not captive (members
can exercise choice at any time) the incentives to collude are weakened.
Accountability of funds is promoted via consideration of past performance in future
tenders (including potential exclusion), and a contract with government including ongoing
communication and disclosure. These enforcement mechanisms are on top of expected
compliance of fund trustees with their fiduciary and statutory obligations along with other
existing laws and regulations (chapter 4). While breaches may be obvious in some cases,
disputes may be more difficult to resolve in cases where the fund has followed the ‘letter’
but not the ‘spirit’ of the commitment. Ongoing disclosure would act as an incentive to
avoid reputational damage. Nonetheless, a residual risk remains that a fund alters the terms
of the product after the designated period, negatively affecting members.
Integrity is also enhanced by the requirement for funds to have at least a five year history
of providing a superannuation or like product, which removes the possibility of new
suppliers with no performance history gaining access to default members.
All considered, the Commission cannot make an unequivocal judgment on how the
multi-criteria tender performs relative to the baseline on integrity grounds. While the
baseline may see many members leaving important decisions to potentially conflicted
individuals, there is no absolute certainty that the assessment panel will itself be
completely free of conflicts and able to make difficult choices that are in the best interests
of members. However, the accountability and transparency mechanisms of the tender are
likely to be more effective in holding funds to account than unassisted active choice.

DRAFT REPORT
Stability
The tender process should not create a degree of instability in the superannuation system
that would lead to any significant systemic risks (chapter 2). To this end, the tender should
not lead to excessive concentration in the superannuation product market or upstream
supplier markets, volatile movement of members and assets, or widespread adoption of
unsustainable loss leading strategies that cause disruption in later periods.
The tender is unlikely to lead to excessive concentration in the superannuation product
market. There are currently 115 MySuper products (APRA 2017c) and some MySuper
providers may seek to merge with each other if they lose access to default members for
four years. Many unsuccessful funds would in such circumstances nevertheless increase in
size and should be able to also attract members through the choice segment of the market.
Indeed, those funds would likely increase their efforts to attract members in the choice
market.
Given the large number of funds in Australia, even a significant number of mergers is
unlikely to threaten system stability. Well-informed stakeholders have suggested there is
room for consolidation (for example, APRA, sub. 33, p. 3; Chant West 2015).
Consolidation could even improve stability if it leads to a more efficient allocation of
resources.
There will be some increase in movement of members and assets under the tender. With up
to 10 winners, there is scope for the set of winners to differ from one tender to the next.
However, this would not constitute volatile movement and would not put at risk the
long-term stability of the system.
Nor do the incentives under a multi-criteria tender favour widespread short-term
behaviour, such as loss leading practices or performance clustering. The inclusion and
relatively high weighting given to long-term net returns provides incentives for funds to
take a long-term approach to investment and to outperform their competitors, which
provides greater scope for diversity in investment approaches. Also, some funds may
choose to remain outside the tender process, sacrificing opportunity to be a default fund for
benefits already captured by their current membership and success in the choice market.
These funds will have a different strategy to those competing for default status, and so
provide diversity in performance. With choice clearly available to all employees at any
time, the likelihood of performance clustering is lowered.
Neither the tender nor the baseline pose a material risk to long-term stability. Both
scenarios are unlikely to lead to excessive concentration or volatile movement of member

DRAFT REPORT
185
funds. Under the baseline individual decisions by employees are likely to lead to small and
predictable member flows. While the tender fuels greater movement of members and assets
in the short term, this can be readily absorbed by the system and thus falls well short of the
levels of volatility that could compromise system stability.
System-wide costs
Minimising system-wide costs involves minimising the total costs across members,
employers, funds, and government (chapter 2). An assessment of this criterion requires an
examination of whether the tender is likely to increase or reduce the costs on each of these
four parties, and a judgment on the overall cost impact.
Total costs to members and employers under the multi-criteria tender are likely to be very
low. Default members incur no costs associated with making a decision given the role of
allocation is assumed by government. The relative performance revealed by the tender
would also reduce search costs for members who do wish to make a choice. And
employers will be relatively detached from the process with no obligations other than to
direct contributions to funds.
While participating in the tender would impose costs on funds, such as the costs of
preparing proposals, overall these costs are likely to be low. The more material costs to
funds will be in marketing. Funds that did not participate or are not successful will face
higher marketing costs in trying to capture active choice members. Ultimately, these costs
will impact on members and (in the case of for-profit funds) shareholders.
Total costs to government of the multi-criteria tender will be low compared to the
contributions directed and the costs of poor decision-making in the future (that is, age
pension costs), but nevertheless may aggregate to $2.5 million or more over each four year
cycle. The more criteria that are included the more costly it will be to run the process and
monitor the winners. Establishing a system for sequentially allocating members to default
products — which could be an extended function of the central information service
(chapter 3) — would be an additional government cost.
At the same time, if appropriate products are selected and more people are in products that
maximise long-term net returns, the tender may reduce reliance on the Aged Pension, and
with it the deadweight losses from raising tax revenue to fund it.
Overall, system-wide costs are likely to be slightly lower under the multi-criteria tender.
This is because new members search (including advice) costs are replaced by a central

DRAFT REPORT
process, and marketing costs to funds will be lower than under the baseline given the
existence of a default option. The large number of members and thus the large aggregate
search costs under the baseline means that the higher costs to government and funds are
outweighed and the tender produces lower system-wide costs.
7.4 Overall assessment
Table 7.2 summarises the costs and benefits of the multi-criteria tender model.
The multi-criteria tender outperforms the baseline on the Commission’s assessment
criteria, but with additional risks that would require mitigation. Some of the key challenges
to designing an effective and workable multi-criteria tender include:
• identifying criteria and weightings that best reflect the needs of default superannuation
members
• establishing effective governance arrangements — including the selection and
oversight of the assessment panel — that ensure tender evaluations are independent,
objective and consistent, and creating rules that better ensure winning funds keep their
commitments
• choosing a number of winners that creates competitive tension in the tender process
without excluding relatively well-performed funds, and then allocating default
members across those funds.
The Commission welcomes participants’ views on whether the tender model assessed
addresses these challenges, and is effective and workable.
DRAFT FINDING 7.1
Relative to the baseline, the multi-criteria tender model would:
• promote member benefits by focusing competition on member satisfaction and
long-term net returns
• focus competition on aspects of performance that matter to members, and more
generally through the winning bid providing a market-wide performance benchmark
• create risks for integrity due to its vulnerability to subjective judgments, yet on the
other hand promote integrity through stronger accountability mechanisms
• not create any material risks to stability, since it is unlikely to lead to excessive
concentration or volatile movement of members and assets
members.

DRAFT REPORT
187
Table 7.2 Multi-criteria tender model: summary
Member
benefits
The selection criteria, in particular the results of member satisfaction surveys and
long-term net returns, help to directly promote member benefits.
These past performance measures provide ongoing incentives for good performance.
These benefits are likely to extend into the choice part of the superannuation product
market through the winning bids providing a market-wide performance benchmark.
Competition The tender would help achieve the competitive tension that would exist in a market with
active and informed users.
The tender opens up the competitive process to a broad range of funds.
By issuing non-prescriptive criteria with no pre-defined solutions, the tender process
provides some incentives for innovation.
Requiring a performance record of at least five years in any superannuation (or like)
product would create a barrier to entry.
Any market consolidation would be unlikely to reduce competition. Consolidation could
bring benefits, through scale economies and greater competitive pressure on upstream
providers.
Integrity There are risks to integrity due to vulnerability to subjective judgments. Constituting a
truly independent panel may encounter difficulties.
The tender has strong accountability mechanisms in the form of contractual obligations
and the threat of future exclusion, and also more broadly from fiduciary and statutory
obligations on trustees and other laws and regulations.
Stability The tender is unlikely to create instability, since it would not lead to excessive market
concentration, volatile movement of members and assets or a short-term focus.
System-wide
costs
The high costs to government and funds are offset by lower costs for members and
employers.
Which aspects of member services should be included in the tender criteria? Do
default members value the same services as choice members?
How should default members be allocated across the set of winning products? Are
there problems with sequentially allocating members into products with different
investment strategies, fees and services, as is implied by this model? What role could
a best-and-final-offer stage play in providing a select number of funds the opportunity
to improve their offer?
Regarding transparency, are there any problems with only withholding information that
can reasonably be shown to be commercially sensitive? What information would fall
under this exemption?

ASSISTED EMPLOYER CHOICE (WITH EMPLOYEE PROTECTIONS)
DRAFT REPORT
189
8 Assisted employer choice (with
employee protections)
Key points
• This chapter describes and evaluates a model enabling employers to choose any default
product offered by an approved provider, subject to that product meeting the minimum
criteria imposed by administrative filtering.
• The diverse needs of individual employers and employees suggest a ‘dual-list’ approach,
incorporating the use of both light and heavy filters, would provide the best outcomes.
− A mandatory light filter would determine eligibility to be chosen as a default product
(a ‘mandatory minimum standards’ list). A heavier filter applying stricter criteria and
employing higher performance benchmark hurdles would identify a narrower ‘preferred
default’ list of the best performing products as additional ‘opt in’ guidance for some
employers. The preferred default list would focus on ‘no frills’ products with higher
suitability for default members.
− The mandatory minimum standards list would provide well-resourced employers
flexibility to negotiate good deals for their workforces, while still providing a safeguard
against very poor choices. The optional preferred default list would help smaller to
medium sized businesses choose a high-performing product for their employees. This
approach would drive improved performance, as funds would compete for their products
to make the preferred list or to be chosen by large employers.
• The light filter would need to be stronger than the current MySuper authorisation criteria,
and provide for the exit of underperforming products. The requirement to demonstrate
adequate performance would be a more robust test than has been applied until now.
• The bodies involved in administrative filtering should be genuinely independent, free of
conflicts of interest and accountable for their decisions.
• Relative to the unassisted active choice baseline, assisted employer choice (with employee
protections) would:
– enhance member benefits by increasing the likelihood of members being placed in higher
quality products, and reducing the likelihood of them being placed in a poor product
– promote healthy competition by presenting product providers with incentives to perform
strongly against the preferred default list selection criteria or compete for corporate
tenders, but there would still be some scope for wasteful marketing to employers
– increase the potential for agency problems given the involvement of employers, although
the risk to the integrity of the system would lie substantially in the appointment process
for the preferred default list selection panel
– reduce search costs for employees, while increasing search costs for employers and
regulatory costs for funds and government.

DRAFT REPORT
This chapter describes and evaluates a model enabling employers to choose any default
product offered by an approved provider, subject to that product meeting the minimum
criteria imposed by administrative filtering. Although a model of this form is currently
used to determine default status, consistent with the approach discussed earlier in the
report, the Commission is developing this model, and the other models, as alternative
models distinct from the current default determination arrangements.
This chapter first explains why the Commission has chosen this model for assessment
(section 8.1). It then discusses how it would work in practice, including the selection
criteria applied and other design features of the model (section 8.2), and the supporting
institutional arrangements required (section 8.3). Finally the model is evaluated against the
Commission’s assessment criteria (outlined in chapter 2) relative to the unassisted active
choice baseline (section 8.4).
8.1 Why the Commission chose this model for
consideration
This model builds on the current regulatory settings which impose a legal obligation on
employers to make superannuation guarantee payments on behalf of their employees.
Employer involvement in defaults is an extension of those obligations — employers bear a
penalty if there is no product into which contributions can be made.
A further rationale for this alternative model stems from the hypothesis that an
intermediary — such as the government or an employer — might be better placed to
choose a superannuation product than an employee, given the difficulties many employees
have making active choices (chapter 4). Involving employers in default product choices
can also assist in generating competition and contestability in the sector, particularly where
disengaged members would otherwise not exert competitive pressures on product
providers.
There are also potential efficiency benefits associated with employers choosing products,
such as economies of scale from employers incurring search costs on behalf of many
employees, and ongoing reductions in transaction costs for employers by minimising the
number of providers they have to deal with.
A number of participants suggested that employers were best placed to choose a default
fund for staff, given they were well placed to understand staff needs. For example,
Workplace Super Specialists Australia (sub. 11, p. 3) stated:
We believe that the employer is the entity that is best positioned to select the default super fund
on behalf of its employees. The employer best understands the culture, demographics and
requirements of its own workforce.

DRAFT REPORT
191
AMP (sub. 42, p. 3) stated there were advantages from employer-chosen defaults:
AMP believes that regardless of the size of employer, employer chosen defaults provide
significant benefits to employees. The product provider works with the employer to drive
stronger member outcomes … There are also benefits to the employer from their choice of
default product. Efficiencies are gained with the ability to select one default product to be used
for all employees and to choose a product provider with more efficient administration services
and functionality.
Some participants did not consider that employers were well placed to choose default
products for their employees. For example, the Australian Chamber of Commerce and
Industry (sub. 27, p. 8) said:
There is no particular reason why as a class employers are well placed to identify the best
default fund and good reason to suspect that they are not. … many employers are not equipped
to identify the best default, or defaults, for their employees (the make-up or circumstances of
which may change over time), [there is the] the principal-agent problem and the fact that the
selection of funds is constrained by modern awards.
The Australian Securities and Investments Commission (ASIC) (sub. 41, p. 3) noted that
many employers were not well placed to choose a superannuation product:
‘Employers’ for superannuation purposes will cover a very wide spectrum of organisations and
individuals with differing level of financial literacy. While large employers may have dedicated
staff, others may be time poor or unaware of the types of issues that can arise in relation to the
choice of a default superannuation fund. … Employees may … perceive the employer to be
more knowledgeable about superannuation … and interpret the employer’s choice as an
implied recommendation.
Many smaller to medium sized (SMEs) employers in particular are not well placed to
absorb the administrative costs of superannuation and to make choices on behalf of their
employees. They may not have sufficient time and resources to effectively guide
employees to make good choices, or they may not feel comfortable doing so (box 8.1).
Why filter products for employers?
An administrative filter providing guidance to employers on the most suitable products for
their employees could assist them in making default product selections. Many employers
see default superannuation selection as imposing a significant burden and would appreciate
the task being made easier. The Restaurant and Catering Industry Association (sub. 10,
p. 1) stated:
Our industry is made up of a large number of small businesses, for whom superannuation is not
core business. Like their employees, they do not have the resources or the expertise to devote to
selecting quality superannuation funds. For these reasons we support the existing transparent,
independent process in which funds are shortlisted by experts based on their performance and
other factors.

DRAFT REPORT
Box 8.1 SMEs and employee superannuation
The industry superannuation fund REST recently commissioned research on entrepreneurship
in Australia and superannuation and retirement adequacy. In total 1701 employers from around
Australia completed online quantitative surveys in mid-2016 including 184 entrepreneurs,
282 SME owners/managers and 64 sole traders. For the purposes of the study, SMEs were
businesses with up to 199 employees.
In relation to SMEs and employee superannuation, the study found that:
• four in five SME owners (82 per cent) and managers (79 per cent) believe that it is the
responsibility of their employees to choose the fund that is right for them
• two in three SME owners (68 per cent) and two in five SME managers (39 per cent)
surveyed agreed that paying superannuation contributions is a chore and want this task
managed as quickly and cheaply as possible
• about four in five SME owners (78 per cent) and two in three SME managers (67 per cent)
want to ‘do the right thing’ by their employees in terms of the superannuation, but there is a
limit to how much time and effort they are willing to spend on it.
Source: REST (2016).
An administrative filter also seeks to reduce the potential for bad outcomes for members
stemming from principal–agent problems associated with employers choosing products.
The Commission has previously identified a number of reasons why employers might not
have the incentive to make a decision that is in the best interests of their employees:
• employers do not receive the direct benefits of a well-performing product, and therefore
might have little incentive to invest time and effort in making choices that are in their
employees’ best interests
• employers might choose a product with the least onerous administrative arrangements
for the employer
• search costs might be high particularly if choosing from a large number of products,
especially where an employer lacks information and expertise
• some funds might have additional benefits specific to employers, such as access to
financial products
• employers might be driven by the legal ramifications of their choices, rather than the
best interests of members (PC 2012, p. 60).
The Australian Institute of Superannuation Trustees (AIST) (sub. 28, p. 36) saw default
product selection by employers as appropriate only after a transparent filter had been
applied:
It is appropriate that employers exercise choice of fund from within a range of eligible
products. Whilst it is acknowledged that many employers do not have the skills or time or
resources to choose from an extensive list of funds, it is appropriate that employers select a
fund that after transparent consideration has been deemed appropriate for the general industry
or occupation that the employer operates within.

DRAFT REPORT
193
The Financial Planning Association of Australia (sub. 29, p. 8) discussed the advantages
and disadvantages of employing a filter:
The advantage of using an administrative filter is that it reduces the cost of selecting defaults
because funds that seem obviously inappropriate don’t have to be rigorously assessed. The
disadvantage is that, without interpretation, there is a risk that temporary characteristics of
funds may make them ineligible to be a default fund (at least until the next assessment).
In summary, an administrative filter can assist employers in choosing a default product that
is in their employees’ best interests by addressing information complexity and asymmetry.
Narrowing the range of default options can make choice more tractable for employers
(as discussed in earlier chapters, too many choices can be overwhelming), and it can
(by imposing minimum standards) protect employees from bad choices by employers
while injecting some trustee-level competitive pressure.
8.2 How would administrative filters operate?
This section discusses the design and technical features of the proposed model, including
broad criteria that could be used to determine the list of eligible default products. While the
Commission’s model is developed anew, the Commission has drawn on existing features
of the MySuper authorisation process and the criteria in the Fair Work Act 2009 (Cwlth)
(Fair Work Act). These provide valuable lessons and reflect relatively recent consultation
exercises (including previous work by the Commission). The discussion also draws on the
2010 Cooper Review, the Commission’s stage 1 work on developing criteria to assess the
competitiveness and efficiency of the superannuation system, submissions to this inquiry
and various other material.
A heavy or light filter, or both?
Inquiry participants generally suggested one of two types of administrative filters:
• The first variation involves the application of a mandatory ‘light’ filter, designed to
eliminate poor quality or underperforming products from the list of eligible products.
This variation of the model is effectively a market-based (or decentralised) model,
leaving employers with a large degree of discretion to choose default products.
• The second variation of the model involves a ‘heavy’ filter, which would be designed
to ensure employers can choose only from products assessed to be of relatively high
quality. This would confine employer choices of default products to a relatively
narrower list of options.

DRAFT REPORT
In practice, there are pitfalls associated with simple notions of the ‘light’ and ‘heavy’
filters operating on a standalone basis, as noted by Industry Super Australia (ISA) (sub. 40,
p. 53):
A weak filter that covers the entire system would be of little utility. A strong filter pursuant to
which few funds are eligible, and that covers the entire system could result in no appropriate
options or an insufficient number of options for certain segments of the population.
PwC (sub. 12, p. 6) noted that the choice could influence perceptions of the government’s
responsibility for outcomes:
If you apply an administrative model, you are basically saying that any fund that meets this bar
is open to being a default fund. For example, this could simply be any MySuper fund as they
have already met a certain minimum standard. … This will likely result in less responsibility
for the Government should one of these funds fail, than if you actively try to select a fewer
number of “best” on offer through a mixture of quantitative and qualitative tests and one of
these funds fail.
The Commission’s preferred approach for employer default product selection is the
‘dual-list’ model: a mandatory light filter determining absolute default product eligibility
(a ‘mandatory minimum standards list’), and the heavy filter providing employers with
guidance about products (a ‘preferred default list’). The Commission considers employing
such an approach would attain the best features of the light and heavy filters, while also
minimising the more negative aspects associated with each when used alone.
To be on the preferred default list, products would be required to meet more stringent
selection criteria and higher performance benchmarks. They would also need to be more
suitable as default products for the broader population — that is, they would be relatively
‘no frills’ products with a simple set of features that are easily comparable across products
and particularly suited to relatively disengaged members. The default selection process
would be designed so that products should primarily compete on long-term net returns and
costs for a given level (threshold) quality of service, not on the quality or range of ancillary
services offered per se (chapter 3).
The process for inclusion on the preferred default list could be considered in some ways
similar to the process discussed in chapter 7 relating to the multi-criteria tender model.
However, the process would not require the same tender-style governance arrangements
that are designed to accompany a mandatory choice outcome as in chapter 7, and would
avoid the cost of a formal allocation mechanism post-tender, with employers playing that
allocative role.
A dual-list approach could help meet diverse needs
Many businesses (particularly SMEs) that lack the expertise or resources to choose a
suitable product would prefer a relatively short list of highly suitable products to choose
from to simplify their choice, and better ensure that they choose a relatively good product.

DRAFT REPORT
195
Employees of these businesses are therefore likely to benefit from a ‘heavier’ filter because
their risks of being placed in a poorly performing product are diminished.
Many larger employers, on the other hand, are well placed to negotiate good deals for their
employees, and might employ methods such as corporate tenders to choose products.
Being required to choose from a preferred default list (which is likely to be a relatively
small proportion of eligible products) would potentially restrict the capacity of these
employers to negotiate good deals (and could compromise the competitiveness of in-house
tender processes). Providing the capacity to knowledgeable employers to choose a default
product from a longer minimum standards list is therefore likely to promote the best
interests of employees and promote competition.
The Commission envisages that the preferred default list would, over time, be populated by
relatively generic products suitable for all employees. It is likely that some employees,
based on their workplace or occupation, would benefit from products not contained on the
preferred default list, and that some providers would only be seeking a segment of the
default market. It would therefore be important to ensure flexibility to provide employees
with the best outcomes.
UniSuper (sub. 20, p. 9), for example, noted it has previously only applied for its MySuper
product to be listed in higher education awards. Whether or not it was included on the
preferred default list, UniSuper should be free to offer its products as defaults, and tertiary
sector employers should still be free to choose them, provided the products continued to
meet the light filter.
The dual-list approach would also promote competition and drive improved performance
as some funds contained on the mandatory minimum standards list would be likely to
compete for their products to make the preferred default list.
The Commission is also supportive of a dual-list approach in this form of model because it
would be one way to avoid capping the overall number of listed products while still
providing a relatively high level of filtering for employers that preferred this. It would
effectively make the ‘cap’ standard based. CHOICE (sub. 31, p. 15) expressed support for
a standard-based approach:
… a cap on the number of qualifying products which is too low may unnecessarily limit
competition. It would be better to base the system on a standard which products must meet,
then progressively lift this standard in an attempt to drive competition.
Arbitrarily capping the number of products that could potentially be listed could impose
artificial constraints on competition and restrict access to default products that might
benefit particular employees.
There is, however, a case for specifying a minimum number of products to be included on
the eligibility list under the heavy filter in order to ensure meaningful competition and that
incumbent funds do not take their membership base for granted.

DRAFT REPORT
In summary, the Commission considers that given the diverse needs of individual
employers and employees, a ‘dual-list’ approach, incorporating the use of both light and
heavy filters, is likely to provide the best outcomes under the assisted employer choice
(with employee protections) model.
• A mandatory light filter would be used to determine eligibility to be chosen by any
employer as a default product (a ‘mandatory minimum standards’ list), while a heavier
filter, applying stricter criteria and employing higher performance benchmark hurdles,
would identify a narrower ‘preferred default’ list of the best performing products as
additional ‘opt in’ guidance for some employers. The preferred default list would also
focus on ‘no frills’ products with higher suitability for default members.
• Maintaining an optional preferred default list would help SMEs to choose a good,
highly suitable product. Maintaining a broader mandatory minimum standards list
would provide employees with a safeguard against very poor choices by their
employer, while giving well-equipped employers flexibility to negotiate good deals for
their workforces.
How could it be ensured employers acted in good faith?
A number of stakeholders expressed concern about employers either being subject to
inducements from product providers, or choosing a product not in the best interests of
employees because of benefits to the employer (concerns which would be exacerbated in a
model providing employers with a high degree of freedom to choose products). For
example, AIST (sub. 28, p. 24) stated:
We are concerned the problem of direct and indirect inducements to employers is a clear
market impediment. We note that this is a contentious subject where proof is often difficult to
obtain. Commercial incentives and sales pressures have not been removed from the system.
There is a need to examine evidence of principal-agent issues in default fund selection,
including any possible direct or indirect incentives provided to employers.
ASIC (sub. 41, p. 2) noted there were no current requirements for an employer to select a
fund that is in the member’s best interests:
Although employers currently have the legal responsibility to make a decision in relation to the
default superannuation product, employers are required to neither select a fund that is in the
best interests of their employees nor to put their employees’ interests ahead of their own in
selecting the fund.
However, there are existing regulatory prohibitions on providers offering inducements to
employers on the condition that their employees join the fund. Section 68A of the
Superannuation Industry (Supervision) Act 1993 (Cwlth) (SIS Act) currently prohibits such
inducements, although contravention only gives rise to the possibility of future civil action
by aggrieved individuals (ASIC, sub. 41, p. 3).

DRAFT REPORT
197
To a significant degree, a well-calibrated and enforced mandatory minimum standards
filter would provide a safeguard against any mal-intentioned behaviour by employers, and
this is the most effective mechanism to address those risks. However, whenever it falls on
employers to choose default products there is an inherent risk some will not act in the best
interests of their employees. There is also an inherent risk of inertia, where employees
might be left in products that are not ideal for their needs. Attempting to prevent these
problems entirely through regulation would impose a large red tape burden on employers.
As such, the potential for principal–agent problems remains a residual risk under this
model.
The Commission has considered and decided against several options for imposing
additional duties on employers (because, as explained below, this is unlikely to be the best
way to deal with the inducements issue).
The first relates to employees being able to take legal action in the event that employers
make poor default product choices. Legal action against employers over various aspects of
pension plans is relatively common in the United States (box 8.2), but not in Australia.
With a mandatory minimum standards filter in place, the Commission does not see it as
purposeful that employers should be subject to subsequent legal action for choosing a
product within that filter. In the 2012 inquiry (PC 2012, p. 175), participants raised
concerns that a risk of legal liability for employers if they choose a product which in
hindsight turns out to be unsuitable might prevent employers from exercising any degree of
discretion at all. This is reinforced given that employees in defaults will generally have not
taken the option of exercising their own choice, preferring to defer it to their employer.
The Commission is also not attracted to softer options that involve mandatory employer
disclosure to their employees of reasons for choosing a particular fund and of not having
any unresolved conflicts of interest. It is unlikely that such requirements would serve an
effective purpose, and they could lead to perverse outcomes. A genuinely mal-intentioned
employer might be seen to have complied fully with the letter of their requirements (with
no adverse consequences). Conversely, another employer making a wholly explicable
choice that nevertheless led to bad outcomes due to factors outside their control could incur
a liability of uninsurable value. As discussed earlier, the Commission anticipates that some
employees would benefit from their employers shopping around and tailoring products for
their workplaces. From both a competition and stability perspective in this model, it would
also be important to ensure products not on the preferred default list did not have their
viability excessively compromised by ill-considered regulations.

DRAFT REPORT
Box 8.2 Pension plan related legal action against employers is
relatively common in the United States
Under US law, employees are able to take legal action against their employers (as their
retirement plan administrator or fiduciary) for not acting in employees’ best interests in
accordance with 401(k) plans. Section 404(a) of the Employee Retirement Income Security Act
1974 (US) (ERISA) requires fiduciaries to discharge their duties with respect to a plan solely in
the interests of participants and their beneficiaries and for the exclusive purpose of providing
benefits and defraying reasonable administration expenses. Legal action can include matters
such as appealing denied claims for benefits and recovering benefits owed, clarifying rights to
future benefits, accessing retirement plan documents and to address breaches of fiduciary
duties.
Over the past decade, a significant number of lawsuits have been filed in the United States
against retirement plan sponsors for failing to properly administer plans. Allegations have
included a failure to monitor excessive fees, directing employee savings toward
affiliate-overseen investment products, and favouring more expensive retail mutual funds over
less costly alternatives. Employers are exempt from fiduciary liability if employees are provided
with individual investment options under a pension plan and given sufficient opportunity to
exercise control over assets in their account (ERISA, s. 404(c)). Essentially this provides a ‘safe
harbor’ for plan fiduciaries where plan participants self-direct their investments, ensuring
employers won’t be legally responsible for their employees’ investment mistakes.
Sources: FindLaw (2016); Shepherd Smith Edwards and Kantas (2015).
Finally, some participants have also suggested a ‘better off overall test’ (BOOT) could be
employed either in conjunction with an administrative filter, or as an alternative to
employing one. The Commission’s view is that a BOOT is not a satisfactory alternative to
an administrative filter (box 8.3). Further, such an approach would be unnecessary with a
dual filter as the latter would be designed to safeguard employee interests, while also
providing employers with a reasonable level of flexibility.
Ultimately, it is implausible to impose an obligation on employers to act in their employees
best interests, not least because the best interests of employees are likely to vary
(sometimes significantly) across the workplace, and employers cannot know what they are
(especially given they are typically not superannuation experts). Rather, the most effective
way of ensuring employers’ actions are consistent with the best interests of employees is to
place obligations on superannuation fund trustees (that is, to ensure inducements are not
offered in the first place). Providers should be prevented from offering employers
inducements (including preferable deals on non-superannuation products) to choose their
products as defaults, and regulators should actively enforce this. The Commission
considers the existing provisions in the SIS Act, and their enforcement, would require
strengthening if the model presented in this chapter were to be adopted.

DRAFT REPORT
199
Box 8.3 Could a measure such as a ‘better off overall test’ be an
alternative to a filter?
One option for allocating default products, either as an alternative to a filter or in conjunction
with it, would be the imposition of a ‘better off overall test’ (BOOT) as is often applied in other
areas of workplace relations. Effectively, it would mean that an employer could remain with their
existing product or switch to a new default product provided an independent decision maker
determined that this arrangement would, on balance, leave employees better off overall than
previous arrangements.
A BOOT would be less straight-forward on an issue such as superannuation fund choice
relative to other areas of industrial relations because future returns cannot be known with
certainty and the respective merits of products could be contentious.
The Commission’s 2012 default superannuation inquiry (PC 2012) considered the merits of
employing a BOOT-style arrangement. Concerns raised about this proposal included that such
a test would not necessarily cover a situation where the majority of employees might be better
off but some worse off, that it could be burdensome for employers to demonstrate the test had
been met, and that employers would not necessarily know for certain whether employees would
be better or worse off. It was also suggested employers might be reluctant to use such
provisions (which might require employers to obtain legal or financial advice).
What would the selection criteria be?
A key aspect of administrative filtering is determining the selection criteria for both the
mandatory minimum standards and the preferred default list.30 A number of participants
suggested that the current criteria employed by the Australian Prudential Regulation
Authority (APRA) when granting MySuper authorisation represented a suitable mandatory
minimum standards filter (box 8.4). Reliance on these criteria would effectively mean that
all authorised MySuper products would be eligible to be chosen as defaults.
30 This discussion is similar to that in chapter 7 regarding assessment criteria under a multi-criteria tender
model.

DRAFT REPORT
Box 8.4 Would allowing any authorised MySuper product represent
an effective minimum standards filter?
A number of participants suggested that the existing MySuper authorisation process
represented an adequate minimum standards filter to ensure employees were being placed in a
quality product. For example, Colonial First State (sub. 25, p. 1) said:
Given the rigorous and costly process of designing MySuper products and then achieving APRA
authorisation, we believe this is the simplest and most effective framework for default arrangements in
the future.
AMP (sub. 42, p. 2) expressed a similar view:
The competitive process to determine which products are eligible to be defaults should be based on
APRA’s strict requirements for a MySuper licence and consistency and transparency of product
information provided. All compliant MySuper products should therefore be able to compete equally as
default products eligible for selection. An open market with employer choice will put pressure on
incumbent funds to compete, which will result in continued downward pressure on fees, increased
product innovation and therefore better outcomes for members.
APRA (sub. 33, p. 2), however, suggested the current authorisation criteria need strengthening:
… both the current legislative MySuper authorisation criteria and the scale test could be strengthened
with a view to MySuper products being of overall better quality. … since 1 July 2013, there has been
considerable variation in net returns and fees for different MySuper products, leading to a wide range
of outcomes … While there has been some evidence of reductions in fees and costs since MySuper
products were introduced, particularly for products with previously very high fee levels, there is clearly
room for further improvement. Further, while many MySuper products have achieved their net return
targets over the past few years, some have fallen well short.
Australian Super (sub. 19, p. 4) raised issues about the suitability of the current MySuper
authorisation process for use as a minimum standards filter:
The MySuper licensing regime has no direct filter to prevent long term low performing funds from
becoming default funds, or ceasing to be default funds.
The Corporate Super Association (sub. 35, p. 2) also expressed concern about opening the
selection process to ‘any MySuper’ product:
… we understand that many of the default funds currently used, particularly in the not-for-profit sector,
have been performing very well in services and investment, whilst there are MySuper products in the
wider market that do not perform so well … To open the market to every MySuper fund provides
employees and small employers with decisions they may be ill-equipped to make … The outcomes
may end up worse than in the existing default system.
A number of participants suggested ways in which the existing MySuper authorisation criteria
could be strengthened to make them more suitable for use as a minimum standards filter. For
example, the Financial Services Council (sub. 38, pp. 22–23) suggested that scale, fees and
long-term net investment returns should be the main criteria for assessing products, together
with ancillary criteria such as capital adequacy, governance and whether income stream
products are offered in retirement. The Financial Services Council also suggested the
Australian Government should benchmark fees and net returns for MySuper products, and that
products seen to be performing poorly over the medium-term could have their MySuper
authorisation revoked (after APRA worked with trustees initially to determine the reasons for
underperformance).

DRAFT REPORT
201
Many other participants suggested the criteria employed currently by the Fair Work
Commission (FWC) (or some variation of these criteria) to determine the eligibility for
MySuper products to be included in awards represented a good basis for developing
selection criteria. Some other participants simply listed those factors they considered
should be applied without reference to the MySuper authorisation or FWC processes.
In its 2012 Default Superannuation inquiry, the Commission did not support moving to an
‘any MySuper’ filter to determine eligibility for default status due to concerns about the
likely variations in the quality of MySuper products that would be offered initially, the
inherent principal–agent issues between employers and employees, and the stated lack of
interest and expertise on the part of many employers when it comes to choosing a default
product. The Commission did not rule out in 2012 the current MySuper authorisation
criteria filter defining suitable minimum standards for default eligibility in the future if the
average quality of MySuper products increased over time and there was an observed
narrowing in the quality differential between them. However, these circumstances have not
eventuated. APRA (box 8.4) has noted there is still ‘considerable variation’ (sub. 33, p. 2)
between the performance of MySuper products, and many have failed to achieve their
target returns. Further, there have been questions raised by some commentators about the
long-term viability of some products exhibiting net outflows (Boyd 2017).
The Commission’s 2012 report proposed additional non-prescriptive criteria for default
status eligibility on top of the MySuper authorisation criteria. The Australian Government
subsequently legislated criteria for the FWC to consider when determining eligibility to be
listed in awards (box 8.5) that were largely based on the Commission’s recommendations.
One criterion included in the current process that requires reconsideration is insurance. As
discussed in chapter 3, there are challenging trade-offs in considering the merits of default
superannuation products, without also considering insurance products. The criteria applied
for both lists in the administrative filter should relate to the superannuation product only.
This would ensure that the filter was consistent with the proposed objective of
superannuation, where insurance has at best only an indirect relevance to that objective.
Clear rules for minimum standards of insurance product quality and level of cover should
be determined separately by regulation. Decision makers should then ensure these
regulated standards and levels of cover in the selected default products are delivered by
products that make it onto the filter.

DRAFT REPORT
Box 8.5 Current criteria for default products in awards
Following the recommendations from the Commission’s 2012 Default Superannuation inquiry,
the Australian Government introduced a quality filter in the Fair Work Act 2009 (Cwlth) for
products that would be eligible for inclusion as defaults in modern awards. Under section 156F
of the Act, the Fair Work Commission is required to compile a ‘Default Superannuation List’ on
the basis of a set of criteria, which include:
• appropriateness of the product’s long-term investment return target and risk profile
• the expected ability of the fund to deliver on the product’s return target, given its risk profile
• the appropriateness of the fees and costs, given:
– its stated long-term investment return target and risk profile
– the quality and timeliness of services provided
• the net returns on contributions invested in the MySuper product
• governance practices, including mechanisms to deal with conflicts of interest
• appropriateness of insurance offered in relation to the MySuper product
• quality of advice given to a member
• the administrative efficiency of the fund.
The Fair Work Commission may also consider any other factors considered relevant. As a
consequence of a Federal Court decision, the Fair Work Commission has been unable to
conduct the review of default superannuation funds. It is awaiting new appointments to its
expert superannuation panel (FWC, sub. 51).
The Commission’s 2012 report also emphasised that judgment would be required to weigh the
importance of each criterion, to determine whether to limit selection criteria to those mentioned
in legislation or include others, and to assess which criteria best respond to the characteristics
and needs of the employees covered by each award (PC 2012, p. 134).
The Commission has also considered criteria developed for similar default selection
processes elsewhere in the world (appendix B). The KiwiSaver criteria applied in New
Zealand (box 8.6) notably incorporate a number of measures relating to size, capacity and
track record of institutions (including a ‘corporate strength’ criterion). Some commentators
in New Zealand (Littlewood 2012) have suggested this has led to choices that might be
‘safe’ for the NZ Government rather than choosing providers that are necessarily focused
on the best interests of members.

DRAFT REPORT
203
Box 8.6 Selection criteria for KiwiSaver default providers
In New Zealand, selection of KiwiSaver default providers is via a multi-criteria tender. However,
the criteria used could apply equally to an administrative filter. The current criteria can be
summarised as:
1. investment capacity — includes a track record of capability and performance, the ability to
meet the NZ Government’s default specifications, and the ability to offer a full range of
KiwiSaver funds in addition to the default
2. corporate strength — a reputable organisation with sufficient financial capacity and
infrastructure
3. administrative capability — able to process a large number of transactions, and flexible
enough to respond to frequent changes
4. track record and stability — commitment to accept the KiwiSaver default mandate and to
carry it out at the fees agreed
5. investor education and advice — the ability to provide investor education and impartial
financial advice, including to proactively contact default-enrolled members to ensure they are
invested appropriately.
Source: NZCO (2013).
The Commission’s proposed criteria for filtering products
The current MySuper authorisation process in principle provides a good basis for a ‘light’
minimum standards filter. However, as noted above, there are some weaknesses with the
current process. Data indicate that since implementation in 2013, there has been
considerable variation in net returns and fees for different MySuper products, leading to a
wide range of outcomes for members across these different products (APRA, sub. 33, p. 2).
The Commission considers that strengthened MySuper authorisation criteria would be a
suitable basis for a mandatory minimum standards filter.
In particular, the Commission considers the minimum standards filter criteria should
consider the performance of products to a greater degree than the current authorisation
process does. Funds should be required to self-report against predetermined targets, with
this self-reporting to be backed up through independent auditing by, or on behalf of,
APRA. Products displaying persistent underperformance against their targets should be
excluded from default eligibility. Product providers considered unlikely to meet their
trustee obligations should be similarly excluded.
The current legislated MySuper authorisation process deals with these matters to a degree.
However, the process for removing authorisation of poorly performing products is
convoluted and difficult and the consequences of such removal for existing members have
not been articulated by Government or the regulators. The Commission sees this as failing
to ensure default members are in products of a reasonably high standard. The imposition of
an enhanced minimum standards filter is a better way of ensuring this.

DRAFT REPORT
The current scale test employed under MySuper also appears to have had little impact,
particularly in view of barriers to fund mergers (such as bulk transfer rules). To meet the
Commission’s proposed minimum standards filter, providers would have to demonstrate
— and have independently verified — that scale was not an impediment to their
performance. The Commission intends that this not be a cursory review.
The Commission also assessed the current legislated criteria contained in the Fair Work
Act, and the modifications suggested by participants. On balance, the Commission has
concluded the current Fair Work Act criteria provide sufficient coverage of relevant
considerations, and represent an adequate basis for developing the heavy filter to determine
the preferred default list. However, consideration could be given to refining some of the
language around the criteria, such as the use of the word ‘appropriate’ which would not
provide sufficient guidance to the decision maker.
The Commission envisages that to qualify for the preferred default list, products would
need to be considered the most suitable for default members in view of their ‘no frills’
nature, and would need to exhibit high governance standards and consistently superior
performance relative to rival products in delivering long-term net returns to members. It is
expected that inclusion on the preferred default list would be difficult to achieve.
A comparison of how the Commission’s proposed criteria for filtering products relates to
the current MySuper authorisation process is contained in table 8.1. Product providers
would be required to self-report against criteria on an ongoing basis under both the
mandatory minimum standards list, and the preferred default list.
Many of the suggestions for additional criteria raised by stakeholders are factors the
Commission assumes would be taken into account by the relevant decision maker under
the proposed criteria. Many of these areas are also covered by separate legislation with
which superannuation funds are required to comply.

DRAFT REPORT
205
Table 8.1 Comparison of MySuper and administrative filter criteria
Characteristic Current MySuper
authorisation process
Mandatory minimum
standards criteria
Preferred default list criteria
Need for
accreditation
Funds must obtain
MySuper authorisation to
receive default
contributions
Funds must comply with
minimum standards criteria
to receive default
contributions
Voluntary for funds
Accumulation
product features
Simple set of core
services including
administration,
investment and intrafund
advice
Must include core services.
Limited to products
considered suitable for
default members, with
relatively homogenous and
easily comparable ‘no frills’
features
Must include core services.
Limited to products
considered the most suitable
for default members, with
intuitive, effective and easily
comparable ‘no frills’ features
Investment
strategy
Single diversified
strategy or set of
life-cycle strategies
Same as MySuper Single diversified strategy
only
Fee structure MySuper product can be
offered at a single price
or fee structure, or at
different prices
Same as MySuper Same as MySuper
Investment fees Investment fees can
include a performance-
based fee subject to
limitations
Same as MySuper Same as MySuper
Exit and
switching fees
Limited to cost recovery Same as MySuper Same as MySuper
Commissionsa Prohibited Prohibited Prohibited
Investment
performance
Funds must report return
and risk targets, but no
specific performance
standard
Products that persistently
underperform (over multiple
years) relative to pre-stated
targets to be excluded
Products must demonstrate
consistently higher
performance relative to the
average of all other default
products
Fund scale Funds must demonstrate
that the scale of the fund
does not disadvantage
members relative to
members of other
MySuper products
Funds must regularly
demonstrate, and have
independently verified, that
fund scale is not an
impediment to performance
Same as minimum
standards criteria
Governance
standards
Funds must be likely to
comply with trustee
obligations
Same as MySuper Funds must demonstrate
consistently high
governance standards and
evidence of meeting trustee
obligations
Insurance Life cover and total and
permanent disability
cover must be provided
on an opt‑ out basis
Insurance offered as part of
default product must meet
minimum regulatory
standards (chapter 3), and
details must be clearly and
accessibly disclosed
Same as minimum
standards criteria
a Paid by trustees or fund providers to financial advisers or third parties.
Sources: Australian Government (2011); Superannuation Industry (Supervision) Act 1993 (Cwlth).

DRAFT REPORT
Should there be multiple lists segmented by industry or occupation, or
a single approach for the entire workforce?
Historically, the Australian system for allocating defaults has been linked to awards,
meaning lists of available default funds have varied across industries and, to some degree,
occupations (chapter 1). The question is whether maintaining a segmented approach with
multiple lists is warranted for the alternative model outlined in this chapter. A number of
participants commented on this (box 8.7).
Box 8.7 Participant views on whether lists should be segmented by
industry or occupation
ISA (sub. 40, p. 29) suggested there were good reasons to differentiate default allocation by
industry:
… there are good reasons for allocating different kinds of workers not just to different kinds of
products, but also to different providers. Different workers operate under different employment
conditions and environments. For example, although impersonal or electronic communication-based
approaches are perhaps the norm, it remains the case that some workers are on ships, or in other
remote worksites. Providers that are familiar with and connected to industries can and do
accommodate industry-specific needs in ways that a generic provider cannot …
The Corporate Super Association (sub. 35, p. 7) expressed concern about having only one
filter:
[The CSA] would be concerned about a single filter covering the entire system. Not every work force
has the same needs. It would be prejudicial to use the same assessment tools in respect of everyone
… homogeneity needs to be avoided in superannuation.
AMP (sub. 42, p. 2) pointed to benefits of enabling employers to consolidate default accounts
across a workplace:
Product providers can work with the employer to tailor services to their workforce, such as education
seminars and advice services which encourage positive member actions including consolidation of
superannuation accounts. Consolidation can result in significantly higher average balances …
ASFA (sub. 24, p. 9) noted that even if products are tailored towards cohorts of workers, this is
still very different to tailoring them towards individual workers:
By construction, tailored products for each cohort, based on a set of assumptions, would be tailored for
the ‘average’ individual in each cohort, not to the range of individuals. This means that not all members
would be allocated to the product that bests suits their circumstances.
The Australian Chamber of Commerce and Industry (sub. 27, p. 11) noted the current system
can lead to account proliferation.
On one hand, there are likely to be variations for some groups across the workforce that
would mean they could benefit from tailoring of products, such as young people (who
would typically benefit from higher risk investments), or construction workers (to whom
some insurance might be more important). However, these benefits appear on closer
examination to be limited to relatively few segments of the workforce (and are also largely
related to insurance offerings attached to the default product, rather than the default

DRAFT REPORT
207
superannuation product per se). Linking defaults to awards has enabled this specialisation
and ensured a default system when none was otherwise available.
But today, 25 years after the inception of a relatively low rate of super contribution (3 or
4 per cent depending on employer size) Australia has a system with in excess of three
times that original minimum level as a proportion of earnings and with wider workforce
coverage. The ability to innovate and constantly reflect on performance is a matter
primarily left to trustees and superannuation fund management. However, with guaranteed
inflows under some industrial agreements, the incentives are not as strong as when a fund
is reliant on competing openly for members. And innovation, especially in investment
strategy, matters, particularly with lower growth outlooks being the new reality in the
western world. With the Australian superannuation system now exceeding $2 trillion and
growing rapidly through compulsion and tax incentives, relying on good intentions —
evident though they are — and trustee standards, may not be sufficient to see ongoing and
meaningful innovation. Default members, many of whom are very disengaged, are most at
risk of failing to observe a lack of innovation until late in life, possibly too late.
Currently employers with staff employed under different awards might be compelled to
place staff in a range of default products due to differences in product eligibility across
awards. Varying default eligibility based on awards, or an alternative mechanism, can
therefore make it difficult to obtain benefits that could accrue to members (and employers)
from having products selected as tailor-made for default members.
On balance, one single ‘minimum standards list’ and one ‘preferred default list’ across the
workforce would likely be sufficient. Those industry-based funds that specialise in default
members are generally well-placed to meet the minimum standards filter requirements, and
on most criteria should be able to readily adapt to potentially meet requirements of the
preferred default filter. While individual default products that maintain an element of
specialisation might be particularly suitable (or unsuitable) for specific employee cohorts
(characterised either by age or occupation), having employer choice enables those products
to be developed and selected.
Related to the question of tailoring is the question of how corporate superannuation funds
would be dealt with under an administrative filtering model. The Corporate Super
Association (sub. 35, p. 2) noted the exclusion of corporate funds from default
arrangements could adversely affect some employees.
The retention of corporate funds as default funds in awards, including the retention through
grandfathering, has been the result of … bargaining and negotiations. To exclude these funds
from awards would result in significant adverse effects for employees because many provide
benefits that exceed [superannuation guarantee] minimum, … tailored … insurance
arrangements, and are otherwise generously supported by the employer-sponsor …
The Commission agrees that arrangements for corporate funds (or defined benefit
arrangements) can be relatively favourable to workers in many cases. In principle, there

DRAFT REPORT
should be no impediment to corporate funds being eligible as default products provided
they meet the criteria for the minimum standards list.
For both industry funds and corporate funds, the scale test may be the real challenge and,
as noted earlier, the ability to demonstrate a capability to sustain performance is intended
under the filters to be a more robust test than it has proved to be until now.
8.3 Institutional arrangements
Who would apply the filters?
Under current arrangements, APRA undertakes MySuper authorisation and the FWC31
determines eligibility for listing in awards. The role given to the FWC reflects that
superannuation entitlements have largely emerged from the workplace relations system and
that choice of fund has been a key focus of industrial parties (chapter 1).
As discussed in chapter 3, the Commission considers an organisation suited to selecting
default products would possess the following characteristics:
• be dispassionate and free of conflicts of interest
• be accountable for its decisions such that, in the presence of poor process or
unanticipated but evident conflict, Ministerial ability to cease a process and/or replace a
panel remains a sanction readily available.
Some participants have suggested APRA as a suitable body. APRA (sub. 33, p. 5) itself
has suggested that, as the prudential regulator for the entire industry, this would not be a
suitable role for it:
As the prudential regulator for the superannuation industry, and given our role in authorisation
of MySuper products offered by RSE [registrable superannuation entity] licensees, APRA is a
key stakeholder in the development and implementation of any default allocation model. …
However given APRA’s broader role in relation to the superannuation system, it would be
inappropriate for APRA to have any decision-making role in relation to the selection of a
sub-set of MySuper products that may be eligible default funds under any default allocation
model.
31 As a consequence of a Federal Court decision, the FWC has been unable to conduct the review of default
superannuation funds. It is awaiting new appointments to its expert superannuation panel (FWC, sub. 51).

DRAFT REPORT
209
ISA (sub. 40, p. 33) also suggested a role in default product selection could conflict with
APRA’s prudential regulation role:
… adding default fund selection to the responsibility of one of the financial services regulators
in Australia would be inconsistent with the “twin peaks” model. The “twin peaks” model,
which separates conduct regulation and prudential regulation, is based on the Tinbergen
principle, which requires focused instruments of policy. Adding a new policy goal – default
super fund selection – could place tension on the regulators’ capacity to perform their core
mission.
The Commission shares this view to a degree. In order to be seen to be treating all funds
equally from a prudential regulation perspective, it is important that a prudential regulator
is not involved in making relative quality judgments to avoid any perception that those
products nominated as suitable defaults are ‘backed’ by the regulator and regulated
differently to others. In short, APRA would be unsuitable for applying the heavy
administrative filter for the preferred default list.
The more subjective the decision, and the closer it is to picking winners and losers, the
greater is the case for a specialised Ministerially-accountable process. And the harder it is
for an independent regulator to then fill that role.
APRA would be potentially more suitable for enforcing the mandatory minimum
standards, in the same way it already does with MySuper. However, to ensure APRA was a
suitable body for this purpose, the criteria employed would need to be relatively objective
and not require subjective judgements about the relative merits of products.
The role equivalent to compiling the Commission’s proposed preferred default list is
currently undertaken by the FWC, following advice from a Default Superannuation Panel.
At best, the justification for the Panel’s location within the FWC is that awards specify
default funds. In other respects, the FWC, as currently constituted, is ill-suited to the role.
As in some of its other functions, the FWC’s capacity to make merit-based assessments of
default funds is limited because it does not have the in-house expertise to deal with the
issues. Moreover, the FWC’s decisions have generally been based on precedent and the
generally harmonious representations of unions and employers in favour of the status quo
(PC 2012).
The Commission has previously commented that the process of making appointments to
the FWC as a whole is in need of reform to ensure that the decision makers are expert and
impartial (PC 2015b). Such a capability would be essential in any default superannuation
system that determined default products on merit. Reforms would thus be an essential
precursor for any ongoing FWC role in superannuation. In particular, the appointment and
accountability structures would need significant adaptation. These are discussed below.

DRAFT REPORT
How should those applying the filters be appointed?
Any entity given the task of determining default selection has the same problem: those
people with the expertise required to undertake default selection are likely to have been
involved in the financial services industry and may at times then be perceived to have
potential biases or conflicts of interest. The only practical offset to this (other than to
perversely favour appointees with weak capabilities) is to adopt the relatively simple (if
somewhat brutal) strategy of ensuring that the panel is accountable to a Minister such that
poor process or evident (real or perceived) conflict can be dealt with immediately and in a
manner that leaves no doubt about integrity. Existing structures (the FWC, as above, but
also APRA or ASIC) are designed largely to promote independence and consequently do
not incorporate this accountability to ministers.
The Commission’s 2015 Workplace Relations report recommended that Australian, State
and Territory Governments create an expert appointments panel which would provide a
merit-based shortlist of candidates for the appointment to the FWC (and the proposed
Workplace Standards Commission). The Australian Minister for Employment would then
make appointments from the shortlist, where they were satisfied the person was
sufficiently unbiased and credible, for a fixed term (box 8.8). The involvement of multiple
jurisdictions (for example, the process to appoint the Australian Competition and
Consumer Commission Chair) encourages a focus on quality appointees. Although the
appointment model was designed for a larger purpose, it still fits this process as well.
The same principles would apply to any alternative body given the task of determining
eligibility for preferred default status, except that the Australian Treasurer should be the
responsible Minister.
The Commission envisages the panel determining eligibility for preferred default status
would meet only for the period it was required and then cease operations. Its decisions
about listing could provide significant commercial advantages to providers with products
included on the preferred default list, and the Commission considers it should therefore be
established as a statutory decision-making body. Decisions about inclusion on the preferred
default list would be made by the panel and its decisions would be made public. The
Minister would have powers to request the panel to review its decision, to initiate a new
review before the term of the current preferred list expires, or to disband the panel at any
time. However, the Minister would not have direct powers to change the decision of the
panel once it is finalised and announced.
How often should the list of eligible products be updated?
Based on the discussion in the Commission’s issues paper for this inquiry (rather than in
direct response to the detailed model proposed in this chapter), participants expressed
differing views on the frequency with which lists of eligible products should be updated.
The Association of Superannuation Funds of Australia (ASFA) (sub. 24, p. 5) noted the
trade-offs regarding the timing of running the process:

DRAFT REPORT
211
Greater frequency of running the filtering process could improve the flexibility and accuracy of
the filter, as well as increasing competition and the opportunity for new products to qualify.
However, this would need to be weighed against higher compliance and administrative costs,
reduced certainty for funds and employees, and potentially placing an unhelpful emphasis on
short-term performance.
Box 8.8 The Commission’s 2015 Workplace Relations Framework
report
In its 2015 Workplace Relations Framework report, the Commission noted history and
precedent play too big a role in some of the Fair Work Commission’s (FWC’s) key economic
and social functions, particularly award determinations. The Commission also noted the FWC’s
practices tend to give greatest weight to the evidence put by contesting parties, rather than on
better evidence that it has actively sought.
The Commission therefore recommended the Australian Government create a separate
institution — the Workplace Standards Commission — to undertake wage determination, with
most of its members to have professional capabilities in economics, the social sciences and
commerce, with legal experts primarily used to ensure enforceable awards. The Commission
envisaged the Workplace Standards Commission would proactively undertake its own data
collection and systematic high-quality empirical research as the key basis for its award
decisions and wage adjustments.
The Commission recommended the FWC continue to be responsible for its pre-existing
quasi-judicial functions, such as hearing matters relating to unfair dismissals and various
administrative functions, such approval of agreements, right of entry and authorisation of
protected action ballots. The processes for appointing members to the FWC should be reformed
(with these reforms to also apply to the Workplace Standards Commission). The Commission
recommended Australian, State and Territory governments create an expert appointments
panel, which would provide a merit-based shortlist of candidates for the two bodies.
The Australian Minister for Employment would then choose members from the shortlist for a
fixed tenure. Both the panel and the relevant minister would need to be satisfied that a
candidate for appointment had (and was seen to have) an unbiased and credible framework for
reaching conclusions and determinations. Appointments would be made for a period of ten
years, or to the age of 70, whichever comes first, with no reappointments beyond this term.
Source: PC (2015b).
APRA (sub. 33, p. 3) said:
Depending on the nature and frequency of the default allocation process, RSE licensees may be
encouraged to adopt a more short-term, narrow focus with respect to investments and product
features and their associated costs. Superannuation, by its nature, is a product where the long
term financial interests of members should be central to decision making by RSE licensees. …
a regularly repeating competitive default allocation process under which there is a relatively
short period of exclusive or shared default status may discourage [long-term] investments. It
would clearly be undesirable if the default allocation model did not align with, or even
undermined, RSE licensees’ meeting their obligations in relation to achieving adequate
retirement outcomes for members over the long term.

DRAFT REPORT
A number of participants noted the trade-off between the need for stability and the
potential benefits from a competitive process. Frequent updating of the list could also
encourage short-term performance targeting by funds which may be inconsistent with the
long-term interests of members given the nature of the superannuation product.
In weighing up these factors, as discussed in chapter 3, the Commission considers the
process of compiling the preferred default list should be undertaken every four years, based
on a strictly long-term performance focus. That is, inclusion on the list should be based on
performance over the long-term (preferably ten years or more).
The determination of the mandatory minimum standards list should be ongoing (again,
based on long-term performance). APRA would routinely monitor governance standards
and compliance with trustee obligations. Product providers failing to meet trustee
obligations would have their products removed from both the minimum standards list and
the preferred default list when this failure becomes apparent, rather than several years later.
The Commission considers it might be worth reviewing the criteria about 10 years after
their commencement to ensure they are still fit for purpose. However, beyond this the
Commission does not think there is a strong case for regular reassessment.
The key features of the proposed assisted employer choice (with employee protections)
model are summarised in table 8.2.
8.4 Evaluation of the model
This section is an assessment of the costs and benefits of the model against criteria relative
to the baseline (unassisted active employee choice).
Member benefits
As discussed in other chapters, members’ best interests are met where long-term net
returns are maximised and employees are allocated to the products that best suit their needs
(chapter 2).
Employer choice could improve outcomes relative to the unassisted active choice baseline
if employers were better placed than employees to choose a default product. Some
employers will have relevant expertise or the ability to run a corporate tender, while others
might have neither the expertise nor incentive to choose wisely on behalf of their
employees. However, employer choice introduces principal–agent issues between
employers and employees that are not at play under the baseline.
By restricting or influencing choices, filters lead to more employees being placed in better
products, whereas the unassisted active choice baseline relies on employees making this
choice for themselves when they might not be well placed to do so. The model should,

DRAFT REPORT
213
therefore, lead to better outcomes for members than would be achieved under the baseline,
although the improvement in member benefits stems more from the application of the
filters than from employers making choices.
Table 8.2 Summary of key features of assisted employer choice (with
employee protections) model
Main design elements Design features
Dual-list approach Apply both ‘light’ and ‘heavy’ filters to develop a mandatory minimum
standards list and a shorter ‘preferred default’ product list to help employers
reduce search costs, while also driving competition.
Preferred default list would employ stricter criteria and higher performance
benchmark hurdles. It would focus on ‘no frills’ products with higher
suitability for default members.
No cap on the number of eligible products listed but consider mandating a
minimum number for the preferred default list.
Employers maintain flexibility to choose any product from the mandatory
minimum standards list.
Only one minimum standards list and one preferred list applied across the
entire workforce for simplicity.
Assessment criteria Long-term investment return target and risk profile.
Expected ability to achieve target.
Comparison of fees and costs against benchmark.
Governance practices.
Quality of service/advice.
Institutional and governance
arrangements
Panel appointed with relevant expertise and knowledge to assess preferred
default status. Focus on best interests of members.
Determination of minimum standards list to be ongoing.
Preferred default list to be compiled every four years.
Both lists subject to revision at any time where necessary.
Transparency and
accountability
All funds seeking to have products listed as preferred defaults apply with
submissions made publicly available.
Decisions of panel clearly articulated and transparent.
Ongoing enforcement and
reassessment
Ongoing monitoring and enforcement of minimum standards for fund
performance by APRA.
Annual self-reporting by fund trustees backed up by independent auditing
by, or on behalf of, APRA.
Periodic review of the criteria to assess default status (every 10 years).
The use of selection criteria relating to returns, fees and governance also provide incentives
to lift performance in these areas. These incentives are less strong in an unassisted active
choice world as, without the aid of filters, many employees might not have the capacity to
interpret elements of risk and return, understand fees or governance arrangements, and
make investment decisions based on this information.
But it may not be a major difference: funds may pay less attention to benchmarks if there is
no guarantee of a reward through achieving preferred status, as happens under the assisted
employer choice (with employee protections) model.

DRAFT REPORT
A potential benefit (albeit possibly minor) of employer chosen defaults, relative to an
unassisted active employee choice world, is that some employers (for example, those
facing high labour or skills demand) may have an incentive to work with approved default
product providers to tailor services to their workforce, including information and advice.
This would include encouraging employees to consolidate products where possible.
Competition
Administrative filters may at first be perceived as more likely to constrain competition than
the baseline because filtering would somewhat narrow the range of eligible products to
choose from. However, a large number of suppliers is not always necessary to have
effective competition. Further, filtering could generate competition to the extent that funds
compete to be included on the preferred default list of default products by performing well
against the selection criteria, and compete to be the default fund for large employers
(including through competitive processes such as corporate tenders). In that sense, there is
more likely to be competition on aspects that are valuable to members, like net returns and
fees, though there is still some scope for unhealthy or wasteful marketing to both
employers and employees.
The application of administrative filters would be likely to lead to some degree of industry
consolidation which, given the importance of scale (Deloitte 2010) in terms of lowering
costs and achieving higher returns, should lead to better outcomes for members. This
would be less than in some other models but likely to be greater than the baseline.
Integrity
As noted by Kinetic Superannuation (sub. 45, p. 2), a market-based model must be
transparent and credible to ensure that default fund trustees remain accountable for the
outcomes delivered to members. The same principles apply to a model with employer
choice.
The model with employer choice introduces additional principal–agent issues relative to
the unassisted active employee choice baseline. Principal–agent problems may in theory
arise between the employer and employee, the organisation responsible for applying the
filters and the employee and, potentially, other parties (such as industrial parties) and the
employee.
As such, this model — particularly given the implausibility of imposing a regulatory
obligation upon employers to act in their employees’ best interests — might in principle
raise more integrity issues than the baseline. However, in practice, much of the substantive
risk lies in the appointment process for the default preferred list panel and the transparency
and accountability of the decision-making process.

DRAFT REPORT
215
Implementation of effective appointment processes and governance rules relating to the
body made responsible for administering the filters is essential. This model could then
achieve higher integrity than that likely to be observed under the baseline.
Stability
The stability impacts of the assisted employer choice (with employee protections) model
would be minimal. If a product was removed from the preferred default list, it would still
be able to receive default contributions and existing members would be able to remain in it.
If a product was removed from the mandatory minimum standards list, the Commission
considers it should be ineligible to receive default contributions. To minimise disruption,
the Commission considers APRA should oversee a process that would ensure existing
members in the now-ineligible default product were shifted to a different product (unless
they actively decided to stay in the original product as choice members). The Commission
considers this outcome necessary because of the nature of the minimum standards filter:
the implication of not meeting the filter is that a product is no longer suitable for default
members. Such a scenario would be expected to be very infrequent and only a small
proportion of employers or members would be likely affected at any one time.
System-wide costs
Under the assisted employer choice (with employee protections) model, search costs would
be low for defaulting employees (as their choice would effectively be made for them by
their employer), and lower than under the unassisted active choice baseline (chapter 4).
Total costs to employers would be somewhat higher than under the baseline. These could
be quite high for a few employers, especially if they chose to use tender processes or
relatively expensive ways of determining the most suitable products for their employees.
However, compared with the more diverse range of products available to employees under
the baseline, employers would choose from a far reduced number of potential preferred
default products. Further, given there are relatively few employers compared with the
number of employees, it is likely that, all else equal, the overall search costs associated
with default selection would be lower than under the baseline.
The total cost to funds will be higher under the administrative filter model compared with
the baseline, as they would incur the costs associated with participation in the filtering
process and, assuming the filtering process was sufficiently competitive, they might also
choose to market heavily to employers to gain members. Most of the costs incurred would
relate to ongoing compliance with the minimum standards filter. However, the total
operating costs for funds would also potentially rise under the baseline, as they competed
for members.
The costs to government would be higher than the baseline because government would
incur the costs associated with the administrative filters. There would also be the costs

DRAFT REPORT
related to the centralised information service (chapter 3). Against this, if the filters
succeeded in encouraging better product choices, this could reduce future Age Pension
liabilities. This would potentially reduce the inefficiency (sometimes called the
‘deadweight loss’) associated with raising revenue to fund these liabilities (which is an
additional benefit to that of the higher returns obtained directly by those members in more
suitable products following the application of the filters).
Table 8.3 Assisted employer choice (with employee protections)
model: assessment summary against the baseline
Member
benefits
Presence of administrative filtering should result in more employees in better performing
products.
Use of selection criteria relating to long-term net returns, fees and governance should
provide incentives to improve performance.
Greater involvement of employers could enable better tailoring of products to workplaces
in some cases.
Competition Narrows range of eligible products which could potentially reduce competition.
However, can also generate ready comparability and competition as funds compete to
have their products included on the preferred default list, and to be the default fund for
large employers. Disclosure and improved comparability could promote active member
engagement.
Will focus competition on aspects of value to members, like long-term net returns and
fees, though there is still some scope for unhealthy and wasteful marketing to employers.
Could provoke some market consolidation, which could benefit members by leading to
more efficient providers and better member returns.
Integrity The role of employers and the panel determining the preferred default list introduces
principal–agent issues, meaning this model can be seen as raising more integrity issues.
In practice, the main integrity issue would relate to the process for appointing the panel
determining preferred default eligibility. This requires effective appointment processes
and governance and accountability arrangements.
Stability Unlikely to create instability, as the model would not lead to excessive market
concentration, sudden movement of balances or a short-term focus by trustees.
System-wide
costs
Search costs lower for defaulting employees as they effectively have their choice of
product decision made for them.
Total costs to employers increased, particularly if they use relatively expensive ways of
choosing a product (such as via tender). However, their ongoing costs might be reduced
if they subsequently dealt with fewer providers.
Overall search costs likely to be lower than under employee choice as there are fewer
employers relative to the total number of employees.
Cost to funds higher as they incur the cost of participating in the filtering process and
would be likely to incur costs from marketing their products to employers (although these
latter costs may be lower than under the baseline).
Total costs to government may be higher as they incur the implementation and
administration costs of the administrative filter (although if the filter leads to higher
balances, this is likely to reduce future Age Pension liabilities).

DRAFT REPORT
217
Overall evaluation
The Commission concludes that assisted employer choice (with employee protections) is a
workable model which is likely to provide superior outcomes compared with the unassisted
active choice baseline.
Most of the benefits associated with the model stem from use of the filters to guide
employees to better products, rather than through the choice of product being placed with
the employer. Some employers will be well placed to choose products, through corporate
tenders or because they have particular expertise, but others may lack the expertise or
willingness to incur expense but will still receive some assistance from the filtering.
Potential benefits directly relating to employer choice (as distinct from those related to the
filtering) would include the opportunity for tailoring of products to workplaces.
DRAFT FINDING 8.1
Relative to the baseline, assisted employer choice (with employee protections) —
employing both a light filter for mandatory minimum standards and a heavy filter for a
preferred default list — would:
• enhance member benefits by increasing the likelihood of employees being placed
in higher quality products, and reducing the likelihood of them being in a poor
product
• promote healthy competition by presenting product providers with incentives to
perform strongly against the preferred list selection criteria or compete for
corporate tenders and facilitating greater comparability, but there would still be
some scope for unhealthy and wasteful marketing to employers
• increase the potential for agency problems given the involvement of employers,
although the risk to the integrity of the system would lie primarily in the
appointment process for the preferred default list selection panel
• create few stability concerns
• reduce search costs for employees, while increasing search costs for employers
and regulatory costs for funds and government.

DRAFT REPORT
What are the main drivers of costs to employers in selecting default products on behalf
of their employees? Would a shortlist of ‘no frills’ default products make this task easier
for employers? Is there an ideal minimum number of products that should be
nominated on the preferred default list?
Are there other specific criteria in addition to those proposed under the minimum
standards criteria that default products should meet to protect members and help to
achieve better outcomes for them in the long term?
Would a dual-list approach, allowing employers to select a product from one of two
lists, provide them with sufficient flexibility to select tailored default products that best
meet the needs of their employees?
Which types of employers prefer to retain a role in default product selection? To what
extent are default products or corporate fund offerings considered important benefits
offered to prospective employees in competitive labour markets?

PUBLIC CONSULTATION
DRAFT REPORT
219
A Public consultation
In keeping with its standard practice, the Commission has actively encouraged public
participation in this inquiry.
• Following receipt of the terms of reference on 17 February 2016, an advertisement was
placed in a national newspaper and a circular was sent to identified interested parties.
• An issues paper was released on 20 September 2016 to assist those wishing to make a
written submission. A total of 52 submissions were subsequently received (table A.1).
Public submissions are available on the Commission’s website: www.pc.gov.au.
• As detailed in table A.2, consultations were held with government departments and
agencies, academics and superannuation industry participants.
The Commission thanks all parties who have contributed to this inquiry and now seeks
additional input for its final report. The Commission welcomes further submissions to
discuss the substance of the draft report, including responses to the information requests
and draft findings and recommendations.

DRAFT REPORT
Table A.1 Public submissions
Individual or organisation Submission number
Accommodation Association of Australia (AAA) 16
Ai Group 21
AMP 42
Association of Superannuation Funds of Australia (ASFA) 24
Australian Chamber of Commerce and Industry (ACCI) 27
Australian Council of Trade Unions (ACTU) 34
Australian Hotels Association (AHA) 6
Australian Institute of Superannuation Trustees (AIST) 28
Australian Prudential Regulation Authority (APRA) 33
Australian Securities and Investments Commission (ASIC) 41
AustralianSuper 19
Centre for International Finance and Regulation (CIFR) 7
Centre for Market Design (CMD) 18
CHOICE 31
Club Plus Super 32
Colonial First State 25
Community Clubs Victoria (CCV) 9
Corporate Super Association (CSA) 35
Fair Work Commission (FWC) 51
Financial Planning Association of Australia (FPA) 29
Financial Services Council (FSC) 38, 49
First State Super 26
Independent Fund Administrators and Advisers (IFAA), QIEC Super and
Club Super
13
Industry Super Australia (ISA) 40
Institute of Public Affairs (IPA) 17
Kinetic Superannuation 45
Land, Richard 22
Linacre, Andrew 52
LUCRF Super 30
Mair, Peter 1
Mercer 15
Mine Wealth + Wellbeing 46
NESS Super 47
Police Federation of Australia (PFA) 14
PricewaterhouseCoopers (PwC) 12
Restaurant and Catering Australia (RCA) 10
Retail Employees Superannuation Trust (REST) 23
Rice Warner 43
Sweeney, Phillip 2, 3, 5, 8
UniSuper 20
Vision Super 4
Workplace Super Specialists Australia (WSSA) 11

PUBLIC CONSULTATION
DRAFT REPORT
221
Table A.2 Meetings and teleconferences
Individual or organisation
AUSTRALIA
Ai Group
AMP
Association of Superannuation Funds of Australia (ASFA)
Australian Chamber of Commerce and Industry (ACCI)
Australian Communications and Media Authority (ACMA)
Australian Council of Trade Unions (ACTU)
Australian Institute of Superannuation Trustees (AIST)
Australian Prudential Regulation Authority (APRA)
Australian Securities and Investments Commission (ASIC)
Australian Taxation Office (ATO)
AustralianSuper
Chant West
ClearView
Professor Gordon Clark (University of Oxford)
Department of Employment
Department of Finance
Department of Prime Minister and Cabinet (DPMC)
Fair Work Commission (FWC)
Financial Services Council (FSC)
First State Super and StatePlus
Grattan Institute
Industry Super Australia (ISA)
Mercer
Northern Territory Department of Treasury and Finance
Rice Warner
Treasury
UniSuper
Pauline Vamos
CHILE
Professor David Bravo (Pontificia Universidad Católica de Chile)
Superintendent of Pensions
Assistant Professor Felix Villatoro (Universidad Adolfo Ibáñez)
NEW ZEALAND
ANZ
Booster Investment Management
Commission for Financial Capability
Financial Markets Authority
Financial Services Council
Inland Revenue Department
Kiwibank
(continued next page)

DRAFT REPORT
Table A.2 (continued)
Individual or organisation
Ministry of Business, Innovation and Employment
New Zealand Treasury
Retirement Policy and Research Centre (Auckland University)
Simplicity
UNITED KINGDOM
Professor Nicholas Barr (London School of Economics)

SELECT DEFAULT MODELS OVERSEAS
DRAFT REPORT
223
B Select default models overseas
Australia is not alone in grappling with the challenge of allocating retirement savers that do
not make an active choice into default products. While Australia’s unique superannuation
system limits the insight that can be gained from other retirement income systems, there is
nonetheless value in considering how other countries deal with the allocation of default
members. This is reflected in the Commission’s terms of reference, which require the
Commission to consider the strengths and weaknesses of competitive processes used
internationally.
This appendix provides background on retirement income systems and default
arrangements in Chile, New Zealand, the United Kingdom and Sweden. These countries
were identified by the Commission in its preliminary research as those most likely to
provide useful insights for informing the inquiry. They also share some similarities with
the Australian system, including the use of defined contribution schemes in which
employees or employers can choose between funds to allocate their contributions.
The sections that follow provide context on each country’s overall retirement income
system, the mechanisms for allocating savers to default products, and any relevant
retirement and default allocation system outcomes.
B.1 Chile
The Chilean retirement income system was restructured in 1981 based around three pillars:
a publicly-financed poverty prevention (or solidarity) pillar, a mandatory contribution
pillar and a voluntary savings pillar. For the mandatory contribution pillar — which is the
focus of this appendix — the previous ‘pay-as-you-go’ scheme in which active employees
financed the benefits of those who had retired was replaced by an individual accumulation
scheme.
In 2006, a study revealed that a significant portion of the population lacked sufficient funds
to support themselves during old age, primarily due to a low contribution rate. This was
largely explained by labour market informality and periods of self-employment (given
self-employed people did not have to make contributions at this time). In the case of
Chilean women, there were typically long periods of time during which they did not
contribute (Superintendence of Pensions 2010).

DRAFT REPORT
The system was therefore significantly reformed in 2008 to increase participation and
reduce administration fees in the mandatory contribution pillar and to bolster the solidarity
pillar (Shelton 2012).
The mandatory contribution pillar
The core of the mandatory pillar is a mandatory contribution rate of 10 per cent (subject to
a ceiling) on monthly earnings. Employers must deduct an employee’s contribution from
their pay and forward it to the employee’s chosen pension fund (Administradoras
de Fondos de Pensiones, or AFP). Employees may choose any AFP and switch at any time,
except for new employees (for whom arrangements are discussed below) (Shelton 2012,
pp. 2–3).
In addition to the 10 per cent mandatory contribution, employers deduct additional
contributions to cover administration fees and insurance premiums (death and disability),
which each represent about 1.4 per cent of earnings (Chant West 2016, p. 2). Employees
therefore need to contribute at least 12.8 per cent of their earnings towards their mandatory
pension.
AFPs must purchase death and disability insurance for their members from a life insurance
company through a public tender process (Superintendence of Pensions 2010, p. 95).
The nature of AFPs
AFPs are single-purpose institutions that charge fees as a percentage of an employee’s
salary. They must collect pension contributions, take action in case of payments being in
arrears, invest resources, manage pension processes and, in some cases, disburse pensions.
They are also generally responsible for all processes that are requested by members
involving their own funds (Superintendence of Pensions 2010, pp. 17–18).
AFPs are subject to meeting minimum investment performance provisions which require
each of their investment returns for the previous 36 months to be comparable with other
AFPs. Otherwise, the difference in performance must be made up through transfers from a
cash reserve equivalent to 1 per cent of each AFP’s value or, should this be inadequate, by
a government guarantee to make up the difference that, if required, would trigger the
liquidation of the AFP (Superintendence of Pensions 2010, p. 60). This means AFPs
typically invest in a very similar manner to ensure they are not caught out by these
provisions (Vásquez 2004).
Election of the boards of directors of AFPs is regulated. AFP boards must include at least
five directors, two of whom must be independent. An independent director shall be defined
as one having no ties with the AFP, the other companies in the business group it belongs
to, its controller, or with the main executives of any of these, any of which could

DRAFT REPORT
225
potentially generate a conflict of interest or otherwise threaten the director’s independent
judgment (Superintendence of Pensions 2010, p. 58).
If an AFP is dissolved for any reason, the Superintendence of Pensions is responsible for
liquidating the company and managing its pension funds. The liquidation of an AFP only
affects the company and not the pension funds it manages. The individual accounts are
transferred to a new AFP chosen by each member (Superintendence of Pensions 2010,
p. 98).
Fees
AFPs are free to set their own administration fees, but must charge the same fee (that is,
the same percentage of salary/contributions) for all members. Only contributors pay this
fee, and these represent about half of all members (Chant West 2016, p. 2).
There are limits on the investment fees AFPs can pay to external fund managers (which
vary according to the asset class, and whether management is passive or active), and the
fees paid must be disclosed to regulators on a monthly basis. The Superintendence of
Pensions subsequently publishes these data on its website (although not broken down by
types of investment or asset classes) (Chant West 2016, p. 10).
Employee contributions are tax deferred in that contributions and earnings are tax exempt
until withdrawal. Retirees pay regular income tax on their pension income (Chant
West 2016, p. 2).
Fund options
Since 2002, AFPs have been able to offer up to five investment funds, named A to E, with
differing proportions of investment in equities. The assumption is that more equities equals
greater risk. All funds must offer funds B to E while fund A (the riskiest) is voluntary.
In practice, all AFPs offer all five funds. Members may invest in two of them, although
older members are restricted from investing in the riskiest funds (Superintendence of
Pensions 2010, pp. 34–35). All funds are heavily regulated regarding the proportion of
various investments they are allowed to make. To simplify the system, fee structures are
uniform across investment options (Superintendence of Pensions 2010, p. 38).
Where members do not make a choice of fund, they are, by law, allocated by their AFP to
funds based on their age (table B.1). The shift from one fund to another is also gradual,
with 20 per cent of assets shifted at the time of the change, and a further 20 per cent
transferred each year until the transfer is complete (Superintendence of Pensions 2010,
p. 35). Fund C had the largest share of total assets under management in 2016 (figure B.1).

DRAFT REPORT
Table B.1 Default assignment by age where no investment fund option
chosen, Chile
Men and women up to
age 35
Men aged between 36
and 55
Women aged between
36 and 50
Men 56 and over
Women 51 and over
Pensioners
Fund A – Most Risky
Fund B – Risky
Fund C – Intermediate
Fund D – Conservative
Fund E – Most
Conservative
Source: Superintendence of Pensions (2010, p. 35).
Figure B.1 Assets under management by investment option ($USbil),
Chile
Data sources: Chant West (2014b, p. 5, 2016, p. 4).
The tender process for new employees
Rationale for tender
The 2008 reforms included the introduction of a tender system, enabling AFPs to bid for
new entrants to the workforce for a two-year period. The process was designed to increase
competition between AFPs and therefore reduce fees. Prior to the introduction of the tender
system, all new employees had to choose an AFP, leading to concern about high costs
associated with marketing by AFPs. In particular, concern focused on the use of large
0
10
20
30
40
50
60
70
A B C D E
2014 2016

DRAFT REPORT
227
commission-driven sales forces to encourage people into funds (driving up administration
costs), and about arguably questionable incentives such as electrical appliances being
offered to members.
The tender also sought to reduce barriers to entry for new AFPs who could now, if
successful in the tender, be guaranteed new customers without incurring high marketing
costs (Shelton 2012, p. 8). All new entrants to the workforce, including the self-employed,
are required to join the AFP successful in the tender.
The focus on new employees is based on the relatively large influence of fees on
employees with low balances. It is assumed fees are more important than investment
performance to people newly entering the system. When initially implementing the tender,
it was also considered the pool of new workers over two years would be of sufficient size
to meet the minimum efficient scale requirements for a new AFP (Superintendence of
Pensions 2010, p. 166).
The tender process
Under the tender process, the AFP with the lowest administration fee (which must be less
than the fee in effect at the time of the bidding process) becomes the new default provider.
New entrants to the workforce are obliged to join that AFP and stay with it for two years
(except under certain circumstances, such as the fees not being the lowest available, or not
being satisfactory to make up for a higher yield from another AFP). The successful default
AFP must maintain fees at the same level for the two years of the tender period, beyond
which it is assumed competition between AFPs will ensure fees remain competitive
(Superintendence of Pensions 2010, pp. 166–167). Details of losing tender bids are
published, including the name of the tenderer and the fee level tendered.
The AFP successful in the tender must offer the same price to all members (old and new).
The fee must be a fixed percentage of salary (that is, contributions not balances). It covers
administration costs, distribution costs and the cost of the internal investment team.
It excludes the investment fees charged by external fund managers, which amount to about
0.25 per cent per year as a proportion of total assets (Chant West 2016, p. 1).
Tender outcomes
The first tender in 2010 was won by a new AFP, Modelo, with a bid of 1.14 per cent of
salary. It subsequently won the second tender with a bid of 0.77 per cent. PlanVital won
the third tender in 2014 with a bid of 0.47 per cent. PlanVital also won the fourth tender in
2016 with a bid of 0.41 per cent. PlanVital was the only participant in this tender, and was
successful as it met the requirement to reduce its fee below the level determined in the
previous tender (Superintendence of Pensions 2016b).

DRAFT REPORT
When PlanVital won the third tender, its administration fee was reduced from 2.36 per cent
of salary to 0.47 per cent. Chant West has questioned the long-term sustainability of this
fee, noting that profits have been very low for PlanVital (Chant West 2016, p. 9). That
said, PlanVital was able to reduce its fee in the most recent tender.
The aim of increasing competition between AFPs was partially achieved initially when a
new AFP won the first tender. However the number of AFPs has only increased from five,
before the tender was introduced, to six. The number of AFPs was 12 in 1981, before
increasing to 21 by 1994, and subsequently declining to 5 by 2008 due to mergers and
closures (Shelton 2012, p. 8).
The two AFPs to have successfully won the tender, Modelo and PlanVital, are much
smaller than the other four AFPs (roughly one-tenth of the size of the largest, based on
assets under management). This suggests established funds with pre-existing large pools of
relatively high-margin clients do not find the default tender attractive, and prefer to operate
solely in the choice sector.
PlanVital, the most recent winning tenderer, has had inferior investment performance
relative to other AFPs over the past three years (Superintendence of Pensions 2016a).
However, this is not to say that new default members are worse off with PlanVital
(particularly as this group would typically have low balances, increasing the relative
importance of fees to them). They may achieve higher net returns with PlanVital after all
fees (including administration fees) are accounted for, at least in the short to medium term.
B.2 New Zealand
KiwiSaver — an automatic-enrolment, voluntary retirement savings scheme — was
introduced in July 2007 with the objectives of encouraging saving and increasing living
standards in retirement. The introduction of KiwiSaver followed concerns about saving
rates and particularly about the potential for middle class New Zealanders to have large
reductions in their living standards post-retirement. Cabinet papers acknowledged the
potential for this to merely subsidise existing saving (St John, Littlewood and Dale 2014,
p. 10).
KiwiSaver — which sits alongside New Zealand’s universal old-age pension scheme
(known as New Zealand Superannuation) — was set up as an automatic-enrolment scheme
for new employees, with employees given eight weeks to opt out. Anyone not opting out in
these eight weeks is required to stay in the scheme for at least 12 months.
There have been a number of revisions to KiwiSaver since its introduction, particularly due
to concerns about the cost of the scheme and doubts about its impact on national savings
rates (box B.1).

DRAFT REPORT
229
Box B.1 KiwiSaver has been subject to a number of changes since its
introduction
Initial arrangements for KiwiSaver included a $NZ1000 kick start payment from the New
Zealand Government and a subsidy of $NZ40 to cover membership fees. Workers initially
contributed either 4 or 8 per cent of their gross income. Matching contributions from employers
were tax-exempt. KiwiSaver also included a first home-owner deposit subsidy that remains a
feature of KiwiSaver.
KiwiSaver II was announced in May 2007, before the initial scheme had begun, with
government matching contributions of about $NZ1040 a year, and partly offsetting employer tax
credits for employer contributions. Compulsory matching employer contributions were
introduced, starting at 1 per cent from 1 April 2008, and then rising by a further 1 per cent each
year to a peak of 4 per cent from 1 April 2011.
KiwiSaver III was announced in 2008. Effective from April 2009, the $NZ40 fee subsidy was
dropped and minimum employee contributions were lowered from 4 to 2 per cent. Employer
contributions were capped at 2 per cent and the employer tax credit removed.
KiwiSaver IV was announced in 2011. From 1 July 2012, employer contributions became
subject to tax (the ‘employer superannuation contributions tax’). The maximum tax credit was
halved to $NZ521.43 (while the contribution required to qualify stayed the same). The minimum
employee and employer contributions increased from 2 to 3 per cent from 1 April 2013.
The $NZ1000 kick start payment was discontinued from 21 May 2015.
The Commission for Financial Capability recently recommended that minimum employee and
employer contributions increase to 4 per cent by 2021.
Sources: CFFC (2016); Littlewood (2012); St John, Littlewood and Dale (2014, pp. 10–12).
KiwiSaver as it currently operates
Employees can choose to contribute either 3, 4 or 8 per cent of their gross pay.
Self-employed or unemployed people can tell their provider how much they wish to
contribute and make payments directly to the provider (Inland Revenue 2016b). KiwiSaver
savings are normally locked in until a member turns 65 or, if they joined after 60, five
years after joining.32 Although changes to future old-age pension eligibility have been
announced recently by the NZ Government, these will not affect arrangements for
KiwiSaver (Joyce 2017).
Contributions are deducted from the member’s pay and sent by the employer to Inland
Revenue for distribution to the chosen provider. Inland Revenue’s role as a clearing house
for contributions is widely considered to be an important feature of KiwiSaver (box B.2).
32 Early withdrawals are possible in some circumstances, like buying a first home, significant financial
hardship or a serious illness (Inland Revenue 2014).

DRAFT REPORT
Box B.2 Inland Revenue’s role as a clearing house
Inland Revenue uses New Zealand’s pre-existing PAYE income tax system for its clearing
house role. Employers deduct employee contributions, make compulsory employer
contributions and remit these directly to Inland Revenue. Inland Revenue transfers KiwiSaver
contributions to each employee’s relevant KiwiSaver scheme provider (Inland Revenue 2017).
The clearing house is widely considered to be an important and successful part of KiwiSaver.
The Commission consulted with KiwiSaver funds, New Zealand government organisations and
academics, most of whom highlighted what they consider to be the clearing house’s benefits:
• ensuring each individual has only one KiwiSaver account
• low administration costs for employers and funds
• the ease with which Inland Revenue can monitor employers’ compliance with mandatory
contributions
• its simplicity for funds facilitates market entry and competition between providers.
One drawback noted by some parties is that Inland Revenue does not have up-to-date contact
details for all employees, making it harder for providers to communicate with their members.
(However, this issue would likely still exist under a different system.)
Employers must make compulsory contributions equivalent to 3 per cent of gross pay
(regardless of the contribution rate of the employee). They can voluntarily make further
contributions above this level. Employers need not make contributions to KiwiSaver if
they are contributing to another eligible superannuation scheme, if the staff member is
under 18 or over 65 (or if they joined aged over 60 and have been a member for 5 years) or
if the staff member is not contributing (for example if they are opted out on a contributions
holiday or leave without pay) (Inland Revenue 2016d).
Auto-enrolled members are allocated to a default KiwiSaver product by Inland Revenue on
a sequential basis. Their initial allocation is considered provisional and confirmed as final
where the member fails to make a specific choice. If an auto-enrolled member chooses a
provider within three months of being enrolled, their contributions will go to that provider
rather than the provider they were provisionally allocated to.
Members can change scheme providers anytime by applying directly to the provider of the
scheme they want to join. The new provider will then arrange for their savings to be
transferred from the old scheme to the new one and notify the member. A member can
only be in one scheme at any given time (Inland Revenue 2016a). If a KiwiSaver provider
closes down, members may transfer to a new scheme or, if they do not do this, they will be
allocated by Inland Revenue to a default scheme.
Anyone who has been a member of KiwiSaver for 12 months or more can take a
contributions holiday of between three months and five years. There is no limit to the
number of times a contributions holiday can be taken.

DRAFT REPORT
231
As at 31 March 2016 there were 2.6 million KiwiSaver members, with about
$NZ33.8 billion in scheme assets (FMA 2016, p. 13). KiwiSaver providers are regulated
by the Financial Markets Authority (Inland Revenue 2016c). At 31 March 2016 there were
24 retail funds and 8 restricted funds. Ten schemes represented nearly 87 per cent of
members (FMA 2016, p. 5).
Investment strategy
Default members must be allocated to a conservative investment option. Investment
exposure to property and shares is low. Between 15 and 25 per cent of the portfolio must
be in growth assets. Given the need to stay within these parameters, a ‘safety margin’ is
generally maintained. Ministerial approval is required for any moves outside these limits.
The default fund in KiwiSaver has been described as a ‘transitional parking space where
members resided temporarily before they made an active choice’ (MBIE 2012, p. 6).
A significant proportion of members do move away from their default fund. For example,
about 34 per cent of members who were initially in a default fund in 2008 had transferred
to a new provider by 2012 and another 5 per cent had switched investment options within
the same default provider (MBIE 2012, p. 34).
However, about 56 per cent were still with their initial default provider in 2012
(MBIE 2012, p. 36).
Staying in a conservative investment product could be very costly to members in the
long term. MacDonald, Bianchi and Drew (2014, p. 31) noted that the conservative
approach to default investing meant ‘KiwiSaver members are inadvertently exposing
themselves to the hidden risk of failing to reach a retirement target that will enable them to
enjoy standards of living in retirement similar to those in pre-retirement’. They also found
that increased investment in equities would enable KiwiSaver defaulters to increase the
upside potential of their investments without significantly increasing risk.
Fees
Non-default KiwiSaver providers are relatively free to set their own fees, with the
KiwiSaver Act stating that fees relating to KiwiSaver must not be ‘unreasonable’.
As is generally the case with pension plans involving multiple providers, making fee
comparisons under KiwiSaver is difficult. For the 21 KiwiSaver providers for which
published investment performance data were available in 2014, there were more than
165 investment options, ranging from investment solely in cash, to solely in shares. Fees
would be expected to vary widely across these options (St John, Littlewood and Dale 2014,
p. 18).
A 2015 NZ Treasury paper found that KiwiSaver’s operating costs were about the median
for global pension systems, noting that countries with defined contribution schemes and

DRAFT REPORT
large numbers of funds are likely to have higher costs than countries with few funds. The
paper also noted that KiwiSaver fees tend to be lower than superannuation fees in Australia
(Heuser et al. 2015, pp. 45–47). The Grattan Institute has also noted this (Minifie, Cameron
and Savage 2014, pp. 25–26). However, in looking at these comparisons, it is relevant that
many KiwiSaver schemes employ low cost investment strategies. The Grattan Institute noted
‘the [NZ] default asset allocation has been criticised for over-emphasising low-risk assets
and giving growth assets too little weight’ (Minifie, Cameron and Savage 2014, p. 25).
The KiwiSaver default tender process
To determine eligible default KiwiSaver providers, the NZ Government commenced a
tender process in 2006. The government appointed an expert panel to review the
submissions received and create a shortlist for interview. The panel then made
recommendations to the minister.
A number of factors were included in the selection criteria, including:
• security and organisational credibility
• organisational capability
• the proposed design of the provider’s default KiwiSaver scheme
• administrative capability
• competitive fee levels
• investment capacity and capability (Cullen and Dalziel 2006).
It is widely believed that the expert panel recommended appointing either four providers or
all of those satisfying the minimum criteria, although six funds were ultimately appointed
(Littlewood 2012, p. 11). It has been suggested that this followed direct approaches to the
NZ Government from tender participants not initially listed in the preferred four. This
highlights the possibility (and risk) of ministerial intervention in tender processes, and
particularly multi-criteria tenders where the weighting given to the various considerations
(and therefore the basis for the final decision) might be unclear. This risk needs to be a
major consideration when designing a tender model.
The six funds appointed in the 2006 tender process were:
• AMP Services (NZ) Limited
• ASB Group Investments Limited
• ING (NZ) Limited
• Mercer Human Resource Consulting Limited
• National Mutual Corporate Superannuation Services Limited (trading as AXA New
Zealand)
• Tower Employee Benefits Limited (Cullen and Dalziel 2006).

DRAFT REPORT
233
A number of criticisms have been made of the 2006 tender process. For example,
Littlewood (2012, pp. 11–12) suggested the process:
• was not transparent, with reasons for government decisions not being adequately
disclosed
• resulted in decisions being made on a ‘least risk to government basis’, rather than being
based on the best interests of members
• resulted in too many default providers being appointed given the size of the market.
Concern has also been expressed about the competitive advantage attained by funds provided
KiwiSaver default status (Littlewood 2012, pp. 10–11). This advantage is even greater if
people assume the government has done ‘due diligence’ on the funds, or if they erroneously
believe they are government guaranteed. Cabinet papers prepared ahead of the 2014 tender
process noted that, at that time, about 37 per cent of KiwiSaver members had entered
through default arrangements, and 23 per cent of KiwiSaver members remained as default
members (having never changed their provider or investment option) (NZCO 2013, p. 1).
The first tender covered the period 2007 to 2014.
A similar process was carried out for the second tender, covering a seven-year period from
1 July 2014. A panel of evaluators with relevant industry expertise evaluated proposals
according to ‘technical’ criteria (assigned a 70 per cent weighting) such as organisational
capability, member education and investment capability; and the providers’ pricing levels
(assigned a 30 per cent weighting) (MBIE 2013, p. 3).
Cabinet documents relating to the second tender concluded that the initial selection criteria
had served default members well in ensuring well-governed and competitively priced
funds. The major item of consideration between the first and second tender related to the
appropriateness of the investment strategy, with consideration being given to shifting from
a conservative to a balanced or life-cycle approach.
It was determined that KiwiSaver default funds should continue to be designed as ‘parking
spaces’ before people shifted to more suitable funds rather than being intended as the most
appropriate long-term investment vehicle for individuals. As such, it was determined the
conservative approach to investment should continue (NZCO 2013, p. 7).
In view of the importance of educating people to shift from the default fund, one change
was made to the selection criteria between the first and second tender processes. This was
the introduction of a requirement that prospective default providers demonstrate how they
will offer investor education to encourage default members to make more active choices
about investment options (box B.3). Otherwise, the selection criteria employed were the
same as those for the initial tender (English and Foss 2013).

DRAFT REPORT
Box B.3 The push for financial literacy
In keeping with the priority of ensuring KiwiSaver members choose the most suitable products
for their needs, the Financial Markets Authority has a three year plan to increase financial
literacy. Default providers must address the financial literacy of their default members. They
must report on their commitments to encourage members to choose an investment within their
scheme, and the number of members who do this.
The Financial Markets Authority highlighted the approach of Grosvenor (now known as Booster)
to contacting default members, as this is seen as the most successful approach (with a 22 per
cent success rate). The approach consists of:
• data clearing and checking to ensure all staff can call members
• three follow up calls if the first call is unsuccessful
• ascertaining the extent of the member’s knowledge
• tailoring conversations
• taking members through questions about their risk profile, and switching them to another
option while they are still on the phone.
Source: FMA (2016, p. 12).
The cabinet papers also show that consideration was given to the treatment of default
providers and their members (from the first tender) who were unsuccessful in the second
tender. It was considered that default members should continue to benefit from the
protections provided by the instruments of appointment that followed the tender, meaning
they should be transferred to a new default fund. However, as some members might wish
to choose to stay in the original fund, members to be transferred should be given an opt-in
election form providing them with this opportunity (NZCO 2013, p. 14).
Consideration was also given to the seven-year appointment process, which was seen as
providing a reasonable period of incumbency, while not allowing for complacency on the
part of providers (NZCO 2013, p. 15).
The 2014 KiwiSaver tender process broadly involved two steps: a shortlisting of suitable
KiwiSaver default funds, followed by a fee negotiation stage designed to place downward
pressure on KiwiSaver default fees. The Commission understands this stage led to fee
reductions from some providers.
Following that tender process, there are currently nine default providers. Although there
have been name changes, mergers and acquisitions, none of the initial default providers
lost their default status. The current providers are:
• AMP Services (NZ) Limited
• ANZ New Zealand Investments Limited
• ASB Group Investments Limited
• Booster Investment Management Limited

DRAFT REPORT
235
• BNZ Investment Services Limited
• Fisher Funds Management Limited
• Kiwi Wealth Limited
• Mercer (NZ) Limited
• Westpac New Zealand Limited (Inland Revenue 2016a).
In the year ending 31 March 2016, the share of members contained in the nine default
funds fell from 18 to 17 per cent, while the share of total assets in default funds fell from
14.3 to 12.5 per cent (FMA 2016, p. 13).
2014 evaluation of KiwiSaver
Inland Revenue (in conjunction with other government organisations) undertook an
evaluation of KiwiSaver in 2014, based on the seven-year period between 2007 and 2014.
It was noted that this was a relatively short time in which to evaluate a program designed
to have a long-term impact, and that the impact of the Global Financial Crisis was likely to
have influenced program outcomes.
Key findings of the evaluation included that:
• both employees and employers had been broadly supportive of KiwiSaver, and
employers did not consider it to be particularly burdensome
• KiwiSaver knowledge is associated with income and asset accumulation of assets. The
more assets individuals have, the higher their likely knowledge about KiwiSaver
• KiwiSaver contributions appear to be increasingly made at the expense of paying off
mortgage or other debt (Inland Revenue 2015).
B.3 United Kingdom
In 2008 the UK Government, concerned about inadequate retirement incomes, introduced
major reforms to its retirement system. The centrepiece of the reforms was a requirement
for employers to automatically enrol their workers in an occupational pension scheme,
with workers able to subsequently opt out of the system. The new system is still being
phased in, and is expected to be fully in place in 2018. While this limits the scope to draw
lessons from the reforms, there is nonetheless value in considering the system’s policy
design and early outcomes.
The UK retirement system
The United Kingdom’s retirement system has three parts: a means-tested state pension,
individual retirement savings plans, and occupational pension schemes.

DRAFT REPORT
The focus of the UK Government’s 2008 reforms was on the occupational pension system.
Employers must automatically enrol workers aged over 22 and earning more than £10 000
into an occupational scheme.33 Employers must contribute an amount equivalent to at least
3 per cent of a worker’s after-tax earnings into the scheme, with workers contributing
4 per cent and the Government another 1 per cent. Contributions are largely exempt from
tax.
Employers choose which pension scheme (or schemes) they use to enrol their workers into.
When workers change jobs they can consolidate their pension accounts into their new
employer’s pension scheme. Employers are not required to make contributions into
pension schemes that they do not use to meet their enrolment obligations.
Most occupational pension schemes are provided by large insurance companies, and
include a mix of defined contribution and defined benefit schemes (although the share of
the latter is declining).34 The market is relatively concentrated, with four large providers
holding the majority of schemes, assets and members (OFT 2013).
The default pension scheme
When the UK Government introduced its reforms, it was concerned that some small
businesses may be unable to secure the services of a private pension provider. To remove
this risk the government created a default pension scheme — the National Employment
Savings Trust (NEST). NEST has a public service obligation to admit any worker
automatically enrolled by their employer, even if the cost of administering the member’s
account is greater than the revenue derived from member charges (UK DWP 2016).
NEST is a non-departmental public body, operating at arm’s length from government with
a duty to act in the interests of scheme members. The trustee (NEST Corporation)35 is the
main governing body. NEST and NEST Corporation are regulated by The Pensions
Regulator, which is responsible for ensuring that pension schemes are adequately funded,
run in the best interests of members, and that employers meet their enrolment and
contribution obligations.
To pay for the costs of administering and managing NEST, members pay a 0.3 per cent
annual charge on total assets under management. A temporary 1.8 per cent charge on each
contribution is also levied to help repay a loan from the UK Government to establish the
scheme.
33 These schemes are not required to provide insurance.
34 In certain circumstances individuals can transfer from defined benefit schemes to defined contribution
schemes, the latter of which offer individuals greater flexibility in accessing their savings.
35 The Secretary of State for Work and Pensions named the initial NEST trustees, along with members of
advisory panels representing workers and employers. NEST is now largely self-governing though
accountable to parliament through the UK Department for Work and Pensions.

DRAFT REPORT
237
Employers are not required to use NEST to meet their enrolment obligations. However,
any scheme used instead must meet a number of regulatory requirements (table B.2). These
requirements differ depending on the employee; those already in an occupational pension
can be enrolled in a scheme registered as a ‘qualifying scheme’, while those without an
occupational pension must be enrolled in a scheme registered as an ‘automatic enrolment’
scheme. Competition with the private sector is intended to ensure NEST provides value for
money.
Workers can continue contributing to their NEST account if they stop working or change
jobs. To help members consolidate their pension accounts, from April 2017 it is intended
that all members will be able to transfer pension accounts into or out of NEST
(NEST 2015).
NEST’s investment and administration services
NEST provides a series of retirement date funds to deliver its default investment strategy.
At any time there are up to 50 retirement date funds (the ‘NEST Retirement Date Funds’).
Each fund has an asset allocation consistent with the expected amount of risk that is
suitable for that stage in a scheme member’s savings career.36
NEST’s investment strategy is relatively conservative. In contrast to many other retirement
date funds, NEST invests the savings of workers in their twenties in low-risk investments. It
then invests in higher-risk investments in the middle years, before transitioning back to
low-risk investments as retirement age approaches. NEST uses this strategy to reduce the
risk that early losses in a person’s savings lifetime could discourage them from saving
(Sass 2014).
NEST contracts out its investment and administration services to the private sector.
• NEST Corporation delegates much of its investment strategy to an investment
committee. The committee assembles a set of funds in house from low-cost funds
offered by major investment houses. There are currently 14 fund managers appointed to
invest on behalf of NEST members.
• Administration services are provided by Tata Consultancy Services. NEST recently
extended Tata’s contract to provide administration services until 2023. State Street
Corporation is the fund administration partner.
It appears from NEST’s procurement ePortal that it uses both tenders and auctions to
procure services. Funds and investment managers are selected using a competitive tender
process and the advice of independent external investment advisers (NEST 2016, p. 38).
36 Scheme members may choose one of the following alternatives to the default investment strategy: the
Higher Risk Fund, Lower Growth Fund, Ethical Fund, Sharia Fund and Pre-retirement Fund.

DRAFT REPORT
Table B.2 Regulatory requirements for UK pension productsa
Scheme type Can be used for Requirements
Qualifying scheme Employees already in
the scheme (to be
registered as a
qualifying scheme)
• An occupational or personal pension scheme
• Tax registered
• Meets minimum ‘quality requirements’ for defined
contribution schemes
– Employer contributions at least 3 per cent
– Total contributions at least 8 per cent
• Personal pension schemes must be regulated by the
Financial Conduct Authority, and have their operations
carried out in the United Kingdom by a person
authorised or exempt under the Financial Services and
Markets Act 2000 (UK)
Automatic enrolment
scheme
Employees not in a
scheme
• The requirements above
• No barriers to enrolment into the scheme (such as age
restrictions)
• No requirement for members to express a choice or
provide information to retain membership
• Main administration in the United Kingdom or another
European Economic Area country
• No consultancy payments to third parties under an
agreement with the employer
a The minimum requirements to certify an occupational pension scheme differ depending on the type of
scheme. The table includes some of the main requirements for defined contribution schemes. If employers do
not meet the criteria, they can still have their scheme certified if they meet various other requirements that relate
to minimum contribution rates.
Sources: The Pensions Act 2008 (UK); TPR (2016).
Impacts of the 2008 reforms
Since the UK introduced its reforms, the number of people with occupational pension
accounts has grown considerably, from about 27 million in 2011, to over 30 million in
2014 (PPI 2016). By May 2016 nearly 6.3 million employees had been automatically
enrolled into an occupational pension (UK DWP 2016).
Fewer people are opting out of their pension schemes than the government expected.
Overall opt-out rates are about 9 per cent, although rates vary across different groups.
Some research suggests that women, older workers and part-time workers are more likely
to opt out (PPI 2015).
NEST has a large share of the market. In 2016 it had over 3.2 million members and more
than 125 000 participating employers. By contrast, NEST’s two major competitors,
NOW: Pensions and The People’s Pension, had about 0.9 million and 2 million members
respectively (Fernyhough 2016). NEST’s large market share is likely to be partly due to it

DRAFT REPORT
239
being the only scheme required to accept all employers, regardless of whether it is
profitable to do so.
It is claimed that NEST has influenced the automatic enrolment marketplace by driving up
standards and best practice (UK DWP 2016). The UK Government is now considering
whether to allow individuals to opt into NEST (that is, employees would not have to rely
on their employer using NEST to meet their enrolment obligations), and whether to reduce
the earnings threshold for automatic enrolment (UK DWP 2016).
B.4 Sweden
The Swedish retirement income system
In 1999 Sweden reformed its national retirement income system, replacing its generous
universal pension and supplementary earnings-related pension with a new three-tier
system.37
• The main source of retirement income today is a notional defined contribution scheme,
called the income pension. The scheme is notional in the sense that contributions are
notionally accumulated according to a rule based on earning trends, rather than actual
returns. As such it shares some similarities with defined benefit schemes. The income
pension is funded by contributions worth about 16 per cent of salary and other taxable
benefits. Retirement payments (an annuity) are financed out of current members’
contributions, and are based on the life expectancy of the recipient.
• The income pension is supplemented by the premium pension (mandatory defined
contribution individual accounts).38 Individuals can select which provider or providers
invest their contributions (up to five funds can be chosen), or otherwise be allocated to
a default fund. The purpose of the premium pension is to provide individuals with some
control over their investments, and the scope to increase returns and diversification
(SIFA 2015). The pension is funded by contributions worth about 2.5 per cent of salary
and other taxable benefits. At retirement, pension assets are converted into an annuity.
• Poverty relief is provided for low-income individuals through the guarantee pension.
The guarantee pension is funded out of general taxation, and is unaffected by returns in
the premium pension system — it only accounts for the income pension.
Administration of the national pension system is centralised within the Swedish Pensions
Agency (SPA) (box B.4).
37 The current national retirement system operates alongside occupational pensions (stipulated in some
collective agreements, covering over 90 per cent of the workforce), and voluntary private pensions.
38 There are no requirements to provide insurance.

DRAFT REPORT
Box B.4 The Swedish Pensions Agency
The Swedish Pensions Agency (SPA) administers the national pension system. It plays the role
of a central clearing house, distributing the income pension and premium pension contributions
it receives from the National Tax Authority (which collects contributions from employers) to
relevant funds. The central clearing house was established to minimise administration costs
(Hagen 2013).
The SPA deposits contributions once a year, after complete wage records for a calendar year
are available from the tax authorities. The SPA keeps all records of individual accounts, and
has one aggregate account for each fund. Funds have no information on individual participants.
All individual transactions for a given fund are aggregated at the end of each day and
transmitted as a net aggregate to the relevant fund. The SPA also administers benefit payments
(annuities).
The SPA is headed by a board appointed by the Swedish Government. It is financed by pension
savers through a fee on balances. In the absence of a direct relationship between savers and
funds, the SPA aggregates member transactions and negotiates fees for choice products
(Minifie, Cameron and Savage 2014).
The premium pension system
Individuals face a significant amount of choice when selecting a premium pension — there
are several government-managed funds and over 850 registered private products operating
in the market.39 To help individuals choose, the SPA provides detailed information on
funds, including time series data on returns, risk and fees. It also provides a tool that
individuals can use to simulate future returns on the basis of different input variables. The
Swedish Investment Fund Association also provides online tools to facilitate fund
comparisons and evaluations.
The default fund
When the premium pension was launched, the government established a default fund for
individuals who did not choose a fund. Workers are only defaulted once when they join the
workforce.
Initially, reformers had suggested that the default should be a low-risk fund. However, policy
makers were concerned that such a strategy would mean that low-income earners, considered
more likely to be in the default fund, would earn lower returns than high-income earners
(Hagen 2013). As such, the goal of the current default fund, AP7, is to achieve a rate of
return that exceeds the average return of private funds in the premium pension system.
39 Funds must sign a contract with the SPA that specifies reporting requirements and the fee structure. The
total fee in the premium pension consists of two parts: a money-management fee and a fixed
administrative fee. In 2015, fees in the premium pension system were capped at 0.62 per cent of assets for
balanced funds and 0.42 per cent for fixed income funds.

DRAFT REPORT
241
AP7’s default investment is a life-cycle product, which automatically adjusts the risk
composition of an individual’s portfolio with their age. Until an individual turns 55, all
contributions are invested in equities (with a mix of active and passive management), to
allow for a high rate of return. As retirement draws nearer, the risk level is automatically
adjusted downwards by letting the share of assets invested in fixed income securities rise.
AP7’s investments are managed either in-house, or by external asset management firms.
External managers are contracted to provide either passive management of global equities,
pure active management, or private equity fund management.
AP7 also competes in the choice market, offering three other investment portfolios that
individuals can choose from. These funds have different mixes of equity and fixed income
(an aggressive fund, a balanced fund and a conservative fund). Individuals can also opt to
have their entire investment in the equity fund or the fixed income fund.
Impact on competition
When the premium pension was launched, the Swedish Government encouraged people to
choose their own fund. An information campaign suggested that a person’s opportunity to
affect their pension benefits was surrendered by investing in the default. As a result, more
than two-thirds of new savers chose a fund.
Since then the share of new savers that actively choose their own fund has steadily
declined, with most now allocated to the default. In 2011 about 98.5 per cent of new savers
joined the default, and there is evidence that many do so by making a deliberate passive
choice (Barr 2013). That said, about 70 per cent of savers aged between 40–60 eventually
move their savings from the default fund (SIFA 2015).
A number of reasons have been offered for the decline in new savers making an active
choice. These include an excessive amount of fund choice, the outbreak of the dot-com
bubble immediately after the launch of the premium pension, and the strong performance of
AP7 compared with other premium pensions (Hagen 2013). AP7’s management fees have in
the past been less than half that charged by the average premium pension fund (Barr 2013).
From 2000 to 2015, AP7 Safa averaged an annual return of 4.91 per cent, compared with
2.57 per cent for choice funds (Better Finance 2016).
In general, premium pensions have achieved higher returns on average than income
pensions. Between 1995 and 2014, premium pensions averaged 6.4 per cent annual return
on investments, compared with 2.5 per cent for income pensions (SIFA 2015). And while
the income pension system has more assets under management, the relative size of the
premium pension system has been growing.

REFERENCES
DRAFT REPORT
243
References
Abaluck, J. and Gruber, J. 2011, ‘Choice inconsistencies among the elderly: Evidence from
plan choice in the Medicare Part D program’, American Economic Review, vol. 101,
no. 4, pp. 1180–1210.
ABS (Australian Bureau of Statistics) 2013, Labour Mobility, Australia, February 2013,
Cat. 6209.0, Canberra.
—— 2016, Labour Force, Australia, November 2016, Cat. 6291.0.55.003, Canberra.
ACCC (Australian Competition and Consumer Commission) 2008, Merger Guidelines,
Canberra.
—— 2015, Information and Informed Decision-Making: A Report to the Australian Senate
on Anti-Competitive and Other Practices by Health Insurers and Providers in Relation
to Private Health Insurance, Canberra.
Actuaries Institute 2016, Submission: Product Dashboards and Disclosure,
www.actuaries.asn.au/Library/Submissions/Superannuation/2016/20161108Submission
TreasuryProductDashboardFnl.pdf (accessed 30 November 2016).
AEMC (Australian Energy Market Commission) 2012, Power of Choice: Giving
Consumers Options in the Way They Use Electricity, Directions Paper, Sydney.
—— 2016, Retail Competition Review, Final Report, Sydney.
Ainsworth, A., Akhtar, S., Corbett, A., Lee, A. and Walter, T. 2016, Superannuation Fund
Performance and Fund Fees, Working Paper no. 115, Centre for International Finance
and Regulation, Sydney.
Ali, P., Anderson, M., Clark, M., Ramsay, I. and Shekhar, C. 2014, Superannuation
Knowledge, Behaviour and Attitudes in Young Adults in Australia, Research Report,
Centre for International Finance and Regulation, Sydney.
Ang, A., Goetzmann, W. and Schaefer, S. 2010, The Efficient Market Theory and
Evidence: Implications for Active Investment Management, now Publishers Inc., Delft,
The Netherlands.
ANZ (Australia and New Zealand Banking Group) 2015, ANZ Survey of Adult Financial
Literacy in Australia, Melbourne.
APRA (Australian Prudential Regulation Authority) 2012, Instruction Guide Application
Form – Authority to Offer a MySuper Product, Sydney.
—— 2016, Draft SPG 227 – Successor Fund Transfers and Wind-ups, Prudential Practice
Guide, Sydney.

DRAFT REPORT
—— 2017a, Annual MySuper Statistics: June 2016, Sydney.
—— 2017b, Annual Superannuation Bulletin 2016, Sydney.
—— 2017c, Quarterly MySuper Statistics Back Series, www.apra.gov.au/Super/
Publications/Pages/Quarterly-MySuper-statistics.aspx (accessed 7 March 2017).
—— 2017d, Annual Fund-level Superannuation Statistics 2016, Sydney.
ASFA (Association of Superannuation Funds of Australia) 2012, Submission to the
Productivity Commission on its Draft Report on Default Funds in Modern Awards,
Submission, Sydney.
—— 2014, Response to the Financial System Inquiry Interim Report, Sydney.
ASIC (Australian Securities and Investments Commission) 2013, Consumer Testing of the
MySuper Product Dashboard, Report 378, Sydney.
—— 2014, MySuper Product Dashboard Requirements for Superannuation Trustees,
Information Sheet 170, Sydney.
—— 2016, FOFA: Background and implementation, www.asic.gov.au/regulatory-
resources/financial-services/future-of-financial-advice-reforms/fofa-background-and-
implementation (accessed 14 December 2016).
Athey, S., Levin, J. and Seria, E. 2011, ‘Comparing open and sealed-bid auctions:
Evidence from timber auctions’, The Quarterly Journal of Economics, vol. 126, no. 1,
pp. 207–257.
ATO (Australian Taxation Office) 2015a, Superannuation Basics for Employers,
https://www.ato.gov.au/General/Other-languages/In-detail/Information-in-other-
languages/Superannuation-basics-for-employers/ (accessed 6 December 2016).
—— 2015b, Superannuation (Super) Standard Choice Form, www.ato.gov.au/Forms/
Superannuation-(super)-standard-choice-form/ (accessed 22 June 2016).
—— 2016a, Self-Managed Super Fund Statistical Report June 2016: Population and Asset
Allocation Tables, www.ato.gov.au/about-ato/research-and-statistics/in-detail/super-
statistics/smsf/self-managed-super-fund-statistical-report-june-2016/?page=3 (accessed
18 October 2016).
—— 2016b, Single Touch Payroll: ATO Consultation Paper, Canberra.
—— 2016c, Super Accounts Data Overview, www.ato.gov.au/About-ATO/Research-and-
statistics/In-detail/Super-statistics/Super-accounts-data/Super-accounts-data-overview
(accessed 2 March 2017).
Australian Government 2011, Stronger Super, Information Pack, Canberra.
—— 2016, Superannuation Legislation Amendment (Choice of Fund) Bill 2016, Canberra.
AustralianSuper 2014, AustralianSuper Submission to the FSI – Interim Report,
Melbourne.

REFERENCES
DRAFT REPORT
245
Ausubel, L.M. and Milgrom, P. 2005, ‘The lovely but lonely Vickrey auction’, in Cramton,
P., Shoham, Y. and Steinberg, R. (eds), Combinatorial Auctions, MIT Press,
Cambridge, Massachusetts, pp. 22–26.
Barr, N. 2013, The Pension System in Sweden, Report to the Expert Group on Public
Economics, Regeringskansliet, Stockholm.
Basu, A. and Andrews, S. 2014, ‘Asset allocation policy, returns and expenses of
superannuation funds: Recent evidence based on default options’, The Australian
Economic Review, vol. 47, no. 1, pp. 63–77.
Bateman, H., Dobrescu, I., Newell, B.R., Ortmann, A. and Thorp, S. 2016a, Flicking the
Switch: How Fee and Return Disclosures Drive Retirement Plan Choice, Working
Paper 2016/15, ARC Centre of Excellence in Population Ageing Research, Sydney.
——, ——, ——, —— and —— 2016b, ‘As easy as pie: How retirement savers use
prescribed investment disclosures’, Journal of Economic Behavior and Organization,
vol. 121, pp. 60–76.
——, Eckert, C., Geweke, J., Louviere, J., Satchell, S. and Thorp, S. 2016c, ‘Risk
presentation and portfolio choice’, Review of Finance, vol. 20, no. 1, pp. 201–229.
Benartzi, S. and Lehrer, J. 2015, The Smarter Screen, Piatkus, London.
—— and Thaler, R.H. 1999, ‘Risk aversion or myopia? Choices in repeated gambles and
retirement investments’, Management Science, vol. 45, no. 3, pp. 364–381.
—— and —— 2001, ‘Naive diversification strategies in defined contribution saving
plans’, American Economic Review, vol. 91, no. 1, pp. 79–98.
—— and —— 2002, ‘How much is investor autonomy worth?’, Journal of Finance,
vol. 57, no. 4, pp. 1593–1616.
—— and —— 2007, ‘Heuristics and biases in retirement savings behavior’, Journal of
Economic Perspectives, vol. 21, no. 3, pp. 81–104.
Better Finance 2016, Pension Savings: The Real Return, The European Federation of
Investors and Financial Services Users, Brussels.
Bhandari, G., Hassanein, K. and Deaves, R. 2008, ‘Debiasing investors with decision
support systems: An experimental investigation’, Decision Support Systems, vol. 46,
pp. 399–410.
Boyd, T. 2017, ‘APRA data shows super fund cash crisis’, Australian Financial Review,
2 February.
Butt, A., Donald, S., Foster, D., Thorp, S. and Warren, G. 2015, Delegation, Trust and
Defaulting in Retirement Savings: Perspectives from Plan Executives and Members,
Working Paper no. 65, Centre for International Finance and Regulation, Sydney.
Canstar 2016, Forget the RBA: Get your own rate cut!, www.canstar.com.au/home-
loans/no-rba-cut-but-still-plenty-of-money-to-save (accessed 20 January 2017).

DRAFT REPORT
Carroll, G.D., Choi, J.J., Laibson, D., Madrian, B.C. and Metrick, A. 2009, ‘Optimal
defaults and active decisions’, The Quarterly Journal of Economics, vol. 124, no. 4,
pp. 1639–1674.
CFFC (NZ Commission for Financial Capability) 2016, 2016 Review of Retirement Income
Policies, Auckland.
Chan, C., Laplagne, P. and Appels, D. 2003, The Role of Auctions in Allocating Public
Resources, Productivity Commission Staff Research Paper, Canberra.
Chant West 2014a, Active and Passive Management: Australian Super Industry
Experience, Research Paper, Sydney.
——2014b, Chilean Pension System: Relevance for Australia, Research Paper, Sydney.
——2015, Financial System Inquiry Final Report: Response to Superannuation
Recommendation 10, Sydney.
——2016, Chilean Pension System: Updated Comparison with Australia, Research Paper,
Sydney.
——2017, Super Ratings Methodology, www.chantwest.com.au/fund-ratings/super-
methodology (accessed 19 June 2016).
Chernev, A., Böckenholt, U. and Goodman, J. 2015, ‘Choice overload: A conceptual
review and meta-analysis’, Journal of Consumer Psychology, vol. 25, no. 2, pp. 333–
358.
CHOICE 2016, Junk Health Insurance, www.choice.com.au/money/insurance/health/
articles/junk-health-insurance (accessed 20 January 2017).
CIFR (Centre for International Finance and Regulation) 2015, Competition in Financial
Services, Research Report, Sydney.
Clare, R. 2010, Choosing to Choose, Association of Superannuation Funds of Australia,
Sydney.
Clark, G., Duong, H., Gerrans, P., Lajbcygier, P., Moulang, C., Strydom, M., Vaz, J. and
Wickramanayake, J. 2013, A Review of Retirement Savings Investment Behaviours:
Theory and Evidence, Working Paper, CSIRO-Monash Superannuation Research
Cluster, Melbourne.
——, Strauss, K. and Knox-Hayes, J. 2012, Saving for Retirement, Oxford University
Press, Oxford.
Cooper, J., Casey, K., Evans, G., Grant, S., Gruen, D., Heffron, M., Martin, I. and Wilson,
B. 2010, Super System Review Final Report – Part One: Overview and
Recommendations, Canberra.
CPP Investment Board 2016, CPP Investment Board: Who We Are, CPP Investment
Board, www.cppib.com/en/who-we-are (accessed 2 March 2017).

REFERENCES
DRAFT REPORT
247
Cullen, M. and Dalziel, L. 2006, KiwiSaver On Track with Default Providers Selected,
Media Release, Beehive, www.beehive.govt.nz/release/kiwisaver-track-default-
providers-selected (accessed 5 January 2017).
Cummings, J. and Ellis, K. 2011, Risk and Return of Illiquid Investments: A Trade-Off for
Superannuation Funds Offering Transferable Accounts, Working Paper, Australian
Prudential Regulation Authority, Sydney.
—— and —— 2015, ‘Risk and return of illiquid investments: A trade-off for
superannuation funds offering transferable accounts’, Economic Record, vol. 91,
no. 295, pp. 463–476.
Deloitte 2010, Super System Review: Default Fund Costs Under the MySuper Proposals,
Sydney.
——2014, Shaping the future – Financial System Inquiry: Deloitte Submission to the
Interim Report, Sydney.
Deloitte Access Economics 2014, Competition in Retail Banking, Sydney.
Delpachitra, S. and Rafizadeh, S. 2014, ‘The switching decision: Are members of
superannuation funds rational and informed investors?’, Australasian Accounting,
Business and Finance Journal, vol. 8, no. 4, pp. 45–62.
Department of Finance 2014, Commonwealth Procurement Rules: 7. Accountability and
Transparency in Procurement, www.finance.gov.au/procurement/procurement-policy-
and-guidance/commonwealth-procurement-rules/cprs-accountability-transparency-
procurement.html (accessed 17 January 2017).
Drew, M., Stanford, J. and Veeraraghavan, M. 2002, ‘Efficiency with costly information:
A study of Australian wholesale superannuation fund performance’, Economic Analysis
and Policy, vol. 32, no. 1, pp. 35–47.
——, Walk, A.N. and West, J.M. 2015, The Role of Asset Allocation in Navigating the
Retirement Risk Zone, Financial Services Institute of Australasia, Sydney.
English, B. and Foss, C. 2013, KiwiSaver Default Provider Review Completed, Media
Release, Beehive, www.beehive.govt.nz/release/kiwisaver-default-provider-review-
completed (accessed 3 January 2017).
Fear, J. and Pace, G. 2008, Choosing Not to Choose: Making Superannuation Work by
Default, The Australia Institute and Industry Super Network, Sydney.
Fernandes, D., Lynch Jr, J.G. and Netemeyer, R.G. 2014, ‘Financial literacy, financial
education, and downstream financial behaviors’, Management Science, vol. 60, no. 8,
pp. 1861–1883.
Fernyhough, J. 2016, The People’s Pension Overtakes NEST as Biggest AE Scheme, FT
Adviser, www.ftadviser.com/2016/05/23/pensions/the-people-s-pension-overtakes-nest
-as-biggest-ae-scheme-eA1OMrnXsB78DA1ATWzvyM/article.html (accessed
20 February 2017).

DRAFT REPORT
FindLaw 2016, ERISA Section 404(c) FAQs, http://employment.findlaw.com/wages-and-
benefits/erisa-section-404c-faqs-.html (accessed 20 February 2017).
FMA (NZ Financial Markets Authority) 2016, KiwiSaver Annual Report 1 July 2015 –
30 June 2016, Wellington.
Frost, J. 2016, ‘Super funds accelerate the shift to digital’, Australian Financial Review,
4 December.
FRTIB (US Federal Retirement Thrift Investment Board) 2016, Summary of the Thrift
Savings Plan, Birmingham, Alabama.
FSC (Financial Services Council) 2014, Financial System Inquiry – Phase Two: Chapter
One: Superannuation, Investment Management and Retirement Policy, Sydney.
FWC (Fair Work Commission) 2016, Default Superannuation List,
www.fwc.gov.au/awards-and-agreements/modern-award-reviews/superannuation-fund-
reviews/overview/default (accessed 1 January 2017).
Gerrans, P. and Yap, G. 2013, Sophistication in Retirement Savings Investment Choices,
Working Paper, Australian Centre for Financial Studies, Melbourne.
Goda, G.S., Levy, M.R., Manchester, C.F., Sojourner, A. and Tasoff, J. 2015, The Role of
Time Preferences and Exponential-Growth Bias in Retirement Savings, Working Paper
no. 21482, National Bureau of Economic Research, Cambridge, Massachusetts.
Government of Canada 2016, Canada Pension Plan: Overview, www.canada.ca/
en/services/benefits/publicpensions/cpp.html (accessed 2 March 2017).
Hagen, J. 2013, A History of the Swedish Pension System, Working Paper 2013:7, Uppsala
Center for Fiscal Studies, Uppsala.
Heuser, A., Kwok, J., Snethlage, D. and Watts, D. 2015, Review of the KiwiSaver Fund
Manager Market Dynamics and Allocation of Assets, New Zealand Treasury,
Wellington.
Huberman, G. and Jiang, W. 2006, ‘Offering versus choice in 401(k) plans: Equity
exposure and number of funds’, Journal of Finance, vol. 61, no. 2, pp. 763–801.
Inland Revenue 2014, KiwiSaver: Get your Money, www.kiwisaver.govt.nz/already/get-
money/early/ (accessed 17 December 2015).
—— 2015, KiwiSaver Evaluation: Final Summary Report, Wellington.
—— 2016a, KiwiSaver: Choosing your KiwiSaver Scheme Provider,
www.kiwisaver.govt.nz/new/providers/ (accessed 19 December 2016).
—— 2016b, KiwiSaver: KiwiSaver in a Nutshell, www.kiwisaver.govt.nz/
new/about/summary/ (accessed 19 December 2016).
—— 2016c, Kiwisaver: Who’s Involved in Kiwisaver, www.kiwisaver.govt.nz/
new/about/who/ (accessed 19 December 2016).
—— 2016d, KiwiSaver: Your Employer’s Contributions, www.kiwisaver.govt.nz/
already/contributions/employers/ (accessed 19 December 2016).

REFERENCES
DRAFT REPORT
249
—— 2017, Working with Inland Revenue (How it works), Inland Revenue,
www.ird.govt.nz/kiwisaver/employers/how/working-with-ird/ks-emp-how-ird.html
(accessed 1 February 2017).
Internal Revenue Service 2017, 401(k) Plan Overview, www.irs.gov/retirement-plans/plan-
sponsor/401k-plan-overview (accessed 2 March 2017).
Iyengar, S.S., Huberman, G. and Jiang, W. 2004, ‘How much choice is too much?
Contributions to 401(k) retirement plans’, Pension Design and Structure: New Lessons
from Behavioral Finance, Oxford University Press, Oxford, pp. 83–95.
Johnson, B., Boccellari, T., Bryan, A. and Rawson, M. 2015, Morningstar’s Active/Passive
Barometer: A New Yardstick For An Old Debate, http://global.morningstar.com/
activepassiveJune2015/ (accessed 7 March 2017).
Johnson, E.J., Hassin, R., Baker, T., Bajger, A.T. and Treuer, G. 2013, ‘Can consumers
make affordable care affordable? The value of choice architecture’, PLoS One, vol. 8,
no. 12, pp. 1–6.
——, Shu, S.B., Dellaert, B.G.C., Fox, C., Goldstein, D.G., Häubl, G., Larrick, R.P.,
Payne, J.W., E. Peters, Schkade, D., Wansink, B. and Weber, E.U. 2012, ‘Beyond
nudges: Tools of a choice architecture’, Marketing Letters, vol. 23, pp. 487–504.
Jones, R. and Wermers, R. 2011, ‘Active management in mostly efficient markets’,
Financial Analysts Journal, vol. 67, no. 6, pp. 29–45.
Joyce, S. 2017, Lifting NZ Super Age the Right Thing To Do, Media Release, Beehive,
https://beehive.govt.nz/release/lifting-nz-super-age-right-thing-do (accessed 16 March
2017).
Kahneman, D. 2011, Thinking, Fast and Slow, Allen Lane, New York.
—— and Tversky, A. 1979, ‘Prospect theory: An analysis of decisions under risk’,
Econometrica, vol. 47, no. 2, pp. 263–292.
Kaufmann, C., Weber, M. and Haisley, E. 2013, ‘The role of experience sampling and
graphical displays on one’s investment risk appetite’, Management Science, vol. 59,
no. 2, pp. 323–340.
Keim, D.B. and Mitchell, O.S. 2016, Simplifying Choices in Defined Contribution
Retirement Plan Design, Working Paper no. 21854, National Bureau of Economic
Research, Cambridge, Massachusetts.
Keller, P.A., Harlam, B., Loewenstein, G. and Volpp, K.G. 2011, ‘Enhanced active choice:
A new method to motivate behavior change’, Journal of Consumer Psychology,
vol. 21, pp. 376–383.
Ketcham, J.D., Kuminoff, N.V. and Powers, C.A. 2016, Estimating the Heterogeneous
Welfare Effects of Choice Architecture: An Application to the Medicare Prescription
Drug Insurance Market, Working Paper no. 22732, National Bureau of Economic
Research, Cambridge, Massachusetts.

DRAFT REPORT
Kling, J.R., Mullainathan, S., Shafir, E., Vermeulen, L. and Wrobel, M.V. 2012,
‘Comparison friction: Experimental evidence from Medicare drug plans’, Quarterly
Journal of Economics, vol. 127, no. 1, pp. 199–235.
Krosnick, J.A. and Alwin, D.F. 1987, ‘An evaluation of a cognitive theory of response-
order effects in survey measurement’, Public Opinion Quarterly, vol. 51, no. 2,
pp. 201–219.
Littlewood, M. 2012, KiwiSaver: Changing the Default Provider Regime, University of
Auckland Business School, Auckland.
Luckling-Reiley, D. 2000, ‘Vickrey auctions in practice: From nineteenth-century philately
to twenty-first-century e-commerce’, The Journal of Economic Perspectives, vol. 14,
no. 3, pp. 183–192.
MacDonald, K.L., Bianchi, R.J. and Drew, M.E. 2014, Equity Risk Versus Retirement
Adequacy: Asset Allocation Solutions for KiwiSaver, Discussion Paper no. 2014-02,
Griffith Business School, Gold Coast.
Mayer, R.E. and Moreno, R. 2003, ‘Nine ways to reduce cognitive load in multimedia
learning’, Educational Psychologist, vol. 38, no. 1, pp. 43–52.
MBIE (NZ Ministry of Business, Innovation and Employment) 2012, Review of KiwiSaver
Default Provider Arrangements, Wellington.
—— 2013, Questions and Answers: KiwiSaver Default Scheme Providers, Wellington.
Minifie, J., Cameron, T. and Savage, J. 2014, Super Sting: How to Stop Australians Paying
Too Much for Superannuation, Grattan Institute, Melbourne.
——, —— and —— 2015, Super Savings, Grattan Institute, Melbourne.
Mitchell, O.S. and Utkus, S.P. 2004, ‘Lessons from behavioral finance for retirement plan
design’, in Mitchell, O.S. and Utkus, S.P. (eds), Pension Design and Structure: New
Lessons from Behavioral Finance, Oxford University Press, Oxford, pp. 3–41.
Morrin, M., Inman, J.J., Broniarczyk, S.M., Nenkov, G.Y. and Reuter, J. 2012, ‘Investing
for retirement: The moderating effect of fund assortment size on the 1/N heuristic’,
Journal of Marketing Research, vol. 49, pp. 537–550.
Mullainathan, S. and Thaler, R.H. 2000, Behavioural Economics, Working Paper no. 7948,
National Bureau of Economic Research, Cambridge, Massachusetts.
Murray, D., Davis, K., Dunn, C., Hewson, C. and McNamee, B. 2014, Financial System
Inquiry: Final Report, Canberra.
NEST (National Employment Savings Trust) 2015, Key Facts: Auto-enrolment and NEST,
London.
—— 2016, Nest Pension Scheme Annual Report and Accounts: For the Year Ended
31 March 2016, London.
Northern Territory Government 2014, Selection of Successor Fund,
www.treasury.nt.gov.au/SuperannuationOffice/SuperannuationOfficeNews/Lists/Posts/

REFERENCES
DRAFT REPORT
251
Post.aspx?List=05af09e0-e46e-47d4-aad0-bfab58a76092&ID=31&Web=bbece4d7-
208b-441f-93a2-849829d258d4 (accessed 6 January 2017).
NZCO (New Zealand Cabinet Office) 2013, Future KiwiSaver Default Provider
Arrangements, Wellington.
O’Dwyer, K. 2016a, Government Moves to Give More Employees Choice, Media Release
no. 026, Treasury, http://kmo.ministers.treasury.gov.au/media-release/026-2016
(accessed 7 March 2017).
—— 2016b, Superannuation Legislation Amendment (Choice of Fund) Bill 2016,
Explanatory Memorandum, Canberra.
OFT (UK Office of Fair Trading) 2013, Defined Contribution Workplace Pension Market
Study, London.
PC (Productivity Commission) 2012, Default Superannuation Funds in Modern Awards,
Report no. 60, Canberra.
—— 2015a, Superannuation Policy for Post-Retirement, Commission Research Paper,
Canberra.
—— 2015b, Workplace Relations Framework, Report no. 76, Canberra.
—— 2016, How to Assess the Competitiveness and Efficiency of the Superannuation
System, Research Report, Canberra.
PHIAC (Private Health Insurance Administration Council) 2015, Barriers to Entry in the
Australian Private Health Insurance Market, Canberra.
PJCCFS (Parliamentary Joint Committee on Corporations and Financial Services) 2007,
The Structure and Operation of the Superannuation Industry, Parliament of Australia,
Canberra.
PPI (Pensions Policy Institute) 2015, The Future Book: Unravelling Workplace Pensions,
London.
—— 2016, Pension Facts: October 2016, London.
QSuper 2017, Choice of Fund for Queensland Government Employees, QSuper -
Employers, https://qsuper.qld.gov.au/employers/payments-and-employment-changes/
queensland-government/how-to-pay/choice-of-fund/ (accessed 2 March 2017).
Rainmaker nd, Harness the Power of Your Super Fund Today, Selecting Super,
www.selectingsuper.com.au/learning_centre/harness_the_power_of_your_super_fund_
today (accessed 2 March 2017).
RateCity 2017, Rates of the Nation Report: January 2017, www.ratecity.com.au/home-
loans/mortgage-news/rates-nation-report-january-2017 (accessed 20 January 2017).
REST Industry Super 2016, Bridge the Gap, White Paper, Sydney.
Rowell, H. 2016, ‘Governance and Culture in Superannuation’, presented to the AFR
Banking and Wealth Summit, Sydney, 5 April.

DRAFT REPORT
—— 2017, Proof Committee Hansard: Economics Legislation Committee, Australian
Prudential Regulation Authority, Senate Estimates, 2 March, Canberra.
Samuelson, W. and Zeckhauser, R. 1988, ‘Status quo bias in decision making’, Journal of
Risk and Uncertainty, vol. 1, pp. 7–59.
Sass, S. 2014, The UK’s Ambitious New Retirement Savings Initiative, Brief, 14–5, Center
for Retirement Research at Boston College, Boston.
Shelton, A. 2012, Chile’s Pension System: Background in Brief, Congressional Research
Service, Washington DC.
Shepherd, Smith Edwards & Kantas LTD LLP 2015, Pimco, Allianz, Sued in 401K Class
Action Case Alleging $2.5M in Excessive Fees, Institutional Investor Securities Blog,
www.institutionalinvestorsecuritiesblog.com/2015/10/pimco_allianz_sued_in_401k_cl
a.html (accessed 7 March 2017).
Shorten, B. 2011, Superannuation Legislation Amendment (MySuper Core Provisions) Bill
2011, Explanatory Memorandum, Canberra.
—— 2013, Super Funds to Disclose Dollar Value of Fees, Media Release no. 030,
Treasury, http://ministers.treasury.gov.au/DisplayDocs.aspx?doc=pressreleases/2013/
030.htm&pageID=&min=brs&Year=&DocType=0 (accessed 22 November 2016).
SIFA (Swedish Investment Fund Association) 2015, More Facts About the Premium
Pension 2.0, Stockholm.
Simon, H.A. 1955, ‘A behavioural model of rational choice’, The Quarterly Journal of
Economics, vol. 69, no. 1, pp. 99–118.
Skelton, R.A. 2015, ‘The Impact of Home Loan Key Facts Sheets on Borrowers’
Comparisons of Loan Costs’, PhD thesis thesis, Queensland University of Technology,
Brisbane.
St John, S., Littlewood, M. and Dale, M. 2014, Now We Are Six: Lessons from New
Zealand’s KiwiSaver, Working Paper 2014-1, Retirement Policy and Research Centre,
University of Auckland, Auckland.
Superintendence of Pensions 2010, The Chilean Pension System, Government of Chile,
Santiago.
—— 2016a, Financial Statistical Bulletin, July 2016, Government of Chile, Santiago.
—— 2016b, Superintendencia de Pensiones Adjudico a Planvital Cartera de Nuevos
Afiliados, Government of Chile, Santiago.
SuperRatings 2017, MySuper Methodology, www.superratings.com.au/ratings (accessed
14 June 2016).
Tasmania DTF (Tasmania Department of Treasury and Finance) 2014, Guidelines on
Tender Evaluation Using Weighted Criteria for Building Works and Services, Hobart.
Thaler, R.H. and Sunstein, C.R. 2009, Nudge: Improving Decisions About Health, Wealth
and Happiness, Penguin, London.

REFERENCES
DRAFT REPORT
253
Thrift Savings Fund 2015, Financial Statements, December, Washington DC.
TPR (The Pensions Regulator) 2016, Detailed Guidance for Employers: Pension Schemes:
Pension Schemes Under the New Employer Duties, Detailed Guide for Employers
no. 4, UK Government, Birmingham.
Treasury 2016, Development of the Framework for Comprehensive Income Products for
Retirement, 15 December, Canberra.
Tversky, A. and Kahneman, D. 1974, ‘Judgment under uncertainty: heuristics and biases’,
Science, vol. 185, no. 4157, pp. 1124–1131.
UK DWP (Pensions) 2016, NEST: Evolving for the Future, London.
US Department of Labor and Internal Revenue Service 2013, Automatic Enrollment 401(k)
Plans for Small Businesses, Publication 4674, Washington DC.
Vásquez, J. 2004, Comportamiento Manada en las Administradoras de Fondos de
Pensiones, University of Chile, Santiago.
Weaven, G. 2016, ‘Workers’ Capital: The story of Industry Funds and Australia’s
Superannuation Revolution’.
Weinberg, A. 2004, IPO Dutch Auctions Vs. Traditional Allocation, Forbes,
www.forbes.com/2004/05/10/cx_aw_0510mondaymatchup.html (accessed 8 December
2016).
White, A. 2011, World’s Largest Defined Contribution Plan to Tender Investments,
Top1000Funds.com, www.top1000funds.com/profile/2011/08/31/us-federal-employees
-plan-embarks-on-giant-investment-tender/ (accessed 6 December 2016).
Willis Towers Watson 2016, The Shift to Digital: How Far Have We Come? Technology-
Based Financial Education in the Super Industry, Sydney.
Yeates, C. 2015, ‘APRA backs more super fund mergers’, The Sydney Morning Herald,
29 April.

Draft Superannuation Alternative Default Models

More Related Content

Similar to Draft Superannuation Alternative Default Models (20)

More from Jonathan Steffanoni (20)

Recently uploaded (20)

Draft Superannuation Alternative Default Models