Adversarial and cooperative models in contracting for public services
1 like607 views
This document discusses contracting for public services and the transition from government to governance. It identifies several confounding factors that add environmental uncertainty and complicate the principal-agent relationship in contracting. Specifically, it examines claims that contracting leads to improved economy, efficiency and effectiveness but finds limited empirical evidence supporting this. Research designs studying contracting often have flaws and do not account for all contextual factors. The level of cooperation between principals and agents may depend on environmental turbulence, which could determine contract types. A proposed study would use a standard set of variables and a typology matrix to categorize contracts based on cooperation levels and turbulence across different contexts.
![12
foisted into a new role with no time or life experience to formulate preferences, let alone a contract
or monitoring strategy [e.g., the new parents of a critically ill newborn (Heimer & Staffen 1998)],
the asymmetry of power shifts from the principal to the agent. The absence of institutional memory
and professional capacity in a public agency can only exacerbate this shift of power.
The increasingly contested political environment of contracting can produce overly rigid and
excessive regulation of contract agencies (Smith and Lipsky 1993; Gutch 1992; Bernstein 1991).
Service providers are asked to submit large amounts of information on their operations, even though
government agencies do not have the capacity for reviewing it. Further, some of the requested
information is not germane to evaluating program quality. For many nonprofit agencies, especially
smaller community organizations, programmatic regulations can lead to major changes in mission,
services, and clients. Service costs also rise because compliance with these regulations requires
contract agencies to add administrative staff and develop new internal monitoring systems. Without
the requisite institutional knowledge, public agencies are unable to enforce their rights of a principal
and remain hostage to contractors’ opportunism without adequate evaluation and enforcement
capacity.
Proposition: Even if a competitive market were to exist, the limited contract management capability
of public agencies would continue to tilt the principal-agent balance in favor of the agent, since
there would be no effective statutory regulator; consequently, the principal’s transaction costs
would continue to rise without any concomitant rise in 3E and accountability of contractors’
services.
D. Regulation of Contractors
Are the internal controls in contracted agencies, upon which evaluation information is
provided to public agencies, adequate? Dicke studied a state agency on disabilities (AOD) and its
contracted private providers for the delivery of residential services to AOD clients in a single
western state (502-16). The 12 provider organizations in the study received more than75%of the
AOD’s contracted services budget for the fiscal year of 2000, and each provider organization
received on average more than 95% of its residential service revenue either directly or indirectly
from a government source. AOD case managers reported that monitoring was overwhelmingly the
most helpful review procedure used by their organizations to ensure quality in services, followed by
outcomes-based assessments (52%), auditing (31%), and provider self reports (14%), and other
reviews (21%). However, on the overall efficacy of these methods, 29% of respondents felt that they
were “just about right,” 6% said they were “too strict and required too much paperwork,” 13%](https://image.slidesharecdn.com/adversarialandcooperativemodelsincontractingforpublicservices-100719050411-phpapp02/85/Adversarial-and-cooperative-models-in-contracting-for-public-services-13-320.jpg)
Adversarial and cooperative models in contracting for public services
- 1. PA 762 ORGANIZATION THEORY Instructor: Dr. Branda Nowell GOVERNMENT TO GOVERNANCE: ADVERSARIAL & COOPERATIVE MODELS IN CONTRACTING FOR PUBLIC SERVICES SHANTANU BASU North Carolina State University Raleigh, NC 27695 Date of Submission: April 23, 2008 Date of Presentation: May 5, 2008
- 2. 1 CONTENTS ABSTRACT _____________________________________________________ 2 I. Introduction ____________________________________________________ 3 II. Transition from Government to Governance __________________________ 3 III. Confounding Factors: creating or compounding uncertainty?_____________ 4 A. Economy, Efficiency and Effectiveness (3E) __________________________ 4 B. Does a Competitive Market Exist? __________________________________ 6 C. Is there a Market Regulator? _______________________________________ 9 D. Regulation of Contractors ________________________________________ 12 IV. Role of Environmental Uncertainty/Turbulence ______________________ 14 A. Government to Governance Model_________________________________ 19 B. Typology Matrix _______________________________________________ 21 C. Operationalizing the Independent Variables __________________________ 21 D. Applicability of Model __________________________________________ 23 VI. Directions for Future Research ___________________________________ 24 VII. REFERENCES_______________________________________________ 25
- 3. 2 ABSTRACT The extant literature on contracting for public services shows a clear dichotomy. On the one hand, are ranged the apologists for contracting; on the other, are the detractors who however, have provided only limited empirical evidence of the pitfalls of contracting, thus leaving an empirical research gap between the perceived benefits and pitfalls of outsourcing. Much of this research gap is attributable to the absence of uniform and accurate measures for the variables used in the study apart from homogeneity of physical areas, target population, etc. Nor do such research designs and findings permit of generalizability. Evidently there are complex issues and linkages involved in transposing private sector principles into a governmental environment that, unless measured and controlled uniformly and accurately, would continue to retard the credibility of research findings. Therefore what are the major confounding factors adding to/creating environmental uncertainty that impact contracting for public services? Does environmental uncertainty determine the structure and/or the quality of a contract? This proposal addresses these research questions in the principal- agent perspective and hypothesizes that the level of cooperation between the principal and the agent is related to the level of environmental turbulence that, in turn, determines the nature and/or types of contracts entered into between them. The proposal proposes a research design that would be flexible of use at multiple levels of analysis and across different sets of programs/organizations using a standard set of eight independent variables and their constituent control variables that dovetail into a typology matrix of contracts. The matrix juxtaposes the level of cooperation with environmental turbulence and categorizes contracts as falling into four distinct categories.
- 4. 3 I. Introduction The transition from government to governance over the last two decades has witnessed governments all over the world contracting out many essential services, particularly municipal and human services, to contractors. However, not only has extensive literature been produced on this subject but there is limited empirical evidence that points to the success or otherwise of outsourcing in a context or set of contexts. Nor do such research designs and findings permit of generalizability. Evidently there are complex issues and linkages involved in transposing private sector principles into a governmental environment that, unless accounted for uniformly and accurately, would continue to retard the credibility of research findings. Therefore what are the major confounding factors adding to/creating environmental uncertainty that impact contracting for public services? Does environmental uncertainty determine the structure and/or the quality of a contract? This proposal addresses these research questions in the principal-agent perspective and hypothesizes that the level of cooperation between the principal and the agent is related to the level of environmental turbulence that, in turn, determines the nature and/or types of contracts entered into between them. The proposal proposes a research design that would be flexible of use at multiple levels with a standard set of eight independent variables and their constituent control variables that dovetail with a typology matrix of contracts. The matrix juxtaposes the level of cooperation with environmental turbulence and categorizes contracts as falling into four distinct categories. II. Transition from Government to Governance What does this transition entail? Gerry Stoker (18-19) puts forth five propositions. Governance refers to a set of institutions and actors that are drawn from outside government. Second, it identifies the blurring of boundaries and responsibilities for social and economic matters. Third, it identifies the power dependence involved in the relationships between institutions involved in collective actions. Fourth, it is about autonomous self-governing network of actors. Fifth, it recognizes the capacity to get things done which does not rest on the power of government to directly command or use its authority. These propositions are complementary and not exclusive. However, Stoker also points out some dilemmas that these propositions carry. First, there is a divorce between the complex reality of decision-making associated with governance and the normative codes used to explain and justify government. Second, the blurring of responsibilities may
- 5. 4 lead to scapegoating. Third, power dependence exacerbates the problem of unintended consequences for government. Fourth, the emergence of self-governing networks raises issues of accountability. Fifth, even where governments operate in a flexible way to steer collective action, governance failure may occur (20-22). Contracting has come mainly in tandem with gradual downsizing in public agencies and the injection of private sector principles of economy, efficiency and effectiveness. At the same time the peculiar nature of such services has rendered public agencies vulnerable to malpractices by contractors and raised serious doubts about the efficacy and desirability of contracting as Stoker has pointed out. New practices evolved by public agencies have added to such doubts. Evidently, there are confounding factors that complicate the principal-agent relationship and add to/interact with environmental turbulence and obfuscates the real benefits/losses of outsourcing. III. Confounding Factors: creating or compounding uncertainty? A. Economy, Efficiency and Effectiveness (3E) Proponents of contracting argue that it is more cost-efficient and better stimulates innovation than direct service delivery. On the other hand, direct service delivery has been found to promote political accountability, stability, and equality of treatment (DeHoog 1984; Donahue 1989; Kettl 1993), although the relative strengths of direct versus contract service delivery appear to vary across circumstances (Brown and Potoski, 2006; Morgan and England 1988; Sclar 2000). The popular perception of contracting leading to 3E has been debunked by George Boyne. First, Boyne argues that public choice theory does not imply that transferring responsibility for a publicly funded service without competition enhances efficiency (475). Without curtailing the services, efficiency would be tested by declining costs of providing the service. Second, contracting is a prerogative of government and does not necessarily provide greater say for consumers. In fact, consumers may be saddled with an efficient service they never asked for. Third, if savings are indeed generated by contracting, municipal budgets should decline; since they don’t evidently the taxpayer reaps no benefit of economy. Last, if such savings do not accrue to the governments, their spending nonetheless on other services would continue to rise. For instance, Boyne reviews Berenyi & Stevens and Kemper & Quigley and concludes that there is little difference to costs or efficiency, regardless of population size where garbage removal was contracted out (Boyne, 477). Similarly, the empirical literature is selective among geographical areas and favors cases where contracting proved
- 6. 5 successful to the exclusion of other areas where it may not have proved economic and efficient. Critiquing McGuire and vanCott, Boynes concludes that there is little difference between public and private suppliers in the provision of school transport and that competition was associated with higher costs (Boyne, 479)! Even where municipalities have contracted with counties for services, the empirical literature does not show that economy moved in tandem with efficiency (Boyne, 478-80). Boyne therefore rightly concludes that research designs suffer from major flaws. First, such designs do not take into account local ideological preferences. Thus a traditional desire for small government may lead to lower expenditure and higher contracting; not as a conscious 3E measure (Boyne, 481). Second, the studies do not mainly control for scale effects. Population size may not therefore be the cause of economy; rather the scale of operations of the provider may be of greater import. Equally, it may not be fair to judge a contractor on the basis of its operations in a limited locality when it operates over a much larger geographical area (Boyne, 481). Third, none of the studies contain any direct measure of competition. It is probable that the contractor may replace the state agency’s monopoly or a state agency may become more efficient by the mere threat of contracting its services. Fourth, unusually small sample sizes are not representative of the ground realities (Boyne, 475). Fifth, the geographic spread of the empirical studies is such that they do not bring out the true economy of contracting such as Ahlbrandt (1973) who compares the actual cost of a contracted service with the hypothetical cost of the same service had it been rendered by a public agency (Boyne, 475). Is contracting therefore another ‘management fad’ and/or are the empirical studies inspired by such ‘fads’ and seek to justify contracting? (Boyne, 482) It is apparent that the design of the proposed research would have to control for all these factors in arriving at a make or buy decision which is what this proposal attempts to arrive at. On the other hand, Cohen (435) states that contracting should not be resorted to functions that regulate or prohibit free movement of people as this a strict sovereign function. In fact he likens private management of prisons as punishable for kidnapping! He says that contracting should not also be resorted to for activities that have no obvious customer with the resources to provide a profit to the organization that performs it, e.g. insurance. He eventually falls victim to stereotyping as he states that if the goals are innovation and efficiency, then contracting with a private vendor may be more desirable, because private employees operate with higher-powered, compensation-based, and profit oriented incentives. If the goal is more government control over service provision, then internal production may be preferred, because government employees’ motivations are typically
- 7. 6 better aligned with the agency’s mission. That efficiency is not the necessary fallout of contracting is shown by Boyne in his analysis of empirical articles where in 7 categories of services, four had unclear improvement in efficiency; three provided lower quality of services, three had held service quality constant while the improvement of two others remained unknown (Boyne, 479). Notwithstanding Boyne’s telling remarks about the make or buy decision, services with inherently lower transaction costs remain more favorable contracting targets, freeing resources to lower costs or to purchase more service quality. On the other hand, services with higher transaction costs pose greater contracting problems, consuming more resources and inhibiting a manager’s ability to optimize competing values. The principal-agent problem seems to grow as do transaction costs even when they can be minimized with state discharge of services. Proposition: 3E is not necessarily achieved by contracting with private parties; neither does contracting reduce transaction costs nor shifts the balance between the principal and agent in favor of the principal. B. Does a Competitive Market Exist? Since public choice theory centers on market competition to drive down costs of contracting, a central presumption is that a market exists. Effective markets provide managers with important information about prices and service quality of vendors and facilitate disciplining vendors who fail to meet contracts (Hart and Moore 1999; Niskanen 1971). In well-functioning markets, competition for contracts can help overcome principal–agent problems. Strong and effective markets, however, require some fairly strict conditions. They need large numbers of buyers and sellers, participants need to be well informed about products and each others ’ preferences, and actors must be able to enter and exit the market and exchange resources at low costs. If a perfect market does exist how does it explain the emergence of contract monopolies? Van Slyke addresses this issue effectively (297) and concludes that there is lack of competition and diminished quality, bearing out Boyne’s findings supra. County managers who were interviewed for this study cited environmental constraints, actions of nonprofit organizations, networked relationships and government enacted barriers as the major causes for the lack of competition. Downsizing has caused a lack of capacity on the part of public agencies to support and sustain a competitive market that, in turn, has prevented the emergence of 3E alternatives for service delivery. This issue has been dealt with at length subsequently in this proposal. Secondly, nonprofits by virtue of their expertise (e.g. mental health) created niches and did not allow competition to grow.
- 8. 7 Such steadfastness to a geographical area was sustained by the nonprofits by supporting social service legislation and strong advocacy for these changes. In addition, the nonprofits erected entry barriers such as accreditation that was, in turn, strengthened by legislation that mandated such accreditation, and kept competitors away (Van Slyke, 303). Added to this was the pressure from politicians on agency managers to contract with specific nonprofits. This has ensured that services continue to be rendered even when and where they are not required, that perpetuates the life of these nonprofits and drives out competition (Van Slyke, 303). Lastly, by contracting between themselves to reduce the costs of contracting (dealt with subsequently in detail in this proposal) government agencies have also severely limited competition (Van Slyke, 305). Markets can fail because of high transaction costs, limited information, uncertainty about the future, and the prospect that people or organizations will behave opportunistically in their interactions (Coase 1937; Williamson 1981, 1991, 1996). Of particular importance are the varying transaction costs inherent in different market and service arrangements. The type of vendor becomes more relevant when contracting for high transaction cost services in thin markets — for example, when a small government needs to deliver a service that requires large asset specific investments beyond what it can afford. Again, the literature suggests that governments already understand this; they more frequently choose nonprofit and other governments when contracting for high transaction cost services and in thin markets (Brown and Potoski 2003b). Yet even contracting with vendors that purportedly share the same goals is not without risks. Van Slyke (2003), for example, finds that governments often establish long-term contract relationships with nonprofits for social services, but then they neglect oversight and monitoring responsibilities. Nonprofits, increasingly reliant on public sector contracts, may begin to behave like conventional monopolists to maintain their resource streams. Perhaps more troubling are circumstances in which a government contracts in a thin market out of ideological, political, or nefarious motives and then neglects contract management ( Romzek and Johnston 2002 ). In the case of contracting, because the parties cannot fully predict all possible future scenarios, contracts are typically underspecified (i.e., incomplete) and may allow opportunistic vendors to exploit contracts to their own advantage at the expense of the contracting government’s goals. To minimize such opportunism, the contracting agency must incur transaction costs by clearly specifying the values sought in performance measures, writing more detailed contracts, monitoring vendors ’ performance, and enforcing sanctions when necessary. Importantly, the type of
- 9. 8 vendor may be less important for low transaction cost services in competitive markets. This is the case because the risk that vendors will become monopoly service providers is low, performance can be easily measured, and contracts are easier to enforce. Indeed, prior research suggests that governments understand this already. They tend to contract more with for-profit vendors in such circumstances (Brown and Potoski 2003b) exploiting the latter’s competitive zeal and relatively lax legal requirements through contract specifications and enforcement mechanisms (Brown and Potoski 2004). For governments that are willing and able to invest insufficient contract management capacity, however, certain types of long-term contractual relationships can foster mutual support and sharing (Artz and Brush 2000; Hart and Moore 1999; Levin 2003). In so-called relational contracts, public managers work with vendors to build a long-term relationship based on trust, reciprocity, and joint involvement in developing and implementing the contract. Over time, though, effective relational contracts may lower transaction costs through reduced bidding, monitoring, and legal costs (Hart and Moore, 1999; Tadelis, 2002 ). A major source of transaction cost is asset specificity that refers to the need for physical infrastructure, technology, or knowledge, skills, and abilities that can only be acquired through on- the-job experience or highly specialized investments (e.g., high-cost investments in computer technology). For winning and losing vendors, investing in an asset-specific service that cannot be readily translated to other economically valuable activities (i.e., used for other organizational purposes or marketed to others) leaves them vulnerable to a single (i.e., monopsonic) service purchaser. This not only raises the costs for vendors to compete in the market, but more importantly makes it unlikely they will remain in the market for future rounds of contract bidding. Conversely, for contracting governments, asset-specific services can dangerously privilege vendors that win the first contracts, thus constraining future competition. Under such monopolistic conditions, the winning vendor can exploit the contracting government with impunity by raising prices or reducing service quality. This is a major disqualification for the existence of a competitive market. Federal and state administrative procedure acts mandating whistleblower and other employee protections, as well as open records and meetings requirements, create both costs and constraints for government managers (Cooper 1996). Yet legal requirements can also restrict contracting practices (DeHoog 1997), such as those that require some percentage of contracts to be awarded to female- or minority-owned firms. Indeed, private firms and nonprofits sometimes avoid competing for public sector contracts because of time-consuming and procedurally complex legal requirements
- 10. 9 (MacManus, 1991; Praeger, 1994). Thus, legal exigencies can lower market competitiveness, thereby diluting the advantages of contracting relative to in-house service delivery. This too then is a major constraint on a competitive market. Proposition: The absence of even an imperfect market and legal complications compound the balance in the principal-agent relation to tilt toward the agent even as transaction costs continue to rise for the principal without any concomitant rise in 3E. C. Is there a Market Regulator? A market presupposes the existence of an effective regulator. Agents have competing interests; indeed the interests of some agents may be more congruent with those of their principals than with the other agents. Some agents are more risk averse than others; incentives work differently on different agents. Some agents may also be free riders. And the existence of multiple principals and multiple agents sometimes increases the informational asymmetries and the difficulties of monitoring. These asymmetries are among the reasons organizational crimes can flourish undiscovered for long periods of time buried in complex structures of action (Shapiro, 267). At other times, multiple parties help to right the imbalance of information, such as when competitive agents leak information to principals in an effort to get an upper hand over other agents (Waterman & Meier 1998). Even if it is assumed that a regulated limited market may be the mid-way answer to this vexatious problem, this would presuppose the existence of a set of regulators with the capability to support and sustain a market. In New York City, government delivers services to homeless individuals and families through shelters run by government and nonprofit contractors. While private firms such as Lockheed, Maximus and America Works provide welfare services, rising costs are engendering competition among these and newer players that are leading to sound business management by these nonprofits (Ryan, 127-36). In Indiananpolis a contractor manages a large share of Welfare-to-Work programs while in New York City, Homes for the Homeless delivers efficient services to homeless families. These agencies invest surpluses into infrastructure for the homeless and are seldom criticized for profiteering (Cohen, 436). Cohen attributes this to the ability of nonprofits to recruit motivated staff without creating an expensive bureaucracy. A major reason for the absence of private companies in the sphere of social services delivery is the need for customer subsidy and political sensitivity to poverty profiteering. Cohen makes a compelling case for privatizing the Emergency Assistance Unit of New York City’s department of Homeless Services primarily for reducing the
- 11. 10 rate of rejections of homeless people since nonprofits are seen to be more caring and motivated than government staff in accommodating such people apart from the reduced cost of delivery of service (Cohen, 439). Brown et al state that nonprofit organizations are thought to share similar missions with government, and thus may be more reliable contract partners (Hansman 1987; Salamon 1995). Rather than behaving opportunistically, a nonprofit might draw on its own private philanthropic resources (e.g., volunteers and endowments) to augment services it delivers under government contract. Yet nonprofits’ goals may not always be aligned with public objectives and may instead channel residual revenue from contracts into subsidizing their other programs. Nonprofits, in turn, are regulated as tax-exempt organizations and, as such, are prohibited from distributing profits to their employees or volunteer boards. Consequently, there may be fewer incentives for them to engage in opportunistic behavior, at least in comparison to private firms. Then why is there a need for a market regulator? Once agencies select a vendor and turn their attention to contract implementation, managers face more decisions regarding trade-offs among public, institutional, and service-market values. Perhaps the most central among these involves monitoring and evaluating the performance of vendors working under contract (Kettl 1993; Praeger 1994). Well-monitored vendors are more likely to perform according to contract specifications, thereby improving returns from contracting, whereas legal, institutional, and service-market constraints can increase transaction costs (Brown and Potoski 2003a). These monitoring activities, of course, vary in their costs and efficacy depending on the nature of the service and existing service-market conditions. For example, in cases in which governments face vendor opportunism stemming from difficult measurement, managers cannot simply purchase and monitor service outcomes alone. Public managers therefore have to be fully conversant with the service process. Sophisticated contract management therefore presupposes the existence of highly-skilled contract managers to counter contractor opportunism at all levels. Analyzing the causes of the accident of the space shuttle Columbia, Alasdair Roberts says that by 2002, downsizing of NASA’s regular workforce created a situation in which contract workers outnumbered civil servants by nine to one (Roberts, 627). With experienced engineers superannuating/leaving, NASA became increasingly reliant on an inexperienced engineer workforce provided by United Space Alliance, the contractor. Thus NASA had no “constructive and
- 12. 11 reinforcing” redundancy left to deal with uncertainty in its system (Landau, 356). Van Slyke’s findings are not dissimilar. Recruitment being a lengthy and slow process, an agency finds it less cumbersome to claim budgets for services than for new regular staff (Van Slyke, 305). Secondly, public hostility to expansion of government makes recruitment a touchy subject and dissuades agencies from filling up even mission-critical vacancies, including contract managers. Politicians too back public sentiments for electoral reasons for which reason legislative support for new posts or filling vacancies is often not available (Van Slyke, 305). This, in turn, causes the agencies to do without adequate contract administering personnel to develop detailed requests for proposals, evaluate bids, award contracts and provide technical assistance to contractors and evaluating performance outcomes. In one case Van Slyke quotes a staff of 12 manages more than 500 providers (Van Slyke, 306). Although reporting requirements increase, there is a corresponding increase in program administration costs. The absence of qualified staff leaves the agency vulnerable to contractor opportunism. The compounding effect of staff shortage also shows in loss of institutional memory and policy expertise (Van Slyke, 306). The nonprofits therefore dictate policy formulation. Similarly, the loss of staff without replenishment whittles down institutional knowledge that is required to be a smart buyer of goods and services. The loss of staff also deprives the agencies of any calculated slack that they have to deal with emergencies; contractors being generally risk- averse maintain the bare minimum staff and have no surplus capacity to deal with emergencies (Van Slyke, 307). In specifying a contract, public managers decide on and implement a bid process, select a vendor, and craft contract terms. Within the contours of legal requirements, public managers typically have discretion in contracts to specify vendor tasks (e.g., the nature and scope of work), outcome measures (e.g., performance-based criteria), vendor qualifications (e.g., licensing or accreditation issues), vendor compensation (e.g., time and materials versus cost plus fee), payment schedules, contract duration (e.g., short versus long), incentives and sanctions (e.g., rewards versus punishments), renewal provisions, reporting requirements. Vendors may favor reducing expenses, particularly if this means pursuing their own (profit) goals at the expense of the government’s objectives. Shapiro (267) says the assumption that principals are in the driver’s seat—specifying preferences, creating incentives, and making contracts that agents must follow—is also problematic (Heimer & Staffen 1998, Sharma 1997). When principals seek out agents for their expert knowledge, when principals are ‘one-shotters’ and agents ‘repeat players’, when principals are unexpectedly
- 13. 12 foisted into a new role with no time or life experience to formulate preferences, let alone a contract or monitoring strategy [e.g., the new parents of a critically ill newborn (Heimer & Staffen 1998)], the asymmetry of power shifts from the principal to the agent. The absence of institutional memory and professional capacity in a public agency can only exacerbate this shift of power. The increasingly contested political environment of contracting can produce overly rigid and excessive regulation of contract agencies (Smith and Lipsky 1993; Gutch 1992; Bernstein 1991). Service providers are asked to submit large amounts of information on their operations, even though government agencies do not have the capacity for reviewing it. Further, some of the requested information is not germane to evaluating program quality. For many nonprofit agencies, especially smaller community organizations, programmatic regulations can lead to major changes in mission, services, and clients. Service costs also rise because compliance with these regulations requires contract agencies to add administrative staff and develop new internal monitoring systems. Without the requisite institutional knowledge, public agencies are unable to enforce their rights of a principal and remain hostage to contractors’ opportunism without adequate evaluation and enforcement capacity. Proposition: Even if a competitive market were to exist, the limited contract management capability of public agencies would continue to tilt the principal-agent balance in favor of the agent, since there would be no effective statutory regulator; consequently, the principal’s transaction costs would continue to rise without any concomitant rise in 3E and accountability of contractors’ services. D. Regulation of Contractors Are the internal controls in contracted agencies, upon which evaluation information is provided to public agencies, adequate? Dicke studied a state agency on disabilities (AOD) and its contracted private providers for the delivery of residential services to AOD clients in a single western state (502-16). The 12 provider organizations in the study received more than75%of the AOD’s contracted services budget for the fiscal year of 2000, and each provider organization received on average more than 95% of its residential service revenue either directly or indirectly from a government source. AOD case managers reported that monitoring was overwhelmingly the most helpful review procedure used by their organizations to ensure quality in services, followed by outcomes-based assessments (52%), auditing (31%), and provider self reports (14%), and other reviews (21%). However, on the overall efficacy of these methods, 29% of respondents felt that they were “just about right,” 6% said they were “too strict and required too much paperwork,” 13%
- 14. 13 offered “some other response,” and 52% said they were “not adequate to oversee performance.” The percentage of respondents who noted inadequacies in the methods was lowest among case managers with less than 12 months of service (25%) and highest— more than 54%—among those with more than 3 years of service. Surprisingly, only 2% of AOD case managers reported no difficulties in evaluating provider performances using the current accountability methods. A total of 58% questioned the validity of the information obtained from monitoring, the most useful method cited by them primarily because quality assessments were based on infrequent visits (18%) and evaluations were based on snapshot depictions (40%) of on-site conditions that could be manipulated by providers. An additional 15% of respondents reported problems due to time constraints and caseload size, which caused a decline in the number of monitoring visits made and the time spent on-site. The veracity of the information gained through audits and provider self reports was doubted by 21%. Staff members in 40% of provider agencies reported they had personally been asked to prepare bogus documents by their supervisors or witnessed others doing so. Several provider employees noted that inadequate training, incomplete documentation and lack were other reasons. Dicke says that other problems relate to the absence of a baseline to assess quality improvement or decline and a rapid variation in program quality attributed to staff turnover (461). Case managers noted that positive findings of quality may be inspite of services provided by poor residential services providers. Some case managers indicated they were “not familiar enough” with the tools used by their agency for ensuring quality. A lack of political will or support from hierarchical superiors also reduced the effectiveness of external controls. For example, Dicke says “controls are compromised when problems are uncovered, but the authority to execute sanctions is lacking or political support for their use is absent.” (Dicke, 462) The overriding consensus on social services contracting in state and local government is that oversight is virtually nonexistent due to a lack of time and resources (Kettl, 1993; Rehfuss, 1989). Kettl found that state and local government agencies typically devote new staff members or other resources. In Massachusetts, oversight was so neglected that “the state embarrassed itself by paying contractors who delivered no services at all” (Kettl, 1993, 175). Similar findings in other states have been cited by DeHoog (1984) and Milward, Provan, and Smith (1994). Most AOD case managers reported meeting face-to-face with their clients monthly or less than monthly (55%) (Dicke, 462).
- 15. 14 Only 23% reported meeting with clients weekly or more frequently (Dicke, 462). Evidently, the limited capacity of the principal to check cheating by the agent is often subverted by the agents. Measurement of performance thus no longer remains a reliable measure of the quality/quantity of work done by a contractor. Proposition: Weak contract management capability of public agencies would be compounded by equally or weaker internal control systems of contractors, both by act of omission or commission; accountability would be adversely affected and there is unlikely to be any gain on 3E of services rendered and accountability therefor, despite transaction costs rising for both the principal and the agent. IV. Role of Environmental Uncertainty/Turbulence The limited assumptions in public administration literature of the principal-agent theory of information asymmetry and goal conflict between principals and agents appear simplistic. While moral hazards and adverse selection are two fallouts of a principal-agent relationship, yet the multiplicity of principals and agents in a market may cause unlikely alliances to build up between principals and agents. Both principals and agents often essay dual roles. While awarding contracts, a public agency is the principal but while evaluating on behalf of a legislature, the same agency is transformed into an agent of the legislature. Similarly, contractors are not always agents. While discharging their agency role by providing services they are agents; but when recruiting staff they are principals. Confounding these complex networks are an entire regime of incentives and punishments in myriad of permutations and combinations that move risk from a principal to an agent along a sliding scale. The lower the risk, the lower is the propensity of the principal to transfer its functions to the agent. The higher the risk, the higher is the propensity for the agent to shirk. Thus an extension of the traditional limited principal-agent model to arrive at a more equitable solution for both principal and agent, appears to be the most viable option. Romzek and Johnston studied five cases of contracts in Kansas for selected welfare (HCBS), Medicaid and foster care and adoption services (FCA) (423-53). Their most important finding related to the sharing of risk between the government and the contractors in high uncertainty situations such as those for the above human services (Romzek & Johnson, 443). With frequent changes in contract terms on account of a dynamic environment, contractors were apprehensive of sustaining losses if the entire risk was transferred to them. The Kansas government shifted the financial risk associated with service delivery on an agency-by-agency basis and proportionate to the
- 16. 15 environmental uncertainty. While for the EPS the state retained the financial risk, for the MMC and FCA programs, the state shifted significant levels of risk to the contractors (Romzek & Johnson, 443). Where the state shifted the risk to the contractors in the case of MMC and FCA, the contractors faced major financial problems. Evidently, the level of environmental uncertainty conditions the extent of shirking by the agent at macro and micro levels of analysis. Dicke and Ott identify ten instruments for enforcing contractual obligations that are presently used (506-7). They conclude that in human services contracting, what is required is a match between the levels of explicit and implicit accountability, rather than the existing bureaucratic hierarchical, legal, professional and moral control standards. Thus the distinction between accountability and control has been obfuscated (Dicke & Ott, 511-12). The flexibility required to deal with human services is not available with the present controls (Dicke & Ott, 512). They advocate Outcome Based Assessments (OBA) for program evaluation that “attempts to determine or verify accountability by measuring and evaluating the end results of government programs rather than monitoring inputs and processes.” (Dicke & Ott, 507) OBA treats contracts as treaties that “provide mechanisms for the parties to adapt to a changing environment, communicate altered expectations, and modify their behaviors based on their learning-on approach favorable for both flexibility and innovation (Dicke & Ott, 513). What is therefore required is a proactive and open relation between the principal and agent that would create trust between them and extend the scope of the contract beyond the written word. In sum, what is being suggested by implication is a cooperative model rather than an adversarial one. Smith suggests that a cooperative model is already in vogue. He states that: As the contracting regime develops, ……………..the unpredictability of public organizations is reduced through the norms and personal ties of the contracting regime…………. The norms guiding these networks can provide a way for both government and provider to lower the transaction costs of contracting and cope with the vagaries of the political world and the market for contract services. (Smith, 117) Smith gives another interesting interpretation of the emerging cooperative networks between principals and agents. He states that in order to build political support for their programs, public agencies, directly or indirectly build, political associations of contractors. Governments have been sponsoring formation of interest groups in the last twenty years (Walker 1991; Salisbury 1990). Government purchasing agencies also occasionally provide seed money for the establishment of these political associations While Federal and state purchasing agencies are prohibited from lobbying legislators, these associations are permitted to lobby legislators directly for more money or
- 17. 16 new contract programs (Smith, 119). Agencies also release grants to research and consulting institutions to certify a government position and help in building political support for programs (Smith, 120) Other instrumentalities of overcoming uncertainty are modification of regulations, new funding; re-bidding contracts; and re-writing a contract on terms favorable to the provider. However, this need for compromise to build or maintain political support explains why many regulations governing contract agencies work against effective performance. For example, one federal program prevents its agencies from raising money with staff paid with federal funds. This regulation is a concession to some interest groups and members of Congress concerned with the spending by contract agencies. However, this regulation is very destabilizing to contract agencies which need to raise outside resources in order to compensate for reduced federal funding (Smith, 120). Agents thus also attempt to rewrite the rules and control the uncertainty in the environment by lobbying for legislation or impose/remove contract conditions in some of which they are aided by the public agencies themselves. This, in turn, widens both the definition and numbers of principals and agents and sometimes creates a cooperative environment characterized by convergence of interest, at least in the short-term. Smith’s view agrees with the tenor of Romzek & Johnson and the regulation of uncertainty in the race to achieve an optimal fit between the principal and agent thus comes to center stage. Both the principal and agent therefore have vested interests in controlling for uncertainty as they strive to achieve their respective optimal fit. There are certain elements in the operating environment that breed uncertainty. In Herbert Simon‘s behavioral model an organization simplifies its options according to its computational capacity in arriving at an optimal benefit decision (Simon, 467-82). The organization considers the alternatives, its own objectives and then arrives at the optimal pay-off by ordering the available options – it uses information available to it while weighing its capacity against environmental factors (“manageable proportions”). An organization has an “aspirational” level to reach which it gathers information and then ranks the options open to it vis-à-vis a win, draw or lose final result. Thus no organization adopts “global” rationality; rather stick to “limited” rationality while also considering the future prospects of its current optimal decision. Applied to contractors and public agencies, both being viewed as organizations, each of them would attempt to arrive at an optimally beneficial (to them) decision. In arriving at optimal decisions, the principal and agent would rely on prices, competition and contracts as governance mechanisms (Williamson 548-77). In sum, the principal and agent would play out the game in a market place and arrive at an optimal decision, relying, inter
- 18. 17 alia, on prices and competition for contracts and the confounding factors mentioned elsewhere in this paper. However, the market for public services contracts being ill-defined or even non-existent, would prices and competition be the sole criterion for determining the relationship between the principal and agent? Astley’s community ecology approach conceptualizes population forms in terms of their functional roles vis-à-vis other populations within technologically interdependent communities (Astley, 224-41) – in this case the public agencies and the contractors. With downsizing in government, the agencies have had, per force, to rely on contractors to deliver services to citizens. This has resulted in the creation of demand for services and the emergence of a limited or quasi- market for services. There is thus an apparent connect between two communities – agencies and contractors, each of which has to survive on a common reservoir of ideas that binds them together, albeit in a principal-agent relationship. As organizations (contractors) diversify into new areas (niches) they establish new populations. New organizational mutations evolve from new technology and seek environmental space that may require only a “tolerable fit” rather than “optimal fit”. Translated in terms of the principal-agent perspective, this effectively implies that the tolerable fit is conditioned by the level of environmental uncertainty, i.e. budgets, govt. regulation, political and ideological differences, risk-sharing, nature of functions, management capabilities of agencies, and other confounding factors, some of which may be peculiar to certain categories of services/contracts, and sundry other key independent variables (that researchers may identify subsequently) that affect such a fit (the end result is to achieve 3E). Any research design should therefore factor these as independent variables and derive conclusions about the ‘tolerable fit’. Proposition: Environmental turbulence created partly by the above discussed confounding factors, in addition to many others, and the need to minimize transaction costs, determines the level of cooperation between principal and agent; higher the uncertainty, higher should be the level of cooperation between principal and agent, leading, in turn, to more cooperative contracting models and lower transaction costs. V. Research Design It is therefore apparent from the foregoing discussion that 3E contracting of public services and their accountability is specific to a context. Evidently, not all public services make for 3E and accountability of outsourcing. Thus there exist major gaps in the identification and measurement of independent and control variables used in research on 3E and accountability of outsourcing. The
- 19. 18 transposing of private sector principles that carry with them key economic concepts such as the existence of markets and perfect competition, profit making motive and all-powerful and non- transparent Boards of Directors do not seem appropriate to standardized public services with equally standardized populations whose preferences change, if at all, over a long span of time. Such services are also often politically volatile and thus generate pressures from citizens and politicians alike, each for their different reasons. Legal and government departmental limitations and the absence of general operational autonomy confound the principal-agent relationship further. This makes accurate measurement of 3E and accountability of outsourcing of public services a much more complex exercise. Therefore the key to any meaningful study of the benefits or otherwise of outsourcing hinges on identification of similar contexts, populations, geographical areas, etc., definition of independent variable boundaries with reference to standardized measures and measure their constituent control variables accurately and uniformly. Unless the control variables for each independent variable are closely analyzed (as Romzek & Johnson have attempted to do) by professional study, such erroneous and even ‘loaded’ conclusions about outsourcing would remain common occurrences and thereby severely undermine the credibility of research findings. Understandably, from a methodological perspective, arriving at a research model would indeed be a complex exercise, partly on account of being able to correctly identify the constituent control variables for each independent variable and devise uniform yardsticks for measuring them. A major challenge lies in defining the range and scope of each control variable, of which there would be profusion for each independent variable. This proposal therefore, in its final form, would therefore use a mix of qualitative and quantitative research on data sourced from similar agencies spread over states (selecting 3-5 counties in each state) and then grouping them into geographical regions rounded off by inter-regional comparison for level of analysis. The sample choice and size of counties would be determined with reference to similar population characteristics and size, availability of similar public agencies, extent of contracting, problems of public agencies in discharging this role, etc. while determining the target sample. Thus, for instance, Wake, Durham and Chatham counties, each with a hypothetical population of greater than 200,000 that have contracted child and/or geriatric care/welfare services
- 20. 19 (field selected for its relatively high uncertainty), downsized their City Council managers and personnel from 1998-2007, incurred similar expenditure in each of the five years, etc. Three other programs that have been outsourced but with varying levels of environmental uncertainty may also be chosen for comparative purposes. Going further, the design would source its quantitative data from complaints received from residents; consumer satisfaction surveys; excess payments made but not recovered from contractors; fuel consumption by dumpers; adverse inspection reports of landfills, short deployment of labor, misuse of incentives, results of repeat tenders for contractors, etc. Following this, the proposal would analyze the quantitative data comparing each sub-category across counties and all states in the SE United States within the broad framework of the confounding factors (independent variables) shown in Fig. 1. The level of analysis could be further augmented to inter-regional comparisons and finally national trends, using the same variables. A. Government to Governance Model Fig. 1 shows the research design to be adopted by this proposal. The design is explained in the paragraphs following Fig.1. Fig. 1 Government to Governance Model (L-R boxes in Fig. 1 are independent variables) Extent of Competition C Incentives & Penalties C O A Downsizing/Contract O N D Local Ideological Preferences Management Capability of O T V Public Agencies P R E E A R R C S 3E of Outsourced Service vis-à- Quality of A T A Outsourced Service vis Govt. provider T R I T I Levels of Accountability – V Y A Extent of Risk-sharing Explicit & Implicit E P L between Principal & Agent E 8 14 20 26 32 Control-Adversarial-Government Cooperative-Accountability-Governance (High) TRANSACTION COSTS (Low)
- 21. 20 To elaborate further, each independent variable shown above would have an overall score ranging from 1 to 4. Using a four-point Likert scale (corresponding to four ranges 8-13, 14-20, 21-25 and 26-32 on the Fig. 1 scale) the sum of the variables in a contract would range from 8 to 32 (maximum score of 4 per variable for 8 variables, 4 being most cooperative and 1 being most adversarial). The more the score is towards 8, the more adversarial and control-based the principal- agent relation would be. Administering such contracts would be difficult as the propensity of the agent to shirk would increase, that would ultimately require ever-increasing voluminous contracts and rising transaction costs. Matched by the overall decline in contract management capacity of public agencies over the last decade (upon downsizing), contract management would favor the agent more than the principal as Boyne has pointed out elsewhere in this proposal. This, in turn, may not only give rise to subjugation of the public interest to the agent’s advantage but also result in provision of services the 3E of which are doubtful. Similarly, the more the score moves toward 32, the resultant cooperation between the principal and the agent, the more accountable the agent would become and the propensity to shirk would decline, making for more cooperative contract management. In sum, the higher the level of uncertainty/environmental turbulence; higher the cooperation between the principal and agent, in the process minimizing transaction costs and asset specificity. What the scale essentially measures is the balance in the principal-agent relationship and consequently its impact on the quality of the contractual relationship – a score toward 8 loads the dice against the principal and in favor of the agent while a score toward 32 is vice-versa. Thus in a control-cooperation model, government control may move toward 8 on the above scale and toward 32 as the level of involvement of the agent in sharing of risks, costs, resources and information asymmetry, etc. increases. This may be reflected adversely in the 3Es of service. It is not the intent of this proposal to promote a model of pure cooperation without control considering the large involvement of public funds. The ultimate objective is to derive a standard typology of contracts appropriate to the context of application that promotes 3E with accountability in the individual and collective interests of both the principal and the agent. In effect, in the scale at Fig. 1 ratings from 8-16 may be deemed to be adversarial while those from 20-32 may be expected to fall in the cooperative range of the same scale. By this classification, Type I and II contracts would be mainly cooperative while Types III and IV would be mainly adversarial. It is also possible that a
- 22. 21 principal-agent relationship may have elements of both adversarial and cooperation that this design factors in by way of the control variables. B. Typology Matrix A typology matrix is proposed at TURBULENCE/UNCERTAINTY Fig. 2 to dovetail the scores from Fig. 1 into categories of contracts HIGH LOW C as shown in Fig. 2. The scale in O H I II O I Fig. 1 is divided into four equal P G 26-32 21-25 E H ranges as shown in Fig. 2 to derive R four categories of contracts (I, II, A T III IV L III & IV). Based on the typology I O O 8-13 14-20 matrix, sorting the categories of W N contracts in projects (by Fig. 2 Typology Matrix agency/object) would generate conclusions. For instance, road maintenance or a state motor vehicle repair contract may figure in Type IV whereas a drug de- addiction service may figure in Type I or II. Type III may include a contract that provides for emergency import and distribution of food to tide over a drought/flood situation but without the required degree of coordination between the principal and agent. Equally, an executive order raising road maintenance contractors’ pre-qualification annual turnover for a national highway construction tender to, say $250,000 from $100,000 may cause a shift from Type I to Type III and a corresponding shift toward 8 on the scale in Fig. 1 as the risk is shifted to the agent; the converse would also be true. This typology would therefore clearly bring out the appropriateness of control and cooperation models specific to fields of their operation. For instance, while road maintenance which is a routine work may require routine regulation since the uncertainty level is low, child care services may require more cooperation because of higher environmental uncertainty/turbulence as Romzek and Johnson have shown in the case in Kansas. C. Operationalizing the Independent Variables Due to space constraints, it is not possible to suggest operationalization parameters for all the variables except for four; researchers may quantify the balance in the light of the literature cited in
- 23. 22 this paper. Nor is it possible to create specific control variables without physically reviewing specific categories of contracts and related documents and deriving a list of variables therefrom. For purely illustrative purposes let us assume the researcher wishes to operationalize the extent of risk-sharing between principal and agent. For this purpose, the researcher’s primary sources of information would be the financial and legal aspects of contract documents such as the extent of reimbursement of expenses, risk-sharing clauses, speed of reimbursement of costs by the agency, extent of damages payable for change in contract design or any other set of parameters that show the financial interaction between the principal and agent, outward and inward flow of reimbursement bills to and from the contractor, number of complaints lodged by the contractor with the agency for delayed or non-payment, etc. Each of these criteria would be marked on a Likert scale of 1 to 4 as shown in Fig. 1 and the total weighted score for this independent variable worked out from 1-4 (the total score shall not exceed 4 for any independent variable; thus each control variable may have to be appropriately weighted). Thus reimbursement of certain expenses by the public agency to the contractor that takes more than 30 days after the contractor prefers their bills to the agency may be assigned a weight of 0.25 while a period of less than 7 days may be assigned 1, per the judgment of the researcher who would have to decide his own yardsticks of scoring. In other words, at the micro- level also the above model may also be used to predict agent behavior in conditions of high/low uncertainty. Aggregated, over a series of similar agencies, the model would pinpoint the quality of supervision and communication between the agency as principal and contractor as agent. Again, aggregated over, say 50 counties in S-E USA, this may provide important leads in the processes of award and quality of management of contracts peculiar to processes, functional sectors and geographical areas. Similarly, for operationalizing the extent of competition, the researcher would have to go through the entire process of tendering. The extent of publicity given to the notice inviting interest (number and geographical and industry spread of advertisements, etc.) followed by the financial and experience norms in short listing contractors for empanelment, differences in financial bids, actual award of the contract and terms and conditions (including incentives, penalties and reporting/oversight arrangements), comparison of retender and new bids, would have to be examined as control variables and slotted into the Likert scale before working out an overall score for this independent variable as a whole ranging from 0-4. Aggregated over functional sectors and
- 24. 23 geographical areas, this variable may throw up leads on irregular agency practices, contractor cartels, pricing patterns (to judge competition) and the extent of risk-sharing, etc. In operationalizing yet another variable – ideological preferences – the researcher may choose to create a local typology of service contracts from 1998-2007 and analyze it with electoral changes in local government or even changes in voting patterns at local and state level and changes in major electoral issues raised by political parties before elections, pattern of legislation corresponding to electoral promises, etc. as control variables. A major consideration here would be to analyze the continuation of contracts entered into by previous local governments and the results on the target clientele – this would bring out whether contracts were continuing without any demand for them, indicative of extraneous considerations in a control-government model. Operationalizing incentives and penalties would start from the financial and administrative details of such clauses in such a manner that such measures regulate environmental uncertainty at the micro-level. In other words, a contract clause providing for a cash incentive of $50 per waste bin cleared per day (over minimum contracted quantity) provides an opportunity to reduce the financial risk of the contractor. However, the same incentive, if related to the number of trips made by dumpsters could easily lend itself to cheating by contractors. While the former incentive reduces the propensity for the contractor to cheat, it also moves the quality of the contract towards 32 on the government-governance scale. However, the latter incentive would prompt more controls to be imposed by the principal resulting in the scale sliding toward 8. Viewed in the matrix at Fig. 2, at the micro-level the type would move from Type II to Type III or IV. D. Applicability of Model Since this design would invariably involve decomposing the contents of each variable, it would also make for useful comparison at the institutional and local political level. It may be argued that a study of this magnitude would be expensive to sustain and require several years to complete. Nonetheless, the flexibility of this design would permit comparison by different sets of researchers at unit, local, regional, inter-regional, inter-state and national levels over a span of time by providing a standardized framework so that the findings can be related again at all levels using identical benchmarks shown in Fig.1. This would obviate any apprehensions of a long drawn out study, since it would be open to researchers to define their level of analysis at different points in time. The above model (Figs 1 & 2) would thus serve a wide range of contracts/services since it provides for
- 25. 24 flexibility to add/subtract independent variables without substantially modifying the scoring method and the matrix that would assure uniformity of treatment of subject contracts within a standard framework. However, this model, like all others, would necessarily depend on a representative sample as indicated elsewhere in this paper for accuracy of results, unlike the empirical studies reviewed by Boyne (474-82). VI. Directions for Future Research The suggested research design would involve a multi-disciplinary approach wherein organizational theorists would share the dais with accountants, lawyers, contract managers, engineers and other professionals without whose active involvement it would be difficult to conduct such a technical study and avoid arriving at conclusions the same way studies analyzed by Boyne and others have. A standard research design with flexibility of application at multiple levels would enable more systemic reviews to be carried out with relatively less normative standards and applied within a standard one–size-fits-all framework (and standard independent variables). The benefits or of outsourcing or otherwise remain virtually unknown as yet. Such a study would also systematically relate theory to practice and prove to be useful for public managers and legislators and also perhaps attract funding for studies. Considering the large federal budget outlays on contracts for diverse purposes, from defense to social welfare and homeland security, such a study (ies) may indeed be worth the moneys and research effort expended. The gap between the true benefits or loss of 3E by outsourcing remain mainly a matter of academic conjecture while outsourcing itself remains a fad without any established credentials as a preferred method of governance. Government may have physically shrunk but outsourcing has extended government vastly, although not necessarily in the direction of good governance.
- 26. 25 VII. REFERENCES Artz , Kendall W . , and Thomas H Brush (2000): Asset Specificity, Uncertainty and Relational Norms: An Examination of Coordination Costs in Collaborative Strategic Alliances . Journal of Economic Behavior and Organization 41 (4 ): 337 – 62 . Astley, WG (1985): The Two Ecologies: Population and Community Perspectives on Organizational Evolution. Administrative Science Quarterly, 30, 224-41. Boyne, George A (1998): Bureaucratic Theory Meets Reality: Public Choice and Service Contracting in US Local Government. Public Administration review, Nov/Dec 1998, Vol. 58, No. 6 Brown, Trevor L , and Matthew Potoski (2003): 2003b . Managing Contract Performance: A Transaction Cost Approach. Journal of Policy Analysis and Management 22 ( 2 ): 275 – 97 . ___________________________________________ (2003b): Managing Contract Performance: A Transaction Cost Approach . Journal of Policy Analysis and Management 22 (2): 275 – 97. _____________________________________________(2004): Managing the Public Service Market . Public Administration Review 64 (6): 656 – 68 . ____________________________________________ (2006): Contracting for Management: Assessing Management Capacity under Alternative Service Delivery Arrangements. Journal of Policy Analysis and Management 25(2): 323–46. Brown, Trevor L , Matthew Potoski & David M. Van Slyke (2006): Managing Public Service Contracts: Aligning Values, Institutions, and Markets Public Administration Review, May-June 2006, 323-31 Coase , Ronald . 1937 . The Nature of the Firm. Economica 4 : 386 – 405 . Cohen, Steven (2001): A Strategic Framework for Devolving Responsibility and Functions from the Government to the Private Sector. Public Administration review, July-Aug 2001 Vol. 61 No. 4 Dicke, Lisa A.: Ensuring Accountability in Human Services Contracting: Can Stewardship Theory Fill the Bill? The American Review of Public Administration 2002; 32; 455 Dicke, Lisa A & J Steven Ott (1999): Public Agency Accountability in Human Services Contracting. Public Productivity & Management Review, Vol. 22, No. 4. (Jun., 1999), pp. 502-516. eLeon , Peter (1995): Democratic Values and the Policy Sciences . American Journal of Political Science 39 : 886 – 905 . Hansman , Henry (1987): The Role of Nonprofit Enterprise. Yale Law Journal 89 (5): 835 – 901. Hart, Oliver , and Jonathan Moore (1999) . Foundations of Incomplete Contracts. Review of Economic Studies 66: 115 – 38 . Heimer CA, Staffen LR. 1998. For the Sake of the Children: The Social Organization of Responsibility in the Hospital and the Home. Chicago: Univ. Chicago Press in Shapiro, Susan P.: Agency Theory. Annual Review of Sociology 2005: 31: 263-84. Kettl, D. (1993). Sharing power. Washington, DC: Brookings Institution. Morgan , David R . , and Robert E England (1988): The Two Faces of Privatization . Public Administration Review 48 (4): 979 – 86 . Niskanen, William A (1971): Bureaucracy and Representative Government. Chicago: Aldine Atherton . Rehfuss, J. A. (1989). Contracting Out in Government. San Francisco: Jossey-Bass.
- 27. 26 Cooper , Phillip J . 1996 . Understanding What the Law Says about Public Responsibility. In Handbook of Public Administration, edited by James L . Perry , 115 – 35 . San Francisco: Jossey-Bass . DeHoog, R. H. (1984). Contracting out for human services: Economic, Political and Organizational Perspectives. Albany: State University of New York Press. _____________ (1997): Legal Issues in Contracting for Public Service: When Business Does Government . In The Handbook of Public Law and Public Administration , edited by Phil Cooper and Chester Newland , 528 – 45 . San Francisco : Jossey-Bass . Donahue , John D (1989): The Privatization Decision: Public Ends, Private Means. New York :Basic Books . Hart , Oliver , and Jonathan Moore (1999): Foundations of Incomplete Contracts. Review of Economic Studies 66 : 115 – 38 . Hart , Oliver , Andrei Shleifer , and Robert W . Vishny (1997): The Proper Scope of Government: Theory and an Application to Prisons. Quarterly Journal of Economics 112 (4): 1126 – 61 . Levin, Jonathan (2003): Relational Incentive Contracts. American Economic Review 93 (3): 835 – 57 MacManus, Susan (1991): Why Businesses Are Reluctant to Sell to Governments. Public Administration Review 51 (4): 328 – 43 . McCubbins , Matthew , and Thomas Schwartz (1984): Congressional Oversight Overlooked: Police Patrols versus Fire Alarms. American Journal of Political Science 28: 165 – 79. Milward, H. B., Provan, K., & Smith, L. J. (1994): Human service contracting and coordination: The market for mental health services. In J. Perry (Ed.), Research in Public Administration (Vol. 3, pp. 231-279). Greenwich, CT: JAI. Moe , Ronald C (1996): Managing Privatization: A New Challenge to Public Administration. In Agenda for Excellence 2: Administering the State, edited by B . Guy Peters and Bert A Rockman. Chatham, NJ : Chatham House . Praeger , Jonas (1994): Contracting Out Government Services: Lessons from the Private Sector . Public Administration Review 54 (2): 176 – 84 . Romzek , Barbara S . , and Jocelyn M . Johnston (2002): Effective Contract Implementation and Management: A Preliminary Model . Journal of Public Administration Research and Theory 12 ( 3 ): 423 – 53 . Ryan, William, P (1994): The New Landscape for Nonprofits. Harvard Business Review 77: 127-36 Salamon , Lester M (1995): Partners in Public Service: Government-Nonprofit Relations in the Modern Welfare State . Baltimore: Johns Hopkins University Press. Sclar , Elliott D (2000): You Don’t Always Get What You Pay For: The Economics of Privatization. Ithaca, NY: Cornell University Press. Seidman , Harold (1998): Politics, Position, and Power: The Dynamics of Federal Organizations. 5th Ed. New York: Oxford University Press. Shapiro, Susan P.: Agency Theory. Annual Review of Sociology 2005: 31: 263-84. Simon, HA (1962): The Architecture of Complexity. Proceedings of the American Philosophical Society, 106(6): 467-82 Smith, Steven R (1996): Transforming Public Services: Contracting for Social and Health Services in the US Public Administration Vol. 74, Spring 1996 (113-127) Stoker, Gerry (1998): Governance in Theory – Five Propositions. ISSJ 155/1998 UNESCO, 1998
- 28. 27 Swindell , David , and Janet M . Kelly (2000): Linking Citizen Satisfaction Data to Performance Measures: A Preliminary Examination . Public Performance and Management Review 24 ( 1 ): 30 – 52 . Van Slyke , David M (2003): The Mythology of Privatization in Contracting for Social Services . Public Administration Review 63 ( 3 ): 277 – 96 . Tadelis , Steven (2002): Complexity, Flexibility, and the Make-or-Buy Decision . American Economic Review Papers and Proceedings 92 ( 2 ): 433 – 37 Van Slyke , David M . , Christopher S . Horne , and John C . Thomas (2005): The Implications of Public Opinion for Public Managers: The Case of Charitable Choice . Administration & Society 37 ( 3 ):321 – 44 . Waterman RW, Meier KJ (1998): Principal-agent models: an expansion? J. Public Admin. Res. Theory 8: 173–202 Williamson, Oliver (1981): The Economics of Organization. American Journal of Sociology, 87: 548 – 77. ___________________(1991): Comparative Economic Organization: The Analysis of Discrete Structural Alternatives . Administration Science Quarterly 36: 269 – 96. __________________ (1996): The Mechanisms of Governance. New York: Oxford University Press.