CASE Network Report 86 - Income and Distribution Effects of Migration and Remittances: an Analysis Based on CGE Models for Selected CIS Countries

The views and opinions expressed here reflect the authors’ point of view and
not necessarily those of CASE Network.
This paper constitutes Deliverable 18 of the ENEPO research project supported
by the European Union’s 6th Framework Program (EU Eastern Neighbourhood:
Economic Potential and Future Development - http://enepo.case.com.pl/). This
paper summarizes the findings of ENEPO Work Package 8: Analysis of Outward
Migration in Selected CIS Countries coordinated by Kiel Institute for the World
Economy. The country case studies in this paper are authored by the respective
country teams as indicated in each section. The overview sections have been writ-ten
by Matthias Luecke with input from all participants. Without implication, the
authors thank Marek Dabrowski and Maryla Maliszewska for their insightful and
detailed comments. The content of this publication is the sole responsibility of the
authors and can in no way be taken to reflect the views of the European Union,
CASE, Kiel Institute for the World Economy or other institutions the authors may
be affiliated to.
Keywords: CIS countries, labor migration, CGE model, system transformation
JEL codes: C68, F22, F24
© CASE – Center for Social and Economic Research, Warsaw, 2009
Graphic Design: Agnieszka Natalia Bury
EAN 9788371784842
Publisher:
CASE-Center for Social and Economic Research on behalf of CASE Network
12 Sienkiewicza, 00-010 Warsaw, Poland
tel.: (48 22) 622 66 27, 828 61 33, fax: (48 22) 828 60 69
e-mail: case@case-research.eu
http://www.case-research.eu

The CASE Network is a group of economic and social research centers in Po-land,
Kyrgyzstan, Ukraine, Georgia, Moldova, and Belarus. Organizations in the
network regularly conduct joint research and advisory projects. The research cov-ers
a wide spectrum of economic and social issues, including economic effects of
the European integration process, economic relations between the EU and CIS,
monetary policy and euro-accession, innovation and competitiveness, and labour
markets and social policy. The network aims to increase the range and quality of
economic research and information available to policy-makers and civil society,
and takes an active role in on-going debates on how to meet the economic chal-lenges
facing the EU, post-transition countries and the global economy.
The CASE network consists of:
• CASE – Center for Social and Economic Research, Warsaw, est.
1991, www.case-research.eu
• CASE – Center for Social and Economic Research – Kyrgyzstan,
est. 1998, www.case.elcat.kg
• Center for Social and Economic Research – CASE Ukraine, est.
1999, www.case-ukraine.kiev.ua
• CASE –Transcaucasus Center for Social and Economic Research,
est. 2000, www.case-transcaucasus.org.ge
• Foundation for Social and Economic Research CASE Moldova, est.
2003, www.case.com.md
• CASE Belarus – Center for Social and Economic Research Belarus,
est. 2007.

Atamanov, Luecke, Mahmoud, Mogilevsky, Tereshchenko, Tourdyeva, Uzagalieva, Vavryshchuk
Contents
1. Introduction and Overview ............................................................................. 8
2. The Macroeconomic Context......................................................................... 13
3. Country Studies: Net Emigration Countries................................................ 19
3.1. Overview over the IFPRI standard CGE model ....................................... 20
3.2. Moldova.................................................................................................... 23
3.3. Ukraine .................................................................................................... 26
3.4. Georgia .................................................................................................... 32
3.5. Kyrgyzstan ............................................................................................... 41
4. Country Study: Russia ................................................................................... 47
4.1. Data availability and quality..................................................................... 47
4.2. Experiment design .................................................................................... 56
4.3. Results ...................................................................................................... 60
4.4. Conclusions .............................................................................................. 64
5. Conclusions ..................................................................................................... 66
References............................................................................................................. 69
CASE Network Reports No. 86 4

INCOME AND DISTRIBUTION EFFECTS OF MIGRATION AND REMITTANCES…
The Authors
Aziz Atamanov is a researcher at CASE-Kyrgyzstan. His interests are centered
on public policy, including issues of social protection, fiscal decentralization, pub-lic
expenditure management, macroeconomics, international trade, remittances and
financial sector. Currently, he is a Ph.D. student at the Maastricht Graduate School
of Governance (http://www.governance.unimaas.nl).
Matthias Luecke is a senior research economist at the Kiel Institute for the
World Economy (www.ifw-kiel.de), where he specializes in development econom-ics,
international trade, and international labor migration. He was a senior econo-mist
in the IMF’s European II Department from 2000 until 2003 and has taught,
among other institutions, at Kiel University and the Joint Vienna Institute
(www.jvi.org).
Toman Omar Mahmoud is a doctoral research fellow at the Kiel Institute for
the World Economy, where he specializes in development economics. His work is
focused on microeconomic issues, in particular migration, rural development, and
HIV/AIDS.
Roman Mogilevsky is an executive director and senior research fellow at
CASE-Kyrgyzstan (www.case.elcat.kg), where he works on a broad spectrum of
problems of transition economies including foreign trade, migration and remit-tances,
effectiveness of foreign aid, and social policies.
Kseniya Tereshchenko is a Ph.D. student at the Center for Economic Re-search
and Graduate Education – Economic Institute (CERGE-EI) in Prague
(www.cerge-ei.cz).
Natalia A. Tourdyeva is a research economist in the Center for Economic and
Financial Research (CEFIR, www.cefir.org). She specializes in international trade,
regional and environmental economics, assessing different policy changes with the
help of computable general equilibrium models. She works a lot with input output
(IO) tables and social accounting matrices (SAM) and has contributed the Russian
IO table to the GTAP 7.0 database.
5 CASE Network Reports No. 86

Ainura Uzagalieva is a post-doctoral research fellow at the Center of Applied
Economics Studies of the Atlantic (CEEAplA), the University of the Azores
(www.uac.pt) and the Economic Institute of the Czech Academy of Science
(www.cerge-ei.cz). She specializes in labor economics and economics of innova-tion.
Previously, she was an economist at the central bank of Kyrgyzstan, CASE
and KPMG Consulting LLC.
Vitaliy Vavryshchuk is an associated researcher at Center for Social and Eco-nomic
Research CASE Ukraine (www.case-ukraine.kiev.ua). He graduated from
EERC program at the National University of "Kyiv-Mohyla Academy" and ob-tained
MA degree in economics. His fields of specialization include foreign trade
and WTO, European integration, and macroeconomics.

Abstract
This paper analyzes the direct and indirect income effects of international labor
migration and remittances in selected CIS countries. The analysis is based on
computable general equilibrium (CGE) models for Moldova, Ukraine, Georgia,
Kyrgyzstan, and Russia. All net emigration countries would experience a sharp
contraction of private consumption in the absence of remittances. In Russia, the
main effect of immigration has been to hold down the real wage (as potential capi-tal
stock adjustments in response to immigration are not reflected in our compara-tive-
static modeling framework). The paper concludes that because of the impor-tant
contribution of migration and remittances to stabilizing and sustaining in-comes
in many CIS countries, enhanced opportunities for legal labor migration
should figure prominently in any deepening of bilateral relations between CIS
countries and the European Union under the European Neighborhood Policy.

1. Introduction and Overview
In many member countries of the Commonwealth of Independent States (CIS),
international labor migration and remittances received by relatives at home now
play a large role in financing private consumption and in shaping the labor supply
and education decisions of households. Remittances received range from 8 percent
of gross domestic product (GDP) in Ukraine to more than 30 percent in Moldova
according to the latest internationally comparable estimates (see Section 2 for de-tails).
At the same time, migration and remittances matter not only at the level of in-dividual
households. The large size of remittances suggests that they have proba-bly
affected output and income distribution not only directly at the level of remit-tance-
receiving households, but also through general-equilibrium or indirect chan-nels.
For example, in many CIS countries, a large share of government revenues
derives from taxes on imports (especially VAT), which grew rapidly as a result of
growing remittance inflows. With higher revenues, governments were able to
maintain and expand social transfers so that transfer-receiving households may
have benefited indirectly from migration and remittances although they received
no remittances of their own. Furthermore, the growth of remittances since ap-proximately
the year 2000 has coincided in most CIS countries with the recovery
of GDP from its transition-induced precipitous fall during the 1990s. This coinci-dence
of remittances and GDP growth raises the question of how labor migration
and remittances may have contributed to the economic recovery in most CIS coun-tries
since 2000.
This paper presents case studies for several CIS countries that seek to assess
the indirect effects of labor migration and remittances on income distribution and
structural change systematically. The case studies are based on single country
computable general equilibrium (CGE) models. While the direct impact of migra-tion
and remittances at the household level has been studied through household-level
analyses in many countries, much less work has been done on indirect (or
general equilibrium) effects. The case studies cover several net emigration coun-tries
– Moldova, Ukraine, Georgia, and Kyrgyzstan – as well as Russia, which has
recently experienced both emigration and immigration.
Section 2 describes the overall macroeconomic context in which the sharp in-crease
in labor migration and remittances since approximately the year 2000 oc-

curred. In terms of the main categories of macroeconomic demand, household
final consumption has grown more strongly than GDP since 2000 in the five sam-ple
countries. By contrast, the evolution of government consumption was rather
diverse. Fixed capital formation only began to grow with a lag of several years
after the pickup in GDP in most countries, although the transition-induced fall had
reduced fixed investment to a level insufficient to prevent a gradual depletion of
the existing capital stock. This overall picture is consistent with a view of GDP
growth as initially driven from the demand side, with households’ disposable in-comes
growing because of remittances, a subsequent real appreciation and expan-sion
of non-tradable sectors, and a late response from investment.
Section 3 presents the case studies on net emigration countries. Section 3.1
contains a brief technical description of the standard IFPRI CGE model that is
used in all country studies. Section 3.2 describes the case study for Moldova
where the best available estimates suggest that remittances reached one third of
GDP in 2006 while the number of migrants abroad amounted to about one quarter
of the working population (including migrants). These estimates count only those
migrants who are still part of a household in Moldova (i.e. contribute to household
income and share in expenditures); remittances may include transfers from indi-viduals
who have left Moldova permanently. Labor migration from Moldova is
remarkable in that poorer households are more likely to send a migrant abroad
than richer ones, contrary to the situation in many other countries where barriers to
emigration are more difficult to overcome for poor, credit-constrained households.
In the case of Moldova, the poor and low-skilled have the option of taking up em-ployment
in Russia, where travel is visa-free and cheap, although working condi-tions
and wages are often poor. By contrast, the EU (especially Italy and Portugal)
would be preferred as a host country by many migrants because of better pay and
conditions, but high up-front costs for illegal travel make this a difficult proposi-tion
for many poor households.
Our simulation results suggest all household groups in Moldova would lose
substantially in the absence of migration and remittances. In relative terms, the
losses would be largest for small farmers because (i) migration, including for sea-sonal
work, is very widespread in the countryside, and (ii) higher disposable in-comes
in the population at large are strengthening demand for local food products.
As expected, private consumption would be one third lower, with a smaller reduc-tion
in GDP of approximately one tenth. The only sector that would gain signifi-cantly
is light industry, mostly through much higher exports; this simulation result
is in line with a conventional Dutch disease effect.
In Ukraine (Section 3.3), remittances are rather smaller in relation to GDP than
in the three smaller CIS countries with net emigration. Plausible estimates put
remittances in the order of one tenth of GDP and the number of labor migrants

substantially below one tenth of the labor force. Still, our CGE-based simulations
reveal quite notable effects of migration and remittances. For example, the coun-try’s
hypothetical economy would have lost up to 7% of its potential without mi-gration
and remittance-induced effects. All types of households benefit from re-mittances
substantially: their overall consumption would have been lower by 14 to
21 percent in the hypothetical “migration and remittance free” economy. Rich
urban households are set to win the most, while households with income coming
mostly in form of government transfers gain the least from remittances and their
economy-wide effects. On production side, light and food industry are the key
beneficiaries of remittance-driven demand effects. On the other hand, local ma-chinery,
construction and public administration sectors appear to be quite remit-tance-
neutral.
In Georgia (Section 3.4), labor migration apparently started earlier than else-where
in the region, with substantial remittances already in the mid-1990s. While
the varying estimates are difficult to reconcile, remittances were probably below
one fifth of GDP. The number of labor migrants is estimated at less than one mil-lion;
even if many of these are in fact long-term emigrants, their number is large
relative to an economically active population in Georgia of approximately
2 million. Labor migration and remittance inflows have a strong macroeconomic
growth effect at the aggregated level; however, not all sectors and residents are
affected symmetrically. The positive effect of remittances is pronounced in manu-facturing,
large-scale agriculture, construction, and service sectors concentrated
mostly in the urban areas of the country. The impact on the production of house-hold
farmers (or small agriculture) depends on the geographical location and iden-tity
of households. In distant regions with high transaction costs, for example,
farm production increases substantially, while in regions with lower transaction
cost farm production decreases once farmers have access to remittance incomes.
As a result, remittances have a rather limited overall impact on poverty reduction
and income inequality, especially in rural areas. Their impact on consumption is
smallest for the group of poor and middle-income rural households and largest for
urban households with higher incomes. Consequently, the wealthier members of
the society gain from remittances more than poorer household categories. Better
access to labor markets, on the contrary, would improve the welfare states of
many, especially, of the rural poor at the outset.
These simulation results for Georgia suggest that government and donor poli-cies
should prioritize a pro-poor approach in improving institutions within the
country, especially by improving access to labor and credit markets. With the fo-cus
on the inclusion of the rural poor in the financial sector, for example, policies
could be designed for meeting the production needs of farmers in distant regions.
This would include opportunities to link remittance flows with family-based mi-

crofinance mechanisms focused on promoting saving, insurance and investment
within a given region, as well as decreasing transaction costs across regions.
In Kyrgyzstan (Section 3.5) remittances also play a very important role in the
economy; by official estimates for 2007, remittances exceeded one quarter of the
country’s GDP. The number of labor migrants is estimated at 15-20% of total la-bor
force in the country; their main destination countries are Russia and Kazakh-stan.
Kyrgyz labor migrants represent all segments of the society, but majority of
them originate from labor abundant and land scarce rural areas and small towns in
the southern part of the country. While wealthier households receive more remit-tances
than poor ones because of better education and access to information, the
role of remittances is larger for poor households. Private consumption and gov-ernment
revenues (through taxation of remittance-driven imports) depend substan-tially
on remittances. In the absence of migration, all types of households would
be worse off, with considerable losses for GDP, private and government consump-tion.
The situation in Russia (Section 4) mirrors in some way the emigration coun-tries
as Russia is the predominant destination for labor migrants within the CIS.
Inward labor migration has come to play a significant role in the Russian econ-omy.
The data situation is less than satisfactory as Russian official statistics usu-ally
quote migration flows, while the World Bank estimates the stock of immi-grants,
rendering comparisons difficult. Thus official sources put the number of
immigrants arriving to Russia in 2006 at less than 200,000 individuals, mostly
from CIS countries. By contrast, the total immigrant population is estimated by the
WB at 12 million individuals in 2005. Remittances make up only a small part of
Russian outward capital flows, with estimates ranging from US$ 6 billion (Central
Bank of Russia) to US$ 12 billion (World Bank) US$.
Our CGE-model-based simulations are designed to assess the direct and indi-rect
effects of migration on the sectoral structure of the Russian economy. An
increase in labor migration increases the supply of labor for all industries, pushes
wages down, and raises rent (capital income). The fall in the wage rate drives do-mestic
prices down and stimulates exports. The price ratio of tradables to nontrad-ables
increases, along with the real and nominal exchange rates, all implying a real
depreciation of the Russian currency. Both the direct and the indirect effects work
in the same direction, with the indirect effects dominating.
Section 5 concludes that, while these country case studies do not analyze spe-cific
policy measures, they demonstrate large potential benefits from labor migra-tion
and remittances for migrants’ home countries. These potential benefits, as
well as the attending risks, depend in important ways on government policies re-lated
to migration in both, home and host countries. Against the backdrop of the

migration-related policies currently pursued in the CIS region and in the EU, three
major policy implications emerge.
First, some CIS country governments faced with large migrant outflows have
been reluctant, for political reasons, to even acknowledge that emigration is taking
place on a large scale. Consequently, they have failed to provide support services,
such as effective consular support, job placement through recognized agencies, or
advocacy for more legal migration opportunities with host country government.
By providing such services, sending country governments would help to limit the
risks faced by migrants and enhance the benefits of migration to migrants and their
families as well as sending country economies overall.
Second, for economic recovery to take hold in the smaller, natural-resource-poor
CIS countries, fixed investments need to be sustained and increased further.
Remittances could help to pay for such investment. However, the business and
investment climate in many of these countries is so poor that, currently, remit-tances
are only rarely used for productive investment. Government efforts to
channel remittances into investment, which are debated in many CIS countries,
will succeed only when all investors can expect to receive an adequate return on
productive investments that is not diminished by parasitic public institutions.
Third, destination countries will increasingly find themselves competing not
only for high-skilled migrants, but also for those willing to perform jobs that are
otherwise difficult to fill (such as seasonal work in agriculture, construction, and
social services). Russia, the most important host country for migrants from the CIS
region, is currently offering legal employment on a fairly broad basis, even though
the number work permits has recently been cut because of the effects of the finan-cial
crisis. By contrast, legal employment opportunities in the European Union are
still limited, although legalization programs such as in Italy and Portugal will soon
create pockets of legal migrants that will become the hubs of future migrant net-works.
As these networks will attract more family-based and other immigration
from CIS countries in the future, it would be in the interest of both migrants and
EU host countries to replace haphazard legalizations with a forward-looking strat-egy
for admitting migrants with good job prospects in the EU. Since the EU func-tions
as a single labor market, such programs should be coordinated at the EU
rather than the national level. For the benefit of both, CIS countries and the EU,
the deepening of bilateral relations under European Neighborhood Policy should
include enhanced opportunities for legal labor migration.

2. The Macroeconomic Context
The stylized facts of macroeconomic development in the CIS countries during
the last two decades are fairly straightforward. After the disintegration of the for-mer
Soviet Union in 1991, GDP fell sharply in all five countries covered by this
study (Figure 2.1; all data in Figures 2.1 through 2.4 are from the World Bank’s
Word Development Indicators Database). This precipitous fall was followed by
prolonged stagnation at a low level during the second half of the 1990s, with some
further losses as a result of the Russian financial crisis in 1998. A sustained recov-ery
began around 2000 in most countries and still continues as of mid-2008. While
total output is still below its pre-independence level in most CIS countries, house-hold
final consumption expenditures have recovered much better (Figure 2.2); in
several countries, they had regained or even surpassed their 1990 levels by 2006.
By contrast, fixed investment recovered much more slowly (Figure 2.3) and the
picture for government final consumption is rather mixed across the CIS countries
covered by this study (Figure 2.4).
Figure 2.1. Total GDP, 1990 to 2006
300
250
200
150
100
50
0
1990 1992 1994 1996 1998 2000 2002 2004 2006
Georgia
Kyrgyz Republic
Moldova
Russia
Ukraine
Note. Year 2000 set as 100.

Figure 2.2. Household final consumption expenditures, 1990 to 2006
250.0
200.0
150.0
100.0
50.0
0.0
1990 1992 1994 1996 1998 2000 2002 2004 2006
Georgia
Kyrgyz Republic
Moldova
Russia
Ukraine
Figure 2.3. Gross fixed capital formation, 1990 to 2006
600
500
400
300
200
100
0
1990 1992 1994 1996 1998 2000 2002 2004 2006
Georgia
Kyrgyz Republic
Moldova
Russia
Ukraine
The driving forces behind this development are somewhat less clear. Indeed,
they have been the subject of contentious debates that evolved considerably as
events unfolded. The post-independence output collapse is probably explained
largely by the rapid collapse of institutions that regulated trade during the Soviet
period, while initially the necessary institutional infrastructure for market-based
economic relations (functioning currencies, hard budget constraints, enforceable
contracts) did not exist. An additional role was played by the sharp reductions in
government procurement, which particularly affected the military industrial com-plex.

Figure 2..4. General government final consumption, 1990 to 2006
300
250
200
150
100
50
0
1990 1992 1994 1996 1998 2000 2002 2004 2006
Georgia
Kyrgyz Republic
Moldova
Russia
Ukraine
The gradual emergence of market-enabling institutions helps to explain why
trade among the CIS countries, along with output, stabilized during the mid-1990s.
However, observers continued to note many persistent shortcomings regarding
corporate governance, the business climate, and the investment environment.
Therefore, the sustained recovery since 2000 is more difficult to explain. Indeed,
investment appears to have lagged, rather than led, the output recovery. One factor
that has clearly played a role in the recovery is the resurgence of Russian import
demand for CIS products on the heels of rising world market prices for energy
materials since the late 1990s. Higher Russian import demand would have led to
higher exports by the other CIS countries and higher rates of capacity utilization.
Another factor that contributed to the recovery has been the growing demand for
non-tradable goods and services (a Dutch-disease-style effect) throughout the re-gion.
It resulted from rising oil and gas revenues in the case of Russia and from
sharply rising migrant remittances since the late 1990s in the remaining countries
(Figure 2.5).
The available time-series data on migrant remittances are based on balance of
payments statistics (as compiled by the International Monetary Fund in its Balance
of Payments Statistics Database), representing the sum of credit items from com-pensation
of employees (income account) and workers’ remittances (transfer ac-count).
Data quality varies widely across countries, especially with respect to
whether they include estimates of cash transfers through informal channels in ad-dition
to transfers through the banking system and money transfer operators. For
every country, however, there is a substantial increase in migrant remittances from

about 2000. The picture for Georgia is more nuanced in that there were substantial
remittances as early as the mid-1990s; notably, Georgia’s GDP also began to re-cover
at that time, earlier than elsewhere in the region. The 2007 data point for
Ukraine clearly represents a break in the underlying time series; indeed, the low
level of remittances suggested by the pre-2007 estimates is not plausible, given
Ukraine’s large migrant population abroad.
Figure 2.5. Migration remittances, 1997 to 2007 (million US$)
5000
4500
4000
3500
3000
2500
2000
1500
1000
500
0
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Georgia
Kyrgyz Republic
Moldova
Ukraine
Figure 2.6. Russia: Migrant remittances credit and debit, 2001 to 2007 (billion US$)
18
16
14
12
10
8
6
4
2
0
2001 2002 2003 2004 2005 2006 2007
Remittances - credit
Remittances - debit
Balance of payments data for Russia, which hosts most labor migrants from
CIS countries, represent the mirror image of these developments. Remittances by
incoming migrants (“debit” items) rose to almost US$ 17 billion in 2007 from

around US$ 1 billion in 2001 (Figure 2.6). Remittances received by Russian expa-triates
also increased, but remained below US$ 3 billion in 2007.
While balance of payments statistics provide at least a starting point for assess-ing
the evolution of migrant remittances over time and across countries, the data
situation with respect to the number of labor migrants is even more problematic.
Official data on population movements are often not helpful because a large pro-portion
of labor migration occurs illegally. Furthermore, labor migration occurs in
many different settings, ranging from seasonal movements to immigrants becom-ing
citizens of the host country. Standard definitions of migrant status based on
foreign birth or foreign citizenship may therefore lead to very different estimates.
Furthermore, these estimates may be difficult to interpret when a country like the
Soviet Union, where internal migration was extensive, is split up into several
countries while new national identities and allegiances are only evolving.
A comprehensive country-by-country matrix of the number of foreign-born
residents has recently been compiled by the World Bank; the IFAD remittances
database uses these migrant numbers to estimate migrant remittances in 2006 (Ta-ble
2.1). Although both data sets are subject to many uncertainties, they represent
the best combined estimates of labor migration and remittances available across a
wide range of countries, drawing on a wide variety of national data sources. For
the countries covered by this study, remittances are estimated at close to one third
of GDP in Kyrgyzstan and Moldova, and at about one fifth in Georgia. Although
much larger in absolute terms, remittances in Ukraine amount to less than one
tenth of GDP. In Russia, remittances received from emigrants are small by com-parison
at less than 2% of GDP.
The large number of emigrants estimated for Russia and Ukraine reflects in
part large groups of ethnic Russian- and Ukrainian-born residents in other coun-tries
of the Former Soviet Union, rather than recent labor migrants to richer coun-tries.
Many of these individuals have probably resided there since before the disin-tegration
of the Soviet Union in 1991; their motivations for staying or leaving in
their host countries are therefore bound to be rather different from recent labor
migrants motivated by higher potential incomes abroad. Accordingly, also, the
remittances received by Russian and Ukrainian as reported in Table 2.1 are likely
to be overestimated somewhat.
Overall, this brief review of macroeconomic developments and remittances in
the CIS countries covered by this study suggests that, other than in Russia, remit-tances
have sustained private consumption and contributed to the resumption of
GDP growth since 2000. Migration and remittances are large enough not only to
improve the welfare of remittance-receiving households, but also to affect macro-economic
variables such as the real exchange rate and wage levels and thus alter

the sectoral structure of the economies, thereby generating further indirect effects
on the welfare of all households, whether or not they receive remittances. These
effects will be analyzed further in Sections 3 and 4 below through simulations
based on computable general equilibrium (CGE) models for the selected CIS
countries.
Table 2.1. Selected CIS Countries: Migrants and Remittances, 2006
Remittances
US$ million percent of GDP
Migrants
(thousand)
Georgia 1,525 20.2 1036
Kyrgyzstan 846 31.4 597
Moldova 1,027 31.4 663
Ukraine 8,471 8.0 5878
Russia: immigrants n.a. n.a. 11977
Russia: emigrants 13,794 1.4 12099
Source: IFAD remittances database http://www.ifad.org/events/remittances/maps/brochure.pdf;
http://www.migrationdrc.org/research/typesofmigration/global_migrant_origin_database.html.

3. Country Studies: Net Emigration
Countries
The following sub-sections present country studies on the effects of migration
and remittances in selected CIS countries that experienced net emigration of work-ers:
Moldova, Ukraine, Georgia, and Kyrgyzstan. Each country study involves the
application of a standard single-country computable general equilibrium (CGE)
model to a social accounting matrix that was developed under the ENEPO project.
The underlying CGE model is the well-documented IFPRI (International Food
Policy Research Institute) standard model (Lofgren et al. 2002) which has been
widely applied to developing countries. Its straightforward basic structure with a
choice of standard neoclassical and neo-structuralist assumptions, its user-friendly
and well-documented code (in GAMS software), and its easy adaptability to na-tional
circumstances (different levels of aggregation for households, the agricul-tural
sector, etc.) render it well-suited for the present analysis.
We also explored the feasibility of using a recursive-dynamic version of the
IFPRI standard CGE model to complement our comparative-static simulations. A
more explicitly dynamic structure would reflect the investment process more accu-rately
and thereby provide additional insights into the growth effects of remit-tances.
However, the additional assumptions required to implement a recursive-dynamic
model turned out to be far-reaching. Overall, we would have introduced a
high level of arbitrariness into the analysis such that the more detailed description
of the investment process in the recursive-dynamic model would ultimately have
been meaningless.
In constructing the social accounting matrices, we combine input output tables,
other national accounts information, household budget surveys, labor force statis-tics,
and fiscal statistics, among other data sources. A key difficulty is that official
data tend to understate migration and remittances in some countries because a
large proportion of remittances are transferred as foreign exchange cash and much
migrant employment is informal. For the databases to reflect realistic orders of
magnitude, various data sources are drawn upon and appropriate adjustments
made. The level of aggregation (number of commodities, sectors, factors of pro-duction,
and household types) differs slightly across the country studies.

The following section summarizes some key features of the IFRPI standard CGE
model. Each of the subsequent country studies starts with a discussion of the data
situation and goes on to report simulation results that seek to answer the question
of what each economy would look like without migration and remittances. These
effects are disaggregated by household type and followed through the economy.
3.1. Overview over the IFPRI standard CGE model1
In general, computable general equilibrium (CGE) models are used to analyze
the effects of policy changes and other shocks (such as an opening to labor migra-tion)
throughout the economy. CGE models may be viewed as an extension of
multi-sector input-output and fixed price models and incorporate the indirect ef-fects
and price effects of policies. They apply to the time period it takes for an
economy to move from one equilibrium to another, in response to a policy change
or other shock. In this sense, a static CGE model as used in the following country
studies generates a medium-term solution - a situation where the initial disequilib-rium
after the shock has disappeared, but before dynamic effects (such as addi-tional
investment or disinvestment) set in.
The database for a CGE model is the social accounting matrix (SAM) for a
given year, a square matrix that describes all commodity and monetary flows
among the economic agents in an economy at a suitable level of aggregation (pro-duction
sectors, households, enterprises, government, the “rest of the world”).
Depending on the chosen level of aggregation, the SAM combines information
from input-output accounts, national income and product accounts, household
budget surveys, labor force statistics, and fiscal statistics, among other data
sources. Although the case studies in this paper are all based on single-country
CGE models, CGE models are also implemented for several countries simultane-ously
or for the global economy disaggregated by regions and countries, requiring
multi-country SAMs.
Naturally, simulation results based on CGE models depend heavily on assump-tions
about functional forms (such as Leontief vs. CES production functions), un-derlying
parameters (such as substitution elasticities in production and demand
functions), and macroeconomic balancing mechanisms. Traditional neoclassical
CGE models are based on Walrasian general equilibrium theory: Firms maximize
profits, and wages and prices adjust to equate supply and demand in factor and
1 This section draws extensively on Fagernäs (2004).

product markets, with factors of production fully employed. By contrast, models in
the structuralist tradition use different assumptions about macroeconomic balanc-ing
mechanisms and the way markets clear, incorporating features of short-run
macroeconomic models with wage or price rigidities and unemployment. Many
CGE models (including in this paper) are real side models without asset markets
and with neutral money2.
The IFPRI standard CGE model on which our case studies are based allows for
a range of alternative assumptions that reflect both the structuralist and the neo-classical
tradition. It is a real-side model without explicitly modeled asset markets
or inflation3. Production is carried out under perfect competition by sectors (“ac-tivities”)
that maximize their profits, subject to a multi-level production function
and given the prices of their inputs, outputs and factors. Within each activity, the
top level of the production function consists of a Leontief function that combines
an aggregate of the factors of production (value-added) on the one hand and inter-mediate
inputs on the other hand. Factors of production (high-skilled workers,
low-skilled workers, capital, etc.) are combined according to a constant elasticity
of substitution (CES) function, while the various intermediate inputs are used in
fixed proportions (Leontief function).
Domestic output of a given good (say, food) may be produced by different sec-tors
(say, small farms and commercial agriculture). Aggregate domestic output is
allocated to exports and domestic sales based on profit maximization with given
prices, a given quantity of total output and imperfect transformability between
domestic sales and exports in line with a constant elasticity of transformation
(CET) function. Export supply is therefore determined by the domestic price of
exports relative to domestic sales, where the domestic price of exports is the given
world price of exports adjusted by export taxes and the exchange rate. Thus the
model assumes a small, open (i.e. price-taking) economy.
Domestic demand for a given good is the sum of demand for private (house-hold)
consumption, government consumption, investment, and intermediate in-puts.
If a commodity is imported, domestic demand is for a composite commodity
of imports and domestically produced goods, with the optimal mix determined
through cost minimisation via a CES aggregation function. This so-called Arming-ton
assumption allows for some decoupling of domestic from world market prices
and ensures that simulated export and import responses to policy changes will be
broadly realistic. The supply of imports is infinitely elastic at given world prices.
2 For an overview over the main CGE modeling approaches with references to appropriate
literature, see Robinson (2003).
3 For a detailed description of the IFPRI standard model see Lofgren et al. (2002).

The domestic price of imports is the given world price adjusted by the exchange
rate and tariffs.
Households receive income from factors of production supplied to production
activities (e.g. labor) and transfers from institutions, particularly the government
and other households. Households save, consume, and pay direct taxes and trans-fers;
direct tax rates and the propensity to save are determined by the chosen clo-sure
rules for the government and savings-investment balances (see below).
Household consumption is allocated across different products according to a linear
expenditure system (LES), which implies in this case that the consumption of in-dividual
commodities is a linear function of total household consumption expendi-ture.
Household types may differ in their demand elasticities and consumption
shares for each commodity.
The macro closure rules for the CGE model define the mechanisms by which
the three macroeconomic balances are determined: (i) the current government
balance; (ii) the current account balance, and (iii) the savings and investment bal-ance.
First, the government receives its income from taxes and transfers from the
rest of the world and decides on the level of current spending (government con-sumption
plus transfers to households) vs. government saving. The IFPRI standard
model allows us to choose between two basic closure rules for the government
balance. Either direct tax rates are fixed, but government saving (the real fiscal
balance) adjusts in response to changes in government revenue; or direct tax rates
adjust to maintain a given level of government saving.
Second, the savings-investment balance states that total investment is the sum
of private investment, government investment, and foreign savings. In this static
model (comparative-static with simulation exercises), investment is not driven by
the rate of return on capital but by the availability of savings. The standard IFPRI
model allows for two basic approaches: (i) private savings are investment-driven
such that the marginal propensity to save adjusts to a given level of investment;
(ii) investment adjusts to the level of savings, given a fixed marginal propensity to
save.
Third, foreign savings are equivalent to the current account deficit and hence
define the external balance (i.e. the balance of inflows and outflows of foreign
exchange). Transfers between the rest of the world and domestic institutions and
factors (including migrant remittances) are exogenous to the model and given in
foreign currency. Thus the current account balance is driven by the balance of
exports and imports. The alternative closure rules in the standard IFPRI model
leave either the real exchange rate or the current account balance constant, requir-ing
the other variable to adjust.

Factor market closures determine the mechanisms that equilibrate the supply
and demand of each factor of production. In line with general equilibrium theory,
each activity uses a set of factors up to the point where the marginal revenue prod-uct
of each factor equals its wage. There are two wage variables: the economy-wide
wage, and an activity-specific wage that is the product of the economy-wide
wage and an activity specific wage or distortion term. There are basically three
possible closure rules: (i) a factor is fully employed and mobile, giving rise to a
uniform economy-wide wage; (ii) a factor is fully employed and immobile, giving
rise to sector-specific wage rates; or (iii) a factor is mobile but may be unem-ployed,
allowing the wage to be set as a policy parameter.
3.2. Moldova4
Migration and remittances play a key role in the Moldovan economy, with ap-proximately
one quarter of the working-age population working abroad for at least
part of the year, and remittances equivalent to one third of GDP in 2006. These
estimates count only those migrants who are still part of a household in Moldova
(i.e. contribute to household income and share in expenditures); remittances may
include transfers from individuals who have left Moldova permanently.
Labor migration from Moldova is remarkable in that poorer households are
more likely to send a migrant abroad than richer ones, contrary to the situation in
many other countries where barriers to emigration are more difficult to overcome
for poor, credit-constrained households. In the case of Moldova, the poor and low-skilled
have the option of taking up employment in Russia, where travel is visa-free
and cheap, although working conditions and wages are often poor. By con-trast,
the EU (especially Italy and Portugal) would be preferred as a host country
by many migrants because of better pay and conditions, but high up-front costs for
illegal travel make this a difficult proposition for many poor households.
Our social accounting matrix is based on the year 2004; available data include
the national accounts through 2004, including an input-output table, annual house-hold
budget surveys through 2004, quarterly labor force surveys through 2005,
and a special household survey on migration and remittances conducted in 2004
and 2006. As is the case in many developing countries, coverage of household
income by the household budget survey is far from complete, particularly for re-mittances.
Therefore, the national accounts data, which are internally consistent,
4 Authors of this section: Matthias Luecke and Toman Omar Mahmoud.

have been used as the starting point in compiling the social accounting matrix.
Subsequently, the household sector and labor income have been disaggregated
using percentage shares for different household and labor types derived from the
household budget survey. The agricultural sector is subdivided into small-scale
(household) agriculture and agricultural enterprises.
Our simulations seek to describe what the Moldovan economy would have
looked like in 2004 without labor migration and remittances. The first simulation
hypothetically eliminates the recent strong growth in total factor productivity
(TFP). That TFP growth is apparent from the fact that GDP grew by about one
third from 2000 through 2004, while fixed investment remained modest and the
labor force declined. In part, TFP growth may have been a natural result of the
recovery from the transition-induced crisis, for example due to the emergence of
market-supporting institutions as systemic transformation takes hold. To this ex-tent,
TFP growth might have occurred even in the absence of migration. However,
we consider it plausible that most of the apparent TFP growth results from higher
utilization rates for existing production capacity that arose as a consequence of
remittances-induced demand growth.
The second and third simulations separately describe the impact of a sharp re-duction
in remittances and a larger domestic labor supply (if there is no labor mi-gration).
The fourth simulation combines lower remittances and a larger labor
supply, and the fifth simulation adds lower TFP. It turns out (Table 3.1) that the
combined effects under the fourth scenario are very similar to the sum of the sepa-rate
effects under the second and third scenario; similarly, the combined effects for
all simulated shocks under the fifth scenario are very similar to the sum of the
separate effects under the first, second, and third scenario. Therefore, in discussing
the results, we focus on the fifth scenario.
Table 3.1. Moldova: Simulation Results (base values and percentage changes in real
terms)
Base
run
Scenario
1
Scenario
2
Scenario
3
Scenario
4
Scenario
5
Macro variables
Domestic absoprtion 412 -13.8 -13.2 9.0 -4.1 -21.6
Private consumption 276 -20.6 -19.7 13.4 -6.1 -32.1
Fixed investment 67 0.0 0.0 0.0 0.0 0.0
Government consump-tion
52 0.0 0.0 0.0 0.0 0.0
Exports 155 -26.1 38.0 18.4 58.0 22.4
Imports -246 -16.4 2.0 11.6 14.6 -7.8
GDP at market prices 320 -17.8 -0.2 11.6 11.6 -10.9
Real exchange rate 95 -4.1 4.8 -0.4 2.6 2.7

Base
run
Scenario
1
Scenario
2
Scenario
3
Scenario
4
Scenario
5
GDP at factor cost
A_AGR_L 19 -42.0 -20.7 5.3 -14.7 -42.0
A_AGR_S 38 0.0 0.0 20.0 20.0 -4.0
A_FOOD 22 -24.2 -7.6 11.2 1.3 -21.8
A_LIGHT 6 -46.5 226.4 51.6 299.0 199.9
A_WOOD 4 -18.9 -2.1 7.9 4.6 -13.5
A_CHEM 7 -19.4 -2.2 9.7 6.7 -14.0
A_MASH 4 -19.7 0.1 7.9 5.8 -10.1
A_ELEC 7 -19.8 -6.9 11.8 4.4 -17.6
A_CONSTR 11 -4.0 -0.9 2.4 1.5 -3.3
A_TRADE 32 -19.3 2.2 12.1 14.2 -9.3
A_REST 3 -22.2 -6.4 12.5 5.1 -19.1
A_TRANS 19 -20.9 -2.3 11.4 8.5 -14.8
A_COMM 17 -21.8 -9.4 10.6 1.6 -22.9
A_FIN 33 -19.0 -6.7 10.8 4.3 -17.9
A_PUBLIC 14 -5.3 -4.4 3.4 -1.0 -7.5
A_PUB_SERV 35 -12.2 -9.6 7.9 -1.7 -16.6
TOTAL 271 -17.3 -0.1 11.7 11.9 -10.5
Household consumption (equivalent variation)
HH_SMALL_FARM 75 -16.7 -41.7 11.6 -31.9 -49.9
HH_OTH_RUR 69 -24.1 -9.4 15.9 7.7 -25.0
HH_OTH_URB 51 -23.9 -10.7 13.8 4.1 -26.2
HH_RICH_URB 28 -22.1 -9.9 14.9 6.2 -23.7
HH_PUBLIC_SECTOR 17 -23.8 -17.7 14.9 -2.2 -32.3
HH_TRANSFER 35 -14.7 -14.9 10.0 -4.6 -23.3
Note. Scenario 1 – TFP reduced by 20 pc (except in small-scale agriculture). Scenario 2 –
Remittances reduced by 70 pc. Scenario 3 – Labor supply increased by 20 pc (except high-skilled
and non-agriculture self-employed: 10 pc). Scenario 4 – Remittances reduced and
labor supply increased. Scenario 5 – TFP and remittances reduced, labor supply increased.
Our simulation results suggest that household consumption would be reduced
sharply in the absence of migration and remittances. This applies both to total
consumption, which would fall by 32.1 percent, and to every household group. In
relative terms, the losses would be largest for small farmers (with consumption cut
in half) because (i) migration, including for seasonal work, is very widespread in
the countryside, and (ii) higher disposable incomes in the population at large are
strengthening demand for local food products.
GDP would fall by approximately one tenth under the fifth scenario and thus
by much less than private consumption. A comparison of the GDP effects under
the first, third, and fifth scenario demonstrates that GDP would decline mainly
because we assume that without migration, total factor productivity (TFP) would
not improve as much as it actually did. As discussed above, total factor productiv-

ity should be thought of in this context as predominantly reflecting higher rates of
capacity utilization, rather than technological change. The positive output effect of
migrant workers returning home and adding to the local labor force (under the
third scenario) would be insufficient to compensate for the fall in TFP.
Without migration and remittances and the associated TFP growth, the Moldo-van
currency would depreciate in real terms, adjusting to lower foreign exchange
inflows as remittances are reduced. Our closure rules for these simulations imply
that foreign savings, investment, and government consumption are exogenous.
While foreign savings are adjusted to reflect lower migrant remittances under the
second, fourth and fifth scenario, investment and government consumption remain
constant in real terms relative to the base run.
The only sector whose output would grow significantly in the absence of mi-gration
and remittances is light industry, whose exports would also increase
sharply. This simulation result is in line with a conventional Dutch disease effect
as a result of the inflow of foreign currency through remittances. It is interesting to
compare these simulation results with recent sectoral changes in the Moldovan
Economy. Textile and clothing exports from Moldova to the European Union have
in fact expanded substantially since 2007 even though migration and remittances
have grown and the Moldovan currency has appreciated further since 2004. These
recent developments suggest that barriers to trade, rather than high and rising pro-duction
costs because of a Dutch-disease-style real appreciation, impeded the ex-pansion
of the Moldovan textile and clothing sector in the past. With Romania’s
accession to the European Union, Moldova became a direct EU neighbor, sepa-rated
by only one border from the European single market; it appears that the re-sulting
reduction in informal trade barriers was sufficient to set off the recent wave
of foreign direct investment and output and export growth in the textile and cloth-ing
industry.
3.3. Ukraine 5
3.3.1. Migration data
According to several alternative studies, the overall stock of Ukrainian labor
migrants working abroad during the late 1990s and early 2000s ranged from 0.8 to
5 Author of this section: Vitaliy Vavryshchuk

2 million persons. The latest ILO survey reveals about 780,000 Ukrainian labor
migrants in other countries (about 3.5 - 4% of total labor force) (IOM 2006).
These numbers contrast sharply with much lower official data on employment
permits for Ukrainians working abroad (only 61,400 permits were granted by resi-dent
employment intermediaries in 2006). Of these, more than two thirds are typi-cally
for EU countries, especially Cyprus, Greece, and the UK.
The latest full-fledged research of labor migration trends in Ukraine was com-pleted
in 2001 when the State Statistics Committee conducted a survey of 18,000
households in 8 regions (oblasts). The number of labor emigrants in the selected 8
oblasts was estimated at 380,000 persons. Extrapolation to the whole country
gives about 800,000 labor migrants. According to the survey about 60% of all
migrants were employed in countries which are currently the members of the
enlarged EU. The most attractive destinations for Ukrainians at that time were
Poland (hosting about 18% of Ukrainian labor migrants), Czech Republic (17%),
Italy (8.5%), and Portugal (3.8%) (Poznyak 2002).
Another survey of Ukrainian households in eight Western regions of Ukraine in
2005 indicates that ranking of most popular destinations among Ukrainian mi-grants
changed somewhat. Italy topped the list of most desired destinations with
60% of votes, followed by Portugal (31%), Spain (24%), and Poland (23%). Out
of ten most frequently mentioned countries, seven were the EU member states
(Starodub & Parkhomenko 2005).
The true scale of Ukrainian labor migrants’ presence in some European coun-tries
was revealed during regularization programs. In 2002 the Italian government
ran a two-month regularization program for housemaids and contract workers. Out
of 341,000 of applications from housemaids, 27% were submitted by Ukrainians.
During the regularization program in Portugal running from January 2001 till
March 2003, more than 62,000 temporary work permits (out of a total of 180,000)
were granted to Ukrainians (Poznyak 2006).
The share of Ukrainian labor emigrants coming from small cities is estimated at
42% while villagers account for about 29% and people from big cities make 25%
of total number of labor migrants. Migrants from the capital accounted for just
3.0% of overall migrants stock abroad while Kyiv population exceeds 6.5% of the
country population. In terms of professional structure most of Ukrainian men
working abroad are construction or agricultural workers. At the same time, most of
women choose to be employed as housemaids.

3.3.2. Remittances data
Statistics on migrants’ remittances in Ukraine are fragmentary and data from
different sources difficult to reconcile. According to WB study, migrants’ remit-tances
to Ukraine and Moldova (the countries are treated as one sub-region in the
study) totaled some USD 0.44bn (Mansoor & Quillin 2006)6. Although the abso-lute
numbers on remittances seem to be strongly underestimated under the WB’s
approach, the geographical distribution of remittance inflow is worth attention.
About 50% of remittances to the region come from EU-15 countries, while 37%
originated from Russia and other CIS resource-rich countries. Another 5% are
attributed to remittances from new EU member states.
Figure 3.1. Migrants’ remittance to Ukraine in 2003-07 by BoP categories
(USD billion)
6
5
4
3
2
1
0
Labor income
Migrants' remittances
Other transfers to private sector
2003 2004 2005 2006 2007
Source: NBU.
Another recent study by the International Fund for Agricultural Development
(IFAD) puts Ukraine among the top world nations receiving the largest amount of
migrants’ remittances in absolute terms. The IFAD estimated total remittances
transferred to Ukraine in 2006 at USD 8.5bn, or 8.0% of the country’s GDP that
year. To compare, inward remittances are estimated at 1.4% of GDP for Russia,
3.9% for Romania, and 1.4% for Poland (IFAD 2006).
Ukrainian official statistics on total workers’ remittances and labor income
abroad are far from complete and reliable. According to the balance of payments,
6 Statistics on migrants’ remittances are estimated by summing up workers’ remittances
and compensation of employees categories from the Balance of Payments statistics.

Ukrainian workers received USD 171 million of factor income abroad and trans-ferred
USD 193 million as remittances in 2004. This is unrealistically low, given
large-scale labor out-migration from the country. In order to upgrade statistics we
classify transfers to “other sectors” as workers’ remittances in line with the OECD
recommendations. This gives us more realistic overall USD 2.4bn of transfers into
Ukraine, or 7% of total household consumption.
Remittances are crucial for many Ukrainian households and regions. Anecdotal
evidence suggests that remittances-induced domestic demand was the key factor
behind dynamic development of local manufacturing in Western Ukraine. Mi-grants’
remittances are spent in a number of ways: purchases of real estate, repair-ing
of dwelling, purchases of cars, material aid for relatives, and payment for high
education. Obviously remittances have many indirect effects which are in most
cases difficult to quantify. First, remittances spent to pay for higher education
enhance formation of the country’s human capital contributing to higher potential
GDP. Second, some returning migrants invest money in new business start-ups
increasing economic potential of regional economies in the long-run. Remittances
are also an important source of foreign exchange to Ukraine, which is especially
important in view of growing deficit of merchandise trade and income balances
observed in the country since 2005. Thus, inflow of remittances helps to partially
compensate for otherwise increasingly negative balance of current account.
3.3.3. Database for CGE model
The household budget survey that we used to disaggregate income among types
of households does not distinguish remittances as a separate type of income
source. Presumably, respondents counted remittances as a part of “other income”.
However, the distribution of “other incomes” across types of households does not
correspond to the likely pattern of remittances in Ukraine. For example, “rich”
(top 2 deciles in terms of per capita income) urban households have more than
twice as much “other income” as “normal” urban households although it is
unlikely that “normal” households benefit from migrant’s remittances to a lesser
extent than “rich” households. Therefore, we have adjusted the data so as to reflect
a plausible distribution of remittances across different types of households.
Our social accounting matrix for Ukraine (input data for standard IFRPI model)
is based on input-output tables at basic and consumer prices, National Accounts of
Ukraine, the balance of payments, the Statistical Appendix to the November 2005
IMF country report on Ukraine, and Household Budget Survey raw data for the
forth quarter of 2004. The quarterly household survey covers a sample of about
10,060 households and 25,700 household members. We distinguish 16 sectors

(including small-scale and large-scale agricultural production) and 6 factors of
production (including low-, medium-, and high-skilled labor). In disaggregating
the household sector, we deviate from our practice for the more agricultural coun-tries
(such as Moldova) by not including agricultural smallholders separately be-cause
they account for less than 1 percent of all households in Ukraine.
3.3.4. Simulation results
Simulations results for the Ukrainian CGE model are qualitatively similar to
those for Moldova. However, we use a different set of parameter changes to esti-mate
basic macroeconomic indicators for the hypothetical “remittances-and-migration
free” Ukrainian economy, given that migration and remittances in
Ukraine are less predominant than in Moldova. TFP is reduced by 10% (scenario
1), remittances are reduced by 70% (scenario 2), and labor supply increased by 5%
(scenario 3) – see Table 3.2.
Results of simulations reveal that the “pure” effect of remittances was quite
modest. However, indirect effects of migration appear to be much more substan-tial.
First, the assumption that TFP would be lower without migration and remit-tances
implies a substantial indirect welfare gain through migration and remit-tances.
Second, in the case of “no migration”, the economy would have had a lar-ger
labor force and higher GDP. The net effect of remittances (accounting for both
direct and indirect effects) proves to be impressive. The 2004 hypothetical econ-omy
would have lost about 7.1% of its potential without migration and remittances
induced effects.
Light and food industry are the key beneficiaries of demand effects due to re-mittances.
These sectors would have contracted by about 17% and 14% if the
economy did not benefit from workers’ transfers. On the other hand, machinery,
construction and public administration services seem to be quite remittance-neutral
sectors.
All types of households benefit substantially from remittances: their overall
consumption would have been lower by 14 to 21% in the hypothetical “remittance
and migration free” Ukrainian economy of 2004. Rich urban households appear to
win the most, while households with most of income coming as government trans-fers
gain the least from remittances and remittance-induced effects. Noteworthy, in
case of increase in supply of all types of labor (by 5%), rich urban households may
gain an extra 8% (the most) in terms of overall consumption. This reflects the fact
that households of this type possess the highest share of skilled, well-paid labor.

Table 3.2. Moldova: Simulation Results (base values and percentage changes in real
terms)
Base
run
Scenario
1
Scenario
2
Scenario
3
Scenario
4
Scenario
5
Macro variables
Domestic absoprtion 320 -10.2 -2.9 3.9 1.0 -10.5
Private consumption 186 -17.5 -5.0 6.7 1.7 -18.0
Fixed investment 77 0.0 0.0 0.0 0.0 0.0
Government consump-tion
61 0.0 0.0 0.0 0.0 0.0
Exports 211 -9.5 2.9 3.6 6.5 -4.0
Imports -185 -10.9 -1.5 4.1 2.6 -9.2
GDP at market prices 345 -9.5 -0.2 3.6 3.5 -7.1
Real exchange rate 94.6 0.6 1.8 -0.3 1.4 2.3
GDP at factor cost
A_AGR_L 26.6 -18.6 0.5 3.7 4.2 -9.7
A_AGR_S 24.3 0.0 0.0 5.0 5.0 -5.5
A_FOOD 12.3 -14.7 -3.5 5.6 2.1 -14.4
A_LIGHT 2.7 -21.3 -1.4 6.1 4.8 -16.9
A_WOOD 5.1 -10.8 -0.7 4.2 3.5 -8.3
A_CHEM 28.1 -9.3 3.0 3.7 6.7 -3.7
A_MASH 13.4 -6.3 3.0 2.2 5.1 -0.6
A_ELEC 12.6 -11.1 -0.5 4.3 3.8 -8.5
A_CONSTR 14.1 -0.7 -0.1 0.3 0.2 -0.6
A_TRADE 38.6 -11.2 -0.1 4.2 4.1 -8.3
A_REST 2.2 13.1 -0.9 4.9 3.9 -10.2
A_TRANS 29.8 -11.1 -0.5 3.8 4.3 -7.5
A_COMM 11.3 -13.9 -2.7 5.0 2.4 -12.8
A_FIN 42.8 -9.8 -0.3 3.6 3.3 -7.4
A_PUBLIC 14.8 -2.9 -0.2 1.2 0.9 -2.3
A_PUB_SERV 31.3 -5.3 -1.3 2.1 0.8 -5.2
TOTAL 310 -9.2 0.0 3.5 3.5 -6.8
HH_O_RUR 40 -19.7 -5.8 7.4 1.6 -20.3
HH_O_URB 40 -18.3 -4.8 6.7 1.9 -18.4
HH_R_URB 20 -21.9 -5.3 8.4 3.1 -21.3
HH_PUB 20 -17.7 -6.7 6.6 -0.2 -19.7
HH_TRANS 60 -13.5 -3.9 5.6 1.8 -13.8
Note. Scenario 1 – TFP reduced by 10 pc (except in small-scale agriculture). Scenario 2 –
Remittances reduced by 70 pc. Scenario 3 – Labor supply increased by 5 pc. Scenario 4 –
Remittances reduced and labor supply increased. Scenario 5 – TFP and remittances re-duced,
labor supply increased.

3.4. Georgia 7
Georgia is a small country that has seen a significant outflow of migrants and,
at the same time, a large inflow of foreign currency in the recent years. While the
available data only provide an incomplete picture, accumulated net migration since
the beginning of the 1990s exceeded 880,000 individuals (with some return mi-grants
in 2004 and 2005; Statistical Yearbook of Georgia 2007). Inward remit-tances
to Georgia amounted to more than US$ 800 million in 2006, equivalent to
about 10.2% of GDP and 72% of the incoming foreign direct investments (US$
1,100 million)8. The size of unofficial remittances is also large, about US$ 315
million or 39.4% of the total amount of remittances9. Meanwhile, according to the
official sources, about more than a third of population (35%) is below the national
poverty line: Georgia is ranked 97th in the list of countries by human development
index in 2006.
The main purpose of this study is to analyze the poverty effect of remittance
flows in Georgia through direct and indirect channels within the context of the
social accounting matrix based CGE model. The main questions of interest are
whether and to what extent a large size of the remittance flows contributes to the
production and consumption pattern of the poor. Two aspects of poverty reduction
are emphasized in this study: (1) the impact of remittances on the aggregated and
sectoral production growth; and (2) the impact of remittances on poor households,
their production and consumption patterns across regions. In addition, this study
pays particular attention to regional differences in terms of market access and
transaction costs, apart from households’ factor endowments and consumption
patterns reported elsewhere.
The available Georgian data cover the national accounts, including the input-output
table, detailed balance of payments, annual report on household surveys10,
and raw data on household budget surveys for 2004. These surveys were con-ducted
in 3551 households inhabiting the capital city (Tbilisi) and 9 regions
through the questionnaires “Shinda 04” for household expenditures, “Shinda 05”
for private and state transfers to households, and “Shinda 05-1” for households
7 Author of this section: Ainura Uzagalieva. The author thanks Kseniya Tereshchenko for
extracting the raw data on household budget surveys from the Access databases and Levan
Gogoberishvili for providing the Georgian dataset.
8 Georgian-European Policy and Legal Advice Centre (GEPLAC): Georgian Economic
Trends, Quarterly Economic Trends, February, 2008.
9 Georgian-European Policy and Legal Advice Centre (GEPLAC): Georgian Economic
Trends, Quarterly Economic Trends, October, 2007.
10 SDSG: “Households of Georgia”, 2003-2004

income from employment and self-employment which are used in this study11. The
source of the data is the State Department for Statistics of Georgia (SDSG).
3.4.1. General macroeconomic and institutional environment in Georgia
Georgia is a mountainous country with population of 4.5 million and area of
69.7 thousands sq. km. The topographical features of the country’s territory are
very contrasting including the Great Caucasian chain (5,068 meters above the sea
level), the medium height mountains (about 3000 meters) and inner lowlands (e.g.
Kolkheti and Alazani) which are used predominantly for cultivating tea, citrus,
grapes and other agricultural products (the arable area is about 11% of the terri-tory).
There is a total of 12 regions in the country, including a capital region (Tbi-lisi),
two autonomous republics and 9 ordinary regions, which are geographically
and economically very diverse. The macroeconomic structure of the economy, in
terms of the average shares of value added and total output by regions (Table 3.3),
shows that industry and service activities are concentrated mostly in the capital
city Tbilisi and a few other regions located predominantly at inner lowlands (e.g.
Region 4). Agriculture is more widespread across the regions and plays a crucially
important role as a source of production and employment. It accounts for about
21% of the gross value added and represents itself the largest employer of domes-tic
labor (54%).
The macroeconomic situation in Georgia is characterized generally by high
volatility caused by both external and internal factors. A slowdown in the annual
growth rate, from 11.7% in 2003 to 7.5% in 2006, stemmed from the trade em-bargo
imposed by Russia on Georgia in 2005 and 2006 as well as political instabil-ity
and inflation pressures existing within the country. According to the annual
reports of the Georgian central bank, the large sizes of current account (1.2 bn.
USD) and trade (2.0 bn. USD) deficits in 2006 originated from the fall of exports,
followed the Russian trade embargo. High prices for the imported mineral prod-ucts
have also amplified inflation; it rose to 9% in 2006 from the average rate of
5% during the period from 2000 to 2002. Additional inflation factors became the
large inward remittances and capital flows in foreign currency. With underdevel-oped
capital markets, the central bank could not sustain effectively large sterilized
interventions.
11Shinda stands for the Georgian abbreviation of households observation (see State De-partment
for Statistics of Georgia: “Households of Georgia, 2003-2004”).

Table 3.3. The regions and main activities of Georgia
Re-gions
Industry
Hotels
and res-taurants
Transport
and
communi-cations
Construc-tion
Agricul-ture
Georgia, total
including:
100.00 100.00 100.00 100.00 100.00
Kakheti Reg. 1 3.62 0.52 0.11 1.77 14.32
City of Tbilisi Reg. 2 43.47 77.83 78.99 63.00 0.10
Shida Kartli Reg. 3 7.51 1.67 0.07 4.61 7.33
Kvemo Kartli Reg. 4 21.36 1.64 1.87 4.44 18.67
Mtskheta-Mtianeti Reg. 5 3.71 5.32 0.10 2.38 4.71
Samtskhe-Javaketi Reg. 6 3.98 0.62 0.04 0.81 7.66
Adjaria Reg. 7 2.64 5.76 8.04 3.74 5.27
Guria and Racha-
Lechkhumi Reg. 8 1.28 0.79 0.08 1.36 6.88
Samegrelo-Zemo
Svaneti Reg. 9 1.83 2.10 10.32 14.51 16.62
Imereti Reg.10 10.59 3.76 0.37 3.39 18.45
Source: SDSG.
The local market conditions and institutions in Georgia are characterized gen-erally
by fragmented labor and credit markets, while the number of people living
below the poverty line is large. In order to demonstrate the poverty situation in
Georgia, households are grouped into three categories based on the sample of 6754
respondents available in the household surveys. These categories include: 1) the
group of poor households with monthly incomes less than 75 GEL12; 2) the group
of middle-income households with incomes varying from 76 GEL to 200 GEL;
and households whose incomes are higher than 200 GEL are assumed in this study
as high-income or rich ones. According to the country-wide household data, about
43% of respondents lived below the poverty line in 2004. Rural areas have a
higher poverty incidence (52%) than the urban ones (35%).
The poverty profile of household groups by major economic activities (Figure
3.4.2) is analyzed in terms of a head count ratio calculated within each group.
Workers engaged in self-employment, family based business enterprises and farms
have the highest poverty incidence (about 70%). One has to remark that the share
of self-employed workers is very large, equivalent to about 50% of economically
active population. Among the groups considered, private employers have the low-est
poverty incidence of less than 10%.
12 This threshold is chosen because it corresponds to the minimum subsistence level (75
GEL) in Georgia (see SDSG: Statistical Yearbook of Georgia for 2006).

A comparison of the regions in terms of individual income reveals a large dif-ference
in intra-regional poverty (Table 3.4and Figure 3.2). The median level of
household incomes is lower than the mean level and this is true for all regions,
confirming again a very high overall poverty incidence. Moreover, both median
and mean levels vary largely from one region to another, that is from 53 GEL and
74 GEL in Region 7 (Adjaria) to 132 GEL and GEL 177 in Region 2 (Tbilisi).
From the standard deviation values and the shapes of income distributions traced
on the basis of individual incomes for every region, one can observe that differ-ences
in terms of poverty gap are also very large between the regions.
Figure 3.2. The distribution of households by income level
70,00
60,00
50,00
40,00
30,00
20,00
10,00
0,00
Wage employed Private employer Self-employed
outside agriculture
Self-employed in
agriculture
Working in family
enterprises/farms
Total
High: above 200 GEL Middle: 75 GEL-200 GEL Low: below 75 GEL
Note. Axis – households share by income level (%).
Table 3.4. Distribution of labor incomes by regions
Reg.
1
Reg.
2
Reg.
3
Reg.
4
Reg.
5
Reg.
6
Reg.
7
Reg.
8
Reg.
9
Reg.
10
Maximum 1054 2000 500 950 767 3000 600 1000 850 667
Minimum 3 3 5 10 5 7 3 10 3 3
Mean 101 177 92 134 103 143 74 93 123 117
Median 68 132 70 109 70 100 53 70 87 80
Standard
deviation 108 172 78 105 109 210 66 92 113 103
Source: the author’s calculations.

Due to a high incidence of poverty as well as fragmented credit and labor mar-kets,
commercial banks are reluctant to extend loans to clients with low-income
whose land and other assets are considered inadequate as collateral. In addition,
the capital markets and pension fund systems are underdeveloped, while the insur-ance
market is very small (0.3% of GDP). As a result, the poor members of the
society especially in distant regions have limited or no access to credit markets or
employment opportunities. According to Figure 4, households borrow funds more
from physical persons (or other households) rather than from financial institutions
and banks. One should remark also that the size of the borrowed funds varies
largely from one region to another, implying a very limited access to these funds
in some regions (e.g. Region 4, 5, 7 and 10).
Taking into account some of the above-mentioned features of the Georgian
household sector as well as its macroeconomic and institutional environment, the
direct and indirect channels between remittance flows and households well-being
are analyzed. The standard IFPRI CGE model is modified by incorporating the
regional dimension of market access, apart from differences in households’ factor
endowments and consumption patterns. The main questions of interest, in this
respect, are whether and to what extend the poor households groups whose access
to markets is limited can benefit from larger inward remittance flows and, thus,
higher disposable incomes at the national levels.
Figure 3.3. Incomes across and within the regions
70,00
60,00
50,00
40,00
30,00
20,00
10,00
0,00
Total 1 2 3 4 5 6 7 8 9 10
Income level by 10 regions
High: above 200 GEL Middle: 75 GEL-200 GEL Low: below 75 GEL
Note. Axis – Percentage of households (%).

3.4.2. CGE model with regional differences: data calibration and simulation
results
The Georgian aggregated social accounting matrix is based on the standard IF-PRI
approach, constructed on the economy-wide data. It represents 13 production
activities from 18 sectors reported in the original input-output tables. The domestic
production generates 12 commodities. The production of agricultural commodities
is separated between large and small agricultural enterprises. Transaction costs
among institutions, including households, enterprises and government originate in
domestic sales, exports and imports activities. Production factors, which are capi-tal,
labor and self-employment, are decomposed between agriculture and other
production sectors. Labor is split to high-, medium- and low-skill components
using the sub-classification of employed by major work positions.
The original social accounting matrix has been modified in this study by disag-gregating
the small agricultural enterprises into three groups of regions. This was
done in order to enable the regional dimension of the market access and transac-tion
costs. A basic intuition behind this is that farmers located in the remote or
mountainous areas of the country face higher transportation and marketing mar-gins
than other regions. In this respect, three types of household farms with the
highest poverty incidence are distinguished in the model. These are the farm fami-lies
located geographically in regions with high- medium- and low-transaction
costs. The grouping of regions is based on the topographical features of the coun-try’s
territory. In particular, regions located at the mountainous parts are consid-ered
of having high-transaction costs, regions with small cities and arable land
face the medium-transaction costs, and the capital city (Tbilisi) with its surround-ing
is the third, i.e. low-transaction cost area. The regional disaggregation of small
agricultural enterprises into three groups by transaction costs and household
groups in the social accounting matrix is presented in Table 3.5. Clearly, urban
households face lower transaction costs with about 88% of their production activi-ties
concentrated in the regions with low and medium transaction costs. As for
rural households, about 40% of their production activities take place in the regions
with high transactions costs.
In examining the poverty profiles, the household accounts are of particular im-portance
because the flows of income and expenditures need to be adequately
reflected in the social accounting matrix. Therefore, based on the level of incomes
and geographical location, the households of Georgia are classified into six groups
including rural-rich, rural-middle income, rural-poor, urban-rich, urban-middle
income, and urban-poor.

Table 3.5. The distribution of small farmers by transaction costs and household
groups
Household
Urban households Rural households
Com-modi-ties
in
agricul-ture
To-tal
To-tal
Rich Middle
income Poor To-tal
Rich Middle
income Poor
Small agri-culture
with: 1121 582 42 26 11 6 539 51 158 331
- low TC 345 173 1 1 0 0 172 16 51 106
- moderate
TC 347 192 36 22 10 5 155 15 45 95
- high TC 429 217 5 3 1 1 212 20 62 130
Source: Input-output tables and household surveys (SDSG).
Five illustrative scenarios are set out in Table 3.6 for Georgia. The macroeco-nomic
impact of remittance inflows applied homogeneously across all sectors is
strongest on the private household consumption and negligible on the GDP growth
rate. Remittances lead to higher domestic absorption, larger imports and lower
exports. The combined effect of the hypothetical lack of remittance inflows and
emigration is negative with respect to all variables considered, with the strongest
impact on the private consumption, domestic absorption and GDP growth rates
which would be lower by 24.7%, 13.6% and 13.3%, respectively.
At the level of individual sectors, a simulated hypothetical stop in remittance
inflows has a strongest influence on the manufacturing output, which decreases by
14.9% and large-scale agricultural production by about 8.7%. The impact of remit-tances
on the production by household farmers (or small agriculture) is two-fold.
In regions with low and high transaction costs, the production increases by 2.8%
and 1.3%, correspondingly, while in the medium transaction cost regions it falls
by 2.5%. Presumably, moderate transaction costs allow these farmers to decrease
the farm related activities substantially and to get involved into other kind of mar-ket-
related activities, once they receive remittances. The positive effect of remit-tances
is pronounced in the construction (4.1%) and service (e.g. hotel and restau-rants)
sectors (2.6%). And the negative impact is on the electricity sector (-2.8%).
The impact on the remaining sectors is negligible. The combined effect of remit-tance
inflows and emigration is strongest in the small agriculture. Namely, the
production of farms in regions with low, moderate and high transaction costs falls
by 13%, 26% and 17%, correspondingly. Only the large agricultural sectors gain
by about 14.3%.

Table 3.6. The results of SAM based CGEM simulations for Georgia
Base
run
Scenario
1
Scenario
2
Scenario
3
Scenario
4
Scenario
5
Macro variables
Domestic absoprtion 11.3 -2.6 -4.7 7.1 2.4 -15.9
Private consumption 7.3 -4.0 -7.3 11.0 3.7 -24.7
Fixed investment 2.8
Government consumpt. 1.1
Exports 2.6 -12.7 11.9 8.7 20.5 -1.6
Imports -4.2 -1.3 -4.3 5.5 1.2 -12.8
GDP at market prices 9.8 -5.9 -0.3 8.2 7.9 -13.3
Real exchange rate 97.1 -2.9 3.3 1.2 4.3 3.7
GDP at factor cost
Large agriculture and
other primary sectors 0.6 3.8 -8.7 -6.3 -14.3 -40.5
Small agriculture
- low transaction cost 0.2 -9.6 -2.8 17.1 13.5 -12.4
- medium transaction
cost 0.4 -10.9 2.5 22.5 25.5 3.1
- high transaction cost 0.3 -9.0 -1.3 18.7 17.3 -7.5
Manufacturing 0.7 -12.5 14.9 6.6 21.5 5.4
Electricity 0.3 -9.8 2.8 11.6 14.3 -12.7
Processing of products
by households 0.4 -6.3 -1.9 10.1 8.4 -19.2
Construction 0.6 -3.3 -4.1 7.6 3.6 -16.5
Trade and repair of moto
vehicles 1.0 -5.2 -1.1 8.1 7.0 -13.6
Hotels and restaurants 0.3 -6.8 -2.6 11.6 9.1 -18.9
Transportation 0.9 -7.6 0.4 9.5 10.0 -16.4
Communication services 0.4 -7.0 -1.2 10.7 9.5 -16.6
Financial, professional,
other private, services 0.7 -6.4 -0.8 9.4 8.6 -14.7
Public administr./ NGOs 0.7 -3.7 -0.4 5.2 4.7 -7.9
Public services and pri-vate
households 0.8 -4.5 -0.4 6.6 6.1 -9.2
Total 8.4 -6.1 0.0 8.5 8.5 -13.1
Rural poor HH 0.9 -10.5 -1.0 15.5 14.9 -25.6
Rural middle-income HH 1.5 -6.8 -0.9 10.9 10.2 -20.3
Rural rich HH 1.4 -3.4 -7.8 10.5 2.6 -26.0
Urban poor HH 0.6 -4.0 -7.4 11.0 3.4 -23.9
Urban middle-income HH 1.2 -5.8 -5.0 11.8 6.8 -22.1
Urban rich HH 1.8 2.1 -16.9 8.9 -8.6 -29.0
Note. Scenario 1 – TFP reduced by 20 pc. Scenario 2 – Remittances reduced by 70 pc.
Scenario 3 – Labor supply increased by 20 pc. Scenario 4 – Remittances reduced and labor
supply increased. Scenario 5 – TFP and remittances reduced, labor supply increased.

In terms of households groups included in the model, the results reveal that
emigration and remittance flows do not affect all residents symmetrically, but
depend on the identity of households. In urban areas, remittances contribute to the
increase of household incomes and consumption smoothing, while in rural areas
the effect is positive, but rather week. For example, the groups of rural poor and
middle-income households gain about 1% of their private consumption each,
while in urban areas these groups gain 7.4% and 5.0%, correspondingly. One can
observe also that the magnitude of this impact is smaller compared to that on rich
households with the pure effect of remittances equivalent to 16.9% and 7.8% of
private consumption, respectively, in urban and in rural areas. Consequently, re-mittances
are beneficial to the wealthier members of this society (i.e. rich house-holds)
in both urban and rural areas. An increase in the supply of labor by 20%, on
the contrary, would improve the welfare of households in all groups. The rural
poor would benefit most, with 16% higher private consumption, while this effect
would be less pronounced for rich urban households, with about 9% higher private
consumption.
The conclusion to be drawn from this exercise is that, while having a strong
macroeconomic growth effect at the aggregated level, emigration and inward re-mittance
flows do not affect all sectors and residents symmetrically. Moreover,
they have a rather limited impact on reducing poverty and income inequality. In
urban areas, for example, remittances contribute to the increase of household in-comes
and consumption smoothing, while in rural areas the effect is two-fold.
Namely, in regions with low and high transaction costs, remittances are beneficial
to small farmers, while in the medium transaction cost regions the effect is oppo-site.
Presumably, the moderate level of transaction costs allows these farmers to
shift from subsistence to market-oriented activities, once they have access to re-mittances.
The magnitude of the impact caused by remittances on the consumption
pattern is smallest for the group of poor and middle-income rural households
(1.6% and 1.0%, respectively) and largest for rich urban households. Under the
absence of remittances, rich households would suffer a loss of about 16.9% in
their private consumption. Consequently, the wealthier members of the society
gain more from remittances than poorer household categories. Better access to
labor markets, on the contrary, would improve the welfare of many, especially, of
the rural poor at the outset.
Policy priorities, in these circumstances, should be given to a pro-poor ap-proach,
especially, in improving institutional mechanisms through which the poor
members of the society can have access to labor and credit markets within the
country. With the focus on the inclusion of low-income and rural households in
the financial sector, for example, the policies could be designed for meeting the

needs of household farmers in distant regions. This would include also enabling
various possibilities for linking remittance flows to the microfinance based
mechanisms focused on promoting savings, insurance and investment within re-gions,
as well as decreasing transaction costs across the regions.
3.5. Kyrgyzstan 13
Migration in the Kyrgyz Republic has become very intensive since the collapse
of the Soviet Union. Between 1990 and 2005, around 500,000 people left the
country permanently. Many of these were Russian-speaking Kyrgyz citizens who
left Kyrgyzstan for permanent residence in Russia and elsewhere. However, dur-ing
the last five years, non-permanent labor migration has increased sharply, espe-cially
in rural areas with high unemployment. The information on the scope of
labor migration is fragmentary and based mostly on the information from the Kyr-gyz
embassies abroad. According to the conservative estimates of State Committee
on Migration, about 250,000-300,000 individuals from Kyrgyzstan are currently
working in Russia and around 100,000 in Kazakhstan.
The National Bank of the Kyrgyz Republic (NBKR) is responsible for compil-ing
the balance of payments and consequently for statistics on remittances at the
macro level. Since NBKR does not have all required information, it has to rely on
several sensitive assumptions and existing data on money flows to estimate the
magnitude of remittances. In particular, it has the following information on cross-border
money flows involving participants in the Kyrgyz Republic:
1. Data on repatriated wages and salaries of foreign employees, who are not
residents of the Kyrgyz Republic and are working in large joint ventures
in the Kyrgyz Republic such as Kumtor Operating Company, the largest
gold-mine enterprise;
2. Data on international money transfers to and from the Kyrgyz Republic
done by individuals through banking accounts including card accounts
(transactions below or equal to $3,000 are considered as workers' remit-tances),
money transfer systems (e.g., the Western Union), as well as the
postal system; and
13 Authors of this section: Aziz Atamanov and Roman Mogilevsky

3. Data on the number of permanent migrants to and from the Kyrgyz Re-public
and the estimated average value of the property that they bring
with them to the country of destination.
Important limitation which NBKR faces is the lack of the information on the
status of workers abroad and the economic nature of the transactions they made
(e.g., intra-family transfer, payment for goods/services, person-to-person loan
disbursements, etc.). Hence the current official statistics of remittances represent a
mix of money flows of different economic nature.
There were several surveys aimed at estimating the magnitude of remittances
and the profiles of labor migrants. Till recently on the micro level the household
budget survey conducted by the National Statistical Committee did not distinguish
remittances as a separate source of income. The most recent report prepared in the
framework of the Asian Development Bank’s project on remittances and poverty
and financial sector development in Kyrgyzstan (Mogilevsky and Atamanov 2008)
contains the most comprehensive analysis of the impact of remittances on poverty
and financial sector in Kyrgyzstan. 3995 households were interviewed in the
framework of this project, providing detailed information on the profile of labor
migrants and remittances' recipients.
According to this survey, there were 251.5 thousands labor migrants working
abroad that account for 5% of the total population and 8.1% of the working-age
population. The regional distribution of the migrants reflects the regional level of
economic development: About 70% of all migrants are from rural areas, 10% are
from Bishkek and 21% are coming from other urban areas. The vast majority of
the labor migrants choose Russian Federation (82.5%), followed by Kazakhstan
(12%) and other countries (5.5%).
Most migrants are employed in the private sector dealing either with construc-tion
(45% of the total number) or trade activities (30.4%). Almost half of the mi-grants
are seasonal workers. However, migrants from Bishkek differ from other
regions of the country. More of them are employed in sectors requiring higher
education and higher qualifications (financial intermediation, public administra-tion,
education, health care, etc.).
According to the survey, the magnitude of remittances coming to the country
differs significantly from the official NBKR estimates which were growing expo-nentially
during last years ($48.6 million. in 2002 and $730 million in 2006).
However, the in-depth analysis shows that this rapid growth results mostly from
increase in large and very large transactions, which cannot necessarily be attrib-uted
to labor migration.
ADB's estimates on remittances are significantly lower than official numbers
and range from US$ 223.7 to US$ 287.1 million, with a confidence interval of

95%. Most remittances are going to the rural areas. The average amount of cash
remittances per household in 2006 is US$ 1,331, accounting for 50% of the total
average household income.
Remittances come to Kyrgyzstan mainly through banks and MTOs (78.5%), on
the second place are households themselves (25.6%), on the third place are friends
and relatives as a source of transfer; the role of postal services is negligible. All
these show that the role of informal intermediaries is insignificant.
Simple correlation analysis of the workers' remittances (NBKR data since this
is the only source of time series on remittances) and various macroeconomic indi-cators
reveals that the growth of remittances contributes to some growth in private
consumption and GDP, has a positive impact on imports and indirectly on gov-ernment
revenues, can be associated with some employment growth in informal
sector and does not produce measurable impact on inflation and real exchange
rate14.
To have a more detailed picture on the impact of remittances on the Kyrgyz
economy and the welfare of its citizens, a CGE model was used. The underlying
social accounting matrix (SAM) for Kyrgyzstan is based on the national accounts
for 2000-2004, input-output tables for 2003, and the household budget surveys for
2003-2004. It was also possible to draw on an existing SAM prepared by the
World Bank for 2003, which has been adjusted to the required format for the stan-dard
IFPRI model and extended by disaggregating the household sector for com-parable
analysis with the other countries in this study. Specifically, the existing
SAM has 91 sectors and households are divided into deciles for urban and rural
areas. We aggregated it up to 14 sectors and constructed representative household
groups comparable with the groups used in Moldova’s SAM, using data from the
household budget survey.
Our household categories are (i) agricultural smallholders with more than half
of their total income from small-plot farming; (ii) other rural households; (iii) rich
urban households (top 2 deciles by consumption); (iv) other urban households; (v)
public employee households which draw more than half their income from public
administration, health and social services; and (vi) pensioners with more than half
of their total income from state transfers. Labor income is disaggregated into (i)
income from low-skilled labor: head of household has general secondary or lower
education; (ii) income from medium-skilled labor: head of household has special
secondary or incomplete higher education; (iii) income from high-skilled labor:
14 However, the situation has changed dramatically in 2007, when inflation was about 20%
which is more than cumulative inflation during the last 4 years. Such a sharp increase was
triggered by external price shocks, but the impact of monetary factors cannot be neglected.

head of household has higher education; (iv) income from non-agricultural em-ployment:
household head is self-employed.
One could identify the following channels of labor migration’s influence on the
Kyrgyz economy: (i) increase in remittances, (ii) outflow of excess labor from
agriculture creating conditions of some restructuring in the sector (e.g. some in-crease
in farm size, more commercial orientation of farms run by households
somewhat insured by remittance flows), probably leading to an increase in total
factor productivity in this sector; and (iii) brain drain or, in terms of this model,
outflow of high-skilled labor. Importantly, outflow of lower-skilled labor is con-sidered
to have no effect on labor supply as this labor is effectively unemployed
on domestic market.
This considerations allow for formulating four scenarios for simulation of eco-nomic
impact of labor migration:
1. Increase in remittances by 40%;
2. Increase in remittances by 40% accompanied by increase in total factor
productivity in agriculture by 5%;
3. Increase in remittances by 40% accompanied by reduction in high-skilled
labor supply by 10%; and
4. Increase in remittances by 40% accompanied by increase in total factor
productivity in agriculture by 5% and reduction in high-skilled labor
supply by 10%.
The simulation results (Table 3.7) suggest that higher remittances (Scenario 1)
positively influence private consumption and production in sectors oriented
mainly on domestic market (e.g. agriculture, food processing, trade). Higher remit-tances
also allow financing increased imports. However, the inflow of foreign
exchange into the country leads to real exchange rate appreciation with negative
consequences for Kyrgyz exports and production in export-oriented (e.g. mining
or machine production) and import-competing (e.g. light industry) sectors. Under
this scenario, the overall effect of remittances on GDP appears to be near neutral.
Accounting for a TFP increase in agriculture along with higher remittances
(Scenario 2) results in a more favorable performance of several GDP components,
with even stronger (6%) growth of private consumption and smaller reduction in
exports. On the contrary, assessment of combination of remittances and brain drain
(Scenario 3) leads to much more modest private consumption growth and severe
deterioration in exports (7.5% decline). Finally, under the last scenario of a simul-taneous
40% increase of remittances, 5% improvement in farm’s total factor pro-ductivity
and 10% outflow of high-skilled labor, GDP increases by 0.6% due to
growth in private consumption by 4.6% and despite of 5.9% reduction in exports.

Table 3.7. Kyrgyzstan: Simulation Results (base values in mln Kyrgyz soms and per-centage
changes in real terms)
Base run Scenario 1 Scenario 2 Scenario 3 Scenario 4
Macro variables
Domestic absoprtion 99402 2.5 4.2 1.5 3.2
Private consumption 68956 3.6 6.0 2.2 4.6
Exports 30766 -6.4 -4.8 -7.5 -5.9
Imports 38627 1.6 2.9 0.8 2.0
GDP at market prices 91541 -0.1 1.7 -1.2 0.6
GDP at factor cost
Agriculture 28094 1.5 4.4 0.5 3.3
Food processing 1575 2.2 4.7 1.0 3.5
Light industry 281 -0.7 4.4 -2.0 3.0
Wood, furniture and
printing 263 -3.3 -2.3 -5.9 -5.0
Mining and chemical
industry 8459 -7.7 -6.8 -7.9 -6.9
Machine production 1265 -4.9 -4.4 -8.0 -7.5
Electricity and gas
distribution 3576 0.0 1.5 -1.1 0.4
Construction 2301 -0.1 0.2 -0.4 0.0
Wholesale and retail
trade 12286 2.4 4.3 1.3 3.2
Hotels and restaurants 1217 2.9 4.7 1.6 3.4
Transportation and
communication 4210 -1.2 -0.2 -3.1 -2.0
Financial services 3161 -1.2 -0.3 -5.0 -4.1
Public administration 3839 0.0 0.2 -0.3 -0.2
Public and other ser-vices
5692 0.7 1.2 0.1 0.5
TOTAL 76219 0.0 1.8 -1.1 0.7
Farming HH 7167 5.7 6.9 3.9 5.0
Other rural HH 19735 2.6 5.4 0.6 3.3
Rich urban HHs 12727 3.5 5.6 2.2 4.3
Other urban HH 11131 4.3 6.9 2.8 5.4
Public sector employee
HH 8704 5.5 8.1 6.1 8.7
Transfer receiving HH 9300 3.6 6.1 1.9 4.4
Note. Scenario 1 – Remittances increased by 40%. Scenario 2 – Remittances increased by
40% and TFP in agriculture increased by 5%. Scenario 3 – Remittances increased by 40%
and high-skilled labor supply reduced by 10%. Scenario 4 – Remittances increased by
40%, TFP in agriculture increased by 5% and high-skilled labor supply reduced by 10%.

All types of households win from increase in remittances. Depending on sce-nario,
some of them win more and others win less. Under the last, “summary”
scenario public sector employee households, farming households and other (e.g.,
more poor) urban households win more, and transfer-receiving and other rural
households win relatively less from migration. These results reasonably probably
reflect the fact that farming and poorer urban households have more migrant
members, while transfer-receiving (e.g. pensioners) households send fewer mi-grants
to work abroad.

4. Country Study: Russia
4.1. Data availability and quality
The data on immigration into Russia is notoriously unreliable due to the preva-lence
of illegal migration. State agencies report the number of the so-called “regis-tered
immigrants,” i.e. those who comply with the laws that require that both Rus-sian
citizens and foreigners register with the authorities upon arrival to Russia (or
to a new location). Alternatively, there is official data on the number of work per-mits
issued by the migration authorities and the number of foreigners legally em-ployed
as reported by the companies. Thus, official sources put the number of
immigrants arriving to Russia in 2006 at slightly over 0.186 million (see Table
4.1), with the CIS countries accounting for the most of this volume (0.177 mil-lion).
Among them Kazakhstan, Uzbekistan and Ukraine sent over 30 thousands
migrants each. The leading sources of immigrants from outside of the CIS were
Germany and Israel. This number holds relatively constant starting with 2001,
with the exception of 2003-2004 when it declined to roughly 0.120 million. In
2000, however, the total number of immigrants was almost twice as high (0.359
million), but already the next year it dropped to its current level.
Overall, according to the Russian State Statistical Agency, there were 1.014
million legally employed foreign nationals in Russia in 2006 (see Table 4.3), up
from 0.7025 million a year earlier and 0.213 million in 2000. The largest sources
of foreign workers in Russia, according to that data, were China (0.2108 million),
Ukraine (0.1713), Uzbekistan (0.1051 million), Turkey (0.1014 million), and Taji-kistan
(0.0987 million). Overall, over half of all foreign workers arrived from the
CIS countries (0.5377 million). Foreign workers are predominantly male (0.8585
million in 2006), while women account for only a small fraction of the total (0.156
million). The single most popular destination for immigrant workers is the con-struction
industry employing 0.4141 million foreign workers in 2006 up from
0.2721 a year earlier. Next comes the retail and wholesale trade (including repair
services) with 0.2709 million foreign workers. Agriculture and transportation em-ploy
over 40 thousand foreigners each, while “other services” employ less than 30
thousand. The presence of foreign workers in other industries is marginal. Nearly
half of all foreign workers are concentrated in Moscow.

Table 4.1 Inflow of migrants into Russia (number of people)
1997 2000 2001 2002 2003 2004 2005 2006
Total 597651 359330 193450 184612 129144 119157 177230 186380
CIS countries 571903 346774 183650 175068 119661 110374 168598 177657
Azerbaijan 29878 14906 5587 5635 4277 2584 4600 8900
Armenia 19123 15951 5814 6802 5124 3057 7581 12949
Belarus 17575 10274 6520 6841 5309 5650 6797 5619
Georgia 24517 20213 9674 7128 5540 4886 5497 6806
Kazakhstan 235903 124903 65226 55706 29552 40150 51945 38606
Kyrgyzstan 13752 15536 10740 13139 6948 9511 15592 15669
Moldova 13750 11652 7569 7562 6391 4816 6569 8649
Tajikistan 23053 11043 6742 5967 5346 3339 4717 6523
Turkmenistan 16501 6738 4402 4531 6299 3734 4104 4089
Uzbekistan 39620 40810 24873 24951 21457 14948 30436 37126
Ukraine 138231 74748 36503 36806 23418 17699 30760 32721
Non-CIS coun-tries
25748 12556 9800 9544 9483 8783 8632 8723
Australia 57 27 25 22 30 42 30 28
Afghanistan 208 288 171 107 82 55 60 86
Bulgaria 750 245 255 238 212 125 118 109
Germany 2379 1753 1627 1962 2692 3117 3025 2900
Greece 183 182 124 150 224 182 200 176
Israel 1626 1508 1373 1670 1808 1486 1004 1053
Canada 73 50 74 70 103 87 99 77
China 2861 1121 405 410 346 212 432 499
Cuba 110 37 42 22 23 12 17 12
Latvia 5658 1785 1283 990 906 819 726 766
Lithuania 1785 945 758 722 535 339 360 371
Poland 247 61 56 53 39 48 55 48
Syria 483 358 260 144 101 56 68 67
USA 668 439 432 455 484 518 396 411
Turkey 176 164 184 144 112 77 86 172
Finland 140 83 97 136 125 141 129 137
Sweden 32 14 28 19 22 16 23 32
Estonia 3483 786 535 534 445 446 432 347
Other 4829 2710 2071 1696 1194 1005 1372 1432
Source: Table 5.9. International Migration, “Russia in Figures 2007”, Russian State Statis-tical
Agency
Table 4.2 Outflow of migrants from Russia (number of people)
1997 2000 2001 2002 2003 2004 2005 2006
Total 232987 145720 121166 106685 94018 79795 69798 54061
CIS countries 146961 82312 61570 52099 46081 37017 36109 35262
Azerbaijan 4302 3187 2170 1704 1771 1336 1274 1366
Armenia 2578 1519 1362 1114 1098 654 620 686
Belarus 18928 13276 11175 8829 7016 5671 6034 6318

1997 2000 2001 2002 2003 2004 2005 2006
Georgia 3286 1802 1339 964 939 740 691 593
Kazakhstan 25364 17913 15186 13939 14017 12504 12437 11948
Kyrgyzstan 6296 1857 1333 1080 959 656 473 605
Moldova 5715 2237 1660 1385 1234 907 786 636
Tajikistan 2474 1158 993 827 922 549 434 424
Turkmenistan 1532 676 352 272 251 168 125 112
Uzbekistan 7370 3086 1974 1400 1130 717 595 648
Ukraine 69116 35601 24026 20585 16744 13115 12640 11926
Non-CIS coun-tries
86026 63408 59596 54586 47937 42778 33689 18799
Australia 297 176 184 144 146 167 209 167
Afghanistan 146 25 18 7 17 2 11 11
Bulgaria 668 180 163 133 156 160 124 116
Germany 48363 40443 43682 42231 36928 31876 21458 8229
Greece 886 314 204 190 186 157 155 139
Israel 12873 9407 4835 2764 2048 1733 1745 1408
Canada 1333 841 812 725 701 783 628 552
China 1222 658 156 151 86 154 456 196
Cuba 89 27 15 6 8 8 2 3
Latvia 636 365 311 256 259 226 211 223
Lithuania 1162 376 262 293 268 282 213 228
Poland 376 135 84 80 72 57 76 84
Syria 256 54 60 66 58 55 54 42
USA 9087 4793 4527 3134 3199 2919 4040 3109
Turkey 356 104 96 80 88 60 85 78
Finland 923 1142 980 1110 737 910 737 695
Sweden 151 195 148 162 151 158 110 132
Estonia 702 385 402 321 351 265 225 270
other 6500 3788 2657 2733 2478 2806 3150 3117
Source: Table 5.9. INTERNATIONAL MIGRATION, “Russia in Figures 2007”, Russian
State Statistical Agency.
This official data on “registered migration,” however, seems to significantly un-derestimate
the volume of migration flows, especially the volume of short-term mi-gration
of low-skilled workers from the CIS countries. On the one hand, citizens of
most of the CIS countries do not require visas to travel to Russia which encourages
migration. On the other, the system of registration in force in Russia throughout the
2000s was extremely burdensome. In order to obtain a registration an immigrant had
to satisfy a number of conditions, most importantly, he had to have a place of resi-dence
and, moreover, to obtain a permission from his landlord to be registered at this
particular address. The very fact of registration made a migrant (and his tax-evading
landlord) “visible” and thus exposed him or her to extortion by the corrupt officials.
Obtaining work permits was similarly burdensome. Employers also had all the in-

centives to encourage illegal immigration, for it gave them significant leverage over
their unregistered and thus legally vulnerable workers.
Table 4.3. Number of foreign nationals employed in Russia (according to the Federal
Migration service data (thsd people)
2000 2001 2002 2003 2004 2005 2006
Total 213.3 283.7 359.5 377.9 460.4 702.5 1014.0
Non-CIS countries 106.9 135.1 154.9 197.4 238.5 358.7 476.1
Vietnam 13.3 20.1 26.7 35.2 41.8 55.6 69.1
China 26.2 38.6 38.7 72.8 94.1 160.6 210.8
North Korea 8.7 9.9 12.7 13.2 14.7 20.1 27.7
USA 1.8 2.0 1.5 1.8 1.9 2.9 3.7
Turkey 17.8 20.9 15.4 37.9 48.0 73.7 101.4
From CIS countries 106.4 148.6 204.6 180.5 221.9 343.7 537.7
Azerbaijan 3.3 4.4 15.0 6.0 9.8 17.3 28.3
Armenia 5.5 8.5 12.6 10.0 17.0 26.2 39.8
Georgia 5.2 5.0 6.8 3.2 3.8 4.3 4.9
Kazakhstan 2.9 3.6 7.6 4.0 4.3 4.1 5.0
Kyrgyzstan 0.9 1.7 6.4 4.8 8.0 16.2 33.0
Moldova 11.9 13.3 40.7 21.5 22.7 30.6 51.0
Tajikistan 6.2 10.0 16.8 13.6 23.3 52.6 98.7
Turkmenistan 0.2 0.1 7.0 0.2 0.3 1.5 0.7
Uzbekistan 6.1 10.1 15.5 14.6 24.1 49.0 105.1
Ukraine 64.1 91.9 61.0 102.6 108.6 141.8 171.3
Source: Russian State Statistical Agency.
As a result, even the authorities themselves acknowledge the presence of a
huge number of illegal immigrants. The magnitude of illegal migration is also
underlined by the fact that after the registration procedure was reformed in 2007
(now foreigners don’t need to ask the officials for registration, it is enough for
them to simply notify the authorities about their new place of residence) the num-ber
of foreigners registering with the Federal Migration Service jumped, according
to some accounts, threefold. Still, it is hard to get a reasonable estimate of the vol-ume
of migration. At the end of 2006 the head of the Federal Migration Service
estimated the number of illegal immigrants in Russia at 10 million, while the
United Nations Population Division (World Migrant Stock) estimated the number
of international migrants in Russia at mid-year (both sexes) in 2005 at 12 million,
about 6 million of them female; this number, according to the UN Population Di-vision
holds roughly constant throughout 1990s-2000s. However, independent
experts in Russia criticize these estimates as overblown. According to the esti-mates
by the Centre for Migration Studies in Moscow, the stock of immigrants
present in Russia at a given moment is about 7 million. Roughly 0.5 million for

them are expatriate workers from the developed countries, while 5 to 6 million are
“migrant labor.” No more than 0.7 million of them come through official channels,
while the rest arrive on their own. Among these about 20% follow the necessary
legal procedures obtaining both registration and work permit, 30% register, but work
without permit, and the rest (i.e. 50%) have neither registration nor permit (these
estimates describe the situation before the 2007 reform of registration system).
Table 4.4. Foreign nationals employed, by sector (Federal Migration Service data)
Thousand people Percentage of the
total by occupation
2005 2006 2005 2006
Total 702.5 1014.0 100.0 100.0
Agriculture and hunting 33.4 44.1 4.8 4.4
Forestry 17.8 28.5 2.5 2.8
Fishing, fish farms 0.7 1.2 0.1 0.1
Mining and quarrying 12.6 19.9 1.8 2.0
Manufacturing 48.7 72.2 6.9 7.1
Construction 272.1 414.1 38.7 40.8
Wholesale trade and commission trade;
repair of motor vehicles, motorcycles;
213.9 270.9 30.4 26.7
personal and household goods
Transport 29.5 43.3 4.2 4.3
Communication 3.5 3.7 0.5 0.4
Financial intermediation 1.9 4.2 0.3 0.4
Real estate, renting and business activi-ties
0.5 0.9 0.1 0.1
Computer-related activities 0.6 0.8 0.1 0.1
Science and research 0.3 0.3 0.0 0.0
Geophysical, geological exploration 3.1 2.3 0.4 0.2
Land-surveying and cartographic activi-ties
0.0 0.4 0.0 0.0
Education 1.0 1.0 0.1 0.1
Health and social work 2.2 2.8 0.3 0.3
Other community, social and personal
service activities 13.6 24.1 1.9 2.4
Other activities 46.8 79.0 6.7 7.8
Source: “Russia in Figures 2007”, Russian State Statistical Agency.
Official data on emigration from Russia should also be treated cautiously, for it
accounts only for those who have chosen to register with the Russian authorities as
residing or working abroad; however, citizens have no incentives to do so, not
there are any enforcement mechanisms. With all that said, the number of depar-tures
from Russia in 2006 is officially put at 54,061, with 35,262 among them
going to the CIS countries. This is a marked decrease compared to the level of

2000 (145,720). Top destinations include Kazakhstan, Ukraine, Germany, and
USA. However, the official data on the number of Russians that left the country in
2006 in order to work abroad is put at 65,747.
Table 4.5. Stock of migrants according to WB Migration and Remittances Fact book
Russian Federation ( Europe & Central Asia; Upper middle income)
Population (millions, 2006) 142 Surface area (1,000 sq. km, 2006) 17098
Population growth (avg. annual %,
1997–2006) -0.4 GNI ($ billions, 2006) 958
Population density (people per sq.
km, 2006) 9 GNI per capita, Atlas method ($,
2006) 5780
Labour force (millions, 2006) 73 GDP growth (avg. annual %,
2002–2006) 7
Urban population (% of pop., 2006) 72.9
Age dependency ratio 0.4
Poverty headcount ratio at national
poverty line (% of pop., 2004) 0
Emigration, 2005
Stock of emigrants 11,480,137
Stock of emigrants as percentage of population: 8.0%
Top 10 destination countries: Ukraine, Kazakhstan, Belarus, Israel, Uzbekistan, United
States, Latvia, Germany, Moldova, Estonia.
Skilled Emigration, 2000
Emigration rate of tertiary educated 1.3%
Emigration of physicians
1,875
as % of physicians trained in the country
0.3%
Immigration, 2005
Stock of immigrants 12,079,626
Stock of immigrants as percentage of population 8.4%
Female as percentage of immigrants 57.8%
Refugees as percentage of immigrants 0.0%
Top 10 source countries: Ukraine, Kazakhstan, Belarus, Uzbekistan, Azerbaijan, Georgia,
Armenia, Kyrgyz Republic, Tajikistan, Moldova
Source: Migration and Remittances Factbook is compiled by Dilip Ratha and Zhimei Xu,
Migration and Remittances Team, Development Prospects Group, World Bank. More in-formation
on other countries and regions are available at
www.worldbank.org/prospects/migrationandremittances.
There is increasing attention on international migrant remittances as a devel-opment
policy and financial stability issue. Official statistics on remittances are
primarily collected and reported through the balance of payments statistics, but
they are incomplete in many countries, may underreport remittance flows, and are
often not comparable15.
15 IMF Remittances Statistics portal (http://www.imf.org/external/np/sta/bop/remitt.htm).

The primary source of information about trans-border money transfers, includ-ing
remittances is the Central Bank of Russia (CBR). According to the CBR16 total
money transfers to Russia amounted to US$ 7,500 million in 2006, with most of
the funds transferred in favor of residents - 66% (US$ 5.0 billion) compared with
55% (US$ 3.5 billion) in 2005. Total outward money transfers were estimated at
US$ 18,800 million in 2006.
Cross-border remittances via money transfer systems accounted for 32% of
outward money transfers from Russia in 2006 (28% in 2005) and for 17% of in-ward
money transfers to Russia (16% in 2005).
The flow of remittances via money transfer systems has risen considerably in
the past few years. It amounted to US$ 7.3 billion in 2006. Between 2003 and
2006, the volume of remittances grew at an annual rate of 150-160%. The high
growth rate was largely maintained by remittances from Russia.
The value of remittances from Russia increased by 69% in 2006 on annual ba-sis
up to US$ 6.0 billion. Compared with 2004, they grew by 190%. Remittances
to Russia grew slower. Their value increased by 25% in 2006 on a year-to-year
basis up to US$ 1.3 billion. Compared with 2004, they grew by 68% (Table 4.6).
The CIS countries accounted for most of remittances from Russia in 2006
(90%). Their value totaled $5.4 billion. Money transfers to Russia came both from
non-CIS countries (US$ 0.7 billion) and CIS countries (US$ 0.6 billion) in 2006.
The average sum of one remittance from Russia increased by 19% in 2006
(comparing to 2005) to US$ 546 per transaction. The average remittance to non-
CIS countries was twice as much as the sum of one remittance to CIS countries
(US$ 1,349 and US$ 511, respectively).
In 2006, Uzbekistan (16.7%), Tajikistan (15.9%), Ukraine (15.4%), Armenia
(10.1%) and Moldova (8.7%) were the main recipients of money remittances from
Russia carried out via money transfer systems. Ukraine registered the smallest
average remittance (US$ 364) from Russia.
The main remitting countries were the United States (14.6%), Kazakhstan
(11.8%), Uzbekistan (7.0%), Ukraine (6.8%) and Germany (4.8%).
The geography of inward remittances was wider than that of outward remit-tances.
In 2006, the top three remitting countries contributed to one-third of money
received in Russia whereas the top three recipient countries accounted for almost
50% of remittances from Russia.
16 http://www.cbr.ru/eng/statistics/CrossBorder/Cross-border_06_e.pdf.

Table 4.6. Remittances according to the Central Bank of the Russian Federation
2003 2004 2005 2006 2007
Total, $ million
Money Transfers from Individuals in
the Russian Federation 1,310 2,070 3,549 6,005 9,444
to non-CIS countries … … 323 622 868
to CIS countries … … 3,226 5,382 8,575
Money Transfers in favour of Indi-viduals
to the Russian Federation 588 777 1,041 1,304 1,681
from non-CIS countries … … 645 746 813
from CIS countries … … 396 559 868
Balance … … -2,507 -4,700 -7,763
with non-CIS countries … … 322 123 -55
with CIS countries … … -2,829 -4,824 -7,707
Average remittance, $
Money Transfers from Individuals in
the Russian Federation … … 457 546 623
to non-CIS countries … … 1,084 1,349 1,407
to CIS countries … … 432 511 589
Money Transfers in favour of Indi-viduals
to the Russian Federation … … 427 480 579
from non-CIS countries … … 506 536 583
from CIS countries … … 340 422 575
Source: Cross-border Remittances via Money Transfer Systems Reports of the credit insti-tutions
"Data on Transfers of the Individuals from the Russian Federation and Transfers to
the Russian Federation for the Benefit of the Individuals and Transactions of the Resident
Individuals with the Non-residents on the Territory of the Russian Federation", other re-ports
of the credit institutions, reports of money transfer systems and of the government
post service "Pochta Rossii".
Remittances without quid pro quo17 accounted for 27% of money transfers from
abroad in favor of individuals in 2006. Their value increased by 29% (US$ 0.3
billion) to US$ 1.4 billion in 2006. 80% (US$ 1.1 billion) of this amount was
channeled via money transfer systems. Transfers in favor of non-residents totaled
$2.5 billion in 2006 compared with US$ 2.9 billion in 2005. Their share in the
structure of transfers dropped from 45% in 2005 to 34% in 2006.
As for the country profile of money transfers, Kazakhstan was the only CIS
country, which remitted more funds to Russia than it received. The transfer of
money from the sale of property by migrants leaving Kazakhstan for permanent
residence in Russia played a noticeable role in this balance.
17 Remittances without quid pro quo include grants, donations, compensation payments,
scholarships, pensions, alimonies, inheritance payments, gifts, and also all remittances
made via money transfer systems.

The largest amount of money transfers from Russia went to China (US$ 2.5 bil-lion)
compared with other countries. These operations were largely conducted by
non-residents. In terms of value, they accounted for one-third of all money trans-fers
by non-residents from Russia. The average transaction was relatively large
and equaled about US$ 20,000, exceeding by 30 times the amount of one money
transfer by a non-resident from Russia to CIS countries. Over 60% of these trans-actions
were performed by non-resident individuals from the Far Eastern and Urals
Federal Districts of Russia, with the average size of a transaction equaling US$
50,000. According to estimates, these transactions are related to payments for
goods imported into Russia without proper customs clearance rather than to remit-tances
without quid pro quo.
More than half of money transfers to Turkey were made by residents. Payments
for the import of consumer goods and the services of travel agencies accounted for
one-third of that amount. According to the CBR estimates, remittances by non-residents
to Turkey largely included the transfer of part of wages received by
Turkish citizens working in Russia to their families.
Among ten main recipient countries, the largest average transfer went to Swit-zerland
(US$ 22,071) and the United Kingdom (US$ 26,256) while the smallest
sums went to CIS countries: Ukraine ($491), Uzbekistan ($666), Tajikistan ($648)
and Armenia ($754). According to CBR estimates, money transfers to the first
group of countries were related to transactions conducted by residents from their
accounts with Russian banks to their accounts with foreign banks. Meanwhile,
transfers to the second group of countries were linked to transactions carried out
by individuals temporarily working in Russia. An average transaction was larger
in Uzbekistan and Tajikistan than in Ukraine because money was consolidated by
one trusted person.
According to CBR estimates, the average fee per remittance abroad via money
transfer systems was 3.7% of the remittance sum in 2006 (Figure 4.1). At the same
time, the average fee was $20 for the average remittance of $500-$600. The size of
remittance fee has decreased considerably in the past few years. The largest reduc-tion
was registered for remittances ranging from $100 to $200. In 2006, the remit-tance
fee fell 7% on average compared with 2005 and 13% compared with 2004.
According to Word Bank estimates, the volume of inward remittance flow to
Russia in 2006 was US$ 3,091 million, with the stock of emigrants estimated at
11.5 million. As one can see WB’s estimates are twice as big as CBR’s are. This is
the case of inconsistency in statistical methods or incorrect terminology. CBR’s
figures on remittances include only personal trans-border transfers, contrary to the
WB methodology which includes trans-border compensation of employees.

Figure 4.1. Average Fee for Remittances from Russia via Money Transfer Systems
(% of remittance)
Source: CBR’s survey of cross-border transactions of individuals:
http://www.cbr.ru/eng/statistics/CrossBorder/Cross-border_06_e.pdf.
The volume of remittances sent home by foreign workers in Russia is estimated
by the World Bank (Migration and Remittances Factbook) at US$ 11,436 million
(2006), or 11.2% of GDP, of which US$ 4,587 million are workers’ remittances,
and US$ 6,038 million is defined as compensation of employees. The volume of
remittances in 2006 increased sharply compared to the previous year ($6,989) and
even more so, compared to 2000 (US$ 1,101 million). These estimates are made
on the assumption that the stock of immigrants in Russia in 2005 was 12 million,
as estimated by the UN Population Division.
Statistical base for social accounting matrices (SAM) in Russia is a “System of
Input-output tables of Russia” published by Goskomstat. The last publication con-tains
2003 data. In order to build a Russian SAM for 2004 we had to update data
using the available information from the National Accounts and Russian Statistical
Yearbook for 2004. Structure of this database is close to the data requirements for
the standard IFPRI model (Lofgren et al. 2002), with one exception: there is no
data on transport and trade mark-ups.
4.2. Experiment design
Designing experiments for this model we kept in mind two possible channels of
influence of an increase in migration on Russian economy: a direct one, account-

ing for a hypothetic increase in immigration; and an indirect one, steaming from
an increase in total factor productivity in selected industries which are usual em-ployers
of immigrants.
We conducted three series of simulations, exploring direct and indirect effects
of an increase in migration. In the first series consisting of ten experiments, we
look at direct consequences of a 10% increase in migration and remittances rela-tive
to a previous simulation. Thus in the last simulation a cumulative increase
relative to the benchmark is two-fold. The GAMS code and some additional in-formation
on the design of the first set of simulations are in the Table 4.7.
Table 4.7. Design of the first set of simulations, an increase in immigration and remit-tances
GAMS CODE: Experiment 1
Parameter SHRMIGR /0.14/;
trnsfrSIM(H,SIMNTBASE(SIM)) = (1 + (ord(SIM)-1)/10) * trnsfr(H,'ROW');
QFSSIM('LAB',SIMNTBASE(SIM)) = ((1 + (ord(SIM)-1)/10)*SHRMIGR + (1-
SHRMIGR)) * QFS0('LAB');
In each simulation households transfers to ROW (interpreted as remittances) and stock
of immigrants in labour force increase by 1% relative to previous simulation. A first
simulation is equivalent to benchmark. Migration and remittances increase twofold in last
simulation relative to benchmark.
Total labour force, thousand people
(Source: Таблица 1.11.; ТРУД И ЗАНЯТОСТЬ В РОССИИ 2007) 74 146 200.00
Stock of immigrants (2005)
(Source: WB Migration and Remittances Factbook ) 12 079 626.00
Share of migrants in the total labour force (SHRMIGR) 14%
We assume that the share of remittances in the labor income of immigrants is
constant. Thus the volume of remittances in local currency units (LCU) increases
by the same factor as immigration does.
The second set of simulations aims at assessing indirect effects of increasing
migration. According to Table 4.4 most officially registered immigrants are em-ployed
in construction (40% in 2006), trade (26%), transport (4.3%), agriculture
and hunting (4.4%). We enlarged this list to cover all service sectors of the model.
Indirect effect of an increase in migration is modeled as an increase in total factor
productivity in service sector. In the second set of experiments we conducted ten
runs. In each run total factor productivity in services increases by 5% relative to a
previous experiment.
The GAMS code and some additional information on the design of the second
set of simulations are in the Table 4.8.

Table 4.8. Design of the second set of simulations, an assessment of an indirect effects
of an increase in migration (an increase in TFP in services)
alphavaSIM(AASERV,SIMNTBASE(SIM)) = (1 + (ord(SIM)-1)/20) * al-phava0(
AASERV);
AASERV(A) services
/
A_sec14 Construction
A_sec16 Transport and communication
A_sec17 Trade
A_sec18 Other services
A_sec19 Housing
A_sec20 Health sports social security education culture and arts services
A_sec21 Science
A_sec22 Finance administration defense and civil organizations
/
In each simulation total factor productivity in services increases by 5%
Table 4.9. Parameter alphavaSIM – a shift parameter in CES activity production
function, by industry for each simulation of the second set
3304 PARAMETER al-phavaSIM
shift parameter
CES activity production
function
Bench
-mark
level
SIM
2
SIM
3
SIM
4
SIM
5
SIM
6
SIM
7
SIM
8
SIM
9
SIM
10
SIM
11
A_sec1 Electricity and
heat 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931
A_sec2
Products of Oil
extraction and
refinery
1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45 1.45
A_sec3 Coal 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931 1.931
A_sec4 Peat 1.998 1.998 1.998 1.998 1.998 1.998 1.998 1.998 1.998 1.998 1.998
A_sec5 Ferrous metals 1.741 1.741 1.741 1.741 1.741 1.741 1.741 1.741 1.741 1.741 1.741
A_sec6 Nonferrous metals 1.821 1.821 1.821 1.821 1.821 1.821 1.821 1.821 1.821 1.821 1.821
A_sec7
Products of
Chemical industry
and petrochemical
industry
1.989 1.989 1.989 1.989 1.989 1.989 1.989 1.989 1.989 1.989 1.989
A_sec8
Machinery and
equipment, metal
works
1.827 1.827 1.827 1.827 1.827 1.827 1.827 1.827 1.827 1.827 1.827
A_sec9
Products of For-estry,
wood-processing
and
paper-pulp indus-try
1.08 1.08 1.08 1.08 1.08 1.08 1.08 1.08 1.08 1.08 1.08
A_sec10
Construction
materials (includ-ing
glass, china
and delftware)
1.901 1.901 1.901 1.901 1.901 1.901 1.901 1.901 1.901 1.901 1.901
A_sec11 Products of Light
industry 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952

3304 PARAMETER al-phavaSIM
shift parameter
CES activity production
function
Bench
-mark
level
SIM
2
SIM
3
SIM
4
SIM
5
SIM
6
SIM
7
SIM
8
SIM
9
SIM
10
SIM
11
A_sec12
Products of Food-processing
Indus-try
1.867 1.867 1.867 1.867 1.867 1.867 1.867 1.867 1.867 1.867 1.867
A_sec13 Products of all
Other industries 1.871 1.871 1.871 1.871 1.871 1.871 1.871 1.871 1.871 1.871 1.871
goods 1.973 2.072 2.171 2.269 2.368 2.467 2.565 2.664 2.763 2.861 2.96
A_sec15
Agricultural goods
and services in
agriculture and
forestry
1.577 1.577 1.577 1.577 1.577 1.577 1.577 1.577 1.577 1.577 1.577
A_sec16
Transport cargo
and
communication
1.999 2.098 2.198 2.298 2.398 2.498 2.598 2.698 2.798 2.898 2.998
A_sec17 Trade (including
catering) 1.338 1.405 1.472 1.539 1.606 1.673 1.74 1.807 1.874 1.941 2.007
A_sec18 Other services 1.877 1.971 2.064 2.158 2.252 2.346 2.44 2.534 2.628 2.721 2.815
A_sec19 Housing 1.918 2.014 2.11 2.206 2.302 2.398 2.494 2.59 2.686 2.782 2.878
A_sec20
Health, sports,
social security,
education, culture
and arts services
1.206 1.266 1.327 1.387 1.447 1.508 1.568 1.628 1.689 1.749 1.809
A_sec21
Science and scien-tific
services,
including geology
and meteorology
services
1.017 1.068 1.118 1.169 1.22 1.271 1.322 1.373 1.423 1.474 1.525
A_sec22
Finance, banking
and insurance
services, govern-ment
and civil
organizations
1.768 1.856 1.944 2.033 2.121 2.21 2.298 2.386 2.475 2.563 2.651
The third set of experiments (Table 4.10) combines design of the first and the
second sets, i.e. in each experiment of this set there is an increase in labor immi-gration,
remittances, and TFP in services.
Table 4.10. Design of the third set of simulations
Parameter SHRMIGR /0.14/;
trnsfrSIM(H,SIMNTBASE(SIM)) = (1 + (ord(SIM)-1)/10) * trnsfr(H,'ROW');
QFSSIM('LAB',SIMNTBASE(SIM)) = ((1 + (ord(SIM)-1)/10)*SHRMIGR + (1-
SHRMIGR)) * QFS0('LAB');
* Increase in productivity in service sectors
alphavaSIM(AASERV,SIMNTBASE(SIM))=(1 + (ord(SIM)-1)/20) * alphava0(AASERV);

AASERV(A) services
/
A_sec16 Transport and communication
A_sec17 Trade
A_sec18 Other services
A_sec19 Housing
A_sec20 Health sports social security education culture and arts services
This is a combination of the set of simulations 1 and the set of simulations 2: In each
simulation households transfers to ROW (interpreted as remittances) and stock of immi-grants
in labour force increase by 10% relative to previous simulation. First simulation is
equivalent to benchmark. Migration and remittances increase twofold in last simulation
relative to benchmark. In each simulation total factor productivity in services (set
AASERV) increases by 0.05%
Simulation number (The first set of simulations)
Parameter 02 03 04 05 06 07 08 09 10 11
Capital rent (%
to benchmark) 0.31 0.62 0.92 1.21 1.49 1.77 2.04 2.31 2.57 2.82
Wage (% to
benchmark) -0.31 -0.62 -0.92 -1.21 -1.49 -1.77 -2.04 -2.31 -2.57 -2.82
4.3. Results
Comparing results of all simulations we can conclude that both, direct and indi-rect
effects are working in the same direction. As for numerical values, the indirect
effect of migration, as it is modeled in the second set, dominates the direct effect
of an increase in immigration and remittances.
Experiment 1: doubling the stock of immigrants and remittances
The consequences of an increase in the labor migration in the model could be
characterized in the following way: as inward labor migration increases, supply of
labor increases for all industries, since we do not have any labor market segmenta-tion
in the model. The present model set-up does not account for unemployment,
either. Thus increase in the labor force pushes wages down in the whole economy.
With the supply of capital being fixed, capital rent rises as wage goes down (see
Table 4.11 for details).
We used the standard macro closure of the IFPRI model for all three sets of
simulations. In this closure a country under investigation is assumed to be a small
open economy with flexible exchange rate, fixed foreign savings and fixed capital
formation.

Table 4.11. Results of the first set of simulations
Percentage increase of LF and remittances
Experiment 1 relative to benchmark. Description of a
parmeter
10 20 30 40 50 60 70 80 90 100
QABSTOT real absorption (LCU
at base prices) 0.54 1.06 1.57 2.07 2.56 3.04 3.51 3.97 4.41 4.85
QHTOT
real household con-sumption
(LCU at
base prices)
1.14 2.25 3.34 4.4 5.44 6.46 7.45 8.43 9.38 10.3
QETOT total real exports
(LCU at base prices) 0.54 1.07 1.6 2.12 2.64 3.15 3.66 4.17 4.67 5.17
QMTOT total real imports
(LCU at base prices) 0.52 1.03 1.53 2.02 2.51 2.98 3.45 3.9 4.35 4.8
REXR PPP real exchange
rate (LCUs per FCU) 0.25 0.49 0.72 0.94 1.16 1.37 1.57 1.77 1.96 2.15
NEXR nominal exchange rate
(LCUs per FCU) 0.04 0.07 0.1 0.13 0.16 0.18 0.2 0.22 0.24 0.26
PDIND
domestic (non-tradables)
price index
(100 for base)
-0.2 -0.4 -0.6 -0.8 -1 -1.2 -1.4 -1.5 -1.7 -1.9
INVGDP investment (% of
nominal GDP) -0.1 -0.2 -0.3 -0.4 -0.6 -0.7 -0.8 -0.9 -0.9 -1
PRVSAVGDP
private (household +
enterprise) savings (%
of nominal GDP)
-0.5 -0.9 -1.3 -1.7 -2.2 -2.5 -2.9 -3.3 -3.7 -4
FORSAVGDP foreign savings (% of
nominal GDP) 0.04 0.08 0.12 0.16 0.2 0.24 0.28 0.31 0.35 0.39
TRDDEFGDP trade deficit (% of
nominal GDP) -0.1 -0.1 -0.1 -0.2 -0.2 -0.3 -0.3 -0.4 -0.4 -0.5
GOVSAVGDP government savings
(% of nominal GDP) 0.3 0.59 0.87 1.14 1.4 1.66 1.9 2.14 2.38 2.61
IMPTAXGDP tariff revenue (% of
nominal GDP) 0.01 0.01 0.02 0.02 0.03 0.03 0.04 0.04 0.05 0.05
DIRTAXGDP direct tax revenue (%
of nominal GDP) 0.01 0.02 0.04 0.05 0.06 0.07 0.08 0.08 0.09 0.1
The wage decrease drives domestic prices down and stimulates exports. Price
ratio of tradables to nontradables goes up, as real and nominal exchange rate de-preciate.
Import is part of an aggregate commodity which is demanded by house-holds
and government for final consumption and by firms for intermediate use (see
Figure 4.2 below.)
Thus increase in volume of production gives rise to intermediate use and de-mand
for composite commodity. In a due course, this tendency gives rise to im-ports.

Figure 4.2. Commodity flows in the standard Ifpri model (Longfren et al, 2002)
Note. CES is constant elasticity of substitution, CET is constant elasticity of transforma-tion.
There is only one household in our version of the model. Thus we can not dif-ferentiate
between residents of the country and labor migrants. An increase in mi-gration
affects the income of the representative household in two ways. First, the
wage rate declines, but the labor force increases. Second, households are assumed
to be the owners of the capital, thus all capital rent goes to the household budget.
With an increasing labor force and declining wage rate, plus an increase in capital
rent, household income increases as a result of an increase in immigration. An
increase in household income and a decrease in domestic prices give rise in private
consumption in all three sets of simulations.
Experiment 2: increasing total factor productivity in services
The indirect effects of immigration through higher TFP in services, as they are
modeled in the second experiment, are much larger than the effects of immigra-tion-
induced increase in the labor force (cf. Experiment 1). For example, with a
hypothetical 10 percent increase in TFP in services, total absorption increases by 8
percent and household consumption by 16 percent (vs. 1 percent and 2 percent,
respectively, in Experiment 1). While it is difficult to establish a tight quantitative
link between immigration and higher TFP in services, it is clear that the impact of
any significant productivity gains would be large compared to the more conven-tional
effects assessed in Experiment 1.
Experiment 3: doubling the stock of immigrants and increasing total factor
productivity in services
Experiment 3 combines the immigration-induced increase of the labor force
and outward remittances (as in Experiment 1) with the TFP increase in the services

sector (as in Experiment 2). The results are unsurprising, with labor force and TFP
effects generally working in the same direction and adding to each other.
Table 4.12. Results of the second set of simulations
Experi
Description of a
SIM
SIM
SIM
SIM
ment 2
parmeter
2
3
4
5
SIM
6
SIM
7
SIM
8
SIM
9
SIM1
0
SIM
11
QABSTOT
real absorption
(LCU at base
prices)
3.92 7.74 11.46 15.09 18.64 22.12 25.53 28.88 32.17 35.4
QHTOT
real household
consumption
(LCU at base
prices)
8.34 16.45 24.35 32.06 39.61 47 54.25 61.36 68.34 75.21
QETOT
total real exports
(LCU at base
prices)
2.17 4.29 6.38 8.44 10.48 12.5 14.51 16.5 18.47 20.44
QMTOT
total real imports
(LCU at base
prices)
4.13 8.17 12.14 16.06 19.93 23.76 27.56 31.33 35.07 38.79
REXR
PPP real exchange
rate (LCUs per
FCU)
0.93 1.63 2.11 2.4 2.52 2.48 2.32 2.06 1.69 1.26
NEXR
nominal exchange
rate (LCUs per
FCU)
0.96 1.74 2.35 2.82 3.14 3.35 3.46 3.46 3.39 3.25
PDIND
price
index (100 for
base)
0.02 0.1 0.24 0.41 0.61 0.85 1.1 1.38 1.67 1.97
nominal GDP) -0.61 -1.17 -1.68 -2.15 -2.59 -2.99 -3.36 -3.71 -4.03 -4.34
PRVSAVG
DP
private (household
+ enterprise)
savings (% of
nominal GDP)
-2.25 -4.3 -6.19 -7.93 -9.54 -11.1 -12.5 -13.8 -14.99 -16.15
FORSAVG
DP
foreign savings
(% of nominal
GDP)
0.32 0.63 0.94 1.25 1.55 1.84 2.12 2.4 2.67 2.93
TRDDEFG
DP
trade deficit (% of
nominal GDP) 0.39 0.77 1.14 1.5 1.85 2.19 2.52 2.84 3.15 3.45
GOVSAVG
DP
government sav-ings
(% of nomi-nal
GDP)
1.31 2.49 3.56 4.53 5.41 6.22 6.97 7.65 8.29 8.88
IMPTAXG
DP
tariff revenue (%
of nominal GDP) 0.05 0.1 0.14 0.18 0.21 0.24 0.27 0.3 0.32 0.34
DIRTAXG
DP
direct tax revenue
(% of nominal
GDP)
0.14 0.27 0.38 0.49 0.59 0.68 0.77 0.85 0.93 1

Table 4.13. Results of the third set of simulations
Experiment
Description of a
SIM
SIM
SIM
3
parmeter
2
3
4
SIM
5
SIM
6
SIM
7
SIM
8
SIM
9
SIM1
0
SIM
11
QABSTOT
real absorption
(LCU at base
prices)
4.47 8.83 13.1 17.29 21.4 25.45 29.44 33.36 37.24 41.06
QHTOT
real household
consumption
(LCU at base
prices)
9.49 18.76 27.83 36.73 45.47 54.07 62.53 70.88 79.11 87.24
QETOT
total real exports
(LCU at base
prices)
2.71 5.38 8.02 10.63 13.22 15.8 18.37 20.92 23.46 25.99
QMTOT
total real imports
(LCU at base
prices)
4.65 9.23 13.74 18.2 22.62 27.01 31.37 35.7 40.02 44.31
REXR
PPP real exchange
rate (LCUs per
FCU)
1.17 2.06 2.72 3.17 3.44 3.54 3.52 3.38 3.15 2.84
NEXR
nominal exchange
rate (LCUs per
FCU)
0.99 1.78 2.41 2.87 3.21 3.42 3.54 3.56 3.51 3.4
PDIND
price
index (100 for
base)
-0.18 -0.28 -0.31 -0.29 -0.22 -0.12 0.02 0.17 0.35 0.54
nominal GDP) -0.73 -1.39 -1.99 -2.55 -3.07 -3.55 -4 -4.41 -4.8 -5.17
PRVSAVG
DP
private (household
+ enterprise)
savings (% of
nominal GDP)
-2.68 -5.12 -7.36 -9.41 -11.3 -13.1 -14.7 -16.2 -17.6 -19
FOR-SAVGDP
foreign savings (%
of nominal GDP) 0.36 0.72 1.07 1.42 1.76 2.09 2.41 2.72 3.02 3.31
TRDDEFG
DP
trade deficit (% of
nominal GDP) 0.35 0.69 1.02 1.35 1.66 1.97 2.27 2.56 2.84 3.11
GOVSAVG
DP
government sav-ings
(% of nomi-nal
GDP)
1.59 3.01 4.29 5.44 6.48 7.43 8.29 9.08 9.81 10.49
IMP-TAXGDP
tariff revenue (%
of nominal GDP) 0.06 0.11 0.15 0.2 0.23 0.27 0.3 0.32 0.35 0.37
DIR-TAXGDP
direct tax revenue
(% of nominal
GDP)
0.15 0.29 0.41 0.53 0.64 0.74 0.83 0.92 1 1.07
4.4. Conclusions
In this country study we analyzed direct and indirect consequences of increase
in migration to Russia using standard IFPRI computable general equilibrium

framework. The benchmark CGE dataset – the social accounting matrix for Russia
is unique and was created for this project. We presented a report on data availabil-ity
on migration and remittances for the Russian economy. We presented three sets
of CGE model runs simulating direct and indirect effects of immigration.
The most important driving force behind all results in the presented sets of
simulations is a significant economy-wide wage rate decrease as a direct conse-quence
of an increase in immigration. This effect is justified by the model settings
but is not supported by the mainstream empirical literature on immigration.
From this literature we know the importance of personal characteristics, such as
skill level, experience, major occupation, etc. General equilibrium analysis would
be much more precise and rich with all these details incorporated in the model.
This exercise could be done if there were empirical estimates of the relevant pa-rameters
for Russia. But empirical literature on the consequences of labor migra-tion
for Russia is constrained by the availability of data, which is very scarce to
say the least.

5. Conclusions
Our case studies of the impact of labor migration and remittances in several net
emigrant countries and in Russia aim to provide a systematic comparative analy-sis,
notwithstanding inevitable concerns about data quality and restrictive model-ing
assumptions. With this qualification, our simulation results demonstrate that in
all net immigrant countries (Moldova, Ukraine, Georgia, Kyrgyzstan), private
consumption is much higher than it would be in the absence of migration and re-mittances.
The simulated decline is between 18 percent for Ukraine and 32 percent
for Moldova (where the share of migrants in the labor force and the ratio of remit-tances
to GDP is especially large). All income groups would be strongly affected,
although the impact on the poorest groups varies across countries. All net emigra-tion
countries would also experience a modest real devaluation of below 3 percent
and associated changes in sectoral output patterns.
In Russia, which hosts most migrants from the net emigration CIS countries,
immigration mainly leads to an economy-wide wage decline. Although this result
is conditioned in part by our assumptions, particularly the comparative-static na-ture
of the CGE model with no additional capital investment in response to immi-gration,
it is plausible that a large increase in the number of immigrants as in Rus-sia
will limit real wage growth. At the same time, Russia has actively encouraged
some immigrants, especially of Russian ethnic background, to settle in Russia
permanently, with a view to compensating for some of the population decline that
Russia is experiencing. It remains to be seen whether these policies will be modi-fied
in response to the end-2008 financial crisis and the consequent slow-down in
GDP growth and construction activity in particular.
These simulation results demonstrate the large benefits from labor migration
and remittances for migrants themselves as well their home countries. Indirectly,
at least, host countries such as Russia and the European Union will also benefit
from higher incomes and greater economic security in their neighboring net emi-gration
countries. Whether these potential benefits are realized depends impor-tantly
on government policies related to migration in both, home and host coun-tries.
Against the backdrop of the migration-related policies currently pursued in
the CIS region and in the EU, four major policy implications emerge.
First, some CIS country governments faced with large migrant outflows have
been reluctant, for political reasons, to even acknowledge that emigration is taking

place on a large scale. Consequently, they have failed to provide support services
to migrants where such services would enhance the benefits from migration, limit
the risks, and strengthen migrants’ attachment to their home country. Such ser-vices
include job placement into legal work abroad through official employment
agencies, high-quality consular services for migrants abroad, advocacy with part-ner
governments for limited-term work opportunities for their residents, etc. The
absence of such support has made migration more costly to households, without
offering attractive alternatives, and alienated migrants from their home country.
By contrast, a forward-looking policy strategy for home countries would be to
support migrants where they are most at risk, such as when seeking employment
and dealing with host country authorities. This would render it more likely that
migrants would favorably consider employment or investment opportunities at
home in the future.
Second, for economic recovery to take hold in the smaller, natural-resource-poor
CIS countries, fixed investments need to be sustained and increased further.
Remittances could help to pay for such investment. However, the business and
investment climate in many of these countries is so poor that, currently, remit-tances
are only rarely used for productive investment. Government efforts to
channel remittances into investment, which are debated in many CIS countries,
will succeed only when all investors – migrants and non-migrants, politically well-connected
or not – can expect to receive an adequate return on productive invest-ments
that is not diminished by parasitic public institutions.
Third, to promote social coherence in emigration countries, prudent govern-ment
policies are called for to ensure that the income gains due to migration are
shared, to some degree at least, by all households. Taxes on remittances are usu-ally
considered counterproductive as income from legal employment is already
taxed in the host country and, in any case, remittances might simply be driven
underground. However, since many CIS country governments rely on taxes on
imports (especially VAT, but also import duties) for much of their revenue, gov-ernment
revenue typically increases along with remittance-driven imports (which
are bought over-proportionately by migrant households). The extra government
revenue can be used to maintain public infrastructure, provide social services and
education (including to the children of migrants left at home or with relatives), and
provide targeted income support.
Fourth, destination countries will increasingly find themselves competing not
only for high-skilled migrants, but also for those willing to perform jobs that are
otherwise difficult to fill (such as seasonal work in agriculture, construction, and
social services). Russia, the most important host country for migrants from the CIS
region, is currently offering legal employment on a fairly broad basis, but mi-grants’
living conditions are frequently poor and harassment by authorities is en-

demic. Extending legal residence and employment to a larger share of migrants
already in Russia, and strengthening the rule of law and ensuring fair treatment for
migrants by authorities, would help to attract the growing numbers of immigrants
that Russia will want to rely on as its economic growth continues.
In EU countries, legal employment opportunities for CIS country migrants are
still severely limited but growing. Legalization programs in countries such as Italy
and Portugal also create pockets of legal migrants that will probably become the
hubs of migrant networks that will attract more family-based and other immigra-tion
from CIS countries in the future. It would be in the interest of both migrants
and EU host countries to replace these haphazard legalizations with a forward-looking
strategy for admitting migrants with good job prospects in the EU. Since
the EU functions as a single labor market, such programs should be coordinated at
the EU rather than the national level. For the benefit of both, CIS countries and the
EU, the deepening of bilateral relations under European Neighborhood Policy
should include enhanced opportunities for legal labor migration.

References
Asian Development Bank (2008a). Country Report on Remittances of Interna-tional
Migrants and the Financial Sector in the Kyrgyz Republic. Remittances
and poverty in Central Asia and South Caucasus. Project number 40038.
Asian Development Bank (2008b). Country Report on Remittances of Interna-tional
Migrants and Poverty in the Kyrgyz Republic. Remittances and poverty
in Central Asia and South Caucasus. Project number 40038.
Fagernäs, Sonja (2004). Analysing the Distributional Impacts of Stablisation Pol-icy
with a CGE Model: Illustrations and Critique for Zimbabwe. ESAU Work-ing
Paper 4, Overseas Development Institute. London.
Georgian-European Policy and Legal Advice Centre (2007): Georgian Economic
Trends, Quarterly Economic Trends, October
Georgian-European Policy and Legal Advice Centre (2008): Georgian Economic
Trends, Quarterly Economic Trends, February
IFAD (International Fund for Agricultural Development) 2006. Sending Money
Home. Worldwide Remittance Flows to Developing and Transition Countries.
IOM (International Organization for Migration in Ukraine) 2006. “Human traf-ficking
survey: Belarus, Bulgaria, Moldova, Romania and Ukraine”. Prepered
by GFK Ukraine. Project manager: Tetiana Sytnyk.
Lofgren, Hans, Rebecca Lee Harris, and Sherman Robinson (2002). A Standard
Computable General Equilibrium (CGE) Model in GAMS. Microcomputers in
Policy Research 5, International Food Policy Research Institute. Washington
D.C.
Mansoor A. and Quillin B. (2006). Migration and Remittances. Eastern Europe
and the Former Soviet Union, The World Bank.
OECD. 2004. Trends in International Migration, Continuous Reporting System on
Migration //Annual report.
Poznyak O. (2002). “Mashtaby Zovnishnih Trudovyh Migratsii Naselennia
Ukrainy” (Scale of External Labor Migration of Ukrainian Population ). Jour-nal:
Ukraina: Aspekty Pratsi (Ukraine: Labor Aspects) #6, pp 37-40.
Quillin, B., Segni, C., Sirtaine, S., and Skamnelos (2007). Remittances in CIS
countries: A Study of Selected Corridors. Europe and Central Asia Chief Eco-nomist’s
Regional Working Paper Series, Finance and Private Sector Devel-

opment Department (ECSPF), Vol.2, No.2. The World Bank, Washington D.C.
http://siteresources.worldbank.org/ECAEXT/Resources/publications/454763-
1190042265575/FINAL_RemittancesStudy_July23_1.pdf.
Robinson, Sherman (2003). Macro Models and Multipliers: Leontief, Stone,
Keynes, and CGE Models. International Food Policy Research Institute (IF-PRI),
mimeo, September.
www.ifpri.org/events/seminars/2003/20031014/robinson_thorbecke_EPIAM.pdf.
Simai, M. (2006). Poverty in Inequality in Eastern Europe and the CIS Transition
Economies. DESA Working Paper, No. 17.
Stiglitz, J. (1994). Whither Socialism? Cambridge, Ma.: The MIT Press.
Starodub, A., N. Parkhomenko (2005). Ukrainska Trudova Migratsia Do Krain ES
v Dzerkali Sotsiologii (Ukrainian Labor Migration to the EU Countries in the
Mirror of Sociology). Center for Peace, Conversion and Foreign Policy of
Ukraine, Kyiv, Ukraine.
Statistical Yearbook of Georgia (2007). Web site: www.statistics.ge.

CASE Network Report 86 - Income and Distribution Effects of Migration and Remittances: an Analysis Based on CGE Models for Selected CIS Countries

More Related Content

What's hot (19)

Viewers also liked (9)

Similar to CASE Network Report 86 - Income and Distribution Effects of Migration and Remittances: an Analysis Based on CGE Models for Selected CIS Countries (20)

More from CASE Center for Social and Economic Research (20)

Recently uploaded (20)

CASE Network Report 86 - Income and Distribution Effects of Migration and Remittances: an Analysis Based on CGE Models for Selected CIS Countries