Migration, Employment and Poverty Reduction

Manufactured in Moldova

Printed by Art Grup Brivet Cover and design by Alexandru Burdila, Nicolai Verbiţki Editing by Mr. John Weeks (English), Mr. Igor Guzun (Romanian), Mr. Anatol Gudym (Russian)

Disclaimer: The analysis and policy recommendations of this Report, fi ndings, data collections, interpretations and conclusions expressed here are those of the author(s) and do not necessarily refl ect the views of the United Nations Development Programme, its Executive Board or its Member States, including the Government of the Republic of Moldova, nor those of the co-sponsors of the Report, the Swedish International Development Agency (Sida), the Department for International Development of the United Kingdom (DFID) and the International Labour Organization (ILO).

This report is an independent report produced by a team of international and national experts organized by the Bureau for Development Policy of UNDP, New York. The main objectives of the report are: 1) to broaden the dialogue in Moldova on economic policies and their impact on growth, employment and poverty reduction; and 2) to off er to the Government of Moldova some practical recommendations on how to use economic policies to promote such objectives.

The Bureau for Development Policy of UNDP has initiated a broad range of global, regional and country reports designed to help fashion more pro-poor economic policies. This set of activities has broadened the dialogue on pro-poor policies at the national, regional and international levels. While each report stands on its own, addressing the specifi c circumstances of each country, taken together they provide a consistent analysis and vision of the poverty reduction process.

This Report constitutes the fi nal product of an 18-month long project intended to foster a national dialogue on growth and poverty reduction policy options. A fi rst version of the Report was presented to the National Review Committee in the fall of 2004 and a revised draft of March 2005 was discussed at a set of roundtables before a fi nal, extended National conference in June 2005. The study was developed by a team of international and national experts:

Mr. John Weeks (team leader) Mr. Giovanni Andrea Cornia, Mr. Selim Jahan Mr. Terry McKinley Mr. Siddiq Osmani Mr. Michael Reynolds Mr. Max Spoor Ms. Felicia Izman

Other contributors: Mr. Roman Ladus, Ms. Ala Mandicanu, Mr. Valeriu Prohnitchi, Ms. Galina Selari, Mr. Philip Martin

Steering Committee Government of the Republic of Moldova Mr. Sergiu Sainciuc, Prime Deputy Minister of Labor and Social Protection Ms. Larisa Stucalov, Head of Social Development Department Ms Marina Semeniuc, Deputy Head of Department Finances in Health and Social Protection, Ministry of Finance Mr. Andrei Rotaru, Head of Monetary Policies and Research Department, National Bank of Moldova Non-Governmental Organizations Ms. Aliona Niculita, Executive Director, CONTACT Mr. Victor Moroz, National Association of Farmers Ms. Inna Gutium, Deputy Director, ADEPT Mr. Anatol Gudym, Executive Director, CSSR International Organizations Ms. Nina Orlova, National Programme Coordinator, Sida Ms. Alla Skvorţova, Head of DFID Section, British Embassy in Moldova Ms. Viorica Ghimpu, National Programme Coordinator, ILO IPEC Ms. Jana Costachi, National Programme Coordinator, ILO MIGRANT Ms. Liudmila Barcari, Programme Associate, UNDP Moldova Mr. Mihail Peleah, Programme Associate, UNDP Moldova

National Review Committee Mr. Anatol Gudym, Executive Director, Centre of Strategic Studies and Reforms (CSSR), Moderator Ms. Raisa Dogaru, Ministry of Labor and Social Protection Mr. Ion Holban, Ministry of Labor and Social Protection Mr. Andrei Rotaru, National Bank of Moldova Ms.Emila Prujanschi, Ministry of Finance Mr.Iurie Torcunov, Ministry of Economy Ms.Galina Savelieva, Ministry of Economy Ms. Elizaveta Ivanova, Ministry of Labor and Social Protection Ms. Liliana Agarcov, NGO “IDEA” Mr. Vladislav Kutirchin, Social Bank

Authors’ acknowledgements: The help of any people in Moldova made this report possible. Special thanks go to the UNDP country offi ce and, in particular, to Ms. Liudmila Barcari for her invaluable support. Thanks also go to Sida, DFID and the ILO for their aid and technical support. Table of Contents

Prologue 6 Executive Summary 7

Chapter 1: INTRODUCTION 29 1.1 Purpose of this Study 29 1.2 Gender Analysis of the Transition 29 1.3 Four Fundamental Transitions 30 1.4 Moldova’s Poverty Reduction Strategy 34 Annex: Common Analytical Mistakes about Transition Societies 39

Chapter 2: ECONOMICS OF THE TRANSITION 44 2.1 Moldova in Comparative Context 44 2.2 Per Capita Income and Development Assistance 46 2.3 Macroeconomic Performance and Policy, 1990-2004 48 Annex 1. A Chronology of Moldova’s External Debt 76

Chapter 3: PUBLIC FINANCE FOR GROWTH AND POVERTY REDUCTION 80 3.1 Introduction 80 3.2 The Mobilization of Domestic Revenue 80 3.3 External and Internal Debt Burden 87 3.4 The Composition of Public Expenditures 90 3.5 Poverty, Inequality and the Budget 93

Chapter 4: POVERTY, INEQUALITY AND POLICY IN MOLDOVA 94 4.1 Introduction 94 4.2 Trends in Poverty 95 4.3 Characteristics of Poverty 98 4.4 Changes in Inequality 103 4.5 Inequality and poverty alleviation 105 Annex 1. Statistical Sources 106 Annex 2. Analytical Framework for Calculating Poverty Reduction 114

Chapter 5: MIGRATION, EMPLOYMENT AND POVERTY REDUCTION 115 5.1 Migration from Moldova 115 5.2 Migration, Growth, Employment and Poverty Reduction 120 5.3 Possible Future Patterns of Migration 124 5.4 Policy and Possible Future Directions 126 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

4 5 Chapter 6: EMPLOYMENT, PRODUCTIVITY AND PROVERTY 129 6.1 Employment and Poverty 129 6.2 Employment, Wages and Productivity 131 6.3 Generating Wage Employment 138 6.4 Labour Market Policies 142 6.5 Conclusions 144

Chapter 7: AGROINDUSTRIAL GROWTH AND SUSTAINAIBLE PROVERTY REDUCTION 147 7.1 Agriculture’s Role in Moldova’s Growth Strategy 147 7.2 Agriculture on the Eve of Independence 147 7.3 Post-Independence Agrarian Reform and Economic Performance 149 7.4 The Underdevelopment of Agricultural Markets 153 7.5 Rural Labour Markets 154 7.6 Agro-Industrial Growth 155 7.7 Policy Recommendations 156

Chapter 8: SOCIAL POLICIES FOR HUMAN DEVELOPMET 160 8.1 Introduction 160 8.2 The Human Development Collapse of the 1990s 163 8.3 Basic Needs and Social Protection 171 8.4 Social Policy for Poverty Reduction 174 References 175 Short biographies of the main contributors 181 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

4 5 Prologue

Almost ten years down the road [from independence] it is clear that the road map [for Moldova’s transition] was at best inadequate and that the so-called transition in most respects has been a failure. (Ronnas & Orlova 2000, 15)

In 1989, Moldova was a small province of a large authoritarian state; fi f- teen years later it was an independent country that had passed through several legitimately representative elections, with prospect to do so in the future. While like many former Soviet republics Moldova suff ered from separatism (the Transdniestre region), no Moldovan soldier had died in confl ict for a decade.

The period of notable political achievements had its profoundly dark aspect. In March 1995, The Economist magazinemagazine describeddescribed MoldovaMoldova asas a ‘model of sound reforming…a perfect laboratory of reform’. If this judgement were correct, it would be a terrible indictment of the transition process. In 1989 the head count poverty rate in Moldova was less than fi ve percent, and in 1993 had risen to two-thirds of the population by one measure, close to half by another, as per capita income fell by half. In 1989 Moldovans had access to medical care comparable to that in Western Europe, and public education of high quality and open to all. A few years later, much of the population found social services beyond their means, accessible only with long delays, or of low quality.

A major tragedy of the collapse in Moldova was its diff erential impact on men and women. The both suff ered an actual fall in life expectancy, though more for men than women (UNDP 1998b, p. 214). In the labour market the diff erential impact was the reverse, more women lost jobs than men, reversing their employment gains during the Soviet period. In the words of a UNDP report,

…[W]omen face[d] increased hardship in many countries. Confronted with decreasing numbers of child care facilities, women have often been forced to abandon work. (UNDP 1998b, p. 7).

In the words of a UN offi cial in Moldova at the time, representatives of external agencies, insulated from the eff ects of the collapse, ‘like the fi rst class passengers of the Titanic in their lifeboats could only look on as their fellow voyagers sank into the abyss’. No transition country in Europe or Central Asia suff ered a weaker recovery from the depths of collapse than Moldova. This report is not merely about another country that suff ers

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction from severe poverty, but an investigation of a society taken to the brink of disintegration.

6 7 Executive Summary

1. INTRODUCTION

1.1. Background to the Report 1.2. The Framework for the Report This report addresses macroeconomic aspects of employment and poverty This report supports and complements reduction. Sectoral and micro issues are the long term strategy of the Moldovan treated in as far as they directly relate government as presented in the Economic to the macroeconomic framework for Growth and Poverty Reduction Strategy growth and poverty reduction. Even document (EGPRS), completed towards the in these cases, specifi c policies are not end of 2004. The objectives of the EGPRSP treated, since they typically involve and the policies to achieve them have been quite technical expertise. Thus, the vast further developed and clarifi ed in several majority of the suggestions refer to subsequent documents: the European macro policy, and are found in chapters 2 Union Action Plan, the Medium Term and 3. The other chapters are to provide Expenditure Framework, and the Action the analytical and empirical basis for Plan of April 2005 presented to parliament these pro-poor macro policy proposals. after the election of March 2005. Successful It is the raison d’etre of this report that achievement of the goals of the EGPRS without a macroeconomic framework requires three necessary developments that accommodates robust and equitable in the economy: narrowing the trade growth, micro policies, no matter how gap, raising aggregate investment, and appropriate, will have little impact. reducing inequality. The framework in which those three developments can be The policy recommendations seek to fos- realised can be summarised as follows: ter pro-poor growth, by which is meant a growth path in which the income of poor, a. At the beginning of the twenty-fi rst however the poor are measured, increas- century, Moldova was a predominantly es more than the income of the non-poor. agricultural economy, but an agriculture- That is, the annual increment in growth led growth strategy is not the route to would be more equally distributed than sustainable poverty reduction. the initial income. Eff ective and feasible b. A viable and sustainable long term policies must be based on rigorous theo- growth strategy must be based on the retical analysis and an understanding of emergence of an internationally com- the characteristics of a country. There petitive manufacturing sector. are no correct policies in the abstract. Therefore, this summary includes a brief c. Rising productivity in agriculture and explanation of the analytical and empiri- manufacturing will require a purpose- cal basis of the policy suggestions, which ful public investment programme that ‘crowds in’ private investment, both do-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction is presented in greater detail in the full report. mestic and foreign.

6 7 d. Expansion, and in some cases, reorganisa- The statistics necessary to incorporate tion of existing poverty reduction pro- adequately gender into the analysis grammes, including the social protection of the transition and pro-poor growth system, is essential, based on the principle do not for the most part exist. This of targeted universal provision supported is a serious oversight that should be by specifi c programmes that are income- addressed immediately by the Moldovan tested; that is, universal programmes for government; all the more immediately targeted groups such as pensions for because there would certainly be funds the elderly, with programmes based on from donors to cover the cost of collecting falling below a designated poverty line and analysing the information. Given the playing a secondary role. lack of statistics necessary for a rigorous gender analysis, this report is reduced If in the medium term, the agricultural sector to the inadequate approach of using the can help narrow the trade gap, a major step available information and attempting to towards sustainable growth would be taken. draw reasonable inferences. Simultaneously, the rate of investment should rise. Trusting to a ‘favourable business climate’ is unlikely to stimulate investment 2.2 Four Fundamental to the level necessary for sustained poverty Transitions reducing growth. This problem is feasibly solved by using public investment to ‘crowd- For all of Central and Eastern Europe the in’ private investment, as well as creating twentieth century brought war and dicta- projects that directly reduce poverty. Along torship. The experience of Moldova was with narrowing the trade gap and increasing unimaginably grim. It is more than remark- investment, poverty reduction requires that, able, that after independence the country at a minimum, income and wealth inequality was relatively free of civil strife, even when not increase. the economy collapsed catastrophically. The absence of major civil strife might be explained in part by the overwhelming support of the vast majority of Moldovans 2. MOLDOVA IN for independence. At independence, Moldovans faced four fundamental social TRANSITION and economic transitions, each of which would have been a formidable challenge 2.1 Gender Analysis on its own: and the Transition 1. From nation to country, which required the construction of the state institutions At the beginning of the 1990s, Moldova appropriate to manage the aff airs of a entered a process of transition from central country, the countryhood transition; planning to market regulation. To under- 2. From an authoritarian to a democratic stand this transition, gender cannot be ig- regime, which necessitated the crea- nored. Gender inequality plays an important tion of the space in which open debate role in growth, employment and poverty and contestation of power could occur reduction. The fundamental policy changes peacefully, the political transition; associated with the transition process are likely to aff ect men and women diff erently 3. From central planning to the regulation because of their diff erent positions, roles, of markets, which rendered most of the and responsibilities in society. The develop- institutions of economic management ment of eff ective economic policies requires anachronistic, the regime transition; and mainstreaming of gender into the design of Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 4. From social delivery of basic needs policies, including the budgeting process. to market provision, which implied a

Executive summary

8 9 revolution in the manner by which the 2.3 External Support population was employed, provisioned for Pro-poor Growth with basic necessities, educated, and provided with health care, the socio-eco- Moldova is a low income country with lim- nomic transition. ited fi scal resources. Success in achieving pro-poor growth requires a disciplined The countryhood transition in Moldova fi scal eff ort by the government and from involved institution building, beginning donors and lenders budget support and with the creation of basic state institutions debt reduction. The government’s record that are the mainstays of public adminis- of economic management qualifi es it for tration in countries: ministries of foreign both, especially when Moldova is com- aff airs, defence, fi nance, and international pared to other low-income countries that trade. Even existing institutions whose currently receive budget support and basic function did not change, such as de- have qualifi ed for debt relief, through the partments of health, education and social HIPC initiative or other debt reduction welfare, required reorganisation because mechanisms. of the basic change in their status from conduits and implementation agencies The case for debt reduction is strong. In for decisions made in Russia, to policy and the early and mid 1990s, Moldova was decision making ministries. classifi ed by the World Bank as a middle income country, not eligible for conces- The problem of national policy manage- sionary loans on terms that characterise ment after 1990 was not due to lack of credits from the International Develop- capable administrators and professions. ment Association (IDA). Toward the end The management diffi culty arose because of the decade on the basis of a corrected administrators and professions found estimation of GDP per capita by the World themselves in institutions that were not Bank, this classifi cation was changed, designed to carry out the tasks of a country. to low income status, and further loans The government’s proposals in 2005 for a provided on IDA terms. Restructuring the comprehensive reorganisation of ministries multilateral debt would have a substantial and their functions represent a major step impact on fi scal resources; reimbursing towards correcting this problem. the excess interest would have a dramatic eff ect. There would be no basis for mak- Across the countries in transition, one of ing this restructuring conditional upon the most intractable aspects of institution policy, since the restructuring would building has been creating the mecha- represent correcting a mistake made by nisms to regulate markets out of the agen- lenders, rather than motivated by the cies and personnel of central planning. The insolvency of the borrower. The funds problem goes far beyond establishing the resulting from the restructuring could be ‘rule of law’ and ‘fostering private sector used to expand the pro-poor economic development’. The period of transition is and social programmes that are discussed by defi nition one in which laws and regula- in this report. The case for action on the tions change. Uncertainty about the legal lending mistake is strengthened because framework is inherent in the transition, during 1992-1998 Moldova consistently since laws are implemented by institu- received less development assistance as tions that are being created. In established a portion of GDP than the average for market economies, implementation of the transition countries, and did not reach legal framework involves interpretation of that average until 1999. Had it received laws, drawing heavily on legal precedent the average, in some years it would have which, by defi nition, does not exist in tran- been able to cover its external current ac-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction sitional economies. Corruption, public and count defi cit, and substitute for loans on private, should be analyzed in this context. commercial terms.

Executive summary

8 9 We recommend: ble result of rendering domestic production non-competitive with imports. The 1. It is appropriate that World Bank and IMF shock of rapid trade liberalisation was loans on non-IDA terms be immediately magnifi ed by exchange rate policy in and unconditionally changed to those the early years of the transition. The con- terms, and the country reimbursed for tinuous appreciation of the Leu from early the excess interest paid; 1994 through the fi rst quarter of 1998 re- 2. Donors and lenders accept the EGPRS as sulted in a decline in exports. The Russian the basis for providing budget support fi nancial crisis provided the shock that led to the government; and to the collapse of the Leu-dollar rate. 3. Greater eff orts should be made by To make matters worse, loans contracted donors and lenders to base their part- from multilateral and private lenders at nerships with the government on the commercial or near-commercial rates principle of national ownership of deve- were used to cover recurrent public ex- lopment policy. penditure. Economic analysis tells one that loans should be contracted only if 2.4 Economic Performance they can be serviced. Servicing debts requires either that 1) the loan creates and Policy, 1990-2004 assets to generate foreign exchange to cover debt service, or 2) the borrower When the Russian economy entered into anticipates a growth of exports that will precipitous decline in the early 1990s, service the debt. Borrowing to fund cur- Moldova followed it down. While the col- rent expenditure violates the fi rst condi- lapse of the Russian economy dictated tion, and borrowing when exports were a collapse in Moldova, it did not dictate falling violates the second. Thus, faced that Moldova’s decline would be so much with an unsustainable balance of pay- greater than the average for all Soviet ments and an enormous budget defi cit, republics. Moldova’s extreme collapse in as a result of the designation as a middle part followed from a misdiagnosis of the income country, the governments of the country’s characteristics at the outset of time had no alternative but to pursue an the transition, which resulted in a series unsound debt accumulation policy. of inappropriate policies. The mistake in calculating per capita income was part In summary, the late and slow recovery of of a larger mistake of presuming the the Moldovan economy was substantially country to be more developed than it magnifi ed by the pursuit of unsustain- was, and more capable of adjusting to able policies: premature trade liberalisa- the demands of a market economy than tion and an unsound fi scal strategy that was the case. The fundamental problem resulted in a contractionary dead-weight was not to open a closed economy, but of debt servicing that continues to the to re-orient the openness from the Soviet present. administrative trading system to market- based international commerce. The recovery and growth of the economy after 1999 was driven primarily by the This reorientation of trade required es- infl ow of remittances. Inherent in remit- tablishing the state institutions of inter- tance-driven growth are several possible national commerce, as well as facilitating problems: 1) the level of remittances investments in modernisation of produc- may not be sustainable; 2) the tempo of tion. Before it established these condi- growth beyond the management of the tions, the government of Moldova was government; and 3) large scale emigra- encouraged to pursue a policy of general

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction tion may negatively eff ect the quality trade liberalisation, which had the inevita- of the domestic labour force. Employing

Executive summary

10 11 fi scal, monetary and exchange rate policy European Union. Was the government to to foster faster growth, with an emphasis maintain the current level of revenue, al- on public investment, would complement lowing expenditure to rise toward a three the expansionary eff ect of remittances, percent defi cit would provide ‘fi scal space’ and create a strategy for a transition to a for pro-poor public investment. Such a development framework more securely modest defi cit would not have infl ation- based on domestic economic potential. ary eff ects in a small, open economy.

While offi cials at the National Bank of Fiscal Policy Moldova recognise this, their main defi cit concern, which they share with colleagues Centrally planned economic systems de- in the Ministry of Finance, is possible rived their government revenue by taking ‘crowding out’ eff ects on private invest- a share of the operating surplus of state ment. Fiscal defi cits result in crowding out enterprises, which in most of the coun- if they put upward pressure on interest tries accounted for virtually all of produc- rates, and private investment is interest tion and distribution. When transition rate sensitive. Even if upward interest rate policies wiped out the surpluses of most pressure resulted from fi scal defi cits, actu- enterprises, massive government defi cits al interest rates are so high that the eff ect resulted. Thus, the loss of fi scal control of the fi scal defi cit would be insignifi cant. was structural, not the result of mistakes Further, the crowding out eff ect is partly in policy or lack of competence of civil dependent upon how the expenditure servants. that generates the defi cit is spent. If the defi cit results from public investment that While narrowing the fi scal gap repre- complements private investment, this sented a necessary step towards macro- could nullify the crowding out. economic stability, the result was a more serious gap, in social provision, a major It is realistic to conclude that while a contributor to the most fundamental crowding out eff ect of the fi scal defi cit is a gap of all, between the level of depriva- possibility, given the level of interest rates tion and meeting basic social needs. It is in Moldova and the likely use of increased essential that future fi scal defi cit targets public expenditure for infrastructure, derive from the goal of poverty reduction, there would be little practical impact. as well as the needs of macroeconomic stability. This means using fi scal policy actively as an instrument of growth pro- Monetary Policy motion. and the Exchange Rate

Sustainable rates of growth are unlikely The National Bank of Moldova (NBM) to return Moldovans to their standard of is assigned the task of managing the living of 1990 in less than a half century. exchange rate. Nowhere in its terms of Thus, poverty reduction requires that the reference is it explicitly required that the public sector play a major role in social NBM should target infl ation. The team provision. Emphasis on private provision was informed that in 2004 the NBM ‘nar- of education and health care has and will rowed’ its interpretation of its mission, continue to be a vehicle to perpetuate to one primarily focussed on infl ation poverty and inequality. targeting. The primary instrument it has used to realise this has been the NBM Since 1998, Moldovan governments have base interest rate. This has resulted in not merely brought the fi scal defi cit un- consistently high real values for the NBM

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction der control, but maintained it signifi cantly rate, and damagingly high private com- below the three percent criterion of the mercial rates.

Executive summary

10 11 These real interest rates represented a defi cit during these thirty-two months. major relative price misalignment in the Since capital infl ows were relatively small, economy. The ‘Golden Rule’ from growth the only credible explanation for the ap- theory states that long term real interest preciation is the infl ow of remittances, a rates should not exceed the sustainable large part of which go unrecorded. rate of real per capita growth. Based on growth during 2000-2004, that sustain- The remittance eff ect on Moldova’s ex- able rate can be estimated as four per- change rate has been similar to that mani- cent. By this rule, private borrowers faced fested in countries passing through an commercial interest rates misaligned by a export commodity price boom and ones factor of four. receiving large infl ows of concessionary or commercial capital. In eff ect, the While the private rates probably refl ect exchange rate disconnects from export market manipulation, the NBM rates refl ect competitiveness, to rise or fall in reaction a concern over the return of unmanage- to purely fi nancial fl ows. For Moldova, able infl ation. The NBM base rate is an in- with its large trade defi cit, this is a very appropriate instrument to use to deal with serious matter. Further real appreciation this anxiety, because hyper-infl ation was could undermine export performance by a transitory phenomenon of the change rendering marginal producers unprofi t- from central planning to market regula- able, and discouraging the entry of new tion. Since the hyperinfl ation of the early producers. It is quite clear that in Moldova 1990s, Moldova has displayed considerable exchange rate movements that under- price stability, except during the shock of mine the production of tradables are not the Russian fi nancial crisis when exchange pro-poor. We conclude that prudent and rate depreciation again provoked an infl a- pro-poor exchange rate policy would tionary burst. All evidence suggests that involve more aggressive intervention to infl ation in Moldova is a structural phe- manage the exchange rate. nomenon resulting from domestic supply constraints and the infl ow of remittances, The NBM is a competent and professional and is not sensitive to interest rates. organisation. Its operations suff er from the absence of formalised mechanisms The Moldovan government and its mone- of accountability which exist in almost all tary authorities could safely pursue a low European countries. Its functions would interest rate policy without provoking in- be enhanced by mechanisms that created fl ation. The eff ect of such a policy would parliamentary and public oversight, to fos- be to stimulate growth, which would enter transparency through accountability. hance poverty reduction. Even should this be accompanied by moderate infl ation in the low teens, and there is little reason to 2.5 Macro Policy think it would, research suggests no link for Pro-poor Growth between growth rates and infl ation up to about forty percent. Abandoning infl ation Based on our analysis of the Moldovan targeting would prove to be a pro-poor economy and objectives defi ned by the policy option. government, we recommend:

As serious as the distorted interest rates 1. Since independence, fi scal policy has exis the behaviour of the exchange rate. cessively constrained growth, and a new After a long period of stability from 1999 fi scal strategy is needed to correct this. through 2003, the Leu appreciated in the 2. Increased government expenditure second half of 2003 and throughout 2005, could be growth-enhancing and pro-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction by fi fteen percent. In part due to the ap- poor, and is likely to ‘crowd-in’ private preciation, the country ran a massive trade investment by Moldovans.

Executive summary

12 13 3. The increased government expenditure toward poverty reduction. The fi scal space could be funded through a larger fi scal for fostering pro-poor growth will remain defi cit, fi nanced either through bonds limited until the government can refi nance sales to the public or by monetisation its remaining multilateral and bilateral debt. of the public debt. Neither method of To do so, it needs cooperation from its mul- fi nance would have a substantial impact tilateral and bilateral creditors. Its successful on infl ation. eff orts to reduce its external debt place the government in a stronger position when it 4. As great as the social expenditure defi cit negotiates with these creditors. In addition is in Moldova, increased government ex- to debt, the government has some fl exibil- penditure should focus on public invest- ity to expand its range of options. In part, ment. its success hinges on its ability to mobilize 5. Private sector interest rates are exces- domestic resources. It has already reformed sively high in Moldova, and represent a its tax system, and more can be done to major price distortion that has contribu- improve revenue generation. ted to decline, stagnation and artifi cially delayed recovery. The NBM should con- The government should assign a larger sider further interest rate decreases, share of public expenditures to economic aiming for a real rate in the range of four services, and to public investment in eco- to fi ve percent. For this to result in lower nomic infrastructure. Since much of the commercial rates, stricter regulation of country’s infrastructure has deteriorated, private market power is required. this investment is badly needed. It could impart a stimulus to economic growth, be- 6. The government should consider amen- cause appropriately designed public invest- ding the law regulating the relationship ment can ‘crowd-in’ private investment. between it and the NBM, to include mechanisms for direct accountability of Mobilizing tax revenue is crucial for imple- the NBM to parliament and the public; menting an investment-led growth and the Federal Reserve System of the United poverty reduction strategy. However, re- States provides one model of how this venues of the consolidated budget were could be done; about a quarter of GDP in 2003, down 7. The NBM guidelines could be altered to from a third of GDP in 1997. They are mandate actions to maintain a competitive projected to decline further, to 22.5 per- exchange rate. Because of the creeping cent of GDP in 2007. Expenditures in the over-valuation of the Leu due to infl ows budget have been reduced even more of remittances, more aggressive manage- drastically than revenues, from forty per- ment of the exchange rate is required. cent of GDP in 1997 to less than a quarter in 2003. It is clear from the low budget defi cits that the government has implemented a relatively tight, if not contractionary, fi scal policy in recent years. 3. PUBLIC EXPENDITURE Moldova has been advised to increase its reliance on indirect taxes, mainly the AND TAXATION Value Added Tax (VAT). In 2005, indirect taxes accounted for over forty percent of state revenue. However, indirect taxes 3.1 The Public Budget are projected to decline as a ratio to GDP and revenue generated from the VAT is The public budget of Moldova should be expected to rise marginally, according to used to accelerate economic growth, gen- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction the Medium Term Expenditure Framework erate employment and direct resources (MTEF). These trends in indirect taxes

Executive summary

12 13 place a burden on direct taxes to raise they would reduce tax evasion by reduc- re venue. At the same time that revenues ing the ‘shadow economy’, and that the are expected to decrease, contributions tax burden contributes to poverty by to social and medical insurance are reducing growth. With regard to the fi rst, programmed to increase. Increases in the rather narrow tax base and the size of expenditures for social protection can be the ‘shadow economy’ in Moldova are not pro-poor and are badly needed. But as primarily the result of tax rates, nor due to long as debt servicing remains high, public government regulations, though the lat- investment and the provision of essential ter could be reformed. The ‘shadow econo- public services will be constrained. my’ is a structural phenomenon, which characterises all low-income countries in The government lowered the tax rate on which a large portion of the labour force corporate profi ts to make Moldova an at- is self-employed and casually employed. tractive tax haven for foreign investors. At With regard to the second, increasing the the same time that the government low- progressivity of the personal income tax ered the tax rate, it projected that corpo- would reduce the burden on the poor rate tax revenue will increase slightly from while maintaining aggregate revenue. just under eleven percent of all revenue in Without a reliable estimate of the gains in 2004 to slightly over twelve in 2007. How- tax compliance as a result of rate reduc- ever, if fi rms have successfully avoided a tion in Moldova, embarking on the policy higher tax rate, there is no guaran tee that courts fi scal disaster. they will abandon avoidance at a lower rate of fi fteen percent. Nor is it certain that foreign investors will move into Moldova 3.2 External Debt principally because of low corporate tax rates. More crucial than the tax rate are The external debt is a tragic legacy of the the profi t opportunities that the economy early transition period, when average in- generates, which in part depend on the come dropped precipitously to the level of ability of the government to use public a low-income country while international expenditures, especially public invest- fi nancial institutions did not off er conces- ment, to stimulate the economy. As the sional terms for lending. The external debt economy grows and corporate profi ts burden continues to constrain the govern- increase, moderate rates on profi t income, ment’s ability to allocate public revenue such as twenty-fi ve percent, should repre- for growth and poverty reduction. sent no signifi cant disincentive to foreign investment. Measured by present value, Moldova’s external debt was 126 percent of its exports The challenge facing the Government is in 2002, placing it in the ranks of other low- that direct taxes have fallen as a percent- income transition economies that have a age of total revenue since 1997. Given substantial external debt burden. Despite the decline, it is ill advised to lower tax promises, the international development rates on personal and corporate incomes. agencies have not launched initiatives This report recommends that the 2003 to provide signifi cant relief of the debt structure for personal income taxation burdens of these transition countries, and be maintained, and that corporate profi t this should be a priority. The Moldovan taxes return to twenty-fi ve percent, which government managed to pay off its trade would contribute to a sustainable fi scal credits, renegotiate its Eurobond debt, and structure consistent with growth and conclude agreements on bilateral debts poverty reduction. with Romania and Turkey. It is not without irony that these middle income countries

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Two major arguments have been ad- have been more generous in debt relief vanced in favour of lower tax rates: that than much richer countries.

Executive summary

14 15 In order to fi nance its budget defi cit in the both education and health increased du- 1990s, generated primarily by external ring this period, those on agriculture, trans- debt servicing, the government had little port and communication declined. Overall, alternative but to increase domestic debt. the share of social services in GDP increased The external debt burden forcing more from twenty percent to twenty-one, while domestic debt at high interest rates em- the share of economic services declined phasises the problematical nature of the from 4.3 to 4.0 percent. Economic services original IBRD-terms loans, which created are programmed to decline further, to 3.2 no assets that would generate an income percent in 2007, while social services will stream suffi cient to repay them. remain virtually constant at 21.4.

Since the government has endured much Given the importance of agriculture and of the pain in reducing the commercial agro-industry to future growth and ex- and non-Paris Club bilateral components ports, the lack of expenditures on eco- of its external debt, it should be in a nomic services for the sector, particularly stronger position to lobby for favourable public investment, is not optimal. Moreo- terms for the remaining part. The starting ver, the lack of public investment in infra- point for such negotiations is the recogni- structure for transport, communication tion that Moldova should have been pro- and energy indicates an erosion of the po- vided with concessional lending begin- tential for long-term sustainable growth. ning with the fi rst loans it contracted in Rural and agricultural infrastructure is the early 1990s. The accumulated interest essential for Moldova’s industrial develop- rate diff erential between the two should ment as well as its agricultural develop- immediately be forgiven. In addition to ment. Even compared to neighbouring the interest resulting from the diff erence transition economies, Moldova allocates between IBRD and IDA terms, the maturi- relatively less to economic services and ties of the non-concessional debt should relatively more to social services. Such an immediately be rescheduled. The result allocation of public resources suggests would be that the net present value of that the full recovery from the catastroph- Moldova’s debt would drop substantially, ic decline of its economy in the 1990s will facilitating the prospects for growth and be more protracted than anticipated. poverty reduction. 3.4 Fiscal Policies 3.3 Public Investment for Pro-poor Growth

The Moldovan government has no Public Based on the analysis above, the report Investment Programme to rationalize its proposes the following policy measures: small capital budget. This absence and reliance on external assistance to fi nance 1. While we support the government’s public investment contributes to the commitment to reducing tax evasion, re- lack of coherence of public investment. ducing tax rates is unlikely to contri bute Moldova needs a consolidated and co- to this goal. Reducing evasion of the ordinated Public Investment Programme personal and company taxes could be that is more ambitious, more geared to achieved through mechanisms that have investing in growth, and more focused on proved successful in other European poverty reduction. countries. The government should con- centrate on raising tax revenue within a An examination of trends in social and more progressive structure. economic expenditures during 2000-2004 2. The government could strengthen the

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction underscores the relative neglect of the effi ciency of the VAT, and design it to economic services. While expenditures on have a progressive structure.

Executive summary

14 15 3. The rate on corporate profi t tax should pre-transition levels. Poverty in Moldova return to twenty-fi ve percent to make it combines a) low income, b) high mortality consistent with the top rate for personal and morbidity, c) declining access to sub- income tax. sidized health and education, d) growing social exclusion, and e) a disempowering 4. Two concrete measures would facilitate lack of participation by the poor in deci- pro-poor budgeting. First, the Ministry of sion-making. Demographic factors, such Finance could institutionalise a po verty as gender, old age and disability, were impact assessment of each annual budg- not major guides to the poverty of house- et. Equally important would be a gender holds, and diff erences in poverty inci- impact assessment of the budget fo- dence are not signifi cantly related to level cused on, but not limited to, the impact of education, except for those with higher on poor households. education whose rate is almost zero. 6. Without delay creditors should enter into negotiations with the government aimed An important characteristic of poverty in at restructuring external debt with the Moldova is the volatility of household sta- applicable rescheduling and grace pe- tus; that is, the movement of households riods for repayment and interest rates. into and out of poverty over short periods of time. This has major implications for 7. The IMF should facilitate and give full policy, since it implies that the identifi ca- support to the government in its nego- tion of poverty households in one time tiations with Paris Club bilateral creditors. period may not apply to the same house- The starting point for such negotiations holds in a subsequent period. is the general recognition that Moldova should have been provided with concessional, instead of non-concessional, 4.2 Characteristics of Poverty lending beginning with the fi rst loans it contracted in the early 1990s. For people in rural areas, the relative risk 8. The reduction in expenditure on eco- of poverty and extreme poverty was very nomic services should be reversed in high compared to households in cities. order to increase the growth capacity of Poverty in rural areas resulted from low ag- the economy. ricultural productivity and limited employment opportunities outside agriculture. 9. The government should establish a pub- The overwhelming majority of rural resi- lic investment programme (PIP) to ratio- dents farmed the plots assigned to them nalise and increase desperately needed by the land reform, or hired themselves out capital expenditures. as rural labourers on larger farms, with little opportunity for non-farm earnings.

Poverty among pensioners was extensive, but not signifi cantly diff erent from the national average, and their risk of extreme 4. POVERTY, INEQUALITY poverty was a bit lower than the national AND POLICY IN average. The lower level of poverty among pensioners in Moldova is due to the al- MOLDOVA most universal coverage of the state pension inherited from the socialist era, which survived the fi scal collapse of the fi rst 4.1 Introduction decade of transition. The enhancement of this universal pension programme should Both income-poverty and human-poverty

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction be a major element in the government’s sharply increased after 1991 in Moldova, poverty reduction strategy. and by 2005 had not recovered to their

Executive summary

16 17 Evidence from most other countries reversed when the economy contracted shows that female-headed households during the Russian fi nancial crisis. In 2001 a frequently face a higher than average risk sustained economic recovery fi nally began. of poverty. While gender discrimination Inequality declined during 2001-2003, with was robust in Moldova, the data indicate a rise in the income share of the bottom that female headed households faced a quintile. Part of the increase in inequality in slightly lower l levelevel ofof povertypoverty thanthan malemale the early 1990s was unavoidable, given the headed ones. If valid, this might be ex- changes in economic system and output plained by three factors: women’s level collapse, and the subsequent protracted of education being equivalent to men’s stagnation. Part was dysfunctional, due to (a legacy of the Soviet era); the indus- the weakness of the legal and institutional tries that collapsed primarily hired men; framework of the fi rst years of transition, and female headed households may be which was unable to prevent accumula- households receiving remittances. tion of wealth in the hands of a few. These inequality changes were an important con- The statistics suggest that widening ac- tributor to the rise in poverty. Policy seeking cess to higher education would at best to reduce poverty must aim at reducing the create the supply side conditions for pov- level of inequality, such as increasing the erty reduction, for there must be jobs for progressivity of taxation. the educated to take. The fundamental employment problem in Moldova, at all In the report we ask what would be the levels of education, is demand, not qualifi - poverty outcome if in the 1990s policy cations. While the highly educated are less makers introduced policies to restrict likely to be unemployed, a substantially the growth of inequality. The results of larger number of tertiary graduates are the calculations show that to cut pov- unlikely now to do more than raise the av- erty in half by 2007 requires generating erage qualifi cations of the unemployed. a sustained and rapid growth of GDP in the range of eight percent, without a As one would expect, in Moldova the further deterioration in the distribution unemployed face a higher risk of poverty of income. A similar reduction in poverty and extreme poverty. Workers employed would be achieved if Moldova grew at a in the private sector have a somewhat more achievable and more egalitarian four higher risk of falling into poverty than percent a year. Both growth and greater those working in the public sector, prima- equality are essential for poverty reduc- rily because the majority of private sector tion: either growth or greater equality employees is employed in agriculture. alone would be insuffi cient. .

4.3 Inequality and Poverty 4.4 Policy Implications

Catastrophic economic collapse was re- The statistics on poverty suggest several im- sponsible for well over half of the increase plications that are the basis of policy recom- in poverty over 1991-1996. Poverty also rose mendations found elsewhere in this report: because of a marked worsening of the distribution of income. The Gini index of the 1. The ‘poor’ is not a stable category, be- distribution of gross earnings rose phenom- cause households move in and out of it enally in a short period of time, from 0.25 in over short periods of time. 1989 to 0.39 in 1992, to stabilize near that 2. Inequality has been a major cause of level during 1993-95. During 1997 and the poverty. fi rst half of 1998, output recovered slightly,

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction and inequality increased further, to a Gini 3. The universal state pension is a major coeffi cient of 0.42. The modest recovery was mechanism of poverty reduction.

Executive summary

16 17 4. Widening access to education, while fi nance Alliance estimated that between desirable in itself, would not have a sub- 265 and 285 thousand Moldovans had stantial poverty reducing eff ect, because gone in search of work. All of these are far employment is limited by aggregate de- less than the media reports of one million mand; and Moldovans working abroad. 5. Growth and greater equality are essential The majority of migrants, almost sixty for poverty reduction, and any policy percent, go to Russia and about thirty aiming at this objective must include percent go to EU member states. Of EU both mechanisms. countries, Italy is by far the most common destination, accounting for almost twenty percent. Female migrants are more likely to go to Mediterranean countries than to northern Europe or members of the 5. MIGRATION, CIS. This pattern conforms to the Soviet period when Moldovans sought seasonal EMPLOYMENT AND work in agriculture and construction in POVERTY REDUCTION other parts of the Soviet Union, particularly in Russia and Ukraine.

5.1 Migration from Moldova 5.2 Migration and Through remittances, migration served Poverty Reduction as a major survival mechanism for households during the 1990s, preventing an The underestimation of remittances in even worse economic collapse, and has household budget surveys leads to some been the primary mechanism for the overestimation of poverty. In spite of this, recovery since 2000. In addition, remittan- the surveys show that the ex-post iimpactmpact ces reduced the government’s depend- of remittances in reducing poverty is very ence on conditionality-based borrowing, large. Many more families would have re- allowing for fl exibility in macro policy. mained below the poverty line if the pos- However, the infl ow of remittances has sibility of migrating had not been there. A the eff ect of appreciating the exchange rough estimate is that migration reduced rate, and making exports less competi- the incidence of poverty in 2002 by some tive. The social costs of migration have 20-25 points, and that these sharp gains in been unacceptably high, the impact on income-poverty were accompanied how- the domestic labour force negative, and ever by a worsening in poverty. The inde- the level of remittances unsustainable. terminacy in these numbers indicates the diffi culty in measuring aggregate poverty There are substantial diff erences between and identifying poverty households. offi cial statistics and unoffi cial estimates of migration from Moldova. The data on legal Remittances have the possibility of in- emigration suggest that yearly outfl ows creasing future consumption through during 1992-2003 range from fi ve to nine income generating investments. The vast thousand. Estimating illegal migration is majority of respondents intended to use extremely diffi cult. In 2000 the Department the assets accumulated from migration for of Statistics and Sociology estimated that consumption. Business investment seems were 234,000 Moldavians working abroad a priority for very few. Any secondary legally and illegally. An estimate close to impact of investment on income poverty this was obtained by the Labour Force reduction through future consumption

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Survey of 2003. In 2004 an independent therefore seems limited in Moldova. As in survey conducted by the Moldova Micro- many other high migration countries, one

Executive summary

18 19 of the greatest challenges for government 5.4 The Social Impact policy makers is to change this and facili- of Migration tate the fl ow of remittances into employment generating activities. Traffi cking in human beings has been an outrageous aspect of migration in Moldo- va. This traffi cking involves various me- 5.3 Migration and thods of misleading potential migrants, the Labour Market especially females into the sex trade. The root causes for the supply of Moldovan The dynamics of the educational sector women to the sex industry are poverty, and the migration process have seriously social exclusion (especially for the Roma aff ected the domestic supply of skills in population), and domestic violence. Moldova. The enrolment of children from the poorest households at the upper sec- Traffi cking of people fragmented com- ondary level has declined to fi fty percent, munities and families, depleted human compared to seventy-one for the richest capital, deprived households of expected quintile. At the university level, the situa- remittances, and caused child neglect. tion is worse. The enrolment rate for the Migrants working illegally in host countries poorest women and men has consistently were subject to exploitation and abuse declined over time to about four percent from criminals organising the migration, in the mid-2000s. For the richest quintile, employers and local authorities. Illegal enrolment not only improved, but was migration contributed to the expansion more than four times than that of the of crime networks in Moldova, which also poorest quintile. Various social illnesses, deal in arms, drugs, and other contraband. some of which are by-products of poverty, alcoholism, drug dependency, and Migration has adversely aff ected children. a high criminality rate, further depleted When children of migrants remain at the quality of the labour force. home under the care of relatives or neighbours, they fall risk to leaving school early, Migration from Moldova strongly aff ected moving away from home, and ending up the domestic supply of skilled labour. Al- as street children. Children themselves most fourteen percent of the population became subject to human traffi cking. Ac- and twenty-three percent of the work- cording to the data from the Ministry of force worked abroad in 2003. Over one Interior, Moldova became a major source half of the workers abroad were under in supplying underage girls to Russia for age thirty, and eighty-eight percent un- sexual exploitation, with an estimated fi ve der forty-fi ve. During 2000-2004, thirty-six thousand traffi cked there annually. thousand doctors left the healthcare system, and twenty-eight thousand teachers left the education system to work abroad. 5.5 Policy Suggestions By any measure, these represent a serious brain drain. Migration of unskilled work- Moldova is faced with a number of pos- ers led to higher than normal depend- sible migration scenarios and it may not ency ratios and to severe family problems, be able to choose the one it would prefer. with children left by their parents growing The status quo is unlikely to continue. The up without care and adequate education. social costs of traffi cking and associated Remittances would benefi t the supply of negative aspects of irregular migration skilled labour if they were used for edu- are not acceptable for society. Moldova cation. In other countries migration has requires a policy framework which builds brought benefi ts from new skills of the on the positive aspects of migration and

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction returning workers, but there is little evi- minimises the negative. Designing this po- dence of this in Moldova. licy would be a political exercise. A public

Executive summary

18 19 consultative process is needed to create a 1. Encouraging commercial banks to off er consensus on the costs and benefi ts of mi- foreign currency accounts; gration. The fundamental issue is whether 2. Measures to reduce the costs of remit- Moldova would be a society dependent on ting funds; migration and remittances, in return for the possibility of a quicker return to pre- 3. Ensuring a well-regulated and eff ective transition living standards, or whether the fi nancial sector, including non-banking vulnerability of such a society requires a fi nancial institutions such as micro-fi - progressive reduction in the importance of nance mechanisms; and migration and remittances. 4. Developing a foreign currency bond with transparent taxation. The Migration Policy Concept of the Re- public of Moldova was adopted by parliament in October 2002. The Migration Department is supported by the donors and lenders to improve migration management, develop and strengthen strate- 6. EMPLOYMENT, gies and policies and develop appropriate agreements with destination countries. PRODUCTIVITY Given the importance of migration and AND POVERTY remittances, it is surprising that the EGPRS has little to say about the subject. It does, however, identify broad directions for the 6.1. Employment and Poverty future, all of which are important, and deserve the fi nancial and technical support Equitable economic growth is essential for of the international community. reducing poverty in a sustained manner, and does so through employment that First, the government should bring le- generates livelihoods above the poverty gislation on migration in accordance with line. The main problem in Moldova is that the international standards. The other well-paid jobs are not created in suffi cient PRSP actions are closely linked and relate numbers to lift the working population to conducting a comprehensive study out of poverty rapidly. Of the people who of the migrant population, creating an were in poverty in 2002, seventy-fi ve per- information system, and improving the cent had work. Their problem was that the management of migration processes. work did not pay suffi ciently to escape poverty. Thus, the challenge for Moldova Perhaps the most important economic is the creation of employment opportuni- issue associated with migration is how to ties associated with increased productivity encourage and facilitate households to so that the employed population can ob- direct part of remittances to productive tain incomes suffi cient to escape poverty. investment. International experience suggests that compulsory measures are only Unemployment has fallen, but employ- eff ective when employment is organised ment has not risen. From 1999 to 2004, the through employment contracts with the rate of unemployment fell from eleven to government of the migrants as one party. seven percent, and the absolute number The mechanisms to foster investment must of unemployed people also declined. The match the characteristics of private and fall in unemployment in great part republic fi nancial institutions. The design of sulted from a steady decline in the work- these is beyond the expertise of this report, ing age population, due to out-migration which can only suggest general guidelines. and an increase in the numbers of people

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Voluntary mechanisms to foster invest- becoming economically inactive, the ‘dis- ment from remittances could include: couraged worker’ phenomenon.

Executive summary

20 21 6.2 Employment, Wages impinged on enterprises, enterprises and Productivity shed this labour. As a result, employment growth lagged behind output growth. As In 2003, the level of employment stood a result, the pressure to migrate abroad eighteen percent below the 1996 level continued, keeping the measured rate of in agriculture, sixteen percent below in unemployment down, but at consider- industry and nineteen percent below able social cost. in services. After the Russian crisis, employment recovered in both agriculture and industry, but not in services. While 6.3 Women’s Employment employment has stagnated or declined and Remuneration in most sectors, output and value-added have increased. The diff erence between Changes in the structure of employment GDP growth and employment growth is in the economy have diff erent impacts on even greater after 2000. Between 2000 men and women, as a result of embed- and 2003, overall GDP grew at four per- ded discrimination. As in other countries, cent per annum, and employment de- women dominate in health and social clined by 2.4 percent. work, education, hotels and restaurants and, surprisingly, fi nance. The latter pays The employment-output relationship has the highest monthly salaries across sec- implications for the growth of producti- tors, but accounts for only two percent vity and labour income. By defi nition, the of women’s employment. Health, social failure of employment to keep pace with work and education, on the other hand, production meant that labour producti- account for over forty percent of women vity increased. During 1996-2003 aggre- workers and are among the lowest paid gate productivity grew at 4.8 percent per sectors in the country. Approximately annum. Along with the slow growth of fi fty-seven percent of women worked in employment and rising productivity went sectors with wages below the national an increase in real wages beginning in the average, while for men the fi gure was mid-1990s. This reversed during the Rus- forty-seven percent. sian crisis but rose again afterwards. The apparently high rate of increase of real The anti-discrimination legislation of wages, almost eight percent per annum, Moldova is quite strong on paper, and, if must be placed in the context of the cata- eff ectively implemented, could reduce the strophic collapse in the early 1990s. Given pay diff erential between men and wo men. the decline in the real value of money and However, enforcement is not eff ective, the social wage during the fi rst half of the perpetuating the inequality despite en- 1990s, it is quite possible that the living lightened legislation. There is widespread standards of workers dropped to such a insensitivity to the discrimination against low level that worker productivity was women in Moldova. Legislation against undermined. That real wages seem to have pay discrimination, for example, is easily outpaced the growth of labour productivi- evaded within a discriminatory culture by ty could be a process of adjustment toward relegating women to low-paid jobs. a more sustainable ‘effi ciency wage’ level. Strong leadership by the government To some extent, the failure of employment is necessary to overcome gender dis- to recover during the recovery should be crimination, and the public sector can expected in a transition economy. As in be a positive example. Moldovan women other transition countries, enterprises in perform an enormous volume of unpaid Moldova carried more workers before the work within the family, and play a leading

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction transition began than would be justifi ed role in many community organisations. At under market criteria. As market forces the same time they are absent from deci-

Executive summary

20 21 sion-making, partly because of the lack of 1. An employment strategy that covers a coherent and supportive state frame- policy actions at the macro, sectoral and work and partly because of the patriarchal micro levels. Macro policies can create mentality which still aff ects the whole of more rapid growth, sectoral policies can society, including women themselves. make growth more broad-based and micro policies can enhance workers’ access to economic opportunities. 6.4 Generating Wage 2. Macro policies would include active fi s- Employment cal policy based on public investment. The most important constraints on em- The rate of fi rm closure in Moldova during ployment growth in Moldova are lack the 1990s was one of the lowest among of demand and productive investment. European and Central Asian transition An active fi scal policy would be growth- economies. Enterprise level data for 2001 enhancing through public investment, show that expanding enterprises created supported by low real interest rates, and jobs at a slower rate than the contracting a modestly undervalued exchange rate fi rms lost them. Thus, it was not so much to foster exports. the severity of job destruction as the sluggishness of new job creation that proved 3. Sectoral Policies should seek to raise pro- the main drag on employment in Moldo- ductivity in agriculture and link agricul- va, and this sluggishness can be explained ture to agro-industry in order to create by demand-constraining macro policies. widespread employment that generates In the face of severe demand constraints, decent incomes. The government should it is not surprising that new enterprises direct public investment and provide in- found it diffi cult to emerge and the exis- centives to the commercial banking sys- ting ones found it hard to expand. tem to provide credit to these purposes. 4. Micro policies should focus on provi ding Whatever may be the problems with the education and skills to workers and labour market in Moldova, it is not ‘infl e- supporting trade unions in all sectors, xible’ with regard to wage ‘adjustment’. including agriculture. Because labour Real wages collapsed in the fi rst half of markets are ineff ective and ineffi cient, the 1990s, indicating excessive down- it is through trade unions that productiv- ward wage ‘fl exibility’. Falling wages did ity increases will translate into real wage nothing to reduce employment decline, increases. The Ministry of Labour and contrary to the orthodox model of labour Social Welfare follows this approach. markets. The simultaneous collapse of 5. The government can also stimulate pri- wages and employment in the 1990s in- vate investment by making regulation dicated the weakness of the bargaining more transparent. While changes in the power of workers, which resulted from regulation of enterprises and markets can at least three factors: 1) the absence of an contribute to employment creation, the independent trade union movement at necessary underlying condition is still an the beginning of the transition; 2) lack of accommodating macroeconomic frame- bargaining power due to growing unem- work. Without such a framework, institu- ployment; and 3) ineff ective regulation of tional reforms would have little impact. employers by the public sector. 6. Given the level of un- and under-employment, market forces alone are unlikely to 6.5 Labour Market Policies foster job security and wage increases that would lift works out of poverty. To increase employment that would re- Therefore, the ability of workers to bar-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction duce poverty, the following policy direc- gain for conditions and wages through tion would be appropriate: unions should be enhanced.

Executive summary

22 23 . AGRICULTURE sectors, improved statistical information 7 by gender is an urgent priority in order to AND SUSTAINABLE fully understand the role of women. POVERTY REDUCTION 7.2 Post-Independence 7.1 Agriculture’s Role in Agriculture Moldova’s Growth Strategy The restructuring of collective and state farms proceeded unevenly after inde- Agricultural development is critical to pendence. Rights to land and other assets Moldova’s future. It could contribute to were distributed among households, and closing the trade gap, increase rural in- could be exchanged. Because so many comes, and possibly reduce migration. It households were eager to leave collective represents an important medium-term and state farms, the November 1994 law vehicle for assuring the recovery of the halted the process of privatisation. A year Moldova economy and laying the neces- later the Constitutional Court declared sary basis for the eventual re-industriali- the limitations on leaving collective and sation of the country. However, without state farms to be illegal, and the process forging strong links to agro-processing, of privatization began anew in 1996. agriculture will be unable to grow either rapidly or in a sustainable manner. By October 1996, some ninety thousand private farmers, individually or in groups, There are structural bottlenecks in the worked one hundred and thirty thousand sector, and between agricultural produc- hectares, fi ve percent of Moldova’s arable ers and processing companies, that re- land. The remaining farm population require policy action in order to make rapid mained in collective farms that had been and sustainable growth possible. The nominally restructured, or in new farm principal problems are underdeveloped associations and production cooperatives markets for inputs, output and fi nance. that had been created from the collective Also weak or absent are supportive in- farms. These new institutions, ‘corporate stitutions, such as output marketing and farms’, essentially functioned as slightly input purchasing cooperatives (so-called smaller kolkhozes. The National Land Pro- ‘single-purpose cooperatives’), which can gram (NLP) of 1998, supported by external increase the bargaining power of small funding agencies, accelerated the process farmers. Further, most of the agricultural of land reform. In 1999, a law on farm sector has suff ered from years of insuffi - debt provided additional incentives for cient investment. As a result, rural roads, dissol ving collectives by arranging debt market infrastructure, irrigation systems, restructuring and consolidation. At the and communication networks have de- end of 2000, nearly 1.7 million hectares teriorated. and the majority of state farm assets were transferred from over one thousand bank- Gender aspects of the land reform have rupt kolkhozes and sovkhozes to about one been poorly documented. Rural women million former farm employees. Orchards made up a majority of the land proprie- and vineyards were also distributed to tors because men migrated to work in households according to the latter’s size. construction, technical services and com- The result was to fragment a large share of merce, losing their right to title under the land holdings into over three million small land reform. However, there appeared to privatised parcels. be few female managers of cooperatives or large farms although defi nitive statis-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction On average, each owner received three tics on this issue are lacking. As in other parcels of land, with an average parcel

Executive summary

22 23 being one-half hectare. The program fi ciency in agriculture may not be small included the establishment of a register farm size, but the underdevelopment of of land titles, and an extensive education markets for both inputs and outputs. campaign for farmers on their legal rights and market opportunities. Some of the As a result of the structural changes and new landowners chose to farm the land general economic collapse, agriculture with family labour. Others leased their contracted dramatically during the 1990s. land and other assets to the newly cre- In 2003, agricultural production was only ated corporate farms, which had been at forty-fi ve percent of its 1990 level. Live- formed out of the collective farms, and stock production proved much more vul- either worked on them as members or nerable to structural changes than crops, employees or received lease payments. decreasing by nearly two-thirds during Thus, the privatization process led to a 1990-2003. In 2001, after the second phase dual agrarian structure polarized between of redistributive land reform, agriculture one part based on a myriad of small-scale registered strong growth, which might farms of one to fi ve hectares and another have continued into 2003 but for a severe part based on a few large corporate farms, drought that destroyed the wheat harvest. such as limited liability companies, joint stock companies, Despite the decline of output, and the constraints of the 1990s, there have been Agricultural markets and related services positive developments. A recent study remain weakly developed. In the after- showed that commercial producers of math of the redistributive land reform, high value products using advanced pro- many of the more than one million ben- duction technologies, including irrigation, efi ciaries did not have the necessary skills registered higher yields for some crops for effi cient agricultural operations. It was than the average in Romania, Bulgaria and expected that the emergence of agricul- Hungary. This suggests that Moldovan tural markets would transfer land from producers have the potential to achieve less effi cient to more effi cient producers international competitiveness. and would lead to a more competitive agricultural sector. The government has viewed fragmented landownership and 7.3 Agricultural Markets the predominance of small farm size as and Poverty a bottleneck for agro-industrial development. Hence, it launched several policy The underdevelopment of markets for proposals to consolidate land and create fertilizers, seeds, machinery and credit is production cooperatives. These initiatives a serious problem. Access to inputs de- have encountered both domestic and in- clined during the transition to a market- ternational opposition because they are based economy and new input markets seen as a step towards reversing some of did not take the place of state provision. the main outcomes of the redistributive Because of the dramatic reduction of live- land reform of the late 1990s. The more stock, organic fertilizer is scarce in rural important of these initiatives were can- areas. Small agricultural producers have celled in 2005. virtually no access to commercial lending. Since land prices remain low and other Although the statistics on land distribu- farm assets are scarce, farmers possess lit- tion should be treated with caution, they tle collateral for borrowing. do suggest that small farms are economically effi cient. Further, land consolidation With the support of external donors, in- has already taken place through lease stitutions have arisen to off er micro-credit

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction markets, and is likely to continue. The to farmers. The total amount is small and principal explanation for the lack of ef- directed mostly to fi nancing working

Executive summary

24 25 capital. While such credit is important, 4. Developing physical marketing infra- loans for long-term productive invest- structure could signifi cantly strengthen ment are more important, but are in short and diversify supply to markets and supply. The same shortage of credit ap- boost farm-gate prices. plies to rural non-farm enterprises, which 5. The weak bargaining position of small are very weakly developed in rural areas. peasant farmers could be partially over- The lack of credit is a major constraint to come by promoting single-purpose both agricultural and non-agricultural de- cooperatives and associations, which velopment. This is a constraint that public focus on input purchase and output policy can do a great deal to remove by marketing. Tax policies and legislation providing various incentives to fi nancial that encourage the formation of such institutions to off er more loans for such organizations could make an important development. contribution. 6. The main bottleneck for agricultural pro- 7.4 Policy Suggestions ducers in Moldova is the under-development of markets. Since these markets The following are recommendations that interlock, an integrated policy response would contribute to the development of towards developing them is needed. agriculture: 1. The government should focus its policies on helping agriculture regain its position of comparative advantage in high value crops and facilitating its linkages to the agro-industrial sector. A central 8. SOCIAL POLICIES ingredient of this policy focus is public investment in roads, irrigation infrastruc- FOR HUMAN ture, rural markets and communication DEVELOPMENT networks, which would ‘crowd in’, or stimulate, private investment. 2. The development of agro-industry is 8.1 Policy Framework central to the revitalization of the agri- and Priorities cultural sector, particularly in ensuring the latter’s linkages to export markets. The central issue of social policy is the basis The institutional and policy environ- on which these social services and ben- ment should encourage new entry into efi ts should be distributed. The recommen- agro-industry, and institute transparent dations made in this chapter for social poli- import and export procedures. cy derive from the following guidelines: the system for the distribution of social services 3. Agricultural producers, particularly small and benefi ts should be operationally prac- ones, need training and extension ser- tical, aff ordable within the national budget, vices in order to produce more and transparent to the benefi ciaries, adminis- higher-quality commodities. This could tratively feasible at reasonable cost, and be facilitated by agro-processors and ex- not require excessive government intrusion porters, who are familiar with the require- into the private realm of households. ments of their target markets. 4. The government can use fi scal mea- The distribution mechanism that most sures to stimulate the establishment of conforms to these guidelines is universal out-grower schemes that can promote eligibility with categorical targeting. ‘Univer- sal eligibility’ meansmeans thatthat therethere isis nono re-re-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction improved product quality and increase farmers’ income. striction on access (sometimes called the

Executive summary

24 25 ‘citizenship principle’). ‘Categorical target- commendation for low income countries ing’ means that a programme is directed is that on equity grounds, health expendi- to an objectively and simply identifi able ture should focus on preventative treat- group of the population, such as children ment and non-hospital care. While both under a certain age. An alternative me- are important, due to the relatively high thod of distribution is needs testing. In this life expectancy in Moldova, the poor are approach, one or more criteria are used to as likely, or more likely, to suff er from mal- discriminate between households which adies requiring hospitalisation as the non- are ‘needy’ and those which are not, with poor. In general, it is important for social the benefi t programme restricted to the policy to recognise that Moldova is not a former. As an operational matter, needs typical low income country attempting to testing is most frequently based on in- combat poverty with limited resources. It come level; a household qualifi es for the is a country which has fallen into low-in- benefi t if its income lies below the rele- come status over a short period of time. A vant level (i.e., the household is ‘poor’). majority of Moldovans, only fi fteen years before this report was written, enjoyed While not opposed in principle to distri- education services, health care, and social buting benefi ts based on means testing, protection close to a First World standard. this report recommends that its use be In the short run, temporary measures are limited in Moldova. First, poverty surveys necessary to alleviate the worst social reveal that there is a high year-to-year problems. The long term goal would be movement of households in and out of to regain the level of services that was lost poverty. This implies that identifi cation during the transition. of the poor would require annual repeti- tion, implying a high administrative cost. Second, the accurate measurement of 8.2 Human Development incomes is especially diffi cult in Moldova. Indicators The country has the diffi culties common to all low-income societies, e.g. a Moldova entered its transition with robust high proportion of self-employed, plus a human development outcomes. By the characteristic of migration countries, e.g. mid-1990s, these achievements were a a quite large level of unrecorded remit- memory. During the fi rst half of the 1990s, tances. Third, the required frequency of the health of Moldovans seriously dete- measuring incomes and seeking informa- riorated, because of eroding standards tion on matters known only to the house- of living, poor access to health care, and hold itself would imply a high degree of its declining quality. Life expectancy de- government intrusion into private space. clined to sixty-fi ve from sixty-nine years and infant and maternal mortality rates We conclude that while needs testing should increased. Sexually transmitted diseases not be ruled out as a method of distribution, took on almost epidemic proportions it should not be the basic framework for and half of the population did not receive education, health, and social benefi ts. More treatment for these diseases. practical and transparent would be categorical targeting. One such programme, the During 2000-2004 the decline in social basic old age pension, has shown itself in indicators was reversed: life expectancy many countries to be an eff ective protection for males and females improved, infant of households against poverty. mortality rate declined, the child mortality rate fell, and the maternal mortality Our recommendations on social policy rate also declined. Despite these improve- must conform not only to the nature ments, health outcomes in Moldova re-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction of poverty in Moldova, but also to the main unsatisfactory. Life expectancy of country’s characteristics. A common re- sixty-eight years is one of the lowest in

Executive summary

26 27 Europe; the infant mortality rate is more portunities because of skewed access than twice that of high human develop- to opportunities. The diff erential access ment countries; and the maternal morta- implies greater inequality in the future lity is four times the European average. and a lower rate of poverty reduction for any rate of economic growth. To improve In the household survey of 2002 one in the access of the poor to education, it is every fi ve Moldovan considered herself or essential to reduce out of pocket costs, himself to have ‘bad’ or ‘very bad’ health. both formal and informal. Reduction of The self-defi ned health status deteriorates the incidence of informal payments and considerably with age, with more than bribes in higher education may require a half of the population over sixty years review of entrance procedures. of age reporting bad or very bad health. Two-fi fths of the population considered Poor and rural families, for whom quality itself to suff er from a chronic condition. education in lyceums and universities has been out of reach, are likely to be more dependent on general secondary and 8.3 Enhancing Human vocational education. The demand for Development secondary vocational education appears to be falling, and graduates of these insti- Social policy to enhance human develop- tutions face diffi culties in fi nding employment in Moldova should focus on reducing ment in the domestic labour market. The human poverty. The effi ciency and eff ec- policies in this area need to be reviewed. tiveness of policy is dependent on the overall macroeconomic framework and broad In order to increase the value of educa- poverty reduction strategy. Across the tion to households, particularly for poor sectors, there are synergies among policies and rural households, improved functio- that should be maximized. Policy should ning of the labour market, and improved not be limited to strategies, but should also labour market opportunities for rural and address institutional strengthening. poor populations are needed. Policies to address these needs lie, for the most part, The government has shifted health ex- outside the education sector but are ad- penditure to primary care. It has been sug- dressed in elsewhere in the report. gested that a large part of the Moldovan population’s health needs can be met, if well funded and supplied, by an eff ective 8.4 Social Policy for primary care system. If so, then a basic Poverty Reduction package of services, fi tting within the available resources, should be made avail- A pro-poor social policy in Moldova would able to all on the basis of medical need. have the goal of reducing poverty, rather Suggestions have been made that this than merely alleviating it. This implies that can be done at a relatively modest cost social policies facilitate the poor in acquir- and will improve utilization of health care ing the means to permanently exit pov- by lower income groups, thus resulting in erty. From this perspective, social policy overall better health outcomes. However, does not provide ‘safety nets’ that catch it would not be a pro-poor strategy in the few or many that fall into poverty. So- health to have a two-tiered health system, cial policy is part of a growth strategy that in which the poor are limited to primary progressively reduces the conditions that care, and access to hospital treatment de- generate poverty. pendent on payment. In a low-income country, social policy

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction In education in Moldova, there is a per- faces many trade-off s, and there are inhe- sistent and widening inequality in op- rent limits to its capacity to reduce poverty

Executive summary

26 27 and provide equitable access to services 3. Progressive taxation is the key to funding and benefi ts. The measures listed below pro-poor health and education systems. recognise the trade-off s and limits, guided 4. Constraints on revenue limit the qual- by the principle that policy should not ity and extent of public health services. endorse the division of Moldova between This requires clear priorities for health the poor and the non-poor, but should expenditures. Policies should be avoided foster systems that progressively eliminate that would institutionalise a two-tiered that division. That objective would be health system, in which free access is to a achieved by combining social insurance minimal system which is income tested, and assistance programmes with a more and other health services are obtained equitable access to health and education, through private payment. Institutiona- within a macro framework of progressive lising a two-tiered system would institu- taxation and public investment. tionalise poverty. 1. Because of the nature of the Moldovan 5. Primary and secondary education should economy, its structure (relatively few be provided to all without charge. formal sector wage earners) and characte- This requires increased funding for tea- ristics (importance of remittances), there chers’ salaries, education materials, and are serious practical diffi culties to using construction of facilities. The capital income testing for access to benefi ts. The expenditure would be part of a public inherent arbitrariness of income testing investment fostering growth strategy. that results from this structure and characteristics suggests that the its use be 6. Unequal access to university education avoided when alternatives are feasible. cannot be eliminated in the foreseeable future; however, it can be moderated by 2. A benefi t system based on targeting cate- income tested funding of university fees gories of the population with universal within a system of access on the basis of access has proved poverty reducing in ability and school performance. Moldova, with the old age pension an example. This approach should be extended to other population groups. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Executive summary

28 29 C h a p t e r 1 : Introduction

1.1 Purpose of this Study transition process. Further, gender equality plays an important role in growth, This report is part of a larger UNDP project employment and poverty reduction, and on the macroeconomics of poverty reduc- cannot be ignored. The fundamental tion, which at the time this study began policy changes associated with the transi- had covered six East and Southeast Asian tion process are likely to aff ect men and and several Central Asian countries. The women diff erently because of their diff er- Bureau for Development Policy of UNDP ent positions, roles, and responsibilities in has helped to initiate a broad range of society. The unequal treatment of men global, regional and country programmes and women cannot only be questioned designed to help fashion more pro-poor on human and equity grounds but is economic policies. The primary purpose also economically ineffi cient. As a result, of this set of activities has been to broad- the development of eff ective economic en the dialogue on pro-poor policies at policies requires gender analysis and the national, regional and international le- the mainstreaming of gender into the vels. While each report stands on its own, development of the policies themselves, addressing the specifi c circumstances of including the budgeting process. each country, taken together they provide a consistent analysis and vision of the pov- It is unfortunately the case that the statis- erty reduction process. The main fi n dings tics necessary to adequately incorporate of the project are presented in series of gender into the analysis of the transition, synthesis reports organised by topic. Of and into pro-poor growth, do not for the particular relevance to this study is the fi s- most part exist in Moldova. With some cal synthesis paper (Roy and Weeks 2004), exceptions, the data that disaggregate whose principles are applied here. between men and women are the common-place statistics that one fi nds in even the least gender-sensitive countries: mortality rates, life expectancy, school enrolment, for example.1 This is a serious 1.2 Gender Analysis oversight that should be addressed im- of the Transition mediately by the Moldovan government; all the more immediately because there Prior to considering the aspects of the would certainly be funds from donors to transition, it is essential to recognise that cover the cost of collecting and analysing gender is central to understanding the the necessary numbers.

1 The Department of Statistics and Sociology (DSS) published Women and Men of Moldova in 2004 (in Romanian and Russian but not English). Gaps in disaggregated data include, among many others, disaggregation on credit and school results. Other gaps include information on violence against women and sexual harassment. There is a chapter entitled ‘Gender-based Policies’ in the Annual Social Report 2003, of the Ministry of Labour and Social Protection that contains disaggregated statistics. The DSS labour market Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction studies, including the Labour Force Survey, have basic gender disaggregation, though not to the extent recommended by ILO statistical guidelines.

Chapter 1: Introduction

28 29 Striking evidence of the lack of gender which country suffered most, it is be- perspective on the transition in Moldova, yond controversy that the experience of and other countries passing through a Moldova was unimaginably grim. In the similar process, is the virtual absence of first decade of the century a great eth- any rigorous or even through treatment nic pogrom left communities divided of the conditions of women under cen- and hostile; to be followed by a second tral planning and the subsequent marin the bleak summer of 1941, when the ket regulation. A few points are beyond country fell under military occupation, dispute: 1) before 1990, women’s access which would last for over three years. to employment was considerably bet- The end of the Second World War ter than after that date; 2) child care was brought peace, but a peace that many more available, without ‘user fees’, facili- Moldovans viewed as a new foreign oc- tating women’s employment; and 3) the cupation. No visitor to the country can deprivations suff ered by the Moldovan for long remain ignorant of the social population resulted in a dramatic increase divisions created by this history. There- in women’s unpaid work. Because the fore, it is more than remarkable, that af- central planning system ended less than ter independence the country has been two decades ago, it would not be diffi cult so free of civil strife, even when the to collect the information to systemati- economy collapsed catastrophically. cally analyse the impact of the transition from a gender perspective. Again, it is Perhaps, the absence of major civil strife likely that external funds, from research can be explained in part by the over- foundations if not from development whelming support of the vast majority agencies, could be obtained to collect the of Moldovans for independence. How- information and analyse it. All is necessary ever, upon independence, Moldovans is for the government to give the task its faced four fundamental social and eco- appropriate priority. nomic transitions, each of which would have been a formidable challenge on its Given the lack of statistics necessary for own: a rigorous gender analysis, this report is 1. from nation to country, which required reduced to the inadequate approach of the construction of the state institutions using the available information and at- appropriate and adequate to manage tempting to draw reasonable inferences the aff airs of a country, the countryhood about what that information does not di- transition; rectly reveal. As highlighted in the policy summary, it is essential that the Depart- 2. from an authoritarian to a democratic ment of Statistics and Sociology intro- regime, which necessitated the crea- duce immediate and thorough policies tion of the space in which open debate to ensure that future economic reports and contestation of power could occur on Moldova do not suff er from the funda- peacefully with political stability, the po- mental weakness that this one does. litical transition; 3. from central planning to the regulation of markets as the regime of economic management, which rendered most of 1.3 Four Fundamental the institutions of economic manage- Transitions ment anachronistic, the regime transition; and For all of Central and Eastern Europe 4. from social delivery of basic needs the twentieth century brought the ra- to market provision, which implied a vages of war and dictatorship. While

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction revolution in the manner by which the it is not possible to assess objectively population was employed, provisioned

Chapter 1: Introduction

30 31 with basic necessities, educated, and in Russia, to policy and decision making provided with health care, the socio-eco- ministries. In this context, a sharp distinc- nomic transition.2 tion must be made between institutions and personnel. The problem of national In other words, it was necessary at the policy management after 1990 was not same time to design an appropriate state due to lack of capable administrators and apparatus, create a civil society, transform professions. The management diffi culty the functions of the state, and construct arose because these administrators and market mechanisms which could provide professions found themselves in institu- livelihoods and basic needs. No post-So- tions that were not designed to carry out viet society has succeeded in carrying out the tasks of a country. all four of these transitions. The countries having greatest success, or least disaster Because institutions evolve from practice in transition, did not face the need for as well as from statute, the transforma- the countryhood transition, or bordered tion of these institutions into eff ective on or were incorporated into a Western ma nagement units could not be done European country (Poland, Czech Repub- quickly;5 tough small changes can have lic, Hungary, Slovenia and the German a major impact. In the context of building Democratic Republic).3 appropriate institutions, it is the impres- sion of the authors of this report that some The countryhood transition in Moldova representatives of international agencies, involved concrete and specifi c institution and many Moldovans, do not fully appre- building, beginning with the creation of ciate the diffi culty of institutional transi- basic state institutions that are the main- tion. This is manifested in a tendency to stays of public administration in countries, attribute all the problems of policy imple- but do not exist in provinces: ministries mentation to bureaucratic ineffi ciency, of foreign aff airs, defence, fi nance, and lack of commitment by public servants, international trade.4 Even provincial in- and a culture of corruption. While no large stitutions whose basic function did not organisation, public or private, is free of change, such as departments of health, these maladies, the public management education and social welfare, required re- problem in Moldova is much more fun- organisation because of the basic change damental. Recent research has demon- in their status from conduits and imple- strated that institutions, including public mentation agencies for decisions made institutions, are essential for sustained

2 These transitions provide the basis for an analytical division of countries in the context of the general transition process after 1990. We refer to three categories: the ‘Central and Eastern European Countries’, which includes Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic, and Slovenia; ‘Albania and the former Yugoslavian republics’, Albania, Croatia, Bosnia & Herze- govina, Macedonia, and Serbia and Montenegro; and ‘the former Soviet republics’, Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyz Republic, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, and Uz- bekistan. The fi rst group includes those countries which have been formally independent since the end of the Second World War, and those which quickly integrated into western Europe (the Baltic states and Slovenia joined the European Union in 2004. The second and third groups include newly independent countries that were previously components of a formally federated unit, with the exception of Albania. The latter is included in this group because of its previous non-integration with either Comecon. This division is similar to that used in the tables in EBRD reports (see EBRD 2004). 3 See Chapter 2 for a discussion of the transition across countries. 4 As a republic of the Soviet Union, Moldova was a member of Comecon, with the bureaucracy associated with the trading block. However, this was quite diff erent form a ministry of external trade. 5 In his edited volume, In Search of Prosperity, Rodrik writes, ‘Determinants of development such as institutions and geography change slowly, or hardly at all,’ (Rodrik 2003, p. 9), adding ‘moderate changes in country-specifi c circumstances (policies and institutional arrangements), often interacting with the exter- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction nal environment, can produce discontinuous changes in economic performances, which in turn set off virtuous or vicious cycles’.

Chapter 1: Introduction

30 31 growth (see Rodrik 2003), and arise form war over the other for a decade. No other a historical process. Frustration over the country of the Commonwealth of Inde- ineff ectiveness of Moldovan institutions pendent States has been at peace with is understandable and can be a source a democratically election government.6 of positive change; to attribute the blame Therefore, it is surprising that some de- for the ineff ectiveness to the institutions velopment agencies refer to the 1990s as themselves or those that serve in them a decade of political instability and eth- abandons historical perspective, and en- nic fragmentation. While there were fi ve ters into a reductionist ‘black box’. Less changes of government during the dec- than fi fteen years after independence, it ade, each was achieved without violence, is not possible to judge whether the proc- through elections that judged by internal ess of building eff ective institutions has and external observers to be open and proceeded quickly or slowly. representative. This would seem a more commendable record than that of other These observations are not to deny or ex-Soviet republics in which the absence justify corruption, either in the public or of changes in governments represented private sector. They place corruption in ‘stability’ hardly consistent with demo- its context: corruption manifests itself in cratic processes. individual behaviour, but its causes are systemic, inherent in the transition proc- Across the countries in transition, one ess. Reducing corruption is a diffi cult task of the most intractable aspects of insti- that is part of the democratic transition to tution building has been creating the open and transparent government opera- mechanisms to regulate markets out of tions. the agencies and personnel of central planning. The problem goes far beyond What should be beyond controversy is establishing the ‘rule of law’ and ‘foster- that Moldovans have excelled at the crea- ing private sector development’. The fi rst tion of political space in which confl icts problem with the terms is that they are so can be managed with limited civil strife, vague as to include any policy measure. an achievement that has allured several The second, equally serious problem is of the former Soviet states. We have been that the period of transition is by defi ni- surprised to discover that in practice there tion one in which laws and regulations is considerable disagreement about this change. It is probably the case that un- conclusion, especially with regard to the certainly about the legal framework is intentions of the current government. To inherent in the transition, since laws are an extent this is understandable among implemented by institutions that are be- Moldovans, who for over fi ve decades ing created. Further, in established market lived under governments dedicated to economies, implementation of the legal restricting political rights. Governments framework involves interpretation of laws, rarely resist the temptation to enhance drawing heavily on legal precedent which, their political strength and undermine by defi nition, does not exist in transitional their political opposition, and Moldova economies. is not an exception to this general rule. However, Moldova has seen many chan- Third, ‘the rule of law’ in a country with ges in government through elections, it established market institutions evolves has resolved one potential internal con- over time, adapting to changing circum- fl ict over secession, and has not been at stances. In general circumstances change

6 See DFID (2004, 8), which characterises the countries as follows: tension between Armenia and Azerbaijan over the Nagorno-Karabakh territory; Kyrgyz Republic, Tajikistan and Uzbekistan involved in confl ict over the Ferghana Valley; in Georgia there separatism in the South Ossetia territory and the spill over from the Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction war in Chechnya, both providing violence; and civil wage raged in Tajikistan during 1992-1997. No more than two of these countries can be said to have democratically elected governments.

Chapter 1: Introduction

32 33 slowly, allowing for adaptation to keep It is diffi cult to generalise about the im- pace. However, in transitional countries, pact of corruption on growth and poverty dramatic and fundamental changes oc- reduction. It is frequently asserted that cur over a few years, and the legal system the poor suff er most from corruption, must be created quickly, usually based on either directly through the cost of bribes frameworks created in mature market so- for services, or indirectly through loss of cieties. The probability that a legal frame- employment and earnings because arbit- work so created will be fully adequate for rary and mercenary offi cials discourage the transitional country is very low. This private sector development. Informants implies that the inconsistencies among, stressed to the mission members the ne- frequent changes in and ambiguities gative impact on the poor of such corrup- of regulatory statutes are unavoidable. tion in Moldova, but the reports remained Those opposed to the government for at an anecdotal level whose quantitative other reasons will tend to interpret these importance could not be assessed. For inconsistencies, changes and ambiguities example, it could not be determined as evidence of malign intent. While this whether the quantitative importance of interpretation may have validity, it must government corruption on the poor ex- be placed in the context of the unstable ceeded or was less than the impact of ad- institutional dynamics of the transition ministered prices of key commodities by process. private companies, which represents the private equivalent of public sector corrup- A concrete case of misinterpretation of tion. Further, the legacy of authoritarian public sector behaviour due to failure to and capricious rule in Moldova has, quite appreciate the complexity of the transi- understandably, produced a tendency for tion process is the emphasis in Moldova parts of the population to attribute the by many external and domestic actors on worst of motives to public offi cials. That corruption. Few governments are free of the current name of the governing party corruption, and attempts to make relative is ‘the Communist Party’ reinforces this judgements about the degree of corrup- reaction among its opponents, both do- tion are doomed to failure or to stating mestic and foreign. the obvious, except when comparing extremes. In the case of Moldova, writ- Closely related to the issue of corruption ings on corruption focus on reporting it is the much-used and abused metaphor, as a problem and off ering dubious indices ‘the business climate’, or more dubious of its severity (see for example Obreja, still, ‘the investment climate’. Using them Efi m, Gasca, and Potirniche 2002; and is more likely to obscure than to clarify Chirtoaca, Nicolae, Gudym and Petrushin discussions of private sector develop- 2000). While description and attempts to ment, especially when the terms are de- measure may be informative, the central fi ned from the point of view of business issues for this report are the causes of interests. To keep within the meteorologi- corruption and its impact on the develop- cal metaphor, the fl ora and fauna that will ment process. A wealth of research shows thrive in one climate may or may not do that the political parties in power in coun- so in another. Carrying on a discussion tries aff ect the form corruption takes, but that suggests there be a single ‘climate’ only to a limited extent are the cause of in which undiff erentiated ‘business’ will it. It may appear that the government is thrive leads to less in-sight into problems corrupt, but this appearance provides lit- that over-sight of crucial distinctions. Fur- tle explanatory value if each succeeding ther, the metaphorical weather advocated government is corrupt. In other words, by ‘business’ is unlikely to incorporate corruption arises from the institutions and adequately the interest of employees.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction mores of society itself, and prevailing eco- Thus, rather than ruminating over the nomic conditions. ‘business climate’, those dedicated to

Chapter 1: Introduction

32 33 pro-poor development should encourage It is the view of this report that successful the Moldovan government to design poli- achievement of the goals of the EGPRS cies of practical compromise that simul- requires three necessary developments taneously serve the interests of private in the economy: narrowing the trade gap, employers, their employees, and those raising aggregate investment, and redu- that consume the commodities that the cing inequality. employers and employees produce. Such a compromise requires each party to be The framework in which those three sensitive to the interests of the others, and developments might be realised can be would be implemented through institu- summarised as follows: tions that develop over time. a. At the beginning of the twenty-fi rst century, Moldova was a predominantly Directly relevant for poverty reduction is agricultural economy because of the the fourth transition, from social delivery collapse of its manufacturing sector due of basic needs to market provision. With to the ravages of the transition and the regard to employment, and the incomes dispute with Transnistria; however, an derived from it, the fi rst decade of the agriculture-led growth strategy is not the transition to a market economy was a route to sustainable poverty reduction. dismal, if not catastrophic failure for the overwhelming majority of Moldovans. b. While the agricultural sector must Indeed, it is diffi cult to fi nd a modern provide the exports to close the trade example of such an atrocious economic gap and employment opportunities collapse in peace time under any eco- to reduce unemployment in the short nomic, political or social system, except and medium term, a viable and sustain- from another transitional country. Equally able long term growth strategy must be catastrophic was the collapse of the social based on the emergence of internatio- sectors, education, health, and provision nally competitive manufacturing sector. for the elderly. Reasonable people can c. Rising productivity in agriculture and have diff ering views on progress in coun- manufacturing will require a purpose- try building, institutionalising democracy, ful public investment programme that and creating a functioning market society. ‘crowds in’ private investment, both do- On the provision of basic necessities the mestic and foreign. judgement is in: the transition was a disaster that challenges the imagination. d. Within this long-term framework, expansion and in some cases reorganisation of existing poverty reduction programmes, including the social protection system, is 1.4 Moldova’s Poverty essential, based on the principle of universal provision rather than means-testing. Reduction Strategy Some, perhaps many, would disagree with our judgement that Moldova’s long The Framework for this Report term development strategy should be based on industrialisation. The argument Through its application to Moldova of the is commonly encountered that Moldova experience gained from other studies, is blessed with a mild climate and particu- this report supports and complements larly fertile soil, and, in any case, has de- the long term strategy of the Moldovan monstrated agriculture to be its so-called government as presented in the Eco- comparative advantage throughout the nomic and Growth and Poverty Reduc- twentieth century. None of these points tion Strategy document (EGPRS), which

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction makes a compelling case for an agricul- was completed towards the end of 2004. tural, or even an agro-industry, based

Chapter 1: Introduction

34 35 growth strategy for the long run. Whether by using public investment to ‘crowd-in’ its climate is relatively favourable is a sub- private investment, as well as creating jective judgement, and there are many projects that directly reduce poverty. European countries with fertile soil that, Along with narrowing the trade gap and none-the-less, do not have predominant- increasing investment, poverty reduction ly agricultural economies. requires that, at a minimum, income and wealth inequality not increase. Each rise There is some superfi cial validity to the ar- in inequality, however measured, reduces gument that Moldova has demonstrated the poverty reducing eff ect of growth. an advantage, comparative or absolute, Fifteen years after independence, Moldo- in agriculture and agro-industry. Up to va’s income and wealth distribution were the Second World War, Moldova was over- among the most unequal among transi- whelmingly rural and agricultural, as were tion countries. Poverty reduction requires almost all of the Central and Eastern Euro- that the trend towards greater inequality pean countries. This represented a phase stop. of underdevelopment, not an outcome based on natural or economic advanta ges. After that war, Moldova became a part of The EGPRS and Donor the Soviet Union, without even nominal & Lender Response sovereignty in economic decision ma king. To each part of the Soviet Union was as- As is generally recognised, poverty reduc- signed a production role, and Moldova’s tion strategy papers (PRSPs) are written was agricultural, with the exception of the by governments to access funding, in Transnistria region. No one would argue particular from the World Bank and the that these production assignments were International Monetary Fund. The former solely or even primarily based on either cannot initiate a substantial programme economic or natural endowment criteria. without receiving and reviewing this In 1991 when Moldova became independ- document. Most of IMF lending is not ent, it was one of the most rural and agri- dependent on a PRSP, but a government’s cultural of the ex-Soviet republics because access to the concessional-terms Poverty it had been assigned those characteristics Reduction and Growth Facility (PRGF) is. by decisions into which its people had lit- In November 2004, the Economic Growth tle input, and which were made with little and Poverty Reduction Strategy paper of concern given to economic effi ciency. the Moldovan government was reviewed by the Executive Boards of the World Bank If in the medium term agricultural sector and the IMF. The World Bank issued a posi- can narrow the trade gap, a major step tive assessment of the document, but as- towards sustainable growth would be serted that ‘slow progress in implement- taken. Simultaneously, the rate of investing structural reforms…limit the scope for ment should rise. Trusting to a ‘favourable IDA assistance’.7 The IMF assessment was business climate’ is unlikely to stimulate similar, ‘over time, the EGPRSP could serve investment to the level necessary for as a basis for concessional lending by the sustained poverty reducing growth. For- Fund, provided the authorities demon- tunately, this problem is feasibly solved strate a genuine commitment to market

7 There was some ambiguity in the World Bank commentary on the document: The World Bank’s Board of Executive Directors today discussed Moldova’s Poverty Reduction Strategy Paper (EGPRSP)…and the Joint IDA[1]-IMF[2] Staff Advisory Note…The Bank’s Board supported the three-pillar EGPRSP…The Board recognized that the proposed strategy, underpinned by a sound diagnostic of the poverty situation and transition experience, is comprehensive and can, over time, foster growth and reduce poverty. The Executive Directors noted with concern several critical areas where current policies do not match Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction those articulated in the EGPRSP…Directors noted that these [inconsistencies] limit the scope for IDA concessional assistance. (World Bank 2004b)

Chapter 1: Introduction

34 35 reforms and establish a credible track institutions to show a strong commit- record in its [sicsic! their] implementation’ ment to supporting the strategy. It would (IMF 2004a).8 An unusual, if not unique, appear that the functional status of the aspect of the mandated joint review by EGPRSP at the end of 2004 was to serve as the World Bank and the IMF, was the ab- a government-generated list of perform- sence of a fi rm commitment to conces- ance conditionalities, in some cases in ad- sional lending. This joint review, called dition to ones specifi ed in other IMF and an ‘Advisory Note’ rather than the usual World Bank agreements. It is unlikely that ‘Assessment’ represents one of only four the government produced the document cases in which the two institutions had for this purpose, and the Advisory Note made a positive review without a com- conforms to assertions by PRSP critics that mitment to funding, a commitment that the documents represent a form of lender heretofore was the sine qua non of PRSP conditionality. joint staff assessments. Moldova is the only one of the four in which concession- At the end of 2004, the reasons for the ary funding is explicitly denied.9 reluctance of the Bretton Woods institutions to support the EGPRSP were not The ‘Joint Staff Advisory Note’ leaves the clear. With regard to the IMF, there would EGPRSP in limbo. On the one hand, it was seem to have been no substantial issue judged to be ‘comprehensive and [to] of disagreement within its ‘core area contain critical policy actions that…could of expertise, for monetary, fi scal and help foster growth and reduce poverty’, exchange rate policy all conformed to ‘has substantial merit’ in its treatment conditio nalities.10 While IMF-World Bank of recent economic developments, pov- guidelines require ‘adding new conditio- erty trends, improvement of the business nality during a review in areas previously environment, and strengthening gover- without any conditionality would require nance (IMF & IDA 2004, p. 9). It was also a clear justifi cation as to why the related judged that it ‘aligns the proposed prio- issues had become critical to program ob- rity programs to the MTEF [Medium Term jective’ (World Bank 2002). In early 2005, Expenditure Framework], and is based on the World Bank seemed to overcome extensive consultation with a broad range its reservations, though its position on of stakeholders’ (Ibid.)..). OOnn tthehe ootherther hand,hand, budget support was still not entirely clear. these laudatory characteristics were not The IMF position remained unchanged, suffi cient to induce the Bretton Woods which was considerably more important

8 In February 2004 document, the IMF listed the following ‘disturbing’ policies on the part of the Moldovan government: 1) ‘in agriculture, recollectivization eff orts are reportedly underway’; 2) ‘formerly privatized companies have been renationalized’; 3) ‘formal and informal restrictions have been imposed on exports of some agricultural goods’; 4) ‘the May 2003 territorial-administrative reorganization…reduced local autonomy’; 5) ‘the Law on Veterans reintroduced a number of privileges and subsidies’; and 6) commitments to the WTO to liberalize telecommunications were reversed (IMF 2004, p. 50). These issues would not seem to fall into the ‘core expertise’ of the IMF (see footnote 4). In the source there is no discussion of their importance to the successful implementation of IMF programmes, though such an explanation is mandated. 9 The other countries were Bhutan, Laos and Niger. 10 ‘In PGRF’s, specifi c [conditionalities] should focus on the Fund’s core area of expertise: monetary, fi scal and exchange rate policies; the institutional arrangements underlying these policies; and structural aspects closely related to them, including those in areas of overlapping responsibilities, such as fi nancial sector, tax administration and governance. In general, conditionality should not extend beyond these areas, except for measures deemed critical for maintaining macroeconomic stability’ (IMF 2003). Through interviews and review of relevant documents, the Mission Team identifi ed the following issues of contention: 1) pre-shipment inspection, alleged breach of the ceiling for public energy-related debt; 3) the government’s introduction of scrap-metal export licensing; and 4) the government’s decision to Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction merge the energy and telecom regulatory agencies. These can be compared to the defi nition of ‘core area of expertise’ in the previous paragraph.

Chapter 1: Introduction

36 37 to the government, since access to Paris In the conditions they set, donors and Club debt reduction depended upon lenders must distinguish between actions reaching an agreement with the Fund. required by the recipient government in A long term view of the performance of order that the funds be used eff ectively, Moldovan governments with respect to and those policies that the donors and relations with the IMF would suggest that lenders have the bargaining power to im- the Fund would favourably review the pose, be they crucial to success or not. An EGPRS not withstanding past and present essential characteristic of the ‘donorship tensions with the government.11 Table 1, regime’, which prevailed at the time of the taken unchanged from an IMF document independence of Moldova, was an omnis- (except for the percentages), shows that cient external judgementalism. The donor for ‘performance criteria’,12 government or lender reserved the right to pass unilat- compliance after 2000 with Fund condi- eral judgement on the appro priateness tionalities was considerably better than of and commitment to recipient gover- or as good as the previous seven years. nment policies. In a national ownership Measured by ‘prior actions’13 and ‘structu- regime, assessment of policies passes to ral benchmarks’,14 performance was quite the recipient government, with consulta- similar throughout the ten years. tion with the donor. National ownership does not require external development With regard to bilateral donors, the Swe- agencies to suspend all judgements; ra- dish International Development Agency ther, it implies that those judgements has been the most consistent supporter arise out of an interactive process with of Moldova’s development, in level of national stakeholders. funding and commitment to partner- ship. Its approach in Moldova represents A central characteristic of donorship was a model that could be usefully followed the presumption that if development as- by other bilateral agencies, refl ecting sistance failed in their goals, the blame the agency’s commitment to recipient lay with the recipient government. This ownership of the national development approach can be seen in a document agenda (see Weeks, et. al., 2002). The UK from the Overseas Development Institute, Department of International Develop- which repeats a view commonly found ment has also stressed the importance in donor documents: ‘[recipient] govern- of ownership, though its programme is ments need to be convinced of the need considerably smaller. In the UN system, for sound policies, rather than coerced’ the UNDP is the appropriate organisa- (Foster 2000, 7). This type of comment is tion to give leadership to the ownership often accompanied by invoking the need agenda, and it is the intention of this for the recipient government to show report to provide further content to that ‘political will’. While these comments leadership towards ownership and away may seem bland and non-controversial, from donorship. they are in the donorship tradition. Not

11 The tensions at the time of this mission were not new. The Moldova Economic Trends April 1999 refers to the ‘freezing of relations with IMF and WB [World Bank]’ (page 4). A representative of a bilateral development agency commented in an interview that ‘the policy errors by the World Bank and the IMF [in Moldova] were suffi ciently serious and well-remembered that it is not surprising that the government does not eagerly take the advice of the IFIs’. 12 ‘A performance criterion is a variable or measure whose observance or implementation is established as a formal condition for the making of purchases or disbursements under a Fund arrangement’ (IMF 2003). 13 ‘A member may be expected to adopt measures prior to the Fund’s approval of an arrangement…when it is critical for the successful implementation of the program that such actions be taken’ (IMF 2003).

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 14 ‘Structural benchmarks are similar to structural [performance critieria], except that individual benchmarks are less critical for meeting the program’s objective’ (IMF 2003).

Chapter 1: Introduction

36 37 withstanding the laudable phase, ‘rather absence of concrete specifi cation of the than coerced’, the fi rst comment makes political context in which a government a number of presumptions inconsistent operates, the term ‘political will’ has lim- with a dialogue based on ownership: ited meaning. ‘Political will’ and ‘political commitment’ obscure understanding 1. it presumes that there exists a set of of policy constraints in a country, rather sound policies which the donors know than facilitate it. Were it possible to defi ne and recipient governments do not (the the concepts rigorously, it is doubtful that recipient is ignorant); one could specify a way to judge the de- 2. that recipient governments must not gree of will and commitment independ- only be informed of the sound policies of ently of the outcome they are alleged to which they are ignorant, but require con- determine. More basic, the terms ‘will’ vincing of the need to implement these and ‘commitment’ fi t well into the donor- (in the absence of donor advocacy, the ship mode, precisely because they are recipient lacks the judgement to distin- black boxes and tautological. An external guish good policies from bad ones); and, agency could explain project or programme failure by ‘lack of political will’, 3. in the past development failures arose with no fear of being refuted. That this from the mistakes of omission or com- approach is increasingly unacceptable mission of recipient governments, not in in the twenty-fi rst century, which is to whole or part the result of unsound poli- Moldova’s benefi t. Following PRSP guide- cies of the donors (development failures lines, the development strategy of the are recipient government failures). country should be established through the democratic process, which has func- These presumptions tended to guide tioned relatively well in Moldova since donors and lenders in their relations with independence. Moldova in the 1990s. A variation on the sound policies criticism of recipient go- It is the view of this mission that while vernments is that they may be aware of relations between the government and the policies, and aware of the need for some donors and lenders have been them, but fail to implement them because strained, there exists scope for construc- of special interests within or outside of tive dialogue. This dialogue requires the government. In such circumstances, the active endorsement by donors the argument goes, donors are justifi ed and lenders of the principle of national in their criticism of policy choices, and the ownership of development strategy, criticism may strengthen domestic sup- facilitated in the case of Moldova be- porters of sound policies. This argument cause of the democratic legitimacy of is also in the tradition of donorship, for it the government. Putting the principle implicitly suggests that institutional, po- of national ownership into practice, an litical, and economic interests do not mo- essential feature of the PRS process, in- tivate donors. The ownership equivalent volves acceptance of a further principle, of the ‘sound policies’ statement would that economic policies and reforms are be: ‘donors and recipient governments not purely technical matters, but issues should engage in dialogue to identify involving trade-off s and political choice. sound policies on the part of each, and In other words, if donors and lenders each should be convinced of the need take national ownership seriously, it to implement these’, and this approach implies that in some cases they must ad- would be productive of good policy in just their views to those of the recipient Moldova. government. In the Moldovan context, two more Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction terms, ‘political will’ and ‘political commitment’, require deconstruction. In the

Chapter 1: Introduction

38 39 Table 1: Moldova and IMF Conditionalities, 1993-2003

Performance Criteria Met: Loan type Period Yes No % Met SBA 12/93-3/95 32 10 76 SBA 3/95-3/96 29 5 85 EFF 5/96-5/00 72 47 61 PRGF 12/00-12/03 47 9 84

No. Prior Actions & Structural Benchmarks Met: Loan type Period On time Late Not met % Met SBA 12/93-3/95 11 - 3 79 SBA 3/95-3/96 9 - 4 69 72 EFF 5/96-5/00 51 3 17 [76] 70 PRGF 12/00-12/03 14 4 2 [90]

Source: IMF 2004. NOTES: The loan types are ‘Stand-by Agreement’ (SBA), ‘Extended Fund Facility’ (EFF), and ‘Poverty Reduction and Growth Facility’ (PRGF). The fi rst two categories are, respectively, ‘met on time or with some delay’, and ‘and ‘met with considerable delay’. The fi rst percentage in the last column of the second part of the table includes only the former category, and the number in brackets includes the latter.

Annex: sitional economies, the fi rst step is to clarify terms. It is unfortunate that in the Common Analytical literature on transitional societies, terms Mistakes about Tran- which previously had straight-forward, dictionary-based meanings, have to come sition Societies15 to assume heavy ideological implications. One of the most important of these is ‘reform’. Concepts and Confusion Strictly speaking, a ‘reform’ is a policy Considering growth and poverty in change of any type. For example, a trade Moldova requires prior discussion of regime can be reformed through tariff s or common analytical fallacies about tran- quotas to alter the level and composition

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 15 The points made in this annex have been developed further in a UNDP report on poverty reduction policies in Vietnam (Weeks, et. al., 2004).

Chapter 1: Introduction

38 39 of imports; similarly a trade regime can be public sector and numerous regulatory in- reformed by cutting tariff s and elimina- terventions. The two types of economies ting quotas. In the discourse of current are not on the same analytical spectrum. economics, ‘reform’ always mean actions The purpose and outcome of centrally of the former type, and is by defi nition a planned economies was to eliminate ‘good’ thing. Those in favour of the latter private ownership and, by doing so, to are described are called ‘reformers’, and eliminate the market as a regulating force viewed as progressive (typically judged in society to the extent possible. In such to refl ect the ‘national interest’), while economies, prices were not ‘distorted’, those in favour of the former are branded because they had no allocative function. as ‘anti-reform’ and reactionary (typi- They served merely as accounting tools. cally judged as motivated by ‘special inte- The purpose and outcome of regulated rests’). These value-laden terms are mis- market economies was and is to manage leading, because, for example, at times it and control private markets. The eff ect of may be rational to favour tariff reduction, regulatory interventions can accurately and other times rational to oppose it. This be described as ‘distortionary’ only if one study avoids use of ‘reform’, in favour of accepts neoclassical general equilibrium specifi c terms: tariff reduction, enterprise welfare economics.16 restructuring, fi nancial sector reorganisation, etc. The qualitative distinction between centrally planned and regulated market Other ideologically-abused terms are economies is essential to understanding ‘open’ and ‘closed’ economies, ‘controlled’ transitional societies. Transitional socie- and ‘liberalised’ economies, and perhaps ties pass through two phases of policy most value laden of all, ‘market-friendly’. change before a market based economy To take but one example, the Moldovan is institutionalised: fi rst, there is the radi- economy is extraordinarily open if one cal shift from central planning (the term defi nes openness by the ratio of trade ‘controlled’ economy strictly applies) to to gross domestic product. However, market based (a ‘regulated’ economy); some spokespeople for external agencies and, second, public policy then addresses judge it as relatively closed because of its the question of what degree of public non-tariff barriers to trade. Attempting to intervention is required for eff ective func- encapsulate a country’s trade position as tioning of the new, market economy. We ‘open’ or ‘closed’ is an essentially ideo- refer to these two phases as the regime logical exercise that provokes controversy change transition and the regulatory tran- without providing insight. Controversy sition. with insight can be achieved by debating specifi c policy measures without assign- The fundamental policy diff erence can ing them value-laden labels. be illustrated by referring to public sector enterprises (SOEs), of which a number re- Perhaps the most fundamental ana- main in Moldova. Within central planning, lytical mistake in analysing transitional enterprises (the modifi er ‘state’ is redun- economies is to treat centrally planned dant) are part of an integrated system of economies as if they were more extreme non-market production, in which market versions of market economies with a large demand, even household demand, is

16 The term ‘distortion’ has meaning only in relation when one specifi es a non-distorted outcome. To take an analogy, one can say that mirrors in carnivals distort one’s refl ection because a true image can be defi ned scientifi cally. Similarly, a price is said to be distorted if it varies form the price which would prevail under conditions of full employment general equilibrium. However, if that equilibrium is not unique, or if one does not accept the analytical framework that generates it, then the term ‘distortion’ loses objective Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction meaning. Rejecting the term ‘distortion’ does not stop one from describing prices and allocation to be dysfunctional, but the judgement is empirical and specifi c, not general and theoretical.

Chapter 1: Introduction

40 41 largely irrelevant. In a regulated market its intended outcome because the instru- economy, state enterprises (now the ad- ments available to the government have jective becomes necessary) are part of a limited impact on private decisions. market system. The central government may grant SOEs monopoly positions, pro- Further, there is the fundamental issue of tect them through trade policy, and ex- the stochastic element in policy. Assume tend them subsidies, but such enterprises that the government seeks to achieve an remain market-based. infl ation rate not higher than fi ve percent. It sets monetary policy such that the mean The mistake of not distinguishing be- outcome will be two percent. If the stan- tween regime change and regulatory dard deviation associated with the policy change gives rise to a closely related one, outcome is three percentage points, then which is concluding that if regime change about one-third of the time the outcome stimulated growth, then a rapid pro- will be ‘bad’, even though, technically, the gramme of deregulation will bring higher policy was ‘good’. When this report makes growth. When the system of central plan- its recommendations, it should not assign ning collapsed in Europe and Central Asia, to the government more infl uence over these systems were defi ned as ineffi cient the economy that its policy instruments, in all aspects, and all problems during the exogenous environment, and effi ciency central planning period were the result of markets allow it. of central planning ineffi ciencies. It was a short step to conclude that all subsequent problems of transitional countries Inequality in Transitional result from the lingering eff ects of central Societies planning. The dominant tendency was to presume that post-central planning prob- As in all policy analysis, the goal sought lems would not arise from inappropriate by this study is the enhancement of hu- changes in regulatory policies (‘reforms’), man welfare. The most important out- or from changes made too hastily or in the comes aff ecting this welfare are levels of wrong order. This approach ignores the household income, social provision, and distinction between regime change and inequality. Growth, trade, and structural regulatory change. Market deregulation change are instruments to achieve those is not a continuation of the dismantling outcomes. In the context of human wel- of central planning. Each deregulation fare, per capita income provides limited policy must be inspected empirically for information, especially since the more its economy-wide eff ects. unequal the distribution of income, the less accurately it approximates a measure Related to the above, one must avoid the of the welfare of the majority of the popu- error of ‘policy optimism’, which is espe- lation.17 cially virulent in transitional countries. At many points in time the growth rate of a Our policy judgements and recommen- country may be only loosely related to its dations on the desirable level of inequa- policies, unless those policies are unusu- lity derive from the context of Moldova as ally dysfunctional. Many factors beyond a transitional economy. Centrally planned the control of the government infl uence economies had low levels of inequality, growth, such as the state of the interna- because prices and incomes were set ad- tional economy, weather, and the effi - ministratively to generate that outcome. ciency of national markets. With regard to Major items of household subsistence the latter, ‘good policy’ may fail to achieve were provided outside mechanisms of ex-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 17 The more skewed the distribution of income, the greater the diff erence between mean and median income, by defi nition.

Chapter 1: Introduction

40 41 change, notably housing, education, and a major impact on all other social policies health care.18 Further, employment was and choices; and, second, the temporary guaranteed for all, though much of it may and transitional nature of distribution not have been productive. By contrast, makes it amenable to public policy. exchange based on private ownership of assets is the fundamental distribution Since the institutions of society determine mechanism of market societies, even ones the income distribution in any country, with a large public sector. one can treat the alternative growth paths as converging to a ‘steady-state’ Therefore, one should expect asset and degree of inequality. The key question for the associated income inequality to in- social policy, and for this study, is which crease during the transition from central path of inequality and growth is implied planning to a market society. Part of this by the institutions of Moldovan society. If increase refl ects the inherent inequalities those institutions would result in a steady- of market societies, and is essential to state degree of inequality that is contrary their eff ective functioning. Another part to social consensus, then the policy issue arises from an emerging diff erential ac- is how they might be changed. cess to assets, services, information, political power, and, it should be added in light For policy purposes, it is important to dis- of corporate behaviour in the developed tinguish between poverty and inequality, countries, fraud and theft. The increased because diff erent, but complementary, inequality associated with the transition policies are needed to reduce both. Pov- from central planning to the market can erty reduction programmes that are be divided conceptually between that targeted at those below a given income which is functionally necessary and that level (or some other criterion) may be which is not. The increased inequality successful in reducing the number of which is not necessary for the operation poor households, but have little or not of a market system can itself be divided impact on the level or trend in inequality. between that which is judged by society Redu cing inequality requires, in addition to be acceptable and that which is not. For to raising the incomes of the poor, pre- example, it might be decided by explicit venting excessive concentration of wealth or implicit social consensus that inherited in the hands of a few, and fostering the wealth is acceptable, even though much growth of a middle class. of it might not be necessary on allocative grounds. Dynamics of Poverty in Among the many social choices unavoid- Transitional Societies ably presented to Moldovan society is the degree of inequality that would be ac- The distinction between the phases of ceptable. All societies face this choice. In transitional countries, regime change and established market societies the division regulatory change, provides insights into of political power associated with a coun- the dynamics of poverty in Moldova. Du- try’s degree of inequality may restrict ring the central planning period, Moldova the debate over the issue. In transitional was a middle income country, with relative- countries, the situation is quite diff erent ly high indicators of human development. for two reasons: fi rst, asset and income The poverty in that period was not of the distribution undergo rapid change, with type that a market economy manifests.

18 That the level of income inequality in socialist economies was quite low is not controversial. Relative magnitudes are indicated by calculations that show that the Gini coeffi cients in the early 1990s for twelve European and Central Asia countries in transition were all lower that for thirty-eight developing Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction market countries, with the exception of only two, Sri Lanka and Rwanda (Dagdeviren, van der Hoeven & Weeks 2002, 406-407).

Chapter 1: Introduction

42 43 Market-based economic growth gener- On the positive side, growth reduces ates poverty, especially in transitional basic poverty by raising incomes in exist- countries, as was demonstrated clearly ing economic activities, and creates new in Moldova. Unlike in low-income mar- activities into which people are drawn. ket societies, prior to the regime change On the negative side, inherent in the Moldova had little or no poverty result- creation of new activities are allocative ing from landlessness or unemployment. changes that ge nerate unemployment Fundamental to the transformation of the and landlessness. It was unfortunately the economy to market-regulated was a real- case that in place of growth, the emer ging location of assets on the basis of private market economy of Moldova suff ered ownership. On the one hand, this cre- a devastating collapse. The rapid rise in ated new mechanisms for private income poverty during the 1990s was the result of generation; on the other it produced a this economic collapse combined with the phenomenon new for Moldova, po verty loss of the entitlements associated with a resulting from loss or lack of access to planned economy. Whether poverty can productive assets. Gainful employment be reduced at a sustained rate depends on became dependent upon private owner- the balance between market-based em- ship of land; wage employment depend- ployment creation and market-generated ent upon those who owned productive landlessness and unemployment. Increa- assets; and self-employment on access to singly, the relationship between poverty credit. The poverty resulting from unem- reduction and growth will refl ect the ba- ployment and landlessness can accurately lance between employment growth and be called ‘market generated’ poverty. poverty-generating structural change. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 1: Introduction

42 43 C h a p t e r 2 : Economics of the Transition

2.1 Moldova in Compara- When such extreme economic declines occur, precise measurement is impossible. tive Context The declines for most of the countries in the table were associated with the des- The problems of the Moldovan economy truction of entire industries, collapse of in the early years of the twenty-fi rst centu- social service provision, massive changes ry cannot be understood without looking in relative prices, and hyper-infl ation. back to the last decade of the twentieth These factors make it diffi cult if not im- century, and to the other countries ma king possible to fi nd a set of price and quantity the transition from central planning to weights to calculate changes in real in- market regulation. Like Moldova, se veral comes. Further, even were accurate calcu- of the countries made the independence lation possible, it is not clear what ‘less’ or transition simultaneously with the change ‘more’ severe decline means, if anything, of economic and social system.19 when one society’s production falls by half or more. Sixteen of the twenty-three European and Central Asian transition countries suf- With these warnings made, a cautious fered declines of output of fi fty percent or interpretation of Table 2.1 provides some more. The seven countries with declines insights into the experience of Moldova less than thirty percent were in Central (also see Figure 2.1, which shows the same or Eastern Europe (Slovak Republic, Bul- statistics graphically by country catego- garia, Romania, Hungary, Czech Republic, ries). First, most catastrophically devas- Slovenia, and Poland, in order of severity), tated were the republics of the former So- while eleven of the fourteen members of viet Union, which make up the fi rst twelve the former Soviet Union had declines of countries in this not-to-be-envied league fi fty percent or more (the exceptions be- table. The eleventh country is Russia itself, ing Belarus, Uzbekistan and Estonia, all followed by fi ve more ex-Soviet Repub- in the minus forty percent range). These lics, with only confl ict-aff ected Croatia economic and social disasters are summa- intruding into this group of post-Soviet rised in Table 2.1, in which the countries states at number thirteen.21 Of the fi fteen are ordered by the size of decline from ex-republics, only relatively ‘unreformed’ 1990 to the lowest point of each.20 Uzbekistan is closer to the Central Euro-

19 In the tables and charts in this chapter that compare across countries, EBRD statistics are used through 2002 or 2003. A country-by-country up-date for 2003 and 2004 would run the risk of non-comparabi lity. When national data from Moldova are used, the tables and charts include 2003 and, in some cases, into 2004. Diff erences appear between EBRD and Moldova national statistics, arising in some cases from defi nitions used. 20 The tables and charts with cross country comparisons have diff erent terminal years because of problems of statistical consistency. Statistics through 2004 and into 2005 for Moldova are presented later in the chapter. 21 That the Slovak Republic, another newly independent country, is next after Uzbekistan further empha- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction sises the diffi culty of combining social and economic transition with construction of institutions appropriate to a country.

Chapter 2: Economics of the Transition

44 45 pean countries in the size of its decline Moldova’s absolutely and relatively weak than to the cohort of newly independent recovery is shown graphically in Figure ex-Soviet states. The newly independent 2.2, where the other twenty-two countries country that declined least was Slovenia, are divided into analytical groups and the formerly a Yugoslavian republic. This is diff erence between their average growth the exception that proves the rule, in rates and Moldova’s is charted. For over that it quickly achieved a relatively close ten years, from 1991 to 2002, there is not integration with the much larger Italian a positive value; that is, Moldova’s growth economy. Second, if there is any correla- rate was less than each of the group aver- tion between declines and the extent to ages in every year. which national economies liberalised, it is overwhelmed by other eff ects. The extent of Moldova’s decline and, equally important, the weak recovery Third, and most relevant to this report, from that decline, have been underscored statistics show that Moldova suff ered the in order emphasise the enormity of the third worst decline of the twenty-three developmental task facing any Moldovan countries and had the smallest recovery. government. Were the economy to In 2001 its per capita income was mea- sustain the 2000-2003 average of fi ve sured to be only nine percent above its percent, it would take over twenty-fi ve lowest value, in 1999.22 Only one country years to regain the level of GDP and per came close to such a small improvement, capita income of 1991.23 Due to the se- Uzbekistan, whose decline was less than vere worsening of income distribution, it one-third of Moldova’s. The extent of would take much longer for the bottom Moldova’s decline and the weakness of quintiles to regain their standard of living its recovery are shown in detail in Table in 1991, perhaps fi fty years were income 2.2. In this table, percentage changes in distribution not to worsen from its degree gross national product are accumulated in of inequality in the early 2000s. each successive column. For each country, the year in which the cumulated change While it is obvious that falls in per capita turned positive is highlighted in bold, and income drive households into poverty, recovery to the level of 1991 set in a bor- less obvious is the extent to which in- dered cell. Sixteen of the other twenty- creases in poverty are magnifi ed by seven countries showed a positive growth increased inequality. Table 2.3 seeks to rate during 1991-1996. Moldova did not quantify this eff ect. Using statistics from have a positive growth rate until 1997, and a study by Milanovic of the World Bank, that weak increase of two percent turned the table reports estimates of headcount sharply negative during 1998-1999, pri- poverty and the Gini measure of inequal- marily due to the Russian fi nancial crisis. ity for transition countries. A regression It was not until 2000 that recovery as such equation was estimated, using the abso- began, at which point twenty of the coun- lute change in the poverty index as the tries had begun their recoveries. Posi- dependent variable (reported in column tive growth over four years, 2000-2003, 3), and the absolute change in the Gini brought Moldova’s GDP to less than thirty coeffi cient (column 6) and the decline in percent of what it had been in 1991. Only per capita income (from Table 2.2, above) two other countries came remotely close as the independent variables (the statisti- to this dismal level of recovery, Georgia cal result is reported in the notes to Table and Ukraine (about fi fty and forty per- 2.3). The last column reports the result cent, respectively, of their 1991 levels). of using this cross country regression

22 Given problems of accurate measurement, this small increase lies within the margin of measurement

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction error; that is, one cannot reject the hypothesis that by 2001 there had been no recovery. 23 This assumes, as during 1991-2003, no population growth.

Chapter 2: Economics of the Transition

44 45 in which the percentage change in the resulted from redistribution from the low- poverty share is a function of the change est quintile to the highest (see Table 2.4). in the Gini coeffi cient and per capita in- Moldovan households in the lowest quin- come. The coeffi cients are used to calcu- tile had the smallest expenditure share of late the poverty impact of each variable. any of the fourteen Central and Eastern This exercise estimates that changes in European countries, with the exception distribution, measured by the Gini coef- of Russia. When eight Central Asian coun- fi cient, accounted for forty-fi ve percent tries are included, only in Uzbekistan and of the calculated change in poverty across Russia are the expenditure shares of the all the countries. For Moldova inequality lowest quintile signifi cantly lower. At the accounted for twenty-eight percent of the other end of the distribution, the richest increase in poverty. The inequality eff ect twenty percent in Moldova enjoyed the for Moldova is below the cross country fourth highest expenditure share among average because its fall in output was so the twenty-four countries. As well as the large, not because inequality rose less (it extreme changes at the ends of the dis- rose more than the average). tribution, the shares of the second and third deciles for Moldova were notably Lack of relevant data makes it impossible low, indicating that relative income losses to assess the extent to which the increase extended into the middle class. This grim in poverty was concentrated in specifi c conclusion points to the importance of groups and regions of the country. Central implementing policies in Moldova that to the analysis of poverty incidence would have a broad-based impact on growth be to investigate the extent to which and employment as well as directly on households headed by women fell into poverty. poverty. The limited statistics avai lable are presented in Chapter 4, but these are insuffi cient for the task. As suggested in Chapter 1, the Department of Statistics 2.2 Per Capita Income and Sociology should give priority to the and Development collection of data which would allow an analysis disaggregated between men and Assistance women. Up to this point, the catastrophic decline Tables 2.2 and 2.3 indicate the absolute of the economy and the rise in poverty and relative severity of Moldova’s decline: have been discussed without reference in none of the other countries for which to Moldova’s per capita income. This has there are comparable data did the meas- been possible because the relative mag- ure of poverty increase so much (sixty- nitude of the decline is generally agreed, fi ve percentage points, compared to fi fty while there has been controversy over for the next highest), or even rise within the level from which that decline began. fi fteen percentage points of Moldova’s.24 It may seem that in the context of such a It is possible that no country at peace catastrophic collapse, debating levels is has ever suff ered such a phenomenal in- largely irrelevant to the point of pe dantry. crease in poverty over such a short period But, quite to the contrary, identifying of time. the appropriate per capita income for Moldova for the early 1990s emerges as More than in any Central and Eastern Eu- an extremely practical issue aff ecting the ropean country and six of the eight Cen- country’s debt liabilities. Because Moldo- tral Asian countries, inequality in Moldova va was offi cially considered to be a middle

24 These poverty measures diff er from those analysed in Chapter 4. Those in Table 2.2 are probably a better Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction estimate, because the poverty line used by the poverty monitoring unit in the Ministry of the Economy is extremely limited in what it includes as basic needs. This is discussed in detail in Chapter 4.

Chapter 2: Economics of the Transition

46 47 income country in the early 1990s by the in 1991, almost exactly half the number World Bank, Bretton Woods lending was reported in the World Development Report not on concessional terms. The debt ser- 1993. As a result, Moldova had dropped vicing and fi scal implications of not being into the low-income category by 1993; a low income country in the early 1990s i.e., below US$800 income per capita. continue to burden the country. Among Thus, the World Bank’s own correction the international agencies, the respon- of the Moldovan statistics implies that sibility for defi ning the income status from 1993 onwards, its loans, and those of a country lies with the World Bank. A of the IMF, should have been on conces- low-income country is in eff ect a country sional terms. However, it was not until that is eligible to borrow from the World late in the de cade that loans were made Bank’s concessional lending branch, the on IDA terms. From 1993 into 1998, how- International Development Agency. In ever, Moldova borrowed almost US$ 250 the fi rst half of the 1990s, the per capita million at near commercial rates from the income ceiling for IDA loans was about World Bank alone (see the Annex to this US$ 800. Chapter, ‘Debt Chronology’).

The second data column of Table 2.5 The per capita income of Moldova, no shows the per capita income attributed later than 1993, placed the country in the to Moldova during 1991-1999 in the World World Bank’s ‘low income’ category, and Bank’s offi cial annual report on develop- this is now accepted by the Bank itself. ment, the World Development Report. InIn However, Moldova was not extended 1991 the World Bank reported a statistic that privilege by the Bank, the Fund or of US$ 2173, which was reported in sub- the European Bank for Reconstruction sequent reports to fall sharply over the and Development until 1997, after which next eight years to US$370. If one accepts new multilateral funding has been small. this reported time series as accurate, An obvious policy action follows: the then Moldova dropped into low income Moldovan government’s request that its status in 1997, at which time the World multilateral debt be restructured with Bank fi rst made loans to Moldova on IDA concessional terms should be accepted. terms. However, it is clear that the time Indeed, the use of a miscalculated per series cannot be correct. First, it implies capita income by the multilateral lenders a decline far greater than the generally justifi es reimbursing the excess interest accepted estimate. Second, beginning paid over more than a decade--perhaps in 1993, when national income estimates directly through debt reduction. Restruc- became, it diff ers from the much lower turing the multilateral debt would have a number of the IMF and the National Bank substantial impact on the commitment of of Moldova.25 Third, and most important, fi scal resources over time; reimbursing the the World Bank time series on per capita excess interest would have a dramatic im- income is inconsistent with the bank’s mediate eff ect. reported annual changes of Moldova’s income over the same years. There would be no basis for making this restructuring conditional upon policy One can presume that the 1999 fi gure changes, since the restructuring would for income per capita is correct, since it represent correcting a mistake made by was reported after the others, and use lenders, rather than motivated by the the percentage changes to calculate and potential insolvency of the borrower. The correct earlier years. The result is a per funds resulting from the restructuring capita income of slightly over US$ 1000 could be used to expand the pro-poor

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 25 Apparently it is highly unusual for the national income statistics used by the IMF and the World Bank to diff er. Moldova may be the only case in which a diff erence persisted for several years.

Chapter 2: Economics of the Transition

46 47 economic and social programmes that terms, the aid would have substantially are discussed later in this report. During reduced annual debt service payments. discussions with relevant multilateral representatives in Chisinau, two arguments The lower level of development assistance were made as to why this could not be to Moldova cannot be explained by the done forthwith. First, it was argued that failure of the various governments dur- it need not be done: since the Moldovan ing the 1990s to meet the conditionalities governments of the time willingly agreed of donors and lenders. On the contrary, to the non-concessional loans, the lenders Moldovan governments were judged at bear no responsibility for the mistake. It is the time as implementing a ‘model of not likely that this argument represents sound reforming…a perfect laboratory of the offi cial position of either the World reform’ (from the Economist inin 1995).1995). Bank or the IMF. If such an argument were valid, it would exclude debt relief in any While there were notable exceptions, form, and the institutions are committed most obviously Sida, during the 1990s the to such relief through several mecha- international development agencies did nisms, including the HIPC. Second, by not serve Moldova well,26 treating it as a charter, terms of reference and practice, middle income country for lending when the Bretton Woods institutions are not it was not. This resulted in less conces- commercial banks, but institutions that sional assistance and the accumulation of a act in the interest of the borrower, not heavy debt burden. It could be argued that their own self-interest. Therefore, it is the international agencies, like the Moldovans role of these institutions to make the bor- themselves, were faced with an unprec- rower aware of the most favourable terms edented crisis, resulting in mistakes of upon which it can contract loans (indeed, judgement made in good faith. Since those commercial banks have the same obliga- mistakes have been identifi ed as such by a tion by the laws in most developed coun- range of commentators foreign and do- tries). This implies that when a mistake is mestic,27 it is appropriate for development made by a Bretton Woods institution, it is agencies to redress their eff ects. the obligation of the lender to correct it in the interest of its client, the borrower.

The case for action on the lending mis- 2.3 Macroeconomic take is strengthened by considering more Performance and ge nerally the support by development agencies for Moldova during the early Policy, 1990-2004 years of the transition. Figure 2.4 shows that over the years 1992-1998 Moldova consistently received less development Policy Framework during assistance as a portion of GDP than the av- Moldova’s Transition erage for transition countries, and did not reach the average of other country groups The foregoing comparative discussion of until 1999. Had it received the average, in transition countries lays the basis for a some years it would have been able to close consideration of Moldova’s econom- cover its external current account defi cit. ic performance. The central issue is how to By substituting for loans on commercial explain the relatively poor performance of

26 ‘[Moldovans] were coached and guided by an increasing number of international experts, who appeared to have a road map and prescribed the course ahead…[T]he road map was at best inadequate… ’ (Ronnas & Orlova 2000, 15). 27 The mistakes of the international agencies on lending terms and advice on transition in general have Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction given rise to some degree of disillusionment among Moldovans: ‘Our common expectations that [the] market would arrange everything did not come true’ (CISR, Moldova in Transition July 2001, 4).

Chapter 2: Economics of the Transition

48 49 the country compared to other transition administrative trading system to market- economies, and draw out the implied les- based international commerce. sons for current policy. To review performance, Moldova’s transition is divided into This reorientation of trade required es- three periods, based on straight-forward tablishing the state institutions of inter- defi nitions using a three year moving national commerce, as well as facilitating average of the annual growth rate: 1990- investments in modernisation of produc- 1994 when the rate of decline worsened tion and marketing in the newly created in each successive year (the ‘collapse’); private sector in order to achieve com- 1995-2000, when growth remained nega- petitiveness, both in manufacturing and tive, but rose towards zero (‘stagnation’); agriculture. However, prior to establish- and 2001-2004, when growth rates were ing these conditions, the government of positive (‘recovery’). These three periods Moldova was encouraged to pursue a po- are shown in Figure 2.5, with both annual licy of general trade liberalisation, which values and three year averages charted. had the inevitable result of rendering domestic production non-competitive with Prior to 1990, Moldova was a small ap- imports, especially those from Western pendage of the Soviet production system; Europe. While a more gradual liberalisa- ‘a small unit of a huge enterprise’, as one tion of the trading regime would have still Moldovan economist has put it. When the resulted in eventual bankruptcies of both Russian economy entered into precipitous state and privatised enterprises, it would decline, it was to be expected that Moldova have provided the opportunity for some would follow it down. However, while the enterprises to adjust to the new pressures collapse of the Russian economy dictated of international commerce. a collapse in Moldova, it did not dictate that Moldova’s decline would be so much The shock of rapid trade liberalisation was greater than the average for all republics. magnifi ed by the exchange rate policy followed in the early years of the transition. Moldova’s extreme collapse in part fol- As discussed later in this chapter, after the lowed from a misdiagnosis of the coun- government created the Moldovan Lei in try’s characteristics at the outset of the 1993 the new currency was pegged to the transition, which resulted in a series of US dollar, an arrangement that was part of inappropriate policies derivative from this on-going IMF programmes.28 The combi- misdiagnosis. The mistake in calculating nation of dollar movements and domestic per capita income was part of a larger infl ation resulted in a continuous appre- mistake of presuming the country to be ciation of the Lei from early 1994 through more developed than it was, thus, more the fi rst quarter of 1998. This policy was capable of adjusting to the demands of maintained despite market signals of its a market economy than was the case. inappropriateness, as the current account Closely linked to this was the view that defi cit, after declining from 1993 to 1994, the Moldovan economy was ‘closed’, and rose from less than ten percent of GDP in would reap substantial static and dynamic early 1994 to twenty percent in early 1998. benefi ts from trade. In fact, the economy The over-valuation of the Lei was further was highly open since exports and im- indicated by a decline in exports over the ports represented a large portion of na- same period. It was to be expected that tional income. The fundamental problem an external shock would provoke a col- was not to open a closed economy, but to lapse of the exchange rate. The Russian re-orient the openness from the Soviet fi nancial crisis provided that shock, and

28 From the second quarter of 1995 through the second quarter of 1998, the average Lei-dollar rate was Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 4.61, with a quarterly coeffi cient of variation of slightly less than two percent. Over the same period, the consumer price index rose by fi fty percent.

Chapter 2: Economics of the Transition

48 49 from the second quarter of 1998 to the During 1900-1994, the economy was in second quarter of 1999 the Lei-dollar rate ‘free-fall’, the rate of decline increasing rose from 4.8 to 10.7. under the fi rst IMF programme, which began in 1993. Stabilisation and structural The instability generated by over-valua- adjustment loans, at IBRD terms through tion that brought on a currency collapse 1997, were not associated with positive was exacerbated by other closely associat- growth: decline continued to the end of ed policy miscalculations. As noted above, the decade. Belatedly, in 2001 growth the loans that the government obtained turned positive, after a decade of negative from multilateral and private lenders were rates except from the modest 1.6 percent at commercial or near-commercial rates,29 rate in 1997. and were used to cover recurrent public expenditure during a period when the In summary, one can conclude that the economy fell into precipitous decline. All late and slow recovery of the Moldovan economic analysis (and commonsense) economy was in part caused by the pur- tells one that loans should be contracted suit of unsustainable policies: premature only if there is a realistic prospect that they trade liberalisation that generated a large can be serviced. Servicing debts requires defi cit on goods and services, an over-va- either that 1) the loan involved directly or lued exchange rate that exacerbated that indirectly creates assets which will gener- defi cit, and an unsound fi scal strategy ate a fl ow of income at least suffi cient to that resulted in a contractionary dead- cover debt service, or 2) the borrower an- weight of debt servicing that continues to ticipates a growth of exports that will ser- the present. vice the debt. Borrowing to fund current expenditure violates the fi rst condition, The recovery and growth of the economy and borrowing when exports are falling after 2000 was fostered primarily by the violates the second. Given that even the infl ow of remittances, discussed later in IMF’s optimistic growth targets for the this report. The large infl ows fostered 1990s were below the rate of interest at growth by relieving the import constraint. which the Moldovan government was In addition, remittances reduced the go- contracting debt (and the target growth vernment’s dependence on conditiona- rates were far above outcomes), the debt lity-based borrowing, thus potentially accumulation policy was unsound, both allowing for fl exibility in macro policy. in the short and long run. It is diffi cult to However, inherent in remittance-driven regard the debt accumulation policy as growth are several possible problems: anything other than unsound, both in the 1) deriva tive as they are from the immi- short and long run. It was unsound in the gration policies of the destination coun- short run because the debts were accumu- tries, the level of remittances may not lated contrary to responsible fi scal policy, be sustainable; 2) there are no eff ective and unsound in the long run because of short-term policy instruments to infl u- the unsustainable fi scal burden the com- ence remittances, causing the tempo of mercial interest rates perpetuated.30 growth to outrun the management of

29 In 1992 the European Union lent Moldova US$ 31.6 million at LIBOR plus three percentage points, and the IBRD in 1993 lent US$ 86 million a variable rate of 4.85 to 7.8 percent. In 1994, the EBRD extended two loans totalling almost thirty million dollars at the LIBOR plus one percentage point. Debt on commercial terms from all lenders continued to accumulate in 1995 and 1996, to US$ 41 and US $300 million, respectively. See the annex to this chapter showing a chronology of debt accumulation. 30 The exception to these rules was the debts incurred by the government for the import of energy. While these debts to Russia were, strictly speaking, unsound in economic terms, the government had no choice, since the country was almost wholly dependent on imported energy. The dilemma facing the Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction government was either to contract onerous debt, or have the economy and society suff er even more. The same argument does not apply to multilateral debt.

Chapter 2: Economics of the Transition

50 51 the government; and 3) large scale emi- as in Vietnam and China, was not pursued. gration may negatively aff ect the quality The manufacturing sector was also nega- of the domestic labour force. Employing tively aff ected by the de facto separation fi scal, monetary and exchange rate policy of the Transnistria region, which had been to foster faster growth, with an emphasis considerably more export oriented than on public investment, would complement the rest of the country. the expansionary eff ect of remittances, and create a strategy for a transition to a Table 2.6 shows the fi scal consequences development framework more securely of the economic collapse across transi- based on domestic economic potential. tion countries, and Figure 2.5 compares Moldova to the Newly Independent States (NIS) and Central and Eastern Euro- Fiscal Policy pean (C&EE) country groups. From a position of no defi cits in the late 1980s, the Moldova’s fi scal performance since inde- cross-country average rose to over eleven pendence must be placed in the context percent of GDP during 1992-1993. While of the transition from central planning to greater than the average, the Moldovan market regulation. Centrally planned eco- defi cit was far from the worst, a grim dis- nomic systems derived their government tinction that goes to its fellow ex-republic, revenue by taking a share of the opera ting Armenia. Previous discussions stressed surplus of state enterprises, which in most the distinction between Central European of the countries accounted for virtually and Baltic countries, on the one hand, all of production and distribution. When and the former republics on the other. the transition process wiped out the sur- The diff erence between the two groups pluses of most enterprises, which was the is striking in Table 2.6 (see also Figure 2.5). case in all these countries except Vietnam During the years of exploding defi cits, not and China, massive government defi cits one country in the former group had an resulted. average defi cit of ten percent or more, while only three Soviet ex-republics, Az- In general, the more closely integrated the erbaijan, Kazakhstan and Turkmenistan, country into the Soviet trading system, did not. Comecon, the more severe was the impact on its production and, therefore, govern- Especially for the former Soviet republics, a ment revenue. Since this integration was coherent fi scal policy, in the sense of using qualitative as well as quantitative, it is dif- taxation and expenditure instruments to fi cult to assess its degree across countries. manage the economy, was impossible in However, there can be no doubt that for the early 1990s. These were years during over forty years the Moldovan economy which fi scal systems were in the process essentially operated as a cog in the ma- of construction. Until tax and expenditure chine of the Soviet economy. Manufactur- administration was established, purpose- ing was disastrously aff ected during the ful implementation of policy could not transition because much of its produc- be done.32 Thus, it makes little economic tion provided direct inputs to Russian sense to refer to fi scal policy as ‘loose’ factories.31 The possibility that Moldovan in the 1990s since the means by which factories could be fi nancially supported policy could be implemented were weak. to convert to new markets and products, By the mid-1990s, defi cits had been sub-

31 Ronnas and Orlova write, ‘The plants were invariably designed with the Soviet rather than the Moldovan market in view and many were entirely dependent on their supply of inputs from outside Moldova as well as markets for their output in other republics’ (Ronnas & Orlova 2000, 28). Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 32 ‘The Republic of Moldova is a new state of Europe, with unconsolidated laws, institutions, and procedures of political, economic and public life’ (CISR, Moldova Economic Trends July 1998, 5).

Chapter 2: Economics of the Transition

50 51 stantially reduced in all former republics, every transition government presided except for those whose defi cits had not over an annual average defi cit of less than been large. None-the-less, these were still fi ve percent of GDP except four, only one not years of fi scal implementation, but of of which was a former Soviet republic continued fi scal construction and strug- (Kyrgyz Republic). The reductions in the gles to reduce unsustainable defi cits. fi scal defi cits across countries over the Fiscal tools were unusable for economic eleven years provide strong and obvious management.33 Expenditure reduction in evidence that the defi cit levels were pri- Moldova was draconian in the extreme. marily a systemic problem of transition, Although government expenditure in which government policy could aff ect 1999 was the same share of GDP as it only to a limited degree. The fi scal posi- had been in 1993, GDP itself had fallen by tion of Moldova corresponded to the more than half. pattern across countries. After defi cits ranging from six to eleven percent of GDP Throughout most of the 1990s it would during 1993-1997, the average fell to less be a mistake to treat the governments than two percent during 1998-2004 (for of these countries as having fi rm control Moldova, see Figure 2.6), which refl ected over their fi scal systems, particularly taxa- the slow adaptation of the economy to tion. This loss of control was structural, market regulation. not the result of mistakes in policy or lack of competence of civil servants. The Unacceptable human suff ering lay be- new taxation systems, based on sales, hind these statistics on defi cit reduc- income and company taxes, were rarely tion, for they were achieved in most of accompanied by adequate enforcement the countries through draconian real mechanisms, in part due to lack of expe- reductions in social expenditure. While rience with them, and the low morale of narrowing the fi scal gap represented a woefully under-paid civil servants new to necessary step towards macroeconomic their tasks. Analysis of the problems of fi s- stability, the result was a more serious cal sustainability in Moldova by reference gap, namely, that in social provision, to the government’s failings or yielding which has been a major contributor to to special interests underestimates the the most fundamental gap of all, be- systemic diffi culties of fi scal policy during tween the level of deprivation and meet- the transition process.34 ing basic social needs. This deprivation gap is enormous in Moldova and unlikely Perhaps in no newly independent country to be substantially narrowed in the fore- were the problems of fi scal management seeable future through market processes and low morale of civil servants worse alone. It is essential that future defi cit than in Moldova. The non-sustainability targets derive from the goal of poverty of the fi scal position of the country in the reduction, as well as the needs of macro- 1990s was inherent in the transition pro- economic stability. This requires using cess, not the result of lack of government fi scal policy actively as an instrument of ‘will’ or ‘commitment’. By the early 2000s, growth promotion.

33 ‘There are some internal constraints…which impede a stimulatory fi scal policy…These constraints are the following: political constraint (uncertainty linked with Transnistria region, or low viability of executive bodies); obsolete social infrastructure...; external and internal…public debt servicing; [and] signifi - cant share of the shadow economy…’ (CISR, Moldova in Transition July 1998, 9) 34 An example of such an emphasis is the following passage from the CASE report on the impact of the Russian fi nancial crisis: ‘Moldovan fi scal policy in recent years was driven by inertia and pressure from [interest groups]. This statement is refl ected [sic! supported] by the slow path of structural reforms and the general weakness of the state. Loose fi scal policy in turn reduced the determination in reforming the Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction state structures’ (Radziwill, Serbatchi & Zaman 1999, 20). This statement assumes that systemic obstacles to the changes required to achieve fi scal sustainability did not exist.

Chapter 2: Economics of the Transition

52 53 Further, the statistics showing the de- tion in Moldova, embarking on the policy cline in social provision do not provide of reducing rates courts fi scal disaster. A the information necessary to assess the more reliable method for reducing tax relative impact on men and women. It is evasion might be improved enforcement known that redundancies in the public of the tax laws. One presumes that the sector aff ected women more than men, evasion-of-taxes argument goes as fol- and the collapse of social services pro- lows: within the existing tax code, house- vided through enterprises, such as child holds and businesses calculate the trade- care, reduced female access to the labour off between the benefi ts of evasion and market. the costs of being apprehended; if the tax rate were reduced, more taxes would be The extent to which the Moldovan gover- paid because the expected benefi ts of nment might use fi scal policy as an instru- evasion fall. It follows from this argument ment of growth and poverty reduction is that more taxes would be paid were the discussed in the next chapter. Relevant probability of being apprehended as a at this point is the fundamental social is- tax evader to rise, and if the penalties sue of the appropriate level of revenue for evasion were increased. Developed and expenditure to achieve the social country experience would suggest that goals for the government, especially po- the mechanisms and technology of moni- verty reduction. In the previous section toring tax payers provides a powerful way it was concluded that sustainable rates of to reduce evasion. Here, the question is, growth are unlikely by themselves to re- whether the increase in public enforce- turn the poor to their standard of living of ment expenditure necessary to reduce 1990 in less than a half century. Therefore, evasion would result in a net increase in poverty reduction requires that the public public sector resources. This would seem sector play a major role in social provision. the more promising approach for the Emphasis on private provision of educa- government, given the resource needs for tion and health care has and will continue poverty reduction. to be a vehicle to perpetuate poverty and inequality. This is not an ideological Closely related to the evasion argument judgement, but a fact of Moldova’s low and specifi c to the company tax, it is ar- income status. gued that since the Moldova rate is above the Russian rate, transfer pricing by com- With this context in mind, one can consi der panies shifts profi ts out of Moldova. While the goal of the government to move the this is a serious problem, lowering tax average tax share in GDP towards twenty rates is not a sustainable policy vehicle to percent, from thirty percent during 1994- resolve it, since Moldova’s neighbours, in- 2003. Two major arguments have been cluding Russia, could retaliate by lowering advanced in favour of this policy: that low- theirs. A more viable strategy would be to er marginal and average tax rates would initiate discussions among governments reduce tax evasion and that Moldovans in the region to harmonise tax policy and are so poor that the current tax burden cooperate on enforcement. Further, were contributes to their poverty. Both argu- the government to take as its corporate ments require close inspection. It may or tax benchmark the Russian rate, this may not be the case that lower tax rates would be quite damaging to Moldova’s would reduce evasion, but lower marginal prospects for EU entry. As pointed out be- rates for the majority of households could low, countries with tax rates far below the be consistent with the same average tax EU average are unlikely to receive serious share within a progressive personal and consideration for entry. In eff ect, emula- corporate income tax structure. Further, ting Russian corporate tax rates would

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction without a reliable estimate of the gains in be part of a policy of staying out of the tax compliance as a result of rate reduc- European Union.

Chapter 2: Economics of the Transition

52 53 A very practical consideration casts into share proportionately in the benefi ts of doubt the wisdom of reducing the tax growth. Therefore, a key element of a pov- share. The Moldovan government and, erty reduction strategy must be provision apparently, a majority of the population, of a substantial ‘social wage’36 Given the support the country’s application to enter government’s other expenditure commit- the European Union at some point in the ments, for external and internal debt serv- future. To achieve this goal, any Moldovan ice, national defence and security, public government would have to provide a infrastructure, and general administration, level of social provision, consistent with it is extremely unlikely that suffi cient rev- its per capita income, to some minimum enue would remain to cover even a mini- stan dard. Since no country in the Euro- malist poverty reduction programme were pean Union, including the new entrants in revenue to be twenty percent of GDP. 2004, has a tax share as low as twenty percent, it is unlikely that this share would be A fi nal issue of fi scal policy, which is dis- viewed by EU members as a satisfactory cussed further in the next chapter, is the basis for candidacy. At the time this report appropriate size of the central government was written, several of the new entrants defi cit. Since 1998, Moldovan governments into the EU had come under criticism by have not merely brought the fi scal defi cit pre-expansion members for their alleged under control, but maintained it signifi - ‘tax-haven’ status. It is quite unlikely that cantly below the famous three percent of in the future new entrants would be al- the ‘Maastricht criterion’ of the European lowed both to avail themselves of the Union. Were the government to maintain subsidy programme for underdeveloped the current level of revenue, allowing ex- regions, and have tax rates that could penditure to rise consistent with a three potentially induce companies to relocate percent defi cit would provide ‘fi scal space’ from existing members to new entrants.35 for pro-poor public investment, as discussed in the next chapter. The argument Most important of all, the degree of in- that such a modest defi cit would have income inequality in Moldova and its per fl ationary eff ects has no empirical support, capita income imply that earnings from especially in a small, open economy.37 Of- employment will not be suffi cient to fi cials at the National Bank of Moldova rec- achieve substantial poverty reduction ognise this, and their main defi cit concern, in the future, even should the poor only which they share with colleagues in the

35 See the discussion in Judt (2005). 36 This commonly used term refers to basic needs that could be purchased in the market, but are provided by the public sector, such as subsidised housing and health care without user fees. 37 The high rates of infl ation in Moldova and other transition countries in the 1990s were not the result of fi scal defi cits. Both infl ation and the defi cits were the result of the collapse of the old economic system before the new one was in place. With few exceptions, the transition countries would seem no more infl ation prone than the Western European countries. This is suggested by a simple hypothesis test, in which the infl ation rate is a function of the fi scal defi cit as a portion of GDP (proxy for money creation), and the rate of growth of GDP (real demand pressure). The statistical estimation covers 1998-2004, after the hyper-infl ation associated with the period of regime change of the transition countries. Only Bela- rus, Serbia and Montenegro and Uzbekistan, out of the twenty-seven transition countries, suff ered high infl ation rates after 1997. These are assigned dummy variables. The statistical result is: ln[infl ation] = .096 + .299 ln[Fiscal Defi cit] + -.167 ln[GDP growth] + .189 [Romania] + .223 [Srb&Mntgr] + .531 [Belarus] + .167 [Uzbek] R2 = .466, F = 25.18, DF = 173 The defi cit and GDP growth are non-signifi cant. All of the country dummies are signifi cant at less than one percent probability. The equation suggests that ‘structural’ infl ation is slightly less than ten percent (the intercept). Across countries, the infl ation rates were signifi cantly lower for 2002-2004 than for 1997- 2001. If a dummy variable is used to divide these periods, the post-2001 structural infl ation rate is 3.2 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction percent across countries. While this is a crude exercise, it supports the widely-held view that infl ationary pressures subsided after the turn of the century, replaced, some would argue, by defl ationary pressures.

Chapter 2: Economics of the Transition

54 55 Ministry of Finance, is possible ‘crowding public expenditure. For example, if pub- out’ eff ects on private investment. lic investment in infrastructure, such as roads, lowers private sector costs, this As government offi cials are well aware, would raise the rate of return on private defi cits themselves do not generate investment more than the increase in the crowding out. Fiscal defi cits result in interest rate. crowding out if they put upward pressure on interest rates, and private investment Finally, one presumes the crowding out is interest rate sensitive. As we see below refers to domestic private investors, since in the discussion of monetary policy, foreign investors have greater opportuni- nominal and infl ation-adjusted interest ty to raise funds in money markets outside rates have been quite high in Moldova, of Moldova. There are several infl uences above the so-called Golden Rule (the that might crowd out domestic private real interest rate equal to the long term investment, of which the interest rate ef- sustainable rate of per capita income fect of defi cits may be a relative minor growth). Therefore, even were there up- component. One source of crowding out ward interest rate pressure resulting from of domestic private investment is foreign fi scal defi cits, actual interest rates are so investment.38 Empirical evidence suggests high that the eff ect of the fi scal defi cit that this is considerably more important would seem insignifi cant. Further, the than the fi scal defi cit. The ‘crowding out’ crowding out eff ect is partly dependent eff ect of FDI on domestic investment is upon the purpose for which the govern- strongly suggested by Figure 2.7.39 It is ment expenditure that generates the realistic to conclude that while a crowd- defi cit is spent. If the defi cit results from ing out eff ect of the fi scal defi cit is a pos- public investment that complements sibility, given the level of interest rates in private investment this could nullify the Moldova and the likely use of increased crowding out caused by the increase in public expenditure, the practical impact the interest rate as a result of increased would be insubstantial.

38 Basic theory predicts this. Assume that domestic and foreign investors have equal access to all investment possibilities. At any moment, there is a fi nite set of investment opportunities, and businesses chose them in descending order of the rate of return. For each rate of return, the more that are taken by foreign investors, the fewer are left for domestic investors, and vice-versa. If due to technological access, skills or other reasons, some opportunities are more accessible to foreign investors and others more accessible to domestic investors, then crowding out is reduced. 39 Using data from the NBM, the following regression was estimated for 1994-2004:

[DPI/GDP]t = αo + α1[FsDf/GDP] t + α2[FDI/GDP]t [What is FsDF??] Defi ning domestic private investment (DPI) as gross domestic investment minus public investment and direct foreign investment, the following results were obtained:

[DPI/GDP]t = 15.378 + -.266[FsDf/GDP]t + -.946[FDI/GDP]t T-statistics (13.25) (-1.51 nsig) (-4.60 sig @ .01) Adjusted R-square = .814, F = 17.53, sig. @ .01 The coeffi cients are of the predicted sign, but that for the fi scal defi cit is non-signifi cant. The elasticity of domestic private investment with respect to foreign investment is not signifi cantly diff erent from unity; that is, the hypothesis that crowding out of domestic investments by foreign investments is total is not Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction rejected. This is not necessarily an argument against FDI, since it may in some sectors create more effi cient production units than domestic investors would.

Chapter 2: Economics of the Transition

54 55 Box 2.1 Defi cits and Infl ation This report concludes that fi scal policy has been too restrictive in Moldova, and that the government could run a small defi cit without threatening macro stability. Why does the report make this argument? Government expenditure, when there is no increase in public revenue, increases aggregate demand, whether it causes a defi cit or not, and this can be infl ationary under specifi c conditions. How much it increases aggregate demand depends on the famous ‘multiplier’, the amount national income must increase in order to generate ‘leakages’ (saving, imports and taxes) that equal the increase in expenditure. In a small, open economy like Moldova’s, increases in expenditure of a few percentage points of GDP are unlikely to be infl ationary, because the excess demand generated will be satisfi ed by imports. The same analysis and conclusion apply to any form of expenditure, public or private. Fiscal defi cits can have an infl ationary eff ect depending on how they are fi nanced and the supply response of the economy. If the government sells bonds to the private sector, and spends the money from those sales, a defi cit results or the existing defi cit increases. However, no increase in the money supply results, so the eff ect is not infl ationary, though it can have a negative eff ect on private investment (see Box 2: Crowding Out). If the government fi nances the defi cit by selling bonds to the central bank or to a government institution, such as a pension fund, the money supply is increased by the amount of the increase in the defi cit (the defi cit is ‘monetised’). The increase in the money supply cannot exceed the increase in the defi cit. The quantity theory of money, the basis of orthodox monetary analysis, states that the maximum value of the ratio of the increase in the price level to the increase in the money supply is one. Thus, an increase in the fi scal defi cit of one percentage point, totally monetised, would increase the money supply by one percentage point, and have a maximum infl ationary eff ect of no more than one percentage point. This is a very small increase compared to Moldova’s rate of infl ation during 2000-2005. As with an increase in aggregate demand, the excess demand generated by the increase in the money supply would be partially satisfi ed by an infl ow of imports. Thus, the net infl ationary eff ect would be less than the theoretical maximum Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

56 57 Box 2.2: Crowding out and Crowding In The emphasis placed on public investment in this report requires consideration of one of the major arguments against it, so-called crowding out. The argument that public investment in Moldova would crowd out private investment is somewhat surprising. It is typically the case that those who make this argument also urge governments to undertake major policy changes to encourage infl ows of foreign investment, usually without expressing strong concern that the latter might crowd out private domestic investment, though the same analysis applies to all forms of expenditure, public or private. The mechanism by which crowding out occurs is through the interest rate. Under some circumstances, government borrowing to fi nance its investment can push up the commercial interest rate. Given the high commercial rates of interest in Moldova, government borrowing is unlikely to have a substantial eff ect. In general, crowding out occurs when an economy is near full employment. When there are unutilised resources, there is resource space for an increase in all types of expenditure, public and private. Even if government borrowing did push interest rates up, this is unlikely to have a negative eff ect on economic growth. Crowding out of private investment is unlikely to be complete, either in theory or practice. That is, the elasticity of private investment with respect to government expenditure of any type will be less than minus one. As a consequence, public investment would be growth- inducing both in its demand and capacity eff ects, unless the return on the marginal private component of investment (the part ‘crowded out’) were suffi ciently higher than on the public component. In this case the growth impact would be negative. This can be shown formally using the simple Harrod-Domar model, where y is the rate of growth, v is the incremental capital-output ratio, and i is the shareshare of investmentinvestment in output. Let the subscripts pr and pu be private and public investment, respectively. Without public investment, the warranted (potential) rate of rate of the economy is: yo = [vpr][ipr] Let the ‘crowding out’ ratio be a (the faction by which public investment reduces private investment). Then, the new growth rate with public investment is: y1 = [vpr][ ipr - aipu] + [vpu][ipu] Subtracting y1 from yo, one gets: yo – y1 = ipu[avpr – vpu] Crowding out will reduce the rate of growth if and only if, vpu > avpr. If the capital- output ratioratio forfor public investmentinvestment is smaller than forfor privateprivate investment,investment, public investment never reduces the growth rate, no matter what the value of a, assuming its upper limit to be unity (‘total crowding out’, one hundred percent). If crowding out is total, the growth rate falls only if public investments are more capital-using than private ones. Thus, the negative impact of public investment on the capacity-creating source of growth occurs only under the very restrictive conditions which crowding out is total and private investments use less capital per unit of output. The former is unlikely and the latter can be avoided by judicious public choice of investment projects. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

56 57 Monetary Policy and In line with this statement of policy, the the Exchange Rate NBM during 1998-2004 engaged in what a NBM offi cial called a ‘managed fl oat’. The National Bank of Moldova, the insti- However, this intervention appears to be tution responsible for the execution of minimalist, with the primary focus of the monetary policy, is explicitly assigned NBM being price stability, implemented the task of managing the exchange rate through adjustments in interest rates. In under the law creating it; indeed, ‘The 2004 the NBM ‘narrowed’ its interpreta- main objective of the National Bank is to tion of its mission to one primarily fo- achieve and maintain the stability of the cussed on infl ation targeting. However, national currency’.40 Nowhere is it expli- seeking price stability in this manner is citly required that the NBM should target unlikely to be a success, since offi cials infl ation. Indeed, in the provisions of the of both the IMF and the NBM expressed law dealing with objectives (Chapter I, scepticism as to the eff ectiveness of in- Article 4), general functions (Chapter I, Ar- terest rates in infl uencing the price level. ticle 5), monetary and foreign exchange In the Moldovan context, a more growth- policy (Chapter II), and the functions of fostering objective of the NBM would be the governing body (Chapter III, Articles purposeful management of the exchange 25 and 26), the words ‘infl ation’ and ‘price rate. stability’ do not appear. However, the current interpretation of the function of the From a theoretical perspective, one can NBM places emphasis on domestic price treat Moldova as a ‘small, open economy’; stability: that is, as a ‘price taker’ in its external trade.42 If one assumes the operation of From the very beginning the main objec- the Law of One Price, then the eff ect of in- tive of the NBM is to realize and maintain terest rate adjustments can be considered the stability of the national currency that by using the Mundell-Fleming model, the ultimately provides the reduction of the IMF’s standard framework for analysing infl ation rate, the ensurance of the gen- external adjustment. In this approach, eral level of prices and of the exchange domestic prices of traded commodities rate of the MDL (NBM website, ‘Monetary cannot systematically diff er from external Policy’).41 prices, except due to tariff s (which aff ect

40 This and the other documents establishing and regulating the NBM can be found on the NBM website (http://www.bnm.org/english/index_en.html). It is useful to quote in detail from the Law on the Na- tional Bank of Moldova (Chapter I, Article 4): The main objective of the National Bank is to achieve and maintain the stability of the national currency. In order to achieve this objective the National Bank establishes and maintains monetary, credit and exchange market conditions conducive to the orderly, balanced and sustained economic development of the Republic and especially of the market-based fi nancial and foreign exchange system. The fi rst function of the Bank listed under Article 5 is ‘to formulate and to promote the state monetary and foreign exchange policy’. It is given authority to use its assets to achieve this. In Chapter V, Article 53 the foreign exchange that it holds is defi ned as one of those assets. Chapter I, Article 15 empowers it to enter into domestic open market operations and Article 16 grants it the right to ‘buy, sell and deal in foreign exchange’ and ‘determine the rate at which it will buy, sell or deal in foreign currencies’. Chapter V, Article 51 gives it the power to intervene in a variety of ways to regulate private dealers in foreign exchange. The document, ‘Regulation on Foreign Exchange Regulation [‘regulation’ a second time??]on the Ter- ritory of the Republic of Moldova’, defi nes the exchange rate as follows: ‘the price of foreign currency expressed in [an]other foreign currency set, as in accordance with the provisions of this Regulation, by the National Bank of Moldova (offi cial exchange rate) or by authorized dealers…’ 41 Comparing this to the text in Romanian, it would appear that ‘provides’ would be better translated as ‘requires’, and the word ‘ensurance’ should be ‘maintain’.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 42 This assumption does not exclude the possibility that private market power manages prices of imported commodities.

Chapter 2: Economics of the Transition

58 59 the margin between the two, but not their nominal rates down into the teens, but relative movements). It follows, therefore, this still resulted in punishingly high in- that domestic infl ation will tend to refl ect fl ation-adjusted rates during 2001-2004 international infl ation and the prices of (except in 2003; again, see Figures 2.8 and non-traded goods if the exchange rate is 2.9). During 1995-2000, after the hyper- constant. infl ation had ended and when nominal NBM rates were over twenty percent (ex- In these circumstances, theory predicts cept for 1997), the infl ation adjusted NBM that the main result of changes in inte rest rate was fi ve percent, which commercial rates is refl ected in capital fl ows. If the banks passed on to borrowers at a mul- domestic rate of interest lies above the tiple of over three, namely, at an average ‘international’ rate, capital will fl ow into of sixteen percent. Facing such a punitive a country, as private agents seek cheaper borrowing rate, it is not surprising that borrowing abroad, and external specula- private sector investment was weak--af- tors seek to gain from holding their funds fordable to those companies that could in the high interest rate country. However, borrow externally. in Moldova, the most important source of fi nancial infl ows is remittances by citizens After 2000, the NBM base rate averaged abroad, which have little sensitivity to the approximately half what it had been over domestic interest rate. Thus, high interest the previous six years, yet the infl ation rates in Moldova could achieve neither of adjusted base rate rose by a percen tage their objectives, neither reducing infl ation point, and the commercial lending rate nor attracting capital. remained the same. These real rates represented at the time of this report a However, the infl ow of remittances has major relative price misalignment in the the eff ect of appreciating the exchange economy. The famous ‘Golden Rule’ from rate, as occurred during 2003 and 2004, growth theory states that long term real thus making exports less competitive. interest rates should not exceed the sus- The higher interest rate would discourage tainable rate of real per capita growth.43 investment by Moldovan companies but Based on growth during 2000-2004, that have little impact on foreign direct invest- sustainable rate can be estimated as four ment. Though there is no research on the percent. Thus, private borrowers faced strength of these eff ects in Moldova, their commercial interest rates misaligned by impact may have been quite strong in the a factor of four. To place this in perspec- late 1990s and early 2000s. Both nominal tive, one can speculate on the eff ects on and real interest rates set by the NBM have employment were the average wage level been extremely high since independence in Moldova to increase by four hundred (see Table 2.8 and Figures 2.8 and 2.9), percent. except during years of high infl ation for the latter. Attempts to moderate infl ation An equally serious misalignment has been after independence by raising nominal the enormous gap between deposit and interest rates to three-digits proved fruit- lending rates. On an annual basis, lending less; indeed, infl ation fell once interest rates exceeded deposit rates by over forty rates began to decline from the punitive percent during 1996-2004. The tiny annu- levels of the early 1990s (again see Table al variation in this interest rate diff erential 2.7). Subsequently, the NBM brought (its coeffi cient of variation being less than

43 It became fashionable in the 1980s and 1990s for some economists to ‘trump’ the Golden Rule’ with the argument that high rates of interest in developing countries represented ‘risk premiums’, which could allegedly be justifi ed on effi ciency grounds. However one justifi es interest rates above the long term Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction sustainable rate of per capita growth, it remains true that they represent a barrier to reach that rate of growth.

Chapter 2: Economics of the Transition

58 59 0.10) strongly suggests commercial banks think it would--research suggests no link exert considerable market power.44 The between growth rates and infl ation up policy implication is that in so far as the to about forty percent (Bruno & Easterly NBM cannot aff ect the rate diff erential, 1999). Infl ation may or may not be ‘bad stimulating private investment requires for the poor’,47 but abandoning infl ation much lower base rates. If the ratio of the targeting would prove unambiguously base rate to the commercial lending rate pro-poor. is taken as a guide,45 bringing real lending rates down close to the Golden Rule rate, A further reason for the NBM to consider say six percent, would require a base rate initiating a more purposeful exchange of two percent.46 rate policy is the impact of remittances, to which we alluded above. In this chap- There can be little doubt that the NBM ter we do not analyse remittances in base rate and private sector interest rates detail, but restrict the discussion to the were seriously misaligned in Moldova exchange rate eff ect. After a long period during the 1990s and the fi rst half of of relative real stability from 1999 through the 2000s, as were the relative prices of 2003, the Leu appreciated in the second lending and holding deposits. While the half of 2003 and throughout 2004 (see private rates probably refl ect market Figure 11). Measured by its dollar rate, its manipulation, the NBM rates refl ect a value adjusted by domestic infl ation rose concern over the return of unmanageable by fi fteen percent. The generally accepted infl ation. As argued above, the base rate eff ects of exchange rate appreciation are is an inappropriate instrument to deal that it renders tradable commodities less with this anxiety, because hyper-infl ation profi table and thus discourages both was a transitory phenomenon during the production of import substitutes and change in the economic regime from cen- exports. Since, as we see in the next sub- tral planning to market regulation. Since section, the country ran a massive trade the hyperinfl ation of the early 1990s, defi cit during these thirty-two months, Moldova has displayed considerable price the appreciation must have resulted from stability, except during the shock of the fi nancial infl ows. Since capital infl ows Russian fi nancial crisis, when exchange were relatively small, the only credible rate depreciation again provoked an infl a- explanation for the appreciation is the in- tionary burst (see Figure 2.10). fl ow of remittances, a large part of which go unrecorded. Evidence suggests that the Moldovan government and its monetary authori- The remittance eff ect on Moldova’s ex- ties could safely pursue a low interest change rate has been similar to that mani- rate policy without provoking infl ation. fested in countries passing through an ex- The eff ect of such a policy would be to port commodity price boom or countries stimulate growth, which would enhance receiving large infl ows of concessionary poverty reduction. Even should this be or commercial capital. In eff ect, the ex- accompanied by moderate infl ation in change rate disconnects from export the low teens-- and there is little reason to competitiveness, to rise or fall in reaction

44 It is the view of NBM professionals that commercial banks in Moldova are extremely profi table, as the rate diff erential would suggest. 45 Calculation from Table 2.7 yields a ratio for the nominal rate during 1995-2000 of 2.00, and during 2001- 2004 of 2.19. The standard deviation across all ten years is quite low, .3. This suggests that commercial banks practiced mark-up pricing of loans. 46 The ratio of the real base rate to the real lending rate for 1995-2000 was three, and for 2001-2004 it declined slightly to 2.73. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 47 The evidence from other countries does not support the general conclusion that infl ation hurts the poor disproportionately (van der Hoeven 2004).

Chapter 2: Economics of the Transition

60 61 to purely fi nancial fl ows. For Moldova, to eff ective exchange rate policy. Nominal with its large trade defi cit, this is a very devaluations could also have negative serious matter, even were remittances welfare eff ects, particularly on the poor, sustainable at the level of the fi rst half of through their impact on energy prices. the 2000s, (which they probably are not). This problem could be mitigated by ap- Most of Moldova’s exports are sold in plying energy subsidies, as was done in highly competitive international markets, the 1990s. and especially its agricultural exports. The eff ect of a real appreciation could under- Our comments indicate that pro-poor mine export performance by rendering growth would be fostered if the parlia- marginal producers unprofi table, and dis- ment and the government gave the NBM couraging the entry of new producers. It is a broader mission than currency stabi- quite clear that in Moldova, exchange rate lity. This broadening of mission would movements that undermine the produc- be consistent with the realities of the tion of tradables are not pro-poor. Moldovan economy. Our review of 1990- 2005 suggests that infl ation is a structural It is generally agreed that remittances can- phenomenon resulting from supply bot- not be monitored, and can be regulated tlenecks created by the collapses of the only to a limited degree. This means that economy, and the enormous infl ow of the use of ‘open-market operations’ to remittances. In consequence, the inter- sterilise remittance fl ows, even if the do- est rate is an ineff ective tool of infl ation mestic bond market were suffi cient deep management. to do so, would always occur in response to exchange rate movements, rather than The broader mission of the NBM would in anticipation of them. Since after-the- call for accommodating growth in the event intervention is a fact of economic short and medium term, and treating cur- life in this sector of the economy, a more rency stability as a medium term rather eff ective post-facto intervention would than a short term goal. This broader mis- be to operate directly on the exchange sion could be part of changes that would rate through the buying or selling of Leu. bring central bank regulation more into Thus, for a range of reasons, we conclude line with European Union practice. The that prudent and pro-poor exchange rate NBM is a competent and professional or- policy would involve more aggressive in- ganisation. However, its operations suff er tervention to mange the exchange rate. from the absence of formalised mechanisms of accountability which exist in al- The government should consider a policy most all European countries. Its functions of managed undervaluation of the ex- would be enhanced by mechanisms that change rate, in which the actual trading create parliamentary and public oversight rate for the Leu stays ahead of the internal in order to foster transparency through price adjustments occurring in response accountability. to the Law of One Price. In other words, exchange rate policy would use the lag in the adjustment of internal prices to Trade Policy and external prices to keep the Leu margi- Poverty Reduction nally undervalued. This recommendation is complicated by the negative impact it The international trade position of Moldo- would have on external debt servicing. va involves a simple story: the balance This makes the debt relief by external of exports and imports is inconsistent agencies, discussed above, all the more with long term development, and must necessary. As for fi scal policy, the debt be corrected as quickly as possible. The

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction burden resulting from miscalculation of enormous trade defi cit as a share of GDP Moldova’s per capita income is a barrier has been sustained through remittances.

Chapter 2: Economics of the Transition

60 61 If remittances begin to fall, the country ing 1994-1996 to twenty-fi ve percent for will suff er yet another severe compression 1997-1999, to close to fi fteen percent dur- of the economy, causing falling incomes ing 2000-2002 (when the data end). In ef- and rising poverty. If remittances are fect, Moldova is substituting remittances sustained, the result is hardly better, how- for net agricultural exports. ever, for it will imply that Moldova will be relegated to the category of an impove- Such a large trade gap is unlikely to be rished exporter of cheap labour. closed or reduced through market processes. Quite the contrary is likely: remittances Table 2.8 provides the basic statistics, and would prevent exchange rate depreciation, Figures 2.12 and 2.13 provide information further discouraging export production in on the external balances. During 1994- agriculture and manufacturing. Even the 1995, the export-GDP ratio was about exchange rate policy suggested above fi fty-nine percent, and the import-GDP would be insuffi cient to the task of sub- ratio about sixty-seven, producing a stantially reducing the current account trade defi cit that averaged eight percent. defi cit. Economic theory, as well as expe- Over the years 1996-2003, the former was rience, suggests that relative price move- fi fty-two percent and the latter seventy- ments bring about marginal changes, but fi ve, for a trade defi cit moving towards in international trade Moldova requires one-fourth of GDP. The statistics for the much more than this. Given the apparent fi rst half of 2004 suggest an even higher low elasticity of the current account with trade defi cit, though this may fall in rela- respect to the exchange rate,49 the change tive terms depending upon the realised in the latter necessary to have a major im- rate of GDP growth. Except in a very few pact on the former would be so large that cases,48 a country with a trade defi cit it would destabilise the economy long be- of this magnitude courts disaster. One fore it corrected the balance of payments. source of this disaster is that the country The price movements arising from such is highly vulnerable to changes in prices a massive structural imbalance give rise of imports, especially petroleum, with the to catastrophic changes, as the events of benefi ts from increased prices of exports 1991-1994 and 1998 already demonstrate. much less pronounced. In addition to energy, the trade defi cit arises from the However, several countries, most notably net imports of manufactures, which re- Vietnam, have corrected trade defi cits this fl ect the collapse of much of Moldova’s large without a severe depression. The industry. policies to achieve this are well-known, and can be implemented in Moldova. Over the medium term the most competi- They involve providing strong and con- tive sector for exports will be agriculture, crete incentives to export and in some yet net exports of food commodities cases targeting these to sectors with declined sharply from thirty percent of export potential. A fi nal observation on the average of imports and exports dur- remittances is necessary. Experience from

48 One case was Panama in the 1950s and 1960s, which had a large surplus on the service account because of the Canal and an oil pipeline. Even in this case, after two decades of rapid growth, the secular decline in the use of both resulted in intractable problems of adjustment. 49 Assume that the current account is determined by the real exchange rate and the interest rate on dollar deposits (which proxies for the diff erence between domestic and ‘world’ interest rates). Since the current account balance shifts between positive and negative, logarithms cannot be used to estimate the model for thirty quarters, 1997.1 through 2004.2 (and there are no quarterly GDP statistics to allow use of a relative measure). The ordinary least squares statistics are: [crr acc] = -.269 – 1.189 [Real Exchange Rate]t - .269 [US$ deposit rate] t t=stats = (-2.82) (-4.53) (-.06) Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction The interest rate variable is not signifi cant. The exchange rate is of the predicted sign and signifi cant at less than .01, though the calculated elasticity is less than .10.

Chapter 2: Economics of the Transition

62 63 other countries shows that worker remit- suff ered from a precipitous drop in remittances arise from specifi c circumstances, tances when the petroleum price fell in and are unlikely to be sustainable in the the early 1980s. Reducing the trade gap long run. For example, those countries (thereby reducing dependency on remit- that provided large numbers of workers tances) would be the only way to prepare to the oil-exporting countries in the 1970s for a similar development for Moldova.

Table 2.1: Transition Declines & Disasters 1990-2001, Levels & Changes in GDP (lowest value = 100)

Percent change Highest in the 1990 Lowest 2001 Decline/ Country 1990/ low/ value year value =100 value high/low recovery* low 2001 Georgia 480 1985 361 1994 146 -131 -113 37 -94 Tajikistan 401 1988 353 1996 126 -120 -112 23 -97 Moldova 292 1989 284 1999 109 -98 -96 9 -89 Turkmenistan 275 1988 253 1997 156 -93 -87 44 -50 Ukraine 252 1989 236 1998 118 -86 -81 17 -70 Armenia 225 1990 225 1993 156 -77 -77 43 -33 Kyrgyz Rep 205 1990 205 1995 128 -69 -69 25 -44 Latvia 196 1989 195 1992 148 -65 -64 39 -26 Azerbaijan 191 1992 na 1996 146 -63 na 38 -25 Lithuania 176 1990 176 1994 135 -55 -55 30 -25 Russia 178 1989 172 1998 122 -56 -53 20 -36 Kazakhstan 169 1989 160 1995 138 -51 -46 32 -19 Croatia 156 1990 156 1993 121 -44 -44 19 -25 Belarus 153 1990 153 1995 145 -42 -42 36 -6 Estonia 154 1989 143 1994 149 -43 -36 40 -3 Uzbekistan 138 1989 137 1995 115 -32 -31 14 -18 Slovak Rep 135 1989 131 1993 137 -29 -27 31 2 Bulgaria 141 1988 127 1997 121 -34 -24 19 -15 Romania 141 1987 124 1992 112 -34 -21 12 -23 Hungary 121 1989 117 1992 133 -19 -16 29 10 Slovenia 116 1990 116 1992 118 -15 -15 17 2 Czech Rep 114 1990 114 1992 121 -13 -13 19 6 Poland 108 1990 108 1992 154 -8 -8 42 35

Source: EBRD 2003; World Bank, World Development Indicators 2004. NOTE: For comparability, the countries are in order of their declines from 1990 to their lowest point.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction *Fall or rise in per capita income in 2001 compared to 1990.

Chapter 2: Economics of the Transition

62 63 no no no no no no no no no no no no no no 1996 2000 2002 1999 1994 1998 1998 1996 1996 2003 2003 2002 2002 year of:year to 1991: to recovery 1994 1994 1995 1994 1996 1995 1994 1994 1994 1994 1994 1995 1994 1998 1994 1994 1996 1999 1996 2001 2000 1997 2002 2002 2003 2000 1997 First growth positive positive 8 2 7 4 5 -6 -9 -1 -2 11 19 43 16 28 47 84 -12 -13 -43 -11 -52 -27 -73 -21 -35 -60 -13 1991-03 8 3 0 4 -5 -2 -2 16 40 12 26 41 81 -10 -18 -15 -16 -45 -20 -60 -11 -32 -79 -23 -66 -27 -43 1991-02 6 8 0 -3 -7 -5 13 38 23 36 77 -15 -25 -22 -10 -17 -49 -15 -31 -66 -32 -86 -31 -52 -28 -71 -20 1991-01 2 9 4 -9 -4 37 20 29 73 -19 -32 -29 -16 -11 -12 -55 -24 -41 -10 -71 -34 -36 -62 -40 -37 -92 -80 1991-00 4 2 -1 -8 33 15 22 67 -25 -17 -39 -33 -17 -14 -17 -60 -30 -52 -73 -43 -42 -94 -46 -71 -57 -86 -15 1991-99 0 1 -1 29 10 13 57 -27 -16 -42 -31 -16 -13 -42 -34 -61 -19 -76 -46 -46 -91 -53 -74 -74 -86 -21 -12 1991-98 0 6 0 -5 -3 24 42 -31 -21 -47 -38 -11 -16 -25 -44 -41 -71 -27 -79 -44 -48 -47 -81 -63 -83 -15 -84 1991-97 0 2 7 5 -5 -9 18 -25 -30 -10 -45 -22 -26 -54 -44 -39 -46 -58 -86 -49 -83 -51 -80 -17 -55 -77 -89 1991-96 -4 -9 -2 -2 12 -16 -11 -59 -15 -28 -27 -62 -50 -78 -41 -46 -65 -80 -45 -79 -45 -70 -19 -34 -50 -81 -100 1991-95 5 -7 -39 -58 -53 -26 -66 -31 -38 -60 -79 -41 -66 -37 -57 -18 -19 -10 -13 -16 -21 -15 -35 -68 -57 -102 -102 1991-94 -1 -21 -12 -37 -16 -60 -43 -20 -26 -13 -24 -41 -24 -70 -62 -46 -18 -91 -26 -40 -48 -29 -47 -20 -35 -14 -102 1991-93 Cumulative Changes in GDP, European and Central Asian Transition Economies, 1991-2003 Economies, Transition Asian Central and European GDP, in Changes Cumulative Country Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Bulgaria Czech Rep Estonia Hungary Latvia Lithuania Poland Romania Rep Slovak Slovenia Albania Bos&Herz Croatia Macedonia Serb&Monte Armenia Azerbaijan Belarus Georgia Kazakhstan Rep Kyrgyz Moldova Russia Tajikistan Turkmenistan Ukraine Uzbekistan Table 2.2: Table 2003. and WDI Annex 3.1, 2003, EBRD Sources: 19991. growth;rstyear of positive shadedrecovery year of of to – level NOTES: bold – fi

Chapter 2: Economics of the Transition

64 65 Table 2.3: Changes in Headcount Poverty and the Gini Coeffi cient, European and Central Asian Transition Countries,1987-1994

% Change Headcount Poverty Gini Coeffi cient Country Change Change due to 1987-88 1993-94 1987-88 1993-94 Gini*: Georgia na na na na na na na Tajikistan na na na na na na na Moldova 4 65 61 24 36 12 28 Turkmenistan (3) 12 48 36 26 36 10 34 Ukraine(2) 4 41 37 23 33 10 38 Armenia (3) na na na na na na na Kyrgyz Rep(1) 5 50 45 26 35 9 37 Latvia 1 19 18 23 27 4 20 Azerbaijan (4) na na na na na na na Lithuania (3) 1 49 48 23 37 14 46 Russia 2 43 41 24 36 12 52 Kazakhstan (3) 5 50 45 26 33 7 39 Croatia (3) na na na na na na na Belarus 1 11 10 23 22 -1 29 Estonia 1 33 32 23 40 17 58 Uzbekistan (3) 24 29 5 28 33 5 44 Slovak Rep 0 1 1 20 20 0 31 Bulgaria 2 17 15 23 31 8 56 Romania 6 32 26 23 33 10 52 Hungary 1 2 1 23 28 5 51 Czech Rep (3) 0 2 2 19 27 8 69 Slovenia (3) 0 1 1 24 29 5 65 Poland (3) 6 20 14 26 31 5 65 Average 4 28 24 24 32 8 45

NOTES: *Calculated from a cross-country regression using the countries with data, by substituting each country’s change in the Gini, holding per capita income (PCY) constant. The regression result is, with T test result in parenthesis under coeffi cient (ns notes non-signifi cance at .10): ln(d[Pov]) = .184 + 2.765[ln(dGini)] – 3.058[ln(dPCY]) (ns) (.090) (.009) Adjusted R-square = .532 F-statistic = 10.11 (signifi cant @ .004) DF = 14

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCES: Milanovic 1995, p. 14; Milanovic 1998, pp. 41, 68; UNDP 1998; EBRD 2003, Annex 3.1; and World Development Indicators 2003.

Chapter 2: Economics of the Transition

64 65 Table 2.4: Distribution of Expenditure by Quintiles, European and Central Asian Transition Countries, 1996-1999 Central and Eastern European Countries country year Q1 Q2 Q3 Q4 Q5 Belarus 1998 11.4 15.2 18.2 21.9 33.3 Bulgaria 1997 10.1 13.9 17.4 21.9 36.8 Croatia 1998 8.8 13.3 17.4 22.6 38.0 Czech Rep 1996 10.3 14.5 17.7 21.7 35.9 Estonia 1998 7.0 11.0 15.3 21.6 45.1 Hungary 1998 10.0 14.7 18.3 22.7 34.4 Latvia 1998 7.6 12.9 17.1 22.1 40.3 Lithuania 1996 7.8 12.6 16.8 22.4 40.3 Moldova 1997 5.6 10.2 15.2 22.2 46.8 Poland 1998 7.8 12.8 17.1 22.6 39.7 Romania 1998 8.0 13.1 17.2 22.3 39.5 Russia 1998 4.4 8.6 13.3 20.1 53.6 Slovenia 1998 9.1 13.4 17.3 22.5 37.7 Ukraine 1999 8.8 13.3 17.4 22.7 37.8 average 8.5 13.0 17.0 22.1 39.4 std dev 1.8 1.7 1.3 0.7 5.2 Central Asian Countries country year Q1 Q2 Q3 Q4 Q5 Armenia 1996 5.5 9.4 13.9 20.6 50.6 Azerbaijan 1995 6.9 11.5 16.1 22.3 43.3 Georgia 1996 6.1 11.4 16.3 22.7 43.6 Kazakhstan 1996 6.7 11.5 16.4 23.1 42.3 Kyrgyz Rep 1999 7.6 11.7 16.1 22.1 42.5 Tajikistan 1998 8.0 12.9 17.0 22.1 40.0 Turkmenistan 1998 6.1 10.2 14.7 21.5 47.5 Uzbekistan 1998 4.0 9.5 15.0 22.4 49.1 average 6.4 11.0 15.7 22.1 44.9 std dev 1.3 1.2 1.0 0.8 3.7

SOURCE: World Bank, Income distribution data base. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

66 67 Table 2.5: Moldova’s Per Capital Income from the WDR, 1991-1999 World Bank Year for PCY & title PCY Reported IMF Low-income PCY implied NBM date of WDR: Reported in Change in MRED 1996 Status? by 1999 website WDR PCY 1991 (1993) no 2173 – 1077 1992 (1994) no 1300 -16.0 905 1993 (1995) no 1060 -29.1 641 310 310 1994 (1996) no 870 -1.2 634 325 377 1995 (1997) no 920 -32.3 429 392 400 1997 (1998/99) yes 540 -4.2 411 528 1999 (2000/01) yes 370 -9.9 370 321

NOTES AND SOURCES: PCY Reported in WDR: per capita income reported in World Development Report. Reported Change in PCY: source as for PCY Reported. Change from the previously listed year. PCY implied by 1999: uses the reported changes and works backwards from 1999. IMF MRED: IMF (1996), photocopy of data page supplied by IMF. NBM: National Bank of Moldova

Table 2.6: Fiscal Defi cits as Percent of GDP, Transition Countries, 1992-2002

Countries: Averages: All yrs 1992-93 1994-97 1998-02 Croatia -3.2 -2.4 -.8 -5.5 Czech Rep -2.5 -.3 -1.7 -4.3 Estonia -.6 na -.5 -.7 Hungary -5.4 -6.1 -6.0 -5.1 Latvia -2.5 na -2.5 -2.5 Lithuania -4.1 -5.3 -3.9 -4.1 Poland -3.6 -3.7 -2.9 -4.4 Slovak Rep -5.6 -9.0 -1.9 -7.2 Slovenia -.6 .5 -.6 -1.6 Albania -12.3 -19.3 -11.9 -8.1 Bos&Herz -3.9 na -1.7 -5.3 Bulgaria -3.7 -6.9 -6.0 -.5 Macedonia -3.8 -11.6 -1.4 -2.6 Romania -3.3 -2.5 -3.3 -3.6 Serb&Monte -2.2 na na -2.2 Armenia -10.0 -34.3 -6.6 -4.6 Azerbaijan -3.8 -6.3 -5.2 -1.8 Belarus -2.1 -3.8 -2.1 -1.4 Georgia -8.5 -25.8 -6.7 -4.0 Kazakhstan -4.8 -5.7 -5.8 -3.1 Kyrgyz Rep -10.5 -14.4 -11.9 -8.6 Moldova -6.8 -17.1 -8.2 -2.8 Russia -5.9 -13.1 -8.7 -.9 Tajikistan -8.9 -26.8 -6.3 -1.6 Turkmenistan -1.0 -6.8 .6 -.8 Ukraine -6.6 -20.8 -5.9 -1.5 Uzbekistan -5.8 -18.4 -4.5 -2.3 average -4.9 -11.3 -4.5 -3.4 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

SOURCE: EBRD 2003.

Chapter 2: Economics of the Transition

66 67 Table 2.7 Nominal and Infl ation-Adjusted Interest Rates in Moldova, 1993-2004

Nominal Infl ation Adjusted Year NBM CBL CBB NBM CBL CBD 1994 150 160 90 -337 -327 -397 1995 21 42 33 21 42 33 1996 18 36 24 -6 13 0 1997 16 33 23 4 21 11 1998 33 30 21 25 22 13 1999 31 36 27 -8 -4 -12 2000 27 33 25 -4 2 -6 2001 13 28 21 3 19 11 2002 10 24 15 5 19 10 2003 14 19 13 2 7 1 2004* 14 21 15 14 21 14 AVERAGES 1995-00 24 35 25 5 16 7 2001-04 13 23 16 6 16 9

SOURCE: National Bank of Moldova. NOTES: NBM is the National Bank of Moldova base rate; CBL is the commercial bank lending rate; CBD is the commercial bank deposit rate. The infl ation adjustment is the relevant interest rate minus the GDP defl ator, or the general consumer price index for 2004. *Through August 2004.

Table 2.8 Exports and Imports, 1994-2004 2004 Item: 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 (2 qtrs) US $ MILLIONS Exports g&s 651 884 929 1057 796 615 644 740 874 1057 592 Merchandise exports 618 739 823 890 644 469 477 569 659 806 466 Imports g&s 751 1006 1249 1434 1230 781 971 1101 1248 1718 934 Merchandise imports 672 794 1076 1238 1032 597 783 882 1038 1428 648 PERCENTAGES Exports share of GDP 56.0 61.2 54.8 54.7 47.1 52.5 50.0 50.0 52.6 54.0 na Food/Exports g&s 66.8 71.7 72.0 72.5 75.4 68.0 61.6 62.9 63.0 na na Manuf/Exports g&s 27.9 22.7 23.2 25.5 22.6 27.1 33.3 33.7 33.0 na na Imports share of GDP 64.6 69.7 73.7 74.1 72.8 66.7 75.4 74.4 75.1 87.8 na Food/Exports g&s 6.5 8.1 10.7 11.1 7.8 5.9 13.1 14.3 14.0 na na Manuf/Exports g&s 35.4 41.8 44.2 48.6 57.6 52.1 51.0 55.8 57.0 na na TRADE BALANCES [X-M]/GDP -8.6 -8.5 -18.9 -19.5 -25.7 -14.2 -25.4 -24.4 -22.5 -33.8 -30.0* Food/Trade 31.8 32.3 28.2 26.3 24.0 24.2 14.8 15.6 16.2 na na Manuf/Trade -5.6 -11.4 -16.8 -19.4 -26.6 -15.7 -18.7 -20.3 -22.5 na na

SOURCE: National Bank of Moldova, website, Moldova, Economic Trends, various issues, and World Development Indicators 2002. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction NOTES: The last two rows of the table give net exports of food and manufactures as a proportion of the average of the total imports and exports

Chapter 2: Economics of the Transition

68 69 Figure 2.1:

GDP Growth Rates of Transition Countries, 1988-2003 (3-year moving averages)

NOTES: ‘TC’ stands for ‘transition countries’ (23 in all). ‘NIC’ stands for ‘newly independent countries’, and ‘frmr USSR’ includes the republics of the former Soviet Union.

Figure 2.2:

GDP Growth, Moldova minus Country Groups, 1988-2003 (3 year moving averages)

SOURCE: World Bank, World Development Indicators 2002. NOTES: See Figure 2.1 for legend. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

68 69 Figure 2.3

NBM and Various World Bank Estimates of Moldova’s Per Capita Income, 1990-2003 (logrithms)

SOURCES: World Bank, World Development Indicators 2002 and National Bank of Moldova Notes: WB is World Bank; PPP is purchasing power parity; US$ is constant United States dollars; and NBM is the National Bank of Moldova.

Figure 2.4

Aid Per Capita to Moldova as Percentage of Country Groups, 1992-2001

SOURCE: World Bank, World Development Indicators 2002. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

70 71 Figure 2.5

Moldova: GDP Growth, Annual & 3 Year Moving Average, 1990-2004

1990-1994 1995-2000 2001-2004 Collapse Stagnation Recovery & Growth IMF & WB First IMF program Feb 1993; IMF programme ‘off -track’ from IMF programmes policy IBRD structural adjustment late 1997, ‘revived’ in early 1999; (PRGF) expire Dec at LIBOR +; IMF calls mid-1998-2000 period 2003; of ‘intense reform’; WB: US$ 13 mn for No new IMF programmes initiated sectoral projects; after Dec 2000; 10.6 mn for structural in 1997 WB makes fi rst IDA terms adjustment (all IDA loans; terms) Revelant Transnistria establishes Remittances become largest forex End of 2004 IMF de- events separate government; Leu source; clines to support the introduced in Nov 1993; Russian fi nancial crisis 1998; EGPRSP with concessional lending Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

70 71 Figure 2.5:

Government Fiscal Balance, Moldova and Country Groups, 1991-2002

SOURCE: World Bank, World Development Indicators 2002.

Figure 2.6

Moldova: Public Sector Revenue & Expenditure as % of GDP, 1993-2003 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: National Bank of Moldova.

Chapter 2: Economics of the Transition

72 73 Figure 2.7

Moldova: Gross Fixed Capital Formation & Net Foreign Investment as Percent of GDP, 1990-2004

Figure 2.8

Molodva: Nominal Interest Rates, 1993-2004 (thru August)

SOURCE: National Bank of Moldova. NOTES: NBM is the National Bank of Moldova base rate; CBL is the commercial bank lending rate; and CBD is the commercial bank deposit rate. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

72 73 Figure 2.9 Moldova: Interest Rates minus the Rate of Infl ation, 1995-2004 (thru August)

SOURCE: National Bank of Moldova. NOTE: P* is the infl ation rate. See Figure 2.8 for the other abbreviations.

Figure 2.10

Moldova: End of Year Rate of Infl ation, 1994-2003

SOURCE: National Bank of Moldova.

Figure 2.11 Moldova: Quarterly Nominal and Infl ation – adjusted Exchange Rate, 1994-2004 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: National Bank of Moldova.

Chapter 2: Economics of the Transition

74 75 Figure 2.12

Trade Balances as a Share of GDP, 1994-2004

SOURCE: National Bank of Moldova. NOTES: Xpt&Mpt is exports and imports; and Manuf is manufactured commodities

Figure 2.13

External Balances, Annual, 1993-2004 (thru 3rd qtr of 2004)

SOURCE: National Bank of Moldova. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

74 75 Annex 1. A Chronology of Moldova’s External Debt

Amount Lender Purpose Amount Terms Outstanding 10/7/04 Romania 1991 Import of goods 10100 Without interest 0 USA (Program PL-480) 2% - fi rst 7 years; 3% Import of cereals 9854 8212 1992 - after 7 years

FMC, Belgium 1991 Import of agricultural combines 1470 no information 0

European Community Import of goods; cereals, drugs, food for 31590 LIBOR+3% 0 1992 children

Transformed into human- China 1992 Import of goods for children 3758 4% itarian aid 31/8/2001

Romania 1993 Import of oils and fuel 14400 Without interest 9400 European Community Balance of Payments 54000 LIBOR+0.1% 15798 1994 USA (Program PL-480) 2% - fi rst 7 years; 3% Import of cereals 9904 8666 1993 - after 7 years 88900 Russia 1993 Reorganization of debt LIBOR+1% 0 Restructured on 17/9/97 22955

IBRD Two 60000 CQB+0.5% variable 45143 Rehabilitation Rehabilitation 1993 25823 from 4.85 to 7.8% 13187

USA (Program PL-480) 2% - fi rst 7 years; 3% Import of cereals 19869 18213 1994 - after 7 years IBRD (SAL) 1994 Structural Adjustment 60000 CQB+0.5% (1.37%) 47022 4.5 LTPLR variable from Japan 1994 Rehabilitation 39933 25035 6.7 to 3.5% Promotion of export of Moldovan wines 17933 5249 EBRD 1994 LIBOR+1% “Vininvest” LTD 11543 3857 Plafond “C” variable from AKA Germany, 1994 Equipment for pharmaceutical complex 7922 6101 7.5 to 6.5% Import of technological equipment for Sparex AG, 1995 4605 7.9% 0 production of glass Plafond “C” variable from AKA Germany, 1995 Equipment for production of glass 6867 1513 6.5 to 7.87%

Financing the private sector “Moldova 12000 0 EBRD, 1995 LIBOR+1% Agroindbank” JSC 6960 0

Financing the energy sector “Termocom” 4649 2534 Republic of Moldova: EBRD,Economic Policies for Growth, 1995 Employment and Poverty Reduction 8.95% JSC 4698 2560

Chapter 2: Economics of the Transition

76 77 Annex 1 (continued) Amount Lender Purpose Amount Terms Outstanding 10/7/04 USA (Program PL-480) 2% - fi rst 7 years; 3% Import of cereals 9599 8799 1995 - after 7 years German Bank Kreditanstalt, I part, Medical Equipment 6483 LIBOR+0.875% 0 Restructured on 7/6/2002 1995 Commerzbank, LIBOR+0.4%; Financing the budgetary defi cit 15000 0 Germany, 1995 LIBOR+0.5% European Community Budgetary Defi cit 19000 LIBOR-0.125% 10937 1996 USA (Program PL-480) 2.25% - fi rst 7 years; Import of cereals 12472 11979 1996 3.25% - after 7 years Russia 1996 Payment for the import of coal, gas 15000 8.156% 0 Restructured in 1999-2000 IBRD (I agricultural Development of agricultural sector 9826 LIBOR+0.5% 8476 project) 1996

IBRD (PSD I) 1996 Private Sector Development 32250 LIBOR+0.5% 28899

German Bank Kreditanstalt, II part, Medical Equipment 6151 LIBOR+0.875% 0 Restructured on 7/6/2002 1996 EBRD 1996 Rehabilitation of roads 11240 LIBOR+1% 6837 IBRD (Energy Project) CQB+0.5% (approx. Reform of energy sector 9165 7786 1996 1.37%)

Dresdner Bank, 1996 Financing the budgetary defi cit 10400 LIBOR+2% 0

Commerzbank, 1996 Financing the budgetary defi cit 15000 LIBOR+0.4% 0

Merrill Lynch, 1996 Financing the budgetary defi cit 30000 LIBOR+2% 0

Gazprom’140, 1996 Paying back the debt to Gazprom 140000 7.5% 0

91676 - Moldova; 30388 - Russia 1996 Restructuring the debt from 1992-1993 91676 LIBOR+1% Transnistria

Import of equipment for processing sugar Plafond “C” varies from AKA Germany, 1996 3821 2523 beets 6.0 p.a. to 6.5%

Dresdner Bank AG, Financing import of petroleum products 23610 LIBOR+1.75% 0 1996

IDA (SAC-II) 1997 Structural Adjustment 45163 0.75% 48493

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction IBRD (SAC-II) 1997 Structural Adjustment 25000 LIBOR+0.5% 23377

Chapter 2: Economics of the Transition

76 77 Annex 1 (continued) Amount Lender Purpose Amount Terms Outstanding 10/7/04

EUROBOND 1997 Financing budgetary defi cit 75000 9.875% 32740

Hewlett Packard 1997 Electronic equipment for schools 18426 6% 0

Import of technological equipment for Drezdner Bank AG, LIBOR+2%, 6% p.a. construction of pharmaceutical complex 3107 2431 1997 - after restructuring (“Perfuzon” JSC)

Procurement and installation of TMC Padovan, Italy, technological lines for wine processing 1980 8.5% 894 1997 “Chirsova” JSC

Rehabilitation of Chisinau City water EBRD, 1997 21021 LIBOR+1% 16725 supply, “Apa Canal Chisinau” JSC Procurement of microbuses “Taxi-Service” Hunday, Korea, 1997 620 10% 0 JSC Financing of the project on petroleum EBRD, 1997 15296 LIBOR+1% 17985 terminal

Equipment and payment for services in Plafond “C” Varies from AKA, Germany 1997 9366 10235 production of ceramics 5.75 to 6.5%

IBRD (Educational CQB+0.75% Variable Development of education 9745 9010 Project) 1998 from 5.75 to 1.36% German Bank Kreditanstalt, III part, Medical equipment 6103 LIBOR+0.875% 0 1998 IDA (Educational Development of education 4682 0.75% 5143 Project) 1998

IDA (PSD-II) 1998 Development of managerial capacity 7379 0.75% 8131

IDA (Cadastre) 1998 Development of land surveys 12690 0.75% 13790

IDA (Rural) 1998 Rural Development 4905 0.75% 5418 Tureximbank 1998 Water supply 15000 LIBOR+2.25% 8333

Institutional restructuring of “Moldova- EBRD 1998 8000 LIBOR+1% 0 Agroindbank” JSC

Financing the project on modernization of EBRD 1998 8843 LIBOR+1% 4960 Chisinau airport Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

78 79 Annex 1 (continued) Amount Lender Purpose Amount Terms Outstanding 10/7/04 IDA (Moldovan Social Investment Fund) Development of social services 14231 0.75% 16017 1999

IDA 1999 Administration of social protection 3942 0.75% 4396

IDA (SAC II) 1999 Structural adjustment 44390 0.75% 49375 63000 7.5% 0 Gazprom’90 2000 Payment back of Gazprom debt 27000 7.5% 0

Rural fi nancing and development of small IFAD 2000 6291 0.75% 6690 enterprises

IDA, Health Care, 2000 Restructuring of health care system 6961 0.75% 7137

Kreditanstalt fur Restructuring 0 EURIBOR+0.875% 16229 Wederauf, 2002 Kreditanstalt fur Promotion of micro, small and medium 2572 0.75% 2734 Wederauf, MEC 2002 enterprises Kreditanstalt fur Wederauf, Sudzucker Creation of agro-technical service centres 320 0.75% 343 2002

IDA (SAC III) 2002 Structural adjustment 10648 0.75% 11903

IDA (Rural) 2002 Rural development 10490 0.75% 11119

IDA (Energy II) 2003 Development of energy sector 1378 0.75% 1369

IDA (Trade&Transport), Development of transport and tourism 526 0.75% 520 2003

IDA (Water) Water supply 525 0.75% 540 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 2: Economics of the Transition

78 79 C h a p t e r 3 : Public Finance for Growth and Poverty Reduction

3.1 Introduction it could impart a signifi cant stimulus to economic growth, because appropriately This chapter examines the role of public designed public investment can ‘crowd fi nance in supporting a development in’, or stimulate, private investment. Pub- strategy for Moldova that would accele- lic investment can also be used to focus rate economic growth, generate employ- more resources on economic services that ment and direct resources toward poverty disproportionately benefi t poor house- reduction. Although the Government of holds. This would be the case, for exam- Moldova has already succeeded in redu- ple, for public investment in irrigation and cing its external debt burden, the fi scal rural roads. space it has in which to manoeuvre will remain limited until it is able to refi nance its remaining multilateral and bilateral Paris Club debt. In order to do so, it needs 3.2 The Mobilization of greater cooperation from its multila teral Domestic Revenue and bilateral creditors, as discussed in Chapter 1. Because of its own eff orts to Mobilizing tax revenue is crucial for im- reduce its external debt, the Government plementing an investment led Growth is in a stronger position when it begins and Poverty Reduction Strategy. Howe- negotiations with these creditors. ver, revenues of the consolidated budget were 24.2 percent of GDP in 2003, down Independently of its debt problem, the by about twenty fi ve percent from their government has some fl exibility and the highpoint of a third of GDP in 1997. ability to expand its range of options. In Moreover, they are projected to decline part, its success hinges on its ability to further, to 22.5 percent of GDP in 2007. mobilize domestic resources. It has al- Expenditures in the consolidated budget ready done a great deal to reform its tax have been reduced even more drastically system, and more can be done to improve than revenues, from 40.5 percent of GDP revenue generation. Some of the major in 1997 to 22.7 percent of GDP in 2003, a recommendations of this chapter focus forty percent decrease. During 2000-2002, on ways to increase revenue and make the the budget defi cit ranged between zero structure of taxation more progressive. and minus one percent of GDP. In 2003 the budget ran a surplus of 1.6 percent This chapter also makes recommenda- of GDP. Thus, the Government has been tions on the re allocation of public expen- implementing a relatively tight, if not con- ditures. It argues that the Government tractionary, fi scal policy in recent years. should assign a larger share of public expenditures to economic services, and to At the same time that revenues are expect- public investment in economic infrastruc- ed to decrease, contributions to social and ture in particular. Since much of the coun- medical insurance are programmed to in- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction try’s infrastructure has deteriorated, such crease. Contributions to state social insur- investment is badly needed. Moreover,

Chapter 3: Public fi nance for Growth and Poverty Reduction

80 81 ance rose from 6.4 percent of GDP in 1999 tive to streamline and rationalize the re- to 7.3 percent in 2003, and are projected venue system. These revenues arise from to increase steadily, reaching 8.3 percent an amalgam of various charges, such as of GDP in 2007. Contributions to medical road fees, charges for using state owned insurance, which were introduced in 2004, trademarks, fees for extracting natural re- are also projected to increase by 2007.50 sources, and payments for licenses. They The rise in total contributions to the two also include net income from the National insurance funds projected for 2007 con- Bank, which is variable from year to year. trasts with the decline in other forms of However, if non tax revenue is declining, tax revenues. then tax revenue should be boosted to compensate for this decline. Revenues for social protection are increasing but not revenues to fi nance Like other countries, Moldova has been other essential public functions, such as advised to increase its reliance on indi- providing essential public services and in- rect taxes, mainly the Value Added Tax. frastructure. Moreover, if contributions to Indirect taxes now account for about such funds are excessive, they can create forty two percent of all state revenue. a disincentive for employers to hire labour However, according to the Medium Term because they incur higher direct labour Expenditure Framework (MTEF), indirect costs. In any economy in which there is taxes are projected to decline as a ratio to widespread underemployment of labour, GDP, from 15.2 percent in 2003 to 13.8 per- such a system might be inadvisable. This is cent in 2007. The revenue generated from especially the case if more of the fi nan cing the Value Added Tax, which accounts burden is placed on workers. A preferable for about two thirds of all indirect taxes, alternative is often to rely on general tax should be on the rise if it is an effi cient revenue to fi nance social protection. form of taxation, but its revenue is projected to increase only marginally. These Increases in expenditures for social pro- trends place a greater burden on direct tection can be pro poor and are certainly taxes to raise revenue. needed. However, as long as the combination of debt servicing and declines in tax Value added taxes are often recom- revenue continues, public investment and mended as a substitute for trade taxes. But the provision of essential public ser vices a recent study of VAT taxation and trade will be constrained, as will the growth liberalization (Baunsgaard and Keen 2004) of income and human development. fi nds that value added taxes have been Sustained growth of the economy is criti- successful in compensating for the loss of cal for the fi nancing of increased social trade taxes, as a result of trade liberaliza- protection. Although increases in current tion, only in high income countries. Mid- expenditure, such as on social and medi- dle income countries have been able to cal insurance, can stimulate aggregate de- compensate for only about thirty fi ve-fi fty mand, they do not expand the productive fi ve percent of the revenue lost from trade capacity of the economy. liberalization. The most troubling fi nding, however, is that indirect taxes in low In 2004, non tax revenues represented income countries have recovered none of about eighteen percent of all revenue. the revenue lost from trade liberalization. However, they are projected to decline, from 6.4 percent of GDP in 2004 to 5.7 percent in 2007. The Ministry of Finance The incidence of the VAT in Moldova has regards the decline as a desirable objec- shifted increasingly towards imported

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 50 If social and medical insurance contributions are added to the calculations, the total revenue to GDP ratio was 31.5 percent in 2003, and projected to be 31.9 percent in 2007.

Chapter 3: Public fi nance for Growth and Poverty Reduction

80 81 goods and away from domestically pro- import tariff (computed as a weighted duced goods. In 2003, VAT revenue on mean) was 3.9 percent in 2001 (World imports was 8.7 percent of GDP, while Development Indicators 2004, TableTable 6.6).6.6). the revenue from domestically produced By comparison, the weighted mean tariff goods was 4.3 percent. This was a reversal was 7.3 percent in Romania, 7.5 percent of the situation in 1997, when the ratio in Hungary (for 2002) and 12.4 percent for imported goods was 1.8 percent and in Albania. For the purposes of generat- that for domestic goods 8.8 percent. This ing badly needed revenue, rather than is partly explained by a change in 1999 for trade protection, some import tariff s to a destination method of VAT taxation, could be raised moderately in Moldova which was applied to imports from CIS without violating WTO rules. countries as well as others. The Government has attempted to give Another part of the explanation could the VAT a progressive structure by setting be VAT refunds, which amounted to 2.8 three rates below the standard twenty percent of GDP 2003. Also, compared to percent. These are: eight percent for trade taxes, sales taxation on domestic bread and dairy products, fi ve percent for goods and services are limited in their agricultural and unprocessed products, revenue raising potential in an economy and zero percent for heat, electricity and with a large informal sector. Another drugs. The revenue attributable to the reason for the lack of revenue is that the fi ve and eight percent rates represents no minimum threshold for obligatory regis- more than one fi fth of the total. However, tration of VAT taxpayers has been raised, it is diffi cult to administer diff erential rates purportedly to make the VAT more effi - for the VAT, especially in an economy with cient. A 2003 study of taxation in thirteen a large unrecorded informal sector. transition countries in Eastern Europe, the Caucasus and Central Asia found that in Some tax specialists justify lower tax rates 2000 Moldova had a ten percent eff ective as a means to encourage informal sector VAT rate (i.e., actual VAT collections as a activities to become part of the registered ratio to the VAT base) (Stepanyan 2003, formal economy. But such provisions, by p. 15). This rate compared favourably to themselves, are unlikely to provide a suf- those in other countries, especially Russia, fi cient motivation for such a change. Eco- Azerbaijan and Georgia, but, nonetheless, nomic growth or the reduction of unneces- it had declined from a high of seventeen sary government regulation is likely to be percent in 1998. A similar trend character- as, if not more, important. The lack of eff ec- ized the VAT effi ciency rate (the eff ective tive administration probably implies that VAT rate as a ratio to the statutory rate): it the VAT is not as progressive as its nominal had declined from eighty four percent in structure would suggest. While in theory 1998 to fi fty two percent in 2000. consumption taxes can be made more progressive, the diffi culty of collection makes it Because VAT is levied on imports, foreign is easier in practice to render a tax system trade taxes as such have remained low, more progressive by placing more weight accounting, for example, for 1.7 percent on direct taxes on income and wealth. of GDP in 2003. Part of the reason that VAT and trade taxes together are low is Excise taxes in Moldova represent a sub- the establishment of free trade arrange- stantial share of total tax revenue, between ments with countries in Southeast Europe fi fteen and eighteen percent of the total af- and the CIS. Another part of the reason is ter 2000. But much of the revenue that they that the Government has refrained from generate (over forty percent) comes from raising some import tariff s, even though taxes on gasoline and diesel fuel. Another

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction it is allowed to do so under the WTO ac- thirty percent comes from taxes on im- cession agreement. Moldova’s average ported vehicles, and twelve percent from

Chapter 3: Public fi nance for Growth and Poverty Reduction

82 83 wine and liquor. As a ratio to GDP, excise on personal income and corporate profi ts taxes conform to the patterns of decline of has no discernible impact on increasing other taxes, falling from 4.5 percent in 1997 tax revenue. There is no reason to believe to 3.3 in 2003, largely because of a drop in that savings will respond to the lowering revenue from wine and liquor. Excise taxes of tax rates on personal income because are projected to decline even further in individuals in countries with low incomes coming years, to 2.2 percent of GDP in 2007. or depressed incomes save mostly for pre- Boosting revenue by levying excise taxes cautionary reasons. The study cannot fi nd on luxury or non essential consumption any evidence that business investment items should be a priority of tax policy. The has increased because of lowered rates direct impact on the poor is not likely to be on corporate profi ts although one might substantial if the items that are taxed, such expect at least some modest increase as imported vehicles or diesel fuel, are not (other factors remaining constant). Lastly, directly used or consumed by the poor. in countries where unemployment and underemployment are already high, low- The Government’s strategy for boosting ered tax rates on personal income do not direct taxes is to radically reduce tax rates, induce individuals to supply more labour. on the assumption that such a reduction will help broaden the tax base (see the dis- This study tends to put more emphasis cussion in Chapter 2). Such a ‘supply side’ on simplifying tax systems, reducing un- approach to taxation will not only diminish necessary exemptions or strengthening the progressive structure of the tax system, administration and compliance as a basis but also run the risk of lowering total reve- to raise more revenue. Since tax rates on nue. This is generally recognised: the IMF personal income and corporate profi ts has warned, for example, against reducing have already been substantially reduced the rates for direct taxes unless other forms in many of the transition economies, fur- of revenue can be generated to off set the ther reductions are more likely to have potential losses (IMF 2002, p. 13). negative eff ects on revenue collection.

Various reasons are given for lowering the The government has already lowered the rates of direct taxes: increasing dispo sable tax rate on corporate profi ts. This rate was income, encouraging informal sector a moderate twenty eight percent in 2001, activities to become registered, and cre- down from thirty two percent in 1997. It fell ating a tax haven for foreign investment. to twenty percent in 2004. By 2006, within Lower tax rates can, in theory, have posi- just two years, it is programmed to drop to tive incentive eff ects, but these are not only fi fteen percent; namely, less than half likely to compensate fully for a substantial its 1997 level. This strategy is designed, in loss of revenue, especially from a radical part, to induce enterprises to enter the for- and swift reduction of rates. Such supply mal economy and declare their full profi t in- side experiments have proven themselves comes. It is also designed to make Moldova to be ineff ective in other countries (a no- an attractive tax haven for foreign investors, table example being the United States in as allegedly happened in Ireland. the 1980s and in the early 2000s). Balloon- ing government defi cits are often the im- At the same time that the government mediate and painful result. These neces- sharply lowers the tax rate, it projects in sitate inevitable cuts in social protection the MTEF that corporate tax revenue will and essential public services. increase from 10.8 percent of all revenue in 2004 to 12.2 percent in 2007. However, if fi rms have successfully avoided a 28 The study of taxation in transition econo- percent tax rate, there is no guarantee mies cited above (Stepanyan 2003) fi nds

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction that they will abandon such avoidance at that lowering the top marginal tax rates a lower rate of twenty or fi fteen percent.

Chapter 3: Public fi nance for Growth and Poverty Reduction

82 83 Nor is it certain that foreign investors will This change to very low marginal rates move into Moldova principally because makes the system for direct taxes mark- of low corporate tax rates. More crucial edly less progressive. While low income than the tax rate is the profi t opportu- persons will receive a three percentage nities that the economy is generating, points drop in their income tax rate, me- which in part depend on the ability of the dium income persons will benefi t by a fi ve government to use public expenditures, percentage points drop, and high income especially public investment, to stimulate persons by ten, making the structure the economy. As the economy grows and considerably less progressive in a country corporate profi ts increase, moderate rates with high inequality. In 2003, the ratio of on profi t income, such as twenty fi ve the middle income tax rate to the low percent, should represent no signifi cant income rate was 1.5, and that of the high disincentive to foreign investment. income rate to the low income rate 2.5. In 2007, the fi rst ratio will drop to 1.43 and Part of the problem with collecting corpo- the second to 2.14. rate income tax is the extent of tax exemptions. According to 2003 tax returns, tax These changes will weaken the vertical exemptions amounted to 315 million lei, and equity of the Moldovan tax system. Those foreign investors received about 50 million of taxpayers with a greater ability to pay will this total. About 145 million lei in exemptions enjoy relatively lower taxes. In relative were extended to domestic corporations to terms, lower income taxpayers will be encourage investment in fi xed assets and worse off . Personal income taxes as a ratio construction. International experience indi- to GDP are already low in Moldova com- cates that such tax incentives usually have a pared to other transition economies. In weak eff ect on private investment, and serve 2000, this percentage was 1.5 whereas the mainly to erode the tax base. unweighted average for other transition economies was over twice as high, i.e., 3.2 percent (Stepanyan 2003, p. 16) The reforms projected for the personal income tax are much the same as for the corporate income tax. In 2003, personal If the government reduces its highest income up to 12,180 lei was taxed at a ten marginal tax rate for personal income, this percent rate, income in the range of 12,180 rate will be well below the comparable 16,200 lei at fi fteen percent, and higher rates for most other transition economies income at twenty fi ve percent. The top in Central and Eastern Europe and the CIS rate refl ects a reduction from the previous (Table 3.2). For example, in 2003 the high- high of thirty two percent in 1997. The in- est marginal tax rate in Bulgaria was twen- tention is that by 2007 the tax rate for low ty nine percent, in Romania forty percent, incomes will fall to seven percent, for me- and in Slovenia fi fty percent. Only in dium incomes to ten percent, and for high Bulgaria would the highest rate be within incomes to fi fteen percent (see Table 3.1). twenty percentage points of Moldova’s. In Western Europe the marginal tax rates tended to be even higher: forty fi ve per- Table 3.1: Personal Income Tax Rates and Targeted Changes Income Level Tax Rate Income Level Tax Rate Percentage Point 2003 2003 2007 2007 Drop Low Income 10 Low Income 7 3 Middle Income 15 Middle Income 10 5 High Income 25 High Income 15 10 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: MTEF 2004.

Chapter 3: Public fi nance for Growth and Poverty Reduction

84 85 cent for Italy, forty nine percent for Ger- Table 3.2: Highest Personal and Corporate many, and fi fty nine for Denmark (Table Marginal Tax Rates by Country, 2003 (per- 3.2). Compared to other countries in the centages) region, Moldova has programmed a potentially counter productive ‘supply side’ Country Personal Corporate experiment with its tax system. Moldova 2004 25 20 Table 3.2 also indicates that the highest Moldova corporate tax rate in Moldova in 2003 was Projected 2007 15 15 already the lowest of the thirteen coun- Bulgaria 29 24 tries listed (2003 is the last year for which Poland 40 27 tax rates are comparable across countries). In recent years, many countries have been Romania 40 25 under pressure, or been strongly advised, Slovak Republic 38 25 to lower tax rates on businesses. There are Slovenia 50 25 countries that have explicitly chosen the Ukraine 40 30 path of low corporate tax rates in order to attract foreign investment. Ireland has Belgium 50 39 a tax rate of sixteen percent on corporate Denmark 59 30 income; Lithuania and Macedonia FYR Germany 49 27 fi fteen percent. However, the government Greece 40 35 should carefully study these experiments in order to determine whether Moldova Italy 45 34 enjoys some of the same advantages as Netherlands 52 35 Ireland, and how successful some of the transition economies, such as Lithuania, SOURCE: World Development Indicators 2004, have been. Table 5.6.

The Stepanyan study found that tax reve- If Moldova joins the European Union, it will nue from corporate profi ts in Lithuania fell need a tax system that generates revenue from a high of 5.8 percent of GDP in 1992 that can fi nance public expenditures that to 1.9 percent in 1996, and fell further to are comparable to the EU norm. As poin- only 0.7 percent in 2000 (Stepanyan 2003, ted out in the previous chapter, Moldova’s p. 18). In 1991, the rate on corporate prof- entry into the EU would be undermined its in Lithuania was twenty nine percent if it were viewed by existing members as and in 2000 it was reduced to twenty a tax haven to which companies would four percent. Exemptions, such as those relocate. In 2000, general government tax on foreign investment, were also part of revenue was about twenty two percent of the problem. This same study found that GDP in Moldova (Stepanyan 2003, p. 20). in Moldova, such tax revenue fell from a This ratio compared favourably to per- high of 4.9 percent of GDP in 1995 to 1.5 centages in other low income countries percent in 2000. The unweighted average such as Kyrgyz Republic (about twelve percentage for all countries in this study percent), Georgia (fourteen percent) and was 2.5 percent in 2000. Azerbaijan (14.5 percent). But the unweighted average percentage for OECD countries was about thirty eight percent. Moldova’s strategic objective should be to reach this level of taxation, not reduce it to the level of other low income transition economies. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 3: Public fi nance for Growth and Poverty Reduction

84 85 Reducing the corporate income tax rate large farming enterprises. The govern- further, to fi fteen percent, would place it ment has considered raising land taxes well below those prevailing in most other as a substitute for other forms of taxes, European and CIS countries. Moldova such as on farm incomes or agricultural could probably return its highest mar- products. This policy should be evaluated, ginal corporate tax rate back up to twenty in its own right, as an effi cient means of fi ve percent, maintain its highest marginal raising revenue and shifting more of personal income tax rate also at twenty the incidence onto richer households. fi ve percent, and still have a tax system Land taxes are typically a more effi cient relatively less progressive than those in method for raising revenue than taxes on most other countries in the region. farm incomes, which are more diffi cult to measure. The challenge facing the Government is that direct taxes (excluding social and In summary, the government should con- medical insurance contributions) have centrate on raising more tax revenue and fallen as a percentage of total revenue doing so with a more progressive struc- since 1997. While corporate income taxes ture, instead of risking loss of revenue as accounted for 8.3 percent of total revenue a result of making the tax structure less in 1997, they fell to 7.7 percent in 2003. vertically equitable. To this end, the gov- Similarly, personal income taxes fell from ernment could initiate a concerted eff ort 9.6 percent to 8.3 percent between 1997 to strengthen the effi ciency of the VAT, and 2003. Given these declines, it is ill which it has chosen as the mainstay of its advised to lower tax rates on personal tax system, and ensure that its impact is and corporate incomes. This report re- indeed progressive. In order to guarantee commends that the 2003 structure for the buoyancy of the tax system as average personal income taxation (namely, rates per capita incomes rise, the government of ten, fi fteen and twenty fi ve percent) should maintain its current rate structure be maintained, and that corporate profi t for personal income and corporate profi t taxes return to twenty fi ve percent. This income. Eventually, the rate on corporate would contribute to a sustainable fi scal profi t income should return to twenty fi ve structure consistent with growth and percent to make it consistent with the top poverty reduction. rate for personal income.

Wealth taxes, such as real estate and land In the name of making the tax system taxes, have also declined. Together, real more efficient (by lowering rates), some estate and land taxes fell from 1.3 percent orthodox tax reformers have sacrificed of GDP in 1999 to 0.7 in 2003. As the econ- progressivity and, in the process, weak- omy grows and asset values increase, re- ened prospects for raising revenue. venues from wealth taxes should increase But the radical version of such advice in importance. This depends on better comes from ‘supply side’ economists registration and more market based who believe; without much empirical valuation of property, which the govern- evidence, unfortunately, that lowering ment is already attempting to institute. tax rates on personal income and cor- This also depends on raising property tax porate profits will lead to a dramatic rates. Such an eff ort could make the entire increase in revenue. The justification is tax system more progressive, since prop- that the tax base will be broadened if erty taxes disproportionately aff ect richer tax rates are lowered, especially for high households. income earners. However, available evidence suggests that the base for taxa- Options worthy of serious consideration tion is more likely to be broadened by

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction are concerted eff orts to register and tax sustained and broad based economic urban real estate and raise land taxes on growth and improved tax administra-

Chapter 3: Public fi nance for Growth and Poverty Reduction

86 87 tion than by the lowering of corporate Table 3.3: External State Debt, 1992–2004 and personal income rates. Moldova’s (Million US$) stated intention to radically reduce such rates is courting trouble. Year Total Debt Year Total Debt 1992 16.5 1999 661.9 1993 188.5 2000 781.3 3.3 External and Internal 1994 345.1 2001 699.8 Debt Burden 1995 401.3 2002 724.3 1996 540.2 2003 751.4 The government of Moldova continues to 1997 709.1 2004 660.7 bear a substantial external debt burden 1998 719.6 that constrains its ability to allocate public revenue to stimulate growth and reduce SOURCE: Government of Moldova, Ministry of poverty. The external debt is a tragic Finance legacy of its early transition period, when its average income dropped precipitously In the wake of government’s eff orts, mul- to the level of a low income country while tilateral institutions now hold sixty three international fi nancial institutions did not percent of Moldova’s external debt, and off er it concessional terms for lending (see bilateral lenders thirty three percent. Chapter 2). The remaining fi ve percent is commercial debt. Moldova will continue to bear In 1992 the external state debt was only a substantial debt burden as long as the US$16.5 million; but by 1997 it had bal- International Monetary Fund does not looned to over US$ 709 million (Table reach agreement with the government 3.3). It reached a peak of US$ 781 million on a Poverty Reduction and Growth Fa- in 2000. Since 2000, the government has cility (see Chapter 1). Without such an worked to reduce this burden. For exam- agreement, the government cannot be- ple, it bought back debt owed to Rus- gin negotiations with the Bretton Woods sia’s Gazprom and the Hewlett Packard Institutions on its multilateral debt nor Corporation. At the same time, it sought start negotiations on its Paris Club bila- to reschedule commercial credits held teral debt. Endorsement of a nationally by banks and the remaining Eurobonds, owned Poverty Reduction Strategy Paper and reschedule the debts owed to bilat- should be a suffi cient condition for such eral creditors that were not part of the an agreement. Paris Club. As a result, external debt fell to about US$ 661 million in 2004, which The government has secured temporary was 25.7 percent of Moldova’s GDP. breathing room by accumulating debt arrears and rescheduling its Eurobond The government paid a high price by allo- debt. For full and timely payment in 2002, cating funds to debt reduction that could it would have been necessary to allocate have been used to stimulate economic sixty two percent of total public revenue growth and foster poverty reduction. to debt servicing; but actual allocation Because there were no real choices, this was twenty eight percent. It should be trade off was probably unavoidable. The clear that a burden of over sixty percent government can be congratulated on was unacceptably large given the other its success in reducing the debt to more legitimate claims on public revenue. In manageable proportions but its external 2003, full debt servicing would have taken lenders could have facilitated this process half of public revenue, in contrast to the by off ering a substantial degree of debt twenty three percent that was actually Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction rescheduling. allocated. As a result, total arrears to ex-

Chapter 3: Public fi nance for Growth and Poverty Reduction

86 87 ternal creditors rose from US$ 3.4 million these countries qualify as ‘severely indebt- in 2000 to US$ 69.9 million in 2003. Since ed’ under the HIPC defi nition. Only Kyrgyz its multilateral and large bilateral credi- Republic, with a debt to export ratio of 221 tors refused to renegotiate their loans, the percent, would qualify. The debt problems government had no alternative to accu- of this group of low income countries mulating arrears, unless it had instituted have received scant attention. Despite draconian cuts in social expenditure. promises, the international development agencies have not launched initiatives to Once negotiations on the multilateral signifi cantly relieve the debt burdens of debt begin, the government has a strong these transition countries although such case to argue that such debt, mistakenly an initiative should be a priority. set on IBRD terms, should be refi nanced to refl ect more favourable IDA terms. One Faced with the policies of multilateral and 2001 estimate calculated the savings on bilateral creditors, the Moldovan govern- interest payments alone at US$ 157 mil- ment was forced to fi nd its own solutions. lion (Olortegui 2001). The government It managed to pay off its trade credits, can also seek rescheduling of its bilateral renegotiate its Eurobond debt, and con- Paris Club debt and use the new terms clude agreements on bilateral debts with for this debt as a basis to renegotiate its Romania and Turkey. It is not without irony remaining commercial debt. that such countries have been more generous in debt relief than much richer countries. The government also bought back its Table 3.4: Debt to Export Ratios of Selected RAO Gazprom debt owed to the Russian Low Income Transition Economies Federation, retiring in 2004 an outstanding debt of US$ 140 million, and bought back its debt to Hewlett Packard, retiring Debt an outstanding debt of US$ 11 million. Country Debt/Exports Classifi cation Despite the absence of support from Armenia 111 L international fi nancial agencies, the Georgia 144 M Government has succeeded in making Kyrgyz Rep 221 S its total debt burden, both external and domestic, more manageable. In 2004, its Moldova 126 M total state debt represented about thirty Mongolia 107 M seven percent of GDP, down from fi fty Tajikistan 124 S seven percent in 2002 (Table 3.5). The de- Uzbekistan 136 M cline in external debt, from 44.4 percent in 2002 to 25.7 percent in 2004, accounted SOURCE: World Development Indicators 2004, for almost all of the fall in total debt. Table 4.17. ‘L’ (less indebted); ‘M’ (moderately indebted); ‘S’ (Severely Indebted). Table 3.5: Total State Debt, External and In present value terms, Moldova’s exter- Domestic (% of GDP) nal debt represented 126 percent of its exports in 2002 (Table 3.4). This places it 2007 in the ranks of other low income transi- Type of Debt 2002 2003 2004 tion economies (Armenia, Georgia, Kyrgyz (Proj.) Republic, Mongolia, Tajikistan and Uz- Domestic Debt 12.5 10.6 11.6 8.3 bekistan) that have a substantial external External Debt 44.4 36.0 25.7 25.9 debt burden. While such debt burdens Total Debt 56.9 46.6 37.3 34.2 severely hamper a country’s ability to ac- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction celerate economic growth, almost none of SOURCE: MTEF 2004 and Ministry of Finance.

Chapter 3: Public fi nance for Growth and Poverty Reduction

88 89 Once discussions with its Paris Club bilateral Receipts from privatisation have not been, creditors and its multilateral creditors begin, and will not be, adequate to reduce the the government should be in a stronger po- domestic debt. There are few state owned sition than previously to negotiate debt relief assets to be privatized; and, in any case, it or refi nancing. Russia, a Paris Club member, is not clear whether privatization would is Moldova’s largest creditor, followed by the be the best policy choices. Another op- United States. Germany, Italy and Japan hold tion, which the government has begun to relatively small outstanding loans. Some re- pursue, is to lengthen the maturity of its lief on multilateral debt is appropriate since domestic securities by issuing bonds with Moldova was forced to accept commercial 2-3 year maturities instead of Treasury Bills, terms on its borrowing when it should which have maturities up to one year. have received concessional terms. Moldova became eligible for concessional terms only In late 2004, there was an unexpected in 1997, well after its deep economic crisis. surge in demand for Government T bills, Some multilateral institutions have argued perhaps due to remittances, even though that earlier in the 1990s the Moldovan gov- their interest yield was below the infl ation ernment did not want concessional terms rate. As long as the maturities of securities for their loans. This story is indeed diffi cult lengthen and the demand for them main- to believe. Even if it were true, this does not tains itself, servicing the domestic debt absolve the same multilateral institutions, as should be manageable. The problem is that public institutions, of the responsibility of such fi nancing is being used, in eff ect, to having provided the government with bet- reduce the external debt instead of fi nanc- ter technical advice on the kind of loans that ing public investment; in other words, the it should have assumed. policy of increasing indebtedness without asset creation continues in another form. In order to fi nance its budget gap in the 1990s, which was generated primarily by In the mid 1990s, the government had to external debt servicing, the government face a signifi cant external debt problem that had little alternative but to increase do- was caused by a rise in the cost of energy mestic debt. This causality, namely, the ex- imports from Gazprom, a privately owned ternal debt burden forcing the shoulder- Russian company. This debt had its origins ing of more domestic debt at high interest in 1994 when Gazprom raised its gas prices rates, emphasises the problematic nature to world levels, and began charging more of the original IBRD loans, since they did for its gas supplies to Moldova Gas, which not create the assets that could generate is also a private company. Because the go- an income stream suffi cient to enable the vernment refused, on equity grounds, to government to repay them (see Chapter raise user prices to households and also to 2). The government has subsequently some productive activities, it began to incur reduced the domestic debt, from 12.5 large losses. As a result, the government percent of GDP in 2002 to 11.5 in 2004, had to assume the debt owed to Gazprom and projects a further reduction to 8.3 (which rose to 11.2 percent of GDP in 1994). percent in 2007 (Table 3.5). This would be This debt increased alarmingly in 1998 and accomplished by reducing security issues, 1999. However, the government bought lengthening the maturities, and repaying back this debt in 1999, paying US$ 36.4 outstanding credits to the National Bank million for US$ 140 million worth of outstan- of Moldova. Domestic debt became a seri- ding debt stock. When this debt emerged ous problem during the Russian fi nancial again after 1999, the government paid US$ crisis when the government was forced 50.1 million in 2004 in order to retire the new to borrow from the central bank in order outstanding debt stock of US$ 114.5 million. to fi nance the expenditures that were not

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction covered by revenue. In 1998, the domestic In summary, the government has managed debt hit a peak of 17.2 percent of GDP. to bring its external and domestic debts

Chapter 3: Public fi nance for Growth and Poverty Reduction

88 89 under control. It has done so at a high cost, of both to growth and poverty reduction. It one that could have been substantially les- would benefi t the Moldovan people be- sened had offi cial lenders been more forth- cause the funds that have been earmarked coming with assistance. The Bretton Woods for debt servicing could have been released Institutions could have easily been more to fi nance badly needed public investment generous in their dealings with Moldova. in economic and social infrastructure and As noted in Chapter 1, they delayed ap- the provision of essential public services. proval of the Economic Growth and Pov- erty Reduction Strategy until late 2004, and when the endorsement came it did not allow for access to key concessionary in- 3.4 The Composition of struments. Should full support be received, especially from the IMF, the government Public Expenditures could seek reduction and rescheduling of the external debt owed to multilateral in- As stated above, the government of Moldo- stitutions and Paris Club bilateral creditors. va has implemented a relatively tight fi scal Given the recent history, the mutual lack of policy, with a defi cit near zero during 2001- confi dence between the government and 2002. In 2003, it ran a surplus of 1.6 percent its multilateral creditors certainly cannot be of GDP, and it is projected to run a surplus attributed only to the former. of over one percent of GDP in 2004. Taking into account expenditures for the social in- Since the government has endured much surance and medical insurance funds, total of the pain in reducing the commercial public expenditures were, according to the and non Paris Club bilateral components MTEF, 33.8 percent of GDP in 2003. The ratio of its external debt, it should be in a of current expenditures to GDP rose during stronger position to lobby for favourable 2000 2003, to a peak of 27.2 percent (Table terms for rescheduling the remaining 3.6), and is projected to rise further to 29.6 part. The starting point for such nego- percent in 2007 (Table 3.7). Conversely, tiations is the general recognition that the ratio of capital expenditures to GDP is Moldova should have been provided slated to decline from 5.1 percent in 2000 with concessional lending, instead of non to three percent in 2007 (Table 3.7). This concessional lending, beginning with does not bode well for sustained economic the fi rst loans it contracted in the early growth in Moldova, because well defi ned 1990s. Thus, the accumulated interest public investment is a powerful stimulus to rate diff erential between the two should private investment. Moreover, if it is too low, immediately be forgiven. In addition to private investment tends to be depressed. the interest rate diff erential between IBRD and IDA terms, the maturities of the non Part of the problem is that one third of concessional debt should immediately be public capital expenditures have been rescheduled. Houston terms for resche- fi nanced by external grants. But Moldova duling, which are hardly generous, would has no Public Investment Programme to involve, for example, stretching the debt coordinate these investments. In order to over 20 years with a 10 year grace period. advance national development priorities, The result would be that the net present public investment should be reliant on value of Moldova’s debt should drop sub- the mobilization of domestic revenues. stantially, greatly facilitating the prospects This depends in turn on the mobilization for growth and poverty reduction. of domestic savings, which remain very low in Moldova. Beyond revenue genera- Reaching a settlement would be in the in- tion, there is a more general problem of terests of both the government of Moldova the lack of accumulation of capital, both

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction and its creditors, and serve the commitment private and public. Moldova’s heavy re-

Chapter 3: Public fi nance for Growth and Poverty Reduction

90 91 liance on external assistance to fi nance vestment (Ibid, p. 25), leaving rural areas, public investment contributes to the lack where infrastructure is sorely lacking, with of coherence of public investment, partly a small share of the total. driven, as it is, by domestic priorities and partly by donor priorities. Moldova needs Within current expenditures, the public a consolidated and coordinated Public In- sector wage bill is projected to increase vestment Programme that is more ambi- only slightly, from 7.7 percent of GDP in tious, more geared to investing in growth 2000 to 8.1 percent in 2007. Between 1998 and more focused on poverty reduction. and 2002, there was a fi fteen percent reduction in the number of employees, with Table 3.6: Composition of Public Expendi- eighty percent of this reduction comprising employees in education and health. tures (% of GDP) This is surely not a pro poor approach to Expenditure 2004 public sector reform. However, the Gov- 2000 2001 2002 2003 Category (app.) ernment did raise the wages of these low Total income employees. Nonetheless, overall 36.4 31.4 33.9 33.8 34.9 Expenditures wages in the public sector remain low. Debt Service 6.4 4.2 2.2 2.1 2.9 Other small categories, such as subsidies Net Lending 0.2 0.1 0.3 0.2 0.1 to enterprises, are expected to decline or Current remain roughly constant in coming years. 25.1 23.8 27.1 27.2 28.1 Expenditures Two categories expected to increase are Capital the general expenditures on goods and 5.1 3.6 5.0 4.6 4.0 Expenditures services and social transfers. Expenditures for medical insurance went from zero in SOURCE: MTEF 2004. 2003 to 3.5 percent of GDP in 2004. This is the main explanation for the rise in ex- Not all of the capital investment that is penditures on goods and services from 5.9 allocated in the budget ends up being percent in 2002 to 8.5 percent in 2007. The disbursed. In 2003, only sixty one percent projected increase in social transfers from of allocated public investment funds were 8.8 percent of GDP in 2000 to 11.0 percent disbursed (IDIS Viitorul 2004, p. 25). A in 2007 is explained by the rise in its major further problem is that Chisinau receives subcategory, social insurance expendi- about seventy percent of all capital in- tures, from 8.2 percent to 9.7 percent.

Table 3.7: Current and Capital Expenditures (% of GDP)

Expenditure 2004 2005 2006 2007 2000 2001 2002 2003 Category (App) (Proj) (Proj.) (Proj.) Current 25.1 23.8 27.1 27.2 28.1 29.9 29.7 29.6 Expenditure Capital 5.1 3.6 5.0 4.6 4.0 4.0 3.3 3.0 Expenditure Total 36.4 31.4 33.9 33.8 34.9 36.4 35.0 34.3 Expenditure

SOURCE: MTEF 2004. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 3: Public fi nance for Growth and Poverty Reduction

90 91 Juxtaposing the increases in expendi- Table 3.8: Public Expenditures by Sector tures for social and medical insurance (percent of GDP) with the decline in capital expenditures highlights current budget priorities. Budget 2004 2000 2001 2002 2003 The government has been forced to Category (Appr’d) favour social protection and welfare General 5.0 4.9 6.0 5.9 5.8 over growth inducing public investment Services because of the disastrous collapse of Social 19.6 18.1 21.2 21.2 21.3 social provision in the 1990s. The low Services capital expenditure is partly due to a Education 5.7 6.0 6.8 6.7 6.4 projected decline in external grants and Health 3.2 3.2 4.0 4.0 4.5 credits for capital expenditures. How- Social 10.0 8.4 9.6 9.7 9.8 ever, if external assistance increases, it Assistance is likely to favour, in line with common Economic 4.3 3.2 3.7 3.7 4.0 donor priorities, expenditures on health Services and education rather than on economic Agriculture 1.3 0.6 0.8 1.1 1.0 services and investment. The offi cial in- Transp. 0.9 0.5 0.5 0.5 0.8 tention of the government is to maintain &Commun. a constant level of domestic fi nancing Communal 0.9 1.3 1.7 1.3 1.1 of public investment during 2005 2007, Services and direct much of it to critically needed construction activities. However, the SOURCE: MTEF 2004. general downward trend of fi nancing for investment portends an inevitable go- Given the importance of agriculture vernment focus on the short run amelio- and agro industry to future growth and ration of poverty instead of its long run exports, the lack of expenditures on substantial reduction. economic services for these sectors, particularly public investment, is not optimal. An examination of trends in expendi- Moreover, the lack of public investment in tures on social and economic services infrastructure for transport, communica- during 2000 2004 underscores the relation and energy indicates an erosion of the tive neglect of the economic services. basis for long term sustainable growth. Expenditures on both education and Rural and agricultural infrastructure is es- health increased during this period sential to stimulate Moldova’s industrial (Table 3.8). By contrast, expenditures development as well as promote its agri- on agriculture, transport and commu- cultural development. Even compared nication edged downward. Overall, the to neighbouring transition economies, share of social services in GDP increased Moldova is allocating relatively less atten- from 19.6 percent to 21.3 percent while tion to economic services and relatively the share of economic services declined more to social services. While this is due in from 4.3 to four percent. Economic serv- part to the depth of the recession that the ices are programmed to decline further, country has endured, such an allocation to 3.2 percent in 2007, while social serv- of public resources suggests that the full ices will remain virtually constant at 21.4 recovery from such a catastrophic decline percent.51 will be more protracted than anticipated.

51 Public investment trends refl ect the general priorities accorded to social services. In 2003, for example, when public investment was 4.6 percent of GDP, a 13.5 percent share was allocated to agriculture. For the same year, 11.6 percent was allocated to healthcare and 17 percent to education. Hence, together Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction health and education accounted for 28.6 percent of all public investment. Another 11.5 percent went to community services and housing.

Chapter 3: Public fi nance for Growth and Poverty Reduction

92 93 3.5 Poverty, Inequality ment, parliament might pass legislation mandating that such an assessment ac- and the Budget company the presentation of each year’s budget. The poverty assessment could Public fi nance, both expenditure and taxa- include an evaluation of the likely impact tion, can be used as a powerful and eff ec- of the budget on income distribution. tive instrument of poverty reduction in Moldova. The country has the institutions Also important would be a gender impact and the human capacity to use the public assessment of the budget, which would budget to make growth more pro poor. be focused on, but not limited to, the Among the principal constraints, however, impact on poor women. Such an assess- are: 1) the low morale of civil servants, which ment would make more transparent any over time should be raised by improved sal- necessary expenditure and taxation trade aries, and 2) the need for a greater focus by off s involved in directing benefi ts to men go vernment on more pro growth and pro and women. One reason that discrimina- poor expenditures and taxes. tion against women in budgets continues is that the gender impact of expenditures Two concrete measures would facilitate and taxes is not transparent, making pub- pro poor budgeting. First, the Ministry lic debate diffi cult. There are standard me- of Finance could institutionalise a pov- thodologies for such assessments, which erty impact assessment of each annual are mandated for several donors and budget. At the initiative of the govern- lenders in their assistance programmes. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 3: Public fi nance for Growth and Poverty Reduction

92 93 C h a p t e r 4 : Poverty, Inequality and Policy in Moldova

4.1 Introduction the review of existing studies points to a picture observed in several other econo- Poverty in Moldova shares features with mies in transition. Poverty is largely a rural other transitional economies of the region, problem, seventy percent of the total, but has specifi c aspects that require coun- aff ecting agricultural labourers and small try-specifi c analysis and policies. The main farmers. Households in small single-in- fi ndings of the analysis suggest, fi rst, that dustry towns were aff ected even more both income-poverty and human-poverty severely, though their number is compar- sharply increased since 1991. With few excep- atively small. Demographic factors, such tions, they still have not recovered their pre- as gender, old age and disability, were transition levels. While most of the current not dominant causes of poverty, though offi cial and academic literature on poverty they matter for some groups, for example, in Moldova goes back to 1997, a complete families with three or more children. Also, understanding of the trends, causes and poverty diff erentials by level of education responses of the present predicament, and are not signifi cant except for those with of the ground that needs to be recovered, re- higher education. This fi nding, typical of quires placing the poverty problem into the the economies in transition, is not charac- proper historical context, i.e., 1989-2004. teristic of other developing countries with similar income levels per capita. Second, unlike in the developing countries of Latin America aff ected by the debt Fifth, the causes of such huge dete- crisis of the 1980s, but similar to what rioration in well-being between 1991 and observed in other members of the former 2000, and of the modest improvements Soviet Union, the trends in diff erent indi- recorded subsequently, resulted from cators of well-being show a marked co- the collapse during 1992-1996, the sub- variation. This suggests that the increase sequent crisis of 1998-99 and the marked in poverty is of a deep structural and mul- growth rebound that began in 2001. It has tifaceted character. Third, this strong co- frequently been ignored in the poverty lit- variation among indicators of well-being erature on Moldova that the initial rise in shows that poverty in Moldova is com- income poverty resulted from an increase bines 1) low income and consumption in inequality during that period (1991- levels, 2) high mortality and morbidity, 3) 1996, see Chapter 2). Likewise, future declining access to subsidized health and prospects for poverty alleviation would education, 4) growing social exclusion, depend on the ability of policy to foster a 5) limited participation in decision-mak- more equitable pattern of growth. ing, and 6) disempowerment of the poor. The problem cannot be solved only by In light of the multiplicity of the factors policies that focus on reduction of income aff ecting living standards, a serious at- poverty, but requires a broader approach. tack on poverty requires interventions on several fronts, through pro-poor growth Fourth, the poverty profi le that emerges Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction characterized by with greater investment from the analysis of available data and in agriculture (see Chapter 7), higher em-

Chapter 4: Poverty, Inequality and Policy in Moldova

94 95 ployment intensity of growth (discussed During 1991-1996 the country was devas- in Chapter 6), supportive labour market tated by a severe depression (see Chapter policies (again Chapter 6), and the mobili- 2), the collapse of Comecon which pro- zation migrant remittances (Chapter 5). voked a large increase in energy prices, the internal confl ict of 1992, and fl oods in 1992 and 1994 that had a severe eff ect on agricultural and food processing. This 4.2 Trends in Poverty period was characterized by a sharp and protracted decline in GDP per capita and Moldova experienced poverty before the a marked increase in income inequality introduction of the market reforms. Be- (again, see Chapter 2). Although no precise cause of full-employment, an egalitarian information is available, one can tentative- distribution, and price stability during the ly estimate that as a result of the economic socialist era, income-poverty was low and collapse,52 hyperinfl ation and a sharp rise mainly aff ected marginal groups. Atkin- in inequality, the poverty headcount ratio son and Micklewrigth (1992) suggest that (PHR) rose during this period, to reach in 1989 the proportion of workers with a sixty-fi ve to seventy percent by 1995, and wage below the ‘offi cial minimum’ was possibly even higher levels during the ear- about fi fteen percent. Closer inspection ly years of hyperinfl ation. The catastrophic reveals that the situation was less positive collapse of output was responsible for well than suggested by the income statistics, over half of the increase in poverty over because the distribution of real consump- 1991-1996. During this period, however, tion was aff ected by queuing and the poverty rose also because of a marked rationing of most goods, dual distribution worsening of the distribution of income.53 systems, parallel markets and regional dif- The Gini index of the distribution of gross ferences in the supply of consumer goods. earnings rose from 0.250 in 1989 to 0.392 in These factors had a marked inequality ef- 1992, to stabilize at broadly that level over fect on the distribution of consumption 1993-95. Hyperinfl ation, a marked shift poverty and family well-being, though in relative prices and the compression of pre-transition poverty was minor com- public social expenditure also contri buted pared to the situation in the mid-1990s. pushing upwards poverty.

Poverty in Moldova started to rise in 1990- During 1997 and the fi rst half of 1998, 91, just before the offi cial onset of the tran- output recovered slightly (growth of 1.6 sition. Its increase over the subsequent fi f- percent in 1997), while inequality reached teen years is indisputable, though the level a Gini coeffi cient of .42. Though no di- and trend for 1989-96 can be estimated only rect data are available, indirect estimates approximately, through indirect me thods. (based on the method in footnote one), Despite a severe deterioration in living it is likely that during these two years po- conditions during 1992-1996, poverty was verty declined, from about sixty percent recognized politically only in 1997 when to the low fi fties. The modest recovery the fi rst random HBS was introduced. was reversed during August 1998 to the

52 A tentative way to reconstruct the poverty trend consists in making use of the relation derived from Bourguignon (2002) who showed that any percentage variation in the poverty headcount ration (DPHR/ PHR) can be mathematically decomposed in the weighted algebraic sum of the percentage change in mean income (approximated by GDPc/GDPc) and the percentage change in the distribution of income (approximated by Gini/Gini). The weights (i.e the poverty-growth and poverty-inequality elasticities) can be drawn from Table 3 while the 1997poverty incidence can be used to work out retrospectively poverty incidence in 1992.1996. Valeriu Prohnitski elaborated an interesting alternative method that leads, however, to results not compatible with the observed PPMU fi ndings for 1997-2202. 53 Though the income distribution data based on randomized surveys are available only since 1997, the Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction data on the distribution of earnings (that over 1997-2002 tend to covariate considerably with the income and consumption inequality indexes obtained from income survey data.

Chapter 4: Poverty, Inequality and Policy in Moldova

94 95 end of 1999, when the economy con- ment of the poverty status carried out in tracted by six percent in 1998 and three August 2002. This survey indicated that percent in 1999. This coincided with the sixty-eight percent of households inter- Russian fi nancial crisis, and the ensuing viewed consi dered themselves in poverty, one hundred percent devaluation of the a much higher rate than that estimated by rouble that aff ected Moldavian exports the PPMU (40.4) or the World Bank (48.5). and emigration to Russia. During 1997- These diff erences can be interpreted in 1998 poverty and extreme poverty rose several ways. The fi rst observation is that to over seventy and sixty percent, respec- the poverty lines used by the PPMU and tively (World Bank 2004b, p.80). the World Bank are too low, and therefore underestimated the real level of poverty as After 2000, a sustained economic reco- perceived by households. A second expla- very fi nally began. Inequality declined nation is that the interviewees associate somewhat between 2001 and 2002, and the term ‘poverty’ with a ‘low income sta- markedly in 2003, as Table 4.1 shows. The tus’ that diff ers from the PPMU and World 2003 improvement was characterized by Bank poverty line, that focus only on se- a sharp rise in the income share of the vere deprivation. However, contrary to the bottom quintile, despite the decline of ‘objective’ measures that showed higher agricultural output in the same year (Eco- poverty rates in rural areas, the subjective nomic Statewatch Q2/2004), confi rming poverty responses were higher in urban the favourable eff ect of a broad based areas (sixty six percent of the households) output expansion (6.3 percent) on in- than in rural (fi fty eight percent). This come distribution and poverty. As a result might be explained by the lack of access to of these eff ects, as well as price stability, land by many urban residents, or the result the poverty headcount ratio fell, and at of higher expectations about the minimum the end of 2002 reached forty percent, level of consumption by urban residents. according to PPMU estimates and 48.5 percent according to those by the World Bank. In 2003 the PPMU poverty rate fell Table 4.1: The quintile distribution of dis- to around twenty-six percent, and ex- posable income, 1998-2003 treme poverty to fi fteen percent, showing a poverty alleviation elasticity of growth of over fi ve, a value that warrants some 1998 1999 2000 2001 2002 2003 1st quintile probing even taking into account the 3.9 4.5 4.6 4.3 4.6 5.8 clear improvement in inequality that took (Q1) 2nd quintile place in 2003. As consumption inequality 9.7 9.8 10.4 10.4 10.4 11.2 fell by three points (from 39.5 to 36.6 over (Q 2) 3rd quintile 1999-2002, see Table 4.2), some thirty per- 15.0 14.7 15.1 14.8 15.0 15.5 cent of the poverty decline was induced (Q 3) 4th quintile by the improvements in the distribution 22.6 22.0 21.9 21.7 21.6 21.8 of consumption expenditure. This esti- (Q 4) 5th quintile mate may in part be a statistical illusion 48.8 49.0 48.0 48.8 48.3 45.7 due to the better accounting of food (Q 5) stocks on farms. ratio Q5/Q1 12.1 10.9 10.4 11.3 10.5 7.9 SOURCE: Bulletin Statistic 2004, ‘Aspectele pri- These preliminary conclusions about vind de trai al populatiei’, p.27, DSS. Numbers may the fall in poverty incidence are cast into diff er from those in Chapter 2 due to diff erences question by the results of the self-assess- in sources used. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 4: Poverty, Inequality and Policy in Moldova

96 97 Table 4.2: Inequality for various measures raise the probability of its intergenera- of income (Gini coeffi cients) tional transmission, and give rise to a culture of poverty and dependence. 1997 1998 1999 2000 2001 2002 Income The panel component in the Moldavian inequality 44 44 44 42 43 42 HBS makes it possible to identify the po- (disp. inc) verty dynamics for individual households Consumption 40 40 38 38 36 and, by aggregation, to evaluate whether inequality poverty in Moldova is predominantly Wage 43 44 39 39 43 transient or permanent. During 1997-2002 inequality eighty six percent of the population experienced poverty for at least one year. For SOURCE: author’s compilation; Note: Income and many of them poverty was a transitory consumption inequality (computed on the bases of HBS) are from PPMU (2003), while the wage phenomenon: 15.1 percent of households inequality data are taken from the Transmonee were in poverty for one year, 20.7 percent database. for two, 24.1 for three, and 26.5 percent remained in poverty all four years. This The poverty gap shows even bigger group of persistent hard core poor con- variations over time than the poverty and sisted primarily of families with many chil- extreme poverty rate. In 1999 it reached dren, with a head with less than college thirty percent, very high, observed only in education, or of elderly unrelated peo- severely deprived developing countries, ple living together (World Bank 2004b). then fell to almost one third this value by However, the incidence of year-to-year 2002. Even greater improvements were repeated poverty showed a decreasing observed for the extreme poverty gap, trend: forty-four percent of poor over two indicating major improvements among consecutive years in 1997-98 to forty two those remaining in poverty. While a poin 2000-01, and forty in 2001-02. verty gap of thirty percent may increase the risk of death and higher morbidity The data about widespread ownership among the poorest population groups, of land and dwellings, levels of educa- one of ten to fi fteen is not likely to raise tion and the broad coverage of basic death rates and severe under-nutrition, social services would suggest that a class though it may still reduce access to health of ‘hardcore poor’ had not yet arisen in care, school enrolment and access to Moldova to any great extent. Even the other essential services. 26.5 percent estimated to be permanently poor during 1997-2002 may not From a programme and policy perspec- have been be asset-poor, education-poor tive, it is essential to know whether over and health-poor, if the poverty gap is not time poverty aff ects diff erent or the same large. Some indications to the contrary people. If the latter, the solution to the emerge. For example, in small cities and poverty problem is more complex, since remote rural areas, home appliances and by defi nition it is structural in nature. In- heating systems fell out of order, dwel- deed, the ‘permanently poor’ may adopt lings were not maintained, educational unsustainable coping strategies such as enrolments and health expenditures on the distress sale of assets (land, houses, newborns were reduced, and assets were working animals and consumer durables), sold. The persistence of low income and the withdrawal of children from school, high inequality might cause long term and socially and humanly destructive apathy among the poor, loss of trust on measures such as prostitution, traffi cking the possibility of exiting ‘the vicious circle in organs of the body, and child labour. of poverty’. The rise in remittances since

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction These carry considerable risks, entrench 1999 might have reduced this fatalism and poverty among those who adopt them, the process of poverty entrenchment.

Chapter 4: Poverty, Inequality and Policy in Moldova

96 97 4.3 Characteristics overwhelming majority of rural residents farmed the plots assigned to them by the of Poverty land reform, or hired themselves out as rural labourers on larger farms. The statis- A major feature of poverty in Moldova tics in Table 4.3 show that while the share was its concentration in rural areas and of poor and extremely poor among the small towns, including in 2002, eighty nine rural labourers fell, because of the decline percent of the poor. This concentration of in the rural population, the opposite was poverty in rural areas and towns rose stea- observed for farmers (see Chapter 7). dily, during both the collapse and recovery. In 2002, people living in small towns had a However, among the three main regions 2.2 higher risk of falling into poverty, and a (North, South, Centre), poverty was not 2.9 greater risk of being in extreme poverty uniformly spread. The results show conside- than the residents of the large cities (Table rable diff erences in the incidence of poverty 4.3). The poor were also disproportionately among districts (judetsjudets).). T Therehere i iss e evidencevidence located in small towns characterised by a (PPMU 2002) that the poverty and extreme failed or diffi cult-to-restructure industry. poverty incidence may vary considerably The ratio of the poverty gap between towns also between the cantons (raions) ofof thethe and cities is even larger (fi ve times greater). same judets, wwhichhich wouldwould mmakeake thethe tasktask ofof This meant that towns had proportionately geographical targeting of public expendi- more poor than the cities, and that the poor ture for anti-poverty interventions complex, in towns are poorer than the cities. detailed and expensive. Given this intra- regional and intra-judetjudet heterogeneity,heterogeneity, thethe Poverty in small towns were linked to the identifi cation of the major poverty groups pattern of industrialisation during the Soviet at the national level is not suffi cient for the period, in which a few medium-large indus- design of eff ective poverty alleviation poli- tries were established in each town, which cies. Further eff ort at mapping poverty at resulted a highly un-diversifi ed structure the local level may be required. similar to that found in Western European countries and the United States during the In many countries, the poor are often same period. With market liberalisation, found in ‘incomplete’ families, in families these mono-industries showed limited with unfavourable dependency ratios, in ability to compete, not least because of lack families headed by pensioners, or inclu- of investment for restructuring. In addition, ding the disabled. This is only partly true while the people in the rural areas could to in Moldova, whose demographic fea- some extent sustain their food consump- tures that diff er markedly from those of tion through subsistence agricultural, this other countries with the same income per was not possible for people in small towns. capita. In Moldova, the bulk of poverty is not explained by high dependency ratios. While poverty was most acute in small Table 4.3 suggests that the incidence towns, the majority of the poor lived in of poverty and extreme poverty rose rural areas. The share of population in steadily with the increase in the number the countryside is close to two thirds of of children, and that poverty headcount the total (Table 4.3), and the incidence of ratio is higher than the national average po verty among this group is the second starting with families with two children. highest. For people in rural areas, the rela- The incidence of poverty appears to tive risk of poverty and extreme poverty in fall steadily with the age of the person. relation to the cities was very high, 1.8 and However, some of these results are am- 2.2 in 2002. The causes of poverty in rural biguous, as they were obtained by the areas resulted from three factors. The fi rst World Bank (2004) making use of ‘per was the extremely limited employment capita equivalence scales’.54 This assumes opportunities outside agriculture. The that children, adults and elderly have the

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 54 We could not locate similar data from the PPMU which uses in the computation of the poverty rates the standard OECD equivalence scales

Chapter 4: Poverty, Inequality and Policy in Moldova

98 99 same consumption needs, and that there These two assumptions are not generally are no economies of size in consumption. accepted in the poverty literature.55

Table 4.3. Incidence of poverty and extreme poverty by main socio-economic characterstics

Share Poverty rate Extreme poverty rate popul. Share of poor Share of extremely poor 1997 1999 2000 1997 1999 2000 2001* 1997 1999 2000 1997 1999 2000* National 47.4 71.2 48.5 37.5 61.3 37.8 100 100 100 100 100 100 100 1. Location rural 49.1 75.8 51.6 39.1 65.4 4 .5 .627 .638 .668 .678 .665 .698 .682 small towns 61.0 8 .3 62.5 49.9 72.0 52.4 .162 .157 .183 .200 .202 .190 .254 large cities 31.6 5 .4 28.5 23.3 4 .8 18.3 .210 .206 .148 .122 .133 .112 .064 2. Regions North 74.2 66.7 .301 .314 .328 Centre 66.2 53.4 .569 .529 .496 South 86.0 83.0 .130 .157 .176 3. Hh size children 35.6 6 .5 36.8 26.2 49.2 26.8 .401 .263 .341 .347 .235 .321 .324 one 47.9 73.0 53.0 38.0 64.5 41.9 .283 .292 .290 .300 .292 .298 .304 two 55.5 81.3 6 .9 45.7 72.5 48.1 .232 .292 .266 .241 .303 .274 .244 three 61.3 87.9 75.9 51.2 75.8 65.4 .066 .105 .081 .090 .111 .081 .100 ≥ four 71.7 89.7 87.5 59.9 84.6 83.3 .017 .048 .022 .031 .049 .023 .037 4. Age 0-15 54.0 78.6 59.2 43.9 69.4 48.1 24.7 .294 27.2 .264 .301 .280 .276 16-59 45.7 69.1 46.4 36.1 59.7 36.5 61.6 .589 59.8 .616 .588 .600 .622 60 & over 41.7 67.4 42.4 31.1 54.5 29.7 13.7 .116 13.0 .120 .110 .121 .102 5. Hh head male 47.5 72.6 49.6 37.3 62.5 38.8 73.3 .737 .747 .703 .731 .747 .705 female 46.8 67.3 46.2 38.1 58.0 35.4 26.7 .262 .252 .297 .269 .253 0295 6. Education hh head illiterate .509 .781 …. .010 .017 .011 …. primary .461 .732 .527 .108 .130 .111 .116 secondary .507 .743 .442 .759 .776 .793 .744 higher .289 .490 .244 .123 .075 .085 .058 7. Work employed 47.0 71.1 49.6 37.3 61.8 39.5 farmer 49.1 74.0 53.9 36.8 62.5 45.0 .103 agr laborer 53.5 79.9 66.2 42.6 7 .9 54.0 .251 .075 .107 .207 .073 .105 .217 n-agr labor self 43.1 65.6 4 .9 34.3 56.8 3 .9 .373 .318 .282 .207 .319 .290 .260 emp 34.8 51.8 33.1 29.3 44.3 24.8 .022 .321 .346 .294 .320 .346 .278 other 38.9 68.5 25.4 31.4 63.1 21.1 .020 .015 .016 .015 .016 .016 .014 unemployed 64.4 73.6 57.0 58.2 7 .0 49.7 .015 .018 .020 .010 .015 .020 .011 pensioner 46.5 71.2 46.8 36.6 58.4 33.9 .231 .253 .230 .252 .251 .220 .229 8. Sector agriculture 51.6 78.2 59.1 4 .2 68.6 48.7 .336 industry 43.3 59.6 38.0 33.4 52.2 29.7 .069 construction 45.6 72.7 44.6 41.0 65.0 35.9 .055 trade 46.3 65.3 43.6 38.5 54.4 33.3 .064 trans&com 45.9 7 .0 37.8 35.0 57.8 25.6 .036 other 24.3 53.7 32.5 19.3 44.9 24.3 .015 state admin 36.4 67.1 27.8 27.1 55.0 21.8 .028 education 35.6 62.5 33.8 27.8 53.1 23.7 .039 pub.services 44.1 62.7 52.2 37.4 57.3 41.2 .051 SOURCE: derived from WB (2004b) NOTES: *Computed using the population shares for 2001 as those for 2002 were unavailable.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 55 Furthermore the sensitivity test presented in WB (2004, p.75) is not entirely satisfactory as it fi xes the total poverty rate rather than letting it change with the change in poverty scale.

Chapter 4: Poverty, Inequality and Policy in Moldova

98 99 Nevertheless, the evidence is strong that higher risk of falling into poverty compared there was a greater incidence of poverty to households with school-age children. The and extreme poverty in families with three design of child allowances should take this or more children. The members of these into consideration. families are poorer than the average, because either the child allowance system In contrast, the incidence of poverty in did not reach all households, because the families with two children might have allowance was insuffi cient to lift the family been overestimated by the World Bank. out of poverty, and because the parents Many of these families were in poverty, would be poor even if they had no chil- but possibly with no greater probability dren. These families are a comparatively than the national average. It is unclear small proportion of the total (8.3 percent). whether the main cause of this problem The average number of children per fa- is the low income received by the adults, mily fell dramatically after independence, due to a low participation rate or low pay, in line with a drop in the total fertility rate or the demographic factor represented by from 2.3-2.4 in 1989-1991 to 1.2 in 2002. two children. Greater poverty incidence among large families may mean that poverty is associ- Poverty among pensioners was high, ated with unemployment or part-time but not signifi cantly diff erent from the employment among women who must national average (Table 6), while their risk care of their children. This problem might of extreme poverty was a bit lower than be solved by child care and labour market the national average. Similar conclusions policies that facilitate women’s employ- emerged from Table 6, panel 4, that in- ment. Be as it may, the elimination of po- dicates the risk of poverty and extreme verty in large families, an issue that should poverty was lower among the over-sixties receive priority, would not solve the bulk than on average.56 These results are con- of the poverty problem in Moldova. fi rmed by logit regression analysis done by PPMU (2002). That the level of poverty Many of children face a higher than average among the pensioners in Moldova is not risk of poverty due to the characteristics of particularly high is due to the almost the families in which they live, whatever universal outreach of the state pension may be the size of thee family. Children li- inherited from the socialist era, which sur- ving in single-parent families accounted for vived the fi scal collapse of the fi rst decade about six percent of all children. The share of transition. The future of such system increased due to changes in divorce and has been questioned, and as a result the adult mortality rates and, especially, because government has introduced (but imple- of the migration. PPMU (2002) suggests that mented in a limited fashion) a compulsory the share of such families among the long- pension insurance system. term poor grew, and accounted for over one-fi fth of the persistently poor in 2002. This average picture could be mislea ding, Infants in rural areas represented another as pensioners were a heterogeneous group particularly exposed to the risk of po- group. Pension levels varied considerably verty. Because of the greater time needed for among recipients, with war veterans, long child care, feeding and parental supervision, service and other privileged categories households with pre-school children faced a enjoying higher than the average, and

56 This is true a fortiori if one considers that the poverty estimates of the World Bank were obtained on the basis of ‘per capita equivalence scales’ while one could argue that the caloric consumption of an elderly non-active person are somewhat lower. On the other side, the WB (and PPMU) poverty line have not – or only partially - corrected for the increased health expenditure incurred by the households – and particularly by the elderly who are more disease prone than children and adults – following the cuts in public Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction expenditure on health that intervened between 1997 and 2001 and that unloaded a considerable part of the cost of health on the families.

Chapter 4: Poverty, Inequality and Policy in Moldova

100 101 those on minimum and survivor pension- rises steeply with their age as, likely, their ers receiving much less. As a whole, the income earning ability falls. The available pension transfers are mildly regressive. In data from the 2003 PPMU poverty diag- addition, the incidence of poverty among nostic for 2001 suggest that the incidence pensioners varies between the rural and of poverty among the disabled is not urban areas. While rural pensioners face a greater than for the general population, lower than average risk of poverty, as they a rather surprising result requiring further often can count also on a plot of land, the verifi cation. Their inability to earn income opposite is true in the cities where the was compensated in part by a disability pensioners that take part time work were payment similar to the old age pension, less numerous. Finally, there is evidence and covered than 115 thousand people that the risk of poverty among pensioners in 2002.

Table 4.4. Average pension per month & number of pensioners by types of pensions

1998 1999 2000 2001 2002 1998 1999 2000 2001 2002 Type of pension Number of pensioners Average pension/month (current Lei) Old age 555.970 535186 516861 495841 472556 87.6 85.0 86.1 143.3 169.8 - minimum 21824 63331 59097 9 0 60.4 64.9 64.9 65.0 …. - privileged 2 13176 13348 12807 12 5 102.9 100.5 101.2 165.2 208.0 Disability 112374 111735 113022 115526 115220 80.6 80.3 82.1 139.6 163.1 Survivor 35963 34493 37470 36012 33515 65.4 64.4 63.1 88.3 107.3 Long service 787 803 774 778 716 126.0 127.3 142.8 171.2 225.4 Chernobil acc. 1375 1595 2049 2178 2175 393.5 396.6 577.3 576.9 559.9 Memo Item Poverty lines WB 128.9 179.2 234.8 257.3 270.7 PPMU 104.7 130.0 175.2 193.1 201.8 Average wage 250.4 304.6 407.8 555.8 691.9

SOURCE: Ministry of Labour and Social Protection (2003): Annual Social Report

International experience shows that fe- legacy of the Soviet era), the industries male-headed households are frequently that collapsed primarily hired men; and face a higher than average risk of poverty. women headed households may be a Female headed household, and women proxy for households receiving remit- more generally, face much higher risks of tances. poverty because of a variety of discrimi- nations aff ecting them in the labour mar- In most developing countries, the level of ket and in access to some social services education is a good predictor of the risk such as education. However, while gender of poverty, but was less so in the econo- discrimination was robust in Moldova, the mies in transition. Until 1991 education in HBS and other data indicate that women Moldova was provided universally, free of headed households face a slightly lower charge to all those until sixteen years of level of poverty than male headed house- age. In 1999 the illiterate were only one holds. If this fi nding is valid, it might be percent of the population, those with

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction explained by three factors: women’s level primary education 10.8 percent, those of education was equivalent to men’s (a with diff erent types of secondary educa-

Chapter 4: Poverty, Inequality and Policy in Moldova

100 101 tion 75.9 percent, and those with higher but this was not true for those employed education 12.3 percent (Table 4.3, panel in ‘other public services’. Generally speak- 6). While secondary education is typically ing, workers employed in the private a route out of poverty in a low income sector have a somewhat higher risk of countries, this cannot be the case in falling into poverty than those working Moldova, where the share of people with in the public sector, primarily because the this degree level was quite high, while majority of private sector employees is the demand low because high unem- employed in agriculture. ployment and low-paying jobs. The risks of poverty only fall only for those with In principle, ownership of land, other higher education. physical and fi nancial assets, durable goods, and access to credit should reduce Widening access to higher education poverty, as the household could use them would at best create the supply side con- to smooth its consumption over time. In ditions for poverty reduction, for there Moldova, asset ownership was signifi - must be jobs for the educated to take. cantly correlated with poverty. Multilogit The fundamental employment problem regression analysis conducted by PPMU in Moldova, at all levels of education, is (2003) on the 2001 HBS indicated that demand, not supply of qualifi cations. while ownership of farmland itself and of While the highly educated were less likely simple household appliances were not as- to be unemployed, a substantially larger sociated with a lower risk of poverty, the number of tertiary graduates are unlikely size of housing, ownership of costly ap- to do more than raise the average qualifi - pliances, and bank savings were. In other cations of the unemployed and create an words, expensive appliances and size of ‘infl ation of qualifi cations’ in the labour land holding were a proxy for income lev- market. el. Lower poverty risk was also correlated with borrowing from banks and access to As one would expect, in Moldova the electricity. From a policy perspective, the unemployed face a higher risk of poverty poverty-reducing impact of credit should and extreme poverty. The unemployment be further investigated, to determine data upon which Table 6 is based must whether bank borrowing is cause or eff ect be viewed as highly unreliable.57 Where of lowering poverty risk. If it is cause, crea- Table 6 implies an unemployment rate tion of programmes that facilitate access of less than three percent, the DSS unem- to borrowing could be pursued. ployment survey found it to be over eight percent for the same year (see Chapter 6). Several studies have shown that migrant Further, both sources did not attempt to remittances can have a benefi cial eff ect on capture ‘discouraged works’, those who income, poverty, though they may nega- abandon the search for work due to lack tively aff ect other aspects of well-being. of success. Moldova illustrates well this trade-off , as it had one of the world’s highest propor- An analysis of the poverty incidence data tions of the labour force working abroad, by sector of employment (Table 6, panel migrant remittances on GDP and private 8) confi rms that poverty is highest in ag- transfers on disposable income. Migration riculture. It also reveals that households eff ects are discussed in detail in the next whose head was employed in the sector chapter. Here it should be noted that the and, including education, were generally generally recognised underestimation of better-off . Their poverty risk was fi fteen remittances at the household level and in percent lower than the national average, the aggregate may lead to some overesti-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 57In 2002, for instance, the unemployed had a 25 percent higher risk than the employed of being extremely poor.

Chapter 4: Poverty, Inequality and Policy in Moldova

102 103 mation of poverty, because the HBS show Second, high inequality reduces access to that the ex-post impact of remittances in education and health care for a large por- reducing poverty was large; as many more tion of the population. Third, when capital families would have remained below the markets are imperfect and the poor can- poverty line if the possibility of migrating not borrow, investment opportunities are had not been there. As a rough estimate, concentrated in the hands of the rich who migration reduced the incidence of pov- invest in activities with decreasing returns erty in 2002 by some twenty to twenty- while the poor, who would engage in fi ve percentage points. more productive activities, cannot realize their productive potential. Fourth, high inequality may reduce taxation and the access to key public services that are es- 4.4 Changes in Inequality sential for growth and social cohesion, as has happened in Moldova. Finally, a high Inequality in Moldova has attracted rela- degree of social stratifi cation can create tively little attention either at the offi cial political instability, undermining the ef- level or in public discussion. As noted fectiveness of public policy. in Chapter 2 and in this chapter, above, from 1991 to 1997, income inequality Part of the inequality rise of the early rose rapidly. The relationship between in- 1990s in Moldova was unavoidable, given equality and poverty can be shown quite the changes in economic system and simply for given a constant poverty line. output collapse, and the subsequent pro- The percentage change over time in the tracted stagnation. But part of it was dys- proportion of the population living be- functional, related to the weakness of the low the poverty line, ( PHR/PHR), can be legal and institutional development of the decomposed into the percentage change fi rst years of transition, which was unable in GDP per capita (DGDPc/GDPc),), aandnd tthehe to prevent the absolute and relative accu- percentage change in income distribution mulation of wealth in the hands of a few. (measured by an appropriate index, such As mentioned above, these inequality as the Gini, DGini/Gini), plus an interac- changes were an important contributor to tion term IT. ByBy defidefi n inition:tion: the surge in poverty rates in the fi rst decade of transition. Policy seeking to reduce DPHR/PHR = - DGDP/c + DGini/Gini + IT poverty must aim at reducing the level of inequality. As discussed in Chapter 3, taxa- The headcount poverty rate (PHR) de- tion and public expenditure are powerful clines with rises in GDP/c, and the elastici- instruments to achieve this. ty varies widely across countries, and even within countries over time, depending on The transition to the market economy the degree of inequality of the income induced large shifts in the structure of distribution, the extent of its clustering household income, which aff ected the around the poverty line, and the ratio of overall distribution of income. Indeed, the the poverty line to the average income income components characterized by low per capita (Table 4.5). Not incorporated in inequality (social transfers and wages) fell, the equation is the possibility that a rise in and those characterized by high inequa- inequality, or its persistence at high levels, lity (profi ts, rents, entrepreneurial incomes as in the Moldavian case, may aff ect the and incomes from the sale of agricultural rate of growth of GDP per capita itself, products) rose. The most dramatic shift thus depressing further the prospects for was the collapse of the wage economy, reducing poverty. There are several rea- with a fall in the wage share between sons for this ‘growth–dampening eff ect’. 1991 and 1997 of twenty-seven points,

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction First, when inequality increases substan- and by another two points by 1999, some tially, the work incentives of the poor fall. recovery during 2000-2002. The share of

Chapter 4: Poverty, Inequality and Policy in Moldova

102 103 social transfers on total disposable in- In parallel, the privatization of state assets come dropped from 19.2 in 1991 to 11.6 and the land, liberalisation of markets, and percentage points in 1997 (Table 4.5). the spread of commercial activities per-

Table 4.5. Structure of household income , by main income types, 1989-2002 (percentages)

Gini Income type 1989 1991 1993 1995 1997 1999 2002 range Wages and salaries 66.3 57.1 46.7 59.0 36.7 33.8 36.2 .35-.45 Self employment 14.9 15.1 26.2 12.4 36.7 39.6 32.0 .45-.60 Social benefi ts total 11.3 19.2 8.3 7.3 11.6 10.7 13.4 .25-.30 - pensions na na na 4.8 10.7 9.9 12.1 na - child allowances na na na 1.7 0.4 0.3 0.3 na - maternity na na na na na na na na - unemployment na na 0.1 0.1 0.1 0.1 0.1 na - social assistance na na na 0.7 0.4 0.4 0.9 na Property income na na na 0.1 0.2 0.1 0.1 .50-.60 Other(incl. 7.5 8.5 18.5 20.0 14.6 15.7 15.5 .45-.55 transfers) Memo item: no. of households 1325 1325 1327 1325 5876 7651 6159

SOURCE: author’s estimates on Transmonee database

mitted created opportunities for income ings inequality focuses on the adoption earning, which were successfully seized of new wage settlement rules that, unlike by a small portion of the population. All during the socialist era, put a premium this raised the share of informal sector in- on qualifi cations and skills; i.e., a relative comes, mixed incomes, profi ts and other wage argument. It emphasises also that types of capital income. For example, the the rising demand for new skills (com- share of self-employment income dou- puter specialists, accountants, foreign bled between 1991 and 2000. The income language specialists, fi nancial analysts, from property and rents probably rose as bankers, etc.) was not matched by sup- well, although it is obvious that these ca- ply, leading to scarcity and rapidly rising tegories are understated in the statistics wages (Vecernik 1994, World Bank 1994). in Table 4.5. In addition, the share of trans- In addition to these, overall wage disper- fers (of which remittances were a major sion in Moldova increased because of a component) doubled between 1991 and rise in inter-industry wage inequality that 2002. was hardly justifi ed by diff erences in skills and experience. Sectors in which wages Prior to the transition Moldova had an rose relatively were fi nance, followed at egalitarian wage scale, with a Gini coef- considerable distance by electricity, gas fi cient for the distribution of gross earn- and water, public administration, and ings in 1989 a low 0.25. Wages diff ered industry while the sectors that suff ered little within and among sectors. Wage the largest losses were agriculture, health, inequality rose sharply in 1992 reaching a education and, in recent years, trade (see Gini index of .40, and between 1992 and Chapter 6). 1999 the distribution of wages worsened further, reaching a Gini value of .44 (to de- Had the general level of wages fallen only cline to .39 in 2001). The standard explana- moderately, it might be possible to lend Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction tion off ered for such a large shift in earn- some credence to the standard explana-

Chapter 4: Poverty, Inequality and Policy in Moldova

104 105 tion. Real wages fell by sixty-to-seventy Table 4.6 calculates poverty incidence over percent in the fi rst half of the 1990s, and the years 2003 to 2007 for various combina- considerable imagination is required to tions of GDP growth and income inequality. attribute this to a relative change in the The results of this calculation show that the demand for skills. A much more credible fastest way to cut poverty in half by 2007 explanation presents itself: the collapse of requires in generating a sustained and rap- the demand for labour, with wage levels id growth of GDP in the range of eight per- further depressed by the virtual absence cent, without a further deterioration in the of trade unions that could protect the bar- distribution of income (see columns 2 and gaining power of workers. 3 of the table).58 Though not impossible, this outcome might be diffi cult to achieve because of the imbalances it might generate in the balance of payments, investment 4.5 Inequality and fi nancing, and institutional capacity. Similar levels of poverty, 26.8 versus 25.6 percent, poverty alleviation would be achieved if Moldova grew at a more achievable and more egalitarian four We can now consider how much faster percent a year. This would require rapid poverty would fall if Moldavian policy broad-based growth in rural areas and makers had introduced policies to restrict small cities that was highly employment the growth of inequality. This can be illus- generating, as well as in the two major ur- trated with a simple arithmetical exercise ban areas of the country. Poverty incidence that calculates the impact of diff erent would fall by almost forty percent from growth rates and income inequality lev- the 2002 level (fourth column), if the four els on the poverty incidence. Using them percent growth of GDP was achieved in a most recent distribution data, 2002 is tak- suffi ciently equitable way, i.e. with levels en as the base year, with the calculations of inequality similar to those of the UK and generated through 2007. Three growth Italy. This an objective would be implicitly rates are assumed: four percent (the aver- requested of Moldova as part of the acquis age rate recorded during the recovery of communautaires to be satisfi ed for a possi- 2000-2002); an ‘Asian Tiger’ growth rate of ble application to and eventual entry into eight percent per year; and slow growth the European Union. Under this scenario of two percent. The exercise assumes two poverty incidence would fall to 30.6 per- alternative levels of inequality of the dis- cent by 2007. tribution of net disposable income, the early 2000s Gini value of .42, and a lower Poverty incidence would remain pretty value of .35, that is close to that observed high in the low growth scenario of two in Poland, Hungary and Vietnam. Also, percent, and even a moderate level of in- it is similar to that of relatively unequal equality would contribute substantially to Western market economies, the United poverty alleviation. The benefi cial eff ects Kingdom, the United States and Italy, and of lower inequality would be signifi cantly higher than that of egalitarian developed amplifi ed by robust growth, and remain market economies, Austria, Belgium and modest with slow growth. Thus, both Sweden. Among developing countries, growth and inequality are essential for this value is close to those in South Korea poverty reduction, and that any policy and Taiwan, where the distribution of as- seriously aiming at this objective must sets and access to education have been include both mechanisms; growth alone egalitarian, credit markets effi cient, and would be insuffi cient; greater equality the labour force fully employed. alone would also be insuffi cient.

58Even more impressive results could be achieved if this ‘Asian tiger’ rate of growth could Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction be combined with a decline of the Gini coeffi cient from 0.42 to 0.35, i.e. from rather high to medium levels of inequality.

Chapter 4: Poverty, Inequality and Policy in Moldova

104 105 Table 4.6 Poverty incidence for diff erent assumptions on growth rates and inequality levels (percentages) Initial Headcount Poverty Rate (2002) = 48.5%

Growth = 2% Growth = 2% Growth = 8% Growth = 8% Growth = 4% Growth = 4% Growth = 4% High ineq High ineq Ineq = .42 Ineq = .35 Ineq = .42 Ineq = .35 Ineq = .30 = .42 = .35 PAEG* 1.5 2.2 1.5 2.2 2.8 1.5 2.2 2003 42.7 40.0 45.6 44.2 43.1 47.0 46.3 2004 37.5 32.9 42.8 40.3 38.2 45.6 44.3 2005 33.0 27.1 40.3 36.8 34.0 44.3 42.3 2006 29.0 22.3 37.8 33.5 30.1 42.9 40.4 2007 25.6 18.4 35.6 30.6 26.8 41.6 38.6

SOURCE: author’s calculations, using poverty alleviation elasticities of growth taken from Kakwani and Song. NOTES: PAEG is the ‘poverty alleviation elasticity of growth’. In the table, the elasticity is kept constant, though with growth the ratio (z/m) of the poverty line (kept constant) to the average income per capita should fall, leading to an increase in paeg. This eff ect may be negligible for growth rates of two and four percent and over a few years, but not so much in the case of a growth of 8% over 5 years. In this specifi c case (columns 1 and 2) – that has been neglected for simplicity – one would expect paeg to rise by about a third in relation to the value indicated in the third line of the table.

Annex 1. Statistical series of independent studies undertaken in recent years by a few United Nations Sources agencies, the IMF, World Bank, bilateral development agencies and several Most of the information used in this chap- Moldovan research centres. ter comes from micro data on poverty incidence and utilisation of social services With a few exceptions, the statistical infor- derived from the Household Budget Sur- mation available in Moldova is compara- veys conducted since 1997 by the Depart- tively abundant and of adequate quality, ment of Statistics and Sociological Studies and allows one to reconstruct relatively and elaborated in the form of Reports on accurately the poverty trends and fea- Poverty in Moldova (2000, 2001 and 2002) tures of this country. In particular, thanks by the Poverty and Policy Monitoring to the assistance provided by the UNDP Unit (PPMU) of the Ministry of Economy. and World Bank, the country can count on Several data are also drawn from the 2003 a comparatively large amount of informa- Public Expenditure Review and the 2004 tion on poverty and inequality trends and Poverty Assessment of the World Bank, on their immediate causes. Random sur- the UNICEF-IRC Transmonee database59, veys of relatively good quality have been the 2003 Statistical Yearbook, and a long taken since 1997 while smaller, non–ran-

59The Transmonee database ( see www.unicef-icdc.org/) reports standardized data for the period 1989- 2003 for all countries of the former European socialist bloc. The information covered concerns (i) indicators of human well-being (mortality, mobility, education, nutrition, child abandonment, crime incidence, etc.), (ii) their determinants (income, income structure and distribution, social transfers, family formation, social expenditure and service provision, relative prices, deciles tabulations of Household Budget Sur- veys, and so on) and (iii) background information on the macroeconomy, population and so on. For every Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction year and country, the database includes about 1500 data points that are provided to Unicef-ICDC on a regular basis and according to a standardized template by the Central Statistical Offi ces of the region.

Chapter 4: Poverty, Inequality and Policy in Moldova

106 107 dom and non comparable (but still useful) is a major source of intellectual and policy information exists for the prior years. The enrichment. routine data compiled regularly by the Department of Sociology and Statistics Prior to 1997, when a randomized HBS are also relatively abundant and detailed comprising about 6.000 households has and help in reconstructing the poverty been introduced and taken since then picture in Moldova. The main problems every quarter Moldova did not dispose aff ecting them derive from the division of of a regular randomized system for col- the country, following the confl ict of 1992. lecting data on household incomes and While some indicators refer to the whole expenditures. It could count, however, on country, others relate only to the area yearly surveys (aff ected by sampling bias) controlled by the Moldavian government, of about 1350 households that provide where the majority of the population live. some information, not immediately com- Though the division of the country oc- parable however with that generated by curred in 1992, the population series were the new household surveys. It was impos- adjusted only in 1998. sible to measure the degree of inequality of the distribution of income or to con- The analysis of the policy and non-policy struct a socio-economic profi le of poverty determinants of income-poverty and on the basis of randomly selected micro- non-income poverty is, in contrast, more data. All this poses a serious problem as a tentative in part because of data problems proper evaluation of poverty trends and (household surveys were introduced only anti-poverty measures and objectives in 1997) are and in part because diff erent should use 1989/91 as a baseline and not models of poverty emphasize diff erent 1997 (as currently done) as all indicators aspects of the problem (e.g. individual in that year were already much worse characteristics of the households aff ected than in the immediate pre-transition pe- vs. public policies). The area where the riod. Whenever possible, the analysis goes analysis is most defi cient concerns the im- back to 1989/91. pact of diff erent policies (especially macroeconomic and institutional but also, if to The country does not have an offi cially a lesser extent, development and social sanctioned poverty line. Several national protection ones)60 on the immediate de- and international agencies and research- terminants of poverty rates and various so- ers have produce poverty data, starting cial indicators. The methodologies for this from the PPMU, the World Bank, the DSS kind of macro-micro analyses are not yet and some researchers. All these institu- fully establishe d and required the ability tions and scholars recognize that poverty to combine diff erent fi les. Finally, reasons is a multidimensional phenomenon en- of space, do not allow this chapter to deal compassing numerous aspects of well-be- adequately with the ‘perceived indicators ing and that requires therefore a variety of of well-being’ derived from consultations gauges, income or consumption-based, with the poor’ through anthropological assets-based, capabilities-based and so studies and sociological surveys. This is on. Among the income/consumption- a major drawback as emphasized by the poverty concepts that have been used in current ‘qualitative-quantitative’ debate the comparatively large literature in this on poverty measurement (Kanbur 2003), area one fi nds the following concepts: (i)

60The panel component in the HBS makes it possible to analyse the evolution of household well-being and the long-term incidence of income poverty and vulnerability and to gauge the poverty impact of shocks and policy changes. However, the usefulness of these data for policy analysis would increase substantially if they could be couples with the routine data collected by various branches of the public administration (about service supply, coverage, utilisation, costs, benefi ts incidence and so on). Such Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction combined data would especially be needed for the evaluation of existing or planned social protection programs.

Chapter 4: Poverty, Inequality and Policy in Moldova

106 107 absolute (normal and extreme) poverty in order to standardize the consumption lines (ANPL and AEPL), that have been of diff erent family members into adult built in normative, empirical or participa- equivalent units. The ANPL assumes the tory manners, (ii) relative poverty lines presence of a non-food basket, which is (also in this case to denote normal or determined on the basis of the eff ective extreme poverty, RNPL, REPL that in this goods consumed by the population close case were by necessity built in a norma- to the extreme poverty line. tive way), and (iii) subjective poverty lines (SPL). The method of construction of these The AEPL developed by the World Bank poverty lines is briefl y reviewed hereafter, (2004b) corresponds to the cost per adult/ so as determine their applicability and day of 2100 calories (multiplied by thrity to usefulness in analyzing poverty changes standardize it on a monthly basis) where in the Moldavian context. the foods included in this caloric basket correspond to ninety items eff ectively There are three main producers of po- consumed by the poor i.e. people in the verty estimates, all making use of the HBS third to fi fth deciles of the distribution of data produced by the DSS, i.e. the Policy per capita consumption expenditure. This Poverty Monitoring Unit (PPMU) of the national poverty line amounted in 2002 Ministry of the Economy, the World Bank to 171.6 Lei per person/month. No equiva- (WB) and the Department of Statistics and lent scales were used to standar dize the Sociology (DSS). Poverty is calculated by consumption expenditure per capita into two absolute measures, ‘normal’ and ‘ex- adult equivalents, while diff erent unit treme’. The Department of Statistics and values are used to take into account rural- Sociology has developed in a normative urban price diff erentials. way following Government Decision 902 of August 28, 2000, involving a broad The World Bank ANPL was obtained by ‘minimum consumption standard’ that adding an allowance for tobacco and includes all the goods and services that a drugs, clothing and footwear, health care, normal family ought to consume to con- education, transport and communica- duct a decent life. The ‘absolute normal tions and entertainment (but not utilities, poverty line’ (ANPL) is then fi xed as 50 housing and consumer durables) corres- percent of such ‘minimum consumption ponding to some average of the eff ective standard’, after standardizing all mem- consumption of households in the third to bers of the households in adult equivalent fi fth deciles (Signoret 2004 in World Bank terms using the OECD equivalence scales 2004b). For 2002, this added another 30.2 that assigns a weight of 1 to the head of Lei per person/month to the poverty line. the family, 0.7 to all other adults and 0.5 Also in this case, no equivalence scale was to children. used to standardize the consumption expenditure into adult equivalent units, In the PPMU approach the absolute ex- while the same rural-urban price adjust- treme poverty line (AEPL) proposed is ments were carried out. set equal to the cost of 2282 Kcal a day generated by the food products eff ec- Though relative poverty lines are no- tively consumed by the people in the toriously of little use for tracking the second to fourth consumption deciles. evolution of material deprivation in low The prices used to determine the Lei income societies and though they can cost of the poverty line are an average of lead to wrong policy conclusions (as in the unit values paid by the poor for the case of sharp recessions accompanied by quantities purchased. Adjustments are a decline in inequality or, as happened in made for diff erences in rural and urban Moldova in 2002, in case of a rapid expan-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction prices. Unlike the World Bank approach, sion accompanied by a deterioration in PPMU used the OECD equivalence scales income distribution) both the World Bank

Chapter 4: Poverty, Inequality and Policy in Moldova

108 109 and the PPMU have ventured in the deve- tors such as the national economic trend, lopment of such poverty lines that were expectations about the future, climatic used, inter alia, in offi cial documents such changes, observation of ‘economic disas the PPMU (2003) ‘Poverty in the Repub- tance’ in relation to other households or lic of Moldova 2000’: individual emotional changes, i.e. factors that are independent from the changes The ‘relative poverty line’ used by the in material circumstance of the people in- PPMU was set at the level of fi fty percent terviewed. In any case, such approach can of the average household consumption be a good complement to the dominant expenditures per adult equivalent. The ANPL or AEPL one. equivalence scales used are the same as in the case of the absolute poverty lines. While the broad trends and features of In turn the ‘extreme relative poverty line’ poverty in Moldova identifi ed on the is the value of forty percent of the ave- basis of the numerous poverty lines are rage equalised household consumption broadly consistent and allow a relatively expenditure. This line was set at the level clear analysis, and while the literature of fi fty percent of the mean household produced in this fi eld by the PPMU and consumption expenditure per capita, in other national and international bodies is terms of adult equivalents, or 210 Lei. truly impressive, the current approach to poverty measurement in Moldova is still In turn, the ‘relative poverty line’ used by perfectible in a few regards. In particular, the World Bank (2004b, p.51) was set at it might be possible to remove a number sixty percent of the median equalised per of methodological problems concerning capita expenditure per year. The PPMU the determination of the level of poverty has also calculated a ‘subjective poverty and the ranking of poverty incidence and line’ (SPL) following the subjective ap- intensity among diff erent groups. Fur- proach developed by Bernard van Praag thermore, the data for Moldova suff er at the University of Leiden. In this ap- from measurement and quality problems proach the households are asked where that may weaken some of the conclusions they can make ends meet easily, not so arrived at on the basis of the present data- easily, with diffi culty and so on. In addi- sets and procedures. The main problems tion, various groups of households are can be summarized as follows: asked the amount of money they think is suffi cient not to be in poverty. This infor- Too many concepts of consumption mation allowed PPMU to estimate several poverty are being used in the analyses subjective poverty lines (see Table A4.1).61 on Moldova produced by national and international institutions.62 Table 4A.1 This approach has a number of infor- includes twelve poverty lines, and a few mational advantages, as it allows to more could be listed. As noted, some of understand the people (rather than the these measures (as the RNPL and REPL) government‘s or World Bank’s or that of area not useful in the Moldavian context the UN) perceptions of what it means to and should be abandoned. Likewise the be poor and can therefore be relevant for international poverty line of US$ 2.15 a the development of anti-poverty policies. person/day is not useful as it was esti- Its main problem is that the perception of mated on the basis of information from poverty may be infl uenced by several fac- about 20 developing countries that are

61 Brought to the extreme, this approach implicitly entails the development of ‘individual poverty lines’ as the perception of poverty and minimum consumption needed to escape it, may diff er substantially from one individual to another.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 62 PPMU computes and publishes also a large number of synthetic measures of inequality (decile ratios, Gini, enthropy, Atkinson, Theil, etc.) that are not used in the subsequent analysis.

Chapter 4: Poverty, Inequality and Policy in Moldova

108 109 Table A4.1. Poverty lines, incidence of poverty, extreme poverty and the poverty gap.

1997 1998 1999 2000 2001 2002 2003 A. Poverty lines (all in Lei per capita/month in current prices) (i) absolute poverty line (food share in parenthesis) - normal poverty line PPMU …. 128.9 (78) 179.2 (78) 234.8 (78) 257.3 (78) 270.7 (78) - - normal poverty line WB 99.3 (86) 104.7 (84) 130.0 (83) 175.2 (84) 193.1 (85) 201.8 (85) - - normal poverty line DSS - - - - ...... - internat. Poverty line US$2.15 WB 85.0 (?) 91.6 (?) 127.6 (?) 167.5 (?) 183.9 (?) 193.7 (?) - - extreme poverty line PPMU … 101.0 (100) 140.4 (100) 183.9 (100) 201.5 (100) 212.0 (100) - - extreme poverty line WB 85.5 (100) 88.4 (100) 107.8 (100) 147.5 (100) 163.4 (100) 171.6 (100) - (ii) relative poverty line - relative poverty line PPMU - - relative extreme pov line PPMU 210 - - relative poverty line WB - (iii) subjective poverty line - subjective poverty line PPMU 1 person household - - - 301.2 - - - 3 persons household - - - 279.0 - - - 5 persons household - - - 269.3 - - - B. Poverty Incidence - normal poverty line PPMU … 52.0 73.0 67.8 54.6 40.4 26.0* - normal poverty line WB 47.4 61.6 71.2 70.6 62.4 48.5 .... - normal poverty line DSS - - - - 51.2 42.2 .... - international poverty line WB 43.4 57.7 67.6 67.1 58.4 44.6 .... -extreme poverty line PPMU .... 37.4 59.7 52.2 38.0 26.2 15.0 -extreme poverty line WB 37.5 51.7 61.3 60.9 52.0 37.8 ....

- relative normal pov. line PPMU - 23* 23* 23.0* 23.8 22.9* .... - relative extreme pov. line PPMU - - - - 13.5 - - - relative normal pov. line WB ...... - subjective poverty line - - - - - 84.2 - - self evaluation by the population - - - - - 68.0 - C. Poverty Gap - normal poverty line PPMU .... 19.5 32.3 27.0 19.3 12.4 .... - normal poverty line WB 16.5 25.1 29.7 28.9 24.0 16.5 .... - international poverty line WB ...... - extreme poverty line PPMU .... 12.4 22.7 17.6 11.6 6.6 3.1 - extreme poverty WB 12.0 19.3 23.0 22.2 18.0 11.6 ....

SOURCE: World Bank (2004a, 2004b), PPMU (2003), GoRoM (2004), Notes: * author’s estimate. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 4: Poverty, Inequality and Policy in Moldova

110 111 very diff erent from Moldova and for this In Moldova, the absolute extreme poverty reason cannot be used as a reference for lines, based on the cost of 2100 or 2278 analysis in this country (Pogge and Reddy calories, are meaningless. A food–only 2003). Reference to this poverty line too poverty line may make some sense in can be dropped. a sub-tropical country with a large and relatively young population. However, in As for the ANPL and AEPL estimated on Moldova where winter temperatures can national data, it is quite evident that the fall as low as minus twenty degrees cen- trends identifi ed on the basis of the PPMU, tigrade, and the elder constitute a high World Bank and DSS approaches identify share of the population, survival requires – unsurprisingly – similar trends.63 Yet, the access to a modicum of shelter, heating, levels of the diff erent estimates can diff er electricity and life-saving drugs. While in in a non negligible way. For instance, the the past utilities and drugs were provided 2002 poverty headcount ratio estimated at highly subsidized prices, this is no lon- on the basis of the ANPL was equal to ger the case. In Moldova’s winter, people 48.5 according to the World Bank es- consuming 2100 or 2278 calories per day, timate and 40.4 according to PPMU. and nothing else, face much higher risks Second, while similar the WB, PPMU and of death, as confi rmed by the Moldavian DDS trends are not parallel, they tend to dismal mortality statistics. Thus, ‘food- converge in period of crisis (e.g. 1998) only’ poverty lines do not seem conceptu- and diverge in years of recovery. Third, ally helpful to identify a minimum survival the diff erent equivalence scales used in package. the three above approaches makes that the value of the PHR changes and, even The ANPL of PPMU and especially the more important, that the ranking of pov- World Bank are low, leading in this way to erty among diff erent groups changes , an underestimation of the number of the as shown by the recurrent problem con- poor, assuming all incomes in kind are ac- cerning the ranking of poverty among curately accounted for, see below. The im- families headed by pensioners versus plicit food share of the World Bank’s ANPL families with two children (see later). ranges between 83 and 86 (see Table The latter group appears to be on top A4.1), well in excess of the usual standards of the poverty list if one uses per capita by which the food share never exceeds equivalence scales and no economies of 75-80 percent of the consumption of the size in consumption, but not if one uses extreme poor (as well illustrated by the OECD or other equivalence scales. This is 80-80 ’Lipton’s rule’). In turn, the food an important issue, as it can infl uence the share of PPMU’s ANPL is 78 percent, still priorities of the anti-poverty policies and high but within the international range. transfer policies in particular. The 2001 HBS indicates for instance that the food share of the poor and extreme This proliferation of measures is quite con- poor is respectively 80.5 and 81.3, entail- fusing – as the results of analyses using ing that the food share at the level of the diff erent standards that are not compara- poverty line should be lower. ble. For policy purposes, there is a need to stick to single ANPL and AEPL (both set The high value of these food shares of the at an appropriate level see below) and to ANPLs estimated by the World Bank and a few measures of inequality, and rather PPMU depends on the method used for focus on the analyses of the relation be- constructing them. While the food com- tween poverty, policy changes and struc- ponent of the poverty line is ‘normative’ tural factors. (in the sense that the number of calories

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 63This should not be surprising as the years under investigation were characterized by marked swings in output

Chapter 4: Poverty, Inequality and Policy in Moldova

110 111 pre-established assures proper nutrition), as a percentage of GDP and in real terms the non-food allowance is derived in a (Table A4.2). This sharp reduction in social ‘non-normative’ (i.e. empirical-behaviour- spending was accompanied by grow- al) way, that basically measures what people ing inequality in the access to services, consume at around the poverty line level deterioration in their quality and the in- (actually below it, i.e. the third to the fi fth troduction of user charges in health and deciles for the World Bank and the second education. The introduction of offi cial to the fourth for PPMU). However, these fees at public institutions together with ‘eff ectively consumed quantities’ do not a rise in informal payments, have de facto necessarily guarantee adequacy in rela- raised the price of health services to a tion to a minimum normative standards. much higher level than before. Table A4.2 Also, the approach followed confl icts with seem to indicate that the average Molda- consumer theory. In addition, in the case vian citizen lost almost half of the health of ANPL developed by the World Bank, and education subsidies received from no allowance was made for expenditures the government in 1997. While some of on utilities and housing (that the 2001 this consumption may have been simply HBS shows account for no less than for foregone, its most essential part was ab- 8.5 and 8.1 percent of the total expendi- sorbed into household expenditure that ture by the poor and extreme poor) thus had to bear in this way a higher cost of the excluding from the computation of a minimum consumption basket. minimum standard an essential input for the production of a minimum of well-be- This means that the real value of the ing. This practice is in sharp contrast with non-food allowance component of the direct fi eld observation where concern for poverty line should have increased in the access to these consumption items is 1998-99 in relation to 1997 so as to refl ect paramount. For instance, in an interview the de-subsidisation of essential public with this author, the mayor of Floresti (a consumption on health and education, as typical de-industrializing town of 30.000 well as utilities and transport. On the basis people) indicated that the 3500 poor pen- of the documents available, it is not clear sioners living in his city allocated some whether this correction, which should forty percent of their meagre consump- have led to a rise in the real value of the tion expenditure to gas, electricity and no poverty line over time, was carried out by longer subsidized drugs. For the winter either PPMU or the World Bank. months, the extremely poor can apply for a monthly heating subsidy of 16 lei per month, but such subsidy is low in relation The poverty measurements are aff ected to the need and, as it is non-universal, it by whether all incomes recorded, and, may cause self-exclusion from the benefi t ff not, whether the under-recording because of the stigma and high informa- bias remains constant, declines or in- tional costs attached to the application. creases. It is well known that the ratio of Indeed, the data in Table 11 in section 3 average consumption expenditure per suggest that this kind of subsidy is allo- capita computed on the basis of the HBS cated in a very regressive way. tends to be a fraction (ranging between forty and 150 percent, with a median of Reductions in subsidized health and edu- around sixty) of the consumption ex- cation imply the need to revise upwards penditure per capita derived from the the poverty line. The World Bank public national accounts (Ravallion 2001). A se- expenditure review of 2003 shows that cond issue in this regard, well illustrated the overall public expenditure on health, by the analysis of the Indian data, is that education and social protection declined such ratio can change sharply over time, thus casting doubt on the consistency of

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction markedly over 1997-2001, following the drastic fi scal adjustment of 1998-99, both the poverty trends identifi ed. This prob-

Chapter 4: Poverty, Inequality and Policy in Moldova

112 113 lem is well known to DSS and PPMU, as in The estimated poverty alleviation elas- Moldova the cash and remittances com- ticity of growth seems rather high for ponent of national income is the lowest credibility. The elasticity was tentatively in Europe. The recording and imputa- estimated by PPMU at around 3, while tion of agricultural incomes is therefore the values estimated by the World Bank complex and could underestimate the (2004b, p. 94) are about the same (though eff ective consumption expenditure per with very high variations from one year capita. Likewise it is unclear to this au- to the next). However, such value looks thor the extent to which a (growing vol- incompatible with the comparatively high ume) of migrant remittances enter the level of income inequality prevailing in income and consumption accounts of the country (Gini of 0.42-0.44) and with the HBS. The evidence mentioned later the ratio of the poverty line to the average on in this paper suggests tentatively that income per capita of about 0.5 in 2002, as remittances are only partially recorded in the PPMU yearly poverty line was 3248 Lei the HBS, possibly leading in this way to while the GDP per capita was around 6100 an undercounting of overall household Lei (Ministry of Labour and Social protec- consumption. tion, 2003).

Table A4.2. Budget aggregates and the index of public expenditure on health, education and social protection

Years 1996 1997 1998 1999 2000 2001 2002 2003 2004 A. Percentage of GDP Total Revenue 26.6 33.0 30.5 25.1 25.6 22.7 22.5 24.2 23.2 Total Expenditure 36.2 40.4 33.9 28.7 26.6 22.7 23.0 22.6 22.1 Defi cit/surplus -9.6 -7.5 -3.3 -3.2 -1.0 0.0 -0.5 1.6 1.1

Social Expenditure 19.9 22.0 16.2 11.8 11.8 10.8 12.4 12.3 11.4 -Education 10.3 10.0 7.0 4.7 4.5 4.8 5.5 5.5 5.3 -Health 6.7 6.0 4.3 2.9 2.9 2.8 3.5 3.4 2.6 -Culture, sport, youth 0.8 0.9 0.9 0.5 0.5 0.4 0.6 0.6 0.6 -Social protection* 2.1 5.1 4.0 3.8 3.8 2.6 2.9 2.7 2.9 -Social protection** na 8.4 8.4 6.1 6.2 5.7 na na na B. Index of expenditure in constant prices (1997=100) Social Expenditure na 100 77.5 57.2 57.2 57.0 na na na -Education na 100 69.7 50.3 47.7 54.7 na na na -Health na 100 67.6 45.1 48.8 52.2 na na na -Culture, sport, youth na na na na na na na na na -Social protection* na na na na na na na na na -Social protection** na na na na na na na na na SOURCE: elaboration on World Bank (2003) and budget data provided by the Ministry of Finance provided UNDP-Kishinew (e-m by Liudmila Barcari). * includes only the transfers from the state budget to the social insurance fund, ** includes the outlays of the social security fund fi nanced with payroll taxes Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction and not included in the state budget.

Chapter 4: Poverty, Inequality and Policy in Moldova

112 113 Indeed, according to Table A4.3 that repro- not smooth and shows considerable bunch- duces the estimates of Kakwani and Song ing (a hump) around the poverty line, so (2003) based on a large sample of ‘growth- that even comparatively modest variations poverty episodes’ for countries with diff e- in consumption expenditure per capita rent levels of inequality, the Moldavian pov- move many people in and out of poverty. erty alleviation elasticity of growth should be in the 1.4-1.6 range. The reasons of the While the latter explanation may account discrepancy between the PPMU and the for the high poverty-growth elasticity estimates in Table A4.3 ought to be clari- observed over the last few years in rela- fi ed. The discrepancy may be due to recent tion to changes in consumption expen- improvements in collecting information on ditures within its specifi c 1997-2002 range food consumption and food inventories in of variation, it is unlikely that such high the HBS, that have caused an increase in ru- elasticity will hold for wider ranges of ral incomes faster than what it would have the distribution. This means that it would been the case with no changes in metho- be unreasonable to assume the same po- dology. A second explanation is that the verty –growth elasticity for projecting the distribution of consumption expenditure is poverty impact of future growth.

Table A4.3. Poverty Elasticities Poverty-growth elasticity Poverty- inequality elasticity z/m Gini coeffi cients 0.3 0.4 0.5 0.6 0.3 0.4 0.5 0.6 0.33 -3.9 -2.1 -1.3 -0.8 5.2 3.3 2.4 2.0 0.50 -2.8 -1.6 -1.0 -0.7 2.5 1.7 1.3 1.2 0.67 -2.0 -1.2 -0.8 -0.5 1.2 0.9 0.8 0.8 1.00 -1.2 -0.8 -0.5 -0.4 0.2 0.2 0.3 0.4 SOURCE: Kakwani and Song (2003) *The left panel is the elasticity between GDP growth and poverty, the right panel the poverty- inequality elasticities under diff erent assumptions about the Gini coeffi cient (shown by column) and the ratio of the poverty line (z) to average income or consumption per capita m (again, by by column)

Annex 2. Analytical Where Cit is the concentration coeffi cient of the i-th income component (wages, Framework for or entrepreneurial incomes, or capital Calculating Poverty income or transfers),64 at time t, and si its share in total income. Changes over time Reduction in Gini coeffi cient can be decomposed as:

We consider the specifi c factors explain- DG = Gt+n - Gt = SDsiCi + SDCi si + SDsiDCi ing the rise of income inequality in Moldova by using the Kakwani decompo- Where the Dsi and DCi refer to changes sition (1980) that shows that at any point over the period t - t+n in the shares of in time, the Gini coeffi cient (Gt) can be the various income components and their written as follows: concentration coeffi cients. This decompo- sition requires data for the initial and fi nal Gt = sit Cit year, the shares of each type of income, and their concentration coeffi cients.

64The concentration coeffi cient C is similar to the Gini index except that the ranking of individuals is by

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction it the total income and not the i-th income components. As a result Cit can ranks between 1 and –1, instead of 0 and 1 as the Gini coeffi cient.

Chapter 4: Poverty, Inequality and Policy in Moldova

114 115 C h a p t e r 5 : Migration, Employment and Poverty Reduction

5.1 Migration from in late 2004 and sponsored by the Inter- national Organisation for Migration, the Moldova IMF and the European Union. The resulting Moldova Remittances Study (MRS) There are many estimations of the number estimated 398,000 Moldovans working of migrants from Moldova, and substan- abroad coming from approximately tial diff erences between offi cial statistics 315,000 households. If the number of and unoffi cial estimates from a variety of people who were in Moldova, but have sources. The offi cial data on legal or regu- been abroad and reside temporarily (the lar emigration (Statistical Yearbooks, 2003/ exact phrase is ‘or a certain time’) in the 4) suggest yearly outfl ows between 1992 country, then the migrant contingent and 2003 ranging between fi ve and nine rises to approximately 567,000. Just over thousand. Capturing illegal or irregular twenty percent of Moldovan households migration is considerably more diffi cult. had one or more members aboard with In 2000 the Department of Statistics and the vast majority of these, almost eighty Sociology estimated that were 234,000 percent, having one. Moldavians working abroad legally and illegally (Sleptova 2003). An estimate The following description of Moldova’s very close to this was obtained by the migrants comes draws heavily on MRS. Labour Force Survey of 2003. In 2004 an Moldovan migrants are generally young independent survey conducted by the with over seventy percent under 40 years Moldova Microfi nance Alliance estimated old and nearly forty percent under 30. that between 265 and 285 thousand There is a signifi cant gender diff erence Moldovans went in search of work abroad with female migrants much more likely to (Ghencea, B & I Gudumac 2004). The me- be in the 31-40 year old category than the dia has reported as many as one million 21-30 year old category (35 percent versus Moldovan citizens working abroad. 28 percent of female migrants). The percentage for men is the reverse (28 percent The most recent estimate comes from a in the 31-40 range versus 38 percent in survey of over 3,700 families undertaken the 21-30).

Table 5.1: Age distribution and gender (percentages)

Under 20 21 – 30 31 – 40 41 – 50 over 51 Total years years years years years Male 5.5 38.3 28.3 22.5 5.5 100.0 Female 4.1 27.9 35.1 29.1 4.0 100.0 Total 5.0 34.6 30.7 24.8 4.9 100.0 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: Moldova Remittance Study (January 2005)

Chapter 5: Migration, Employment and Poverty Reduction

114 115 Although we do not have adequate infor- to Spain or Greece have higher qualifi ca- mation on the poverty status of migrants tions compared to only eight percent of the MRS reveals that the amount of mo- migrants to the Ukraine or nearly twelve ney required to become a migrant is not percent of migrants to Russia (MRS). large although it varies signifi cantly with the destination country. More than half The majority of migrants from Moldova, claim that migration expenses are less over fi fty eight percent, go to Russia (MRS) than US$100 with a quarter claiming that and less than thirty two percent go to EU they are less than fi fty. The majority of the member states. Of these countries, Italy migrants spending the small amount of is by far the most popular destination ac- funds migrate to CIS countries. Migrants counting for over eighteen percent of to the EU will spend considerably more, migrants. Female migrants are more likely for example more that US$2000 for Italy to go to Mediterranean countries than to or Spain and over US$440 for Turkey. northern Europe or members of the CIS (MRS). This pattern is not surprising given Migrants tended to be generally well- the tradition from the Soviet period when educated. Only ten percent had no edu- Moldovans sought seasonal work in ag- cational qualifi cations, and over forty one riculture and construction in other parts percent have reached secondary educa- of the Soviet Union, particularly in Russia tion qualifi cations. A further twenty eight and Ukraine. Today most reportedly work percent have post high-school qualifi ca- there as labourers in the oil, manufactur- tion while twenty percent have higher ing, and construction industry; others work education qualifi cations. There is also a as drivers, cleaners and small-scale retailers strong relationship between destination (Dudwick and Sethi). Although the number and education with migrants to the EU of respondents is small, just over half of the having higher levels of qualifi cation than migrants appear to be employed in the those to the CIS or other transition coun- construction industry. No other category tries. For example about half of migrants comes close to this proportion (MRS). Table 5.2: Main countries of destination

of the families with migrants migrants Spain 1.4 1.5 Italy 20.2 18.9 Russia 56.4 58.2 Turkey 1.3 1.3 Romania 1.5 1.9 Portugal 4.7 5.0 Cyprus 0.5 0.5 Germany 1.2 1.0 Greece 3.2 2.7 The Czech Republic 1.1 0.8 Belgium 0.6 0.5 Israel 1.7 1.4 Ukraine 1.7 2.2 France 1.0 0.9 Other countries 3.5 3.0 Total 100.0 100.0 EU 33.9 31.8 CIS 58.1 60.4 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: Moldova Remittance Study (January 2005)

Chapter 5: Migration, Employment and Poverty Reduction

116 117 Table 5. 3: Main countries of destination by gender

Country Male Female Gap Cyprus 15.60 84.40 68.80 Greece 29.60 70.40 40.80 Turkey 32.20 67.80 35.60 Italy 36.50 63.50 27.00 Spain 41.40 58.60 17.20 Romania 46.70 53.30 6.60 France 50.80 49.20 -1.60 The Czech Republic 51.70 48.30 -3.40 Israel 56.30 43.70 -12.60 Ukraine 63.20 36.80 -26.40 Belgium 63.90 36.10 -27.80 Portugal 67.90 32.10 -35.80 Russia 74.30 25.70 -48.60 Germany 78.20 21.80 -56.40

SOURCE: Moldova Remittance Study (January 2005)

Table 5.4: Spheres of activity of the Moldovan migrants abroad:

The percentage of the total number of the interviewed migrants (214 people) Agricultural works 6.9 Constructions, reparations 51.3 Employees at factories, plants, fi rms 7.2 Employees at private people 10.9 Housekeeping, social care 7.8 Commerce 7.3 Services 7.3 Something else 0.9 No answer 0.4

SOURCE: Moldova Remittance Study (January 2005)

Just over twenty-eight percent worked Fewer than thirty percent were helped by on an employment contract leaving over friends, neighbours and acquaintances and seventy percent outside the regular sys- approximately twelve percent were helped tem. Very few had assistance from the by family. Only 4.3 percent and 3.8 percent government or migrated through offi cial were helped by tourist agencies and em- /regular channels. When asked who had ployment agencies respectively. The high helped them or organised their departure proportion of migrants who leave without

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction for working abroad, respondents revealed any help is explained by the fact that most that over forty six percent had no help at all. of these were going to Russia. Over eighty

Chapter 5: Migration, Employment and Poverty Reduction

116 117 one percent of migrants to Russia (over fi fty While obtaining adequate information eight percent of total migrants from Moldo- about migrants is problematic, the situ- va) went without any support. The vast ma- ation is mirrored when trying to iden- jority of migrants also found a job within tify information about remittances. In one month (over eighty two percent) with a similar way there are offi cial statistics half fi nding work when they arrived. Only that do not capture fl ows outside formal a very small percentage paid something to channels and many unoffi cial estimates. obtain a job (less than six percent). Again the media presents very high numbers in this respect. Moldova is not In terms of the way of departure, the major- alone with this problem and the quality of migrants left using an identify card (fi f- ity of data on remittances is gene rally ty four percent). This roughly corresponds quite poor (Kapur 2004). Over half of with the number of migrants working in respondents to the MRS send money the CIS. Just over thirteen percent used a home at least once a quarter with about work visa while just over fourteen percent 25 percent sending it home on a regular used a tourist visa. Only a small number basis. The MRS team estimate that over used a false visa or were illegally transport- half a billion dollars was transferred in ed by interlopers (just over fi ve percent in remittances over the last twelve months each category) although these fi gures may (October 2004). obviously be under represented.

Table 5.5: Reason for Migration ( of total respondents)

Tourist visa 14.3 Work visa 13.4 Offi cial trip from the working place 1.0 Sport delegation 0.8 With the identity card 54.3 With foreign citizenship 2.4 Cultural exchange (dance, theatre, orchestra troupe etc.) 0.4 Business relationship 0.2 Marriage 0.7 Something else 0.7 With false visa 1.3 Illegally transported by the interlopers 5.2 No answer 5.1 Total 100.0

SOURCE: Moldova Remittance Study (January 2005)

It should also be noted that Migrant and household assets (MRS). Only a third families also send money to Migrants al- of transfers are made through formal though the amounts are very small (MRS). channels (bank transfers, rapid transfer Just over one third of Moldovan migrants and the post offi ce) on a regular basis also sent home goods in addition to although more than half have used them

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction money, mostly clothes, food products at least once.

Chapter 5: Migration, Employment and Poverty Reduction

118 119 Table 5.6: Methods used by the migrants for sending money home

They used it at least one time They used it more often Bank transfers 25.7 13.0 Rapid transfers 28.8 19.5 Post offi ces 4.4 1.9 Train conductors 12.7 7.1 Bus, mini-bus conductors 11.4 6.8 He brought it personally 38.9 17.3 Through relatives, friends, acquaintances 29.1 11.7 In parcels with the other things you receive 8.7 1.3 Something else 9.4 0.9 No answer 20.6

SOURCE: Moldova Remittance Study (January 2005)

In terms of using the remittances received mulated abroad, over forty fi ve percent of over seventy percent say they used funds respondents reported that their priority is to buy food and clothes while over sixty current consumption (food, clothing etc). percent used funds to pay for household Only a small percentage prioritised busi- utilities. In terms of using the funds accu- ness investment.

Table 5.7: Planned use of money earned abroad ‘in the next 12 months’

First Priority Second priority percent percent The debt repayment 11.8 5.1 Current consumption (food, clothing, current 45.4 21.7 household expenditures, services and utilities) Special consumption (education, health, household 10.6 24.6 durables, lending the money etc.) Household investment ( to buy a car, house/ apartment, house/apartment renovation, weddings , 15.5 22.8 funerals , bank deposits) Business Investment (to buy land, agricultural 1.5 4.8 machinery, mini-bus, cattle/fowl/Sheep ) Something else 0.7 0.2 No answer 9.8 2.4 The number of people who answered the question 95.3 81.7

SOURCE: Moldova Remittance Study (January 2005)

It is also important to recognise the diver- ferent impacts. For example, the impact sity of remittances (COMPAS) in the sense of intra-family transfers may be diff erent

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction that the label remittances relates to a vari- from transfers controlled by the migrant ety of diff erent transfers that will have dif- to collective transfers.

Chapter 5: Migration, Employment and Poverty Reduction

118 119 5.2 Migration, Growth, order to get a better overview of the positive and negative impacts of migration. Employment and Poverty Reduction As noted in Chapter 2 the recovery and growth of the economy after 1999 was driven primarily by the infl ow of remit- Migration and Poverty tances which fostered growth through relieving the import constraint. In addi- Given the diffi culties with obtaining tion, it was noted that remittances have adequate information about migration reduced the government’s dependence levels, migrants and remittances, accu- on conditionality-based borrowing, thus rately identifying their impact on growth, potentially allowing for fl exibility in macro employment and poverty reduction is not policy. However, the infl ow of remittances surprisingly problematic. It should also be has the eff ect of appreciating the ex- noted that the impacts that fall into the change rate, as occurred during 2003 and categories of positive or negative may not 2004, making exports less competitive, be substitutes. For example, it is impossi- while the higher interest rate would dis- ble to say that the brain drain or reduction courage investment by Moldovan com- in human capital can be countered by in- panies, and have little impact on foreign creased fi nancial fl ows from remittances: domestic investment. Though there is no They are not substitutes (Kapur 2004). research on the strength of these negative This section sets out what we believe to eff ects, their impact may have been quite be the various impacts of migration on strong in the late 1990s and early 2000s. growth, employment and poverty reduction. Some of the impacts are direct, for Remittances can have a direct impact example increased consumption for poor on poverty reduction though increasing households from remittances. Others are levels of consumption. The draft study less direct, present consumption from revealed the importance of remittances previous investment or increased know- to the welfare of those left behind and ledge from the experience abroad leading families with migrants are generally bet- to higher earning potential. Many of the ter off than those without. Table 5.8 below issues have been raised in other chapter illustrates the before and after migration and these are referred to in this section in perception of households with migrants.

Table 5.8: Household Perception of Eff ect of Migration Good Very No Very good Diffi cult (adequate) diffi cult answer Before the migration 1.6 57.1 34.3 5.6 1.4 Food After migration 7.4 71.6 18.5 1.0 1.5 Before the migration 1.6 52.1 38.2 6.3 1.9 Dwelling After migration 4.7 70.4 21.5 1.6 1.8 Before the migration 1.1 46.9 45.0 5.3 1.7 Clothes After migration 4.5 68.1 24.3 1.4 1.7 Before the migration 1.6 50.7 38.3 7.6 1.8 Health After migration 2.8 57.6 33.2 4.7 1.7 Before the migration 1.9 47.4 35.5 4.6 10.6 Education After migration 3.3 60.1 20.8 1.8 14.0 Before the migration 1.0 31.8 46.7 12.9 7.5 Entertainment After migration 2.3 47.3 35.6 7.1 7.7 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: Moldova Remittance Study (January 2005)

Chapter 5: Migration, Employment and Poverty Reduction

120 121 In terms of poverty reduction the major migrate are neither the rich, who are not decline in the “very diffi cult” category and impelled to move by material deprivation, the signifi cant decline in the “diffi cult” cate- nor the poor who are credit-constrained, gory across all basic needs illustrate the but a middle class of fairly young people contribution of remittances to household with medium-high levels of skills and welfare among households with migrants schooling. This high-remittances group abroad. Several studies have shown that is however under-represented in the HBS. migrant remittances can have a benefi cial This last argument may not hold as we eff ect on income–poverty, though they know that most migrants work in Russia may also aff ect negatively other aspects and only need very small funds to get of well-being. Moldova illustrates well there. Of course they also earn conside- this trade-off , as it has one of the highest rably less. proportions of her labour force working abroad, migrant remittances on GDP and The situation seems to be one in which private transfers on disposable income. If the underestimation of remittances in properly accounted, all these three ratios HBS both at the household level and in are likely to near thirty percent of the total. the aggregate possibly leads to some overestimation of poverty. In spite of this, In accounting terms, the HBS includes the HBS show that the ex-post impact of questions also on ‘private transfers’ with- remittances in reducing poverty is very out distinguishing whether these origi- large, as many more families would have nate from abroad or within the country. remained below the poverty line if the In addition, the relation between remit- possibility of migrating had not been tances and poverty cannot be studied there. A rough estimate is that migration thoroughly as only part of them are reduced the incidence of poverty in 2002 accounted in the HBS, in which they by some twenty-twenty fi ve points, and represent 5.5 percent of total household that these sharp gains in income-poverty income against the seventeen percent were accompanied however by a worse- of GDP estimated on the basis of the na- ning in social-poverty. tional accounts.

Three factors may explain this discre- Remittances have the possibility of in- pancy. To start with, remittances used for creasing future consumption through in- the construction of houses may not be come generating investments. As already considered as discretionary income and noted the vast majority of respondents not be declared in the survey. Second, intend to use the assets accumulated especially in the countryside, lax security from migration primarily for consump- and fear of theft may induce households tion. Business investment seems a prio rity living in isolated areas to ‘hide’ the remit- for very few. The secondary impact of tances received for fear that this informa- investment on income poverty reduction tion may leak out and induce some rob- through future consumption therefore ber to steal these moneys. Third, the HBS seems limited in Moldova. Like in many undersamples the families of emigrants. other high migration countries, one of the As the ‘hump theory of migration’ sug- greatest challenges for government poli- gests, migration is costly and only those cy makers is to change this situation and with resources or access to family or infor- to facilitate the channelling of resources mal credit are able to fi nance their migra- to sustainable income and employment tion.65 This mean that the most likely to generating activities.

655000 Euros per migrant is the most common estimate of the investment required for the illicit acquisi- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction tion of visas and working permits, travel costs and the connection to networks in the country of destination.

Chapter 5: Migration, Employment and Poverty Reduction

120 121 Migration and had previous experience in medicine, edu- the Labour Market cation, agriculture, construction and small business management (Dudwich & Sethi The dynamics of the educational sector 2004). By any standard, these represent a and the migrations process have seriously serious brain drain. It had immediate im- impacted on the human capital formation pact on societal health, quality of educa- in Moldova. Only one in every fi ve people tion and economic growth. Migration of in Moldova goes for university education. unskilled workers, mainly from rural areas, But more important is the fact that enrolled to higher than normal dependency ment of children from the poorest house- ratio and to several family problems with holds even at the upper secondary level children left by their parents, to grow up has declined in recent times and it is only with no care and no adequate education. fi fty percent, compared to seventy-one for the richest quintile. At the university On the other hand, consumption may level, the situation is even worse. The en- benefi t the supply of human capital espe- rolment rate for the poorest women and cially where remittances are used for edu- men has consistently declined over time cation. Other countries have benefi ted to about four percent in the mid-2000s. from new skills of returnees though there For the richest households, the enrolment is little evidence of this in Moldova, that is not only improved, but at thirty-four per- not to say that it is not happening (i.e., just cent was more than four times than that because it is not captured by surveys) of the poorest quintile. Human capital formation in Moldova had a defi nite bias One of the greatest challenges faced in towards the rich. Various social illnesses, the transition has been to adapt profes- some of which are by-products of po- sional education towards to new econ- verty, alcoholism, drug dependency, and omy and the new types of job that are high criminality rate, depleted the quality available. Moldova is no diff erent form of human capital as well. other FSU countries and faces the challenge of re-organising its Vocational Edu- The migration from Moldova also impac- cation and Training (VET) system in terms ted on the formation of human capital in of facilities, curricula and the subjects it that country. For a country of 4.4 million teaches. Migration can lead to shortages people, an estimated 600,000 people, al- of workers, especially in specifi c sectors most fourteen percent of the population where the skills are required abroad. It is and twenty three percent of the work- widely reported that shortages of skilled force, worked abroad in 2003.66 Over one and non-skilled workers in the agriculture half of the expatriate workers are under sector are largely the result of migration, age thirty, and eighty eight percent under both international and internal, especially forty fi ve, implying that workers in the among the young. It may be that rural ag- prime age group were leaving the coun- ricultural workers are migrating internally try. In the fi ve years, 2000-2004, thirty six to fi ll the jobs left by urban workers who thousand doctors left the healthcare sys- have migrated abroad. Domestic migra- tem, and twenty-eight thousand teachers tion data also needs to be developed. The left the education system. About seven census probably has many of the answers percent of the expatriate labour has a but is not yet available. university education, and one third of the workers abroad had secondary vocational While there are reports of large increases education, compared to just twenty-fi ve in salaries being made to attract construc- percent of the workers at home. Nearly tion workers back to Moldova, there is lit- three-quarters of those working abroad tle evidence of large increases in salaries Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

66The estimates of migration range from 230,000 (LFS 2002) to one million (MISS).

Chapter 5: Migration, Employment and Poverty Reduction

122 123 in the construction and agricultural sec- entered the sex trade, though estimates tors. Offi cial data suggest that the ave rage vary widely, from 10,000 to 100,000 (US nominal monthly salary has increased by Embassy in Sweden 2002). While some 305 percent between 1997 and 2003 women knowingly become involved into while in the agricultural and construction prostitution abroad or were aware of the sector nominal monthly salaries have in- risk, others are lured abroad with false job creased by 270 percent and 265 percent promises and then coerced into prostitu- respectively. tion. There were several root causes for the ready supply of Moldovan women to the sex industry. These included The Social Impact of Migration structural reasons such as poverty, social exclusion (in case of Roma population), An outrageous aspect of migration was and insecurities of the youth that pushed human traffi cking, which was quite wide young people abroad in search for better spread in Moldova. There were three opportunities. A survey of Moldovans be- major features of Moldovan human traf- tween sixteen and thirty years found that fi cking. First, there was the misleading of almost all of them wanted to go abroad if innocent people with the false promise of they could fi nd an opportunity. Domestic employment abroad. Moldovans spent an violence also instigated women to leave average of US$126 to obtain access to em- the country. ployment in CIS countries, 1,987 for Wes- tern Europe and 2,470 for the Middle East. Traffi cking of people fragmented com- Only two-fi fths of the migrants depended munities and families, depleted human on their own resources. Of the rest thirty capital, deprived households of expec- percent borrowed from relatives, thirty ted remittances, and caused child ne- percent from moneylenders (at monthly glect. Migrants working illegally in host interest rates of ten to fi fteen percent in countries were subject to exploitation foreign currency), and almost twenty per- and abuse from smugglers, employers cent from Savings and Credit Associations and local authorities alike. As countries (at interest rates of twenty to twenty).67 In tigh tened immigration restrictions, they addition, people often would sell assets drove greater numbers of people into (Munteanu 2002). illegal migration. Even when illegal migrants could remain working in the Second, in many cases, people who were host country, they would become part traffi cked became bonded labour in the of an underclass without legal rights or host country. This happened when peo- recourse. Finally, illegal migration con- ple could not fi nd work abroad, or their tributed to the expansion of global crime documents were lost, stolen, or seized by networks. Benefi ting from the mismatch the employers, or who could not pay off between legal opportunities and actual debts related to their migration. The last demand for migration, criminal organi- phenomenon was often related to traf- zations dealing in arms, drugs, and pros- fi cking of the poorest. They would arrange titution diversifi ed into migrant smug- that traffi ckers pay money in advance to gling. They benefi ted from the steady their families, with the understanding that demand from migrants, coupled with the sum would be repaid by the traffi cked lax prosecution and enforcement. The persons from their earnings. victims of this traffi c become vulnerable to violence, isolation, and diseases such Third, a signifi cant portion of female as STDs and HIV/AIDS. The end result is migrants, most often those under 25, serious human deprivation.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 67 Moneylenders charged monthly interest rates of ten to fi fteen percent. The Savings and Credit Associa- tions were even more expensive, twenty to twenty-fi ve percent.

Chapter 5: Migration, Employment and Poverty Reduction

122 123 Migration also adversely aff ected chil- 5.3 Possible Future dren. When children of migrants remain at home under the care of relatives or Patterns of Migration neighbours, they fall risk to leaving school early, moving away from home, Both qualitative and quantitative research and ending up as street children, and this reveals similar motivations for becoming process characterised Moldova. Children a migrant. The MRS reveals that the main themselves became subject to human motivation is for current consumption with traffi cking. According to the data from nearly forty four percent of respondents the Ministry of Interior, Moldova became citing this reason as the main factor. Debt a major source in supplying underage repayment and household investments girls to Russia for sexual exploitation, were the second and third main factors. with an estimated 5,000 traffi cked there Combined these three factors were the annually. Children from care institutions, major motivations for over eighty four per- forced to leave at sixteen with few job cent of respondents. A very small number skills or support network, were particu- of migrants (less than one percent) iden- larly at risk, and there were allegations tifi ed business investment as the major implicating an orphanage director in factor for migration. This corresponds to selling girls to traffi ckers (Dudwick & Ethi actual use of remittances and resources ac- 2004). cumulated by migrants described above.

Table 5.9: Factors that determine departure

First Second Debt repayment 21.2 6.5 Current consumption (food, clothes, commodities, household assets, etc) 43.9 32.6 Special consumption (education , health, household durables, lending money, etc) 11.3 21.4 Household investments (car, house/apartment, renovation, weddings, funerals etc) 19.0 28.6 Business investments (land purchase, agricultural machinery, mini-bus, animals etc.) 0.9 3.9 Something else 1.7 0.2 No answer 1.5 0.5 The total number of people who answered the questions 99.5 93.7

SOURCE: Moldova Remittance Study (January 2005)

The qualitative survey undertaken re- ply side there is no reason to suggest that vealed similar results where the key fac- these factors or motivations will change. tors for the departure included (a) insuf- As noted in earlier chapters, signifi cant fi cient money for the bare necessities or levels of growth and some time is going “fi nancial problems”; (b) lack of a place to be required to bring average living for living or to improve the living condi- standards to those achieved by the time tions; (c) lack of a job/ well-paid job; (d) of independence. the necessity to pay the studies of one of the family members, and; (e) the opportu- Two possible future scenarios were raised nity for leaving, the fact that someone is by a number of informants. First there is

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction waiting for them in the host country, the anecdotal evidence that people may stay possibility to leave legally etc. On the sup- in host countries and bring their families

Chapter 5: Migration, Employment and Poverty Reduction

124 125 to them thereby drying up the fl ows of Russia in particular, Moldova’s relation- remittances. This is certainly a scenario ships with these countries may be as that is commonly described by close ob- important. This illustrates the problem servers of the situation in Moldova. The of data. While CIS may be responsible evidence from the MRS does not comply for absorbing the most construction and with this fear, indicating that only a small agricultural workers, the EU migrants may percentage of migrants intend to stay in be responsible for the majority of remit- the host countries with the vast majo rity tances or skill used by returned migrants. wanting the return to Moldova once It is therefore very diffi cult to generalise enough money had been saved or remit- without adequate data, disaggregated ted. The MRS study revealed that most by country of destination. Only then can Moldovan migrants (over sixty seven the pros and cons of changing contexts percent) intend to return to Moldova be analysed and diff erent scenarios de- after making enough money abroad. veloped by government as the basis on While in a static sense this would lead to a which to make policy. fall in migration in a dynamic way it may not as re turning migrants may simply be Moldova has agreements with Ukraine, replaced by new migrants. Nearly sixty Belarus and the Russian Federation but percent of respondents stated that they also two agreements within the frame- or a member of their family intended to work of the CIS. The fi rst from 1995 pro- migrate in the near future. Of these over vides a degree of protection for migrant half intended to migrate to Russia. Finally, workers and their families. The second, the basic question concerning the inten- more recent agreement (2002) relates to tion to settle in the destination country for cooperation on fi ghting illegal migration good revealed that only 11.4 percent said (IPP 2005). Immigration cards were intro- they would. Additional fi n dings from the duced in Russia in early 2003 to better Public Opinion Barometer (August 2000 control migration. Although these meas- through November 2003) conducted by ures will allow migrants to work legally, it the Moldovan Institute of Public Policy is reported that a large number have not (IPP) confi rms the tendency suggesting chosen to register (Dudwick and Sethi). that about eighty percent of the population wants to leave the country on a per- In 2003 the EU introduced the European manent or temporary basis. For the young Neighbourhood Policy (ENP) to enhance the fi gure is higher with almost ninety relations with some of its neighbours percent of young people between ages (Ukraine, Moldova, Belarus and the South- of eighteen-twenty wanting to leave the ern Mediterranean countries). Moldova Republic of Moldova for a period. Other has now developed an action plan the fi ndings seem to confi rm the scenario: framework of the ENP which includes the Barometer found that, if there were the objective of ensuring the effi cient ma- a possibility, about thirty seven percent nagement of migratory fl ows, including would want to leave the country forever. initiating the process towards conclusion Only nine percent of young people want of a readmission agreement between to stay in the country. the European Community and Moldova. Moldova’s relationship with the EU will A second fear is that the demand for mi- change when its neighbour Romania grants may fall over time either through gains membership, likely to be in 2007. labour market changes in host countries Although Moldovans require passports or through increasing diffi culties for to enter Romania, unlike EU countries and migrants in terms of policing. The latter other EU-accession countries, Romania would obviously aff ect irregular migrants. does not require visa from Moldovans.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction While most concerns relate to the EU with At the same time Moldovans may ac- most migrants going to CIS countries and quire Romanian citizenship if they had

Chapter 5: Migration, Employment and Poverty Reduction

124 125 grandparents who lived on Romanian impact of migration. But there will be soil before 1918, when Moldova became tradeoff s and a balance will need to be part of Romania. Martin (2004) reports the achieved. Diffi cult decisions will need to estimate that at least 500,000 Moldovans be made and in some cases, such as the had acquired Romanian citizenship, mak- complex issue of the negative social ef- ing them dual nationals69. In addition, fects of migration, the impact is going to he reports estimates that two million be diffi cult to quantify. What is needed Moldovans, almost half the population, is a wide-ranging consultative process could obtain Romanian passports (allow- of the type that should have been un- ing them to spend up to 3 months in the dertaken during the preparation of the EU without a visa); that 140,000 Molda- PRSP to determine what kind of society vian citizens also have Russian citizenship is needed and what direction Moldova is in 2003; 60,000 to have Israeli citizenship, going in. Future policy directions need and; 20,000 to 30,000 Ukrainian citizen- to be based around a vision of the role ship. Studies of the EU labour Market of migration in Moldovan society. Will suggest that demand for migrant labour Moldova be a society dependent (and will not decrease in the near future and in therefore vulnerable) on migration and fact is likely to rise, although more so for remittances, be willing to put up with skilled rather than unskilled labour. (Dou- some of the downsides of migration deijns and Dumont, 2003) (such as on social poverty) in return for the possibly of a quicker return to pre- Moldova is faced with a number of pos- transition living standards. There also sible scenarios and it may not be able to needs to be recognition that while there choose the one it would like. It is clear are policies that may have an important that the status quo cannot continue. The impact on growth, employment and social costs of traffi cking and associated poverty reduction there are limitations negative aspects of irregular migration of government action not only in the are not palatable for society. A reduc- sense of infl uencing individuals in a soci- tion in migration may slow the economy ety where government is not that trusted even if it improves family cohesiveness but also where exogenously determined and addresses shortages in some parts of policies have a great infl uence on out- the labour market but is unlikely to occur comes. A realistic sense of what can be while Moldova still faces serious domestic achieved needs to be ensured. social and economic problems and while demand is still strong. The likely scenario The Migration Policy Concept of the is one of maintaining high levels of migra- Republic of Moldova was adopted by tion but move to improved management Parliament in October 2002. In Decem- and more regular fl ows. Some negative ber of the same year, Parliament also aspects may remain but policy interven- passed the Law on Migration (IPP 2005) tions can be designed. which included the establishment of a new organisation responsible for migration management. The Migration Department it is being supported by the 5.4 Policy and Possible donor community to improve migration management, develop and strengthen Future Directions strategies and policies and to develop appropriate agreements with destina- It is clear that a policy framework needs tion countries. Although the basic insti- to build on the positive aspects des- tutions are in place but much needs to cribed above and reduce the negative be done. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

69Moldova offi cially allowed its citizens to hold dual nationality in 2003

Chapter 5: Migration, Employment and Poverty Reduction

126 127 Given the importance of migration and re- importance as the impact of migration on mittances it is surprising that the PRSP has planning will be important. An example so little to say about the subject. It does, would be the delivery of health services however, identify three broad directions and education in rural areas where migra- for the future, all of which are important, tion has had a major impact. It should also if a little vague in terms of content, and be noted that since many of the factors deserve the fi nancial and technical sup- that infl uence migration fl ows are exo- port of the international community. genously determined it is important that the government has the tools, capacities First, the government could bring legisla- and information to adapt policies within a tion on migration in accordance with the potentially changing environment. international standards. Concrete actions to implement those conventions already There are other policy measures that can ratifi ed by Moldova are needed. In addi- be used to increase the benefi ts of migration, eff orts could be made to move to- tion and remittances in terms of growth, wards regular migration and the signing employment and poverty reduction. First, of more bilateral agreements with host policies related to the volume, charac- countries is required. In addition more teristics and utilization of remittances. fl exible Memoranda of Understanding Second, eff orts can be made to aff ect may be used for seasonal workers (Martin the volume of remittances. Government 2004). The move towards regularising mi- eff orts to oblige migrants to remit a cer- gration has the advantages already noted tain percentage of income home would in the chapter in terms of the signifi cant be unrealistic even if regular migration social costs of migration and traffi cking, increased. The design of private sector but also in terms of enabling more eff ec- investment instruments that are attractive management and the subsequent tive to migrants is a better approach. For alleviation of other costs. As part of this, example, encouraging commercial banks strengthening migration management is to off er foreign currency accounts and critical and especially ensuring that ad- transactions or encouraging competition equate legislative environment covering in the banking sector to reduce the costs the participating of private sector organi- of remitting funds. Ensuring a well-regu- sations (such as recruitment and tourist lated and eff ective fi nancial sector includ- agencies) involved in the migration proc- ing for non-banking fi nancial institutions ess is in place and implemented. such as micro-fi nance mechanisms is key. A specifi c option may be the develop- The second and third PRSP actions are ment of a foreign currency bond that closely linked and relate to making a ensures the anonymity of the investor comprehensive study on population mi- and would be more attractive to migrants gration and its consequences, and the who are concerned about the tax or other creation of an informational system and implications of remitting funds (Moreno- the improvement of the management of Fontes Chammartin, G. and F. Cantu-Ba- migration processes, respectively. There zaldua 2005). The government can also is need for better information on migra- play a role in ensuring that information on tion for more eff ective policy making. Not fi nancial services are available to migrants only should there be consolidation and and are reliable. reconciliation of the information already available but also better coordination In addition to the macroeconomic im- and cooperation among stakeholders pact of remittances, the utilization for involved in information collection and remitted funds for productive purposes analysis, including international organi- is important for growth, employment n

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction sations. At this stage the collection and poverty reduction. Many of the actions analysis of migration information is of vital described elsewhere in this report with

Chapter 5: Migration, Employment and Poverty Reduction

126 127 regard to the appropriate macro-eco- growth and potential returns from invest- nomic framework, especially investment ment is an example of the former, while patterns, and policies for employment training specifi cally aimed at migrants is promotion and income generation are as a case of the latter. What is important is important for migrants as non-migrants. that these innovations are part of a well Putting these policies in place, especially managed strategy that includes piloting, those related to the agricultural sector, learning through eff ective and evalua- is the priority for encouraging increased tion of the interventions and revision of investment of migrant remittances. In the schemes in light of lessons learned. addition to these specifi c instruments Since the ability of government to infl u- can be designed to attract migrant funds ence the investment choices of individu- for investment or to ensure the more ef- als through such schemes may be limited, fi cient use of investment resources. The what is not needed is an ad hoc sserieseries ooff provision of information on likely sector interventions. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 5: Migration, Employment and Poverty Reduction

128 129 C h a p t e r 6 : Employment, Productivity and Proverty

6.1 Employment and Poverty The challenge for Moldova is the creation of employment opportunities associated For the majority of poor people, the only vi- with increased productivity so that the em- able escape route out of poverty in the long ployed population can obtain incomes suf- term is through employment. Economic fi cient to escape poverty. At the same time, growth is essential for reducing poverty in ways must be found to equip the poor with the requisite skills and assets so that they a sustained manner, and does so through are able to avail of the opportunities for generating employment that provides liveli- productive employment as and when they hoods above the poverty line. When growth arise. This chapter analyses the nature and creates ample opportunities of employment extent of this challenge. and the poor are able to grasp those opportunities, then poverty declines in a sustain- After 1999, the rate of unemployment fell able manner. This is as true in Moldova as from eleven to seven percent, and the ab- elsewhere. However, merely creating jobs solute number of unemployed people also may not be suffi cient. The main problem in declined. Unemployment had risen in the Moldova is that well-paid jobs are not being wake of the Russian crisis, from 167 thousand created in suffi cient numbers to rapidly lift in 1998 to 187 thousand in 1999. Then it de- the working population out of poverty. clined to 110 thousand by 2002, an impressive forty percent decline in a span of just three Of the people who were in poverty in 2002, years (Table 6.2). An unemployment rate twenty percent were outside the labour of seven to eight percent is not high by the force, and only two percent were unem- standards of either the developed countries ployed, implying that about seventy-fi ve of Europe or other transition economies. percent had work of some sort (Table 6.1). The problem was that their work did not A closer look reveals a less positive picture. pay suffi ciently to escape poverty. This The fall in unemployment did not refl ect a need not imply that working brought no corresponding rise in employment, since the benefi ts. Those with a formal sector job had number of employed people remained ap- a lower rate of poverty (thirty seven per- proximately constant at about 1.5 million from cent) compared to the unemployed (sixty 1999 to 2002.69 This refl ected in part a steady percent), and compared to those outside decline in the working age population, pri- the workforce (forty six percent). The last marily a result of out-migration (see Chapter group received pension support from the 5), and partly a refl ection of an increase in the numbers of people becoming economically government. So, a job helped to attenuate 70 the burden of poverty, but for many workers inactive. Thus, the main reason why unem- it was not suffi cient to move them and their ployment has fallen is that many Moldovans have either chosen to migrate abroad, or have households out of poverty. become economically inactive.

69 In 2003, total employment dropped to 1.35 million, with the fall concentrated in agriculture. 70The activity rate declined steadily from 61.3 percent in 1999 to 51.6 percent in 2003. While both the size Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction of the labour force and the number of employed persons declined in absolute terms, the number of inactive persons increased from 1.06 million in 1999 to 1.38 million in 2003.

Chapter 6: Employment, Productivity and Poverty

128 129 Another problem is that the average number It was noted above that the decline in the of hours worked per week is very low, about rate of unemployment was not the result thirty hours for the economy as a whole. This of employment growth, either for men or refl ects substantial ‘hidden’ underemploy- women. After 1999, the employment rate (the ment. There have been signs that the extent employed as a percentage of the working of involuntary underemployment (manifes- age population) remained almost constant at ted in short working time) may be declining: just over forty two percent for men, and forty the length of the working week rose from percent for women. A somewhat longer time 26.6 hours in 1999 to thirty hours in 2002, perspective reveals that the recent stagnation although this is still below the normal level in employment was part of a trend through- in Western Europe. Open unemployment is out the transition in Moldova. In 1996, total more serious in urban areas, but underem- employment was 1.66 million workers. This ployment is the greater problem in rural ar- declined to 1.5 million in 1999 following the Russian crisis, and did not recover thereafter, eas. Statistics for 2002 show that unemploy- even though aggregate output expanded ment was 12.1 percent in urban areas, and substantially during 2000-2004. three percent in rural areas. By contrast, the average number of hours worked per week Table 6.1: Poverty by Work Status, 2002 was only twenty four in agriculture, com- Poverty % working age pared to thirty for the economy as a whole. Employment Status The problem of fi nding gainful employment rate (%) population remains pervasive throughout the country, Formal sector 37.4 25.0 although its manifestation diff ers in urban Informal (non-agri) 47.3 5.0 and rural areas. Informal (agri) 40.2 8.0 Measured unemployment is more of a Multiple jobs 25.0 3.0 problem for men than women. In 2002, Household self- 8.1 percent of the male labour force was 46.9 27.0 unemployed, compared to 5.5 percent of employment Other unpaid the female labour force. The unemploy- 63.8 10.0 employment ment rate declined for both men and women at about the same rate. As a re- Unemployed 59.9 2.0 sult, the rate for men has remained higher Out of labour force 46.5 20.0 than for women since 1999, and the relative gap has not changed. Total 45.2 100.0 SOURCE: World Bank (2004)

Table 6.2: Employment and Unemployment, 1997-2003 1996 1997 1998 1999 2000 2001 2002 2003 Absolute number (thousands) Population 3599 3654 3652 3646 3639 3631 3623 3612 Labour force 1686 1671 1809 1682 1655 1617 1615 1474 Employment 1660 1646 1642 1495 1515 1499 1505 1356 Unemployment 167 187 140 118 110 118 Percentage of labour force Employment 98.5 98.5 90.9 88.9 91.6 92.8 93.0 92.0 Unemployment n.a. n.a. 9.1 11.1 8.4 7.2 7.0 8.0 SOURCE: DSS 2003 NOTES: From 1998, the ILO defi nition of unemployment has been added to employment fi gures to Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction obtain labour force estimates. Thus, statistics for the proportion of employment before and after 1998 are not comparable.

Chapter 6: Employment, Productivity and Poverty

130 131 6.2 Employment, Wages However, the picture changed signifi - cantly in 2003. Agricultural employment and Productivity dropped precipitously, by more than twenty percent compared to the level in the previous year. Changes in the other Employment and Wages two sectors were relatively minor, with the result that overall employment de- To understand the employment problem clined by almost ten percent. Over the in Moldova, it is necessary to analyse the whole post-crisis period from 1999 to structure of employment, its distribution 2003, overall employment declined large- across economic sectors, and the change ly due to a fall in agricultural employment in that distribution over time. In 1996, the (Table 6.5). service sector was the leading employer of the labour force, accounting for a share While employment has either stagnated of forty fi ve percent of the total (Table 6.3). or declined in most sectors, output and Agriculture followed closely behind, with value-added have increased substantially. a forty three percent share, and indus- After 1996, GDP grew at 1.8 cent per an- try came a distant third, with a share of num, while employment declined at the about twelve percent. As the Russian crisis annual rate of 2.9 percent. The diff erence deepened, agriculture became the leading between positive GDP growth and nega- employer, absorbing many workers laid off tive employment growth is even greater from the other two sectors. This trend con- during the post-crisis recovery period. Be- tinued until 2002, when agriculture em- tween 1999 and 2003, overall GDP grew ployed half the labour force, and service’s at four percent per annum, while employ- share fell to forty percent. In 2003, how- ment declined by 2.4 percent (Tables 6.5 ever, the share of agriculture fell sharply to and 6.6). forty three percent, making services once again the largest employer, with forty fi ve Among the three broad sectors of the percent, as in 1996. During these years, the economy, the diff erential between out- share of industry changed very little, fl uc- put and employment was the greatest in tuating between ten and twelve percent. industry, where nearly an eight percent growth of value added after 1999 was The long-term decline in the number of accompanied by a meagre 0.6 percent employed was shared across all three growth in employment. In the service sectors (Table 6.4). In 2003, the level of sector, value-added increased at four employment stood at eighteen percent percent per annum along with stagnation below the 1996 level: eighteen percent of employment (only a 0.2 percent rate of below for agriculture, sixteen below for growth). The contrast was even greater industry and nineteen for services. After for transport: twelve percent annual the Russian crisis, employment recovered expansion of value-added in contrast to in both agriculture and industry, but not negative employment growth (of almost in services. This may indicate that the minus one percent) (see Tables 6.5 and services category included many jobs 6.6). taken only when no viable alternative presented itself. Industrial employment The divergent movement in employment was seven percent higher in 2002 than and output growth in other sectors mani- in 1999 (Table 6.5). Since industry was the fested itself in agriculture as well, but with smallest of the three sectors, the eff ect of a diff erence. Between 1999 and 2002, val- this mild increase on overall employment ue-added in agriculture grew at the rapid was minimal. More signifi cant was the rate of 4.3 percent per annum, whereas tendency for employment to rise slightly

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction employment increased very slowly, at 0.7 in agriculture. percent. This was similar to the output-

Chapter 6: Employment, Productivity and Poverty

130 131 employment growth relationship in the Along with the slow growth of employ- other two sectors. However, the relation- ment and rising productivity went an ship changed dramatically in 2003, when increase in real wages beginning in the measured value-added in agriculture fell mid-1990s. This trend was reversed during by nine percent compared to the previous the Russian crisis, to rise again thereafter. year, and employment dropped by a disa- The apparently high rate of increase from strous twenty two percent. This simulta- 1996 to 2003 (almost eight percent per an- neous fall in production and employment num (Table 6.8)) and during the recovery raises a question of causality: whether period after 1998 (14.7 percent) must be employment fell because of a depression placed in the context of the catastrophic in output, or whether a drop in the rural collapse in the 1990s. To such depths had labour force due to out-migration engen- real wages fallen, due to output decline dered an output collapse. It would appear and infl ation, which their low levels un- that so many agricultural workers migra- doubtedly undermined productivity. ted from Moldova that a labour shortage might have become a critical bottleneck The rise in real wages during 1996-2003 for some activities, thereby constraining occurred at similar rates across all the the expansion of output (see Chapter 5).71 major sectors of the economy: 5.4 percent per annum in agriculture, 6.3 percent in The employment-production relationship industry, and 5.8 in services (Table 6.9). has implications for the growth of pro- Some variation is observed, however, ductivity and labour income. The failure among the sub-sectors within services. of employment to keep pace with pro- In transport and construction, real wages duction meant that labour producti vity grew at ten and eight percent, respec- increased, quite sharply in some cases. tively. The lowest rates of increase were in Over the period 1996 to 2003, producti- trade and the public sector (health, edu- vity of labour grew at 4.8 percent per an- cation, and public administration). num for all sectors (Table 6.7). The most rapid growth occurred in services (6.9 Given the decline in the real value of mo- percent per annum), especially in trade- ney and the social wage during the fi rst related services (11.5 percent). Industry half of the 1990s, it is quite possible that had the slowest growth, at 2.2 percent. the living standards of workers dropped Agriculture had 3.9 percent growth.72 to such a low level that worker produc- Considering only the recovery period tivity was undermined due to ill-health from 1999, the overall picture remains and malnutrition. Thus, that real wages the same although the relative positions seemed to have outpaced the growth of of the sectors change quite a bit. For the labour productivity could refl ect a process economy as a whole, labour productivity of adjustment toward a more sustainable grew at the rate of 6.7 percent per annum. ‘effi ciency wage’ level. While the average The highest rate of growth of productivity real wage for the economy as a whole was observed in agriculture and industry increased at the rate of 7.7 percent from (both just over seven percent per annum). 1996 to 2003, labour productivity grew at Productivity in services grew at the rate the rate of only 4.8 percent (Tables 6.7 and of 4.2 percent. Within services, transport 6.8). Looking separately at the sectors of emerged as the leading sector, with a production, both agriculture and industry growth rate of 12.9 percent, whereas pro- saw real wages grow faster than produc- ductivity in trade declined. tivity. Only in services did the opposite

71 The hypothesis of excessive migration out of agriculture seems eminently plausible in view of the signifi - cantly lower levels of productivity and wages that prevail in agriculture compared to other sectors. See the discussion below.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 72However, if the special case of 2003 is excluded, then up to the year 2002 productivity grew in agriculture only at the rate of 1.8 percent, by far the lowest among all the sectors.

Chapter 6: Employment, Productivity and Poverty

132 133 happen. However, during the recovery and if eff ectively implemented, would period since 1999, all three sectors saw reduce the earnings gap between men real wages growing faster than producti- and women. However, enforcement is vity (Tables 6.7 and 6.8). not eff ective, perpetuating the inequality despite enlightened legislation. A new Agricultural real wages were the lowest draft Law on Gender Equality has been among all sectors, and the gap with wa- prepared and may be approved in 2005. ges in other sectors widened slightly over While this would extend the prohibition time. In 1996, the real wage in agriculture against discrimination, it would not ad- was two-thirds of the national average; by dress the implementation and enforce- 2003, it had fallen to below sixty percent ment problem. As important as formal (Table 6.9). Within the service sector, the legislation is, there is still a ‘culture of lowest wages were in health and educa- sexism’ in Moldova. Many men, as well as tion. The highest real wages; namely, some women, are notably insensitive to those in the fi nancial services sector, discrimination against women. As long as were fi ve times above those in agricul- this continues, substantially reducing dis- ture. These inter-sectoral variations in real crimination against women in the labour wages refl ect corresponding diff erences market will be very diffi cult. in labour productivity. Agriculture had the lowest level of labour productivity among Legislation against pay discrimination, all the major sectors of the economy, and for example, is easily evaded within a dis- the gap with the national average wi- criminatory culture that relegates women dened over time (Table 6.9). to low-paid jobs. Enforcing equal pay within occupations and skill categories Changes in the structure of employment has limited impact on salary diff erentials in an economy tend to have diff erent im- if women are segregated into the lowest pacts on men and women, as a result of positions. Strong leadership by the go- the discrimination against the latter that vernment is necessary to overcome such relegates (if not segregates) them to low- discrimination. The public sector should paying jobs. As in other countries, women set the example for the private sector. in Moldova dominate in health and social Moldovan women perform an enormous work, education, and hotels and restau- amount of unpaid work within the family, rants. Surprisingly, they also dominate in and play a leading role in many commu- fi nance, which pays the highest monthly nity organisations. At the same time, they salaries among sectors but represents a are absent from most decision-making, very small proportion of total women’s partly because of the lack of coherent and employment; namely, less than two per- supportive state policies and partly be- cent. Hotels and restaurants employ a cause of a patriarchal mentality that still similarly small percentage of all female adversely infl uences the entire society, workers. Health, social work and educa- including women themselves. tion, on the other hand, account for over forty percent of women workers although they are among the lowest paid sectors in Productivity and Effi ciency the country. Similar to trends for trends for wages, The average monthly wage in 2003 was labour productivity was the highest in 891 Leu. Approximately fi fty seven per- industry, followed by services, with ag- cent of women worked in sectors with riculture a distant third. In 2003, labour wages below this average (see Table 6.10); productivity in agriculture was one-third while for men the percentage was forty of that in industry and two-thirds of that

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction seven. The anti-discrimination legisla- in the service sector. Correspondingly, tion of Moldova is quite strong on paper, agricultural wages were forty percent

Chapter 6: Employment, Productivity and Poverty

132 133 of industrial wages and sixty percent elsewhere in the economy. As a result, the of wages in services. Within services, pressure to migrate abroad continued, labour productivity was the lowest in keeping the measured rate of unemploy- public administration and the highest in ment down, but at considerable social cost fi nancial services, the same pattern ex- (see Chapters 4 and 5). hibited for real wages (Table 6.11). Given the well-known problems of measuring Stagnant or falling employment coin- productivity in services, especially in the cided with rising labour productivity and public sector where workers typically do increased real wages in all sectors of the not produce a marketed output, these economy. Workers who managed to retain comparative fi gures must be interpreted or fi nd employment gained from these cautiously. The need for caution is sug- trends, though the vast majority of them gested by the calculation that the serv- remained at extremely low, if not poverty- ice sector had the most consistent and assuring, wage levels. Combined with gov- fas test growth of productivity compared ernment policy to protect the pensioners, to both agriculture and industry, but this the rise in real wages helped reduce diff erence was not refl ected in the relative poverty in some years (see Chapter 4). If growth of real wages. Between 1996 and employment does not recover robustly 2003, measured labour productivity grew in the future, it is likely that real wages will in services almost three times faster than cease to rise, and productivity gains will in industry, and yet average real wages in accrue to the owners of capital. However, services grew at a slower pace (5.8 per- if wages rise, the number of poverty-level cent per annum in services compared to households supported by wage labour 6.3 percent for industry). will decline. Thus, for the majority of poor households to benefi t from the growth The preceding analysis suggests that the process, it is essential to substantially raise Moldovan economy failed to raise the le- the overall level of employment. vel of employment suffi ciently to achieve poverty reduction, even when output The Moldovan transition to a market-based recovered. Offi cial unemployment fi gures economy is far from complete. This is high- do not refl ect this failure, because of mas- lighted by the fact that the level of output sive out-migration and a growing propor- remains below the pre-transition level. tion of economically inactive people. The Since enterprises have probably shed most stagnation of employment levels and the of the excess labour inherited from central existence of severe underemployment planning, employment increases depend bear adequate testimony to this failure. now on the diff erential between the growth rates of output and productivity. Analysis To some extent, the failure of employment of enterprise level data reveals a positive to recover fully, even during a period of out- relationship between labour productivity put recovery, should be expected in a tran- and employment creation in Moldova, not sition economy. As in all transition coun- the contrary. Enterprises with higher labour tries, enterprises in Moldova carried more productivity have been the ones that have workers before the transition began than created more jobs (Rutkowski 2004). But would be justifi ed under market criteria. only a small number of enterprises have As market forces impinged on enterprises, such an achievement. These have usually they began to shed this excess labour. As a been newly emerging enterprises in the result, employment growth lagged behind private sector, rather than privatised enter- output growth for some time. However, prises. These new enterprises are typically it appears that the Moldovan economy small and non-agricultural. Poverty reduc- has completed this process. Employment tion requires that this productivity-driven

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction fell in agriculture, and this decline was employment generation spread more not matched by employment expansion broadly across sectors and enterprises.

Chapter 6: Employment, Productivity and Poverty

134 135 Table 6.3: Employment by Economic Sectors, 1996-2003

1996 1997 1998 1999 2000 2001 2002 2003 Agriculture 42.8 41.6 45.7 48.9 50.6 51.0 49.6 43.0 Industry 11.7 11.6 11.1 10.7 10.6 11.0 11.4 12.1 Services 45.4 46.8 43.2 40.4 38.8 38.0 39.0 44.9 Construction 3.3 3.2 3.5 2.9 2.9 2.9 3.1 3.9 Trade 16.3 16.8 12.7 10.0 10.8 10.9 11.6 13.0 Transport 4.0 4.4 4.7 4.7 4.2 4.3 4.1 5.0 Public admin 16.8 17.6 17.5 17.8 16.5 16.1 16.1 18.0 Others 5.0 4.9 4.8 5.0 4.4 3.9 4.1 5.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

SOURCE: This table, Tables 6.4-6.9 and Table 6.11 rely on data, or calculations based on data, from the DSS.

Table 6.4: Index of Employment by Economic Sectors, 1996-2003

1996 1997 1998 1999 2000 2001 2002 2003 Agriculture 100.0 96.2 105.5 102.8 107.7 107.5 105.1 82.0 Industry 100.0 97.9 93.3 82.1 82.6 84.6 87.7 84.1 Services 100.0 102.3 94.2 80.0 78.0 75.6 77.9 80.8 Construction 100.0 94.5 105.5 80.0 80.0 78.2 83.6 96.4 Trade 100.0 101.8 76.8 55.4 60.1 60.5 64.6 64.9 Transport 100.0 109.1 116.7 106.1 97.0 97.0 93.9 103.0 Public admin 100.0 103.9 103.2 95.3 89.6 86.4 87.1 87.5 Others 100.0 97.6 95.2 89.2 80.7 69.9 73.5 81.9 Total 100.0 99.2 98.9 90.0 91.3 90.3 90.7 81.7

Table 6.5: Index of Employment during the Recovery Phase, 1999-2003

1999 2000 2001 2002 2003 Rate of growth Agriculture 100.0 104.8 104.5 102.2 79.8 -5.5 Industry 100.0 100.6 103.1 106.9 102.5 0.6 Services 100.0 97.5 94.5 97.3 101.0 0.3 Construction 100.0 100.0 97.7 104.5 120.5 4.8 Trade 100.0 108.7 109.3 116.7 117.3 4.1 Transport 100.0 91.4 91.4 88.6 97.1 -0.7 Public admin 100.0 94.0 90.6 91.4 91.7 -2.1 Others 100.0 90.5 78.4 82.4 91.9 -2.1

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Total 100.0 101.4 100.3 100.7 90.8 -2.4

Chapter 6: Employment, Productivity and Poverty

134 135 Table 6.6: Index of Sectoral GDP at 1996 Prices, 1999-2003

Rate of 1999 2000 2001 2002 2003 growth Agriculture 100.0 101.9 109.2 115.2 105.2 1.3 Industry 100.0 110.3 117.8 118.1 134.4 7.7 Services 100.0 95.0 99.1 109.3 118.8 4.4 Construction 100.0 69.7 86.7 91.6 112.5 3.0 Trade 100.0 76.4 77.6 80.1 88.1 -3.1 Transport 100.0 121.4 132.4 144.9 157.6 12.1 Public admin 100.0 100.3 99.6 116.6 118.0 4.2 Others 100.0 114.9 121.6 137.5 155.5 11.7 Total 100.0 100.3 106.1 113.0 117.4 4.1

Table 6.7: LLabourabour ProductivityProductivity byby SectorsSectors andand AnnualAnnual GrowthGrowth Rates,Rates, 1996-20031996-2003 (1996 = 100)

Rate of 1997 1998 1999 2000 2001 2002 2003 growth Agriculture 116.5 99.8 98.8 96.1 103.2 111.4 130.3 3.9 Industry 91.1 81.0 88.7 97.3 101.3 98.0 116.4 2.2 Services 99.0 106.8 135.2 131.7 141.7 151.8 159.1 6.9 Construction 88.9 67.9 93.1 64.9 82.6 81.6 87.0 -2.0 Trade 92.6 143.3 285.0 200.5 202.3 195.7 213.9 11.4 Transport 96.0 82.4 88.2 117.1 127.7 144.2 143.1 5.3 Public admin 106.3 100.1 86.4 92.2 95.0 110.2 111.1 1.5 Others 101.5 108.3 139.3 176.7 216.1 232.4 235.7 13.0 Total 102.3 96.5 107.5 106.4 113.7 120.6 139.1 4.8 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 6: Employment, Productivity and Poverty

136 137 Table 6.8: Real Wages by Sectors, 1996-2003 (1996 = 100):

Rate of 1997 1998 1999 2000 2001 2002 2003 growth Agriculture 99 96 84 94 107 127 144 5.3 Industry 111 120 106 107 117 136 154 6.3 Manufacturing 111 117 104 109 119 136 152 6.2 Mining/quarry 104 103 90 85 103 129 135 4.4 Utilities 112 130 109 97 108 132 159 6.9 Services 102 106 88 86 106 130 148 5.8 Construction 118 121 102 99 114 133 170 7.8 Trade 104 102 89 85 91 119 135 4.4 Transport 112 121 105 112 138 161 198 10.3 Finance 109 138 147 158 139 149 152 6.1 Real estate 103 112 105 102 122 142 162 7.1 Health/educ 97 95 72 69 86 113 133 4.1 Public admin 99 109 88 80 104 132 133 4.2 Others 99 112 83 80 96 118 141 5.0 Total 105 110 97 99 120 146 168 7.7

Table 6.9: Relative Real Wages by Sectors, 1996-2003 (1996 prices, weighted sectoral average = 100)

1996 1997 1998 1999 2000 2001 2002 2003 Agriculture 65.2 61.5 56.3 56.6 61.9 57.9 57.0 56.0 Industry 155.6 165.4 169.3 169.7 167.8 151.9 145.0 142.5 Manufacturing 150.8 160.3 159.3 161.2 166.6 149.3 140.7 136.4 Mining/quarry 164.7 163.1 153.7 151.8 142.0 140.9 145.9 132.3 Utilities 181.3 194.5 213.6 203.4 176.9 163.3 164.3 172.0 Services 113.7 110.3 109.0 102.9 98.6 100.1 101.7 100.3 Construction 132.6 149.0 144.5 139.3 132.7 125.4 121.3 133.9 Trade 111.8 110.7 103.4 102.0 95.6 84.5 91.2 89.7 Transport 138.0 147.6 150.5 148.8 156.0 158.1 152.7 163.0 Finance 362.6 375.4 453.1 547.2 578.1 418.4 371.1 327.2 Real estate 132.1 129.8 134.1 142.6 136.1 133.7 128.8 127.3 Health/educ 85.0 78.8 73.1 62.5 59.5 60.4 65.9 67.3 Public admin 158.3 149.0 156.5 143.6 127.3 136.3 143.2 125.7 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Others 89.8 84.7 90.6 76.9 72.7 71.8 73.1 75.4

Chapter 6: Employment, Productivity and Poverty

136 137 Table 6.10: Wages and Employment by Sector and Gender, 2003

Average Women as % % of Sector Monthly of total workers women workers Wage (Lei) by sector by sector Financial Activities 2926 63 1.4 Transport and Communications 1454 31 7.3 Industry 1271 47 18.8 Construction 1194 17 2.4 Real estate activities 1133 44 3.9 Public Administration 1050 37 8.5 Hotels and Restaurants 827 76 1.0 Trade 795 54 5.2 Other public 671 56 2.8 Education 610 75 18.7 Health and Social Work 579 80 9.5 Agriculture and Forestry 499 43 20.6 Average 891 Total 100%

SOURCE: DSS (2003a).

Table 6.11: Relative Labour Productivity by Sectors, 1996-2003 (Weighted sectoral average = 100)

1996 1997 1998 1999 2000 2001 2002 2003 Agriculture 73.3 83.4 75.8 67.37 66.2 66.5 67.7 68.7 Industry 224.4 199.7 188.2 185.12 205.2 199.9 182.4 187.7 Services 93.0 90.0 102.9 116.97 115.2 115.9 117.1 106.4 Construction 131.7 114.4 92.7 114.05 80.3 95.7 89.1 82.4 Trade 63.2 57.2 93.9 167.58 119.2 112.4 102.6 97.2 Transport 161.4 151.4 137.7 132.33 177.7 181.2 193.0 166.0 Public admin 86.9 90.3 90.2 69.79 75.3 72.6 79.4 69.4 Others 130.9 129.9 146.9 169.57 217.6 248.9 252.3 221.9

6.3 Generating Wage and this rate has slowed down in the fi rst half of the 2000s. As should be expected, Employment the rate of growth dropped sharply after the Russian crisis, but it did not recover Moldova has experienced a slow rate of thereafter. In 1997, new enterprises grew

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction growth of new enterprises, even by the at ten to fourteen percent, depending on standards of other transition countries, the legal status of the fi rm; in 2002, the

Chapter 6: Employment, Productivity and Poverty

138 139 rate fell to fi ve to seven percent (Table The transition to a market-based eco- 6.12). These rates can be contrasted with nomy throughout Central and Eastern average rates of sixteen to eighteen per- Europe resulted in devastating job decent in the transition economies of Cen- struction, in notable contrast to transi- tral Europe (Eurostat 2000). tion in Vietnam and China. This difficult transition led to loss of output on the The rate of fi rm closure in Moldova has one hand and continued hardship of the been one of the lowest among European working class on the other. The extent and Central Asian transition economies. of job destruction in Moldova was typi- The sluggishness in the growth of new cal of European transition economies. enterprises, rather than the closure of Atypical was the slow rate of new job exis ting enterprises, has contributed growth, either in the form of the crea- more to the stagnation in employment. tion of new firms or the expansion of However, enterprise-level data for 2001 existing ones. show that expanding enterprises created jobs at a slower rate than the contrac- Identifying the major impediments to ting fi rms lost them (Table 6.13). While employment growth is essential to de- average employment increased by thir- signing eff ective employment policies. teen percent in the expanding fi rms, it Rutkowski (2004) explored alternative declined by sixteen percent in the con- hypotheses to explain the slow rate of tracting ones, and the contracting fi rms job creation in Moldova. In particular, outnumbered the expanding ones by a he assessed the importance of so-called ratio of four to three. ‘labour market rigidities’,73 or the costs of hiring and fi ring workers, which orthodox The preponderance of contracting en- analysts argue has been responsible for terprises over expanding ones means the relatively slow pace of job creation in that in the formal sector job destruction Western Europe. He found that since the outpaced job creation. In 2001, expand- rate of job destruction in Moldova was no ing enterprises added 6.7 percent to total less than in other transition economies, employment, while the contracting ones there is no reason to believe that the subtracted 11.2 percent. The net result costs of hiring and fi ring posed a serious was that employment shrank by 4.5 per- problem. The Moldovan labour codes cent. The relative preponderance of con- appear to make dismissals diffi cult, by tracting fi rms is not unique to Moldova. requiring, for example, the approval of The imbalance, however, has been much trade unions. Surveys of employer opi- greater, because of a low proportion nions also indicated that they consider of expanding enterprises. For example, the procedural and monetary costs of while the proportion of contracting en- dismissals an important factor in redu- terprises in 2001 was virtually the same cing incentives to create new jobs (ProEra in Moldova as in Lithuania, the proportion 2002). However, the high rate of job de- of expanding enterprises was only thirty struction suggests that the provisions of percent in Moldova compared to thirty the Labour Code are honoured more in seven percent in Lithuania. As in the case the breach than in implementation. For of enterprise turnover, so for enterprise example, the requirement that employ- size: it is not so much the severity of job ers consult with workers’ representatives destruction as the sluggishness of new on dismissals has had little practical con- job creation that has proved the main sequence because of the weakness of the drag on employment in Moldova. trade union movement.

73This concept suff ers from a one-sided interpretation in practice, in which virtually all ‘rigidities’ are those Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction labour practices and regulations that are in the interests of employees, and not ‘rigidities’ arising from ineffi cient management.

Chapter 6: Employment, Productivity and Poverty

138 139 In addition to labour market ‘rigidities’, collapse in the real incomes of its people, Rutkowski also examined the importance and recovery has been slow since then of other factors, such as access to credit, (see Chapter 2). Foreign demand also export orientation, ownership structure, collapsed in the wake of the disintegra- capital density, and investment. He found tion of the Soviet trading system, and has no systematic relationship between these not yet been regenerated by tapping into factors and an enterprise’s ability to cre- new export markets. In the face of severe ate new jobs. While allowing for the demand constraints, it is not surprising possibi lity that any relationship may be that new enterprises found it diffi cult to hidden by limitations of the data, he ac- emerge and the existing ones found it dif- cepted the hypothesis that the main bot- fi cult to expand. This demand constraint tleneck in the creation of new jobs lies in must be addressed as a priority if the rate an unfavourable regulatory environment of new job creation is to exceed the high resulting in highs costs of ‘doing business’. rate of job destruction. This conclusion accords well with the prevailing perception of private employers Since job creation rates have varied in Moldova about the problems of ‘gover- among diff erent types of fi rms, it is im- nance’ in business regulations. portant to identify which were more likely to create employment. The evidence for The argument that the regulatory envi- 2001 indicates that, fi rst, each of the three ronment has not been friendly to business major sectors of the economy; agricul- in Moldova, while true, may not be deci- ture, industry and services, created fewer sive, however. The important question jobs than it destroyed (Table 6.14. The is the relative signifi cance of its impact. problem was most severe in agriculture, Rutkowski’s own cross-country evidence and only moderate in services. Though suggests that while the costs of ‘doing services performed relatively well on business’ in Moldova are far higher than net employment losses, in most other in Western Europe, they are not out of transition economies, by contrast, this line with those in other transition econo- sector absorbed the workers displaced mies. The cost of ‘doing business’ cannot, elsewhere in the process of restructu- therefore, be the major explanation of ring. However, this did not happen in Moldova’s massive failure to create new Moldova. employment compared with other transition economies. It would appear that the form of ownership aff ected job losses and creation. We One problem with Rutkowski’s other- consider three types of ownership: state, wise useful analysis is that he ignores collectives, and private.74 In both state the importance of the demand side in and collective enterprises, the rate of job explaining the slow rate of employment destruction exceeded the rate of job crea- creation in Moldova. Supply-side factors, tion. Only the private enterprises were such as production and distribution costs, net creators of jobs, but this category have played a part and reducing these accounted for less than ten percent of costs could form an integral part of any the workforce outside agriculture in the comprehensive employment policy. But early years of the 2000s. It is not unex- Moldova’s failure relative to other transi- pected that new enterprises would be the tion economies cannot be fully under- greater net creators of employment, since stood without underscoring demand-side the statistics include only the successful factors. Domestic demand contracted ones, not those that failed to establish severely in Moldova, due to a massive themselves.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 74 We use the Moldovan statistical categories: collectives were formerly state-owned enterprises that had been either fully or partially privatized; while ‘private’ refers to new enterprises.

Chapter 6: Employment, Productivity and Poverty

140 141 It also appears that enterprise size aff ected stics of an economy on the medium and employment creation. Large and medium long term potential for small enterprises size enterprises showed negative net job to become viable (Weeks 2003). Public creation, while small and micro-enter- policy in Moldova should seek to foster prises, and especially the latter, generated employment in all ways that are effi cient, more jobs than they destroyed. These eff ective and sustainable. A major part smaller enterprises accounted for about of this eff ort would be a vigorous public sixteen percent of the work force in the investment programme that reduces the non-agricultural sector. However, their operating costs of enterprises regardless employment creating record is due to the of their size of ownership. fact that they hardly existed before 1990. Thus, while their record on employment Table 6.12: Enterprise* Openings was encouraging, it is not strictly compa- and Closures, 1997-2002 rable to the record of state enterprises or collectives, which did exist before the tran- Enterprise 1997 1998 1999 2000 2001 2002 sition and took the brunt of adjustment. category Openings Rutkowski repeats the ubiquitous view Individual that small enterprises face discrimination 14.3 11.2 4.1 3.4 3.6 4.8 owners in most developing economies in both public policy and private markets. There Incorporated 10.3 9.3 8.1 8.1 7.8 7.4 is no systematic evidence to support this Closures allegation; on the contrary, there have Individual 1.6 1.6 0.8 2.4 2.8 2.5 been notable examples of public policy owners seeking to foster small enterprises—but Incorporated 1.6 1.1 0.8 2.4 0.7 0.5 with mixed success (Katarak 2002). A study of Bulgaria found that the rapid Turnover growth of small enterprises resulted more Individual 15.9 12.8 4.9 5.8 6.4 7.3 from fi lling the gap left by state enter- owners prises than from any disciplining process Incorporated 11.9 10.3 8.9 10.5 8.4 7.9 of competition.75 More generally, cross country evidence suggests that although SOURCE: Rutkowski 2004 public policy is important, it is much less NOTES: The term ‘natural persons’ and ‘legal per- important than the structural characteri- sons’ are used in the source.

Table 6.13: Expanding and Contracting Enterprises, 2001

Share in enterprises Share in employment Average enterprise Change in Enterprise category (%) (%) size (persons) employment (persons) Contracting 42.0 61.7 84.7 -16.1 Stable 28.5 6.6 13.4 0.0 Expanding 29.5 31.7 61.9 13.0

SOURCE: Rutkowski 2004

75 In the conclusion of the study, the author writes, “It was the initial ‘push eff ect’ of the launch of system reform, and the initial liberalization of the legal framework, that had the most marked impact on the numbers of new start-ups, though this must be viewed in the context of the “hidden privatization” made Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction possible as a result of the disorder following the withdrawal of the state from control of the economy.” (Kassayie 2002, p. 179)

Chapter 6: Employment, Productivity and Poverty

140 141 Table 6.14: Job Creation and Job Destruction by Types of Enterprise, 2001

Gross job Gross job Employment share Sector Net job creation creation destruction (%) All ownership units State 3.0 8.4 -5.4 30.0 Collective 7.0 13.2 -6.2 60.9 Private 16.5 8.6 8.0 9.1 Non-agricultural sector State 2.6 7.0 -4.5 42.1 Collective 8.9 11.6 -2.7 46.4 Private 11.6 7.5 4.1 11.4 Enterprise size Micro (1-10) 37.3 9.5 27.9 3.4 Small (11-50) 13.5 12.6 0.8 12.9 Medium (51-250) 6.8 15.4 -8.6 34.0 Large (250+) 2.8 9.0 -6.2 49.7 Sector Agriculture 3.8 14.0 -10.2 34.0 Manufacturing 7.4 11.9 -4.5 33.3 Services 9.0 9.2 -0.2 32.8

SOURCE: Rutkowski (2004).

6.4 Labour Market Strategy on Labour Force Employment in the Republic of Moldova for 2002-2008. Policies This strategy was meant to fulfi l four over- arching objectives: improving the skills of Mindful of the need for new and more the labour force; developing entrepre- productive employment creation, the neurial capacity; increasing the ‘fl exibility’ government of Moldova has pursued ac- of workers and enterprises; and providing tive labour market policies. Parliament ap- equal opportunities. proved a new version of the Labour Code in 2003, and other laws on employment, The Strategy was to be implemented in two remuneration, and the settling of labour phases, with the fi rst phase in 2002-2004 and disputes were also introduced. Article the second in 2005-2008. In December 2004, 43 of the Constitution of the Republic of the Ministry of Economy provided the fi rst Moldova ensures the right to work and assessment of the progress made in imple- to labour protection for every Moldovan menting the fi rst phase of the strategy. This citizen. In order to improve the labour evaluation indicated that while progress had

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction market and reduce the level of unem- been made, the record on reducing unemployment, the government adopted the ployment and poverty was disappointing.

Chapter 6: Employment, Productivity and Poverty

142 143 Previously, the National Programme for Em- In order to develop and facilitate youth employment, which had been initiated by the ployment, the Government adopted the Ministry of Labour and Social Protection in National Programme for Prevention and 1994, had included labour market analysis Reduction of Youth Unemployment. With and forecasting, special measures for job the fi nancial support of the World Bank, the creation, employment opportunities for National Net for Youth Employment was young people, and training and re-training established, with four regional centres and for the unemployed. The National Employ- eleven resource centres. These resource ment Agency and its territorial bodies were centres provide consulting services, profes- assigned with the task of supplying infor- sional training, and cultural activities. Dur- mation on existing vacancies and short- ing 2002-2003, more than fi fty thousand listing applicants by matching job off ers people took advantage of the consulting to the qualifi cation, aptitudes and interests services and nearly eighty percent of them of applicants. The services provided by the were sixteen-twenty nine years of age. agency also included guidance on occupa- tional choice, support to new enterprises, In order to foster entrepreneurship and promotion of labour mobility, and training. facilitate access to funding for new and However, these eff orts produced little suc- existing small and medium enterprises, cess in fostering job creation. According the government adopted the State Pro- to the Ministry of Economy, only about gramme of Small Business Support for fi ve percent of Moldovan businesses took 2002-2005. The Programme led to the fol- advantage of government incentives, such lowing reforms: as reduced income taxes or social contribu- 1. Simplifi cation of the registration proce- tions, and those that did were enterprises dure for enterprises; that generated mostly low-paid jobs. 2. Streamlining of the list of activities sub- The agency’s training programmes have ject to licensing; also had a limited impact. The government 3. Reduction of the tax burden for small usually contracted out training courses to and medium enterprises;76 and private institutions and fi nanced the costs from the Employment Fund, which has 4. Establishment of a Fund for Entrepre- neurship Support and Small Business been part of the government’s Social Fund. 77 The training was intended to be based on Development; existing and projected demand for diff e- 5. Adoption of the Law on Micro-Finance rent types of labour. In practice, however, Organisations in order to facilitate access there was considerable mismatch between by small and medium enterprises to non- supply and demand. To correct this, the bank micro-fi nance services. National Employment Agency changed the programme to establish a more demand The ‘Entrepreneurial Patent’ has been an led system and create labour supply more additional public measure designed to adapted to enterprise needs. Evidence encourage entrepreneurship in specifi ed indicates that the National Employment activities. It has involved a simplifi ed regis- Agency and its regional offi ces have made tration and taxation system for obtaining progress in ensuring that the unemployed a patent. These patents can be obtained benefi t from job search and guidance serv- by nationals or foreigners from local tax ices at an early stage. authorities or, in certain cases, from local

76 According to the new version of Article 49 (Tax Code, Title II) a small and medium enterprise could be exempted from income tax for the fi rst 3 years (“tax holidays”) if its staff is fewer than 20 persons and its annual turnover no more than leu3 million. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 77In 2004, one million Leu were transferred to the Fund from the state budget and used for guaranteeing credits received by SMEs from commercial banks.

Chapter 6: Employment, Productivity and Poverty

142 143 administrations. In 2003-2004, this pro- agriculture. Because labour markets are gram reached more than eighty thousand ineff ective and ineffi cient, it is through people, of who twenty four percent were trade unions that productivity increases in Chisinau. can be translated into real wage increases, working conditions can be improved and There is lack of information, even under- job security fostered. On-the-job haras- standing, on the part of workers and employ- sment of workers, especially women, ers about the nature of the labour market. can also be reduced. A policy of actively One consequence is that the system of wage strengthening trade unions is especially setting does not encourage workers to ac- important, because Moldova has never quire new skills or facilitate their movement had an independent workers’ movement, between jobs. Many workers have only a much less one with the ability to protect narrow range of technical skills, often specifi c workers’ rights in an eff ective manner. to industries in decline. The National Employ- Research covering almost twenty coun- ment Strategy is trying to address this prob- tries, including transition economies, has lem by enlisting the help of employers in the shown ‘[c]ompanies that score high in design and delivery of training and requiring terms of the human development enter- them to share the costs and responsibilities. prise index tend to do better in terms of economic and employment performance’ The ineffi ciency of the labour market in (Standing 1997a).78 Moldova in allocating workers among jobs should not be interpreted as market The Ministry of Labour and Social Welfare ‘infl exibility’ in the orthodox sense of the is committed to the protection of working term. Whatever may be the problems with conditions and workers’ right to organise, the labour market, it is not ‘infl exible’ with but lacks suffi cient funding to enforce the regard to wage ‘adjustment’. As shown in this country’s progressive labour legislation. chapter and Chapter 2, real wages collapsed Adequate funding for the ministry would in the fi rst half of the 1990s, indicating, be an essential element, perhaps a neces- if anything, excessive downward wage sary condition, for the growth of employ- ‘fl exibility’. However, this real wage fl exibility ment providing earnings that would lift was not matched by employment fl exi bility: wage earner households out of poverty. the wage collapse was associated with a simultaneous collapse of employment. That is, falling wages did nothing to boost employment, contrary to the predictions 6.5 Conclusions of an orthodox model of labour markets. The simultaneous collapse of wages and Moldova’s employment problem does employment in the 1990s refl ects, in part, not arise principally from defects in the the weak bargaining power of workers, labour market, such as so-called ‘rigidi- which has resulted from at least three factors: ties’, but from a broader problem of in- 1) the absence of an independent trade complete and painfully slow transition union movement at the beginning of the of its economy. Growth is not only not transition 2) rapidly growing unemployment rapid enough but also not broad-based and 3) ineff ective regulation of employers by enough. Moreover, many poor workers do the public sector. not possess the skills, resources or assets needed to take advantage of economic A poverty reducing employment strategy opportunities if they become available. in Moldova needs the active support of Hence, an eff ective employment strategy trade unions in all sectors, especially in needs to be comprehensive, covering a

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 78 The index includes twenty-four indicators of enterprise characteristics, including labour turnover and working conditions. It is explained in detail in Standing 1997b.

Chapter 6: Employment, Productivity and Poverty

144 145 range of policy actions at the macro, sec- At the sectoral level, special attention toral and micro levels (see Chapters 2 and should be given to improving produc- 3). Macroeconomic reforms are needed to tivity and incomes in agriculture. When spark more rapid growth, sectoral reforms the Moldovan economy collapsed in the are needed to make growth more broad- early stage of transition, people fell back based, and microeconomic reforms are on agriculture in order to eke out a living. needed to enhance workers’ access to But the precipitous decline in agricultural expanding economic opportunities. production undercut the sector’s prospects for generating poverty-reducing The most important constraints on em- employment. Hence, an agriculture-led ployment growth in Moldova are related growth strategy in Moldova would quick- to insuffi cient demand and the lack of ly exhaust its poverty reducing potential. productive public and private investment Productivity and real wages are lower in in particular. While changes in the regula- agriculture than in industry and services. tion of enterprises and labour markets can Based on the current open trade regime, contribute to employment creation, the the development of a vibrant agricultural fundamental starting point is designing sector that is competitive in both domes- a more growth-oriented macroeconomic tic and external markets implies that few- framework. This framework would have er agricultural workers will be needed in several components: expansionary fi s- the long run. Already, many workers have cal policy focused on public investment, migrated out of the sector in search of monetary policy geared to fostering low higher incomes, either in Moldova’s larger real interest rates, and a modestly under- cities or abroad. valued exchange rate designed to boost exports. It is a strategic priority to revive growth in the agricultural sector. Chapter 6 pro- The composition of fi scal expansion is as vides several recommendations on how important as its aggregate demand ef- to achieve this objective. Development fect. While there is excess capacity in the is not possible in Moldova as long as private sector, much of it is in enterprises agriculture is a moribund sector. But ag- whose expansion and rejuvenation are riculture cannot grow and develop on its unlikely to occur. So, stimulus of demand, own. Moreover, it has absorbed too much by itself, is not likely to increase capacity surplus labour. Employment also needs utilization. New productive capacity needs to be created elsewhere in the economy. to be created. Hence, public investment in In this respect, agro-industry provides a economic and social infrastructure, which critical engine of job creation. The rela- would lower the costs of private-sector in- tive absence of linkage eff ects from ag- vestment, is a necessary condition for sus- riculture to manufacturing has resulted, tainable employment growth at wages in fact, in deep recession in small towns, high enough to reduce poverty. which suff er from the highest incidence of poverty in the country (see Chapter Making regulation of the private sector 4). Hence, building up the agro-industrial more transparent, changing aspects of sector will be critical not only for creating labour market governance, and reducing broad-based productive employment in bureaucracy and corruption can comple- Moldova but also for contributing to comment a pro-growth, employment-foster- mercializing agriculture and making it ing macroeconomic framework. However, more productive and competitive. Public without a conducive macroeconomic investment should be directed towards framework, such institutional reforms are agro-industry and private credit from the likely to have only a modest impact on commercial banking system should also

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction growth. be encouraged to fl ow to this sector.

Chapter 6: Employment, Productivity and Poverty

144 145 The acceleration of growth throughout household in Moldova have been losing the economy and the stimulation of sec- opportunities for such educational attain- toral growth in both agriculture and agro- ment because inequality in access has in- industry will generate more decent-pay- tensifi ed, especially for higher education. ing employment. Based on this increased demand, the supply-side interventions of In the short to medium term, agriculture has the government to improve workers’ skills to be one of major sources of growth in the and mobility, develop entrepreneurship Moldovan economy. But this growth will and innovation, and enhance equal op- depend not only on raising labour producti- portunity, such as for women and youth, vity in agriculture but also on expanding the will become much more successful. An agro-industrial sector so that it can re-absorb integral component of a comprehen- the workers who had to resort to agricultural sive programme to improve the supply livelihoods because of the collapse of in- of skilled workers is to ensure that the dustry during the transition. Agro-industry educational system off ers broad-based can provide workers with the remunerative access to general secondary, specialized employment that is not possible on a wide- secondary and higher education. Poorer spread basis in agriculture itself. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 6: Employment, Productivity and Poverty

146 147 C h a p t e r 7 : Agro-Industrial Growth and Sustainaible Proverty Reduction

7.1 Agriculture’s Role in require policy action in order to make rapid and sustainable growth possible. The Moldova’s Growth principal problems are underdeveloped Strategy markets for inputs, output and fi nance. Also, weak or absent are supportive institutions, such as output marketing and Agricultural development is critical to input purchasing cooperatives (so-called Moldova’s future because it would in- ‘single-purpose cooperatives’), which can crease rural incomes and job opportuni- increase the bargaining power of small ties, and contribute to reducing migration. farmers. Confi dence in the public admi- It represents an important medium-term nistration and in commercial companies vehicle for assuring the recovery of the is low, and enforcement of legislation is Moldova economy and laying the neces- lacking, in particular when lack of enforce- sary basis for the eventual re-industrialisa- ment adversely aff ects small producers. tion of the country. However, developing Finally, most of the agricultural sector has the agro-industrial sector should go hand suff ered from years of insuffi cient invest- in hand with the revival of the agricultural ment. As a result, rural roads, market in- sector. Without forging strong links to frastructure, irrigation systems, and com- agro-processing, agriculture will be un- munication networks have deteriorated, able grow either rapidly or in a sustain- undermining the sustainable recovery of able manner. the sector. The agro-industrial sector of Moldova is still recovering from the eff ects of the 1998 Russian crisis. This sector was im- 7.2 Agriculture on the portant during the Soviet era, refl ecting Moldova’s absolute and relative advan- Eve of Independence tages compared to other republics of the former Soviet Union. As the economy col- During the 1980s, agriculture accounted lapsed after independence, a large part of for over thirty percent of Moldova’s GDP the agro-industrial sector also collapsed. and provided employment for thirty fi ve The remaining part has restructured and to forty percent of the labour force.79 modernized, and is seeking to compete in Agro-processing contributed an additio- export markets and with imports. nal ten to fi fteen percentage points of GDP, to make agricultural-based activities While agriculture may have growth po- total about forty percent of GDP. Agro- tential, there are structural bottlenecks processing also employed another eight in the sector, and between producers and to ten percent of the labour force. The processing companies in particular, that share of the rural population declined

79The data presented in section 6.1 are based on two sources: Annual Statistical Books issued by the De- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction partment of Statistics and Sociology (DSS) of the Republic of Moldova and the World Bank Statistical Handbook 1993 – States of the Former USSR, Washington, D.C., September 1993.

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

146 147 during the 1980s, but remained quite in two major types of centrally controlled high compared to other Soviet republics. large-scale socialized farms, the collective In 1990, this share was fi fty three percent, kolkhozes and state-owned sovkhozes. The compared to the FSU average of thirty four system was complemented by inter-farm percent. Because non-agricultural jobs enterprises; large cooperatives, jointly grew slowly and agricultural salaries were owned by several farms, that comprised similar to those in other sectors, the rural large livestock and feed complexes. Du- labour force showed little job mobility. ring the 1980s, these enterprises provided about eighty percent of the country’s The importance of agriculture in Moldova’s pork and ninety percent of its eggs and economy derived in part from the coun- poultry. Overall, they accounted for about try’s natural resource base and fertile soils. fi fteen percent of gross agricultural out- Rich black soils cover about eighty percent put. However, since inter-farm enterprises of the total land area. Agriculture’s impor- were not self-suffi cient in feed, Moldova tance also resulted from political factors had to rely on imports of over one mil- that assigned Moldova primarily an agri- lion tons of grain per year from other FSU cultural role within the Soviet division of republics. labour. There were no production or trade decisions made within this division solely, In addition to the state farm sector, rural or even primarily, on economic and natural private household plots and garden plots endowment grounds (see Chapter 1). of urban households accounted for an important share of agricultural produc- The crop sub-sector accounted for over tion. The average size of these plots was sixty percent of gross agricultural output small, averaging three-tenths of a hectare. (GAO) in the 1980s. Moldova was the They relied on part-time family labour source of thirty percent of the USSR’s to- and produced primarily for subsistence, bacco, twenty percent of its grapes and although part of the output found its way wines, thirteen percent of its fruits, and ten to markets in nearby towns or Chisinau. percent of its vegetables. These comprised While holding only seven percent of to- the so-called high value agricultural (HVA) tal agricultural land, private households products (Economist Intelligence Unit produced up to eighteen percent of ag- 2003). Moldova recorded some of the high- ricultural output. This can be explained est crop yields in the FSU. For example, the partly by these households’ specialization 1985 yields of Moldovan HVA crops were in high-value commodities, their access to higher than the current yields in Romania subsidized inputs from the large farms, or Bulgaria (USAID 2004). While labour and price diff erentials between private productivity in Moldova ranked below that markets and the administrative system. in the Baltic republics and other European republics, it was signifi cantly above that For the state sector, the government gave in the Transcaucasian and Central Asian highest priority to meeting production republics, whose agricultural employ- targets set by the central plan. Centrally ment shares were similar to Moldova. The controlled farms had mechanisms to con- most remunerative crops were tobacco, trol costs or surpluses (profi ts) because: (i) vegetables, fruits, and grapes, followed by their inputs were delivered at non-negoti- potatoes, sugar beet, and (in much smaller able prices; (ii) their capital investments amounts) sunfl ower and grains. Moldova’s were set by the production plan; (iii) their crop diversity is a distinct advantage and credit was supplied by the government remains relevant to its current policy op- without regard to repayment capacity; tions and economic opportunities. and (iv) sales were guaranteed at prices to cover production costs. Profi tability

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Agricultural production in Moldova, as in at the farm level was sustained through the rest of Soviet Union, was organized subsidies. The share of agricultural sub-

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

148 149 sidies in total government expenditures opposition to land reform. This led to the in Moldova grew from sixteen percent in November 1994 law, ‘Suspension of Some 1980 to thirty fi ve percent in 1990. How- Articles of the Land Code’, which halted ever, by 1992, the fi rst full year of Moldo- the process of privatisation. However, in va’s independence, agricultural subsidies November 1995 the Constitutional Court dropped to 3.4 percent of government of Moldova declared the limitations on expenditures. leaving collective and state farms to be unconstitutio nal, and the process of pri- The state farms provided a range of so- vatization began anew in 1996. cial services in villages and small towns, which were absorbed into operating By October 1996, some ninety thousand costs. It was the state’s responsibility to private farmers, individually or in groups, ensure that these farms remained viable. worked one hundred and thirty thousand This objective was accomplished through hectares, fi ve percent of Moldova’s arable a wide range of fi nancial instruments, land. The remaining farm population re- which enabled the farm enterprises to mained in collective farms that had been fulfi l their mission of both producing food nominally restructured, or in new farm for the cities and providing social services associations and production cooperatives for their workforce. Under this system, that had been created from the collective the rural population was supported at farms. These new institutions, ‘corporate a standard of living above poverty. The farms’, essentially functioned as slightly average wage in agriculture was similar smaller kolkhozes. The National Land Pro- to the economy-wide average, and rural gram (NLP) of 1998, supported by external households had the additional benefi t of funding agencies, accelerated the process being able to cultivate individual plots, of land reform. In 1999, a law on farm which met most of their food needs and debt provided additional incentives for provided some cash from sales. dissol ving collectives by arranging debt restructuring and consolidation. At the end of 2000, nearly 1.7 million hectares and the majority of state farm assets were 7.3 Post-Independence transferred from over one thousand bank- rupt kolkhozes and sovkhozes to about one Agrarian Reform and million former farm employees. Orchards Economic Performance and vineyards were also distributed to households according to the latter’s size. The result was to fragment a large share of Restructuring of Ownership land holdings into over three million small privatised parcels. The restructuring of collective and state farms proceeded unevenly after inde- On average, each owner received three pendence. The fi rst phase began in 1991 parcels of land, with an average parcel with the formal distribution of the house- being one-half hectare. The program hold plots. In 1992, the Land Code and included the establishment of a register the Law on Peasant Farms were adopted, of land titles, and an extensive education initiating large-scale land redistribution. campaign for farmers on their legal rights Land and other collective and state farm and market opportunities. Some of the assets were distributed among house- new landowners chose to farm the land holds as paper shares, which could be with family labour. Others leased their exchanged for land or non-land assets. land and other assets to the newly cre- However, the large wave of households ated corporate farms, which had been

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction clamouring to leave collective and state formed out of the collective farms, and farms and privatize their land prompted either worked on them as members or

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

148 149 employees or received lease payments. market infrastructure, including market- Thus, the privatization process led to a ing, input and credit supply; (ii) improving dual agrarian structure polarized between the legal and regulatory framework; and one part based on a myriad of small-scale (iii) facilitating access to information and farms of one to fi ve hectares and another technical advice. part based on a few large corporate farms, such as limited liability companies, joint Despite these measures, agricultural stock companies, production coopera- markets and related services remain tives, or farm associations. weakly developed in Moldova. In the aftermath of the redistributive land The gender aspects of the land reform reform, many of the more than one have been poorly documented. Rural million beneficiaries did not have the women made up a majority of the land necessary skills for efficient agricul- proprietors because men worked in tural operations, especially because a constructions, technical services and large proportion of the recipients were commerce, losing their right under the pensioners at the time of privatization. land legislation to obtain land. However, It was expected that the emergence despite this apparent advantage, women of agricultural markets would transfer did not end up playing a leading role in land from less efficient to more efficient the land reform. There appeared to be producers. Greater land consolidation few female managers of cooperatives or and concentration were the ultimate large farms although defi nitive statistics objective. Such a trend was expected, on this issue are lacking. As in other sec- in turn, to lead to a more competitive tors, gender-focused statistics for agricul- agricultural sector. ture are an urgent priority in order to fully understand the role of women. The latest data of the Cadastre Agency indicate that 53.6 percent of total agri- Privatization and farm restructuring, cultural land is utilized by relatively large- completed in early 2001, were the major scale corporate farms, mainly limited components of the fi rst stage of land liability companies, with a much smaller reform. Stage two, the post-privatization share utilized by production cooperatives program, began immediately afterwards and joint stock companies that also oper- and continues with the support of inter- ate on the basis of leasing land and assets national donors. The post-privatization from small landowners. The corporate program has focused on facilitating the farms are often headed by former manag- adjustment of private farms to a ‘market ers or agricultural specialists of the collec- environment’ through: (i) developing tive enterprises.

Table 7.1 Private Land Tenure in Moldova as of January 1, 2004

Average Form of enterprise Number of units Land used (ha) % of land used Farm size (ha) Limited liability 1188 639,000 42.5 538 companies Production 111 105,700 7.0 952 cooperatives Joint stock companies 95 47,100 3.1 496 Kolkhozes 4 13,700 0.9 3425 Peasant farms 555,000 698,200 46.4 1.26 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: Cadastre Agency, Land Balance Data - 2004

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

150 151 In a relatively short period, 2001 to 2004, ternational opposition because they are a rapid process of land consolidation and seen as a step towards reversing some of farm restructuring took place through the main outcomes of the redistributive lease arrangements between individual land reform of the late 1990s. landowners and large corporate farms. Individual ownership of land was not The government proposals have encou- directly aff ected. However, next to these raged the creation of ‘consolidated enter- corporate farms, the small scale sector, prises’, which would function essentially composed of 555,000 peasant farms (reg- as production cooperatives. The govern- istered and unregistered), suff ered from ment’s reasons for undertaking land con- a very small size of land holdings. The solidation have been based on the follow- average holding was 1.26 hectares, often ing assumptions: (i) land fragmentation is further subdivided into tiny fragmented excessively high in Moldova; (ii) small scale land plots. This small farm sector cultiva- agriculture is ineffi cient and uncompeti- ted 46.4 percent of the agricultural land. tive; (iii) consolidated agricultural enter- Some of these small farmers have tried prises, because they represent the optimal to cope with the market environment by size, will be able to produce a high-quality forming partnerships among themselves and competitive output; and (iv) the envi- but this phenomenon has been limited.80 saged consolidated enterprises will create new jobs, increase labour productivity and Compared to large-scale units, small reduce migration from rural areas. farms appear to have higher productivity, measured by value added per unit of land. Some of these assumptions have been This can be explained largely by the pro- called into question. First, although the duction structure of the small scale sec- statistics on land distribution mentioned tor, which grows most of the high value earlier should be treated with caution, crops, such as fruits, grapes, vegetables they do suggest that small farms are eco- and potatoes. A recent IFAD study (2003) nomically effi cient. Second, the concept of highlighted the other typical advantage ‘optimal size’ is problematic. Even for the of small-scale growers; namely, that they same product, the desirable size can vary have a competitive edge over larger in accordance with diff erent soil charac- operators because of their lower costs. teristics and production techniques. Third, However, the Moldovan agricultural sec- land consolidation has already taken tor has yet to achieve the necessary con- place through lease markets, and is likely ditions for international competitiveness. to continue. Such market forces cannot be The development of various factors, such stopped but rather can be directed into as agricultural services and markets for more socially desirable directions. Finally, inputs, outputs and credit as well as sup- the principal explanation for the lack of portive government policies, will have an effi ciency in agriculture may not be small important impact on this prospect. farm size, but the underdevelopment of markets for both inputs and outputs. The government has viewed fragmented landownership and the predominance of A recent study attempting to gauge farm- small farm size as a bottleneck for agro- ers’ attitudes toward the government’s industrial development. Hence, it has proposed form of land consolidation launched several policy proposals aimed found that seventy six percent of land- at land consolidation and the creation of owners (or leaseholders) and eighty two production cooperatives. These initiatives percent of lessees would not agree to have encountered both domestic and in- transfer their land to a ‘consolidated en-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 80The category of peasant farms (or ‘the individual sector’, as it is called) also includes about 340 farms whose size is over 50 hectares.

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

150 151 terprise’ (CISR 2003). Typical reasons given after the second phase of redistributive for not wanting to do this were: ‘won’t be land reform, there was strong growth of six able to pass my land over to my heirs’, ‘the percent, followed by three percent in 2002. collective farm will not be more effi cient This growth might have continued into and the employees will be irresponsible’, 2003 but a severe summer drought de- ‘can sell my land anytime now’, or ‘am sa- stroyed the wheat harvest, which normally tisfi ed with my land lease contract’. accounts for fi fteen percent of agricultural value added. As a result, growth was a ne- As this report was being fi nalized, the gative fourteen percent. A simulation that government proposal for forming ‘con- nets out the negative eff ect of the drought solidated enterprises’ was dropped. But on agricultural output indicates that the the debate on how to improve the agrar- growth rate of gross agricultural output for ian structure, in order to promote both ef- that year could have been 2.2 percent. fi ciency and equity, continues in Moldova. One way out of this impasse is to encou- During the period of transition, dramatic rage small-scale farmers to join ‘single- changes in the structure of agricultural purpose cooperatives’ for buying inputs production accompanied contraction. or marketing outputs. This would lead Demonetization, demand contraction and to greater consolidation of land holdings the risk-avoiding behaviour of small farm- and higher levels of productivity without ers prompted increases in the land under pushing some farmers off the land and low-value staple crops, such as wheat, corn impoverishing them. However, while most and sunfl ower, and the concomitant reduc- individual farmers are not required to pay tion of the area under high-value crops (see value added taxes because their turnover Figure 7.2). By 2003, grains and sunfl ower is below 200,000 leu per year, they would occupied about seventy fi ve percent of have to pay such taxes if they joined co- cultivated area, whereas in the past they operatives. In order to rectify this situation occupied fi fty to fi fty fi ve percent. and actively encourage farmers to form such associations, the government should The largest contraction in the crop sub- provide them with tax breaks. sector was in such high-value products as tobacco, fruits (except grapes) and vegetables. As a result, the share of high Agricultural Performance value crops in total value added declined from sixty six percent in the pre-transition As a result of the structural changes and period to thirty percent in 2000-2003 (see general economic collapse, agriculture Figure 7.3).81 Because potatoes are grown contracted dramatically during the 1990s. mainly for subsistence, their production In 2003, agricultural production was only at remained constant over the years and forty fi ve percent of its 1990 level (see Fig- thus their relative share in GAO increased. ure 7.1). Livestock production proved much more vulnerable to structural changes than Large declines in yields characterised the crop production. Over thirteen years, it de- 1990s (see Figure 7.4). The yield declines creased by nearly two-thirds. As a result, were higher for the machinery-intensive the share of livestock in total agricultural crops, and lower for staples (wheat, corn, value added decreased to thirty percent sunfl ower and potatoes). The low and from the forty percent that was typical declining yields of fruits and vegetables82 during the Soviet period. However, in 2001, were the result not only of insuffi cient and

81 The so-called High Value Agriculture sector comprises tobacco, vegetables and fruits (including berries and grapes). Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 82The year 2003 was an exception in this respect: the registered yields were higher due to excellent climatic conditions for fruits and vegetables.

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

152 153 inadequate inputs. There were two other im- of the high subsidies it had received during portant constraints: the lack of re-planting of the Soviet period, and, even more, due to ageing orchards and vineyards, and the lack of the catastrophic decline in urban incomes rehabilitation and modernization of irrigation (which lowered demand for meat pro- systems, particularly small-scale systems ad- ducts). The sector was devastated by the justed to the needs of the recently emerged combination of price and trade liberalisa- private far mers. The Ministry of Agriculture tion. Now, the livestock herds are concen- reported that only half of existing vineyards trated on small farms (which account for and orchards were productive in 2004. over ninety percent of all animals). These farms rely on livestock for family consump- Also, only 10,000 hectares, or fi ve percent tion needs as well as for earning income of the previously irrigated land, was still from market sales. Cattle, which generally irrigated in 2003. Water use for irrigation require long-term investments, registered had dropped in 2003 to only ten percent the greatest decline among livestock, of its level in the mid 1990s. Repairing from over twenty percent of total livestock the old irrigation system is not likely to during 1985-1990 to only seven percent be advisable since it was highly energy now. Lack of genetic resources, which intensive and has been largely destroyed would enable livestock production to be in any case. Most grain production could competitive in price and quality,83 limited be supported by a rain-fed system. But ir- cattle selection, and inadequate breeding rigation is certainly of great importance to technology continue to be major impedi- the productivity of high value agricultural ments to livestock development. crops. A 2002 IFAD study estimated that net farm income could be boosted by as much as US$ 700 per hectare on the basis of irrigating high value crops. Irrigation 7.4 The Underdevelop- is such a high priority for agricultural de- ment of Agricultural velopment that the government should off er tax incentives or preferential credit Markets to help farmers purchase irrigation equipment. Small-scale irrigation systems could The underdevelopment of markets for be run effi ciently by farmer organizations, fertilizers, seeds, machinery and credit is such as Water Use Associations. a serious problem for Moldovan farmers. Their access to inputs has radically de- Along with the negative trends and con- clined during the transition to a market- straints of the 1990s, there were some based economy and new input markets positive developments. According to have yet to take the place of state provi- the High Value Agriculture Study (CNFA, sion. Because of the dramatic reduction 2004), commercial producers of high of livestock, organize fertilizer is scarce. value products that used advanced pro- The use of chemical fertilizers, which was duction technologies, including irrigation, extensive in the Soviet period, dropped in registered signifi cantly higher yields for 1999 to one-tenth of its 1990 level. This some crops than the average in Romania, represents a severe constraint, in par- Bulgaria and Hungary. This suggests that ticular, on the growth of high value agri- Moldovan producers have the potential to cultural crops. While seeds have become achieve international competitiveness. more available in Moldova, their quality remains variable because of diffi culties in The livestock sub-sector shrank sharply importing suffi cient quantities of certifi ed over the transition period due to removal seeds.

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 83As of July 2003, the number of pedigree cattle was seven to ten times lower than the normative requirements. See Agricultura Moldovei, 2/2004.

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

152 153 There has been a rapid deterioration in verty and Policy Monitoring Unit (PPMU) the existing stock of farm machinery. of the Ministry of Economy reported that The number of tractors has declined, for agricultural workers were the poorest example, by over twenty percent and group of the rural population (66.2 per- another ten percent of tractors are used cent of them were poor and fi fty four per- exclusively for transport. The stock of trac- cent extremely poor), and peasant farmers tors is now quite old: half of them have were the next poorest (53.9 percent were been in use for over fi fteen years. poor and 45 percent extremely poor).

Small agricultural producers have virtually More recent data, for 2003, indicate slight- no access to commercial lending. Since ly reduced levels of poverty among farm land prices remain low and other farm as- households. The data show that access to sets are scarce, farmers possess little col- land contributed to lowering the poverty lateral for borrowing. The main lenders to headcount. On average, 46.9 percent of agriculture are Moldova Agroindbank and farmers were poor and 32.6 percent ex- Victoria Bank, which together account for tremely poor. But those with less than about half of all lending. However, loans two hectares of land had a higher poverty still account for only a small fraction of incidence. So, consolidating land holdings gross agricultural output. is essential for agricultural development and poverty reduction. However, since With the support of external donors, Sa- most agricultural workers, who tend to vings and Credit Associations have arisen be the poorest group, are not employed to off er micro-credit to farmers. However, on small farms, reducing rural poverty total volume is small and directed mostly requires a broad-based growth strategy to fi nancing working capital. While such that benefi ts both large and small pro- credit is indeed important, loans for duction units. But this does not imply that long-term productive investment in ag- land should become more concentrated, riculture are in short supply. The same i.e., consolidated under the control of shortage of credit applies to rural non- large corporate farms that wield excessive farm enterprises, which are very weakly market power. developed in rural areas. The lack of credit is a major constraint to both agricultural In the fi rst half of the 2000s, over half of the and non-agricultural development. This country’s population lived in rural areas, is a constraint that public policy can do with agriculture being the main source a great deal to remove by providing vari- of their employment. While the share of ous incentives to fi nancial institutions to agriculture in GDP declined during the off er more loans for such development. 1990s, its share in total employment grew Helping direct such loans to private in- slightly (see Figure 7.5). This implies that vestment in agriculture and agro-industry during the economic collapse after 1991, would complement the spearheading agriculture absorbed some, although cer- role of public investment in these sectors. tainly not most ,of the labour shed by other sectors. The falling share of agriculture in GDP implies that the infl ow of labour into farming did not spur agricultural growth, 7.5 Rural Labour Markets but in most cases was absorbed in subsistence production that reduced average As shown in Chapter 4, poverty in Moldo- sector-wide productivity. va’s rural areas is most serious amongst those who have little land. However, Low agricultural returns and lack of jobs poverty is most prevalent in small towns, outside agriculture stimulated labour mi-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction where people have few employment op- gration, in many cases to other countries portunities and no land. In 2002, the Po- (see Chapter 5). Non-farm employment

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

154 155 has remained insuffi cient to reduce un- markets for agricultural services, fi nal deremployment in rural areas, in spite of products, fi nance and labour. Another various projects supported by external major problem is the lack of long-term donors and lenders. This strengthens the investment, both domestic and foreign, case for focusing on increasing agricultural that could enhance agricultural producti- incomes, through augmented output vity and foster internationally competitive and higher prices. This could be fortifi ed agro-food products. These investments through links with an expanding agro- are needed at the production level; in processing industry. Agricultural prospe- replanting orchards and vineyards, resto- rity will be the necessary foundation for ring irrigation, and developing processing the growth of the non-farm economy. facilities. During 2000-2004, agriculture attracted not more than US$ six to ten Estimates indicate that hundreds of thou- million annually for investment in fi xed sands Moldovans work abroad, the majority assets ,only about fi ve percent of all such of them originating from villages and small investment. In contrast, the investment towns (see Chapter 5). According to DSS needed to revitalize high value added data, this out-migration has contributed to crops is estimated to be about US$ two the feminization of the rural labour force billion (USAID 2004). since most migrants have been men. It is clear that skilled labour has become increa- Linking agricultural to agro-processing singly scarce in the countryside since the can help stimulate growth and increased late 1990s. The loss of such a large propor- investment. The agro-processing indus- tion of the working age population imposes try accounted for sixty percent of gross serious constraints on agricultural develop- industrial output in 2003. Since 2000, this ment. This implies that the development industry has registered impressive annual of human resources in rural areas is still es- growth rates of fi fteen to eighteen percent. sential. But the problem is much larger than The wine industry dominates this industry, this. Consolidation of land holdings will help with a forty percent share in its output, fol- spur development. Strengthening links with lowed by fruits and vegetables (eight per- agro-industry will also improve productivi- cent), dairy products (seven percent), oils ty. Special measures to channel remittances (six percent), and tobacco (fi ve percent). from rural workers labouring abroad into Almost all agro-processing enterprises productive investment in agricultural deve- were private in 2005, with only tobacco lopment could also contribute. and some wineries still state-owned.

Moldovan agriculture and agro-industry 7.6 Agro-Industrial Growth continue to have strong export potential. Processed and unprocessed agricultural There are many constraints impeding ag- products accounted for about sixty per- ricultural development in Moldova. They cent of Moldovan exports (Table 7.2). arise, in part, from the underdeveloped In spite of the economic collapse of the

Table 7.2. Agro-Food and Primary Agricultural Exports (1995-2003) (US$ millions & percentage) 1995 1996 1997 1998 1999 2000 2001 2002 2003 Agric, non-processed 144 128 151 106 97 89 98 122 120 Agric, processed 393 457 487 354 200 202 260 285 343 Total Agric 538 585 638 460 297 291 358 406 463 Total Exports 746 795 874 632 463 472 570 666 790 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Agric /Total, percent 72 74 73 73 64 62 63 61 59

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

154 155 early 1990s, Moldova remained a net ag- mills, productivity is low. Therefore, even ricultural exporter, but still specializing in with the temporary protection provided wines and other alcoholic drinks, tobacco by parliament, the sugar-processing sec- products, and fruits and vegetables. tor may not survive, especially given the over-production in the European Union Export volumes fell drastically over the and the subsidies that it gives to its own course of the 1990s. In 2004, Moldova ex- producers. A pervasive problem through- ported only ten percent of the fresh pro- out Moldova is the monopoly power of duce that it had supplied to the USSR in processing companies. They can make 1985 (USAID 2004). Russia and, to a lesser large profi ts by being able to pay farmers extent, other CIS countries continued to low prices for their output. absorb over seventy percent of Moldovan agro-food exports, while the CEE coun- Sustainable broad-based economic growth tries absorbed ten to fi fteen percent and and poverty reduction in the Republic of the European Union absorbed about ten Moldova depend on revitalizing agriculture percent. Despite the 1998 Russian fi nan- and strengthening its linkages to agro-in- cial crisis, which exposed the vulnerabi- dustry. The high productivity of agriculture lity of concentrating on a single market, registered in the pre-transition period testi- Moldova has not altered its trade pat- fi es to the sector’s long-term potential. But terns. Several factors have continued to in order to achieve this potential, develo- determine the destinations of Moldovan ping well-functioning agricultural markets agro-food exports: (i) the large size of and services is a key requirement. the CIS market and the growing demand for Moldovan products in this market; (ii) the familiar and positive consumer image of Moldovan products, particularly wines, fruits and vegetables; (iii) easier 7.7 Policy access to CIS markets through free trade Recommendations agreements and common standards for produce; and (iv) a familiar business en- A strategy for agro-industrial growth and vironment, which facilitates establishing sustainable poverty reduction in Moldova partnerships and trading relations. would need to have several components. Public investment will play a central role, Agro-industry is the fastest growing eco- especially in rebuilding agriculture’s nomic sector in Moldova. In 2003, the comparative advantage in high value manufacturing sector as a whole grew crops and linking their production to the at a rate of 13.7 percent, with food and agro-processing sector. Public policy has beverages representing around seventy to resolve the polarizing impact of Moldo- percent of its value added. In 2004, it is va’s dual agrarian structure, which is expected that food and beverages output dominated by a few large, and mostly in- would increase at a rate of 16.7 percent effi cient, corporate farms and a myriad of (Center for Economic Policies 2004). small under-resourced subsistence farms. Particularly for poverty reduction pur- There are clear signs of modernization poses, public support needs to be geared in various sub-sectors, such as fruit and to small farmers, whose tiny fragmented vegetables, sunfl ower oil processing and plots cannot produce enough commer- the wine industry. However, there are also cially viable commodities. Small farmers negative trends, as in the sugar beet indus- need to be grouped into larger associa- try, where there is substantial over-capa- tions and the monopoly power of corpo- city coupled with a decreasing domestic rate farms and agro-processors needs to

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction supply of raw materials. Since technology be reduced. The government of Moldova is out-dated in these Soviet-made sugar- has limited capacities and resources. Thus,

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

156 157 it needs to focus them where they are of 4. Developing physical marketing infra- most strategic importance. Following are structure (collection, grading, sorting, a few recommendations that we believe processing, storing, packaging and trans- are strategically crucial. porting commodities) could signifi cantly strengthen and diversify the supply of 1. The government of Moldova should output to markets and boost farm-gate focus its policies on helping agriculture prices. Facilitating and stimulating pri- regain its position of comparative advan- vate investment, both local and foreign, tage in high value crops and facilitating in such infrastructure should be a key its linkages to the agro-industrial sector. feature of Moldova’s agricultural deve- This can lead to the improvement of lopment strategy product quality and productivity and the transfer of advanced technologies down 5. The weak ‘bargaining position’ of the to the farm level. Such eff orts would help multitude of small peasant farmers (who expand exports to both traditional (CIS) possess very limited ‘market power’) markets and new (European) markets. can be overcome by promoting single- A central ingredient of this policy focus purpose cooperatives and associations, is public investment in roads, irrigation which focus exclusively on input pur- infrastructure, rural markets and com- chase and/or output marketing. This munication networks, which would is not a recommendation for full-scale ‘crowd in’, or stimulate, the broad-based production cooperatives. Tax policies private investment that the sector sorely and legislation that encourage the for- lacks. mation of such organizations can make an important contribution to accelerat- 2. The development of agro-industry is ing their emergence. central to the revitalization of the agricultural sector, particularly in ensuring 6. The main current bottleneck for agricul- the latter’s linkages to export markets. A tural producers in Moldova is the under- growing agro-processing sector will also development of markets (such as the create alternative employment opportu- seed, fertilizer, agricultural machinery, nities that will relieve agriculture of the labor, fi nance and output markets) and heavy burden of continuing to absorb their corresponding institutions (such excess, unproductive labor. The institu- as banks, microfi nance organizations tional and policy environment should and other fi nancial institutions). Since encourage new entry into agro-industry, these various markets are interlocking, institute straightforward and transparent an integ rated policy response towards import/export procedures, and actively develo ping them is needed. reduce its current monopoly powers. 7. The Government of Moldova has limited 3. Agricultural producers, particularly small fi nancial resources to address the host ones, need training and extension services of problems and bottlenecks hindering in order to produce more and higher- agriculture and agro-industry. However, quality commodities. This could be facili- it can also create a legal, regulatory and tated by their major clients, agro-proces- policy environment that could encou- sors and exporters, who are familiar with rage the private sector to make greater the requirements of their target markets. investments in these sectors. This should The government can use fi scal measures be a priority of public policy. An example to stimulate the establishment of mutu- would be guarantees or other incentives ally benefi cial out-grower schemes that given to the commercial banking system can promote improved product quality in order to induce it to lend more to agri- and increase farmers’ income. culture and agro-industry. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

156 157 Figure 7.1: Evolution of Gross Agricultural Output (1990)

Figure 7.2: Production Structure by Value: Crops and Livestock (Avg. 2000-2003)

A. Crops B. Livestock

Figure 7.3: Crop Production Structure by Area (2000-2003) Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

158 159 Figure 7.4: Comparative Yields (1985-1990; 2000-2002 and 2003)

Figure 7.5: Share of Agriculture in GDP and Employment, 1993-2003 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

158 159 C h a p t e r 8 : Social policies for Human Developmet

8.1 Introduction argues that part of the population should be excluded. Human poverty results from inadequate access to health services, safe water, ba- Across countries, the attempt to achieve sic sanitation, and educational facilities. universality is carried out through vari- Income may be an important means to ous concrete mechanisms, with varying enhance such access, but not a suffi cient degrees of success. What might be called one. People may not have access to these the ‘polar positions’ are, on the one hand, services, because they do not exist, or universal provision by the public sector, they may be discriminated against be- without charge for all major social serv- cause of race, religion, ethnicity, caste and ices; and, on the other, private provision colour. Under such circumstances, public through making social services marketed provision of basic social services becomes commodities. Universality through mar- critically important. kets suff ers from two major economic problems, well recognised in the litera- As discussed in Chapters 1 and 2, following ture. First, while markets may be open in the collapse of the Soviet Union, Moldova principle to all members of society, the passed through a deep and prolonged income and wealth distribution of the so- depression. The Russian fi nancial crisis, ciety will exclude a portion, perhaps large, which struck in August 1998, brought from eff ective access. Second, in an argu- another major blow. As a result, human ment made famous by Nobel Prize winner development outcomes regressed, social Kenneth Arrow, there is asymmetrical in- safety nets broke down, and institutions formation. The buyers of health care, for collapsed. To a great extent, migration example, are unable in practice to acquire emerged as perhaps the major ‘safety net’ even a fraction of the knowledge of the of households. These crises were partly commodity that the provider has. There- due to transition itself, partly to adverse fore, they lack the information to make a eff ects of policies, and partly to external rational choice. factors. Beginning in 2000, Moldova reversed these disastrous trends, but the As a result of these and other, non-eco- poverty situation remains serious. nomic considerations, in every country there is a mixture of private and public It is important to stress the obvious prin- provision, and within the public sector ciple that the purpose of social provision a range of delivery mechanisms. One is to achieve a healthy, educated popula- delivery mechanism is income testing (in tion. In pursuing this objective, the social some countries called ‘means testing’), provision system should ensure access in which those members of society with by the poor. Social provision is not an incomes below a specifi ed level receive anti-poverty programme; by its nature a specifi c service without charge or via it is universal. We use ‘universal’ to mean a reduced charge, while those above the

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction that all members of society have access income level pay considerably more. The to it. To our knowledge, no government same delivery mechanism is used in many

Chapter 8: Social Policies for Human Development

160 161 countries for benefi ts such as child allow- Second, a large proportion of households ances. It should be noted that income tes- receive remittances from family members ting need not divide households between working abroad, about twenty percent, poor and non-poor, except tautologically. and sixty-fi ve percent of those making re- There is no reason, for example, that each mittances did so by means that would not income tested service or benefi t need use appear in offi cial statistics.85 Even if it were the same income level. the case that the overwhelming proportion of remittances entered the country We recommend against excessive use of through formal channels, it would not fol- income testing as a delivery mechanism low that these could be traced to specifi c in Moldova. This not based on principle, households. These two characteristics of but practical and technical considerations. household income sources imply that es- First, for the income testing mechanism to timating income would either have to rely be eff ective, it must be possible a) to estab- on the reporting of the household itself, lish the relevant income level below which or on direct investigation of households, special measures, such as reduced char- which could potentially be quite intrusive, ges, will begin; b) to measure individual as well as subject to corruption at the lo- or household income with appropriate ac- cal level. curacy, and c) to achieve the measurement at an acceptable administrative cost. The Perhaps more fundamental than measu- identifi cation of a non-arbitrary income rement itself, is the implicit assumption for a benefi t or service is itself problemati- in income testing that incomes remain cal, though the wide practice of doing so stable over time. If a large portion of lends it a superfi cial credibility. the population is clustered close to the poverty line, as is the case for Moldova Much more problematical in any coun- (see Chapter 4), then there will be a try is the measurement of household tendency for many households to move incomes, because of the diff erences in above and below the line over the eco- household compositions, and prices be- nomic cycle. This creates two problems: tween regions and rural and urban areas, fi rst, even if technically feasible in light to name but two diffi culties. The particu- of the discussion above, the administra- lar circumstances of Moldova make the tive task of identifying the poor becomes measurement task diffi cult to the point of so expensive as to be unmanageable, futile. The fi rst of these problems, which because households that are defi ned as Moldova shares with many other deve- poor in one period are not in another. loping countries, is that the structure of Second, as a result of the fi rst, the entire employment does not lend itself to ad- purpose of the income testing exercise is ministrative verifi cation of income levels rendered invalid if the ‘poor’ are not a sta- of households. In developed countries, ble administrative or practical category. this verifi cation is made by inspecting the Attempts to deliver benefi ts by income applicant’s employment records, or if em- testing will result in an arbitrary delivery ployed, registration with unemployment system with ‘leakages’ to the non-poor. In centres. In Moldova, only about thirty such circumstances, income testing fails percent of the economically active popu- its own test: were the identifi cation of the lation is employed in the so-called formal poor administratively possible at a point sector, which would have the information in time, the identifi cation is not valid over necessary to estimate earnings.84 time. This represents an extreme example

84The statistic rises to thirty-seven percent if those in unpaid employment are excluded from the economically active. See Chapter 6, Table 6.1. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 85This does not imply that sixty-fi ve percent of the value of remittances were sent by such methods; the value may have been a larger or smaller proportion (see Chapter 5, section 1).

Chapter 7: Agro-Industrial Growth and Sustainable Poverty Reduction

160 161 of the ‘borderline problem’, in which be- ees and employers, progressive taxation nefi t delivery on the basis of income test- can be used to cancel out the ‘leakage’ to ing systematically fails. the non-poor.86

The practical diffi culties with income tes- Further, it appears from the experience ting in Moldova do not negate that fi scal re- of developed countries that income test- source constraints limit the extent to which ing can undermine the political sustain- the government can fi nance social pro- ability of social programmes because grammes. Therefore, some mechanism is those excluded from benefi ts have no required to establish expenditure priorities. direct material incentive to support in- An alternative to income testing is ‘univer- creased funding for such programmes. sal provisional within categorical targeting’, Targeting easily identifi able groups with defi ned as directing benefi ts to groups of universal programmes has the political the population that can be administratively advantage of giving the non-poor an identifi ed (categorical targeting), without incentive to support poverty-reduc- dividing the group between benefi ciaries ing programmes.87 It also avoids the and non-benefi ciaries (universal provision). potential social stigma of dividing the Examples of this approach are universal old population between the poor and the age pensions and child allowances. The non-poor. great advantage of this approach is its administrative feasibility, and the population On the basis of this discussion of the categories can be chosen to maximise the characteristics of Moldova and their poverty-reducing eff ect. implications for the access mechanisms for social services and benefits, we An argument against this approach is that suggest that the government follow a some benefi ts accrue to those who do not strategy that would be both feasible need them (so-called leakages), and as a and effective: result of making payments to the non- 1. the basis of access would be targeting poor, the level of benefi ts to the poor is clearly identifi able groups with pro- reduced. Considering fi rst the trade off in grammes of universal access (universal level of benefi ts, for any level of benefi t, access within categorical targeting); the universal programme is likely to have lower unit costs (the benefi t plus the ad- 2. the system of universal access pro- ministrative cost) than the income tested grammes would be augmented by programme. This is for the obvious reason income tested programmes for specifi c that the universal programme would not and narrowly defi ned benefi ts, such as need a bureaucracy to verify and police subsidies for tertiary education; and qualifi cation for benefi ts. Therefore, un- 3. these programmes be funded in part der a universal programme each benefi t through programme specifi c taxes (such unit accruing to the non-poor does not as mandatory pensions contributions), represent a full unit that would accrue to and partly through progressive income the poor if a shift to income testing oc- taxation. curred. Second, for formal sector employ-

86Eff ective collection of income taxes is largely limited to the formal sector for the same reason that it is in this sector that incomes can be most easily verifi ed: the government can enforce accounting regulations that require formal sector employers, including the public sector, to keep records of income payments. Placing the burden of income taxation on formal sector employees and employers is pro-poor. As shown in Chapter 6, the poverty rate in the formal sector in 2002 was thirty-seven percent, compared to forty- seven percent in the informal sector, and forty-fi ve percent for the entire labour force (see Table 6.1). 87The social security retirement pension system of the United States, created by act of Congress in 1935, Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction was designed by the Roosevelt Administration to be universal for the explicit purpose of ensuring its long term political sustainability (see Shieber & Shoven 1999, Chapter 3 ).

Chapter 8: Social Policies for Human Development

162 163 8.2 The Human The increase in mortality rates during the transition represented a reversal from Development the previous decade. The mortality rate Collapse of the 1990s for males aged 40-45, 45-50 and 50-55 increased by about twenty-fi ve, forty- fi ve and forty-fi ve percent, respectively Declines in Health from 1990 to 1995, then improved in 1996 and 1997. However, the Russian crisis of and Demographic Changes 1998 brought another increase, of about fi ve to ten percent. Thus, the rise in male Moldova entered its transition with high mortality rates clearly correlates with human development indicators. In 1990, the collapse of the Moldavian economy the life expectancy was about 69 years, during the 1990s. In 2002 the mortality and there was one doctor for every 250 rates of men forty to fi fty-fi ve remained people (UNDP 1993). Adult literacy was ten to twenty percent higher than in the almost total, with high participation and pre-transition years. A similar trend could completion rates at all levels of education. be observed for the over sixty group, for But by mid-1990s, many of these achieve- both men and women, whose death rates ments eroded. During the fi rst half of the rose by some twenty-fi ve to thirty percent 1990s, the health of Moldovans seriously during the fi rst fi ve years of transition, to deteriorated, because of eroding stand- recover thereafter, and increase again af- ards of living, poor access to health care, ter 1999 to a level twenty percent higher and its declining quality. Life expectancy than prior to the transition.88 declined to 65 years, infant and maternal mortality rates increased. Sexually trans- The rise in male mortality rates resulted pri- mitted diseases took on near epidemic marily from stress-related affl ictions: heart proportions, and half the population did ailments, circulatory diseases, alcoholism, not receive treatment for these diseases. suicide, and even homicide. Dealing with Only after 2000 did improvements in these requires not only better medical some indicators occur. care, but a pro-poor policy programme to reduce unemployment, uncertainty and The decline in life expectancy resulted in a social marginalisation. Evidence from other shift in the sex ratio of those of reproduc- countries in transition, for example, Russia tive age, due to higher mortality among and Latvia, shows that those most aff ected men. This shift was increased by the higher by stress related affl ictions were urban rate of emigration by men. These factors men, possibly unemployed, with lower caused a drop in birth rates to such an ex- levels of education (Shkolnikov and Cornia tent that the rate of natural increase of the 2000), and men from minority or migrant resident population declined. The large backgrounds and incomplete families. drop in birth rates implied an increase in Other causes of the increase in male mor- the number of one-child families. In 2002 tality, which also aff ect women to some fi fty-two percent of the births recorded degree, were strokes, cancer and diabetes, were the fi rst by the parents. The decline even among people under forty. Mortality in birth rates was accompanied by an from chronic liver diseases and cirrhosis increase in the proportion of births to un- exceeded the average for the European married mothers, increasing the number Union by a factor of six (DFID 2003). of children at risk of abandonment, institu- tionalization and poverty. These risks were Data on disease incidence confi rm the exacerbated by the growing mortality mortality data. While respiratory diseases, rates among men aged 30-49. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

88These statistics are from DFID (2003), and the Ministry of Health.

Chapter 8: Social Policies for Human Development

162 163 digestive diseases and tumours showed the government did in 2002 and 2003, stagnant or slightly declining trends, in- when the share of GDP allocated to fectious diseases and cardiovascular prob- public expenditure on health rose from lems showed a sharp upward trend (Table 2.9 to 3.4. However, public resources for 8.2). The increase in infectious diseases health were not used in the most rational resulted from mutually off setting tenden- way. Hospitals took over forty percent of cies. As a result of diphtheria outbreaks the public health budget, though some in 1994 and 1995, the immunisation pro- operated at less than potential capacity. gramme in Moldova was strengthened Improved priorities in public health re- with the help of external donors, and sources across levels of care and parts of the vaccination coverage reached over the country could allow gains in eff ective- ninety-fi ve percent. Consequently, the ness. Some essential health activities have incidence of several infectious diseases, improved, as shown by the high coverage for example, diphtheria, measles and viral of hospital-based deliveries and vaccina- hepatitis, decreased after the mid-1990s tion programs. To a limited degree, these (DFID March 2003). The incidence of sexu- measures preserved a decent health sta- ally transmitted diseases peaked in 1996 tus among children during the transition. after a twenty-fold rise compared to the As a result, the infant mortality rate fl uc- 1980s, and decreased subsequently. In tuated around twenty per thousand be- contrast, the incidence of tuberculosis, tween 1990 and 2000, and showed clear HIV/AIDS and drug addiction rose, espe- improvement in 2001 and 2002. cially among marginal groups. Community health care institutions were The decline in service availability in rural funded from the local budgets, using areas, the nation-wide fall in expenditure taxes levied at the local level, while hos- on drugs and other medical supplies, the pitals, preventive medicine, teaching in- transfer of formal and informal health costs stitutions and health administration were to households, and the decline in family funded from the central government incomes lead to a fall in the use of health budget. This fi nancing was severely af- facilities. This is confi rmed by the decline fected by the collapse of 1991-95 and Au- in inpatient and outpatients contacts per gust 1998. As a result, by 2002, real public thousand people, and the duration of hos- health expenditure per capita fell to US$ pitalisation and occupancy rate in acute twelve per capita, fi fty-two percent of its wards (DFID 2003). This decline was par- 1997 level, and less than twenty percent tially reversed after 2001, but household of its pre-transition level. Budgetary cuts budget survey evidence (HBS) shows that led to the introduction of user fees and to the recovery in contacts was mainly by the de facto transfer of the cost of drugs better off households (World Bank 2004, to households. Patients’ payments for Figure 5). Thus, being poor means a dra- hospital services equalled almost twenty matically lower access to health facilities percent of total health spending in 2002 (PPMU 2002). For example, the duration of (DFID 2003). Estimates by the World Bank hospitalisation of the poor was less than a indicate that private payments rose from tenth that of the non-poor. Exclusion from twenty-seven percent of total health ex- health care services was more pronounced penditure in 1997 to almost half in 2001- in rural areas, for larger households, and for 2002 (World Bank 2004a). households whose head had lower educational achievements. Especially acute is the The health fi nancing reform introduced problem of elderly people with chronic in January 2004, from funding with ge- diseases and low pensions. neral revenue to health insurance, may represent an additional constraint to the

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction A poverty-reduction strategy needs to access by the poor, most of whom are not increase the health expenditure, which employed in enterprises from which the

Chapter 8: Social Policies for Human Development

164 165 compulsory premiums could be collected. intake. About twenty-eight percent of While the state will contribute the health children under fi ve and twenty percent insurance premium for children and the of women of reproductive age suff ered elderly, so far no solution has been found from iron defi ciency anaemia. Diets were on how to ensure the access to health not only poor, but unbalanced, leading to care by the poor working in the informal excessive calorie intake resulting in obe- sector and small scale agriculture (World sity, with about half of adults overweight Bank 2004). (World Bank 2004).

The fall in health indicators refl ected a Health conditions varied across regions and simultaneous decline in nutrition. In the socio-economic classes. For example, while early 2000s, one in every nine children the infant mortality rate in Taraclia Judet was under fi ve, and about eight percent of about fi fteen per thousand live births, the children under twelve, were stunted, fi gure for Tighina was twenty-two (World exhibiting manifestations of chronic Bank 2003). The incidence of infectious and malnutrition. There were widespread de- parasitic diseases was six per 100,000 people fi ciencies in micronutrients and calcium in Tighina, but four times higher in Orhei.

Table 8.1: Health Indicators for Moldova, 200-2004

Indicators 2000 2001 2002 2003 2004e Male life expectancy at birth (years) 63.9 64.5 64.4 na na Female life expectancy at birth (years) 71.2 71.7 71.7 na na Mortality rate (per 1,000 inhabitants) 11.3 11.0 11.6 11.9 11.4 Morbidity rate (per 1,000 inhabitants) 357.6 338.3 352.3 na na Infant mortality rate (per 1,000 live births) 18.3 16.3 14.7 14.3 12.3 Child mortality rate (per 1,000 live births) 23.3 20.3 18.2 17.8 na Maternal mortality rate (per 100,000 live births) 27.1 43.9 28.0 21.9 na TB incidence (per 100,000 inhabitants) 68.5 89.4 97.3 na na TB-related mortality (per 100,000 inhabitants) 16.9 15.0 15.8 na na HIV/AIDS morbidity rate (per 100,000 inhabitants) 0.1 0.2 0.5 1.2 na HIV carriers (per 100, 000 inhabitants) 4 6 5 5 No. of new HIV/SIDA cases 176 234 206 na na

SOURCE: DSS, Ministry of Health (quoted in EGPRSP). na – not available.

Table 8.2: Number of new cases per year of major groups of diseases (per 1000 inhabitants)

1995 1996 1997 1998 1999 2000 2001 2002 Infectious 35.3 43.1 49.1 43.9 38.0 36.7 35.9 37.8 Respiratory 171.0 118.1 144.9 135.3 131.1 130.7 107.4 111.9 Tumours 3.5 3.8 3.9 3.9 3.4 3.6 3.0 3.2 Digestive 22.6 20.8 21.9 21.5 16.9 15.9 20.9 19.4 Cardiovascular 7.0 6.9 8.0 8.9 7.6 10.0 13.0 12.5

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction SOURCE: Statistical Yearbook of the Republic of Moldova, 2003

Chapter 8: Social Policies for Human Development

164 165 Inequality in Education despite a sharp contraction in birth rates that reduced the number of children of Despite the economic hardships of the pre-school age. Considerable damage transition, Moldova maintained high might result in child socialization, peer educational participation and attain- interaction and school preparedness. ment, particularly compared to other These are life skills in which the family countries at similar income levels. How- can be of limited relevance, particularly in ever, education became less accessible one-child families, single parent families, to the poor. Far from being a source of and for children of poor families whose opportunity, the education system be- parents have less time to stimulate their came a vehicle for the enhancement of intellectual development. inequality. Superfi cially, the population of Moldova appeared well-educated during The fall in enrolments resulted from a the transition. Less than two percent of variety of factors. Shortage of funding the adult population reported that they resulted in closures of kindergartens were unable to read or write. On average managed by state enterprises. In the Moldovans between eighteen and thirty mid-2000s, almost one hundred percent had completed twelve years of education of pre-school institutions were still run and ten percent of Moldovans older than by state institutions (DSS 2003). While eighteen completed university educa- the new policy on pre-school education tion. However, persistent and widening aimed at transferring these services to disparities in educational attainments municipal authorities, decentralization characterised the transition. For exam- suff ered from administrative and fi nancial ple, the average resident of Chisinau and problems, and several local authorities Balti completed two more years of formal were forced to close their kindergartens schooling than the average rural resident, due to lack of funds. Cost-cutting diluted and the non-poor about one year more the quality of the services that were pro- than the poor. The gross enrolment rates vided. There is evidence that the contrac- at the university level in 1997-98 were tion in the supply of kindergarten places twenty-seven percent in large cities and was particularly marked in urban areas. seven in rural areas. In 2001-02, the com- Demand factors seem to have dominated parable fi gures were thirty-seven and ten the contraction in kindergarten enrol- percent, respectively. Table 8.4 presents ment rates, as suggested by the decline in the disparities in terms of poverty status, occupancy rates during the 1990s, when where the same point is can be made. the demand fell more quickly than supply However, educational attainment and en- (Table 8.4), especially in rural areas. Con- rolment ratios for girls and women were fronted with a mounting fi nancial crisis, higher than boys and men at all levels of municipalities introduced or raised fees education (Chacin & Nayar 2004). for school meals, uniforms, heating and bus services that represented a substan- The most strikingly anti-poor develop- tial share of the declining average wage ment in education during the transition (DSS 2003). The conclusion that demand was the slump in the kindergarten at- factors tended to dominate, especially tendance (Table 8.4). The developmental for poor households and in rural areas, and psychosocial benefi ts of pre-school implies that public policy ought to reduce education have long been recognized. drastically the private costs for attending However, enrolment rates for the three kindergarten. through six year olds fell from sixty-one to thirty-two percent from 1989 to 2000. Trends in basic education, i.e. primary Despite recovery, in 2002 the enrolment education, for children seven to ten, and

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction rate was still below its pre-transition lower secondary education, eleven to level. This slump in enrolment took place fi fteen, suggests that public policy ma-

Chapter 8: Social Policies for Human Development

166 167 naged to broadly sustain at high levels new students, reduction in the number enrolments in primary and lower second- of teachers,89 and reductions in the real ary education during the diffi cult initial wages of teachers and other school em- years of transition (1991-95), and in the ployees. While this policy contributed to aftermath of the 1998 crisis. These were deterioration in the quality of education, years during which indicators of pov- its impact on enrolments was less severe erty and social marginalisation showed than had the cuts been a proportional marked deteriorations. Indeed, except reduction in the number of teachers. for a fi ve point decline between 2000 and Fourth, educational reforms allowed pri- 2002 (Table 8.5), over the entire transition vate schools to open, the reform of the period the primary enrolment rate oscil- pre-reform curricula, and the reallocation lated around ninety-fi ve percent despite of students from professional to grammar strains on family incomes and massive and vocational schools. While the latter cuts in public expenditure per capita, that policy may have primarily benefi ted the declined by 2001 to only fi fty-fi ve percent children of the middle class, it contributed of its 1997 level, and twenty percent of its to a recovery of upper secondary enrol- pre-transition level. In turn, enrolments in ments. lower secondary education remained in the low ninety percentages throughout In contrast, the enrolment rate in non- the entire period. compulsory upper secondary education, pupils fi fteen to eighteen, declined stea- To a considerable extent, the success dily until 2000, though recovering rapidly achieved in sustaining enrolment rates during 2001-2004 (Table 8.5). This reco- in basic education was due to policy very resulted from two mutually off set- changes introduced in the educational ting tendencies. The enrolment decline sector. Though the share of GDP assigned was marked among the pupils of voca- to public education declined sharply after tional schools, whose number fell from the pre-transition years, it remained at fi fty-nine to twenty-three thousand from a level higher than most countries with 1990 to 2002, then stabilized during 2003- similar GDP per capita (5.5% of GDP in 2004. It is likely that the decline aff ected 2002). Second, the principle of free provi- the children of low and middle income sion of basic compulsory education was groups, especially from rural areas,90 who not abandoned for pupils six through disproportionately attend these schools. fi fteen, or for non-compulsory education. In contrast, after a thirty-fi ve percent fall However, tuition fees were introduced for between 1989 and 1995, the number of children of better off families attending students enrolled in secondary education non-compulsory classes. While private that leads to university rose steadily to ex- expenditure on education rose, free ceed its pre-transition levels. education for most households avoided a collapse in enrolments among the chil- While the deteriorating quality of educa- dren of the poor, which occurred in other tional institutions played a role, the drop economies in transition that introduced in enrolment rates in upper secondary fees in secondary education. Third, the education can be attributed to three large expenditure cuts were translated demand factors. First, the sharp fall in into an almost complete cancellation of real household income for many families capital investment in education, a policy disproportionately raised the opportu- facilitated by the fall of the cohorts of nity cost of the time of adolescents. The

89For instance, the number of students enrolled in basic education fell from about 650,000 in 1989-91 to 518,000 in 2002. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 90The World Bank (2004) indicates that seventy-fi ve percent of the youth not enrolled in upper secondary education in 2001-2 came from the rural areas.

Chapter 8: Social Policies for Human Development

166 167 ‘opportunity cost’ of the child and ado- qualifi cations to the graduates of voca- lescent’s time grows with the decline of tional schools, declined abruptly both in family income, particularly for families the fi rst part of the transition, and even in which the child or adolescent can fi nd more so from 1997 to 2002 (Table 8.5). employment in on the family farm or in The decline in enrolment was primarily family businesses. In rural areas, part of due to demand factors. The proportion this decline and the increase in drop- of students paying fees in colleges rose out rates seems to be explained by the from sixteen percent in 1997 to twenty- redistribution of land to farmers and seven percent in 2002, with the eff ect of the proliferation of small family shops. excluding potential low-middle income Second, private expenditure on education applicants. The low labour market returns increased considerably. In 2002, school to this type of training was an additional fees were paid by fi fteen percent of the factor in the drop in college enrolment students enrolled in vocational educa- rates. tion (DSS 2003). These two factors proved particularly devastating for the families of The quality of education seems to have the bottom deciles, especially those with deteriorated, which is closely associated two or more children in school-going age. to the remuneration of teachers. In 1995 In 2002 expenditure on education was the average wage in education was 83.4 almost nine percent of the total consump- percent of the national average, and de- tion of the poorest households, and only clined to 60.7 by 2000, then recovered four percent for the richest twenty percent slightly to 67.0 in 2002. Of the major (World Bank 2004). Third, the perceived sectors of employment, only agriculture loss of relevance of vocational education had lower wages than education. Wages contributed to the decline in enrolments, and salaries in the health sector were ap- as wages for this type of skill diminished. proximately the same level as education. The government addressed this problem This resulted in the resignation of some through educational reforms, but increa- of the best and younger teachers, greater sing the relevance of education remained absenteeism due to second jobs, and a problem, particularly for adolescents low morale. Few long–term statistics are in rural areas and in small cities suff ering available for the pre-transition period on from a rapid industrial decline. student performance, so it is impossible to quantify the deterioration in school There was a decline in the total number of performance. Recent data suggests that university enrolments from 1989 to 1994, the educational reforms introduced after but subsequently the number rose, even 1996 improved the literacy and numeracy in the aftermath of the 1998 crisis. By 2002 scores of pupils in basic education (World enrolments had doubled compared to Bank 2004), thought the performance of the pre-transition level. Almost half the Moldavan pupils in relation to other coun- increase was in private universities, aff or- tries remains low. dable only to well off families. Tuition fees were introduced in public universities, In the past, education was freely provided and the proportion of students paying by the state and the income bias in educa- them rose from twenty-seven to sixty- tion was limited. During the transition the six percent between 1997 and 2002 (DSS decline in public expenditure per capita, 2003). The fees do not seem to have dis- the introduction of school fees for non- couraged enrolments, suggesting to the compulsory education and the transfer of emergence of a social class with suffi cient part of the costs of compulsory education income to pay for university. to households has fi rmly established an income bias. In 2002, private expendi-

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction In contrast, enrolment in short-term col- tures on education accounted for about leges, off ering non-academic tertiary two percent of GDP, compared to zero

Chapter 8: Social Policies for Human Development

168 169 during the pre-transition years (World In conclusion, primary and secondary Bank 2004). About half of this expendi- education would be more pro-poor by ture was incurred for the nominally free integrating it into a growth strategy of basic education, refl ecting costs borne by expansionary macro policies, progressive households for contributions to parents- taxation, and public investment. For the teachers associations, school supplies, foreseeable future, even a progressive tax school meals, textbooks and optional structure generating a tax share in GDP of activities. Another forty percent of private about twenty-fi ve percent, would make expenditure on education was on upper universal access to tertiary education secondary and tertiary education. based on ability possible only for a small portion of the population. Therefore, It is necessary to introduce measures that there is no alternative to income-tested reduce the private cost of primary and scholarships to give access to university secondary education. First, this involves an for low-income households. Such a pro- integrated set of policies that achieve an gramme would suff er from arbitrariness increase in the quality of public education, in selection due to the practical prob- such that the skills acquired at the second- lems of income testing discussed at the ary level give access to wages above the start of this chapter. Its pro-poor impact poverty level. This would induce families would also be reduced by what can be to keep their children in school, rather than called the ‘pre-selection problem’. Stu- taking them out of school and putting them dent performance in secondary school is to work at poverty wages. This increase in strongly aff ected by family income, which quality can only be achieved through al- implies that those low-income applicants locating more funds to public education. to university that demonstrate high Second, the incidence of general upper academic performance will not tend to secondary and tertiary education could come from the poorest households. No be made more progressive by eliminating developed country can claim a tertiary formal and informal fees, through public education system which does not dis- funding of all compulsory school activities. criminate against the poor, though some Third, public revenue to improve quality come considerably closer than others. and eliminate charges paid by households Given Moldova’s low per capita income, would come from a more progressive tax limited public resources, and the coping structure. Fourth, expansionary monetary strategies of households (see the chapter and fi scal policy are required that stimulates on migration), substantially reducing the growth, in part through public investment bias against the poor in tertiary educa- in labour-using infrastructure projects. tion is a task for the future. Table 8.3: Net Enrolment Rates by Levels of Education, Moldova, 1997-2002, Percentages

Level 1997 1998 1999 2000 2001 2002 Preschool 39.2 38.4 34.3 70.0 74.0 77.0 Primary na na 94.0 93.5 92.4 92.7 Lower na na 87.0 87.0 86.8 83.9 Secondary Upper na na 29.0 31.5 34.5 38.8 Secondary Higher 18.3 19.0 20.1 20.1 20.4 20.4

SOURCE: Department of Statistics (2003) Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction NOTE: The numerator consists of enrolment data based on school surveys, and offi cial population pro- jections are used for the denominator. Higher education includes non-university and university.

Chapter 8: Social Policies for Human Development

168 169 Table 8.4: Places, Occupancy and Enrolment Rates in Kindergartens, 1989-2002

Variable 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Children in kindergarten 353 336 311 213 202 182 161 147 138 126 101 94 97 104 (‘000) N. of places 294 268 270 218 216 205 194 182 177 166 152 143 145 150 (‘000) … … … … … … … … 83 80 74 65 64 64 - urban … … … … … … … … 94 86 78 78 81 86 - rural Occupancy rate (%) 120 125 115 98 93 89 82 81 78 76 66 65 66 69 - urban … … … … … … … … 85 81 72 73 74 78 - rural … … … … … … … … 73 71 61 58 60 63 Enrolment rate (net) … … … … … … … … 50 49 44 44 47 54 -All … … … … … … … … 73 66 53 49 50 54 -males … … … … … … … … 66 60 48 45 46 50 -females

SOURCE: Elaboration on Trans MONEE Database (for the data 1989-1996), and DSS (2003).

Table 8.5. Enrolment in Primary, Secondary and Tertiary education, 1989-2003

Year 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 A. Output indicators Primary - net na na 95 95 96 96 95 95 96 95 94 93 92 88 - gross na na na na na na na na 99 100 100 99 99 95 Lower secondary na na na na na na na na na na 87 87 87 84 - net 98 95 94 97 93 93 92 93 89 88 91 90 91 88 - gross

na Upper 2ry, gross na na na na na 46 46 47 47 48 40 42 44 48 82 -gen ed (000) 78 76 65 50 50 51 53 57 57 62 59 65 72 80 23 -voc ed (000) 61 59 55 41 37 35 34 34 34 32 23 23 23 23 3ary, gross 40.4 41.8 41.2 38.9 39.6 42.4 - college (‘000) na na … … … … … … 32.7 29.6 25.4 19.9 17.0 15.2 18.7 - univ (000) 55.4 54.7 52.5 47.2 46.8 49.3 54.7 58.3 65.5 72.7 77.3 79.0 86.4 95.0 104 % private …. … …. … …. …. 4.2 6.4 9.4 12.4 16.8 19.6 22.8 23.8 23.1 % children in 31 28 23 14 14 12 10 10 8 7 4 4 5 7 7 after school care B. Input indicators Total education expendit.,% of na na na 7.8 6.0 8.7 7.6 10.3 10.0 7.0 4.7 4.5 4.8 5.5 4.8 GDP

SOURCE: Elaboration on Trans MONEE Database. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 8: Social Policies for Human Development

170 171 8.3 Basic Needs and However, during winter a four-member household living in a three-bedroom Social Protection apartment paid more than one thousand Leu for housing, representing two-thirds of The transition has been associated with se- the offi cial minimum consumption basket. vere declines in access to and quality of the In such conditions, access to other basic most basic needs of households: potable social services, health and education, were water, electricity and other energy sources, diffi cult (Prohnitchi 2004). The poor state and housing. Improving these would have of roads impeded access of children to re- a strong poverty reducing eff ect, and could mote schools and of farmers to market out- be implemented in a pro-poor manner, be- lets. With little expenditure on operations cause for the most part they do not require and maintenance, around one hundred income testing or needs testing at the and fi fty kilometres of national and local household level. The measures to achieve roads deteriorate each year. broader and better quality of access to water, energy and housing for the poor would The social protection system in Moldova be a central part of the public investment consists of a set of social insurance andand social led government expenditure programme assistance p programs.rograms. InIn 2002,2002, socialsocial insur-insur- which this report recommends as the basis ance benefi ts, primarily pensions and short- for a pro-poor growth programme. term benefi ts to those enrolled to social security, accounted for three quarters of the In 1995, ninety-seven percent of the social budget, 7.7 percent of GDP. Because of Moldovan urban population were estima- falling revenue in the aftermath of the Rus- ted to have access to ‘improved’ water, sian crisis, the budget for social assistance though the proportion was less than twen- benefi ts shrank from 2.3 percent in 1999 to ty percent for rural households. The same 1.4 percent in 2002 (see Box 8.1). Social insu- percentage in urban areas had access to rance benefi ts were subject to previous en- basic sanitation, and ten percent in rural rolment into social security of the recipient areas (Czaki & Tuck 2000). By 2000, there or her or his dependent. Social assistance had been a substantial improvement for programs target specifi c groups (categori- rural households, with thirty percent ha- cal targeting): children, disabled, elderly, ving access to ‘improved’ water (Prohnitchi veterans, single pensioners, single mothers, 2004). However, only eighty percent piped and families with many children. These pro- water was judged as sa tisfactory for con- grams were targeted at specifi c groups and, sumption without some form of fi ltering. within the group, income tested. Part of the water problem arose from the severe deterioration of infrastructure, re- The main goal of social assistance pro- sulting in huge losses in central systems of grammes was poverty alleviation, and thirty to sixty percent, due to bad instal- that of the social insurance programmes lations and connections. Related to this, is to compensate for income loss due to access to heating was li mited both for the sickness, maternity, unemployment, or poor and the non-poor, because of dete- death of a dependent of an income earning rioration of the centralized distribution spouse. At the time this report was written, system in ci ties. As a result, in the 2000s, twenty-fi ve percent of the population re- about thirty percent of urba nites and ceived direct social insurance recipients, ninety percent of rural Moldovans used and about six percent received social as- wood stoves to warm their homes. sistance benefi ts. Including the recipients and members of their households, social Access to housing worsened as a result of programmes reached forty-seven per- sharp price increases after 1995. Since 2000, cent of the population. Table 8.6 presents with increases in household purchasing

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction the coverage of the social safety nets in power, housing conditions have improved. Moldova by age groups.

Chapter 8: Social Policies for Human Development

170 171 Box 8.1: Social protection, main cash benefi t programmes Social Insurance benefi ts include:* Old age, disability, and survivors pensions (60) Sick leave benefi ts (5) Maternity Leave benefi ts Unemployment benefi ts (2) Social Assistance benefi ts include: Gas and energy compensations (16) Social pensions (2) Children’s benefi ts (4) In replacement of unpaid separated parent’s child support To single mothers of children between 1.5 and 16 years of age (income tested) To mothers of children under 18 months old To families with many children (income tested) To each mother having given birth Benefi ts to orphan children at education institutions and benefi ts to adopted children Compensations and supplements to rehabilitated people Death Allowances Local material assistance and other transfers Women in full-time employment receive a maternal allowance, funded from the social insurance contributions. The payment is quite small, about US$ ten per month. Besides these cash transfers, the public social safety net is defi ned to include administration of institutions for disabled children and elderly and special transport for the disabled, as well as other programmes administered by the Ministries of Labour such as active labour market programmes.

*The fi gures in the bracket indicate the percentage of social safety funds going tothat category. The fi gure on the sick leave benefi ts also includes maternity leave benefi ts.

SOURCE : Andrews (2003)

Taken as a whole, the social protection quintile was forty-fi ve percent, nine times system of Moldova is pro-poor, espe- more than the three percent enjoyed by cially the old age pension, which has the bottom income quintile. Households been highly successful in protecting the with children less than fourteen years old, elderly from poverty. However, it remains and pensioners benefi ted least. These quite limited in what it can achieve. About problems could be resolved within a uni- seventy-three percent of extremely poor versal entitlement by a simpler, standard recipients had less than three quarter of payment per household. their needs covered by social assistance benefi ts. Gas and energy compensation The government changed the pension benefi ts would seem the least eff ective system in 1999, introducing new minimum programmes. The compensations for payments, which was fully applied in 2003. gas and energy and payments for wood In August of that year, the government and coal for heat were not distributed in passed a new law on pension indexation,

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction a pro-poor manner. In 2003, the share of and eliminated privileged pensions and such benefi ts going to the top income early retirement. As a study for the World

Chapter 8: Social Policies for Human Development

172 173 Bank concluded, the present universal ergy. At the same time, the government entitlement of the old age pension is well- introduced income testing for some child suited to Moldova’s conditions,91 and has a benefi ts. It has been argued that eliminat- substantial poverty reducing eff ect. ing payments for energy using the funds to increase benefi ts to families that have With regard to social assistance, in 2000 three or more children would have a net a range of small benefi ts was eliminated, positive impact on poverty (Lerman 2001, and new categories of vulnerable groups World Bank 2001). The government could defi ned. A new benefi t was introduced to commission a review to assess this pro- compensate the cost of heating and en- posal.

Table 8.6: Social Payments Recipients, Percent of age group, 2002

Early Young Middle Age Retirement Elderly Retirement Social Insurance 3.7 10.2 27.7 94.5 97.7 Social Assistance 5.8 6.6 11.0 9.4 11.0 Pensions Old-age 0.0 0.0 4.1 86.0 92.6 Disability 2.2 8.5 15.4 6.6 3.5 Survivor 1.4 0.0 0.0 0.2 0.8 Privileged 0.1 1.6 8.0 1.6 0.8 Social 1.2 0.2 0.4 0.2 0.8 Energy Subsidies 2.7 6.5 10.8 8.8 9.9 Child Benefi ts 2.4 0.0 0.2 0.2 0.0

SOURCE : Department of Statistics (2002) NOTES: Young, 14-44 for males and females; Middle age, women 45-51, men 45-56; Early retirement, women 52-56, men 57-61; Retirement, Women 57-64, men 62-64; Elderly, 65 and older, men and women.

Table 8.7: Pensioners and Pensions in Moldova, 1999-2004

1999 2000 2001 2002 2003 2004 Av. monthly pension, Leu 83.9 82.8 85.1 135.8 160.8 210.5 Annual real increase, % 46 13 17 na No. of pensioners, 1 Jan 758 728 706 684 658 628

SOURCE: Department of Statistics

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction 91 Comparing a ‘unitary’ system that is not income tested to an income tested ‘two-tiered’ system, Mabbett concludes, ‘Moldova probably does not fulfi l the three criteria [for a two-tiered system]’ (Mabbett 1996).

Chapter 8: Social Policies for Human Development

172 173 8.4 Social Policy for income testing for access to benefi ts. The inherent arbitrariness of income testing Poverty Reduction that results from this structure and cha- ra cteristics suggests that the its use be A pro-poor social policy in Moldova would avoided when alternatives are feasible. have the goal of reducing poverty, rather than merely alleviating it. This implies that 2. A benefi t system based on targeting cate- social policies facilitate the poor in acquir- gories of the population with universal ing the means to permanently exit pov- access has proved poverty reducing in erty. From this perspective, social policy Moldova, with the old age pension an does not provide ‘safety nets’ that catch example. This approach should be ex- the few or many that fall into poverty. So- tended to other population groups. cial policy is part of a growth strategy that 3. Progressive taxation is the key to funding progressively reduces the conditions that pro-poor health and education systems. generate poverty. 4. Constraints on revenue limit the qua lity and extent of public health services. In a low-income country, social policy This requires clear priorities for health faces many trade-off s, and there are expenditures. Policies should be avoided inherent limits to its capacity to reduce that would institutionalise a two-tiered poverty and provide equitable access to health system, in which free access is to a services and benefi ts. The measures listed minimal system which is income tested, below recognise the trade-off s and limits, and other health services are obtained guided by the principle that policy should through private payment. Institutiona- not endorse the division of Moldova be- lising a two-tiered system would institu- tween the poor and the non-poor, but tionalise poverty. should foster systems that progressively eliminate that division. That objective 5. Primary and secondary education should would be achieved by combining social be provided to all without charge. This insurance and assistance programmes requires increased funding for teach- with a more equitable access to health ers’ salaries, education materials, and and education, within a macro framework construction of facilities. The capital of progressive taxation and public invest- expenditure would be part of a public ment. investment fostering growth strategy. 6. Unequal access to university education 1. Because of the nature of the Moldovan cannot be eliminated in the foreseeable economy, its structure (relatively few future; however, it can be moderated by formal sector wage earners) and characte- income tested funding of university fees ristics (importance of remittances), there within a system of access on the basis of are serious practical diffi culties to using ability and school performance. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

Chapter 8: Social Policies for Human Development

174 175 References

Addison, Tony, and Alan Roe Caraşciuc, L, Efi m Obreja, Viorelia Gasca, Na- 2004 (eds.) Fiscal Policy for Development: Poverty, talia Izdebschi, Constantin Lazari, and Marin Reconstruction and Growth (London: Palgrave) Mazur 2002 National Integrity System Indicators: Re- Andrews, Emily public of Moldova (Chisinau: Transparency 2003 Social Protection Programme Recommen- International) dations (Chisinau: World Bank, ms) 2004 Participation, Employment and Unemploy- Carling, Jørgen ment (Washington,(Washington, DC:DC: WorldWorld Bank,Bank, ms)ms) 2004 ‘Policy Options for Increasing the Benefi ts of remittances’ COMPAS Working Paper No. 8, Araujo, Caridad Centre on Migration, Policy and Society (Ox- 2004 Poverty Mapping Mission to Moldova: Final ford: University of Oxford) Report (Chisinau: UNDP) 2005 ‘Migrant Remittances and development Berdaga, Viorica, Svetlana Stefanet, Octavian Cooperation’ PRIO Report 1/2005 (Oslo: Interna- Bivol tional Peace Research Institute) no date Access of the Populaiton of Moldova to Castel, Paulette Health Services (Chisinau: UNICEF) 2004 Public and Private Social Safety Net in Balmori, Helena Hofbauer Moldova (Washington: World Bank, ms) 2003 ‘Gender and Budgets – Overview Report,’ Castel, Paulette, and Lousie Fox (Brighton: IDS Bridge Project) No date Gender Dimensions of Pension Reform Baunsgaard, Thomas and Michael Keen (2004). in the Former Soviet Union (Washington: World Tax Revenue and (Or?) Trade Liberalization. Fiscal Bank) Aff airs Department, International Monetary Center for Economic Policies (IDIS ‘Viitorul’) Fund, September 20 Draft. 2004 “Economic Statewatch: Quarterly Anal- Brundenius, Claes, and John Weeks ysis and Forecast.” Quarter 2. (Chisinau: IDIS) 2001 (eds.) Globalization and third World Socialism (London: Palgrave) Centre for Strategy Studies and Reforms, Counterpart International, Inc. Bruno, M., and W. Easterly 2002 Evaluation of the Social Assistance through 1998 ‘Infl ation Crises and Long-Term Growth,’ the Nominal Targeted Compensations Pro- Journal of Monetary Economics 41 (3-26) gramme: Findings of the Study (Chisinau:(Chisinau: CISR)CISR) Business Consulting Institute and Institutului Chacin, Maria, Eugenia Bonila and Reema Naţional de Economie şi Informaţii Nayar No date Dezvoltaria regionala: Republica Moldo- 2004 Education and Poverty in Moldova: 1997- va vis-a-vis Uniunea Europeana (Chisinau) 2002 (Washington: World Bank, ms) Caraşciuc, L, M. Ciubotaru, N Patiuc, and L Centrul de Investigatii Strategice si Reforme Bejan (CISR) 2003 Corruption and Tax Evasion: Economic dimen- [Various years] Moldova in Transition (Chisinau: sions (Chisinau: Transparency International) CISR) Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

174 175 2003 ‘Lease of agricultural lands – 2003,’ (Chisi- Dudwick, Nora and Akshay Sethi nau: CISR) 2004 Migration: Coping with Economic Hardships 2003 Report on Transnistre (Chisinau & Tiraspol: in Recession, Recovery and Poverty in Moldova. CISR) Vol. 2. no date Evolution of the Transnistre Economy: Ellerman, David, and Valdimir Kreacic Critical appraisal (Chisinau:(Chisinau: CISR)CISR) 2002 ‘Transforming the Old into a Foundation Chirtoaca, Nicolae, Anatol Gudym and Oleg for the New: Lessons of the Moldova ARIA Petrushin Project,’ World Bank Policy Research Working 2002 Nations in Transition 1999-2000, Civil Soci- Paper 2866 (Washington: World Bank) ety, Democracy and Markets in East Central Eu- Economist Intelligence Unit rope and the Newly Independent States: Moldova 2003 Moldova – Country profi le (London: The (New York: Freedom House and Transactions Economist) Publishers) European Bank for Reconstruction and Devel- Chowdhury, Anis opment (EBRD) 2004 ‘Fiscal Policy – Infl ation, Ricardian equiva- 2003 Transition Report 2003: Integration and re- lence, Crowding-out and –in,’ note prepared for gional cooperation (London: EBRD) the Second Workshop on Macroeconomics of Poverty Reduction (Jakarta: UNDP/UNSFIR) Eurostat 2000 New Enterprises in Central European Coun- Czaki, Czaba and Laura Tuck ( (2000).2000). Rural tries in 1998. (Luxembourg: Eurostat) Development Strategy : Eastern Europe and Cen- tral Asia. World Bank Technical Paper No. 484 Fantom, Neil, Lynn MacDonald, Angela Me, (Washingotn: World Bank) and José Cervera-Ferri 2004 PARIS21: Task Team: Improved Statistical Departamentul Statistica si Sociologie (DSS) Support for Monitoring Development Goals, Re- al Republicii Moldova port on the country case study Moldova (Wash- 2003a Labour Market in the Republic of Moldova ington: World Bank) (Chisinau: DSS) 2003b Statistical Yearbook of Moldova 2003 Foster, M. (Chisinau: DSS) 2000 ‘New Approaches to Development Co- 2004a Buletin Statistic 2003/IV (Chisinau:(Chisinau: DSS)DSS) operation: What can we learn from experience 2004b Women and Men of Moldova (Chisinau: with implementing Sector Wide Approaches?’ DSS) (London: Overseas Development Institute) 2004c Statistical Yearbook of Moldova 2003 Department of Statistics (Chisinau: DSS) 2002a Household Budget Survey of Moldova, Department of International Development, 2002. (Chisinau: DSS) Health Systems Resource Centre 2002b Moldova Economic Trends (Chisinau: Gov- 2003 ‘Republic of Moldova,’ Country Health ernment of Moldova) Briefi ng Paper (London:(London: DDFID)FID) 2003a Labour Market in the Republic of Moldo- 2004 Central Asia, South Caucasus & Moldova: va (Chisinau: DSS) Regional Assistance Plan (London: DFID) 2003b Moldova Statistical Yearbook. 2003 (Chisi- nau: Government of Moldova) Development Alternatives, Inc. 2000 Small and Medium Enterprises Baseline Sur- Ghencea, B and I. Gudumac vey (Chisinau:(Chisinau: DevelopmentDevelopment Alternatives)Alternatives) 2004 Labour Migration and Remittances in the Republic of Moldova. Marketing and Research Doudeijns, Marco and Jean-Cristophe Dumont Department, Moldova Microfi nance Alliance 2003 ‘Immigration and Labour Shortages: Eval- (Chisinau: MMA ) uation of Needs and Limits of Selection Policies in the recruitment of Foreign Labour,’ Economic Huppi, Monika and Jakoby, Hanan and Social Aspects of Migration, Conference 2004 Poverty and Health Care (Washington: Jointly organised by The European Commission World Bank, ms) Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction and the OECD, Brussels, 21-22 January 2003

176 177 International Crisis Group Lerman, Zvi 2004 Moldova: Regional Tensions over Transdni- 2001 Moldova: Baseline Farm Survey October-No- estria, Europe Report No. 157 (Brussels:(Brussels: ICG)ICG) vember 2000, Draft Analytical Report. (Washing- ton: World Bank) Institute for Public Policy (Moldova) 2005 (DRAFT) “MigrationMigration PoliciesPolicies inin thethe RepublicRepublic Mabbett, Deborah of Moldova” (Chisinau: IPP) 1996 ‘Social Insurance in a Transition to a no date National Report on the phenomenon of Market Economy: Theoretical Issues with traffi cking in children for sexual exploitation and Application to Moldova,’ World Bank Policy work in Moldova (Chisinau: IPP) Research Working Paper 1588 (Washington: World Bank) International Monetary Fund 1996 Republic of Moldova: Recent Economic Martin, Philip Developments IMF Staff Country Report 96/117 2004 ‘Migration and Poverty in Moldova,’ (Chisi- (Washington D.C.: IMF) nau, ms) 1999 Republic of Moldova: Recent Economic Milanovic, Branko Developments IMF Staff Country Report 99/110 1995 ‘Poverty, Inequality, and Social Policy in (Washington D.C.: IMF) Transition Countries,’ Policy Research Working 2002 Staff Report for the 2002 Article IV Consulta- Paper 1530 (Washington: World Bank) tion (Washington D.C.: IMF) 1997 Income, Inequality, and Poverty during 2003 Operational Guidance on the New Condi- the transition from Planned to Market Economy tionality Guidelines (Washington: IMF) (Washington: World Bank) 2004a ‘IMF Executive Board Reviews the Republic of Moldova’s Poverty Reduction Strategy Paper Mindicanu, Ala and Joint IMF-World Bank Staff Advisory Note,’ 2003 ‘Women’s social and economic rights in Press Release No. 04/256 (Washington: IMF) Moldova’ (Chisinau: ms) 2004b Republic of Moldova: 2003, Article IV Con- 2004 ‘Mainstreaming Gender Policies in the sultation, Staff Report ((WashingtonWashington D.C.:D.C.: IMF)IMF) Republic of Moldova’ (Chisinau: ms) International Organisation for Migration and Ministerul Economiei al Republicii Moldova Swedish International Development Agency 2003 Saracia in Republicia Moldova 2000 [Pov- 2003 ‘Migration Management Assessment’ erty in the Republic of Moldova 2000] (Chisinau: (Chisinau: IOM & Sida) Tipografi a Centrala) International Monetary Fund and the Interna- Ministry of Education tional Development Association 2002 The Pilot on Assessing Pupils’ Basic Skills 2002 Joint Staff Assessment of PRSP Preparation : The VIth Grade. Analytical Report (Chisinau: Process (Washington: IMF & World Bank) Education Assessment Division) 2004 Joint Staff Advisory Note for the Poverty Reduc- Ministry of Finance tion Strategy Paper (Washington:(Washington: IMFIMF & WorldWorld Bank)Bank) 2004 (May) ‘Politica fi scala şi administrarea fi s- Judt, Tony cala,’ (Chisinau: Ministry of Finance ms) 2005 ‘Europe versus America,’ New York Review Ministry of Labor and Social Protection, Re- of Books 52, 2 (10 February) public of Moldova Kapur, Devesh (2002). Yearly Social Report 2002 (Chisinau: Min- 2004 ‘Remittances: The new development man- istry of Labour and Social Protection) tra?’ G-2424 DiscussionDiscussion PaperPaper SeriesSeries No.No. 2929 (Geneva: 2003 Annual Social Report (Chisinau:(Chisinau: MinistryMinistry ofof UNCTAD) Labour and Social Protection) Lanjouw, Peter, Branko Milanovic and Stefano Moldova, Republic of, Ministry of Finance Paternostro 2002 Interim Poverty Reduction Strategy Paper 1998 ‘Poverty and Economic Transition: How do (Chisinau: Government of Moldova) changes in economies of scale aff ect poverty 2004a Economic Growth and Poverty Reduction rates of diff erent households?’ paper presented to Strategy Paper, 204-2006 (Chisinau: Government Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction the Western Economic Association, Lake Tahoe, July of Moldova)

176 177 2004b Medium Term Expenditure Framework 2002b Report on Poverty in Moldova 2000 (Chisi- (2005-2007) (Chisinau: Ministry of Finance) nau: Ministry of the Economy) 2003a Poverty in Moldova 2001: Preliminary results, Moldova, Government of the Republic of, De- an internal note (Chisinau:(Chisinau: MinistryMinistry ofof thethe Economy)Economy) partamentul Statistica si Sociologie 2003b Poverty in Moldova 2002 (Chisinau: Minis- 2004 Buletin Statistic 2003/IV (Chisinau:(Chisinau: StatisticaStatistica try of the Economy) Moldovei) Prekker, A., M. Jakab and M. Schneider Moreno-Fontes Chammartin, G. and F. Cantu- 2002 Erosion of Financial Protection in Health Bazaldua Systems of ECS Transition Economies. World Bank 2005 Identifi cation of potential growth targets Working Paper. (Washington DC) from remittances for the creation of jobs and productive investments in Albania, Moldova ProEra Gup and Ukraine ILO International Migration papers 2003 Report on the Regulatory Costs Assessment 75 (Geneva: ILO) Survey of the Republic of Moldova during the 2000-2002 Period (Chisinau:(Chisinau: ProEra)ProEra) Muravschi, Alexandru 2004 Cost of Doing Business Survey, Draft Report 2004 ‘Current Development Status of the Ag- (Chisinau: ProEra) ricultural Value Chain’, background paper prepared for the UNDP Mission (Chisinau: ms) Prohnitchi, Valeriu 2003 Poverty in Moldova: A synthesis of literature Murrugarra, Edmundo, and Jose Signoret (Chisinau, ms) 2003 ‘Vulnerability in Consumption, Education 2004 Reform Priorities of the Social policy in and Health: Evidence from Moldova during the Moldova (Chisinau, ms) Russian Crisis,’ World Bank Policy Research Work- ing Paper 3010 (Washington: World Bank) Radziwill, Artur, Octavian Serbatchi and Con- stantin Zaman Obreja, Efi m, and L Caraşciuc 1999 Financial Crisis in Moldova – Causes and 2002 Corruption in Moldova: facts, analysis and Consequences (Warsaw: Centre for Social and proposals (Chisinau: Transparency Interna- Economic Research) tional) Rodrik, D. Obreja, Efi m, Viorelia Gasca, Natalia Calenic 2003 (ed.) In Search of Prosperity (Princeton:(Princeton: 2003 The Customs Activity in the Republic of Princeton University Press) Moldova and Corruption (Chisinau: Transpar- ency International) Ronnas, Per, and Nina Orlova 2000 Moldova’s Transition to Destitution (Stock- Obreja, Efi m, Viorelia Gasca, and Teodor Potirniche holm: Sida) 2002 Public Procurement and Public Ethics: a focus on fi ghting corruption (Chisinau: Transpar- Roy, R., and J. Weeks ency International) 2004 ‘Making Fiscal Policy Work for the Poor,’ UNDP Asia-Pacifi c Regional Programme on Mac- Olortegui, Armando roeconomics of Poverty Reduction (New York: 2001 “Moldova: The External Debt Problem, UNDP) Causes and Likely Remedies — A Summary Review.” Paper prepared for the Department Rutkowski, Jan of Public Debt, Ministry of Finance of Moldova 2004 ‘Firms, Jobs and Employment in Moldova,’ (Chisinau: ms) World Bank Policy Research Working Paper 3253 (Washington: World Bank) Pace, Pierella 2002 Gender in Transition (Washington DC: Scanlan, Shivan World Bank) 2002 Report on Traffi cking from Moldova: Irreg- ular labour markets and restrictive migration Poverty and Policy Monitoring Unit, Ministry policies in Western Europe (Chisinau: ILO) of the Economy 2002a ‘Poverty in the Republic of Moldova: Schieber, S. and John Shoven Diagnostics, Determinants, Conclusions’ (Chisi- 1999 The Real Deal: The History and Future of So- Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction nau: Ministry of the Economy) cial Security (New(New Haven:Haven: YaleYale UniversityUniversity Press)Press)

178 179 Sleptovia, Evghenia United Nations Development Fund 2003 ‘Labour Migration in Europe: Special Focus 1998a National Human Development Report: Re- on the Republic of Moldova’ (Chisinau: IPP) public of Moldova 1998 (Chisinau: UNDP) 1998b Poverty in Transition? (New(New York:York: UNDP)UNDP) Solonari, Validmir 1999a Development Cooperation, Republic of 2003 ‘The Political Economy of Moldova,’ Paper Moldova 1999 (Chisinau: UNDP) prepared for the Lucerne Conference of the CIS-7 1999b National Human Development Report: Initiative, 20-22 January (Chisinau) Republic of Moldova 2000, Transition and Human Standing, Guy Security (Chisinau: UNDP) 1997a ‘The Human Development Enterprise,’ 2000 National Human Development Report: ILO Enterprise Forum 96 (http://www.ilo.org/ Republic of Moldova 2000: Towards a Culture of public/english/employment/ent/entforum/ Peace (Chisinau: UNDP) forum96/sessiond.htm) 2002 ‘Gender Equality in the Republic of 1997b ‘What Would be a ‘Good’ Firm? The “Hu- Moldova,’ Sociological Research 2003a Millen- man Development Enterprise”,’Industrial’Industrial Rela-Rela- nium Development Goals: Eradicating Extreme tions Journal 28,28, 4 Poverty and Hunger, Implementation Perspective Stepanyan, Vahran and Monitoring Policy Impact (Chisinau:(Chisinau: UNDP)UNDP) (2003) ‘Reforming Tax Systems: Experience of 2003b National Human Development Report the Baltics, Russia, and Other Countries of the 2003, Republic of Moldova: Good Governance and Former Soviet Union,’ IMF Working Paper 03 173 Human Development (Chisinau:(Chisinau: UNDP)UNDP) 2003c The Roma in Central and Eastern Europe: Swedish International Development Agency Regional Human Development Report, Avoiding 2002 Support for Economic Reforms for 2003 the Dependency Trap (New York: UNDP) – Moldova (Stockholm: Sida) 2003d The UNDP in Moldova, Special Issue: Millen- 2004 Strategy for cooperation with Moldova, nium Development Goals (Chisinau: UNDP) 2004-2006 (Stockholm: Sida) no date Dreams into Reality (Chisinau:(Chisinau: UNDP)UNDP) no date The case for a bridge fi nancing operation (Stockholm: Sida) United States Agency for International Devel- opment (USAID) Tibi, Claude, Sue E. Berryman, and Michael 2002 Challenges for Development, Moldova 2002: Peleah USAID Fiscal Reform Project Policy Analysis (Chisi- 2002 Moldova’s Education Sector: A fi nancing nau: USAID) strategy to leverage system-wide improvement 2004 ‘Moldovan High Value Agriculture Export (Washington: World Bank) Competitiveness Study’, Chisinau: USAID) Transparency International United States of America, Embassy 2002 Corruption and access to justice (Chisinau: 2002 Novoye Vremya. No. 39, October 25. Transparency International) 2003 Small business and corruption: a guide for Weeks, John entrepreneurs (Chisinau: Transparency Interna- 2001 ‘A Tale of Two Transitions: Cuba and Vi- tional) etnam,’ in C. Brundenius and J. Weeks (eds.) Globalization andThird World Socialism (London: United Nations Children’s Fund Palgrave) 2002a Access of the Population of the Republic of 2003 ‘Small Manufacturing Establishments in Moldova to Health Services (Chisinau: ms) Developing Countries: An empirical analysis,’ 2002b The Nutritional Status of the Population in International Review of Applied Economics, Moldova (Chisinau: UNICEF and the Ministry of Health) Weeks, John, David Anderson, Chris Cramer, Alemayehu Geda, Degol Hailu, Frank Muhe- United Nations Country Team reza, Matteo Rizzi, Eric Ronge and Howard 2000 Common Country Assessment ( (Chisinau:Chisinau: Stein (Weeks, et. al.) UNCT) 2002 Supporting Ownership: Swedish Develop- 2001 The United Nations Development Assistance ment Cooperation with Kenya, Tanzania and

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction Framework (Chisinau:(Chisinau: UNDP)UNDP) Uganda (Stockholm: Elanders Novum)

178 179 Weeks, John, Nguyen Thang, Rathin Roy and 2003c Moldova Trade Diagnostic Study ( (22 V Vol-ol- Joseph Lim umes) (Chisinau: World Bank) 2004 Viet Nam: Seeking Equity within Growth 2004a Moldova: Investment Climate Report (London: Centre for Development Policy & Re- (Chisinau: World Bank) search) 2004b ‘Moldova: World Bank Supports Eco- nomic Growth and Poverty Reduction Strategy,’ World Bank Press Release 11 November (Washington: World 1993 Statistical Handbook 1993: States of the Bank, www.worldbank.org) Former USSR (Washington, DC: World Bank) 2004c Poverty Assessment in Moldova 2004 1999 Memorandum of the President of the Inter- (Washington: World Bank) national Bank for Reconstruction and Develop- 2004d Recession, Recovery and Poverty in Moldo- ment, the International Development Association, va (Washington, D.C.: World Bank) and the International Finance Corporation to the 2004e The World Bank in Moldova (Chisinau: Executive Directors on a Country Assistance Strat- World Bank) egy for the Republic of Moldova (Washington: no date Costs of Doing Business Survey: A World World Bank) Bank country study ( (Washington,Washington, D.C.:D.C.: W Worldorld 2000 Making Transition Work for Everyone: Pov- Bank) erty and Inequality in Europe and Central Asia 2004e World Development Indicators (Washing- (Washington: World Bank) ton D.C.: World Bank, CD rom) 2001 Agricultural Strategy for Moldova: Accel- 2004f World Development Report 2004 (New erating Recovery and Growth (Washington DC: York: Oxford University Press) World Bank) *no date Poverty in Moldova (1997-2002): Pov- 2002 Staff Guidance Note: Operationalizing Bank- erty Assessment Concept Paper ((Chisinau:Chisinau: WorldWorld Fund Collaboration in Country Programs and Bank) Conditionality (Washington:(Washington: WorldWorld Bnak)Bnak) 2003a Moldova Health Policy Note: The Health Zghibarta, Nina Sector in Transition (Washington: World Bank) No date ‘Position of Women on the Labour 2003b Moldova: Public Economic Management Market in Moldova,’ SEELINE Project (Chisinau: Review (Washington,(Washington, D.C.:D.C.: WorldWorld Bank)Bank) ms) Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

180 181 Short biographies of the main contributors

John Weeks (team leader) is Professor of advice and inter-agency collaboration on Development Economics and Director of these issues. Holding a Ph.D in Economics the Centre for Development Policy and from McGill University, Montreal, Canada, Research at the School of Oriental and Mr. Jahan is the author of six books and African Studies, University of London, UK. more than 150 articles in national and Prof. Weeks has served as a consultant international academic journals. He con- to numerous development agencies and tributes regularly to newspapers and is author of many articles and books on periodicals the issues facing developing countries. His research interests include migration, Terry McKinley is also a Senior Policy structural adjustment and stabilisation Adviser on Poverty Reduction in UNDP’s programmes, the eff ects of hyperinfl a- Bureau for Development Policy in New tion and the classical theory of value and York. He holds a Ph.D. in Development growth. His teaching interests are political Economics from the University of Califor- economy and development studies. nia, USA. His specializations are poverty reduction, growth, inequality, employ- Giovanni Andrea Cornia is Professor ment and human development. He has of Economics at the University of Flor- contributed to many of UNDP’s global ence, Italy. He has also lectured in many Human Development Reports, as well as research centres in both developing and to several National Human Development developed countries. Prof. Cornia’s work Reports and is the principal author of UN- has focused primarily on macroeconomic, DP’s global Poverty Report (Overcoming distributive, poverty and human develop- Human Poverty). He has organized and ment issues in developing countries and, contributed to many studies on economic starting in 1989, transition economies. He policies and poverty reduction and au- is the author or co-author of 10 books and thored books including Implementing a 16 chapters in books edited by others and Human Development Strategy ( (withwith KeithKeith has published 50 articles in refereed jour- Griffi n), The Distribution of Wealth in Rural nals or other serials. China and Macroeconomic Policy, Growth and Poverty Reduction. Mr. McKinley’s Selim Jahan is Senior Policy Adviser, Em- current research and policy interests in- ployment for Poverty Reduction in the clude promoting consistency between Bureau for Development Policy (BDP), economic policies and poverty reduction United Nations Development Programme strategies; employment generation and in New York. Mr. Jahan is engaged, on the poverty reduction; and the implications one hand, in research on such issues as for economic policies of linking poverty

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction pro-poor economic and social policies, reduction strategies to the Millennium and on the other, in country-level policy Development Goals.

180 181 S. R. Osmani is Professor of Develop- as for the Dutch Ministry of Development ment Economics at the University of Cooperation. Ulster, United Kingdom. He holds a Ph.D in Economics from the London School of Felicia Izman graduated from the Central Economics and has worked at the Bang- European University (Budapest, 1997) ladesh Institute of Development Studies with an M.A. in Economics. On the job, in Dhaka, Bangladesh and at the World she has specialized in the agriculture Institute for Development Economics Re- sector: policy analysis, farm surveys and search in Helsinki, Finland. Prof. Osmani studies, various aspects of agriculture and has published widely on issues related food sector development. Ms. Izman has to poverty, inequality, hunger, famine, worked as an agricultural economist in nutrition, rights-based approach to de- Chisinau, Moldova for agriculture-related velopment, and development problems projects fi nanced by the World Bank and in general. USAID. She is the author or co-author of various publications, one of the most re- Michel Reynolds is an independent decent being the USAID/CNFA “Moldovan velopment consultant with substantial High Value Agriculture Export Competi- experience in the transition countries of tiveness Study” (2004). Eastern Europe, the Former Soviet Union and Asia. He has been involved in the Dr. Anatol Gudym is Executive Direc- development and implementation of tor of the Center for Strategic Studies national poverty reduction strategies for and Reforms (a Moldovan NGO, http: over ten years and has worked as a con- //cisr-md.org), Chişinău/Moldova. Areas sultant to the World Bank, Asian Develop- of research: macroeconomics, transitional ment Bank, UNDP and UN country teams, processes, regional development. For- Save the Children Fund, the EU etc. merly he worked at Institute of Econom- ics, Moldovan Academy of Sciences Max Spoor is Associate Professor of (researcher 1965-1985), State Institute of Transition Economics and Coordinator of Planning (director 1986-1990), hold the the Centre for the Study of Transition and post of Vice-minister of Economy 1994- Development (CESTRAD), at the Insti- 1997, worked in UNDP and World Bank tute of Social Studies (ISS) in The Hague, Project “Strategy for Development” (na- Netherlands and Professor Extraordinary tional coordinator 1997-1999). at the Centre for International Relations and Development (CIDOB) in Barcelona, Roman Ladus has macroeconomic and Spain. Prof. Spoor has published widely private sector consulting experience. He on rural economics, environment and worked as business consultant and was agricultural policies of developing and also advisor for the USAID Fiscal Reform transition countries. In the past decade and SME Development Projects. He is his research has focused on the trans- the Director of the Center for Assistance formation of the rural economy and en- to Public Authorities, a think tank that vironmental problems in the Post-Soviet provides development solutions to local states (such as in Central Asia, the Cau- and central public authorities and civil casus and Moldova). He has worked as a society. He holds an M.Sc. degree in business administration and economics from

Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction development consultant and economist for many international agencies as well Stockholm University.

182 183 Ms. Ala Mandicanu is President of the Women’s Christian Democratic League National Council of Women and of the US and chair of the Sub-commission for Equal Embassy Alumni Board, Senior Lecturer Opportunities, she promotes the princi- and Head of the Department of Public ples of gender equality and the rights of Communication at the Free University of women in her university activities and in Moldova (ULIM), where she is also image projects, as well as in her written work. Dr. and communications adviser. While she Mundicanu has published numerous sci- holds a PhD in Economics, Ms. Mandicanu entifi c works, articles on political analysis, is a journalist and a politician. A former editorials and interviews with prominent Member of Parliament for Moldova’s personalities in diplomacy and politics. Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

182 183 Republic of Moldova: Economic Policies for Growth, Employment and Poverty Reduction

184