Macro - Poverty and Inequality in Cambodia

Copyright 2004, Asia-Pacific Regional Programme on The Macroeconomics of Poverty Reduction, United Nations Development Programme.
. This publication presents some of the principal findings from the UNDP's Asia-Pacific Regional Programme on The Macroeconomics of Poverty Reduction. DISCLAIMER The responsibility for opinions in this publication rests solely with its authors. Publication does not constitute an endorsement by the United Nations Development Programme or the institutions of the United Nations system.
Design and layout by Graphic Roots - graphicroots@online.com.kh Printed by Ponleu Khmer Printing House
THE MACROECONOMICS OF POVERTY REDUCTION
IN CAMBODIA
Melanie Beresford Nguon Sokha Rathin Roy Sau Sisovanna Ceema Namazie
MARCH 2004
Contents
Contents.............................................................................................................................................................................4 List of tables......................................................................................................................................................................7 List of figures....................................................................................................................................................................8 Abbreviations...................................................................................................................................................................9 Special Message............................................................................................................................................................11 Preface..............................................................................................................................................................................12 Summary of the report...............................................................................................................................................14
1. THE DIMENSIONS OF THE PROBLEM: AN OVERVIEW OF POVERTY AND INEQUALITY IN CAMBODIA...........26

1. Introduction............................................................................................................................................................26 2. Changes in Incidence and Depth of Income Poverty...............................................................................27 3. Level and trends in keys human poverty indicators.................................................................................30 3.1 3.2 3.3 3.4 Basic health indicators..............................................................................................................................30 Education indicators..................................................................................................................................33 Social Exclusion...........................................................................................................................................36 Vulnerability to external shocks.............................................................................................................37
4. Rates and changes in inequality......................................................................................................................38 4.1 The distribution of consumption..........................................................................................................38 4.2 Functional distribution.............................................................................................................................40 5. Levels and changes in the distribution of productive assets.................................................................44 5.1 5.2 5.3 5.4 Land.................................................................................................................................................................45 Other Assets..................................................................................................................................................46 Asset Markets...............................................................................................................................................46 The credit market........................................................................................................................................47
6. The Demographic Challenge............................................................................................................................48 6.1 The impact of the baby boom................................................................................................................48 6.2 Internal Migration.......................................................................................................................................49 7. Poverty and the rural economy........................................................................................................................50 8. Conclusion...............................................................................................................................................................52
2. THE MACROECONOMIC FRAMEWORK..................................................54

1. The Legacy of Cambodia's historical experience.......................................................................................54 1.1 Population and Labour Force.................................................................................................................56 1.2 Governance...................................................................................................................................................59 2. The sources of growth since 1993...................................................................................................................62
The macroeconomic of poverty reduction in Cambodia Contents
2.1 Sectoral pattern of growth......................................................................................................................62 2.2 Financing of growth...................................................................................................................................65 3. Recent focus of macroeconomic policies.....................................................................................................67 3.1 The donor perspective..............................................................................................................................67 3.2 Implementation in Cambodia................................................................................................................68 3.3 An alternative perspective.......................................................................................................................69 4. Conclusion...............................................................................................................................................................75
3. MONETARY AND EXCHANGE RATE POLICY...........................................78

1. Introduction............................................................................................................................................................78 2. Basic Environment Surrounding the Conduct of Monetary Policy......................................................80 2.1 Institutional constraints............................................................................................................................80 2.2 Dollarization of the Economy.................................................................................................................81 2.3 The role of market expectation..............................................................................................................83 3. Constraints on Monetary Policy Instruments..............................................................................................86 3.1 3.2 3.3 3.4 3.5 3.6 Responsiveness to interest rate changes............................................................................................86 Standing facilities........................................................................................................................................87 Minimum reserve requirement..............................................................................................................87 Exchange rate policy and foreign reserves management............................................................89 Design of monetary policy under the PRGF program....................................................................90 Policy implications......................................................................................................................................92
4. Costs and Benefits of Dollarization for Economic Development and Poverty Reduction.............95 4.1 Advantages of dollarization....................................................................................................................95 4.2 Disadvantages of dollarization..............................................................................................................96 5. Dollarization and Cambodia's Foreign Trade ...........................................................................................100 5.1 Implications for exports.........................................................................................................................101 5.2 Implications for imports.........................................................................................................................103 5.3 Policy implications...................................................................................................................................105 6. The Choice of Monetary System....................................................................................................................106 6.1 Dollarization as an interim measure..................................................................................................106 6.2 Steps toward de-dollarization.............................................................................................................107 6.3 Policy implications...................................................................................................................................109 7. Conclusion............................................................................................................................................................110
4. FISCAL POLICY......................................................................................114
1. Introduction.........................................................................................................................................................114
2. The macro-fiscal situation: an overview.....................................................................................................115 2.1 Institutional and macroeconomic strategic initiatives: fiscal components.........................117 3. Analysis of Cambodia's Fiscal Stance...........................................................................................................124 3.1 3.2 3.3 3.4 3.5 3.6 3.7 Domestic resource mobilization.........................................................................................................124 Dependence on donor finance............................................................................................................127 Costs associated with the chosen strategy......................................................................................128 A form of Dutch disease.........................................................................................................................132 Fiscal benefits from civil service reform............................................................................................133 Additional payments to civil servants by donors..........................................................................135 Disbursement of resources...................................................................................................................135
4. Conclusions..........................................................................................................................................................138 4.1 Concluding remarks................................................................................................................................138 4.2 Summary of principal analytical conclusions.................................................................................140
5. TRADE LIBERALIZATION......................................................................144
1. Introduction.........................................................................................................................................................144 2. Review of policy to date...................................................................................................................................145 3. Impacts of trade liberalization.......................................................................................................................147 3.1 3.2 3.3 3.4 3.5 Structure of exports.................................................................................................................................149 Employment impact................................................................................................................................156 Enclave development.............................................................................................................................159 Value added................................................................................................................................................159 Impacts of liberalization on import growth....................................................................................160
4. Developing Cambodia's Exports Further...................................................................................................163 4.1 Potential for future growth of exports..............................................................................................163 4.2 Where are the major constraints?.......................................................................................................168 5. Conclusion: policies for growth and redistribution under trade liberalization.............................171
6. FINANCIAL LIBERALIZATION...............................................................172
1. Introduction.........................................................................................................................................................172 2. Financial Sector Reform in Cambodia.........................................................................................................173 2.1 Transformation of structure and function.......................................................................................173 2.2 Legal infrastructure..................................................................................................................................175 3. Banking Sector Development........................................................................................................................176 3.1 Structural reform of the sector.............................................................................................................176 3.2 Liberalization of interest rates..............................................................................................................178 3.3 Present structure and performance of Cambodia's banks.........................................................181
4. The Structure and Performance of Rural Finance....................................................................................182 4.1 4.2 4.3 4.4 Regulation and supervision..................................................................................................................183 Sources of rural financial services.......................................................................................................184 Demand for rural financial services....................................................................................................188 Savings mobilization...............................................................................................................................189
5. Conclusion............................................................................................................................................................190
7. PRIVATIZATION....................................................................................192
1. 2. 3. 4. 5. Introduction.........................................................................................................................................................192 The Foreign Trade Bank....................................................................................................................................193 Rubber plantations............................................................................................................................................194 Privatization of land and common resources...........................................................................................195 Conclusion............................................................................................................................................................197
List of references........................................................................................................................................................200 Persons consulted.....................................................................................................................................................204 Appendix: project level interventions................................................................................................................206
List of Tables
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 1.13 1.14 2.1 2.2 2.3 2.4 2.5 2.6 3.1 3.2 3.3 Cambodian head-count indices...............................................................................................................29 Early childhood mortality rates.................................................................................................................31 Population aged 25 or more by educational level..............................................................................33 School enrolment by consumption quintile and region, 1997......................................................34 Male and Female enrolment rates, 1997................................................................................................35 Consumption inequality profile...............................................................................................................39 Contributions of income from different sources, 1999.....................................................................40 Main source of income of household heads in previous year by size of landholdings (%)........................................................................................................................41 Composition of employment (%) by sector of primary employment and sex, 1993/4 and 1999...........................................................................................................................42 Distribution within households by labour-force category and sex 1999...................................43 Distribution of employees by type of employer and sex, 1997 and 1999..................................44 Rural Distribution of Assets and Income................................................................................................46 Migrants from rural areas 1996 and 1998..............................................................................................49 Rice production trends................................................................................................................................51 Employment by sex.......................................................................................................................................59 Annual average growth rates and shares of real GDP by sector 1999-2001..............................62 Sectoral contributions to growth 1993-2001.......................................................................................63 Employed persons 10 years and over, 2000..........................................................................................64 Investment share of GDP 1993-2001 (%)...............................................................................................65 Balance of payments 1994-2001..............................................................................................................66 Evolution of Foreign Currency Deposits, 1990-June 2002...............................................................82 Interest Rates on Riel and USD Deposits................................................................................................86 Bank Reserves, 1993-2001..........................................................................................................................88
3.4 3A.1 3A.2 3A.3 4.1 4.2 4.3 4.4 4.5 4.6 4.7 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6.1 6.2 6.3 6.4
Estimated Seigniorage Loss for the Year 2001.....................................................................................98 Cambodia - Monetary Survey, 1993 - 2001.........................................................................................112 Cambodia - Exchange Rates, 1990 - 2001...........................................................................................113 Cambodia - Inflation, 1994 - 2001..........................................................................................................113 Macro-fiscal statistics 1993-2002...........................................................................................................115 Cambodia Revenue GDP ratios 1991-2002 (%).................................................................................116 Key macro-fiscal forecasts 2003-2005..................................................................................................121 Baseline and allocable ceilings 2003-2005.........................................................................................122 GDP expenditure shares (current and domestically financed capital expenditure as per cent of GDP) and sector shares in current expenditure...........................122 Functional classification of government current expenditure 1994-2002..............................129 Monthly figures for key fiscal heads, Financial year 2001..............................................................137 Evolution of Trade Policy...........................................................................................................................146 Merchandise Trade, 1994-2002..............................................................................................................148 Merchandise Trade of Exports, 1994-2001, % distribution...........................................................149 Exports by commodity, 1993-2001.......................................................................................................152 Exports and Imports of Milled Rice, 1995-1999................................................................................154 Tourism, 1995-2002....................................................................................................................................155 Key Indicators for the tourism industry, 1995 & 2000.....................................................................158 Merchandise import trade, 1994-2001, % distribution..................................................................161 Import and Export Cost Comparisons, July 2002.............................................................................169 Selected indicators of financial development...................................................................................177 Cambodian interest rates, 1996-2001..................................................................................................178 Distribution of interest rates....................................................................................................................187 Use of loans...................................................................................................................................................188
List of Figures
2.1 2.2 2.3 2.4 2.5 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 5.1 6.1 Per capita monthly household expenditure 1993-94 and 1999....................................................64 Investment by source 1993-2001.............................................................................................................65 Harvested area and yields of wet and dry season rice 1993-2001................................................73 Changes in paddy output and prices 1993-2001................................................................................74 Public and private investment 1993-2001............................................................................................75 Composition of Money Supply, 1993-June 2002..........................................................................82 Money Supply, Net Claims on Government and Inflation, 1989 - 2000......................................84 Government Budget Deficits and Exchange Rates, 1990-2001.....................................................85 Income Velocity of Money,1995-2001....................................................................................................93 Consumer Price Index and Inflation Rates, 1995-2001.....................................................................99 Domestic Exports, Retained Imports and Trade Balance, 1993-2001.......................................101 Domestic Exports, Retained Imports and Exchange rates, 1992-2001.....................................102 Exchange Rate Movements of the US dollar against the riel, Singapore dollar, and Thai baht, 1995-2001......................................................................................104 Foreign Visitors by Age Group, Ministry of Tourism November 1999........................................155 The banking system in Cambodia.........................................................................................................180
Abbreviations
ACLEDA ADB AFTA ASEAN CAR CDC CDRI CEPT cif CPP CSES D-MTEF EMT ESAF EU FDI fob FTB GDP GEM GMAC GRET GSP GTZ HCI HDI IAS IF IFAD ILO IMF I-PRSP ISA KFW KHR M2 MEF MFI Association of Cambodian Local Economic Development Agencies Asian Development Bank ASEAN Free Trade Area Association of South-East Asian Nations Council on Administrative Reform Cambodia Development Council Cambodia Development Resource Institute Common Effective Preferential Tariff customs, insurance and freight Cambodian People's Party Cambodia Socio-Economic Survey Draft Medium Term Expenditure Framework Ennatien Moulethan Tchonnebat (Rural Credit Project) Enhanced Structural Adjustment Facility European Union Foreign direct investment free on board Foreign Trade Bank Gross Domestic Product Gender Empowerment Measure Garment Manufacturers' Association of Cambodia Group de Recherche et d'Echanges Technologiques Generalized System of Preferences Gesellschaft fur Technische Zusammenarbeit Head Count Index Human Development Index International Accounting Standards Integrated Framework International Fund for Agricultural Development International Labour Organization International Monetary Fund Interim Poverty Reduction Strategy Paper International Standards of Auditing Kredittanstalt fur Wiederaufbau Khmer Riel Aggregate Money Supply Ministry of Economy and Finance Micro-Finance Institution
The macroeconomic of poverty reduction in Cambodia Abbreviation
MFN MOP MRD MTEF MWVA NBC NCG NCP NDADMBs NDANBC NFADMBs NFANBC NGO NIS ODA OIN PAP PER PGI PIP PNBC PRGF PRSP RDB RGC SEDP II SME UNDP UNICEF UNTAC USAID VAT WB WFP WTO
Most favoured nation Ministry of Planning Ministry of Rural Development Medium Term Expenditure Framework Ministry of Women's and Veterans' Affairs National Bank of Cambodia Net Claims on Government Net Claims on the Private Sector Net Domestic Assets of Deposit Money Banks Net Domestic Assets of the NBC Net Foreign Assets of Deposit Money Banks Net Foreign Assets of the NBC Non-government organization National Institute for Statistics Official Development Assistance Other Items Net Priority Action Program Public Expenditure Review Poverty Gap Index Public Investment Program People's National Bank of Cambodia Poverty Reduction and Growth Facility Poverty Reduction Strategy Paper Rural Development Bank Royal Government Of Cambodia Socio-Economic Development Plan II - 2001-2005 Small and Medium Enterprise United Nations Development Programme United Nations Children's Fund United Nations Temporary Administration of Cambodia United States Agency for International Development Value-added Tax World Bank World Food Program World Trade Organization
10
The macroeconomic of poverty reduction in Cambodia Abbreviation
Special Message
Macroeconomic policies affect both the rate of growth of the economy and the distribution of the benefits of growth across economic sectors and social groups. The process of structural change that is currently underway in many Asian countries adds new dimensions to the linkage between macroeconomic policies and poverty reduction. With these considerations in view as well as in the context of supporting the achievement of the Millennium Development Goals, UNDP's Regional Bureau for Asia and the Pacific launched in early 2002 the Asia-Pacific Regional Programme on the Macroeconomics of Poverty Reduction. The objective is to identify and promote more pro-poor macroeconomic policies in the region. With two-thirds of the world's poor living in the region such pro-poor approaches are critical to the process of poverty reduction. Fostering greater consistency between country's macroeconomic framework and its national poverty reduction strategy and broadening the policy options and dialogue on poverty reduction issues have been the pillars of the programme. As part of the programme, policy-oriented research in Bangladesh, Cambodia, China, Indonesia, Mongolia, Nepal and Vietnam have been carried out by teams of international and national consultants together with policy advisors from the UNDP's Bureau for Development Policy. These reports help to identify practical policy options to foster more pro-poor stabilization, economic restructuring and growth. These draft reports were presented and discussed in the first regional workshop on the Macroeconomics of Poverty Reduction in January 2003 in Kathmandu, Nepal. The representatives from governments and other relevant stakeholders gave various comments and suggestions in the context of increasing the usefulness of the reports for policy formulation. After incorporating these comments and suggestions into the country reports, a series of national workshops were organised during March-November 2003 in partnership with respective governments to discuss these reports with wider audiences. After refining these reports in line with the comments and suggestions received in these national workshops, six thematic case studies have now been completed on Fiscal, Monetary & Exchange rate Policies, Trade & Financial Liberalisation and Privatisation / De-regulation. I would like to convey my appreciation to the authors, to the focal points of the programme from UNDP country offices and Bureaus as well as the UNOPS-Asia office for their hard work and diligence. I would also like to record my special appreciation to the Programme Coordinator and staff of the Regional Programme based in Kathmandu, Nepal for their support and substantive contributions to the case studies. I hope that the independence of views and the professional competence of the authors ensure that the conclusions and recommendations will have the greatest possible audience and that the thematic case studies will be read with great interest. Hafiz A. Pasha
UN Assistant Secretary General & UNDP Assistant Administrator and Regional Director of the Bureau for Asia and the Pacific
The macroeconomic of poverty reduction in Cambodia Special message
11
Preface
This study focuses on how macroeconomic policy tools have impacted poverty. The set of macroeconomic issues that are the focus of this study are; monetary policy, fiscal policy, trade liberalization, financial liberalization and privatization. Most poverty alleviation theory and practice focuses on micro-level and sectoral interventions without regard to the ways in which the macroeconomy influences the outcomes of such interventions. Since much economic theory teaches that the macroeconomy is no more than an aggregation of its micro-components, this approach is hardly surprising. However, the numerous interdependencies between individual economic agents make simple aggregation impossible and this means that micro-level actions can have outcomes which are quite different from those intended. This study therefore approaches the subject from a different viewpoint - that in which the whole is not the sum of its parts. We start from the premise that macroeconomic policy settings, even if they produce respectable economic growth and stability, often have distributional impacts that can constrain the best efforts of individuals to climb out of poverty. As we shall show, there is evidence that this has happened during the last decade in Cambodia. Some elements of the study which were not part of our direct concern need to be mentioned here as they provide important context to the study. First, we note that poverty itself encompasses a much wider concept of deprivation than the purely material. However in this study the main focus is on how macroeconomic policies can affect income poverty. The most direct route, in the absence of a social safety net, is through increasing sustainable employment opportunities. The labour market is thus the main link between macroeconomic policy and poverty reduction. Wider dimensions of deprivation that are often incorporated in a capabilities approach to poverty (Sen 1999) are addressed through the supply constraints that inhibit agents at the microeconomic level. Although the final objective focuses on alleviating income poverty, the study also addresses non-monetary constraints that inhibit agents from participating in a economic and social
12
The macroeconomic of poverty reduction in Cambodia Preface
development and thus indirectly incorporates a wider concept of deprivation. Second, governance or, more accurately, political economy plays an important role within the Cambodian context, and influences economic and social policies and their implementation. It is important to appreciate that Cambodia has only recently emerged from a civil war that destroyed much of the human and physical infrastructure in the economy. The peace agreement of 1991, the 1993 election, renewed instability in 1996-97, and the election of 1998 mark the process of re-establishing central control over the entire country and have heralded an as yet short-lived period of relative stability. Many aspects of the economy are shaped by this legacy. The current structure of the government incorporates an expanded civil service to include members from all sides of the political divisions in the decision-making process of the country, as part of the 'peace dividend'. This has contributed to institutional flaws (from an economic viewpoint), which often render policy outcomes ineffective. However, most would claim that the benefits of political and social stability far out-weigh any negative implications of such a structure. Reform of the civil service and other key institutions is currently underway, but the full impact is unlikely to appear for several years. However, the third national election in 2003, which should have consolidated the growing confidence in the present system and the considerable progress towards reform, resulted in political paralysis. At the time of writing this final version of our report, no government had been formed. Ministers were in caretaker positions, indeed some were boycotting meetings of the Council of Ministers, with the result that crucial decisions for poverty alleviation could not be taken.
The consequences of these recent events should be borne in mind when understanding the policies in place. Addressing widespread poverty, such as is experienced in Cambodia, requires long-term coherent strategies that coordinate and mobilize all appropriate resources. Long-term strategies are emerging but are not yet in-place. The establishment of UNTAC marked the beginning of a large presence of external assistance to Cambodia. While it brought much needed financial assistance, the shaping of policy has been influenced to a large extent by outside guidance. Although donors have, in recent years, focused much of their effort on poverty reduction and encouraged the Cambodian government to develop strategic programs, we argue that the link between the macro strategy and the micro or sectoral strategies has yet to be made. The result is that, despite the government's identification of priority areas for development, the approach remains partial: some sectors suffer from neglect while those that have received extensive attention seem unlikely to be fully effective. In carrying out our research for this report we received considerable assistance from the UNDP country office, especially from Ingrid Cyimana and Richard Schiere. Their logistical support as well as their intellectual input to the project are much appreciated. However, we would like to emphasise that the views expressed in this report do not necessarily reflect those of the Cambodia country office. We would particularly like to thank Dr Hang Chuon Naron, Deputy Secretary General of the Ministry of Economy and Finance, our partner organization in Cambodia, for his support and guidance. Others, too numerous to name individually, we also thank for providing us with documents and various other forms of assistance.
The macroeconomic of poverty reduction in Cambodia Preface
13
Summary of the Report

1. Overview of the Poverty Situation
Despite the absence of reliable data, there are signs that economic growth during the past decade in Cambodia has not produced any significant poverty reduction. Indeed, there are some signs that the situation is worsening - reduced per capita consumption (measured in riel); fewer public health facilities and rising infant and child mortality; poor education outcomes; increasing population pressure on cultivable land with rising rural underemployment due to labour-force entry of baby boomers and lack of growth in non-farm employment. Increased vulnerability of some, resulting in part from Cambodia's war legacy and the threat posed by the HIV/AIDS epidemic, are also important factors in poverty in Cambodia. Inequality is also very high, particularly along urban-rural lines. In fact most of the economic growth of the past decade has been confined to urban enclaves, while rural growth, especially in staple food production, has barely kept pace with the increase in population. The problem of tackling poverty and inequality is thus largely one of addressing these inequalities between city and countryside by generating more broad-based growth. This is now possible, given the increasing stability of the countryside and successful macroeconomic stabilisation. The depth of poverty is quite shallow, with the exception of certain especially vulnerable groups, and inequality is lower in the rural areas. Thus a poverty reduction strategy which aims at achieving more egalitarian growth by broadly targeting development of the rural areas, is likely to have a significant and widespread impact. Such a strategy will be based, first,
14
The macroeconomic of poverty reduction in Cambodia Summary of the report
on increasing the dynamism of agriculture and, second, on generating non-farm employment, particularly in the rural areas if excessive stress on urban infrastructure is to be avoided. Both supply-side and demand-side measures will be needed. On the one hand, there is a clear need for demand expansion to promote rural prosperity and to integrate more of the rural population into the market economy. On the other hand, there is a need for great improvements in the supply of human and physical assets (both public and private) in order to improve the capacity of the vast majority of Cambodians to take advantage of opportunities which may arise. Specific actions on the supply side, in education, health, provision of infrastructure and finance, etc., properly belong to the realm of sectoral interventions and largely fall outside the scope of this study. Here, we are concerned with the impact that macroeconomic policy can have on the ability to achieve such sectoral (and more micro-level) changes.
In the last decade, while it has succeeded in producing fairly consistent economic growth, the macroeconomic framework established under guidance from the IMF has not successfully tackled the key elements of poverty in Cambodia. To be sure, it is difficult to disentangle the impacts of macroeconomic policy from those of the political instability that affected the economy before 1999. The sharp fall in investment and output in the mid-1990s, for example, is more likely attributable to political events than to macro policy as such. However, even during high growth periods, poverty reduction either did not occur or was minimal. The question that we wish to pose in this report, therefore, is what changes could be made to the macropolicy settings that would be conducive to a more rapid reduction in poverty and improved opportunities for the vast majority of Cambodians to participate in the benefits of this growth? Growth in recent years has been very strongly influenced by expansion of the export-oriented garment industry. There are reasons to suppose that the rapid expansion of this sector is unlikely to continue. It is imperative, particularly in the context of the current uncertainty in the global environment to find ways of diversifying the sources of economic growth (including export growth) in order to absorb more of the excess capacity which presently exists. Land, labour and finance can all be used more effectively than they are now. Up to now, the policy framework has served the country well, in the sense that the focus on tight monetary and fiscal policy has
2. The Macroeconomic Framework

In the developing world, Cambodia is special on account of the historical legacy of genocide. This affected its progress towards poverty reduction through destruction of social support networks and shortages of skilled and intellectual human resources. During the 1980s, the country remained isolated from all but Soviet and Vietnamese assistance and, while formal peace was established under UN auspices in 1991, real stability was not achieved until late in the decade.
15
helped to eliminate excessive rates of inflation and generate some confidence in the authorities. Dollarization has also assisted in this process. In more recent years, however, there has been deflation and in the context of global recession, the stabilization package would appear to have outlived its usefulness. Even with global recovery in 2002, inflation has remained too low. At the meso-level, sectoral policies implemented under donor guidance have largely addressed the problem of human capital, particularly the health and education sectors. These policies do go some way towards addressing poverty reduction, in the sense that they are generally gender sensitive although the education policy is unlikely to assist the large proportion of older women in the workforce - and will be helpful in overcoming the extremely damaging legacy from the years of war and genocide. They will, however, have a very limited impact on income poverty unless there is a concurrent expansion of investment and employment opportunities for the newly trained and healthy workforce to take up. A policy of selective and carefully targeted demand expansion seems to us to be a good solution to this problem. We note especially that there is considerable underutilized capacity in Cambodia's rural areas. There are no simple solutions, however. In particular, there are legal and institutional obstacles to a demand expansion policy, partly due to the historical legacy and partly due to the policy framework established under the IMF structural adjustment program. Many of these policy-driven obstacles are well founded, especially as confidence in the authorities and in the financial sector is
only gradually improving. They should not be undone in haste, but they should be undone over time.
3. Monetary and Exchange Rate Policy

Low public confidence, weak financial markets, lack of transmission mechanisms, and a high degree of dollarization of the economy create obstacles to the effective conduct of monetary and exchange rate policy. Instruments such as repurchase agreement and statutory reserve requirement have proved to be inapplicable in managing money supply and promoting growth. A managed floating exchange rate system supported by prudent foreign exchange interventions, on the other hand, has been found to be more effective in stabilizing the movement of the exchange rate, while also taking into account the target set for international reserves. Under the IMF's Poverty Reduction Growth Facility, tight monetary and fiscal policy has been pursued. The restrictions imposed on fiscal deficit financing by the banking system, especially the central bank, coupled with the ceiling on net domestic assets of the NBC and a floor on its net international reserves have been successful in promoting macroeconomic stability. The mix of the two policies has provided stability in the overall price level and the purchasing power of the riel, while contributing to GDP growth of about 6-7% per annum. This achievement has to be credited also to dollarization, which emerged as a consequence of low confidence and a persistent feeling of uncertainty.
16
Dollarization has provided both benefits and drawbacks for economic development and poverty reduction. The benefits of dollarization include switching public practice from gold to banknotes denominated in US dollars thereby promoting monetization of the economy, preventing capital flight, promoting financial deepening, lowering the risk of currency devaluation, and facilitating Cambodia's international integration process. All those issues facilitate and foster economic growth and have thus an indirect impact on poverty. On the negative side, dollarization means undermining the effective conduct of monetary and fiscal policy, damaging the pride of the nation, reducing the opportunity to earn income from seigniorage, and widening the gap between the urban dollar-based economy and poor rural areas where riel are normally used. The recent increase in circulation of riel, without pressuring macroeconomic stability, seems to reflect efforts of the government and NBC in encouraging the use of the riel and indicates that there is no compelling reason to go to full dollarization. Although the recent low or negative inflation represents a remarkable achievement of the existing policy settings, such settings no longer seem appropriate to address the issue of poverty reduction. To be more pro-poor, monetary and exchange rate policy and also fiscal policy need to be more pro-active and there is a strong case to review them in order to move faster to achieving the government's ultimate goal. In other words, stimulating more broad-
based growth, ensuring more efficient and development-oriented resource mobilization, and restoring monetary sovereignty to form a broader basis for poverty reduction should now be the focus of monetary and fiscal arrangements. In this context, some policy options have been suggested, including in particular the following: i. Depart from tight policy stance and allow some inflation to 'grease' the economy; ii. Establish an appropriate balance between accumulation of international reserves and the level of net domestic assets, which should stimulate domestic demand. iii. Address distortions to financial intermediation by undertaking a systematic and properly sequenced banking and financial reform; iv. Adopt some immediate mechanism to bring banks and potential savers more closely together, especially the introduction of a public education campaign to provide the general public with knowledge on banking services and to reduce uncertainty. v. Undertake a stepwise approach to de-dollarizing the economy by avoiding a forced conversion, promoting the money function of the riel and providing rieldenominated assets, especially treasury bills. In sum, from the point of view of poverty reduction, monetary and exchange rate policies need to support increasing economic integration of the rural areas
17
into the national market, in particular, by supporting higher growth that is more equitably distributed - which means more employment opportunities, increased banking intermediation, and a reduced gap between rich and poor.
the private sector in terms of contributing to GDP growth, the tighter will be the fiscal constraint and the lower the chances of redressing low private sector growth through fiscal stimulation or, indeed, direct public investment. ii. Given the resource envelope constraint, Cambodia's pro-poor macro-fiscal strategy seeks to secure growth through a highly specific expenditure switching policy from general services to social services, i.e. health and education. On the face of it, these sectoral expenditures have been designed to optimise their pro-poor impact, and most official documents highlight these benefits. They do not, however, highlight the implicit and explicit costs which need to be taken into account when assessing the net pro-poor benefits of the chosen strategy. It is imperative to recognise that increased investment in social services in Cambodia has come at the expense of low investments in agriculture and rural development. While the precise opportunity cost of this trade-off can and should be debated, there is no doubt that lack of investment in agriculture and rural development in a country like Cambodia with significant rural poverty cannot be held to be optimal from the point of view of alleviating income poverty. Again, this strategic choice is based on the assumption that private sector growth will be the engine of overall economic growth. If private sector growth is lower than anticipated then there is no cushion left for boosting growth and alleviating income poverty through investment in economic services.
4. Fiscal Policy
In relation to fiscal policy we make four major points: i. Cambodia's resource envelope is limited in the medium term. Only modest increases in public resources can be expected through growth in the tax base and efficiency improvements in the medium term. Domestic resource mobilization through domestic borrowing is also not feasible in the medium term due to IMF restrictions, but more importantly due to low confidence in the authorities. Cambodia is heavily dependent on concessional donor finance for its development investments. This raises concerns regarding future debt service ratios. The present and projected grant: loan portfolio make it imprudent for government to significantly seek additional development assistance without reckoning on a concomitant rise in its future debt service obligations. Any future adverse change in the grant: loan ratio would exacerbate this fiscal constraint. This limitation on Cambodia's resource envelope means that Cambodia runs the risk of falling into a severe low level equilibrium trap with negative consequences for poverty reduction, if the chief growth engine in Cambodia's future economy private sector development - fails. The greater the shortfall in expectations from
18
iii. Downsizing the civil service will provide little payback from a fiscal perspective, either by releasing resources to pay civil servants better to reduce corruption and increase efficiency or to reallocate these resources away from general services to social and economic services. Further, a large sum of donor assistance in fact serves as indirect budget support by significantly increasing the average remuneration of skilled Cambodian civil servants This indicates that the principal incentive to work in public employment is the prospect of access to external salary supplements. Any analysis that relies on civil service downsizing either to improve the efficiency of public service delivery or to release resources for pro-poor expenditures would therefore be misplaced. From a fiscal perspective much of the emphasis on civil service reform is therefore not relevant. What needs to be tackled urgently is the incentive and capacity distortions caused by donor supplementary remuneration to skilled civil servants. iv. The crowding out effect of aid on government's revenue efforts appears to be limited in Cambodia. However, generous concessional terms for investment resources from donors can act as a disincentive when RGC considers measures to generate domestic resources. The existence of excess liquidity in the domestic banking system negates the possibility of crowding out and offers RGC potentially attractive finance for development. The incentive for RGC to access such finance is, however, limited given the potential availability of easy
concessional money from donors who accept RGC's bounded fiscal sovereignty (albeit with conditionalities) unlike domestic economic agents who, without recourse to conditionalities, are reluctant to do the same. In such a situation there is little incentive for either RGC or the donors to speed up the financial reform process that would enhance RGC's credibility and remove the bounds on its financial sovereignty. It is this sense that there exists a Dutch disease in Cambodia's fisc. The major policy implication here is that while donors can live with RGCs bounded fiscal sovereignty since they exercise countervailing power through conditionality-based safeguards, private economic agents cannot since they have no countervailing safeguards. This leads to sub-optimal utilisation of investible resources in Cambodia and makes the monotonic emphasis on private sector growth as the engine of Cambodia's development less credible. This problem has political, institutional and technical dimensions all of which need to be researched in depth and urgent policy measures taken v. The key to unlocking the constraint to pro-poor fiscal policy in Cambodia is not directly macroeconomic in nature. It has to do with the poor disbursement of resources allocated for pro-poor expenditures that result in weak budget implementation. There are political, institutional and economic reasons for this phenomenon. Donor designed technical assistance to resolve this issue (even in their chosen priority sectors) has been piecemeal in approach and conditionality driven. There is as yet no comprehensive appraisal of the problem.
19
There is need for a comprehensive and frank policy assessment of the reasons underlying poor disbursement, followed by the design of a comprehensive, sustainable and practical disbursement mechanism that would overcome this problem. If such a mechanism were made operational successfully, then this would increase the efficacy of pro-poor expenditures, improve confidence in RGC's fiscal sovereignty and, therefore, alleviate several fiscal constraints. There are currently five major policy documents outlining a pro-poor strategy for growth: the RGC's National Poverty Reduction Strategy (NPRS 2002), the World Bank Integrated Fiduciary Assessment and Public Expenditure Review (IFAPER 2003), the IMF's 2002 PRGF and RGC's SocioEconomic Development Plan II (SEDP-ii 2002) and MEF's Draft Medium Term Expenditure Framework (D-MTEF 2002, 2003). None of these documents contains either a clear enunciation of macro-fiscal policy instruments that explicitly relate fiscal spending to poverty reduction, or an assessment of the impact of macro-fiscal policy changes on poverty alleviation. Most of these strategic macroeconomic documents in Cambodia, seek to enhance fiscal performance by enhancing the efficiency of revenue and expenditure policy implementation and the macro-fiscal strategy deals almost exclusively with this issue. They address the issue of weak governance and improvement in the efficiency, governance and effectiveness of public revenue and expenditure performance. The SEDP-ii document, apart from presenting projections, contains no further macrofiscal analysis of strategic information.
Thus, the major pro-poor content of the existing macro-fiscal strategy lies in its emphasis on the 'priority sectors'. Between them, they present cogent, costed arguments outlining the pro-poor benefits of spending on the social service priority sectors, namely, education and health. These arguments are backed by detailed health and education medium-term sector spending strategies in the D-MTEFs. With agriculture and rural development however, they are much weaker on quantifiable argument. There is no costed quantitative strategy that provides a link between outcome indicators in the rural sector and poverty alleviation targets. Rural development receives short shrift, with a statement of three pious intentions: (a) to promote decentralization, (b) to facilitate an integrated participatory rural development approach and (c) to develop a rural infrastructure strategy. There is no detailed medium-term expenditure plan for these sectors in the D-MTEFs. It may appear from the above discussion that RGC is more serious about prioritising government spending on social services and the mention of agriculture and rural development is gestural. This would be disastrous in an economy where the bulk of the poor live in rural areas and work in agriculture and agriculture related activities. Fortunately this is not the case. Our investigations revealed ample evidence that political authorities recognise the critical importance of development of agriculture and rural development as a priority goal for Cambodia and the political payback from providing such services particularly rural roads, rural employment, enhanced agricultural productivity and access to common and private agricultural
20
inputs (water, fertilizer, seeds, etc.) - is acknowledged to be significant in government circles. Why, then, this Cinderella treatment at the strategic level? The answer lies not in an examination of government priorities but rather donor priorities. It is in the nature of advanced public finance processes, such as that sought to be instituted by the D-MTEFs, that resource allocations to particular sectors depend significantly on the effectiveness with which the ministries and agencies responsible for formulating sector strategies make their case to government. In Cambodia, the four priority sectors compete intensively for the extremely limited recurrent and capital funds available. A major determinant of access to these resources is the effectiveness with which the case for using the funds is made. The Cinderella treatment that agriculture and rural development (not to mention other economic services) have received in this respect is an outcome of the fact that technical assistance for strategic preparation of sector priorities in these areas has been much less than that received by the health and education sectors. Thus the choice of social sector spending as the spearhead of pro-poor development can be viewed as a consequence of asymmetry in access to donor technical assistance rather than strategic selection. Urgent attention therefore needs to be paid to ways in which this asymmetry can be addressed. As a consequence Cambodia lacks policy capacity to assess the pro-poor impact of fiscal policy changes. This reduces national ownership of macroeconomic strategy, discourages participation, and inhibits the acknowledgement and redressing of political
issues that retard fiscal performance, with negative attendant macroeconomic implications. Donors and multilateral organisations that claim to have poverty reduction as their paramount objective therefore need to pay more attention to this issue when providing technical assistance. While it requires more attention to political economy issues, the payback in terms of effectiveness would be sizeable. Well designed technical assistance for policy building can enhance gender equity, and compensate for the lack of attention to income poverty.
5. Trade Liberalization
Trade liberalization, combined with dollarization, since 1993 has resulted in significant expansion of exports and narrowing of the current account deficit. Currently, however, the benefits derived from export growth are too narrowly based to produce any major impact on poverty reduction. The main beneficiaries of trade liberalization have been those directly involved in the activities, which are largely in the urban areas, and the limited auxiliary service activities that have developed around these activities. Even in the provinces, due to the lack of infrastructure there has been limited redistribution from urban to rural areas. By far the most important export commodities have come from the garment industry, although tourism has also shown recent rapid growth. However, these have remained restricted to urban enclaves and, due to high import content, they have contributed relatively little value added. The absence of backward and forward linkages has also
21
meant that employment generation as a result of rapid export expansion has been minimal. To enable equitable distribution of the benefits of trade, there need to be opportunities for the rural population to participate in trade-oriented activities. There are currently many obstacles preventing such participation. Although current policy statements emphasize pro-poor trade, actual implementation does not appear to address these intentions. The delay in addressing these constraints will further increase the already existing inequities in access to resources and work against poverty reduction. Trade liberalization in Cambodia will not contribute to poverty reduction if other macroeconomic policies are not supportive. In reality, the impact so far has been rather narrow, in terms of its impact on employment, income distribution and contribution to GDP (in Keynesian accounting terms, X-M). As well as the narrow range of products being exported, the high import content needs to be addressed since it restricts the benefits to the local economy that come with trade liberalization. Trade liberalization will benefit the poor if policy is structured to develop local markets that will then eventually have the capacity to compete in international markets. The division of labour, domestically as well as internationally, should benefit from increasing the extent of the domestic market. Addressing the immediate constraints in the country in the areas of; governance, infrastructure and poor human capital, will go further in addressing pro-poor trade growth since it will expand the number of products that can be exported and widen the proportion
of the population that can benefit from trade liberalization, and thus establish a conduit for the re-distribution of growth.
6. Financial Liberalization
Following financial liberalization, market determination of interest rates should result in modestly positive real interest rates which should, in theory, increase the resources available to the financial system. However, this effect has only applied in Cambodia to the dollarized urban economy. The perceived high cost and high risk of lending to the rural areas has resulted in most banks refusing to accept riel deposits, with the effect that formal financial intermediation is simply not available to the poor. Re-regulation of the sector, after the initial chaos in the wake of dismantling the monobank system, has succeeded in improving the performance and to some extent confidence of the public in the banking sector. Again, however, the impact has been restricted to the urban economy. Despite the high growth of GDP and the reduced number of banks after restructuring, this market fragmentation has resulted in inability to achieve financial deepening and in excess liquidity of the banking system. The financial system that does exist in rural areas is therefore based on NGOs, with limited coverage, and the informal sector. Micro-finance is based on the recognition that small business and employment generation, not least engaging women in more profitable economic activities, would be encouraged by the availability of credit.
22
Despite substantial efforts and the best of intentions, the results, however, have been below expectations. It seems likely that the introduction of dedicated financial institutions in the rural areas would help promote a more efficient and stable rural financial market than the current system, based as it is on NGOs whose primary efforts are in projects with different objectives. However, expanded formal intermediation is unlikely to happen in the absence of a more conducive macro and sectoral environment which improves the profitability of rural activities, reduces risk for clients, improves information flows and reduces legal impediments. An improved economic environment is also likely to promote the forging of better links between formal and informal financial institutions and the diffusion of other financial instruments such as deposit accounts in riel or crop insurance. Financial liberalization in the absence of a good legal and institutional framework is indeed dangerous, as the recent history of Indonesia demonstrates (although banking crises have also occurred in Cambodia itself in the past decade). However, reforms aimed at upgrading institutions alone are unlikely to be very successful, particularly in terms of poverty reduction. A case in point is the question of land titling, the absence of which is, for many commentators, a major obstacle to provision of medium and long-term credit. At present, only 10 percent of rural families hold legal land titles. Nevertheless, titling is not a sufficient condition to improve access to credit: factors such as larger farm size, higher levels of human capital, and proximity to
major consumer markets appear to be more important in determining access to formal credit. Other reforms, for example, in the legal environment of the financial system including prudential supervision, registration, licensing, standardised accounting practices, transparency, a secured transaction framework and enforcement provisions - can reduce the risk and cost of rural financial intermediation and are essential to the process or restoring confidence in the banking system. But they are more likely to benefit the non-poor, who already participate in the formal credit system, in the short to medium term. The vast majority of the poor will continue to rely heavily on informal sector loans because of their greater flexibility. As noted in this chapter, the vulnerability of the poor to exogenous shocks is itself a major obstacle to the extension of formal financial services to them. Institutional reforms need to be accompanied by policies directly aimed at generating higher growth and employment in the rural economy. These include expenditure on education and health, which are already receiving priority from donors, and on rural infrastructure, intensification of agricultural production and extension of economic services more generally to the countryside. Increasing rural incomes and productivity will both lower the marketing risk of financial institutions and their intermediation costs. It will provide an enabling environment in which the current fragmentation of rural markets, which in many ways contributes to the fragmentation, petty scale and high costs of financial markets, can be eliminated.
23
7. Privatization
On the whole it would seem that privatization has tended to encourage economic growth in Cambodia, but it has not served the objective of poverty alleviation. Privatization in the banking sector has, if anything, reduced the access of the rural population to the formal financial sector while it has enhanced growth in the urban areas. For other sectors there is less information available, but the record of a lagging rural economy would suggest much the same pattern. Rubber plantations, which have yet to be privatized, offer an opportunity for increased investment and employment in the rural economy as well as a means of improving export diversification, but the private sector, with good reason, has shown a reluctance to take on the task. In other areas, the privatization of forest and water resources has proved controversial due to competing claims and has largely failed to provide the intended sources of commercial development and government revenue. At the same time, it has restricted access of the poor to resources upon which they have previously relied. Privatization of remaining public assets does not appear to be high on the government's
agenda at the present time and there would seem to be good reasons to reconsider the policy with regard to these assets. Public ownership of some assets creates opportunities to pursue developmental objectives, provided that, in the absence of market discipline, government is also able to impose discipline on the firms. The enterprises might eventually be privatized, but their long-term profitability must also be assured by complementary macro-fiscal policies which are aimed at growing the rural economy as a whole. Finally, the case of forestry, agricultural and fishery concessions involves political economy issues which have no simple technical solution (such as land titling). The recent resumption of some concession areas by the state is a positive step forward in this regard, as it increases the access of the poor to resources. While the granting of concessions could indeed be a positive step towards development of commercial activities in the countryside, policy needs to proceed with far greater caution in future. Particularly if political stability is to be preserved, it needs to take into account local land tenure systems, which may conflict with conceptions of private property being promoted by a number of donor agencies.
24
1
26
Dimension of the Problem:

Overview of Poverty and Inequality in Cambodia1
1. INTRODUCTION
Poverty in Cambodia is widespread and severe. The importance of the problem is fully recognized by the Royal Government of Cambodia and poverty reduction was incorporated as a major element of Cambodia's Socio-economic Development Plans for 1996-2001 and 2001-2005. Recently, the government has also produced, in collaboration with the World Bank and other donors, a Poverty Reduction Strategy Paper which seeks 'to more clearly define and operationalise the Government's approach to poverty reduction' (NPRS 2003: 6). However, we will argue that Cambodia does not yet have a comprehensive poverty reduction strategy. As a background to the main part of the report, this chapter surveys issues of poverty and inequality in Cambodia, focusing on: 1. Changes in incidence and depth of income poverty based on a fixed poverty line.
1 This chapter comprises a cut down and revised version of a more detailed paper prepared by Dr Sau Sisovanna. Persons interested in reading the full version may obtain it from him at sssovanna@hotmail.com. The macroeconomic of poverty reduction in Cambodia The dimension of the problem: an overview of poverty and inequality
2. The level and trends of key human poverty indicators. 3. Levels and changes in inequality based on expenditure, functional distribution and spatial distribution 4. Level and changes in the distribution of productive assets. 5. The challenge posed by Cambodia's special demographic characteristics. 6. The centrality of agricultural and rural development. The chapter is thus intended as a poverty mapping exercise, which attempts to identify some of the patterns and trends in poverty and inequality over the past decade. It may be read in conjunction with the main part of the report, but we do not attempt at this stage to draw any conclusions about which policies or other factors may be responsible for these patterns and trends.
Unlike most of its neighbouring countries, Cambodia has not yet established an official poverty line. However, a baseline poverty profile has been compiled to provide estimates of poverty in Cambodia, which are comparable to those developed recently by the World Bank for a number of other Southeast Asian countries (Prescott & Pradhan 1997). Per capita consumption was calculated, on the somewhat unrealistic assumption that each household member receives an equal share of the household's total consumption, and these estimates were compared to two alternative poverty lines - a food poverty line and an overall poverty line - to determine an individual's poverty status. The food poverty line was calculated using data from the CSES 1993-94 and updated with data from the subsequent surveys. Some key points to bear in mind are as follows: 1. Food poverty is measured in terms of calories. The benchmark adopted is a 2,100 calories minimum energy requirement per person per day (MOP/UNDP 2002). 2. A reference food basket was constructed by taking average values of the reported quantities consumed of each food item by the 3rd quintile of the population, in which 69% of the calories obtained are from cereals, especially rice. The reference food basket derived in this way actually corresponds to a calorie content of 2,298. Thus, in order to identify the proper composition of the minimum food basket, quantities of all items are is proportionally scaled down (MOP/UNDP 2002).
2. CHANGES IN INCIDENCE AND DEPTH OF INCOME POVERTY

Standardized quantitative measures provide a starting point for understanding poverty. In order to identify the proportion of the population that should be counted as income poor, their consumption expenditure is compared to the value computed and defined for the poverty line. Consumption expenditure has been used instead of income data since not only do they tend to be more reliable than reported income data, but income data was not in fact collected by the Cambodia Socio-Economic Survey (CSES) before 1999.
The macroeconomic of poverty reduction in Cambodia The dimension of the problem: an overview of poverty and inequality
27
3. The reference basket was valued using market prices in Phnom Penh, other urban areas, and rural areas. Separate price estimates were obtained for each of the three strata, leading to three different food poverty lines. A non-food minimum allowance was then calculated and added to the food poverty line to provide the total poverty line. The minimal allowance for non-food goods was based on the typical non-food spending of those who can just afford the reference food requirement. In practice, it is not easy to calculate the cost of this minimum non-food allowance, and consequently the overall poverty line, because no unit value for non-food items were recorded in any of the CSES surveys. Further, CSES 1997 used a different method for calculating the cost of this allowance from that employed in the 1993-94 and 1999 surveys and there were also differences in territorial coverage in the three surveys and there were substantial unexplained differences in the level of per capita household consumption between the 1999 survey's two rounds (NPRS 2002: 32). Estimates of poverty for Cambodia over the three CSES periods are therefore difficult to interpret in any meaningful comparative way. The most basic measure, the head-count index (HCI) provides an initial view of the prevalence of poverty. The proportion of the Cambodian population which is defined as poor according to this analysis is shown in Table 1.1. In 1999 two rounds of the survey were undertaken because the first round results were so dramatically different from
those in preceding years. The table thus highlights the unreliability of the data. As a result, any changes over time in the level of absolute poverty could be seen as a statistical error2. Nevertheless, the data suggest that there has probably been zero or minimal reduction in absolute poverty between surveys. Broadly speaking, probably between 40 and 45 percent of the Cambodian population subsist below the poverty line and there is no evidence that this situation is improving (MOP/UNDP 2002). Moreover, in real terms, average per capita consumption by all strata of the population fell between the 1993-94 and 1999 surveys, from 2,260 riels per day to 1,800 (if calculated in dollar terms this represents a much more substantial fall). This is a striking result, which should form the starting point for reflection on the macroeconomics of poverty reduction. (MOP/UNDP 2002: 29) The poverty gap index, expressing distance below the poverty line, indicates that the consumption by poor people is concentrated at about 15% below the poverty line (NPRS 2002: 33). However, the gap is significantly higher in rural areas and provincial urban centres than in Phnom Penh.
2 See WFP 2001for a broader discussion of CSES 1999.
28
Table 1.1: Cambodian head-count indices (per cent of population) CSES 1993-94 National Phnom Penh Other Urban Rural % under food poverty line
Source: MOP/UNDP 2002.
CSES 1997 (adjusted) 36.1 (47.8 unadjusted) 11.1 29.9 40.1 17.9
CSES 1999 (round 1) 51.1 64.4 19.4 57.3 70.0 47.7
CSES 1999 (round 2) 35.9 9.7 25.2 40.1 11.5
39.0 11.4 36.6 43.1 20.0
Indeed, in the rural areas poverty is almost four times higher than in Phnom Penh, and significantly higher than in other urban areas The headcount poverty rate is highest for those engaged in agriculture and they comprise the overwhelming majority of the headcount poor. (MOP 1999: 51). The use of the food poverty line yields considerably lower levels of poverty, but it does not alter the conclusion that poverty is the highest in rural areas and the lowest in Phnom Penh. Life cycle factors are clearly associated with poverty in Cambodia, especially given the baby boom that followed the Khmer Rouge period. In the poorest per capita consumption quintile the average household size is 5.8 people compared to 4.2 people in the richest quintile. Poor households tend to have more children, elderly and disabled persons as well as younger heads than do richer households. They are, however, less likely to be headed by a woman, which may be explained by the fact that the average age of women household heads is over 39 and such households are also smaller. It seems probable that this factor is also related to life-cycle factors - the older women
are more likely to head households with childless adult labour and few dependants (MOP 1999). In the case of extreme poverty, there is definite evidence of the effect of landmines and war. Just over 30% of people living in households headed by a person with a war- or mine-related disability are food poor, three times the rate for households whose head is not disabled. While this group represents a small proportion of those below the poverty line, their average distance below the food poverty line is 27%. They are therefore a particularly vulnerable group. (MOP 1999). Although the available analyses of CSES data do not distinguish ethnicity as a variable, we should also note here that poverty is a major problem for upland ethnic minority groups such as the Tumpuon, Phnong, etc. Ethnicity becomes a critical issue given the possibilities for social tensions and the fact that there are likely to be significant agricultural development prospects in the upland areas. Most ethnic minorities lack access to all-weather roads, assistance in controlling flash flooding, means for dry season management of scarce water resources, electricity supply and public health facilities.
29
Besides the fact that certain groups are particularly vulnerable to poverty on account of location (in the rural/upland areas), life-cycle factors, war-related disability and ethnicity, it is important to note that the distribution above and below the poverty line is highly sensitive to changes in the poverty line itself. In reality, most of the population is clustered around the poverty line. This result has two important implications. First, a small change in the poverty line will produce large changes in the HCI. Given that this poverty line is rather low, the meaning is that most Cambodians, especially those outside Phnom Penh, are poor. Second, the shallowness of the PGI plus the fact that most 'non-poor' are close to the poverty line, indicates that broad-based strategies aimed at the rural sector are likely to have significant impacts on poverty reduction3.
Cambodia's HDI, as a technical change occurred in the computation algorithm of UNDP in 1999, making the time series too short to draw meaningful conclusions.
3.1 Basic Health Indicators

The CSES 1999 data showed that about 12% of the sample villages had khum clinics. As many as 70% and 27% of sample wards in Phnom Penh and other urban areas had the services of a medical doctor in the locality itself, but in the rural sector this percentage dropped to 4%. The services of a nurse were available in 22% of the sample villages in the rural sector and those of a trained midwife in 23%. Moreover, the number of public medical establishments at the commune level has declined steadily since 1989, to 68% of its previous level by 2000 (NIS 2002a). This decline has been at least partly compensated for by the rise in mostly unlicensed private establishments, though it is not known to what extent these reach rural populations. Thus, physical access to health services appears to be an important and growing problem in Cambodia. 3.1.1 Infant and Child Mortality Table 1.2 presents neonatal, post neonatal, infant, child, and under-five mortality rates for several five-year periods prior to the survey carried out in 2000. Figures for mortality for the period 20-24 years before the survey (1975-80) give a sense of the impact of the Khmer Rouge regime on levels of mortality among even the youngest members of Cambodian society.
3. LEVEL AND TRENDS IN KEYS HUMAN POVERTY INDICATORS

According to the Human Development Report for 2001 (UNDP 2003), the HDI for Cambodia is 0.556, which ranks the country at 130 out of the 175 countries. The country exhibits one of the poorest human development performances of its region. Replicating the pattern for income poverty, the HDI score of urban Cambodia in 1999, at 0.604, was well above that for rural Cambodia (0.500). The richest 20% of Cambodians had an HDI score of 0.623, which was 40% above that of the poorest 20% (0.445) (MOP/UNDP 2000). It is not possible to look for a temporal trend in
3 Nevertheless, many components of the HDI, such as life expectancy and literacy, do not change much from year to year. As such, the HDI is a relatively stable indicator that changes rather slowly over time.
30
Neonatal mortality in the most recent period is 37 per 1,000 live births. However, the risk of dying for any Cambodian child who survived the first month of life increases during the next 11 months. Thus, almost one in every ten babies born in Cambodia does not survive to his or her first birthday. These figures earn Cambodia the unenviable position as the country with the highest infant mortality in East and Southeast Asia. Extremely high mortality was registered for the Khmer Rouge years. It seems inevitable that afterwards, there would be a decrease in mortality, as is the case in the period 10-19 years before the survey4. However, for most recent decade the survey registers an increase in mortality, both for infants and under-fives. This is another indication that poverty may not have decreased over the past decade.
Table 1.2: Early childhood mortality rates Years preceding the survey 0-4 5-9 10-14 15-19 20-24
Note:
Neonatal mortality 37.3 40.5 44.2 48.1 53.7
Post-neonatal mortality 57.8 50.4 34.5 36.4 75.0
Infant mortality 95.0 91.0 78.8 84.6 128.7
Child mortality 32.5 31.3 39.1 53.3 129.7
Under-five mortality 124.4 119.4 114.7 133.4 241.7
Childhood mortality rates are expressed as numbers per thousand and are defined as follows: Neonatal mortality: the probability of dying within the first month of life. Post-neonatal mortality: the probability of dying between the first month of life and first birthday. Infant mortality: the probability of dying between birth and the first birthday. Child mortality: the probability of dying between ages one and five. Under-five mortality: the probability of dying between birth and the fifth birthday. Source: NIS 2001.
3.1.2 Drinking Water and Sanitation A large proportion of the diseases in Cambodia are water-borne and caused by lack of access to safe drinking water and sanitation. Piped water (to the dwelling) or water from public taps is a luxury for a very small, mainly urban, segment of the population. Urban residents also have access to water vendors. The CSES 1996 data showed that about two-thirds of the country's population obtained water from
unprotected wells or from ponds, rivers and streams. Water from these sources is often contaminated. Cambodia Demographic and Household Survey (CDHS) 2000 data showed that half of all households were still using these two sources and, while the proportion with access to tube wells had increased slightly, rural households were far less likely to have access to safe water sources (NIS 2001: 20).
4 Figures may, however, be compromised for this period by lack of complete coverage.
31
Only 18% of the population use adequate sanitation facilities (MOP/UNDP 2002). Data from the CDHS (NIS 2001: 20) showed that 86% of the rural population and 40% of the urban population had no toilet facilities at all. While 13% had a flush toilet (over half were unconnected to either a sewer or septic tank), this proportion rose to 51% in the urban areas. Other types of toilet facilities are rare. As a result, people tend either have a flush toilet or no toilet of any type. 3.1.3 Nutrition Rice is the staple food of Cambodians. In 1995 Cambodia achieved its first (modest) surplus since the eve of war in the 1960s, and surpluses have been recorded in every subsequent year. However there are many Cambodians who lack access to this basic necessity due to insufficient purchasing power (the price is influenced by export markets in Thailand and Vietnam) and poor transport and marketing systems. Also, the agricultural production system remains highly vulnerable to natural disasters and pest damage, leading to large fluctuations in yields. Despite the current surplus, nearly half of the 24 provinces are food deficit areas and a significant proportion of the population is unable to meet minimum rice requirements, the critical periods generally being mid-July to mid-October (pre-harvest) (RGC 2002a). In addition, consumption of rice alone is insufficient to meet dietary needs. For much of the population, rice accounts for over 70% of calorie intake (RGC 2002a). In Prey Veng, Kampong Speu, and Kandal provinces, rice constitutes over 80% of the
calorie intake (RGC 2002a). The non-poor gain more of their calorie intake from the more expensive, non-rice foods. It has been estimated that if protein deficiency alone were reduced by half, 20,000 lives could be saved each year (MOP/UNDP 1997). The predominance of rice in the local diet is a major maternal and child health issue making women and young children, particularly, susceptible to diseases like anaemia and beriberi. The CDHS (NIS 2001: 173) data show that chronic malnutrition among Cambodian children is high, with 45% of children moderately stunted and more than one in five children (21%) severely stunted. The level of stunting increases rapidly with age from 15% among children under six months of age to about 50% among children age three years and older. The problem is spread equally between boys and girls. However, children with a birth interval less than 24 months had the highest level of stunting (55%). As expected, rural children are more likely to be stunted than urban children. Mother's education impacts children's nutritional status positively, with 3% of children of highly educated mothers stunted, compared with 51% of those whose mothers had no education. Malnutrition rates are much lower in the first year of life, but increase sharply in the second year of life, after which they remain more-less constant. For many children, malnutrition sets in during weaning when breast milk intakes decline sharply and adequate complementary feeding is crucial for growth. That problem may be further complicated by premature introduction of weaning foods.
32
HIV/AIDS deserves special mention. In 2002, it was estimated that 2.6% of the adult population was infected with HIV, that many tens of thousands have already died, and that possibly two hundred thousand people including children will develop AIDS within the next 5-10 years (UNCT 2001). Given that the infection rate is one of the highest in Asia, the economic implications for the country in the near future could be staggering.
mid-1990s there were 6,100 schools in the country, with some 62,000 teachers and 2.3 million pupils (Chan Sophal et al. 1999). Nearly 62% of the Cambodian population reported having an operational primary school in the village, and 82% were within one kilometre of a primary school. Thus physical access to primary schools does not appear to be a significant problem for the vast majority of Cambodians. Only Mondulkiri province seemed inadequately endowed with primary schools. However, the depth of education still leaves much to be desired and most literate adults have only received very rudimentary formal education (in the rural areas, 73% of women and 55% of men did not complete primary school) (NIS 2002a: 62). In the rural areas, moreover, over half of adult women have never even been to school (more than double the proportion of males). Educational attainment of the working age population is thus very low by regional standards, particularly for women (Table 1.3).
3.2 Education Indicators

Cambodia has a particularly difficult legacy in the area of education, as the entire educational system was effectively destroyed during the reign of the Khmer Rouge. At the time of the fall of the Khmer Rouge in 1979, all formal education in the country had practically ceased to function. There were virtually no trained teachers left, the majority of the schools had been destroyed, and there was a total lack of books and educational materials (Chan Sophal et al. 1999: 114). The recovery has been remarkable: by the
Table 1.3: Population aged 25 or more by educational level Nationwide Both sexes No schooling Primary (grades 1-6) Secondary (grades 7-9) High school (grades 10-12) Technical/vocational University Other Not stated Total
Source: Chan Sophal et al. 1999.
Rural areas Female 48.9 36.6 9.7 2.5 0.1 0.1 0.5 1.6 100.0 Both sexes 40.4 41.5 12.7 2.8 0.1 0.1 1.0 1.4 100.0 Male 25.7 47.6 19.1 4.6 0.2 0.2 1.6 1.0 100.0 Female 52.2 36.3 7.6 1.3 0.0 0.0 0.6 1.7 100.0
Male 23.2 44.4 21.1 7.6 0.6 0.9 1.3 1.0 100.0
37.5 40.0 14.7 4.8 0.3 0.5 0.9 1.3 100
33
Data from the CSES of 1997 (Table 1.4) showed that 67% of children aged 6-11 - the official primary school age group in Cambodia - were enrolled in primary school. Beyond this age, enrolment drops off sharply, with only 14% of children aged 12-14 enrolled in lower-secondary school and 7% of the 15-17 group enrolled in uppersecondary. Repetition and dropout rates were very high, particularly in the lower grades. In the 2000-01 academic year, half the pupils in grade one did not successfully complete the grade (GSCSD/PMATU 2003). About 510,000 of the 1.59 million children aged 6-11 are not enrolled in school. Among all school aged children (6-17), there are
about 1.1 million not enrolled, a huge waste of human resources for the future (MOP 1999). As the table shows, there are large disparities in enrolment between urban and rural areas, especially at higher levels of schooling. For instance, Phnom Penh's net primary enrolment ratio is 25% higher than in rural areas, while its net lower-secondary enrolment ratio is 4 times higher and the net upper-secondary ratio 10 times higher. Almost 86% of the children aged 6-11 who are not in school live in the countryside and only 4% in Phnom Penh. Thus policies to raise enrolments should be targeted mainly to rural areas.
Table 1.4: School enrolment by consumption quintile and region, 1997 Expenditure quintile
1-poorest
2 64.7 6.0 1.2 0
3 70.5 13.0 5.0 0.1
4 73.2 17.1 6.9 0.6
5-richest
Nation- Phnom Other wide Penh urban 67.0 14.1 6.6 0.7 112.8 34.7 11.6 1.2 68.0 77.4 42.0 9.9 81.3 67.9 31.6 5.8 118.3 76.5 49.4 10.7 86.6 84.8 68.9 21.7 69.8 23.8 7.7 0.6 109.2 48.2 15.1 1.1 71.6 79.9 44.0 10.3
Rural areas 64.9 9.5 3.1 0.1 112.7 27.1 6.1 0.2 65.3 76.1 38.1 8.6
Net enrolment Primary Lower-secondary Upper-secondary Post-secondary Gross enrolment Primary Lower-secondary Upper-secondary Post-secondary Agespecific enrolment Ages 6-11 Ages 12-14 Ages 15-17 Ages 18-24
Source: MOP 1999.
55.4 5.6 1.6 0 94.0 13.6 2.4 0.2 55.5 69.3 33.4 6.9
77.5 31.0 18.1 2.3 124.2 72.9 32.0 4.0 80.8 85.0 55.2 15.7
112.0 119.5 122.8 20.4 31.8 39.6 2.5 8.0 12.3 0.0 0.4 0.8 65.3 74.1 37.3 8.3 71.0 80.5 39.3 8.1 74.4 79.1 44.3 8.8
Gender differences are also marked, as shown in Table 1.5. At the primary level net enrolment for boys is only 3.5% greater than that for girls. The discrepancy in lower-secondary education is much larger, but the biggest disparity is in higher schooling. Roughly 69% more boys are in upper-secondary school and 4 times more are in post-secondary education.
34
Thus although the enrolment ratios for both boys and girls drop considerably after primary school, the gender gap also grows larger.
Table 1.5: Male and Female enrolment rates, 1997 Gross enrolment Males Females Primary Lower secondary Upper secondary Post secondary
Source: MOP 1999.
Net enrolment Males Females 68.3 15.5 8.4 1.2 65.7 12.5 5.0 0.3
Age specific Males Females 69.2 84.2 54.4 13.3 66.7 69.8 30.4 7.0
119.5 40.9 15.2 2.2
106.0 27.1 8.2 0.4
On one widely accepted indicator of school quality, pupil-teacher ratios, Cambodia's average of 63 students for each primary school teacher is much higher than for other countries in the region. There is also a big rural-urban difference in primary school pupil-teacher ratios - 35 in Phnom Penh and 39 in other urban areas, but 69 in rural areas. Moreover, in primary schools attended by the poorest quintile, the ratio is a staggering 98.6. Secondary school pupil-teacher ratios are much lower (MOP 1999). Inadequate quality of education is largely accounted for by lack of funds. The proportion of recurrent government expenditure allocated to education decreased from 18.4% in 1989 to 9.9% in 1997 (Chan Sophal et al. 1999), though it had risen again to 13% by 2001. In absolute terms, the allocation in 1997 amounted to a mere $9 per pupil. This level of expenditure does not even allow for maintenance of the educational system built up in the 1980s. Teachers' salaries are barely at the subsistence level and most schools receive little or no funds for other recurrent expenditure.
The public sector runs all but a handful of schools in Cambodia but private spending accounts for a large share of education expenditures. Total annual spending of the Ministry of Education in 1996-97 was $33 million, of which $3.6 million was financed through school fees paid by parents. So the public subsidy to education was $29.4 million. Donors and NGOs contributed $57 million. The 1997 survey data indicate that household spending on schooling (books, materials, transportation, uniforms, and so on) was $56 million. This means that the government provided only a fifth of all education resources. Households financed about two fifths, and donors and NGOs accounted for the rest (MOP 1999). Private tutoring, often by the same teacher as at school, is another major expense for parents. Among the rich, it is frequent and nearly obligatory, for teachers have to supplement their low salaries. But among the poor it is negligible, leading to further educational disadvantage. Finally, students and their families have to contribute almost the entire costs of school construction, equipment, furniture and maintenance.
35
Affordability of public schooling can best be assessed by looking at the ratio of average schooling expenditure per student to household non-food spending per capita, with the latter serving as a proxy for discretionary household income. For example, the quintile-specific affordability ratio of 26% for primary schools for the poorest 20% of Cambodians is calculated by dividing their average schooling cost per student of $9.5 by non-food expenditure per capita of $36. Thus the cost of one child in primary school takes up a quarter of all non-food spending per capita in the poorest quintile, while a child in lower-secondary school takes 57%. In reality, however, the share of total schooling expenditure in household consumption expenditure could well be lower for the poor because of their lower rates of enrolment. Across all quintiles only 1-1.5% of total consumption expenditure is devoted to education (MOP 1999). Poor (and apparently declining) quality of schools may be a factor in lower enrolment rates among the poor as parents might not consider it a worthwhile investment. Moreover, poor-quality schools mean poor student performance, increasing the likelihood of repeated grades and dropouts and reducing the likelihood that poor children will be able to move on to higher education. However, the high private cost and unaffordability of schooling would appear to be a far more important reason. The CSES 1996 revealed that the main reason for dropping out of school among children aged 5-17 years lay not on the supply side, but on the demand side: children dropped out 'to help with household chores' (15.1%).
This cost affected girls far more than boys among the female dropouts, economic reasons were stated in 43.0% of the cases. This clearly suggests that poverty is, in itself, a main reason behind the high dropout rates and that poverty, especially among females, is perpetuated as it is the poor who dropout of school prematurely. While the supply of quality schools is clearly in need of improvement, educational attainment is unlikely to improve significantly until increased prosperity arrives.
3.3 Social Exclusion

Three aspects of social exclusion are important in Cambodia: (i) illiteracy; (ii) access to decision making; and (iii) corruption and poor governance (RGC 2002a). Of the adult population, only 37% is functionally literate (RGC 2002a). The link between illiteracy and poverty is demonstrated by the CSES 1997 data: the incidence of poverty among households whose heads were reportedly illiterate was 42%, compared with 34% for those with literate heads. However, differences were even greater between those households whose heads were primary school leavers and those whose heads were secondary school leavers. Farmers, fishing families, housewives, labourers and highland minorities are groups with high levels of illiteracy. Gender imbalances are also severe: 47.6% of men are literate, while only about 22% of women can read and write. People aged 25 to 40 have a higher illiteracy rate than other age groups because they were deprived of education during the decades of war and political instability.
36
Illiteracy is the basis of long-term social exclusion in the wider society. The lack of access to decision-making has prevented the poor from actively participating in the formulation of policies, plans, programs, and projects, which in turn has inhibited efficient and effective implementation, and exacerbated tensions between local authorities and local populations. Many existing laws and regulations were adopted without consultation with local communities. New problems are emerging, such as land disputes and HIV/AIDS, in which the interests of the poor need to be protected. The poor - particularly the most disadvantaged groups such as internally displaced people and refugee returnees, war widows, orphans, street children, squatters, and people with disabilities - have little or no influence. Women are particularly affected by this lack of access to decision-making. The UNDP's Gender Empowerment Measure (GEM) is a measure of the relative participation of women and men in political and economic leadership roles, relative to their representation in the general population. Cambodia's GEM, at 0.283 is lower than those of three other Asian countries for which calculations are available -Bangladesh, Sri Lanka, and the Philippines (MOP/UNDP 2000). Corruption is a major problem that adversely affects the poor; both by its negative effect on government revenue, and by the direct costs imposed on households. In the latter regard, rent seeking in the form of illegal fees for education and health services is especially burdensome. The poor are also victims of (i) illegal logging that
destroys forests as a community resource, (ii) arbitrary reallocation of fishing resources to private commercial interests, and (iii) land grabbing by influential officials and business interest groups. Feelings of powerlessness are an inevitable result of this social exclusion. The quality of public administration is low, especially at the provincial level, and makes for poor delivery of basic services. Again, the poor find themselves in powerless to claim those services. Governance reform and decentralization efforts are intended to address the problems of corruption and access to decisionmaking. In particular the newly-elected Commune Councils have gender quotas intended to improve the situation of women, but it is too early (the elections were held in early 2002) to assess the impact of these reforms.
3.4 Vulnerability to External Shocks

A key aspect of the lives of the poor is vulnerability to natural shocks (some of disastrous proportions). A large share of the population is dependent on rain-fed agriculture and lives alongside waterways prone to seasonal flooding. The 2000 floods affected some 444,000 hectares of agricultural land, in over half of all provinces. Damage occurred to rice and other crops, to rural provincial infrastructure and to housing. Other natural disasters including drought, crop failure, and fire also have differential impacts on the provinces.
37
Flood, drought, and crop failure are more frequent in poor villages than in those inhabited by the rich. In the 12 months preceding the 1997 CSES survey, for example, 82% of the poorest villages experienced some form of natural disaster compared to only 26% of the richest (MOP 1999). Not only are poor villages more likely to experience natural disasters, they are also less able to cope because of their lower wealth and extremely limited disaster management capabilities. Disasters that destroy assets can force borrowing and asset sales that undermine the prospects of climbing out, and staying out, of poverty. Bad health and illness, aggravated by chronic malnutrition, can generate major crises for the poor, also. The most common complaints in rural areas are malaria, dengue fever, tuberculosis, diarrhoea, hepatitis, typhoid, cholera, and birth complications. Health shocks induced by these complaints often lead to indebtedness and asset liquidation. There is also the opportunity cost from labour days lost by both the ill and their caregivers. Once people living with AIDS reach the terminal stages, health care expenditures and labour days lost is likely to be even greater. Animal mortality and disease is systemically linked to other crises impacting upon rural Cambodians (Murshid 1998: 54). Raising livestock is important for most rural households. While they are unlikely to own draught animals, the poor - especially women - try to raise pigs and poultry, to act as a savings mechanism for times of need. The loss of livestock due to illness or theft means that many rural households have no way to meet the expenditure necessary for
crises associated with natural disasters or the illness or death of household members.
4. RATES AND CHANGES IN INEQUALITY 4.1 The Distribution of Consumption

Growth with increasing inequality limits the extent of poverty reduction and conceivably could even increase the incidence of poverty. Table 1.6 provides a comparison over the six-year period 199399 based on the three CSESs. Measured by several alternative indicators, inequality is consistently higher in Phnom Penh and declines outside the urban areas. The table also shows a consistent pattern along social lines: the poorest decile of the population shared 3% of the total consumption and the richest decile over 30%. This inequality is slightly less marked in rural areas with the corresponding figures rounding to 4% and 20% respectively.
38
Table 1.6: Consumption inequality profile Indicators Inequality in consumption: Gini Index Strata National Phnom Penh Other Urban Rural CSES 1993-94 0.38 0.39 0.44 0.27 3.4% 2.5% 2.7% 4.4% 32.8% 31.2% 36.7% 22.9% CSES 1997 (adjusted) 0.42 0.46 0.44 0.33 3.0% 2.6% 2.7% 3.7% 35.3% 40.3% 37.6% 27.1% CSES 1999 0.35 0.38 0.35 0.24 3.2% 2.8% 2.8% 4.0% 30.7% 30.0% 28.4% 20.3%
Consumption share of the poorest 10%
National Phnom Penh Other Urban Rural
Consumption share of the richest 10%
National Phnom Penh Other Urban Rural
Sources: MOP/UNDP 2002.
Rural households in 1997 spent as much as 71% of total monthly consumption expenditure on food (NIS 1998). The indications of low average consumption (see Section 2) mask wide differences in real consumption due to spatial cost of living differences. Real per capita consumption expenditure in 1993-94 was more than twice as high in Phnom Penh as in the country as a whole and the same pattern is true for 1997. Border provinces in the far west near Thailand, and in the east near Vietnam, have the lowest average consumption levels. In fact food consumption expenditures fall below 70% of per capita household expenditures only among the richest quintile of the population. Non-food expenditures rise above 20% only in that quintile. Rice purchases represent 20-30% of expenditures for the poorest half of the population, thus any policy that has a high impact on rice prices should have a high impact on poor household's consumption. Keeping the other prices and the sources of income constant, a reduction, say, of 10% in the
price of rice would imply an extra 3% of disposable income for people in the poorest decile and 0.6% for people in the richest decile (NIS 1999). This pattern of inequality is replicated along urban-rural lines. For households in Phnom Penh, the share of food in total expenditure is below 55% for all deciles, and as low as 18% for the 10th decile. Thus any policy that affects food prices, everything else constant, will have a greater impact on households' welfare outside the capital, and a smaller impact - albeit still important for people in the poorest deciles - on households in Phnom Penh (NIS 1999). Membership of different consumption quintiles is also strongly related to such factors as household size, number of dependent children and education levels. Those among the wealthier sections of the population have smaller families, fewer children and higher levels of education. However, even among the richest 20% of the population, levels of literacy and educational attainment
39
remain rather low. For the poorest quintile adult literacy is as low as 56%, rising to 74% for the richest. Years of schooling for those aged 15 and over in the poorest quintile are 3.5 and 2.1 for males and females respectively, but still only 6.2 and 4.1 for the richest quintile (NIS 1999).
4.2 Functional Distribution

Table 1.7 shows the contribution to total income from different sources by per capita adult equivalent consumption decile. Earnings from self employment are on average equal to 61% of total income, 2/3 of which come from activities related to cultivation, livestock, fish raising, and forestry and hunting. Income from wage employment was on average only 20.5%, slightly higher than the 16.5% coming from such sources as rental income, interest received and imputed value of houses. Income from remittances represents on average only about 2.1% of total income.
Table 1.7: Contributions of income from different sources, 1999 Consumption decile Source of Income Self employment from cultivation from rice cultivation from other crops from livestock from aquaculture from forestry and hunting from non farming activities Other sources from wages from remittances Others (rents, dividends, etc) Total 1 63.8 27.8 21.4 6.4 16.6 6.1 8.3 8.9 32.2 19.0 1.2 12.1 100 2 71.4 29.7 24.3 5.4 14.3 7.1 10.2 10.0 28.6 15.0 1.9 11.7 100 3 71.7 31.1 25.3 5.8 14.2 5.8 5.5 9.7 11.2 28.3 15.9 10.9 100 4 70.3 30.3 23.6 6.7 13.0 8.7 7.3 10.9 29.7 14.5 2.8 12.4 100 5 73.6 31.9 25.7 6.2 12.4 6.9 7.7 14.7 26.4 14.3 1.3 10.8 100 6 68.5 32.1 25.0 7.1 11.9 6.8 7.9 9.8 31.5 18.5 2.4 10.6 100 7 71.2 30.6 22.7 7.9 11.6 8.7 8.1 12.2 28.8 17.6 1.6 9.6 100 8 70.7 32.3 20.4 11.9 12.0 7.1 6.1 13.2 29.3 16.5 1.7 11.1 100 9 63.3 19.8 9.7 10.1 8.3 5.2 4.9 25.1 36.8 21.0 2.3 13.4 100 10 36.3 3.2 2.1 1.1 1.5 2.1 0.6 29.0 63.9 30.2 2.6 31.1 100 av. 60.9 22.4 16.3 6.1 9.4 5.6 5.7 17.7 39.1 20.5 2.1 16.5 100
Source: RGC-Ministry of Commerce 2002.
The table shows that the composition of income is similar for the 8 poorest deciles and only different for the top two deciles. The latter show a higher importance of self-employment in non-farming activities, of wage income, and, for the tenth decile, a higher proportion of income coming
from capital stock ownership. Wages are, predictably, a more important source of income for Phnom Penh residents than in the rural areas. Regarding the distribution of employment of the household head by industry, 75%
40
of Cambodian household heads have agriculture, hunting and forestry as their main occupation, with a slightly higher proportion for female headed households (78%). Employment in the public, education and health sectors is next in importance as a source of main income (9% average) followed by employment in the wholesale and retail trade sectors (7.3% average) (MOP 1999). Income inequalities are sharp and this is particularly true of agricultural and related incomes. To some extent this is offset by earnings from the labour market. For the poor, the role of women in the labour market is particularly striking. The other major feature relates to the role of hunting/ gathering activities and access to common property resources. Common property is a significant source of income, especially for the poor. However, important changes are underway that will tend to restrict access.
Much of the extractive activities in the forest is 'illegal' in nature, requiring the payment of bribes to guards in addition to the risks faced from land mines, malaria and fever. It is thus the most desperate who will choose this option. Leasing of water bodies to powerful business interests and ever increasing restrictions on free access to fisheries is already evident in places, like Kandal, where the poorest depend most on hunting/gathering for their livelihoods. What we observe is a precarious balance in terms of sustaining livelihoods and trends that appear to have potentially adverse consequences for the poor. As things currently stand, there is a serious shortage of productive non-farm employment (Table 1.8) The near landless, who show a very high dependence on cultivation, are particularly vulnerable. Less than half a hectare of land does not suffice to provide an adequate living for a household.
Table 1.8: Main source of income of household heads in previous year by size of landholdings (%) Source of income Cultivation Livestock, fisheries, forestry Non-farm Wages/salaries
Source: CSES Database1997
landless 20 9 38 33
>0.5 ha 81 6 7 6
0.5>1.0 ha 82 7 5 6
>1.0 ha 84 7 4 5
Average 74 7 10 9
41
Table 1.9: Composition of employment (%) by sector of primary employment and sex, 1993/4 and 1999 Male 1993/4 Agriculture, hunting, forestry Fishing Manufacturing Electricity, gas, water Construction Wholesale, retail trade Hotels, restaurant Transport, storage, communications Public admin, defence, etc. Education Health, social work Private households International organizations Others Total
Source: Godfrey et al. 2001.
Female 1993/4 75.3 1.0 3.2 0.0 0.1 15.6 0.3 0.9 0.8 1.0 0.6 0.1 0.0 1.3 100 1999 77.1 1.0 6.5 0.1 0.3 10.6 0.4 0.4 0.6 1.9 0.5 0.6 0.2 0.2 100
Both sexes 1993/4 70.6 1.5 3.7 0.1 0.9 11.1 0.2 3.4 4.3 1.7 0.6 0.3 0.1 1.5 100 1999 73.9 2.1 5.3 0.1 1.4 7.4 0.4 2.3 3.3 1.5 0.5 0.6 0.3 0. 100
1999 70.3 3.3 3.9 0.2 2.6 3.9 0.4 4.3 6.1 2.2 0.4 0.7 0.4 1.3 100
65.6 2.0 4.2 0.2 1.6 6.2 0.2 6.1 8.2 2.4 0.6 0.5 0.1 2.1 100
In fact the sectoral structure of the labour market has changed comparatively little since the early 1990s. According to Table 1.9, the proportion of workers in agriculture actually increased during the decade. Gradual settlement of land ownership rights after privatization in 1989, rather than a shift of people to agriculture because of rising labour demand, is perhaps one reason for the observed proportional increase of workers in agriculture as well as the return of refugees and demobilization of the armed forces. The other sectors to show increases in their share were manufacturing, construction, and, from very low levels, hotels and restaurants, private households and international organizations. The increase in the proportion of women working in manufacturing, reflecting the rise of the garment industry, apparently at the expense of their involvement in trade,
is remarkable. So is the fall in the proportion of men engaged in manufacturing, trade, transport and communications, and public administration, defence and security. For obtaining a clearer analysis of the relationship between the labour market and poverty, it is necessary to get away from the concept of 'household head' and to look, rather, at the composition of households (Table 1.10). Leaving aside those below working age, the most common employment categories within the richest expenditure quintile in 1999 were: (1) economically inactive adults; (2) ownaccount workers and (3) employees. These three groups accounted for 71% of adult household members. For the poorest quintile, on the other hand, the largest categories were: (1) unpaid family workers; (2) economically inactive adults and (3) own-account workers,
42
accounting for 91% of the total. The proportion of wage and salary earners was three times greater in the richest than in the poorest households, a pattern repeated across genders. The percentage of economically inactive adults was also 50% higher while that of unpaid family workers was only 63% of that in the poorest quintile. The percentage of unemployed members is actually slightly higher in the richest households, as might be expected since they can afford to wait for the job of their choice.
Table 1.10: Distribution within households by labour-force category, sex and consumption quintile 1999 Household per capita consumption quintile The poorest quintile M Employees Employers Own-account workers Unpaid family workers Other employed Unemployed Economically-inactive adults Under working age Total h/m members
Source: CSES Database 1999.
The richest quintile M 10.5 0.0 10.1 4.7 0.1 0.7 12.6 9.3 48.0 F 5.5 0.1 8.1 10.7 0.0 0.7 17.6 9.1 52.0 F+M 16.0 0.1 18.2 15.4 0.1 1.3 30.4 8.5 100
F 2.1 0.0 5.4 17.2 0.0 0.6 11.5 14.3 51.1
F+M 5.3 0.0 17.2 24.4 0.0 1.0 22.4 29.6 100
3.2 0.0 11.8 7.3 0.0 0.5 10.8 15.2 48.9
These data suggest that unemployment and economic inactivity are luxuries that richer households can better afford, rather than that they are causes of poverty. They are also consistent with the suggestion that the way out of poverty is for household members to move out of unpaid family work into wage employment (for both males and females). Better forms of selfemployment, especially for women, may also be warranted. However, it should be noted that, according to the CSES 1997, the self-employed - mainly farmers - contributed 74.3% to total poverty. We note the importance of non-farm self-employment in the incomes of the top two deciles shown in Table 1.7.
Within the small wage and salary earning group, data on the distribution of employees by type of employer in 1997 and 1999 show that the proportion of private sector wage earners rose substantially, especially in the case of women (Table 1.11). This development is largely attributable to the expansion of the garment industry.
43
Table 1.11: Distribution of employees by type of employer and sex, 1997 & 1999 1997 Male % Government State enterprise Joint venture Private International organization NGO Total
Source: CSES database 1999
1999 Total % 41 5 7 44 2 2 100 Male % 44 1 2 49 1 1 100 Female % 16 4 3 73 2 2 100 Total % 32 2 3 59 1 2 100
Female % 27 4 9 54 1 4 100
47 5 5 40 2 1 100
Average wages of both male and female workers rose by more than 50% between 1996 and 1999. Even with price inflation around 29% over this period, there has been a substantial real wage increase. Women's wages increased particularly rapidly for those with some secondary schooling, while men's wages rose very fast for the few with post-secondary education. Differentials between men and women generally increased, though they declined for those with no schooling or some secondary schooling. Moreover, the proportion of females in each earning group declined as the amount earned rose. Women form 40% of the lowest earning category, and only 13% of the highest (Gorman et al. 1999). Analysis of the CSES data, controlling for age and education in six occupations, found that men, on average, earned 50% more than women (Gorman et al. 1999). The lower earnings of women thus reflect discrimination. The vast majority of Cambodia's poverty, over 71% according to the CSES 1997, is in households whose heads work in agriculture. Other sectors in which the incidence is high include construction and non-specified industries. However, because of their small
size, the overall contribution to poverty of the latter groups is low. In spite of low wages in the public sector, households headed by people working in government, education and health services are among the least poor. Public sector employees are able to generate extra incomes on the strength of their education and placement. The policy implication clearly is that poverty alleviation needs to focus on the selfemployed and unpaid workers, particularly those in agriculture. Only about 4% of the poor work in private sector non-family businesses, so policies such as minimum wage legislation will have little effect on the poor. Rather, the poor will benefit most from policies that improve the functioning of rural markets, including those for agricultural commodities, inputs and rural credit as well as expansion of non-farm waged employment.
5. LEVELS AND CHANGES IN THE DISTRIBUTION OF PRODUCTIVE ASSETS

An obvious cause of poverty is inadequate endowment of, or access to, productive resources. Not only land, but also physical
44
and human resources, are important in this respect.
5.1 Land
With little more than 50 inhabitants per square kilometre, Cambodia is one of the more sparsely inhabited countries in developing Asia. Yet it does not have an abundance of agricultural land relative to its population. Much of the rural population is concentrated on 30% of the country with only a quarter of the area cultivated. First, the population growth implies a diminution of cultivable land per capita and, if non-farm employment does not increase, a reduction in the land per farmer. According to the Cambodia Land Cover Atlas of 1994 (Chan Sophal et al. 1999), there was already in 1992 less than 0.3 hectares of cultivated land per inhabitant. Second, there are large regional differences in the land/labour ratio. In many areas there is scope for increasing the cultivated area, either through cultivation of new land or double cropping. Third, the most worrying and least documented factor is the unequal access to land and common property resources within individual communities in rural areas. In 1989 the Council of Ministers issued Sub-Decree No. 3 on the 'Principles for Possession and Use' by each household of up to five hectares of land for cultivation. There seems to be an implicit assumption that this reform solved once and for all the problem of access to and distribution of land. However, people received varying amounts of land with different soil quality: for example, those in densely-populated
provinces received smaller amounts of land than those in the north-west where, in some provinces, the soil quality was also higher. Moreover, although during the Khmer Rouge period nobody could own anything privately, some people were able to hang on to valuables and use them afterwards, while others received resources from relatives in third countries. These groups have started out with an advantage over others. Today, average agricultural land size per household and average incidence of landlessness vary significantly across provinces. Some major provinces tend to have a relatively-high incidence of landlessness without showing a significant relationship between land availability and landlessness. This might be due to competition for land in those provinces or to the availability of alternative income opportunities. For example, Koh Kong province had the highest degree of landlessness (50%), but it also has alternative opportunities in fishing, cross-border trade with Thailand and logging. Distress sales are one of the most important causes of increasing landlessness, one study reporting that 40% of new landlessness was the result of health-related distress sales (NPRS 2002: 18). Inequality of access is another factor affecting both landlessness and poverty among landholders. Over the last two decades, some people have been able to expand their land holding by buying more land while others have had to sell. High population growth and the poor land supply system have exacerbated this process. Current land market distortions not only cannot prevent
45
land concentration, but accelerate it in rural areas. Unfortunately, there are no comprehensive data on land concentration in Cambodia, but Gini coefficients, ranging from 0.47 to 0.66, have been calculated based on a number of surveys - all showing a high level of inequality. Moreover, the surveys5 came up with the fairly uniform result that about 20-30% of landowners held roughly 70% of the land, while the poorest 40% occupied only 10%. The problem of security of land tenure is a serious one. Even farmers who own land often have no legal titles to the land they cultivate. It is estimated that only 10% of applicants for titles have so far been granted certificates of ownership, and even these certificates are of a temporary nature. Sometimes possession of a title is insufficient
protection. This not only causes land disputes but also discourages farmers from investing in the long-term quality of their land, as their tenure is insecure.
5.2 Other Assets

Animals are an important source of wealth in rural Cambodia, second in importance to land. They are used for draught power and transportation (oxen) as well as being a source of income through fattening and trade (pigs and poultry). However, animal raising appears to be hazardous as a result of high mortality rates. Land ownership and that of other assets, such as animals, transport and machinery, capital goods and consumer durables appear to go hand in hand. Indeed, in every case, distribution was found to be even more adverse than in the case of land (Table 1.12).
Table 1.12: Rural Distribution of Assets and Income Share of bottom 20% Land (cultivated) Animals Transportation Durables Machinery Income per adult
Source: Murshid 1998.
Share of top 10% 33.0 30.4 46.0 36.3 47.1 25.0
Gini ratio 0.47 050 0.61 0.49 0.65 0.33
3.7 0.4 2.3 6.9 1.6 10.0
5.3 Asset Markets

There is significant land market activity in Cambodia. At one end of the scale there are large institutional and quasi-institutional landowners that include the army (more accurately, individual officers) and international agro-business interests. Tracts of both agricultural and forest lands have been taken over by these groups, taking advantage of a weak regulatory framework, poor enforcement of property rights and widespread
5 For example, Household Socio-economic Survey, MRC 1995-96; CSES, NIS 1997; Baseline Survey, WFP 1998; and PET, WFP1998.
46
corruption. This has led to widespread landgrabbing by powerful vested interests, pointing to an emerging dual ownership pattern in the rural areas: a small-scale subsistence sector and a nascent commercial sector. Little has been forthcoming by way of investment in the latter, so that the principal motivation here seems to be timber extraction and land speculation. Although this sector is believed to be large, its actual size is not known. The speed and scale of land transactions within the small-scale sector are significant. During a five-year reference period, between 11 to 13% of the total land changed hands, the rich and well-off being net buyers, the poor being net sellers (Chan Sophal et al. 1999). Distress sales were the most common cause of concentration in land holdings. Ownership of and trade in animals seem to be concentrated in the hands of the better-off households, who account for over 80% of turnover value, dealing mainly in pigs and cattle. Poor households are mostly restricted to the ownership of and trade in chickens and ducks (Murshid 1998). The market for animals does not appear to provide a significant avenue of income or profit for the poorer households whose scale of operations is tiny. Indeed, for the poorest, profitability appears to be negative (ratio of sale price to purchase price was 0.77) (Murshid 1998). The tenancy market for land, animals or trees also appears to be restricted to non-poor households. In the case of land, the predominant form of contracts was the fixed-rent type, in cash, kind or both. Rental of palm trees was found to be extremely
important in Kompong Speu, where the incidence was 50%, but less so elsewhere (70% were owned by the well-off). Contracts are again of the fixed type, generally payable in kind, i.e. sugar. Animals are also shared, particularly among female-headed households which, again, tend to be among the non-poor (Murshid 1998). From the point of view of poverty alleviation, the increasing landlessness and nearlandlessness is very serious indeed. It would appear that the advantage of near universal and fairly equal access to land is rapidly being lost. Combined with a loss of access to common property resources it implies that an increasing number of rural households remain without any productive resources except their own labour. These households will be unlikely to improve their poverty status unless there is an overall intensification and improvement in agriculture which generates more rural, especially non-farm, employment opportunities. Preventive measures, focusing on halting the process of pauperization of parts of the rural population, are likely to be much more effective than any subsequent attempts to revert loss of land and to bring the landless and near landless back into the mainstream of development.
5.4 The Credit Market

Despite the poor development of formal credit institutions, most households are involved in the credit market. There is a great variety of informal credit proliferating in rural Cambodia, mediated in cash, in kind or in a combination of the two. In some types of credit, there is an explicit interest
47
rate involved, while in others it may take indirect forms. Rural areas are, however, inadequately serviced by development institutions. Government ministries have little visibility at the local (commune) level and NGOs working in development are spread thinly and unevenly throughout the country. The financial system is weak and rural banking virtually non-existent. Only about 10% of the population is estimated to have access to formal credit. The absence of land titles is one of the main factors limiting land holders' access to such credit. When one talks of credit markets in the Cambodian context, one is referring essentially to informal markets: these tend to have high costs to participants related to high levels of distress borrowing (see Chapter 6).
has one of the youngest populations in the world. This distorted age structure has important implications for the economy. To begin with, the high proportion of children in the population implies a high dependency ratio (although this has begun to fall recently). In other words, each income earner has to support a high number of dependants. This has a generally depressing effect on income per head. At the household level, the ratio of economically active to non-economically active people, which in the Cambodian context reflects the number of children in the household, exerts a strong influence on poverty. The youthfulness of the population also implies a high rate of population growth, which exerts a pressure on the economy as a whole and on agriculture, in particular, to expand at the same rate lest per capita incomes and per capita food production decrease. Population growth has been rapid (2.5 average annual population growth rate from 1962 to 1998) from its nadir at the fall of the Khmer Rouge regime. Despite recent falls in fertility rates, it will continue to grow fast and is expected to reach 15.5 million by 2010 and 19.3 million by the year 2020 (NIS, 2002a). Even more important for macroeconomic policy, the labour force will increase at in unprecedented rate in the coming decades as the 'baby-boomers' enter the labour market, while few leave it. Indeed, this process has already begun. The apparent contradiction in the past few years between the high growth rates of GDP on the one hand and the fall in productivity on the other is primarily a consequence of the
6. THE DEMOGRAPHIC CHALLENGE 6.1 The Impact of the Baby Boom

Cambodia's demographic features are important for an analysis of poverty. Perhaps the most important relates to the age structure of the population. As a consequence of the country's violent past, the age groups born prior to 1980 are relatively small, reflecting the suffering during the war and the Khmer Rouge period. On the other hand, the age groups born since 1980, that is those who are at present aged 22 or less, are relatively large, as a result of the baby boom which followed the fall of the Khmer Rouge. With about 43% of the population under the age of 15 in 2000 (NIS 2001), Cambodia
48
rapid increase in the labour force. Generating productive employment opportunities for the large number of the new labour force entrants will, more than any other factor, determine Cambodia's economic prospects. In the future, growth must be based on raising productivity in rural areas, and in particular in agriculture, where the large increases in the labour force will be concentrated. The demographic trend creates both a critical challenge and an opportunity. The challenge is to generate employment for the rapidly increasing labour force without compromising the level of productivity. In other words, the additional employment created must be at a level of productivity that is at least as high as that of the already existing jobs. The opportunity lies in the implication that if the challenge to create productive employment is successfully met, the improved dependency ratio will add to the
growth momentum through automatic improvements in per capita incomes and, as a consequence, in savings which can be turned into productive investments.
6.2 Migration
The combined effects of rural underemployment and more rapid urban economic growth have resulted in a growing tendency for rural Cambodians to leave their villages. Comparison of the 1996 sample survey with the 1998 Census data, provides an indirect way to estimate recent trends in internal migration, which shows that rural out-migration is increasing. The 1998 Census counted a total of 881,400 persons who had moved from a rural area within the five years prior to the Census. As Table 1.13 shows, of these, almost one-third (31%) had moved within the previous 12 months.
Table 1.13. Migrants from rural areas 1996 and 1998 1996 Demographic Survey 'recent' migrants' 'very recent' migrants
Source: Godfrey et al. 2001.
1998 Population Census 881,439 273,534
2-year increase 38.9% 58.7%
634,786 172,305
As a result of increasing geographic mobility, most places in Cambodia now include significant and fast-rising proportions of 'newcomers'. This is especially noticeable in urban areas. In 1998, 10.5% of the rural population consisted of persons who migrated within the previous five years and this figure was 29.6% for the urban areas. Such striking figures raise questions not only about the adequacy of urban infrastructure to accommodate newcomers, but also loss of contact with old community networks and values (Godfrey et al. 2001). Census data on net out-migration at the provincial level show that in 1998, just five provinces accounted for over one-half of all 'recent' out-migrants. Four of these were provinces with the highest rural population densities (all relatively near Phnom Penh) and small average farm sizes (Godfrey et al. 2001).
49
Provincial in-migrations are just as geographically concentrated, with five provinces receiving over 60% of in-migrants in 1998. Of course, the top destination was Phnom Penh, which alone received about one-third of all inter-provincial migrants. Two rural provinces, Koh Kong and Banteay Meanchey, with the highest average farm size and low population densities, also figured in this list (Godfrey et al. 2001). Youth and young adults are disproportionately represented among migrants. Persons aged 15 to 29 years, who make up 24% of the total population, accounted for 43% of the 'very recent' migrant flows. Men make up just a little over one-half of the total number of migrants, but females made up 56% of 'very recent' migrants to Phnom Penh. This reflects the opening up of numerous garment and shoe factories in the capital in the late 1990s. Conversely, men made up well over 60% of 'very recent' migrants to more remote provinces and over 80% in some provinces along the Thai border, suggesting a strong movement of single males, possibly employed with the army, into the 'frontier areas' (Godfrey et al. 2001). Of the total migrants who had left their villages less than one year before the census date, 29% stated that their principal reason for moving was the need to search for employment. Another 25% stated that they needed to follow their families. These movements thus appear to reflect strongly the prevailing underemployment in the countryside (Godfrey et al. 2001). In addition to the voluntary migration discussed above, Cambodia also suffers
from a serious problem of trafficking in women and children. It is reported that about 88,000 women have been trafficked into Thailand and around 14,000 local sex workers were also tricked or forced into the business (NPRS 2002: 39).
7. POVERTY AND THE RURAL ECONOMY

It should be clear enough by now that whether or not Cambodia will successfully face up to the challenge ahead will be to a large extent determined by developments in the agricultural sector and the rural sector more generally. Agricultural production will have to increase on a sustained basis at least in line with the increase of the labour force, which in real terms means by at least 4-5% a year. Creation of non-farm employment, primarily in rural areas, is essential for the economy to progress. However, such diversification will only take place in the context of dynamic development in agriculture and with increases in real production beyond subsistence level. At the same time, a dynamic development of agriculture is likely to serve as a demand-based catalyst for a development of the non-farm sectors and thus for a diversification of the rural economy. Past trends in agriculture do not, at first glance, seem to give much ground for optimism (Table 1.14). Production recovered quite quickly after the abysmal situation during the Lon Nol and Khmer Rouge periods. However, between 1980-81 and 1994, the trend in rice production was one of slow growth with large yearly fluctuations. Growth was largely due to an increase in the cultivated area in the first decade to
50
1989. However, yields actually fell and per capita production declined to pre-1986 levels during the first few years of the new administration. Although data are not available, it seems safe to conclude that labour productivity in agriculture declined throughout this period, and by 1994 it must have been considerably below its peak levels in the late 1960s.
Table 1.14: Rice production trends Year 1967-69 av. 1980 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
Note:
Since 1995 there has been a marked improvement in rice production. Output increased by 55% between 1994 and 1995 and has since grown roughly in line with population. The sharp increase in production in 1995 was partly due to unusually adverse climatic conditions in 1994 and under-reporting of production in earlier years (Chan Sophal et al. 1999).
Production (000 tonnes) 2,500 1,717 1,812 2,093 1,815 2,500 2,672 2,500 2,400 2,221 2,383 2,223 3,448 3,404 3,415 3,510 4,030 4,050 4,026
Area (000 ha) 2,513 1,441 1,462 1,535 1,378 1,879 1,932 1,890 1,910 1,853 1,857 1,924 2,086 2,171 2,076 2,095 2,158 2,158 2,319
Yield (t/ha) 1.00 1.19 1.25 1.36 1.32 1.33 1.38 1.32 1.25 1.20 1.28 1.16 1.65 1.64 1.65 1.67 1.87 1.88 1.74
Production per capita (kg) 385 262 242 272 230 309 322 291 273 246 256 225 334 329 319 307 347 331 307
Per capita availability data are particularly unreliable since no actual census was carried out between 1962 and 1998. Source: Calculated from NIS 2002a: 5, 109.
However, there are also clear indications that a process of intensification of agricultural production has begun. The most concrete evidence is the increase in the cultivation of dry season rice, which is entirely dependent on irrigation. Although virtually insignificant a decade ago, the area under higher-yielding dry-season rice expanded by 30% in 199495 and by 6% a year during the next two years. By 1997 the dry season crop accounted
for 22% of the total production. Another indication of improved cultivation practices is the increase in the yields of wet season rice in the past five years, which can hardly be put down to increasingly favourable climatic conditions. However, in the absence of consistent studies of cultivation practices over time, it is difficult to pinpoint exactly the reasons behind the increase (Chan Sophal et al.1999).
51
Comparison with neighbouring countries shows that future rice production potential is considerable. Current yields per hectare are less than half those in neighbouring Vietnam and three quarters of those in Thailand. Those two countries use 214 and 71 kg/ha respectively of chemical fertilizer, compared to Cambodia's 2.3 kg. Similarly, the use of high-yielding varieties of rice would appear to be much lower in Cambodia than in neighbouring countries. While the overall production of rice is adequate, there are wide fluctuations in supply depending on the weather (including natural calamities like typhoons) and the flood regime from the major or rivers. Besides rice, almost all of the other crops, except vegetables, are also grown without irrigation and therefore subject to the same wide fluctuations. In the absence of good buffer stocks at the regional level, these fluctuations create periods of food insecurity and rising prices for people in certain regions. At the heart of the problem lies the need for a transformation from a primarily subsistence-oriented agriculture, characterized by low use of cash inputs and low returns to land, to a more commercially-oriented, more intensive agriculture. The scope for such a transformation remains large, but there are also major obstacles, ranging from the poor physical infrastructure resulting in high transport costs and poor market integration, to low levels of education, training in new techniques and the poor state of health. Overcoming these bottlenecks will be the main key for achieving sustained rapid economic growth based on expansion of increasingly productive employment opportunities. Like markets and economic services, economic infrastructure in the form of
roads and electricity can raise productivity in both farm and non farm activities and thus improve incomes. Yet in the CSES sample, only 43% of households in villages had access to electricity. Access to a motorable road was much better, with nearly 81% reporting a motorable road entering the village. However, only 9.3% of households in the poorest quintile of villages had electricity in their homes, compared to 81% in the richest quintile. Likewise, only 70.2% of the poorest quintile of villages, but 94.6% of the richest, had a motorable road entering the village. Economic services, especially agricultural extension, are also poorly supplied. This is a relatively new concept in Cambodia. There were no specialized courses in the field until 1991, when Australian Catholic Relief introduced the subject through training existing agricultural agents and technicians. Now there are university graduates, posted mainly in ministerial and provincial departments. Too few of them are at the district or commune level, where they can provide consultation directly to the farmers. Their mobility is limited by the poor infrastructure, while their salaries are low. Thus farmers have limited access to extension services and continue to rely on traditional farming techniques.
8. CONCLUSION
Despite the absence of reliable data, there are signs that economic growth during the past decade has not produced any significant poverty reduction. Indeed, there are some signs that the situation is worsening - reduced per capita consumption (measured in riels); fewer public health facilities and rising infant and child mortality; poor education outcomes; increasing population pressure
52
on cultivable land with rising rural underemployment due to labour-force entry of baby boomers and lack of growth in nonfarm employment. Increased vulnerability of some, resulting in part from Cambodia's war legacy and the threat posed by the HIV/AIDS epidemic, are also important factors in poverty in Cambodia. Inequality is also very high, particularly along urban-rural lines. In fact most of the economic growth of the past decade has been confined to urban enclaves, while rural growth, especially in staple food production, has barely kept pace with the increase in population. The problem of tackling poverty and inequality is thus largely one of addressing these inequalities between city and countryside by generating more broad-based growth. This is now possible, given the increasing stability of the countryside and successful macroeconomic stabilization (see Chapter 2). We have noted that the depth of poverty is quite shallow, with the exception of certain especially vulnerable groups, and inequality is lower in the rural areas. Thus a poverty reduction strategy which aims at achieving more egalitarian growth by broadly targeting the rural areas, is likely to have a significant and broad impact. Such a strategy will be based, first, on increasing the dynamism of agriculture and, second, on generating non-farm employment, particularly in the rural areas if excessive stress on urban infrastructure and living conditions is to be avoided. As will be argued in the next chapter, both supply-side and demand-side measures will be needed. On the one hand, there is a clear need for demand expansion to promote rural prosperity and to integrate more of the rural population into the market economy. On the other hand, there is a
need for great improvements in the supply of human and physical assets (both public and private) in order to improve the capacity of the vast majority of Cambodians to take advantage of opportunities which may arise. Specific actions on the supply side, in education, health, provision of infrastructure and finance, etc., properly belong to the realm of sectoral interventions and largely fall outside the scope of this study. Here, we are concerned with the impact that macroeconomic policy can have on the ability to achieve such sectoral (and more microlevel) changes. Having laid out the broad dimensions of the problem of poverty in Cambodia, we now turn therefore to the macroeconomic issues.
53
2
54
Photographer: Mak Remissa
The Macroeconomic Framework

This chapter poses two fundamental questions. The first relates to the role of macroeconomic policy in Cambodias rather poor performance in poverty reduction during the past decade. To what extent can the situation be attributed to policy decisions? Given the constraints imposed by the legacies of historical events, have policy settings been broadly pro-poor or have they been neutral or even anti-poor? The second relates to changes that might be made to the existing policy settings in order to achieve a better outcome. Even if, for example, current policies are found to be pro-poor, can we find ways to enhance them in ways that the currently rather dismal achievement in poverty reduction can be turned into a outstanding one?
1. THE LEGACY OF CAMBODIAS HISTORICAL EXPERIENCE

The first point to make is that it is difficult to draw any conclusions about the impact of macroeconomic policy on poverty reduction in recent years for three connected reasons. First link in the chain is that the country has suffered from three decades of war and political instability that continued even
The macroeconomic of poverty reduction in Cambodia The macroeconomic framework
after the signing of the peace agreement in 1992 until at least 1998. In practice it is only since December 1999 that all areas of the country have been truly pacified and accessible. It is, therefore, only possible to talk about a stable political environment since 1999. Before that, attempting to disentangle the effects of policy from those of political instability is very difficult. Moreover, we take it as given that a fairly high level of political and social stability is a pre-condition for sustained economic growth and development. Thus while macroeconomic policy between 1993 and 1998 could be applied in secure areas of the country, many of the poor lived outside these areas. Some of the economically destabilising events - notably the bank-financed budget deficit in the election year of 1998 and the flight of capital in 1997-98 - need to be seen in the context of politics. In those years factional fighting affected all areas of the country. Given that a real peace has been established for the first time only after 1998, we have a very short time span, and an absence of good data (see Chapter 1) in which to assess the impact of policy on poverty reduction. Second, the prevalence of warfare, shading into banditry, until the late 1990s meant that any development that did take place was heavily biased towards the safer urban areas. Many rural areas were not only subject to sporadic outbreaks of violence, but were littered with mines which made moving about risky. An extensive de-mining effort has succeeded in clearing all but a few areas. Nevertheless we note that in this years budget for rural development, one third continues to be devoted to mine clearing (MRD interview).
Third, data collection is very unevenly developed. The immediate focus of UNTAC after the signing of the peace accord was on collecting population data. Since the second half of the 1990s technical assistance has been available for improvement of economic data, but these remain poor, particularly in relation to socio-economic data for the rural areas where most of the poor live. Access to rural areas, especially the more remote ones, continues to be hampered by poor transport and communications infrastructure. In the case of trade, smuggling is still rife and much data has to be estimated from partial surveys. Problems with data are discussed at relevant points throughout this report. From the point of view of the macroeconomy, the legacies of decades of war, the Khmer Rouge regime and the breakdown of society that these entailed have resulted in several major challenges. First they have led to an abnormal demographic structure that affects the prospects of many citizens to obtain gainful employment and renders others vulnerable to a vicious downward spiral into poverty. Second, as noted above, they have contributed to urban-bias in economic development that has further reduced the prospects of many to benefit from rapid economic growth. Third, a combination of lawlessness, the lack of a tradition of political accountability and a large influx of funds from donors, have created an atmosphere in which governance is generally weak and there are, as yet, few possibilities for the poor to effectively influence policy decisions. Fourth, one of the hypotheses that we test in this report is that donor dependence resulting from these decades of destruction has led to a case of Dutch disease which
55
further hampers poverty reduction (see Chapter 4). Some of these legacies are canvassed below. Others are dealt with in subsequent sections.
1.1 Population and Labour Force7

Cambodias demographic structure is special on account of the genocidal period of the Khmer Rouge regime, 1975-78, and it has large implications for macroeconomic development. The first general population census carried out since the 1960s was conducted in 1998 and showed a population of 12.4 million. Projected population for 2001 was 13.1 million, based on an estimated growth rate of 2.5% per year. The impact of the Khmer Rouge years shows up very clearly in the almost hour-glass shape of the population pyramid which indicates that only 6.5% of the population were born in the second half of the 1970s, a smaller number than were born during the worst years of the war (1970-75). In contrast 11.8% were born in the 5 years following establishment of the Vietnamese-backed regime in 1979. Due to a post-1979 baby boom, at least half of the population is under 18 years of age and 42% are below the accepted working age of 15. While the birth rate now seems to be declining, these figures give a dependency ratio of 83.2% and mean that over the next decade a very large number of young people will be entering the workforce. In rural areas, where most of the poor are located, the dependency ratio rises to 89.6%, compared with 69.1% in urban areas. Under current
economic conditions, underemployment of the rural labour force is already significant. Large numbers of young rural people entering the workforce are therefore faced with the prospect of unemployment or, more likely, underemployment. Policies that create the conditions for employment generation are therefore an important issue to be faced in the near future. Aside from the economic benefits of growth with full employment, such policies are also important from the point of view of maintaining Cambodias hard won and still fragile political and social stability. Something that is not visible in the demographic data, but becomes apparent upon a short acquaintance with the country, is the impact of the genocide on the quality of human resources. Other 20th century genocides were ethnically and/or politically motivated, but Cambodias genocide was mainly cultural. Although ethnic Vietnamese were targeted, the main victims were Cambodians. These included most intellectuals, especially teachers, and buddhist monks who might hold ideas at odds with the isolationist and peasantoriented revolutionaries. Urban people were turned into farmers and all vestiges of cosmopolitanism and commercialism ruthlessly suppressed. While the decade of Vietnamese occupation resulted in reconstruction of schools and extensive training of professionals and skilled workers (in Eastern Europe), resources for primary and secondary education remained limited. In the 1990s a few former exiles have returned home. Nearly three decades after the genocide, however, Cambodia still
7 Unless otherwise stated, all statistics in this section come from NIS 2002a.
56
suffers more than other developing countries from an acute shortage of people with high levels of technical and professional skills. This shortage is a unique feature of Cambodian society that cannot be solved in the short-term.8 Another reason for Cambodias peculiar demographics has been war. Many men who were of military service age during the years of war, especially the Vietnam War during 1970-75, are simply missing from the population. In 1998, 53.6% of the working-age population (15-64) was female. Female preponderance is more pronounced in the rural areas (54%) compared to 52% in urban areas. Among the under-15s, however, males are 51.1% of the population.9 Women are also 58.7% of the over-64s. The sex ratio in fact deteriorates sharply from the 40-44 age group. Among the approximately 40% of the working-age group who would have been of military age during the war years (1970-75), 59% are female. Even allowing for normally greater life expectancy of women, this is a high figure. The preponderance of women in the older age groups is economically significant because while labour force participation rates begin to drop for both sexes from age 50 onwards, nearly half the relevant age group, including about 35% of women, is still in the workforce at age 70. From the point of view of poverty reduction, therefore, a focus on gender equity is extremely important. Participation in the productive labour force is equal between men and women, at 66% in 1998. Participation rates rise very steeply
from age 12 onwards, so that over 60% of both men and women are working by the age of 20 (rising at a faster rate for women). By age 30 over 80% of women and nearly 100% of men are participating. However, these figures severely distort the picture with regard to womens contribution to the economy. Peak labour force participation rates for women occur during their child-bearing years (which in practice for Cambodian women start at 15). After that they are often engaged in caring for grandchildren while mothers are in paid work. In addition to paid work, they are therefore simultaneously engaged in biological and social reproduction of the labour force. Improved income opportunities for the poor in general, and for women in particular, are likely to reduce the currently high fertility rates. Higher incomes and fewer children will also reduce demands on children for household income-earning activities and provide both the financial capacity and incentive to keep children in school longer. Womens health and nutritional status, education levels and skills development will, moreover, have an important impact on quality of Cambodias future labour force via the socialisation and improved health that they impart to their children. Improving the quality of life for these women is probably the shortest way to reduce underemployment, raise productivity and grow the economy. Investment in womens health and education therefore constitutes an investment of incalculable value to the national economy.
8 Anecdotal evidence suggests that virtually all skilled workers (mechanics, metal workers, carpenters, etc) in Cambodia today are Vietnamese immigrants. 9 In most populations females form a slight majority, but at the previous census date in 1962 males also formed a slight majority, so the current youth-sex distribution appears to be normal. The macroeconomic of poverty reduction in Cambodia The macroeconomic framework
57
However, this value is never included in the GDP because it is not monetised. Where its value to society is recognised, there is a tendency to focus effort on womens role in child-raising at the expense of providing them with equality of opportunity and conditions in the paid workforce. There is a tendency to assume that women are not the main breadwinners (an assumption made, for example, in analyses of the Cambodian socio-economic surveys (CSES), even though this is often not the case. Womens economic contribution is usually invisible, not only for the reason that their work does not appear in GDP statistics, but because they are too busy to participate in the social and political networks and community organisations through which economic decisions are made. Unfortunately for Cambodia there are as yet no time use surveys which would enable at least a time-based assessment of the relative contributions of men and women to the household economy, including reproduction of its labour force. In other LDCs, however, such surveys indicate that women bear a disproportionate burden of work. In farming communities, for example, they are required to exhibit qualities of endurance which are generally counted as less important than the shorter bursts of physical strength required from men. Moreover, in the household economy, especially in farming, it is impossible to disaggregate the contributions of men and women to household income production. Neither men nor women can earn any income at all without the contribution of the other. The assumption that the
household head is the main breadwinner constitutes pure discrimination. In Cambodia, women appear to enjoy an especially low economic status compared to other countries in the region. This is indicated by the CSES in which workers are classified according to their status. In the 1999 CSES, for instance, 58.5 percent of rural men were classified as either employers or own-account workers (nearly all were own-account, reflecting the undeveloped nature of the rural economy), compared with 34.7 per cent of women. Forty-six per cent of women, on the other hand, were classified as unpaid family workers as against 19% of men. Data from urban areas do, however, show less of a bias against women. While the presentation of these women as unpaid family workers may simply be a statistical convenience, it reinforces the often mistaken notion that male heads of these households are the main breadwinners10 and obscures problems of intra-household distribution (of income and labour) which can severely affect opportunities for women to improve their socio-economic status. Women are also very restricted in their access to higher-paid wage employment. In rural areas 9.8% of men are employees compared to 3.7% of women. In urban areas, where most wage work is concentrated, the discrepancy is also very high: 14.9% of women are wage and salary earners compared with 36.6% of men. Men also have more diversified occupations available to them as shown in Table 2.1.
10 Predictably, no such interpretation is made for female household heads, a large proportion of whom are thought to be simply the oldest member of the household.
58
This remains true, even in the little-diversified rural economy, where women are currently restricted to four types of occupation.
Table 2.1: Employment by sex National Women Agriculture and fisheries Service, shop, market sales Elementary occupation Craft & related trades Professionals Technicians & associated professionals Plant & machine operators and assemblers Legislators, senior officials and managers Clerical Other occupation n.e.c. Not stated Total
Source: NIS 2002a.
Rural Men 72.8 4.2 5.0 4.3 2.6 3.5 2.6 1.1 0.2 1.6 2.1 100 Women 87.7 4.1 2.7 2.2 0.6 0.2 0.3 0.0 0.0 0.0 2.2 100 Men 83.5 2.2 3.6 2.7 1.9 1.2 1.2 0.7 0.0 1.0 2.1 100.1
79.0 9.5 3.4 3.1 1.1 1.0 0.3 0.1 0.1 0.1 2.4 100
In summary, Cambodias peculiar demographics are important for this study because they carry major implications for the prospects of achieving growth with equity in terms of both employment generation and gender equality. While lifting people out of poverty is an important end in itself, this study emphasizes that it is also an important objective purely in terms of achieving sustained economic growth. Growth with employment equity can not only ensure that the country does not slide back into a state of social instability that discourages investment, but also maximise increases in productivity by making the best use of Cambodias major productive resource, labour. Most important of all, rising prosperity of the rural areas where 80% of the population live can stimulate demand-led growth, providing a strong
domestic market which will in turn lead to a sustainable export platform. Putting it bluntly, our interest in writing this report is not in viewing Cambodians as cheap labour (the sustainable poverty or trickle-down approach), but in viewing them as potentially prosperous citizens of an emerging democracy with the capacity to influence their own destiny. We think that the latter objective will not be achieved by a strategy that creates an unlimited supply of labour being fed too slowly into low-wage, urban-based industrial and service sectors.
1.2 Governance
Years of war and the Khmer Rouge regime left Cambodia with a broken social structure. During the 1980s, under Vietnamese occupation, considerable progress was
59
made in restoring a degree of social normalcy. However, the economic system established in that period proved unsustainable and a number of reforms were made in an effort to stimulate growth. The Khmer Rouge continued to wage guerrilla warfare throughout this period, in alliance with anti-communist forces, aided and abetted by western and other aid in the context of Cold War politics and Sino-Vietnamese conflict. At the end of the decade the Vietnamese withdrew and the Soviet Union, which had supplied the only significant aid to the regime, collapsed. It was in this context that a UN-sponsored peace agreement came about, a United Nations Temporary Administration for Cambodia (UNTAC) was installed and elections were held. The outcome of these elections was a shared administration between the Cambodian Peoples Party (CPP) and Prince Ranariddhs Funcinpec two parties which, until recently, had been at war with each other. The Khmer Rouge continued their activity, though on a diminished scale and, in light of diminished aid, were compelled to turn increasingly to commercial activities such as export of logs and gemstones. Thus not only did the country remain at war, but the political system proved extremely unstable. Since 1997 the CPP, led by Prime Minister Hun Sen, has been able progressively to consolidate power and, as part of this consolidation, it has also made peace with the Khmer Rouge. Most of our interlocutors expect the stability of the present regime to continue and this is encouraging from an economic point of view. It is worth noting, however, that as in
other countries, the periodic election cycle is also expected to produce some destabilising effects as resources are diverted for election purposes from established development programs. The recent difficulty of forming a government after the 2003 election indicates that the problem is an ongoing one. A number of governance-related issues arise from this complicated political and social environment that affect both the formulation and implementation of macroeconomic policy. 11 While the terms of reference for this study explicitly exclude a focus on governance, we believe that it affects macroeconomics so profoundly that some discussion must be included here in order to provide necessary background to the study of policy making. First the inability of central government to impose its authority in large areas of the country, particularly those bordering Thailand, during the years of war gave rise to widespread smuggling. Illegal exports of raw logs and gems were especially notable, but second-hand right-hand drive cars purchased in Thailand and driven, tax free, across the border are still very commonly found on Cambodian roads. In addition, since neighbouring Vietnam suffered from a similar shortage of goods during the 1980s, a thriving re-export trade in smuggled cigarettes, beer, motorcycles and cars grew up, with Phnom Penh acting as an entrept between Thailand and Vietnam. Efforts to control smuggling have so far been only partially successful. A total ban on export of raw logs has, together with monitoring by the NGO Global Witness, succeeded in
11 Since hard data are difficult to come by, information in this section comes mainly from interviews with donors and government officials.
60
reducing illegal timber exports, while the entrept trade has also diminished in recent years.12 However, smuggling activities continue to deprive the government of much needed revenue and will continue to remain a problem in the foreseeable future due to vested interests of some government officials (see Chapter 5). Second, the large influx of foreign money since the arrival of UNTAC has created massive opportunities for rent seeking among political figures, government officials and others with access to these resources. Real estate prices in Phnom Penh rose very rapidly. Although the average wage of civil servants is currently only $28 a month well below the minimum paid to garment workers, for example (Ngo Hongly interview) - the CSES data indicate that civil servants are generally among the highest income group in Cambodia. Below-subsistence salaries forced civil servants into moonlighting, corruption and other rent-seeking activities which detracted from their effectiveness in carrying out their duties. (In Chapter 4 we also discuss the impact of donor salary supplementation on elite civil servant incomes.) Third, the so-called peace dividend entailed guaranteeing former Khmer Rouge fighters positions in the bureaucracy. More than a few of these relatively recent additions to the civil service are not effective. They are either insufficiently educated or trained or there is insufficient work for them to do. Many of them only turn up on pay day. Politically, the issue of rationalising the civil service remains sensitive. So one option,
being pressed by certain important donors, of switching that portion of recurrent expenditure which goes to civil services salaries to more development-oriented expenditures is currently impossible. In any case, as shown in Chapter 4, its actual impact would be of minimal importance. Similar strictures apply to demobilisation of the armed forces, which now contain many former Khmer Rouge, although it poses a less delicate problem than the civil service. Gradual demobilisation of the armed forces is underway, though it must be emphasised that the savings from demobilisation are largely offset by the redundancy packages being paid to the former soldiers. Fourth, the existence of a large number of weapons in the community has made law enforcement difficult. Tax collectors, for example, have been threatened when attempting to collect taxes. The result is that the domestic private sector is largely untaxed. Most domestic revenue comes from foreign-invested companies. Fifth, the prevalence of rent-seeking among government officials has created difficulties in disbursement of government monies (see Chapter 4).
12 Official estimates of illegal raw log exports fell from $128 m in 1997 to $2.4 m in 2001. The reliability of these estimates is unknown. The macroeconomic of poverty reduction in Cambodia The macroeconomic framework
61
2. THE SOURCES OF GROWTH SINCE 1993

The growth performance of Cambodia over the period 1993 to 2002 has been mixed. Measured in riel at constant 2000 prices, the average annual growth rate of GDP was 6.7% and the per capita figure rose from 907,000 riel to 1,110,000 riel. Measured in dollars at 2000 prices, however, the growth rate was far less impressive, averaging only 2.6% per annum and in per capita terms it fell from $330 to $283 (NIS 2003). The reason for the dollar decline in per capita GDP is related to both the political crisis of the mid-1990s and the wider regional crisis of 1997-98. Because of its high dollarisation, the Cambodian economy was less immediately affected by the regional crisis than other economies, such as Thailand. The riel devalued by a 27% in 1998 alone - most likely related to internal political instability - and was accompanied by a temporary increase in inflation (15%) in that year. Since the poor are relatively insulated from
the dollar economy, the per capita GDP decline in dollar terms will during the mid1990s affected them less than it affected the urban-based dollar economy. While real per capita GDP in riel remained stable during 1996-98, it fell by 43% in dollar terms (NIS 2003). Since 1999 both economic and political conditions have stabilised and growth has resumed. The average real growth rate 1999-2002 was 7.2% in riel and 6.2% in dollars. The global recession has, however, affected the economy with growth slowing in 2001 and 2002 (NIS 2003). The riel/dollar exchange rate has remained relatively stable and, if anything, the post-crisis period has been characterised by deflation.
2.1 Sectoral Pattern of Growth

Most of the growth since 1999 has been accounted for by industry, in fact by the extremely rapid expansion of the exportoriented garment industry. Real sectoral rates of growth (measured in riel) are depicted in Table 2.2.
Table 2.2: Annual average growth rates and shares of real GDP by sector 1999-2001 Source of income Agriculture, fisheries & forestry of which: crops fisheries forestry & logging textiles, wearing apparel and footwear Growth % p.a. 0.3 2.9 3.0 -20.6 20.0 34.4 24.8 6.3 15.6 9.2 43.7 17.6 12.9 6.2 16.8 5.1 3.6 35.6 3.0 5.9 Shares 1998 % 2002 % 33.5 14.9 11.0 2.2 26.3 12.5 6.6 34.3 4.0 6.3
Industry of which: Services of which: hotels & restaurants transport & communications
Source: NIS 2003.
construction
62
Aside from the rapid expansion of the apparel sector, which has increased its share of GDP from 1.1% in 1993, other sectors which have shown above-average growth rates in 1999-2002 include construction, transport and communications, and the tourism-related sector of hotels and restaurants. Taken together, these four sectors have increased their share of GDP from 17.6 to 29.4 per cent of GDP. Table 2.3 shows the sectoral contributions to GDP growth between 1993 and 2001. In the period after 1998, the data show that industry accounted for the majority of GDP growth. Within this sector, the apparel-related industries accounted for 64% of the growth. Services also made a respectable contribution, boosted chiefly by growth in tourist-related services such as hotels and restaurants, transport and communication. Agricultures contribution, on the other hand, was very small and has declined over time.
Table 2.3: Sectoral contributions to growth 1993-2001 percentage points 1993-2002 1999-2001 Agriculture Industry of which: textiles, apparel, footwear Services
Note:
Cambodian development. The rapid growth of garments, in particular, has been a plus for export income and non-farm employment generation - both direct and indirect. There are, however, two caveats which need to be made here. First is that growth has been remarkably narrowly based. Over a third of all growth in the period 1999-2002 came from the garment industry. In addition, there has been a rising dependence on expansion of garment production over time. This industry is currently in a rather fragile state, as discussed in Chapter 5, and, as presently constituted, may not be able to continue its growth-leading role. Moreover, the industry is heavily concentrated in and around three urban centres of Phnom Penh, Siem Reap and Kompong Som (Sihanoukville) and thereby increases the urban bias which Cambodias pattern of growth has exhibited so far. While tourism, construction and transport and communications have also shown above average growth rates, they are similarly heavily concentrated in a handful of urban centres. Second, agriculture, forestry and fisheries continue to provide a large share of GDP at 33% and they also employ over 70% of the population. As noted in chapter 1, poverty is overwhelmingly concentrated in these rural sectors, particularly agriculture. In the latter case, growth has been unstable. The shares of agriculture, livestock, fisheries and forestry in GDP have all fallen markedly since 1996. In the case of forestry, this has been the result of policy directed at ending the mostly illegal trade in raw logs, but growth rates in farming (crops and livestock) have shown a high degree of instability.
18.5 43.4 27.4 28.7
1.8 56.0 35.7 30.5
These sectoral contributions do not add to 100% because growth in net taxes and imputed bank charges are not included. Source: Calculated from NIS 2003: 437.
Such a pattern is, of course, consistent with historical development patterns in all developing countries, including todays developed countries. In this respect it is a positive from the point of view of overall
63
These sectors are severely affected by natural conditions due to low levels of water control. In 2000, for example, there were severe floods. Other years, including 2002, were affected by drought conditions or late arrival of the monsoon. Increased investment in agriculture - in raising productivity, rendering output less vulnerable to natural conditions, encouraging expansion of market-oriented production (especially through infrastructure improvements) and increasing value added in rural production by developing processing and related services - will both widen the base on which growth occurs and have maximum impact on the poor. As noted above, economic growth since 1993 has been heavily biased towards urban areas. There are no available statistics on the breakdown of GDP by region, but using per capita household expenditure data from the socio-economic surveys as a proxy, it can be seen that not only are incomes in Phnom Penh much higher than rural incomes, but the gap has widened over time. Whereas average monthly expenditure per head was 3.1 times higher in Phnom Penh than in the rural areas in
1993-94, by 1999 it was 3.6 times higher (Figure 2.1). The gap between Phnom Penh and the other urban centres has also increased over time.
Figure 2.1: Per capita monthly household expenditure 1993-94 and 1999
1993-4 1999
250000 200000 Riel 150000 100000 50000 0

Phnom Penh Other urban Rural
Source: CSES 1993-94, 1999.
Underemployment is a more serious problem in Cambodia than unemployment. Table 2.4, based on the 2000 Labour Force Survey, gives some indication of the extent to which labour resources are under-utilised. It shows that 5.3 million persons aged 10 and over were employed. Of these, 1.66 million (31%) indicated that they were available for extra work and over half a million were employed for less than 30 hours during the survey week.
Table 2.4: Employed persons 10 years and over, 2000 Employed Total 000s Urban Rural Total 614.7 4660.5 5275.2 Female % 48.6 52.3 51.9 Working <30 hours Total 000s 58.7 491.2 549.9 Female % 59.9 51.0 51.9 Available for extra work Total 000s 94.2 1567.2 1661.4 Female % 39.3 44.8 44.5 'Unemployed' housekeepers Total 000s 101.2 315.4 416.5 Female % 92.8 89.7 90.5
Source: NIS 2002a: 44-45.
64
2.2 Financing of Growth

Economic growth comes chiefly from investment that leads to more employment and from changes in the capital-output ratio. The latter can be achieved by the introduction of more productive technologies or through better utilisation of existing labour resources - for example, by better organisation without the addition of extra capital equipment. Total investment (both public and private) has grown at 12.3% a year since 1993 (RGC
2002b: 13). As a result, as Table 2.5 shows, investment as a share of GDP has risen fairly steadily, with private sector investment in particular showing improvement over time, despite the crisis of the late 1990s. Foreign private investment has tended to grow slightly faster (13.6%) than domestic private investment (10.1%). Capital-output ratios are rather high for an economy with surplus labour: $460 of investment were required to produce $100 increase in output during 1994-2001.13
Table 2.5: Investment share of GDP 1993-2001 (%)
1993 Total Public Private Domestic Foreign 13.9 3.9 10.0 7.7 6.3 303
1994 17.8 5.4 12.4 10.0 7.8 429
1995 19.9 6.8 13.1 9.0 10.9 614
1996 25.4 7.5 17.9 10.9 14.4 800
1997 20.5 4.9 15.6 11.6 8.8 623
1998 21.9 5.8 16.1 12.6 9.4 612
1999 20.8 6.3 14.5 12.2 8.6 635
2000 22.1 7.4 14.7 13.0 9.0 682
2001 23.3 8.0 15.3 12.3 11.0 769
Total ($US M)
Source: RGC 2002b: 13.
Figure 2.2: Investment by source 1993-2001
RGC 2002b: 13.
Foreign savings provided the majority of investment until 1996, shared roughly 50:50 between public (foreign aid) and private investment (Figure 2.2). Since the crisis of 1997-98, however, domestic investment has increased its share. The level of domestic private investment stagnated during 199699, but has grown quickly since then, while foreign private investment recovered less rapidly. Domestic public investment has also grown since 1998 as a result of the small surplus on the recurrent account.
13 Calculated from RGC (2002: 13).
65
The investment data show that dollarisation has been successful in stemming capital flight. Table 2.6, showing financial aspects of the balance of payments illustrates this net private transfers and investment have been positive every year since 1994. However, the table also shows the fragility of confidence in the Cambodian economy as net private financial inflows have reduced substantially since the mid-1990s. The current account deficit, which grew rapidly in 1994-96 and then dropped away after the 1997 crisis, has been largely offset by official transfers and loans. The private sector contribution has been far smaller although a rising proportion of the financing of the deficit prior to 1997, it has dropped away again since then. We should note here that net private sector transfers include only those recorded by the National Bank of Cambodia. An unknown amount is transferred in and out of the country by illegal means.
1994 -368 -254 -87 -47 2 -49 -15 20 283 16 13 3 69 -66 36 23 296 7.8 1995 -443 -332 -74 -57 10 -67 -18 20 335 110 61 49 151 -101 -13 69 396 17.4
Both official and private sector transfers and investments showed the loss of confidence in Cambodia in 1997. Official flows have revived, however, as stability has been restored, but private sector inflows have continued to diminish, possibly in response to global recession. Since most FDI comes from within the Asian region, it is also likely that much investment in the early 1990s was a response to the bubble economy developing elsewhere in the region. Cambodia is not an inherently attractive country for foreign investors suffering as it does from poor infrastructure, a low-skilled workforce and a recent history of instability. For this reason, one could not reasonably expect post-crisis private FDI to return to pre-crisis levels in the near future. It is important, therefore, for Cambodia to raise its domestic investment effort if povertyreducing growth is to be sustained.
Table 2.6: Balance of payments 1994-2001
Current account Trade balance Net services Net income receipts payments of which: interest Net private transfers Net official transfers Capital account Official sector loans* Net non-official investment FDI net Other investment net Overall balance Net private transfers and FDI Net official transfers and loans Ratio of private to official %
Source: Calculated from NIS 2002: 252.
1996 -546 -428 -52 -86 13 -98 -19 20 437 259 90 170 294 -124 72 190 527 36.1
1997 -268 -230 -43 -53 16 -69 -13 60 288 164 41 123 168 -46 33 183 329 55.6
1998 -207 -173 -66 -33 18 -51 -17 66 199 145 43 101 121 -20 21 167 242 69.0
1999 -259 -275 -11 -41 20 -62 -15 70 208 124 44 80 144 -64 51 150 252 59.5
2000 -257 -263 -12 -51 32 -82 -19 72 271 99 75 24 112 -87 89 96 346 27.7
2001 -217 -226 -20 -41 25 -66 -17 70 274 75 68 7 113 -106 71 77 342 22.5
66
3. RECENT FOCUS OF MACROECONOMIC POLICIES 3.1 The Donor Perspective

It has become widely accepted, particularly within the Bretton Woods institutions, the UN and WTO, that poverty reduction is best addressed by economic growth. This argument follows that presented in a paper by David Dollar and Aart Kraay (Dollar and Kraay (2000) that incomes of the poorest quintile rise in line with overall growth. Their conclusion was that policy-induced growth is as good for the poor as it is for the overall economy - encapsulated in the phrase a rising tide lifts all boats. Moreover, their paper argued that the openness of the economy benefits the poor to the same extent that it benefits the whole economy. The Dollar and Kraay thesis has had a significant impact on donor policies in less developed countries like Cambodia. First, there has been renewed emphasis on policies which are believed to be growth promoting. These largely involve embracing economic liberalization in order to promote opportunities for the private sector to grow. We note here that the implied association between liberalization and growth is not well established for less developed countries.14 The literature on East Asian development (Amsden 1989; Wade 1990) has posed a significant challenge to this assumption, at least for the early stages of development and there is certainly no consensus among economists as to the universal growthinducing benefits of liberalization. Nevertheless, in the hands of the Bretton
Woods institutions, particularly the IMF, this type of approach has led to a policy focus on liberalization accompanied by macroeconomic stabilization, especially elimination of inflation and cutting budget deficits both to reduce inflation-promoting expansion of the money supply and to reduce public debt to eliminate the drain on public development resources posed by high debt service levels. To complement this focus on stabilization, recent literature has focused on the need for effective institutions as a necessary prerequisite for liberalization to generate growth. In this view, institutional reform becomes a major focus of technical assistance, seen largely as a technocratic problem of training and developing computerised systems. The approach is well illustrated in the Cambodian context by the technical assistance currently being provided to the Ministry of Economy and Finance, Customs Department and Treasury, to improve systems of revenue collection and expenditure disbursement. Civil service reform, aimed at improving the efficiency of civil service operations is another major focus of donor policy in Cambodia. Second, there is a recognition that liberalization alone can lead to growth without significant poverty reduction, if the poor are unable to take advantage of the opportunities arising. Poverty reducing interventions have therefore tended to focus on human resource development and improving access of the poor to capital markets. In particular, the strong association between poverty and low education levels
14 Lbker et al. (2002) have pointed out that the Dollar and Kraay paper lacks a theoretical basis, in particular, there is 'no obvious causal mechanism by which average income would transmit its increase to one of its components'. Since, in the absence of major policy changes or external shocks, income distribution tends to change only slowly, the result is also not very interesting. The most disturbing element they find, however, is the inference that because 'growth is good for the poor', so-called 'growth promoting' policies of the Bretton Woods institutions are universally applicable.
67
as well as poor health has led to an emphasis on concentrating expenditures on improving access to education and health services. To complement this, there has been increased emphasis on micro-credit schemes which are designed both to mobilize the savings of the poor into the quasi-formal financial system and to provide credit to the poor on terms that are more favourable than those available either from the formal banking system or informal money-lenders.
3.2 Implementation in Cambodia

In Cambodia, these policies have been implemented in two ways. First, at the macro level policy has focused on stabilization measures. This focus has been justified on the grounds that there are inflationary pressures present in a resource-constrained Cambodian economy which have been and also could be detrimental to growth and particularly to the living standards of the poor. The policy has responded to two inflationary shocks to the economy in 1990 - caused by inflationary deficit financing in the wake of Soviet aid cancellation - and again in 1997 - caused by loss of confidence during the political crisis and sharp devaluation of the riel. One thing that remains unclear is the extent to which policy in the intervening years was successful in stabilizing prices or these shocks would have had temporary effects anyway. Certainly the inflation of 1998 was not excessive by developing country standards. A question which is addressed in this report is whether the focus on stabilization has outlived its usefulness. As discussed in Chapter 3, the use of monetary policy is
restricted by high levels of dollarisation of the economy, but a number of elements of the IMF policy package adopted in Cambodia are additionally biased against growthinducing policy. In particular, increases in the money supply have been used to build up international reserves rather than to increase domestic assets. Although international reserves are now in a healthy state, the policy package implemented under the PRGF prevents further increases from being diverted to domestic asset accumulation. The central bank is unable to lend money to the government and the government is unable to borrow from the private sector. While there is excess liquidity in the commercial banking system, these funds cannot be mobilised for domestic investment - for reasons discussed in Chapters 3, 4 and 6. Anti-inflationary monetary policy thus places an important restriction on the potential for using fiscal policy as a means for inducing economic growth. As a complement to the stabilization measures, donor policy has focused on technical assistance in the area of institutional reform as the chief means for improving revenue-raising and disbursement capacity, reducing rent-seeking and establishing the rule of law. Second, since 1999 budget allocations to health and education have been substantially increased (see Chapter 4) and, at the sectoral level, the Ministries of Education and Health have received extensive technical assistance in establishing sector-wide development programs. While the government has also declared agricultural and rural development to be priority areas, in practice these two sectors have received
68
few funds or technical assistance. Non-governmental organisations have been extensively mobilised to provide interventions at the grassroots level. A large majority of NGOs operating in Cambodia are also providing assistance in the health and education sectors. To date, the micro-credit sector has received less attention. There are, however, several NGOs operating such schemes, chiefly as adjuncts to their health and education programs (see Chapter 6), and the recent establishment of the Rural Development Bank (RDB) reflected an intention of the government to channel funds towards further development of this sector. It should be noted that these interventions are generally gender-sensitive, thus addressing one of the major issues posed by Cambodias special demographics outlined above. (However, in the case of education, which is targeted at the young, the needs of poor working-age women are not addressed.) They are also appropriately targeted at increasing the skills base of the population, which deals, over the longer-term, with the unique impact of the Khmer Rouge genocide. Improved health and education are, moreover, strongly associated with reductions in non-income poverty by providing people with access to information and ideas, enabling their more active participation in social and political affairs thereby improving the quality of democratic institutions, and generally increasing self-esteem and self-confidence. There are also impacts on reduction of income-poverty: a healthier workforce is a more productive workforce, an educated population has better access to new technologies and the ability to access key market-related information.
3.3 An Alternative Perspective

3.3.1. The Potential for Demand Expansion A major concern for the authors of this study is that the policy package focuses too strongly on the supply side and not enough on the demand side. In other words, growth is expected to follow, with the poor benefiting in tandem, from a combination of improvements in market facilitating interventions (liberalization), better provision of government and financial services (institutional reform) and human resource development (a healthier and better educated labour force). Without challenging the need for such supply-side changes, we suggest that more consideration should, at this stage of Cambodian development, be given to the alternative view emphasising. Keynesian-type demand expansion as the principal means to achieve growth with significant poverty reduction. Too much focus on the supply side poses a danger that employment creation will be insufficient for the population to take advantage of improved health and education and that those best able to use these advantages will be the already much wealthier urban population. Improved supply of health centres and schools are of little use if the poor cannot access them through either lack of income or lack of transport facilities. High repetition rates in primary schools at present are one indication that, while there is a theoretical demand for education, practical needs of income generation are leading to poor attendance rates of enrolled children. Young girls are especially likely to be kept out of school to
69
help their over-burdened mothers with income-earning work, care of infants and household chores. Low rural incomes mean that urban in-migration, which has accelerated in the recent past, will continue at a high rate and pose problems of inadequate urban infrastructure, expansion of slum areas and downward pressure on wages. We have already seen that growth since 1993 has not produced significant poverty reduction and has increased inequality particularly along urban-rural lines (Chapter 1). Cambodias only major export commodity - garments - has developed with limited employment impact in enclaves within three major urban areas and on their outskirts (Chapter 5). During the 1990s, 40% of the growth in global demand was accounted for by growth in US demand. While the worlds largest economy has a generally strong domestic market, most developing countries, on the other hand, are dependent on external demand. This dependence is well illustrated in the Cambodian case, where external demand for garment exports to the US and EU provided the majority of economic growth in the 1990s. Strong growth of domestic demand would help to limit, though of course it would not eliminate, the impacts on Cambodia of external shocks such as recession in the US. We can confirm that Cambodias good growth rates, with the exception of the 1996-98 period, have been mainly driven by external demand. Growth of domestic demand has been minimal. In particular, real household final consumption expenditure has stagnated during 1993-2001, growing at only 2.0%
which means that it has actually declined in per capita terms.15 This figure is a further indication that significant poverty reduction has not occurred. Indeed, since consumption expenditures of some sections of the population have clearly increased, it suggests that the situation of those who remain poor has significantly worsened. Trickle down effects of expansion in external demand have clearly been insufficient to produce poverty-reducing growth. Policies which are specifically targeted at stimulating domestic demand might, therefore, have a more significant and rapid impact on both poverty reduction and rising inequality. Most importantly, they could lift the poorest sections of society out of poverty at a faster rate. Reasons for introducing more expansionary policies at this stage include the increasing political stability of the country, the impacts of external recession and the absence of serious inflationary pressures. At the time of writing Cambodia is suffering from the impacts of global recession. Although growth has been too slow so far to produce significant poverty reduction la Dollar and Kraay, it has slowed further under the impact of recession in the United States and elsewhere in 2001. Inflation would appear to have all but disappeared. After rising to 8% and 14.7% in 1997 and 1998 (year averages), inflation subsided to 4% in 1999, - 0.8% in 2000 and 0.6% in 2001 (NBC data). In 2002 the year on year rate peaked at nearly 4% as commune council elections increased the flow of cash
15 Assuming the population growth estimates of 2.5% or more are correct.
70
to the population. The national election in 2003, however, did not generate inflation and prices have remained very stable since mid-2002 (NBC data). The consumer price index is only calculated for Phnom Penh, but a study by the Cambodia Resource Development Institute found a decline of 1.6% in the rural areas (calculated in riel) for 2001 (CDRI 2002: 17). From a growth point of view, such a rate can scarcely be considered dangerous and indeed deflation or zero inflation is not conducive to economic growth. Further, from the point of view of poverty reduction, the price stability of recent years has not been helpful. For example, from December 2000 to December 2001 the price of rice in Phnom Penh markets dropped by 19% and those of pork and chicken by 7 and 15% respectively. Food prices fell by 3% (CDRI 2002: 17). These price falls reflected lower prices paid to producers, that is, the farmers who constitute most of the poor. 3.3.2. Potential Obstacles to Demand Expansion Policies A number of questions arise in relation to any proposal to increase domestic demand by direct intervention. The first of these is the argument that demand expansion policies in a resource-constrained developing country would immediately come up against supply constraints and lead to inflation. In the Cambodian case this is a dubious argument, although care would need to be taken to achieve demand expansion in the right sectors, as outlined below. Cambodia is not a country that is producing at its production possibility frontier. The single
factor of production that is severely constrained is physical capital. Land is generally under-utilised in the sense that far greater output could be achieved by increasing yields per hectare. In rice production, for example, average yields are only 2 tonnes/ha compared to 4.4 in more densely populated Vietnam. Intensification of production could go far towards increasing output. Labour is also an under-utilised resource. Under-employment is prevalent in the rural areas and is likely to increase in coming years due to the impact of the post-Khmer Rouge baby boom. Expansion of demand, therefore needs to focus on increasing the utilisation of land and labour, particularly by inducing labour-intensive development of agriculture and providing labour-intensive non-farm employment for the rural surplus labour. Increased use of excess labour supply requires a focus on labour-intensive employment creation. In our extremely limited experience of village-level Cambodia, this also happened to be a major demand of the rural population. When we visited a drought-affected village in Kompong Speu, the meeting of villagers which we attended was overwhelmingly in agreement that their primary development need was water, for both irrigation and drinking. Water control would enable them to use their available farm labour more intensively for both crops and livestock. When we asked about roads, they said they wanted employment in road construction. In other words, they were not thinking about roads as a means of transport, but as a means of putting cash in their pockets. Only when they have money to spend, will they be able to think about market-based opportunities to expand their income further.
71
Labour-intensive road construction and maintenance are currently being supported in Cambodia through projects operated by WFP (food for work) and ILO. These programs focus on tertiary and sub-tertiary roads, rather than the highly capital-intensive works being constructed by the Japanese (often also with ADB funding). While the latter are obviously needed, especially for exportoriented growth, the former can have a more immediate impact on poverty-reduction. The WFP program has been criticised for the poor quality of road construction (laterite roads which deteriorate quickly in the wet season), but this is a criticism which again focuses on the supply side rather than demand. In the food deficit areas where the WFP program is carried out, the payment of wages in the form of food, can significantly affect the health and productivity of workers, enabling them to increase their incomes in other ways. They can also shift expenditure from food to other items such as clothing and shelter or even farm inputs. While it is a short-term solution - poverty alleviation rather than poverty reduction - it is far better for the poor than no increase in their income at all. To date, these programs have only been able to reach a small proportion of the population and, while rural development programs have been designated by the government as a high priority area, little funding has been available. In practice, funding for such priority programs is donor-driven. The reasons given to us, by all donors as well as officials at MEF, for the increase in funding to the health and education sectors was that these ministries had demonstrably better capacity to design and implement sectoral programs and budgets. However,
the principal reason for this better capacity is that the ministries have been targeted for technical assistance by donors (see Chapter 4). By contrast, the Ministry of Agriculture, Fisheries and Forestry and the Ministry of Rural Development have received far less technical assistance and correspondingly lower budget allocations. Improved water control is another way in which the labour intensity of rural output could be rapidly increased. At the present time, few Cambodian farmers have access to irrigation systems. Most farmers are therefore highly dependent on rainfall and vulnerable to drought and floods. They can grow only one crop a year, during the monsoon season. Yields are low by regional standards. They have, however, shown a tendency to increase in recent years. Figure 2.3 compares the performance of wet and dry season rice production and yields. The area devoted to dry season rice has risen from 8 to 14% of the harvested area since 1993. While yields of rain-fed wet season rice have increased during the period, the harvested area has fluctuated resulting in variable production. For those farmers able to use irrigation and produce a second crop, the yields on dry season rice are not only higher, but provide a more stable output. Thus, while total production of wet season rice rose by 45% over the period, that of dry season rice rose by 142% (NIS 2003: 212).
72
Figure 2.3: Harvested area and yields of wet and dry season rice 1993-2001
Source: NIS 2003: 212.
The second argument against demand expansion, is that putting cash in the pockets of the rural population will increase the demand for food and, given the low productivity, subsistence nature of most agriculture, there will be an insufficient supply response. Rice prices will therefore rise, leading to further inflationary pressures and an adverse impact on the poor since food will become less affordable. There are two arguments to counter this. First, generally speaking food security is not a major issue in Cambodia. Per capita output of paddy has averaged over 300 kg since 1998. While there are areas of hunger, particularly in the pre-harvest months, the distribution of food is a more important question than the total supply. In recent years, food surplus farmers have actually begun to export rice - mainly to Vietnam and
Thailand. There is nothing to stop Cambodia from exporting its high quality, traditional varieties to the world market and importing cheaper low quality rice to ensure food security for the poor. Second, it is not at all clear that increased demand from consumers would produce an insufficient supply response from farmers. There is, in Cambodia, little evidence to support this argument, as shown in Figure 2.4. This shows that output of paddy has been reasonably responsive to changes in price. It appears that output responds better to price increases, suggesting that for food security reasons, farmers do not decrease output when prices fall. The picture does not suggest, however, that there would be no supply response to an increase in prices brought about by rising demand.
73
Figure 2.4: Changes in paddy output and prices 1993-2001

Output (1000 t) 200 180 160 140 120 100 80 60 40 20 0
1993 1994 1995 1996 1997 1998 1999 2000 2001
Price
made, farm-gate prices are correspondingly reduced and it is the farmers who bear the brunt of the extra costs (estimated by CDRI at about $10-15 per tonne) (CDRI 2001: 40-41). At the moment, nearly all exports are of unmilled paddy to neighbouring Vietnam and Thailand, two countries which have an important effect on world market prices. Cambodian exports therefore remain vulnerable to supply and demand conditions in these two countries. Increasing the value added by milling rice in the country and removing the barriers to export posed by irregular customs, inspection and handling costs would further improve the prospects for development of the export trade and enable the export of high quality rice in exchange for lower value rice for domestic food security. We conclude from this discussion that demand expansion policies are not likely to create an inflationary supply bottleneck which would adversely affect the food security of the poor. Instead farmers are likely to respond positively to higher prices, at least maintaining per capita food security. We should be careful, however, to qualify this conclusion by pointing out that, given the current surplus rice production, further increases in output are likely to be affected by export prospects. Demand expansion policies therefore need to be carefully co-ordinated with improvements on the supply side, including measures which produce less fragmentation of the rice market (linking current deficit areas to the national market), improved technology in farming and processing and reduced barriers to production for export.
Source: NIS 2002a; NIS 2002b.
Indeed domestic rice prices have, with the exception of the crisis period of 1997-98, broadly followed the downward trend in world market prices (Bangkok f.o.b.) since 1995 (CDRI 2001: 36). Thus, although there may be areas of subsistence farming (in the sense that farmers produce only for their own consumption), at the national level Cambodian agricultural output is affected by external conditions. Between 1995 and 1999 Customs data recorded the export of over 1.4 m. tonnes of paddy. The figures ranged from a low 2.9% of domestic output in the crisis year of 1998, to 11.7% in 1999. During the same period, 116,203 tonnes of milled rice (or 187,424 tonnes of paddy equivalent) were legally imported, creating a surplus to total consumption requirements. A conference of rice millers held in May 2000 identified irregularities in the management of the rice trade as one of the main constraints on rice exports from Cambodia and it must be assumed that this results in significant evasion measures by rice and paddy exporters. Illegal trade is widely believed to be much higher than the recorded trade. Moreover, where irregular payments are
74
Index (1993 =100)
The third argument against demand expansion involves the balance of payments constraint: given the supply constraints, increased incomes would simply suck in more imports and increase the current account deficit. We have already suggested that the supply constraints might not be as important as this argument suggests, provided the increased demand is concentrated in rural areas where both poverty and lack of supply bottlenecks are most evident. There is great potential to increase the export of raw and semi-processed agricultural products in the near term. Policies targeting development of rural areas, especially agriculture and agricultural processing, would also help to reduce the countrys current over-dependence on apparel exports (see Chapter 5) and simultaneously reduce income poverty in the areas where 90% of the poor are concentrated. Finally, demand expansion is often regarded as a dangerous course because it would crowd out private sector investment. The crowding outhypothesis is a key assumption of the IMF program, but there is also a possibility that public investment may crowd in private investment if it is well-targeted to increase investment opportunities in areas currently operating below capacity. In theory, crowding out will only occur when the economy is operating at or near full capacity and this is not the case for Cambodia. Three key points are worth flagging here: 1. There would appear to be no evidence that public expenditure has, in the past, crowded out private domestic investment. Figure 2.5 shows that public and private investment have in fact moved in line with each other.
Figure 2.5: Public and private investment 1993-2001
Source: RGC 2002b: 13.
2. Crowding out is more likely to occur if government spends on consumption rather than investment. 3. There is excess liquidity in the banking system. Public investment would not, therefore, divert savings from private sector uses. These points are discussed in considerably more detail in Chapter 4.
4. CONCLUSION
To date, the macroeconomic framework established under guidance from the IMF has not succeeded in addressing the key elements of poverty in Cambodia. To be sure, it is difficult to disentangle the impacts of macroeconomic policy from those of the political instability that affected the economy before 1999. The sharp fall in investment and output in the mid-1990s, for example, is more likely attributable to political events than to macro policy as such. However, the evidence of Chapter 1 suggested that even during high growth periods, poverty reduction either did not occur or was minimal. The question that we wish to pose in this report, therefore, is what changes could be made to the macro-policy settings that would
75
be conducive to a more rapid reduction in poverty and improved opportunities for the vast majority of Cambodians to participate in the benefits of this growth? We showed that growth in recent years has been very strongly influenced by expansion of the export-oriented garment industry. There are reasons, which we discuss at length in Chapter 5, to suppose that the rapid expansion of this sector is unlikely to continue. It is imperative, particularly in the context of the current uncertainty in the global environment to find ways of diversifying the sources of economic growth (including export growth) in order to absorb more of the excess capacity which presently exists. Land, labour and finance can all be used more effectively than they are now. Up to now, the policy framework has served the country well, in the sense that the focus on tight monetary and fiscal policy has helped to eliminate excessive rates of inflation and generate some confidence in the authorities. Dollarization has also assisted in this process. In more recent years, however, there has been deflation and in the context of global recession and uncertain recovery, the stabilization package would appear to have outlived its usefulness. At the meso-level, sectoral policies implemented under donor guidance have largely
addressed the problem of human capital, particularly the health and education sectors. These policies do go some way towards addressing poverty reduction, in the sense that they are generally gender sensitive although the education policy is unlikely to assist the large proportion of older women in the workforce - and will be helpful over the long term in overcoming the extremely damaging legacy from the years of war and genocide. They will, however, have a very limited impact on income poverty unless there is a concurrent expansion of investment and employment opportunities for the newly trained and healthy workforce to take up. A policy of demand expansion seems to us to be a good solution to this problem. There are no simple solutions, however. In particular, there are legal and institutional obstacles to a demand expansion policy, partly due to the historical legacy and partly due to the policy framework established under the IMF structural adjustment program. Many of these policy-driven obstacles are well founded, especially as confidence in the authorities and in the financial sector is only gradually improving. They should not be undone in haste, but they should be undone over time. In subsequent chapters, we suggest a number of ways in which these issues might be addressed.
76
3
78
Monetary and Exchange Rate Policy

1. INTRODUCTION
Within the RGC, sustainable growth is seen as a crucial means of increasing peoples incomes and attaining nationwide provision of basic public services, particularly health care and education. While, on the one hand, it seems to be obvious that policies such as public expenditure management can have a direct impact on poverty alleviation through, among others, the pro-poor targeting of government resources, the issue of monetary policy and exchange rate regimes is often seen as only very indirectly related to the issue of how to be pro poor and how to balance equity and efficiency goals at the national level. However, the recent history of huge swings in exchange rates that provoked the Asian regional crisis, has directed thinking toward the close connection of the monetary system to social outcomes, as the costs of the collapsing exchange rate regime fell heavily on the poor. Consider the case of Thailand. According to Thailands National Economic and Social Development Board the countrys poverty level had fallen to 12% before the onset of the crisis in 1997. Nonetheless, as a result of the
The macroeconomic of poverty reduction in Cambodia Monetary and Exchange Rate Policy
crisis, poverty has re-emerged as a central development concern. In Cambodia, Economic Watch-Indicators produced by the CDRI showed that average daily earnings of workers who normally earn in local currency, such as cyclo drivers, small vegetable sellers, rice-field workers, skilled and unskilled construction workers, dropped significantly after the crisis and have taken several years to begin to recover (Cambodia Development Review, July-September 2002). Observations of various economies showed that economic growth is associated with financial development, although such a correlation does not identify the causation. Therefore, the role of a central bank to create a sound, efficient, and integrated banking system as well as to mobilize financial resources to finance productive investment cannot be overemphasized. While banks play a very important role in creating money for an economy to run efficiently, central banks are the primary agent of monetary policy, through their actions to influence the flows of money and credit and, therefore, the entire economic environment. Price stability is not only important for an optimal allocation of resources, but, from the point of view of poverty reduction, it is especially vital to the livelihoods of the people, in particular the poor. If inflation is high the poor will be the first group to suffer from loss in purchasing power. The point is not that the authority has to fight inflation at any price and to value only anti-inflation policy. Rather, it has to face the tough question of how far to go with fighting inflation, knowing that with ongoing deflation an economy might face greater risk of entering into the chain of rising unemployment,
falling demand, and reduction in the level of national income. Money, on the other hand, is the key to economic activities as it forms the basis for various financial and business transactions. It is prerequisite to stability of daily life and economic growth that people can use money every day with full confidence in its value. Few will dispute the claim that there will be catastrophic economic consequences if people stop exchanging goods and services because the pieces of paper they receive have no real economic value. In this context, therefore, the issue of the credibility and effectiveness of the existing monetary and fiscal policy emerges in the foreground. Furthermore, there is large uncontrollable use of the US dollar in the Cambodian economy, which has major impact on governments macroeconomic policies, business environment and well-being of the general public. While large business transactions are conducted in foreign currency and the economy is largely based on the US dollar, the riel remains the currency of the poor. Maintaining the value of the riel is thus very important for the living standards of the poor. All these factors imply that, from the point of view of poverty reduction, government policy needs to maintain a stable macroeconomic environment, while on the other hand it has to look for a mechanism to ensure efficient mobilization and allocation of resources, which will in turn generate the basis for the distribution of wealth. Monetary and exchange rate policy play a very important contribution in this and
79
should be thus mainstreamed into the macroeconomics of poverty reduction. This chapter aims to highlight key issues related to the conduct of monetary policy and the implications of dollarization from the point of view of economic stabilization, growth and poverty reduction.
iii. A low level of intermediation. As measured by the ratio of banking sector assets to GDP, the degree of intermediation was 18% on average during 1999-2001. The provision of banking services is thus relatively weak in regional comparison and indicates the need for further development of the banking sector in Cambodia. iv. A constraint on efficiency and profitability. Figures on bank deposits and loans illustrate that the ratios of total customer deposits and of total credit outstanding to GDP remain comparatively low at 9.6% and 6.1% respectively as of the end of 2001 (NBC Bulletin 2002). The banking sector thus operates with excess liquidity that could be utilized more effectively to facilitate further growth in the private sector, or to support the governments policy of poverty alleviation. Lack of credit culture and weak legal and regulatory foundation for extending credit have been identified as a major cause for the slower credit growth behind deposit growth (ADB and Kingdom of Cambodia 2001). At the same time, due to the absence of an interbank market, banks are holding a high percentage of reserves (cash plus total deposits with the NBC), reaching about 29% of total bank assets in 2001. v. A large intermediation spread. The average spread between deposit and lending rates for the banking system as a whole ranges between 12%-15% due largely to high risk and operating costs. This inhibits potential investment growth and represents a great barrier for the poor. vi. Inadequate geographical distribution of services. Banking services are available only in very few existing business centres where
2. BASIC ENVIRONMENT SURROUNDING THE CONDUCT OF MONETARY POLICY 2.1 Institutional Constraints
As a legacy of prolonged internal strife, the history of high inflation and instability in the foreign exchange markets, the banking sector in Cambodia suffered a low level of public confidence and the economy therefore operates largely on a cash basis. Main characteristics of the current banking sector can be summarized as follows: i. General lack of confidence. As a legacy of devastation, including abolition of the banking system, and civil strife during the past three decades, confidence in the banking system was low and further complicated by lack of trust in the national currency. ii. Relatively high concentration and low competition. There are vast differences in size among the 14 commercial banks that now operate in Cambodia. For illustration: as at the end of June 2002, the 5 top banks represented about 70% of total bank assets, absorbed 80% of customer deposits and provided almost 70% of total credit.
80
they offer a narrow range of services and products to a narrow portion of the population. The mass of the population has little knowledge about banking services and limited outreach to the poor remains an important outstanding issue. All these factors contribute to the high degree of cash transactions. Between 1999 and 2001 the share of domestic currency in circulation was on average more than 80% of the total issued. The very thin financial market, coupled with dependence on cash in the real economy, clearly indicate a restraint on the transmission mechanisms of any monetary policy, since monetary policy normally affects the economy through banks.
2.2 Dollarization of the Economy

Between 1975 and 1979 the Cambodian economy fell under the control of the genocidal regime of Pol Pot, a regime that overwhelmingly destroyed the infrastructure of an already weak economy. Under this regime there was no market, no trade. Money and banking were abolished. The NBC building was blown up. In the early period after the Pol Pot regime was ousted in 1979, commercial transactions were mainly conducted in the form of barter. The rest was settled using Vietnamese dong and gold. The latter was used as unit of account, medium of exchange and savings instrument. Subsequent to the re-establishment of the NBC in 1980, the riel was re-issued and circulated as the countrys legal tender. However, confidence in the riel was low, as the political structure and security situation remained unsettled.
Use of the US dollar and Thai baht was originally restricted by the closed nature of the centrally planned economy and by the fact that the country was linked politically to the socialist bloc. However, the use of foreign currencies has increased gradually since the country opened its door to the outside world. US dollars and, to a lesser extent, Thai bahts flooded the Cambodian market when the government stimulated foreign investment in 1989 in conjunction with its free market-oriented policy. More important than this, however, was the presence of the United Nations Transitional Authority in Cambodia for the peace-keeping process during 1991-1992. The operation in Cambodia was reported to be one of the largest and most expensive UN operations at a cost of USD 1.7 billion. Not surprisingly, the banking system started to record a remarkable increase in foreign currency deposits in 1992 (Figure 3.1, Table 3.1). Although the concept of dollarization is widely used, currently Cambodia can be characterized as a multiple currency zone: the Vietnamese dong co-exists with the riel near the border with Vietnam (especially in Prey Veng and Svay Rieng provinces), the Thai baht circulates in Battambang and Koh Kong, while the US dollar commands the greatest share in domestic use. Gold is still used in the rural areas; for example, it is used for land purchase and as a store of value. In addition to the large amount of dollar denominated assets that is captured by the official statistics, an unknown volume of dollars is widely circulating alongside the riel. The riel is used by the government and the central banks for paying their employees
81
and is preferred by the general public for small transactions in both the urban and rural areas. The dollar, on the other hand, serves all the functions of money in Cambodia. It is used as:
a valuation instrument (prices of goods and services are very often indicated in dollars); a settlement instrument (transactions are mostly made in US dollar cash); a saving instrument for Cambodian people.
Figure 3.1: Composition of Money Supply, 1993-June 2002

COMPOSITION OF MONEY SUPPLY 3,000 2,500 2,000 1,500 1,000 500
-
Riel-denominated deposits Foreign currency deposits Currency outside banks

1993 1994 1995 1996 1997 1998 1999 2000 2001Jun-02
The US dollar is used by both residents and non-residents of Cambodia. Of course, foreign firms, aid agencies, embassies, and non-government organizations spend in dollars, including the payment to their employees. Nonetheless, labour contracts, including those with the domestic labour force, are also in dollars in most local firms. Banks too prefer to carry out foreign currency transactions, many of them do not even accept deposits denominated in local currency, although the riel is declared by law as the countrys legal tender. Banks have declared that this is because they have difficulties finding a suitable asset form for riel-denominated liabilities. Thus, not surprisingly, most of their lending operations are in dollars. To illustrate the extent of dollarization in banking operations, we can look at the structure of bank deposits and lending services at the beginning of 2002. Out of 18 banks currently in the system (14 commercial banks and 4 specialized banks),
Source: National Bank of Cambodia
Table 3.1: Evolution of Foreign Currency Deposits, 1990-June 2002 End of Period Foreign currency deposits (A) (bn riel) 0.4 0.7 65.3 121.1 232.6 365.6 574.8 664.9 667.0 878.8 1245.0 1538.6 1831.4 M2 (B) (bn riel) 61.5 79.0 248.1 333.5 449.9 649.1 911.6 1062.9 1230.1 1442.5 1830.5 2203.9 2648.5 Ratio (A)/(B) in % 0.7 0.9 26.3 36.3 51.7 56.3 63.1 62.6 54.2 60.9 68.0 69.8 69.1
Riel millions
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Jun-02
82
only 8 are prepared to open deposit accounts in riel and only 5 offer fixed 12-month deposit accounts. All provide USD deposit accounts, while only 2 of the 18 accept deposits in foreign currency other than US dollar. Regarding lending operations, only 5 banks issue riel-denominated loans, and 3 of these are specialized banks. Among the 18 banks, 3 banks extend foreign currency loans other than in US dollars. Foreign currency deposits constitute a significant share of officially recorded broad money, reaching over 60% during the last six years, except for the crisis period in 1998 when their share fall to 54%. Since the mid- 1990s the foreign currency share, especially the US dollar, represents about 95% of commercial bank deposits, and 98% of loan portfolios.
2.3 The Role of Market Expectation

A major unsolved problem in economics is how to deal with the uncertainty the future holds, especially when individuals have different perceptions of that future. In the Cambodian context, people usually refer to anticipation of future events that guides present economic acts as the psychological factor. To understand the intensity of this psychological factor, one needs to look at the macroeconomic historical record and the dynamics of dollarization. Certainly dollarization is not the result of a policy decision. It emerged and remained because confidence of the public in the national currency and in government monetary and fiscal policy was eroded and this was in turn intensified by security issues arising from political crisis.
It should be recalled that fiscal policy deteriorated seriously as a result of the loss of external assistance that used to be provided by the former Council of Mutual Economic Assistance up to the late 1980s and in the aftermath of the First General Election in March 1993. Due to limited foreign financing and the inability to issue debt instruments, the government had recourse to bank financing that caused increase in net claims on government by the banking sector by more than four-fold in 1993, compared with its level in 1990. During the same period the volume of the national currency circulating outside the banking system rose at about the same pace as net claims on government, accompanied by a three-digit hyperinflation and a sharp devaluation of the riel. While already having a substantially high growth rate of 70% between 1988 and 1989, money supply (M2) swelled by 241% in 1990, slowed down to a rate of 29% in 1991 and accelerated to above 200% again in 1992. The pattern of money supply growth in this period was actually a reflection of the movements in net claims on government, which increased by 117 %, 408 %, 34% and 184 % in those four years respectively (Figure 3.2). The introduction of the fiscal reform program subsequent to the formation of the new government in 1993, coupled with substantial external financial support for rehabilitation and reconstruction made bank financing of the government budget deficit unnecessary between 1994 and 1998. Cambodia is a highly but not a fullydollarized economy. The government, through the MEF still remains the core institution that injects
83
riel into circulation. Its unchecked deficit spending behaviour can still exercise a destabilizing impact on the economy. In 1998 on account of increased political tension and the associated pressure on government expenditure, the government had recourse again, for the first time since 1994, to bank financing.
Figure 3.2: Money Supply, Net Claims on Government and Inflation, 1989 - 2000
MONEY SUPPLY, NET CLAIMS ON GOVERNMENT & INFLATION (ANNUAL PERCENT CHANGE)
M2
NCG
CPI
400
150 130
300 M2 & NCG
110 90 70 C PI
200
100
50 30
0 10 -100 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 -10
Net credit to the government by the banking sector rose by 126 billion riel in 1998, that is, to more than three times the 1997 level, while riel currency outside banks increased by 153 billion riel or 43%. This put great pressure on the exchange rate, which plunged by 27% (Figure 3.3). The large reduction in the share of foreign currency balances held by the banks in 1998 may not necessarily signal a reduction in degree of dollarization because the reduction may have originated entirely from funds being transferred internally from banks to the non-bank sector. All these facts give evidence that political instability greatly affected confidence.
Partial dollarization has imposed some discipline on fiscal operations, but not totally. The commitments of the government to establish political and macroeconomic stability were and remain a crucial factor for a reform program to be a success. It is also in this context that both the autonomy of the central bank in carrying out its functions and the accompanying accountability must be recognized. As stipulated in the central bank law, which was promulgated in 1996, the principal mission of the NBC is to determine and direct monetary policy aimed at maintaining price stability in order to facilitate economic development within
84
the framework of the economic and development policy of the Kingdom. Thus if the central bank is to accept responsibility for price stability (and the stability in the external value of the national currency) it has to have the operational autonomy and authority within the larger political system to discharge that responsibility. Since growing public preference for dollars is closely link to uncertainty and unpredictable action by the government, it is difficult to evaluate the degree of dollarization that might be thought of as normal under conditions of low inflation and reasonable
predictability regarding movement in the exchange rate and changes in government policy. Since 1999 the country has been praised for its political and macroeconomic stability. But during the period from 1999 until now the ratio of foreign currency deposits to officially recorded broad money remained firmly within the range of 65% and 70%, despite the deflationary trend and a remarkable strength of the riel. This suggests that the degree of dollarization is determined not only by the extent of the uncertainty but also by how long the feeling of uncertainty and negative expectation persists.
Figure 3.3: Government Budget Deficits and Exchange Rates, 1990-2001
Sources: National Bank of Cambodia & Ministry of Economy and Finance
85
3. CONSTRAINTS ON MONETARY POLICY INSTRUMENTS

In light of low domestic currency demand, policy tools that aimed at changing the total magnitude of claims against the central bank available to the NBC are limited. With the transformation from a mono-banking system to a two-tier banking system the monetary authority also tried to introduce more marketoriented instruments of monetary policy such as repurchase agreements and the reserve requirement. However, those instruments were found not to be effective.
local currency deposits with commercial banks were negative in real terms. For example, in 1993, while riel term deposits earned 2.5% interest per month or 30% per annum, the annual average inflation rate reached 144%. Yet, despite the sharp decline in inflation, or the return to positive real interest rates, observed especially since 1999, riel term and savings deposits increased relatively slowly compared with US dollar deposits. The 2.5% share of riel term and savings deposits in broad money at the end of 2001 was little changed from that of 2.4% measured at the end of 1993. At the onset of dollarization in 1993, the total volume of this type of deposit was around R8.5 billion (equivalent to $3.6 million at the prevailing exchange rate) and rose to R54.4 billion by the end of 2001 (equivalent to about $14 million at the 2001 exchange rate). Thus over a period of eight years riel term and savings deposits rose by about 6.5 times, but only by about 3.9 times when valued in US dollars.
3.1 Responsiveness to Interest Rate Changes

In the normal case, interest rate is the price paid for the use of money and credit and is thus an important determinant of the demand and supply of funds. But empirical analysis for the Cambodian case showed that local currency savings are relatively interest inelastic. Due to hyperinflation during the early 1990s, interest rates on
Table 3.2: Interest Rates on Riel and USD Deposits Dec-1997 12-month riel deposits 12-month USD deposits Growth in riel deposits (USD equivalent)
Dec-1998 11.09 6.45 1.4
Dec-1999 9.84 6.20 3.2
Dec-2000 7.20 5.97 3.3
Dec-2001 8.60 4.72 2.2
11.13 6.36 0.9
Another factor that badly affected riel savings was the riels rapid devaluation on the foreign exchange markets. In 1998 for instance, a 4.6% interest difference in favour of riel deposits was more than offset by a 27% depreciation of the riel. It is thus very obvious that dollar deposits were preferred by market participants.
86
Third, and most important, was the expectation that is induced by the memory of hyperinflation and devaluation, the short-term effect of which has long term impact on various economic variables, including riel savings. Consumer expectation may explain the slow annual growth of time and savings deposits since 1999. The interest rate inelasticity in Cambodia is self-explanatory since the anticipated yield on domestic currency denominated investment depends not only on the interest rate being paid, but also on the anticipated foreign exchange rate, against the US dollar, expected to prevail when the investment matures, given the continuous strength of this currency against the local currency as well as against many other currencies in the international foreign exchange markets over recent years.
hardly face an urgent need for local currency liquidity. On the whole, due to lack of eligible financial instruments (bills of exchange, promissory notes etc.), such central bank services have not been used so far. To date, there is no government security that can be used as proper collateral in accordance with the requirement of the regulation on the repurchase agreement/credit window. Nor can banks rely on other banks credit facility, since an inter-bank market has not yet developed. Therefore, banks are forced to maintain a large cash reserve, which affects their efficiency. To implement standing facilities the authorities need to scrutinize the existing legal/regulatory framework and to encourage the issuance of negotiable instruments. Recently progress has been made in improving legal infrastructure for the banking sector, which is reflected in the efforts of the NBC to draft the Law on Negotiable Instruments and Payment Transactions.
3.2 Standing Facilities

Pursuant to the Law on the Organization and Conduct of the NBC, that allows the central bank to provide rediscount facilities for banks and financial institutions, in the mid-90s the NBC issued regulations for the operation of repurchase agreements. The eligible instruments for this operation should be denominated in local currency. In addition, the legal basis is also in place for the NBC to grant advances to commercial banks in order to support banks with shortterm liquidity needs. However, the law requires that the credit window should be local currency-based and that banks must have government paper as security. Theses two features proved later to be the main obstacle for effective use of such a facility. Since about 90% of banks transactions are denominated in US dollars, generally banks
3.3 Minimum Reserve Requirement

Regulation on the maintenance of required reserves at the central bank without any remuneration was in place in early 1994. The reserve ratio was initially established at 5% of commercial bank deposit liabilities and other borrowings. The reserve requirement can be maintained in both riel and dollars. Due to the dominance of the dollar in their operations, however, banks have maintained more than 90% of their reserve deposits with the central bank in that currency. This reserve should be held in accounts
87
with the central bank, while cash in vault denominated in riel can be included in the reserve requirement as well. With the introduction of this instrument, the NBC was hoping to influence the commercial banks ability to create deposits and, by changing the reserve ratio, money supply. However, reserve requirements have so far proved to be an ineffective instrument of monetary policy under present conditions. The reasons were twofold: i the large volume of excess reserves;
of effective means to redistribute reserve balances among financial institutions, prompted banks to hold large cash reserves. Banks have simply to use their excess reserves to meet the extra demand of their depositors and it is difficult for the central bank to mop up those excess funds since they are also unevenly distributed among banks. Second, heavy dollarization undermines the ability of the NBC to use reserve requirements to steer deposits and the money creation of the commercial banks in so far as its affects the limited existence of local currency deposits. Even though increase and decrease in reserve ratio might have some effects on US dollar deposits, it would not help the central bank to absorb riel liquidity from the market. This means that the central bank can affect the money supply in riel only through a direct type of instrument, namely printing money or the reverse.
ii the high degree of dollarization of the economy. First, as can be seen in Table 3.3, the volume of banks excess reserves was higher than the compulsory reserves throughout the period under review. The predominance of cash transactions, combined with the lack
Table 3.3: Bank Reserves, 1993-2001 1993 Reserve Requirements Free Bank reserves
Sources: National Bank of Cambodia
1995 26,332 32,004
1997 43,767 135,691
1999 82,786 337,218
2001 131,944 614,306
30,438
Under these circumstances, the NBC understands that the most efficient contribution the reserve requirements can make to the stability of the economy is to serve as prudential instrument in banking regulation rather than as an instrument for monetary purposes. As such, and in the wake of the outbreak of the Asian financial crisis in July 1997, upon the National Banks circular of 26 December 1997 the reserve ratio for all commercial banks has been raised from 5% to 8%. This move aims to ensure that bank has enough funds to meet depositor withdrawals and hence to strengthen confidence of the general public in the banking sector.
88
3.4 Exchange Rate Policy and Foreign Reserves Management

The current exchange rate regime in Cambodia may be characterized as a dual regime consisting of an official and a parallel market rate. The official rate is mainly used for transactions between the NBC and the public sector. The parallel market rate is a market determined, freely floating exchange rate, whereas the official rate is determined by the the NBC in view of developments in the parallel market rate. This means that Cambodia has adopted a market-oriented exchange rate policy with the official exchange rate adjusting to movements in the parallel market rate with a lag. The government narrowed the spread from 5% originally to no more than 2% on a daily basis from the beginning of 1994. Since 1995, the official exchange rate usually has not differed by more than +/- 1% from the market rate. This measure has been taken with the aim of protecting the balance of payments and ensuring that foreign reserves are not utilized to defend an unrealistic exchange rate. Many economists argue that in a dollarized environment monetary policy is the same as exchange rate policy, and this is true in Cambodia, where the objective of the exchange rate policy is confined to price stability rather than supporting export industry. It is quite certain that uncertainty about government policies and future business environment would cause sharp foreign reserve movements for countries with pegged exchange rates and sharp exchange rates fluctuations for those with floating exchange rates. Because, the riel
exchange rate stands largely under the influence of political and psychological factors, it is important for the monetary authority to step in into the local foreign exchange market. That is where the authority normally employs its official rate. If for instance the riel depreciates on account of psychological factors by 5%, the authority - instead of following closely the downward movement of the market rate has kept the officially determined value of the riel a bit stronger. Therefore the spread sometimes surpassed the normal benchmark of 1%. The official exchange rate plays in this context the role of a signal to the market that the monetary authority does not support what it thinks an unfounded move in the market exchange rate. The NBC has also resorted to foreign exchange market interventions from time to time, when market sentiment and psychological factors caused disorder in the market. US dollar auctions were the most frequent instruments that have been used so far. Buying and selling of US dollars by the NBC has proved to be an effective form of open market operations. It has been used carefully, however, and not to deal with depreciation arising from fundamental imbalance or to resist global pressure. In general, the central bank has taken the opportunity provided by the relative strength of the local currency to buy US dollars from the market with a view to accumulating international reserves. US dollar auctions were conducted during the period 1993-98. The volume of dollars sold by the NBC was very small in 1993, reflecting the low level of the countrys gross international reserves at that time,
89
equivalent to only 0.6% months of import. However, the supply of dollars to the market rose remarkably in 1994, when the reserves were able to cover 1.6 months of imports, while the value of imports also expanded by 1.6 times within that one year. Since 1999 no more dollar auctions have been conducted, as the value of the riel has largely stabilized and even gained some strength against the greenback. The central bank can thus be active in the local foreign currency market to the extent that it accumulates reserves, while it has at the same time to observe the floor on these reserves set in a reform program agreed with the IMF.
transactions outside the banking sector, monetary policy has been oriented toward restricting growth of officially recorded liquidity. The NBC, as mentioned earlier, lacks appropriate indirect instruments to serve its restrictive monetary objectives. The present monetary policy, designed by the IMF with the agreement of the government under the reform program supported by the Poverty Reduction and Growth Facility, is mainly implemented through careful control over the movements of net domestic assets of the central bank, a quantitative restriction on net bank credit to the government and a floor on the countrys official international reserves. The design of the policy is derived from the relationship in the monetary survey that can be illustrated as follows:
3.5 Design of Monetary Policy under the PRGF Program

Facing difficulties in assessing liquidity growth due to unmanageable dollar cash
M2 ( Total liquidity of the banking sector )
( Total Net Foreign Assets ) NFADMBs + NFANBC
( Total Net Foreign Assets ) NDANBC + NDADMBs
Net Internatinal Reserves
NCG
NCP
OIN
90
a. Ceiling on net bank credit to the government. Ceiling on net bank credit to the government has been used as an intermediate target with the rationale of restricting credit from the banking system to the government, which was the main cause of macroeconomic instability in the past. The expected implications of this measure were interpreted in two possible ways: Restriction on bank financing would eliminate the crowding out effect of government borrowing and would thus provide more room for private sector credit.
In addition, fiscal policy of the government has started to emphasize the importance of building a surplus on current balance, while relying for the financing of capital expenditures largely on external sources in the form of concessional loans. b. Ceiling on net domestic assets of the NBC. This measure is intended to serve two purposes:
Limitation on the recourse to central bank financing of the budget would reduce excess money supply, thereby lowering inflation.
Crowding-out effect occurs on the assumption that to the extent that the money supply is fixed, additional loanable funds are not forthcoming to finance the governments additional expenditure, unless the level of private sector spending is reduced. As mentioned earlier, however, the banking sector in Cambodia is over-liquid. Hence the assumption is no longer valid for Cambodias current circumstances.
A cut in credit to the government (through expenditure reducing policy) would lower aggregate demand by reducing the absorption capacity of the economy, which may improve the current account balance.
Refrain from refinancing commercial banks would help correct the current account balance through reducing credit expansion and also reduce the pressure on drawing down the countrys international reserves.
On the latter ground, since 1999 the government has struggled to enhance revenue collection and rationalize its expenditures, while making efforts to reduce commercial bank credit to public sector companies and to restrict itself from recourse to bank financing.
c. The floor on the countrys net official international reserves. Net official international reserves are defined by the IMF as a difference between total net foreign assets of the NBC and unrestricted deposits of banks held with it. Such a floor determines how much of a current account deficit Cambodia can afford. It is intended to ensure high levels of international reserves in order to be able to cope with the economic consequences of a free capital mobility and associated short-term pressure on the value of the riel. Perhaps business communities as well as the general public may feel more assured, if they know that the monetary authority has built up a reserve fund in anticipation of financial sector emergencies, etc.
91
3.6 Policy Implications

In the monetary sector, monetary development reflects the restricted feature of the monetary and fiscal measures described above. Since the end of 2000, claims on the government by the banking sector have been virtually eliminated. Credit from the central bank to the remainder of the banking system remains absent. The NBC has constantly accumulated its international reserves. Growth in the countrys gross international reserves has kept pace with the continuous increase in total imports of goods and services and reached more than three months cover in recent years. Inflation has been kept within the target and in fact turned negative during recent period, while the value of the riel has largely been stabilized The existing restricted monetary policy and fiscal policy has served the countrys economy well so far in terms of promoting macroeconomic stability and generating growth. However, growth remains fragile. The recent stability relies heavily on external assistance. Many sectors, especially public investment will suffer greatly when external funds discontinue, while slow growth of foreign direct investment is also another matter of concern. The monetary authority thus needs to be more proactive. To be more pro-poor, monetary policy has to address two main challenges: 1. To stimulate higher economic growth through a departure from tightening monetary and fiscal policy;
2. To encourage the intermediation process by reducing the cash-based economy to underpin the efficient transmission of monetary policy and improve the link between savers and investors. Meanwhile, the monetary and fiscal policy may need to be more flexible than before. Presently, given that the monetary policy is restricted, increase in international reserves represents almost the sole source of liquidity growth, while domestic credit to the economy rises at a very slow pace. Broad money growth was brought down from more than 200% in 1992 to an average rate of around 35% during mid-1990s. It grew by 27% in 2000 after rising by about 16% during the crisis period and recorded a growth rate of 21% in 2001. However, this restricted policy should not be seen as the only way out, without flexibility. The inflation rate, as measured by the annual average consumer price index, has remained low since 1995. We can even say that it is too low, showing a deflationary trend throughout the period from mid-1999 to 2001. And during this time growth performance, after facing a low record between 1996 to 1998, bounced back to a normal level for Cambodia of 6%-7%. However, as shown previously, this rate is insufficient to have a clear impact on poverty reduction. Different studies provide that the threshold inflation for developing countries that maximize output is about 7%-12% (Khan and Senhadji 2000). Indeed a double-digit inflation like 15% might not be good for Cambodia, given its inflation history and high degree of dollarization. But, the
92
targeted inflation of about 4%, as mentioned in the governments SEDP II, seems to be low if the aim is to create employment and generate growth. Moreover, even this targeted rate has not been reached during the past 4 years. Furthermore, as can be seen from Figure 3.4, the velocity of money has declined steadily since 1998, which may explain the low inflation during that period and reflect dampened aggregate demand. Declining velocity might also suggest that the public ignores or is simply unaware of the opportunity cost of holding money, as deposits grew relatively slowly despite the return to a real positive interest rates. To offset the fall in velocity the central bank may thus opt for a loosening monetary policy. Central messages in the above paragraphs are: a. Containing inflation should not be regarded as the NBCs primary concern at this moment. Higher consideration should be given to the stimulation of domestic demand. Cambodia should allow some inflation to grease the economy. b. Instead of focusing predominantly on making increase of net foreign assets of the banking system the sole source of liquidity growth, the authorities may need now to balance between the two determinants of money supply by shifting toward a stronger increase in net domestic credit.
Figure 3.4: Income Velocity of Money, 1995-2001
Source: National Bank of Cambodia; Velocity of money is calculated as a ratio of nominal GDP to average stock of broad money M2.
Actually, as the macroeconomic imbalance and distortions are reduced, particularly the unmanageable budget deficit, the negative real interest rates, and the unstable economic conditions, there should be increasing opportunity for the public to avail itself of depository and payment services of banks. The economy should have the chance to mobilize idle cash in the shape of both, outside-bank savings and bank-cash reserves, which should contribute to enhance credit generation. But this does not happen in the reality of Cambodia. Thus stable economic conditions and a sustainable fiscal position are desirable but not sufficient to develop a deposit and credit culture. Consequently central messages here are: a. Distortions to financial intermediation have to be addressed. To achieve this, besides maintaining a reasonable inflation rate, the monetary authority has to conduct other important auxiliary functions, notably promoting the development of the money market, safeguarding the payments and clearing system, restructuring the banking
93
system, and performing bank regulation and supervision, to support the main function. A well-functioning financial system is the key to ensuring macroeconomic stability, mobilizing savings, allocating resources for productive purposes, promoting private sector development, and in turn generating employment opportunities and reducing poverty. To develop the financial system, the government has formulated a long-term vision and development strategy for the financial sector with the technical assistance from ADB. The blueprint called Vision and Financial Sector Development Plan for 20012010 was adopted on 24 August 2001 as part of an overall government reform agenda toward establishing a more market-based, open, and private-sector led economy. The blueprint focuses, among other things, on issues such as establishing a sound legal framework to protect property rights and enforce contracts; establishing the rule of law through legal/judicial reform to underpin financial and commercial activities; sequencing efforts to develop financial infrastructure, particularly interbank and money market, payments and clearing system, accounting and audit system, credit information system, and regulatory framework; structuring the regulatory and policy framework to foster competition in the financial markets and to encourage financial institutions to realize economies of scale and scope. With the blueprint in place, it is now up to the government to materialize the vision set out in this ten-year plan. b. Nonetheless, the development of such institutional and legal framework is a complex process that usually requires considerable time to be completed. In the meantime, improved intermediation process may be
started with putting in place mechanisms to bring banks and potential savers closer. One way the authorities might initiate this process by requesting the payment of wages and salaries through bank accounts. To ensure the success of this measure, it is an essential prerequisite to ensure that bank deposits in riel are fully and easily convertible into cash. The government should of course start first with all salaries of the public sector. The amount involved is about 4% of GDP. Although we can assume that savings associated with this figure will be small, the move would have more impact on conventional practices and thinking by showing to the public that the government has confidence both in the national currency and in the banking system. Obviously, the measure would contribute somewhat to reducing cash transactions, but more important than this is that it establish the connection between at least some higher proportion of the general public and the banking system. When implementation in the case of public sector salaries functions smoothly, the authorities could move to the next step with the private sector. Also for the same objective, the authority may consider seeking co-operation from all international institutions having presence in Cambodia to pay their employees through bank accounts. Associated with this initiative, a public education campaign will play an important role in explaining the advantages of entrusting savings to banks and to work against the grip of past memories and expectations. Until now only a very narrow portion of the population and enterprises have any understanding of how to use banking services. Public awareness building is thus a need for Cambodia.
94
4. COSTS AND BENEFITS OF DOLLARIZATION FOR ECONOMIC DEVELOPMENT AND POVERTY REDUCTION 4.1 Advantages of Dollarization
Using the US dollar brought a wide range of advantages to Cambodia as follows: Switching public practice from gold to banknotes denominated in US dollars. The use of US dollars helped enlarge the sphere of the monetary economy by discouraging the accumulation of wealth in the form of unproductive physical assets such as gold which was a common practice in Cambodia until the early 1990s. The process of monetization of the economy in turn enables small savings, in particular within the middle income class, which was not possible with the use of gold. Monetization implies a reduction in relative size of subsistence activities in total activities. This process of monetization manifests itself in phenomena such as increased output and the increase in cash transactions (rather than payment in kind) in labour, land, and credit markets. However, it should be noted that in some remote areas gold continues to be accepted as payment instrument, for example, for the purchase of land, and as store of value. Lack of transport, connection to market, and market information are the main obstacles to the monetization process.
Preventing capital flight and promoting financial deepening. Dollarization provides the opportunity for the domestic financial system to enhance domestic intermediation by eliminating the incentive to place foreign currency savings in accounts abroad. Asset
substitution consequently permits the growth of the financial system in Cambodia. Domestic savings as measured in terms of foreign currency deposits with the banking system have increased substantially since Cambodia opened its economy in 1993 (see Table 3.1). For comparison, savings in local currency also rose during that period, but at a much slower pace (see Table 3.2). The seriousness of the problem would be much more visible, had the authorities prohibited residents from having foreign currency bank accounts or otherwise banned the use of the US dollar.
Lowering the risk of currency devaluation. Besides the fundamental causes for a given currency to devalue such as in-house inflation or persistent, large current account deficits, fluctuations of a currency value can be inflamed by contagious effects of inflation and devaluation abroad. Since demand for the riel remains relatively low and the market for it is very small (currency in circulation accounts on average for only about 30% of the total officially reported liquidity of the banking sector), one can argue convincingly that there is little or no incentive for speculators to try to gain from short-term changes in the price of riel. The high degree of dollarization also provided a cushion against the risk of contagion as Cambodias economy has shown resilience in the face of the Asian crisis. The stronger riel relative to the Thai baht during the currency crisis that erupted in July 1997 in Thailand was an example from recent history. Strong demand for the US dollar brought down the value of the Thai baht by 86% between 1996 and 1997, while the riel depreciated by only 27%. Dollarization sustained confidence of investors in their
95
operations in Cambodia, hence relieving pressure on the exchange rate. A relatively low level of inflation during the crisis period (14%) was, without doubt, the most important social achievement of dollarization, although we should also emphasize that other factors, especially the small and undeveloped financial sector and the long-term nature of private capital inflow played an important part in alleviating the crisis effects, especially strong export and tourism performance that helped add some dynamism to economic activities. We should also note that dollarization in the Lao People Democratic Republic did not create such a buffer against currency crisis. All these imply that dollarization by itself is not a sufficient condition for eliminating devaluation risk. Facilitating Cambodias integration process. The US dollar is a currency with high purchasing power and has relatively lower exchange rate risk compared with many other currencies in the world. Stable currency promotes macroeconomic stability and a predictable business environment, which is of utmost importance for Cambodia, particularly from the point of view of international investors. In addition, the use of the dollar helps reduce the number of currency conversions, and hence the transaction costs that would be necessary if investors, both local and foreign, had to conduct business in local currency. With the move toward a market economy supported by the official permit to residents and non-residents alike to transact and to hold accounts in foreign currency, Cambodia has been able to encourage domestic investment and also to attract remarkable investment flows from abroad. With respect to poverty
reduction, we should take note of the boom in the garment industry that is worth over $1000 million (gross revenue) and provides employment opportunities for numerous women, who would otherwise be unpaid family farm workers, contributing to increased underemployment in the agricultural sector (see Chapter 5).
4.2 Disadvantages of Dollarization

The drawbacks of dollarization can also be numerous:
Undermining the effective conduct of monetary and fiscal policy. Low demand for domestic currency means low domestic financial liabilities of the central bank that issues that currency. This in turns means a small margin for monetary policy to be active. Since commercial banks loans and deposits are largely denominated in foreign currency, there is little room for the NBC to develop instruments of monetary policy. Hence the NBC cannot play its role as a lender of last resort for banks facing short-term liquidity problem as banks requirements are for dollars and the NBC is not able to provide dollar support. Furthermore, low demand for domestic currency impedes the central bank in its efforts to build up foreign reserves. Overall, the existence of such a dual currency limits the central banks ability to manage liquidity and restricts the use of the monetary aggregate M2 as a good intermediate target to affect the general price level.
There is also limited room for fiscal adjustment because dollarization puts constraint on the mechanism of deficit financing other than
96
external debt or assistance or domestic borrowing at market-determined interest rate. In other words, dollarization may deprive fiscal benefit that can occur when the government runs a fiscal deficit through the central banks expansion of money supply. In principle, this is inflationary only to the extent that the general public does not want to hold the additional local money. Assume that the economy uses only the national currency, the riel, and suppose that real income is expanding by 10%. The extra transactions associated with the higher Cambodias production and income would call forth a new demand for holding the riel of 10%. If the increase in riel money stock does not exceed the increase in demand for it, meaning that the fiscal deficit is not inflationary, the government will then benefit from an excess of expenditures over revenue. Given the large preference of the general public for US dollar, however, fiscal policy tends to be stringent.
Damaging the pride of the nation. Deprived of monetary and exchange rate flexibility as well as of an independent fiscal policy, Cambodia tends to surrender to the United States economic policy and their economic health, although Cambodias economy is structurally fully different from that of the United States. This will leave Cambodia limited room to strike its own balance between inflation and unemployment objectives in the determination of macroeconomic policy. The key point is that Cambodia cannot isolate itself from external shocks and that those shocks require a domestic adjustment, one way or another. Further, despite much talk about globalization and integration, divergence in the economic and political space still dominates. It is therefore important
to retain tools of economic sovereignty unless there are compelling arguments to the contrary. As well, the current situation in Cambodia does not give the feeling of a total loss of confidence in the local currency. Reducing the opportunity to earn income from seigniorage. Revenue from seigniorage can be defined as that resulting from the difference between the cost of producing and distributing money and the eventual opportunity costs (or their greater purchasing power). It is generally agreed today that dollarization represents seigniorage loss for a dollarized economy. The question is only on the extent of this loss. To the extent that the public want to hold foreign currency in place of local currency, the US Government will gain added seigniorage benefits. It is estimated that the annual cost to the US government of creating and servicing currency is now less than 0.1% of the face value of currency outstanding. However, dollar notes, apart from those entering into Cambodia in the form of financial assistance, buy goods and financial assets in Cambodia at face value. Thus, the loss in seigniorage for Cambodia is actually less than the value of total dollar-denominated money stock because part of this stock is held in the form of financial assets that bear interest income. But the part that Cambodian people hold for transaction purpose, the dollar currency in circulation, does not earn interest nor dividend. Information received from discussions with money changers, commercial banks and the private sector indicated that the amount of US dollar currently circulating inside Cambodia as medium of exchange is estimated to be about 5 to 10 times the amount of riel currency in circulation. Given that riel circulating outside banks in 2001
97
was equivalent to USD 150 million, the dollars in circulation would be in the range USD 750-1500 million. If we suppose that out of the population of 12 million, one million have access to dollars, this would mean that on average each of them held about USD 750-1500 in cash. This seems to be reasonable given that a basic skilled worker can earn about USD 100 a month. (We should note that a recent IMF study estimated dollars circulating outside banks at USD 2.9 billion in 2001 (Zamarczy and Sa 2002)).16 To estimate seigniorage loss we develop three scenarios to reflect different estimates of dollar notes in circulation. For each
scenario, we compute the interest foregone associated with the use of dollar notes for transaction purposes based on an estimated average return rate of 3% net of the eventual costs of printing and distributing the riel in exchange for those dollars. Further, we consider the distributed opportunity costs that would occur to the concerned agencies, knowing that the Article 18 of the Law on the Organisation and Conduct of the NBC provides for the use of net income of the monetary authority as follows:

20% to build up the general reserves; 5% bonus for the National Bank Staff; 75% to transfer to the national budget.
Table 3.4: Estimated Seigniorage Loss for the Year 2001 (in USD million, unless otherwise specified) USD notes in Scenario circulation I II III I70 1,000 1,500 General reserve 20% 4.5 6.0 9.0 NBC personal 5% 1.1 1.5 2.3 Forgone. Forgone. National government government income to budget income to actual 75% GDP revenue 16.8 4.3% 0.5% 22.5 5.7% 0.7% 33.7 8.5% 1.0%
Interest forgone 22.5 30.0 45.0
If Cambodia succeeded in reducing currency substitution, the central bank would have the possibility to improve its reserve positions and thereby enhance stability and confidence, and strengthen the banking system. The government, on the other hand, would have more room for its social and economic spending and would have also resources to compensate its increasing need to spend in foreign currency as integration into the world economic system gathers pace.
Widening the gap between rich and poor. We know that purchasing power of a currency fluctuates with its exchange rate. The existence of co-currencies will imply differences in purchasing power brought about by the different denomination of nominal incomes. Firstly, since the US dollar is generally a strong currency, there might be a benefit for those segments of the Cambodian population who received income in hard currency. And the
16 In this case the estimate was derived from a model.
98
reverse effects can be expected for those who earned in local currency. We now look at this issue by trying to determine social conditions behind the relationship between the two macroeconomic variables in Figure 3.5 below. If we look at the time series from 1995 to 2001 on the changes in exchange rates and price levels for Cambodia, we quickly see that changes in inflation are closely associated with the fluctuations of the riel against the dollar. The most interesting picture is found for the period of currency and financial crisis in 1997 and 1998, where the rate of currency devaluation was much larger than that of the change in the consumer price index. While the riel experienced an acceleration in its depreciation from 12.4% in 1997 to 26.7% in 1998, the changes in the general price level were 9% and 13.6% for these two years respectively. Depreciation of the riel leads to a rise in price and consumers need more riel to buy the same goods and services. The question that follows is thus who does and does not get more riel? This price and exchange rate development can exercise different effects on different classes of population depending on the currency they hold. And it is the practice now in Cambodia that income in the form of corporate profits, rental, and wages for employees in many private companies, international organizations and NGOs is denominated in US dollar. The riel depreciation of 27% in 1998 means that with the same income, US dollar earners
have more riel in hand. After adjusting for the inflation rate of 13.6%, this group still earns a net positive benefit of some 13%. Through the same mechanism gains can also be shown for this group during 2000 and 2001. Thus US dollar earners tend to benefit when the devaluation rate exceeds the inflation rate and they tend to see their real incomes shrink with the appreciation of the riel against the dollar, other things being constant.
Figure 3.5: Consumer Price Index and Inflation Rates, 1995-2001
Note: Upward movements of the riel/USD line point to depreciation of the riel. Sources: NBC & NIS, annual average figures
Earners of riel cannot multiply their income through the foreign exchange markets regardless of the direction of movement in the exchange rate. Unlike the group discussed above, they cannot use devaluation to compensate the price increase and they lose purchasing power with the rise in inflation. In the case of government employees, since the government budget constraint generates relatively strong rigidity in their nominal wages, this fact implies that their material standard of living worsened during the high inflation (disregarding those civil servants who may have a second source of income). This is also true for those self-employed in the informal sector such as motorbike and tricycle drivers, small vendors and rural poor
99
in the agricultural sector that make up the vast majority of the population. Small farmers in Cambodia usually work in family production and operate in local currency. The reasons for this could be three-fold. First, their money transactions are small. Second, due to lack of familiarity with the dollar and concern about counterfeit currency, the majority of rural people still prefer to use riel. And third there is no foreign exchange market in the rural areas. Exchange rate changes themselves may affect, in addition to the riel value of USD income, the prices of some tradable goods and services that enter into commodity basket defining the price level. However, there are big differences in price movements between Phnom Penh and the rural areas that are not reflected in the current measurement of price level. And this complicates the assessment of the social conditions of the poor. The National Institute of Statistics currently produces a CPI only for Phnom Penh (NIS 2001). Each item in the index is weighted according to its relative importance in total expenditure of consumers in Phnom Penh. Due to large income gap, wide difference in the needs and price differential between urban and rural areas the current CPI is not truly representative for Cambodia as a whole. The CSES showed, for example, that expenditure on food and medical care weighed much larger in the rural areas than in Phnom Penh and other urban areas. A study conducted by the Cambodia Development Resource Institute to assess the social impact of the 1997 Asian crisis indicated that there were larger fluctuations of prices in the provinces than in Phnom Penh (CDRI 1997). It was also reported that during that period prices in the provinces
increased more rapidly than those in Phnom Penh, probably due to the increase in transportation costs of gasoline, whose riel-denominated prices increased sharply following the depreciation of the riel in the second half of 1997. High price fluctuations in the provinces might also be caused by limited access to markets due to bad road conditions, inadequate transportation, and disconnection to the markets through flooding. Therefore in order for the CPI to provide better picture of the purchasing power of the rural poor, the concerned authority may have to review the current methodology for CPI compilation with regard to the coverage of price collection and the weighting pattern.
5. DOLLARIZATION AND CAMBODIAS FOREIGN TRADE

Since the opening of its economy in the early 1990s Cambodia has actively participated in the process of international division of labour and promoting economic relations with other countries not only in South-East Asia but also in the entire Asia-Pacific region. In this context the government adopted an outward-oriented development strategy aimed at attracting foreign capital and improving the countrys exports. Recently with mainstreaming the trade sector into the national economic development plan efforts will be focused, among other things, on policies directed towards export-led poverty reduction. In this framework, it would be appropriate to consider whether the use of the US dollar as a means of payment is relevant to the objectives.
100
5.1 Implications for Exports

When reviewing Cambodias external trade performance during the last decade, one can clearly see that external trade activities have expanded rapidly (Figure 3.6).
Figure 3.6: Domestic Exports, Retained Imports and Trade Balance, 1993-2001
Source: Customs Department & NBC's estimates
Cambodias domestic exports (omitting re-exports) increased more than elevenfold during 1993-2001, although from a very low base. The domestic export to GDP ratio increased from 4.3% in 1993 to 35.2% in 2001. The country is now, therefore, a very open economy. As a result of rapid export growth, the current account deficit (excluding official transfers) as a share of GDP has approximately halved since 1997 compared to its previous level (NIS 2002: 252, 262).
From the perspective of monetary and exchange rate policy, the interesting question to ask is whether riel/dollar exchange rate movements have played a role in the export boom. The value of the riel against the dollar, as measured by the annual average market-buying rate, has moved sharply (but temporarily) since early 1993. In early 1993, amidst the First General Election in the country, it fell to 2,906 riel/$ from 1,363 in 1992. It then appreciated and remained relatively stable (at around 2,570 riel/$)
101
during 1994-1996, as relative macroeconomic and political stability was restored and the credibility of banking and financial policies improved. However, the exchange rate remained subject to a large range of shocks, and the political and financial crisis in neighbouring countries in 1997-1998 contributed to a further depreciation. Between July 1997 and October 1998 it depreciated very sharply, by about 27 per cent. Facing the strength of the US currency, the riel weakened further, but slowly, and traded at around 3,930 riel/$ in December 2001 Figure 3.7). International economic theory suggests that devaluation tends to improve international price competitiveness (as long as any changes in the domestic price level or the foreign price level do not offset the exchange rate change). At the same time, imports tend to decrease as the domestic currency price of imported products rises. Through this mechanism one can expect thus an improvement in the trade balance.
Figure 3.7: Domestic Exports, Retained Imports and Exchange rates, 1992-2001
DOMESTIC EXPORTS, RETAINED IMPORTS & EXCHANGE RATES

1600 1400 1200 Millions of USD 1000 800 600 400 200 0
1992 1993 1994 1995 1996 1998 1999 2000 1997 2001
1000 1500 2000 Riels/USD 2500 3000 3500 4000 4500
Domestic exports
Retained imports
Exchange rates
Sources: National Bank of Cambodia and Customs Department
However, this mechanism is hardly applicable to the case of highly dollarized Cambodia. Under such circumstances, the riel/dollar exchange rate is scarcely applicable to the countrys exports and imports. It can not be denied that the nominal exchange rate is widely reported in statistical reports and financial press and is regularly quoted in
the local foreign exchange market. It is also true that the value of this exchange rate reflects, to some extent, demand and supply conditions in the market for riel and US dollar. The fact of the matter is that this foreign exchange market is very thin and this exchange rate is determined, above all, by psychological and political conditions,
102
by international economic conditions and, last of all, by Cambodias international trade position. We know that textiles and garments dominate Cambodias exports, reaching about 80% of total export value in 2000 and 2001. Therefore change in external import demand for Cambodias garment will greatly affect Cambodias external position. We also know that all inputs and labour employed for the production of garments are paid for in US dollars. Since prices of Cambodian exports are determined in world markets in US dollar terms, and Cambodian exporters also pay local producers in US dollar, then movement in the value of the riel/US dollar exchange rate will not affect price. Of course, as shown in Figure 3.7, the riel depreciated more or less constantly throughout the 1990s, the period in which Cambodias exports also boomed. This appearance is clearly misleading because due to dollarization Cambodias export remains largely inelastic to the riel depreciation. Exports have been pushed up by expansion in direct investment of foreign firms in the textile industry who see the return in political stability after the election in 1998 and a comparative advantage of Cambodia in that sector. The comparative advantage of the industry depends not only on its productivity relative to foreign competitors, but also on the domestic wage rate relative to the foreign one. It is reported that workers in the Cambodian garment industry earn on average about $45 a month. This comparative advantage based on low wages compared to other garment export countries such as Thailand and China has been further enhanced by Cambodias receipt of Most Favoured Nation and Generalized System
of Preference status from 28 developed countries, especially the United States and European Union, and by other incentives that the government provides to exportoriented industry. In this context, export competitiveness of Cambodia can be jeopardized not by appreciation of the riel, but by the rising value of the dollar. Cambodia is a price taker and has to accept the price the world offers for its products. Since local costs are priced in US dollars, the country may lose its export competitiveness, unless the costs are absolutely lower than elsewhere. Due to dollarization, prices are sticky in dollar terms; hence neither the nominal exchange rate nor the real effective exchange rate is an appropriate measurement of the countrys export competitiveness. This is clearly a limitation on the Cambodian authorities access to policy instruments because it is not possible for them to make adjustments in light of exchange rate movements of competitor countries.
5.2 Implications for Imports

For the purpose of analysing implications of dollarization on Cambodias imports and consumer well-being, we look at movements of the riel and the currencies of Cambodias two largest suppliers, Singapore and Thailand, against the US dollar. Figure 3.8 shows that all three fell against the dollar, although to varying degrees. The Singapore dollar remained stronger than the riel throughout the period 1995-2002, while the riel largely firmed against the Thai baht. We should note that Cambodia has a very limited production capacity. The country relies heavily on imports, not only of inputs
103
for domestic production but also of goods for domestic consumption. The strength of the US dollar against the Singapore dollar and baht contributed to ensure cheap imports, and thus tends to encourage higher imports. This could be a factor in Cambodias continuing current account deficit.
Figure 3.8: Exchange Rate Movements of the US dollar against the riel, Singapore dollar, and Thai baht, 1995-2001
EXCHANGE RATE MOVEMENTS OF THE US DOLLAR AGAINST SELECTED CURRENCIES 80 90 100 110 120 130 140 150 160 170 180 190
1995 1996
Riel
1997
1998
SinDollar
1999
2000
2001
need to distinguish between the sector of the economy that operates in US dollar, and are thus sensitive to change the exchange rates of the US dollar against that of Cambodias foreign suppliers, and the sector that uses the riel. Lack of data on the size of each sector forces us to assume that because the volume of US dollars in circulation is several times larger than that of riel, changes in the US dollar relative to the Singapore dollar and the baht are more relevant to Cambodias imports than is the rate of the riel to either of these. This argument is undoubtedly true for imported intermediate goods such as inputs for the garment factories. The rise of 4% in the US dollar against the Singapore dollar in 2001 means that dollar-denominated price of Singapore cloth declined by 4%. An export industry with high import content can thus compensate to some extent their inability to benefit from depreciation of the local currency, by obtaining cheaper imports. What would happen to Cambodias trade if the authorities decided to de-dollarize the economy in the current context? In the short run, export companies would benefit. We argue this on the assumption that de-dollarization would translate into devaluation of the national currency. We would not expect the riel to become stronger against the dollar, given the still relatively weak capacity of the economy and the authorities to support a strong riel. Throughout the whole history of the riel exchange rate movements we see that it has tended to devalue, especially in face of shocks, including both external and internal political shocks, and a stability or slow depreciation otherwise. With devaluation domestic costs are expected to decline.
Index Numbers (1995=100)
Thai Bath
Note: A downward movement indicates depreciation against the US dollar. Source: Derived from IMF, International Financial Statistics
Cambodian imports from Singapore, the largest supplier, are charged in Singapore dollars. When sold in the domestic market these goods are priced in US dollars for large value products, but also in riel for goods of small value. Consumers can pay in either US dollars or the local currency at the prevailing exchange rate. The examination of dollarization impacts on Cambodias imports is complicated by the fact that one cannot maintain with assurance that the movements in the value of the riel/US dollar exchange rate do not enter into pricing decisions for imported products in the local markets. There is a
104
So garments exports can be expected to become cheaper in the short run. Nonetheless, there are several things that may turn the garment industry into a long term loser. First, devaluation can result in a generalized domestic price increase, annihilating any artificial price competitiveness gained through devaluation. We must take note of the dynamics of Cambodias labour market. Even now, with wages paid in hard currency, we often hear that garment workers have struck for higher pay. Besides, there are many instances in which competitiveness has been partially offset by high interest rates, given that presently the manufacturing sector is the largest debtor to the banking system. Thus, devaluation and inflation will cause chaos, force firms to adjust to their costs and make longer-term planning difficult. Thus a large devaluation may not be effective to gain competitiveness and could be costly in terms of output loss. Second, unlike traditional exports, Cambodias garments exports rely heavily on imported intermediate goods, which is less of an issue for many of its competitors who can take advantage of both cheap local resources and devaluation. In a small open economy such as Cambodia, the depreciation of the exchange rate does not result in a sustained reduction in costs to the export sector. Instead, depreciation in the context of de-dollarization would serve to increase the cost of intermediate goods used in the productive sector. There is another question as to what should happen to the export proceeds? If the authorities do not allow exporters to hold foreign exchange accounts, they face a
foreign exchange risk which can lead to capital flight. In contrast to the garment industry, exports of agricultural products such as rice and fish have either low or no import content. Cambodias agricultural and fisheries exports are therefore largely unaffected by dollarization. This sector may react to changes in exchange rates of the riel relative to other currencies as in the case of normal sovereign economies without dollarization. From this point of view, agriculture seems to be a long-term winner from de-dollarization.

The trade issues discussed above indicate that, on the one hand, dollarization has made it difficult for Cambodias exporters to adjust their costs to changes in their competitors prices. Certainly, it is not an attractive position to be in, but the idea that the authorities must de-dollarize for the sake of increasing competitiveness is a fallacy. It would not be advisable for the authorities to take action to support competitive devaluation by a sudden move to a full riel economy. Moreover, the authorities would not want to risk their efforts to restore public confidence in macroeconomic stability by adopting such a policy. On the other hand, dollarization has helped improve the cost structure of intermediate goods for domestic production, and is thus very supportive of an economy that remains highly dependent on imports. Because the US dollar, rather than the riel, dominates in terms of determining demand for imports, changes in its exchange rate against the currencies of supplier countries tend to impact the volume of Cambodias imports.
105
Finally, devaluation is only a short-term tool. A sustainable export-led economy cannot be achieved through devaluation. A sustainable solution must be found, inter alia, in improving productivity, reducing unnecessary additional costs to exporters and removing other barriers to trade (see Chapter 5).
indicates that confidence in the riel has improved. What is clear to everyone is that dollarization is not the result of a policy decision, but is a consequence of a low confidence in the political and economic atmosphere, which drove economic agents toward voluntarily replacement of their local currency holdings with US dollars. Without confidence in the environment, people can neither act efficiently in the short run, nor can they make proper long-term plans. Restoring confidence and providing a stable macroeconomic environment has been very important for Cambodia, given the years of instability in its recent history. Dollarization has so far delivered this change to Cambodia while taking away the flexibility and independence of the countrys monetary policy and complicating the income distribution and several other things as discussed above. While there is comfort in having the flexibility to choose instruments to fit the countrys needs, experience so far has shown that such flexibility could not ensure the stability that was needed even more. Therefore, the stability and predictability that are requisites for sustainable economic development and poverty reduction may be worth bearing the costs of dollarization, at least as an interim measure. Interim is the key word here: de-dollarization is acceptable as a long term policy direction but should not be implemented immediately. We should, then, consider this as a transitional period during which the authorities should make the best use of the opportunity offered to build the foundations for a future economy without dollarization. Policy directions have to be directed toward putting in place further
6. THE CHOICE OF MONETARY SYSTEM 6.1 Dollarization as an Interim Measure

The problems we have considered in relation to dollarization in Cambodia clearly do not admit of an easy solution. Analysis of its costs and benefits for economic development and poverty incidence in Cambodia is not very conclusive and any solution hinges essentially on political considerations. However, a conclusion from the above discussion is that unless a country has been so devastated by successive monetary and fiscal policy failures and its population has totally lost confidence in the capacities of the government and the monetary authority to provide macroeconomic stability, the adoption of a full dollarization does not seem to be an option. In the Cambodian case, local currency in circulation has risen remarkably in recent years, while inflation remained low or turned negative and the exchange rate of the riel largely stabilized. Growth of currency outside the banking system was 17% during 2001 and accelerated to 23% during the first half of 2002 compared to a negative increase of -4% and a 1% increase in 1999 and 2000 respectively. There has been increasing demand for local currency which in turn
106
pillars to support private sector development. This would include strengthening the rule of law, increasing discipline and governance, building a strong banking system and sound banking practices etc.
to switch to the riel. The de-dollarization process should be implemented, without harming growth of economic activities, by making efforts to increase demand of the general public for the riel. The current increase in demand for local currency could be a result of the recent efforts of the government and the central bank to encourage the use of the riel. Various measures have been so far adopted. These include the following:
Since the beginning of 2000 an increasing portion of public sector transactions have been required to be effected in local currency. Payments by consumers for the use of public services and public utilities such as water and electricity are made in riel. Furthermore, companies and other tax payers are also requested to settle their tax liabilities in local currency. Within the framework of reforming the operations of the National Treasury, in November 2001 the MEF introduced a requirement for payment of tax directly to the National Treasury accounts held with the NBC. Prior to that, tax collection had been effected through payment to the Treasury itself. National Treasury data indicate that tax revenue in riel rose to more than 97% of total tax revenue in 2001, from some 95% in 2000. Based on the current success of the above measure, the government may consider expanding its local currency requirement to the collection of non-tax revenue: by contrast with tax revenue, the riel share of total non-tax revenue decreased from 35% in 2000 to 26.5% in 2001. On the expenditure side, the local currency share also declined from about 77% of the total in 2000 to 74% a year later. This may reflect
6.2 Steps toward De-dollarization

Based on what has been demonstrated above, de-dollarization should be seen as a medium- to long-term goal of the government policy in the monetary sphere. The length of the interim period will depend on the ability of the government to move forward with the reforms needed to increase the confidence of the public in the local currency. Removing the use of the dollar cannot proceed by forced conversion of foreign currency deposits into domestic currency. While this may at first seem attractive and appear to have quick results, there is a high probability that such a measure would stimulate capital flight and drive the dollarized economy underground. This would re-create all of the rent-seeking opportunities that Cambodia with its structural reforms has tried so hard in recent years to eliminate. Furthermore, associated underground activities are likely to lead to large distortions and inefficiencies in money and financial markets, seriously inhibiting the development of financial intermediation, and ultimately investment and growth. Implementation of a de-dollarization policy is not, therefore, a matter for legislation. It implies the creation of an economic and financial environment in which it is clearly to peoples advantage
107
the increasing need of the government to purchase goods and services, such as capital inputs and foreign experts associated with donor-financed projects, from abroad. If the increasing need is justified, the switch from disbursements in US dollars to the riel appears to be unjustified. On the whole, however, to move toward a de-dollarized economy, the government should consider effecting all domestic market expenditures in local currency. Furthermore, in order to expand the scope of riel-denominated transactions the authorities may as well consider requesting all international organisations in Cambodia to use the riel for financing all their local costs.
The NBC has also made an effort to replace old, dirty bank notes with new ones. Because cash based transactions are dominant in the local markets and because there are limited channels by which notes can return to the central bank, a massive volume of very, very old bank notes remained in the circulation. Currency traders and other market participants claim that this has been a factor discouraging riel preference among the population. Because there is a demand for new notes, traders have developed a business of exchanging old notes against new ones while charging a margin of about 5%. Obviously, the many poor cannot afford to enter into such a business. Therefore, besides contributing to increasing the demand for riel, another implication of requesting the direct payment of tax to the central bank is that it greatly facilitates the collection and removal of old, worn and torn bank notes from circulation and replacing them with the new ones.17
The NBC has made efforts to maintain the value of the riel. It has actively exercised its right to perform foreign exchange interventions to attenuate excessive fluctuations of the exchange rate and to influence its evolution with the view to reducing uncertainty. The NBC has entered the foreign exchange market to purchase foreign currency as and when market conditions allow. Direct sales of US dollars have been continuously undertaken to accommodate the demand of the Electricit du Cambodge (EDC), the state-owned electricity company, since this company collects income in local currency, but a very large proportion of its total expenditure is in US dollars. Without this NBC accommodation of the EDCs need, the company would have to rely on the foreign exchange market to purchase dollars, putting pressure on the riel. In general, taking into account its target for international reserves, the NBC is able to meet around of the annual demand of the EDC.
Alongside those efforts, in the context of the de-dollarization process, the authorities might take the following further options into consideration.
To stimulate riel savings there is a need to review the provision of the law on taxation that imposes a tax rate of 5% on interest income on deposits. Although the Law was adopted in 1997, the provision on taxation of interest income was only scheduled to come into force in August 2002. This measure might impact adversely on savings, especially small savings. To ease the situation the government might consider establishing a
17 There is a view among some market participants that removing very dirty notes from circulation should impact positively on the health of holders, though there is no way to measure the effect.
108
threshold, below which small depositors could enjoy a tax exemption. It might also consider differentiating between earnings generated from deposits in riel and in foreign currency. A tax rate in favour of local currency deposits might be helpful in stimulating the acceptance of riel as a store of value.
To move towards de-dollarization, the authorities need to establish a mechanism to develop intermediation in local currency by the commercial banking system. Banks would be more willing to accept riel deposits, if they could find riel-denominated financial assets against which to match their riel liabilities. In this regard, the introduction of treasury bills seems to provide an appropriate solution. In particular in the current circumstances that commercial banks have excess liquidity, treasury bills could serve as a tool to mop up excess funds. In addition, by the purchase or sale of treasury bills the central bank would have the opportunity to put open market policy into effect, while the bills would also help to promote the development of money/interbank markets. As for the government, raising money through such a domestic debt instrument would avoid inflationary pressure and the funds could be used independently of donor-funded resources which are usually tied to projects the donors themselves have approved. Consultation with commercial banks and representatives of the private sector is very crucial in defining the modality, settlement rules etc., for the Treasury bills. Initially the bill could be designed for subscription by commercial banks and only later offered to the public once it has been seen to function well and the credibility of the government increases. It is important
that the instrument guarantees savers the availability of a real positive interest rate.

Dollarization is not the result of a policy decision, but a consequence of low confidence that is exacerbated by prolonged uncertainty. So far dollarization has been very useful in helping to stabilize the economic system and prevent capital flight. It gives Cambodia a breathing space but is not a sufficient step towards achieving sustainable economic development and equitable distribution of growth. Dollarization intensifies the issue of income inequality, since it breaks up the society into a group with strong currency and strong purchasing power and another group that is more vulnerable to exchange rate risk. The standard of living of vulnerable and poor segments of the Cambodian population can suffer with the fall in the value of the riel relative to the group that has access to the strong currency, unless the authorities adopt appropriate measures for income redistribution. Dollarization also limits the outreach of formal financial intermediation into the rural and poor areas, as the poor operate in local currency, while financial intermediaries, including NGOs and MFIs prefer USD operations (see Chapter 6). However, immediate or forced de-dollarization could make even more people worse off. In that case, Cambodia would risk jeopardizing the advantages of the monetary economy. The return of inflation and unstable money triggered by renewed uncertainty could lead people to fall back on barter trade or use gold in place of currency. Business activities could decline and with them Cambodias total national income. And this
109
would mean an increase in poverty and not its reduction. Consequently, as the macroeconomic situation has been stabilized, the authorities need to take the opportunity to improve the functioning of the financial sector by strengthening bank supervision, developing financial market infrastructure, and ensuring the availability and enforceability of necessary legislation. Improved financial intermediation with enhanced public confidence in the financial system and in the domestic currency would expedite the de-dollarization process.
rate, while also taking into account the target set for international reserves. Under the PRGF, tight monetary and fiscal policy has been pursued. The restrictions imposed on fiscal deficit financing by the banking system, especially the central bank, coupled with the ceiling on net domestic assets of the NBC and a floor on its net international reserves have been successful in promoting macroeconomic stability. The mix of the two policies has provided stability in the overall price level and the purchasing power of the riel, while contributing to GDP growth of about 6-7% per annum. This achievement has to be credited also to dollarization, which emerged as a consequence of low confidence and a persistent feeling of uncertainty. Dollarization has provided both benefits and drawbacks for economic development and poverty reduction. The benefits of dollarization include switching public practice from gold to banknotes denominated in US dollars - thereby promoting monetization of the economy, preventing capital flight, promoting financial deepening, lowering the risk of currency devaluation, and facilitating Cambodias international integration process. All those issues facilitate and foster economic growth and have thus an indirect impact on poverty. On the negative side, dollarization means undermining the effective conduct of monetary and fiscal policy, damaging the pride of the nation, reducing the opportunity to earn income from seigniorage, and widening the gap between the urban
7. CONCLUSION
Poverty reduction is a daunting challenge in Cambodia given its very difficult history of political, social and economic development. To enhance the well-being of Cambodians, macroeconomic policies must conducted in such a way that they foster sustained and broadly distributed economic growth - by creating conditions that favour rising output, employment, incomes and a sustainable macroeconomic environment. Low public confidence, weak financial markets, lack of transmission mechanisms, and a high degree of dollarization of the economy, however, create obstacles to the effective conduct of monetary policy and exchange rate policy. Instruments such as repurchase agreement and statutory reserve requirement have proved to be inapplicable in managing money supply and promoting growth. A managed floating exchange rate system supported by prudent foreign exchange interventions, on the other hand, has been found to be more effective in stabilizing the movement of the exchange
110
dollar-based economy and poor rural areas where riel are normally used. The recent increase in circulation of riel, without pressuring macroeconomic stability, seems to reflect efforts of the government and NBC in encouraging the use of the riel and indicates that there is no compelling reason to go to full dollarization. Although the recent low or negative inflation represents a remarkable achievement of the existing policy settings, such settings no longer seem appropriate to address the issue of poverty reduction. To be more pro poor, monetary and exchange rate policy and also fiscal policy need to be more proactive and there is a strong case to review them in order to move faster to achieving the governments ultimate goal. In other words, stimulating more broad-based growth, ensuring more efficient and developmentoriented resource mobilization, and restoring monetary sovereignty to form a broader basis for poverty reduction should now be the focus of monetary and fiscal arrangements. In this context, some policy options have been suggested throughout the chapter, which include in particular the following: i. Depart from tight policy stance and allow some inflation to grease the economy; ii. Establish an appropriate balance between accumulation of international reserves and the level of net domestic assets, which should stimulate domestic demand. iii. Address distortions to financial intermediation by undertaking a systematic and properly sequenced banking and financial reform;
iv. Adopt some immediate mechanism to bring banks and potential savers more closely together, especially the introduction of a public education campaign to provide the general public with knowledge on banking services and to reduce uncertainty. v. Undertake a stepwise approach to de-dollarizing the economy by avoiding a forced conversion, promoting the money function of the riel and providing riel denominated assets, especially the treasury bill. In sum, from the point of view of poverty reduction, monetary and exchange rate policies need to support increasing economic integration of the rural areas into the national market, in particular, by supporting higher growth that is more equitably distributed - which means more employment opportunities, increased banking intermediation, and a reduced gap between rich and poor.
111
APPENDIX
Table 3A.1 : Cambodia - Monetary Survey, 1993 - 2001 (billion riel, end of period ) 1993 Net foreign assets Net foreign assets (MA) Foreign assets (MA) Foreign Liabilities (MA) Net foreign assets (DMBs) Foreign assets (DMBs) Foreign Liabilities (DMBs) Net domestic assets Domestic credit Net claims on government Claims on government Deposits of government Non government State enterprises Private sector Other Restricted deposits Capital & Reserves Others Broad money Money Currency outside banks Demand deposits Quasi-money Time and savings deposits Foreign currency deposits 119 37 58 -21 82 240 -158 215 343 175 206 -31 168 6 162 -128 -27 -181 79 333 204 190 14 130 9 121 1994 391 229 305 -76 162 326 -165 60 386 143 215 -72 243 6 237 -327 -29 -484 187 450 200 176 23 250 18 233 1995 550 304 485 -181 246 412 -166 99 446 148 217 -69 299 5 293 -347 -29 -472 153 649 278 251 28 371 5 366 1996 881 535 723 -188 346 507 -162 31 567 128 214 -86 440 5 435 -536 -82 -656 202 912 329 300 29 583 8 575 1997 1178 814 1039 -225 363 564 -200 -115 697 54 212 -158 643 6 637 -811 -47 -995 231 1063 385 356 29 678 13 665 1998 1726 1422 1676 -254 304 528 -224 -496 839 179 289 -110 660 6 655 -73 268 1230 543 509 34 687 20 667 1999 2019 1649 1924 -275 370 585 -214 -576 876 103 283 -180 773 10 763 -80 265 1443 532 490 42 911 32 879 2000 2589 2104 2389 -285 486 659 -173 -759 904 3 272 -269 901 3 898 -86 214 1831 540 495 45 1291 46 1245 2001 3081 2429 2740 -311 651 847 -196 -877 868 -75 271 -346 943 7 936 -100 314 2204 610 578 32 1594 56 1539
-1335 -1453 -1663 -1744 -1529 -1638 -1791 -1959
112
Table 3A.2 : Cambodia - Exchange Rates, 1990 - 2001 Market exchange rate year average 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 484 856 1363 2906 2576 2478 2660 2990 3789 3826 3868 3931 Y-Y change 76.9 59.2 113.2 -11.9 -3.8 7.3 12.4 26.7 1.0 1.1 1.6 Offical exchange rate year average 412 704 1267 2689 2544 2451 2758 2946 3744 3808 3845 3919 Y-Y change 70.9 80.0 112.2 -5.4 -3.7 7.1 12.3 27.1 1.7 1.0 1.9
Table 3A.3 : Cambodia - Inflation, 1994 - 2001 CPI 1994 1995 1996 1997 1998 1999 2000 2001
CPI: July, August, September 1994 = 100
Y-Y change (End of period) 1.06 10.03 9.22 13.29 -0.54 -0.81 -0.61
Y-Y change (Average period)
104.23 105.34 115.91 126.60 143.42 142.64 141.49 140.62 7.15 9.03 13.7 4.03 -0.79 -0.6
113
4
114
Fiscal Policy
1. INTRODUCTION
This chapter reviews the possibilities for using fiscal policy in order to induce a more pro-poor pattern of economic growth. It is divided into four sections. In the second section, we review Cambodias macro-fiscal record, as well as institutional and macroeconomic strategic documents, highlighting major fiscal trends and policy positions. In the third section we present a series of deductions regarding aspects of Cambodias current macro-fiscal strategy and point out the implications for the macroeconomics of poverty reduction. In the fourth section we present a summary of principal analytical conclusions together with some policy implications that can be derived from these conclusions. We then present some concluding comments on aspects of the overall macro-fiscal strategy currently in place in Cambodia which could serve as inputs for making that strategy more pro-poor.
The macroeconomic of poverty reduction in Cambodia Fiscal policy
2. THE MACRO-FISCAL SITUATION: AN OVERVIEW

Table 4.1: Macro-fiscal statistics 1993-2002 (Billion riel) Item Total revenue Of which: central government Tax revenue Direct taxes Of which: Profit taxes Indirect taxes Of which: Excise taxes VAT (net VAT refunds) Trade taxes Nontax revenue Capital revenue Total expenditure Current expenditure Wages Civil administration Defence and security Nonwage Of which: Operating expenditures Interest Capital expenditure Locally financed Externally financed Current balance Overall fiscal balance Overall fiscal balance (including grants) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 290 590 643 749 534 27 18 164 57 ... 344 176 39 1,441 813 354 125 229 459 345 881 597 44 35 206 74 ... 347 271 12 1,260 808 382 132 250 426 322 943 1,338 1,316 679 966 56 83 42 61 248 433 76 92 316 376 433 230 357 33 15 1,571 1,927 941 1,119 451 525 154 194 297 331 490 572 372 410 1,442 1,408 1,055 136 101 500 113 371 390 356 31 2,119 1,223 512 211 301 677 497 1,561 1,520 1,128 140 113 571 155 403 376 424 1,834 1,790 1,279 127 96 686 220 452 430 530 25 2,367 2,682 1,391 1,620 488 524 259 265 1,052 676
234 365 445 9 21 14 8 19 75 104 3 9 ... ... 160 281 321 56 225 190 0 1 8 608 1,009 1,206 373 674 695 293 326 101 110 192 216 380 368 324 290
1 2 13 10 15 22 21 25 235 335 511 629 452 630 808 896 1,062 23 79 57 62 110 120 303 303 362 212 257 454 567 342 510 504 592 700 -83 -85 -59 -103 61 -32 205 188 160 189 -318 -419 -563 -692 -379 -628 -589 -677 -806 -848 -69 -105 -113 -38 -286 -247 -293 -410 -369
Source: Data provided by IMF Country Office, Cambodia
The history of Cambodias macro-fiscal development across the 1990s can be divided into two phases (a) the period from 1993-1997 and (b) 1997-2002. 1997 marks a structural break because of the east Asian crisis and domestic disturbances. IMF staff estimates, reproduced in Table 4.1 above, provide consistent and comparable data spanning both these periods from 1993 onwards. The first phase of Cambodias development marked attempts by a new administration to take control of fiscal operations in a
country still divided by internal war, with only rudimentary fiscal administrative structures in place, inherited from the UNTAC administration. Following elections in May 1993, the government began formulating a comprehensive macroeconomic and structural reform programme with support from the international community. Major customs reforms were instituted and a taxation structure established. Non-tax revenues began to increasingly flow into government coffers rather than being appropriated by private agents. The bulk of revenues were generated by taxes on
115
international trade and turnover taxes, with domestic excise accounting for a relatively small share. Cambodias domestic revenue as a share of GDP doubled between 1991 and 1994. However, the revenue/GDP ratio stagnated from 1996 to 1998, falling sharply in 1998 as a consequence of external and internal disturbances (Table 4.2). Thus, a promising start on the resource mobilization front can, with hindsight, be judged to have been unsustainable due to adverse developments exogenous to the fiscal system. The reestablishment of stability since 1999 has seen domestic revenues recover and grow slowly, chiefly as a result of growth in non-tax revenues.
Table 4.2: Cambodia Revenue GDP ratios 1991-2002 (%) Year 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Tax Non-tax Revenue/G Revenue/G revenue/G DP DP DP 4.4 2.3 2.1 6.2 5.4 9.6 8.9 8.1 9.7 8.1 10.6 11.2 11.7 12.1 4.4 4.3 5.9 6.2 6.5 6.6 5.9 7.7 8.2 8.4 8.5 1.8 1.0 3.7 2.6 2.2 3.0 1.9 2.8 2.8 3.2 3.5
In the first phase, current expenditures accounted for over half total expenditure and the government consistently incurred a current fiscal deficit (Table 4.1). This deficit was financed largely by foreign aid receipts from 1993 to 1997. Domestic public investment was therefore completely financed by foreign savings. A significant majority of recurrent expenditure was on military spending and civilian wage expenditures, with some donor financed current expenditures providing for basic non-wage operating expenses. Since 1997, Cambodia has enjoyed a period of relative political stability that has allowed RGC to focus on achieving medium-term fiscal sustainability. This requires RGC current expenditures to be financed entirely out of domestic revenue. As the data in Table 4.1 demonstrates, since 1999, Cambodia successfully secured a small surplus of current revenues over current expenditures, in contrast with many developing countries. Fiscal discipline is adequate to ensure that despite macro-economic shocks (such as the 2001 floods), there is no recurrent deficit. This effort is complemented by active donor assisted capacity building initiatives to reform public expenditure management and improve fiscal and administrative governance. However, Table 4.1 also reveals that revenue mobilization is inadequate for providing sustainable domestic financing of growth-enhancing and poverty-reducing investments. The period 1999-2002 in particular has also been one in which several key changes have been initiated in Cambodias macrofiscal structure. First there is a concerted
Source: World Bank 1999: 113; World Bank 2003: 137.
116
attempt to change the revenue structure, to reduce the share of taxes on foreign trade and to increase the share of taxes on domestic goods and services. This change is broadly revenue neutral. Tax collections from international trade activities are declining very marginally in absolute terms while there are significant increases in VAT collections and non-tax revenues as well as direct tax collections although the last contributes a small share. Thus, the modest growth in the revenue/GDP ratio comes largely from increasing absolute collections of direct taxes and non-tax revenue, with VAT growth offsetting the decline in international trade tax collections ratio. In the case of public expenditures, Table 4.1 indicates that current expenditures have risen slightly. In terms of GDP share, such expenditures have risen from under 8% in 1997 to 11.3% in 2002. This increase has been marked by a fall in the share of wages and salaries and a rise in the share of non-wage operating expenditures. As we will show later, the main beneficiary of the rise in the expenditure/GDP ratio has been the social services sector, principally health and education, especially since financial year (FY) 2000. RGC fiscal policy is thus to sustain macroeconomic and external sector stability by ensuring that current expenditures do not exceed current revenues and to avoid recourse to bank financing of the budget deficit. The strategy for doing this in recent years has been to compress civilian operating expenditures so as to ensure that current expenditures match current revenues.
2.1 Institutional and Macroeconomic Strategic Initiatives: Fiscal Components

There is no dearth of strategic and policy documentation, on macroeconomic and macro-fiscal issues in Cambodia. Since 2000 there have been several major policy documents on the subject, each prepared with generous donor assistance. Each of these documents claims to have poverty reduction as the overarching goal. The macro-fiscal content of these documents are reviewed below as these constitute important strategic milestones linking fiscal policy with poverty reduction. 2.1.1 The World Bank Public Expenditure Reviews The first of these (World Bank 1999) provided a comprehensive assessment of the major macro-fiscal challenges in Cambodias development history in the 1992-1998 period and set the agenda, in many ways, for Cambodias future development strategy and the key assumptions underlying the strategy. While acknowledging that Cambodia made considerable progress towards obtaining a stable macroeconomic structure in the 1993-96 period, the report found significant regress since 1996. The cause was clearly pinpointed. Weak physical infrastructure and, more important, inadequate capacity of human resources severely constrain Cambodias development potential (World Bank 1999: ii). Recommendations for future macro-fiscal strategy sought explicitly to address these constraints. The major recommendations were:
117
1. To increase resource mobilization by improving the tax collection effort (World Bank 1999: iii-iv). The report estimated that regulatory and technical reforms could substantially improve Cambodias tax effort from 8% in 1998 to close on 13% by 2002; 2. To improve the efficiency of non-tax revenue administration; 3. To reduce defence expenditure which crowded out spending on development activities; 4. To substantially increase outlays for social sectors financed from revenue enhancement and also savings from military demobilization; 5. Implement civil service reform to improve operational efficiency of public service delivery so as to improve the quality of public services without increasing expenditures on general services; 6. Increase outlays on non-wage operating expenditures; 7. Enhance public expenditure effectiveness by improving the quality of public expenditure management. The above recommendations were basically repeated in the 2003 public expenditure review (World Bank 2003) which stressed the need to raise tax revenues, switch spending from low to high priority areas, improve civil service reform and enhance public expenditure management. Notably, the later document remains silent on the issue of enhanced resource mobilization via domestic borrowing an issue that we consider needs urgently to be discussed
at the policy level (see sections 3.1 and 3.4 below). However, the 2003 PER makes two additional recommendations that were contained in the widely circulated first draft of this report and with which naturally we concur. The first is that much more attention needs to be given to two of the governments four priority sectors which have so far lagged in terms of having clearly costed targets that are integrated within the budget framework (World Bank 2003: xii). (Note, however, that in this document RGCs priority area of rural development is renamed as road transport.) The second is that Donors will also be challenged to rationalize their salary supplementation practices as the Government moves forward with its civil service reform program (World Bank 2003: xiv). We discuss the latter issue in section 3.6 below. 2.1.2 The PRGF Memorandum of Understanding The international monetary fund and RGC signed a memorandum of economic and financial policies and technical memorandum of understanding on 26 December 2001 that underpinned IMF financial support for RGC under the poverty reduction and growth facility (PRGF) which aims to accelerate growth and reduce poverty. As a loan document, this memorandum also contains annexes outlining a series of policy conditionalities. The document (IMF 2001) emphasizes two urgent fiscal priorities:
The need to improve revenue collection through better revenue administration;
118
The need to optimize expenditure efficiency.
2.1.3 The NPRS (2002) The National Poverty Reduction Strategy (NPRS 2002) for Cambodia emphasizes two issues in the macro-fiscal context: 1. Increasing investments in health and education to improve human development achievements and enhance human capabilities; 2. Maintenance of macroeconomic stability and investment by enhancing the efficiency of fiscal processes so as to maximize revenue and optimize expenditure allocation decisions so as to maximize growth and human development. The NPRS presents a macroeconomic framework for the period 2003-2005 which seeks to secure modest increases in government revenue and expenditure/GDP ratios while maintaining an overall fiscal deficit under 5% of GDP. Increases in the revenue/GDP ratio are sought through fiscal consolidation, i.e. improved fiscal management (NPRS 2002: 129-130), while expenditure increases are required to be targeted mainly at securing increases in social sector spending, in terms of both share in total spending and social sector spending as a percentage of GDP. The assertions underlying the macroeconomic framework used are not analytically justified (for example, those concerning the desirable size of the fiscal deficit). The policy matrix for 2003-2005 contains a series of prescriptions for fiscal reform which deal almost exclusively with expenditure management and revenue administration. While some concerns are expressed about
Thus the accent of the PRGF is on technical efficiency and not on policy issues. Following from the PER (1999) it assumes that the main objective of macro-fiscal policy is to secure technical improvements to increase the efficiency of fiscal policy implementation. Given the lack of policy content the main interest in this document remains its coercive content, i.e. the conditionalities. These, again are mainly technical in nature requiring improvement in cash management procedures, customs codes etc. (IMF 2001: 13-14). However, certain benchmark conditionalities continue to be enforced. The most important ones are (IMF 2001: 8):
a ceiling on net domestic financing of the budget; zero foreign commercial borrowing; minimum maintenance of foreign currency reserves; a ceiling on National Bank of Cambodia credit to government.
Thus this document is both technocratic and prohibitory. The underlying strategy is to improve fiscal efficiency through technical improvements in fiscal governance and broad but restrictive prohibitions on credit acquisition to maintain a conservative fiscal stance.
119
the need to study actual distributional impacts of VAT and land tax exemptions currently being applied, explicit mention of policy goals for revenue growth or expenditure allocation are generally absent. The NPRS identifies four priority sectors from a poverty reduction perspective. These are: health, education, agriculture and rural development. It is noteworthy, however, that only health and education are slated to receive substantial increases in expenditure to 2005. Their combined share of current expenditure will increase from 22.4% in 2000 to 32.9% in 2005. Agriculture and rural development, on the other hand, will increase their share from 2.9% to 4.4% over the same period (NPRS 2002: 134). Proposed expenditures are integrated with the Medium-Term Expenditure Framework (see 2.1.5 below). 2.1.4 The SEDP-ii (2002) The Socio-Economic Development Plan II (SEDP-ii 2002) is presented by RGC as the Royal Government of Cambodias national economic growth and poverty reduction strategy (SEDP-ii 2002: 1). The strategic message of the Plan is that economic growth is a pre-requisite for poverty reduction and the key to growth is private sector development, which will be achieved largely through improvements in the governance environment. (SEDP-ii 2002: 3). This very specific strategy ought to be backed by a complementary macro-fiscal strategy designed to maximize private sector development and improve governance. Unfortunately the SEDP-ii contains no such strategy and the intention apparently is
to rely on other strategic documents to provide a complementary strategy. While the SEDP-ii does provide a public investment programme (PIP) consisting of a list of priority projects in different sectors, these projects have not been costed or justified within an overall capital resource envelope. In the words of a senior government economist the PIP is nothing but a shopping list. 2.1.5 The D-MTEF In the 2002 Budget circular RGC stated its intention to introduce multi-year budgeting which better links development plans, programmes and budgets within a Medium term expenditure framework (MTEF). This framework is to be produced by the finance ministry with long term peripatetic advice from an expatriate principal fiscal adviser. The original intention was for the first exercise to be completed by 2003 (ADB 2001). However, the process was advanced by a year and, in June 2002, RGC produced a draft MTEF document incorporating a three year macro-fiscal medium term expenditure framework for 2003-2005 (D-MTEF, 2002). The D-MTEF incorporates sectoral MTEFs for two priority sectors, health and education. MTEFs are technical exercises that derive their policy inputs from the wider policy process and seek to reconcile the trade-offs between competing demands for resources within an overall agreed feasible resource envelope, and in accordance with strategic priorities that are decided at the highest level. Thus, MTEFs require a strategic mandate with which to prioritize resource allocation. They further require demonstrable evidence from each sector that the sector spending plans are in accordance with sector priorities
120
and have been designed to maximize operational efficiency i.e. at minimum possible cost. A MTEF demonstrates its effective incorporation of governments strategic concerns by incorporating a policy goals and strategies paper which seeks to provide a critical review of existing strategic documents (such as those described above) and producing, after consultation a statement of government policy goals and strategies. This document is then approved at the highest political level (in Cambodia, the Cabinet). The document then binds all actors, and serves as the main justification for enforcing the MTEF. Section 2 of the draft Cambodia MTEF (D-MTEF, 2002: 2) incorporates the policy goals and strategies component. It aims to achieve a balanced and sustained economic growth rate of 6%. It identifies the following macroeconomic and institutional strategies for achieving this growth rate 1. Implementation of fiscal efficiency reforms and maintenance of stable macro environment; 2. liberalization of trade and investment policies and accelerated private sector development; 3. financial sector development; 4. improved governance and regulatory environment, including improving the access of the poor to social and economic services by removing anti-poor distortions in product and factor markets.
The MTEF also identifies four priority sectors in line with the NPRS and the SEDP-ii. These are health, education rural development and agriculture. Each sector is expected to produce a sectoral MTEF, which can then be used to work out sector medium term expenditure. For the 2002 D-MTEF a matrix of objectives, constraints and policy directions (costed expenditure-generating activities) exists for health and education only. Section 3 of the D-MTEF presents a macroeconomic and medium term fiscal framework. This provides a range of macro fiscal projections for GDP growth, inflation the exchange rate revenue, and current and capital expenditure. Based on these projections the D-MTEF also derives figures for current surplus and overall (fiscal) deficit, for the period 2002-2005. (Table 4.3). Thus, this section effectively provides an estimate of total resources available to RGC for various recurrent activities and the overall fiscal sustainability targets
Table 4.3: Key macro-fiscal forecasts 20032005 (% of GDP) Item Total revenue Tax revenue Non-tax revenue Capital expenditure Current surplus Overall deficit
Source: D-MTEF 2002: 4
2002 2003 2004 2005 12.8 8.9 3.7 7.4 1.3 -5.9 13 9.3 3.7 11.5 6.8 1.5 -5.3 13.6 9.7 3.7 12 6.8 1.6 -5.2 13.9 10.1 3.6 12.4 6.6 1.5 -5.1
Current expenditure 11.3
Section 4 of the D-MTEF presents the Medium term Budget framework. This section tries initially to calculate the baseline budget for each sector i.e. the total expected expenditure
121
on wages and salaries, and operating costs, as well as expected commitments to social and economic subsidies (which the D-MTEF assumes will be maintained over the period at 2002 levels) which are considered to be unaffected by future policy changes. In Cambodia, given that donors are responsible for almost all capital spending, most capital expenditure is also taken to be part of the baseline budget. The total baseline budget is then deducted from the projected total expenditure for each year to provide the total available allocable expenditure for each year. (Table 4.4 below provides the calculations). The allocable ceiling thus indicates the quantum of resources which the government can switch between sectors depending on its strategic priorities.
Table 4.4: Baseline and allocable ceilings 2003-2005 Item Baseline ceiling million riels Allocable ceiling million riels Allocable ceiling (per cent of GDP) Total expenditure (per cent of GDP)
Source: D-MTEF 2002: 5.
2003 2410582 453551 2.9 18.3
2004 2564500 642592 3.77 18.8
2005 2752825 780035 4.2 19.0
Finally the D-MTEF proves a distribution of resources by sector (Table 4.5). It seeks to switch allocable resources away from defence and security towards health and education and to a lesser extent towards
agriculture and rural development This it does both in terms of the share of these sectors in total current expenditures and in terms of expenditure on these sectors in per cent of GDP.
Table 4.5: GDP expenditure shares (current and domestically financed capital expenditure as per cent of GDP) and sector shares in current expenditure (figures in brackets indicate forecasted share of that sector in total current expenditure) per cent Item General administration Defence and security Education Health Agriculture Rural development 2003 1.54 2.53(22.0) 2.13(18.5) 1.38 (12) 0.26 (2.3) 0.13 (1.1) 2004 1.45 2.5 (20.8) 2.34 (19.5) 1.5 (12.5) 0.3 (2.5) 0.15 (1.3) 2005 1.36 2.4 (19.4) 2.5 (20.0) 1.6 (12.9) 0.35 (2.8) 0.20 (1.6)
Source: Compiled from D-MTEF 2002: Tables 4 and 5.
122
Thus the D-MTEF provides a comprehensive and reasoned statement of macro-fiscal priorities and intentions for the 2003-2005 period based on an assessment of historical macro fiscal trends, a clear statement of assumptions (summarized in Box 4.1 below) and a sectoral allocation of free resources. Given the sparse policy input in the macro-fiscal area in other strategic documents like the SEDP-ii and the NPRS, it is likely that the MTEF will in effect provide the strategic blueprint for fiscal policy in the medium term18.
Box 4.1: D-MTEF assumptions
The D-MTEF makes the following general assumptions. While these are quantified in the actual document the actual numbers are confidential, with good reason as the precise numbers are market sensitive and could trigger anticipatory action by private economic agents. Hence only the generic assumptions are presented here: New tax measures and expansion of coverage of existing taxes will generate rises in revenues in each of the fiscal years 2003-2005.
Modest expansion of existing non-tax revenue sources will increase non-tax revenues in each of the three years 2003-2005. However, the growth in non-tax revenues is expected to be equal to or less than the forecasted GDP growth rate.
Administrative and efficiency improvements will raise tax revenues modestly in 2003 and 2004 Thus growth, rather than improved revenue effort will drive revenue/GDP increases.
Capital expenditures will not grow faster than GDP thereby resulting in a small fall in the capital expenditure/GDP ratio.
Current expenditures will rise as non-wage recurrent expenditure allocations in the priority sectors are enhanced. In addition current expenditure increases can be anticipated in 2004 and 2003 due to increased demobilization costs, some wage adjustments and higher transfers to communes.
18 Since the second Cambodia PER was not ready by the time the D-MTEF was presented to the National Assembly as a budget annex in December 2002, it appears that it will only influence policy marginally in the medium term. Clearly any new advice in the 2003 IFAPER could only influence the MTEF exercise for 2006-2008 without leading to policy confusion.
123
3. ANALYSIS OF CAMBODIAS FISCAL STANCE

In this section we present a series of deductions regarding aspects of Cambodias current macro-fiscal strategy and point out the implications for the macroeconomics of poverty reduction. Our first analytical conclusion, explained below, is that expansion of Cambodias resource envelope is limited in the medium term. This is so for two reasons:
Only modest increases in public resources can be expected through growth in the tax base and efficiency improvements in the medium term. Domestic resource mobilization through domestic borrowing is not feasible in the medium term.
public sector and so does not directly increase the countrys liabilities to the rest of the world. It can also increase the propensity to save to the extent that the borrowing reduces domestic consumption. It is, to this extent, a free transfer. It can, however, crowd out private investment and have distributional implications depending on the precise nature of the borrowing exercise. It can also have inflationary consequences, if domestic absorption increases significantly before public investment generates increased output. Finally, borrowing for government consumption has negative implications for growth if it reduces national saving by financing the borrowing out of private saving.
By borrowing internationally. This type of borrowing results in an increase in international debt and creates liabilities for the country. It represents a temporary increase in domestic resources, that can be amortized only if the pay-back from the investment of these resources results in an increase in GDP greater than the net present value of the loan. The greater the degree of concessionality in such borrowing (zero, is the limit when the international transfer is a full grant) the lower the net present value of the debt.
3.1 Domestic Resource Mobilization

A country can mobilize resources in the following ways: By increasing tax and non-tax revenues. This source of revenue is the most preferable from governments perspective, as it generates no future fiscal liabilities. However, taxation can have a variety of macroeconomic and sectoral effects on the spending and investment decisions of economic agents, which can have implications for equity, efficiency and growth.

By borrowing from the domestic private sector. This source of revenue is attractive since it represents a transfer from the domestic private sector to the domestic
In Cambodia, the bulk of resources accruing to government for recurrent expenditure are generated through domestic revenues. As we saw in table 4.1, in recent years Cambodia has not been borrowing directly for meeting recurrent expenditure needs. Current revenues pay for current expenditures and, in fact, generate a small surplus for investment. Hence, the greater the prospects for increases in Cambodias current revenues, the greater the potential
124
for increasing domestically financed public investment. In comparative terms, Cambodias revenue/ GDP ratio is lower than in other countries in ASEAN at comparable levels of development (IMF, 2001). This indicates that there is considerable potential to increase the revenue/GDP ratio. However, the various policy documents cited in the previous section are fairly pessimistic about the potential to increase Cambodias revenue/ GDP ratio. As the D-MTEF (2002: 4) states, It is assumed that the significant increase in fiscal revenues in absolute terms will be driven mainly by high and sustained growth of the Cambodian economy rather than a dramatic improvement in revenue effort. Revenues are expected to increase by one percentage point from 12.8% of GDP in 2002 to 13.9% of GDP in 2005. (The macroeconomic framework presented in SEDP-ii (2002: 98) is even more modest assuming revenue GDP ratios of 13.5% for each year in the 2002-2005 period.). Our discussions indicated that several foreign economists and donors feel that there is still considerable scope for revenue increases through administrative and technical improvements in the taxation structure and this view is reflected in the latest draft of the MTEF (August 2003: 47) in which revenue is projected to increase to 14.5% of GDP by 2007. The projected increase in the tax/GDP ratio in the 2003-2005 period is expected to be generated through increases in taxation of domestic goods and services and some modest increase in direct tax collections (SEDP-ii 2002: 99; also see Box 4.1 above). Non-tax revenues are expected to remain
constant in terms of their share of GDP. These modest improvements will be secured through strengthening tax collections and by reducing tax exemptions. At the same time there is a projected fall in revenues from customs duties (due to trade liberalization prior to accession to WTO in 2005). The net tax burden is therefore not slated to rise significantly. There is a simple explanation for this moderate forecast for Cambodias revenue growth even if it somewhat contradicts the assertions about private sector development in the policy documents. This explanation has two elements.
First, Cambodias direct tax collections account for about one per cent of GDP. Eighty per cent of this in 2000 came from taxes on profits (IMF 2001: Table 4). Unless there is rapid and sustained growth in corporate private activity, there will be no significant rise in the direct tax base. Hence, the implicit assumption in the policy documents cited must be that future increases in private sector activity, touted in so many policy documents, will occur mainly in the non-corporate private sector. If this is so, then, such increased activity will have a negligible fiscal impact, since increases in corporate and non-corporate tax rates are ruled out to encourage domestic private sector growth and to make Cambodia an attractive destination for private investment in a highly competitive environment.
Second, taxes on international trade have historically accounted for a large segment of total tax collections in Cambodia. In the 2000-2002 period, such taxes accounted for
125
33% of total tax revenue (over 60% if VAT on international goods and services is added on). With customs duties slated to reduce in the run-up to WTO accession, the best that can be hoped for is a revenue neutral increase in taxes on domestic goods and services. Customs duties and VAT are roughly equal in scope at the present time. The strategy contemplated is to increase excise duty collections from luxuries marginally and to increase VAT collections by improving administration. This provides little scope for revenue increase. It is therefore clear that the even the modest increase in the tax/GDP ratio planned for the medium term is, given the preceding discussion, based on fairly brave assumptions regarding the non-quantifiable benefits of current technical assistance to improve tax administration. There are no planned specific economic events or policy interventions that can enable a rise in the tax/GDP ratio. Hence, conservatism in revenue forecasting is well justified. In these circumstances, it would be natural to look at resource mobilization through domestic borrowing for public investment. Here, Cambodia is severely constrained by IMF conditionalities which prohibit government from any substantive recourse to domestic borrowing. While it is tempting to see this as yet another example of unwarranted Bretton Woods conservatism, a closer look at the facts reveals a rather complex picture. Leaving distributional considerations aside, domestic borrowing adds to gross investment and to growth if it does not substitute for
resources available for private investment and if the returns from public investment are sufficient to justify borrowing costs.19 In Cambodia the public savings/GDP ratio has risen quite sharply from 7.4% in 1995 to 13.2% in 2001 (IMF 2001: Table 4.1). This rise in savings reflects less an increase in the propensity to save than the increased channelling of domestic savings into the formal sector. Our discussions revealed that this is an extremely low ratio, for two reasons:
Underdeveloped financial structure (see Chapter 6).
Lack of trust in government. In many developing countries with underdeveloped private financial markets, sovereign borrowing for public investment has often been used as an instrument to extract savings kept in forms that do not permit them to be used for productive investment (chiefly non monetary tradable assets like gold etc. and money kept at home or entrusted to informal agents). This process is known as the realisation of savings. Our discussions revealed that a considerable proportion of savings are presently stored in this form, in addition to deposits abroad. It is difficult to obtain precise quantitative volumes for such savings but there is enough indication that it would not be unreasonable to suppose that such savings would be around 5% of GDP. If realised for productive investment this would be a substantial amount, and would considerably boost governments resource mobilization efforts. However, given the absence of any track record of such mobilization by
19 Note that in poor countries domestic public debt increases can have negative distributional implications, which have to be off set against the benefits of increased domestic resource mobilization. For details see Roy (1994).
126
government and a constant fear of political uncertainty, it is doubtful whether government would be able to access such savings, account for them, and deploy them without considerable technical assistance and a visible transparent accounting mechanism for public debt. Without such a mechanism visibly in place confidence in governments capacity to use these resources effectively and repay national debt will remain low, and such savings continue to remain unrealised by the public sector. The lack of confidence in government as a credible borrowing agent is perhaps the biggest obstacle to effective mobilization of domestic resources for development. In part this has to do with the recent record of political instability and government solvency as well as doubts about the safety of private resources in government hands. Many doubts were expressed during our discussions about the capacity of government to handle domestic resources and to abide by the commitment to use these resources exclusively for investments that yield tangible returns. This lack of confidence in government as a financial agent implies that RGC must be seen as an institution whose financial sovereignty is bounded in the perception of private savers. The reasons underlying this bounded sovereignty have to do with governance issues that lie outside the scope of this report. The implication, however, is unmistakable without a major improvement in the credibility of its financial sovereignty, RGC cannot expect to be able to use domestic borrowing as an instrument to improve savings realisation until financial markets are developed sufficiently.
The above institutional constraint is particularly ironic given the existing situation with bank credit. In the past year, deposit liabilities of commercial banks have been around 9% of GDP, while lending by commercial banks has amounted to around 6% of GDP. Thus, even with realised savings, commercial banks are unable to lend their total funds at prevailing rates of interest. In effect three per cent of GDP is idle, being neither consumed nor saved. Information we gathered indicates that a significant portion of these resources are placed on deposit with the National Bank of Cambodia, which then, inter alia, invests them in the Singapore inter-bank money market. Given the institutional background, this action (if information received is correct) is a perfectly understandable banking decision, though its sub-optimality from a development finance perspective is quite apparent.
3.2 Dependence on Donor Finance

Cambodia is heavily dependent on concessional donor finance for its development investments. This raises concerns regarding future debt service ratios. RGC is effectively prohibited from recourse to commercial international borrowing by a conditionality agreed with the IMF. Given the conditions set out in the previous paragraphs this seems quite reasonable given the governments bounded financial sovereignty. Hence the main sources of international finance are concessional grants and loans from multilateral and bilateral donors. These have traditionally dominated
127
public investment financing in Cambodia and are likely to do so in the medium term. Thus net foreign financing in 2002 is estimated to be 5.8% of GDP, while externally financed capital expenditure in the same year was 4.9% of GDP and locally financed investment was 2.4% of GDP (calculated from Table 4.1). Thus, around two thirds of total investment is financed by external concessional assistance. The RGCs reports on development cooperation (RGC 2001; RGC 2002) provide an estimate of the past and projected structure of such concessional finance. Generally disbursements are around 75% of pledges and while grants currently provide the bulk of development assistance to the country as a whole, project loans to government as a recipient have been increasing quite sharply since 2001. They presently account for 2.1% of GDP compared with 1.6% of GDP in 1998. This indicates that while the stock of the countrys overall external debt is rising quite slowly, governments direct debt service commitments are rising fast enough to cause concern regarding the contingent current expenditure commitments the rising debt stock generates. Hence, the present and projected grant:loan portfolio would make it imprudent for government to seek significant additional development assistance without reckoning on a concomitant rise in its future debt service obligations. Any further adverse change in the grant:loan ratio would exacerbate this fiscal constraint.
seeks to secure growth through a highly specific expenditure switching policy from general services to social services, i.e. health and education. On the face of it, these sectoral expenditures have been designed to optimize their pro-poor impact, and most official documents highlight these benefits. They do not, however, highlight the implicit and explicit costs, which need to be taken into account when assessing the net pro-poor benefits of the chosen strategy. To establish the above proposition we need to take a close look at the structure of government expenditure in the period 1997-2002, and its evolution from 2002 to 2005. Table 4.6 below presents data on the structure of government expenditure by functional classification in the period 1994-2002
3.3 Costs Associated with the Chosen Strategy

Given the resource envelope constraint, Cambodias pro-poor macro-fiscal strategy
128
Table 4.6: Functional classification of government current expenditure 1994-2002 (% of expenditure; figures in bracket are % of GDP) Secto General Defence and Social Economic Rural Education Health Agriculture Year Services Security Services services development 13.3 59.1 19.9 9.3 4.5 7.6 1.9 0.2 1994 (0.49) (0.83) (0.2) (0.02) (1.44) (6.38) (2.15) (1.01) 14.7 58.5 20.2 10.0 3.5 6.5 1.8 0.3 1995 (0.36) (0.66) (0.18) (0.03) (1.5) (6.0) (2.06) (1.03) 15.4 51.5 22.7 10.2 5.4 8.8 2.0 0.3 1996 (0.52) (0.84) (0.2) (0.03) (1.47) (4.93) (2.18) (1.0) 16.7 51.4 23.2 10.2 5.6 7.5 1.8 0.2 1997 (0.5) (0.67) (0.16) (0.02) (1.5) (4.61) (2.08) (0.92) 21.3 48.5 21.8 10.9 4.7 6.7 1.6 0.2 1998 (0.42) (0.6) (0.15) (0.02) (1.89) (4.31) (7.5) (0.97) 16.6 42.7 27.4 13.7 6.9 8.6 1.9 0.3 1999 (1.31) (0.66) (0.82) (0.18) (0.03) (1.59) (4.08) (2.62) 18.9 37.1 30.3 10.7 8.4 9.3 1.9 0.6 2000 (1.36) (0.83) (0.92) (0.19) (0.06) (1.88) (3.69) (3.02) 19.5 30.5 35 16.4 9.9 11.8 2.5 1.0 2001 (1.77) (1.07) (1.27) (0.25) (0.10) (2.11) (3.29) (3.78) 14.5 24.7 37 18.2 10.9 11.0 2.5 1.3 2002 (2.0) (1.2) (1.21) (0.28) (0.14) (1.6) (2.72) (4.07)
Note: 1994-2000 figures are final outcomes; 2001 figures are revised estimates; 2002 figures are provisional budget figures. Source: based on data supplied by Ministry of Economy and Finance.
The above table marks a clear pattern. Since 1997, the share of civil administration in total expenditure has been roughly constant through to 2001, even rising slightly in GDP terms. The biggest fall has been in defence expenditure, often called the peace dividend. This has resulted in a significant rise in the share of social services, particularly education and health. The Economic services share has also risen slightly though not for agriculture. Rural development continues to account for a small proportion of current expenditure. This macro-fiscal change in the structure of current public expenditure has not happened accidentally. Most major policy documents celebrate this change. The NPRS, SEDP-ii and the PRGF, in line with the World Banks IFAPER, all emphasize the
virtues of a poverty reduction strategy that improves human development by improving access to, and quality of publicly funded education and health services. As we saw in the previous section the D-MTEF explicitly seeks to continue with this strategy. It is clear from Tables 4.4 and 4.5 that while there are four priority sectors, it is health and education that are slated to secure the lions share of incremental resources in the 2003-2005 period These documents do not, however, directly point out the trade-offs involved in the choice of such a strategy in Cambodia. With limited scope for additional revenue mobilization, public expenditure on development benefits from reduced expenditure on defence and, to a lesser extent, on general services. The bulk of the resources thus released in the
129
past five years have been allocated to social services especially health and education. Key economic services have received relatively little by way of increased allocations. In particular, agriculture which provides livelihoods for over 70% of the population and rural development have received insignificant amounts of resources compared to general and social services, even though these have been identified as priority sectors in strategic policy documents like the NPRS and the D-MTEF. Far from being a strategy of balanced growth, the macro-fiscal picture reveals a clear decision to trade-off expenditure on economic services in favour of social development expenditures. The contemporary discourse on public goods within the world of development economics (World Bank 1997; Atkinson and Stiglitz 1987) contends that, to the extent possible, the state should provide for those public goods which the private sector is unlikely to be able to deliver and let the private sector deliver public goods which it is able and willing to do. In practice countries are encouraged to experiment with private provision even of non-basic health and education services, and it is rarely, if ever, the case that provision of public goods in economic services is encouraged. Irrespective of the pro-poor consequences of this strategy for the developing world at large, in Cambodia, the stark question is whether the private sector is at all able and willing to provide public goods, particularly in areas like agriculture and rural development. Given the narrow base of the domestic private sector, inadequate development of a legal and regulatory framework for private sector development, limited FDI in labour intensive economic sectors, and the
acceptance by most development agents that the process of private sector development is likely to be a long drawn out affair in short the very reasons which justify the acceptance of a pessimistic scenario regarding the use of public debt as a resource mobilization instrument by RGC it is doubtful that private sector development of economic services will complement the development of social services by government. The current strategy therefore implies that Cambodia sacrifices public investment in economic services, since private sector development is likely to be limited by several binding constraints (including those that limit the public resource envelope). The D-MTEF (2002) provides evidence that the present strategy is contemplated in the future as well. It forecasts a fall in the defence and general administration/GDP spending ratios of 0.18 and 0.13 percentage points respectively in the period 2003-2005. In contrast there is a planned rise in spending on social services of 0.57% of GDP (calculated from Table 4.3 above). The total forecast rise in contingency and debt service commitments is 0.40% of GDP (D-MTEF 2002: 12). The overall rise in the allocable resource envelope, is 1.3% of GDP (Table 4.3). Deducting debt service and contingency provisions, this leaves 0.9% of GDP as increased allocable resources. Almost two thirds of total re-allocated resources are earmarked for health and education. In contrast agriculture and rural development collectively receive just 0.16% (of GDP) which is less than a fifth of the total reallocated resources. Thus, the D-MTEF presents an explicit expenditure switching policy whereby the
130
entire projected fall in spending on general services and over half the projected increase in revenue is absorbed by social services, with only a small residual available for investment in current economic services and locally financed capital expenditure. This strategic choice is further reinforced when reference is made to the current and forecast allocation of capital resources from the SEDP-ii. At first blush it appears that transport and rural development secure the largest share of capital spending. However, when the social development, health and education sector capital spending programmes are aggregated, they collectively constitute 39.5% of total donor disbursements, with rural development and transport accounting for 26% of total disbursements. In future years too, it appears that there will be a sharp rise in education and health funding although the transport sector continues to receive generous capital investment and technical assistance from specific donors (particularly the ADB and Japan). RGC in fact commits significant resources to the social sectors, relying on donors to fund projects in transport, water and sanitation, and other infrastructure.20 Agriculture continues to receive a low share of total capital investment resources. Thus the importance accorded to public current spending on education and health finds a resonance in capital spending as well However, it is certainly not the case that all of the projected capital spending is expected to be delivered in social services. The infrastructure sector receives almost 50% of the total spending on new projects (SEDP-ii 2002).
This raises another serious issue: the strategy does not address the issue of adequate resources for servicing the recurrent expenditure commitments generated by donor financed capital investment projects. This is particularly a problem for economic services and infrastructure. The existing education sector development strategy and the proposed health sector development strategy both benefit from donor budget support for non-wage recurrent expenditure. Indeed the conditionalities underpinning the release of the ADB education sector programme loan incorporate a commitment from RGC to provide adequate recurrent support for the investment programme (ADB 2001). Thus while donors provide (or influence RGC to provide) budget support for non-wage recurrent expenditures on social services, there is no corresponding support for economic services. While, there is no doubt that the design of the current strategy is pro-poor in nature, particularly if one examines the carefully crafted sector development plans for the education and health sectors outcomes, it may not be the optimal strategy for income poverty reduction. Its appropriateness can be defended if a broader approach to poverty reduction is adopted, one that stresses improved capability to participate in social life and relative empowerment, particularly in view of Cambodias historical legacy in this regard (see Chapter 2). Devoting public resources to providing inclusive quality basic education to the poor also has undoubted long term positive effects on growth, by improving the skills base of the country and providing the younger poor
20 Particularly electricity supply which will be provided by BOT projects where possible.
131
with access to income generating opportunities. However, it must be clearly understood that investments in education do not in themselves generate such opportunities; at the very least a complementary income generation strategy is necessary for this latter purpose. In addition the benefits of enhanced education spending accrue in the first instance to young people and this limits the immediate and medium-term impact of education spending on the poor as a whole (see Chapter 2). It does not address the needs of poor adults, especially adult women. In the case of health, similar arguments apply though it can be argued that improved health care reduces the vulnerability of the poor, and increases their disposable income by reducing private expenditure on health and increasing labour productivity. Hence, fiscal strategies that focus exclusively on expenditure switching away from general services to social services run the risk of being sub-optimal from the point of view of poverty alleviation since they do not devote sufficient resources to economic services.
the spending effect of donor employment policies on the costs of using skilled human resource inputs in domestic economic activities, including public administration (see sections 3.5-6).
3.4 A Form of Dutch Disease

The scale of external assistance in Cambodia in comparison with domestic resources raises the question of whether the Cambodian fisc is subject to some form of Dutch disease. This question needs to be understood not in the usual context (Davis 1995; White and Wignaraja 1992) of aid inflows in comparison with export earnings but, rather in the context of:
The first issue has been typically posed in the literature in terms of the crowding out effect of aid on governments tax gathering efforts. Since increased tax effort involves higher political and administrative costs, the availability of aid can often reduce the incentive to increase tax effort. In the Cambodian context this does not seem to have happened the determinants of tax effort seem to be independent of aid inflows, largely because revenue resources are used cover current expenditures. Were this to reversed, the stringent conditionalities in the PRGF and other donor aide memoires and memoranda of understanding actually put the government at risk of losing, rather than gaining, aid inflows. In a conditionalitybased lending environment, fiscal discipline in the form of low inflation, limited fiscal deficits and zero current deficits is rewarded with more generous sanctioned assistance and vice versa. Hence the crowding out effect of aid on governments revenue efforts appears to be limited in Cambodia. The perverse incentive in the Cambodian fiscal regime comes when government attempts to secure domestic savings for public investment. For any investment project RGC can, in principle, choose between concessional external assistance and borrowing from the domestic private sector. In a previous section, we concurred with the widely held view within donor and central banking circles in Cambodia that confidence in RGCs capacity to manage
the opportunity cost of accessing a dollar of private sector savings for public investment.
132
domestic debt and deploy it with prudence was low, creating a supply problem. On the demand side too, generous concessional terms for investment resources from donors can act as a disincentive when RGC considers measures to generate domestic resources. While such resources do not generate contingent external liabilities for Cambodia, they do generate public sector liabilities that need to be paid off. With private sector lending rates of 15% and above, the portfolio of viable lending opportunities in Cambodia is limited. There is excess liquidity in the banking system, amounting to some 3% of GDP (see section 3.1 above), despite the fact that formal savings are low and the banking system is not the repository of most savings in Cambodia. Hence there is every incentive for the banking sector to lend to RGC at substantially lower rates of interest on a project by project basis. Even an interest rate of eight per cent could be viable particularly for economic services that generate a clear, costed, benefit stream. However, if RGC is able to access concessional resources for a fraction of that cost for infrastructure, agriculture, rural development, etc. then there is little incentive for it to seek private sector credit and, transitively, little incentive to reform its own image in the financial markets. Even if the resources are not actually forthcoming for such projects the potential access to such finance can act as a disincentive to consider seriously domestic debt financed public investment. Thus, while the existence of excess liquidity negates the possibility of crowding out and offers RGC potentially attractive finance for development, the incentive for RGC to access
such finance is limited given the potential availability of easy concessional money from donors who accept RGCs unbounded fiscal sovereignty (albeit with conditionalities) unlike domestic economic agents who, lacking recourse to conditionalities, seem reluctant to do the same. In such a situation there is little incentive for either RGC or the donors to speed up the financial reform process that would enhance RGCs credibility and remove the bounds on its financial sovereignty. It is this sense that there exists a Dutch disease in Cambodias fisc. We note here that since 2003 a limited experiment with sale of riel-denominated Treasury Bills to commercial banks has been carried out. As far as we can ascertain the experiment has worked smoothly. However, the market for such public debt is extremely limited since most commercial banks do not accept riel deposits and consequently have no interest in lending to the government. The only commercial bank that does accept substantial riel deposits is the state-owned Foreign Trade Bank. It is, therefore, scarcely possible to say that the market for government debt has been tested.
3.5 Fiscal Benefits from Civil Service Reform

The analysis of the spending effect of donor employment policies on the costs of using skilled human resource inputs in domestic economic activities, including public administration is linked with the question of the fiscal benefits from civil service downsizing. Starting with the first World Bank PER (World Bank 1999) several policy documents
133
have argued the fiscal case for a civil service reform, and some donors pursue this agenda quite aggressively with RGC. The argument goes as follows: Cambodias civil service is overstaffed and underpaid. It would make sense in terms of efficiency to downsize the civil service and pay the remainder a better wage, while at the same time releasing resources from general services for investment in social and economic services. Indeed this issue finds mention in the PRGF document as well (IMF 2001: 5-6). This document requires RGC to cap the wage bill, stating, The overall wage bill will be contained below 4% of GDP further staff reductions through contracting out and privatization will be carried out in 2003, contingent on establishing a donor supported safety net (IMF 2002: 6, italics added). There may be doctrinal and efficiency merits in changing the balance of employment between the public and private sectors and paying some civil servants better than others. These do not fall within the scope of this Chapter. The fiscal questions here are:
expenditure, and assuming a constant military wage bill (both heroic assumptions) the total sum available for rewarding civil servants comes to approximately 15 million dollars. If this amount is spread over the 160,000 strong administrative (excluding teachers, health workers etc) staff of the civil service then it provides 9.4 dollars per civil servant per annum. With average civil service pay at around $360 per annum it is clear that the resources significantly to increase civil service pay are simply not available given the IMF cap. By the same token the fiscal rewards from cutting the civil service are limited. Our discussions at the Council on Administrative Reform (CAR) revealed that a 20% cut in civil service size would yield approximately $ 960,000 an insignificant saving for the fisc as a whole and insufficient to significantly increase remuneration of the remainder of civil servants. Our discussions with CAR also revealed that present RGC strategy is to target the moneys available to priority mission groups. These would be clusters of civil servants, retained to perform specific time-bound assignments, who would be rewarded with enhanced pay, tentatively $2400 per annum. One thousand civil servants would be selected to staff these groups. This would generate an incremental expenditure of around $2000 per annum per civil servant (assuming that personnel in these groups would belong to the middle and upper echelons of the civil service), generating a cost of $2 million per annum for such groups. Thus, attempting further downsizing of general services has economic as well as
What kind of fiscal savings would be secured by capping the wage bill at 4% of GDP? To what extent could civil service pay be increased within this cap? The first question is a redundant one. Wages are already below 4% of GDP and are slated to fall to 3.6% of GDP in 2002. This leaves approximately 0.4% of GDP for RGC contributions to safety net measures and for improved civil service pay i.e. a sum of 60 billion riel approximately. Assuming donor grants finance the entire safety net
134
political limits. Downsizing of non-wage operating expenditures would have a detrimental effect on the efficiency of present and future investment. Downsizing of the wage bill will not yield significant incremental resources for investment in economic services or, even, resources to significantly enhance average civil service pay. The present targeted, output driven, strategy of selective remuneration seems to be the only feasible and optimal one for the immediate future, from a macro-fiscal point of view.
RGC a reasonably conservative assumption would be that around 25% of this sum is spent on payments to counterpart project staff assigned to work on such projects from the civil service. This would amount to $55 million, which is almost four times the figure available for remuneration within the PRGF cap. This is a significant macro-fiscal phenomenon that has received insufficient attention in terms of its macroeconomic impact. First it indicates that in practice a large sum of donor assistance in fact serves as indirect budget support by significantly increasing the average remuneration of skilled Cambodian civil servants. Second, it indicates that the principal incentive to work in public employment is the prospect of access to external salary supplements. If these supplements are taken into account then the total expenditures on general services can be seen to be significantly higher than when looking purely at on-budget expenditures. Further, these expenditures are controlled not by RGC but by the donor community and therefore there is little policy action that RGC can take to tackle this, particularly given the strong emphasis on technical capacity building in the various macro-policy documents.21
3.6 Additional Payments to Civil Servants by Donors

A more important macro-fiscal issue in respect of spending on general services is that of additional payments to civil servants involved in donor projects. Godfrey et al. (2000) establish that a large number of technical assistance projects provide additional remuneration to Cambodian counterparts in a variety of ways These additional payments amount, on average, to about $150 a month and therefore far exceed the payments provided as part of regular salary. Leaving aside the micro and incentive impacts of this phenomenon, the sheer aggregate magnitude of these payments is of interest in macro-fiscal terms. Over 45 percent of ODA disbursements over the 1992-2001 period was in the form of free standing technical co-operation (RGC 2001: 14-15) amounting to around US$220 million in 2001. Since the bulk of technical assistance in Cambodia is channelled through
3.7 Disbursement of Resources

The key to unlocking the constraints on pro-poor fiscal policy in Cambodia is not directly macroeconomic in nature. It has to do with poor disbursement of resources allocated for pro-poor expenditures, resulting in weak budget implementation. There are
21 We note that, since we first raised it, the World Bank's IFAPER (World Bank 2003) has attempted to address this issue.
135
political, institutional and economic reasons for this phenomenon. Donor designed technical assistance to resolve this issue (even in their chosen priority sectors) has been piecemeal in approach, and conditionality driven. There is as yet no comprehensive appraisal of the problem. Table 4.7 provides a quantitative flavour of the disbursement problem. We can see that domestic revenue generation is fairly even through the year, while expenditures are extremely compressed in the initial months of the year. Current expenditure in particular merits attention with over a quarter of disbursements occurring only in December. Capital expenditure, on the other hand, is more evenly disbursed with domestic and foreign finance for capital expenditure also being generated fairly evenly throughout the year. Each and every component of current expenditure shows similar disbursement trends. The story repeats itself with respect to provincial expenditures. While it is outside the scope of this report to provide an examination of the various hypotheses seeking to explain the above, we note that during our discussions, RGC officials, donors, independent economists, consultants all agreed that this unevenness in disbursement causes severe problems for the effectiveness of fiscal spending, particularly in the priority sectors. Recognising this problem, RGC has for the past few fiscal years instituted a Priority Action Program (PAP). This program seeks to identify priority current expenditure items within the priority sectors (for example school equipment, or basic drugs) and sanctions release of half the budgeted amount for the
priority item for immediate disbursement. The remaining amount is then disbursed following satisfactory accounting for expenditures incurred from the first tranche of funds. PAP has somewhat ameliorated, but not resolved the disbursement problem. The limited success of PAP makes it clear that the solution to the disbursement problem is not simply a technical one of making cash available at the right time and the right place. There are more complex problems associated with release and disbursement of resources which need to be identified and addressed. Further, PAP in itself is not a long-term sustainable solution to the release problem, as it can only be used to ameliorate disbursement bottlenecks for highly specific expenditures. A comprehensive and sustainable solution to this problem is therefore of the highest importance if the effectiveness of public expenditure in Cambodia is to be improved.
136
Table 4.7: Monthly figures for key fiscal heads, Financial year 2001
Billions of Riels
I. DOMESTIC REVENUE II. BUDGET EXPENDITURE 2451.00 1426.00 542.285 235.285 307.000 883.715 575.970 476.200 99.770 38.315 121.970 122.460 25.000 1025.00 325.000 700.000 58.000 58.000 496.086 128.490 32.472 113.299 101.475 19.795 956.249 267.405 688.844 71.504 69.758 624.576 891.616 1.022 0.791 0.791 0.000 0.000 0.200 0.000 0.030 291.729 0.000 228.757 1.673 520.486 1.673 1412.12 2.694 2368.35 1. CURRENT EXPENDITURE a. Wages (chap. 10) Civil administration Defence - security b. Non wages Operating costs (chap. 11) Civil administration Defence - security Economic transfers (chap. 30, 52) Social transfers (chap. 31) Other current expenditure Interest (chap. 20) 2. CAPITAL - Domestic financing (Project) - External assistance (Project) Provincial revenue* Provincial expenditure* 1651.00 1524.73
2001 2001 Estimated Jan-01 Feb-01 Mar-01 Apr-01 May-01 Jun-01 Jul-01 Aug-01 Sep-01 Oct-01 Nov-01 Dec-01 Budget out-turn
107.523 106.373 152.185 132.747 151.711 115.739 115.811 119.269 124.989 112.857 119.821 165.678
75.595 128.660 161.838 150.197 204.112 157.422 182.837 214.204 168.502 202.846 241.738 480.400
47.270 98.742 79.279 115.925 78.762 105.570 129.267 80.691 116.819 171.970 385.111
31.034 66.463 34.543 59.233 36.089 32.788 54.068 20.561 43.719 54.205 86.109
11.722 24.974 17.419 22.079 17.603 18.248 19.525 15.038 18.965 18.450 43.060 19.313 41.489 17.123 37.155 18.487 14.539 34.542 5.523
24.754 35.755 43.049
16.236 32.279 44.736 56.692 42.673 72.782 75.199 60.130 73.100 117.765 299.002 6.808 5.594 1.214 0.100 1.110 7.010 1.208
15.010 33.283 42.064 25.041 55.751 55.027 39.862 54.048 76.921 219.969
14.515 32.000 29.418 24.402 39.971 40.329 35.723 34.312 63.458 175.570 0.495 1.640 10.753 1.275 3.601 1.283 2.336 7.151 1.081 0.885 12.646 2.701 8.851 2.269 0.807 0.639 1.848 9.383 3.560 2.842 15.780 14.697 2.857 11.456 1.459 1.258 1.091 7.601 9.253 2.227 4.140 4.295 9.243 5.088 1.641
19.736 13.463 44.398 2.079 8.461 6.514 1.997
4.357
9.168
15.082 24.007
19.881 44.085
1.524
1.774
72.900 81.390 63.095 70.918 88.187 78.660 77.268 84.937 87.810 86.026 69.768 95.289 20.526 20.160 16.438 21.721 27.624 3.822
16.493 24.852 28.000 28.273 23.582 35.915
52.374 61.230 46.658 49.197 60.564 74.838 60.775 60.085 59.811 57.753 46.186 59.375 3.516 0.000 9.396 0.312 4.409 3.568 2.513 3.019 8.121 5.511 5.541 5.604 3.053 6.269 5.568 5.139 8.026 3.216 9.136 6.142
4.571
7.653
5.794
25.185
Source: Data provided by Ministry of Economy and Finance.
137
4. CONCLUSIONS
In this concluding section we first present some concluding comments on aspects of the overall macro-fiscal strategy currently in place in Cambodia which could serve as inputs for making that strategy more pro-poor. We then present a summary of our principal analytical conclusions, together with some policy implications that can be derived from these conclusions.
4.1 Concluding Remarks

A pro-poor macro-fiscal policy has two essential elements:
A clear enunciation of macro-fiscal policy instruments that explicitly relate fiscal spending to poverty reduction. An assessment of the impact of macrofiscal policy changes on poverty alleviation
In Cambodia, the documents reviewed in Section 2 are clear on the macroeconomic policy instruments that can most effectively secure poverty reduction. These documents tend to agree in targeting growth (at an annual rate of 6-7%) together with an equitable distribution of this economic growth. The fiscal strategy to achieve these goals at the macro economic level consists of the following steps:
The NPRS, like most strategic macroeconomic documents in Cambodia, seeks to enhance fiscal performance by enhancing the efficiency of revenue and expenditure policy implementation and the first two prongs of the macro-fiscal strategy deal exclusively with this issue. In this sense the NPRS complements the 1999 World Bank PER and the 2002 PRGF. In the PER, the macro fiscal strategy was to increase domestic resource mobilization by improving revenue administration. Thus, [t]he authorities must address weak governance which is the core of the current low revenue performance (World Bank 1999: xvii). Similarly the bulk of recommendations in the PER were to improve the efficiency and effectiveness of public expenditure performance. Ten of the 13 recommendations on public expenditure were measures designed to secure improvements on this front. In the PRGF document the section on macroeconomic and structural policies for 2002 requires government to undertake a range of revenue reforms to improve revenue governance while maintaining fiscal restraint. (IMF 2002: 3-11). The SEDP-ii document, apart from presenting the projections reviewed in section two, contains no further macrofiscal analysis of strategic information. (SEDPii 2002: 99-100) Thus, the major pro-poor content of the existing macro-fiscal strategy lies in its emphasis on the priority sectors. The documents all present cogent, costed arguments outlining the pro-poor benefits of spending on the social service priority sectors. The pressure for a rise in education spending is based on the argument that acute budget constraints create difficulties in access to education for the poor and for girls leading to gender and income inequities. To redress
Strengthening revenue collections Enhancing the transparency of fiscal operations Increasing public investment in, by shifting spending priorities to, health, education, agriculture and rural development
138
this, the NPRS recommends alleviating the budget constraint in education and asserts that spending for education is the key priority (NPRS 2002). In both health and education, these arguments are backed by detailed medium-term sector spending strategies in the draft MTEF. With agriculture and rural development however, the documents are much weaker on quantifiable argument. There is no costed quantitative strategy that provides a link between outcome indicators in this sector and poverty alleviation targets. Rural development receives short shrift, with a statement of three pious intentions: (a) to promote decentralization, (b) to facilitate an integrated participatory rural development approach and (c) to develop a rural infrastructure strategy. There are no detailed medium-term expenditure plans for these sectors in the MTEF drafts. It may appear from the above discussion that RGC is more serious about prioritising government spending on social services and the mention of agriculture and rural development is gestural. This would be disastrous in an economy where the bulk of the poor live in rural areas and work in agriculture and agriculture related activities. Fortunately this is not the case. Our investigations revealed ample evidence that political authorities recognize the critical importance of development of agriculture and rural development as a priority goal for Cambodia and the political payback from providing such services particularly rural roads, rural employment, enhanced agricultural productivity and access to common and private agricultural inputs (water, fertilizer, seeds, etc.) is acknowledged to be
significant in government circles. Prime Minister Hun Sen himself is very particular about emphasising the importance of investments in rural roads rural infrastructure and agricultural development, in many of his policy speeches 21 Why, then, this Cinderella treatment at the strategic level? The answer lies not in an examination of government priorities but rather donor priorities. It is in the nature of advanced public finance processes, such as that sought to be instituted by the draft MTEF, that resource allocations to particular sectors depend significantly on the effectiveness with which the ministries and agencies responsible for formulating sector strategies make their case to government. In Cambodia, the four priority sectors compete intensively for the extremely limited recurrent and capital funds available. A major determinant of access to these resources is the effectiveness with which the case for using the funds is made. The Cinderella treatment that agriculture and rural development (not to mention other economic services) have received in this respect is an outcome of the fact that technical assistance for strategic preparation of sector priorities in these areas has been much less than that received by the health and education sectors. Thus, the education sector is slated to receive $197,280,000 in technical assistance in the period 2002-2004 (including ongoing projects). Agriculture receives $170,944,000 dollars in the same period, of which $100 million is for de-mining operations (covering a tiny minority of the land area). In terms of investment in actual policy formulation capacities in the two ministries, agriculture has no donor support while education is
21 We note that, since we first raised it, the World Bank's IFAPER (World Bank 2003) has attempted to address this issue.
139
expected to receive about $32 million in technical assistance. The health sector has $84 million worth of such technical assistance in the same period. (RGC 2002). Two conclusions can be deduced from the above story: The approach to macro-fiscal issues in Cambodia in most strategic documents has been largely technocratic, apart from the pro-poor expenditure switching policy discussed above. As a consequence there is no policy capacity to assess the pro-poor impact of fiscal policy changes. This reduces national ownership of macroeconomic strategy, discourages participation, and inhibits the acknowledgement and redressing of political constraints and challenges that retard fiscal performance, with negative attendant macroeconomic implications. Donors and multilateral organizations that claim to have poverty alleviation as their paramount objective, or super-goal, therefore need to pay more attention to this issue when providing technical assistance. This would require paying more attention to political economy issues, but the payback in terms of effectiveness would be sizeable. Well designed technical assistance for policy building can enhance gender equity, and compensate for the lack of attention to income poverty. Within the technocratic exercise, education and health are better equipped to argue their case for enhanced funding on pro-poor grounds as compared to the other to Cambodian priority ministries, agriculture and rural development. This is because the first two sectors have benefited disproportionately from donor-financed technical
assistance to equip them for the exercise. Hence, the poverty reduction impact of education and health spending are clear and costed. However the same is not true of the other priority sectors or other economic sectors more generally. Thus the choice of social sector spending as the spearhead of pro-poor development can be viewed as a consequence of asymmetry in access to donor technical assistance rather than strategic selection based on full information. Were this asymmetry to be redressed, Cambodias pro-poor fiscal strategy would arguably be very different from the present one.
4.2 Summary of principal analytical conclusions

4.2.1 Resource Mobilization Cambodias resource envelope is limited in the medium term. Only modest increases in public resources can be expected through growth in the tax base and efficiency improvements in the medium term. For a number of reasons, including low private sector confidence, IMF conditionalities and lack of incentive (given availability of concessional external finance), domestic resource mobilization through domestic borrowing is also not feasible in the medium term. Limitations on the domestic resource envelope exist despite rather high excess liquidity in the banking sector, together with significant unrealized domestic savings. Further, Cambodia is heavily dependent on concessional donor finance for its development investments. This raises concerns regarding future debt service ratios. The present and projected grant-loan portfolio make it
140
imprudent for government to significantly seek additional development assistance without reckoning on a concomitant rise in its future external debt service obligations. Any future adverse change in the grant-loan ratio would exacerbate this fiscal constraint. This limitation on Cambodias resource envelope means that Cambodia runs the risk of falling into a severe low level equilibrium trap with significantly negative consequences for poverty reduction, if the chief growth engine in Cambodias future economy private sector development fails. The greater the shortfall in expectations from the private sector in terms of contributing to GDP growth, the tighter will be the fiscal constraint and the lower the chances of redressing low private sector growth through fiscal stimulation or, indeed, direct public investment. 4.2.2 Public Expenditure Allocations Given the resource envelope constraint, Cambodias pro-poor macro-fiscal strategy seeks to secure growth through a highly specific expenditure switching policy from general services to social services, i.e. health and education. On the face of it, these sectoral expenditures have been designed to optimize their pro-poor impact, and most official documents highlight these benefits. They do not, however, highlight the implicit and explicit costs which need to be taken into account when assessing the net pro-poor benefits of the chosen strategy. It is imperative to recognize that increased investment in social services in Cambodia has come at the expense of low investments in agriculture and rural development. While
the precise opportunity cost of this trade-off can be debated, there is no doubt that lack of investment in agriculture and rural development in a country like Cambodia, with significant rural poverty and underutilization of capacity, cannot be held to be optimal from the point of view of alleviating income poverty. Again, this strategic choice is based on the assumption that private sector growth will be the engine of overall economic growth. If private sector growth is lower than anticipated then there is no cushion left for boosting growth and alleviating income poverty through investment in economic services. 4.2.3 Civil Service Downsizing Downsizing the civil service will provide little payback from a fiscal perspective, either by releasing resources to pay civil servants better to reduce corruption and increase efficiency or to reallocate these resources away from general services to social and economic services. Further, a large sum of donor assistance in fact serves as indirect budget support by significantly increasing the average remuneration of skilled Cambodian civil servants. This indicates that the principal incentive to work in public employment is the prospect of access to external salary supplements. Any strategy that relies on civil service downsizing either to improve the efficiency of public service delivery or to release resources for pro-poor expenditures would therefore be misplaced. From a fiscal perspective much of the emphasis on civil service reform is therefore not relevant. What needs to be tackled urgently is the incentive and capacity distortions caused
141
by donor supplementary remuneration to skilled civil servants. 4.2.4 Aid and Dutch Disease While there appears to be little crowding out effect of aid on governments revenue efforts in Cambodia, the generous concessional terms for investment resources from donors can act as a disincentive when RGC considers measures to generate domestic resources. The existence of excess liquidity in the domestic banking system negates the possibility of crowding out and offers RGC potentially attractive finance for development. The incentive for RGC to access such finance is, however, limited given the potential availability of easy concessional money from donors who accept RGCs bounded fiscal sovereignty (albeit with conditionalities), unlike domestic economic agents who, without recourse to conditionalities, are reluctant to do the same. In such a situation neither RGC or the donors are encouraged to speed up the financial reform process that would enhance RGCs credibility and remove the bounds on its financial sovereignty. It is this sense that there exists a Dutch disease in Cambodias fisc. Therefore, while donors can live with RGCs bounded fiscal sovereignty since they exercise countervailing power through conditionality-based safeguards, private economic agents cannot since they have no such safeguards. This leads to an immensely sub-optimal utilization of investible resources in Cambodia and makes the monotonic emphasis on private sector growth as the engine of Cambodias development less credible. This problem has political, institutional and technical dimensions all of which
need to be researched in depth and urgent policy measures taken. We have already suggested in Chapter 3 that one possible measure is a pilot experiment in issuing Treasury Bills. 4.2.5 Expenditure Disbursement The key constraint to pro-poor fiscal policy in Cambodia is not directly macroeconomic in nature. It has to do with poor disbursement of resources allocated for pro-poor expenditures, resulting in weak budget implementation. We have noted that there are political, institutional and economic reasons for this phenomenon. Donor designed technical assistance to resolve this issue (even in their chosen priority sectors) has been piecemeal in approach and conditionality driven. There is as yet no comprehensive appraisal of the problem. In this study we have not been able to undertake such an appraisal (see, however, World Bank 2003). The design of a comprehensive, sustainable and practical disbursement is urgently needed and, if operational successfully, it would increase the efficacy of pro-poor expenditures, improve confidence in RGCs fiscal sovereignty and, therefore, alleviate several fiscal constraints. By way of a final remark here, we reiterate that the disbursement issue is not a macroeconomic policy issue. Indeed in all the policy documents reviewed above, there has been a strong focus on technical and sectoral issues the desirable macroeconomic policy framework has tended to be taken as given. This is so, despite the fact that performance in poverty reduction could be improved. A serious policy debate would raise some of
142
the questions that we have discussed here: the implied costs (in addition to the benefits) of current policy settings, the possible benefits of a shift in donor sectoral priorities, so that they include all four of the a governments areas of concern, and ways the country might move towards a situation in which public confidence in the authorities is sufficient that it would be able to mobilize a larger proportion of existing excess bank liquidity for development purposes. As in the case of monetary policy, we do not advocate sudden changes in direction, but a step-wise approach in which the pro-poor goals of expanding rural purchasing power and making meaningful changes to their quality of life are given serious attention.
143
5
144
Trade Liberalization
1. INTRODUCTION
Since the early 1990s, Cambodia can be classified as a small open economy. The success of this openness can be seen in increasing trade flows and, in particular, in achievements in the garment sector which is currently worth over $US 1 billion in annual export revenue. However Cambodias exports are largely dominated by one sector and, with even this sector having a significant component of imported inputs, there is much need, and great potential, to diversify the export base. Even though many of the natural resources have already been largely exploited (timber, fishery), some even exhausted (certain gems), the potential from other sectors, especially agriculture and tourism, can be enhanced. The RGCs decision to apply for WTO accession prompted the development of a pro-poor trade strategy, in conjunction with technical assistance from various donor organizations. The culmination of this assistance, which involved several stages, is embodied in the Integrated Framework (IF) which attempts to integrate trade with national development strategies, see
The macroeconomic of poverty reduction in Cambodia Trade Liberalization
(RGC and Ministry of Commerce 2002). The strategy for pro-poor trade policy, as defined in the IF report, is consistent with policy objectives and strategies for national development in the National Poverty Reduction Strategy (NPRS) and the Second Socio-economic Development Plan (SEDP II), see (RGC 2002b). The RGC is pursuing trade policy reform with the objective that it will enable all people in urban and rural areas to enter into domestic and international trade and is seen as an important step towards improving economic performance and poverty reduction. The policy reforms include building the appropriate institutions to facilitate transactions and market function, judicial and public sector reform, investment in social sectors and rural development. Issues of trade competitiveness through exchange rate movements are a significant part of trade policy. In Cambodia the prevalence of multiple currencies in circulation has important implications for assessing competitiveness and the trade balance. The high dollarization of the economy means that Cambodias export competitiveness depends on movements in the US dollar on international markets, rather than the price of the riel. The impact of the domestic political turmoil and the regional crisis of 1997 were to a large extent cushioned by the effects of dollarization, causing a less than expected negative impact to the economy. In parts of the country bordering Thailand and Vietnam, the values of the Thai baht and Vietnamese dong are important. The implications for trade of monetary and exchange rate policy are not presented here but have been canvassed already in Chapter 3.
Instead, this chapter reviews Cambodias trade structure, the potential and constraints to increasing trade under the liberal regime for which the government has opted, and an assessment of its capacity to meet a pro-poor objective within this framework.
2. REVIEW OF POLICY TO DATE

Table 5.1 presents, in summary form, the main policy changes in the area of foreign trade. Two features stand out in the process of reform since the 1980s. First, a phase that had been largely achieved by 1994, comprised the process of dismantling the central planning system. All quantitative restrictions on trade had been removed, although a number of import tariffs were introduced in their place. Second, during the late 1990s there was a more deliberate phase of taking positive steps towards creating a highly liberal trade regime. Accession to ASEAN in 1999 meant that Cambodia joined the Asian Free Trade Area (AFTA) and committed itself to progressive tariff reductions over the following decade. Application for full membership of WTO then became a government priority and means that the RGC intends to extend this liberalization process further in coming years. It has received strong encouragement in pursuing this direction from major donors, who have designated it, along with Mauritania and Madagascar, as one of three pilot countries to combine rapid global integration with a pro-poor trade strategy (the IF).
145
Table 5.1: Evolution of Trade Policy Year Policy
1960s Exported agricultural products, largely rice, rubber and corn 1970s Virtual collapse of foreign trade New trading system controlled the level and composition of trade through quantitative restrictions and state owned trading bodies. Tariffs and trade taxes applied merely for revenue purposes. 1980s Market-oriented liberalization in the late 1980s, abolishing state monopoly of foreign trade. Foreign investment law was promulgated in 1989 enabling private companies to engage in foreign trade From 1993 restrictions limiting the ability of firms and individuals to engage in international 1990 trade were largely removed 1994 1996 1997 All quantitative restrictions on trade were eliminated in 1994. Cambodia gains MFN status from the US. Cambodia gains GSP status. IF Program commenced to promote the integration of least developed countries into the global economy (includes also Mauritania and Madagascar). April, Cambodia becomes member of ASEAN. Committed to a gradual reduction in most tariff rates by 2010 for trade with other ASEAN members. US quota imposed on 12 broad categories of garments. End of Year, Cambodia gained MFN/GSP status from 28 countries. The application for membership of WTO has led to the rationalization of the tariff structure. April, Cambodia designated as one of the three pilot countries (along with Madagascar and Mauritania) for pro-poor trade sector strategy formulation under the Integrated Framework. Reform of the tariff structure with the number of tariff categories falling from 12 to 4, and the top tariff rate reduced to 35%. July, Pro-poor Trade Sector Strategy at Cambodia's Third Consultative Group meeting in Tokyo. Chins accession to WTO. Cambodias accession to WTO U.S. and E.U. export market quotas are set to expire. Quantitative restrictions on imports of textiles and apparel products originating in China expire end of year.
1999 2000
2001
2003 2005 2008
Source: RGC and Ministry of Commerce 2002.
The discussion of this chapter therefore takes it as given that Cambodia will have a more, rather than less, open and liberalized trade regime. In other words, we assume that problems of access to markets, at least in so far as they depend on legal frameworks, will be resolved and that its increased openness will leave Cambodia more vulnerable to external shocks such as global recession. We examine first of all
(in Section 3) the impacts that opening up the economy has had on the structure and rate of growth of trade. Second, in Section 4, we examine supply-side issues; that is, domestic opportunities for and obstacles impeding the generation of pro-poor trade within the liberalized framework. Since it is by no means automatic that Cambodias poor will benefit from a more liberalized trade regime, the policies adopted to
146
strengthen the countrys competitive advantages will be the key determinants of whether trade liberalization at the macro level leads to poverty reduction.
trade balance in 1996. Nevertheless, export coverage of the import bill rose from only 61% in 1994 to around 85% in 2000-01. On the import side, however, the most remarkable feature of this table is the rapid rise in duty free imports. Customs data on imports subject to duty (c.i.f ) rose by only 6.5% a year from their low point in 1998 and were still below their 1997 levels in 2002 and coverage adjustments (estimates of unrecorded trade) showed a similar pattern, duty free imports rose by 35% a year. This rise is particularly associated with increasing imports of inputs for the garment export industry. When both legal and illegal duty-free imports are taken together, they form a sharply rising proportion of the total import bill as well as of total export income. By far the largest proportion of smuggled imports consists of petroleum products that have been rising in price in recent years and would otherwise attract high tariff payments. These accounted for 71% of the estimated unrecorded trade for 2002 (NIS 2003: 425). However, depending upon the extent to which they can be relied upon, the data also suggest that slow progress has been made in the governments ability to collect tariffs. Estimates of tariff-evading trade fell from 33 to 31% of all dutiable trade over the period 1998-2002.
3. IMPACTS OF TRADE LIBERALIZATION

Table 5.2 presents the flow of trade between 1994 and 2002. The lifting of quantitative restrictions on trade at the end of 1994 had a positive impact on the trade flow, with the total value of exports (f.o.b), particularly re-exports, increasing by three-quarters and imports (c.i.f ) by 60% within the first year. During 1994-2000, a $1 increase in GDP was associated with a $1.40 increase in exports and a $1.30 increase in imports. The gaining of GSP status in 1997 is also associated with a rise in domestic exports as a proportion of the total. However, the figures for 1998-2002 also include large upward revisions based on estimates of unrecorded trade that are absent from the earlier figures. Upward revisions for fish products and agricultural products (especially paddy) are particularly noteworthy, as they indicate that considerable incentives exist to avoid government export procedures. We discuss this problem later in the chapter. Although there was an overall worsening of the trade balance between 1994 and 1996, it has improved again through an increase in domestic exports, in particular the growth in the garment sector. Despite a fall in imports in 1996 and increase in domestic exports, the decrease in re-exports of 40% contributed to a worsening of the
147
Table 5.2: Merchandise Trade, 1994-2002 (in $US millions, unless otherwise stated) Trade Balance 1994 -254 1995 -332 1996 1997 1998 1999 2000 -263 2001 -226 2002
Domestic Exports 262 268.2 Logs 142.3 111.6 Sawn Timber 81.3 73.1 Fish products 2.4 1.9 Rubber 25.6 41.2 GSP Exports 2 27.5 Agricultural products 5 10 Other domestic exports 0.1 0.1 NR carrier procurements 3.4 2.8 Re-exports 227.7 586.1 Total Exports (f.o.b.) 489.8 854.3 Customs (c.i.f.) Cigarettes Motorcycles Beer VCRs Televisions Audio cassettes Gold Vehicles Construction materials Cloth Clothing Petroleum products Sugar Cement Steel Other Duty free imports (c.i.f) Coverage adjustments Total Imports (c.i.f ) Total Imports (f.o.b) of which: retained imports Exchange rate (riel /USD) 643.9 923.6 140.3 192.5 32.9 36 14.6 14.7 22.6 14.9 21.3 17.3 4.2 4 78.9 305 11.3 22 24.8 19.2 10.8 17.2 35.2 12.6 61.4 99.3 9.3 6.4 5 12.3 2.4 4.3 169 146 165.2 340.7 809.1 1264.3 744.4 1187 503.0 2545 574.0 2451
-428 -230 -173 -275 Exports 295.2 534.3 693.7 883.9 52.5 128.1 90.8 65 95.9 95.6 91.1 73.4 3.1 2.9 37.6 42.0 31.9 22.8 41.3 48.8 101.8 278.5 392.4 564.3 7.6 2.5 24.3 71.9 0.1 0.2 0.1 0.9 2.2 3.7 5.8 5.4 348.4 327.3 121.8 132.1 643.6 861.6 815.5 1016.0 Imports 722.1 780.5 565.7 715.7 210 187.9 143.9 119 18.5 17.3 44.1 35.7 13.7 5.2 3.1 2.2 6.1 0.8 2.4 0.9 13.3 5.7 4.6 6.1 2.5 1.5 1.9 2.5 41.2 136 3.5 28 24.7 25.6 14.9 27.2 19.6 18.9 7.8 13.4 21.4 17.6 21.8 33.5 10.1 4 3.7 3.5 128.7 135.8 146.6 151 13.5 15.3 13.7 21.8 15.2 16.3 12.7 19.6 5 0.8 6 18.5 178.6 191.8 145 237.8 437.1 612.8 438.8 394.1 273.4 283.1 1160.9 1174.6 1276.2 1611.6 1071.6 1092.4 1179.3 1489.6 723.0 2624 765.0 2946 777.0 3744
1277.9 1415.5 1655.9 46.8 37.6 21.7 53.7 30.4 15.9 43.8 42.3 60.2 60.0 52.4 62.7 1012 1141.5 1403.4 36.7 81.2 50.9 3.5 6.1 15.1 7.3 7.6 8.1 117.8 109.2 110.9 1395.6 1524.7 1766.8 740.6 710.7 728.7 70.1 70.1 68.1 30.6 21 27.9 2.5 2.3 1.5 1.9 1.7 1.1 5.6 4.8 4.7 2.5 2.5 1.9 34.6 11.5 10.3 23.1 25.7 38.2 13.1 14.4 12.9 46.1 27.9 38.6 3.3 11.2 5.3 155.6 174.8 149.6 9.6 25.4 25.9 27.4 31.1 37.3 20.8 18 22.6 302.1 277.2 281.1 1075.7 1214.3 1444.4 278.6 299.7 327.3 2094.8 2224.7 2500.4 1934.7 2052.3 2311.1 1425.0 3922 n.a. 3925
987.0 1354.0 3808 3845
a: Not recorded separately prior to 1998. Source: NIS 2002a: 252-253; NIS 2003: 423-425.
148
3.1 Structure of Exports

As Table 5.3 shows, the structure of exports has shifted markedly over the last seven years. There has been a large rise in the share of domestic exports, as opposed to re-exports. The high proportion of re-exports in the past largely reflected Cambodias role in the illegal trade between Thailand and Vietnam, but this has declined, probably as a result of Vietnams own increasing openness and expanding domestic output of the principal goods involved (such as beer, cigarettes and motorbikes). Access to the GSP system for Cambodia is also linked to the rapid rise in the share of domestic exports.
Table 5.3: Merchandise Trade of Exports, 1994-2001, % distribution 1994 Domestic Exports (f.o.b) Logs Sawn Timber Fish products Rubber GSP Exports Agricultural products Other domestic exports NR carrier procurements Re-exports Total Exports (f.o.b) Exchange rate (riel /USD) 53.5 29.1 16.6 0.5 5.2 0.4 1.0 0.0 0.7 46.5 100.0 100 1995 31.4 13.1 8.6 0.2 4.8 3.2 1.2 0.0 0.3 68.6 100.0 96.3 1996 45.9 8.2 14.9 0.5 5.0 15.8 1.2 0.0 0.3 54.1 100.0 1997 62.0 14.9 11.1 0.3 2.6 32.3 0.3 0.0 0.4 38.0 100.0 1998 85.1 1.1 1.1 4.6 5.1 48.1 3.0 0.0 0.7 14.9 100.0 1999 87.0 6.4 7.2 4.1 4.8 55.5 7.1 0.1 0.5 13.0 100.0 2000 91.6 3.4 3.8 3.1 4.3 72.5 2.6 0.3 0.5 8.4 100.0 2001 92.8 2.5 2.0 2.8 3.4 74.9 5.3 0.4 0.5 7.2 100.0 2002 93.7 1.2 0.9 3.4 3.5 79.4 2.9 0.9 0.5 6.3 100.0
107.1 112.3 127.1 101.7 101.0 102.0 100.0
Source: Calculated from same sources as for Table 5.2.
In the early 1990s timber represented a significant share of overall exports, over 40%, with the next largest share being rubber at only 5%. Since the mid-1990s, there has been significant growth in the garment sector, reflected in the GSP (Generalized System of Preferences) category in the table. Indeed, the export share of garments more than doubled in the space of just three years after 1997, while the value of textile and garment exports almost doubled in just one year, 1999-2000. Since
then the total has been over $US1 billion. In 1999 GSP exports comprised largely apparel (74%) and a small component (5%) of footwear. The main markets for apparel were the US (76%) and the EU (23%), while the main buyers of footwear were the EU (60%) and Japan (35%) (RGC and Ministry of Commerce 2002). Between 1999 and 2001, garment exports to the US increased by 50%, while exports to the EU tripled over the same period, (Sok Hach et al. 2001).
149
3.1.1 Textile and Garment Sector Textile and garment factories were established in the 1950s in Kompong Cham, Battambang and Phnom Penh, but many were damaged during the war in the 1970s. During the 1980s, although there was some recovery in this sector, production was largely for limited domestic purposes. It was not till the mid-1990s that the export garment industry began to develop (Sok Hach et al. 2001). The production of garments has been a significant source, not only of revenue for Cambodia, but also overall growth in the economy. The speed of growth in this sector is a sign of the potential for private sector investment when the conditions are favourable. By 2000, garments were exported to 38 different countries, although the US and EU are the principal destinations. Extensive tax exemptions on garments and textiles under the Investment Law, the extent of import content, and widespread foreign ownership have limited the overall contribution of this sector to state revenues and the Cambodian economy (see Section 3.4). The RGC has actively promoted foreign investment in the garment-manufacturing sector using various incentives. The Letter of Invitation from the RGC, stipulating US and EU quotas, various tax holidays and duty exemptions resulted in an inflow of foreign investment in garment manufacturing. Much of the investment is from Taiwan, Hong Kong, China, Singapore, Malaysia, Korea, Thailand and Indonesia. As a result of Cambodia obtaining most favoured nation (MFN) and GSP status, Normalized Trade Relationship agreements were first
signed with the EU in 1996 and with the US in 1997. By 1999 the growth in exports to the US resulted in the US introducing quotas on 12 categories of garment exports. While Cambodias trade regime has been substantially liberalized, this does not yet apply to its chief export markets. Garment manufacturers considered quotas, tariffs and labour standard requirements the main obstacles to the US market. In reality Cambodias competitive advantage lies in areas considered sensitive and much output is therefore excluded from GSP access. MFN rates apply and 12 categories of garments are subject to restrictive quotas, with the result that the average duty paid on exports to the US is 17%. The situation was similar for the EU markets, but with, in addition, the constraint of local content requirements under the everything-but-arms initiative, with the result that substantial quantities of goods are not eligible for duty free access and must enter at MFN rates. The low level of development of the Cambodian textile sector results in the rules of origin requirements imposing an 8.8% duty for exporters. Footwear firms have been established primarily through Taiwanese investment, with exports destined largely for the EU and Canada. Again, little of the input is locally sourced. Given their restricted entry, for the time being, to the main export markets, Cambodias garment manufacturers identify transport costs (due to the high import content) and bureaucratic payments as the main constraints to further expansion.23
23 Interestingly one manufacturer we spoke to had begun planting raw cotton, not as a means of meeting European local content requirements, but as a means of reducing transport costs of his inputs.
150
As in many countries, trade flows fell as a consequence of global recession in 2001-02. Garment sector export growth slowed markedly in 2001, largely due to falls in US sales. However, the industry may also benefit somewhat from recent developments, since many US companies have switched their sourcing to countries that they believe are not connected to terrorism. Cambodia, being largely a nonMuslim country, has benefited from recent shifts by buyers who normally operate in neighbouring countries. Although this increase may be short lived, it can be seen as providing an opportunity for presenting Cambodias business potential. 3.1.2 Forestry and Fisheries The decline in wood-based exports is a result of legal constraints imposed by the government to address illegal logging activities. These activities have taken place on either common property resources or areas of land granted as concessions. The result has been the depletion of forests to 35% of their original area. The widespread occurrence of illegal logging and smuggling has meant that little of the revenue from this natural resource has accrued to government. The Forest Crime Monitoring Unit has been set up to enforce the forest management polices intended to monitor illegal logging. A new forestry law provides the framework for sustainable management of the 23 protected areas. Table 5.4 shows export data supplied by the Customs Department as well as official estimates of unrecorded trade across different categories. As noted above, the 2003 Statistical Yearbook contains substantial
upward revision of unrecorded trade estimates from 1998 onwards. Earlier figures have not been adjusted, but they are also likely to be higher than shown in Table 5.4. The estimates of coverage adjustment for forestry products form the major proportion of the overall value of exports of this commodity and illustrate to some extent the severity of the problem. Reports of illegal transportation of fish to the Thai border, where the fish are unloaded into smaller containers that meet the Thai regulations are also given substance by the Table. Such transactions are facilitated by under-the-counter payments to Customs and checkpoint guards (RGC and Ministry of Commerce 2002). It is noteworthy that in 2002 no less than 94% of fish exports came under the illegal category.
151
Table 5.4: Exports by commodity, 1993-2001 Million $US Logs (wood) Customs Coverage adjustment Sawn Timber Customs Coverage adjustment Fish products Customs Coverage adjustment Rubber Customs Coverage adjustment Agricultural products* Customs Coverage adjustment na na 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 36.3 3.2 33.1 44.5 19.0 25.5 1.4 0.1 1.3 27.5 12.6 14.9 142.3 111.6 100 42.3 81.3 63 18.3 2.4 2.4 0 25.6 25.6 0 5.0 na na 51.6 60 73.1 15.5 57.6 1.9 1.9 0 41.2 37.2 4 10.0 na na 52.5 0.02 52.5 96.0 21.0 75 3.1 3.1 0 31.9 27.7 4.2 7.6 na na 128.1 0 128.1 95.6 11.2 84.4 3.0 3.0 0 22.8 19.1 3.7 2.5 na na 90.8 0 90.8 91.1 6.1 85.0 37.6 2.5 35.1 41.3 25.3 15.9 na na 24.3 65.0 3.6 61.4 73.4 29.5 43.9 42.0 3.4 38.6 48.8 27.4 21.4 na na 71.9 46.8 3.7 43.1 53.7 20.9 32.7 43.8 4.5 39.3 60.0 31.5 28.9 na na 36.7 37.6 5.0 32.6 30.4 5.6 24.8 42.3 4.1 38.2 52.4 24.6 27.9 na na 81.2 21.7 0 21.7 15.9 0.8 15.0 60.2 3.7 56.4 62.7 29.6 33.1 na na 50.9
* Customs data for these products for 1998-2002 are included under 'other domestic exports' in the 2003 statistical yearbook (they are therefore rather small - see Table 5.2). In the 2002 yearbook, no separate data are given for unrecorded trade. Source: NIS 2002a; NIS 2003.
3.1.3 Rice Production Rice production has historically been an important source of export revenue. At the end of the 1960s annual exports were up to 0.5 million tonnes of milled rice. In the 1950s and 1960s it was the most important source of foreign exchange, as much as 1/3 of the total, and 60% in the mid-1960s, falling to 40% by the end of the decade. Two decades of civil war broke this success and destroyed much of the infrastructure resulting in Cambodia becoming a net importer of rice, see (Sok Hach et al. 2001). Production in rice has only started to increase again since 1995. 1997 saw the end of subsidized fertilizer prices, though fixed prices ended with the 1980s reforms.
In 1999/2000 production reached a new peak of 4 million tonnes of paddy. Due to problems of food security in some areas, there is a restriction on the export of rice, making it necessary to obtain a permit before exporting. Although the licence is officially free, the limited number means that unofficial payments and time-consuming procedures are usually involved. This problem has led to an increase in illegal exports of rice, especially paddy; recorded exports are far below the available surplus production. Between 1996-98, real exports were about 30% below the licensed quantities, worsening to 50% in 1999. Farmers sell their produce to paddy collectors and traders or millers. The small scale
152
fragmented family production system leads to low bargaining power by the farmer when it comes to selling the unmilled paddy at the end of the harvest (Dec-Feb) and low farm-gate prices. Farmers often have to borrow from moneylenders to cover the cash shortfall for hiring labour or to purchase inputs, which usually occurs during or very shortly after the harvesting season. If fertilizer is purchased this also has to be repaid at harvest time. Weddings and ceremonies and the start of schooling are also held at harvest time, which makes this time of year particularly costly. Financial constraints are deepened with the fall in price of rice at harvest time, when all farmers are trying to sell their produce since many cannot afford storage facilities for the paddy, (Sok Hach et al. 2001). Although farmers do not benefit from good returns from paddy growing, the mark-up earned by intermediaries can be around 10-15%. There are various intermediaries in the chain from paddy production to rice export; the farmer, paddy collector/trader, miller and exporter. The bargaining power of the different agents is reflected in the price they receive for their goods, with the farmer the most vulnerable. The costs of paddy storage, milling, transport (as much as 5-10% or $10-20/tonne) and payment of unofficial fees indicate the level of resources needed in the marketing chain. Lower barriers to entry would reduce the profit to intermediaries. Low incomes, savings mobilization and investment in rural areas are the main barriers, however, implying that policies to raise rural incomes and investment are needed first. The direct economic losses incurred on rice exports between 1996-2001 have been
estimated to range between $1.5-$7 million annually, equivalent to 6-8% of total government expenditure on the agricultural sector (Sok Hach et al. 2001). These figures do not include the sizeable unofficial quantity of paddy exported across the border to Thailand and Vietnam in recent years. Official estimates of these exports show very wide fluctuations in both volume and value, with a declining trend in value per tonne (NIS 2003: 424). Changes in the demand and supply for rice in Thailand and Vietnam have ramifications on prices and exports of Cambodian rice. This uncertainty in prices has implications for farmers in investing in increased production and hence many remain self-sufficient producers only, with commercial production seen as too risky. Commercial rice imports are taxed at 7% while food-aid is tariff free. The main destination of milled rice is ASEAN member countries, with 70%-85% destined for Singapore (Table 5.5).
153
Table 5.5: Exports and Imports of Milled Rice, 1995-1999 (tonnes) 1995 Developed countries Asean Singapore Other Total (actual export) Licensed 1 Developed countries 2 Japan ASEAN 3 Thailand Vietnam Other countries Total 691 90 2197 2030 145 27 2915 11 25977 2981 2531 2645 5625 16310 Imports 9728 8000 16238 16227 7001 7000 12753 8951 3801 2 19756 19773 16533 13644 8951 3801 15 33432 10884 10882 23285 21267 2000 19 34188 1996 Exports 424 2732 2732 667 3823 10957 784 3080 2 782 560 21 2970 2200 120 3111 7390 1997 1998 1999
Notes: 1. Quantity licensed and registered at the Ministry of Commerce 2. It is not economically viable to import rice from developed countries, hence the imports are assumed to be under food aid for the poor 3. As over 90% of rice imports come from ASEAN countries (Thailand and Vietnam) some of the imports were assumed to be commercial. Source: (Sok Hach et al. 2001) using data from Customs Dept 1995-99, Dept. Foreign Trade 1996-99.
3.1.4 Tourism The tourist industry has expanded significantly since 1993, with annual average growth of 20% between 1993 and 1996, and represents an area of potential for further growth. The growth has further increased since the elections of 1998, and the 1997 open skies policy that allowed international airlines to fly into Siem Reap, (Sok Hach et al. 2001). The open skies policy was first introduced in the late 1960s, and saw the opening of the international airport at Siem Reap in 1968, with Chinese
assistance. Flights were ended, however, during the 1970s. By the end of 2000, after the re-introduction of the policy, four international airlines had commenced flights into Siem Reap. Its re-introduction was part of a strategy to increase tourism in both Thailand and Cambodia under the strategy of Two Kingdoms, One Destination slogan, (Sok Hach et al. 2001). The government has widened the number of legal entry points at the border and, together with improvements in road transport and access by sea, this has resulted in an increasing number travelling by boat and overland from neighbouring
154
countries. In addition, improvements in road transport have seen a significant increase in domestic tourism. A survey conducted by the Ministry of Tourism in November 1999 showed that approximately 63% of visitors to Siem Reap were on a group tour, while the number for Phnom Penh is 31%. These figures are also reflected in the age composition of tourists shown in Figure 5.1: there is a high concentration of older tourists visiting the Angkor complex, while younger tourists also tend to take in other parts of the country.
Figure 5.1: Foreign Visitors by Age Group, Ministry of Tourism November 1999
Source: Sok Hach et al. 2001.
Tourist arrivals averaged 15.5% growth per year during 1995-2002 (Table 5.6). They slowed during the events of 1997 resulting in the closure of some hotels and the reduction in working hours of staff. 24 From 1998 to 2000, however, they grew rapidly - at 21% a year while, in the same two years, average receipts from tourism rose at 64% a year, and receipts per tourist by 35%. It seems likely, however, that these impressive figures are mainly the result of rising prices of hotels and other tourist facilities. Anecdotal evidence suggests that much of the money does not remain in Cambodia.
Table 5.6: Tourism, 1995-2002 1995 Visitor arrivals by mode of transportation (000s) Arrivals by air Phnom Penh Siem Reap Other types of arrivals 1 Purpose of visit (000s) Tourism Other 2 Tourism receipts Million US dollars Annual % change Average per tourist Per cent of GDP 327 1.6 53 82 55 422 2.4 68 -17 417 2.0 107 57 465 3.5 212 98 576 6.2 304 43 652 8.5 380 25 628 10.3 454 19 577 11.4 1996 1997 1998 1999 367.7 219.7 219.7 n.a. n.a. 219.7 162.3 57.4 260.5 260.5 n.a. n.a. 260.5 194.4 66.1 218.8 218.8 n.a. n.a. 218.8 163.0 55.8 186.3 175.9 10.4 n.a. 186.3 141.9 44.4 262.9 234.4 28.5 104.8 262.9 199.6 63. 3 2000 466.4 351.7 264.6 87.0 114.7 351.7 209.6 142.1 2001 604.9 408.4 274.7 133.7 196.5 408.4 355.4 73.0 2002 786.5 523.0 320.2 202.8 263.5 523.0 441.4 75.5
Notes: 1. Arrivals by land and boat 2. including business and other purposes, n.a.- not available. Source: NIS 2002a; NIS 2003.
24 After July of 1997 the government exempted hotels from paying the 10% tariff.
155
Despite the increase in the numbers of tourists coming to Cambodia in the 1990s, in particular to Siem Reap and Phnom Penh, the levels are well below those of neighbouring Thailand and Vietnam. Compared to other ASEAN countries, hotel accommodation in Cambodia is relatively expensive, with the result that length of stays and spending levels are lower than in many other comparable countries. Under the current structure of the tourist industry, the airlines, larger hotels and the casinos, are mainly foreign-owned and benefit from privileged tax incentives. This even includes the Angkor Wat complex itself, in Siem Reap. The government is currently reviewing its policy on the tariff structure for hotels and casinos. A negative aspect of the increase in tourism in Cambodia is the increase in the sex tourism industry. There are growing reports of children being exploited. A survey conducted in January 2000 showed that Phnom Penh had the highest rate of prostitution, and that it was on the rise. While 32% of visitors were interested in cultural tourism and 45% came for business or official visits, not less than 22% came for sex tourism (Sok Hach et al. 2001). It is widely believed that expansion of the sex tourism industry has led to the increase of HIV/AIDS infection, although there is no available evidence to support this view (in Thailand, for example, it is the domestic market for sex which has been chiefly responsible).
3.2 Employment Impact

The impact of trade liberalization on employment has been very concentrated and limited. The export sector, including tourism, employs around only 6-7% of the work force and will be required to maintain its high growth rates of the immediate past if it is to make a significant impact on rural underemployment. So far, the main area of employment growth has been in the garment industry. There are varying figures on the numbers employed in this sector, with estimates ranging from 160,000 to 180,000 people, mostly young women.25 This represents around only 3% of total employment, in a workforce aged 10 years and older, approximately half of the total employed in the manufacturing sector. However indirect employment, in auxiliary services and activities around garment factories, could bring the total to as much as 500,000 (interviews with GMAC officers). An unusual feature of Cambodias garment exports is that in order to gain access to US markets, the country was required by the Clinton administration, to adhere to labour standards. These are in some cases higher standards than apply in the US itself, but they have nevertheless been incorporated into the Cambodian Labour Code. Cambodian employers see some of these conditions, such as payment of double time for night shift work, as restricting employment opportunities for those who would accept lower wages. In Vietnam, the night shift allowance is
25 Sok Hach et al. (2001) estimate total employment in the textile and garment industry in 2000 at 200,000, of which 40,000 remained in the traditional sector.
156
reported to be only 30% of the wage (Van Sou Ieng 2002). The rules also prohibit a worker from doing more than 2 hours over time per day, thus exerting pressure on employers to operate a night shift. The employer association estimates that around 20,000 job opportunities have been lost due to the strict labour laws. Implementation of these laws is, however, patchy, as there are insufficient numbers of inspectors and strong incentives for employers to evade the law. Compared with regional competitors, Cambodian garment workers are relatively high cost. Compliance with the Cambodian Labour Code and EU GSP status has meant implementation of a minimum wage, currently $45/month, though average earnings are higher. Progress has been made in a few areas including lack of discrimination in employment, prohibition of forced or bonded labour, and the prohibition of exploitative child labour. Improvements in the application of the labour code have enabled Cambodia to gain an extra 9% quota in exports to the US, rather than the expected 14% (Sok Hach et al. 2001). While these protect workers, they tend to make Cambodian workers more expensive than those in other countries.26 Mandatory double time for night shift work, where applied, also increases production costs. This is a relatively high average wage for Cambodia in general, but not for urban areas where the cost of living is higher. An interesting side effect of the relatively high wages in the garment industry is that many young women who leave the industry
turn to prostitution as a means of maintaining their higher incomes. A survey carried out among prostitutes in one area of Phnom Penh indicated that 60% of them were former garment workers (interview at MWVA). Bargaining positions between workers and employers have been strengthened, in particular in the garment industry but increasingly in other industrial sectors, with growing membership of trade unions. The presence of unions at the workplace provides some protection for the workers through collective bargaining, and offers an avenue for mediation between the workers and employer. Within the garment industry trade union presence is highly varied, with some factories having none while others have several active unions. Collective bargaining has been at least partly responsible for the rise in the minimum wage from $27 to $40 in 1999, and then to $45, and a reduction in hours of work from 48 to 44 per week. However, the increasing militancy of a few trade unions has led to some disruption of production, especially strike action, which is particularly costly to garment employers who face tight delivery deadlines. In 1999 there were 76 strikes, and as many as 35 strikes in the first seven months of 2000, (Sok Hach et al. 2001). There are, however, reports that the frequency of strikes may be on the decline as unions appreciate their limited effectiveness (Calvert 2002). Over 70% of the workforce is engaged in agriculture - the proportion has increased slightly in recent years - with only limited numbers involved in commercial activities. Development of agricultural exports has
26 This calculation does not, however, take into the probability of reduced productivity of the workforce if working hours were significantly extended. The macroeconomic of poverty reduction in Cambodia Trade Liberalization
157
probably led to some reduction in underemployment in those areas where commercial production takes place, particularly through increased labour intensity from double cropping. As noted elsewhere in this report, however, there has been little associated development of non-farm employment associated with agriculture - in processing, transport, distribution, input supplies and extension services. The tourist industry has a potential to create a large impact on employment both within the tourism-related services but also through increasing local demand for handicrafts. Direct employment in tourism in fact doubled between 1995 and 2000, from 31,000 to 60,000 (Table 5.7). While little information is available on indirect employment, hotel construction seems to have been the main area. Since room occupancy rates have
declined, such employment generation may not be stable. We were told that the tourism sector also has a high import content: as yet few other linkages exist to the domestic economy. In particular, the concentration of tourism in Siem Reap/ Angkor on group tours probably limits new employment opportunities. Most groups stay in foreign-owned hotels, use buses rather than small-scale transport and buy souvenirs from a handful of designated shops. Cambodias reputation as a destination for sex tourism does, however, undoubtedly help support employment of both women and children in that sector, a fact which reflects the absence of other comparable income earning opportunities for the poor. Such employment can, however, scarcely be described as pro-poor, since it very often involves trafficking of both women and children into slavery.
Table 5.7: Key Indicators for the tourism industry, 1995 & 2000 1995 Number of hotels Number of rooms Number of visitors by air (000s) Length of stay (average days)1 Daily room rate (average, US$/room) Occupation rate (average %) Daily foreign tourist spending (US$)1 Income from domestic tourism (million US$)2 Total tourism income (million US$)2 % of nominal GDP Cumulative investment in hotels (million US$)3 Employment in tourism (000s)4 % of total labour force
Note:
2000 241 9,800 405 5.5 35 60 80 60 271 8.7 313 60 1.1
136 4,500 212 5.8 27 75 66 75 156 5.3 122 31 0.6
1. tourists that arrived by air, 2. includes hotels and restaurants, 3. based on hotel construction costs in Ho Chi Minh City and Bangkok, excluding land, 4. includes staff of hotels (about 7000 employees in 2000), guides and travel agency staff (3000) and restaurant workers (50,000). Indirect employment has been excluded. Source: Sok Hach et al. 2001, based on Ministry of Tourism, National Institute of Statistics, CDRI interviews with tourism industry professionals conducted in 2000.
158
3.3 Enclave Development

Manufacturing, services and exports are overwhelmingly urban and currently concentrated in three cities: Phnom Penh, Siem Reap and Sihanoukvillle. Tourism is concentrated in two centres: Siem Reap and Phnom Penh. Linkages of all these sectors with the rest of the economy are weak. Investment in the export sector has been heavily concentrated in urban areas due to the poor state of the transport networks spanning the country. The shortage of good roads restricts export production to areas with easy access, via one or two main highways, to the port of Sihanoukville and the two international airports at Phnom Penh and Siem Reap. Roads to the Thai border in the west and to the Vietnamese border in the east are of poor quality, which restricts volumes, but are especially important for the unofficial trade with these two countries. Urban concentration reinforces the bias against development of rural markets. Private investment continues to be attracted to urban agglomerations where transport, energy and communications infrastructure are more highly developed and other external economies can be reaped (for example, access to business services, finance and information). Further, as investment in urban enclaves increases, so does demand for improvements in the urban infrastructure, resulting in a shift of public funding to these areas instead of the less developed rural areas. Although there is an emphasis in government policy pronouncements on strategies to widen the regional export base, this emphasis is currently not reflected in an increase in actual public expenditure on rural development.
Current concentration of the tourist industry is indicated by the extremely rapid rise in tourist arrivals directly into Siem Reap since the re-opening of the international airport there and introduction of the open skies policy. The proportion of tourists flying directly to Siem Reap rose from only 7% in 1998 to about 40% in 2000-02. This rise is an indication that tourists tend to visit the Angkor complex as an extension of their travel to other regional destinations, rather than as part of a visit to Cambodia itself. Yet cultural and environmental tourist attractions in Cambodia are located in several provinces other than those where tourist activity is currently concentrated. There are opportunities to enhance the current tourist destinations, but also to develop others including national parks, beach resorts, such as those around Sihanoukville, and mountain areas. Further developing the transport linkages and the facilities for tourism in these locations offers an opportunity to increase the distribution of economic benefits from this sector more broadly.
3.4 Value Added

Data on the amount of value added in export production is very scarce. Limited data do exist for the garment industry (Sok Hach et al. 2001): they indicate that for each $100 worth of exported garments, $63 dollars is spent on imported raw materials and $4 on utilities and other domestic inputs. Value added is thus approximately one-third of the total sale price. Labour costs are estimated at $13 and bureaucracy costsat another $7, leaving total gross profits at around 13%. Since it is estimated that three-quarters of profits are repatriated,
159
not more than a quarter of the sale price represents value added that is retained in country. While the progress of garment exports and the billion dollar export revenue are much cited as evidence of the recent success of macroeconomic policy, the real contribution to Cambodias economic development is only about $250 million a year. There is little equivalent information on value added for other products. While the mark-up earned in processing, transporting and marketing of rice exports is significant, only a tiny proportion of exports are of milled rice. The vast majority are of unprocessed paddy (see Chapter 2), meaning that Cambodia currently receives very low value added on this commodity. The case of fish is similar: there is very limited processing in Cambodia itself. We have already noted that the tourism industry is both import-intensive and has a high proportion of value added, in the form of profits, repatriated to investor countries.
capital goods in the import structure. Unfortunately, for Cambodia, the published data do not allow us to examine whether any such shift has in fact taken place. The most notable feature of the structure of imports since 1994 (Table 5.8) has been the rise in duty-free imports. There is no doubt some statistical error in this table as poor border security probably means that the unrecorded trade is understated for the earlier period. While the borders remain quite porous, improved security and better co-operation with the authorities of neighbouring countries (especially Thailand) means that the ability of the authorities to estimate the extent of unrecorded trade has improved over time. Nevertheless, the consistently high and rising proportion of unrecorded imports is a strong indication that the incentives to smuggle remain. The extent of unrecorded trade although difficult to capture quantitatively, can be seen in the use of the Thai baht in the northwestern provinces of Cambodia, near the border. The Vietnamese dong also circulates inside Cambodia, though in lesser quantities than the baht, in the border areas. Evidence from our interviews in Cambodia suggest that in practice a large amount of unrecorded imports are of petroleum, estimated to be as much as 25-50% of petroleum consumption. These imports reflect the incentives provided by the rising world price of oil, the tariff structure (see below) and the requirements or unofficial payments to evade customs inspection. They also reflect the growth of the urban economy: few Cambodians in the rural areas make use of petroleum products in either production or consumption (see Chapter 1).
3.5 Impacts of Liberalization on Import Growth

3.5.1 Structure of Imports From the standpoint of a pro-poor growth strategy, the main purpose of expanding export income is to enable increased import of capital goods for economic development. Unless there is an increase in the rate of investment, domestic production levels will continue to remain low and the country will continue to remain dependent on imported consumer and intermediate goods. There should, therefore, be a shift over time from a dominance of consumer goods in the import structure towards intermediate and
160
Table 5.8: Merchandise import trade, 1994-2001, % distribution 1994 Customs (c.i.f) Cigarettes Motorcycles Beer Durablesa Gold Vehicles Construction materials Cloth Clothing Petroleum products Sugar Cement Steel Other Coverage adjustmentsb Duty-free imports Total Imports (c.i.f ) Total Imports (f.o.b)/(c.i.f ) Changes in exchange rate (Riel /USD) 79.6 17.3 4.1 1.8 5.9 9.8 1.4 3.1 1.3 4.4 7.6 1.1 0.6 0.3 20.9 20.4 100.0 92.0 100 1995 73.1 15.2 2.8 1.2 2.9 24.1 1.7 1.5 1.4 1.0 7.9 0.5 1.0 0.3 11.5 26.9 100.0 93.9 96.3 1996 62.2 18.1 1.6 1.2 1.9 3.5 2.1 1.7 1.8 0.9 11.1 1.2 1.3 0.4 15.4 37.8 100.0 92.3 107.1 1997 66.4 16.0 1.5 0.4 0.7 11.6 2.2 1.6 1.5 0.3 11.6 1.3 1.4 0.1 16.3 33.6 100.0 93.0 112.3 1998 44.3 12.3 3.8 0.3 0.8 0.3 1.3 0.7 1.9 0.3 12.6 1.2 1.1 0.5 12.4 21.4 34.3 100.0 92.1 127.1 1999 44.4 9.5 2.8 0.2 0.8 2.2 2.2 1.1 2.7 0.3 12.0 1.7 1.6 1.5 18.9 17.6 38.0 100.0 92.2 101.7 2000 35.4 4.2 1.9 0.2 0.6 2.1 1.4 0.8 2.8 0.2 9.4 0.6 1.7 1.3 18.3 13.3 51.4 100.0 92.2 101.0 2001 31.9 4.0 1.2 0.1 0.5 0.7 1.5 0.8 1.6 0.6 10.1 1.5 1.8 1.0 16.0 13.5 54.6 100.0 92.1 102.0 2002 29.1 2.7 1.1 0.1 0.3 0.4 1.5 0.5 1.5 0.2 6.0 1.0 1.5 0.9 11.2 13.1 57.8 100.0 92.4 100.0
Notes: a. Includes VCRs, televisions and audio cassettes. Source: Calculated from NIS 2002a; NIS 2003.
b. Prior to 1998 includes duty-free imports.
The published data of recorded trade are also not very helpful, showing as they do, only some of the main imported products. Thus it can be seen that the share of certain luxury consumer goods and those that were formerly re-exported has declined, while the share of intermediate goods has increased substantially - particularly reflected in the expansion of the share of duty-free imports for the garment industry. The rising share of dutiable cloth imports, and corresponding reduction in that of clothing, also reflects
the growing importance of garment manufacturing. Fertilizer is a notable absentee from this list. However, in the absence of more detailed data it is difficult to say anything meaningful about the structure of imports. There is no indication, for example, of whether imports of capital goods for development are generally on the increase. Although most quantitative restrictions and licensing of imports were eliminated in 1994, a few goods do remain under some
161
form of control: pharmaceutical products, gold and silver, ornaments, ammunition and various cultural and medical materials. Imports of some goods, including pork meat, used motorbike tyres, right-hand drive motor vehicles27 and used footwear, are banned (RGC and Ministry of Commerce 2002). 3.5.2 Influence of Tariff Structure on Imports Compared with Cambodias export markets, its trade regime is relatively open and the tariff regime has not evolved to protect local industry. The existing tariff structure is based on four broad bands, which correspond to the degree to which products have been processed. There exists a cascading tariff structure with, generally speaking, the highest rates applying to processed goods and lowest rates on raw materials. Such a tariff structure can provide high effective protection even with a tariff ceiling of 35%. A rate of 0% applies to breeding animals, vegetable seed and special medicine, 7% rates are imposed on chemical raw materials, raw materials for textile and clothing (in practice exempted for garment factories), metal ores, base metals, wood in rough, wool and mineral fertilizers. The 15% rates apply to capital goods, locomotive, machinery and electrical appliances, finished industrial products and semi-finished agricultural products. The 35% rates are imposed on finished agricultural products, finished industrial products, alcohol, precious stones, precious metals, motorcycles, cars and petroleum products (Naron 2002).
The Law on Investment provides for exemptions on the import of goods for export production. However the government does charge a specific fee (visa fees) per dozen of garments sold under the quota to the US and EU (Naron 2002). In order to limit exempted imported inputs to goods actually used in export production, exporters are required to complete a Master List of planned imports. This practice leads to high bureaucratic costs for exporters. Duty rebating or duty referral would also cause the same problem and hence zero tariffs on goods most frequently used would be an option to make the system more transparent and lower the administrative costs. Risk-based auditing is another option. Lowering tariffs on all goods in the long run would diminish this problem and would improve the attractiveness of export-oriented investment. Cambodia is aiming to reduce its tariff to 0-5% in line with the Common Effective Preferential Tariff (CEPT) by 2010. Differences in tariff rates may result in some negative outcomes. Lack of information on different tariff rates, leads to scope for unofficial charges, and differences between CEPT/AFTA rates lead to alternative sourcing strategies, encouraging intra-ASEAN trade. So far, tariff adjustment has taken place with no adverse consequences. The amount of duty paid by importers is also affected by valuation procedures. The Pre-shipment Inspection Company proposes a dutiable value based on fair market value, while in other cases the
27 The ban on right-hand drive vehicles, which have been imported in large numbers from Thailand, is quite recent. Casual observation suggests that about 40% of passenger cars on Cambodia's roads have the driver's seat on the wrong side of the car.
162
Ministry of Economy and Finance sets minimum or fixed values for duties and other taxes. The imposition of tariffs has consequences for local consumers, not just producers involved in the production of traded goods. In order to assess the nature of the impact of tariffs on consumers at all levels, further research needs to be conducted in the range of products affected by tariffs, examining demand and elasticity questions as well as the implications for the poor.
External constraints upon export growth are likely to remain, for a small country lacking in influence on the global stage, essentially beyond the governments control. Of particular concern at present, in view of the narrow focus on US and European markets, is the impact of recession in these markets. The problem for Cambodia, therefore, is how to expand it export base and grow its exports in the face of these external constraints and in such a way that this growth will be pro-poor. Cambodia does have significant areas of potential to expand its export base, such as in rice production and processing, niche products such as spices, herbs, speciality tropical fruits, sesame seeds and essential oils, handicrafts, and developing further its tourism industry. A more diverse basket of exports would also decrease reliance on one or two markets. A very notable feature of most late-industrializing economies (South Korea and Taiwan come to mind as recent examples) has been the strength of their domestic markets. In other words, export growth has not come in the form of enclave development with few linkages to other sectors of the economy, but as a result of rising demand in the domestic market, which both encouraged scale economies and efficiency in export industries and provided savings that were extensively mobilized for investment purposes. However diversification away from the garment sector is unlikely to occur in short term given the current supply constraints in the economy and the financial constraints on domestic private investment in particular. Unless there are considerably greater
4. DEVELOPING CAMBODIAS EXPORTS FURTHER 4.1 Potential for Future Growth of Exports
Its narrow export base leaves Cambodia vulnerable to international competition upon accession to WTO. There is, therefore, an urgent need to diversify into other sectors, not only to broaden the revenue base, but also as a source of expanding employment opportunities. The concentration of ownership of garment enterprises by non-Cambodians is also of concern. In particular, widespread Chinese ownership of garment factories leads to the worry that these will relocate to China as that country gains free access to the US and European markets under WTO rules by 2008. There are also concerns about the balance of trade once Cambodias quota allocation expires in 2005. Although its comparative advantage in relatively cheap labour will continue to provide some incentive for foreign investment, Cambodia will face greater competition from large-scale producers such as India and China after 2008 (Naron 2002).
163
incentives for producers to invest outside of the current growth enclaves, trade liberalization will not prove to be effective in achieving pro-poor growth. 4.1.1 Textiles and Garments The most obvious area of immediate export growth is in the garment sector. However with the current business climate and supply-side constraints (discussed below), which have an impact on all aspects of the economy, and the accession of both Cambodia and China to WTO, the potential for growth in the garment sector is unclear. Unless the existing constraints are effectively addressed, it is not evident WTO accession will see further growth in the garment sector. As noted earlier, imported inputs currently account for 63% of the gross value of garment exports. Only 4% are sourced locally and these consist primarily of utilities such as power and water. Some of the major expenses associated with garment production arise from the cost of importing these inputs while the costs faced by Cambodian producers are considerably higher than elsewhere in the region (see Section 4.2). Significant increases in competitiveness could, therefore be achieved, not only by improvements in port and customs efficiencies, but by sourcing more inputs domestically. Some parts of Cambodia have good conditions for growing cotton, although these may need some government assistance if the quality and price of cotton is to meet export standards in short time. Unlike neighbouring Vietnam, which is one of its major competitors in this sector, Cambodia does not have a developed textile sector and therefore needs to rely on
imports of its bulkiest input. Other areas in which linkages could be developed are in the production of accessories such as buttons, zippers, thread, ribbon and in complementary industries like embroidery. These would also assist the development of a stronger domestic-market oriented garment industry and help producers to achieve economies of scale in the longer run. Even if such industries do not at first operate at optimum efficiency (the infant industry argument), there could be substantial savings on transport- and customs-related costs, which would result in efficiency gains to the export industry as a whole. They would also aid the expansion of local employment and locally retained value added which is crucial for continued growth with more equitable distribution. 4.1.2 Tourism The potential for increasing tourist numbers is significant. In addition to Phnom Penh and the historical sites at Siem Reap, there is scope for ecotourism, in Takeo, Kompong Speu, Kratie, Ratanakiri and Mondolkiri provinces, and in the Bokor and Cardamom mountains. Coastal resorts such as those around Sihanoukville could also be developed. These offer the opportunity for wider geographical distribution of growth. The tourism sector can increase demand for products and services from other sectors, such as construction, transportation, petrol and electricity, wholesale and retail trade and agricultural products. Local handicrafts also have potential to grow in response to demand from tourists as well as other overseas markets. At present, however, both the quality and quantity of such products
164
need to be improved, and this is reflected in the prevalence of products from neighbouring countries being sold in local markets, e.g. from Vietnam, Thailand and China in particular. The existing structure of the tourism industry results in a disproportionate amount of the value added accruing outside of Cambodia. Many of the hotels, casinos and airlines are owned by non-Cambodians and, although we could find little hard evidence, the tourism market appears to have become more concentrated. Further, while government revenues from the Angkor park have improved since the visitor franchise was leased to a private company, it may be the case that the distribution of benefits to the surrounding community is less broad than before. Similarly, employment generation in activities like construction and restaurants has increased, but may not be stable, particularly if further linkages to the local economy fail to develop. It is a striking fact that, despite its major tourist industry at the provincial capital, Siem Reap province as a whole remains one of the poorest in the country. The government should give serious consideration to encouraging more backpacker tourism. While backpackers do not spend as much money on a daily basis, they do tend to remain in the country longer (3 weeks compared to a few days) and they are more adventurous and likely to travel to areas which currently have more potential than actual tourism. Experience in hosting backpackers can provide local communities not only with extra income, but a valuable learning experience in foreign tastes and service requirements which will
enable them to develop their facilities. Backpackers are also more likely to spend money in small and micro enterprises, such as retail shops, food stalls, motorbike transport, selling local products and services and thereby creating linkages from directly tourist-related enterprises to the wider local economy. 4.1.3 Handicrafts Handicraft production has traditionally involved the transfer of skills from generation to generation, a process that has been broken by years of civil war. There is potential to further develop the quantity and quality of handicrafts produced locally, for both the tourist industry in Cambodia as well as for international markets. Currently there is a wide availability of handicrafts from the neighbouring countries of Thailand, Vietnam and to a smaller extent China, sold in Cambodia, again representing a lost domestic market for Cambodian producers. Since handicrafts are by definition labour intensive, the encouragement of this industry can make a major impact on employment. In particular, many handicrafts are done by women - usually as a part-time activity in association with farming and raising families. Development of the handicraft industry would thus provide more incomeearning opportunities for these women, in a relatively flexible working environment, as well as allowing more specialized production to develop over time. 4.1.4 Rice Rice production is the single most important economic activity, accounting for more than 80% of estimated agricultural labour
165
input. It also provides an estimated 75% of total energy intake and is an important factor in food security, (Ministry of Planning 1999). In Cambodia, most of the rural population is involved in subsistence farming. However the traditional technologies, lack of income for better quality inputs and, most importantly, the absence of functioning irrigation systems make Cambodian rice production one of the least productive in the region. This low productivity has implications for food security as well as potential revenue from exports. The lack of appropriate infrastructure, and hence difficulties in getting access to markets, has even resulted in the export of surplus rice from some areas, while others suffer from food shortages. However, there is a strong case to be made for the export of high quality rice from surplus areas and importing lower quality, cheaper rice for food security purposes. Cambodia has been a net exporter of rice since 1996, having exceeded its level of selfsufficiency (3.5 million tonnes of paddy) at the national level (Sok Hach et al. 2001). However, due to the lack of transport network and marketing, many parts of the country are unable to access food, and remain food poor. Even during this period paddy was exported to Thailand and Vietnam, and to counter the shortfall in rice, milled rice was imported from those countries during the wet season as donor-funded food aid and for commercial sale (Sok Hach et al. 2001). The low productivity of rice production is a significant obstacle to growth in this sector. Some attempt has been made to introduce commercial links with rice producers, through which marketing and processing of the paddy is carried out. The
scale of the operations varies in size from small-scale family businesses to large mills. However the majority of farmers are small scale, using traditional farming tools and techniques. Much of the marketed rice is from Battambang. Price increases in rice can often be tracked to changes in the baht or dollar prices. Thus, the rice price might increase due to increases in baht price or depreciation of the riel against the Thai currency. Border traders can usually obtain higher prices for Cambodian rice in Thailand. In the eastern region, Vietnamese traders are also willing to pay a higher price in riel than local traders (Sophal et al. 1999). Farmers do not benefit from such increases in price since the transporting and cross border trading of the rice are done by other agents. Moreover, since the vast majority of processing takes place outside Cambodia, the rice export trade currently produces little value added. 4.1.5 Diversification of Farm Activities The agro-industrial sector is expanding, and already includes oil palm, crude palm oil mills, tapioca plantations, tapioca starch processing plants, large-scale internationalstandard cattle & buffalo rearing and feedlot farms, rubber and cashew plantations, rice farming, rice mills amongst other produce (Van Sou Ieng 2002). The high cost of transportation and the high unofficial fees are said by representatives of this sector to increase production costs and reduce competitiveness. They have called for an exemption from container scanning charges at the Sihanoukville port for all export of agri-products and all agriculture
166
and agro-industry inputs, as well as an 80% reduction of road tolls. To increase local content in production and local added value, local producers are keen for quotas to be set for goods imported when local produce is available. Rubber is a highly labour-intensive crop, and encouraging this sector provides opportunities for generating rural employment. The current annual level of production is 45,000 tonnes, only 25% of its 1960s level, largely due to the old age of the trees and a lack of investment in replacement planting. There is scope for increasing the land under cultivation, but a major obstacle is the capital requirement: since rubber trees take 6 years to become productive, poor farmers generally cannot afford the investment (see further in Chapter 7). As well as the incorporation of the seven existing state-owned plantations, small farmers who engage in rubber production have been assured they will not lose their land title. The new market for rubber wood offers further opportunity for rural employment and diversification of crop production away from rice. Limited processing activities do take place at the household level, these include milling rice; smoking, drying, fermenting or salting fish; fermenting cabbage and pickling vegetables. Increased information on new technology and inputs, as well as access to markets via improved infrastructure and to investment finance, are all necessary if the potential for commercialization and growth in these areas is to be realised. Various controls on livestock exports have been implemented covering a variety of concerns including health and security or
quotas and inspection procedures applied to meet requirements of trading partners. The poor health of animals, particularly pigs, and the lack of health and hygiene awareness have resulted in widespread outbreaks of disease amongst livestock. Lack of veterinary services, health and hygiene awareness and affordable treatments affect not only potential for exporting, but have direct implications to household welfare, since livestock is a form of household asset (see Chapter 1). Export taxes are also applied to several classes of goods, such as live animals, fish and raw hides and skins. Although the objective is employment generation, or encouraging value added, in practice this has been found to be at the expense of another activity, with little output gain, encouraging illegal exports and reducing the returns to animal producers from hides and skins. Low quality production has consequences for competitiveness in international markets. 4.1.6 Fishery Cambodia has one of the largest and most fertile freshwater fisheries in Southeast Asia. The annual reflux of the Mekong River up the Tonle Sap into the great lake and flooding of the wetlands that surround it, help to produce particularly productive fishing grounds. These are, however, under threat from environmental degradation. Over-fishing, in particular, is a problem affecting sustainability. Recently, reform of the fishing lot system has reduced the size of commercial concessions by a total of 536,302 hectares, accounting for 56% of the total fishing lots in 12 provinces. This move should both reduce the pressure on fish stocks and increase access for poor fishing
167
families to what was previously a common resource. Development of the fishery sector is hindered by intervention of government agencies in almost all aspects of market transactions. Marketing is dominated by the state-owned enterprise, KAMFIMEX, which has the sole authority to control fish export. All fish destined for legal export are sold through this agency, which in turn licenses five export traders to take physical delivery of the fish and transport them to the Thai border to the Aranyaprathet fish market. The Kamfimex monopoly currently both encourages smuggling because of the associated bureaucratic costs and reduces prices to direct producers. Currently, its monopoly rents appear to provide few incentives for it to develop processing activities that would increase the value-added accruing within Cambodia.
4.2.1 Governance Lack of transparency and the sheer number of bureaucratic procedures present formidable costs to any form of transaction in Cambodia. Table 5.9 gives an indication of the importance of these in accounting for the higher costs borne by Cambodian producers, as compared with those in some other regional countries. Low civil service salaries, inadequate physical support facilities, poor management information systems and the large number of goods subjected to different rates of tariff are at least partly accountable for these extra costs. They result in high unofficial payments, unreliable clearance and processing times, and tariff collection that is significantly lower than is implied by the official rate. In addition to raising costs for existing producers, they can be a deterrent to new investors, both domestic and foreign. Even if unofficial payments to facilitate transactions were known, and thus accounted for in production costs, there is large uncertainty of the service being delivered upon payment and hence trade becomes too risky for many potential private investors. There are also wider implications in loss of revenue due to lower tariff collection and diversion to unofficial payments. Reduced government revenue has ramifications, in turn, for government expenditure. Problems of governance and lack of transparency have been recognised, and the Governance Action Plan sets out to tackle these issues through the establishment of a modern legal and judicial system that can provide for a clear rule of law and modern commercial laws and regulations. It seems more likely, however, that governance will be improved if the incentives to evade regulations are
4.2 Where are the Major Constraints?

The major constraints to export growth and diversification that we have identified in this chapter reflect the fundamental questions concerning macroeconomic policy settings that we have raised in earlier chapters. Most poor Cambodians have few opportunities to engage with the market economy, let alone the international market. While recent political and economic stability have improved the situation, we have also seen (in Chapter 3) that the present deflationary situation is unlikely to enhance opportunities for the poor. Moreover, trade-off decisions related to the limited resource envelope (Chapter 4) have tended to diminish the possibility of reductions in income poverty.
168
reduced. To this extent the elimination of differences in tariff rates, which currently encourage such corrupt practices as misclassification of goods, and imposition of a common low tariff would be advantageous. 4.2.2 Infrastructure Trade facilitation is a problem. Attention needs to be addressed to impediments to the movement of goods which might arise in the form of poor infrastructure (RGC and Ministry of Commerce 2002).
A recognised problem within Cambodia is the absence of a local supply of raw materials and inputs necessary for production, in particular affecting the garment sector. Delays in transporting inputs are thought to be around 10-12 days, compared to countries such as China, India, Thailand or Indonesia where local supply is available. Table 5.9 also gives an indication of the high transportation charges faced by Cambodian exporters and importers.
Table 5.9: Import and Export Cost Comparisons, July 2002 (20 foot container) Cambodia Hong Kong Malaysia Sri Lanka Madagascar ($US) Import Costs Trucking Customs Lifting Terminal Handling Charges (THC) Documentation Scanning of container 0.1% invoice value for Cam Control Other expense sub-total Export costs Trucking Customs Lifting Inspection (at factory) OT charge for inspection Terminal Handling Charges (THC) Documentation customs inspection fee Miscellaneous 135 210 26 60 150 70 15 150 30 150 0 0 0 0 225 15 0 0 45 0 0 0 0 83 13 0 0 52 36 20 0 4 143 0 18 15 50 25 38 0 136 35 0 10 140 190 77 70 15 50 0.10% 50 592 150 0 0 225 15 0 0 0 390 66 0 0 108 13 0 0 0 187 52 4 37 143 0 18 0 21 275 50 25 136 35 0 0 10 256 ($US) ($US) ($US) ($US)
169
sub-total Export Documentation C.P. Inspection C.P. Application C.O. Inspection (per set) C.O. Applicaton Visa Application Visa Export Management Fee (1.28/doz, 5,400 pcs) sub-total Total Export Costs
Source: Presented at the Private Sector Forum, August 2002
846 0 31 15 33 22 31 540 672 1518
390 0 0 0 14 0 25 0 39 429
141 0 0 0 2 0 0 0 2 143
288 0 0 0 3.5 0 3.5 0 7 295
294 0 0 0 20 0 66 0 86 380
The lack of good infrastructure is a significant barrier to the gains from trade. This includes transportation, communication and access to sources of energy, such as electricity or water. The absence of a good quality road network in particular excludes the rural areas from taking up market opportunities in domestic and non-domestic markets, as well as the access to goods and services. Transport costs are also a significant barrier to trade. There are unofficial charges for road use, though some roads do carry user charges or toll operations to fund road maintenance and construction. Port charges are also significant. Shipping charges in Cambodia are the highest in the region and as much as four times those of Thailand, (RGC and Ministry of Commerce 2002). Although Government strategy in several reports, such as the IF, SEPD II and I-PRSP includes rural development, effective budgetary expenditure marked for spending in economic services is insufficient to make effective change to the current situation. 4.2.3 Human Resources Current poverty alleviation policy focuses heavily on education and health. As discussed
in Chapter 4 these do not in themselves created new employment opportunities and in the long run the benefits may be lost if opportunities for more highly skilled work are not available to those completing their education. Moreover, the focus of the education policy is on the future workforce, rather than the current one and this feature is particularly discriminatory towards the significant majority of women who comprise the existing workforce. Women are more likely than men to be involved in trading activities in Cambodia (see Chapter 2). They are also significantly involved in several productive activities with export potential, such as handicrafts and livestock raising. The tourism industry is directly and indirectly a major employer of women. Training in business skills, provision of extension services to farmers and other forms of access to information and improvements in technology, will not only attract investment into the country, but also increase the capacity of local business ventures. Especially if targeted towards women, they will help to reduce the current gender imbalance in the incidence of poverty and influence the attitudes of mothers towards education and training for their daughters.
170
4.2.4 Finance for Rural Development Finance for the development of the private sector comes from two sources, public and private. We have already seen in Chapter 4, that certain policy decisions have restricted the ability of government to allocate resources to agriculture and rural development. The discussion of private finance is left to Chapter 6. Suffice it to say here that the current macroeconomic framework, which leaves around 3% of GDP as idle commercial bank deposits (possibly invested outside the country), is sub-optimal from the point of view of rural development. It is also a strong indication that the private sector is unlikely to take the initiative to invest in diversification and development of the rural economy.
to the lack of infrastructure there has been limited redistribution from urban to rural areas. To enable equitable distribution of the benefits of trade, there need to be opportunities for the rural population to participate in trade-oriented activities. There are currently many obstacles preventing such participation. Although current policy statements emphasize pro-poor trade, actual implementation does not appear to address these intentions. The delay in addressing these constraints will further increase the already existing inequities in access to resources and work against poverty reduction. Trade liberalization in Cambodia will not contribute to poverty reduction if other macroeconomic policies are not supportive. In reality, the impact so far has been rather narrow, in terms of its impact on employment, income distribution and contribution to GDP (in Keynesian accounting terms, X-M). As well as the narrow range of products being exported, the high import content needs to be addressed since it restricts the benefits to the local economy that come with trade liberalization. Trade liberalization will benefit the poor if policy is structured as to develop local markets that will then eventually have the capacity to compete in international markets. Addressing the immediate constraints in the country in the areas of; governance, infrastructure and poor human capital, will go further in addressing pro-poor trade growth since it will expand the number of products that can be exported and widen the proportion of the population that can benefit from trade liberalization, and thus establish a conduit for the re-distribution of growth.
5. CONCLUSION: POLICIES FOR GROWTH AND REDISTRIBUTION UNDER TRADE LIBERALIZATION

There are two ways to impact poverty reduction; 1) directly through participation in the growth process, 2) or indirectly through the redistribution of gains from growth. Currently neither of the above is taking place and the benefits from the current sources of export growth have had a limited impact on the majority of the population, and by implication on poverty reduction. The main beneficiaries of trade liberalization have been those directly involved in the activities, which are largely in the urban areas, and the limited auxiliary service activities that have developed around these activities. Even in the provinces, due
171
6
172
Financial Liberalization
1. INTRODUCTION
Financial liberalization proceeded rapidly in Cambodia from 1992 onwards. The main thrust of the reforms have been the creation of a commercial banking system out of the previous monobank, liberalization of the interest rate regime and a process of bank restructuring aimed at reducing the number and improving the quality of commercial banks in an attempt to instil public confidence, strengthen governance and supervision, and adopt a market-based indirect system of monetary, exchange and credit management. This chapter reviews the process of financial liberalization, the structure and performance of the banking system, its connection (or lack thereof ) to rural finance and to the micro-financing of consumption and investment by the poor. We argue that financial liberalization, while assisting in the stabilization of the Cambodian macroeconomy and promoting financial intermediation within the rapidly expanding urban (dollarized) economy, has so far failed to encourage pro-poor growth. Financial markets remain fragmented and, in the rural areas, poorly managed, involve high costs to clients and
The macroeconomic of poverty reduction in Cambodia Financial Liberalization
disincentives to mobilization of domestic savings, and are unable to meet the demand of the poor for credit to sustain and improve their livelihoods. Part of the problem lies in the bumpy process of liberalization itself, which has induced a number of crises of confidence in the system. Another part lies in the relative absence of accompanying macroeconomic measures, particularly the lack of demand expansion in rural areas. This absence has created a major obstacle to financial deepening in the countryside by ensuring that banks remain unwilling to invest in the rural economy where most of the poor live, thus forcing the poor to rely on informal and highly fragmented markets for credit.
barter or gold, with the latter being the universal means of hoarding. During 198891, the Vietnamese disengagement left an unsettled political situation. Massive central bank financing of recurrent budget deficits during that period resulted in high inflation, in the range of 90-177% a year (end period), and in an erosion of public confidence in the national currency. 2.1.1 Initial Reform in the Banking Sector 1989-1992 The transition from a centrally planned to a market economy meant first of all a change in the allocation mechanism of financial resources through decentralization. Since, in the centrally planned economy, banks had no active role in allocation of financial resources this role was initially taken up by non-bank institutions because neither commercial nor central banking skills had yet been developed. The initial reforms gave state-owned enterprises greater autonomy in directing and managing their own business, including financial autonomy, and permitted private sector activities, supported by the legal restoration of private property. For the first time since 1975 foreign investors were encouraged. Within the financial sector, reforms in 1990 also gave greater autonomy to provincial and municipal branches of the mono-bank, allowing them to extend credit to all economic sectors. While some of them had sufficient deposits, most had borrowed substantial amounts from headquarters. Inefficiencies of administrative lending under the previous policy contributed to
2. FINANCIAL SECTOR REFORM IN CAMBODIA 2.1 Transformation of Structure and Sunction

After 1979 Cambodia literally had to reconstruct its financial sector from scratch. Under the extreme revolutionary experience of the Khmer Rouge during 1975-79 money had been abolished and the central bank physically destroyed. In 1980, once the Khmer Rouge had been driven out of most parts of the country by the Vietnamese, the riel was reintroduced and a state-owned mono-bank (the Peoples National Bank of Cambodia) was established. Credit allocation became dependent on central planning and had a purely accounting character. During the 1980s, therefore, there was only limited monetization of the economy and most non-plan transactions were based on
173
the build-up of non-performing loans in the state banks. In 1991 the Peoples National Bank of Cambodia (PNBC) and the National Treasury were separated. Two commercial joint ventures between PNBC and private banks were established. In 1992, as foreign investors began to arrive, eight additional commercial banks followed. These were either locally incorporated or subsidiaries or branches of foreign banks. In the same year PNBC reacquired its pre-Khmer Rouge name of National Bank of Cambodia (NBC). Under the 1992 central bank law, the authorized capital of NBC was set at five billion riel (approximately USD 3.7 million).28 This law gave it financial autonomy and divested it of commercial banking functions. However, it still lacked independence in formulating and conducting monetary policy, especially in relation to financing of the budget. The result of this deficiency was soon visible. With the collapse of Eastern Europe and the end of financial support from the former Soviet Union, pressure to undertake inflationary financing was aggravated (see Chapter 3). Although the various measures taken throughout 1989-1992 had helped create a new environment in the banking sector, the legal and structural framework of this sector in place by early 1993 was not yet well adapted to the requirements of the market economy. Lack of appropriate skills in the sector and uncertainty at the political level about the issues that needed to be addressed had contributed to this deficiency,
as did exogenous factors such as the dilapidated infrastructure and macroeconomic instability. 2.1.2 Financial Reform Since 1993 Formation of the coalition government following the UN-sponsored general election in May 1993 marked the beginning of a new era in the reform process. Assistance from leading countries in the Paris Club to help Cambodia clear its arrears with the IMF enabled the resumption of the relations with the latter. As a result, a more comprehensive economic program was developed under the Funds Systematic Transformation Facility. The focus of financial and technical assistance under this facility was on development of macroeconomic policy and the establishment of a market-based administrative and structural framework, including restructuring the banking system. In mid-1994 the IMF, together with NBC, produced a Policy Framework Paper containing comprehensive economic reform programs to be supported by the Funds Enhanced Structural Adjustment Facility (ESAF). Within the framework of the overall objectives of the economic reform, the objectives envisaged for the banking sector were the following (Source: NBC): 1. Reducing or eliminating where possible the direct intervention of the government with the aim of strengthening the role of market forces in the allocation of financial resources;
28 Average exchange rate for 1992 was 1,363 riel to a US dollar.
174
2. Developing a modern banking system and promoting competition among commercial banks; 3. Enhancing the effectiveness of monetary policy instruments; and 4. Strengthening the legal and regulatory framework in order to ensure an effective and smooth conduct of central and commercial banking activities.
chiefly by strengthening the regulatory system and prudential supervision. The Law on Banking and Financial Institutions, for example, represents a step towards promoting more orderly financial markets by providing the legal framework for the licensing, organization, operation, and supervision of a broad range of financial service companies. The principal issues covered by the new legislation are the following:
The delineation of business and operations of commercial banks and specialized banks. Despite some restrictions imposed on the business of specialized banks, commercial and specialized banks can deal with security transactions but cannot be involved in insurance services.
2.2 Legal Infrastructure

The current legislative framework has largely been developed in the second half of the 1990s in response to deficiencies evident in earlier measures. These recent reforms have focused largely on re-regulation of a system that had earlier suffered from excessive liberalization following the collapse of the former centrally planned economy. The main elements in the new framework are: 1. The Law on the Organization and Conduct of the National Bank of Cambodia, promulgated in January 1996, replaced the 1992 law. 2. The Law on Foreign Exchange, enacted in May 1997, replaced the Law on the Management of Foreign Exchange, Precious Metal and Stones promulgated in 1991; 3. The Law on Banking and Financial Institutions enacted in November 1999 and a number of subsequent regulations (prakas) issued by the NBC. Thus the Cambodian banking system is still developing. Until now the focus of reform has been on issues relating to the build up of public confidence in the banking sector,
Bank licensing procedures;
The regulatory and supervisory power of the NBC;
Requirements for the integrity of bank management;

Identification of shareholders, and their responsibilities;
Requirements for internal control as well as obligations of external auditors;
The power of the NBC to impose disciplinary sanctions and penalties against noncompliant banks;

Bank liquidation procedures; Provisions for bank re-licensing.
175
In general, however, a good legal framework for the financial sector has yet to develop. At present, for example, articles 11 and 12 of Contract Law No.38 state that security over property is only effective if the object is delivered to the creditor in person. In other words, banks are compelled to act like pawn shops. A new draft contract law, based on the Malaysian Contracts Act 1995, is not yet approved. The proposed law allows for pledge of goods on a non-possessor basis and will create both improved security for bankers and advantages for micro-finance institutions (MFIs). The Malaysian act does not, however, deal with cession of intangible assets such as debts, bank balances or shares, so there remains further room for development. Other laws in Cambodias developing legal framework that are relevant to the financial sector include the Commercial Enterprise Law, which has yet to be passed by the National Assembly. The draft deals with sole proprietorship, partnership, public and private limited liability company, and joint ventures. Co-operatives and trusts, however, are not mentioned. A law on co-operatives will be needed for MFIs wishing to incorporate as credit unions. Drafting of a bankruptcy law has yet to begin. Ownership and legal status of international and local NGOs present a problem. At present local NGOs have to be approved and registered with the Ministry of Interior while international NGOs are under the Ministry of Foreign Affairs. There is a draft law on associations, also yet to pass the National Assembly. It covers only domestic NGOs, but does not deal with NGOs ownership of
their capital funding, legal identity or public liability. Since NGOs in the micro-finance sector often have financial relationships with international sponsors (not dealt with in the law), the legal framework requires further development. Cambodia has yet to adopt common accounting and audit standards or to establish an enforcement mechanism. MEFs Commission for Accounting Reform has endorsed the adoption of International Accounting Standards (IAS) and International Standards on Auditing (ISA) as core principles, and submitted a draft Accounting and Audit Law to the Council of Ministers for consideration. The Central Banking Law also permits NBC to establish accounting standards, distinct from those for taxation purposes, for the banking sector. Currently, most of these date from 1989 and are not in accordance with IAS. However IAS/ISA will be difficult to implement in Cambodia due to the limited capacity and compliance cost to the private sector over the medium term. While many necessary laws are not yet in place, enforcement of existing law is currently weak and financial institutions still operate in an uncertain environment that is detrimental to both efficient operation and customer confidence.
3. BANKING SECTOR DEVELOPMENT 3.1 Structural Reform of the Sector

At present, banks are the only major financial institutions in Cambodia. The system has, however, played a limited role in economic
176
development - it accounts (excluding the central bank) for only 20% of GDP and does not function effectively as a financial intermediary, except perhaps in the urban environment. Although the authorities have intensified efforts to restructure the banking sector over the last few years, the system is far from meeting development needs. Cambodias efforts in modernizing the banking system started from a very low level with the total asset size of all commercial at less than 10% of GDP, and both broad money and private sector credit at only 3-5 percent of GDP in the early 1990s. After some initial success, a deteriorating economic and political environment during 1997-98 slowed progress, although the shortage of financially sound banks was perhaps a more important factor in hindering financial development. Table 6.1 suggests that Cambodias banking system currently has more in common with several poor African countries than with those of its more prosperous ASEAN neighbours. Clearly there is room for further progress.
Table 6.1: Selected indicators of financial development
(In percent of GDP, unless otherwise indicated; 1998-2000 average)
Broad money
Credit to private sector 17.9 12.0 6.8 68.0
Currency/ Interest rate Per capita broad spread 4/ GDP (1999, money (%) (%) in USD) 20.2 33.9 33.2 8.4 9.2 14.7 15.4 2.7 398 335 241 2,242
PRGF Asian countries 1/ PRGF Sub-Saharan Africa 2/ Cambodia ASEAN 3/
28.5 22.5 13.2 82.0
1. Bangladesh, Lao P.D.R, Mongolia, Nepal, Sri Lanka, and Vietnam. 2. Cameroon, D.R. Congo, Ethiopia, Mozambique, Rwanda, Sierra Leone, Uganda. 3. Indonesia, Malaysia, Philippines, and Thailand. 4. For Cambodia, foreign currency, saving rate (both short-term), For other countries, lending rate-deposit rate (both short term or medium term). Source: IMF 2002.
In the process of creating a two-tier banking system, the NBC was divested of its interest in all commercial banks except the Foreign Trade Bank (FTB). Subsequently 80% ownership of the Foreign Trade Bank was transferred to the Ministry of Economy and Finance and a plan has been developed to corporatize and eventually to privatize it. The number of registered banks increased dramatically in the early years of transformation - to 24 at the end of 1994. In addition, a
number of NGOs began operating in rural finance. This wave of new entrants imposed a heavy supervisory burden on a central bank with little experience in the task. Due to a lack of a clear entry policy, a large number of the new banks were poorly managed and prone to taking excessive risks. Accordingly, public confidence in the banking system (as well as mutual confidence among banks) remained very low.
177
The Rural Development Bank (RDB) was established in 1998. It operates as financial wholesaler that provides funding on concessional terms to micro-finance operators. The RDB is financially under the supervision of the Ministry of Economy and Finance. Technically, it is subject to the same prudential and supervisory regulations of the NBC as the other commercial and specialized banks. The main objective of the RDB is to promote the development of agriculture, rural and general economic activities. The 1999 Financial Institutions Law led to a further extensive restructuring of the banking system. Taking into account the scope and risk of their operations, NBC set minimum capital requirements for different types of institutions as follows:
for commercial banks and each of their branches: 50 billion riel (approx. $13 million),29 for specialized banks: 10 billion riel,
Six banks were fully re-licensed, including a specialized state-owned bank, and 13 more were conditionally re-licensed, including another state-owned bank. The conditionally re-licensed banks were required to take corrective actions, agreed with the NBC. Typically, they had to fulfil a number of conditions, including the injection of supplementary capital to meet the minimum capital requirement. By the end of 2001, 8 had fully paid up their required capital. By end-March 2002, 2 more met their capital requirement, one was voluntarily wound up, one was compelled to go into liquidation and one was converted into a specialized bank, with a lower capital requirement. In addition, 2 specialized banks were created in October 2000 and March 2001, respectively. Eighteen banks now operate in Cambodia, compared to the earlier 31. The structure of this banking system is shown in Figure 6.1.
3.2 Liberalization of Interest Rates

Generally speaking, the lower the cost of borrowing money, the more will be demanded by consumers and business. Conversely, the higher the rate of interest, the greater the supply of loanable funds. Onxe of the initial steps in the reform process, aimed at mobilizing domestic savings and encouraging the use of the national currency as a savings instrument, was thus an increase in deposit rates so that they became positive in real terms. On 15 June 1994, NBC issued a circular requesting all commercial banks to set their deposit rates on three-month riel deposits at or
for micro finance institutions: 250 million riel.
At the start of the process, the system comprised 31 banks, including 22 private commercial banks, 7 foreign bank branches, and 2 state-owned banks. As a result of the first phase of a mandatory re-licensing procedure in 2000, the NBC closed 4 banks in July 2000 and another 7 in December 2000. In March 2001, 1 bank elected for voluntary liquidation, bringing the total to 12 closed banks.
29 Exchange rate: 1USD = 3900 riel.
178
above the minimum rate determined by NBC. This measure was taken in accordance with the structural benchmark set by the IMF under the ESAF program. However, it proved to have little impact on savings behaviour since savings deposits at commercial banks grew only slightly over the subsequent period. The NBC therefore decided to liberalize interest rates by issuing a new circular authorizing the commercial banks to set rates on deposits and loans at their own discretion. State-owned banks were subject to the same liberalization. Normally, interest rates on loans are calculated on the basis of the average cost of funds, administrative expenses and other operational costs of the bank, the cost of
Table 6.2: Cambodian interest rates, 1996-2001 (Per cent, end of period)
provision for bad and doubtful debt and an allowance for profit. Different rates are therefore applied to sectors like agriculture and small industry due to their higher operational costs and risk. With regard to interest rate policy, since the NBC does not refinance banks, its does not influence interest rates and therefore cannot use interest rate policy as an effective monetary instrument. The structure of deposit and lending rates is shown in Table 6.2. High interest rate spreads are characteristic of less developed economies, but the presence of dollarization usually enhances the credibility of the exchange rate and reduces inflation, thus real interest rates tend to fall.
1996 1997 1998 1999 Deposit rates 1/ Riel savings deposits 8.8 Riel term deposits 2/ 0.9 Foreign currency savings deposits 2.5 Foreign currency term deposits 3.7 Lending rates 3/ Foreign currency loans: rates charged to private enterprise 18.6
2000 2001 Mar Jun Sep Dec Mar Jun Sep 7.3 0.7 2.2 3.5 7.2 0.7 2.2 3.5 6.6 0.5 2.2 3.5 5.9 0.5 2.3 3.7 5.1 0.5 2.3 3.9 5.1 0.5 2.3 3.9 3.3 3.0 2.0 3.3
7.4 0.9 2.4 3.9
7.5 0.9 2.5 4.0
7.3 0.7 2.2 3.5
18.4 17.6 17.3 17.3 17.3 17.8 17.4 17.1 17.0 15.5
Source: Data provided by NBC. 1. Simple averages of rates reported by the ten commercial banks with the largest deposits. 2. The average rate shown is that reported for three-month riel deposits. The volume of total riel deposits has been small throughout 1996-2001. 3. Virtually all loans to the private sector in Cambodia are denominated in foreign currencies.
179
180
NATIONAL BANK OF CAMBODIA
14 Commercial Banks 4 Specialized Banks Micro Finance Institutions 1 State owned 4 Foreign Branches
- Rural Development Bank
- Credit Agricole Indosuez - First Commercial Bank - Krung Thai Bank - May Bank
Figure 6.1: The banking system in Cambodia
20 NBC Provincial Branches
1 State Owned
9 Locally incorporated
- Foreign Trade Bank
Licensed - Enatien Moulethan Tchonnebat (EMT) - Hatthakasekar - Micro Finance Cambodia
3 Privately owned
- ACLEDA Bank Limited - Specialized Bank Peng Heng S.M.E Limited - Cambodia Agriculture Industrial Specialized Bank
- Advanced of Asia Bank - Cambodia Asia Bank - Canadia Bank Ltd. - Cambodian Commercial Bank (Siam Commercial Bank) - Cambodia Mekong Bank - Cambodia Public Bank (Public Berhad Bank) - Singapore Banking Corporation - Union Commercial Bank - Vattanac Bank
NGOs
Representative Office
- Standard Chartered
- Kampong Cham - Sihanouk Ville - Siemreap - Kandal - Battambang - Prey Veng - Kampong Thom - Takeo - Svay Rieng - Pursath - Kampong Chnaing - Kampong Speu - Kampot - Koh Kong - Preah Vihear - Kratie - Ratanakiri - Mondolkiri - Stung Treng
Exchange Bureaux
Registered 2,294 - Phnom Penh 397 - Provincial 1,915
Registered - Aid Farmers Association - Arun Reah Organization - Buddhism for Development - Cambodia Community Building - Cambodia Credit to Abolish Poverty Organization - Cambodia Health Committee - Cambodia Rural Economic Development Organization - Cambodia Women's Development Agency - Catholic Relief Service - Concern Organization - Handicap International - Help the Widow Organization - Khemara - Khmer Rural Development Association - Maxima Org. for Household Economic Development - Prasac Credit Association - Rural Development Association - Seilanithih - Social Development in RuralArunr - Samakithor Organization - The Lutheran World Federation Organization - World Relief Cambodia - World Vision International Cambodia - Women's Saving and Development Cooperation
3.3 Present Structure and Performance of Cambodias Banks

The profitability of the whole banking system improved following closure of the non-viable banks during 2000-01. The system is now well capitalized, with a capital-asset ratio of about 30% (IMF 2002), although this mainly reflects small asset size due to the limited loan portfolio. The sector is highly concentrated. Of the 18 re-licensed banks, the top five banks represent about of total deposits and 2/3 of total credit. In the deposit markets, the top 3 banks are locally owned (including the FTB). By contrast, in lending markets, foreign bank branches are better represented among the top 5, as they support business activities of their home country customers in Cambodia and do not play as active a role in deposit taking. The system is currently concentrated in the capital city, although 90 percent of the population resides elsewhere. A number of locally incorporated banks have set up branches located mostly in major commercial and tourist areas such as Sihanoukville, Siem Reap and Battambang. Banking transactions reflect the high level of dollarization, as more than 90% of banks assets, liabilities and shareholders equity are denominated in US dollars. As noted in previous chapters, the dollar economy is largely an urban phenomenon. Various institutional factors continue to hinder development of the banking sector. These include: (i) high information cost, (ii) the lack of legal infrastructure to support
the enforceability of financial contracts, and (iii) a weak bank supervisory and regulatory framework. All these factors together lead to high risk and a high cost structure for the banking system, resulting in credit rationing and inefficient financial intermediation. Banks are very cautious about providing loans because of a lack enforceable collateral and limited investment projects, and often prefer placing funds at the NBC. As the result the loan to deposit ratio is low (68 percent)30. While this cautious portfolio management helps to secure a stable source of income, expected returns remain low, and banks have little incentive to mobilize savings. Furthermore, the public itself still does not fully trust the banking system, and people often prefer holding gold or cash to opening deposit accounts at banks. The banking system is thus highly dollarized and very liquid, although FTB has a larger share of local currency denominated lending and deposits. At the end of 2000, banks held more than 90% of total deposits either in foreign exchange accounts abroad or as reserves at NBC31. Foreign banks in particular hold large foreign assets and there is reason to have some concern about the quality of these assets, as some of the banks have significant non-performing loans in their home country. In addition, banks held a large amount of vault cash to prepare for unexpected withdrawals, reflecting the lack of a central bank liquidity facility or an inter-bank market. Furthermore, the excess reserve ratio (excess reserves divided by total deposits) has significantly increased, reaching about 40-50% of total deposits, reflecting a regulation implemented in October 1998
181
preventing the transfer of domestically collected funds abroad. The system also suffers from a high ratio of non-performing to total loans (about 20% at end-December 2000), even with a rather generous definition of non-performing loans. 32 Although these loans are fully provisioned for and do not bring immediate concerns about systemic instability, the high non-performing loan ratio squeezes the profitability of Cambodian banks. As a result, although interest rate spreads shrank slightly since late 1998, they remain at around 15 percentage points. High interest rate spreads reflect various factors including inefficiency of the banks, high risk premiums (e.g., due to high default risks, legal, regulatory and other uncertainties), and high intermediation costs (e.g., due to non-transparency of the market, high information costs, and an inefficient payments system) and high operating costs including treasury costs, salaries, and other administrative expenses of the banks. The most important element in developing the financial system and reducing interest rate spreads in future will be a deepening of the financial system which will increase competition among financial institutions and enhance the efficiency of financial intermediation. At present the high interest rate spread reflects both the fragmentation of the market (leading to loss of economies of scale) as well as institutional shortcomings which, despite recent improvements, continue to constrain public confidence in the system. No doubt the restructuring and interest rate
liberalization has improved the performance of the sector in recent years, but it remains largely confined to a few urban centres and has very little contact with a rural economy which operates mostly in local currency and contains the vast majority of the poor. As we shall see in the next section, the formal banking system at present remains largely irrelevant to the requirements of poverty reduction.
4. THE STRUCTURE AND PERFORMANCE OF RURAL FINANCE

As we have noted frequently in this report, increasing employment, particularly outside the capital city, is crucial to reducing the incidence of poverty. In this regard, the establishment of an efficient, effective and sustainable rural credit system is needed to facilitate the start-up of income generating activities. The objective is to improve access to credit for both farmers, to increase the dynamism of agriculture, and rural SMEs, to provide growth in non-farm employment. At present, the banking system is unable to address the demand for rural finance - either from the poor, who need micro-finance to stabilize consumption, build assets, and develop micro-enterprises, or from the non-poor who have the potential to develop commercial cultivation, trade, and non-farm enterprises. Half of the 9.6 million rural population have no bank branch within their province, and where branches do exist they are neither
30 For example, as economic and political confidence returned in 1999-2000, foreign currency deposits increased sharply. However, an increase in deposits was not accompanied by an increase in loans, and the loans to deposits ratio declined during this period. 31 Excess reserves are defined by commercial bank deposits at NBC minus required reserves and the capital guarantee deposit. 32 The classification of overdue loans is as follows: 'standard', 3 < 6 months; 'doubtful', 6 < 12 months; and 'lost', >12 months.
182
reaching people at the district and village levels nor providing the types of credit needed for long term investment and growth. Only 6% of total banking sector advances are for agriculture or related activities, primarily short-term dollar-denominated loans. The estimated rural finance demand of $120-$130 million per annum, of which one third is for micro-credit ranging from $50-$300 per borrower (NBC 2002), is only fractionally met from institutional resources. Reliable savings facilities are generally not available in rural areas or to the poor. Currently, about 90 NGOs supported by funding from domestic and international agencies provide microfinance to nearly 420,000 poor households, or 15 percent of the total, with most of the borrowers being women. The remainder rely on the informal sector.
reforms have been supported by technical assistance from France and the ADB. Some key elements of the reforms adopted so far include the following: A regulation ( prakas ) preventing the previous common practice among MFIs of charging interest on the full balance of the loan (flat amortization). The prakas stipulates that interest on micro-credit must be calculated on the outstanding balance of a loan, after deduction of principal repayments already made.
Compulsory registration for NGOs with an outstanding loan portfolio of 100 m. riel or more, or more than 1 m. in deposits of > 100 depositors. Licensing is compulsory for those with outstanding loans exceeding 1 bn riel or over 10,000 borrowers, and for those with more than 100 m. in deposits of > 1,000 borrowers. Licensed MFIs must comply with a series of other regulations33 and must submit monthly reports.
4.1 Regulation and Supervision

RGC has adopted a rural credit policy aimed at encouraging the entry of commercial banks into rural finance, building confidence in the banking system and creating a more effective rural financial system. Among the various elements of the policy package are (i) initiation of a Rural Credit and Savings Project and technical assistance for Capacity Building for Rural Financial Services; (ii) creation of a special unit within the NBC for supervision of micro-finance institutions (MFIs); (iii) establishment of the Rural Development Bank (RDB) to provide financing of MFIs and commercial banks operating in rural areas; (iv) compulsory licensing of larger NGOs and MFIs; and (v) intervention in interest rate policies used by MFIs. These
New rules for non-performing loans require MFIs to make provision for these at 20% in the case of loans overdue for 30 days rising to 100% after 90 days for loans with an initial maturity of less than one year and to 100% after 180 days for those with an initial maturity of more than a year.
Simplification of prudential measures (liquidity ratio, compulsory reserve deposits, reporting requirements) in order to take into account the weak capacity of many MFIs.
183
On-site inspections by NBC of institutions applying for a licence, with follow-up inspections every 12 to 18 months together with quarterly reporting requirements for registered MFIs and monthly reporting for licensed MFIs.
Many of these regulations, due to cost considerations, apply only to the larger institutions involved in rural finance. As in the case of the urban banking system, they are unlikely to impact significantly on the poor in areas where, due to market fragmentation, MFIs remain very small.
Kong. There are no bank branches at all in 16 provinces that together account for 48 percent of the total population. Where branches do exist, the average population covered is 300,000 people. However, they are concentrated in provincial capitals, which are difficult for village people to reach, and they offer only limited services. Banks depend on asset-based lending technologies to cope with risk and imperfect information problems. This technology discriminates against small, collateral constrained borrowers and implies high transactions costs for medium - and smallscale entrepreneurs able to access credit. Because of weakness in the legal and regulatory framework, especially surrounding movable collateral, formal intermediaries are further constrained and not eager to expand activities in rural areas (see Section 2.2 above). Borrowers tend to rely on a single source of credit providing a single financial product. Those who access formal credit tend to be a larger farmers and a higher income households and rely primarily on cash loans of relatively long duration. Poor people tend to rely on informal sources and the loans they receive tend to be of short duration. 4.2.2 Semi-Formal Financial Institutions
4.2 Sources of Rural Financial Services

4.2.1 Formal Financial Institutions During the years of the mono-banking system, NBC financed state-owned enterprises through its 20 provincial branches as well as making small loans to thousands of farmers based on mutual guarantees. After 1990, with the privatization of banking and the discontinuance of NBCs role as a commercial bank, farmers and entrepreneurs with small and medium holdings and businesses were deprived of access to formal financial services. At present, access to formal credit and deposit services remains very low. Few bank branches exist outside Phnom Penh. Some banks, including the RDB (state-owned), ACLEDA Bank, Specialized Peng Heng SME Bank and the Cambodia Agriculture and Industrial Specialized Bank, have branches in Battambang, Sihanoukville, Siem Reap, Kampong Cham, Pailin, Kandal and Koh
There are about 90 NGOs supplying credit in the rural areas, providing micro-finance to nearly 420,000 poor households and supported by funding agencies such as KFW and GTZ of Germany, UNICEF, IFAD and USAID. Many international NGOs have
33 Dealing with capital adequacy, liquidity, loans loss provisions, risk diversification, foreign currency exposure.
184
established local counterparts in order to establish organizational continuity. Others are independent, but may receive support from foreign NGOs in the form of management guidance and/or financing. So far only three institutions have been licensed. The two most important ones are EMT and Hattha Kaksekar. A third, Cambodia Microfinance Co. Ltd., was licensed in June 2002 without following the usual procedure in the sense that its loan portfolio is below that required for licensing. Its does not mobilize any savings. The micro-finance sector is dominated by a few important institutions, and it is expected that there will some consolidation in coming years as many of the smaller NGOs are weak. In the end there will likely remain only a handful of large MFIs with extensive networks across the country, following in the footsteps of ACLEDA, which already has 68 branches in 15 of the countrys 24 provinces. Most NGOs provide loans to groups the members of which have the same occupation or type of business. In these cases there are few opportunities for loan diversification. Loans are usually provided on a six-month cycle basis. Smaller MFIs tend to have compulsory savings requirements under which members control the group savings and lend among themselves. A number of NGOs are only marginally viable while some appear to be on the brink of collapsing, putting the savings of their members at risk. Management and staff often have limited capacity in the management of savings and credit schemes: their policies are often not fully clear to all clients and they lack the
benefit of NBC supervision. Many of these are indigenous NGOs that are located in areas difficult to reach and their activities may be too small to warrant the cost of supervision. Frequently, the credit operations of NGOs are integrated with other programs such as health, education and community development. They may be dependent on grants from international organizations whose funds are often provided for a limited duration meaning that the credit operations are not sustainable. Generally, lack of long-term viability can be attributed to a lack of savings and good savings services, excessive reliance on external financing in the form of grants or project financing, lack of professional management, lack of an appropriate legal structure, and lack of external supervision. Where external assistance is either nonexistent or of limited duration, local savings must be the main source of funds. Lack of sufficient resources is a major obstacle to the extension of outreach by the NGO micro-finance sector and, in the past two years, some have had to curtail the expansion of their lending activities for this reason. In this perspective the most expedient approach adopted by NGOs and MFIs to solve their funding problem has been bank loans. Since it is still difficult to borrow from commercial banks, the chief source of such funds has been the RDB. However, because the regulation on risk concentration limits exposures to MFIs to 10% of RDBs net worth,34 this source was until recently limited to US$500,000 per borrower (RDBs net worth is currently $5million).
185
In addition, some institutions like EMT and Hattha Kaksekar have started to borrow on international markets. In the case of EMT this creates a foreign exchange exposure and increased risk since foreign loans are in dollars whereas EMT on-lends in riels. An additional risk, given the low capacity of many MFIs, is that they tend to borrow medium-term and engage in short-term lending. As a result many have borrowed to an extent that threatens their viability. Excessive borrowings on the part of microfinance institutions, the absence of meaningful asset-liability management and increasing foreign exchange exposures illustrate how weak most institutions are in financial management. With the notable exception of ACLEDA, few institutions have well-conceived and balanced funding strategies that provide for adequate capitalization and identify diversified, stable and economical sources of funds. The low skills base of the Cambodian economy shows through in this area of financial management. In addition to the activities of the RDB, other specialized banks and NGOs, the Social Fund created by the government (with financing from the World Bank) is another important source of (grant) finance for rural investment. The Social Fund focuses on rebuilding social sector rural infrastructure in a situation with weak local implementation capacities. Its supports the restoration of essential rural infrastructure through quick, small scale rehabilitation and restoration efforts identified and implemented by the communities
themselves, with the aim of generating rural employment. Although changes in the regulatory environment have made some impact on the micro-finance sector, it continues to face several challenges if it is to provide a sustainable credit service for rural development. First, most MFIs currently operating in Cambodia started out as NGOs in such fields as health care, education or community development and have been more concerned with extending their outreach and ensuring the well being of their beneficiaries than with running credit operations on a long-term viable basis. Second, as a result of poor risk management, they have reportedly been exposed to high delinquency and heavy loan losses. Banking for the poor is inherently a high-risk activity due to the vulnerability of borrowers to external shocks such as natural disasters and economic downturns. But on-site inspections by NBC showed that none of the major MFIs operating in Cambodia were using risk assessment criteria such as cash flow, collateral, personal character, capital base, etc, to assess the viability of loans. Many are, however, using land as collateral which, in Cambodias uncertain legal system, leads to problems in loan recovery and often to damaging confrontations with clients. Third, high dependence on external funding (whether domestic or foreign) has also enabled them to avoid the question of sustainable funding for their activities.
34 A limit of 20% is applicable to commercial and specialized banks.
186
Delivering a large number of small loans is very costly, yet cost control is difficult for institutions used to receiving grants and for which sustained profitability was until recently secondary to social development objectives. The problem of cost control is compounded by weak accounting procedures and expertise as well as the lack of an efficient payments system linking MFIs to the formal financial sector or to their clients. Fourth, training in banking activities for MFI staff is inadequate, mostly one-week followed by on-the-job training in which the pressure to extend new loans and collect dues exceeds that to carefully evaluate loan proposals. 4.2.3 Informal Financial Sources Whereas the formal and semi-formal finance sectors have limited coverage in the rural areas, most poor people have to rely on the informal sector. This sector has absolute information and transaction costs advantages due to proximity, dense social networks, willingness to accept a greater variety of collateral, and superior enforcement capacity, compared to formal rural intermediaries. Nonetheless, informal intermediaries have limited capital bases, tend to exercise market power, and are geographically bound. As a result, identical clients receive different loan terms and conditions. The high interest rates charged cover higher risk faced but also include a monopoly rent. According to the CSES of 1996 (NIS 1997), approximately 813,148 rural households took loans with a total value of about KHR300 billion ($120 million) during 1994-1995. About 7% of them obtained loans from traders,
15% from moneylenders, 51% from relatives, and 15% from friends and neighbours. Of the remainder, 1% took loans from banks and 10% from NGOs, United Nations agencies and 1% from others. Interest rates on loans from informal sources are extremely variable, ranging from zero usually on loans from relatives and friends - to 20% per month or more on those from traders and moneylenders. Interest free loans are used for a variety of reasons including ease of obtaining the loan, flexibility in repayments, personal relationships, and lack of alternatives, but they often carry non-monetary obligations. Table 6.3 shows that the average interest rate works out at over 100% per year, though if zero-interest loans are included, this figure falls to 66%. There is a clear and systematic difference in the rates charged for rich and poor households. These rates can be compared with 4-5% per month from NGOs, which is not cheap, especially when transaction costs in terms of the paperwork and time are taken into account.
Table 6.3: Distribution of interest rates Including zero interest rates Very poor Poor Marginal (negative) Marginal (positive) Well-off Rich Average
Source: Murshid 1998
Excluding zero interest rates 169.7 142.7 103.3 100.4 57.2 52.8 104.7
101.8 99.4 59.0 67.0 33.8 29.3 65.9
187
In common with loans from formal and non-formal financial institutions, loans from informal sources are commonly used for agricultural production (19%) and investment in business (18%). They are also important for covering sickness in the family (26%) and other purposes (16%). Much of the high cost to the poorest classes is related to distress borrowing predominantly for rice purchases or treating illness (Table 6.4). While such borrowing may temporarily alleviate the stresses of poverty, it is not conducive to long-term poverty reduction. Although the richer classes do participate substantially in the credit market, their main
Table 6.4: Use of loans Household type Total borrowers
purpose is more likely to be for working capital or input costs. There is a great variety of credit relations in the informal sector that are often, despite the high cost, advantageous to those who need to borrow under stress. There is flexibility in the repayment system, with cash obligations convertible to kind or labour, though this may entail an even higher cost on the part of the borrower. The implicit price for advance sales of rice or fish, for example, were well below market prices. Although delays are not unusual, outright default seems to be rare, only possible in the event of death or permanent out-migration from the area.
Rice purchases
Health treatment
(percentage) Very poor Poor Marginal(negative) Marginal (positive) Well-off Rich

Source: Murshid 1998.
95.2 66.7 81.8 46.6 71.7 32.0
70.0 62.0 40.0 27.0 11.0
20.0 16.0 17.0 12.0 9.0
Given the high levels of indebtedness, it appears that access to credit is not entirely lacking. Rather the problem is one of excessively high costs in both formal and informal sectors. These high costs are directly related market fragmentation and to poverty and the associated risk of lending to the poor. Moreover, in terms of economic development prospects, the current rural finance system appears to favour the non-poor rather than the poor and very poor.
4.3 Demand for Rural Financial Services

Despite the widespread use of credit among the rural poor, studies undertaken by a number of NGOs suggest that the demand is far from being met. Estimates of untapped potential
188
demand for rural credit have ranged from $45 million to $130 million (Bousso et al. 1997; Prins 1996; NIS 1997), the differences depending on the method of calculation. These estimates suggest that the current aggregate demand for rural credit is in the range of $75-125 million. Since current supply from all sources is of the order of $25-33 million, a credit gap of $50-100 million would appear to be reasonable estimate. Irrespective of the accuracy of calculations, there is clearly a major credit gap. At present, rather little is being done to fill this gap. A new ADB initiative, for example, has offered a line of credit to RDB of around US$9-10 million from mid-2002 until the end of 2006. It is based on the assumption that only large institutions with portfolios outstanding of KHR 1 billion or more will borrow under this facility. These are the ones that must apply for a license; other institutions are too small and unsophisticated to represent attractive credit risks (NBC 2002).
of KHR 1000-2000, or more if members so choose. Although many of the NGOs have incorporated savings mobilization in their credit programs, none have been successful in mobilizing significant amounts. Most commonly, borrowers have been required to save in order to obtain and maintain access to credit. Frequently, because interest earnings were below the rate of inflation, these savings were an additional cost of borrowing. Savers, furthermore, could only gain access to their savings by taking a group-sanctioned loan, in effect borrowing back their savings. Savings mobilization is considered difficult in Cambodia because of the lack of public confidence in the financial sector. As of the end of June 2002, it is estimated that the micro-finance network was mobilizing only KHR 10.2 billion (US$ 2.6 million) from the rural population. Moreover, based on the small amounts of deposits and the frequency of withdrawals, the annual costs of administering such small volumes probably exceeded the savings balance. Group borrowers, with their very small financial transactions, are unlikely to be a cost-effective and attractive market for deposit mobilization. The fundamental elements needed for successful savings mobilization are completely lacking in rural areas. There is, effectively, no real savings service. However, a UNICEF survey found that 80% of respondents had savings prior to borrowing and there was a potential for increased savings mobilization if proper facilities were provided. The case of ACLEDA also provides an indication of that potential: between January 2001 and
4.4 Savings Mobilization

Savings mobilization is an important factor in achieving sustainability of credit markets for economic development and poverty reduction. As noted above, most MFIs need more diversification of funding sources and technical assistance in developing and implementing effective savings mobilization schemes. Most saving programs require members to make a deposit before getting loans. The interest rate on deposit ranges from 1% to 3% per month and savings are used to augment the fund for lending purposes. Required deposits are usually in the range
189
the end of June 2002, it increased its deposits collected from the public from nothing to KHR 16 billion, including KHR 5.2 billion (US$ 1.3 million) from individuals. On the other hand, many institutions do not practice compulsory savings because they consider the measure to be unpopular with clients. Some donors, such as GTZ have provided assistance to MFIs (e.g. Hattha Kaksekar) to mobilize voluntary savings, but these projects have so far met with limited success, mostly due to the publics reluctance to trust financial institutions.
deepening and in excess liquidity of the banking system. The financial system that does exist in rural areas is therefore based on NGOs with limited coverage and the informal sector. Micro-finance is based on the recognition that small business and employment generation, not least engaging women in more profitable economic activities, would be encouraged by the availability of credit. Despite substantial efforts and the best of intentions, the results, however, have been below expectations. It seems likely that the introduction of dedicated financial institutions in the rural areas would help promote a more efficient and stable rural financial market than the current system, based as it is on NGOs whose primary efforts are in projects with different objectives. However, expanded formal intermediation is unlikely to happen in the absence of a more conducive macro and sectoral environment which improves the profitability of rural activities, reduces risk for clients, improves information flows and reduces legal impediments. An improved economic environment is also likely to promote the forging of better links between formal and informal financial institutions and the diffusion of other financial instruments such as deposit accounts in riel or crop insurance. Financial liberalization in the absence of a good legal and institutional framework is indeed dangerous, as the recent history of Indonesia demonstrates (although banking crises have also occurred in Cambodia itself in the past decade). However, reforms aimed
5. CONCLUSION
Following financial liberalization, market determination of interest rates should result in modestly positive real interest rates which should, in theory, increase the resources available to the financial system. However, this effect has only applied in Cambodia to the dollarized urban economy. The perceived high cost and high risk of lending to the rural areas has resulted in most banks refusing to accept riel deposits, with the effect that formal financial intermediation is simply not available to the poor. Re-regulation of the sector, after the initial chaos in the wake of dismantling the monobank system, has succeeded in improving the performance and to some extent confidence of the public in the banking sector. Again, however, the impact has been restricted to the urban economy. Despite the high growth of GDP and the reduced number of banks after restructuring, this market fragmentation has resulted in inability to achieve financial
190
at upgrading institutions alone are unlikely to be very successful, particularly in terms of poverty reduction. A case in point is the question of land titling, the absence of which is, for many commentators, a major obstacle to provision of medium and long-term credit. At present, only 10 percent of rural families hold legal land titles. Nevertheless, titling is not a sufficient condition to improve access to credit: factors such as larger farm size, higher levels of human capital, and proximity to major consumer markets appear to be more important in determining access to formal credit. Other reforms, for example, in the legal environment of the financial system-including prudential supervision, registration, licensing, standardised accounting practices, transparency, a secured transaction framework and enforcement provisions - can reduce the risk and cost of rural financial intermediation and are essential to the process or restoring confidence in the banking system. But they are more likely to benefit the non-poor, who already participate in the formal credit system, in the short to medium term. The vast majority of the poor will continue to rely heavily on informal sector loans because of their greater flexibility. As noted in this chapter, the vulnerability of the poor to exogenous shocks is itself a major obstacle to the extension of formal financial services to them. Institutional reforms need to be accompanied by policies directly aimed at generating higher growth and employment in the rural economy. These include expenditure on education and health, which are already receiving priority from donors, and on rural infrastructure,
intensification of agricultural production and extension of economic services more generally to the countryside. Increasing rural incomes and productivity will both lower the marketing risk of financial institutions and their intermediation costs. It will provide an enabling environment in which the current fragmentation of rural markets, which in many ways contributes to the fragmentation, petty scale and high costs of financial markets, can be eliminated.
191
7
192
Privatization
1. INTRODUCTION
The issue of privatization of state enterprises is no longer considered a very important one in Cambodia. Public enterprises were nearly all privatized in the early 1990s. At present approximately 20 state enterprises remain and, although the intention is to proceed eventually, there appears to be little pressure. Most of the remaining enterprises are rubber plantations. The only major entity for which a privatization plan exists, but has yet to be implemented, is the Foreign Trade Bank (FTB). Land was effectively privatized by the 1989 Land Reform and, since then, there are no longer any elements of state or co-operative ownership of farm land or capital assets. Of more pressing concern at present is the privatization of common resources - the granting of forest, fishing and agricultural lands to concessionaires. This issue has provoked some controversy and, in some cases, reversals of policy have taken place. Past privatizations appear to have produced few adverse effects in terms of economic growth. The distributional effects are more controversial.
The macroeconomic of poverty reduction in Cambodia Privatization
Privatization of land has produced some conflicts. In 1989 the Council of Ministers issued Sub-Decree No. 3 on the Principles for Possession and Use by each household up to five hectares of land for cultivation. Today, an implicit assumption tends to be made that this reform solved once and for all the problem of access to and distribution of land. However, average agricultural land-size per household and average incidence of landlessness vary significantly. Over the last two decades, some people have been able to expand their land holdings, while others have had to sell. Land concentration is high (Chapter 1, Section 5.1). Lack of security of tenure is widely regarded as leading to low rates of investment by farmers. Much of the literature focuses on the need for an effective land titling system to overcome these conflicts, although there is not a lot of evidence of any correlation between lack of legal title and vulnerability of the poor to forced sale or arbitrary land seizure. Land titling is essentially a technical solution for a problem that is political-economic in nature. The issue remains a thorny one because traditional conceptions of land ownership are sometimes at odds with westernized legal concepts. Cambodian society is traditionally matrilocal which, unlike neighbouring Vietnam, gives Cambodian women some advantage in continued land tenure in the case of divorce or widowhood. The legal system, on the other hand, applies more modern concepts of ownership by the household, division of property on divorce and reversion of familial land to the husband in case he becomes a widower. Inter alia, these different conceptions of property may go some way to explaining the low rate of
title issue. Such questions are, strictly speaking, beyond the scope of this chapter. Privatization of state enterprises does appear to have reinforced the urban bias in economic development. As we have noted in previous chapters, there have been few incentives for firms to invest outside the more secure urban areas, which are also better supplied with infrastructure and services. In particular, we noted in Chapter 6, that privatization of the banking system has resulted in reduced access by farmers to sources of credit.
2. THE FOREIGN TRADE BANK

FTB currently operates as a state-owned commercial bank. Unlike the other major state-owned bank, the Rural Development Bank, it does not have a specific developmental objective. Given the impact of privatization of the banking system to date, however, there must be some concern about the likely impact of privatization of FTB on poverty reduction. Under state ownership, the FTB is one of the very few commercial banks that accept deposits in riel. In reality, most riel deposits are held by FTB, while privately owned banks have shown a distinct preference for dollar deposits. Since farmers, who are the vast majority of the poor in Cambodia, use riel for most of their transactions, their ability to access credit in the local currency is bound to be an important element in any poverty reduction strategy. In Chapter 6 we noted that during the years of the mono-banking system, NBC, through its 20 provincial branches, made small loans to thousands of farmers based on mutual guarantees. After 1990, with the privatization of banking and the discontinuance of NBCs
193
role as a commercial bank, farmers and entrepreneurs with small and medium businesses were deprived of access to formal financial services. At present, the RDB (state-owned), ACLEDA Bank (operated by an NGO), Specialized Peng Heng SME Bank and the Cambodia Agriculture and Industrial Specialized Bank, have branches in 7 provinces, while 48% of the population lives in provinces with no bank branch at all. Moreover, these banks are concentrated in provincial capitals, which are difficult for village people to reach, and they offer only limited services. While the FTB is not listed among those banks offering services to the rural population, the newly developed banks have yet to contribute very much to financial deepening. As a state-owned bank that accepts riel deposits, FTB has the opportunity to contribute to the development of rural finance by on-lending to those banks and NGOs that do service the rural sector. Early privatization of FTB, on the other hand, is likely to inhibit the development of rural banking. We conclude that privatization of FTB needs to await more positive development of the rural economy and rural finance.
largely due to the old age of the trees and a lack of investment in replacement planting. Most of the rubber plantations were under state ownership before the 1990s and many have remained so until today. According to our discussions in Cambodia, the reason for the relative lack of privatization in this sector is the age of the currently producing trees. There was a hiatus in investment during the war and Khmer Rouge years and a lack of capital for reinvestment during the central planning period. After two decades of neglect, the rubber plantations require extensive new planting to be able to raise productivity and become profitable operations. Since rubber trees take six years to become productive, this requirement makes them an unattractive proposition for private investors. The government has attempted to expand rubber planting by encouraging small private holdings, but lack of capital among small farmer households similarly makes the proposition an unattractive one. Investment capital is not, however, the only problem facing the rubber industry. World market prices for natural rubber have tended to remain rather high since the two oil shocks of the 1970s, but due to the large devaluations in the currencies of major producers (Thailand, Indonesia and Malaysia) in 1997, the price in US dollars fell steeply (Burger and Smit 2001: 10-11). In December 2001, for example, the Bangkok prices of sheet rubber and latex had fallen to 10 year lows (Bangkok Post, 17 December 2001). While such falls should increase exports, the major producers did indeed run down their stocks, the falls coincided with the onset of recession in major consuming countries
3. RUBBER PLANTATIONS
Rubber was formerly a major Cambodian export, but its value stagnated in the 1990s and its share in total legal exports has consequently declined from 5% of the total to 1.7% in 2001. Illegal exports have recently expanded (Tables 5.2; 5.3). The current annual recorded level of production is 45,000 tonnes, only 25% of its 1960s level,
194
and squeezed the profits of producers, discouraging investment in expansion of production. There would therefore seem little prospect for successful privatization of Cambodias rubber plantations in the near future. Burger and Smit forecast large increases in rubber consumption in the Asian region over the next two decades, but relative stagnation in other markets (2001: 13). Since rubber production is a labour-intensive activity, Cambodia is likely to have a competitive advantage over its higher-wage neighbours, especially Malaysia and Thailand, but this advantage is premised on sufficient investment in new trees. To take market share from competing countries, there would seem, therefore, to be some case for continued state involvement in the rubber industry over the coming period. Assistance to the state plantations, as well as private producers, combined with specific export targets as the new trees come on stream would seem to be a sensible way, not only to kick-start the recovery of an industry with considerable growth potential, but for creating poverty reducing employment in the rural areas.
resources increases the real inequality of land distribution. Moreover, exploitation of natural forestry and fishery resources, which was particularly severe during the 1990s, represents a significant loss of income for many rural households. According to Ministry for Agriculture, Forestry and Fisheries (MAFF) data for 1999, the total land under forest concessions was 6.4 m. hectares (about 2/3 of the total forest area), a figure which is widely believed not to have increased since then (Chan Sophal et al. 2001). The same data suggest, however, that only 2.2 m. ha. were actually under exploitation, a figure reduced to 2.0 m by 2001 according to the MAFF sources. Indeed the MAFF sources cited by Chan Sophal et al. reduce the total area under concession in 2001 to 4.2 m ha. Individual concessions are generally large, exceeding 100,000 ha. and usually have a term of 30 years. However, individual contracts are not accessible to the public. Agricultural concessions were less extensive in 1999, at 662,496 ha, given out to 46 companies.35 They are also concentrated in forested areas, rather than in the main rice-growing areas and the majority were devoted to perennial crops, including cashew, rubber, palm oil, coffee and fruit, with a term of 70 years. While the average size of these concessions was much smaller than the forestry concessions, at 14,095 ha., this average is boosted tremendously by the largest concession, of 315,000 granted to Pheapimex in Kompong Chhnang and Pursat provinces.36 By 2001 the number of concessions had fallen to 39, but the area
4. PRIVATIZATION OF LAND AND COMMON RESOURCES

Land traditionally controlled by the local community has, in a number of cases, been converted to some other use, e.g. logging and fishing concessions, or plantations (Sik Boreak 2000). The de facto privatization of common property resources, such as common grazing grounds, ponds and water
35 In addition there is an unknown amount of land given out as concessions by the military. 36 Excluding the Pheapimex concession, the average size is 7554 ha. The macroeconomic of poverty reduction in Cambodia Privatization
195
had increased to 705,394 ha. (Chan Sophal et al. 2001). Malaysians and Chinese, as well as Cambodians, are the most prominent holders of agricultural concessions. As in the case of forestry concessions, a large proportion of these agricultural concessions are not yet in operation. The most common reason is that many parties claim possession or ownership of the land under consideration. It is often the case that the companies are unaware of local land tenure systems. In a few cases local officials have engaged in land-grabbing. There have also been instances of companies holding onto land for speculative purposes. Fishing lots cover 9,634 km2 of Cambodias of freshwater surface, including the MekongSap river system and the Tonle Sap lake. Concessions over these lots are generally short-term - from one to five years - and about 1 m. ha are given out annually. Large-scale operators operate under two year concessions, medium-scale operators in both open access areas and demarcated lots under licence. Small-scale (household) operators have access to open areas only. Over the 1990s, riparian communities have faced diminishing access to their fishing areas. Additionally, the spread of agriculture around the Tonle Sap and diminution of forested land is thought to be affecting the long-term viability of the fisheries due to siltation and chemical run-off. In mid-2000 the government announced the cancellation of a number of fishing concessions, effective from June 2001 (Chan Sophal et al. 2001). The appropriation of private land for concession use is thus a serious problem for the livelihoods of the rural population,
reducing their access to forest and aquatic resources, creating insecurity of land tenure and restricting their ability to resolve the problem of underemployment by opening up new land for agricultural purposes. While the concession system in is intended to open up commercial opportunities and generate government revenue, it is not clear that it is currently performing this function. As noted above, many forest and agricultural concessions are not operating, which may be good a good thing from the point of view of conservation, but serves to deprive poor communities of the use of resources they have traditionally relied upon to supplement their incomes. The short-term nature of fishing concessions may also encourage over-fishing and deplete the resource over the long term. Currently only a small percentage of the population hold title for their land, including both agricultural and residential land holdings (Sik Boreak 2000). Land entitlement differs across districts, and women headed households have gained less land as distribution is dependent on the number of able-bodied adults in the household. In more densely populated areas, land entitlements have been insufficient. Given the constraints on land use as collateral discussed in Chapter 6, the current insecurity of tenure has at least partly contributed to the lack of financial deepening and inability of farmers to obtain investment funds. A land law is currently being prepared to address issues of land redistribution, land management and land administration. The law is intended to establish systematic land registration; a legal framework to enforce property rights; provincial, municipal and
196
national master plans and zoning, as well as reform the taxation system and develop a housing policy. The systematic land registration scheme has so far resulted in 15% of land being registered, (RGC 2002b; Chan Sophal et al. 1999). It is estimated that with the current level of resources set aside for this scheme, the time taken to complete land registration will be approximately 15 years. Indeed the issue of certificates of possession appears to have slowed considerably in recent years (Chan Sophal et al. 2001: 30) and is severely skewed across provinces. The lack of progress would appear, from our discussions, to reflect considerable controversy over land claims. The CSES survey of 1997 revealed that over 80% of the rural population considered themselves to possess title to their land! Moreover, where titles have not been issued, this has not prevented the development of a market in land. Increasing landlessness in some areas is created less by arbitrary seizure or the granting of concessions than by forced sales due to the need to pay for health services or food (Chan Sophal et al 2001: 45). The process of providing security of tenure would thus seem to be strongly related to issues of political economy rather than simply technical questions. Providing individual households with land titles is unlikely to resolve the competing claims. Nevertheless, land distribution and security of tenure have important implications for rural poverty alleviation, and for increasing Cambodias potential in rice exports and agricultural diversification. The present policy provides little incentive for commercial scale production but reinforces subsistence farming among the majority of the population.
Improvements in rural infrastructure and new job creation in the countryside are therefore unlikely to be forthcoming from private investors. Current expenditure patterns on economic services do not provide encouraging indications on this front (see Chapter 4). This situation is neither consistent with the strategy of rural development for pro-poor growth, nor with the re-distribution of current resources to share the benefits of growth recently experienced. Potentials for growth in the agricultural sector are not being effectively pursued and hence current policy threatens to continue the exclusion of the majority of the Cambodian population from benefiting from economic growth in the country.
5. CONCLUSION
On the whole it would seem that privatization has tended to encourage economic growth in Cambodia, but it has not served the objective of poverty alleviation. Privatization in the banking sector has, if anything, reduced the access of the rural population to the formal financial sector while it has enhanced growth in the urban areas. For other sectors there is less information available, but the record of a lagging rural economy would suggest much the same pattern. Rubber plantations, which have yet to be privatized, offer an opportunity for increased investment and employment in the rural economy as well as a means of improving export diversification, but the private sector, with good reason, has shown a reluctance to take on the task. In other areas, the privatization of forest and water resources has proved controversial due to competing claims and has largely failed to provide the intended sources of commercial development and government
197
revenue. At the same time, it has restricted access of the poor to resources upon which they have previously relied. Privatization of remaining public assets does not appear to be high on the governments agenda at the present time and there would seem to be good reasons to reconsider the policy with regard to these assets. Public ownership of some assets creates opportunities to pursue developmental objectives, provided that,in the absence of market discipline, government is also able to impose discipline on the firms. The enterprises might eventually be privatized, but their long-term profitability must also be assured by complementary macro-fiscal policies which are aimed at growing the rural economy as a whole.
Finally, the case of forestry, agricultural and fishery concessions involves political economy issues which have no simple technical solution (such as land titling). The recent resumption of some concession areas by the state is a positive step forward in this regard, as it increases the access of the poor to resources. While the granting of concessions could indeed be a positive step towards development of commercial activities in the countryside, policy needs to proceed with far greater caution in future. Particularly if political stability is to be preserved, it needs to take into account local land tenure systems, which may conflict with conceptions of private property being promoted by a number of donor agencies.
198
List of References
Acharya, S. 2002. Labour Migration in the Transitional Economies of South-East Asia, Cambodia Development Review 6(1): 12-15. ADB 2001. Education Sector Development Programme Aid Memoire and Memorandum of Understanding. Phnom Penh and Manila: Asian Development Bank. ADB and Kingdom of Cambodia 2001. Financial Sector Blueprint for 2001-2010. Phnom Penh, August. Amsden, Alice 1989. Asias Next Giant. New York and London: Oxford University Press. Atkinson A. B. and Stiglitz, J. 1987. Lectures in Public Economics. London and New York: McGraw-Hill. Bousso, P.P., Daubert, N., Gaither, M., Parent and C. Ziegle, 1997. L impact micro-economique du credit rural au Cambodge. Phnom Penh: Groupe de Recherche et dEchanges Technologiques. Burger, Kees and Hidde P. Smit 2001. Economic growth and demand for commodities - natural rubber and cocoa Paper presented at International Conference on the Future of Perennial Crops, Yamoussoukro, Cote dIvoire, 5-9 November (at www.feweb.vu.nl/esi/bin/pdf/667.pdf ). Calvert, B. 2002. Unions Want Less Strikes, More Talks, The Cambodia Daily, DATE ??: 1 CDRI 2001. Annual Economic Review. Phnom Penh: Cambodia Development Resource Institute, August.
200
The macroeconomic of poverty reduction in Cambodia List of References
CDRI 2002a. Annual Economic Review. Phnom Penh: Cambodia Development Resource Institute, August. CDRI and CICP 1999. The Asian Financial Crisis and Southeast Asian Transitional Economies: The Cambodian Perspective , Regional Conference on The Impact of the Asian Financial Crisis on the Southeast Asian Transitional Economies, Phnom Penh. CDRI 2002b. Cambodia Development Review, issue 3, December 1997 and issue 3, July-September 2002; Phnom Penh. Chan Sophal, Martin Godfrey, Toshiyasu Kato, Long Vou Piseth, Nina Orlova, Per Ronnas & Tia Savora 1999. Cambodia: The Challenge of Productive Employment Creation. Phnom Penh: CDRI, Working Paper 8. Chan Sophal, Tep Saravay and Sarthi Acharya 2001. Land Tenure in Cambodia: a data update. Phnom Penh: CDRI, Working Paper 19. Davis, G.A. 1995. Learnin g to Live with the Dutch Disease: Evidence from the Mineral Economies, World Development, vol. 23, no. 10. D-MTEF 2002. Draft Medium Term Expenditure Framework. Phnom Penh: Ministry of Economy and Finance. Dollar, David and Kraay, Aart 2000. Growth is Good for the Poor. Washington DC: World Bank. Godfrey, M., Sophal, C., Kato, T., Piseth, L.V., Dorina, P., Saravy, T., Savora, T. and Sovannarith, S. 2000. Technical Assistance and Capacity Development in an Aid-dependent Economy: the Experience of Cambodia. Phnom Penh: CDRI, Working Paper 15, August.
Godfrey, Martin, So Sovannarith, Tep Savery, Pon Dorina, Claude Katz, Sarthi Acharya, Sisowath D.Chanto and Hing Thoraxy 2001. A Study of the Cambodian Labour Market: Reference to Poverty Reduction, Growth and Adjustment to Crisis. Phnom Penh: CDRI, Working Paper 18. Gorman, Siobhan, with Pon Dorina & Sok Kheng 1999. Gender and Development in Cambodia: An Overview. Phnom Penh: CDRI, Working Paper 10. ILO 2002. The Upstream Project. Phnom Penh: International Labour Organization IMF 2000. Cambodia: Statistical Appendix. Washington D.C.: International Monetary Fund. IMF 2001. Cambodia: Letter of Intent, Memorandum of Economic and Financial Policies and Technical Memorandum of Understanding. Washington D.C.: International Monetary Fund. IMF 2002. Cambodia: Fourth Review Under the Poverty Reduction and Growth Facility. Phnom Penh: International Monetary Fund, January. IMF 2002a. Cambodia: Statistical Appendix. Washington D.C.: International Monetary Fund. I-PRSP 2001. Cambodia: Interim Poverty Reduction Strategy Paper. Phnom Penh: Royal Government of Cambodia. Kimsong, K. and B. Calvert 2002. Rising Costs Give Investors Smugglers Blues, The Cambodia Daily. Phnom Penh: 1, 18.
201
LIDEE Khmer (The League of Khmer Students from Abroad) 2001. Assessment Survey of Demobilized Soldiers Needs and Vulnerability. Phnom Penh: Royal Goverment of Cambodia. Lbker, Malte, Graham Smith and John Weeks 2002. Growth and the poor: a comment on Dollar and Kraay, Journal of International Development, (14) 555-571. Manon. Jayant 1998. Cambodia: ASEAN Membership and macroeconomic Policy Issues. Technical assistance to Cambodia and Lao PDR to Prepare for Membership in ASEAN, ADB. Mario de Zamarczy and Sopanha Sa 2002. Macroeconomic Adjustment in a highly dollarized Economy: The Case of Cambodia. IMF Working Paper WP/02/92. MEF 2002. Monthly Bulletin of Statistics. Phnom Penh: Ministry of Economy and Finance, no. 74, February. MEF. Monthly Bulletin of Statistics, Various issues, Phnom Penh: Ministry of Economy and Finance. MOP 1999. Cambodia Poverty Assessment. Phnom Penh: Ministry of Planning. Murshid K.A.S. 1998. Food Security in an Asian Transitional Economy: The Cambodian Experience . Phnom Penh: CDRI, Working Paper 6. Naron, Hang Chuon. 2002. The Effects of WTO Membership on the SMEs and Garment Sector of Cambodia. Phnom Penh: Ministry of Economy and Finance.
NBC 2002. Micro-finance Activity in Cambodia. Phnom Penh: NBC Department of Supervision. NBC. Monthly Economic and Statistic Review. Various issues. Phnom Penh: National Bank of Cambodia. NIS 1997. Socio-Economic Survey of Cambodia 1996: Summary Results, 2 vols. Phnom Penh: Ministry of Planning. NIS 1998. Report on the Socio-economic Survey of Cambodia 1997 . Phnom Penh: Ministry of Planning. NIS 2001. Organization and Administration of The General Population Census of Cambodia, 1998. Phnom Penh: Ministry of Planning. NIS 2002a. Cambodia Statistical Yearbook 2001. Phnom Penh: National Institute of Statistics, Ministry of Planning. NIS 2002b. Kingdom of Cambodia, National Accounts of Cambodia 1993-2001. Phnom Penh, National Institute of Statistics, Ministry of Planning, May. NIS 2003. Cambodia Statistical Yearbook 2003. Phnom Penh: National Institute of Statistics, Ministry of Planning. Prescott N. and Pradhan M. 1997. A Poverty Profile of Cambodia. Prins N. 1996. Rural Credit Review, Cambodia. Manila: Asian Development Bank. RGC 2000. Report on the Cambodia Socioeconomic Survey (CSES) 1999. Phnom Penh: Ministry of Planning.
202
RGC 2001a. Development Co-operation Report. Phnom Penh: Council for the Development of Cambodia, Royal Government of Cambodia. RGC 2001b. Financial Sector Blueprint for 2001-2010. Phnom Penh: National Bank of Cambodia. RGC 2002a. Second Five Year Socio-Economic Development Plan (SEDP II) 2001-2005 . Phnom Penh: Ministry of Planning. RGC 2002b. Socio-economic Development Priorities and Official Development Assistance Needs. Phnom Penh: Council for the Development of Cambodia, May. RGC-Ministry of Commerce 2002. Cambodia Integration and Competitiveness Study . Phnom Penh. Roy, Rathin 1994. The Politics of Fiscal Policy. Cambridge UK: University of Cambridge. Sambath, T. and D. Kihara 2002. Demobilizing Effort Delayed by Govt. The Cambodia Daily, Phnom Penh: date 1-2 Sen, A. 1999. Development as Freedom. New York: Alfred A. Knopf. Sik Boreak 2000. Land Ownership, Sales and Concentration in Cambodia. Phnom Penh: Cambodia Development Resource Institute. Sok Hach, Chea Huot and Sik Boreak 2001. Cambodias Annual Economic Review 2001. Phnom Penh: Cambodia Development Resource Institute Toms J. T. Balio, Adam Bennett, Eduardo Borensztein 1999. Monetary Policy in Dollarized Economies. Washington, D.C.: IMF.
Toshiyasu Kato 1998. Cambodia: Learning from the Asian Economic Crisis. Phnom Penh: CDRI. UNDP 1997. Cambodia Human Development Report 1997. Phnom Penh: Ministry of Planning. UNDP 1999. Cambodia Human Development Report 1999: Village Economy and Development. Phnom Penh: Ministry of Planning. UNDP 2000. Cambodia Human Development Report 2000: Children and Employment . Phnom Penh: Ministry of Planning. UNDP 2001. Cambodia Human Development Report 2001: Societal Aspects of the HIV/AIDS Epidemic in Cambodia, Progress Report . Phnom Penh: Ministry of Planning. Van Sou Ieng 2002. Export Process Working Group. 7th Royal Government and Private Sector Forum, Council for the Development of Cambodia, Phnom Penh. Wade, Robert 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton NJ: Princeton University Press. White, H. and Wignaraja, G. 1992. Exchange Rates, Trade Liberalisation and Aid: the Sri Lankan Experience, World Development, vol.20, no. 10. World Bank 1997. World Development Report. The World Bank: Washington D.C. World Bank 1999. Cambodia: Public Expenditure Review . Washington D.C.: The World Bank.
203
Persons Consulted
Mr Russell Peterson, Director, NGO Forum Ms Mia Hyun, Senior Program Officer, Oxfam America Mr Daniel Agiros, Governance Adviser, DFID (UK) Ms Ros Sopheap, Director, Gender Network Ms Eva Mysliwec, Director, CDRI Mr Ita Praing, Secretary of State, Ministry of Industry, Mines and Energy Mr Long Piseth, Assistant, Program/Operation Analysis, ADB Mr Ouch Chamroesn, Poverty Analysis, ADB Mr Robert Hagemann, Resident Representative, IMF Mr Hank Bekedam, Director, WHO Mr Joe Martin, WHO Budget Advisor to Ministry of Health Mr Ben Diokno, ADB Fiscal Adviser, MEF Mr Tony Felts, Senior Program Officer, European Union Mr Roger Tan, Secretary General, GMAC Ms Dominique McAdams, Resident Representative, UNDP
204
The macroeconomic of poverty reduction in Cambodia Persons Consulted
Mr Ladislaus Byenkya, Deputy Res.Rep. (Programs), UNDP HE Sok Siphana, Secretary of State, Ministry of Commerce Mr Bill Ledrew, IMF adviser, Customs Department Mr Praveen Agrawal, Deputy Country Director WFP Mr Sik Boreak, VAM officer, WFP Dr Sarthi Acharya, Research Director, CDRI Dr Sok Hach, Senior Economist, CDRI Mr Aldo dellAriccia, Charg dAffaires of the European Union Mr Marc Paoletti, IMF, Budget Adviser, State Treasury HE Kim Saysamalen, Secretary of State, Ministry of Planning Mr Urooj Malik, ADB Country Office Director Mr Mohammed Elmansi, UNDP Fiscal Decentralisation Adviser Mr Hong Lee, Adviser, Council of Ministers Mr Kong Sophy, Director, Governance Project, CAR Mr Sar Chanty, director, Human Resource Development, CAR Mr Norman Mailhot, Canadian Ambassador Ms Kim Sathavy, Legal Adviser to Deputy Prime Minister Sar Kheng.
Mr Brian Dawe, IMF Adviser, Tax Department Dr Sam Ghanty, Institute of Economics and Finance. Messrs Robert Taliercio, Geoff Dixon and Ms Chorching Goh, World Bank Public Expenditure Review Mission Dr Farid Siddiqui, UNDP Senior Adviser, CDC Mr Zia Abbas, IMF adviser, National Institute of Statistics Dr Hang Chuon Naron, Deputy Secretary General, Ministry of Finance Mr Steven Schonberger, Senior Program Officer, World Bank Mr Faruque Sarkar, Director, Canadian Co-operation Office Mr Ty Yao, Secretary of State, Ministry of Land Management, Construction and Urban Planning Ms Elaine McKay, Gender Adviser, MWVA Ms Nhean Sochetra, Project Director, MWVA
The macroeconomic of poverty reduction in Cambodia Persons Consulted
205
Appendix
Project Level Interventions
There are a large number of civil society organizations, NGOs and multilateral and bilateral donor agencies operating throughout Cambodia. Such schemes include humanitarian assistance in organizing food for work programmes, encouraging SME activities through micro credit schemes, or through the organization of public works schemes to build infrastructure such as national highways. Below is a brief overview of a few of the programmes that engage local labour, and is by no means an exhaustive list. The World Food Programme (WFP) aims to provide emergency humanitarian relief and has been in operation in Cambodia since 1979. The programmes have a primary objective of ensuring food security and include; food-for-work (FFW) activities which began in 1991, School feeding pilots in 1999, supplementary feeding project from 2000 and the expansion of Protracted Relief and Recovery Operation (PRRO) which began in 1999 and which focuses on integrated development and support to the social sector. In 2002 the PRRO expanded to include maternal child health care project. The WFP is also involved in the demobilization of the Royal Cambodian Armed Forces, contributing to the take-home rations of the 30,000 ex-soldiers. A vulnerability analysis and mapping technique (VAM) has also been developed and applied to target the food-insecure and vulnerable areas. WFP operations are located in much of the area between the NW and SE of the country and works in partnerships with a variety of bilateral and multilateral donors, NGOs and numerous Ministries.
206
The macroeconomic of poverty reduction in Cambodia Appendix Project Level Interventions
The FFW activities in 1999 absorbed about 82% of the food distributed and lead to the rehabilitation or construction of rural roads, schools and health centres, as well as clearing land for resettlement. FFW involves the implementation of labour-intensive community-based projects in areas where there is chronic or transient hunger. Food aid offers local employment opportunities, for both men and women, as well as creating village-level physical assets. Daily income transfers of those participating in the FFW programmes is 4,000-4,5000 riels/ day and the rations are based on nutritional needs and designed to ensure basic food intake in terms of energy, fat and protein. The rations consist of rice, canned fish, vitamin-afortified vegetable oil and iodised salt. WFP also supports credit schemes such as rice banks, through other partners. The types of community projects implemented include road improvement and maintenance essential to market integration: water resource development (dykes, dams, cannels) horticulture and fishponds and sanitation projects. There have been some criticisms of FFW programmes, since these programmes operate in areas that are in dire need, and the local community has to work for the food rather than receiving humanitarian assistance. Since 1992 the International Labour Organization (ILO) has been promoting labour based appropriate technology (LBAT) to combine both the need for employment generation and the improvement and maintenance of essential rural infrastructure in Cambodia. The labour-based rural infrastructure works programme, known as the Upstream Project became operational in July 1998 with bilateral assistance, see (ILO 2002), working closely with the Ministry of Rural Development.
The Upstream Project includes a variety of operations, including the training of private sector small-scale subcontractors to carry out rural infrastructure projects, the designing and implementation of engineering programmes to build in-country capacity in applying LBAT approaches, building capacity within the MRD to maximize the pro-poor impact through infrastructure investment and demonstrating LBAT practices in Siem Reap through maintenance and construction works. This latter project involved approximately 575,000 work-days. The project also works in partnership with the WB and ADB towards an integrated rural accessibility planning (IRAP) approach. The IRAP programme aims to mobilize the abundance of unskilled labour in rural areas. Prior to 1998, the ILO operated on a 50% food, 50% cash basis, though now is wholly cash based. The number of small-scale subcontractors has grown to 33 in 2002. Sub-contractors have to bid for the projects, and are given assistance with financial planning. Till now, 120km of road has been built in Siem reap. Labour for projects are advertised within a 3km radius of the road, and selected through a lottery process to ensure random selection for workers, half male and half female. Due to the nature of the work, labourers need to be 18 years and over and require Ids to confirm their age. Salaries are set by the Ministry of Rural Development and for unskilled workers are approximately 4,000 riel/$1 per day in most rural areas. This rate varies across the country depending on the nature of the local labour market. In Siem Riep, and around the airport, labour costs are $5/6 per day and workers do not work at night, while in other Provinces it may be as low as 60-75 cents. Skilled labour can charge $1.5/day and these tend to be the machine operators who are mainly men. Projects tend to be
207
approximately 2 months in length, with a second lottery drawn if more labour is needed. Projects work around the harvesting period, which falls between late December to January. The Food and Agricultural Organization (FAO) have been operating in Cambodia since 1992, with normative activities since 1945 involving dialogue with stakeholders and improvements in the management of natural resources and improving food security at the farm level. FAO is operating in several provinces including Battambang, Kandal, Takeo, Kampot, Siem Repa. Food security programmes are also operating in areas where recently demobilized soldiers reside. The European Union (EU) is the largest European contributor in Cambodia and has provided assistance to the RGC since 1994 through the European Rehabilitation Programme for Cambodia (PERC). Half of PERC funding of 150 million euros is used for the Rehabilitation and Support Programme for Cambodias Agricultural Sector (PRASAC). The main objectives of PRASC include; improving the water supply for domestic use through building water ponds and wells and improving agricultural production though the building or rehabilitation of 140 km of irrigation canals in seven separate areas, covering 20,000 hectares. Training and maintenance of systems was also provided, including agricultural support using improved inputs, agricultural extension, village development committees, access to rural credit leading to the creation of 700 village banks, providing support to 6,000 agricultural micro-enterprises, and the training of local government staff particularly in areas of agricultural and rural development.
Rural banking has been introduced in 24 districts covering 800 villages. In 1998, over 45,000 loans granted, valuing 3,200,000 (euros), average repayment rate is 90%. Since January 1995-December 2003, it is estimated that over 15,000 families will have benefited from the programme. As well as multilateral donor, bilateral donors also provide extensive assistance. DFID, SIDA, CIDA, AUSAID are just few funding organizations that have sectoral interests. The Asian Development Bank (ADB) and the Japanese International Cooperation Agency (JICA) are significant contributors to overall assistance in Cambodia. Many of their projects relate to infrastructure building and maintenance, in the form of schools, medical facilities, and roads. Much of these activities are labour intensive, and so do provide opportunities for employment. Many of the Japanese funded projects have a high content of products that are brought from Japan, and include Japanese labour. Although infrastructure is created, the extent of skills transfer and capacity building is limited. ADB assistance has also been in the area of rural credit. This has resulted in the establishment of a selected number of rural banks in particular areas. However there are still constraints in the provision of rural credit, due to the limited legal structure and bank lending restrictions. The result is rates of interest that reflect the high risk of credit lending in agricultural based businesses and so does not provide a viable option for many potential creditors.
208
UNDP is the United Nations' development network that spans more than 130 countries, connecting knowledge, experience, and resources from one nation to another. Asia-Pacific Regional Programme on the Macroeconomics of Poverty Reduction Programme is a programme of UNDP Regional Bureau for Asia and the Pacific. It addresses poverty reduction within national economic frameworks and presents practical policy options to foster more pro-poor stabilisation, economic restructuring and growth in the region. Influenced by a precedent-setting case study conducted in Mongolia in 2001, the programme currently covers twelve countries: Bangladesh, Cambodia, China, India, Indonesia, Iran, Mongolia, Myanmar, Nepal, Samoa, Sri Lanka and Vet Nam. In order to promote macroeconomic policies that have greatest impact on poverty by achieving equity and growth, the programme has adopted three steps: policy oriented research, capacity building for regional learning and policy development, and policy dissemination, advocacy and advisory services. For further information visit the programme website: www.asiapropoor.net This publication is a part of the series that disseminate the findings of work in progress of the Programme to encourage exchange of ideas on pro-poor economic policies. One objective of the series is to get the findings out quickly to the policymakers and others, even if the studies are not fully polished. The studies carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in these studies are entirely those of the authors, and do not necessarily represent the views of the UNDP.
Asia-Pacific Regional Programme on the Macroeconomics of Poverty Reduction UNDP, UN House, GPO Box No. 107 Kathmandu Phone: 977-1-5542682/5542817 Fax: 977-1-5542683 www.asiapropoor.net

Macro - Poverty and Inequality in Cambodia

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Macro - Poverty and Inequality in Cambodia

Hochgeladen von

Copyright:

Verfügbare Formate

Copyright 2004, Asia-Pacific Regional Programme on The Macroeconomics of Poverty Reduction, United Nations Development Programme.

THE MACROECONOMICS OF POVERTY REDUCTION

1. THE DIMENSIONS OF THE PROBLEM: AN OVERVIEW OF POVERTY AND INEQUALITY IN CAMBODIA...........26

2. THE MACROECONOMIC FRAMEWORK..................................................54

The macroeconomic of poverty reduction in Cambodia Contents

3. MONETARY AND EXCHANGE RATE POLICY...........................................78

The macroeconomic of poverty reduction in Cambodia Contents

The macroeconomic of poverty reduction in Cambodia Contents

The macroeconomic of poverty reduction in Cambodia Contents

The macroeconomic of poverty reduction in Cambodia Contents

The macroeconomic of poverty reduction in Cambodia Abbreviation

The macroeconomic of poverty reduction in Cambodia Abbreviation

The macroeconomic of poverty reduction in Cambodia Special message

The macroeconomic of poverty reduction in Cambodia Preface

The macroeconomic of poverty reduction in Cambodia Preface

Summary of the Report

The macroeconomic of poverty reduction in Cambodia Summary of the report

2. The Macroeconomic Framework

The macroeconomic of poverty reduction in Cambodia Summary of the report

3. Monetary and Exchange Rate Policy

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

The macroeconomic of poverty reduction in Cambodia Summary of the report

Dimension of the Problem:

2. CHANGES IN INCIDENCE AND DEPTH OF INCOME POVERTY

2 See WFP 2001for a broader discussion of CSES 1999.

CSES 1999 (round 1) 51.1 64.4 19.4 57.3 70.0 47.7

CSES 1999 (round 2) 35.9 9.7 25.2 40.1 11.5

39.0 11.4 36.6 43.1 20.0

3.1 Basic Health Indicators

3. LEVEL AND TRENDS IN KEYS HUMAN POVERTY INDICATORS

Neonatal mortality 37.3 40.5 44.2 48.1 53.7

Post-neonatal mortality 57.8 50.4 34.5 36.4 75.0

Infant mortality 95.0 91.0 78.8 84.6 128.7

Child mortality 32.5 31.3 39.1 53.3 129.7

Under-five mortality 124.4 119.4 114.7 133.4 241.7

3.2 Education Indicators

37.5 40.0 14.7 4.8 0.3 0.5 0.9 1.3 100

2 64.7 6.0 1.2 0

3 70.5 13.0 5.0 0.1

4 73.2 17.1 6.9 0.6

119.5 40.9 15.2 2.2

106.0 27.1 8.2 0.4

3.3 Social Exclusion

3.4 Vulnerability to External Shocks

4. RATES AND CHANGES IN INEQUALITY 4.1 The Distribution of Consumption

Consumption share of the poorest 10%

National Phnom Penh Other Urban Rural

Consumption share of the richest 10%

National Phnom Penh Other Urban Rural

Sources: MOP/UNDP 2002.

4.2 Functional Distribution

Source: RGC-Ministry of Commerce 2002.

F 2.1 0.0 5.4 17.2 0.0 0.6 11.5 14.3 51.1

3.2 0.0 11.8 7.3 0.0 0.5 10.8 15.2 48.9

1999 Total % 41 5 7 44 2 2 100 Male % 44 1 2 49 1 1 100 Female % 16 4 3 73 2 2 100 Total % 32 2 3 59 1 2 100

5. LEVELS AND CHANGES IN THE DISTRIBUTION OF PRODUCTIVE ASSETS