- Kalpana Kochhar, Erik Offerdal, Louis Dicks-Mireaux, Mauro Mecagni, Jianping Zhou, Balázs Horváth, David Goldsbrough, and Sharmini Coorey
- Published Date:
- August 1996
There is a reasonable degree of consensus on the basic paradigm underlying structural adjustment: macroeconomic stability along with market-based and outward-oriented economic policies are conducive to sustained faster growth—even if such policies are not always sufficient. The experience of the high-growth economies of East Asia also suggests that getting certain fundamental policies right builds the foundation for sustained growth (Box 7). However, the problems of transition to such a growth-friendly policy environment are often not clear-cut and sometimes require difficult trade-offs between intermediate objectives. These choices can be especially complex when, as is the case for most countries undertaking IMF-supported adjustment programs, the initial position is one of severe external financing constraints, large macroeconomic imbalances, and deeply entrenched structural weaknesses. What lessons does experience suggest about managing the transition? Without attempting a comprehensive survey or setting out in detail estimated quantitative relationships, this appendix summarizes some central messages to emerge from some recent studies (denoted by abbreviations in Box 8).
Do Adjustment Policies Foster Growth?
It is extremely difficult to distinguish the effects of adjustment policies from the many other influences on growth in the short term. Indeed, exogenous factors such as the terms of trade are likely to influence not just outcomes but the policies themselves, making it especially hard to disentangle cause and effect. Moreover, the depth and complexity of adjustment programs can vary widely—from short-term macroeconomic stabilization to efforts to address deep-rooted macroeconomic and structural imbalances—so that the lack of a simple association between “adjustment” and growth is hardly surprising. With these caveats, most studies that have examined the effects of adjustment supported by the IMF or the World Bank typically suggest that, on average, real GDP growth was broadly unchanged (CR) or strengthened, albeit only moderately (RAL-2, RAL-3, and Corbo), in the years immediately following the adoption of adjustment programs.105
Experience also suggests that delaying core reforms until macroeconomic imbalances and price distortions are well entrenched is likely to be harmful to growth. In particular, there is evidence that growth falls substantially during periods of high inflation and recovers after inflation is stabilized.106 However, macroeconomic adjustment is not always sufficient to put countries onto a sustained faster growth path. Especially for the low-income countries, long-run development problems associated with weak social and economic infrastructure and institutions still need to be tackled; resolving these problems will require considerable time (RAL-2, RAL-3, ESAF). Nonetheless, the record of adjusting countries in sub-Saharan Africa suggests that, even in countries with deep-rooted development constraints, progress toward macroeconomic stability helps to promote growth, both through its effects on productivity and on capital accumulation (Hadjimichael and Africa).
In only a few countries (most notably in Asia, such as Korea and Thailand) were episodes of significant adjustment in macroeconomic and structural policies followed by a swift transition to a new growth path with only a short recession. These countries typically had smaller initial macroeconomic problems and structural distortions, possessed reasonably well-developed private sectors, and generally had above-average growth rates even prior to adjustment. In these cases, exports, saving, and, with a short lag, private investment all rose following the implementation of adjustment policies. For most other countries, structural adjustment has taken longer and has typically involved a period of declining output and labor demand before new sources of growth emerged. Higher exports and saving frequently followed after a year or two (although often only weakly in many low-income countries) but private investment generally only responded after a number of years (RAL-2, Chhibber). Even in some of the Asian success cases (for example, Indonesia) the foundations for growth were laid only over quite a long period.107
Box 7.Lessons from the East Asian “Miracle”
The World Bank’s comprehensive study of the factors underlying the success of the eight high-performing Asian economies concluded that there was no single “model” of success.1 The role of public policy varied from Hong Kong, with its hands-off approach, to much more activist policies in Japan and Korea. However, some common threads do emerge, including a commitment to a relatively equal dispersion of the benefits of growth and a reliance on the private sector. Moreover, getting the fundamentals right in several broad policy areas was crucial: (1) prudent fiscal and monetary management; (2) heavy investment in human capital, especially in basic education; (3) creation of effective and secure financial systems; (4) limiting price distortions, especially to ensure that domestic prices of traded goods were never too far from international prices, and promoting flexible labor markets to help supply respond quickly to those prices; (5) ensuring easy access to foreign technology rather than attempting a path of self-reliance; and (6) limiting the bias of price incentives against agriculture. In countries that relied on selective government intervention, notably in the areas of industrial policy and the financial sector, the results of such intervention were mixed. The study finds that some favorable outcomes were achieved by creating competition among firms to meet well-defined economic performance criteria in return for preferential treatment; even then, industrial policies also produced some notable failures. Moreover, the conditions for successful intervention (notably a bureaucracy insulated from political pressures) would be hard to reproduce and were less crucial to growth than the fundamentals emphasized above.1 See World Bank (1993b), Chapter 7.
The usual relationship between real GDP growth and the ratio of investment to GDP typically breaks down during adjustment periods. Growth is faster than one would expect given the decline in investment, either because of efficiency improvements associated with adjustment or because of a recovery in capacity utilization (RAL-2, OED).108
Investment and Saving: The Need for Credibility and Sound Fiscal Policies
Countries need to establish a track record of sound policies—macroeconomic and structural—before a significant strengthening of private investment can be expected, and this takes time (CR, RAL-2, RAL-3). The investment “pause” can typically last from three to five years in middle-income countries and even longer in many low-income countries (RAL-3, Chhibber).
It is not possible to make general statements about the links between public and private investment; effects vary by country, depending upon the extent to which governments compete with the private sector for investment opportunities and access to financing. For example, crowding-out effects predominate in about half of the cases considered in Easterly. However, econometric studies suggest that public infrastructure investment is often positively correlated with private investment.109
The channels for policies to directly influence saving rates are often limited. Increasing government saving is the most direct means to raise national saving. On average, only about half of any change in government saving is offset by opposite movements in private saving (CR, RAL-2, Easterly).110 Haque reflects the presence of liquidity constraints affecting some households. Their results also suggest that developments that ease these liquidity constraints—for example, financial reform that liberalizes access to consumer credit—could be associated with a temporary stimulus to consumption.
Box 8.Principal Studies Referred to in Appendix I
• Bruno: contains case studies of stabilization in a number of high-inflation countries—Argentina, Bolivia, Brazil, Chile, Israel, Mexico, and Turkey; see Bruno and others (1991).
• Chhibber: contains case studies of Chile, Colombia, Egypt, Indonesia, Morocco, Turkey, and Zimbabwe; see Chhibber and others (1992).
• Corbo: symposium on the effectiveness of adjustment lending by the World Bank and on macroeconomic policies to restore sustainable growth; see Corbo, Fischer, and Webb (1992).
• CR: the 1994 review of IMF conditionality; see Schadler and others (1995) and Schadler (1995).
• East Asia: World Bank study of the role of public policy in development in eight economies—Hong Kong, Indonesia, Japan, Korea, Malaysia, Singapore, Taiwan Province of China, and Thailand; see World Bank (1993b).
• Easterly: studies of the links between fiscal adjustment and macroeconomic performance in ten countries—Argentina, Chile, Colombia, Cote d’lvoire, Ghana, Mexico, Morocco, Pakistan, Thailand, and Zimbabwe; see Easterly and others (1994).
• ESAF: the 1993 review of lending under the IMF’s enhanced structural adjustment facility; see Schadler, Rozwadowski, Tiwari, and Robinson (1993).
• Hadjimichael: IMF study on growth, saving, and investment in sub-Saharan Africa; see Hadjimichael and others (1995).
• Kirmani: contains a comprehensive review of the trade policy content of IMF-supported adjustment programs; see Kirmani and others (1994).
• Little: studies of 17 countries—Argentina, Brazil, Cameroon, Chile, Colombia, Costa Rica, Cote d’lvoire, India, Indonesia, Kenya, Mexico, Morocco, Nigeria, Pakistan, Sri Lanka, Thailand, and Turkey; see Little and others (1993).
• OED: the 1980-92 report of the World Bank’s Operations Evaluation Department; see World Bank, Operations Evaluation Department (1995).
• OED-Trade Policy: study by the World Bank’s Operations Evaluation Department; see World Bank, Operations Evaluation Department (1992).1
• WDR: the World Bank’s 1991 World Development Report; see World Bank (1991).
These studies have adopted a wide variety of methodological approaches, including detailed analysis of individual country cases (for example, Africa, East Asia, Bruno, Little); modified control-group comparisons that attempt to control for initial conditions and exogenous shocks (RAL-2, RAL-3, Corbo); cross-country or time-series econometric analysis of the links between policies and performance (most studies); division of countries into groups according to their growth performance in order to identify any systematic policy differences (RAL-2, RAL-3, OED-Trade Policy, WEO); and, for a few specific issues, counterfactual simulations using partial or reduced-form models for particular countries (Easterly, Bruno). Although all of these studies provide insights into the links between adjustment and growth, each is subject to some methodological limitations that must be borne in mind when interpreting the results.2 Nevertheless, a number of broad messages emerge consistently and appear to be robust to different approaches.1 See also Thomas, Nash, and others (1991) and Papageorgiou and others (1990).2 Before-after comparisons are especially likely to be misleading if exogenous influences are important, and division of countries into groups according to their performance is subject to sample selection bias. Modified control-group approaches are, in principle, a conceptually superior way of comparing the outcomes of IMF- or Bank-supported adjustment programs with what would have happened in the absence of such programs, but the policy reaction functions that underlie some of these approaches may be unstable. Model-based simulations of counterfactuals for particular countries are also not without problems: simple models cannot capture all of the important links between policies and outcomes, particularly when structural policy changes are important, whereas more complex models cannot be estimated empirically for most developing countries. Moreover, the problem of parameter instability across regime changes is always an issue.
Most empirical studies have concluded that the real interest elasticity of total savings is not significantly different from zero. However, Ogaki and others (1996) show that the pattern of saving tends to change once subsistence considerations are satisfied. They find positive interest rate effects that vary with income, but are still small. Moreover, there is considerable evidence that positive real interest rates boost the share of saving intermediated through the financial system, with potentially positive effects on investment efficiency.
The correlation of fiscal deficits with any one indicator of macroeconomic imbalance (inflation, real exchange rates, and real interest rates) is low in the short run, because of the great variety of ways in which governments finance their deficits.111 However, there is substantial evidence that reducing large deficits is good for growth.112 The benefits can operate through various channels including reduced monetary financing and hence inflation and reduced issuance of domestic debt, which helps reverse crowding out; as well as effects on the exchange rate (RAL-3, Easterly).
At an aggregate level, the composition of fiscal adjustment in IMF-supported programs was typically tailored to the nature of the imbalances: in terms of country averages, deficit reduction was targeted to come equally from revenue increases and expenditure cuts; moreover, targeted revenue increases were largest in countries with low initial revenue ratios. However, ad hoc, mid-program adjustments fell most heavily on expenditures, especially capital spending (CR). Moreover, in many programs supported by the World Bank there has been an inadequate reallocation of spending to some high priority, growth-oriented activities.113 Critical nonwage operating and maintenance expenditures were often unduly squeezed, while there was typically scant progress in reducing excessive public sector employment (RAL-3).114 More fundamentally, adjustment programs have had limited success in tackling problems such as the lack of public sector accountability and transparency that are at the core of efficient public expenditure management. Such problems require broader, and long-term, efforts to build institutional capacity (Africa).
There is some evidence of a nonlinear relationship between real interest rates and growth. Moderately positive real interest rates (that is, in the range of 0-5 percent) have been a feature of most countries with better than average growth, whereas a larger proportion of countries with negative or very high positive real rates have been in the lower-growth categories. However, a few countries recorded strong growth even though very high ex post real interest rates persisted for long periods (WEO).115 Real interest rates were also considerably less volatile in the successful East Asian economies than in many other regions, reflecting their record of relative macroeconomic stability (East Asia). Cross-sectional studies suggest that the link between moderately positive real interest rates and growth operates largely through influences on the level and efficiency of financial intermediation, rather than on the overall saving rate.
Exchange rate policy typically reflects a tension between two goals: first, anchoring inflation expectations and disciplining fiscal policies; and, second, maintaining or improving competitiveness. Ideally, financial policies would be sufficiently restrictive to allow the exchange rate to be used as an anchor, with adjustment only in the face of real disturbances. In practice, financial policies are usually not this strong, and there is frequently a trade-off between inflation and competitiveness (CR, Bruno). This trade-off is perhaps most evident during exchange-rate-based stabilization where even in “successful” disinflation efforts backed by substantial fiscal consolidation, such as those of Chile and Israel, choices on when or if to relax the anchor can have an important influence on the transition from stabilization to growth—in part because such choices affect the switch of demand to net exports (Bruno). However, similar choices can be required in countries with a record of low inflation that are subject to large external shocks to which fiscal and wage policies are slow to adjust, such as in the CFA franc zone. There-fore, such trade-offs should be recognized explicitly and kept under close review. An exchange rate anchor may have to give way to greater flexibility in the face of adverse terms of trade movements or a protracted loss of competitiveness. However, limits to the effectiveness of nominal depreciations in achieving a real depreciation without appropriate support from fiscal and wage policies must also be recognized. In particular, explicit real exchange rate rules are never a good idea, because of the inflation risks involved in indexing such an important price and the potential for continuous upward pressure on the money supply through the balance of payments if an attempt is made to set the real rate at an inappropriate level (CR).
Stabilization in economies with chronic inflation is likely to be complicated by built-in mechanisms for mitigating inflation that make it easier to live with inflation and more difficult to gain a consensus for disinflation policies. Even when successful, stabilization is likely to be protracted because of imperfect policy credibility (Corbo, Bruno). If a nominal anchor is used to help break inflationary inertia, both theory and empirical evidence suggest that the choice of anchor can have important effects on the dynamics of the disinflation program. In particular, the timing of any downturn in economic activity associated with the disinflation is typically different for money-based (early recession) and exchange-rate-based (late recession) stabilization episodes.116 In the case of exchange-rate-based stabilization, imperfect credibility of the anchor and supporting policies frequently leads to an initial, unsustainable boom in consumption and a surge in imports, followed by a recession. In many cases, an additional important factor was the expansionary effects of a repatriation of assets during the initial stages of stabilization. When disinflation efforts have failed, it has often been because of the lack of supporting fiscal policy or an inability to tackle indexation, especially for wages (CR, Bruno, Corbo).117
External Financing and the External Economic Environment
It is difficult to identify any systematic association between external financing and growth from cross-country studies.118 This is not surprising, since the relationship depends crucially on how policies respond to a greater availability of foreign saving. On average, net resource transfers (as a share of GDP) showed little difference between “high-growth” and “low-growth” groups covering all developing countries (WEO).119 Cross-country evidence provided in RAL-2 suggests that, on average, total official external flows had a positive effect on growth in low-income countries, but no clear effects for middle-income countries. In sub-Saharan Africa, countries that benefited from an increase in external transfers achieved moderately higher growth rates—although differences in policies were a more important factor determining growth performance (Africa).120
Cross-sectional analysis also provides a mixed picture of the effects of capital flows on investment: the significantly positive relationship between changes in countries’ debt-to-GDP ratios and changes in their investment rates that existed during the 1970s suggests that external borrowing tended to raise investment rates; this relationship vanished during 1979-83 and only re-emerged weakly during the rest of the 1980s (WEO). Most econometric results based on cross-section or panel data also support the hypothesis that foreign capital inflows have a negative impact on domestic saving. An IMF staff study of saving behavior suggests that an increase in foreign saving equivalent to 1 percentage point of GDP is associated on average with a decline in national saving of about 0.4 percentage point of GDP.121 Once again, however, both the Hadjimichael and Africa studies suggest that the response of policies is crucial: in sub-Saharan Africa, the negative relationship between external financing and domestic saving was much stronger in countries that were not able to address protracted imbalances; indeed, in the group of sustained adjusters, an increase in foreign aid appears to have been associated with a small rise in domestic saving.
Good policies matter for growth, but so does good luck. For example, the “low-growth” group of countries identified in the WEO were exposed to a particularly unfavorable external environment during 1984-93.122 A deteriorating terms of trade and weak industrial country demand is estimated to have reduced the average annual growth rate of this group by 3/4 of 1 percentage point during this period, whereas more favorable conditions boosted the average growth rate of the “high-growth” group by 1 percentage point; thus, the external environment accounted for more than one quarter of the difference in growth between the two groups. Nevertheless, even in regions that have been most severely affected by a deteriorating terms of trade—such as sub-Saharan Africa—countries that have acted to improve their competitiveness and implement structural adjustment measures have done better than others in the region in terms of saving, investment, and per capita GDP growth (Hadjimichael and ESAF).
The degree of correlation between external financing and investment or growth appears to differ according to the composition of the external financing. Foreign direct investment and private debt-related flows are both more positively correlated with domestic investment than are official flows, whereas foreign direct investment is more positively correlated with real GDP growth than either of the other two types of flows (WDR).123 However, the direction of causation is not clear: faster growth is also likely to stimulate more foreign direct investment.
Structural Policies: Start Early and Reach a Critical Mass Soon
The analytical basis for comparing different sequences for removing structural distortions in a second-best world is limited; the appropriate sequencing depends on characteristics specific to each case and must, therefore, be based on an evaluation of the main obstacles to growth in each country.124 Difficult trade-offs between various intermediate goals for structural reform are often required, especially in countries with limited administrative capacity. Nevertheless, the experiences reviewed by the various studies suggest a few general lessons:
Reform efforts must reach a critical mass to be effective. Small reductions in large distortions do little to raise growth; therefore, sequencing should generally begin by tackling the largest distortions (RAL-2, Corbo, WDR).125
Structural reforms have typically not been successful when they were undertaken without effective macroeconomic stabilization. When macroeconomic conditions are unstable, consumers and producers do not believe the reforms will be sustained; in the resulting uncertainty, their responses can destabilize the reform effort. Moreover, improved resource allocation signals may be lost in the noise created by high inflation (OED, RAL-2, RAL-3).
Reforms that take a long time to be effective should generally begin at an early stage. Moreover, “getting prices right” is not the only determinant of faster growth: supply responses are likely to be muted without comprehensive efforts to address institutional and regulatory bottlenecks, which typically entail long gestation periods (RAL-3, OED). Moreover, there are often close linkages between certain “macro” reforms (such as trade liberalization) and more micro-level, sectoral reforms (such as deregulation of marketing and distribution). Careful attention to the appropriate clustering of such reforms, which can only be determined case by case, can have a major influence on the speed and magnitude of the supply response (RAL-2, Corbo).
The speed of supply response to trade and exchange reforms varies widely from country to country.126 Reducing protection in the context of a medium-term strategy that clearly establishes and preannounces immediate and medium-term objectives provides the right signals for production and investment and reduces uncertainty, thereby enhancing the supply response and helping to counter pressures for an ad hoc reinstatement of protection. Credibility and sustainability also appear to be enhanced if trade liberalization begins with a substantial initial effort; in contrast, programs that began with weak initial steps are less likely to be sustained (RAL-2, Corbo, Kirmani).127 Fiscal considerations have often played a key role in limiting the pace of tariff reform, so program design needs to recognize explicitly the link between the two; complementary domestic tax reforms contribute to the sustainability of trade reforms (Kirmani). However, a well-designed trade reform need not be inimical to revenue generation: the conversion of quantitative restrictions to tariffs is likely to increase fiscal revenues and should be implemented at an early stage of macroeconomic stabilization (Kirmani). Poorer and less-diversified economies, often facing institutional weaknesses and impediments to factor mobility, have the slowest supply response to trade reforms; in these cases, efforts to strengthen the base for exports (through domestic regulatory or institutional reforms) often take much longer than expected and should be given high priority (RAL-2, OED-Trade Policy).
Public enterprise reforms, especially divestiture, have typically been among the slowest of all reforms to be implemented, which suggests that the ground-work needs to be laid at an early stage. The primary objective of the reforms should be to improve efficiency, which should not be overshadowed by the short-term goal of generating fiscal receipts from privatization (OED). In Africa, efforts to reform public enterprises through hard budget constraints without privatization have generally not achieved lasting success, because they tend to be undermined by the myriad channels for soft budget support; in such circumstances, removal of explicit budgetary subsidies encouraged their replacement by less-transparent methods (Africa (World Bank, 1994a, Chapter 4)).
Severe financial repression (that is, large negative real interest rates) is bad for growth and should be removed quickly, taking account of the potential fiscal consequences (WEO). The evidence on the effects of moderate financial repression, provided real interest rates on deposits are positive, is less clearcut. Holding interest rates slightly below market levels, along with bureaucratic direction of credit, may have had some beneficial effects on investment rates in some of the high-performing Asian economies (for example, Korea and Taiwan Province of China); however, it is doubtful whether the conditions for such a beneficial effect can be reproduced in most developing countries (East Asia). In many low-income countries, effective implementation of prudential regulations seems to have been a particular problem, and progress in addressing the problems of troubled financial institutions was especially slow. A general lesson seems to be that broad financial reform needs to be accompanied by changes in the public enterprise sector and by a strengthening of the legal framework to enforce financial discipline. For example, bank recapitalizations in Africa—without a corresponding restructuring of public sector enterprises—have generally failed, with new balance sheet problems quickly reemerging (Africa (World Bank, 1994a, Chapter 4)).
The experience of sub-Saharan Africa suggests that improvements in macroeconomic policy are not, by themselves, sufficient to trigger a substantial supply response from agriculture. However, when accompanied by other, micro-level reforms—most notably marketing board reforms—that raised real producer prices, the countries involved have typically achieved higher agricultural growth (Africa (World Bank, 1994a, Chapters 4 and 5)).128
Labor market data is extremely weak in many developing countries so it is often not possible to make definitive statements about the effects of adjustment policies; in many countries, the major form of quantity adjustment in the labor market has been a shift from formal to informal sector employment. Moreover, labor market reforms have typically been the most difficult of all to implement. Labor market rigidities—notably a slow responsiveness of real wages to labor market conditions—were prevalent in many countries with IMF-supported programs, but were addressed to only limited degrees (CR). However, for those countries where relevant data is available, obtaining an initial decline in real wages at the outset of an adjustment effort appears to have helped to achieve macroeconomic adjustment without major unemployment (Corbo). De facto or explicit arrangements for backward-looking indexation were especially problematic (CR, Bruno).
There is a large and growing body of empirical literature on factors influencing economic growth in the long term.129 In this appendix, evidence drawn from a cross-country empirical analysis of long-term growth is used to examine, in turn, the following three questions:
• What does evidence from a broad panel of countries say about the correlation of particular policy-related variables with growth, capital accumulation, and productivity?
• What is the influence of factor accumulation and technological convergence on long-term growth?
• Taking into account the estimated influence of these long-term determinants of growth, how did the pattern of growth in the eight countries change over the different adjustment periods and how much of these changes can be explained by quantifiable policy-related variables?
Correlations Between Growth and Policy-Related Variables
To address the first question, a growth-accounting framework was used to assess empirically the links between policies and growth for as broad a group of countries as possible. One can then examine how these policy-related factors have fared in the eight countries. Annual real GDP growth, the rate of growth of the real capital stock, and the rate of growth of total factor productivity (TFP) are each regressed on a set of policy-related variables using pooled cross-section and time-series (panel) data for 92 industrial and developing countries over the period 1970-92.130
The evidence (Table 15) suggests that macroeconomic instability (as measured by inflation, fiscal deficits, and the black market exchange rate premium) is significantly negatively correlated with growth and that the links operate by dampening both capital accumulation and productivity—although causality is likely to run in both directions since adverse supply shocks are likely to raise inflation and put upward pressure on fiscal deficits.131 Specifically, high inflation is associated with lower output and TFP growth and with slower capital accumulation. Fiscal surpluses are positively correlated with growth in output and TFP but are negatively correlated with capital accumulation. An examination of the separate influences of government saving and capital expenditure reveals that both are positively correlated with capital accumulation but that the coefficient on capital expenditures is larger than that on government saving. This finding suggests that government capital spending appears to increase overall capital accumulation (that is, private investment is not crowded out one-for-one). Large parallel market exchange rate premiums are negatively correlated with growth. Structural distortions—namely, high trade taxes and an underdeveloped financial system—and low educational attainment are also negatively correlated with growth; the effects appear to operate through both productivity gains and capital accumulation.132 Finally, favorable terms of trade movements are found to be associated with higher growth, with the link operating mainly through their impact on productivity.
|Estimated Correlation with Growth||Estimated Correlation with Capital Accumulation||Estimated Correlation with TFP Growth||World||Developing Countries||Bangladesh||Chile||Ghana||India||Mexico||Morocco||Senegal||Thailand|
|Inflation||− − −||− − −||− − −||22||27||11||90||41||9||41||8||7||7|
|Exchange rate premium||−−−||−−−||…||75||102||121||54||417||20||8||6||2||− −|
|Change in the terms of trade||+++||…||+++||−0.4||−0.5||1.0||−3.4||−2.7||0.9||−0.2||0.4||−0.3||−1.8|
|Primary school enrollment rate||+++||…||++||85||78||61||100||61||72||100||52||43||81|
|Ratio of broad money to GDP||…||++||…||39||32||25||30||19||36||24||43||25||47|
|Average effective trade taxes4||…||− − −||…||13||14||12||16||18||25||10||10||14||8|
The partial correlations are estimated from three panel regressions (across country and time, with individual annual observations) for a large group of countries with growth in real GDP, capital accumulation, and TFP as dependent variables. TFP residuals are calculated as follows: TFP = ZGDP –0.4 ZKAP –0.6 ZLAB, where ZGDP is the rate of growth of real GDP, ZKAP is the rate of growth in the real capital stock, and ZLAB is the growth rate of the labor force. The data for ZGDP and GDP in 1970 are measured at PPP prices and are taken from the Penn World Tables, Mark 5.6; data for ZKAP are derived using data from the Penn World Tables, Mark 5.6, and the methodology proposed by King and Levine ((1994); data for the budget surplus, government capital expenditure and saving, and the average effective import tax rate are from International Monetary Fund, Government Finance Statistics Yearbook, or staff country reports; data for inflation and broad money are from International Monetary Fund, International Financial Statistics; and all other variables are from Fischer (1993). +++ and − − − mean that the correlations are significant at the 1 percent level, and ++(––) and + (–) imply significance at the 5 percent and 10 percent levels, respectively.
World and developing country averages for inflation exclude Argentina, Bolivia, Brazil, and Nicaragua.
The effects of the budget surplus, on the one hand, and capital expenditures and government saving, on the other, were estimated in separate panel regressions. The estimated effects of the other macroeconomic variables did not differ significantly between the two regressions.
Measured as total trade taxes divided by the total value of merchandise trade.
To illustrate how these various factors may have affected long-term growth in the eight countries, Table 15 also presents period averages of the various policy-related variables. Important points to note are, first, Chile, Mexico, and Ghana experienced much higher rates of inflation than the other countries in the study; Ghana as well as Bangladesh also had larger exchange market distortions. In contrast, Morocco, Senegal, Thailand, and India had relatively stable macroeconomic environments (although significant exchange market distortions existed in India). Second, Bangladesh, Ghana, Morocco, Senegal, and, to some extent, India, began with markedly lower human capital endowments, measured by primary school enrollment rates. For these countries, fiscal adjustment that involved squeezing spending on primary education could be expected to be especially harmful to growth. Third, Ghana and India had distinctively more restrictive trade regimes, and Bangladesh and Ghana more underdeveloped financial sectors—although substantial structural changes took place in each of these countries over the period. Fourth, Chile, Ghana, and Thailand experienced the most unfavorable terms of trade movements on average during this period.
Long-Term Cross-Country Comparisons
The second question posed above is examined using cross-section regressions that link per capita real income growth to the accumulation of physical and human capital. Following Barro (1989), and based on a sample of 103 countries and data averaged over the period 1960 to 1992, per capita income growth is regressed on the share of investment in GDP, the growth rate of population, and measures of educational attainment as proxies for the accumulation of physical capital, labor, and human capital, respectively.133 An additional explanatory variable is the relative income gap of each country with the United States in 1960—a term that is meant to capture the prediction of growth models that less technologically advanced economies would tend to catch up with more advanced ones.134 The estimated parameters are then used to conduct a simple decomposition of growth into the part accounted for by factor accumulation and that due to productivity changes other than those arising from technological convergence. The results for the eight countries in this study are shown in Table 16.
The results suggest that the two most important variables contributing to growth are investment in physical capital and the achievement of basic levels of educational attainment, as measured by primary school enrollment rates. Population growth was found to have an insignificant impact on per capita growth. However, like other studies of this type, which generally find that factor accumulation accounts, on average, for between one half and two thirds of long-run growth, the right-hand-side variables explain only part of the differentials in long-term per capita growth. This suggests that, to varying degrees, long-term growth performance is also attributable to other influences, which include productivity developments not related to “catching up” and changes in the quality of factor inputs—both of which are influenced by economic policies.
After taking account of factor accumulation and technical convergence, actual growth is found to be higher than “predicted” growth, for five of the eight countries—Bangladesh, Mexico, Morocco, Senegal, and, most notably, Thailand. In Chile, Ghana, and, to a smaller extent, India, however, predicted growth is found to be higher than actual growth, suggesting that other influences, including policies, have tended to dampen productivity and growth, at least on average during this 30-year period.135 However, it will be shown in the next section that these period averages mask important shifts over time.
|Parameter Estimate||Sample mean||Contribution||Sample mean||Contribution||Sample mean||Contribution||Sample mean||Contribution|
|Actual as a percentage of estimated||550||64||35||86|
|Parameter Estimate||Sample mean||Contribution||Sample mean||Contribution||Sample mean||Contribution||Sample mean||Contribution|
|Actual as a percentage of estimated||159||331||428||175|
Performance During Different Adjustment Periods
Changes in growth performance in the eight countries during different phases of adjustment can be considered on the basis of the cross-country regressions of the long-term determinants of growth. Specifically, the following questions are examined:
• Was growth in the eight countries—after taking into account the long-term determinants of growth discussed above—significantly below or above the “world average” before, during, and after adjustment?
• Do the policy-related factors discussed earlier explain most of these growth differentials and how they have shifted over the different adjustment periods?
Following an approach suggested by Bruno and Easterly (1995), the regression discussed above is re-estimated with the following modifications: for each country under consideration, the 1970-92 period is divided into three (or four) subperiods corresponding to the identified adjustment phases and data are averaged over these subperiods.136 A panel regression is then estimated across these subperiods and over the whole sample of countries (totaling 92) with two sets of dummy variables—one for each subperiod for all countries and one for each subperiod for the country of interest.137 The coefficients on the individual country dummies are reported in Table 17 (growth differential A) and in Chart 10. These “growth differentials” can be interpreted as the extent to which each country’s growth in each subperiod differed from the world average after controlling for the long-run determinants of growth and exogenous shocks common to the growth experience of all countries.
On the basis of this exercise, the eight countries can be classified into several groups according to the growth response over the different adjustment periods:
• Chile, Ghana, and Mexico paid a substantial price, in terms of negative growth differentials, during the first phase of their adjustment. Chile and Ghana, however, achieved a large and sustained turnaround associated with a large positive growth differential later in the adjustment process; the length of the recovery in Chile especially, spanning more than a decade, suggests more than a reversion-to-trend phenomenon. In contrast, the revival in Mexico was much more muted; the growth differential during the second adjustment phase (1988-93), while positive, was relatively small. This suggests that, unlike most other countries (including Chile and Ghana) that achieved substantial reductions in inflation, Mexico did not benefit from a large post-stabilization rebound in per capita growth relative to the world average.
• Bangladesh shifted from a situation where growth was markedly below the world average prior to adjustment to positive growth differentials in the respective adjustment periods.
• In Morocco, Senegal, and India, no sustained shift in growth differentials occurred between the preadjustment and adjustment periods.138 Unlike the other two countries, however, Morocco’s growth performance remained above the world average in all periods, despite major adverse supply shocks.
• Thailand stands out with much stronger growth relative to the world average throughout the period. No decline in the growth differential was observed during the adjustment period, and the post-adjustment growth differential widened markedly.
A second set of growth differentials (line B in Table 17) was estimated by adding to the right-hand side a set of policy-related variables (inflation, the budget deficit, and the parallel market exchange premium).139 This specification provides an alternative approach to measuring growth differentials in which the role of macroeconomic policies is explicitly taken into account.140
An important conclusion from this exercise is that the quantifiable proxies for macroeconomic policies do have additional explanatory power in periods when countries experienced episodes of massive macroeconomic instability (for example, high inflation or extreme distortions in the exchange market): Chile in the early 1970s, Ghana prior to the adoption of the Economic Recovery Program, and Mexico after the debt crisis. Although there are always problems of establishing the direction of causation, since exogenous factors that lower output may also put up-ward pressure on inflation, the results suggest that early action to prevent such episodes is likely to be beneficial to growth.
|Preadjustment Period||First Adjustment Period||Second Adjustment Period||Postadjustment Period||Adjusted R2 of the Associated Regression|
|A. Growth differential after controlling for all long-run determinants and changes in the terms of trade||−4.2***||25***||2.9***||…||0.37|
|B. Growth differential after controlling for long-run determinants, terms of trade changes, and policies||−3.8***||2.1**||2.3***||…||0.43|
|A. Growth differential after controlling for all long-run determinants and changes in theterms of trade||−3.6***||−1.0||3.0***||3.1***||0.31|
|B. Growth differential after controlling for long-run determinants, terms of trade changes, and policies||−1.9**||−1.2||2.4***||2.8***||0.39|
|A. Growth differential after controlling for all long-run determinants and changes in the terms of trade||−24***||1.3***||2.7***||2.3***||0.25|
|B. Growth differential after controlling for long-run determinants, terms of trade changes, and policies||−1.8***||1.7**||2.1***||…||0.36|
|A. Growth differential after controlling for all long-run determinants and changes in the terms of trade||0.5||−0.6||…||…||0.21|
|B, Growth differential after controlling for long-run determinants, terms of trade changes, and policies||0.5||−0.3||…||…||0.34|
|A. Growth differential after controlling for all long-run determinants and charges in the terms of trade||2.1***||−1.4***||0.5||…||0.31|
|A. Growth differential after controlling for all long-run determinants and changesin the terms of trade||—||1.9***||−0.8**||…||0.30|
|A. Growth differential after controlling for ail long-run determinants and changesin the terms of trade||2,1***||2.9***||6.8***||0.36|
|B, Growth differential after controlling for long-run determinants, terms of trade changes, and policies||2.1***||2.7***||6.4***||0.43|
Following an approach suggested by Bruno and Easterly (1995), the growth differentials are defined as the estimated coefficients on the dummy variables for each country (that is, the estimated B’s) in Barrostyle growth regressions using pooled data on the average growth rates over the respective subperiods and across 92 countries. The estimated equations took the following form: In addition, the following policy-related factors were added to the regressors underlying the second set of results reported for each country:
α7 INFL + α8 BUDSUP + α8 PREMIUM + α10 HINFL
where T is the number of subperiods in each case; GRTH is the growth rate of real per capita GDP measured in terms of constant domestic prices; INV is the share of investment in GDP; POPG is the growth rate of the population; GAP70 is the ratio of country i income to that of the United States in 1970; PRI70 and SEC70 are primary and secondary school enrollment rates, respectively, in 1970; INFL is the average annual rate of inflation in the CPI; BUDSUP is the government budgetary surplus as a percent of GDP; PREMIUM is the average premium between the parallel market and official market exchange rates; and ATOT is the change in the ratio of export prices to import prices. Data for GRTH are taken from Bruno and Easterly (1995). Data for all other variables are from the same sources as listed in Tables 15 and 16. Possible nonlinearities in the effects of inflation on growth are taken account of by including the variable HINFL, which takes the value of the inflation rate when it is greater than 40 percent. DUMi are the dummies for each period for all countries and DUMI are dummies for the country being examined, one for each period. Absolute values of heteroskedastic-consistent t-statistics are in parentheses.*** denotes significance at the 1 percent level, and ** and * at the 5 percent and 10 percent levels, respectively.
Fiscal year data.
For countries that experienced long periods of macroeconomic stability, most notably Thailand, the proxies for macroeconomic policies tend to have less significant explanatory power, suggesting that an important part of the variation in growth in some of these countries (notably Thailand) is due to factors (including confidence effects) other than those captured by the macroeconomic policy variables included here.
Assessments of the sustainability and internal consistency of policies by the private sector can greatly influence the response of private investment. An important consideration in any such assessment is how the stance of fiscal policy will affect the public sector debt burden. A path of primary fiscal balances that implied either continued large increases in the ratio of public debt to GDP or continued heavy use of the inflation tax would be less credible. As an indication of the progress made toward fiscal sustainability in each of the eight countries, Chart 23 presents the results of some debt dynamics, based on the intertemporal budget constraint for the public sector.
The actual primary fiscal balance is compared with two notions of a “sustainable” balance: (1) the primary balance that would keep the public debt-to-GDP ratio constant on the assumption that the current rate of inflation (and hence seigniorage) and the current interest rate on domestic debt (and hence the magnitude of financial repression) are maintained in the future; and (2) the primary balance that would keep the public debt-to-GDP ratio constant in a context of low inflation and no financial repression.141 This latter measure is a more comprehensive indicator of fiscal sustainability, since it provides a measure of the deficit reduction that is needed to make fiscal policy consistent with other important goals of an adjustment strategy, namely low inflation and efficient financial intermediation.
Judged by the above criteria, a significant move toward a more sustainable fiscal stance occurred in all eight countries. In the “preadjustment” period, the actual primary deficit was much higher than the sustainable balance in most countries (Bangladesh, India, Mexico, Morocco, and Senegal).142 During the adjustment periods, the actual primary balance was sharply reduced in all countries, and in some (Mexico, Morocco, and Thailand) shifted to a level that significantly exceeded either of the two indicators of a sustainable balance.143 Moreover, financial repression was an important feature of the preadjustment or early adjustment periods in Chile, Ghana, India, Mexico, and Morocco. The substantial convergence in the two sustainability criteria over time indicates the reduced reliance on the inflation tax and on financial repression as means of financing the budget deficits.
One would thus conclude that fiscal adjustments had moved the path of public debt to well within the sustainable range, in the narrow sense used here, in Chile, Mexico, Morocco, and Thailand. A similar unambiguous conclusion is difficult in the cases of Bangladesh, Ghana, India, and Senegal. However, external borrowing on concessional terms is important in these countries, and has contributed to keeping the effective real interest rate on public debt at or below the rate of growth of real GDP. In these circumstances, there is no long-term solvency constraint in the sense that the debt-to-GDP ratio will not grow without bounds.144
Chart 23.Actual and Sustainable Primary Fiscal Balances1
Sources: World Bank, World Debt Tables; and IMF staff estimates.
1 Primary balance includes grants. Sustainable (I) is the balance consistent with constant public-to-debt ratio. Sustainable (2) is the balance consistent with constant public-to-debt ratio, low inflation and no financial repression.
2 Primary balance for Chile includes quasi-fiscal losses of the central bank.
Of course, the measures of fiscal sustainability are partial in that they focus only on the capacity to service and repay outstanding debt. They do not address the more fundamental issue of whether a fiscal balance is also consistent with other macroeconomic objectives, including for the external current account balance. In this latter respect, the fiscal policy responses to recent capital inflows have differed substantially, with potentially important consequences for growth (see Section VII). Moreover, the permanence of fiscal measures used to obtain a sustainable balance can also be a major issue.
The econometric analysis of private investment in the eight countries was guided by two considerations. First, all the time series were expressed either in scaled form or in rates of change to ensure stationarity and avoid spurious correlation.145 Second, the preferred equations were derived from a specification search selecting a parsimonious model from a larger initial set of explanatory variables, by successively eliminating statistically insignificant or implausibly signed regressors.146
The investment equations were estimated by ordinary least squares, for two reasons. First, the full range of factors affecting investment cannot be captured quantitatively (for example, in the case of structural reforms and their impact on private sector expectations). This underscores the importance of selecting an estimation technique relatively robust to misspecification errors in small samples.147 Nevertheless, to reduce the risks of simultaneity bias, the activity variable was specified with a one-period lag.148 Second, unlike instrumental variable techniques, ordinary least squares are not affected by the problem of potential dependence of results on the choice of instruments. Recursive least squares were used to account for the possibility of time-varying parameters, given the potential for shifts in policy regimes and changes in the nature of exogenous shocks.
The results, summarized in Table 18 and discussed in the main text, confirm that estimated equations broadly conforming to standard specifications of investment functions for developing countries are able to account for most of the variation of private investment in each of the eight countries. The estimates generally underscore the important role of accelerator-type effects (except for Bangladesh), indicators of financial policies, and measures of uncertainty and macroeconomic stability in explaining the observed pattern of private investment.
|Country||Sample (Number of annual observations)||Endogenous Variable (In percent of GDP)||Aggregate activity||Real interest rate/user cost||Credit availability||Public investment (In percent of GDP)||Foreign exchange availablility|
|Bangladesh||1979/80–1993/94 (15 obs.)||Priv. inv. (constant prices, logs)||Not signif., omitted||VAR.: real lending rate k COEF:− − −||VAR.: real growth of private sector credit COEF.:+||Not signif., omitted||VAR: reserves in Months of imports (logs) COEF: +++|
|Chile||1975–93 (19 obs.)||Nongovt. inv. (constant prices, logs)||VAR.: Real GDPgrowth, lagged COEF.: +++||(Real lending rate not signif, omitted) VAR.: relative price inv./GDP COEF.:−−||Not signif., omitted||VAR.: govt. inv. (constant prices, logs) COEF.:− − −||Not signif., omitted|
|Ghana||1971–93 (23 obs.)||Priv. inv. (current prices, logs)||VAR.: real GDPgrowth, lagged COEF.:+++||VAR.: real lending rate COEF.:−||Not signif., omitted||VAR.: publ. inv. (current prices, lagged, logs) COEF:− − −||Not signif, omitted|
|India||1973/74–1993/94 (21 obs.)||Priv. inv. (constant prices, logs)||VAR: Lagged growth of industrial production COEF.:++||VAR.: real lending rate COEF.:−||VAR.: real growth of private sector credit COEF: +||VAR.: publ. inv. (constant prices, logs) COEF:− − −|
|Mexico||1972–93 (22 obs.)||Priv. inv. (current prices, logs)||VAR.: real GDP growth lagged COEF.:+++||VAR.: real deposit rate COEF.:−||VAR.: real growth of private sector credit COEF: ++||VAR.: publ. inv. (current prices, logs) COEF:− − −||Not signif, omitted|
|1981–93 (13 obs.)||Priv. inv. (constant prices, logs)||VAR.: real GDP growth, lagged COEF.: ++||VAR.: real deposit rate COEF:− − −||VAR.: real growth of private sector credit COEF: ++||VAR.: publ. inv. (constant prices, logs) COEF.:−|
|Morocco||1972–93 (22 obs.)||Priv. inv. (constant prices, logs)||VAR.: Real GDP growth, lagged COEF: +||VAR.: real 6-month T-bill rate COEF.:−−||VAR.: real growth of nongovt. sector credit COEF: +||VAR.: publ. inv. (constant prices, logs) COEF: +++||Not signif, omitted|
|Senegal||1978–93 (16 obs.)||Nongovt. inv. (constant prices, logs)||VAR.: Real GDP growth, lagged COEF.: +||Not signif., omitted||Not signif., omitted||VAR.: govt. inv. (constant prices, logs) COEF.:−−|
|Thailand||1970–93 (24 obs.)||Priv. inv. (constant prices, logs)||VAR.: Real GDP growth in partner countries COEF.:+++||VAR.: Real lending rate COEF.:−||Omitted||VAR.: publ. inv. (constant prices, logs) COEF:− − −||Not signif, omitted|
|Measures of uncertainty/macroeconomic instability|
|Country||Inflation rate/parallel market premium/REER||External debt/debt service||Sample variable measures (or other)||External shocks||Lagged dependent variable||Estimation Technique and Fit|
|Bangladesh||VAR.: external debt service/exports (logs) COEF:− − −||VAR.: variance REER COEF:− − − (variance of inflation and real interest rate not signif., omitted)||COEF: +++||OLS-White HCSE Durbin’s h-statistic = −0.34|
|Chile||Not signif., omitted||VAR.: external debt service/exports (logs)||COEF:− − −||Not signif., omitted||Not signif., omitted||Not signif., omitted||OLS-White HCSE DW statistic = 1.52|
|Ghana||VAR.: exchange rate premium (logs)||COEF:− − −||Not signif., omitted||(Variance of nominal exchange rate not signif., omitted)||VAR.: export price shock (in % o f GDP) COEF:++||Not signif., omitted||OLS-White HCSE DW statistic = 1.87|
|India||Not signif., omitted||VAR.:Variance of WPI inflation rate.||COEF:− − −||VAR: Variance of real exchange rate changes COEF:−||COEF.:+++||OLS-White HCSE Durbin’s h-statistic = −1.82|
|Mexico||Not signif., omitted||Not signif., omitted||VAR.: variance of industrial production COEF:−||Not signif., omitted||COEF: ++||OLS-White HCSER Durbin’s h-statistic = 0.62|
|VAR.: CPI inflation rate COEF:−−−||Variance of industrial production not signif., omitted||COEF:++||OLS-White HCSE Durbin’s h-statistic = −0.47|
|Morocco||VAR.: REER level (logs) COEF:−−−||Not signif., omitted||Variance of CPI inflation not signif., omitted||VAR.: export priceshock (in % of GDP) COEF:++ VAR.: import price shock (in % of GDP) COEF:−||COEF:+++||OLS-White HCSER Durbin’s h-statistic = −0.74|
|Senegal||VAR.: CPI inflation rate COEF:−−− VAR.: REER level (logs) COEF:−−||VAR.: Ext. debt/GDP (logs) COEF:− − −||VAR.: govt. domestic borrowing/GDP COEF:−||Not signif., omitted||OLS-White HCSE DW statistic = 2.71|
|Thailand||VAR.: RULC level (logs) COEF:−−−||Not signif., omitted||Variance of inflation not signif., omitted||Not signif., omitted||COEF:+++||OLS-White HCSE Durbin’s h-statistic = −0.00|
1+, −: Sign of the coefficient, not statistically significant at 5 percent level, but significant at a marginally higher level.
++, − −: Sign of the coefficient, statistically significant at 5 percent.
+++,− − −: Sign of the coefficient, statistically significant at I percent.
This appendix presents background information referred to in Sections IV and VIII. The material is organized in three summary tables: trade reforms (Table 19), financial sector reform (Table 20), and labor market characteristics (Table 21).
|Country (Initial year of reforms)||Status Prior to Reforms||Major Reform Elements||Status in Mid–1995|
|Bangladesh (1985–86)||High tariffs and extensive use of quantitative restrictions (QRs); high degree of tariff escalation.||Initial sharp reduction in number of tariff bands. Gradual reduction in QRs. Significant progress was only made starting in 1991.||Despite reforms, system remains complex and contains many discretionary elements.|
|Chile (1974)||Severely distorted trade system, with average nominal tariff of 105 percent; high dispersion; extensive QRs; and multiple exchange rates.||Rapid replacement of QRs with tariffs; tariffs unified at 35 percent.||Simplified trade system in place with uniform tariff rate of 11 percent.|
|Ghana (1983)||Severely distorted trade system with highly overvalued exchange rate; virtually all imports covered by QRs; extensive taxation of cocoa exports.||Substantial reduction in QRs; maximum tariff reduced from 50 percent to an essentially uniform tariff of 30 percent. Cocoa taxation reduced. In 1986, uniform tariff replaced with four-tiered system with escalating structure to protect consumer goods and luxury items. Import surcharge levied in 1988 and super sales tax on luxury goods imports in 1990.||Despite some reversals, system is much less restrictive and complex than before reforms.|
|India (1991)||Severely restrictive and complex trade regime with proliferation of QRs and high and dispersed tariffs. Imports of consumer goods virtually banned.||Reforms began with removal of export restrictions and of QRs on intermediate and capital goods. Reductions in the level of dispersion of tariffs.||Average tariffs remain high and large share of domestic production—especially of consumer goods—still protected by QRs.|
|Mexico (1985)||Following debt crisis in 1982, many previously reduced trade restrictions reinstated. Widespread trade and exchange controls in use.||Preannounced program of reductions in coverage of QRs; extensive cuts during the first two years. Coverage of official reference prices reduced significantly. Preannounced schedule of planned tariff reductions; maximum tariff reduced in one step from 100 percent to 50 percent. Joined General Agreement on Tarriffs and Trade (GATT) in 1986.||Considerably more liberal, simplified, and outward-oriented trade regime. Remaining import licensing controls affect mainly agricultural and agroindustrial products, oil and derivatives, cars and trucks.|
|Morocco (1983)||Relatively high tariffs and widespread use of QRs. High discrimination against labor-intensive activities.||Early focus of trade reforms was sharp reduction in QRs and official reference prices. Maximum tariff reduced in one step from 400 percent to 60 percent with subsequent further reductions. In 1987, joined GATT and bound a large share of tariffs.||Although only a moderate reduction in average nominal tariffs, the system is much simpler and less distorted. Virtually all tariffs bound under the Uruguay Round.|
|Senegal (1986)||High tariff protection, widespread use of QRs, and selective trade and tax preferences.||Reform program began with the reduction of QRs and a moderate reduction in the maximum tariff rate. Reforms almost fully reversed three years later owing to revenue shortfalls. Average import taxes, which had declined from 98 percent in 1986 to 35 percent in 1988, were back up to 90 percent in 1992, and the use of official reference prices was broadened. In late 1994, the requirement of prior authorization for certain imports and exports was eliminated.||By 1992, Senegal's trade protection was no lower than before the start of reforms, and possibly was more complex. A new phase of reforms launched in 1994 has reduced the restrictiveness of the system.|
|Thailand||Relatively low levels of protection. Trade regime relatively outward oriented but significant intervention through selective tax preferences. Coverage of QRs very low. Protection biased against agriculture and toward capital-intensive import substituting industries.||Only minor initial tariff reductions with some reversal owing partly to fiscal constraints. Export taxation, especially of rice, eliminated in 1986. Tariff reduction on 4,000 product lines in 1995; average tariff to be reduced to 17 percent by 1997.||Trade system remains relatively unchanged with average nominal tariffs at about 30 percent and a relatively large number of tariff rates.|
|Country (Initial year of reforms)||Status of Financial Sector Prior to Reform||Key Elements of Reforms||Important Outcomes|
|Bangladesh (late 1980s)||Small and relatively underdeveloped financial sector dominated by commercial banks. Complex structure of interest rates subject to controls. Extensive quantitative credit controls. Growing share of nonperforming loans. Weak supervisory framework.||Significant progress with liberalizing interest rates. Some steps to strengthen prudential regulations.||Real interest rates positive though not market determined. Spreads remain high, reflecting lack of competition, defaults, and nonperforming loans. Some evidence of financial deepening, but an insignificant increase in private sector‘s share of total credit; other aspects of financial system remain relatively underdeveloped.|
|Chile (1974 and 1981)||Prior to 1974, domestic financial system was fully publicly owned and heavily regulated through interest rate ceilings, quantitative controls on banks, substantial directed credit, and restrictions on operations of financial institutions. Real interest rates were negative.||Privatization and restructuring of the banking system; interest rates liberalized; quantitative credit controls abolished. However, weaknesses in bank supervision and prudential regulations as well as excessive deregulation led to a massive accumulation of nonperforming loans, which were further fueled by speculative capital inflows in the late 1970s. This was followed by a series of bank failures. Major reforms of supervisory and regulatory framework in 1982–83.||Initial sharp increase in real interest rates that did not narrow until mid-1980s. Financial crisis in early 1980s. Significant financial deepening, measured by the share of broad money, plus other financial assets to GDP and in private sector’s share of domestic credit.|
|Ghana (1988)||Underdeveloped and heavily regulated financial sector. Controls on interest rates led to highly negative real rates; together with pre-emption of financial resources by Government, this led to severe financial repression and inefficiencies in banking system.||Controls on deposit and lending rates and on credit lifted starting in 1988. Institutional reforms of the banking system implemented in 1989–90. Government took over large part of nonperforming portfolios of commercial banks.||Real interest rates turned positive in 1991 (lending rates) and 1992 (deposit rates). Spread between deposit and lending rates widened. No significant change in the ratio of M2 to GDP; moderate increase in the private sector’s share of total credit.|
|India (1991)||Highly regulated and inefficient financial markets with controls on interest rates, quantitative credit allocation, government-owned banking system, and public sector pre-emption of a large proportion of financial resources.||Interest rate structure simplified and rates adjusted. In 1994, lending rates liberalized. Government pre-emptions reduced by lowering reserve requirements and remunerating bank holdings of government securities at market-determined rates. Some reduction in directed credit and prudential regulations strengthened. Substantial liberalization of broader financial system, but banking system remains dominated by government-owned banks.||Since late 1994, when interest rates were liberalized, spreads have narrowed but remain high reflecting the lack of competition and high intermediation costs. Financial deepening, especially if measured by a broader financial aggregate, occurred, although private sector’s share of total credit remained broadly unchanged. Some reduction in the share of bank credit to priority sectors.|
|Mexico (1988–89)||Until the mid–1970s, ceilings on interest rates and extensive use of directed credit. But inflation was low during this period. Capital flight in late 1970s weakened banking system. In 1982 commercial banks were nationalized and there was a forced conversion of foreign exchange-denominated deposits at non-market rates.||In late 1988, financial sector reform program launched consisting of interest rate decontrol; rationalization of reserve requirements; and removal of mandatory subsidized lending to the public sector and restriction on lending to the private sector. Banks reprivatized in 1991–92.||Period of financial disintermediation between 1982–88. Capital inflows that began in late 1980s led to expansion in bank balance sheets. Increasing ratio of nonperforming bank loans since 1988 associated with insufficient ability to assess credit risk. Limits imposed in 1992 on banks’ open foreign exchange position and exposure to exchange risk.|
|Morocco (1985)||Heavily regulated financial sector. Quantitative credit controls. Interest rate controls resulted in significantly negative real interest rates in 1980s.||Interest rates liberalized gradually. Credit ceilings abolished in early 1991. New banking law adopted in 1993 to strengthen bank supervisory framework.||Interest rates are positive but not market determined. Government retains preferential access to financial resources, resulting in high taxation of financial intermediation. Indicators of financial sector development show moderate progress.|
|Senegal (1988)||Regulated financial sector, almost fully publicly owned. Quantitative credit controls. Ceilings on refinanced crop credits abused and excessive government guarantees contributed to overborrowing by public enterprises.||Flexible interest rate and credit policies adopted. Supervisory and regulatory capacity of central bank reinforced. The banking sector was restructured, several public sector banks were closed and there was a significant reduction in government ownership in others. Since August 1994, the process of securitizing the banking system debt assumed by the Government has been under way.||Real lending interest rates turned positive after reforms. Indicators of financial sector development present a mixed picture—the share of private sector credit in total credit has increased but the ratio of broad money to GDP has remained unchanged. Reforms reasonably successful in restricting government intervention in credit allocation and in interest rate determination.|
|Thailand||Relatively oligopolistic banking system until mid-1980s when foreign banks were allowed. Interest rates subject to controls. Some directed credit allocation. Relatively open external capital account, especially for inflows.||Financial reforms undertaken at various times in response to particular problems. Important steps include (1) measures to increase competition in banking system and develop interbank money market in 1979; (2) strengthening supervisory arrangements in 1983 and 1985; (3) overall ceiling on domestic credit abolished in 1984; and (4) administered rates that were adjusted to take account of inflation coupled with gradual iberalization of deposit rates (1989—92) and lending rates (1992).||Selective credit policies still in effect. Until the liberalization of lending rates in 1992, spread between deposit and lending rates was wide, suggesting the presence of distortions in banking system. Indicators suggest that substantial progress has been made in developing an efficient financial system.|
|General:||Highly segmented labor markets. Large role of public sector. Restrictions on retrenchment.||Important reforms took place after 1973. Labor markets relatively flexible following elimination of wage indexation in 1982.||Substantial labor market segmentation. Public sector prominent in formal market, acts as wage leader for private sector.||Substantial labor market segmentation. Formal markets highly regulated through restrictions on retrenchment.|
|Labor market segmentation:|
|1. Size of formal sector||1. About 10 percent of labor force.||1. Formal sector is large relative to informal sector.||1. Formal market small relative to in-formal market. Labor unions powerful in formal market and public sector.||1. Employment in formal sector about 8 percent of total employment.|
|2. Wages in formal sector||2. Unskilled public sector wage 68 percent above unskilled agricultural wage in same region.|
|Minimum wage laws:|
|1. Coverage||1. Minimum wages for different skill levels and sectors recommended by the National Minimum Wage Board. Reviews are on an ad hoc basis. Applies to 38 sectors.||1. Minimum wages regulations cover all workers.||1. Minimum wages apply mainly to unskilled formal sector workers.||1. Coverage is narrow.|
|2. Were they important for wage developments? (e.g., ratio of average to minimum wage)||2. No significant impact on market wages (often below market wages). Government unable to enforce minimum wage regulations.||2. Minimum wage growth in 1982–87 fell below average wage growth, and fell in real terms by 40 percent. By early 1990s, real minimum wage returned to 1982 level.||2. Increases in minimum wages contributed to boosting wage incomes between 1984 and 1990.||2. Minimum wages not effectively enforced and no strong impact on other wages. Ratio of average wage to minimum wage: 7:1 for unskilled workers; 3.5:1 for skilled workers.|
|Job security regulations:|
|1. Social security and unemployment insurance||1. No unemployment insurance.||1. Unemployment compensation program covering only a small portion of labor force. Emergency unemployment subsidies created so that 50 percent of unemployed received some assistance in 1982–83.||1. No unemployment insurance.||1. No unemployment insurance.|
|2. Exit and entry laws||2. Retrenchment is restricted. High severance pay that in practice amounts to a minimum of 60 days of wages for each year of service.||2. Liberal. Restrictions on retrenchment eliminated in 1973.||2. Retrenchment is highly restricted in public sector.||2. Stringent job security regulations; in many cases, firms must obtain prior government consent before laying off workers.|
|Public sector employment|
|I. Size||1. Public sector accounts for 50 percent of value added in manufacturing and one third of formal sector employment.||1. Public sector employment amounted to about 9 percent of total employment in 1986.||1. Government employment accounts for two thirds of total formal sector employment; 21 civil servants per 1,000 inhabitants is high by African standards.||1. About 6 percent of total working age population and over 70 percent of formal sector employment.|
|2. Ratio of public sector to private wages or similar indicator||2. Public sector average wages greater than private sector. Public and private wages highly correlated.||2. Public sector has little influence on private sector wages and employment.||2. Public sector wages and benefits comparable to private sector for most of 1980s, but in 1995 was one and a half times greater (except for business sector). Private sector competitiveness and profitability suffered from the effect of generous public sector pay awards in 1992.||2. Public sector wages appear to exert considerable leverage over private formal sector wages.|
|1. Rules and coverage 2. Were they effective?||1. None||1. From 1973, government regulations provided for backward-looking wage indexation, until eliminated in 1982.||1. None||1. None, although public wages are de facto partially indexed.|
|Reforms undertaken during adjustment:||Adjustment programs envisaged little reform; little reform undertaken.||Starting in 1973, legislation enacted that relaxed restrictions on dismissals, diminished union power, and reduced employer share of social security taxes and other nonwage costs. Mandatory wage indexation abolished in 1982.||Civil service reform in second phase of Economic Recovery Program (1987–91) with view to increasing efficiency. But little net retrenchment in overall government employment. Some progress in linking wage increases to productivity and increasing wage differentials. Modest reduction of state enterprise employment.||Minimal changes; mechanisms established to provide additional compensation for layoffs in public enterprises.|
|General:||Formal labor markets regulated and wages influenced by wage-price pacts.||Formal labor markets highly regulated through well-enforced minimum wage laws, and restrictions on hiring and retrenchment.||Highly regulated formal sector and poor record of reforms.||Labor markets have been generally flexible and less segmented.|
|Labor market segmentation:|
|1. Size of formal sector||1. Formal sector substantial; about 70–75 percent of total employment in 1980–85.||1. Wage employment in 1990:62 percent of the urban labor force and 17 percent of the rural labor force.||1. Modern sector employment (civilservice, industry, finance) 131,000 in 1986.||1. Formal sector about 25 percent of labor force before adjustment phase and 33 percent afterward.|
|2. Formal sector wages relative to informal sector||2. Wages in informal sector about one third of formal sector.|
|Minimum wage laws:|
|1. Coverage||1. Introduced in 1970 to cover all workers.||1. SMIG covers all nonagricultural employees and SMAG covers all agricultural employees including temporary and seasonal workers. SMIG has been about 50 percent higher than SMAG since 1986.||1. SMIG covers essentially unskilled urban employees.||1. Introduced in 1973 to cover unskilled nonagricultural workers. About 20 percent of labor force.|
|2. Were they important for wage developments? (e.g., ratio of average to minimum wage)||2. Diminished importance over time. Ratio of average to minimum wage: 1:0.2 in 1984; 1:0.1 in 1990. Some 16 percent of full-time male workers and 66 percent of female workers paid below minimum wages in 1988.||2. The SMIG is strongly correlated with a lag with average formal sector wage; but more than half of firms paid unskilled workers below minimum wages in 1986.||2. In 1986 the average wage of an unskilled worker was about 20 percent higher than the SMIG.||2. Less than one half of unskilled labor being paid the minimum wage. Ratio of average to minimum wage in Bangkok area is 4:1.|
|Job security regulations:|
|1. Social security and unemployment insurance||1. Social security schemes cover formal sector and public sector employment.||1. About 20 percent of urban labor force are covered by National Social Security Fund. No unemployment insurance.||1. Nonwage benefits set by fixed rules.||1. Public sector pension and medical care but no private sector arrangements. Tripartite social security scheme began in 1991 for firms with more than 10 employees. No unemployment insurance.|
|2. Exit and entry laws||2. Severe restrictions on hiring, firing, and reassignment of workers across plants and production lines.||2. Firing workers is highly regulated and often implies costly severance pay.||2. Prior government authorization required for hiring and retrenchment. Controls on use of temporary workers. Collective bargaining at industry level required.||2. Liberal,|
|Public sector employment:|
|1. Size||1. 1980: government employment was 95 percent of the private urban formal sector employment.||1. 1983: civil service—66,310||1. Four percent of labor force.|
|2. Ratio of public sector to private wages or similar indicator||2. The average wage is higher in public sector than private sector.||2. 1980–85: SMIG (real terms) fell 22 percent; real average earnings of civil service fell 30 percent.||2. Private sector wages greater than public sector. Estimate of differential is (1.5—2): 1.|
|1. Rules and coverage||1. 1987-early 1995: annual wageprice pacts negotiated among large unions, industrial groups and government linking exchange rate action with understandings on wages, employment, and public sector price increases.||1. Law requires adjustment in minimum wages whenever the cost of living index rises by 5 percent; applies to all sectors; adjustments of SMIG are irregular.||1. None.||1. None.|
|2. Were they effective?||2. Only increases in minimum wages and public sector prices announced; over time wage guidelines became less representative of actual labor remuneration.||2. Not applied.|
|Labor reforms undertaken:||Apart from the removal of de facto backward-looking wage indexation, relatively little fundamental reform undertaken.||Apart from laws on firing, most regulations not binding. Adjustment programs envisaged little reform and little reform was undertaken.||1986–87: hiring practices and use of temporary labor partially liberalized, but critical clauses of labor code not modified. 1991: limited public sector reform. 1994: labor code revised to liberalize hiring and firing practices.||Limited.|
Price Distortions and Growth in Developing Countries,” World Bank Staff Working Paper No. 575 (Washington: World Bank,July1983).“
The Labor Market and Economic Adjustment,” IMF Working Paper No. 95/125 (Washington: International Monetary Fund,November1995).“
Macroeconomic Adjustment with Segmented Labor Markets,” NBER Working Paper No. 4769 (Cambridge, Massachusetts: National Bureau of Economic Research,June1994).“
Inflation and the Financing of Government Expenditure: An Introductory Analysis with an Application to Turkey,” World Bank Economic Review, Vol. 3(January1989), pp. 17–38.“
On the Political Sustainability of Economic Reform,” IMF Paper on Policy Analysis and Assessment No. 94/3 (Washington: International Monetary Fund,January1994).“
Stock Markets and Development,” European Economic Review, Vol. 37(April1993), pp. 632–40.“
Economic Growth in a Cross-Section of Countries,” NBER Working Paper No. 3120 (Cambridge, Massachusetts: National Bureau of Economic Research,September1989).“
International Comparisons of Educational Attainment,” Journal of Monetary Economics, Vol. 2(December1993), pp. 363–94.“
Behavior of Nominal and Real Interest Rates,” in IMF Conditionality: Experience Under Stand-By and Extended Arrangements, Part II: Back-ground Papers, ed. by SusanSchadler, (ed)IMF Occasional Paper No. 129 (Washington: International Monetary Fund,September1995).“
The Recent Surge in Capital Inflows to Three Asian Countries: Causes and Macro-economic Impact, South East Asian Central Banks, Occasional Paper No. 15(Kuala Lumpur,1993).
Issues in Managing and Sequencing Financial Sector Reforms: Lessons from Experiences in Five Developing Countries,” IMF Working Paper No. 92/82 (Washington: International Monetary Fund,October1992).“
The Impact of Foreign Aid on Savings and Growth,” Working Paper No. 1265 (London: Center for Economic Performance, London School of Economics,November1994).“
How Does Foreign Direct Investment Affect Economic Growth?” NBER Working Paper No. 5057 (Cambridge, Massachusetts: National Bureau of Economic Research,March1995).“
Bosworth, Barry P., RudigerDornbusch, Laban, Raul eds., (ed)The Chilean Economy: Policy Lessons and Challenges (Washington: Brookings Institution,1994).
Adjustment and Equity in Developing Countries: A New Approach (Paris: Development Center, Organization for Economic Cooperation and Development,1992).
A Framework for Evaluating Policy Adjustment Programs: Lessons from a Cross-Country Evaluation,” in Evaluation and Development: Proceedings of the 1994 World Bank Conference (Washington: World Bank, Operations Evaluation Department,1995).“
Bruno, Michael, S.Fischer, E. Helpman, and N.Liviatan, eds., (ed)Lessons of Economic Stabilization and Its After-math (Cambridge, Massachusetts: MIT Press,1991).
Inflation Crises and Long-Run Growth,” NBER Working Paper No. 5209 (Cambridge, Massachusetts: National Bureau of Economic Research,July1995).“
Inflation Stabilization and Nominal Anchors,” IMF Paper on Policy Analysis and Assessment No. 92/4 (Washington: International Monetary Fund,December1992).“
Exchange-Rate-Based Stabilization Under Imperfect Credibility,” in Open-Economy Macroeconomics, ed. byHelmutFrisch and AndreasWörgötter (ed) (Houndsmill, Basingstoke, England: Macmillan Press,1993).“
Chhibber, Ajay, MansoorDailami, and NematShafik, eds., (ed)Reviving Private Investment in Developing Countries: Empirical Studies and Policy Lessons (Amsterdam, New York: North-Holland,1992).
India: Economic Reform and Growth,IMF Occasional Paper No. 134 (Washington: International Monetary Fund,December1995).
Citrin, Daniel A., and Ashok K.Lahiri, eds., (ed)Policy Experiences and Issues in the Baltics, Russia, and Other Countries of the Former Soviet Union, IMF Occasional Paper No. 133 (Washington: International Monetary Fund,December1995).
Interest Rates, Growth and External Debt: The Macroeconomic Impact of Mexico’s Brady Deal,” CEPR Discussion Paper No. 904 (London: Center for Economic Policy Research,1994).“
IMF Lending Programs: Participation and Impact,” Journal of Development Economics, Vol. 45(December1994), pp. 365–91.“
Public Policy and Saving in Developing Countries,” Journal of Development Economics, Vol. 36(July1991), pp. 89–115.“
Corbo, Vittorio, StanleyFischer, and Steven B.Webb, eds., (ed)Adjustment Lending Revisited: Policies to Restore Growth (Washington: World Bank,1992).
Trade Reforms and Labor Market Adjustment in Morocco,” Labor Markets Workshop (Washington: World Bank,June1994).“
Economic Growth in Latin America,” Journal of Development Economics, Vol. 39(July1992), pp. 59–84.“
Investment and Hysteresis,” Journal of Economic Perspectives, Vol. 6, No. 1(Winter1992), pp. 107–32.“
Investment Under Uncertainty (Princeton, New Jersey: Princeton University Press,1994).
Moderate Inflation,” World Bank Economic Review, Vol. 7(January1993), pp. 1–44.“
Endogenous Growth in Developing Countries with Government-Induced Distortions,” in Adjustment Lending Revisited: Policies to Restore Growth, ed. byVittorioCorbo, StanleyFischer, and Steven B.Webb (ed) (Washington: World Bank,1992).“
Easterly, W., Carlos A.Rodriguez and KlausSchmidt-Hebbel, eds., (ed)Public Sector Deficits and Macroeconomic Performance (New York: Oxford University Press,1994).
On the Sequencing of Structural Reforms,” NBER Working Paper No. 3138 (Cambridge, Massachusetts: National Bureau of Economic Research,October1989).“
Exchange Rates, Inflation and Disinflation: Latin American Experiences,” NBER Working Paper No. 4320 (Cambridge, Massachusetts: National Bureau of Economic Research,April1993).“
Sources of Growth: A Study of Seven Latin American Countries, International Center for Economic Growth (San Francisco: ICS Press,1992).
Saving Trends in Southeast Asia: A Cross-Country Analysis,” IMF Working Paper No. 95/39 (Washington: International Monetary Fund,April1995).“
Workers and Trade Liberalization: The Impact of Trade Reforms in Mexico on Wages and Employment” (unpublished; Cambridge, Massachusetts: Harvard University,1994).“
Role of Macroeconomic Factors in Growth,” Journal of Monetary Economics, Vol. 32(December1993), pp. 485–512.“
Sub-Saharan Africa: Growth, Savings, and Investment, 1986-93,IMF Occasional Paper No. 118 (Washington: International Monetary Fund,January1995).
Public Policies and Private Savings and Investment in Sub-Saharan Africa: An Empirical Investigation,” IMF Working Paper No. 95/19 (Washington: International Monetary Fund,February1995).“
Macroeconometric Model for Developing Countries,” Staff Papers,International Monetary Fund, Vol. 37(September1990), pp. 537–59.“
Dynamic Responses to Policy and Exogenous Shocks in an Empirical Developing Model with Rational Expectations,” in Macroeconomic Models for Adjustment in Developing Countries, ed. by Mohsin S.Khan, Peter J.Montiel, and Nadeem U.Haque (ed) (Washington: International Monetary Fund,1991).“
Horton, Susan, RaviKanbur, and DipakMazumdar, eds., (ed)Labor Markets in an Era of Adjustment,Issues Papers,Vols. 1 and 2,EDI Development Studies (Washington: World Bank,1994).
Husain, Ishrat and RashidFaruqee, eds., (ed)Adjustment in Africa: Lessons from Country Case Studies,World Bank Regional and Sectoral Studies (Washington: World Bank,1994).
International Monetary Fund, “The Experience of Successfully Adjusting Developing Countries,” in World Economic Outlook, October 1992: A Survey by the Staff of the International Monetary Fund,World Economic and Financial Surveys (Washington: International Monetary Fund,October1992).
International Monetary Fund, (1993a), “Convergence and Divergence in Developing Countries,” in World Economic Outlook, May 1993: A Survey by the Staff of the International Monetary Fund,World Economic and Financial Surveys (Washington: International Monetary Fund,May1993).
International Monetary Fund, (1993b), “Domestic and Foreign Saving in Developing Countries,” in World Economic Outlook October 1993: A Survey by the Staff of the International Monetary Fund,World Economic and Financial Surveys (Washington: International Monetary Fund,October1993).
International Monetary Fund, “Why Are Some Developing Countries Failing to Catch Up?” in World Economic Outlook May 1994: A Survey by the Staff of the International Monetary Fund,World Economic and Financial Surveys (Washington: International Monetary Fund,May1994).
International Monetary Fund, World Economic Outlook, May 1995: A Survey by the Staff of the International Monetary Fund,World Economic and Financial Surveys (Washington: International Monetary Fund,May1995).
International Monetary Fund, Fiscal Affairs Department, Unproductive Public Expenditures: A Pragmatic Approach to Policy Analysis,IMF Pamphlet Series, No. 48 (Washington: International Monetary Fund,1995).
Macroeconomic Effects of Fund-Supported Programs,” Staff Papers,International Monetary Fund, Vol. 37(June1990), pp. 195–221.“
Fund-Supported Adjustment Programs and Economic Growth,IMF Occasional Paper No. 41 (Washington: International Monetary Fund,1985).
Capital Flows in the APEC Region,IMF Occasional Paper No. 122 (Washington: International Monetary Fund,March1995).
Khan, Mohsin S., Peter J.Montiel, and Nadeem U.Haque, eds., (ed)Macroeconomic Models for Adjustment in Developing Countries (Washington: International Monetary Fund,1991).
Business Cycle Associated with Exchange Rate-Based Stabilizations,” World Bank Economic Review, Vol. 6(May1992), pp. 279–305.“
Financial Indicators and Growth in a Cross Section of Countries,” Policy Research Working Paper WPS/819 (Washington: World Bank,1992).“
Finance and Growth: Schumpeter Might Be Right,” Quarterly Journal of Economics, Vol. 108(August1993), pp. 717–37.“
Capital Fundamentalism, Economic Development, and Economic Growth,” Carnegie-Rochester Conference Series on Public Policy, Vol. 40(1994), pp. 259–300.“
and others, International Trade Policies: The Uruguay Round and Beyond,Vols. I and II,World Economic and Financial Surveys (Washington: International Monetary Fund,1994).
Thailand: The Road to Sustained Growth,International Monetary Fund (forthcoming).
Macroeconomic Determinants of Growth: Cross-Country Evidence,” Journal of Monetary Economics, Vol. 16(September1985), pp. 141–63.“
The Political Economy of Agricultural Pricing Policy (Baltimore: Johns Hopkins University Press,1992).
A Sensitivity Analysis of Cross-Country Growth Regressions,” American Economic Review, Vol. 82(September1992), pp. 942–63.“
Boom, Crisis, and Adjustment: The Macroeconomic Experience of Developing Countries (New York: Oxford University Press,1993).
Loser, Claudio M., and EliotKalter, eds., (ed)Mexico: The Strategy to Achieve Sustained Economic Growth,IMF Occasional Paper No. 99 (Washington: International Monetary Fund,1992).
and others, Composition of Fiscal Adjustment and Growth,International Monetary Fund, (forthcoming).
A Contribution to the Empirics of Economic Growth,” Quarterly Journal of Economics, Vol. 107(May1992), pp. 407–37.“
Analytical Small-Sample Distribution Theory in Econometrics: The Simultaneous-Equations Case,” International Economic Review, Vol. 23(October1982), pp. 503–33.“
Saving Behavior in Industrial and Developing Countries,” in Staff Studies for the World Economic Outlook, by the Research Department of the International Monetary Fund,World Economic and Financial Surveys (Washington: International Moneary Fund,September1995).“
Measuring the Effect of External Shocks and the Policy Response to Them: Empirical Methodology Applied to the Philippines,” Policy Research Working Paper No. 1271 (Washington: World Bank,March1994).“
Experience with Nominal Anchors,” in IMF Conditionality: Experience Under Stand-By and Extended Arrangements, Part II: Background Papers, ed. by SusanSchadler, (ed)IMF Occasional Paper No. 129 (Washington: International Monetary Fund,September1995).“
Adjustment and Equity in Chile (Paris: Development Center of the Organization for Economic Cooperation and Development,1992).
The Transmission Mechanism for Monetary Policy in Developing Countries,” Staff Papers,International Monetary Fund, Vol. 38(March1991), pp. 83–108.“
A Macroeconomic Simulation Model for India” (unpublished; Oberlin College,February1993).“
Aid and Power: The World Bank and Policy-Based Lending (London: Routledge,1991;2nd ed.,1995).
A New Database on Physical Capital Stock: Sources, Methodology and Results,” Revista de Andlisis Economico, Vol. 8, No. 1(1993), pp. 37–59.“
Ghana, 1983-91,” in Adjustment for Growth: The African Experience,International Monetary Fund (forthcoming).“
Resilience and Growth Through Sustained Adjustment: The Moroccan Experience,IMF Occasional Paper No. 117 (Washington: International Monetary Fund,January1995).
Saving Behavior in Low-and Middle-Income Developing Countries: A Comparison,” Staff Papers,International Monetary Fund, Vol. 43(March1966), pp. 38–71.“
Mexico: Reform and Productivity Growth,” World Bank Report No. 11823-ME (Washington: World Bank,May1994).“
Liberalizing Foreign Trade in Developing Countries: The Lessons of Experience (Washington: World Bank,1990).
Exact Small Sample Theory in the Simultaneous Equation Model,” in Handbook of Econometrics, Vol. I, ed. by ZviGriliches and Michael D.Intrilligator (ed) (Amsterdam: North-Holland,1983).“
Irreversibility, Uncertainty and Investment,” Journal of Economic Literature, Vol. 29(September1991), pp. 1110–49.“
Real Effects of Exchange-Rate-Based Stabilization: An Analysis of Competing Theories,” in NBER Macroeconomics Annual, 1995 (Cambridge, Massachusetts: MIT Press,1995).“
Nominal Interest Rates, Consumption Booms, and Lack of Credibility: A Quantitative Examination,” Journal of Development Economics, Vol. 46(April1995), pp. 357–78.“
Employment and Wage Effects of Trade Liberalization: The Case of Mexican Manufacturing,” Policy Research Working Paper No. 1524 (Washington: World Bank,1995).“
Adjustment and Equity in Ghana (Paris: Development Center of the Organization for Economic Cooperation and Development,1992).
Wages, Profits and Macroeconomic Adjustment: A Comparative Study,” Brookings Papers on Economic Activity(2/1979), pp. 269–332.“
Demographic Dynamics and the Empirics of Economic Growth,” Staff Papers,International Monetary Fund, Vol. 42(June1995), pp. 398–410.“
Nonlinear Effects of Inflation on Economic Growth,” Staff Papers,International Monetary Fund, Vol. 43(March1996), pp. 199–215.“
Private Saving in IMF Arrangements,” in IMF Conditionality: Experience Under Stand-By and Extended Arrangements, Part II: Back-ground Papers, ed. by SusanSchadler, (ed)IMF Occasional Paper No. 129 (Washington: International Monetary Fund,September1995).“
Economic Adjustment in Low-Income Countries: Experience Under the Enhanced Structural Adjustment Facility,IMF Occasional Paper No. 106 (Washington: International Monetary Fund,September1993).
Recent Experiences with Surges in Capital Inflows,IMF Occasional Paper No. 108 (Washington: International Monetary Fund,December1993).
IMF Conditionality: Experience Under Stand-By and Extended Arrangements, Part I: Key Issues and Findings,IMF Occasional Paper No. 128 (Washington: International Monetary Fund,September1995).
Schadler, Susan, (ed) ed., IMF Conditionality: Experience Under Stand-By and Extended Arrangements, Part II: Background Papers,IMF Occasional Paper No. 129 (Washington: International Monetary Fund,September1995).
Public Policy and Private Saving,” in Adjustment Lending Revisited: Policies to Restore Growth, ed. by VittorioCorbo, StanleyFischer, and Steven B.Webb (ed) (Washington: World Bank,1992).“
Serven, Luis and AndrésSolimano, eds., (ed)Striving for Growth After Adjustment: The Role of Capital Formation, World Bank Regional and Sectoral Studies (Washington: World Bank,1994).
Senegal, 1978-93,” in Adjustment for Growth: The African Experience,International Monetary Fund (forthcoming).“
and others, Best Practices in Trade Policy Reform (New York: Oxford University Press,1991).
Thomas, Vinod, AjayChhibber, M.Dailami, Jaime deMelo, eds., (ed)Restructuring Economies in Distress: Policy Reform and the World Bank (New York: Oxford University Press,1991).
Productivity Growth in Thailand” (unpublished; Bangkok: National Economic and Social Development Board and Thailand Development Research Institute,1994).“
Macroeconomic Policies, Crises, and Long-Term Growth in Indonesia, 1965-90,World Bank Comparative Macroeconomic Studies (Washington: World Bank,1994).
World Bank, “Adjustment Lending Policies for Sustainable Growth,” Policy and Research Series, No. 14 (Washington: World Bank,1990).
World Bank, World Development Report 1991 (New York: Ox-ford University Press,1991).
World Bank, “Adjustment Lending and Mobilization of Private and Public Resources for Growth,” Policy and Research Series, No. 22 (Washington: World Bank,1992).
World Bank, (1993a), Ghana 2000 and Beyond: Setting the Stage for Accelerated Growth and Poverty Reduction (Washington: World Bank,February1993).
World Bank, (1993b), The East Asian Miracle: Economic Growth and Public Policy (New York: Oxford University Press,1993).
World Bank, (1994a), Adjustment in Africa: Reforms, Results, and the Road Ahead (New York: Oxford University Press,1994).
World Bank, (1994b), “Mexico: Country Economic Memorandum—Fostering Private Sector Development in the 1990s,” Report No. 11823-ME (Washington: World Bank,May1994).
World Bank, (1994c), “Kingdom of Morocco—Preparing for the 21st Century: Strengthening the Private Sector in Morocco,” Report No. 11894-MOR (Washington: World Bank,June1994).
World Bank, World Development Report 1995 (New York: Ox-ford University Press,1995).
World Bank, Operations Evaluation Department, Trade Reforms Under Adjustment Programs (Washington: World Bank,1992).
This work is referenced in the text and footnotes.
In practice, no single model can capture all of the influences at work. One approach that attempts to generate “representative” estimates, from a panel of developing countries, for key macroeconomic parameters is described in Haque and others (1990) and Haque and Montiel (1991). Models with more complex production structures are needed to analyze many supply-side issues; data limitations prevent the estimation of such models for most developing countries, so a mixture of imposed and estimated parameters are typically used in simulations. This is the approach taken in Montiel (1993) and Bourguignon and Morrisson (1992). Also, see Khan and others (1991).
This study does not address explicitly the causes of the 1994 Mexican financial crisis.
The role of macroeconomic policies and reforms in growth has been the subject of many studies—see Appendix I. An early contribution on the role of IMF-supported adjustment programs is Khan and Knight (1985).
For this reason, no countries were included in which adjustment policies were especially weak. Nor are the transition economies represented, since their growth performance has been subject to a number of quite different influences.
Asilis and Milesi-Ferreti (1994) provide a survey of (typically highly stylized) studies of how initial economic conditions, institutions, and political systems may interact with policy choices to influence the political sustainability of reforms. Uncertainty and conflict over the distributional consequences of adjustment generally play a key role in explaining delays in implementing stabilization in such models, suggesting that adjustment is more likely to be delayed the more polarized the society.
Disentangling the underlying growth in potential output from cyclical developments and supply shocks is always difficult and can be especially complicated in the case of severe recessions (Chile, 1975 and 1982) or where supply shocks are large and frequent (Morocco and Senegal). TFP measures can also be a misleading indicator of underlying productivity developments if major structural changes make part of the capital stock obsolete. The growth accounting exercises involve imposing a common production function on all of the countries as well as a series of assumptions concerning the initial level of the capital stock and rate of depreciation; see the footnotes to Table 2 for details. The estimates for TFP in Chile, Mexico, and Thailand are broadly comparable to those reported in Bosworth and others (1994), Elias (1992), and Tinakorn and Sussangkarn (1994)), respectively. A set of alternative estimates was also prepared using a production function that includes human capital, following an approach similar to Mankiw, Romer, and Weil (1992), but allowing for differences in efficiency because of changes in the age profile of the population. Both of the estimated TFP measures tend to overestimate cyclical movements in productivity (for example, because of lack of data on changes in man-hours worked).
Sarel (1995) estimates an age-related productivity structure which suggests that, on average, workers reach peak productivity between their thirties and fifties.
In Mexico, the current methodology for estimating the national accounts, which has not been updated since 1980, is also thought to introduce a downward bias in GDP growth since oil has an excessive weight in total output.
Because conditions prior to the two adjustment periods in Chile were so different, they are discussed separately. Chile I refers to the adjustment period that began in 1974 following the advent of the Pinochet regime and Chile II refers to the stabilization period following the collapse of the exchange-rate-based stabilization strategy in 1982.
Senegal has access to the common pool of reserves shared by all the members of the CFA franc zone.
There are signs that, prior to the recent crisis, the reforms were beginning to yield significant gains, most notably in the manufacturing sector, where average output per worker rose at an annual rate of 6 percent during the period 1988–94.
These issues are discussed in more depth in Section VIII.
The results of Bruno and Easterly (1995) suggest that growth falls sharply during episodes of high (which they define as above 40 percent per annum) inflation, and generally recovers strongly after inflation is stabilized. Sarel (1996) suggests that the negative influences of inflation on growth begin to be significant at lower inflation rates (that is, above 8-10 percent per annum).
The adjusted R-squared was typically in the range 0.1 to 0.3.
The approach follows Bruno and Easterly (1995). A more appropriate counterfactual than the world average would be the post adjustment performance of each country relative to what it would have been in the absence of adjustment; in practice, however, it was not possible to estimate such counterfactuals with available information. The choice of starting year for particular subperiods was determined by the timing of implementation of major adjustment policies and was often associated with a major adverse shock. Such factors may also have an independent effect on growth; therefore estimates of the effects of policies on growth derived on the basis of the adjustment periods may be affected by reversion-to-trend biases. These issues are likely to be of greater concern for short-run changes in growth than for changes that persist over longer periods.
A positive growth differential in Chart 9 implies that growth was above the world average after controlling for various determinants of growth. Owing to data limitations on proxies for structural policies, only macroeconomic variables were examined in this part of the exercise. The change in growth differential between the two estimates cannot be interpreted as representing the overall impact of policies on growth. Policies also act through investment, which is included independently in the equation. The adjusted R-squared was generally in the range of 0.3 to 0.4.
These findings are confirmed by Bruno and Easterly (1995); they show that in most countries experiencing a high-inflation crisis, growth was much lower than the world average during the period of high inflation and that this differential was worse than in the precrisis period. Furthermore, most countries that reduced inflation experienced a strong rebound in their growth differential after the crisis. Mexico stands apart for the absence of a significant catching up, although its growth rate did recover to about the “world average” (after controlling for other influences).
In the absence of reliable measures of capacity utilization for all countries, the output gap is approximated by the deviation of actual from trend real GDP. Trend GDP is estimated using a Hodrick-Prescott filter—a univariate trend extraction algorithm—on actual real GDP over the period 1970-93. The filter estimates trend GDP by minimizing the variation of actual GDP around a trend, subject to certain assumptions on the variance of the cyclical component relative to the trend component. Such measures can only be approximate and are subject to a number of potential pitfalls. For example, the estimates tend to overstate the size of output gaps in years prior to a sustained acceleration in output. Thus, the size of the output gap in Thailand in 1985–86 may be overstated. Generally, however, the estimates correspond reasonably closely with independent measures of capacity utilization for those few countries (for example, Chile, in Bosworth and others (1994)) where such measures are available.
Two indicators, albeit imperfect, that can be used to examine the direct impact of fiscal policy on aggregate demand are changes in the primary fiscal balance and a fiscal impulse measure. The primary balance avoids attributing demand effects to interest payments on the large stock of external debt. The fiscal impulse measure attempts to distinguish the cyclical and discretionary components of fiscal adjustment, but “cyclical” revenue influences are likely to also reflect supply and terms of trade shocks, particularly in Ghana, Morocco, Senegal, and Mexico (the discovery of oil reserves). In practice, for the period under review, the two measures generally reveal similar patterns. Neither measure captures the potential expansionary effects that may result from expectations of lower public debt burdens in the future.
The fiscal data throughout this study relate to the consolidated central government in all countries except for Mexico (nonfinancial public sector) and Thailand (budgetary central government). For India, the central government accounts are net of tax revenues collected on behalf of the states.
See Mackenzie and others (forthcoming), Section I.
See the discussion in Box 3 in Section V.
In Ghana, under the Economic Recovery Program, inflation was reduced at the same time as the official exchange rate was massively depreciated. This depreciation did not have a large inflationary impact because the bulk of transactions were already taking place at the parallel market rate and because the nature of the economy, with its relatively small formal sector, meant that real wage rigidity was less of a factor.
Chart 13 distinguishes between predominantly market borrowers and other countries because when countries benefit from substantial concessional external financing a rising debt-to-GDP ratio would not necessarily signal a potential problem, provided the concessional assistance could be expected to continue over the medium term.
The assessment is based on whether fiscal policy was likely to be judged sustainable at the time it was adopted, rather than with the benefit of hindsight; the calculation effectively assumes static expectations about the future path of relevant variables.
See, for example, the discussion on Morocco in Nsouli and others (1995).
Real credit is measured in terms of the contemporaneous rate of inflation of the GDP deflator.
Domestic and international interest rates may, however, diverge for reasons other than financial policies, such as exchange rate expectations, and differences in tax systems (see Bennett (1995)). Moreover, when the rate of inflation is high and variable, any measure of the real interest rate can only be approximate.
Real interest rates fell after a substantial increase in the primary fiscal surplus had been achieved and the Brady agreement for the restructuring of external debt was reached.
Public sector wages were an important benchmark for formal sector wages. See Tahari and others (forthcoming).
The comparison of Chile (1974–75) with Mexico (1987) also illustrates the influence of the nominal anchor choice on the timing of any downturn in economic activity associated with the disinflation strategy: money-based disinflation programs are typically associated with an early downturn and exchange-rate-based programs with a late recession. See Calvo and Végh (1992) and 1993), Kiguel and Liviatan (1992), and Rebelo and Végh (1995).
See, for example, Edwards (1993).
The causality between investment and saving (and growth) can, however, operate in both directions.
The concept of investment pause is used here and in recent studies (Serven and Solimano (1994)) to indicate the reaching of a plateau after the initial decline, during which period neither a noticeable recovery nor a further decline takes place. The length of the pause cannot be simply equated to the time between a peak investment rate and the subsequent return to that rate because the initial peak could be associated with unsustainable macroeconomic conditions, as in Chile or Mexico prior to the 1982 crisis.
Private investment in India during the 1970s and most of the 1980s followed a similar pattern.
The dependent variable was the share of private (or non-government) investment in GDP. The estimated equations included as determinants: (1) the growth of aggregate economic activity, lagged to reduce simultaneity problems; (2) financial variables, notably real interest rates and the volume of credit to the private sector; (3) public investment, to assess complementarity versus substitutability with private investment; (4) the availability of foreign exchange; (5) proxies for uncertainty (volatility of selected macro variables) and other indicators of macroeconomic instability such as inflation and debt and debt-service ratios; and (6) measures of external shocks. In general, the equations tracked private investment developments reasonably well, albeit with noticeable differences in goodness of fit—as the adjusted R-squared ranged from 0.6 to 0.9 across countries. The comparability of results is complicated by differences in the treatment of public enterprise investment, which is included in the dependent variable in the cases of Chile and Senegal.
For Thailand, economic activity in partner countries was a major influence on private investment, suggesting that the investment boom that began in the second half of the 1980s was due in part to broader regional developments.
The initial specification of investment equations included both credit volumes—capturing credit rationing effects—and real interest rates since in several instances restrictions affecting both quantities and prices in the market for credit were in place for at least part of the sample. Constraints on the availability of credit proved important in India, Mexico, and less so in Bangladesh and Morocco, while proxies for the user cost of capital—that is, real ex-post interest rates or the relative price of capital goods—were significant in most cases except Senegal. In countries adopting financial liberalization reforms (Mexico, for example) the influence of interest rates gained greater prominence over time.
In particular, the weak response of private investment in Bangladesh appears to have been caused in part by increasing bank spreads and credit rationing, indicating a problem of weak bank balance sheets.
Estimates of the impact of real lending rates on investment proved insignificant in the case of Chile. Expectations of an imminent devaluation and continued lending in support of nonperforming loans appear to have been the dominant factors prior to the 1982 crisis.
In Morocco, the complementarity effects were strongest in the 1980s when the reduction in the share of government investment may have contributed to the pause and weak recovery of private investment. Crowding-out effects appear to have been stronger during the period of rapid public investment expansion in the 1970s.
For example, Easterly, Rodriguez, and Schmidt-Hebbel (1994) found that crowding-out effects predominated in about half of the countries covered, whereas earlier cross-country studies (Serven and Solimano (1994)) found complementarity between public and private investment. Also, estimates of investment equations in the country case studies indicate that crowding-out effects may vary in strength over time. The discussion in the companion study (Mackenzie and others (forthcoming)) points to some evidence suggesting that public infrastructure investment (in particular investment in transport and communication) may complement private sector investment.
A distinction between infrastructure and other public investment was possible for India. When both variables were included in the investment equation, they had negative signs although the infrastructure term was not statistically significant.
For India, estimates of real rates of return to different categories of investment suggest that public sector investments in manufacturing have yielded much lower rates of return than either public infrastructural investment or private investment in manufacturing.
There is also considerable evidence that saving rates rise with per capita income levels, especially during the transition from low income levels once subsistence needs are satisfied; see Ogaki and others (1996).
Empirical studies generally reject full Ricardian equivalence (a proposition implying that changes in public saving lead to equal and offsetting changes in private saving), but find that an increase in public saving results in a partial offset in private saving. Estimates of the offset coefficient are typically in the 0.4–0.6 range; see Savastano (1995) and Corbo and Schmidt-Hebbel (1991).
The issue is also complicated by the difficulty of distinguishing between the inputs (that is, spending) and outputs of the government sector and by intergenerational concerns (for example, changes in unfunded pension liabilities). See International Monetary Fund, Fiscal Affairs Department (1995).
Taxes affect saving by altering the rate of return enjoyed by savers. The evidence on the impact of taxes on saving, discussed in the companion study (see Mackenzie and others (forthcoming)), is that tax policy probably does not have a marked effect on the total but can affect substantially its composition. Tax schemes favoring particular classes of saving can entail substantial distortions.
Masson and others (1995); the panel data covered 64 developing countries over the period 1970–93, and included all countries of the present study except Senegal.
Hadjimichael and Ghura (1995) used a panel of 41 sub-Saharan African countries (including Ghana and Senegal) over 1986-92 to investigate the effectiveness of public policies in stimulating private saving and investment. They found that policies that kept inflation low, reduced macroeconomic uncertainty, promoted financial deepening, and increased public saving were conducive to higher national saving. They also found that policies leading to a reduction in the external debt burden significantly increased measured national saving.
Data limitations complicate the analysis of saving, which is typically measured as a residual and is beset by classification, valuation, and measurement problems. Public saving data for Bangladesh, Chile, India, and Thailand include that of public enterprises; in the four other countries, public enterprise saving is included with that of the private sector.
The reversal of the saving gains in Ghana during the early 1990s reflected, in part, fiscal policy slippages that led to lower public saving.
See, for example, Faruqee and Husain (1995).
The Chilean privatization program resulted in a large reclassification of firms from the public to private sector, which complicates judgments on the relative contributions of each sector to the improvement in national saving. Partial write-offs of private sector domestic debt also complicate the estimates.
The privatized plans contributed the equivalent of 2–3 percentage points of GDP to national saving throughout the 1980s; the net impact is harder to estimate because of reduced saving elsewhere, especially in the public sector. Bosworth and others (1994) contains a detailed discussion of the reform.
At the start of the extended arrangement, a cumulative improvement in private saving of about 1 percentage point of GDP was targeted for 1989–92, whereas the eventual outcome was a cumulative decline of 10 percentage points. Shortfalls from the revised targets prepared under each annual segment of the arrangement were also very large.
In addition, the progress in reducing inflation and the resulting drop in the inflation tax, as well as completion of the Brady debt consolidation deal in late 1989, probably raised expected permanent incomes. The possible factors behind the decline in Mexico’s private saving are discussed at greater length in Savastano (1995).
Some of the available evidence is discussed in Appendix I, section on External Financing and the External Economic Environment.
Increased use of foreign saving would generally raise GDP more than GNP because of the need to service the foreign borrowing; however, a sufficiently long time series for all eight countries is only available for GDP. The correlation between the average per capita growth rate and changes in net external financing as a share of GDP over two periods (1983–88 and 1989–93) was about 0.3 for the eight countries; similar results have been obtained for broader groups of developing countries.
The panel regressions for a broader group of countries discussed in Section V suggest that, on average and after controlling for other influences including the terms of trade, about 40–50 percent of any increase in foreign saving goes to raise domestic consumption.
The periods shown in the chart were chosen because they represent the times of largest changes in net external financing for many of the eight countries. In each case, the computed change is with respect to the average for the preceding five-year period. The correlation between changes in net external financing and changes in domestic investment for the eight countries is much higher (about 0.7 and 0.5, respectively, for the two periods) than the correlations with growth, but causation is likely to run in both directions.
See also the discussion in Loser and Kalter (1992), Section VII. In addition. Claessens, Oks, and van Wijnbergen (1994) found a positive effect on private investment in Mexico of a reduced variance in debt-service obligations following the bank deal.
Growing arrears problems and a rundown in reserves forced a collapse in import volume during 1982–83 to barely half the level of three years earlier. The largest part of the external “financing” during 1980–82 consisted of a buildup in arrears. See Nowak and others (forthcoming).
PAMSCAD represented a comprehensive attempt at addressing social strains created by macroeconomic adjustment over the period 1988-89. It included financing of infrastructure rehabilitation, public works projects, and addressing the basic needs of the poor. PAMSCAD was to be fully funded by foreign aid. However, because of slow disbursements and implementation difficulties less than one fifth of the pledged donor assistance materialized by mid-1992. Mainly because of this financing shortfall, but also as a result of inadequate cost management and targeting, the goals of PAMSCAD were less than fully realized. See Nowak and others (forthcoming) and Roe and Schneider (1992).
In Bangladesh, persistent shortfalls from program targets for external financing occurred in the second adjustment period, but appear to have reflected difficulties in implementing the public investment program rather than a shortage of potential external financing.
Data on the nature of the inflows vary in quality and disaggregation and definitions are not uniform across countries. Recorded FDI was, on average, in the range of 11/2percent to 2 percent of GDP in Chile, Mexico, and Thailand during the recent periods of peak capital inflows. However, Thailand’s data tend to understate substantially the share of FDI, much of which appears to be recorded as inflows into the banking system. See also Bercuson and Koenig (1993). FDI also rose to these levels in Morocco during 1991-93, in part reflecting foreign acquisitions of privatized assets. The capital inflows to India have been dominated by portfolio equity capital (see Chopra and others (1995)).
Available evidence is discussed in Appendices I and II.
Mackenzie and others (forthcoming).
Such classifications involve a considerable measure of judgment. The rankings do not cover less quantifiable factors such as regulatory complexities, which were the most severe in Bangladesh and India. Morocco was not included in the Agarwala sample, but available information suggests that it falls into the category of moderate initial distortions.
Average output per worker in Mexican manufacturing rose at an annual average rate of 6 percent during the period 1988-94.
A detailed analysis of the impact of trade reforms in IMF-supported programs is contained in Kirmani and others, Vol. II (1994).
These tax reforms typically included a broadening of the base of domestic sales taxation by eliminating exemptions and preferential rates and/or through the introduction of a value-added tax; see Mackenzie and others (forthcoming).
Changes in export performance as indicated by the growth in export market shares cannot be attributed exclusively to the reforms unless it is assumed that market shares would have remained unchanged otherwise, which is unlikely to be the case.
The link between more developed financial intermediation and growth has also been attributed to the idea that the former promotes innovative activity. King and Levine (1993) present cross-country evidence that higher levels of financial development are robustly correlated with higher rates of growth, capital formation, and efficiency improvements. The discussion in this section focuses on banking system reforms. Other financial sector reforms, such as the development of bond and equity capital markets, also help to improve the efficiency of financial intermediation. Atje and Jovanovic (1993) find evidence that the development of stock markets has a significant effect on growth.
The efficacy of financial sector reforms in mobilizing financial saving and improving the efficiency with which resources are intermediated is difficult to measure directly. Typical summary indicators include the size of the financial system measured by the ratio of broad money to GDP, the private sector’s share of total credit as a measure of asset distribution, and the spread between deposit and lending rates. King and Levine (1992) and 1993) find a positive association between the allocation of credit to the private sector and the rate of investment and productivity growth.
Financial repression here refers to cases where real interest rates were significantly negative over long periods of time. Distortions in Mexico’s financial system prior to 1982 were relatively limited; however, financial repression emerged after 1982 when inflation surged and interest rate ceilings persisted, and the banking system was nationalized.
In India, reforms of the nonbank financial sector (including equity and insurance markets) have been more extensive. See Chopra and others (1995).
This assessment is based on World Bank (1994b).
Segmentation does not imply independence; shocks in one submarket can have significant wage and employment repercussions in the others.
For example, stringent job security regulations in India (including the requirement in most cases to obtain prior government consent for layoffs) are estimated to reduce formal employment in 35 industries by 18 percent (World Bank ((1995)). In Morocco, temporary employment in manufacturing grew two and a half times faster than permanent employment in 1984-90 largely because of restrictions against layoffs (World Bank (1994))c. See also Tahari and others (forthcoming) for a discussion of regulations in Senegal.
Because of data limitations, this analysis could only be carried out for Chile, Mexico, Morocco, and Thailand.
In terms of relative unit labor costs in manufacturing, the cumulative real appreciation during 1987-93 amounted to about 80 percent—albeit from a level that was substantially depreciated. This returned the real exchange rate to a level at or above that prevailing in 1983. A similar trend is revealed by other measures of the real exchange rate. Assessments of Mexico’s competitiveness during this period have been the subject of considerable controversy, also due to the fact that the marked opening up of the economy, in response to trade and other structural reforms as well as the increasing integration of the North American economies, led to widely divergent movements in trade indicators. Thus, the real exchange rate appreciation was not enough to discourage strong export growth, but did contribute to a very large increase in import penetration.
The improvement in relative unit labor costs was also partly due to broader regional developments, especially rising labor costs in Japan, Korea, and Taiwan Province of China. See Kochhar and others (forthcoming).
These calculations are based on formal sector employment and total manufacturing output. Over time, the share of the formal sector in total employment appears to have increase reflecting changes in the structure of the economy and deregulation. Therefore, productivity growth may be understated and changes in unit labor costs and the wage gap overstated.
For instance, model simulations for Bangladesh indicate that when formal sector wages are rigid, a real depreciation raises real wages in the formal sector and reduces them in the informal sector because unemployed formal sector workers expand the supply of informal sector labor (World Bank ((1995)). Also, there are signs that real wages in India’s informal sector fell during the 1991-92 adjustment period, although observers attribute much of this decline to a relatively poor harvest.
The lack of adequate data makes it difficult to analyze the effect of adjustment policies on employment and unemployment (particularly for Ghana and Senegal). Because data on informal markets are unavailable, measured employment in most cases refers only to formal sector employment.
Evidence on the impact of trade liberalization on labor markets is mixed. In a review of the literature, Agénor (1995) notes that in Morocco, trade reform is estimated to have had a small, but notable, impact on employment and wages in manufacturing during 1984-90, with pronounced sectoral shifts in employment (Currie and Harrison (1994)). The evidence on Mexico, appears inconclusive: according to Revenga (1995), tariff reduction (of about 10 percentage points) between 1985-88 reduced manufacturing employment by 2-3 percentage points and increased average wages, with marked intraindustry shifts in employment; however, Feliciano (1994) finds no employment impact and an increase in wage dispersion rather than an effect on average wages. In both Mexico and Morocco, the increase in wages may have reflected a change in the composition of employment toward highly skilled high-wage workers.
Model-based simulations for Bangladesh indicate that an increase in output-employment elasticities (for example, through labor market reform) would significantly reduce the extent of underemployment (World Bank ((1995)).
For instance, the (World Bank (1995)) estimates that underemployment in Bangladesh is widespread and equivalent to having more than one fourth of the labor force unemployed. In India, Morocco, and Thailand, open unemployment is concentrated on younger and better educated workers (who are supported by other means), while in Chile a large proportion of unemployed in the 1980s were skilled workers.
Mackenzie and others (forthcoming).
Mackenzie and others (forthcoming).
Public enterprise reforms are discussed in the companion study by Mackenzie and others (forthcoming).
A number of other factors, including the fiscal policy stance, were also important; these two episodes are discussed at greater length in Section IV.
Mackenzie and others (forthcoming).
Khan (1990), using a modified control-group approach that attempts to distinguish between the impact of macroeconomic policies and the Fund program per se, concludes that achievement of macroeconomic stability had a positive impact on long-run growth whereas, after allowing for an estimate of the policy changes that would have taken place even without Fund arrangements, the growth rate declined in the program year. This adverse effect diminished when the time horizon was extended beyond the program year. Mosley and others (1991) identify a weak positive association between Bank-supported programs and GDP growth. Using a similar approach, Conway (1994) concludes that participation in Fund-supported programs had a negative same-year effect on growth and domestic investment, but that the lagged effects were both positive and larger than the same-year effects.
For example, see De Gregorio (1992). Bruno and Easterly (1995) show that this negative relationship between growth and inflation is much more robust at high rates of inflation (that is, annual rates above 40 percent).
The long transition from severe distortions to sustained growth in Indonesia is discussed in Woo and others (1994).
Branson and Jayarajah (1995) reach the same conclusion.
Serven and Solimano (1994) and the references therein. The results reported in Hadjimichael suggest that, on balance, government investment has generally complemented private investment in sub-Saharan Africa in recent years. However, investment by public enterprises—potentially the most important source of direct crowding-out effects—were included in private investment for the purposes of the study.
The conclusions that fiscal adjustment induces an offset, but only a partial one, in private saving is supported by a large number of econometric studies. See Savastano (1995) and Masson and others (1995) for a discussion of this evidence. The method of raising public saving also appears to matter, with a cut in current spending inducing a smaller decline in private saving than would an increase in tax revenues; see Corbo and Schmidt Hebbel (1991).
A more extensive discussion of the theoretical and empirical evidence on the role of public sector reform in promoting growth is contained in Mackenzie and others (forthcoming).
Some of this evidence is discussed in Section III.
Countries with Bank-supported structural adjustment programs generally also had arrangements with the IMF.
For example, only a handful of countries in sub-Saharan Africa have succeeded in reducing the number of civil service employees by more than 5 percent since they began structural adjustment (Africa(World Bank, 1994a), p. 123).
A detailed discussion of recent experiences with nominal anchors in IMF-supported programs is contained in Mecagni (1995). See also Dornbusch and Fischer (1993), Edwards (1993), and Reinhart and Végh (1995).
However, the boom-bust cycle has also occurred in countries that ran fiscal surpluses (for example, Chile in the late 1970s and Israel in 1985-86).
See, for example, Boone (1994).
International Monetary Fund (1993b). The results quoted are for the period 1984—93; similar results have been observed for different time periods.
See International Monetary Fund (1995). The Africa and Hadjimichael studies reached a similar conclusion.
Edwards (1989) provides a succinct discussion of these issues.
Easterly (1992) presents an endogenous growth model that illustrates how different magnitudes of distortions can affect growth.
Empirical studies investigating the links between trade liberalization and growth are summarized in Kirmani, Appendix I.
Papageorgiou and others (1990) reach the same conclusion.
See also Krueger and others (1991).
The approach follows Fischer (1993). TFP is derived by subtracting the contributions of capital and labor accumulation from real GDP growth, assuming factor shares of 0.4 and 0.6 for capital and labor, respectively. Various alternative measures of TFP growth were all highly positively correlated.
Given the inherent difficulty of constructing simple measures of policy variables, it is necessary to use proxies that will also be affected by non-policy-related factors. The exercise therefore should not be interpreted as a structural equation explaining growth. Also, the proportion of the variation in growth accounted for by such a regression is quite low—adjusted R2s are typically in the range of 0.1 to 0.3.
The difficulty of deriving simple measures to summarize the complex nature of structural distortions and reforms implies that it is often not possible to identify statistically robust relationships between growth and structural policies. The estimated partial correlations are typically highly sensitive to the proxies used, as well as to the coverage of countries and time periods. See Levine and Renelt (1992). However, microeconomic evidence based on ex-post rates of return on projects supported by the World Bank also suggests that trade distortions dampen the efficiency of investment. See World Bank (1991).
For the purposes of international comparison, growth rates for this exercise are measured at purchasing power parity (PPP) adjusted prices. For most of the eight countries, it does not make much difference whether growth is measured at national or international relative prices. However, for a few countries, there do appear to be sizable differences during selected periods. The differences are often most marked in periods when the trade and exchange system was especially distorted (for example, Bangladesh during 1970-79) because growth is weighted in favor of protected sectors that benefit from high domestic relative prices, or during periods following substantial exchange or trade reforms (for example, Ghana after 1983 and Mexico in 1988-92) as sectors with a comparative advantage expand more rapidly. Since growth at national prices directly affects welfare and is how the effects of policies are usually assessed, it is the measure used for the panel estimates discussed later in this appendix.
This specification, which follows Barro (1989) and the World Bank (1993b), is generally associated with tests of theories of endogenous growth, which are generally predicated on the assumption that “knowledge” and factor accumulation interact to produce increasing returns to scale. Therefore, an economy can be on various possible long-run growth paths, depending upon the level of investment and activity in areas that benefit most from such factors. However, it is also consistent with the transitional dynamics of a neoclassical growth model that incorporates human capital. The results should be interpreted with caution, however, as the estimated equation explains less than one half of the variation in per capita income growth. Moreover, the approach requires the assumption that all countries have the same production function. The quality of the data for several countries is also poor, and it is impossible to identify the direction of biases so introduced.
Because of the severe depth of the 1975 and 1983 recessions in Chile, growth rates in subsequent years were heavily influenced by changes in the degree of capacity utilization. However, such factors would be less likely to influence long-term average growth.
As noted earlier, the subperiods chosen here do not necessarily correspond to the timing of IMF-supported programs.
The constant is omitted from the regression and the change in the terms of trade is added to the list of right-hand-side variables. The inclusion of a broad group of other countries in the regression helps to control for other exogenous influences common to all countries (for example, changes in world economic conditions), but does not correct for country-specific reversion-to-trend effects.
Senegal’s rather variable growth performance appears to be largely the result of the impact of supply shocks. See Tahari and others (forthcoming).
It was not possible to include variables proxying structural policies in this part of the exercise because the available measures were typically computed as period averages or at a limited number of points in time over the sample period. It should also be recognized that the interpretation of this policy-augmented growth regression is not straightforward, as policies also affect investment, which is included as a regressor; one possible interpretation of the coefficients on the policy variables in this regression is that they capture, broadly speaking, the influence of policies on productivity.
Since the differentials A and B are derived from different specifications of the growth equation, a direct comparison of the two is not possible.
The intertemporal budget constraint for the public sector can be described as b = (r–n) b + d – s where b, d, and s represent public debt, the primary deficit, and seigniorage revenues (including the inflation tax), respectively (all as shares of GDP), r represents the average real interest rate on debt, and n is the real GDP growth rate (see Anand and van Wijnbergen (1989)). The aim is to assess whether fiscal policy was sustainable at the time, rather than with the benefit of hindsight. Consequently, the calculations are based on a three-year moving average of actual values of interest rates and growth rates and assume that the real exchange rate is expected to be constant. The estimate of “low-inflation” seigniorage used in the calculations was derived from a quadratic function linking seigniorage to inflation that was estimated by Easterly and others (1994) on the basis of cross-country data and assuming an annual inflation rate of 5 percent. This measure may underestimate the potential for noninflationary seigniorage in countries such as India, where currency holdings are relatively large.
The sustainability calculations for Bangladesh are based on external debt, because of the paucity of data on domestic debt. However, rough estimates of the size of domestic public debt (at about 10 percent of GDP) suggest that its inclusion would not fundamentally alter the results.
In Chile, fiscal sustainability has not really been an issue except in 1982-85, when domestic public debt increased sharply as a result of recapitalization of the banking system. A similar point applies to Bangladesh, where there have been several bank recapitalizations in recent years.
In 1993, the shares of external debt on concessional terms in total external debt were Bangladesh, 98 percent; Ghana, 83 percent; India, 53 percent; and Senegal, 70 percent. If all concessional lending were to be phased out and replaced by borrowing at market terms, then, based on the 1993 debt composition, interest payments would increase as follows: Bangladesh, 1.6 percentage points of GDP; Ghana, 1.4 percentage points; India, 0.5 percentage point; and Senegal, 2 percentage points.
Low or rapidly decaying values of the sample autocorrelation function indicated that nonstationarity was in no instance a problem. The small sample size prevents a reliable use of formal unit root tests.
This modeling strategy potentially minimizes omitted variable bias, although accuracy may be affected by highly collinear regressors. The general-to-specific approach faced inevitable degrees of freedom constraints in instances where only a small sample of observations was available.
Available small sample studies show that when an estimated equation incorrectly omits relevant explanatory variables, ordinary least squares are less sensitive to misspecification than simultaneous estimation methods and therefore may be a preferable estimation technique. See Mariano (1982) and Phillips (1983).
Standard errors were computed using White’s heteroskedasticity-consistent variance-covariance estimator.
Recent Occasional Papers of the International Monetary Fund
139. Reinvigorating Growth in Developing Countries: Lessons from Adjustment Policies in Eight Economies, by David Goldsbrough, Sharmini Coorey, Louis Dicks-Mireaux, Balazs Horvath, Kalpana Kochhar, Mauro Mecagni, Erik Offerdal, and Jianping Zhou. 1996.
138. Aftermath of the CFA Franc Devaluation, by Jean A.P. Clément, with Johannes Mueller, Stéphane Cossé, and Jean Le Dem. 1996.
137. The Lao People’s Democratic Republic: Systemic Transformation and Adjustment, edited by Ichiro Otani and Chi Do Pham. 1996.
136. Jordan: Strategy for Adjustment and Growth, edited by Edouard Maciejewski and Ahsan Mansur. 1996.
135. Vietnam: Transition to a Market Economy, by John R. Dodsworth, Erich Spitäller, Michael Braulke, Keon Hyok Lee, Kenneth Miranda, Christian Mulder, Hisanobu Shishido, and Krishna Srinivasan. 1996.
134. India: Economic Reform and Growth, by Ajai Chopra, Charles Collyns, Richard Hemming, and Karen Parker with Woosik Chu and Oliver Fratzscher. 1995.
133. Policy Experiences and Issues in the Baltics, Russia, and Other Countries of the Former Soviet Union, edited by Daniel A. Citrin and Ashok K. Lahiri. 1995.
132. Financial Fragilities in Latin America: The 1980s and 1990s, by Liliana Rojas-Suárez and Steven R. Weisbrod. 1995.
131. Capital Account Convertibility: Review of Experience and Implications for IMF Policies, by staff teams headed by Peter J. Quirk and Owen Evans. 1995.
130. Challenges to the Swedish Welfare State, by Desmond Lachman, Adam Bennett, John H. Green, Robert Hagemann, and Ramana Ramaswamy. 1995.
129. IMF Conditionality: Experience Under Stand-By and Extended Arrangements. Part II: Background Papers. Susan Schadler, Editor, with Adam Bennett, Maria Carkovic, Louis Dicks-Mireaux, Mauro Mecagni, James H.J. Morsink, and Miguel A. Savastano. 1995.
128. IMF Conditionality: Experience Under Stand-By and Extended Arrangements. Part I: Key Issues and Findings, by Susan Schadler, Adam Bennett, Maria Carkovic, Louis Dicks-Mireaux, Mauro Mecagni, James H.J. Morsink, and Miguel A. Savastano. 1995.
127. Road Maps of the Transition: The Baltics, the Czech Republic, Hungary, and Russia, by Biswajit Banerjee, Vincent Koen, Thomas Krueger, Mark S. Lutz, Michael Marrese, and Tapio O. Saavalainen. 1995.
126. The Adoption of Indirect Instruments of Monetary Policy, by a Staff Team headed by William E. Alexander, Tomás J.T. Baliño, and Charles Enoch. 1995.
125. United Germany: The First Five Years—Performance and Policy Issues, by Robert Corker, Robert A. Feldman, Karl Habermeier, Hari Vittas, and Tessa van der Willigen. 1995.
124. Saving Behavior and the Asset Price “Bubble” in Japan: Analytical Studies, edited by Ulrich Baumgartner and Guy Meredith. 1995.
123. Comprehensive Tax Reform: The Colombian Experience, edited by Parthasarathi Shome. 1995.
122. Capital Flows in the APEC Region, edited by Mohsin S. Khan and Carmen M. Reinhart. 1995.
121. Uganda: Adjustment with Growth, 1987-94, by Robert L. Sharer, Hema R. De Zoysa, and Calvin A. McDonald. 1995.
120. Economic Dislocation and Recovery in Lebanon, by Sena Eken, Paul Cashin, S. Nuri Erbas, Jose Martelino, and Adnan Mazarei. 1995.
119. Singapore: A Case Study in Rapid Development, edited by Kenneth Bercuson with a staff team comprising Robert G. Carling, Aasim M. Husain, Thomas Rumbaugh, and Rachel van Elkan. 1995.
118. Sub-Saharan Africa: Growth, Savings, and Investment, by Michael T. Hadjimichael, Dhaneshwar Ghura, Martin Mühleisen, Roger Nord, and E. Murat Uçer. 1995.
117. Resilience and Growth Through Sustained Adjustment: The Moroccan Experience, by Saleh M. Nsouli, Sena Eken, Klaus Enders, Van-Can Thai, Jörg Decressin, and Filippo Cartiglia, with Janet Bungay. 1995.
116. Improving the International Monetary System: Constraints and Possibilities, by Michael Mussa, Morris Goldstein, Peter B. Clark, Donald J. Mathieson, and Tamim Bayoumi. 1994.
115. Exchange Rates and Economic Fundamentals: A Framework for Analysis, by Peter B. Clark, Leonardo Bartolini, Tamim Bayoumi, and Steven Symansky. 1994.
114. Economic Reform in China: A New Phase, by Wanda Tseng, Hoe Ee Khor, Kalpana Kochhar, Dubravko Mihaljek, and David Burton. 1994.
113. Poland: The Path to a Market Economy, by Liam P. Ebrill, Ajai Chopra, Charalambos Christofides, Paul Mylonas, Inci Otker, and Gerd Schwartz. 1994.
112. The Behavior of Non-Oil Commodity Prices, by Eduardo Borensztein, Mohsin S. Khan, Carmen M. Reinhart, and Peter Wickham. 1994.
111. The Russian Federation in Transition: External Developments, by Benedicte Vibe Christensen. 1994.
110. Limiting Central Bank Credit to the Government: Theory and Practice, by Carlo Cottarelli. 1993.
109. The Path to Convertibility and Growth: The Tunisian Experience, by Saleh M. Nsouli, Sena Eken, Paul Duran, Gerwin Bell, and Zühtü Yücelik. 1993.
108. Recent Experiences with Surges in Capital Inflows, by Susan Schadler, Maria Carkovic, Adam Bennett, and Robert Kahn. 1993.
107. China at the Threshold of a Market Economy, by Michael W. Bell, Hoe Ee Khor, and Kalpana Kochhar with Jun Ma, Simon N’guiamba, and Rajiv Lall. 1993.
106. Economic Adjustment in Low-Income Countries: Experience Under the Enhanced Structural Adjustment Facility, by Susan Schadler, Franek Rozwadowski, Siddharth Tiwari, and David O. Robinson. 1993.
105. The Structure and Operation of the World Gold Market, by Gary O’Callaghan. 1993.
104. Price Liberalization in Russia: Behavior of Prices, Household Incomes, and Consumption During the First Year, by Vincent Koen and Steven Phillips. 1993.
103. Liberalization of the Capital Account: Experiences and Issues, by Donald J. Mathieson and Liliana Rojas-Suarez. 1993.
102. Financial Sector Reforms and Exchange Arrangements in Eastern Europe. Part I: Financial Markets and Intermediation, by Guillermo A. Calvo and Manmohan S. Kumar. Part II: Exchange Arrangements of Previously Centrally Planned Economies, by Eduardo Borensztein and Paul R. Masson. 1993.
101. Spain: Converging with the European Community, by Michel Galy, Gonzalo Pastor, and Thierry Pujol. 1993.
100. The Gambia: Economic Adjustment in a Small Open Economy, by Michael T. Hadjimichael, Thomas Rumbaugh, and Eric Verreydt. 1992.
99. Mexico: The Strategy to Achieve Sustained Economic Growth, edited by Claudio Loser and Eliot Kalter. 1992.
98. Albania: From Isolation Toward Reform, by Mario I. Blejer, Mauro Mecagni, Ratna Sahay, Richard Hides, Barry Johnston, Piroska Nagy, and Roy Pepper. 1992.
97. Rules and Discretion in International Economic Policy, by Manuel Guitián. 1992.
96. Policy Issues in the Evolving International Monetary System, by Morris Goldstein, Peter Isard, Paul R. Masson, and Mark P. Taylor. 1992.
95. The Fiscal Dimensions of Adjustment in Low-Income Countries, by Karim Nashashibi, Sanjeev Gupta, Claire Liuksila, Henri Lorie, and Walter Mahler. 1992.
94. Tax Harmonization in the European Community: Policy Issues and Analysis, edited by George Kopits. 1992.
Note: For information on the title and availability of Occasional Papers not listed, please consult the IMF Publications Catalog or contact IMF Publication Services.