Part IV Macroeconomic and Structural Policies: Review of Experience

Charalambos Christofides, Atish Ghosh, Uma Ramakrishnan, Alun Thomas, Laura Papi, Juan Zalduendo, and Jun Kim
Published Date:
September 2005
  • ShareShare
Show Summary Details

Contents of Part IV


    Abed, George, and others,1998, Fiscal Reforms in Low-Income Countries, IMF Occasional Paper No. 160 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Baldacci, E., and others,2004, “Front-Loaded or Back-Loaded Fiscal Adjustments: What Works in Emerging Market Economies,”IMF Working Paper No. 04/157 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Bubula, Andrea, and InciÖtker-Robe,2002, “The Evolution of Exchange Rate Regimes Since 1990: Evidence from De Facto Policies,”IMF Working Paper No. 02/155 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Calvo, Guillermo, and CarmenReinhart,2002, “Fear of Floating,”Quarterly Journal of Economics, Vol. 117 (May), pp. 379–408.

    Calvo, Guillermo, and CarlosVégh,1994, “Inflation Stabilization and Nominal Anchors,”Contemporary Economic Policy, Vol. 12 (April), pp. 35–45.

    • Search Google Scholar
    • Export Citation

    Calvo, Guillermo, and CarlosVégh,1999, “Inflation Stabilization and BOP Crises in Developing Countries,”NBER Working Paper No. 6925 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation

    Christensen, Jakob E.,2004, “Domestic Debt Markets in Sub-Saharan Africa,”IMF Working Paper No. 04/46 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Daseking, Christina, Atish R.Ghosh, Alun J.Thomas, and Timothy D.Lane,2004, Lessons from the Crisis in Argentina, IMF Occasional Paper No. 236 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Easterly, William R.,1996, “When Is Stabilization Expansionary? Evidence from High Inflation,”Economic Policy, No. 21, pp. 67–107.

    • Search Google Scholar
    • Export Citation

    Fischer, Stanley, RatnaSahay, and CarlosVégh,2002, “Modern Hyper- and High Inflations,”IMF Working Paper No. 02/197 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Ghosh, Atish, TimothyLane, MarianneSchulze-Ghattas, AlešBulíř, JavierHamann, and AlexMourmouras,2002, IMF-Supported Programs in Capital Account Crises, IMF Occasional Paper No. 210 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Ghosh, Atish, Anne-MarieGulde, and Holger C.Wolf,2002, Exchange Rate Regimes: Choices and Consequences (Cambridge, Massachusetts: MIT PRess).

    • Search Google Scholar
    • Export Citation

    Gupta, Sanjeev, MarkPlant, BenedictClements, ThomasDorsey, EmanueleBaldacci, GabrielaInchauste, ShamsuddinTareq, and NitaThacker,2002, Is the PRGF Living Up to Expectations? An Assessment of Program Design, IMF Occasional Paper No. 216 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Hamann, Javier,1999, “Exchange-Rate-Based Stabilization: A Critical Look at the Stylized Facts,”IMF Working Paper No. 99/132 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Hamann, Javier,2001, “Exchange Rate-Based Inflation Stabilization,”IMF Research Bulletin, Vol. 2, No. 3 (September).

    Hamann, Javier, and AlessandroPrati,2002, “Why Do Many Disinflations Fail? The Importance of Luck, Timing, and Political Institutions,”IMF Working Paper No. 02/228 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Hemming, Richard, Michael S.Kell, and SelmaMahfouz,2002, “The Effectiveness of Fiscal Policy in Stimulating Economic Activity—A Review of the Literature,”IMF Working Paper No. 02/208 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Independent Evaluation Office, 2003, Fiscal Adjustment in IMF-Supported Programs (Washington: International Monetary Fund).

    Independent Evaluation Office, 2004, Evaluation of the IMF’s Role in Poverty Reduction Strategy Papers and the Poverty Reduction and Growth Facility (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund, 2001, “Structural Conditionality in Fund-Supported Programs;”available at

    • Search Google Scholar
    • Export Citation

    International Monetary Fund, 2003, World Economic Outlook, September, World Economic and Financial Surveys (Washington).

    International Monetary Fund, 2005, “Review of the 2002 Conditionality Guidelines;”available at

    • Search Google Scholar
    • Export Citation

    Kiguel, Miguel, and NissanLiviatan,1992, “The Business Cycle Associated with Exchange Rate-Based Stabilizations,”World Bank Economic Review, Vol. 6 (May), pp. 279–305.

    • Search Google Scholar
    • Export Citation

    Reinhart, Carmen, and KennethRogoff,2004, “The Modern History of Exchange Rate Arrangements: A Reinterpretation,”Quarterly Journal of Economics, Vol. 119 (February), pp. 1–48.

    • Search Google Scholar
    • Export Citation

    Santaella, Julio, and AbrahamVela,1996, “The 1987 Mexican Disinflation Program: An Exchange-Rate-Based Stabilization?”IMF Working Paper No. 96/24 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Schadler, Susan, AdamBennett, MariaCarkovic, LouisDicks-Mireaux, MauroMecagni, James H.J.Morsink, and MiguelSavastano,1995, IMF Conditionality: Experience Under Stand-By and Extended Arrangements, Part I: Key Issues and Findings, IMF Occasional Paper No. 128, and Part II: Background Papers, IMF Occasional Paper No. 129 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Schaechter, Andrea, Mark R.Stone, and MarkZelmer,2000, Adopting Inflation Targeting: Practical Issues for Emerging Market Countries, IMF Occasional Paper No. 202 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Sobolev, Yuri V.,2000, “Exchange-Rate-Based Stabilization: A Model of Financial Fragility,”IMF Working Paper No. 00/122 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Stone, Mark R.,2003, “Inflation Targeting Lite,”IMF Working Paper No. 03/12 (Washington: International Monetary Fund).

    Truman, E.,2003, Inflation Targeting and the International Financial System: Challenges and Opportunities (Washington: Institute for International Economics).

    • Search Google Scholar
    • Export Citation

    Végh, Carlos,1992, “Stopping High Inflation,”IMF Staff Papers, Vol. 39, No. 3 (September), pp. 626–95.


To include both program and postprogram experience, the sample consists of arrangements approved over the period 1995–2000 and supported by the General Resources Account (GRA)—Stand-By Arrangements (SBAs) and Extended Fund Facility (EFF) arrangements—or by concessional facilities—the Enhanced Structural Adjustment Facility (ESAF) prior to 1999/2000 and the Poverty Reduction and Growth Facility (PRGF) since then. For simplicity, the term PRGF is used to refer to both ESAF- and PRGF-supported programs. A list of arrangements can be found in Appendix I of “Objectives and Outcomes” (see Part II of this occasional paper); individual analyses reported below may use subsamples according to data availability.


However, to control for possible omitted variable bias, where relevant, the regressions reported below include the various policy instruments simultaneously.


In particular, the coefficients on policy variables may be misestimated if regressors are correlated with the policy variable (and the dependent variable), but omitted from the regression (see Appendix IV of “Objectives and Outcomes” (Part II of this occasional paper) for a discussion of alternative methodologies for evaluating the effects of programs).


The Independent Evaluation Office (IEO) comes to a similar conclusion in its assessment of programs in low-income countries (IEO, 2004).


This paper examines performance during the year of program approval and the following year because most programs span more than one calendar year. The average duration of IMF-supported programs in the sample is 17 months for SBAs, and 35 months for programs supported by the EFF as well as by concessional facilities—the ESAF and the PRGF. Further, when an arrangement was approved in the last quarter of the year, for analytical purposes it is treated as having been approved in the subsequent year.


Such up-front devaluations as part of the initial package of policies under the program are rare, however. In the sample (about 130 programs approved during the period 1995–2000), only Mauritania and Ukraine carried out a step devaluation at the beginning of the program.


The results in this section are based on the IMF’s official classification of exchange rate regimes, as reported in the Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER). The IMF uses a de facto classification that combines quantitative and qualitative information, including the authorities’ stated exchange rate policy. While the IMF changed to a de facto classification in 1999, the data for previous years were obtained from Bubula and Ötker-Robe (2002), who constructed the back series using the same methodology. Using a purely de facto classification (for instance, that proposed by Reinhart and Rogoff, 2004) would lead to a larger proportion of countries being classified as having pegged regimes (perhaps reflecting “fear of floating” (see Calvo and Reinhart, 2002)), which would strengthen the relationship between pegged regimes and better inflation performance reported below. The main conclusions of this section, however, are unaltered.


This proportion is similar to the proportion of countries changing their exchange rate regimes outside the context of an IMF-supported program. These statistics may overstate the proportion of cases where the regime was changed as part of the IMF-supported program since, in some cases, the regime change occurred a few months prior to or a few months following the approval of the IMF arrangement and thus may have not been part of the design of the program. For the purposes of these statistics, any change between the eight categories of exchange rate regime presented in the AREAER is counted as a regime shift.


Conventional pegs are more than twice as frequent (37 percent of countries) when the country does not have an IMF-supported program as when it does (15 percent of countries).


Among PRGF-supported countries, members of the CFA zone maintain a pegged regime for long-standing institutional reasons. For non-CFA zone members, there may be hesitation in adopting a peg even in the context of an attempt at disinflation because the institutions and policy discipline necessary to maintain the peg may be lacking.


Ghosh, Gulde, and Wolf (2003) find that the association between low inflation and pegged exchange rate regimes survives the inclusion of other explanatory variables and a battery of robustness tests including possible endogeneity of the exchange rate regime. The association between the exchange rate regime and growth, however, breaks down once endogeneity of the regime is taken into account.


In most cases where no explicit monetary anchor was in place, a ceiling on net domestic assets (NDA) and a floor on net international reserves (NIR) were targeted.


The 1994 Conditionality Review (Schadler and others, 1995) drew a similar conclusion.


However, in contrast, for PRGF-supported programs in non-transition economies, the equation does not show any statistically significant correlations.


A common argument is that the de jure or de facto exchange rate pegs in the Asian crisis countries provided an implicit exchange rate guarantee, encouraging unhedged foreign borrowing.


In “Objectives and Outcomes” (see Part II of this occasional paper) two metrics were employed to assess external adjustment: the comparison with the projected current account balance, and the comparison with the debt-stabilizing balance (when the initial level of external debt is below 40 percent of GDP). There is no statistically significant difference between (subsequent) external adjustment under pegged and flexible regimes using the latter criterion.


To reduce the influence of high-inflation outliers, all figures are transformed to map into the interval (–100, 100) percent prior to taking averages.


An alternative gauge of monetary policy is given by the behavior of interest rates. Problems of availability and comparability of data however make it less useful in cross-country comparisons.


Define a benchmark growth in broad money Δm^ as the growth rate implied by program expectations of inflation and real GDP growth and the expected behavior of velocity: Δm^=πpΔypΔv^. One way to capture the expected behavior of velocity is to use the country’s trend velocity growth: Δv^=Δv¯. Similarly, the programmed increase in broad money growth can be written Δmp = πp + Δyp – Δvp. Subtracting, yields: ΔmpΔm^=Δv¯Δvp so that Δvp>Δv¯ implies that broad money growth envisaged under the program is lower than would be implied by trend velocity (and program expectations of inflation and growth). As such, it can be interpreted as a programmed tightening of the monetary stance. One possibility, however, is that velocity itself depends upon expected inflation. In that case, the appropriate benchmark is not the trend change in velocity but the expected change in velocity, Δv^, taking account of possible remonetization. Although it is difficult to establish how much remonetization should occur, since programs typically target disinflation, this should at least imply Δv^Δv¯. Therefore, if programmed velocity is higher than the historical trend, then this necessarily implies a tighter programmed monetary stance: Δvp>Δv¯Δvp>Δv^Δmp<Δm^. (To the extent that velocity rises relative to trend in the year prior to approval of the arrangement, however, this measure may overstate the degree of tightening.) On the other hand, if programmed velocity is lower than the historical trend, then this need not indicate a monetary loosening since it is possible that Δvp<Δv¯ but Δvp>Δv^.


The appropriate response to large donor inflows will be examined in the forthcoming review of PRGF-supported programs.


The IEO came to a similar conclusion based on a smaller sample of PRGF-supported programs.


For this exercise, the monetary policy stance is measured by velocity (with an increase indicating a tighter stance), instrumented by the programmed velocity. These regressions also include the overall fiscal balance (instrumented by its program projection) to control for possible omitted variable bias. The role of fiscal policy is discussed in Section IV, below.


Monetary policy also has an important role to play—particularly in capital account crises—in stemming capital outflows and achieving a more orderly external adjustment. Empirical evidence on the impact of higher interest rates on capital flows is discussed in IMF-Supported Programs in Capital Account Crises (see Ghosh and others, 2002).


Since the sample here is arrangements approved over the period 1995–2000, the results reported for PRGF-supported programs actually refers mostly to ESAF-supported programs.


These figures pertain to targets specified in the original program. As discussed in Box 2.2 of “Objectives and Outcomes” (Part II of this occasional paper), in a number of capital account crises, especially in East Asia, the initial fiscal targets were revised as it became apparent that activity was turning out significantly weaker than expected.


Other variables that might influence the programmed fiscal adjustment are public debt and either the preprogram inflation rate or the targeted reduction in inflation. Specifications (not shown in Table 4.15) that included these variables did not yield significant coefficients however, and—in the case of public debt—entailed dropping a large number of observations due to the lack of data availability.


The results here are given in terms of overall fiscal balance adjustment. Qualitatively, the results are similar when the primary balance is used in the regressions, albeit with generally lower statistical significance because fewer observations are available.


In the transition economies, by contrast, a larger gap is associated with a larger adjustment effort, though this most likely reflects the difficulties of estimating meaningful output gaps in a period in which potential output growth was changing rapidly.


The Independent Evaluation Office Report on Fiscal Adjustment in IMF-Supported Programs (IEO, 2003) likewise finds that programmed fiscal adjustment is tailored to the country’s specific circumstances.


Defined as cases in which more than one-half of the programmed improvement in the fiscal balance reflects improved revenues.


Defined as the top quartile of programs ranked by the programmed adjustment.


When interpreting these results it is important to bear in mind the possible endogeneity of the regressors, particularly the projection errors for the current account balance and real GDP growth. The results reported below (Table 4.19) using instrumental variable estimation suggest that (actual) fiscal adjustment contributes to current account adjustment, which would imply a positive bias to the coefficient on the programmed current account balance reported in Table 4.16; the results in Table 4.19 suggest that fiscal adjustment has very little effect on growth. An alternative specification is to estimate a probit of cases where fiscal adjustment fell short of the program target, regardless of the magnitude of the slippage. Although the main findings are similar, those results are somewhat stronger in both the GRA- and PRGF-samples, with about 80 percent of the observations correctly classified.


A recent paper that indeed finds that expenditure cuts increase the probability of successful fiscal adjustment (albeit for the short term) is Baldacci and others (2004). They also find that political economy variables capturing social cohesion as well as IMF-supported programs contribute to a higher probability of successful fiscal adjustment.


This distribution is not statistically significantly different from the distribution of cases where fiscal adjustment was more than had been programmed (12 cases, total); 7 out of 12 cases had above-programmed external adjustment and 5 out of 12 cases had below-programmed external adjustment.


As noted in the IEO report on fiscal adjustment (IEO, 2003) and the 2003 World Economic Outlook chapter on public debt (IMF, 2003), consistent time series on public debt (including the domestic component) are often lacking. A recent paper (Christensen, 2004) reports data on domestic debt for a set of 27 sub-Saharan African economies, but issues of coverage and the lack of consistent series on above-the-line fiscal accounts preclude its use here.


Arithmetically, for a given capital outflow, the higher the public sector saving-investment balance, the smaller the private sector’s balance needs to be, but this does not necessarily imply a lower burden of adjustment on the private sector in terms of consumption and investment. If the effects of capital outflows are in the nature of a supply-side shock (for example, the associated exchange rate depreciation raises the price of imported intermediate inputs or leads to widespread bankruptcies because of the private sector’s foreign exchange exposure), then a higher public sector balance indeed reduces the adjustment burden on the private sector. By contrast, if capital outflows represent (or exacerbate) a demand-side shock, and if Keynesian effects are important so that a fiscal loosening has an expansionary effect on activity, then allowing the public sector balance to deteriorate could help achieve the requisite external adjustment with a smaller decline of output and of private consumption and investment. For a fuller discussion, see IMF-Supported Programs in Capital Account Crises (Ghosh and others, 2002).


Although there is an extensive literature on this topic (with various findings), most of it does not focus on countries that have IMF-supported programs. Hemming and others (2002) provide a literature review, which generally asserts that there are significant Keynesian effects between fiscal balances and output, though these results pertain mostly to industrial economies. Consistent with the results reported in this paper, Gupta and others (2002), find instead that strong fiscal consolidation is associated with higher growth in a sample of low-income countries. One explanation may be that fiscal consolidation stimulates growth in countries with weak institutions by a reduction in rent-seeking and the scope for corruption, raising overall productivity and growth.


Some specifications (not reported), for instance using two lags of the fiscal balance, yield a negative coefficient (i.e., a larger fiscal balance is associated with lower growth), but even in these regressions the implied Keynesian effects are small: a 1 percent of GDP improvement in the overall balance would be associated with 0.3 percentage points lower growth two years later. Similarly, using government expenditures rather than the overall balance does not suggest a substantial role for stimulative fiscal policy. Inclusion of other control variables, such as the real effective exchange rate, does not affect the results. Segmenting observations by exchange rate regime suggest a stronger positive impact of fiscal adjustment on growth among countries with flexible regimes.


“Structural Conditionality in Fund-Supported Programs” (see IMF, 2001) documents the increase in the structural content of IMF-supported programs over the period 1987–99. In 2000–02, the IMF undertook a broad review of structural conditionality in IMF-supported programs, culminating in the 2002 Conditionality Guidelines. A review of the 2002 Conditionality Guidelines, which examines the application of the revised guidelines, has recently been completed (IMF, 2005).


Recognition that balance of payments problems may reflect structural weaknesses was part of the rationale for the IMF Executive Board decision to create the Extended Fund Facility (EFF) in 1974. The decision notes that structural policies are required in “an economy experiencing serious payments imbalance relating to structural maladjustments … or … characterized by slow growth and … weak balance of payments” (Decision No. 4377-74/114).


The classification into these three categories is carried out by mapping the eight categories in the MONA database into the three groups specified. The MONA classification is prepared by country teams at the time of approval and following each review. The alignment between structural measures and policy objectives is examined in the Review of the Conditionality Guidelines (IMF, 2005).


For instance, changes to the tax structure may be important to bolster macroeconomic stabilization (category 1, below) but also for increasing economic efficiency (category 2). Likewise, reforms in specific sectors such as agriculture may reduce the cost of untargeted subsidies, but may also have important efficiency and growth benefits as well as raising incomes of farmers by dismantling distortionary state marketing boards.


IMF-supported programs have increasingly included measures geared toward institution building, which are usually included in the “economic efficiency” category. Indeed, the share of conditions that are related to institution building has risen from about 3 percent of all annual conditions in 1995 to more than 10 percent by 2000 (especially in transition economies and PRGF-supported programs), though these figures probably understate the proportion of measures related to institution building as many might be classified elsewhere within the MONA database; for example, measures to improve budget control and expenditure management also aim at improving a country’s institutional framework.


Many goals are sought through privatization. For example, the sale of utility companies is proposed when services are poor and an infusion of capital is needed—the purpose is to improve services and modernize the capital infrastructure. In some cases, this is also an opportunity through which to attract foreign investment. Privatization receipts may also play a fiscal role. Privatization of utilities should be assessed carefully so as to avoid transforming a public monopoly into a private monopoly. In contrast, privatization of state firms in other sectors (from wineries to steel mills) are sought either to redefine the role of the state or to stem the fiscal implications of loss-making state firms. In particular, IMF conditionality is justified when loss-making state firms compromise the sustainability of the fiscal position.


Typically, reforms in the financial sector are divided between measures aimed at strengthening the central bank, such as measures to increase its independence, and measures aimed at strengthening the financial sector more generally. The latter focuses on strengthening banking supervision and dealing with problem banks.


Within GRA-supported programs, therefore, those supported by EFF arrangements are more likely to have structural measures oriented toward enhancing economic flexibility and efficiency than those supported by stand-by arrangements.


Between classic adjustment and PRGF-supported programs, the former might be expected to have a slight bias toward demand management measures, and the latter toward efficiency and growth-enhancing measures. Programs in transition economies are likely to straddle these two groups, since achieving macroeconomic stability and adjustment was a critical objective of these programs as was longer-term structural transformation of the economy.


The three categories reclassified are the “tax and expenditures” category of the MONA database, a category referenced as “other measures,” that also includes fiscal measures, and a category related to trade measures as many of those measures aim at improving the collection of customs taxes.


The results presented are derived using a balanced panel of 100 programs approved in the period 1995–2000 and for which fiscal data (projected and actual) is available for three years after program approval. Data used include fiscal revenues, fiscal expenditures, and the fiscal balance. An unbalanced panel based on available data provides broadly similar results.

Other Resources Citing This Publication