Chapter

V Structural Reforms

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

Since the mid-1980s, structural policies have played an important role in IMF-supported programs.39 These structural measures are intended to complement and buttress macroeconomic policies, raising the likelihood that program objectives will be attained. This section examines whether structural measures included in IMF-supported programs have been geared toward, and have contributed to, achieving program objectives.40 To this end, this section first proposes a simple classification of structural measures according to their primary economic objective—underpinning stabilization efforts, increasing economic flexibility and efficiency, and addressing vulnerabilities.41 It then reviews the structural content of programs to see how various measures, thus classified, have been aligned to the broad objectives of different types of IMF-supported programs. Next, while recognizing the inherent difficulties in establishing the impact of individual structural measures, the section considers experience with two of the most common goals of structural measures in IMF-supported programs: underpinning fiscal adjustment and promoting sustained output growth. This analysis is based on the outcomes observed in the first three years following the approval of each arrangement.

Structural Content of IMF-Supported Programs

Structural reforms in IMF-supported programs range from measures that are very specific to the particular circumstances of the country or the macroeconomic instruments being employed—such as establishing the legal and institutional framework of a currency board arrangement—to those that are more common across programs such as the introduction of a value-added tax to raise revenues. To examine the alignment of structural reforms to program objectives requires classifying these disparate measures. While any classification system inevitably involves an element of arbitrariness—some measures may fit more than one group,42 while others are difficult to assign to any category—it is useful to divide reforms into three categories or groups according to their intended goals.43 These groups are as follows:

  • Measures that underpin a medium-term framework for demand management and for addressing flow imbalances. These policies are designed to underpin stabilization efforts and to enhance the functioning of fiscal, monetary, and exchange rate policies. For example, reducing fiscal imbalances may require underlying reforms to expenditure and revenue in order to be sustained and remain credible. Structural policies in the fiscal area include measures that improve the tax structure—including widening the tax base—and tax administration, as well as policies that strengthen public expenditure management. Deepening financial markets and expanding the menu of instruments available to the monetary authorities can provide for a more stable environment for conducting monetary policy. Finally, other policies aim at strengthening exchange systems; for example, measures that strengthen foreign exchange markets.
  • Measures that enhance economic flexibility and efficiency. These measures often have a combination of objectives, making it difficult to distinguish precisely their flexibility and efficiency goals. Nevertheless, among the flexibility goals are all measures that increase the ability of the economy to adapt to new conditions. Usual examples are trade reforms and policies that affect resource allocation across sectors, such as pricing policies of factor markets (labor and capital) and the institutional features of these markets. They also include pricing policies that transcend individual sectors, such as energy prices. In contrast, the private sector efficiency component refers to impediments to investment and growth and reforms that affect individual sectors, such as pricing policies and marketing arrangements in agricultural markets and institutional changes that affect corporate sector behavior. Privatization of state enterprises and utilities also fall into this category, though often these measures have other objectives as well, including use of privatization receipts for stabilization efforts or to strengthen balance sheets.44 Finally, the public sector efficiency component relates to measures that improve the delivery of public services or redefine the role of the state in the economy.
  • Measures that address economic vulnerabilities, including stock or balance sheet mismatches. These policies may be directed at tackling unsustainable public or external debt dynamics, reducing the vulnerability of domestic balance sheets to sharp swings in the exchange rate or interest rates, as well as structural weaknesses in the financial sector—particularly those that may result in contingent liabilities of the public sector. Strengthening prudential regulations and financial sector supervisory capabilities form an important element of this category.45

As discussed in “Objectives and Outcomes” (Part II of this occasional paper), most IMF-supported programs can be usefully classified as “classic” (current account) adjustment, poverty-reducing and growth-enhancing programs, or capital account crises. In a classic adjustment program, structural policies are expected to center primarily on the first of the above categories, but reforms that increase efficiency and reduce vulnerabilities can also be important.46 The emphasis of structural reforms in PRGF-supported programs is on efficiency measures that improve potential output growth, including measures to enhance human capital, health, and education. However, medium-term demand management measures are also necessary for various reasons, including the role played by macroeconomic stability in strengthening growth and the challenges faced by these countries in mobilizing tax revenues and strengthening expenditure control.47 Transition economies, reflecting the numerous systemic transformation challenges faced by these countries, are a hybrid of these two types of programs; efficiency and growth-oriented measures are critical, though demand management and financial sector reforms are also needed. Capital account crisis programs have more clearly defined reform needs. Specifically, reforms that reduce stock vulnerabilities take center stage among these countries, partly driven by the urgency in improving confidence in the economy. In crises where the capital outflows are primarily from the private sector, this means financial and corporate sector reforms. By contrast, where markets are responding to concerns about public debt sustainability, measures that improve the viability of public finances are required, even if they only have an impact over the medium term. Although the source of the balance sheet imbalances has a bearing on the design of reforms and may reveal structural weaknesses throughout the economy (from weak demand management to efficiency bottlenecks for private sector growth), the core reform efforts of these programs are directed toward addressing balance sheet weaknesses.

How well aligned are structural measures to the broad objectives of the various types of IMF-supported programs in practice? The distribution of structural measures (classified into the three categories described above) is reported in Figure 4.5 for GRA-supported programs (excluding programs in transition economies and capital account crises), transition economy programs, capital account crises, and PRGF-supported programs in low-income countries (again, excluding transition economies). The distribution mirrors, at least to some extent, the expected distribution by type of IMF-supported programs. In particular, measures in GRA-supported programs in nontransition economies are split between macroeconomic management (35 percent) and efficiency and growth-related measures (35 percent), and those aimed at reducing vulnerabilities (30 percent). Relative to this benchmark, programs in transition economies are somewhat more oriented toward growth-enhancing measures (41 percent, a difference that is statistically significant from the nontransition GRA sample). PRGF-supported programs likewise show a somewhat greater preponderance of growth enhancing measures (38 percent, though this difference with GRA-supported programs is not statistically significant). The largest, and statistically significant, difference lies between capital account crisis programs, with their much greater emphasis on reducing sources of vulnerability—60 percent of measures (versus 17 percent on macroeconomic management and 22 percent on growth-related measures), and all other program groups.

Figure 4.5.Distribution of Structural Conditionality in IMF-Supported Programs

(In percent of total number of conditions per program year; average 1995–2000)

Sources: IMF, MONA database; and IMF staff estimates.

1Excludes transition economics.

2For each type of condition, the null hypothesis of no difference from GRA-supported programs is tested; *significant at 10 percent level and ***significant at 1 percent level.

Experience

The inherent problems of quantifying structural policies make it difficult to establish links between specific structural reforms and macroeconomic outcomes. With this limitation in mind, this section takes up two of the most common goals of structural measures in IMF-supported programs—underpinning fiscal adjustment and enhancing economic efficiency and output growth—with a view to shedding some light on whether, or to what extent, structural policies have been useful in attaining these objectives. Given the lack of better alternatives, the analysis is limited to the effects of the number of conditions on the objectives these structural measures seek to accomplish. The analysis also distinguishes between stopped and nonstopped programs in an attempt to identify implementation issues.

Fiscal Adjustment

Part of the impetus for structural reforms in IMF-supported programs was the observation in the early 1980s that fiscal adjustment efforts were often not sustained. To examine whether structural measures help underpin fiscal adjustment, program conditions related to fiscal measures were classified according to their intended effects on revenues and expenditures. The three categories related to tax and expenditure measures in the MONA database, are reclassified into two core revenue categories (tax policy and tax administration), two core expenditure categories (expenditure control and expenditure management), and a number of ancillary revenue and expenditure categories.48 The ancillary group includes measures related to fiscal transparency, debt-related measures, civil service reform, and measures targeting a country’s social security system.

Table 4.20 reports the results of regressions of fiscal adjustment—over the three-year period that begins with the approval of each arrangement—in the overall balance, and of adjustment in revenues and expenditures separately, on the corresponding structural measures. The results suggest that structural measures are related with better fiscal performance, particularly in regard to core revenue measures on revenue adjustment and to core revenue and core expenditure measures on the overall fiscal adjustment. Core expenditure measures do not, however, appear to have a correlation with expenditure reduction—except perhaps among transition economies where the country-type dummy is positive and highly significant. Not surprisingly, among arrangements that did not go off-track, the impact of these measures on fiscal adjustment is stronger.49 In addition, the numerous ancillary fiscal measures that characterize IMF-supported programs (see previous paragraph) are not found to have a correlation with overall balance, revenue, or expenditure adjustment and are not included in the regressions reported in the table.

Table 4.20.Structural Measures and Fiscal Adjustment: Regression Results1
Dependent Variables: Actual Adjustment in2
Fiscal balanceGovernment revenueGovernment expenditure
All programsNonstopped programs3All programsNonstopped programs3All programsNonstopped programs3
Regressors
Number of fiscal measures related to the4
Fiscal balance50.11*0.14*
Revenues0.16**0.32***
Expenditures0.000.02
Dummy variables
Stopped program3–0.27–0.13–0.05
PRGF program–0.05–0.690.05–0.080.180.20
Transition economy0.71*0.31–0.37–0.541.40***1.29***
Capital account crisis0.45–0.090.800.460.080.39
Intercept–0.140.28–0.17–0.17–0.32–0.18
Number of observations/programs100671006710067
R20.110.120.080.160.150.23
F-statistic61.9*1.61.32.2*2.8**3.7***
Sources: IMF, MONA database; and IMF staff estimates.Note. * significant at 10 percent level, ** significant at 5 percent level, and *** significant at 1 percent level.

Regressions based on a dataset of programs approved during the period 1995–2000 for which data on fiscal balance, revenues, and expenditures are available.

Average adjustment in years t, t+1, and t+2, where t is the year the program is approved.

A stopped program is defined as a program that terminates earlier than was originally anticipated.

The number of fiscal measures is normalized by the duration of the program.

Structural measures that affect the fiscal balance includes revenue and expenditure reforms.

F-statistic for null hypothesis that all explanatory variables (other than the constant) are jointly equal to zero.

Sources: IMF, MONA database; and IMF staff estimates.Note. * significant at 10 percent level, ** significant at 5 percent level, and *** significant at 1 percent level.

Regressions based on a dataset of programs approved during the period 1995–2000 for which data on fiscal balance, revenues, and expenditures are available.

Average adjustment in years t, t+1, and t+2, where t is the year the program is approved.

A stopped program is defined as a program that terminates earlier than was originally anticipated.

The number of fiscal measures is normalized by the duration of the program.

Structural measures that affect the fiscal balance includes revenue and expenditure reforms.

F-statistic for null hypothesis that all explanatory variables (other than the constant) are jointly equal to zero.

Output Growth

Both in PRGF-supported programs and, to a lesser degree, in classic adjustment programs, structural reforms may be undertaken to enhance economic efficiency and long-term growth performance, raising the question of the effectiveness of such reforms. Typical reforms include measures aimed at liberalizing the trade regime as well as changes in pricing and marketing policies. Table 4.21 reports the results of a regression of the average change in real GDP growth (between years t–1 and t+2) on growth-related structural measures. Growth, of course, may depend on a number of other factors. To purge the effects of variables that are unlikely to vary significantly over a two- to three-year horizon—such as the stocks of human and physical capital—the dependent variable is specified as the change in real GDP growth. To proxy for macroeconomic variables that are likely to change at higher frequency, both the change in the fiscal balance and the change in the inflation rate are included in the regression as additional explanatory variables.

Table 4.21.Structural Measures and Growth: Regression Results1
Dependent Variables: Average Change in Real GDP Growth Rates in the First Three Years Following Program Approval2
All programsNonstopped programs3
Regressors
Number of growth-related structural measures40.08**0.07**0.12*0.10*
Dummy variables
Stopped program3–0.040.06
PRGF program–0.30–0.26–0.53–0.38
Transition economy1.77***0.93**1.38**0.98*
Capital account crisis0.520.520.770.71
Macroeconomic variables
Average change in the fiscal balance0.28**0.25*
Average change in the inflation rate–0.08***–0.04
Intercept–0.24–0.38–0.19–0.30
Number of observations/programs1001006363
R20.300.400.270.32
F-statistic58.2***8.7***5.4***4.4***
Sources: IMF, MONA database; and IMF staff estimates.Note. * significant at 10 percent level, ** significant at 5 percent level, and *** significant at 1 percent level.

Regressions based on a dataset of programs approved during the period 1995–2000.

Average growth in years t, t+1, and t+2, where t is the year the program is approved.

A stopped program is defined as a program that terminates earlier than was originally anticipated.

The number of growth-related structural measures is normalized by the duration of the program.

F-statistic for null hypothesis that all explanatory variables (other than the constant) are jointly equal to zero.

Sources: IMF, MONA database; and IMF staff estimates.Note. * significant at 10 percent level, ** significant at 5 percent level, and *** significant at 1 percent level.

Regressions based on a dataset of programs approved during the period 1995–2000.

Average growth in years t, t+1, and t+2, where t is the year the program is approved.

A stopped program is defined as a program that terminates earlier than was originally anticipated.

The number of growth-related structural measures is normalized by the duration of the program.

F-statistic for null hypothesis that all explanatory variables (other than the constant) are jointly equal to zero.

From the table, growth-related structural measures are positively and significantly related to better growth performance, especially in programs in which there are no stoppages. At the same time, the effects are not large: from the estimates, each additional measure is associated with 0.1 percentage point higher real GDP growth. Of course, it bears emphasizing that these results should only be viewed as indicative, given the possibility of omitted variables, endogeneity of program participation, and the inherent difficulties of quantifying structural measures. Moreover, it is plausible that the real relationship is nonlinear, possibly as a result of threshold effects, with diminishing returns to the number of structural measures. These more complicated relationships, however, would not be captured by the simple linear regression reported here.

Summary

Structural policies have played an increasingly important role in IMF-supported programs, complementing macroeconomic policies by underpinning stabilization efforts and orderly adjustment, enhancing efficiency and growth, and reducing vulnerabilities to future crises. There is broad alignment between the nature of structural reforms included in IMF-supported programs and the objectives of the program. Thus, classic adjustment programs tend to focus on medium-term demand management issues, PRGF-supported programs include growth and efficiency measures (as is also the case for transition economies), and capital account crisis programs aim at addressing vulnerabilities. Turning to experience, within the inherent limitations of quantitative analysis of the effects of structural reforms, the evidence suggests that structural measures included in IMF-supported programs might have had some positive effects on achieving sustained fiscal adjustment and output growth.

    Other Resources Citing This Publication