Helping Countries Develop

8 The Effectiveness of Government Spending on Education and Health Care in Developing and Transition Economies

Benedict Clements, Sanjeev Gupta, and Gabriela Inchauste
Published Date:
September 2004
  • ShareShare
Show Summary Details
Sanjeev Gupta, Marijn Verhoeven and Erwin R. Tiongson 

1. Introduction

Policy makers are interested in the composition of public spending.1 This attention stems in part from the belief that government spending on education and health care can increase economic growth, promote income equality, and reduce poverty (Barro, 1991; Chu et al., 1995; and Tanzi and Chu, 1998). International financial institutions, donors, and NGOs therefore call for increased government spending on education and health care.2 In addition, an increasing number of studies have documented the adverse economic consequences of corruption; in particular, studies have shown that corruption is associated with higher military spending (Gupta et al., 2001) and lower government spending on education and health care (Mauro, 1998). These studies provide evidence that policies aimed at reducing corruption lead to increased spending on more productive outlays, such as education and health spending.

The justification for public spending on basic education is based on the social rate of return. Studies have found that the social rate of return is highest for primary education, followed by secondary and tertiary education (Psacharopoulos, 1994; and World Bank, 1995).3 At the same time, evidence suggests that spending on tertiary education in many countries is excessively high (see, for example, Sahn and Bernier, 1993; Gupta et al., 1998; and World Bank, 1995).

Public spending on primary health care is justified by disease reduction during the productive years of life. The burden of disease in developing countries could be reduced if governments were to make available a minimum package of essential, cost-effective clinical services (World Bank, 1993). In this respect, secondary health care has been found to provide little health gain. Many studies have concluded that the most cost-effective interventions are often preventive in character, and that in many developing countries public allocations for secondary or curative services are excessive (see, for example, Sahn and Bernier, 1993; and Pradhan, 1996).

Although the studies that focus on social rates of return to education and on the burden of disease provide a compelling reason for policy makers to shift public resources toward basic education and primary health care, they do not yield conclusive evidence that such a reallocation would improve education attainment and health status. It may well be that public spending crowds out private spending on primary and secondary education and primary health care, or that public resources are used inefficiently and inequitably. In fact, the evidence on whether aggregate education and health spending has a beneficial impact on relevant social indicators—taken as a proxy for outputs of public spending on social sectors—is mixed. Many studies show that the relationship between public spending for education and measures of education attainment is weak (Landau, 1986; Noss, 1991; Mingat and Tan, 1992, 1998; and Flug et al., 1998). Instead, other variables have been found to be important in explaining education attainment. These include per capita income (Flug et al., 1998; Mingat and Tan, 1992), the age distribution of the population (Mingat and Tan, 1992), parental perceptions of costs and benefits, and family background or parental education (Appleton et al., 1996). In contrast, Gallagher (1993) shows that, after correcting for its quality and efficiency, spending on education has a positive impact on indicators of education attainment.

Similarly, many studies find that the contribution of public health outlays to health status as measured by infant mortality or child mortality is either small or statistically insignificant (Kim and Moody, 1992; McGuire et al., 1993; Aiyer et al., 1995; Musgrove, 1996; Filmer and Pritchett, 1997; and Filmer et al., 1998). Carrin and Politi (1995) conclude that poverty and income are crucial determinants of health status indicators but fail to find that public health spending has a statistically significant effect on these indicators. Similarly, Filmer and Pritchett (1997) find that cross-country differences in income alone account for 84% of the variation in infant mortality, with socioeconomic variables accounting for 11%, and public spending for less than ⅙ of 1%. These results are confirmed by Demery and Walton (1998), who note that “the conclusion that public spending is a poor predictor of good health is a common one” (p. 26). In contrast, Anand and Ravallion (1993) and Hojman (1996)—with relatively small sample sizes of 22 observations and 10–20 observations, respectively—do find that public health spending has a statistically significant effect on health status. Similarly, Bidani and Ravallion (1997) find for a larger sample of 35 countries that public spending has a beneficial impact on the health condition of the poor.

Although the evidence presented in the above-mentioned studies in general goes against the presumption that higher public spending on education and health is effective in improving social indicators, some relevant issues are overlooked in these studies. As noted earlier, allocations within the sectors are widely believed to be important in explaining changes in social indicators, but these studies typically sidestep this issue.4 In fact, Ogbu and Gallagher (1991) infer from a study of five African countries that enrollment rates are affected by the composition of public education spending. And in a survey of 10 country studies, Mehrotra (1998) concludes that high education attainment is associated with relatively high public spending on education and a relatively high share of primary education in total education expenditures. Unfortunately, neither paper supports its claim about the efficacy of public spending on basic education with statistical analysis.

Filmer et al. (1998) attempt to address the issue of allocations within the health sector by including a measure of government spending on primary health care in their cross-section analysis of the causal factors of infant mortality. As it turns out, they fail to find a statistically significant impact of primary health care spending on infant mortality rates. But their aggregate health sector data are not necessarily consistent with either the overall fiscal or the intrasectoral data. Measurement errors may have been further exacerbated by the use of statistical techniques to create imputed values for missing observations.

Against the background of these empirical results, this paper reassesses whether increased public spending on education and health matters by using a comprehensive, internally consistent, and up-to-date cross-section data set of public spending and social indicators for 50 developing and transition countries. The statistical results indicate that, in education, both the overall level of public spending and intrasectoral allocation matter; in particular, shifting spending toward primary and secondary education is associated with improvements in widely used measures of education attainment. In the health sector, increased overall health spending is associated with reduced infant and child mortality rates. We do not, however, prove that higher spending causes such improvements—although the results that we present provide some, but indefinite, evidence of causality. We also address the issue of the appropriate functional form of the relationship between social indicators and public spending.

The paper is organized as follows. Section 2 discusses the model and the data set; Section 3 presents the results; and Section 4 sets out the policy implications.

2. Model and Data

We use the following equation to evaluate the impact of public spending on education and health care:

where Yi is a social indicator reflecting education attainment or health status for a country i, which is a function of aggregate public spending on education or health care as a share of GDP,5,6 Χ1i; allocations to different programs within the sector (i.e., primary education and primary health care) as a share of total sectoral spending,7 Χ2i; and a vector of socioeconomic variables, Zi.

A range of social indicators is available to gauge performance of education and health care spending by the government. Three considerations guided the choice of indicators. First, to facilitate a comparison of results, indicators used by other authors were selected where possible. Second, because many indicators are collected infrequently and with a lag, the indicators used were those for which the most up-to-date values were available. Finally, as many as possible of the core indicators proposed by the Development Assistance Committee (DAC) of the OECD, the World Bank, and the UN to measure development performance were used.8

Education attainment is proxied by the gross enrollment ratio in primary and secondary education (the number of enrolled students in percent of the total number of school-age persons), the persistence through grade four (percent of children reaching that grade), and the primary-school drop-out rates. Two indicators are used to gauge health status: infant (aged 0 to 1 year) mortality rates and child (aged 0 to 5 years) mortality rates.9

In addition to two expenditure variables, the education regressions include the following control variables.

  • Percent of population in the age group 0-14. It is difficult and costly to expand enrollment rates in countries with low enrollment when the population is relatively young (Mingat and Tan, 1992). A high share of young in the population would be expected to be negatively correlated with enrollment rates.

  • Per capita income. As household incomes rise, the relative cost of enrolling children in school is reduced, suggesting that increasing income would be associated with rising enrollments. Furthermore, at higher income levels, the demand for education increases, if education is a normal good. This effect is captured by GDP per capita in purchasing power parity (PPP) terms.

  • Urbanization. Households in urban areas are more likely to send their children to school, because, among other reasons, access to education is typically better in urban areas (Plank, 1987). In addition, the private cost of education (e.g., transportation costs) may be lower for urban households.

  • Child nutrition. Better nutrition for children makes it easier for enrolled school-age children to continue in school, thereby affecting enrollment and persistence (Glewwe and Jacoby, 1995). This variable is proxied by child mortality.10

Control variables in the health regressions include the following.

  • Per capita income. Empirical evidence suggests that the population’s health status improves as per capita incomes rise.

  • Adult illiteracy rates. As for education, many studies show a strong inverse relationship between adult illiteracy and infant mortality rates (e.g., Tresserras et al., 1992). A number of studies indicate that female literacy affects the health status of infants and children (see, for example, Schultz, 1993). However, due to data limitations, the overall—rather than the female—adult illiteracy rate is used.11

  • Access to sanitation and safe water. A sanitary environment, as reflected by increased access to sanitation and safe water, improves health status. Access to safe water, for example, has a significant effect on infant and child mortality (Kim and Moody, 1992; and Hojman, 1996). Because of data limitations, access to sanitation is used in the regressions.

  • Urbanization. Schultz (1993) finds that mortality is higher for rural, low-income, agricultural households, suggesting that increased urbanization is associated with improved health status of the population.

Data limitations prevent adding other controls for socioeconomic characteristics that may affect indicators of education attainment and health status. In particular, private spending on both education and health is omitted due to a lack of data (evidence on the importance of private spending is provided by Psacharopoulos and Nguyen, 1997). Similarly, data limitations prevent including control variables that capture the factors adversely affecting children’s caregivers (for example, the impact of the AIDS epidemic in Africa).12

Some authors (Bredie and Beeharry, 1998; and Filmer et al., 1998) propose including other demand factors, such as income distribution, in the regressions for both education and health. While data are available for only a small subsample, the regression results are tested for robustness by including the Gini coefficient.

Finally, Mingat and Tan (1998) point to the importance of teachers’ salaries in increasing the cost of education in low-income countries. They estimate that 50% of the difference in education attainment between high-income and low-income countries can be attributed to lower teachers’ salaries in relation to the rest of the economy in high-income countries that release resources for nonwage inputs, such as textbooks. Data on teacher salaries are not available for a sufficient number of countries to use as a control variable. However, as relative teacher salaries are highly correlated with illiteracy and child mortality rates,13 the latter control variables are expected to pick up much of the effect of differences in salary levels.

Data on education and health spending are drawn from a number of sources, including various issues of the IMF’s Government Finance Statistics (GFS) and Recent Economic Development reports. These expenditure data, in general, exclude local government spending. This can be a major deficiency in countries that have devolved expenditure responsibilities to lower levels of government.

For around 50 developing and transition countries, the intrasectoral allocation for education spending (primary, secondary, and tertiary) and health care spending (primary, secondary, and other) are taken from GFS and UNESCO databases, the World Bank’s Poverty Assessments, and Public Expenditure Reviews.14 Most spending and other data are for 1993-1994. The intrasectoral data have been taken from the same source as aggregate allocations to that sector where possible and were checked for consistency with aggregate spending.15 Available data on education spending typically do not distinguish between primary and secondary education. Consequently, spending on these two levels is analyzed as a single item.

A universally accepted definition of primary health care does not exist. As a result, intrasectoral data for health care are not strictly comparable across sources. Primary health care is defined as public spending on clinics and practitioners according to the GFS categorization.16 For countries for which this classification is not available, public spending on primary health care or preventive care, as defined in the World Bank’s Poverty Assessments and Public Expenditure Reviews, is used. Secondary health care is defined as hospital services and curative treatments by medical specialists.

Data for control variables and the indicators of education attainment and health status were drawn mostly from the World Development Indicators database. Data on primary drop-out rates are from Barro and Lee (1996) and Gini coefficients are from Deininger and Squire (1996). To the extent possible, the data for the health and education indicators and control variables were matched with the year of the spending data.17

Figs. 13 present average public spending levels and intrasectoral shares of education and health care spending in the sample countries. These figures show that in the sample countries, the share of education expenditures allocated to primary and secondary education is 79%, whereas the share of health care spending allocated to primary care is 16%. These numbers are broadly consistent with average intrasectoral allocations previously observed by others (World Bank, 1993; Sahn and Bernier, 1993).

Figure 1.Total Education and Health Care Spending

(Percent of GDP)

Sources: UNESCO database; World Bank, Public Expenditure Review (Washington, various issues); World Bank, Poverty Assessment (Washington, various issues); IMF, GFS database; and IMF staff estimates.

Figure 2.The Intrasectoral Shares of Education Spending

(In 50 selected countries)

Sources: UNESCO database; World Bank, Public Expenditure Review (Washington, various issues); World Bank, Poverty Assessment (Washington, various issues); IMF, GFS database; and IMF staff estimates.

Figure 3.The Intrasectoral Shares of Health Care Spending

(In 40 selected countries)

Sources: UNESCO database; World Bank, Public Expenditure Review (Washington, various issues); World Bank, Poverty Assessment (Washington, various issues); IMF, GFS database; and IMF staff estimates.

Empirical Results

Eq. (1) is estimated using OLS (correcting for heteroskedasticity) and two-stage least squares (2SLS) regressions. In the first instance, we follow functional forms used previously in the literature—that is, linear and log-log specifications—for ease of comparison of our results with those previously obtained. For example, Hojman (1996) previously tested a linear specification. Pritchett and Summers (1996) use log-log specifications to estimate the nonlinear relationship between income and health. The log-log specification provides the added convenience of yielding ready elasticity estimates. We employ statistical tests to assess the appropriateness of the functional forms.

This section is set up as follows. The appropriate functional form—whether linear or log-log—for Eq. (1) is tested using both the Mackinnon, White, and Davidson (MWD) and Ramsey’s RESET test of up to three fitted terms. As it turns out, the issue of functional form is difficult to resolve in some cases. In addition, stylized facts pertaining to social indicators, such as diminishing returns to increased spending, suggest alternative specifications in addition to the linear and log-log specifications. In particular, Anand and Ravallion (1993) created an index of health status to reflect diminishing returns to scale. A similar index has been proposed by Kakwani (1993). This issue is taken up in the section on robustness tests.

The 2SLS technique is used primarily to address the problem of reverse causality. For instance, higher spending on primary education may have a positive effect on enrollment, but a higher demand for primary education, reflected in higher enrollment rates, may also provide a push for higher spending. A similar dual relationship may exist between public spending on primary health care, on the one hand, and child and infant mortality rates, on the other. In addition, 2SLS regressions address potential problems of measurement errors in variables.18

However, it is difficult to find the appropriate instruments because it is also not clear a priori which variables determine social expenditures but do not affect social indicators. The literature on the determinants of social spending (for example, Looney and Frederiksen, 1996; Gbesemete and Gerdtham, 1992) suggest that foreign aid and military spending may explain some of the variance in the share of public spending on education and health. These variables are tested as potential instruments, along with other variables, in the 2SLS regression. The results should not be interpreted as providing strong evidence for the existence of a causal relationship between spending and social indicators because this study relies on cross-section data.

In evaluating the regression results, it should be borne in mind that multicollinearity among variables affects the standard errors of coefficients on the control variables. The literature on the determinants of social spending, for example, suggests that per capita income is a significant predictor of social spending.

A final note of caution is related to the number of observations. This can vary, depending on the availability of data for a specific variable. The number of observations for the health care regressions is relatively low.

Education Regressions

Table 1 reports results of the education attainment regressions for the linear specification. Four measures of education attainment are used: gross enrollment in primary and secondary education, gross enrollment in secondary education, persistence through Grade 4, and primary school drop-out rates. The MWD test of functional forms suggests that the linear form is appropriate for regressions for drop-out rates and persistence through Grade 4 (therefore, we do not present results for the log-log specification in the table). This is supported by Ramsey’s RESET test of up to three fitted terms. However, both tests are agnostic as to the preferred functional form for the regressions on secondary enrollment, while the linear form is weakly preferred for the regression on primary plus secondary enrollment. The explanatory variables account for 40–83% of cross-country variation in education attainment. The F-statistic for all regressions is statistically significant at the 1% level.19

Table 1.Regression Results for Education Indicators: Linear Regressionsa
Enrollment Rates
Gross primary and secondaryGross secondaryPersistence Through

Grade 4

Drop-Out Rates
Primary and secondary
education spending0.17**0.080.28***0.200.17**0.25**−0.43*−0.21
(percent of total
education spending)(2.31)(0.82)(3.34)(0.21)(2.19)(2.14)(−1.89)(−0.68)
Education spending1.68*3.20**2.26**2.62**1.592.73−5.19**−6.49**
(percent of GDP)(1.86)(2.46)(2.18)(2.16)(0.29)(1.57)(2.06)(−2.45)
Population aged 0-140.29**0.32**−0.98***−1.03***−0.07−0.191.16***1.13***
(percent of population)(2.25)(2..10)(−4.93)(−5.22)(−0.32)(−0.59)(3.93)(3.58)
Child mortality rate−0.16**−0.18**−0.05−0.05−0.10*−
(per thousand of children 0-5 years)(−2.54)(−2.47)(−1.47)(−1.42)(−1.73)(−1.40)(0.44)(0.52)
Income per capita in PPP termsb0.340.610.15**0.16**0.10−0.25−0.98−0.55
Urbanization (percent of population)0.27***0.35***0.41***0.37***
Adjusted R-squared0.670.640.820.830.500.450.420.39
Number of observations4239454224233838
Sargan’s P-value0.
Source: Authors’ calculations.The instruments used are aid per capita, aid in percent of government expenditures, military spending in percent of government expenditures, share of unallocated education spending, and total government spending.

Except for columns (1) and (2), where regular t-statistics are shown in parentheses, White’s het-eroskedasticity-corrected statistics are shown.

Multiplied by 100.

Indicates significance at the 10 percent level.

Indicates significance at the 5 percent level.

Indicates significance at the 1 percent level.

Source: Authors’ calculations.The instruments used are aid per capita, aid in percent of government expenditures, military spending in percent of government expenditures, share of unallocated education spending, and total government spending.

Except for columns (1) and (2), where regular t-statistics are shown in parentheses, White’s het-eroskedasticity-corrected statistics are shown.

Multiplied by 100.

Indicates significance at the 10 percent level.

Indicates significance at the 5 percent level.

Indicates significance at the 1 percent level.

Total education spending in relation to GDP is significant in all but the regressions for persistence through Grade 4 at the conventional levels of significance. On the other hand, the share of spending on primary plus secondary education in total education spending is statistically significant for the OLS specifications, but not the 2SLS specification, possibly reflecting the weakness of the instruments. Sargan’s test suggests, however, that the 2SLS specifications for all regressions are correct, except possibly for persistence through Grade 4, at the 10% level of significance. The first-stage adjusted R-squared for overall education spending and the intrasectoral spending are 0.51 and 0.29, respectively; the F-statistics are significant at the 1% level.20 Finally, results from regressions with gross primary enrollment and net secondary enrollment (not reported) also suggest that the intrasectoral allocation and total level of education spending matter.21 This contrasts with the findings of a positive but insignificant correlation between spending and enrollment (Flug et al., 1998).

The results show that socioeconomic variables, such as urbanization, the percent of the population in age group 0-14, and per capita income, are important in explaining variances in enrollment rates. Except for percent of population aged 0-14, all the other variables have signs that are consistent with our expectations. Significance varies across indicators. Urbanization is a strong predictor of enrollment rates. Because of multicollinearity, however, the level of significance of the control variables should be interpreted with caution; nevertheless, findings presented here are broadly consistent with the empirical literature on determinants of education attainment.

Robustness Tests

The literature suggests a nonlinear relationship between individual incomes and social indicators.22 This relationship can be taken into account by including a measure of within-country income distribution in the regression of aggregate indicators. When the Gini coefficient is added to the linear education regressions, the sample size drops by about a third and the spending variable loses significance. When added to the log-log regressions, however, the spending variables remain significant.

We also added dummy variables for regions. This addresses, to some extent, the lack of data on some important control variables, such as a proxy for the impact of the AIDS epidemic in Africa. In general, the inclusion of dummy variables for regions did not improve the explanatory power of the regression models, nor did it affect the coefficient estimates and significance levels. Similarly, the results are robust to the inclusion of adult illiteracy rates which are a proxy for parental education. As expected, illiteracy is negatively correlated with educational attainment.

Running the regressions in log-log form yields results similar to those of the linear form (not reported in the paper). The share of spending in primary plus secondary education is significant at the 5% confidence level or better across all education indicators, whereas total education spending in percent of GDP is significant in two of four indicators (gross secondary enrollment regressions and in the primary school dropout rates). Per capita income and urbanization are significant in the primary school drop-out rates but not in other regressions.

Finally, we estimated an ad hoc system of four equations, to allow for the “production” of multiple goods with increased spending, including gross primary enrollment rates, gross secondary enrollment rates, persistence to Grade 4 and primary drop-out rates. Following Barro and Lee (1997) and Wenger (2000), we estimated this system using the seemingly unrelated regression (SUR) technique. We allowed for different intercepts and slope coefficients for each independent variable.23 Both the share of spending on primary and secondary education and total education spending are significant at conventional levels of significance in all education regressions, except gross primary enrollment rate where the adjusted R-squared itself is low. The other control variables vary in significance; in cases where they are significant, however, the direction is as expected. The results are in Appendix B, Table B2.

Several conclusions can be drawn from the education regressions. First, despite the lack of data on some control variables, the regressions explain a large part of the cross-country variation in enrollment rates. Second, the intrasectoral distribution of public spending for education generally has a statistically significant effect on indicators of both access and education attainment. Third, the overall level of education spending has a statistically significant impact on all indicators except for persistence through Grade 4.24

Table 2 reports the results of an analysis of partial variances. For selected regressions, the adjusted R-squared of models of education attainment, with and without the spending variables, are compared. This analysis indicates that including the spending variables increases the explained cross-country variation in education attainment by between 4% and 15%.25

Table 2.Adjusted R-Squared of the OLS Education Attainment Regressions




Grade 4


Excluding spending variables0.780.410.27
Including total education spending0.800.450.37
Including share of primary and secondary education0.820.500.42
Source: Authors’ calculations.
Source: Authors’ calculations.

The magnitude of the impact of education spending on education attainment can be put in perspective by examining some of the relevant coefficient estimates. For instance, based on the estimates of the OLS regression in column (3) of Table 1, a 5 percentage point increase in the share of outlays for primary and secondary education in total public expenditures for education increases gross secondary enrollment by over one percentage point. A 1 percentage point of GDP increase in spending on education increases gross secondary enrollment by more than 3 percentage points. Although this shows that spending and its intrasectoral allocation have an important impact on education attainment, it also indicates that increasing attainment through shifting intrasectoral allocations or increasing total spending on education alone may be very difficult. This illustrates the importance of control variables in explaining education attainment.

Mingat and Tan (1998) suggest another reason why the marginal costs of increasing indicators of education attainment are so high. They demonstrate for a sample of 125 countries that, as primary enrollment increases, resources earmarked for primary education are shifted toward decreasing pupil-teacher ratios (this shift in focus begins to occur at primary enrollment rates as low as 50%). Consequently, these additional resources do not significantly increase enrollment rates or persistence.

Health Regressions

Table 3 reports the results of log-log regressions with infant and child mortality rates as dependent variables. The MWD test of functional forms as well as Ramsey’s RESET test of up to three fitted terms suggests that the log-log form is the appropriate functional form. The log-log form follows the Cobb-Douglas production function described by Filmer et al. (1998). In addition, the empirical literature has used mainly the log-log specification (see, for example, Pritchett and Summers, 1996; Filmer and Pritchett, 1997; Filmer et al., 1998; and Wang, 2001).

Table 3.Regression Results for Health Indicators: Log-Log Regressionsa






Primary health care spending0.
(percent of total health care spending)(0.50)(0.24)(0.85)(0.31)
Health spending−0.31***−0.26*−0.30***−0.29**
(percent of GDP)(−3.65)(−1.91)(−3.73)(−2.11)
Adult illiteracy rate0.31***0.31**0.32***0.30**
(percent of population 15 years or older)(4.18)(2.26)(3.75)(2.08)
Income per capita in PPP terms−0.31***−0.33**−0.36**−0.37*
(percent of population)(−1.77)(1.43)(−2.43)(−1.62)
Access to sanitation0.200.20*0.210.20
(percent of population)(1.69)(1.73)(1.35)(1.31)
Adjusted R-squared0.840.830.830.82
Number of observations22222222
Sargan’s P-value0.790.86
Source: Authors’ calculations.The instruments used are aid per capita, aid in percent of government expenditures, military spending in percent of government expenditures, and total government spending.

Except for columns (1) and (2), where regular t-statistics are shown in parentheses, White’s heteroskedasticity-corrected statistics are shown in parentheses.

Indicates significance at the 10 percent level.

Indicates significance at the 5 percent level.

Indicates significance at the 1 percent level.

Source: Authors’ calculations.The instruments used are aid per capita, aid in percent of government expenditures, military spending in percent of government expenditures, and total government spending.

Except for columns (1) and (2), where regular t-statistics are shown in parentheses, White’s heteroskedasticity-corrected statistics are shown in parentheses.

Indicates significance at the 10 percent level.

Indicates significance at the 5 percent level.

Indicates significance at the 1 percent level.

On average, the explanatory variables account for more than 80% of cross-country variation in infant and child mortality rates. The F-statistic for all regressions is statistically significant at the 1% level. Sargan’s specification test suggests that the instruments are correctly specified.

Total health spending is statistically significant in all regressions, but the share of primary health spending is not.26 As in the education regressions, control variables are important in explaining variances in health care status. For example, both adult illiteracy rate and income per capita are consistently significant in all the regressions. In addition, the income elasticities are comparable with previous estimates. For example, Pritchett and Summers (1996) estimate that the long-run income elasticity of infant and child mortality in developing countries lies between −0.2 and −0.4. The results in Table 1 indicate that the income elasticity in our sample is about −0.3. The elasticity of infant and child mortality with respect to health spending in percent of GDP is thrice that of Filmer and Pritchett (1997). On the other hand, Filmer et al. (1998) estimated higher income elasticities.

Robustness Tests

The health regressions are robust to various specifications. When dummy variables for regions are added, total health spending remains significant, with estimated elasticity of health status with respect to spending roughly about the same. The signs of the dummy variables also suggest that other regions have better health status than sub-Saharan Africa, on average.

While our test of functional form suggests that the log-log specification is appropriate, Anand and Ravallion (1993) have suggested that the proper specification should take into account deceasing returns to scale in the improvement of health. This was noted by Kakwani (1993) as well. This can be done using an index that transforms mortality rates into a new variable that reflects decreasing rates of mortality reduction.27 We replicate this index for our sample and find that overall health care spending remains a significant determinant of infant and child mortality, but with higher elasticities than those reported in Table 3. The estimated elasticities are also higher than those of income. Primary health care spending is not statistically significant.

In order to account for the concave relationship between health attainment and individual incomes, a control for income distribution was added. The results are robust to the inclusion of the Gini coefficient. Owing the lack of Gini data, however, the sample size drops significantly.

An ad hoc system of two equations using the SUR technique accounts for the multiple production of health output with increased spending. The relevant results are in Appendix B, Table B3. We considered different intercept and slope coefficients and also the same intercepts and slope coefficients.28 The results remain essentially unchanged. Total health spending, adult illiteracy, and per capita income are significant at the 1% level. The estimated elasticities are approximately the same as those reported in Table 3.

Table 4 reports the results of partial variances for health regressions. These suggest that the health spending variable may explain as much as an additional 6–9% of cross-country variation in health status. This contrasts with the results of Filmer and Pritchett (1997), who found that the contribution of health outlays to health care status as measured by child mortality rates was almost negligible (less than ⅙ of 1%).

Table 4.Adjusted R-Squared of the OLS Health Status Regressions


Excluding spending variables0.750.77
Including health spending0.840.83
Source: Authors’ calculations.
Source: Authors’ calculations.

Taking the results reported in columns (1) and (3) of Table 3, the coefficient estimates suggest that increasing the share of total health care spending in GDP by 1 percentage point decreases child and infant mortality rates by about 3 death per 1,000. This suggests significant gains from increased health spending.

Conclusions and Policy Implications

We have provided evidence supporting the proposition that increased public spending on education and health care matter for education attainment and health status, although definitive evidence for a causal relationship is lacking. The evidence is strongest for education. The relationship is weaker for health.29

Greater public spending on primary and secondary education has a positive impact on widely used measures of education attainment, and increased health care spending reduces child and infant mortality rates. For example, a 5 percentage point increase in the share of outlays for primary and secondary education increases gross secondary enrollment by over one percentage point. A 1 percentage point increase in health care spending decreases infant and child mortality rates by about 3 per 1,000 live births. If expenditure allocations for education and health care are to boost economic growth and promote the well-being of the poor, policy makers in many developing and transition economies need to pay greater attention to allocations within these sectors. These allocations—both their size and efficiency—are an important vehicle for promoting equity and furthering second-generation reforms.

Some caution, however, is required in using these results for estimating the budgetary resources needed for achieving objectives.30 Education and health are also affected by per capita income, urbanization, adult illiteracy, and access to safe sanitation and water. Private sector spending also matters.

Appendix A

Countries with Intrasectoral Education Spending Data













Czech Republic


El Salvador








Iran, Islamic Republic













Papua New Guinea




Sierra Leone

St. Vincent and the Grenadines

Syrian Arab Republic









Countries with Intrasectoral Health Spending Data










El Salvador

















Netherlands Antilles

Papua New Guinea





St. Vincent and the Grenadines

Syrian Arab Republic



Trinidad and Tobago




Appendix B
Table B1.Summary Statistics(Individual samples as indicated)

Gross primary enrollment96.9100.027.0130.019.945
Gross secondary enrollment50.349.
Gross primary and secondary75.280.020.0105.018.844
Drop-out rates, primary28.721.
Persistence to Grade 489.892.763.7100.09.824
Primary and secondary education spending78.980.219.895.512.850
Education spending3.
Income per capita in PPP terms3,9823,63314413,3702,96450
Population aged 0–1437.639.718.151.69.550
Adult illiteracy rate21.416.91.179.318.248
Urbanization (percent of population)49.751.214.689.819.950
Child mortality rate65.351.59.1269.056.250
Health Care
Infant mortality42.035.59.1137.035.640
Child mortality60.942.011.0233.059.639
Primary health care spending16.312.70.050.615.140
Health care spending2.
Immunization against measles82.286.022.0100.015.739
Access to sanitation69.270.010.0100.026.133
Adult illiteracy rate19.111.60.568.019.838
Income per capita in PPP terms4,2583,17015615,1813,37739
Urbanization (percent of population)
Sources: See text.
Sources: See text.
Table B2.SUR Regression Results for Education Indicators(t-statistics in parentheses)
Gross PrimaryGross SecondaryPrimary Drop-outGrade 4
Primary and secondary education spending0.090.29***−0.46**0.18**
(percent of total education spending)(0.46)(2.61)(−2.34)(2.13)
Education spending(0.70)2.24**−5.44***1.65*
(percent of GDP)(0.42)(2.30)(−2.91)(1.72)
Income per capita in PPP termsa0.0260.15**−0.710.13
Population aged 0–140.35−1.00***1.19***−0.09
(percent of population)(1.00)(−4.66)(3.07)(−0.52)
Urbanization (percent of population)0.020.41***0.060.11
Child mortality rate−0.21***−0.050.04−0.10**
(per thousand children 0-5 years)(3.59)(−1.39)(0.66)(−2.11)
Adjusted R-squared0.240.820.420.50
Number of observations45453824
Source: Authors’ calculations.

Indicates significance at the 10% level.

Indicates significance at the 5% level.

Indicates significance at the 1% level.

Coefficient estimates in linear regressions are multiplied by 100.

Source: Authors’ calculations.

Indicates significance at the 10% level.

Indicates significance at the 5% level.

Indicates significance at the 1% level.

Coefficient estimates in linear regressions are multiplied by 100.

Table B3.SUR Regression Results for Health Indicators: Log-Log Regressions(t-statistics in parentheses)










Primary health spending0.
(percent of total health spending)(0.42)(0.59)(0.51)(0.51)
Health spending−0.31***−0.30***−0.30***−0.30***
(percent of GDP)(−3.71)(−3.22)(−3.58)(−3.58)
Access to sanitation0.
(percent of population)(1.31)(1.22)(1.30)(1.30)
Adult illiteracy rate0.31***0.32***0.31***0.31***
(percent of population(4.84)(4.37)(4.75)(4.75)
15 years or older)
Income per capita in PPP terms−0.31***−0.36***−0.33***−0.33***
(percent of population)(−1.18)(−1.23)(−1.23)(−1.23)
Adjusted R-squared0.840.830.810.77
Number of observations22222222
Source: Authors’ calculations.

Indicates significance at the 1% level.

Source: Authors’ calculations.

Indicates significance at the 1% level.


    AiyerS.JamisonD.T.LondoñoJ.L.1995. Health policy in Latin America: Progress, problems, and policy options. Cuadernos de Economía 3211-28.

    AnandS.RavallionM.1993. Human development in poor countries: On the role of private incomes and public services.Journal of Economic Perspectives 7133-50.

    AndrewsD.BooteA.R.RizaviS.S.SinghS.1999. Debt relief for low-income countries. IMF Pamphlet Series No. 51International Monetary FundWashington, DC.

    AppletonS.HoddinottJ.MackinnonJ.1996. Education and health in sub-saharan Africa. Journal of International Development 8307-39.

    AppletonS.MackinnonJ.1996. Enhancing human capacities in Africa. In: NduluB. and others (Eds.) Agenda for Africa’s Economic Renewal. Overseas Development CouncilWashington, DC109-49.

    BarroR.J1991. Economic growth in a cross-section of countries. Quarterly Journal of Economics 106407-44.

    BarroR.J.LeeJ.W.1996. International measures of schooling years and schooling quality. American Economic Review 86218-23.

    BarroR.J.LeeJ.W.1997. Schooling quality in a cross section of countries. NBER Working Paper No. 6198National Bureau of Economic ResearchCambridge, MA.

    BehrmanJ. R.BirdsallN.SzekelyM.1998. Intergenerational schooling mobility and macro condition and schooling policies in Latin America. IDB Office of the Chief Economist working paper no. 386Inter-American Development BankWashington, DC.

    BennellP.1995. Rates of return to education in Asia: a review of the evidence. Working paper no. 24Institute of Development StudiesOxford.

    BennellP.1996. Rates of return to education: does the conventional pattern prevail in sub-Saharan Africa. World Development 24183-200.

    BidaniB.RavallionM.1997. Decomposing social indicators using distributional data. Journal of Econometrics 77125-39.

    BredieJ.W.B.BeeharryG.K.1998. School enrollment decline in sub-Saharan Africa: Beyond the supply constraint. World Bank discussion paper no. 395Washington, DC.

    CarrinG.PolitiC.1995. Exploring the health impact of economic growth, poverty reduction, and public health expenditure. Tijdschrift voor Economie en Management 40227-246.

    CassenR.1996. Human development: research and policy choicesODC occasional paper no. 3Overseas Development CouncilWashington, DC.

    ChuK. et al.1995. Unproductive public expenditures: a pragmatic approach to policy analysis. IMF pamphlet series no. 48International Monetary FundWashington, DC.

    DeatonA.2001. Health, inequality, and economic development. WHO Commission on Macroeconomics and Health (CMH) working paper no. WG1:3World Health OrganizationGeneva.

    DeiningerK.SquireL.1996. A new data set measuring income equality. World Bank Economic Review 3565-591.

    DemeryL.WaltonM.1998. Are Poverty Reduction and Other 21st Century Social Goals Attainable?World BankWashington, DC.

    FilmerD.HammerJ.PritchettL.1998. Health policy in poor countries: weak links in the chain. Policy research working paper no. 1874World BankWashington, DC.

    FilmerD.PritchettL.1997. Child mortality and public spending on health: how much does money matter?Policy research working paper no. 1864World BankWashington, DC.

    FlugK.SpilimbergoA.WachtenheimE.1998. Investment in education: do economic volatility and credit constraints matter?Journal of Development Economics 55465-81.

    GallagherM.1993. A public choice theory of budgets: implications for education in less developed countries. Comparative Education Review 3790-106.

    GbesemeteK. P.GerdthamU.G.1992. Determinants of health care expenditure in Africa: a cross-sectional study. World Development 20303-308.

    GlewweP.JacobyH.G.1995. An economic analysis of delayed primary school enrollment in a low-income country: the role of early childhood nutrition. Review of Economics and Statistics 77156-69.

    Guin-SuiM.T.YamadaG.CorbachoA.1999. The effects of public spending on education and health on social development indicators: a panel data analysis. MimeoInternational Monetary FundWashington, DC.

    GuptaS.ClementsB.TiongsonE.1998. Public Spending on Human Development. Finance and Development 3510-13. MelloL.SharanR.2001. Corruption and military spending. European Journal of Political Economy 17794-77.

    HojmanD.E.1996. Economic and other determinants of infant and child mortality in small developing countries: the case of Central America and the Caribbean. Applied Economics 28281-90.

    KakwaniN.1993. Performance in living standards: an international assessment. Journal of Development Economics 41307-36.

    KimK.MoodyP.M.1992. More resources better health? a cross-national perspective. Social Science & Medicine 34837-42.

    LandauD.1986. Government and economic growth in the less developed countries: an empirical study for 1960–80. Economic Development and Cultural Change 3535-75.

    LooneyR.E.FrederiksenP.C.1996. Defense expenditures and budgetary patterns in selected Middle Eastern and Mediterranean countries: an assessment. Public Budgeting and Financial Management 893-105.

    MauroP.1998. Corruption and the composition of government expenditure. Journal of Public Economics 69263-79.

    McGuireA.ParkinD.HughesD.GerardK.1993. Econometric analyses of national health expenditures: can positive economics help answer normative questions?Health Economics 2113-26.

    MehrotraS.1998. Education for all: Policy lessons from high-achieving countries. UNICEF staff working papers Evaluation Policy and Planning Series No. 98-005UNICEFNew York, NY.

    MehrotraS.BucklandP.1998. Managing teacher costs for access and quality. UNICEF staff working papers Evaluation Policy and Planning Series No. 98-004UNICEFNew York, NY.

    MingatA.TanJ.-P.1992. Education in Asia: A Comparative Study of Cost and Financing. World BankWashington, DC.

    MingatA.TanJ.-P.1998. The mechanics of progress in education: evidence from cross-country data. Policy research working paper no. 2015World BankWashington, DC.

    MusgroveP.1996. Public and private roles in health: theory and financing patterns. World Bank discussion paper no. 339Washington, DC.

    NossA.1991. Education and adjustment: a review of the literature. Policy Research and External Affairs working paper WPS 701World BankWashington, DC.

    OECD (Organisation for Economic Co-operation and Development)/DAC (Development Assistance Committee)1996. Shaping the 21st century: the contribution of development cooperation. Paris.

    OgbuO.M.GallagherM.1991. On public expenditures and delivery of education in sub-Saharan Africa. Comparative Education Review 35295-318.

    PlankD.N.1987. The expansion of education: a Brazilian case study. Comparative Education Review 31361-76.

    PradhanS.1996. Evaluating public spending: a framework for public expenditure reviews. World Bank discussion paper no. 323Washington, DC.

    PritchettL.SummersL.H.1996. Wealthier is healthier. Journal of Human Resources 31841-68.

    PsacharopoulosG.1994. Returns to investment in education: a global update. World Development 221325-43.

    PsacharopoulousG.NguyenN.X.1997. The role of government and the private sector in fighting poverty. World Bank technical paper no. 346Washington, DC.

    RavallionM.1992. Does undernutrition respond to incomes and prices?Dominance tests for Indonesia. World Bank Economic Review 6109-24.

    SahnD.BernierR.1993. Evidence from Africa on the intrasectoral allocation of social sector expenditures. Cornell Food and Nutrition Policy Program working paper no. 45Cornell UniversityIthaca.

    SarganJ.D.1964. Wages and prices in the United Kingdom: a study in econometric methodology. In: HartP.E.MillsG.WhitakerJ.K. (Eds.) Econometric Analysis for National Economic Planning. Butterworth, London25-63.

    SchultzT.P.1993. Mortality decline in the low income world: causes and consequence. Economic Growth Center discussion paper no. 681Yale UniversityNew Haven, CT.

    SenA.1999. Economic policy and equity: an overview. In: TanziV.ChuK.GuptaS. (Eds.) Economic Policy and Equity. International Monetary FundWashington, DC28-42.

    ShawR.P.GriffinC.C.1995. Financing Health Care in Sub-Saharan Africa Through User Fees and Insurance. World BankWashington, DC.

    TanziV.ChuK. (Eds.) 1998. Income Distribution and High-Quality Growth. MIT PressCambridge, MA.

    TresserrasR.J.C.AlvarezJ.SentisJ. and SallerasL.1992. Infant mortality, per capita income, and adult illiteracy: an ecological approach. American Journal of Public Health 82435-37.

    WagstaffA.van DoorslaerE.2000. Income inequality and health: what does the literature tell us?Annual Review of Public Health 21543-67.

    WangL.2001. Health outcomes in poor countries and policy options: a summary of empirical findings from DHS data. MimeoWorld BankWashington, DC.

    WengerJ.W.2000. What do schools produce? Implications of multiple outputs in education?”Contemporary Economic Policy 1827-36.

    WhiteH.1980. A heteroskedasticity-consistent covariance matrix and a direct test for heteroskedasticity. Econometrica 48817-38.

    World Bank1993. World Development Report 1993: Investing in Health. Oxford University PressNew York, NY.

    World Bank1995. Priorities and Strategies for Education: A World Bank Review. World BankWashington, DC.

This chapter is reprinted from the European Journal of Political Economy, Vol. 18, Gupta et al., “The Effectiveness of Government Spending on Education and Health Care in Developing and Transition Economies,” pp. 717-37, © 2002, with permission from Elsevier.

The views expressed in this paper are those of the authors and do not necessarily represent those of the IMF. The authors wish to thank Benedict Clements, Hamid Davoodi, Luiz de Mello, Robert Gillingham, Henry Ma, Edgardo Ruggiero, Christian Schiller, Gustavo Yamada, and two anonymous referees for their helpful comments on the earlier drafts.

Amartya Sen (1999) has correspondingly proposed that “since premature mortality, significant undernourishment, and widespread illiteracy are deprivations that directly impoverish human life, the allocation of economic resources as well as arrangements for social provision must give some priority to removing these disadvantages for the affected population.” This typically requires improvement in the provision of basic education and primary health care.

Increasing education and health care spending in the poorest countries is a central element of the recently launched initiative granting debt relief to Heavily Indebted Poor Countries (Andrews et al., 1999).

The methodological basis of studies estimating social rates of return of education has however been questioned. For example, Bennell (1995, 1996) does not find support for the proposition that basic education has a higher social return than other levels of education. See also Appleton et al. (1996) and Cassen (1996).

Also, the absence of a measurable impact of public spending on indicators could be due to a differential effect on poor and nonpoor groups, which is not captured by aggregated social indicators (Bidani and Ravallion, 1997).

A consequence of measuring education spending as a share of GDP is that the associated spending per student can vary greatly among countries depending on the level of GDP. The results presented in this paper also hold, however, when education and health care spending is expressed in per capita terms. To capture the impact of income, GDP per capita was included as a control variable (see below). Thus, the effect of per capita spending can be gauged from the coefficients for spending as a percent of GDP and GDP per capita (since the product of these variables equals spending per capita). See also footnote 6.

A simple “production function” for health should yield the same point estimate and significance for the effect of public spending on health outcomes, regardless of which measure of public spending on health is used (Filmer et al., 1998). The difference is the coefficient on income. For example, suppose the production function for health is of the following form: Y=(H/N)α(NH/N)β(e)A, where Y is a measure of health status; H is public spending on health; and NH is the “rest of GDP.” Dividing the numerators and denominators by GDP and taking logs gives the following equation: lnY=α*lnH/GDP+β*lnNH/GDP+(α+β)*lnGDP/N+A.

It should be noted that an increase in public allocations for, say, primary education, while holding all other spending constant, has an effect on education indicators both directly through Χ2i, and indirectly through the overall level of education spending Χ1i.

The list of core indicators for education and health includes: net enrollment in primary education, persistence through grade four, literacy rate of 15- to 24-year-olds, adult literacy rate, infant mortality rate, child mortality rate, maternity mortality ratio, births attended by skilled health personnel, contraceptive prevalence rate, HIV infection rate in 15- to 24-year-old pregnant women, and life expectancy at birth. See

The relevant descriptive statistics are provided in Appendix B.

Other proxies of child nutrition, such as indicators of malnourishment and birth weight, were not available.

In fact, female illiteracy was found to have a weaker effect than overall illiteracy.

An attempt was made to circumvent the problem of missing control variables by adding dummies for regions, under the assumption that the variation of omitted controls within regions is dominated by the variance among regions. The results are robust to the inclusion of these dummy variables.

The correlation coefficient between illiteracy rates and average teacher salaries as a multiple of GDP per capita was −0.80 for 24 countries for which data were available (data on teachers’ salaries are from Mehrotra and Buckland, 1998). The correlation coefficient with child mortality rates was −0.72. The correlation coefficient between income per capita in PPP terms and the relative teacher salaries was also relatively high at −0.48.

The list of countries is included in Appendix A.

If the deviation between the sum of intrasectoral spending and total sectoral spending exceeded 10%, the observation was dropped.

This measure of primary health care, which includes services provided by clinics and medical, dental, and paramedical practitioners, appropriately captures primary-level health care, as it is the “first point of contact” between clients and a facility in a health system (e.g., Shaw and Griffin, 1995). The GFS disaggregation of health spending—into hospitals, clinics and practitioners, and other spending—is also used by others to examine priorities in the health sector (e.g., Appleton and Mackinnon, 1996).

For example, intrasectoral education data for 1994 were matched with enrollment data for 1994, if available. If enrollment data for 1994 were not available, observations in the range of three years before and after the year of spending were used (1991-1997). Potential problems of measurement error were addressed by running two-stage least squares regressions.

The data set includes some outlying observations (for example, Papua New Guinea). However, these outliers did not critically affect the regression results after corrections for heteroskedasticity were made.

To address heteroskedasticity, White’s (1980) corrected covariance and standard errors are used, except for the equation with gross primary plus secondary enrollment as the dependent variable. The latter regression was estimated using the weighted-least-squares (WLS) technique, with adult illiteracy used as a weight. This weight can be interpreted as a scaling factor, indicative of the challenge of achieving targeted levels of education attainment, and yields better results than White’s corrected regression. The use of a consistent set of instruments in the 2SLS regressions was checked for validity using Sargan’s (1964) general misspecification test.

Foreign aid as a proportion of government expenditure is strongly correlated with the endogenous variables; military spending is negatively correlated with the overall level of public spending on education.

The coefficient estimate of the share of spending on primary plus secondary education from the WLS regression with gross primary enrollment as the dependent variable is 0.21. The coefficient estimate from an OLS regression with net secondary enrollment as the dependent variable is 0.19. Both are significant at the 5% level. OLS regressions with the spending and education attainment variables in logs yield similar results for gross primary and secondary enrollment and gross secondary enrollment as dependent variables, but the statistical significance of the intrasectoral spending variable for persistence through Grade 4 regression is reduced.

Deaton (2001) and Wagstaff and van Doorslaer (2000) review the theoretical and empirical literature on the nonlinear relationship at the individual level. See also Ravallion (1992). The literature is based on micro studies of the determinants of health, although a nonlinear relationship has been suggested in education as well (see, for example, Behrman et al., 1998).

A Wald test of equality of the slope coefficients indicates that they are different.

The regressions do not permit drawing up of conclusions about the effect of changes in the level of spending on primary and secondary education—as opposed to the share of such spending in total education expenditure. This issue was addressed by re-estimating the education regressions including spending on primary and secondary education as a percentage of GDP and omitting the variables for intrasectoral spending and the overall spending. In the four regressions for enrollment, this newly defined spending variable was significant at the 1% level; the coefficient estimated ranges between 3.0 and 4.0. In the two regressions for persistence through Grade 4, spending on primary and secondary education as a percent of GDP was only significant at the 10% level, with a coefficient of 2.7 for the OLS regression and 5.5 for the 2SLS regression. These results suggest that, irrespective of the specification, spending for the two sectors matters.

Partial variance analysis only yields accurate results if the underlying assumption on the ordering of casual effects is correct (i.e., partial variance analysis assumes here that public spending affects social indicators only after all other variables have taken effect). Alternatively, the results of partial variable analysis would be correct if spending were to have an effect independent of the other explanatory variables. These are demanding assumptions, and the results presented here should be interpreted with caution.

These results should be interpreted with caution. First, because of the above-noted lack of a uniform definition of primary health care, the intrasectoral distribution is not measured consistently across the sample. Second, the sample size is relatively small. Third, the sample used for the health regressions includes eight observations that have zero spending on primary health care, which could reflect institutional differences in these countries (e.g., all primary health care could be private), or simply measurement error. Finally, in the case of 2SLS regressions, the results may reflect the weakness of the instruments. The test for over-identifying restrictions suggests that the chosen instruments are appropriate; however, while the F-statistic for the first-stage regression is significant at the 1% level for overall health spending, the F-statistic for the share of primary health spending is not. In particular, foreign aid is strongly correlated with overall level of health spending. The adjusted R-squared for the first-stage regression is 0.67.

This index is defined for a given country i as [ln(Max − Min) − lnMRi − Min)]/ln(Max − Min), where Max is the maximum of observed mortality rate, Min is the minimum of observed mortality rate, and MRi is the mortality rate observed in country i. As the mortality rate in country i approaches the minimum of observed mortality rates, the index for country i approaches 1.

However, the null hypothesis that the slope coefficients are equal was rejected in a Wald test.

Research using panel data (Guin-Sui et al., 1999) suggests that there are no apparent gains from longitudinal analysis in terms of stronger or more robust results. As regards the functional form, we find that of the specifications used hitherto in the literature (linear and log-log), there is evidence that linear specification is more appropriate for the education regressions and the log-log specification for the health regressions. The results are robust to alternative functional forms, such as the nonlinear specification for health indicators following Anand and Ravallion (1993) and Kakwani (1993).

The Development Assistance Committee (DAC) of the OECD, building on the results of the 1995 Social Summit in Copenhagen, has established goals that include reaching universal enrollment in primary education and reducing infant and child mortality by two-thirds in all developing countries by 2015 (OECD/DAC, 1996).

    Other Resources Citing This Publication