## _wp06180

## Source details

**Canonical URL:** [_wp06180](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06180.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06180.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06180.pdf.json)

---

### I. Introduction and objectives
- MDGs established by OECD DAC in 1996 and reinforced at the United Nations Millennium Assembly in September 2000; aims include eradicating poverty and hunger, achieving universal primary education, promoting gender equality and maternal health, reducing child mortality, ensuring environmental sustainability, and combating diseases.
- Heavily Indebted Poor Countries (HIPC) Initiative launched in 1996; in mid-2005 supplemented by a multilateral debt relief proposal to eliminate debts of HIPC countries owed to the IMF, World Bank, and African Development Bank at a projected cost of about $55 billion.
- Paper objective: examine historically whether debt-service savings and increased grants translated into rising expenditures on health and education; combine estimates with effects of social expenditures on MDG-related social indicators to assess resource gaps for MDGs.
- Methodological note: using observed changes in debt service avoids assumptions on concessionality but may understate the true effect if changes in debt service are driven by factors other than relief.
- Hypotheses:
  - Middle-income countries (MICs): debt ratio may be a significant determinant of expenditures due to borrowing capacity and sensitivity of spreads to debt levels.
  - Low-income countries (LICs): more likely credit constrained; changes in debt-service likely to have greater effects on social expenditures.

### II. Data sample and stylized trends
- Sample: developing countries publishing health and education expenditures (about 110 countries); data from IMF Fiscal Affairs Department database; time period 1985–2003/2004.
- Social expenditure trends:
  - LICs: social expenditures rose from about 4 percent of output in 1985 to almost 6 percent in 1992, then declined thereafter.
  - MICs: social expenditure ratio rose fairly consistently over time, most recent estimate at almost 8 percent of output.
  - Disparity: narrowed to almost zero in 1992, then rose to about 2 percent of output in 2004.
- Budgetary movements:
  - Budgetary movement defined as change in government budget balance divided by lagged output.
  - Inverse relationship observed between sharp budgetary movements and social expenditures for both LICs and MICs (oil-exporting Arab countries excluded because of volatility).
  - LICs: change in budget balance became sharply negative in early 1990s at over 4 percent per year while social expenditure ratio peaked.
- Debt-service and grants:
  - Debt-service-to-exports declined through 1995 for LICs and MICs; divergence afterward.
  - For LICs, debt-service ratio projected to fall to 5 percent by 2007; debt-service in relation to output projected to decline by 1 percent among HIPC countries through 2007.
  - Grants to LICs more than doubled relative to output between 1985 and 1994, declined thereafter, then trended up with HIPC; grants to MICs gradually declined, currently about 1 percent of GDP on average.
  - Possible explanation: donor fiscal consolidation in mid-1990s reduced overseas development assistance from 0.33 percent to 0.22 percent of donor output between 1992 and 1997.
- Debt stock measures:
  - NPV of debt-to-exports fell since introduction of HIPC; averaged about 200 percent of exports in 2004 for LICs.
  - MICs: debt ratio about 80 percent of exports in 2003.

### III. Econometric approach and validity
- Estimation method:
  - Dynamic panel specification with lagged dependent variable; Arellano-Bond procedure used (first differences and instruments lagged at least two periods).
  - Instruments include second lag of social expenditure ratio, one-period lags of debt-service and debt ratios (interacted with MIC/LIC dummies), lagged levels of aid ratio, youth literacy rate, output per capita, population density; positive and negative changes in budget balance interacted with MIC/LIC dummies.
- Specification tests:
  - Hansen statistic indicates instruments exogenous (over-identifying restrictions insignificant in main specifications).
  - Arellano-Bond test: first-order negative serial correlation present; no second-order correlation.

### IV. Key empirical findings on determinants of social expenditures
- Summary distinction by income group:
  - MICs:
    - Debt-service coefficients insignificant for MICs — suggests MICs are not credit constrained with respect to education and health spending.
    - Debt ratio coefficients significantly negative for MICs — higher debt stock reduces social expenditures. Example: at a debt-to-export ratio of 200 percent, social expenditures are lower by ¼ percentage point of output relative to a country with no debt.
    - Budget consolidation: a 1 percentage point rise in the budget balance in MICs leads to a 0.04-0.05 percent decline in the ratio of social expenditures to output.
  - LICs:
    - Debt-service coefficients significantly negative in terms of both exports and output; stock of debt insignificant — consistent with credit constraints making LIC expenditures sensitive to flow (debt-service) rather than stock.
    - Governments have insulated social expenditures from positive budgetary consolidation, but social expenditures are sensitive to declines in the budget balance with coefficients ranging between -0.08 and -0.1.
- Other significant covariates:
  - Population density significant at or above 90 percent confidence for both debt-service measures.
  - Youth literacy significant at 90 percent in the equation with debt-service measured in output; positive coefficient implies countries with low literacy rates may have curtailed education spending due to resource constraints.
  - Examples reported:
    - High-density countries (Malta and Mauritius) spend about 2 percent of output less on health than the lowest density country (Mongolia).
    - Niger with youth literacy about 20 percent spends over 2 percentage points of output less on health and education than Guyana with youth literacy 100 percent.
- Selected coefficient estimates (as reported):
  - Lagged dependent Variable: 0.856 ***, 0.857 ***, 0.83 ***, 0.852 ***
  - Debt service–export ratio for LICs: -0.018 *
  - Debt service–export ratio for MICs: 0.007, 0.007
  - Debt service–output ratio for LICs: -0.048 **, -0.038
  - Debt service–output ratio for MICs: 0.02, 0.02
  - Debt ratio for LICs: 0.0003, 0.001, 0.0002, 0.001
  - Debt ratio for MICs: -0.001 **, -0.005 *, -0.002 ***, -0.006 **
  - Aid/GDP: 0.01, 0.006, 0.014, 0.006
  - Positive change in budget balance for LICs: 0.045, 0.04, 0.043, 0.031
  - Negative change in budget balance for LICs: -0.096 ***, -0.084 **, -0.087 **, -0.072 *
    - Note: The hypothesis of no effect of changes in the budget balance on social expenditures cannot be rejected for low-income countries after 1995.
  - Positive change in budget balance for MICs: -0.047 **, -0.044 **, -0.026, -0.019
  - Negative change in budget balance for MICs: -0.042 **, -0.046 *, -0.055 ***, -0.058 **
  - Youth literacy rate: 0.004, 0.005 *, 0.003, 0.004
  - Output per capita (logarithm): 0.103, 0.091, 0.153, 0.1
  - Population density (logarithm): -0.097 **, -0.109 **, -0.095 **, -0.099 **
  - Dummy for Middle East oil exporting countries: -0.47 ***, -0.49 ***, -0.61 ***, -0.58 ***
  - Test statistics examples: Hansen test of overidentified restrictions: 75.575, 52.962, 2.2; A-Bond test for AR(1): -4.21 ***, -4.12 ***, -4.44 ***, -4.3 ***; A-Bond test for AR(2): 0.36, 0.38, 0.24, 0.28
  - Number of observations: 771, 772, 771, 772
- Robustness:
  - Inclusion of time dummies: makes debt-service in relation to output insignificant for LICs, but debt-service in relation to exports remains significantly negative and comparable to baseline.
  - Dropping about 150 influential observations did not affect debt-service coefficients.

### V. Fiscal incidence and expenditure composition
- Changes in debt-service have no impact on the total expenditure envelope, suggesting debt-service savings have been targeted toward social expenditures in particular.
- Aid/GDP is significantly positive for total expenditures, suggesting this type of aid appears to be channeled to non-social expenditures.
- Positive changes to the budget balance have not influenced the expenditure ratio among LICs, suggesting the budgetary boost has come from higher revenues; this is also the case for MICs.
- Since 1995, the sensitivity of expenditures to budget declines has increased significantly among LICs; the coefficient varies between -0.85 and -0.95 for the post-1995 period.
- Both social and other expenditures are highly sensitive to declines in the budget balance among LICs; the social expenditure share is insulated from increases in the budget balance for LICs (in contrast to MICs), but the social expenditure share declines for all countries when the budget balance worsens.
- The social expenditure ratio is closely related to per-capita income: richer countries place more resources into achievement of social objectives.

### VI. MDG achievement likelihoods and resource needs
- Child mortality (MDG): target is a 67 percent decline on average between 1990 and 2015 (projection based on maintaining the average annual decline over 1990–2003 through 2015).
- Child mortality and health expenditure indicators (Middle-Income Countries / Low-Income Countries):
  - Average mortality rate for all countries (2003): 32.6, 117.9
  - Average mortality rate for countries projected to miss the MDG target (2003): 98.5, 138.2
  - Percent projected to achieve MDG target 1/: 63.0, 24.1
  - Average health expenditures for all countries (2003): 2.6, 1.8
  - Health expenditures for countries projected to miss the MDG target (2003, in percent of GDP): 2.1, 2.0
  - Increased health expenditures needed to reach target (in percent of GDP, Gupta and others (2002)): 2.1, 2.2
    - 1/ Countries are projected to reach the MDG target if the annual decline in child mortality over 1990–2003 is higher than the annual decline required between 2004 and 2015 to achieve a 67 percent in the rate.
- Elasticities and implications for health:
  - Gupta and others (2003) estimate an elasticity of 0.3 between changes in health expenditure and declines in the infant mortality rate. Using this estimate, health expenditures need to rise by more than 2 percentage points per annum for the low-income countries unlikely to reach the target without additional financing.
  - Filmer and Pritchett (1997) estimate an elasticity of -0.1; applying this would imply health expenditures would need to rise by more than 6 percentage points of output to reach the MDGs.
- Primary education (MDG): full coverage by 2015.
- Primary enrollment and education expenditure indicators (Middle-Income Countries / Low-Income Countries):
  - Average enrollment rate for all countries (in percent): 92.9, 67.8
  - Average enrollment rate for countries projected to miss the MDG target (in percent): 86.6, 57.1
  - Countries at full enrollment (in percent): 31.1, 11.5
  - Countries projected to achieve MGD target by 2015 (in percent) 1/: 60.6, 30.8
  - Average education expenditures for all countries (in percent of GDP): 4.7, 4.1
  - Education expenditures for countries projected to miss the MDG target (in percent of GDP): 5.6, 3.9
  - Increased education expenditures needed to reach target:
    - Gupta and others (2002): 3.3, 8.4 (in percent of GDP)
    - Baldacci and others (2004): 1.3, 6.4 (in percent of GDP)
    - Mignat and others (2003): 3.5 (in percent of GDP)
    - 1/ Countries are projected to reach the MDG target if the annual increase in primary enrollment over 1990–2003 is sufficient to achieve a 100 percent primary enrollment rate by 2015.
- Interpretation of education cost estimates:
  - Gupta and others (2002): a 1 percent increase in education expenditures to output raises enrollment by 3 percent — implies >8 percent of output additional infusion for low-income countries.
  - Baldacci and others (2004): a 1 percentage point increase in education expenditures leads to a 0.16 percentage point increase in enrollment — implies an additional infusion of 6 ½ percent of GDP needed.
  - Mignat and others (2003): country-by-country analysis implies on average about 3 ½ percent of output is needed to achieve full primary enrollment among LICs — consistent with current resource cost of producing a 60 percent enrollment rate.

### VII. Quantitative projections and policy implications
- Projected impacts of debt-service declines on social expenditures:
  - Projected decline in debt-service ratio by 1 percent of GDP on average between 2004 and 2007 in LICs is associated with a long-run boost to social expenditures of about 0.35 percent of output.
  - Projected decline in debt-service relative to exports by 5 percentage points over same interval associated with a long-run rise in social expenditures of about 0.5 percent of output.
  - Using the largest coefficient estimates, the projected 1 percent decline in debt-service in relation to output through 2007 would raise social expenditures by only about 0.4 percent of output.
- Grants versus debt-service:
  - Impact on social expenditures of decline in debt-service is significantly stronger than for increase in grants; coefficient on grants is barely positive, consistent with grants being targeted to projects not overlapping social expenditures and difficulty in targeting marginal aid to social spending.
- Policy-relevant implications:
  - For LICs, reductions in debt-service via debt relief can translate into meaningful increases in social spending; however, grant increases alone may have limited impact on social expenditure shares unless better targeted.
  - For MICs, debt stock reduction appears important for enabling higher social expenditures; budgetary consolidation in MICs tends to compress social spending.
- Overall conclusion:
  - The multilateral debt relief initiative (HIPC) relaxes the budget constraint for HIPC countries and is likely to boost social expenditures and improve MDG indicators, but the magnitude of these effects is dwarfed by the financial resources needed to reach the MDGs by 2015.

*Source: _wp06180 - References..............................................................................................................*

### References..............................................................................................................

### _wp06180 - References..............................................................................................................

### I. INTRODUCTION
- MDGs established by OECD DAC in 1996 and reinforced at the United Nations Millennium Assembly in September 2000; aims include eradicating poverty and hunger, achieving universal primary education, promoting gender equality and maternal health, reducing child mortality, ensuring environmental sustainability, and combating diseases.
- Heavily Indebted Poor Countries (HIPC) Initiative launched in 1996; in mid-2005 supplemented by a multilateral debt relief proposal to eliminate debts of HIPC countries owed to the IMF, World Bank, and African Development Bank at a projected cost of about $55 billion.
- Literature summary:
  - Gupta and others (2002): increased public expenditure on education associated with improvements in school access and attainment; increased public health expenditure reduces child mortality.
  - Bidani and Ravallion (1997): supportive earlier work on health expenditures.
  - Kim and Moody (1992); Filmer and Pritchett (1997): find public health outlays contribution to health status insignificant in some studies.
  - Clements, Bhattacharya, and Nguyen (2003): ratio of debt-service to output is significantly negatively related to public investment; coefficient of 0.2 implies a decline in the debt-service ratio from 10 percent to 5 percent of GDP would raise the public investment rate by about 1 percent of GDP.
  - Kraay and Chauvin (2005): argue debt relief has not contributed to significant change in health and education expenditures; note noisy debt-relief data and need for simplifying assumptions on concessionality.
- Methodological point: using observed changes in debt service avoids assumptions on concessionality but may understate the true effect if changes in debt service are driven by factors other than relief.
- Hypothesis on income groups:
  - Middle-income countries (MICs): debt ratio may be a significant determinant of expenditures due to borrowing capacity and sensitivity of spreads to debt levels (Mody and Saravia (2003); Hilscher and Nosbusch (2004)).
  - Low-income countries (LICs): more likely credit constrained; changes in debt-service likely to have greater effects on social expenditures.
- Paper objective: examine historically whether debt-service savings and increased grants translated into rising expenditures on health and education; combine estimates with effects of social expenditures on MDG-related social indicators to assess resource gaps for MDGs.

### II. DATA SAMPLE
- Sample: developing countries publishing health and education expenditures (about 110 countries); data from IMF Fiscal Affairs Department database; time period 1985–2003/2004.
- Social expenditure trends:
  - LICs: social expenditures rose from about 4 percent of output in 1985 to almost 6 percent in 1992, then declined thereafter.
  - MICs: social expenditure ratio rose fairly consistently over time, most recent estimate at almost 8 percent of output.
  - Disparity: narrowed to almost zero in 1992, then rose to about 2 percent of output in 2004.
- Budgetary movements:
  - Defined as change in government budget balance divided by lagged output.
  - Inverse relationship observed between sharp budgetary movements and social expenditures for both LICs and MICs (oil-exporting Arab countries excluded because of volatility).
  - LICs: change in budget balance became sharply negative in early 1990s at over 4 percent per year while social expenditure ratio peaked.
- Debt-service and grants:
  - Debt-service-to-exports declined through 1995 for LICs and MICs; divergence afterward. For LICs, debt-service ratio projected to fall to 5 percent by 2007; debt-service in relation to output projected to decline by 1 percent among HIPC countries through 2007.
  - Grants to LICs more than doubled relative to output between 1985 and 1994, declined thereafter, then trended up with HIPC; grants to MICs gradually declined, currently about 1 percent of GDP on average.
  - Possible explanation: donor fiscal consolidation in mid-1990s reduced overseas development assistance from 0.33 percent to 0.22 percent of donor output between 1992 and 1997.
- Debt stock measures:
  - NPV of debt-to-exports fell since introduction of HIPC; averaged about 200 percent of exports in 2004 for LICs.
  - MICs: debt ratio about 80 percent of exports in 2003.

### III. ECONOMETRIC ANALYSIS
- Estimation approach:
  - Dynamic panel specification with lagged dependent variable; Arellano-Bond procedure used (first differences and instruments lagged at least two periods).
  - Instruments include second lag of social expenditure ratio, one-period lags of debt-service and debt ratios (interacted with MIC/LIC dummies), lagged levels of aid ratio, youth literacy rate, output per capita, population density; positive and negative changes in budget balance interacted with MIC/LIC dummies.
  - Hansen statistic indicates instruments exogenous (over-identifying restrictions insignificant); first-order negative serial correlation present; no second-order correlation.
- Key empirical findings on determinants of social expenditures:
  - MICs:
    - Debt-service coefficients (Table 2, columns 1 and 2) insignificant for MICs — suggests MICs are not credit constrained with respect to education and health spending.
    - Debt ratio coefficients significantly negative for MICs — higher debt stock reduces social expenditures. Example: at a debt-to-export ratio of 200 percent, social expenditures are lower by ¼ percentage point of output relative to a country with no debt.
    - Budget consolidation: a 1 percentage point rise in the budget balance in MICs leads to a 0.04- 0.05 percent decline in the ratio of social expenditures to output.
  - LICs:
    - Debt-service coefficients significantly negative in terms of both exports and output; stock of debt insignificant — consistent with credit constraints making LIC expenditures sensitive to flow (debt-service) rather than stock.
    - Governments have insulated social expenditures from positive budgetary consolidation, but social expenditures are sensitive to declines in the budget balance with coefficients ranging between -0.08 and -0.1.
  - Other variables:
    - Population density significant at or above 90 percent confidence for both debt-service measures.
    - Youth literacy significant at 90 percent in the equation with debt-service measured in output; positive coefficient implies countries with low literacy rates may have curtailed education spending due to resource constraints.
    - Examples: high-density countries (Malta and Mauritius) spend about 2 percent of output less on health than the lowest density country (Mongolia); Niger with youth literacy about 20 percent spends over 2 percentage points of output less on health and education than Guyana with youth literacy 100 percent.
- Robustness checks:
  - Inclusion of time dummies: makes debt-service in relation to output insignificant for LICs, but debt-service in relation to exports remains significantly negative and comparable to baseline.
  - Dropping about 150 influential observations did not affect debt-service coefficients.
- Quantitative projections based on estimates:
  - Projected decline in debt-service ratio by 1 percent of GDP on average between 2004 and 2007 in LICs is associated with a long-run boost to social expenditures of about 0.35 percent of output.
  - Projected decline in debt-service relative to exports by 5 percentage points over same interval associated with a long-run rise in social expenditures of about 0.5 percent of output.
- Grants versus debt-service:
  - Impact on social expenditures of decline in debt-service is significantly stronger than for increase in grants; coefficient on grants is barely positive, consistent with grants being targeted to projects not overlapping social expenditures and difficulty in targeting marginal aid to social spending.
- Policy-relevant implication:
  - For LICs, reductions in debt-service via debt relief can translate into meaningful increases in social spending; however, grant increases alone may have limited impact on social expenditure shares unless better targeted.
  - For MICs, debt stock reduction appears important for enabling higher social expenditures; budgetary consolidation in MICs tends to compress social spending.

*Source: _wp06180 - References..............................................................................................................*

### introduction, the HIPC initiative was introduced in 1996 emphasizing the importance of

### _wp06180 - introduction, the HIPC initiative was introduced in 1996 emphasizing the importance of

### Sensitivity of social expenditures to budgetary changes
- A dummy for low-income countries for the period since 1995 was interacted with the budget balance; the coefficient was insignificant, so the sensitivity of social expenditures to an increase in the budget balance has remained unchanged.
- Low-income countries have managed to fully insulate social expenditures from the effects of fiscal tightening over the past two decades, protecting social expenditures during budgetary consolidation.

### Regression highlights for social expenditures (selected coefficient estimates as reported)
- Lagged dependent Variable: 0.856  *** 0.857  *** 0.83  *** 0.852  ***
- Debt service–export ratio for low-income countries (LICs): -0.018  *
- Debt service–export ratio for middle-income countries (LICs): 0.007 0.007
- Debt service–output ratio for low-income countries: -0.048  ** -0.038
- Debt service–output ratio for middle-income countries: 0.02 0.02
- Debt ratio for LICs: 0.0003 0.001 0.0002 0.001
- Debt ratio for MICs: -0.001  ** -0.005  * -0.002  *** -0.006  **
- Aid/GDP: 0.01 0.006 0.014 0.006
- Positive change in budget balance for low-income countries: 0.045 0.04 0.043 0.031
- Negative change in budget balance for low-income countries: -0.096  *** -0.084  ** -0.087  ** -0.072  * 1/
  - 1/ The hypothesis of no effect of changes in the budget balance on social expenditures cannot be rejected for low-income countries after 1995.
- Positive change in budget balance for middle-income countries: -0.047  ** -0.044  ** -0.026 -0.019
- Negative change in budget balance for middle-income countries: -0.042  ** -0.046  * -0.055  *** -0.058  **
- Youth literacy rate: 0.004 0.005 * 0.003 0.004
- Output per capita (logarithm): 0.103 0.091 0.153 0.1
- Population density (logarithm): -0.097  ** -0.109  ** -0.095  ** -0.099  **
- Dummy for Middle East oil exporting countries: -0.47  *** -0.49  *** -0.61  *** -0.58  ***
- Test statistics reported (examples): Hansen test of overidentified restrictions: 75.575 52.962 2.2; A-Bond test for AR(1): -4.21  *** -4.12  *** -4.44  *** -4.3  ***; A-Bond test for AR(2): 0.36 0.38 0.24 0.28
- Number of observations: 771 772 771 772
- Sources used for regressions: FAD database; GFS database; World Bank.

### Findings on total expenditures and expenditure shares
- Changes in debt-service have no impact on the total expenditure envelope, suggesting debt-service savings have been targeted toward social expenditures in particular.
- Aid/GDP is significantly positive for total expenditures, suggesting this type of aid appears to be channeled to non-social expenditures.
- Positive changes to the budget balance have not influenced the expenditure ratio among low-income countries, suggesting the budgetary boost has come from higher revenues; this is also the case for middle-income countries.
- Since 1995, the sensitivity of expenditures to budget declines has increased significantly among low-income countries; the coefficient varies between -0.85 and -0.95 for the post-1995 period.
- Both social and other expenditures are highly sensitive to declines in the budget balance among low-income countries; the social expenditure share is insulated from increases in the budget balance for low-income countries (in contrast to middle-income countries), but the social expenditure share declines for all countries when the budget balance worsens.
- The social expenditure ratio is closely related to per-capita income: richer countries place more resources into achievement of social objectives.

### Likelihood of achieving the MDG targets — child mortality and health expenditures
- Child mortality MDG: target is a 67 percent decline on average between 1990 and 2015 (projection based on maintaining the average annual decline over 1990–2003 through 2015).
- Selected indicators (Middle-Income Countries / Low-Income Countries as presented):
  - Average mortality rate for all countries (2003): 32.6 117.9
  - Average mortality rate for countries projected to miss the MDG target (2003): 98.5 138.2
  - Percent projected to achieve MDG target 1/: 63.0 24.1
  - Average health expenditures for all countries (2003): 2.6 1.8
  - Health expenditures for countries projected to miss the MDG target (2003, in percent of GDP): 2.1 2.0
  - Increased health expenditures needed to reach target (in percent of GDP, Gupta and others (2002)): 2.1 2.2
  - 1/ Countries are projected to reach the MDG target if the annual decline in child mortality over 1990–2003 is higher than the annual decline required between 2004 and 2015 to achieve a 67 percent in the rate.
- Elasticity evidence and implication:
  - Gupta and others (2003) estimate an elasticity of 0.3 between changes in health expenditure and declines in the infant mortality rate. Using this estimate, health expenditures need to rise by more than 2 percentage points per annum for the low-income countries unlikely to reach the target without additional financing.
  - Filmer and Pritchett (1997) estimate an elasticity of -0.1; applying this would imply health expenditures would need to rise by more than 6 percentage points of output to reach the MDGs.

### Likelihood of achieving the MDG targets — primary education and education expenditures
- Primary enrollment MDG: full coverage by 2015.
- Selected indicators (Middle-Income Countries / Low-Income Countries as presented):
  - Average enrollment rate for all countries (in percent): 92.9 67.8
  - Average enrollment rate for countries projected to miss the MDG target (in percent): 86.6 57.1
  - Countries at full enrollment (in percent): 31.1 11.5
  - Countries projected to achieve MGD target by 2015 (in percent) 1/: 60.6 30.8
  - Average education expenditures for all countries (in percent of GDP): 4.7 4.1
  - Education expenditures for countries projected to miss the MDG target (in percent of GDP): 5.6 3.9
  - Increased education expenditures needed to reach target (in percent of GDP, Gupta and others (2002)): 3.3 8.4
  - Increased education expenditures needed to reach target (in percent of GDP, Balducci and others (2004)): 1.3 6.4
  - Increased education expenditures needed to reach target (in percent of GDP, Mignat and others (2003)): 3.5
  - 1/ Countries are projected to reach the MDG target if the annual increase in primary enrollment over 1990–2003 is sufficient to achieve a 100 percent primary enrollment rate by 2015.
- Interpretation of education cost estimates:
  - Gupta and others (2002): a 1 percent increase in education expenditures to output raises enrollment by 3 percent — implies >8 percent of output additional infusion for low-income countries (viewed as unrealistically high).
  - Baldacci and others (2004): a 1 percentage point increase in education expenditures leads to a 0.16 percentage point increase in enrollment — implies an additional infusion of 6 ½ percent of GDP needed.
  - Mignat and others (2003): country-by-country analysis implies on average about 3 ½ percent of output is needed to achieve full primary enrollment among low-income countries — consistent with current resource cost of producing a 60 percent enrollment rate.

### Implications of debt-service savings and overall resource gaps
- Debt-service savings appear targeted toward social expenditures rather than the total expenditure envelope.
- Taking the largest coefficient estimates from the paper, the projected 1 percent decline in debt-service in relation to output through 2007 would raise social expenditures by only about 0.4 percent of output.
- Conclusion: the multilateral debt relief initiative (HIPC) relaxes the budget constraint for HIPC countries and is likely to boost social expenditures and improve MDG indicators, but the magnitude of these effects is dwarfed by the financial resources needed to reach the MDGs by 2015.

### Concluding observations on debt, income, and social spending
- Declines in debt-service costs among low-income countries help raise health and education expenditures significantly.
- Higher debt ratios have no bearing on health and education expenditures among low-income countries but adversely impact such expenditures among middle-income countries.
- Possible interpretation: low-income countries are constrained in the financing they can receive and therefore spend donor financing directly on social sectors; middle-income countries can smooth consumption and adjust social expenditures in response to longer-lasting economic shocks.

*Source: _wp06180 - introduction, the HIPC initiative was introduced in 1996 emphasizing the importance of (PDF chapter/section).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06180.pdf_