## 2. External Debt and Economic Reform in Countries with Low Ranking of Economic Freedom

## Source details

**Canonical URL:** [2. External Debt and Economic Reform in Countries with Low Ranking of Economic Freedom](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0750.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0750.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0750.pdf.json)

---

### Motivation and main hypothesis
- Question: If economic reform is the necessary “treatment” for a sick economy, why is it often delayed?
- Central hypothesis: Easily available external financing (foreign borrowing) can delay necessary economic reform by reducing the immediate cost of the status quo and by enabling competing socioeconomic groups to “wait out” concession—external financing acts as a “pain reliever.”
- Theoretical mechanism emphasized:
  - Alesina and Drazen (1991) war-of-attrition framework extended to allow private-sector foreign borrowing.
  - Private and public foreign borrowing lower the short-term cost of delaying reform (reducing distortionary taxation or financing lobbying), increasing the time to concession T, up to the point where servicing debt becomes too costly.

### Theoretical model — key structure and conditions
- Baseline fiscal setup at time t = 0: balanced budget g + r b0 = τ with b0 entirely external.
- For 0 < t < T:
  - Government increases spending gt > g0; share γ < 1 of spending paid by taxes, remainder (1 − γ) financed by external borrowing.
  - Taxes: τ(t) = γ [g(t) + r b(t)] (equation (1) in source).
  - Change of public external debt: ḃ(t) = (1 − γ)[g(t) + r b(t)] (equation (2)).
- Private-sector borrowing li reduces distortionary tax disutility θ(li); li known only to group i and drawn from uniform distribution F(l) on [l, l¯].
- Stock of external debt of group i at time t: li e−rt (equation (6) structure).
- Key comparative and first-order conditions:
  - Inequality (12) in source defines when winner is truly a winner (exact expression preserved in source).
  - First-order condition for concession time T yields expression (18); sign of dT/dli depends on whether the loser is credit-constrained.
  - Condition for loser to be able to borrow enough so reform is delayed is inequality (19) in source (exact expression preserved in source).
- Main theoretical implication:
  - If foreign borrowing is relatively easily available to all groups (i.e., groups are not credit constrained and (19) holds), private-sector foreign borrowing further delays reform.
  - When international capital markets stop financing public and private sectors (as in crises), reform follows quickly because borrowing cannot finance the status quo.

### Empirical strategy and data
- Reform measure: change in Fraser Institute index of economic freedom (index range 1 to 10; average of five sub-indexes: government size, legal system & property rights, sound monetary policy, freedom to trade, regulation strictness & efficiency).
- Sample: all developing and emerging economies with available data, period 1970–2002, total of 81 countries (Annex II).
- Time aggregation: five-year periods (three years for the last observation).
- Main dependent variable: R = change in economic freedom index over each five-year period.
- Key regressors and controls:
  - B: dummy grouping countries in six groups increasing in the change of their external debt-to-GDP ratio (detailed grouping rules preserved in source).
  - F: level of economic freedom index at beginning of period.
  - GR: crisis dummy if average real GDP growth rate below -1 percent in previous period.
  - INF: crisis dummy if average inflation rate above 40 percent in previous period.
  - AID: average aid flows to GDP.
  - OP: trade-to-GDP ratio.
  - Interaction term B*F to test whether impact of foreign borrowing depends on prior reform.

### Empirical findings — reform regressions and robustness
- Simple correlations:
  - Strong negative correlation between change in external debt-to-GDP ratio and change in economic freedom index and its five subindexes; similar pattern for countries with low ranking of economic freedom.
- Main panel regression results (Table 1 and text summary):
  - Estimate of external borrowing (B) is negative and statistically significant at the 1 percent level in all specifications.
  - Switching to the next borrower group is correlated with a reduction in the reform index by about 0.4. This is double the average change in the index over a five-year period in the sample.
  - Level of economic freedom F has a negative and significant coefficient.
  - INF (inflation crises) leads to reform (positive and significant); GR (growth crisis) positive but not statistically significant.
  - Interaction B*F positive and significant: the negative link between external borrowing and reform is stronger in countries with lower prior reform.
  - Adjusted R-squared values reported (Table 1): 0.15, 0.27, 0.30, 0.30, 0.24.
- Lags and IV robustness (Table 2 and text):
  - Lagged external borrowing B(-1) has a negative and statistically significant coefficient (about −0.09 to −0.11 in reported columns).
  - Inclusion of B(-1) suggests countries that borrow a lot in one period reform less in the current and the next periods.
  - Instrumental variable: U.S. short-term interest rate used to instrument external borrowing; IV results indicate estimate of external borrowing remains negative and statistically significant.
  - Adj. R-squared values in Table 2: 0.27, 0.28, 0.21, 0.25, 0.10, 0.12.
- Subindex analysis (Table 3): rising external debt is linked to slower progress in most subindexes:
  - D(F1) government size: B = −0.09*** (t = −5.95)
  - D(F2) legal system & property rights: B = −0.07 (not significant)
  - D(F3) sound money: B = −0.36*** (t = −5.95) — largest effect
  - D(F4) freedom to trade: B = −0.11*** (t = −2.85)
  - D(F5) regulation: B = −0.07** (t = −2.43)
  - Adj. R-squared values in Table 3: 0.18, 0.21, 0.27, 0.19, 0.34.
- Probit and ordered probit (Table 4):
  - Probit columns: B coefficients −0.35*** and −0.42*** (z-statistics −3.41, −3.64).
  - Ordered probit columns: B coefficients −1.68**, −0.24***, −0.28***, −0.48 (varied significance).
  - Lagged B(-1) shows negative significance in several probit specifications (e.g., −0.21*; −0.23**).
  - Adj. R-squared values reported: 0.15, 0.29, 0.28.
- Growth regressions linking debt and growth (Table 5):
  - Dependent variable: average real GDP per capita growth over five-year non-overlapping sub-periods; estimation with fixed effects.
  - External debt/GDP coefficients reported in columns include:
    - −0.012*** (t = −4.162)
    - −0.026*** (t = −4.893)
    - −0.016** (t = −2.450)
  - Change of external debt/GDP coefficients:
    - −0.021*** (t = −3.168)
    - −0.013** (t = −1.909)
  - Other growth determinants (examples from Table 5):
    - Real per capita GDP coefficients: −4.233***, −3.353***, −6.439*** (t-statistics shown).
    - Secondary school enrollment coefficient 0.055** (t = 2.539).
    - Investment/GDP coefficient 0.311*** (t = 7.647).
    - Inflation rate coefficient −0.001*** (t = −3.490).
    - Growth in trading partners coefficient 0.559*** (t = 3.552).
    - Trade/GDP coefficient 0.031** (t = 2.403).
  - Adj. R-squared values in Table 5: 0.28, 0.37, 0.60.

### Interpretation and implications
- Consistent empirical pattern: higher external borrowing (especially increases in external debt) is associated with delayed economic reform across the aggregate index and most subindexes (strongest for sound money; weaker for legal/institutional reform).
- Causality caveats and mitigants:
  - Reverse causality is possible (reforms attract capital inflows).
  - Evidence mitigating reverse causality concerns:
    - Interaction term B*F shows stronger negative effect where prior reform is low.
    - Lagged B(-1) retains negative significance.
    - IV estimation using U.S. short-term interest rate still yields negative significant effect of external borrowing on reform.
  - Available evidence suggests at least part of effect runs from external borrowing to delayed reform.
- Link to growth:
  - Both the level and the change of external debt-to-GDP negatively affect real per capita growth in panel regressions.
  - Proposed channel: easy external financing reduces pressure to reform, which in turn slows long-run growth; change in debt (flow) matters importantly for reform, while both level and change matter for growth.

### Policy-relevant conclusions
- Easily available external financing can postpone necessary macroeconomic and structural reforms by reducing the short-term pain of the status quo.
- When international capital markets stop financing an economy (as in crises), reform often follows quickly because borrowing can no longer sustain the status quo.
- Empirical evidence supports the view that rising external debt is linked to slower reform and that both level and change of external debt matter for growth; therefore, policymakers should be aware that temporary external financing may delay reforms that are growth-enhancing in the long run.

### ANNEX I. Definition of Index of Economic Freedom — major components
- Size of Government:
  - General government consumption as share of total consumption
  - Transfers and subsidies as a share of GDP
  - Government enterprises and investment as a share of gross investment
  - Top marginal tax rate
- Top Marginal Income Tax Rate:
  - Top Marginal Income and Payroll Tax Rate
- Legal System & Property Rights:
  - Judiciary independence
  - Impartial courts
  - Protection of intellectual property
  - Military in Politics
  - Law and Order
- Sound Money:
  - Avg. growth of money (last 5 yrs) minus growth of real GDP (last 10 yrs)
  - Standard deviation of annual inflation (last 5 yrs)
  - Annual inflation (most recent yr)
  - Freedom of citizens to own foreign currency bank accounts (dom. and abroad)
- Freedom to Trade Internationally:
  - Tariffs
  - International trade tax revenues (in percent of trade sector)
  - Mean tariff rate
  - Standard deviation of tariff rates
  - Regulatory Trade Barriers
  - Hidden import barriers
  - Costs of importing
  - Actual vs. expected size of trade sector
  - Difference between official and black market exchange rates
- International Capital Market Controls:
  - Foreign ownership restrictions
  - Index of capital controls among 13 IMF categories
- Regulation (Credit Market, Labor Market, Business Regulations):
  - Credit market items: ownership of banks; competition in domestic banking; extension of credit; interest rate regulations (leading to neg. rates); interest rate controls
  - Labor market items: impact of minimum wage; hiring and firing practices; labor force share with wages set by centralized collective bargaining; unemployment insurance; use of conscripts
  - Business regulation items: price controls; burden of Regulations; time with government bureaucracy; starting a new business; irregular payments

*Source: _wp0750 - 2. External Debt and Economic Reform in Countries with Low Ranking of Economic Freedom (IMF, chapter text provided).*

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

### _wp0750 - References..............................................................................................................

### Tables
- 1. Economic Reform and External Debt, Panel Regressions, 1970–2000, 5-year Periods......18
- 2. Economic Reform and External Debt, Panel Regressions with Lags and Estimation with   
    Instrumental Variables, 1970–2000, 5-year Periods............................................................19
- 3. Economic Reform and External Debt, Panel Regressions with Random Country Effects,  
    1970–2000, 5-year Periods ..................................................................................................20
- 4. Economic Reform and External Debt, Probit and Ordered Probit Regressions 1970–2000,    
    5-year periods.......................................................................................................................21
- 5. External Debt and Economic Growth, Panel Regressions with Fixed Effects, 1970–2000,  
    5-year Periods ......................................................................................................................22

### Figures
- 1. External Debt and Economic Reform ..................................................................................16

*Source: _wp0750 - References..............................................................................................................*

### 2. External Debt and Economic Reform in Countries with Low Ranking of Economic

### 2. External Debt and Economic Reform in Countries with Low Ranking of Economic Freedom

### Motivation and main hypothesis
- Question: If economic reform is the necessary “treatment” for a sick economy, why is it often delayed?
- Central hypothesis: Easily available external financing (foreign borrowing) can delay necessary economic reform by reducing the immediate cost of the status quo and by enabling competing socioeconomic groups to “wait out” concession—external financing acts as a “pain reliever.”
- Theoretical mechanism emphasized:
  - Alesina and Drazen (1991) war-of-attrition framework extended to allow private-sector foreign borrowing.
  - Private and public foreign borrowing lower the short-term cost of delaying reform (reducing distortionary taxation or financing lobbying), increasing the time to concession T, up to the point where servicing debt becomes too costly.

### Theoretical model — key structure and conditions
- Baseline fiscal setup at time t = 0: balanced budget g + r b0 = τ with b0 entirely external.
- At 0 < t < T government increases spending gt > g0; share γ < 1 of spending paid by taxes, remainder (1 − γ) financed by external borrowing.
- Taxes at time t: τ(t) = γ [g(t) + r b(t)] (equation (1) in source).
- Change of public external debt: ḃ(t) = (1 − γ)[g(t) + r b(t)] (equation (2)).
- Private-sector borrowing li reduces distortionary tax disutility θ(li); li known only to group i and drawn from uniform distribution F(l) on [l, l¯].
- Stock of external debt of group i at time t: li e−rt (equation (6) structure).
- Key comparative condition (when winner is truly a winner): inequality (12) in source:
  - rtLW ebTlTl )1( )12()()( γ α − − −+< 
- First-order condition for concession time T yields expression (18); sign of dT/dli depends on whether the loser is credit-constrained.
- Condition for loser to be able to borrow enough so reform is delayed is inequality (19) in source (exact expression preserved in source).
- Main theoretical implication:
  - If foreign borrowing is relatively easily available to all groups (i.e., groups are not credit constrained and (19) holds), private-sector foreign borrowing further delays reform.
  - Conversely, when international capital markets stop financing public and private sectors (as in crises), reform follows quickly because borrowing cannot finance the status quo.

### Empirical strategy and data
- Reform measure: change in Fraser Institute index of economic freedom (index range 1 to 10; average of five sub-indexes: government size, legal system & property rights, sound monetary policy, freedom to trade, regulation strictness & efficiency).
- Sample: all developing and emerging economies with available data, period 1970–2002, total of 81 countries (Annex II).
- Time aggregation: five-year periods (three years for the last observation).
- Main dependent variable: R = change in economic freedom index over each five-year period.
- Key regressors and controls:
  - B: dummy grouping countries in six groups increasing in the change of their external debt-to-GDP ratio (detailed grouping rules preserved in source).
  - F: level of economic freedom index at beginning of period.
  - GR: crisis dummy if average real GDP growth rate below -1 percent in previous period.
  - INF: crisis dummy if average inflation rate above 40 percent in previous period.
  - AID: average aid flows to GDP.
  - OP: trade-to-GDP ratio.
  - Interaction term B*F to test whether impact of foreign borrowing depends on prior reform.

### Empirical findings — reform regressions and robustness
- Simple correlations (Figures 1 and 2): strong negative correlation between change in external debt-to-GDP ratio and change in economic freedom index and its five subindexes; similar pattern for countries with low ranking of economic freedom.
- Main panel regression results (Table 1 and text summary):
  - Estimate of external borrowing (B) is negative and statistically significant at the 1 percent level in all specifications.
  - Switching to the next borrower group is correlated with a reduction in the reform index by about 0.4 (text: "Switching to the next group of borrowers is correlated with a reduction in the reform index by about 0.4. This is double the average change in the index over a five-year period in the sample.").
  - Level of economic freedom F has a negative and significant coefficient (countries already liberalized have less room for measured reform).
  - INF (inflation crises) leads to reform (positive and significant); GR (growth crisis) positive but not statistically significant.
  - Interaction B*F positive and significant: the negative link between external borrowing and reform is stronger in countries with lower prior reform.
  - Adjusted R-squared values reported (Table 1): 0.15, 0.27, 0.30, 0.30, 0.24 (depending on specification).
- Lags and IV robustness (Table 2 and text):
  - Lagged external borrowing B(-1) has a negative and statistically significant coefficient (about −0.09 to −0.11 in reported columns).
  - Inclusion of B(-1) suggests countries that borrow a lot in one period reform less in the current and the next periods.
  - Instrumental variable: U.S. short-term interest rate used to instrument external borrowing (IV results in Table 2 indicate estimate of external borrowing remains negative and statistically significant).
  - Adj. R-squared values in Table 2: 0.27, 0.28, 0.21, 0.25, 0.10, 0.12 (by column).
- Subindex analysis (Table 3):
  - Rising external debt is linked to slower progress in most subindexes:
    - D(F1) government size: B = −0.09*** (t = −5.95)
    - D(F2) legal system & property rights: B = −0.07 (not significant)
    - D(F3) sound money: B = −0.36*** (t = −5.95) — largest effect
    - D(F4) freedom to trade: B = −0.11*** (t = −2.85)
    - D(F5) regulation: B = −0.07** (t = −2.43)
  - Adj. R-squared values in Table 3: 0.18, 0.21, 0.27, 0.19, 0.34 (for D(F1)–D(F5) respectively).
- Probit and ordered probit (Table 4):
  - Probit columns: B coefficients −0.35*** and −0.42*** (z-statistics −3.41, −3.64).
  - Ordered probit columns: B coefficients −1.68**, −0.24***, −0.28***, −0.48 (varied significance).
  - Lagged B(-1) shows negative significance in several probit specifications (e.g., −0.21*; −0.23**).
  - Adj. R-squared values reported: 0.15, 0.29, 0.28 (as in Table 4).
- Growth regression linking debt and growth (Table 5):
  - Panel regressions with fixed effects, dependent variable: average real GDP per capita growth over five-year non-overlapping sub-periods.
  - External debt/GDP coefficients:
    - Column examples: −0.012*** (t = −4.162), −0.026*** (t = −4.893), −0.016** (t = −2.450).
  - Change of external debt/GDP coefficients:
    - −0.021*** (t = −3.168) and −0.013** (t = −1.909) in reported columns.
  - Other growth determinants (examples from Table 5):
    - Real per capita GDP coefficient values: −4.233***, −3.353***, −6.439*** (t-statistics shown).
    - Secondary school enrollment coefficient 0.055** (t = 2.539).
    - Investment/GDP coefficient 0.311*** (t = 7.647).
    - Inflation rate coefficient −0.001*** (t = −3.490).
    - Growth in trading partners coefficient 0.559*** (t = 3.552).
    - Trade/GDP coefficient 0.031** (t = 2.403).
  - Adj. R-squared values in Table 5: 0.28, 0.37, 0.60 (by column).

### Interpretation and implications
- Consistent pattern: higher external borrowing (especially increases in external debt) is associated with delayed economic reform across a broad index and most subindexes (strongest for sound money; weaker for legal/institutional reform).
- Causality caveats:
  - Reverse causality possible (reforms attract capital inflows), but:
    - Interaction term B*F and lagged B reduce concerns by showing stronger negative effect where prior reform is low and lagged borrowing also matters.
    - IV estimation using U.S. short-term interest rate still yields negative significant effect of external borrowing on reform.
  - Available evidence suggests at least part of effect runs from external borrowing to delayed reform.
- Link to growth:
  - Both the level and the change of external debt-to-GDP negatively affect real per capita growth in panel regressions.
  - Proposed channel: easy external financing reduces pressure to reform, which in turn slows long-run growth; change in debt (flow) matters importantly for reform, while both level and change matter for growth.

### Policy-relevant conclusions (from source)
- Easily available external financing can postpone necessary macroeconomic and structural reforms by reducing the short-term pain of the status quo.
- When international capital markets stop financing an economy (as in crises), reform often follows quickly because borrowing can no longer sustain the status quo.
- Empirical evidence supports the view that rising external debt is linked to slower reform and that both level and change of external debt matter for growth; therefore, policymakers should be aware that temporary external financing may delay reforms that are growth-enhancing in the long run.

*Source: _wp0750 - 2. External Debt and Economic Reform in Countries with Low Ranking of Economic Freedom (IMF, chapter text provided).*

### ANNEX I.    DEFINITION OF INDEX OF ECONOMIC FREEDOM

### ANNEX I.    DEFINITION OF INDEX OF ECONOMIC FREEDOM

### Economic freedom index — Major components
- Size of Government
  - General government consumption as share of total consumption
  - Transfers and subsidies as a share of GDP
  - Government enterprises and investment as a share of gross investment
  - Top marginal tax rate
- Top Marginal Income Tax Rate
  - Top Marginal Income and Payroll Tax Rate
- Legal System & Property Rights
  - Judiciary independence
  - Impartial courts
  - Protection of intellectual property
  - Military in Politics
  - Law and Order
- Sound Money
  - Avg. growth of money (last 5 yrs) minus growth of real GDP (last 10 yrs)
  - Standard deviation of annual inflation (last 5 yrs)
  - Annual inflation (most recent yr)
  - Freedom of citizens to own foreign currency bank accounts (dom. and abroad)
- Freedom to Trade Internationally
  - Tariffs
  - International trade tax revenues (in percent of trade sector)
  - Mean tariff rate
  - Standard deviation of tariff rates
  - Regulatory Trade Barriers
  - Hidden import barriers
  - Costs of importing
  - Actual vs. expected size of trade sector
  - Difference between official and black market exchange rates
- International Capital Market Controls
  - Foreign ownership restrictions
  - Index of capital controls among 13 IMF categories
- Regulation
  - Credit Market Regulation
    - Ownership of banks
    - Competition in domestic banking
    - Extension of credit
    - Interest rate regulations (leading to neg. rates)
    - Interest rate controls
  - Labor Market Regulations
    - Impact of minimum wage
    - Hiring and firing practices
    - Labor force share with wages set by centralized collective bargaining
    - Unemployment insurance
    - Use of conscripts
  - Business Regulations
    - Price controls
    - Burden of Regulations
    - Time with government bureaucracy
    - Starting a new business
    - Irregular payments

### Countries included in the panel regression
- Algeria
- Argentina
- Bangladesh
- Barbados
- Belize
- Benin
- Bolivia
- Botswana
- Brazil
- Bulgaria
- Burundi
- Cameroon
- Central African Republic
- Chad
- Chile
- China
- Colombia
- Congo, Dem. Rep.
- Congo, Rep.
- Costa Rica
- Cote d'Ivoire
- Croatia
- Czech Republic
- Dominican Republic
- Ecuador
- Egypt, Arab Rep.
- El Salvador
- Fiji
- Gabon
- Ghana
- Guatemala
- Guinea-Bissau
- Guyana
- Haiti
- Honduras
- Hungary
- India
- Indonesia
- Iran, Islamic Rep.
- Jamaica
- Jordan
- Kenya
- Latvia
- Madagascar
- Malawi
- Malaysia
- Mali
- Mauritania
- Mexico
- Morocco
- Nepal
- Nicaragua
- Niger
- Nigeria
- Oman
- Pakistan
- Panama
- Papua New Guinea
- Paraguay
- Peru
- Philippines
- Poland
- Romania
- Rwanda
- Senegal
- Sierra Leone
- Slovak Republic
- South Africa
- Sri Lanka
- Syrian Arab Republic
- Tanzania
- Thailand
- Togo
- Trinidad and Tobago
- Tunisia
- Turkey
- Uganda
- Uruguay
- Venezuela, RB
- Zambia
- Zimbabwe

*Source: _wp0750 - ANNEX I.    DEFINITION OF INDEX OF ECONOMIC FREEDOM*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2007/_wp0750.pdf_