## _wp1140

## Source details

**Canonical URL:** [_wp1140](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1140.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1140.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1140.pdf.json)

---

### Sample construction and index methodology
- Appendix 3 indicates the year in which each of the SSA countries in the sample became stable/liberalized according to the authors' criteria.
- Index construction:
  - Stable/liberalized index = 100 in 1980 and grows each year at the average growth rate of income per capita of SSA countries that, in that year, have already become stable/liberalized.
  - Rest SSA index = 100 in 1980 and grows each year at the average growth rate of countries not stable/liberalized in that year.
- Exclusions and adjustments:
  - The first three years after stabilization/liberalization are excluded to avoid rebound effects.
  - Sample excludes oil-exporting countries and small island countries; Eritrea excluded because Sachs and Warner index not listed in Wacziarg and Welch (2003).
  - Oil exporters excluded: Angola, Cameroon, Chad, Equatorial Guinea, Gabon, Republic of Congo, and Nigeria.
  - Small countries (population below 1 million by 2009): Cape Verde, Comoros, Equatorial Guinea, Sao Tome and Principe, and Seychelles.
- Definition detail: A country is counted as liberalized one year after the country became open according to Sachs and Warner (example: Ghana open and politically stable by 1985, counted as stable/liberalized from 1986).

### Comparative performance: stable/liberalized vs Rest of SSA and other regions
- Long-run index behavior:
  - Income per capita of the Rest of SSA group remains flat for almost three decades.
  - Income per capita of stable/liberalized countries grows rapidly and monotonically.
- Average growth rates (1980–2008):
  - Average growth rate of income per capita of the stable/liberalized countries is 3.7 percent per year in 1980–2008.
- Cross-region comparisons:
  - Growth in stable/liberalized SSA countries comparable to high-performing regions (including East Asian economies).
  - Figure 4 comparison: stable/liberalized SSA superior to Latin America and slightly better than ASEAN-5 and South Asia.
- Hypothetical extrapolation:
  - If all SSA economies had grown at the average rate of stable/liberalized countries since 1960, their income per capita would now be 90 percent higher than that of ASEAN-5.
  - Implication: regional income per capita near US$5,000 would have fostered industrialization of natural resource–based exports and greater internal market-driven industrial development (regional market described as almost US$4 trillion).

### Robustness, sensitivity, and heterogeneity
- Robustness checks:
  - Performance robust to modifications in the sample, changes in the observation period, and alternative estimation methods.
  - Table 1 columns summarize alternative computations:
    - Column 1: average growth rate in index by country group.
    - Column 2: average growth excluding six outliers (three highest and three lowest post-stabilization/liberalization).
    - Column 3: average growth rates since 1995.
    - Column 4: averages including IMF growth projections up to 2014.
    - Column 5: group average computed as average of country-specific stable/liberalized period growth rates.
  - Conclusion: growth rates of ASEAN-5, South Asian, and stable/liberalized SSA countries are comparable.
- Outliers and country heterogeneity:
  - Best performers post-stabilization/liberalization: Rwanda, Mozambique, Ethiopia.
  - Worst performers: Benin, Kenya, Liberia.
  - Botswana: only case where post-stabilization/liberalization growth lower than pre-stabilization/liberalization (explained by extraordinarily favorable pre-stabilization terms of trade); Botswana’s post-stabilization/liberalization growth still superior to average growth in ASEAN-5.
  - Post-stabilization/liberalization country growth range: 1.49–7.33 percent (all countries experienced positive output per capita growth after stabilization/liberalization).

### Apparent payoff of stabilization/liberalization (descriptive and regression evidence)
- Distributional evidence:
  - Kernel distributions: post-stabilization/liberalization growth rates distribution shifted right relative to rest of sample.
  - Mean growth post-stabilization/liberalization is 2.9 percent higher than for rest of sample.
- Univariate panel regression specification:
  - γi,t = αi + π Si,t + εi,t, where S = 1 in post-stabilization/liberalization years and 0 otherwise; sample: non-oil exporting SSA countries, 1960–2008.
  - π interpreted as growth payoff from political stability and economic liberalization (authors caution on causal interpretation).
- Selected regression estimates (Table 2; coefficients with standard errors in parentheses):
  - Lagged GDP growth coefficients:
    - Column (1): 0.102 (0.031)*** 
    - Column (2): 0.173 (0.030)*** 
    - Column (3): 0.017 (0.081)
    - Alternate set: 0.093 (0.031)***, 0.161 (0.030)***, 0.093 (0.076)
  - Stable (S) coefficient estimates (growth payoff in percentage points):
    - Column (1): 2.771 (0.439)*** 
    - Column (2): 2.718 (0.347)*** 
    - Column (3): 2.370 (0.593)*** 
    - Alternate set: 2.909 (0.442)***, 2.792 (0.351)***, 2.701 (0.631)*** 
  - Terms of trade and lagged terms of trade:
    - Terms of trade growth: 0.009 (0.008), 0.010 (0.008), 0.012 (0.008)
    - Lagged terms of trade growth: 0.018 (0.008)**, 0.018 (0.008)**, 0.105 (0.011)
  - Constant terms:
    - -0.357 (0.207)*, -0.381 (0.190)**, -0.229 (0.350)
    - Alternate set: -0.412 (0.210)*, -0.413 (0.193)**, -0.344 (0.360)
  - Observations and country counts:
    - Observations: 992, 992, 956; alternate: 1002, 1002, 965
    - Number of countries: 36, 36, 36; alternate: 37, 37, 37
  - Significance notation: * significant at 10%; ** significant at 5%; *** significant at 1%.
- Regression summary:
  - Under all specifications reported, estimated payoff from stabilization/liberalization is significant, adding between 2.37 and 2.9 percentage points to growth of output per capita.
  - Authors refrain from strong causal claims; regressions are univariate panel as an indicative exercise.

### Before-and-after comparisons and statistical test
- Comparing average growth before versus after stabilization/liberalization:
  - Average growth after stabilization/liberalization was 2.9 percent higher than before stabilization/liberalization.
  - Student’s t-test: difference in means significant with more than 99 percent confidence; t-statistic = 4.7.
- Table 3 selected country pre/post excerpts (Pre, Post, tDifference):
  - Benin 0.19 1.49 1.30
  - Botswana 9.70 4.71 -4.99
  - Burkina Faso 1.00 2.83 1.83
  - Burundi 0.54 2.36 1.82
  - Ethiopia 0.62 4.88 4.26
  - Gambia, The 1.56 2.33 0.77
  - Ghana -1.36 2.32 3.68
  - Kenya 1.47 1.55 0.08
  - Liberia -1.69 1.91 3.60
  - Madagascar -1.23 2.28 3.51
  - Mali -0.05 2.18 2.23
  - Mauritius -2.70 4.04 6.75
  - Mozambique 0.62 5.37 4.75
  - Namibia 0.08 2.68 2.60
  - Niger -0.48 3.01 3.50
  - Rwanda 0.76 7.33 6.57
  - Sierra Leone -0.49 3.73 4.21
  - South Africa 0.57 2.25 1.68
  - Tanzania 0.72 4.48 3.76
  - Uganda -0.77 4.07 4.84
  - Zambia -0.19 3.25 3.44
  - Average 0.42 3.29 2.87

### Reverse causality exploration and probit analysis
- Objective: assess whether stabilization/liberalization is driven by good pre-existing economic performance or favorable growth determinants.
- Mean-comparison findings (Appendix 5):
  - For very few variables there is a statistically significant difference between means at onset of stabilization/liberalization and other years; no clear overall “superiority” in either group.
  - Onset group had: a lower current account deficit, lower government spending, but also more people living in landlocked areas.
  - Onset group not different in institutional quality, education, and life expectancy.
  - Manufacturing exports were less relevant in onset group.
  - GDP growth was lower in the years before stabilization/liberalization, rejecting the hypothesis that high GDP growth led to stabilization/liberalization.
- Table 4 selected mean comparisons (Unstable/Unliberalized Years vs Onset of Stabilization/Liberalization; means and 95% CIs preserved):
  - Current Account Balance (% of GDP): Unstable/Unliberalized Years -7.716 -8.367 -7.064; Onset -4.499 -6.130 -2.867
  - Central Government Balance (% of GDP): Unstable/Unliberalized Years -6.258 -6.702 -5.814; Onset -3.362 -4.067 -2.656
  - Manufacturing Exports (% of Merchandise Exports): Unstable/Unliberalized Years 19.106 16.863 21.349; Onset 8.933 5.736 12.130
  - Population 100KM from Coast in 2000 (%): Unstable/Unliberalized Years 42.420 39.049 45.791; Onset 29.515 21.705 37.325
  - CFAF Zone Dummy: Unstable/Unliberalized Years 0.305 0.270 0.341; Onset 0.182 0.088 0.276
- Probit specification (Table 5):
  - YS_{i,t} = β0 + βj Z_{i,t,j} + ε_{i,t}; YS = 1 in year of stabilization and in previous three years; 0 otherwise; years after stabilization/liberalization excluded.
- Selected probit coefficient estimates (standard errors in parentheses; significance markers preserved):
  - Current account balance as % of GDP: -0.080 (0.138); -0.256 (0.143)*; -0.267 (0.187); -0.080 (0.199)
  - Government balance as % of GDP: -0.116 (0.169); -0.353 (0.196)*; -0.222 (0.194); -0.094 (0.207)
  - Manufacturing exports as % of merchandise exports: -0.131 (0.042)***; -0.100 (0.075); -0.085 (0.082); -0.097 (0.096)
  - Population 100KM from coast in 2000: 0.010 (0.028); -0.047 (0.049); -0.097 (0.047)**; -0.061 (0.049)
  - Year: 1.110 (0.001)***; 0.849 (0.167)***; 0.970 (0.418)**; 1.409 (0.423)***
  - Constant: -2,217.760 (0.000); -1,694.573 (333.369)***; -1,937.479 (828.816)**; -2,811.646 (839.942)***
  - Observations: 224, 120, 105, 77
  - Number of countries: 26, 17, 14, 9
- Probit interpretation:
  - No robust evidence that the tested independent variables are consistently related to the likelihood of stabilization/liberalization across specifications.

### Conclusions and policy implications (VIII)
- Core conclusions:
  1. SSA countries under peace and after economic liberalization have attained impressively high economic growth; economic liberalization accompanied high growth, unlike in countries remaining in conflict.
  2. SSA countries can grow sustainably without changing their comparative advantage in natural resources; adding value to natural resources recommended.
  3. Beyond peace and market reforms, post-stabilization/liberalization governments boosted investments in health, education, and infrastructure and undertook institutional reforms.
  4. Probit analysis shows SSA countries did not wait until institutions or other growth determinants improved before attaining political stability and launching economic liberalization.
- Interpretive points:
  - Conclusions emphasize observation that stable/liberalized SSA countries have been high growth performers rather than claiming definitive causality from stabilization/liberalization.
  - Achieving high economic growth may not require complex preconditions; political stabilization and market reforms have been crucial contributors.
  - Role of local policymakers, civil society, entrepreneurs, and international organizations in implementing pro-growth policies beyond market reforms is emphasized.

### Data, appendices, and methodological notes
- Data sources and variables (labels preserved as in source): IMF WEO, WDI, PRS database, Freedom House, World Resource Institute, CID (Harvard); variables include Real GDP, PPP conversion factor, Barro and Lee Human Capital, Real exchange rate misalignment (Elbadawi and Soto (2005)), Political rights (Freedom House), Bureaucracy quality (PRS), Manufacturing exports (WDI), Current account (WEO), Government balance (WEO), Terms of trade (WEO), External debt (WDI), Corruption index (PRS), Population 100KM from coast (World Resource Institute), among others.
- Appendix contents:
  - Appendix 2: Data Sources.
  - Appendix 3: Year of Stabilization/Liberalization for countries in sample.
  - Appendix 4: Year-by-year GDP per capita growth rates post stabilization/liberalization and "Average of Stable/Liberalized" series.
  - Appendix 5: Comparison of Means — Unstable/Unliberalized Years vs. Years at the Onset of Stabilization (with 95% Confidence Intervals).

*Source: IMF, World Economic Outlook, October 2009; Authors' construction, 2010.*

### Appendix 3 indicates the year in which each of the SSA countries in our sample became

### _wp1140 - Appendix 3 indicates the year in which each of the SSA countries in our sample became

### Sample construction and index methodology
- Appendix 3 indicates the year in which each of the SSA countries in the sample became stable/liberalized according to the authors' criteria.
- The table in Appendix 4 shows the time-varying composition of the stable/liberalized group and their GDP per capita growth rates in the years in which they had already become stable/liberalized.
- Early 1980s membership: only Mauritius and Botswana; later additions included Ghana, Mali, Uganda, and more SSA countries.
- Index construction:
  - The stable/liberalized index equals 100 in 1980 and grows each year at the average growth rate of income per capita of SSA countries that, in that year, have already become stable/liberalized.
  - The Rest SSA index is 100 in 1980 and grows each year at the average growth rate of countries not stable/liberalized in that year.
- Exclusions and adjustments:
  - The first three years after stabilization/liberalization are excluded from calculations to avoid rebound effects.
  - Sample excludes oil-exporting countries and small island countries; Eritrea is excluded because the Sachs and Warner index was not listed in Wacziarg and Welch (2003).
  - Oil exporters excluded: Angola, Cameroon, Chad, Equatorial Guinea, Gabon, Republic of Congo, and Nigeria.
  - Small countries (population below 1 million by 2009): Cape Verde, Comoros, Equatorial Guinea, Sao Tome and Principe, and Seychelles.
- Definition detail: A country is considered liberalized one year after the country became open according to Sachs and Warner (example: Ghana open and politically stable by 1985, counted as stable/liberalized from 1986).

### Comparative performance: stable/liberalized vs Rest of SSA and other regions
- Long-run index behavior:
  - The income per capita of the Rest of SSA group remains flat for almost three decades.
  - Income per capita of stable/liberalized countries grows rapidly and monotonically.
- Average growth rates (1980–2008):
  - The average growth rate of income per capita of the stable/liberalized countries is 3.7 percent per year in 1980–2008.
- Cross-region comparisons:
  - Growth in stable/liberalized SSA countries was as high as high-performing regions, including East Asian economies.
  - Figure 4 comparison: stable/liberalized SSA vs Latin America, South Asia, and ASEAN-5 — evolution in stable/liberalized SSA is superior to Latin America and slightly better than ASEAN-5 and South Asia.
- Hypothetical extrapolation:
  - If all SSA economies had grown at the average rate of stable/liberalized countries since 1960, their income per capita would now be 90 percent higher than that of ASEAN-5.
  - Implication: regional income per capita near US$5,000 would have fostered industrialization of natural resource–based exports and greater internal market-driven industrial development (regional market described as almost US$4 trillion).

### Robustness, sensitivity, and heterogeneity
- Robustness checks:
  - Performance is robust to modifications in the sample, changes in the observation period, and alternative estimation methods.
  - Table 1 columns:
    - Column 1: average growth rate in index by country group.
    - Column 2: average growth of stable/liberalized SSA excluding six outliers (three highest and three lowest growth rates after becoming stable/liberalized).
    - Column 3: average growth rates since 1995 (excluding early years with few stabilizers/liberalizers).
    - Column 4: averages including IMF growth projections up to 2014.
    - Column 5: group average computed as average of country-specific stable/liberalized period growth rates (within then between calculation).
  - Conclusion: growth rates of ASEAN-5, South Asian, and stable/liberalized SSA countries are comparable.
- Outliers and country heterogeneity:
  - Best performers post-stabilization/liberalization: Rwanda, Mozambique, Ethiopia.
  - Worst performers: Benin, Kenya, Liberia.
  - Of 22 countries that stabilized/liberalized in the period under analysis, Botswana is the only case where post-stabilization/liberalization growth was lower than pre-stabilization/liberalization; explained by extraordinarily favorable pre-stabilization terms of trade. Botswana’s post-stabilization/liberalization growth still superior to average growth in ASEAN-5.
  - Post-stabilization/liberalization country growth range: 1.49–7.33 percent (all countries experienced positive output per capita growth after stabilization/liberalization).

### Apparent payoff of stabilization/liberalization (descriptive and regression evidence)
- Distributional evidence:
  - Kernel distributions separate post-stabilization/liberalization growth rates from the rest of the sample; post-stabilization/liberalization distribution is shifted right.
  - Mean growth post-stabilization/liberalization is 2.9 percent higher than for rest of sample.
- Univariate panel regression specification:
  - γi,t = αi + π Si,t + εi,t, where S is binary (1 in post-stabilization/liberalization years; 0 otherwise); sample: non-oil exporting SSA countries, 1960–2008.
  - π interpreted as growth payoff from political stability and economic liberalization (with caveats on causality).
- Table 2 regression results (selected exact coefficients and statistics preserved):
  - Lagged GDP growth coefficients:
    - Column (1): 0.102 (0.031)*** 
    - Column (2): 0.173 (0.030)*** 
    - Column (3): 0.017 (0.081)
    - Alternate set: 0.093 (0.031)***, 0.161 (0.030)***, 0.093 (0.076)
  - Stable (S) coefficient estimates (growth payoff in percentage points):
    - Column (1): 2.771 (0.439)*** 
    - Column (2): 2.718 (0.347)*** 
    - Column (3): 2.370 (0.593)*** 
    - Alternate set: 2.909 (0.442)***, 2.792 (0.351)***, 2.701 (0.631)*** 
  - Terms of trade and lagged terms of trade:
    - Terms of trade growth coefficients: 0.009 (0.008), 0.010 (0.008), 0.012 (0.008)
    - Lagged terms of trade growth: 0.018 (0.008)**, 0.018 (0.008)**, 0.105 (0.011)
  - Constant terms:
    - -0.357 (0.207)*, -0.381 (0.190)**, -0.229 (0.350)
    - Alternate set: -0.412 (0.210)*, -0.413 (0.193)**, -0.344 (0.360)
  - Observations and country counts:
    - Observations: 992, 992, 956; alternate: 1002, 1002, 965
    - Number of countries: 36, 36, 36; alternate: 37, 37, 37
  - Significance notation: * significant at 10%; ** significant at 5%; *** significant at 1%.
- Summary from regressions:
  - Under all specifications reported, the estimated payoff from stabilization/liberalization is significant, adding between 2.37 and 2.9 percentage points to growth of output per capita.
  - Authors refrain from strong causal claims; regressions are univariate panel as an indicative exercise.

### Before-and-after comparisons and statistical test
- Comparing average growth before versus after stabilization/liberalization:
  - Average growth after stabilization/liberalization was 2.9 percent higher than before stabilization/liberalization.
  - Student’s t-test: difference in means is significant with more than 99 percent confidence; t-statistic = 4.7.

### Reverse causality exploration
- Objective: assess whether stabilization/liberalization itself is a result of good pre-existing economic performance or favorable growth determinants.
- Approach:
  - Examine a list of growth variables commonly cited in the growth literature.
  - Test whether means of these variables in the three years before stabilization/liberalization differ from means in other years without stabilization/liberalization.
  - This identifies whether the three years before stabilization/liberalization are years of relatively better performance along those dimensions.

*Source: IMF, World Economic Outlook, October 2009.*

### Appendix 5 shows that, for very few variables, there is a statistically significant difference

### _wp1140 - Appendix 5 shows that, for very few variables, there is a statistically significant difference

### Appendix 5: comparison of means at onset of stabilization/liberalization vs. other years
- Appendix 5 finds that, for very few variables, there is a statistically significant difference between the means in the years at the onset of stabilization/liberalization and the means in other years without stabilization/liberalization; there is no clear overall “superiority” in either group.
- The group at the onset of stabilization/liberalization had:
  - a lower current account deficit,
  - lower government spending,
  - but also more people living in landlocked areas.
- The onset group was not different in terms of most variables usually quoted as “preconditions to stabilization/liberalization,” including institutional quality, education, and life expectancy.
- Manufacturing exports were less relevant in countries at the onset of stabilization/liberalization.
- GDP growth was lower in the years before stabilization/liberalization, rejecting the possibility that high GDP growth led to stabilization/liberalization (i.e., reverse causality).

### Table 3: GDP Per Capita Growth before and after Stabilization/Liberalization (selected country excerpts and average)
- Selected country entries (Pre, Post, tDifference):
  - Benin 0.19 1.49 1.30
  - Botswana 9.70 4.71 -4.99
  - Burkina Faso 1.00 2.83 1.83
  - Burundi 0.54 2.36 1.82
  - Ethiopia 0.62 4.88 4.26
  - Gambia, The 1.56 2.33 0.77
  - Ghana -1.36 2.32 3.68
  - Kenya 1.47 1.55 0.08
  - Liberia -1.69 1.91 3.60
  - Madagascar -1.23 2.28 3.51
  - Malawi --- (omitted)
  - Mali -0.05 2.18 2.23
  - Mauritius -2.70 4.04 6.75
  - Mozambique 0.62 5.37 4.75
  - Namibia 0.08 2.68 2.60
  - Niger -0.48 3.01 3.50
  - Rwanda 0.76 7.33 6.57
  - Sierra Leone -0.49 3.73 4.21
  - South Africa 0.57 2.25 1.68
  - Tanzania 0.72 4.48 3.76
  - Uganda -0.77 4.07 4.84
  - Zambia -0.19 3.25 3.44
  - Average 0.42 3.29 2.87
- Source: IMF, World Economic Outlook, October 2009.
- Note: grouping divides observations into country/years with stabilization/liberalization and the rest of the sample; the rest includes observations for countries that never stabilized/liberalized (this grouping differs from Table 3 in which countries that never stabilized/liberalized are excluded).

### Table 4: Comparison of Means — Unstable/Unliberalized Years vs. Years at the Onset of Stabilization (selected means and 95% confidence intervals noted in source)
- Current Account Balance (% of GDP)
  - Unstable/Unliberalized Years -7.716 -8.367 -7.064
  - Onset of Stabilization/Liberalization -4.499 -6.130 -2.867
- Central Government Balance (% of GDP)
  - Unstable/Unliberalized Years -6.258 -6.702 -5.814
  - Onset of Stabilization/Liberalization -3.362 -4.067 -2.656
- Manufacturing Exports (% of Merchandise Exports)
  - Unstable/Unliberalized Years 19.106 16.863 21.349
  - Onset of Stabilization/Liberalization 8.933 5.736 12.130
- CFAF Zone Dummy
  - Unstable/Unliberalized Years 0.305 0.270 0.341
  - Onset of Stabilization/Liberalization 0.182 0.088 0.276
- Population 100KM from Coast in 2000 (%)
  - Unstable/Unliberalized Years 42.420 39.049 45.791
  - Onset of Stabilization/Liberalization 29.515 21.705 37.325
- Source: IMF, World Economic Outlook (2009), World Bank World Development Indicators.
- *See Appendix 2 for definition of variables.

### Probit analysis: specification and results (Table 5 summary)
- Empirical model:
  - YS_{i,t} = β0 + βj Z_{i,t,j} + ε_{i,t}
  - YS is 1 in the year of stabilization and in the previous three years and 0 in other years; observations in years after stabilization/liberalization are excluded.
- Key coefficient estimates and standard errors (selected; standard errors in parentheses; significance markers preserved):
  - Current account balance as % of GDP: -0.080 (0.138); -0.256 (0.143)*; -0.267 (0.187); -0.080 (0.199)
  - Government balance as % of GDP: -0.116 (0.169); -0.353 (0.196)*; -0.222 (0.194); -0.094 (0.207)
  - CFAF zone dummy: -3.847 (2.573); -2.459 (2.330); -0.645 (3.037); 1.798 (3.703)
  - Manufacturing exports as % of merchandise exports: -0.131 (0.042)***; -0.100 (0.075); -0.085 (0.082); -0.097 (0.096)
  - Population 100KM from coast in 2000: 0.010 (0.028); -0.047 (0.049); -0.097 (0.047)**; -0.061 (0.049)
  - Corruption index: -0.607 (0.753); -0.539 (0.784); -0.807 (1.093)
  - Debt outstanding % of GDP: 0.016 (0.015); 0.025 (0.022)
  - Real exchange rate misalignment: 0.027 (0.040)
  - Year: 1.110 (0.001)***; 0.849 (0.167)***; 0.970 (0.418)**; 1.409 (0.423)***
  - Constant: -2,217.760 (0.000); -1,694.573 (333.369)***; -1,937.479 (828.816)**; -2,811.646 (839.942)***
- Observations and sample:
  - Observations 224 120 105 77
  - Number of countries 26 17 14 9
- Statistical significance notation:
  - * significant at 10%; ** significant at 5%; *** significant at 1%
- Main interpretation:
  - There is no robust evidence that any of the independent variables is related to the likelihood of stabilization/liberalization across specifications.

### Conclusions and policy implications (VIII)
- Summary conclusions stated in the paper:
  1. SSA countries under peace and after economic liberalization have attained impressively high economic growth. Economic liberalization has been accompanied by high economic growth; the same may not have been true in countries that remained entangled in conflict.
  2. SSA countries can grow sustainably without changing their comparative advantage in natural resources. A natural resource–based model can be consistent with sustained economic growth; countries should seek to add value to their natural resources.
  3. Not only peace and market reforms are needed to achieve high economic growth. Post-stabilization/liberalization, SSA governments boosted and improved investments in health, education, and infrastructure and undertook institutional reforms beyond market liberalization; these actions are not inconsistent with the Washington Consensus paradigm shift.
  4. As shown in the probit analysis, SSA countries did not wait until they could reform institutions or improve other growth determinants before attaining political stability and launching economic liberalization.
- Additional interpretive points:
  - The conclusions do not hinge on proving that stabilization/liberalization generated the growth payoff, but on observing that stable/liberalized SSA countries have been high growth performers.
  - The experience suggests that achieving high economic growth may not require complex, highly specific preconditions; political stabilization and market reforms have been crucial contributors to accelerated economic growth in SSA.
  - The role of local policymakers, civil society, entrepreneurs, and international organizations in implementing pro-growth policies beyond market reforms is emphasized.

*Source: IMF, World Economic Outlook and related tables and text as contained in the provided PDF content unit.*

### 1. Fiscal Discipline. This was in the context of a region where almost all countries had run

### _wp1140 - 1. Fiscal Discipline

### Major policy prescriptions (Washington Consensus ten points)
- 1. Fiscal Discipline. "This was in the context of a region where almost all countries had run large deficits that led to balance of payments crises and high inflation that hit mainly the poor because the rich could park their money abroad."
- 2. Reordering Public Expenditure Priorities. "This suggested switching expenditure in a pro-growth and pro-poor way, from things like non-merit subsidies to basic health and education and infrastructure." The write-up emphasized neutrality about the desirable size of the public sector.
- 3. Tax Reform. "The aim was a tax system that would combine a broad tax base with moderate marginal tax rates."
- 4. Liberalizing Interest Rates. The author notes in retrospect this should have been framed as broader financial liberalization, with recognition of differing views on pace and importance of accompanying prudential supervision.
- 5. A Competitive Exchange Rate. The author acknowledges possible wishful thinking that there was consensus for a competitive exchange rate and notes a tension with the emerging "two-corner doctrine" (fix firmly or float "cleanly").
- 6. Trade Liberalization. The paper acknowledges debate on speed, but consensus that liberalization was the appropriate direction.
- 7. Liberalization of Inward Foreign Direct Investment. The author explicitly did not include comprehensive capital account liberalization.
- 8. Privatization. Noted as the one area where neoliberal ideas had won broad acceptance; cautioned that privatization "can be a highly corrupt process" but "brings benefits (especially in terms of improved service coverage) when done properly, and the privatized enterprise either sells into a competitive market or is properly regulated."
- 9. Deregulation. Focused on easing barriers to entry and exit, not abolishing regulations for safety, environment, or non-competitive industries.
- 10. Property Rights. Primarily about enabling the informal sector to gain property rights at acceptable cost (inspired by Hernando de Soto’s analysis).

### Data description and sources (variables used by the authors)
- Variables assembled (source abbreviations preserved exactly as in the source):
  - Real GDP (in billions of national currency) — WEO
  - PPP conversion factor for GDP (national currency per international U.S. dollar) — WEO
  - Barro and Lee Human Capital: Educational attainment of people aged 15 years and above reported every 5 years — CID website, Harvard University
  - GDP at current prices (in billions of national currency) — WEO
  - Real GDP growth — WEO
  - Real exchange rate misalignemnt — Elbadawi and Soto (2005)
  - Population 100KM from coast in 2000 (as % of total population) — World Resource Institute
  - Political rights score — Freedom House database
  - Civil liberties — Freedom House database
  - Bureaucracy quality index — PRS database
  - Investment profile index — PRS database
  - Manufacturing exports (as % of merchandise exports) — WDI
  - Gross capital formation (as % of GDP) — WEO
  - Current account (as % of GDP) — WEO
  - Government balance (as % of GDP) — WEO
  - Imports of goods (as % of GDP in real terms) — WEO
  - Political risk index — PRS database
  - Value added of mining and quarrying (as % of GDP) — WDI
  - Exports of goods (as % of GDP in real terms) — WEO
  - Life expectancy — WDI
  - Terms of trade — WEO
  - External debt (in millions of U.S. dollars) — WDI
  - Industry value added (as % of GDP) — WDI
  - Corruption index — PRS database
  - Exchange rate (national currency per U.S. dollar) — WEO
- Data construction note: "Source: Authors' construction, 2010."
- Appendix 2 declared as "Data Sources" and lists IMF WEO, WDI, PRS database, Freedom House, World Resource Institute, CID (Harvard) among inputs.

### Empirical comparisons — means for selected variables
(Comparison of "Unstable/Unliberalized Years" vs. "Onset of Stabilization/Liberalization" — mean values preserved exactly as in the source)
- GDP Growth (%)
  - Unstable/Unliberalized Years: -0.134-0.4420.175
  - Onset of Stabilization/Liberalization: -0.399-1.5710.773
- Terms of Trade Annual Growth (%)
  - Unstable/Unliberalized Years: 0.666-0.3511.683
  - Onset of Stabilization/Liberalization: -2.219-3.551-0.887
- Educational Attainment (Years)
  - Unstable/Unliberalized Years: 2.2582.1262.391
  - Onset of Stabilization/Liberalization: 2.6652.2263.104
- Mining Value Added (% of GDP)
  - Unstable/Unliberalized Years: 0.0390.0320.045
  - Onset of Stabilization/Liberalization: 0.0480.0320.064
- Gross Capital Formation (% of GDP)
  - Unstable/Unliberalized Years: 18.36517.51619.370
  - Onset of Stabilization/Liberalization: 17.51614.62320.408
- Current Account Balance (% of GDP)
  - Unstable/Unliberalized Years: -7.716-8.367-7.064
  - Onset of Stabilization/Liberalization: -4.499-6.130-2.867
- Central Government Balance (% of GDP)
  - Unstable/Unliberalized Years: -6.258-6.702-5.814
  - Onset of Stabilization/Liberalization: -3.362-4.067-2.656
- Exports of Goods (% of GDP)
  - Unstable/Unliberalized Years: 23.92822.46125.396
  - Onset of Stabilization/Liberalization: 22.97418.79327.156
- Life Expectancy (years)
  - Unstable/Unliberalized Years: 51.85251.17652.529
  - Onset of Stabilization/Liberalization: 50.71749.09052.345
- Bureaucracy Quality
  - Unstable/Unliberalized Years: 1.4301.3071.553
  - Onset of Stabilization/Liberalization: 1.3130.9641.663
- Investment Profile
  - Unstable/Unliberalized Years: 5.4895.2975.681
  - Onset of Stabilization/Liberalization: 4.8224.3355.309
- Law and Order
  - Unstable/Unliberalized Years: 2.6082.4872.730
  - Onset of Stabilization/Liberalization: 2.5672.3032.832
- Corruption Index
  - Unstable/Unliberalized Years: 2.7932.6662.921
  - Onset of Stabilization/Liberalization: 3.0662.7583.374
- Debt Outstanding (% of GDP)
  - Unstable/Unliberalized Years: 106.74796.987116.508
  - Onset of Stabilization/Liberalization: 138.062108.369167.756
- Industry Value Added (% of GDP)
  - Unstable/Unliberalized Years: 21.00420.32721.682
  - Onset of Stabilization/Liberalization: 23.05820.15725.960
- Manufacturing Exports (% of Merchandise Exports) [label preserved exactly]
  - Unstable/Unliberalized Years: 19.10616.86321.349
  - Onset of Stabilization/Liberalization: 8.9335.73612.130
- Real Exchange Rate Misalignment (>100 is overvaluation)
  - Unstable/Unliberalized Years: 147.797135.677159.918
  - Onset of Stabilization/Liberalization: 198.82469.419328.230
- CFAF Zone Dummy
  - Unstable/Unliberalized Years: 0.3050.2700.341
  - Onset of Stabilization/Liberalization: 0.1820.0880.276
- Population 100KM from Coast in 2000 (%)
  - Unstable/Unliberalized Years: 42.42039.04945.791
  - Onset of Stabilization/Liberalization: 29.51521.70537.325

### Appendix highlights and methodological notes
- Appendix 2 lists data sources used to construct the dataset (IMF WEO, WDI, PRS, Freedom House, World Resource Institute, CID/Harvard).
- Appendix 3 provides "Year of Stabilization/Liberalization" for a set of countries (years listed in the source table).
- Appendix 4 contains year-by-year GDP per capita growth rates post stabilization/liberalization for individual countries and an "Average of Stable/Liberalized" series.
- Appendix 5 presents "Comparison of Means - Unstable/Unliberalized Years vs. Years at the Onset of Stabilization*" with 95% Confidence Interval (labels preserved as in the source).

*Source: Authors' construction, 2010.

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp1140.pdf_