## _wp09274 - References .............................................................................................................

## Source details

**Canonical URL:** [_wp09274 - References .............................................................................................................](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09274.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09274.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09274.pdf.json)

---

### I. Introduction — context, objectives, and focus
- Context and motivation:
  - "In 2009 about three-fourths of African countries are expected to increase their fiscal deficits excluding grants, or to decrease their projected surplus (IMF, 2009a)."
  - Many SSA countries allow automatic stabilizers to operate; some pursue countercyclical policies.
- Paper objectives:
  - (i) Document magnitude and evolution of cyclical patterns in government expenditures in SSA since 1970, correcting for endogeneity.
  - (ii) Identify factors underlying changes in cyclicality over time, emphasizing macroeconomic and institutional conditions.
- Focus and measurement:
  - Government spending chosen as fiscal policy instrument variable (Kaminsky, Reinhart, and Vegh, 2004).
  - Tax rates not used due to data limitations.
  - Outcome measures (overall fiscal balance, tax revenues) considered less appropriate because partly determined by policymakers and affected by output fluctuations.
- Stylized background:
  - Since the late 1990s many SSA countries reached a "post-stabilization phase" with sustainable fiscal and external positions, single-digit inflation, deeper domestic financial markets, and better institutions.

### II. Literature review — stylized facts and hypotheses
- Empirical regularities:
  - Developing countries generally observed to be procyclical; advanced economies acyclical or countercyclical (Braun, 2001; Lane, 2003; Kaminsky, Reinhart, and Végh, 2004; Alesina et al., 2008; Ilzetzki and Végh, 2008).
  - Some studies question the extent of procyclicality in developing countries (Rigobón, 2003; Jaimovich and Panizza, 2007).
  - Evidence on temporal evolution is mixed (Kaminsky et al. vs. Alesina et al.).
- SSA-specific evidence:
  - Thornton (2008): government consumption in low-income African countries (1960–2004) highly procyclical.
  - Diallo (2009): panel regressions (1989–2002) find fiscal policy on average procyclical in low- and middle-income African countries.
- Theoretical explanations:
  - Political/institutional factors: weak controls, corruption, volatile tax bases, competition for windfalls.
  - Financing constraints: limited international access and shallow domestic financial markets produce procyclicality.
  - Aid flows: mixed evidence on whether aid increases or decreases procyclicality.

### III. Empirical strategy and identification
- Core model (notations preserved as in source):
  - tiittiti eZYF ti,,,        (1)
  - ti P ti,10,           (2)
  - tititi vFY ,,,           (3)
  - tiiti e , ,            (4)
- Interpretation and cyclicality metric:
  - F and Y measure growth in the fiscal variable and output; β determines cyclicality: β<0 countercyclical; β=0 acyclical; β>0 procyclical.
- Endogeneity concerns and instruments:
  - Reverse causality between output growth and fiscal policy; correlation with unobserved country-specific effects.
  - Estimators: Difference-GMM (Diff-GMM) and System-GMM (Sys-GMM); internal instruments: past values; external instrument: export-weighted average of trading-partner GDP growth.
  - Diff-GMM preferred as it imposes fewer restrictions on instrument-error correlation.
- Two-stage empirical strategy:
  - Stage 1: assume no evolution (1β = 0) to benchmark average procyclicality.
  - Stage 2: allow 1β ≠ 0 and interact GDP growth with factors P to analyze determinants.

### Data, variables, and measurement
- Data:
  - Annual unbalanced panel, 39 years (1970–2008), 174 countries: 44 SSA, 33 advanced economies, 97 other developing countries.
- Key variables:
  - Dependent variable: growth in real central government spending.
  - Key explanatory variable: growth in real GDP.
  - Detrending: less parametric approach; Hodrik-Prescott not used.
  - No separation of discretionary spending vs. automatic stabilizers.
- Institutional and exposure proxies:
  - Political institutions: Polity 4 dataset (degree of democracy, constraints on the executive, political competition); robustness: ICRG (from 1985) and World Bank Database of Political Institutions (from 1975).
  - Commodity exposure proxy: share of commodity exports in GDP.
- Financing restriction proxies:
  - Domestic: private credit to GDP; real central bank interest rate.
  - External: net capital flows to GDP; official development aid to GDP.
  - Sustainability: debt-to-GDP ratio; inflation rate.
  - IMF program dummy: equals 1 if country has an IMF program in the current year (and transformed versions used).
- Controls and robustness checks:
  - Lagged central government spending growth (instrumented); growth in terms of trade; growth in oil prices; commodity prices; trade openness; democracy measure; dependency ratio; urbanization.

### Methodology and specification tests
- Estimation methods reported:
  - OLS with fixed effects.
  - 2SLS (instrumenting GDP growth with export-weighted trading-partner growth).
  - Two-step Diff-GMM (preferred); instruments: lags of real GDP growth, trading-partner growth, and lags of lagged dependent variable. Windmeijer (2005) correction applied for two-step GMM standard errors.
  - Sys-GMM for robustness.
- Diagnostic tests reported:
  - Arellano-Bond tests for AR(1) and AR(2).
  - Hansen and Sargan test p-values; attention to instrument proliferation and serial correlation (Roodman, 2006).

### Key empirical findings (1970–2008) — magnitudes and statistical evidence
- Overall conclusion:
  - Fiscal policy is procyclical in sub-Saharan Africa over 1970–2008.
- Cyclicality coefficient (β) by estimator (Table 2):
  - Sub-Saharan Africa:
    - OLS: 0.92*** (T-stat 6.3)
    - 2SLS: 2.21 (T-stat 1.23)
    - Diff-GMM: 1.54*** (T-stat 4.04)
  - Other Developing Countries:
    - OLS: 0.68*** (T-stat 6.15)
    - 2SLS: 1.25* (T-stat 1.87)
    - Diff-GMM: 1.16** (T-stat 2.41)
  - Advanced Economies:
    - OLS: -0.18 (T-stat 0.29)
    - 2SLS: -1.85 (T-stat 0.78)
    - Diff-GMM: -0.36 (T-stat 0.64)
- Magnitude interpretation (author statement):
  - "We find that a 2 percentage point increase in the rate of real GDP growth raises the growth rate of real government spending by about 3 points in SSA countries and 2.4 points in other developing countries; it does not affect the growth of real government expenditure in advanced economies."
- Preferred Diff-GMM results:
  - SSA: cyclicality coefficients generally stay in the 1.5–2 range across robustness checks.
  - Other developing countries: coefficients generally in the 1.1–1.2 range.
- Diagnostics (selected p-values from Table 3):
  - Arellano-Bond AR(1) examples: 0.001, 0.000, 0.001.
  - Arellano-Bond AR(2) examples: 0.232, 0.917, 0.207.
  - Hansen test p-values examples: 0.5150, 0.7980, 0.6270.

### Robustness and sensitivity
- Controls for common shocks (oil and commodity price growth) and long-term determinants yield similar cyclicality coefficients (SSA: 1.5–2; other developing: 1.1–1.2).
- Sys-GMM estimates are very close to Diff-GMM preferred estimates.
- External instrument validity concern (export-weighted trading-partner growth) addressed: export taxes have near-disappeared; terms of trade included; results using only internal instruments reported as very similar.

### Evolution by decade (Two-step Diff-GMM estimates, Table 4)
- Sub-Saharan Africa (β by decade):
  - 1970-1979: -1.53
  - 1980-1989: 1.83***
  - 1990-1999: 2.1***
  - 2000-2008: 1.48**
- Other Developing Countries (β by decade):
  - 1970-1979: 1.9
  - 1980-1989: -0.03
  - 1990-1999: 1.17*
  - 2000-2008: 1.73***
- Advanced Economies (β by decade):
  - 1970-1979: -0.08
  - 1980-1989: 0.09
  - 1990-1999: -0.07
  - 2000-2008: 0.04
- Interpretation:
  - Advanced economies acyclical throughout.
  - Other developing countries acyclical in 1970s–1980s; more procyclical since 1990s.
  - SSA acyclical in 1970s (cannot reject acyclicity); procyclical in 1980s and 1990s; coefficient falls in 2000–08 (1.48**) but decline not statistically significant given large standard deviations.

### Determinants — political economy and institutions (Table 5)
- Main finding:
  - No evidence that political institutions (degree of democracy, constraints on the executive, political competition) affect cyclical fiscal behavior in SSA (Table 5, columns 1–3).
- Robustness:
  - Results unchanged when restricting to post-1990, using alternative institutional measures, or interacting political variables with GDP growth during "good times".
- Commodity exports exposure:
  - Interaction of commodity exports-to-GDP ratio with GDP growth: coefficient 7.8 (T-stat 1.62); authors caution data paucity and sensitivity to coding missing observations.

### Determinants — financing restrictions (Table 6 and Appendix D)
- GDP growth unconditional coefficients (examples across specifications): 2.39*** (T-stat 2.76); 2.35*** (4.43); 3.1*** (3.47); 3.33*** (4.13); 3.09*** (4.0).
- Domestic financing proxies (interacted with GDP growth):
  - Private credit to GDP ratio: 0.01 (T-stat 0.34) — effectively zero and imprecisely estimated.
  - Lagged real central bank interest rate: -0.00 (T-stat 0.03) — effectively zero and imprecisely estimated.
- External financing proxies (interacted with GDP growth):
  - Lagged net capital flows to GDP ratio: -1.89 (T-stat 1.45) — not statistically significant.
  - Lagged aid to GDP ratio: -7.90* (T-stat 1.80) and -8.46* (T-stat 1.90) in different specifications — suggests larger aid shares decrease procyclicality (significant at 10%).
- Interpretation:
  - Domestic financial market characteristics appear irrelevant for cyclicality in SSA (median private credit/GDP in SSA is half the developing-country average).
  - Greater exposure to capital flows does not significantly reduce procyclicality.
  - Larger aid-to-GDP shares appear to decrease procyclicality, suggesting concessional finance can mitigate spending cyclicality.

### Aid, capital flows, and procyclicality — synthesis
- Net capital flows historically procyclical in developing countries; relationship weaker in low-income countries.
- Paper reports negative but statistically insignificant coefficient for net capital flows interacting with GDP growth — consistent with capital-scarcity hypothesis.
- Aid to GDP interaction: negative, significant, and large estimate in Table 6 (column 4); contrasts with previous findings (e.g., Thornton, 2008).
- Chauvet and Guillaumont (2008): aid became less procyclical since the 1990s, especially in SSA — consistent with paper's finding that aid may have helped mitigate financial constraints on fiscal policy over the whole period.

### Macroeconomic stability, fiscal space, and procyclicality (Table 7 and Table 8)
- Historical public external debt-to-GDP in SSA:
  - 1970: 18 percent
  - 1982 (start of debt crisis): 40 percent
  - 1996 (HIPC launch): 80 percent
  - 2007: 40 percent
- Empirical findings (Table 7):
  - Lagged inflation: no systematic effect on fiscal cyclicality.
  - Lagged public external debt-to-GDP ratio: smaller (lagged) external debt-to-GDP appears to significantly diminish procyclicality (columns 2 and 3); in part due to debt relief for HIPC countries (column 4).
  - Including proxies for fiscal space and external financial conditions: public external debt impact remains significant.
- Predicted-decade exercise (Table 8) — medians and predicted β:
  - Public external debt to GDP — Median / Predicted β:
    - 1970-1979: 15.5 / 0.73
    - 1980-1989: 43.4 / 1.55
    - 1990-1999: 75.5 / 2.50
    - 2000-2008: 62.5 / 2.11
  - Interpretation: public external debt-to-GDP closely mirrors evolution of cyclicality coefficient; the decline in debt ratios explains a large share of the fall in procyclicality in SSA since 2000.

### IMF programs and procyclicality (selected results from Table 7)
- IMF program indicators:
  - IMF program in current year interacting with GDP growth: 2.65* (t=1.84) — tends to increase procyclicality.
  - IMF program completed interacting with GDP growth: -0.23 (t=0.19) — suggests mild decrease in procyclicality, not significant.
  - IMF program about to start interacting with GDP growth: 2.06 (t=0.74) — not significant.
- Authors' interpretation:
  - Ambiguous role of IMF programs: countries in programs may have more procyclical policies while under restrictive commitments; exit from programs may be associated with slightly less procyclical behavior (not statistically robust).

### Additional caveats and measurement issues
- No attempt to measure automatic stabilizers or separate discretionary policy due to data limitations.
- Commodity export data missing for many observations; coding missing as zero affects statistical significance but not qualitative results.
- External instrument (export-weighted trading-partner growth) may have indirect effects via exports and tax revenues; authors contend this channel is quantitatively small and control for terms of trade growth.

### Conclusions and policy implications
- Main empirical conclusions:
  - Fiscal policy in SSA is strongly procyclical, with an elasticity of government spending to output growth above one (higher than in other developing countries).
  - Procyclicality in SSA appears to have declined in the 2000s relative to the 1980s–1990s.
  - Declining public external debt ratios explain a large share of the decrease in procyclicality since 2000; evidence consistent with increased fiscal space enabling more flexibility.
  - More aid could reinforce recent less procyclical trends despite concerns about aid volatility.
- Policy recommendations:
  - Continue efforts to stabilize debt levels; these efforts facilitate using fiscal policy as a stabilizing tool.
  - Strengthen public financial and debt management and medium-term fiscal frameworks to ensure sustainable debt accumulation and enforce fiscal restraint in good times.
  - Expand tax bases (for example, VAT or corporate tax) to make revenue collection more responsive to the cycle.
  - Reinforce automatic stabilizers with more comprehensive social safety nets to reduce welfare costs of downturns and make spending more countercyclical.

### Suggested avenues for future research
- Investigate regional differences in procyclicality patterns and why other developing countries became more procyclical.
- Study fiscal institutions and rules in SSA over the last two decades to understand their role in promoting less procyclical policies.
- Conduct econometric studies on technical and administrative constraints and the effectiveness of automatic stabilizers in SSA.
- Analyze the size and determinants of fiscal multipliers in SSA; this paper does not estimate fiscal multipliers.
- Assess how changes in procyclicality affect output volatility and growth.

*Source: _wp09274 - Appendix 1 gives more details on the variables used. (PDF chapter/section, 1970–2008 sample as reported in the source content.)*

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

### _wp09274 - References .............................................................................................................

### I. INTRODUCTION
- Context and motivation:
  - In 2009 about three-fourths of African countries are expected to increase their fiscal deficits excluding grants, or to decrease their projected surplus (IMF, 2009a).
  - Many are letting automatic stabilizers operate, and some are actively pursuing countercyclical policies.
  - This contrasts with much more modest increases, and even decreases, in fiscal deficits possible in past global slowdowns (Figure 1).
- Paper objectives:
  - (i) Document facts about the magnitude and evolution of cyclical patterns in government expenditures in SSA since 1970, correcting for possible bias due to endogeneity.
  - (ii) Identify factors underlying changes in cyclicality over time, with emphasis on macroeconomic and institutional conditions.
- Focus and measurement rationale:
  - Focus on government spending as the fiscal policy instrument variable (Kaminsky, Reinhart, and Vegh, 2004).
  - Tax rates not used due to data limitations for the sample.
  - Outcome measures (overall fiscal balance, tax revenues) considered less appropriate because they are partly determined by policymakers and affected by output fluctuations.
- Stylized background:
  - Since the late 1990s many SSA countries reached a “post-stabilization phase” characterized by sustainable fiscal and external positions, single-digit inflation, deeper domestic financial markets, and better institutions (Table 1).
  - The paper aims to close gaps in rigorous econometric analysis of fiscal cyclicality in SSA.

### II. LITERATURE REVIEW
- Empirical regularities:
  - Fiscal policy in developing countries generally observed to be procyclical; advanced economies are consistently acyclical or countercyclical (Braun, 2001; Lane, 2003; Kaminsky, Reinhart, and Végh, 2004; Alesina, Campante, and Tabellini, 2008; Ilzetzki and Végh, 2008).
  - Some studies question the extent of procyclicality in developing countries (Rigobón, 2003; Jaimovich and Panizza, 2007).
  - Evidence on evolution over time is mixed: Kaminsky, Reinhart, and Végh (2004) show decline after 1980; Alesina, Campante, and Tabellini (2008) do not find such evidence.
- Evidence specific to SSA:
  - Time series regressions for low-income African countries (1960–2004) find government consumption on average highly procyclical (Thornton, 2008).
  - Panel regressions addressing endogeneity find fiscal policy on average procyclical for 1989–2002 in a sample of low and middle-income African countries (Diallo, 2009).
  - Regression-based analysis limited (examples: South Africa case studies).
- Theoretical and empirical explanations for procyclicality:
  - Harmful effects: procyclical fiscal policy tends to exacerbate business cycle fluctuations; depends on fiscal multipliers and possibility of crowding out private investment (Lane, 2003a).
  - Two broad factor groups:
    - Political and institutional factors: lack of institutional controls, corruption, limits on executive checks, volatile tax bases, and competition for windfalls drive overspending in good times (Tornell and Lane, Talvi and Vegh, Alesina et al., Calderon et al., Akitoby et al.).
    - Financing constraints: limited access to international capital markets and shallow domestic financial markets make fiscal policy procyclical in downturns (Gavin and Perotti, Caballero and Krishnamurthy, Kaminsky et al., Alberola and Montero).
  - Role of aid flows is inconclusive: Akitoby et al. (2006) find no evidence aid dependency leads to more procyclical spending; Thornton (2008) finds the opposite for SSA.
- Contribution of this paper:
  - Extends endogeneity-corrected regression estimates to benchmark procyclicality in SSA.
  - Estimates evolution of cyclical fiscal patterns across advanced, developing, and SSA economies.
  - Analyzes specific factors explaining changes in procyclicality in SSA over time.
  - Provides econometric evidence on the role of fiscal space in diminishing procyclicality in the region.

### III. EMPIRICAL STRATEGY
- Model specification (as presented):
  - tiittiti eZYF ti,,,        (1)
  - ti P ti,10,           (2)
  - tititi vFY ,,,           (3)
  - tiiti e , ,            (4)
  - Interpretation:
    - F and Y measure growth in the fiscal variable and output; i denotes country; t denotes time.
    - Equation (1) captures fiscal reaction to contemporaneous output changes with controls Z and fiscal shocks e_{i,t}.
    - Cyclicality determined by coefficient : <0 countercyclical; =0 acyclical; >0 procyclical.
    - Equation (2) allows cyclicality to vary by country and over time with factors P.
    - Equation (3) allows fiscal policy to affect output.
    - Equation (4) decomposes fiscal shock into country-specific term i and residual ti,.
- Identification and endogeneity concerns:
  - Two main endogeneity sources:
    - Reverse causality between output growth and fiscal policy.
    - Correlation between output growth and unobserved country-specific effects.
  - Estimation approach:
    - Use difference (Diff-) and System (Sys-) GMM estimators (Arellano and Bond, 1991) to address endogeneity.
    - Internal instruments: past values of explanatory variables.
    - External instrument: (export) weighted average of GDP growth in a country’s trading partners (following Jaimovich and Panizza, Ilzetzki and Végh) to improve correlation with GDP growth.
  - Specification diagnostics:
    - Report p-values for the Hansen over-identification test and Arellano-Bond test for autocorrelation in first differences.
    - Diff-GMM chosen as preferred specification because it imposes fewer restrictions on correlation between instruments and error term.
- Additional modelling concerns:
  - Differentiate cyclical reaction from long-run relationship (Wagner’s Law) by taking first differences to use deviations from fixed long-run trends and rule out linear Wagner’s Law effects.
  - Include long-run determinants of government spending in robustness checks.
- Two-stage empirical strategy:
  - Stage 1: Assume factors driving evolution of procyclicality are absent (i.e., 1 = 0 in (2)); benchmark average procyclicality over entire sample and by decade versus advanced and other developing economies.
  - Stage 2: Allow 1 ≠ 0 and interact GDP growth with variables P to analyze determinants of procyclicality changes over time; taking first differences shows how changes in factors affect procyclicality depending on the sign of 1.
- Data:
  - Annual data in an unbalanced panel covering 39 years (1970–2008) and 174 countries.
  - Sample composition: 44 countries in SSA, 33 advanced economies, and 97 non-sub-Saharan African developing countries.

*Source: _wp09274 - References .............................................................................................................*

### Appendix 1 gives more details on the variables used.

### _wp09274 - Appendix 1 gives more details on the variables used.

### Data and key variables
- Dependent variable: growth in real central government spending.
- Key explanatory variable: growth in real GDP.
- Detrending approach: the paper does not use Hodrik-Prescott detrending because "detrending is highly problematic in developing countries (see Aguiar and Gopinath, 2004)"; instead, a less parametric approach is used.
- No attempt is made to separate discretionary government spending from automatic stabilizers; the aim is to capture overall cyclical behavior of fiscal policy.
- Political institutions: Polity 4 dataset (variables: degree of democracy, constraints on the executive, political competition). Robustness checks use International Country Risk Guide (starts in 1985) and World Bank’s Database of Political Institutions (starts in 1975).
- Commodity exposure proxy: share of commodity exports in GDP used to proxy volatility in tax revenues.
- Financing restriction proxies:
  - Domestic: share in GDP of credit to the private sector; real central bank interest rate.
  - External access: ratios of net capital flows to GDP; official development aid to GDP.
  - Sustainability/stabilization: debt-to-GDP ratio; inflation rate.
  - IMF program dummy: equals 1 if country has an IMF program in the current year (and several transformed versions used).
- Controls: lagged central government spending growth (instrumented using past values); growth in terms of trade; robustness controls include growth in oil prices, commodity prices, trade openness, a measure of democracy, dependency ratio, degree of urbanization.
- Timing and sample: 1970–2008 for main results; sub-Saharan Africa (SSA), other developing countries, advanced economies.

### Methodology
- Main specification: equation (1) with decomposition of fiscal variable into long-run trend and cyclical component.
- Estimation methods:
  - OLS with fixed effects (column 1).
  - 2SLS to address reverse causality (column 2); instruments include growth of trading partners weighted by exports.
  - Two-step Difference-GMM (Diff-GMM) preferred (column 3); instruments include past values of real GDP growth, growth of trading partners weighted by exports, and past values of the lagged dependent variable. Windmeijer (2005) finite sample correction applied for two-step GMM standard errors.
  - System-GMM (Sys-GMM) presented as robustness.
- Tests reported: Arellano-Bond tests for AR(1) and AR(2); Hansen test p-values; Sargan test p-values. Attention to instrument proliferation and serial correlation (Roodman, 2006).
- Addressing reverse causality: lagging factors of interest when appropriate; in column (2) GDP growth instrumented; Diff-GMM and 2SLS typically yield larger coefficients than OLS for developing countries and SSA.

### Key empirical findings (1970–2008)
- Overall result: fiscal policy is procyclical in sub-Saharan Africa over 1970–2008.
- Table 2 cyclicality coefficient (β) summary by estimator:
  - Sub-Saharan Africa:
    - OLS: 0.92*** (T-stat 6.3)
    - 2SLS: 2.21 (T-stat 1.23)
    - Diff-GMM: 1.54*** (T-stat 4.04)
  - Other Developing Countries:
    - OLS: 0.68*** (T-stat 6.15)
    - 2SLS: 1.25* (T-stat 1.87)
    - Diff-GMM: 1.16** (T-stat 2.41)
  - Advanced Economies:
    - OLS: -0.18 (T-stat 0.29)
    - 2SLS: -1.85 (T-stat 0.78)
    - Diff-GMM: -0.36 (T-stat 0.64)
- Magnitude interpretation:
  - "We find that a 2 percentage point increase in the rate of real GDP growth raises the growth rate of real government spending by about 3 points in SSA countries and 2.4 points in other developing countries; it does not affect the growth of real government expenditure in advanced economies."
- Preferred estimates (Diff-GMM / robustness):
  - For SSA, cyclicality coefficients generally stay in the 1.5–2 range across robustness checks.
  - For other developing countries, coefficients generally in the 1.1–1.2 range.
- Diagnostics (selected p-values from Table 3):
  - Arrelano-Bond test for AR(1): examples include 0.001, 0.000, 0.001 across columns.
  - Arrelano-Bond test for AR(2): examples include 0.232, 0.917, 0.207 across columns.
  - Hansen test p-values reported (examples): 0.5150, 0.7980, 0.6270; authors note p-values are high enough to reject endogeneity but below 0.8 in cases.
  - Sargan test p-values vary; for SSA specification authors report satisfactory p-values; for other developing countries some p-values are too low.

### Robustness and sensitivity
- Including controls for common shocks (oil and commodity price growth) and long-term determinants of government spending yields coefficients of expected sign but not statistically significant; estimated cyclicality coefficients remain similar (1.5–2 for SSA; 1.1–1.2 for other developing countries).
- Sys-GMM estimates (columns 3, 6, 9 in Table 3) are very close to Diff-GMM preferred estimates.
- Concern addressed: external instrument validity for low-income countries (export-weighted trading partner growth) — potential indirect effect via exports and tax revenues is considered quantitatively unimportant due to near-disappearance of export taxes; terms of trade shocks included to control for this. Results using only internal instruments are reported as very similar (results available upon request).

### Evolution by decade
- Table 4: Two-step Diff-GMM estimates by decade (dependent variable: growth in central government expenditures)
  - Sub-Saharan Africa:
    - 1970-1979: -1.53
    - 1980-1989: 1.83***
    - 1990-1999: 2.1***
    - 2000-2008: 1.48**
  - Other developing countries:
    - 1970-1979: 1.9
    - 1980-1989: -0.03
    - 1990-1999: 1.17*
    - 2000-2008: 1.73***
  - Advanced economies:
    - 1970-1979: -0.08
    - 1980-1989: 0.09
    - 1990-1999: -0.07
    - 2000-2008: 0.04
- Interpretation:
  - Advanced economies: acyclical throughout the period.
  - Other developing countries: acyclical in 1970s–1980s; increasingly procyclical since the 1990s.
  - Sub-Saharan Africa: acyclical in the 1970s (cannot reject acyclicity), positive and increasing procyclicality in the 1980s and 1990s; coefficient for 2000–08 falls and is lower than for other developing countries, suggesting fiscal policy in SSA has become less procyclical in recent years, but the decline is not statistically significant given large standard deviations.

### Determinants of cyclical fiscal behavior: Political economy and institutions
- Main finding: no evidence that political institutions (degree of democracy, constraints on the executive, political competition) have any effect on cyclical behavior of fiscal policy in SSA (Table 5, columns 1–3).
- Results robust to:
  - Restricting sample to post-1990 years.
  - Using alternative institutional measures (International Country Risk Group; World Bank Database).
  - Interacting political variables with GDP growth during "good times" (dummy when GDP growth is above median) — results unchanged.
- Commodity exports share:
  - Estimated large and positive coefficient for commodity exports to GDP ratio when interacted with GDP growth.
  - Table 5 reports coefficient on "Commodity exports to GDP ratio" (interacted with GDP growth): 7.8 (T-stat 1.62) — authors note data paucity and caution interpreting this result; when missing observations are coded as zero, qualitative results remain but lose statistical significance.
- Interpretation: marginal support that higher reliance on commodity exports (proxy for output/revenue volatility) is associated with more procyclical fiscal behavior in SSA.

### Determinants of cyclical fiscal behavior: Financing restrictions
- Table 6 main results (variables interacted with GDP growth; dependent variable: growth in central government expenditures; two-step Diff-GMM):
  - GDP Growth coefficients across columns (examples): 2.39*** (T-stat 2.76), 2.35*** (4.43), 3.1*** (3.47), 3.33*** (4.13), 3.09*** (4.0).
  - Domestic financing proxies:
    - Private credit to GDP ratio (interacted): coefficient reported as 0.01 (T-stat 0.34) — effectively zero and imprecisely estimated.
    - Lagged real central bank interest rate (interacted): coefficient reported as -0.00 (T-stat 0.03) — effectively zero and imprecisely estimated.
  - External financing proxies:
    - Lagged net capital flows to GDP ratio (interacted): coefficient -1.89 (T-stat 1.45) — not statistically significant.
    - Lagged aid to GDP ratio (interacted): coefficient -7.90* (T-stat 1.80) and -8.46* (T-stat 1.90) in different specifications — suggests larger aid shares decrease procyclicality; significance at 10% level.
  - HIPC decision point dummy: coefficient examples 2.43 and 0.46 (not clearly significant).
- Interpretation:
  - Domestic financial market characteristics appear irrelevant for cyclicality in SSA (coefficients near zero, imprecisely estimated). Possible explanation: underdeveloped domestic financial markets (median private credit/GDP in SSA is half the developing-country average).
  - Greater exposure to capital flows does not significantly reduce procyclicality.
  - A larger share of aid in GDP seems to decrease procyclicality, implying access to concessional finance helps mitigate the tendency for spending to rise and fall with output.

### Additional notes and caveats from the authors
- Measurement limitations: data on automatic stabilizers and policy formulation/implementation capacity in SSA are not examined due to measurement difficulties.
- Commodity export data are missing for many observations; robustness checks coding missing values as zero change statistical significance but not qualitative conclusions.
- Export-weighted trading partner growth as instrument may have indirect effects via exports and tax revenues; authors argue export taxes have quasi-disappeared over the period, reducing this concern; terms of trade controls are included.
- Large standard deviations in procyclicality coefficients imply that detecting systematic changes over time in SSA requires looking beyond arbitrary decade splits to changes in specific factors (addressed in the Factors section).

_Italic: Source: _wp09274 - Appendix 1 gives more details on the variables used. (PDF chapter/section, 1970–2008 sample as reported in the source content.)_

### appendix for variable description.

### appendix for variable description.

### Aid, capital flows, and procyclicality
- Net capital flows are consistently found to be procyclical in developing countries, though the relationship is somewhat weaker in low-income countries because they tend to be less integrated into global financial markets and more capital-scarce than medium-income countries (Kaminsky, Reinhart, and Végh, 2004).
- In circumstances of capital scarcity, larger capital inflows may help to decrease procyclicality by relaxing financial constraints; the paper reports a negative but statistically insignificant coefficient consistent with this hypothesis.
- Previous studies found aid flows to be weakly procyclical (Bulir and Hamman, 2008). Despite a high share of aid in total government spending in many countries studied (Appendix Table A.3), the paper finds a negative, significant, and large estimate of the impact of the aid-to-GDP ratio on procyclicality (Table 6, column 4).
- This finding contrasts with Thornton (2008), who found countries receiving more foreign aid are more procyclical; the negative aid effect holds even among countries that have reached the HIPC initiative decision point (Table 6, column 5).
- Chauvet and Guillaumont (2008) find that (i) aid has had a stabilizing effect on countries vulnerable to external shocks because it is less procyclical than exports, and (ii) the procyclicality of aid flows has declined since the 1990s, particularly in sub-Saharan Africa. This may explain why aid flows appear to have mitigated financial constraints on fiscal policy in SSA over the whole period, given the previous finding that procyclicality has fallen in recent years.

### Macroeconomic stability, fiscal space, and procyclicality
- Historical context:
  - Public external debt to GDP in the region: 18 percent in 1970; 40 percent at the start of the debt crisis (1982); 80 percent in 1996 when the HIPC initiative was launched; 40 percent in 2007.
- Conceptual link:
  - High inflation and debt can constrain a government’s ability to adjust fiscal policy over the cycle because policy is subordinated to avoiding hyperinflation and default.
  - Fiscal space is understood as availability of budgetary room so a government can use resources for desired purposes without prejudicing fiscal sustainability (Heller, 2005).
- Empirical findings (Table 7):
  - Lagged inflation does not seem to affect fiscal cyclicality (column 1).
  - A smaller (lagged) external debt-to-GDP ratio appears to significantly diminish procyclicality over the period (columns 2 and 3), and in part this may be due to debt relief for HIPC countries (column 4).
  - Including proxies for both fiscal space and external financial conditions: the impact of public external debt is affected very little and remains significant, while the remaining factors lose significance (Table 7, column 5).
  - Interpretation: even for a given level of access to international financial resources (including debt relief), countries with less debt are more likely to have less procyclical fiscal policies—possibly because bringing debt down to sustainable levels is less of a binding policy constraint.

### Impact of IMF programs (Table 7 results)
- The role of IMF programs on fiscal space and procyclicality is ambiguous:
  - IMF program in any given year tends to increase procyclicality (column 1).
  - A dummy equal to 1 if the country had an IMF program in the previous year but not in the current year (indicating structural reforms but no longer restrictive fiscal commitments) suggests the country was on average mildly less procyclical, though the coefficient is not statistically significant.
  - A country that will start an IMF program in the next year has more procyclical policies, suggesting countries request programs when they have no fiscal flexibility; again the coefficient is far from significant.
- Table 7 (selected coefficient and significance excerpts):
  - GDP Growth coefficients across specifications: 1.68*** (t=3.57); 0.09 (t=0.06); 0.27 (t=0.22); 2.32** (t=2.97); 0.11 (t=0.04); 1.20 (t=1.12); 2.83*** (t=3.46); 1.77** (t=1.93).
  - Lagged public external debt to GDP ratio interacting with GDP growth: 3.20 (t=1.32); 3.01* (t=1.84); 3.35* (t=1.70).
  - HIPC decision point reached interacting with GDP growth: -3.43** (t=2.10); 2.88 (t=0.74).
  - Lagged net capital flows to GDP ratio interacting with GDP growth: 8.34 (t=0.92).
  - Lagged aid to GDP ratio interacting with GDP growth: -7.98 (t=0.71).
  - IMF program interacting with GDP growth: 2.65* (t=1.84).
  - IMF program completed interacting with GDP growth: -0.23 (t=0.19).
  - IMF program about to start interacting with GDP growth: 2.06 (t=0.74).
  - Observations by column: 1464; 1291; 1291; 1464; 1291; 1464; 1464; 1423.
  - Significance legend: * significant at 10% level; ** significant at 5% level; *** significant at 1% level. Absolute values of T statistics using Windmeijer (2005)'s finite sample correction for standard errors in parentheses.
  - Note: All regressions include a control for terms of trade growth, lagged growth in central government spending, and the respective factor itself. Equation (3) uses the same controls as equation (2) except for external public debt, which is replaced by the HIPC dummy. GDP growth and the lagged dependent variable are instrumented for using lags, and the growth of major trading partners is used as an exogenous instrument.

### Explaining the evolution of procyclicality (Table 8)
- Approach: Assuming constant impacts over the period, predict the cyclicality parameter β for each decade using estimated values and median values per decade for each factor.
- Key result:
  - Public external debt-to-GDP mirrors the evolution of the cyclicality coefficient and appears central to explaining the increase and later decrease in procyclicality in sub-Saharan Africa between 1970 and 2008.
- Table 8 excerpts (median and predicted β by decade and variable):
  - Net capital flows to GDP — Median / Predicted β:
    - 1970-1979: 1.6 / 2.4
    - 1980-1989: 2.9 / 2.4
    - 1990-1999: 2.5 / 2.4
    - 2000-2008: 1.5 / 2.4
  - Aid to GDP — Median / Predicted β:
    - 1970-1979: 4.3 / 2.9
    - 1980-1989: 8.5 / 2.6
    - 1990-1999: 11 / 2.4
    - 2000-2008: 9 / 2.5
  - Public external debt to GDP — Median / Predicted β:
    - 1970-1979: 15.5 / 0.73
    - 1980-1989: 43.4 / 1.55
    - 1990-1999: 75.5 / 2.50
    - 2000-2008: 62.5 / 2.11
- Interpretation: Public external debt-to-GDP is the only variable that comes close to explaining the rise and fall of the procyclicality coefficient over the period.

### Conclusions and policy implications
- Main findings:
  - Fiscal policy in sub-Saharan Africa is strongly procyclical, with an elasticity of government spending to output growth above one (higher than in other developing countries), but procyclicality seems to have been declining in the last decade.
  - The decrease in debt ratios in recent years explains a large share of the fall in procyclicality in sub-Saharan Africa since 2000; there is tentative evidence that lower debt ratios have allowed countries more flexibility in setting fiscal policy objectives (i.e., created fiscal space).
  - More aid to the region could reinforce recent less procyclical trends, despite concerns about volatility in official development aid flows.
- Policy recommendations:
  - Continue efforts to stabilize debt levels; these efforts are paying off and should be furthered by countries that wish to use fiscal policy as a stabilizing tool.
  - Strengthen public financial and debt management and medium-term fiscal frameworks to ensure current debt accumulation is sustainable and to help enforce fiscal restraint in good times.
  - Expand tax bases (for example, VAT or corporate tax) to make revenue collection more responsive to the cycle.
  - Reinforce automatic stabilizers with more comprehensive social safety nets to reduce the welfare costs of downturns and make spending more countercyclical.

### Suggested avenues for future research
- Investigate why patterns of procyclicality differ across regions and why other developing countries may have become more procyclical.
- Study the variety of fiscal institutions and rules adopted in sub-Saharan Africa over the last two decades to understand the role of political and fiscal institutions in promoting less procyclical policies.
- Conduct econometric studies on the role of technical and administrative constraints and the effectiveness of automatic stabilizers in reducing fiscal procyclicality in the region.
- Analyze the size and determinants of fiscal multipliers in sub-Saharan Africa; the current study does not address the other side of the relationship between fiscal policy and growth.
- Assess how changes in procyclicality affect output volatility and growth.

*Source: appendix for variable description.*

### REFERENCES

### REFERENCES

### Bibliographic references
- Adam, Christopher S., and David L. Bevan, 2004, “Fiscal Policy Design in Low-income Countries,” in Fiscal Policy for Development: Poverty, Reconstruction and Growth, ed. by Tony Addison and Alan Roe (New York: Palgrave Macmillan).
- Aghion, Philippe, and Ioana Marinescu, 2007, “Cyclical Budgetary Policy and Growth: What Do We Learn from OECD Panel Data?” NBER Macroeconomics Annual, ed. by M. Gertler and K. Rogoff (Cambridge, Massachusetts: MIT Press).
- Aguiar, Mark, and Gita Gopinath, 2004.“Emerging Market Business Cycles: The Cycle is the Trend” (unpublished; Chicago: University of Chicago).
- Akitoby, Bernardin, Benedict Clements, Sanjeev Gupta, and Gabriela Inchauste, 2006, “Public Spending, Voracity, and Wagner’s Law in Developing Countries,” European Journal of Political Economy, Vol. 22, pp. 908–24.
- Alberola, Enrique, and José M. Montero, 2006, “Debt Sustainability and Procyclical Fiscal Policies in Latin America,” Banco de Espana Working Paper No. 0611 (Madrid: Banco de Espana).
- Alesina, Alberto, Filipe Campante, and Guido Tabellini, 2008, “Why is Fiscal Policy Often Procyclical?” Journal of the European Economic Association, Vol. 6, No, 5, pp. 1006–36.
- Arellano, Manuel, and Stephen Bond, 1991, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,” Review of Economic Studies, Vol. 58, pp. 277–97.
- Balassone, Fabrizio, and Manmohan Kumar, 2007, “Cyclicality of Fiscal Policy,” in Promoting Fiscal Discipline, ed. by Manmohan S. Kumar and Teresa Ter-Minassian (Washington: International Monetary Fund).
- Berg, Andrew, Norbert Funke, Alejandro Hajdenberg, Victor Lledó, Rolando Ossowski, Martin Schindler, Antonio Spilimbergo, Shamsuddin Tareq, and Irene Yackovlev, 2009, “Fiscal Policy in Sub-Saharan Africa in Response to the Impact of the Global Crisis,” IMF Staff Position Note 09/10 (Washington: International Monetary Fund).
- Blundell, Richard, and Stephen Bond, 1998, “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models,” Journal of Econometrics, Vol. 87, pp. 115–43.
- Bratton, Nicholas, and Michael Van de Walle, 1997, Democratic Experiments in Africa, (Cambridge, United Kingdom: Cambridge University Press)
- Braun, M. 2001. “Why Is Fiscal Policy Procyclical in Developing Countries?” (unpublished; Cambridge, Massachusetts: Harvard University).
- Bulir, Ales, and A. Javier Hamman, 2003, “Aid Volatility : An Empirical Assessment,” IMF Staff Papers, Vol. 50, No. 1, pp. 64–89.
- Bulir, Ales, and A. Javier Hamman, 2008, “Volatility of Development Aid: From the Frying Pan into the Fire?” World Development, Vol. 36, No. 10, pp. 2048–66.
- Caballero, Ricardo J., and Arvind Krishnamurthy, 2004, “Fiscal Policy and Financial Depth,” NBER Working Papers 10532 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Calderón, Cesar, Roberto Duncan, and Klaus Schmidt-Hebbel, 2004, “Institutions and Cyclical Properties of Macroeconomic Policies,” Central Bank of Chile Working Paper No. 285 (Santiago: Central Bank of Chile).
- Chauvet, Lisa, and Patrick Guillaumont, 2008, “Aid, Volatility and Growth Again: When Aid Volatility Matters and When it Does Not,” United Nations World Institute for Development Economics (UNU-WIDER) Research Paper No. 20078/78 (New York City: UNU-WIDER).
- Diallo, Oumar, 2009, “Tortuous Road Toward Countercyclical Fiscal Policy: Lessons from Democratized Sub-Saharan Africa,” Journal of Policy Modeling, Vol. 31, pp. 36–50.
- Du Plessis, Stanley, and Willem Boshoff, 2007, “A Fiscal Rule to Produce Counter-cyclical Fiscal Policy in South Africa” Stellenbosch University Working Paper 13/2007 (Stellenbosch, South Africa: Stellenbosch University).
- Du Plessis, Stanley, Ben Smit, and Federico Sturzenegger, 2007, “The Cyclicality Of Monetary And Fiscal Policy In South Africa Since 1994,” South African Journal of Economics, Vol. 75, No. 3, pp. 391–411.
- Frankel, Jeffrey, Ben Smit, and Federico Sturzenegger, 2008, “Fiscal and Monetary Policy in a Commodity-Based Economy,” Economics of Transition, Vol.16, No. 4, pp. 679–713.
- Gali, Jordi, 2005, “Modern Perspective on Fiscal Stabilization Policies,” CESifo Economic Studies, Vol. 51, No. 4, pp. 587–99.
- Gali, Jordi, and Roberto Perotti, 2003, “Fiscal Policy and Monetary Integration in Europe,” Economic Policy, Vol. 18, pp. 533–72.
- Gavin, Michael, and Roberto Perotti, 1997, “Fiscal Policy in Latin America,” NBER Macroeconomic Annual, pp. 11–71 (Cambridge, Massachusetts: MIT Press).
- Greene, Joshua, 1989, “The External Debt Problem of Sub-Saharan Africa,” IMF Staff Papers, Vol. 36, No.,4, pp. 836–74.
- Gupta, Sanjeev, Benedict Clements, Emanuele Baldacci, and Carlos Mulas-Granados, 2005, “Fiscal Policy, Expenditure Composition, and Growth in Low-income Countries,” Journal of International Money and Finance, Vol. 24, (April), pp. 441–63.
- Heller, Peter S., 2005. “Understanding Fiscal Space,” IMF Policy Discussion Papers 05/4 (Washington: International Monetary Fund).
- Ilzetzki, Ethan, and Carlos A. Végh, 2008, “Procyclical Fiscal Policy in Developing Countries: Truth or Fiction?” NBER Working Paper No. 14191 (Cambridge, Massachusetts: National Bureau of Economic Research); available from http://www.nber.org/papers/w14191.
- International Monetary Fund, World Economic Outlook, October 2008, World Economic and Financial Surveys (Washington).
- International Monetary Fund, 2009a, Regional Economic Outlook: Sub-Saharan Africa (Washington, April).
- International Monetary Fund, 2009b, Regional Economic Outlook: Sub-Saharan Africa (Washington, October).
- Jaimovich, Dany, and Ugo Panizza, 2007, “Procyclicality or Reverse Causality?” Working Papers 1029 (Washington: Inter-American Development Bank).
- Kaminsky, Graciela, Carmen Reinhart, and Carlos A. Végh, 2004, “When It Rains, It Pours: Procyclical Capital Flows and Macroeconomic Policies,” in NBER Macroeconomics Annual, ed. by Kenneth Rogoff and Mark Gertler (Cambridge, Massachusetts: MIT Press). Also available as NBER Working Paper No. 10780: http://www.nber.org/papers/w10780.
- Keen, Michael and Mario Mansour, 2009, “Revenue Mobilization in Sub-Saharan Africa : Challenges from Globalization,” IMF Working Paper No 09/157 (Washington: International Monetary Fund).
- Lane, Philip R., 2003a, “Business Cycles and Macroeconomic Policy in Emerging Market Economies”, International Finance, Vol.6. No. , pp.89–108.
- Lane, Philip R., 2003b, “The Cyclical Behavior of Fiscal Policy: Evidence from the OECD,” Journal of Public Economics, Vol. 87, pp. 2661–75.
- Lindert, Peter H., 2004, Growing Public: Social Spending and Economic Growth since the Eighteenth Century, two volumes (Cambridge, United Kingdom: Cambridge University Press).
- Manasse, Paolo, 2006, “Procyclical Fiscal Policy: Shocks, Rules, and Institutions: A View from MARS,” IMF Working Paper No. 06/27 (Washington: International Monetary Fund).
- Marshall, Monty, and Keith Jaggers, 2009, “Polity IV Project. Political Regime Characteristics and Transitions, 1800–2007,” Center for Systemic Peace, www.systemicpeace.org/polity4.
- Nickell, Stephen, 1981, “Biases in Dynamic Models with Fixed Effects,” Econometrica, Vol. 49, pp.1417–26.
- O’Connell, 1988, “Fiscal Policy in Low-Income Africa”, World Bank Working Paper WPS 39 (Washington: World Bank).
- Reinhart, Carmen, Kenneth Rogoff, and Miguel Savastano, 2003, “Debt Intolerance,” NBER Working Paper No 9908 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Rigobón, Roberto, 2003, “Comments on: ‘When It Rains, It Pours’ by Graciela Kaminsky, Carmen Reinhart, and Carlos Vegh,” in NBER Macroeconomics Annual, ed. by M. Gertler and K. Rogoff (Cambridge, Massachusetts: MIT Press).
- Rodrik, Dani, 1998, “Why Do more open Economies Have Bigger Governments?” Journal of Political Economy Vol. 106, pp 997–1032.
- Roodman, David, 2006, “How To Do xtabond2: An Introduction to ‘Difference’ and ‘System’ GMM in Stata,” Center for Global Development Working Paper No 103 (Washington: Center for Global Development).
- Siebrits, Franz K., and Estian Calitz, 2007, “The Legacy and Challenge of Fiscal Policy in Sub-Saharan Africa,” South African Journal of Economics, Vol. 75, No 2, pp 221–35.
- Stawczynski, Michel, and Joseph Zeira, 2007, “Cyclicality of Fiscal Policy in Israel,” Israel Economic Review, Vol. 5, No 1, pp. 47–66.
- Stawczynski, Michel, and Joseph Zeira, 2009, “Cyclicality of Fiscal Policy: Permanent and Temporary Shocks”, CEPR Discussion Papers No. 7271 (London: Centre for Economic Policy Research).
- Talvi, Ernesto, and Carlos A. Végh, 2005, “Tax Base Variability and Pro-cyclical Fiscal Policy in Developing Countries,” Journal of Development Economics, Vol. 78, pp. 156–90.
- Thornton, John, 2008, “Explaining Pro-cyclical Fiscal Policy in African Countries,” Journal of African Economies, Vol. 17, No. 3, pp. 451–64.
- Tornell, Aaron, and Phillip Lane, 1999, “The Voracity Effect,” American Economic Review, Vol.89, No. 1, pp.22-46.
- World Bank, 2009, Global Economic Prospects: Commodities at the Crossroads (Washington: World Bank).

### Appendix A — Countries in the sample
- Sub-Saharan Africa (selected entries): Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Democratic Republic of Congo, Republic of Congo, Côte d'Ivoire, Equatorial Guinea, Eritrea, Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Seychelles, Sierra Leone, South Africa, Swaziland, São Tomé & Príncipe, Tanzania, Togo, Uganda, Zambia, Zimbabwe, Kiribati, Kuwait, Kyrgyz Republic, Laos, Latvia, Lebanon, Libya, Lithuania, Macedonia, Malaysia, Maldives, Mauritania, Mexico, Montenegro, Morocco, Myanmar, Nepal, Nicaragua, Oman, Pakistan, Panama, Papua New Guinea, Paraguay, Peru, Philippines, Poland, Qatar, Russia, Samoa, Saudi Arabia, Singapore, Solomon Islands, Sri Lanka, St. Kitts and Nevis, St. Lucia, St. Vincent & Grenadines, Sudan, Suriname, Syria, Tajikistan, Thailand, Timor-Leste, Tonga, Trinidad and Tobago, Tunisia, Turkey, Ukraine, United Arab Emirates, Uruguay, Venezuela, Vietnam, Yemen, Vanuatu.
- Advanced Economies and Other Developing Countries: Australia, Austria, Belgium, Canada, Hong Kong, Cyprus, Czech Republic, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Israel, Italy, Japan, Korea, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Taiwan, United Kingdom, United States, and a large set of other developing countries listed in the sample.
- Note: The country classification comes from the World Economic Outlook (IMF).

### Appendix B — Definitions and sources of variables (selected entries)
- Growth in terms of trade — WEO, IMF: Price of exports divided by the price of exports.
- Trade openness — WEO, IMF: Sum of total exports and exports divided by GDP.
- Oil price — WEO, IMF: Price in US dollars of a barrel of crude oil.
- Index of commodity prices — WEO, IMF: Price of non fuel commodity exports.
- Urbanization — WDI, World Bank: Ratio of urban population to total population.
- Private credit to GDP — WDI, World Bank: Ratio of credit to the private sector to GDP.
- Net foreign capital flows — WEO, IMF: Capital inflows minus capital outflows.
- Current account balance — WEO, IMF.
- Inflation — WEO, IMF: Growth in the CPI.
- HIPC dummy — Strategy, Policy and Review department database, IMF: Equal to 1 at (if until June) and after the year a country reached the decision point to be considered for HIPC Initiative assistance.
- Real GDP growth — World Economic Outlook (WEO), IMF: Growth in nominal GDP deflated using the CPI.
- Constraints on the executive — Polity4 database, xconst variable: Extent of institutionalized constraints on the decision making powers of chief executives, from 1 (unlimited authority) to 7 (executive parity or subordination).
- Official Development Assistance — Global Development Finance, World Bank.
- Commodity Exports — UN Comtrade database: Based on Collier and Hoeffler (2002)'s definition of commodity exports: goods categories 0, 1, 2, 3,4 and 68 from the SITC4 nomenclature.
- Political competition — Polity4 database, polcomp variable: Degree of insitutionalization of political competition combined with the extent of government restriction on political competition, from 1 to 10.
- Dependency ratio — World Development Indicators (WDI), World Bank: Ratio of dependants to working age population.
- Real central bank interest rate — International Financial Statistics, IMF: Central bank main interest rate deflated using the CPI.
- IMF program dummy — Strategy, Policy and Review department database, IMF: Equal to 1 if there is an IMF program in place in the country during that year.
- Public external debt — Global Development Finance, World Bank: All public debt to foreign creditors.
- Democracy — Polity4 database, polity2 variable: Difference between a democracy index (0 to 10) and an autocracy index (0 to 10).
- Real growth in central government spending — WEO, IMF: Growth in nominal central government total spending deflated using the CPI.
- Real GDP growth of main trading partners — WEO, IMF: GDP growth of main trading partners (each partner weighted by its share of exports in the country's total exports) weighted by the share of exports in GDP.

### Appendix C — Descriptive statistics (Sub-Saharan Africa and Other Developing Countries, 1970-2008)
- Real GDP growth
  - Sub-Saharan Africa: Mean 3.39, Median 3.61, SD 6.8
  - Other Developing Countries: Mean 3.78, Median 4.4, SD 6.3
- Real growth in central government spending
  - Sub-Saharan Africa: Mean 4.15, Median 3.89, SD 45.2
  - Other Developing Countries: Mean 4.29, Median 4.27, SD 20.3
- Real GDP growth of main trading partners
  - Sub-Saharan Africa: Mean 1.11, Median 0.85, SD 0.89
  - Other Developing Countries: Mean 1.43, Median 1.21, SD 1.23
- Growth in terms of trade
  - Sub-Saharan Africa: Mean -0.48, Median 0.7, SD 0.09
  - Other Developing Countries: Mean 0.90, Median 18.6
- Private credit to GDP
  - Sub-Saharan Africa: Mean 33, Median 20.8, SD 71.48
  - Other Developing Countries: Mean 14.1, Median 1.23, SD 58
- Real central bank interest rate
  - Sub-Saharan Africa: Mean -42.8, Median 1.29, SD 749.8
  - Other Developing Countries: Mean -20.8, Median 1.96, SD 386.9
- Net foreign capital flows to GDP
  - Sub-Saharan Africa: Mean 2.97, Median 2.09, SD 13.7
  - Other Developing Countries: Mean 3.32, Median 2.22, SD 43.3
- Aid to GDP
  - Sub-Saharan Africa: Mean 11.02, Median 8.16, SD 11.14
  - Other Developing Countries: Mean 4.32, Median 1.13, SD 13.4
- Current account balance to GDP
  - Sub-Saharan Africa: Mean -5.43, Median -4.65, SD 12.7
  - Other Developing Countries: Mean -3.7, Median -3.1, SD 20.3
- Commodity exports to GDP
  - Sub-Saharan Africa: Mean 14.52, Median 10.31, SD 13.2
- Democracy
  - Sub-Saharan Africa: Mean -2.4, Median -5.9, SD -0.4
  - Other Developing Countries: Mean -2.7, Median 7.35
- Constraints on the executive
  - Sub-Saharan Africa: Mean 2.93, Median 1.9, SD 3.89
  - Other Developing Countries: Mean 2.25
- Political competition
  - Sub-Saharan Africa: Mean 3.82, Median 3.2, SD 4.86
  - Other Developing Countries: Mean 3.5
- Public external debt to GDP
  - Sub-Saharan Africa: Mean 58.74, Median 48.15, SD 42.9
  - Other Developing Countries: Mean 29.16, SD 6.4
- Inflation
  - Sub-Saharan Africa: Mean 50.7, Median 9.27, SD 678
  - Other Developing Countries: Mean 78.25, Median 7.5, SD 500.7
- IMF program completed
  - Sub-Saharan Africa: Mean 0.06, Median 0.0, SD 0.22
  - Other Developing Countries: Mean 0.05, Median 0.0, SD 0.22
- IMF program next year
  - Sub-Saharan Africa: Mean 0.07, Median 0.0, SD 0.25
  - Other Developing Countries: Mean 0.47, Median 0.0, SD 0.21

### Appendix D — Financing constraints and political institutions in good and bad times (Table A.4, selected results)
- Dependent variable: Growth in central government expenditures.
- Estimation: Two-step difference-GMM estimates.
- GDP Growth coefficients (columns indicated in table):
  - Column (1): 2.04**
  - Column (2): 2.53
  - Column (3): 1.72
  - Column (4): 2.35**
  - Column (5): 2.29***
  - Column (6): 3.04***
  - Column (7): 3.2***
- Interaction notes: All variables below are interacted with GDP growth.
- Political institutions (interaction effects reported where present):
  - Democracy: -0.26
  - Democracy*good times: 0.48
  - Constraints on the executive: -0.15
  - Constraints on the executive*good times: -0.15
  - Political competition: 0.07
  - Political competition*good times: 0
- Financing constraints (interaction effects reported where present):
  - Private credit to GDP ratio: 0.01
  - Private credit to GDP ratio*bad times: 0.01
  - Lagged real central bank interest rate: 0.00
  - Lagged real central bank interest rate*bad times: 0.02
  - Lagged net capital flows to GDP ratio: -1.58
  - Lagged net capital flows to GDP ratio*bad times: 2.91
  - Lagged aid to GDP ratio: -8.16**
  - Lagged aid to GDP ratio*bad times: 3.8
- Observations by column: 1295, 1205, 1205, 1216, 1147, 1428, 1387.
- Significance notation: * significant at 10% level; ** significant at 5% level; *** significant at 1% level.
- Definition of "Good times"/"Bad times": 'Good times' is a dummy equal to 1 if growth is above the median for the country over the period considered, and 'bad times' is a dummy equal to 1 if growth is below this median.
- Instrumentation: GDP growth and the lagged dependent variable are instrumented for using lags, and the growth of major trading partners is used as an exogenous instrument. See data appendix for variable description.

*Content derived from: _wp09274 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09274.pdf_