## _wp13153 - References .............................................................................................................

## Source details

**Canonical URL:** [_wp13153 - References .............................................................................................................](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13153.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13153.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13153.pdf.json)

---

### Introduction and research scope
- Focus: political budget cycles in Low-Income Countries (LICs) using panel data covering 68 LICs over 21 years, 1990-2010 (around 51 LICs had at least one election).
- Motivation: LICs are particularly vulnerable to election-related fiscal volatility due to weaker institutions and poor budget transparency.
- Election data source: National Elections across Democracy and Autocracy (NELDA) (Hyde and Marinov, 2012); 191 national elections during the sample period.
- Key research questions:
  - Magnitude and composition of election-year fiscal expansions and post-election adjustments across fiscal variables (government consumption, public investment, tax-revenue composition, budget balance).
  - Do national fiscal rules and IMF programs dampen political budget cycles in LICs?
- Empirical approach: dynamic panel equations estimated with System-GMM (Arellano and Bover, 1995; Blundell and Bond, 1998), Windmeijer (2005) correction for standard errors, Sargan/Hansen and AR(2) tests for instrument validity.

### Data and baseline specification
- Panel: 68 LICs, 1990–2010; panel is unbalanced.
- Main fiscal outcomes: government final consumption (current expenditures), public investment, tax revenue decomposed into direct, indirect, and trade taxes, and overall fiscal balance (percent of GDP).
- Covariates: real GDP growth rate, inflation rate, trade openness, foreign aid, external debt, natural resource rents, agriculture value added, fiscal rules; data sources include IMF WEO, Penn World Table 7.1, World Development Indicators.
- Election indicators: ELEt, ELEt-1, ELEt-2 measure impact during, one year after, and two years after an election.
- Tax revenue descriptive means in the sample:
  - Mean overall tax revenue: 14.8 percent.
  - Indirect taxes (taxes on goods and services): 5.6 percent of GDP.
  - Direct taxes (income, profits, capital gains): 4.4 percent of GDP.
  - Trade taxes: 3.8 percent of GDP.

### Baseline findings — expenditures, revenues, and fiscal balance
- Government consumption (current spending):
  - Election-year increase: 0.8 percentage point of GDP (ELEt coefficient significant).
  - No significant decrease in the two years after elections.
- Government investment:
  - Decrease of almost 0.4 percentage point of GDP the year following the election (statistically significant).
  - ELEt-2 coefficient negative but not significant.
- Tax revenues:
  - Overall tax effort increases significantly in the years following an election.
  - Indirect taxes (broadly defined) show no significant election-related cycle when disaggregated from trade taxes.
  - Direct taxes: no significant variation over the election cycle.
  - Trade taxes:
    - Slight (barely statistically significant) increase of about 0.11 percentage point of GDP during the election year.
    - This increase is maintained and strengthened during the two post-election years.
  - Interpretation: post-election partial rebuilding of fiscal buffers is achieved mainly through increased discretionary trade tax mobilization and cuts to public investment.
- Overall fiscal balance:
  - Fiscal deficit increases by about 1 percentage point of GDP during the election year (driven mainly by increased current spending).
  - Post-election attempts to rebuild buffers are present but not large or balanced enough to fully offset the election-year deviation.
  - Two years after election: some reduction in fiscal deficit (see endogeneity results for magnitude).

### Addressing endogeneity of election timing
- Election timing classification from NELDA: predetermined (fixed by constitution/procedure) vs endogenous (early/late relative to scheduled date).
  - Among 191 elections: 56.5 percent classified as predetermined; 108 predetermined, 38 endogenous, 45 unclassified.
- Re-estimated models with ELEPRE and ELEENDO indicators yield similar results:
  - Government current expenditures deviate during election years, increasing the overall fiscal deficit by about 1.3 percentage point of GDP.
  - Post-election years: partial fiscal buffer rebuilding accompanied by a reduction in public investment of about 0.4 percentage point of GDP.
  - Two years after election: estimates indicate a reduction of the fiscal deficit by about 0.5 percentage point of GDP.

### Domestic and international scrutiny — fiscal rules and IMF programs
- Hypotheses:
  - National fiscal rules can dampen political budget cycles by constraining incumbents’ ability to expand fiscal policy during elections.
  - IMF programs may act as international scrutiny mechanisms reducing election-driven fiscal manipulations via program conditionality.
- Identification concerns and strategies:
  - Fiscal rules: adoption and stability are endogenous. Strategy: interact election dummy with a 5-year lagged fiscal-rule dummy (FR) to mitigate endogeneity; results focused on election year only.
  - IMF programs: addressed via a first-stage pooled probit selection model; selection correction factor (λ) included in outcome equation (Maddala, 1983; Vella and Verbeek, 1999; Keen and Lockwood, 2010).
    - Selection model controls include external reserves, fiscal balance, trade openness, inflation rate, natural resource rents, a dummy indicating whether a national election is scheduled next year, and election × electoral competitiveness.
    - Selection-correction factor computed from probit (using standard normal density and cdf functions).

### Results on fiscal rules
- In LICs without national fiscal rules, political budget cycle on government consumption is around 1 percentage point of GDP (consistent with baseline).
- Interaction of election dummy with lagged national fiscal rule dummy:
  - Interaction coefficient turns negative and statistically significant but of low significance overall.
  - Marginal effect in LICs with national fiscal rules: size of the fiscal deviation during an election year is close to 0.13 percentage point of GDP (1 - 0.87 = 0.13).
- Interpretation and caveats:
  - Dampening role of national fiscal rules is present but weak in LICs, likely because very few LICs have national fiscal rules, enforcement/compliance is limited, and national rules in LICs are predominantly debt rules.
  - Only 4 LICs use national fiscal rules (text remark), limiting statistical power.

### Results on IMF program engagement
- Baseline interpretation:
  - In absence of an IMF program, government consumption deviates by about 1 percentage point of GDP during national elections.
  - In presence of an active IMF program, the size of the deviation drops to 0.34 percentage point of GDP.
- Selection-correction factor (λ) is positive and significant, indicating selection into IMF programs matters and its correction was crucial.
- Interpretation and caveats:
  - IMF programs—and the associated conditionality—appear to reduce the magnitude of election-driven consumption increases.
  - Coefficients on interaction terms with IMF programs exhibit low significance; majority of elections in LICs occur while countries are already under IMF agreements, limiting statistical power.
  - Self-selection bias is only partially controlled for despite the improved selection model.

### Key policy-relevant findings and implications
- Election-year fiscal expansion in LICs:
  - Government consumption increases (0.8 percentage point of GDP baseline; 1 percentage point in some specifications).
  - Overall fiscal deficit increases by about 1 percentage point of GDP during election years.
- Post-election adjustment:
  - Implemented through increased trade tax mobilization (about 0.11 percentage point of GDP at election and maintained thereafter) and reductions in public investment (about 0.4 percentage point of GDP the year after).
  - Adjustment is incomplete: current spending envelopes remain elevated and revenue mobilization is insufficient to fully offset election-year deviations.
- Role of constraints:
  - National fiscal rules: evidence of dampening political budget cycles in government consumption but effect is weak in LICs (marginal effect in rule countries ~0.13 percentage point of GDP).
  - IMF programs: associated with a lower political budget cycle (deviation drops to 0.34 percentage point of GDP under program engagement), but significance is limited and selection issues remain.
- Growth implications:
  - Election-driven volatility and the post-election sacrifice of public investment, together with increased trade taxes, likely harm future economic growth prospects in LICs.

### Concluding remarks
- Elections in LICs are associated with discernible political budget cycles: higher current spending and larger deficits during election years; partial post-election adjustment through trade tax increases and investment cuts.
- National fiscal rules and IMF programs appear to mitigate election-driven fiscal deviations, but their dampening effects are limited in magnitude and statistical precision in the LIC context.
- Policy implication: strengthening fiscal institutions, enforcement of fiscal rules, and effective international oversight may help reduce election-driven fiscal volatility; incomplete post-election adjustments and reduced public investment raise concerns for long-term growth.

*Source: Excerpt from paper content unit titled "_wp13153 - References ............................................................................................................."*

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

### _wp13153 - References .............................................................................................................

### Introduction and research scope
- Focus: political budget cycles in Low-Income Countries (LICs) using panel data covering 68 LICs over 21 years, 1990-2010 (around 51 LICs had at least one election).
- Motivation: LICs are particularly vulnerable to election-related fiscal volatility due to weaker institutions and poor budget transparency.
- Election data source: National Elections across Democracy and Autocracy (NELDA) (Hyde and Marinov, 2012); 191 national elections during the sample period.
- Key research questions:
  - Magnitude and composition of election-year fiscal expansions and post-election adjustments across fiscal variables (government consumption, public investment, tax-revenue composition, budget balance).
  - Do national fiscal rules and IMF programs dampen political budget cycles in LICs?
- Empirical approach: dynamic panel equations estimated with System-GMM (Arellano and Bover, 1995; Blundell and Bond, 1998), Windmeijer (2005) correction for standard errors, Sargan/Hansen and AR(2) tests for instrument validity.

### Data and baseline specification
- Panel: 68 LICs, 1990–2010; panel is unbalanced.
- Main fiscal outcomes: government final consumption (current expenditures), public investment, tax revenue decomposed into direct, indirect, and trade taxes, and overall fiscal balance (percent of GDP).
- Covariates: real GDP growth rate, inflation rate, trade openness, foreign aid, external debt, natural resource rents, agriculture value added, fiscal rules; data sources include IMF WEO, Penn World Table 7.1, World Development Indicators.
- Election indicators: ELEt, ELEt-1, ELEt-2 measure impact during, one year after, and two years after an election.
- Tax revenue descriptive means in the sample:
  - Mean overall tax revenue: 14.8 percent.
  - Indirect taxes (taxes on goods and services): 5.6 percent of GDP.
  - Direct taxes (income, profits, capital gains): 4.4 percent of GDP.
  - Trade taxes: 3.8 percent of GDP.

### Baseline findings — expenditures, revenues, and fiscal balance
- Government consumption (current spending):
  - Election-year increase: 0.8 percentage point of GDP (ELEt coefficient significant).
  - No significant decrease in the two years after elections.
- Government investment:
  - Decrease of almost 0.4 percentage point of GDP the year following the election (statistically significant).
  - ELEt-2 coefficient negative but not significant.
- Tax revenues:
  - Overall tax effort increases significantly in the years following an election.
  - Indirect taxes (broadly defined) show no significant election-related cycle when disaggregated from trade taxes.
  - Direct taxes: no significant variation over the election cycle.
  - Trade taxes:
    - Slight (barely statistically significant) increase of about 0.11 percentage point of GDP during the election year.
    - This increase is maintained and strengthened during the two post-election years.
  - Interpretation: post-election partial rebuilding of fiscal buffers is achieved mainly through increased discretionary trade tax mobilization and cuts to public investment.
- Overall fiscal balance:
  - Fiscal deficit increases by about 1 percentage point of GDP during the election year (driven mainly by increased current spending).
  - Post-election attempts to rebuild buffers are present but not large or balanced enough to fully offset the election-year deviation.
  - Two years after election: some reduction in fiscal deficit (see endogeneity results for magnitude).

### Addressing endogeneity of election timing
- Election timing classification from NELDA: predetermined (fixed by constitution/procedure) vs endogenous (early/late relative to scheduled date).
  - Among 191 elections: 56.5 percent classified as predetermined; 108 predetermined, 38 endogenous, 45 unclassified.
- Re-estimated models with ELEPRE and ELEENDO indicators yield similar results:
  - Government current expenditures deviate during election years, increasing the overall fiscal deficit by about 1.3 percentage point of GDP.
  - Post-election years: partial fiscal buffer rebuilding accompanied by a reduction in public investment of about 0.4 percentage point of GDP.
  - Two years after election: estimates indicate a reduction of the fiscal deficit by about 0.5 percentage point of GDP.

### Domestic and international scrutiny — fiscal rules and IMF programs
- Hypotheses:
  - National fiscal rules can dampen political budget cycles by constraining incumbents’ ability to expand fiscal policy during elections.
  - IMF programs may act as international scrutiny mechanisms reducing election-driven fiscal manipulations via program conditionality.
- Identification concerns:
  - Fiscal rules: adoption and stability are endogenous (self-selection). Strategy used: interact election dummy with a 5-year lagged fiscal-rule dummy (FR) to mitigate endogeneity; results focused on election year only.
  - IMF programs: potential endogeneity of program participation with elections and macro outcomes addressed via a first-stage pooled probit selection model; selection correction factor (λ) included in outcome equation (Maddala, 1983; Vella and Verbeek, 1999; Keen and Lockwood, 2010).
    - Selection model includes controls such as external reserves, fiscal balance, trade openness, inflation rate, natural resource rents, a dummy indicating whether a national election is scheduled next year, and election × electoral competitiveness.
    - Selection-correction factor computed from probit (using standard normal density and cdf functions).

### Results on fiscal rules
- In LICs without national fiscal rules, political budget cycle on government consumption is around 1 percentage point of GDP (consistent with baseline).
- Interaction of election dummy with lagged national fiscal rule dummy:
  - Interaction coefficient turns negative and statistically significant but of low significance overall.
  - Marginal effect in LICs with national fiscal rules: size of the fiscal deviation during an election year is close to 0.13 percentage point of GDP (1 - 0.87 = 0.13).
- Interpretation and caveats:
  - Dampening role of national fiscal rules is present but weak in LICs, likely because very few LICs have national fiscal rules, enforcement/compliance is limited, and national rules in LICs are predominantly debt rules.
  - Only 4 LICs use national fiscal rules (text remark), limiting statistical power.

### Results on IMF program engagement
- Baseline interpretation:
  - In absence of an IMF program, government consumption deviates by about 1 percentage point of GDP during national elections.
  - In presence of an active IMF program, the size of the deviation drops to 0.34 percentage point of GDP.
- Selection-correction factor (λ) is positive and significant, indicating selection into IMF programs matters and its correction was crucial.
- Interpretation and caveats:
  - IMF programs—and the associated conditionality—appear to reduce the magnitude of election-driven consumption increases.
  - Coefficients on interaction terms with IMF programs exhibit low significance; majority of elections in LICs occur while countries are already under IMF agreements, limiting statistical power.
  - Self-selection bias is only partially controlled for despite the improved selection model.

### Key policy-relevant findings and implications
- Election-year fiscal expansion in LICs:
  - Government consumption increases (0.8 percentage point of GDP baseline; 1 percentage point in some specifications).
  - Overall fiscal deficit increases by about 1 percentage point of GDP during election years.
- Post-election adjustment:
  - Implemented through increased trade tax mobilization (about 0.11 percentage point of GDP at election and maintained thereafter) and reductions in public investment (about 0.4 percentage point of GDP the year after).
  - Adjustment is incomplete: current spending envelopes remain elevated and revenue mobilization is insufficient to fully offset election-year deviations.
- Role of constraints:
  - National fiscal rules: evidence of dampening political budget cycles in government consumption but effect is weak in LICs (marginal effect in rule countries ~0.13 percentage point of GDP).
  - IMF programs: associated with a lower political budget cycle (deviation drops to 0.34 percentage point of GDP under program engagement), but significance is limited and selection issues remain.
- Growth implications:
  - Election-driven volatility and the post-election sacrifice of public investment, together with increased trade taxes, likely harm future economic growth prospects in LICs.

### Concluding remarks
- Elections in LICs are associated with discernible political budget cycles: higher current spending and larger deficits during election years; partial post-election adjustment through trade tax increases and investment cuts.
- National fiscal rules and IMF programs appear to mitigate election-driven fiscal deviations, but their dampening effects are limited in magnitude and statistical precision in the LIC context.
- Policy implication: strengthening fiscal institutions, enforcement of fiscal rules, and effective international oversight may help reduce election-driven fiscal volatility; incomplete post-election adjustments and reduced public investment raise concerns for long-term growth.

*Italic: Source — Excerpt from paper content unit titled "_wp13153 - References ............................................................................................................."*

### References

### _wp13153 - References

### Bibliographic references
- Alesina, A., Roubini, N., Cohen, G.D., 1997. Political cycles and the macroeconomy. the MIT Press.
- Arellano, M., Bond, S., 1991. Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations. Review of Economic Studies 58, 277–297.
- Arellano, M., Bover, O., 1995. Another look at the instrumental variable estimation of error-components models. Journal of Econometrics 68, 29–51.
- Block, S.A., 2002. Political business cycles, democratization, and economic reform: the case of Africa. Journal of Development Economics 67, 205–228.
- Blundell, R., Bond, S., 1998. Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics 87, 115–243.
- Brender, A., Drazen, A., 2005. Political budget cycles in new versus established democracies. Journal of Monetary Economics 52, 1271–1295.
- Brender, A., Drazen, A., 2008. How do budget deficits and economic growth affect reelection prospects? Evidence from a large panel of countries. The American Economic Review 98, 2203–2220.
- Combes, J.-L., Saadi-Sedik, T., 2006. How does trade openness influence budget deficits in developing countries? Journal of Development Studies 42, 1401–1416.
- Combes, J.-L., Ebeke, C., and Maurel, M. 2013. Remittances prior to an election. CERDI Etudes et Documents, 2013-07.
- Drazen, A., 2001. The political business cycle after 25 years, in: NBER Macroeconomics Annual 2000, Volume 15. MIT Press, pp. 75–138.
- Drazen, A., Eslava, M., 2010. Electoral manipulation via voter-friendly spending: Theory and evidence. Journal of Development Economics 92, 39–52.
- Dreher, A., 2004. The Influence of IMF Programs on the Re-election of Debtor Governments. Economics & Politics 16, 53–76.
- Dreher, A., and Vaubel, R., 2004. Do IMF and IBRD Cause Moral Hazard and Political Business Cycles? Evidence from Panel Data, Open Economies Review, 15, 5-22.
- Ehrhart, H., 2012. Elections and the structure of taxation in developing countries. Public Choice 1–17.
- Fatás, A., Mihov, I., 2003. The Case for Restricting Fiscal Policy Discretion. The Quarterly Journal of Economics 118, 1419–1447.
- Fatás, A., Mihov, I., 2012. Policy Volatility, Institutions and Economic Growth. Review of Economics and Statistics. doi: 10.1162/REST_a_00265
- Fouda, S.M., 1997. Political Monetary Cycles and Independence of the Central Bank in a Monetary Union: an Empirical Test for a BEAC Franc Zone Member Country. Journal of African Economies 6, 112–131.
- Faye, M., and Niehaus, P., 2012. Political Aid Cycles. American Economic Review, 107, 3516-3530.
- Hyde, S.D., Marinov, N., 2012. Which Elections Can Be Lost? Political Analysis 20, 191–210.
- Hyde, S.D., O’Mahony, A., 2010. International Scrutiny and Pre-Electoral Fiscal Manipulation in Developing Countries. The Journal of Politics 72, 690–704.
- Keen, M., Lockwood, B., 2010. The value added tax: Its causes and consequences. Journal of Development Economics 92, 138–151.
- Khemani, S., 2004. Political cycles in a developing economy: effect of elections in the Indian States. Journal of Development Economics 73, 125–154.
- Maddala, G.S., 1983. Limited-Dependent and Qualitative Variables in Econometrics, Econometric Society Monographs, Cambridge University Press.
- Nickell, S., 1981. Biases in Dynamic Models with Fixed Effects. Econometrica 49, 1417–1426.
- O’Mahony A., 2010. Engineering Good Times : Fiscal Manipulation in a Global Economy, British Journal of Political Science, 41, 315-340.
- Przeworski, A., Vreeland, J.R., 2000. The effect of IMF programs on economic growth. Journal of Development Economics 62, 385–421.
- Rodrik, D., 1998. Why Do More Open Economies Have Bigger Governments? Journal of Political Economy 106, 997–1032.
- Roodman, D., 2009. A Note on the Theme of Too Many Instruments. Oxford Bulletin of Economics and Statistics 71, 135–158.
- Rose, S., 2006. Do fiscal rules dampen the political business cycle? Public choice 128, 407–431.
- Schaechter, A., Kinda, T., Budina, N., and Weber, A., 2012. Fiscal Rules in Response to the Crisis - Toward the Next-Generation Rules. A New Dataset, IMF Working Papers 12/187, International Monetary Fund.
- Schuknecht, L., 2000. Fiscal Policy Cycles and Public Expenditure in Developing Countries. Public Choice 102, 113–128.
- Shi, M., Svensson, J., 2006. Political budget cycles: Do they differ across countries and why? Journal of Public Economics 90, 1367–1389.
- Stotsky, J., G., and WoldeMariam, A., 1997. Tax Effort in Sub-Saharan Africa, IMF Working Paper 97/107, International Monetary Fund.
- Vella, F., Verbeek, M., 1999. Estimating and Interpreting Models With Endogenous Treatment Effects. Journal of Business & Economic Statistics 17, 473–478.
- Vergne, C., 2009. Democracy, elections and allocation of public expenditures in developing countries. European Journal of Political Economy 25, 63–77.
- Windmeijer, F., 2005. A finite sample correction for the variance of linear efficient two-step GMM estimators. Journal of Econometrics 126, 25–50.

### Appendix A — Sample: countries and number of national elections (1990–2010)
- Afghanistan 2
- Armenia 5
- Bangladesh 3
- Benin 4
- Bolivia 6
- Burkina Faso 4
- Burundi 2
- Cambodia 3
- Cameroon 3
- Central African Rep. 4
- Chad 3
- Comoros 5
- Congo, Dem. Rep. 1
- Congo, Republic 3
- Cote d’Ivoire 4
- Djibouti 3
- Ethiopia 3
- Faye, M., and Niehaus, P., 2012. Political Aid Cycles. American Economic Review, 107, 3516-3530. (entry appears in original list)
- Gambia 4
- Georgia 5
- Ghana 5
- Guinea 4
- Guinea-Bissau 5
- Haiti 6
- Honduras 5
- Kenya 4
- Kyrgyz Rep. 5
- Lesotho 3
- Liberia 2
- Madagascar 5
- Malawi 4
- Mali 4
- Mauritania 5
- Moldova 8
- Mozambique 4
- Nepal 3
- Nicaragua 4
- Niger 4
- Nigeria 4
- Papua New Guinea 4
- Papua New Guinea appears in original list with 4 elections
- Papua New Guinea 4
- Papua New Guinea duplication noted in source listing
- Papua New Guinea duplicate retained as in source
- Papua New Guinea 4
- Papua New Guinea duplicate entries in original source omitted hereafter
- Papua New Guinea duplicates are present in source; the appendix list in the source includes a single entry per country.
- Rwanda 2
- Senegal 3
- Sierra Leone 3
- Solomon Islands 1
- Tajikistan 4
- Tanzania 5
- Timor-Leste 1
- Togo 4
- Uganda 3
- Uzbekistan 3
- Yemen 2
- Zambia 5

Notes: Only legislative elections for parliamentary systems and executive elections for presidential systems are included. Source: National Election Around the World (NELDA) dataset.

### Appendix B — Descriptive statistics (LICs sample, 1990–2010) — Table A2 (Obs, Mean, Std. Dev., Min, Max)
- Election dummy: 1330, 0.13, 0.34, 0, 1
- Government consumption ratio: 1145, 15.20, 7.52, 1.53, 62.17
- Public investment ratio: 1044, 7.37, 5.69, 0.08, 59.85
- Total tax revenue ratio: 832, 14.68, 6.51, 1.27, 58.11
- Taxes on goods and services ratio: 666, 5.47, 3.19, 0.04, 16.96
- Direct taxes ratio: 697, 4.31, 3.07, 0.01, 23.89
- Trade taxes ratio: 666, 3.88, 2.87, 0.00, 14.12
- Overall fiscal balance ratio: 1004, -2.48, 6.74, -72.35, 61.83
- Real per capit GDP growth: 1273, 1.23, 7.68, -71.24, 64.20
- Official Development Assistance ratio: 1276, 13.22, 11.74, -2.56, 146.89
- External PPG debt ratio: 1226, 85.47, 110.30, 0.58, 2394.86
- Trade openness ratio: 1179, 76.28, 37.18, 0.19, 213.22
- ln (100+Inflation): 1279, 5.40, 0.65, 4.61, 10.35
- National fiscal rule dummy: 1330, 0.02, 0.14, 0, 1
- ln (Total population): 1330, 15.16, 1.92, 11.13, 18.86
- Total natural resource rents ratio: 1330, 4.85, 12.29, 0, 105.73
- Reserve coverage (in month of imports): 1204, 3.50, 2.58, 0.00, 19.75
- Political globalization (from The KOF Institute): 1313, 46.40, 18.46, 6.59, 90.90
- Change in real percapita GDP growth: 1266, 0.00, 10.00, -60.55, 132.33

Note: All variables expressed as “ratios” denote nominal values normalized by nominal GDP of each country.

### Appendix C — Correlates of participation into Fund programs in LICs (Table A3)
Period: 1990–2010. Dependent variable: Fund program dummy. Models: LPM (country-fixed effects) and Probit.
- International reserve coverage (in months of imports), t-1:
  - LPM: 0.00496 [0.0152]
  - Probit: -0.0425** [0.0178]
- Fiscal balance-to-GDP, t-1:
  - LPM: 0.00286 [0.00349]
  - Probit: -0.00296 [0.00966]
- ln(100+Inflation), t-1:
  - LPM: -0.0494 [0.0867]
  - Probit: 0.00539 [0.0650]
- Official development assistance-to-GDP, t-1:
  - LPM: 0.00882*** [0.00261]
  - Probit: 0.0417*** [0.00547]
- ln(Population), t-1:
  - LPM: -0.0856 [0.313]
  - Probit: 0.146*** [0.0307]
- (Election*Competition), t+1:
  - LPM: 0.0408** [0.0198]
  - Probit: 0.163** [0.0665]
- Election dummy, t+1:
  - LPM: -0.244* [0.133]
  - Probit: -0.790* [0.410]
- Political globalization (KOF Institute index), t:
  - LPM: -0.00115 [0.00382]
  - Probit: 0.0204*** [0.00325]
- Change in real per capita GDP growth, t:
  - LPM: -0.00204** [0.000880]
  - Probit: -0.00654 [0.00499]

Model diagnostics and sample:
- Intercept: LPM 2.085 [4.483]; Probit -3.430*** [0.532]
- Country-fixed effects: LPM Yes; Probit No
- Observations: 916
- R-squared: LPM 0.043; Probit 0.177
- Number of countries: 63
- Note: Robust standard errors in brackets. ***, **, * denote p<0.01, p<0.05, p<0.1 respectively.

### Appendix D — Regression results: Political Budget Cycle estimates and robustness checks

Table 1. Estimates of the Political Budget Cycle across selected Fiscal Variables in LICs: 1990–2010.
- Dependent variables across columns: G (Government consumption ratio), I (Public investment ratio), T (Total tax revenue ratio), TGS (Tax revenues on goods and services ratio), TD (Tax revenues on income), TT (Trade tax revenues ratio), Bal (Overall fiscal balance ratio).
- Key coefficients (Election dummies and controls). Standard errors in brackets. Significance: * p < 0.10, ** p < 0.05, *** p < 0.01.

Election effects:
- Election t:
  - G: 0.841*** [0.258]
  - I: -0.194 [0.210]
  - T: 0.209 [0.152]
  - TGS: -0.075 [0.061]
  - TD: 0.102 [0.083]
  - TT: 0.112* [0.066]
  - Bal: -1.047* [0.542]
- Election t-1:
  - G: -0.059 [0.223]
  - I: -0.371** [0.159]
  - T: 0.340* [0.187]
  - TGS: 0.134 [0.084]
  - TD: 0.023 [0.084]
  - TT: 0.173*** [0.064]
  - Bal: 0.275 [0.330]
- Election t-2:
  - G: -0.049 [0.197]
  - I: -0.081 [0.208]
  - T: 0.290* [0.159]
  - TGS: 0.026 [0.101]
  - TD: -0.046 [0.061]
  - TT: 0.212* [0.114]
  - Bal: -0.262 [0.322]

Selected control coefficients (examples):
- Lagged dependent variable:
  - G: 0.730*** [0.103]
  - I: 0.813*** [0.089]
  - T: 0.893*** [0.168]
  - TGS: 1.029*** [0.057]
  - TD: 0.831*** [0.105]
  - TT: 0.976*** [0.064]
  - Bal: 0.260*** [0.094]
- Real per capita GDP growth:
  - G: -0.032 [0.034]
  - I: 0.027 [0.023]
  - T: 0.041** [0.020]
  - TGS: 0.033*** [0.006]
  - TD: -0.003 [0.010]
  - TT: 0.016*** [0.006]
  - Bal: 0.084*** [0.031]
- Official development assistance-to-GDP:
  - G: 0.073** [0.037]
  - I: 0.038** [0.015]
  - T: -0.001 [0.014]
  - TGS: -0.000 [0.003]
  - TD: -0.002 [0.003]
  - TT: 0.005 [0.003]
  - Bal: -0.019 [0.018]
- Trade openness:
  - G: 0.028** [0.012]
  - I: 0.011** [0.005]
  - T: 0.013 [0.012]
  - TGS: 0.000 [0.002]
  - TD: 0.008* [0.004]
  - TT: 0.002 [0.001]
  - Bal: 0.014 [0.010]
- Fiscal rule dummy t-1:
  - G: 1.625*** [0.470]
  - I: 0.345 [0.327]
  - T: 0.179 [0.541]
  - TGS: -0.036 [0.170]
  - TD: 0.281 [0.206]
  - TT: -0.232** [0.093]
  - Bal: -0.400 [1.603]

Model diagnostics and sample sizes:
- N (observations): G 1234; I 1140; T 815; TGS 679; TD 705; TT 679; Bal 970
- No of countries: G 60; I 57; T 56; TGS 52; TD 53; TT 52; Bal 61
- m1:p-value: 0.004, 0.001, 0.003, 0.000, 0.022, 0.000, 0.017
- m2:p-value: 0.516, 0.416, 0.530, 0.350, 0.294, 0.104, 0.260
- Hansen OID: p-value: 0.829, 0.060, 0.031, 0.839, 0.249, 0.540, 0.114
- No of instruments: 17, 17, 21, 21, 21, 25, 23

Estimation note: All equations are estimated using the two-step System-GMM with Windmeijer (2005) correction of standard errors.

Table 2. Addressing the endogeneity of the election timing in LICs. 1990–2010.
- Predetermined election coefficients (t, t-1, t-2) for same dependent variables (G, I, T, TGS, TD, TT, Bal). Standard errors in brackets.

Predetermined election t:
- G: 0.758** [0.327]
- I: -0.231 [0.183]
- T: 0.095 [0.167]
- TGS: -0.074 [0.089]
- TD: 0.147 [0.116]
- TT: 0.098* [0.055]
- Bal: -1.278** [0.618]

Predetermined election t-1:
- G: 0.224 [0.305]
- I: -0.398** [0.188]
- T: 0.131 [0.255]
- TGS: 0.053 [0.108]
- TD: 0.093 [0.086]
- TT: 0.112* [0.058]
- Bal: 0.253 [0.438]

Predetermined election t-2:
- G: -0.031 [0.218]
- I: -0.159 [0.308]
- T: 0.076 [0.148]
- TGS: -0.020 [0.108]
- TD: -0.057 [0.096]
- TT: 0.171* [0.094]
- Bal: -0.466* [0.278]

Model diagnostics and sample sizes:
- N: G 1131; I 1043; T 758; TGS 633; TD 657; TT 631; Bal 900
- No of countries: 60, 57, 56, 52, 53, 52, 61
- m1:p-value: 0.007, 0.002, 0.004, 0.000, 0.009, 0.001, 0.019
- m2:p-value: 0.653, 0.390, 0.236, 0.269, 0.250, 0.089, 0.210
- Hansen OID: p-value: 0.583, 0.022, 0.031, 0.864, 0.238, 0.705, 0.217
- No of instruments: 20, 20, 24, 24, 24, 28, 20

Estimation note: Two-step System-GMM with Windmeijer (2005) correction. Predetermined election dummies identify elections held on the date fixed by an established constitution or procedure.

Table 3. Do fiscal rules and IMF programs dampen the Political Budget Cycle in LICs?
- Dependent variable: G (Government consumption ratio). Two columns with interaction terms.

Column (1):
- Election dummy: 1.009*** [0.331]
- Election*Lagged fiscal rule dummy: -0.870* [0.511]
- Lagged fiscal rule dummy: 1.986*** [0.662]
- IMF program dummy: -0.712 [0.444]
-  (Predicted selection correction factor): 0.621** [0.253]
- Lagged dependent variable: 0.695*** [0.089]
- Real GDP growth: -0.017 [0.041]
- Official development assistance-to-GDP: 0.073* [0.037]
- External debt-to-GDP: -0.003 [0.003]
- Trade openness: 0.032*** [0.011]
- ln (100+Inflation): -1.189 [0.976]
- Intercept: 7.809 [5.321]
- N: 1234
- No of countries: 60
- Joint significance of election coefficients: P-value 0.007
- m1:p-value 0.002; m2:p-value 0.341; Hansen OID: p-value 0.808; No of instruments 17

Column (2) (controls for self-selection into IMF programs via two-step approach):
- Election dummy: 0.963*** [0.293]
- Election*IMF program dummy: -0.623* [0.342]
- Lagged fiscal rule dummy: 1.621*** [0.376]
- IMF program dummy: -0.712 [0.444]
-  (Predicted selection correction factor): 0.621** [0.253] (included)
- Lagged dependent variable: 0.837*** [0.041]
- Real GDP growth: 0.004 [0.014]
- Official development assistance-to-GDP: 0.045*** [0.017]
- External debt-to-GDP: -0.001 [0.003]
- Trade openness: 0.015*** [0.005]
- ln (100+Inflation): -0.442 [0.344]
- Intercept: 3.488* [1.896]
- N: 864
- No of countries: 59
- Joint significance of election coefficients: P-value 0.002
- m1:p-value 0.019; m2:p-value 0.114; Hansen OID: p-value 0.132; No of instruments 18

Estimation note: All equations estimated using two-step System-GMM with Windmeijer (2005) correction. Column 2 controls for self-selection bias into IMF programs using a two-step approach a la Maddala (1983), Vella and Verbeek (1999), and Keen and Lockwood (2010).

*Source: _wp13153 - References (PDF content supplied).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13153.pdf_