## _wp11171

## Source details

**Canonical URL:** [_wp11171](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11171.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11171.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11171.pdf.json)

---

### Scope and variables analyzed
- Examines five fiscal variables—non-oil revenue; the non-oil primary balance; and total expenditure and its components, consumption and capital expenditure— for the full sample and subgroups divided by their development levels and by correcting the endogeneity bias between the fiscal variables and the output variable.
- Emphasizes dividing the OPCs into groups because OPCs are not a homogenous group and group-specific patterns can inform effective fiscal policy design.
- Notes the importance of examining aggregate fiscal variables and their subcomponents separately, since subcomponents may move in offsetting ways.

### Key empirical findings on cyclicality
- Total expenditure is highly procyclical in:
  - the full sample,
  - the low-income group,
  - the middle-income group.
- The low-income group constitutes a large share of the full sample, therefore weighing heavily in the results of the full sample.
- Total expenditure is countercyclical in the high-income countries — possibly due to their greater accumulation of financial assets, which eases their financial constraints when funds are needed.
- Although expenditure is countercyclical for the high-income group, its components move in different directions:
  - consumption is procyclical,
  - capital expenditure is countercyclical.

### Institutional and financing determinants
- Political and institutional factors, as well as financing constraints, play a role in the cyclicality of fiscal policies in the OPCs.
- For the low-income group:
  - Most of the variables on the quality of institutions and the political structure appear to be significant.
  - Domestic financing constraints seem to matter.
- For the middle-income group:
  - Two institutional variables are significant: the composite institution index and checks and balances.
  - Fiscal policy is affected more by the external financing constraint.
- For the high-income group:
  - None of the institutional variables turns out to be significant.
  - Fiscal policy is affected more by the external financing constraint.
- Possible explanation: middle- and high-income groups may be more integrated into the global financial system than the low-income countries.

### Policy implications and recommendations
- All OPCs face volatile and unpredictable oil revenues, making fiscal management challenging.
- Imperative for OPCs to formulate effective countercyclical fiscal policies to:
  - smooth government expenditure,
  - decouple expenditure from volatile oil revenues,
  - prevent boom-and-bust cycles,
  - sustain long-term growth,
  - keep the safety net that the poor need.
- Sound fiscal policies and discipline require:
  - strong institutions,
  - a higher-level bureaucracy,
  - more transparency.
- Strong institutions and transparency can help reduce the "voracity effect," facilitate the accumulation of financial assets, and build investor confidence to raise funds when needed.

### Suggested avenues for future research
- "It would be useful to look at the institutions and political structure in the high-income group in detail in future research." 21

### Data and variable definitions (selected)
- Independent variable
  - Real non-oil GDP growth — Source: WEO — Description: Growth in nominal GDP deflated using the CPI
- Dependent variables — Source: Villafuerte and Lopez-Murphy (2010) and WEO
  - Real total government spending — Growth in nominal GDP deflated using the CPI
  - Real government consumption
  - Real capital spending
  - Real non-oil primary balance
  - Real non-oil revenue
- Financial constraints variables
  - Domestic: Real central bank interest rate — WEO/IFS
  - Domestic: Private credit to GDP — WDI
  - External: Net foreign capital flows — WEO
  - External: Debt-GDP ratio — WDI
  - Inflation — WEO
- Political and Institutional Variables (ICRG and Polity4)
  - Bureaucracy Quality — ICRG (Rating 0 to 4: low rating, low bureaucracy quality)
  - Composite Risk Rating — ICRG (Rating 0 to 100; 0 is high risk, 100 is low risk)
  - Corruption — ICRG (Rating 0 to 6; 0 is high, 6 is low corruption)
  - Law & Order — ICRG (Rating 1 to 3; low rating, low law obedience)
  - Democracy — Polity4 database, polity2 variable (Difference between a democracy index (0 to 10) and an autocracy index (0 to 10))
  - Constraints on the executive — Polity4 database, xconst variable (Extent of institutionalized constraints on the decision making powers of chief)
  - Political competition — Polity4 database, Polcom variable (Degree of institutionalization of political competition combined with the extent of government restriction on political competition, from 1 to 10)
- Other Control Variables
  - Oil revenue as share of total revenue — WEO

### Descriptive statistics (selected entries)
- Real Non-oil GDP Growth
  - Mean: Low to middle income 3.8, Upper-middle income 4.3, High income 7.2, All countries 5.0
  - Median: Low to middle income 5.5, Upper-middle income 4.5, High income 3.9, All countries 4.8
  - St. Dev.: Low to middle income 12.9, Upper-middle income 9.0, High income 12.0, All countries 12.0
  - Observation: 241, 89, 155, 485
- Real Government Spending Growth
  - Mean: 5.0, 5.5, 6.4, 5.6
  - Median: 6.2, 5.2, 6.1, 6.0
  - St. Dev.: 23.2, 17.1, 25.5, 23.0
  - Observation: 244, 98, 173, 515
- Real Government Consumption Growth
  - Mean: 4.0, 5.9, 5.6, 4.9
  - Median: 7.2, 5.9, 5.1, 6.2
  - St. Dev.: 23.3, 16.4, 13.6, 19.3
  - Observation: 238, 92, 162, 492
- Real Government Investment Growth
  - Mean: 9.3, 9.3, 8.9, 9.2
  - Median: 7.1, 9.3, 4.1, 7.1
  - St. Dev.: 42.2, 33.5, 29.5, 36.8
  - Observation: 227, 88, 153, 468
- Real Non-oil Primary Balance Growth
  - Mean: -0.5, -18.3, 45.2, 10.9
  - Median: 5.6, 0.1, 6.7, 5.8
  - St. Dev.: 216.7, 194.0, 637.9, 402.2
  - Observation: 244, 100, 165, 509
- Real Non-oil Revenue Growth
  - Mean: 4.5, 4.1, 4.7, 4.5
  - Median: 6.2, 5.5, 4.3, 5.6
  - St. Dev.: 24.1, 22.4, 34.8, 27.7
  - Observation: 238, 97, 160, 495
- Countries: 146, 92, 9 (as displayed)

### Correlations (selected)
- Low income (sample correlations with fiscal aggregates shown)
  - Gross international reserve correlation with Total expenditure 0.414, Consumption 0.414, Capital expenditure 0.351, Non-oil revenue 0.356, Non-oil primary balance -0.374
  - Share of oil revenue in total revenue correlation with Total expenditure -0.330, Consumption -0.326, Capital expenditure -0.274, Non-oil revenue -0.417, Non-oil primary balance -0.018
  - GDP per capita correlation with Total expenditure 0.955, Consumption 0.943, Capital expenditure 0.903, Non-oil revenue 0.902, Non-oil primary balance -0.684
- Middle income (selected)
  - Inflation correlation with Total expenditure 0.443, Consumption 0.392, Capital expenditure 0.574, Non-oil revenue 0.548, Non-oil primary balance -0.227
  - Oil wealth correlation with Total expenditure 0.555, Consumption 0.593, Capital expenditure 0.430, Non-oil revenue 0.551, Non-oil primary balance -0.576
- High income (selected)
  - Gross international reserve correlation with Total expenditure 0.827, Consumption 0.788, Capital expenditure 0.932, Non-oil revenue 0.832, Non-oil primary balance -0.837
  - Oil wealth correlation with Total expenditure 0.845, Consumption 0.819, Capital expenditure 0.883, Non-oil revenue 0.859, Non-oil primary balance -0.839

### Estimation results (selected coefficients and statistics)
- Pooled OLS (Table A4, expenditure as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.78***  (0.08); Constant 0.02**  (0.01); Observations 477; R-squared 0.17
  - Low income: Δ(log(non-oil GDP)) 1.18***  (0.09); Observations 233; R-squared 0.43
  - Middle income: Δ(log(non-oil GDP)) 1.24***  (0.16); Observations 89; R-squared 0.41
  - High income: Δ(log(non-oil GDP)) -0.06  (0.17); Observations 155; R-squared 0.00
- Pooled OLS (Table A5, consumption as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.85***  (0.06); Observations 472; R-squared 0.29
  - High income: Δ(log(non-oil GDP)) 0.35***  (0.09); Observations 155; R-squared 0.10
- Pooled OLS (Table A6, non-oil revenue as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.82***  (0.09); Observations 466; R-squared 0.14
  - High income: Δ(log(non-oil GDP)) 0.51**  (0.22); Observations 150; R-squared 0.04
- Pooled OLS (Table A7, capital expenditure as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.89***  (0.13); Observations 465; R-squared 0.09
  - Middle income: Δ(log(non-oil GDP)) 2.03***  (0.35); Observations 86; R-squared 0.29
- Pooled OLS (Table A8, non-oil primary balance as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) -0.24  (1.55); Observations 477; R-squared 0.00
  - Low income: Δ(log(non-oil GDP)) 1.46  (1.11)
  - High income: Δ(log(non-oil GDP)) -4.80  (4.42)
- Fixed Effects (Table A9, expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.55***  (0.17); Δlog(TOT) 0.31***  (0.10); Observations 460; R-squared 0.16; No of countries 29
  - Low income: Δ(log(non-oil GDP)) 1.10***  (0.16); R-squared 0.38; No of countries 14
- Fixed Effects (Table A10, consumption as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.67***  (0.13); Δlog(TOT) 0.15***  (0.05); Observations 452; R-squared 0.24
- Fixed Effects (Table A11, non-oil revenue as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.88***  (0.13); Δ(log(Revenue(t-1))) -0.26***  (0.08); Observations 448; R-squared 0.18
- Fixed Effects (Table A12, capital expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 1.08***  (0.24); Δ(log(Capital Exp.(t-1))) -0.18***  (0.07); Observations 438; R-squared 0.13
- System GMM (Table A19, expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.64***  (0.05); Δ(log(Tot. Expend(t-1))) 0.05***  (0.01); Δ(log(TOT)) 0.24***  (0.01); Observations 444; No of countries 28; AR1 0.0128; AR2 0.151; Hansen test-p 1.000
- System GMM (Table A20, consumption as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.91***  (0.05); Δ(log(Consump(t-1))) -0.11***  (0.01); Δ(log(TOT)) 0.12***  (0.01); Observations 436; No of countries 28; AR1 0.000267; AR2 0.822; Hansen test-p 1.000
- System GMM (Table A21, non-oil revenue as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 1.00***  (0.02); Δ(log(Revenue(t-1))) -0.20***  (0.01); Δ(log(TOT)) 0.05***  (0.01); Observations 432; No of countries 28; AR1 0.00495; AR2 0.0988; Hansen test-p 1.000
- System GMM (Table A22, capital expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 1.61***  (0.16); Δ(log(Capital Exp.(t-1))) -0.15***  (0.02); Δ(log(TOT)) 0.11***  (0.02); Observations 422; No of countries 28; AR1 0.000508; AR2 0.347; Hansen test-p 1.000
- System GMM (Table A23, non-oil primary balance as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) -1.17***  (0.11); Δ(log(Primary Bal.(t-1))) -0.04***  (0.00); Δ(log(TOT)) 1.76***  (0.01); Observations 444; No of countries 28; AR1 0.237; AR2 0.758; Hansen test-p 1.000

### Key empirical patterns
- Positive and statistically significant elasticities of fiscal expenditures, consumption, non-oil revenue, and capital expenditure with respect to Δ(log(non-oil GDP)) in the full sample across OLS, Fixed Effects, 2SLS and System GMM specifications (typical full-sample point estimates range from 0.55 to 1.61 depending on the fiscal aggregate and estimator).
- Non-oil primary balance shows mixed responses: pooled OLS and fixed-effects estimates are often statistically insignificant or imprecise, while System GMM reports a negative and statistically significant coefficient for the full sample: Δ(log(non-oil GDP)) -1.17***  (0.11).
- Lagged dependent variables often enter with negative and statistically significant coefficients for consumption and capital expenditure (e.g., Δ(log(Consump(t-1))) -0.11***  (0.01) in System GMM; Δ(log(Capital Exp.(t-1))) -0.15***  (0.02) in System GMM), indicating dynamic adjustment.

*Source: _wp11171 - REFERENCES (IMF PDF content unit)*

### 2009. It examines five fiscal variables—

### 2009. It examines five fiscal variables—

### Scope and variables analyzed
- Examines five fiscal variables—non-oil revenue; the non-oil primary balance; and total  expenditure and its components, consumption and capital expenditure— for the full sample and subgroups divided by their development levels and by correcting the endogeneity bias between the fiscal variables and the output variable.
- Emphasizes dividing the OPCs into groups because OPCs are not a homogenous group and group-specific patterns can inform effective fiscal policy design.
- Notes the importance of examining aggregate fiscal variables and their subcomponents separately, since subcomponents may move in offsetting ways.

### Key empirical findings on cyclicality
- Total expenditure is highly procyclical in:
  - the full sample,
  - the low-income group,
  - the middle-income group.
- The low-income group constitutes a large share of the full sample, therefore weighing heavily in the results of the full sample.
- Total expenditure is countercyclical in the high-income countries — possibly due to their greater accumulation of financial assets, which eases their financial constraints when funds are needed.
- Although expenditure is countercyclical for the high-income group, its components move in different directions:
  - consumption is procyclical,
  - capital expenditure is countercyclical.

### Institutional and financing determinants
- Political and institutional factors, as well as financing constraints, play a role in the cyclicality of fiscal policies in the OPCs.
- For the low-income group:
  - Most of the variables on the quality of institutions and the political structure appear to be significant.
  - Domestic financing constraints seem to matter.
- For the middle-income group:
  - Two institutional variables are significant: the composite institution index and checks and balances.
  - Fiscal policy is affected more by the external financing constraint.
- For the high-income group:
  - None of the institutional variables turns out to be significant.
  - Fiscal policy is affected more by the external financing constraint.
- Possible explanation: middle- and high-income groups may be more integrated into the global financial system than the low-income countries.

### Policy implications and recommendations
- All OPCs face volatile and unpredictable oil revenues, making fiscal management challenging.
- Imperative for OPCs to formulate effective countercyclical fiscal policies to:
  - smooth government expenditure,
  - decouple expenditure from volatile oil revenues,
  - prevent boom-and-bust cycles,
  - sustain long-term growth,
  - keep the safety net that the poor need.
- Sound fiscal policies and discipline require:
  - strong institutions,
  - a higher-level bureaucracy,
  - more transparency.
- Strong institutions and transparency can help reduce the "voracity effect," facilitate the accumulation of financial assets, and build investor confidence to raise funds when needed.

### Suggested avenues for future research
- "It would be useful to look at the institutions and political structure in the high-income group in detail in future research." 21

*Source: _wp11171 - 2009.*

### REFERENCES

### _wp11171 - REFERENCES

### References (selected)
- Abdih, Yasser, P. Lopez-Murphy, A. Roitman, and R. Sahay, 2010, ―The Cyclicality of Fiscal Policy in the Middle East and Central Asia:Is the Current Crisis Different?‖ IMF Working Papers WP/10/68
- Alesina, Alberto and G. Tabellini, 2005, ―Why is Fiscal Policy Often Procyclical?‖ NBER Working Papers No 11600
- Arellano, Manuel and S. Bond, 1991. ―Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,‖ Review of Economic Studies, vol. 58(2), pages 277–97, April
- Arezki, Rabah and K. Ismail, 2010, ―Boom-Bust Cycle, Asymmetrical Fiscal Response and the Dutch Disease,‖ IMF Working Papers 10/94
- Baldini, Alfredo, 2005, ―Fiscal Policy and Business Cycles in an Oil-Producing Economy: The Case of Venezuela,‖ IMF Working Papers 05/237
- Barro, Robert, 1973, ―The Control of Politicians: An Economic Model,‖ Public Choice, 13, 19–42
- Calvo, Guillermo, A., and C. M. Reinhart, 2000, ―Fear of Floating,‖ NBER Working Papers No 7993
- Ilzetzki, Ethan and Carlos A. Vegh, 2008. ―Procyclical Fiscal Policy in Developing Countries: Truth or Fiction?,‖ NBER Working Papers 14191
- Kaminsky, Graciela, C. Reinhart, and C. Vegh, 2004, ―When it Rains it Pours: Procyclical Capital Flows and Macroeconomic Policies,‖ NBER Working Papers No 10780
- Reinhart, Carmen M., K. S. Rogoff and M. A. Savastano, 2003, ―Debt Intolerance,‖ Brookings Papers on Economic Activity, Vol. 1, 1–74
- Villafuerte, Mauricio and P. Lopez-Murphy, 2010, ―Fiscal Policy in Oil Producing Countries During the Recent Oil Price Cycle,‖ IMF Working Papers 10/28

### Data and variable definitions (Table A1)
- Independent variable
  - Real non-oil GDP growth — Source: WEO — Description: Growth in nominal GDP deflated using the CPI
- Dependent variables — Source: Villafuerte and Lopez-Murphy (2010) and WEO
  - Real total government spending — Growth in nominal GDP deflated using the CPI
  - Real government consumption
  - Real capital spending
  - Real non-oil primary balance
  - Real non-oil revenue
- Financial constraints variables
  - Domestic: Real central bank interest rate — WEO/IFS
  - Domestic: Private credit to GDP — WDI
  - External: Net foreign capital flows — WEO
  - External: Debt-GDP ratio — WDI
  - Inflation — WEO
- Political and Institutional Variables (ICRG and Polity4)
  - Bureaucracy Quality — ICRG (Rating 0 to 4: low rating, low bureaucracy quality)
  - Composite Risk Rating — ICRG (Rating 0 to 100; 0 is high risk, 100 is low risk)
  - Corruption — ICRG (Rating 0 to 6; 0 is high, 6 is low corruption)
  - Law & Order — ICRG (Rating 1 to 3; low rating, low law obedience)
  - Democracy — Polity4 database, polity2 variable (Difference between a democracy index (0 to 10) and an autocracy index (0 to 10))
  - Constraints on the executive — Polity4 database, xconst variable (Extent of institutionalized constraints on the decision making powers of chief)
  - Political competition — Polity4 database, Polcom variable (Degree of institutionalization of political competition combined with the extent of government restriction on political competition, from 1 to 10)
- Other Control Variables
  - Oil revenue as share of total revenue — WEO

### Descriptive statistics (Table A2 — selected entries)
- Real Non-oil GDP Growth
  - Mean: Low to middle income 3.8, Upper-middle income 4.3, High income 7.2, All countries 5.0
  - Median: Low to middle income 5.5, Upper-middle income 4.5, High income 3.9, All countries 4.8
  - St. Dev.: Low to middle income 12.9, Upper-middle income 9.0, High income 12.0, All countries 12.0
  - Observation: 241, 89, 155, 485
- Real Government Spending Growth
  - Mean: 5.0, 5.5, 6.4, 5.6
  - Median: 6.2, 5.2, 6.1, 6.0
  - St. Dev.: 23.2, 17.1, 25.5, 23.0
  - Observation: 244, 98, 173, 515
- Real Government Consumption Growth
  - Mean: 4.0, 5.9, 5.6, 4.9
  - Median: 7.2, 5.9, 5.1, 6.2
  - St. Dev.: 23.3, 16.4, 13.6, 19.3
  - Observation: 238, 92, 162, 492
- Real Government Investment Growth
  - Mean: 9.3, 9.3, 8.9, 9.2
  - Median: 7.1, 9.3, 4.1, 7.1
  - St. Dev.: 42.2, 33.5, 29.5, 36.8
  - Observation: 227, 88, 153, 468
- Real Non-oil Primary Balance Growth
  - Mean: -0.5, -18.3, 45.2, 10.9
  - Median: 5.6, 0.1, 6.7, 5.8
  - St. Dev.: 216.7, 194.0, 637.9, 402.2
  - Observation: 244, 100, 165, 509
- Real Non-oil Revenue Growth
  - Mean: 4.5, 4.1, 4.7, 4.5
  - Median: 6.2, 5.5, 4.3, 5.6
  - St. Dev.: 24.1, 22.4, 34.8, 27.7
  - Observation: 238, 97, 160, 495
- Countries: 146, 92, 9 (as displayed)

### Correlations (Table A3 — selected correlations)
- Low income (sample correlations with fiscal aggregates shown)
  - Gross international reserve correlation with Total expenditure 0.414, Consumption 0.414, Capital expenditure 0.351, Non-oil revenue 0.356, Non-oil primary balance -0.374
  - Share of oil revenue in total revenue correlation with Total expenditure -0.330, Consumption -0.326, Capital expenditure -0.274, Non-oil revenue -0.417, Non-oil primary balance -0.018
  - GDP per capita correlation with Total expenditure 0.955, Consumption 0.943, Capital expenditure 0.903, Non-oil revenue 0.902, Non-oil primary balance -0.684
- Middle income (selected)
  - Inflation correlation with Total expenditure 0.443, Consumption 0.392, Capital expenditure 0.574, Non-oil revenue 0.548, Non-oil primary balance -0.227
  - Oil wealth correlation with Total expenditure 0.555, Consumption 0.593, Capital expenditure 0.430, Non-oil revenue 0.551, Non-oil primary balance -0.576
- High income (selected)
  - Gross international reserve correlation with Total expenditure 0.827, Consumption 0.788, Capital expenditure 0.932, Non-oil revenue 0.832, Non-oil primary balance -0.837
  - Oil wealth correlation with Total expenditure 0.845, Consumption 0.819, Capital expenditure 0.883, Non-oil revenue 0.859, Non-oil primary balance -0.839

### Estimation results — pooled OLS, fixed effects, 2SLS, and System GMM (selected coefficients and statistics)
- Pooled OLS (Table A4, expenditure as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.78***  (0.08); Constant 0.02**  (0.01); Observations 477; R-squared 0.17
  - Low income: Δ(log(non-oil GDP)) 1.18***  (0.09); Observations 233; R-squared 0.43
  - Middle income: Δ(log(non-oil GDP)) 1.24***  (0.16); Observations 89; R-squared 0.41
  - High income: Δ(log(non-oil GDP)) -0.06  (0.17); Observations 155; R-squared 0.00
- Pooled OLS (Table A5, consumption as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.85***  (0.06); Observations 472; R-squared 0.29
  - High income: Δ(log(non-oil GDP)) 0.35***  (0.09); Observations 155; R-squared 0.10
- Pooled OLS (Table A6, non-oil revenue as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.82***  (0.09); Observations 466; R-squared 0.14
  - High income: Δ(log(non-oil GDP)) 0.51**  (0.22); Observations 150; R-squared 0.04
- Pooled OLS (Table A7, capital expenditure as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) 0.89***  (0.13); Observations 465; R-squared 0.09
  - Middle income: Δ(log(non-oil GDP)) 2.03***  (0.35); Observations 86; R-squared 0.29
- Pooled OLS (Table A8, non-oil primary balance as dependent variable, Base Regression)
  - Full sample: Δ(log(non-oil GDP)) -0.24  (1.55); Observations 477; R-squared 0.00
  - Low income: Δ(log(non-oil GDP)) 1.46  (1.11)
  - High income: Δ(log(non-oil GDP)) -4.80  (4.42)
- Fixed Effects (Table A9, expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.55***  (0.17); Δlog(TOT) 0.31***  (0.10); Observations 460; R-squared 0.16; No of countries 29
  - Low income: Δ(log(non-oil GDP)) 1.10***  (0.16); R-squared 0.38; No of countries 14
- Fixed Effects (Table A10, consumption as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.67***  (0.13); Δlog(TOT) 0.15***  (0.05); Observations 452; R-squared 0.24
- Fixed Effects (Table A11, non-oil revenue as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.88***  (0.13); Δ(log(Revenue(t-1))) -0.26***  (0.08); Observations 448; R-squared 0.18
- Fixed Effects (Table A12, capital expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 1.08***  (0.24); Δ(log(Capital Exp.(t-1))) -0.18***  (0.07); Observations 438; R-squared 0.13
- System GMM (Table A19, expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.64***  (0.05); Δ(log(Tot. Expend(t-1))) 0.05***  (0.01); Δ(log(TOT)) 0.24***  (0.01); Observations 444; No of countries 28; AR1 0.0128; AR2 0.151; Hansen test-p 1.000
- System GMM (Table A20, consumption as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 0.91***  (0.05); Δ(log(Consump(t-1))) -0.11***  (0.01); Δ(log(TOT)) 0.12***  (0.01); Observations 436; No of countries 28; AR1 0.000267; AR2 0.822; Hansen test-p 1.000
- System GMM (Table A21, non-oil revenue as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 1.00***  (0.02); Δ(log(Revenue(t-1))) -0.20***  (0.01); Δ(log(TOT)) 0.05***  (0.01); Observations 432; No of countries 28; AR1 0.00495; AR2 0.0988; Hansen test-p 1.000
- System GMM (Table A22, capital expenditure as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) 1.61***  (0.16); Δ(log(Capital Exp.(t-1))) -0.15***  (0.02); Δ(log(TOT)) 0.11***  (0.02); Observations 422; No of countries 28; AR1 0.000508; AR2 0.347; Hansen test-p 1.000
- System GMM (Table A23, non-oil primary balance as dependent variable)
  - Full sample: Δ(log(non-oil GDP)) -1.17***  (0.11); Δ(log(Primary Bal.(t-1))) -0.04***  (0.00); Δ(log(TOT)) 1.76***  (0.01); Observations 444; No of countries 28; AR1 0.237; AR2 0.758; Hansen test-p 1.000

### Key empirical patterns (as reported)
- Positive and statistically significant elasticities of fiscal expenditures, consumption, non-oil revenue, and capital expenditure with respect to Δ(log(non-oil GDP)) in the full sample across OLS, Fixed Effects, 2SLS and System GMM specifications (typical full-sample point estimates range from 0.55 to 1.61 depending on the fiscal aggregate and estimator).
- Non-oil primary balance shows mixed responses: pooled OLS and fixed-effects estimates are often statistically insignificant or imprecise, while System GMM reports a negative and statistically significant coefficient for the full sample: Δ(log(non-oil GDP)) -1.17***  (0.11).
- Lagged dependent variables often enter with negative and statistically significant coefficients for consumption and capital expenditure (e.g., Δ(log(Consump(t-1))) -0.11***  (0.01) in System GMM; Δ(log(Capital Exp.(t-1))) -0.15***  (0.02) in System GMM), indicating dynamic adjustment.

*Italic: Source — _wp11171 - REFERENCES (IMF PDF content unit)_*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11171.pdf_