## _wp10192

## Source details

**Canonical URL:** [_wp10192](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10192.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10192.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10192.pdf.json)

---

### I. Introduction — context and purpose
- Recent boom and bust in commodity prices (rising from 2004 to almost the end of 2008 and reversing from September 2008) raised concerns about impacts on Latin American fiscal positions.
- From 2004 to almost the end of 2008, rising commodity prices boosted government revenues and likely spurred FDI and overall economic activity in many Latin American (LA) countries.
- Some governments saved a large proportion of buoyant revenues and accumulated financial assets; others used windfalls to expand government spending.
- Episodes of commodity price booms and busts generate volatility in fiscal revenues; some governments increase expenditures more than proportionally during booms, making later adjustment difficult.
- Prior literature emphasized fiscal procyclicality with respect to the output cycle; this paper focuses directly on commodity price cycles and their dynamic impact on revenues, expenditures, and GDP.

### II. Methodology
- Estimation method: Vector Autoregression (VAR) with identification via Cholesky decomposition (following Sims (1980)).
- Cholesky ordering: Country-Specific Commodity Price Index (CSCPI) first, GDP second, government expenditures third, and government revenues fourth.
- Data transformation and frequency:
  - Variables expressed as logs of real levels; model estimated in first differences because unit root tests indicate stationarity in first differences.
  - Quarterly data; number of lags selected by the Hannan-Quinn information criterion (1 to 6 lags considered).
  - Series seasonally adjusted using X-12 where required.
- Impulse responses and variance decomposition used to trace propagation and quantify contribution of commodity price shocks.
- CSCPIs:
  - Constructed from IMF monthly commodity price indices for 55 commodities and country commodity export shares from the Standard International Trade Classification (SITC) in WITS.
  - Weights held constant within each year to match monthly price frequency with annual trade data.
  - Weights reflect, on average, 80 to 90 percent of a given country’s total commodity exports.
- Additional notes:
  - All series are in real terms; nominal series deflated (CPI for fiscal series; GDP deflator for GDP).
  - Johansen cointegration tests did not suggest cointegration vectors; results robust to alternatives (HP filter, linear quadratic detrending).

### III. Data, variables, and sample
- Sample: quarterly database for 12 commodity-exporting countries; period varies by country, beginning as early as 1975Q1 and ending 2008Q4.
- Country coverage:
  - Latin American sample (8 countries): Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Peru, and Venezuela.
  - High-income commodity-exporting comparison group (4 countries): Australia, Canada, New Zealand, and Norway.
- Key variable definitions:
  - Revenues: total revenues (aggregation depends on data availability).
  - Primary expenditures: total expenditures minus interest payments (aggregation depends on country definitions and availability).
  - Real GDP: obtained by Haver Analytics using primary data from each country’s authorities; nominal data deflated by each country’s CPI when real data were not available.
  - CPI: obtained by Haver Analytics using primary data from each country’s authorities.
- Country-specific aggregation notes (selection):
  - Argentina: author’s estimations from nonfinancial public sector data published by the Finance Ministry.
  - Australia: central government aggregation data from the Reserve Bank of Australia.
  - Brazil: central bank database at the central government level (including federal government, central bank, and social security administration).
  - Chile: Finance Ministry at the central government level.
  - Colombia: nonfinancial public sector from the Finance Ministry’s CONFIS.
  - Ecuador: nonfinancial public sector from Ecuador’s central bank.
  - Mexico: central government data from the Finance and Public Credit Department.
  - New Zealand: government consumption used as proxy for quarterly primary expenditures; revenue series not available quarterly.
  - Norway: central government level (including the national insurance scheme) from Statistics Norway.
  - Peru: central government level from Peru’s central bank.
  - Venezuela: central government level from Venezuela’s central bank.
- Sample periods (selected, exactly as reported):
  - Argentina: 1993Q1–2008Q4
  - Brazil: 1993Q1–2008Q4
  - Chile: 1995Q1–2008Q4
  - Colombia: 1995Q1–2008Q4
  - Ecuador: 1998Q1–2008Q4
  - Mexico: 1993Q1–2008Q4
  - Peru: 1995Q1–2008Q4
  - Venezuela: 1998Q1–2008Q4
  - Australia (AUT): 1975Q1–2008Q4
  - Canada: 1985Q2–2008Q4
  - New Zealand (NZL)1/: 1988Q1–2008Q4 (1/ Primary expenditure data not available at quarterly frequency; government consumption used as proxy. Revenue series not available quarterly.)
  - Norway: 1991Q1–2008Q4

### IV. Results — dynamic responses to commodity price shocks
- Size of the one-standard-deviation commodity price shock:
  - On average in Latin American countries: around 13 percent.
  - Venezuela: 16.4 percent (largest).
  - Brazil: 7.8 percent (smallest).
  - High-income commodity exporters: Norway’s on-impact response reaches almost 10 percent.
- Commodity composition note:
  - Argentina mainly focuses on soy and beef.
  - Other LA and high-income exporters heavily rely on metals, minerals, and oil.
- Impulse response findings (accumulated responses to a Cholesky one-standard-deviation commodity price shock in a quarter):
  - Government real revenues:
    - Range from around 2 percent in Brazil and Canada to 10 percent in Venezuela and around 14 percent in Ecuador.
    - Peak revenue response typically occurs two quarters after the shock.
  - Real GDP:
    - Range from around 1 percent in Australia to 2.5 percent in Argentina and 3 percent in Norway.
    - Peak GDP response typically occurs two quarters after the shock.
  - Primary expenditures:
    - Highly heterogeneous across LA countries.
    - Range from 0 percent in Chile to an increase of almost 12 percent in Venezuela.
    - Median expenditure response: around 4 percent.
    - Ecuador’s expenditure response: less than 8 percent.
    - High-income countries:
      - Australia and New Zealand show no reaction of primary expenditures to the shock.
      - Canada and Norway show on-impact decreases of -0.2 percent and -1.5 percent, respectively.
- Country-specific impulse responses (selected entries from Table 5; quarterly percentage change):
  - Argentina — Primary Expenditures: Impact Response 0.0; Peak Response 2.0; Quarter of the Peak Response 4th. Revenues: Impact Response 1.3; Peak Response 2.9; Quarter of the Peak Response 2nd. GDP: Impact Response 0.3; Peak Response 0.5; Quarter of the Peak Response 2nd.
  - Brazil — Primary Expenditures: Impact Response 0.0; Peak Response 0.7; Quarter of the Peak Response 2nd. Revenues: Impact Response 0.8; Peak Response 0.3; Quarter of the Peak Response 1st. GDP: Impact Response 0.8; Peak Response 0.8; Quarter of the Peak Response 1st.
  - Chile — Primary Expenditures: Impact Response 0.0; Peak Response 0.0; Quarter of the Peak Response -. Revenues: Impact Response 1.3; Peak Response 2.1; Quarter of the Peak Response 2nd. GDP: Impact Response 0.5; Peak Response 0.5; Quarter of the Peak Response 1st.
  - Colombia — Primary Expenditures: Impact Response 2.0; Peak Response 3.0; Quarter of the Peak Response 2nd. Revenues: Impact Response 0.0; Peak Response 1.7; Quarter of the Peak Response 4th. GDP: Impact Response 0.4; Peak Response 0.5; Quarter of the Peak Response 2nd.
  - Ecuador — Primary Expenditures: Impact Response 3.0; Peak Response 4.8; Quarter of the Peak Response 2nd. Revenues: Impact Response 6.4; Peak Response 7.1; Quarter of the Peak Response 2nd. GDP: Impact Response 0.5; Peak Response 0.5; Quarter of the Peak Response 1st.
  - Mexico — Primary Expenditures: Impact Response 2.1; Peak Response 2.1; Quarter of the Peak Response 1st. Revenues: Impact Response 0.8; Peak Response 2.0; Quarter of the Peak Response 2nd. GDP: Impact Response 0.3; Peak Response 0.3; Quarter of the Peak Response 1st.
  - Peru — Primary Expenditures: Impact Response 0.7; Peak Response 1.0; Quarter of the Peak Response 3rd. Revenues: Impact Response 2.0; Peak Response 2.0; Quarter of the Peak Response 1st. GDP: Impact Response 0.1; Peak Response 0.7; Quarter of the Peak Response 2nd.
  - Venezuela — Primary Expenditures: Impact Response 3.0; Peak Response 5.3; Quarter of the Peak Response 2nd. Revenues: Impact Response 2.6; Peak Response 6.5; Quarter of the Peak Response 2nd. GDP: Impact Response 0.0; Peak Response 1.2; Quarter of the Peak Response 2nd.
  - Australia — Primary Expenditures: Impact Response 0.0; Peak Response -0.6; Quarter of the Peak Response 3rd. Revenues: Impact Response 1.0; Peak Response 1.0; Quarter of the Peak Response 1st. GDP: Impact Response 0.2; Peak Response 0.2; Quarter of the Peak Response 1st.
  - Canada — Primary Expenditures: Impact Response 0.0; Peak Response 0.6; Quarter of the Peak Response 3rd. Revenues: Impact Response 0.0; Peak Response 0.5; Quarter of the Peak Response 3rd. GDP: Impact Response 0.6; Peak Response 0.6; Quarter of the Peak Response 1st.
  - New Zealand — Primary Expenditures: Impact Response 0.0; Peak Response 0.1; Quarter of the Peak Response 2nd. Revenues: Impact Response -; Peak Response -; Quarter of the Peak Response -. GDP: Impact Response 0.0; Peak Response 0.1; Quarter of the Peak Response 2nd.
  - Norway — Primary Expenditures: Impact Response -0.4; Peak Response -0.9; Quarter of the Peak Response 2nd. Revenues: Impact Response 1.8; Peak Response 1.8; Quarter of the Peak Response 1st. GDP: Impact Response 1.5; Peak Response 1.5; Quarter of the Peak Response 1st.
- Variance decomposition (contribution of commodity price shocks to forecast error variance of government expenditures at a 10-quarter horizon):
  - Venezuela: almost 17 percent (highest).
  - Chile: less than 4 percent (lowest among LA; less than one-fourth of Venezuela’s figure).
  - Other six Latin American countries (excluding Venezuela and Chile): average contribution of 9.5 percent at a 10-quarter horizon.
  - High-income commodity-exporting countries: average contribution of 3.9 percent at a 10-quarter horizon.

### V. Fiscal rules, institutions, and cross-country heterogeneity
- Heterogeneous institutional frameworks can explain variation in expenditure responses across countries.
- Evidence and country examples:
  - Corbacho and Schwartz (2007) find fiscal performance did not improve after fiscal responsibility laws in Argentina (1999), Colombia (2003), and Peru (1999) because numerical targets were breached and credibility undermined.
  - Chile: better fundamentals, stronger institutions, and fiscal rules encouraging public saving may have helped avoid procyclical responses; Chile’s behavior resembles high-income commodity-exporting countries.
  - Australia: diversification of revenue sources and a large industrial and service sector reduced volatility from commodity price shocks (Bhattacharyya and Williamson (2009)).
  - Canada: strong fiscal record due to a proven budgetary framework, transparent public finances, and a de facto fiscal rule of “budget balance” between 1998 and 2006; the new government in 2006 targeted a moderate surplus of 0.25 percent of GDP.
  - New Zealand: by the early 1990s policy advice shifted toward fiscal consolidation and a medium-term focus, with institutional framework alterations to support these objectives.
- Norway (fiscal framework specifics):
  - Non-oil structural budget deficit target: 4 percent.
  - Government Petroleum Fund established: 1990.
  - Fund renamed: “Government Pension Fund Global” since 2006.
  - Purpose: preserve assets for future generations and avoid potential crowding-out effects (Dutch disease).

### VI. Main findings and interpretation
- Motivation: the recent boom and bust in commodity prices raised concerns about commodity-price impacts on Latin American fiscal positions; the most recent shock was unprecedentedly large.
- Methodology: dynamic effects of commodity price shocks estimated for Latin American commodity-exporting countries and compared with high-income commodity-exporting countries using a unique quarterly data set and novel CSCPIs.
- Key empirical results:
  - The impact of commodity price shocks on fiscal revenues is not that different between Latin American commodity-exporting countries and high-income commodity-exporting countries.
  - There is a spectrum of primary expenditures responses within Latin American countries; Chile and Venezuela are at extremes of the distribution.
  - Chile behaves similarly to high-income countries, with no positive expenditure response to higher commodity prices.
- Interpretation:
  - A potential explanation for Chile’s behavior: efficient application of fiscal rules, accompanied by strong institutions, political commitment, and high standards of transparency.

### VII. Policy recommendations and suggested further research
- Suggested next research steps:
  - Look for structural breaks in countries that adopted fiscal rules during the last decade to detect changes in fiscal behavior and establish empirical links with such rules.
- Implicit policy considerations drawn from findings:
  - Fiscal rules, credible institutions, and saving of commodity windfalls can potentially limit procyclical expenditure responses.
  - Diversification of revenue sources and stronger budgetary frameworks appear associated with smaller expenditure responses to commodity price shocks.

*Italicized source attribution: IMF working paper content unit _wp10192 (excerpt).*

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

### _wp10192 - References

### References
- References ................................................................................................................................12

### Appendix 1. Data
- Appendix 1. Data .....................................................................................................................14

### Tables
- 1. The Importance of Commodity Exports ..............................................................................15
- 2. Composition of Country-Specific Commodity Price Indices ..............................................16
- 3. Commodity Price Shocks as a Source of Fluctuations in Primary Expenditures ................17
- 4. Sample Period by Country ...................................................................................................18
- 5. Responses to Commodity Price Shocks ...............................................................................19

### Figures
- 1. Country-Specific Commodity Price Indices ........................................................................20
- 2. Commodity-Related Fiscal Revenues ..................................................................................21
- 3. Contribution of Commodity Price Shocks to the Variance of Primary Expenditure  
  Growth ...............................................................................................................................22
- 4. Response of Commodity Prices to One Standard Cholesky Innovation in .........................23
- 5. Accumulated Response of Commodity Prices to One Standard Cholesky ..........................24
- 6. Accumulated Response of Primary Expenditures to One Standard Cholesky ....................25
- 7. Accumulated Response of Total Revenues to One Standard Cholesky ..............................26

*Source: _wp10192 - References (PDF).*

### 8. Accumulated Response of GDP to Commodity Price One Standard Cholesky ..................27

### 8. Accumulated Response of GDP to Commodity Price One Standard Cholesky ..................27

### I. Introduction — context and purpose
- Recent boom and bust in commodity prices (rising from 2004 to almost the end of 2008 and reversing from September 2008) raised concerns about impacts on Latin American fiscal positions.
- From 2004 to almost the end of 2008, rising commodity prices boosted government revenues and likely spurred FDI and overall economic activity in many Latin American (LA) countries.
- Some governments saved a large proportion of buoyant revenues and accumulated financial assets; others used windfalls to expand government spending.
- Episodes of commodity price booms and busts generate volatility in fiscal revenues; some governments increase expenditures more than proportionally during booms, making later adjustment difficult.
- Prior literature has emphasized fiscal procyclicality with respect to the output cycle; this paper focuses directly on commodity price cycles and their dynamic impact on revenues, expenditures, and GDP.

### II. Methodology
#### Vector Autoregression approach
- Estimation method: Vector Autoregression (VAR) with identification via Cholesky decomposition (following Sims (1980)).
- Ordering used in Cholesky decomposition: Country-Specific Commodity Price Index (CSCPI) first, GDP second, government expenditures third, and government revenues fourth.
- Variables expressed as logs of real levels and the model is estimated in first differences because unit root tests indicate stationarity in first differences.
- Number of lags considered: 1 to 6, selected by the Hannan-Quinn information criterion.
- Impulse responses and variance decomposition analyses are used to trace propagation and quantify the contribution of commodity price shocks.

#### Data
- Sample: novel quarterly database for 12 commodity-exporting countries; period varies by country, beginning as early as the first quarter of 1975 and ending in the last quarter of 2008.
- Country coverage:
  - Latin American sample (8 countries): Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Peru, and Venezuela.
  - High-income commodity-exporting comparison group (4 countries): Australia, Canada, New Zealand, and Norway.
- Variables of interest: government primary expenditures, revenues, real GDP, and Country-Specific Commodity Price Indices (CSCPIs).
- CSCPIs combine IMF monthly commodity price indices for 55 commodities and country commodity export shares from the Standard International Trade Classification (SITC) in WITS.
- Weights in CSCPIs reflect, on average, 80 to 90 percent of a given country’s total commodity exports and are held constant within each year to match monthly price frequency with annual trade data.
- All series are in real terms; nominal series are deflated (CPI for fiscal series; GDP deflator for GDP). Series are in logs and seasonally adjusted using X-12 where required.
- Johansen cointegration tests did not suggest cointegration vectors; series are included in the VAR in first differences. Results are robust to alternatives (HP filter, linear quadratic detrending).

### III. Results
#### A. Dynamic responses of fiscal variables to commodity price shocks
- Size of the one-standard-deviation commodity price shock:
  - On average in Latin American countries: around 13 percent.
  - Venezuela: 16.4 percent (largest).
  - Brazil: 7.8 percent (smallest).
  - High-income commodity exporters: Norway’s on-impact response reaches almost 10 percent.

- Commodity composition note:
  - Argentina mainly focuses on soy and beef.
  - Other LA and high-income exporters heavily rely on metals, minerals, and oil.

- Impulse response findings (accumulated responses to a Cholesky one-standard-deviation commodity price shock in a quarter):
  - Government real revenues:
    - Range from around 2 percent in Brazil and Canada to 10 percent in Venezuela and around 14 percent in Ecuador.
    - Peak revenue response typically occurs two quarters after the shock.
  - Real GDP:
    - Range from around 1 percent in Australia to 2.5 percent in Argentina and 3 percent in Norway.
    - Peak GDP response typically occurs two quarters after the shock.
  - Primary expenditures:
    - Highly heterogeneous across LA countries.
    - Range from 0 percent in Chile to an increase of almost 12 percent in Venezuela.
    - Median expenditure response: around 4 percent.
    - Ecuador’s expenditure response: less than 8 percent.
    - High-income countries: Australia and New Zealand show no reaction of primary expenditures to the shock; Canada and Norway show on-impact decreases of -0.2 percent and -1.5 percent, respectively (suggestive of countercyclical fiscal policy).
  - Overall pattern:
    - Revenues and GDP react positively to commodity price shocks across both LA and high-income countries.
    - Primary expenditures display a wide dispersion; Chile resembles high-income countries (limited expenditure response), while Venezuela is an outlier with disproportionate expenditure increases.

- Variance decomposition (contribution of commodity price shocks to forecast error variance of government expenditures at a 10-quarter horizon):
  - Venezuela: almost 17 percent (highest).
  - Chile: less than 4 percent (lowest among LA; less than one-fourth of Venezuela’s figure).
  - Other six Latin American countries (excluding Venezuela and Chile): average contribution of 9.5 percent at a 10-quarter horizon.
  - High-income commodity-exporting countries: average contribution of 3.9 percent at a 10-quarter horizon.

#### B. Fiscal rules and institutional explanations
- Heterogeneous institutional frameworks can explain variation in expenditure responses across countries.
- Evidence and examples:
  - Corbacho and Schwartz (2007) suggest fiscal performance did not improve after fiscal responsibility laws in Argentina (1999), Colombia (2003), and Peru (1999) because numerical targets were breached and credibility undermined.
  - Chile: better fundamentals, stronger institutions, and fiscal rules encouraging public saving may have helped avoid procyclical responses; Chile’s behavior resembles that of high-income commodity-exporting countries.
  - Australia: diversification of revenue sources and a large industrial and service sector reduced volatility from commodity price shocks (Bhattacharyya and Williamson (2009)).
  - Canada: strong fiscal record due to a proven budgetary framework, transparent public finances, and a de facto fiscal rule of “budget balance” between 1998 and 2006; the new government in 2006 targeted a moderate surplus of 0.25 percent of GDP.
  - New Zealand: by the early 1990s policy advice shifted toward fiscal consolidation and a medium-term focus, with institutional framework alterations to support these objectives.

### IV. Key methodological and data notes (concise)
- Identification: Cholesky ordering assumes countries are price takers; CSCPI → GDP → expenditures → revenues.
- Frequency and transformations: quarterly data; logs of real levels; first-differenced to achieve stationarity.
- CSCPIs constructed from IMF monthly commodity prices (55 commodities) and WITS SITC trade shares; weights held constant within each year; weights cover on average 80 to 90 percent of commodity exports for each country.
- No robust evidence of cointegration among system variables under the model considered.

*Source: _wp10192 - 8. Accumulated Response of GDP to Commodity Price One Standard Cholesky ..................27*

### introduction of the country’s Fiscal Responsibility Act (1994), which is based on “principles”

### _wp10192 - introduction of the country’s Fiscal Responsibility Act (1994), which is based on “principles”

### Fiscal frameworks and examples
- Norway
  - A key fiscal rule sets the non-oil structural budget deficit of the central government at 4 percent (allowing for temporary deviations over the business cycle).
  - To manage oil wealth, a Government Petroleum Fund was established in 1990. It receives most of Norway’s petroleum revenue and invests it in financial assets abroad.
  - Purpose of the fund: preserve assets for future generations and avoid potential crowding-out effects (Dutch disease) from rapid spending of oil wealth.
  - Note: Known as the “Government Pension Fund Global” since 2006.

### Main findings (conclusion)
- Motivation
  - The recent boom and bust in commodity prices raised concerns about commodity-price impacts on Latin American fiscal positions; the  most recent shock was unprecedentedly large, motivating detailed study.
- Methodology
  - The paper estimates dynamic effects of commodity price shocks for a group of Latin American commodity-exporting countries and compares VAR impulse response functions with those of high-income commodity-exporting countries.
  - Uses a unique quarterly data set including a novel country-specific commodity price index.
- Key empirical results
  - The impact of commodity price shocks on fiscal revenues is not that different between Latin American commodity-exporting countries and high-income commodity-exporting countries.
  - There is a spectrum of primary expenditures responses within Latin American countries; Chile and Venezuela are at extremes of the distribution.
  - Chile behaves similarly to high-income countries, with no positive expenditure response to higher commodity prices.
- Interpretation and suggested next research steps
  - A potential explanation for Chile’s behavior: efficient application of fiscal rules, accompanied by strong institutions, political commitment, and high standards of transparency.
  - Recommendation for further research: look for structural breaks in countries that adopted fiscal rules during the last decade to detect changes in fiscal behavior and establish empirical links with such rules.

### Data, variables, and sample
- Time span
  - The data go as far back as the first quarter of 1975.
- Country-Specific Commodity Price Indices
  - Author’s estimations based on the IMF monthly commodity price indices and trade data from the Standard International Trade Classification (World Integrated Trade Solution).
- Real GDP
  - Obtained by Haver Analytics using primary data from each country’s authorities; nominal data deflated by each country’s CPI when real data were not available.
- CPI
  - Obtained by Haver Analytics using primary data from each country’s authorities.
- Revenues and Primary Expenditures
  - Revenues: total revenues (aggregation depends on data availability).
  - Primary expenditures: total expenditures minus interest payments (aggregation depends on country definitions and availability).
- Country-specific aggregation notes (selection)
  - Argentina: author’s estimations from nonfinancial public sector data published by the Finance Ministry.
  - Australia: central government aggregation data from the Reserve Bank of Australia.
  - Brazil: central bank database at the central government level (including federal government, central bank, and social security administration).
  - Chile: Finance Ministry at the central government level.
  - Colombia: nonfinancial public sector from the Finance Ministry’s CONFIS.
  - Ecuador: nonfinancial public sector from Ecuador’s central bank.
  - Mexico: central government data from the Finance and Public Credit Department.
  - New Zealand: government consumption (from national accounts) used as proxy for quarterly primary expenditures; revenue series not available quarterly.
  - Norway: central government level (including the national insurance scheme) from Statistics Norway.
  - Peru: central government level from Peru’s central bank.
  - Venezuela: central government level from Venezuela’s central bank.

### Selected key statistics and results
- Fiscal rule (Norway)
  - Non-oil structural budget deficit target: 4 percent.
  - Government Petroleum Fund established: 1990.
  - Fund renamed: “Government Pension Fund Global” since 2006.
- Table 1 (importance of commodity exports, averages 1999-2006; selected entries)
  - Argentina (ARG): As a Share of Total Merchandise Exports 55.6; As a Share of GDP 11.2
  - Brazil (BRA): As a Share of Total Merchandise Exports 41.0; As a Share of GDP 5.4
  - Chile (CHL): As a Share of Total Merchandise Exports 65.4; As a Share of GDP 24.1
  - Ecuador (ECU): As a Share of Total Merchandise Exports 75.7; As a Share of GDP 23.2
  - Mexico (MEX): As a Share of Total Merchandise Exports 17.5; As a Share of GDP 5.3
  - Peru (PER): As a Share of Total Merchandise Exports 53.5; As a Share of GDP 11.4
  - Venezuela (VEN): As a Share of Total Merchandise Exports 80.9; As a Share of GDP 25.6
  - Australia (AUT): As a Share of Total Merchandise Exports 49.6; As a Share of GDP 9.8
  - Canada (CAN): As a Share of Total Merchandise Exports 30.0; As a Share of GDP 12.3
  - New Zealand (NZL): As a Share of Total Merchandise Exports 48.6; As a Share of GDP 15.2
  - Norway (NOR): As a Share of Total Merchandise Exports 56.3; As a Share of GDP 24.4
- Table 5 (Responses to one-standard-deviation commodity price shock, quarterly percentage change; selected entries)
  - Argentina — Primary Expenditures: Impact Response 0.0; Peak Response 2.0; Quarter of the Peak Response 4th. Revenues: Impact Response 1.3; Peak Response 2.9; Quarter of the Peak Response 2nd. GDP: Impact Response 0.3; Peak Response 0.5; Quarter of the Peak Response 2nd.
  - Brazil — Primary Expenditures: Impact Response 0.0; Peak Response 0.7; Quarter of the Peak Response 2nd. Revenues: Impact Response 0.8; Peak Response 0.3; Quarter of the Peak Response 1st. GDP: Impact Response 0.8; Peak Response 0.8; Quarter of the Peak Response 1st.
  - Chile — Primary Expenditures: Impact Response 0.0; Peak Response 0.0; Quarter of the Peak Response -. Revenues: Impact Response 1.3; Peak Response 2.1; Quarter of the Peak Response 2nd. GDP: Impact Response 0.5; Peak Response 0.5; Quarter of the Peak Response 1st.
  - Colombia — Primary Expenditures: Impact Response 2.0; Peak Response 3.0; Quarter of the Peak Response 2nd. Revenues: Impact Response 0.0; Peak Response 1.7; Quarter of the Peak Response 4th. GDP: Impact Response 0.4; Peak Response 0.5; Quarter of the Peak Response 2nd.
  - Ecuador — Primary Expenditures: Impact Response 3.0; Peak Response 4.8; Quarter of the Peak Response 2nd. Revenues: Impact Response 6.4; Peak Response 7.1; Quarter of the Peak Response 2nd. GDP: Impact Response 0.5; Peak Response 0.5; Quarter of the Peak Response 1st.
  - Mexico — Primary Expenditures: Impact Response 2.1; Peak Response 2.1; Quarter of the Peak Response 1st. Revenues: Impact Response 0.8; Peak Response 2.0; Quarter of the Peak Response 2nd. GDP: Impact Response 0.3; Peak Response 0.3; Quarter of the Peak Response 1st.
  - Peru — Primary Expenditures: Impact Response 0.7; Peak Response 1.0; Quarter of the Peak Response 3rd. Revenues: Impact Response 2.0; Peak Response 2.0; Quarter of the Peak Response 1st. GDP: Impact Response 0.1; Peak Response 0.7; Quarter of the Peak Response 2nd.
  - Venezuela — Primary Expenditures: Impact Response 3.0; Peak Response 5.3; Quarter of the Peak Response 2nd. Revenues: Impact Response 2.6; Peak Response 6.5; Quarter of the Peak Response 2nd. GDP: Impact Response 0.0; Peak Response 1.2; Quarter of the Peak Response 2nd.
  - Australia — Primary Expenditures: Impact Response 0.0; Peak Response -0.6; Quarter of the Peak Response 3rd. Revenues: Impact Response 1.0; Peak Response 1.0; Quarter of the Peak Response 1st. GDP: Impact Response 0.2; Peak Response 0.2; Quarter of the Peak Response 1st.
  - Canada — Primary Expenditures: Impact Response 0.0; Peak Response 0.6; Quarter of the Peak Response 3rd. Revenues: Impact Response 0.0; Peak Response 0.5; Quarter of the Peak Response 3rd. GDP: Impact Response 0.6; Peak Response 0.6; Quarter of the Peak Response 1st.
  - New Zealand — Primary Expenditures: Impact Response 0.0; Peak Response 0.1; Quarter of the Peak Response 2nd. Revenues: Impact Response -; Peak Response -; Quarter of the Peak Response -. GDP: Impact Response 0.0; Peak Response 0.1; Quarter of the Peak Response 2nd.
  - Norway — Primary Expenditures: Impact Response -0.4; Peak Response -0.9; Quarter of the Peak Response 2nd. Revenues: Impact Response 1.8; Peak Response 1.8; Quarter of the Peak Response 1st. GDP: Impact Response 1.5; Peak Response 1.5; Quarter of the Peak Response 1st.
- Table 3 (Contribution of commodity price shocks to variance of primary expenditure growth; selected entries)
  - Quarters after the shock (4 quarters shown as rows): Venezuela (VENCOLARGECUMEXBRAPERCHL): 1 quarter after: 15.7 36.8 40.0 24.1 91 0.5 72.6 11.5 60.6  (table formatting in source spans many cells).
  - Additional panel for CANNZLNORAUT (selected): 1 quarter after: 10.3 80.0 0.0 0.8 60.6
  - (Source table includes multi-quarter contributions at 1, 2, 4, 8, 10 quarters after the shock for many countries.)
- Sample periods for series (Table 4)
  - Argentina: 1993Q1–2008Q4
  - Brazil: 1993Q1–2008Q4
  - Chile: 1995Q1–2008Q4
  - Colombia: 1995Q1–2008Q4
  - Ecuador: 1998Q1–2008Q4
  - Mexico: 1993Q1–2008Q4
  - Peru: 1995Q1–2008Q4
  - Venezuela: 1998Q1–2008Q4
  - Austria (AUT): 1975Q1–2008Q4
  - Canada: 1985Q2–2008Q4
  - New Zealand (NZL)1/: 1988Q1–2008Q4 (1/ Primary expenditure data not available at quarterly frequency; government consumption used as proxy. Revenue series not available quarterly.)
  - Norway: 1991Q1–2008Q4

*Italicized source attribution: IMF working paper content unit _wp10192 (excerpt).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10192.pdf_