## _wp16254

## Source details

**Canonical URL:** [_wp16254](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16254.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16254.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16254.pdf.json)

---

### I. Introduction — scope and dataset
- Research questions:
  - What is the impact of crude oil price changes on retail fuel prices?
  - How fast do crude oil price shocks feed through to domestic fuel prices?
  - Are responses to price declines symmetric to responses to price increases?
- Dataset compiled:
  - Monthly retail prices for four petroleum products: gasoline, diesel, kerosene, and LPG.
  - Coverage: 162 countries, monthly data for most countries starting from the early 2000s until December 2014; for some countries data go as far back as the 1970s.
  - Total of more than 65,000 monthly observations across products.
- Contributions:
  - Confirms static pass-through findings from prior literature.
  - Estimates impulse response functions of domestic retail gasoline prices using high-frequency data and Jorda (2005) local projections.
  - Differentiates positive versus negative crude oil price shocks to test asymmetry.
  - Demonstrates importance of within-year price changes for fiscal cost estimates of incomplete pass-through.

### II. The new global database on retail fuel prices — sources and methodology
- Primary sources:
  - Publicly posted data from country authorities (regulatory agencies, ministries, state-owned petroleum companies/refineries, national statistics agencies, central banks).
  - IMF staff-collected data for countries not publishing prices online (~40 percent of countries post prices online; IMF staff account for close to 40 percent of the sample; existing databases ~30 percent).
  - Remaining data from press coverage.
- Sample composition (based on 2014 monthly data):
  - Sub-Saharan Africa: 27 percent of sample
  - Advanced Economies: 21 percent
  - Developing Asia: ~14 percent
  - Latin America and the Caribbean: ~14 percent
  - Middle East, North Africa, and Pakistan (MENA): ~14 percent
  - Emerging Europe, Commonwealth of Independent States: remainder
- Data construction and adjustments:
  - End-user, tax-inclusive prices in local currency units (LCU): gasoline, diesel, kerosene in LCU per liter; LPG in LCU per kg (motor LPG per liter).
  - No extrapolation/interpolation; averages used to convert daily/weekly to monthly series.
  - When governments control prices and multiple changes occur in a month, month averages used, except price changes in last week are reported in following month.
  - Averaged across grades when multiple grades exist to limit breaks.
  - Harmonized units: 1 US gallon = 3.78 liters; 1 Imperial gallon = 4.54 liters. LPG unit price uses most common household canister ~12kg or 25 lbs.
  - Adjusted for currency redenomination and currency changes; fuels priced in USD converted at prevailing official exchange rate.
  - Where national series unavailable, capital city averages reported.
- Limitations noted:
  - LPG data coverage limited (12 percent of observations).
  - Kerosene coverage limited (19 percent of observations).
  - Database excludes discounts/rebates, and fuel sold below pump price for specific consumers; shortages and smuggling not captured when official data absent.

### III. Stylized facts on price levels and trends (January–December, 2014 averages; USD per liter except LPG in USD per kg)
- World average:
  - Gasoline 1.3; Diesel 1.2; Kerosene 0.9; LPG 1.2
- By region:
  - Advanced Economies: Gasoline 1.9; Diesel 1.7; Kerosene 1.5; LPG ..
  - Emerging Europe: Gasoline 1.8; Diesel 1.8; Kerosene ..; LPG ..
  - Developing Asia: Gasoline 1.2; Diesel 1.1; Kerosene 1.0; LPG 1.5
  - Latin America and the Caribbean: Gasoline 1.1; Diesel 1.0; Kerosene 1.0; LPG 1.1
  - Middle East and North Africa: Gasoline 0.7; Diesel 0.6; Kerosene 0.5; LPG 0.2
  - Sub-Saharan Africa: Gasoline 1.4; Diesel 1.3; Kerosene 1.1; LPG 1.4
  - United States (memo): Gasoline 0.9; Diesel 1.0
- Key observations:
  - Advanced economies and emerging Europe have highest average retail fuel prices; United States is an outlier with lowest gasoline and diesel among advanced economies.
  - MENA countries exhibit lowest retail prices—major oil producers heavily subsidize domestic prices.
  - Sub-Saharan Africa has the highest average fuel prices among developing countries (more than twice MENA), reflecting high supply, transport, storage, and refinery costs.
  - Gasoline usually more expensive than diesel; differentials largely driven by taxation and policy objectives.
  - Wide cross-border price disparities fuel smuggling and complicate subsidy reform.

### IV. Static pass-through analysis — median coefficients by region and period
- Pass-through coefficient definition (expressed in percent):
  - PT_i = 100 * (P_t^i * e_t - P_{t-1}^i * e_{t-1}) / (P_t^* - P_{t-1}^*)
- Five international oil price subperiods (2005–2014):
  - Jan 2005–Dec 2006: gradual increase to peak July 2006.
  - Jan 2007–Jun 2008: dramatic rise, peak 2008 crisis.
  - Jul 2008–Feb 2009: sharp decline at global financial crisis onset.
  - Mar 2009–Jun 2014: rebound then relative stability.
  - Jul 2014–Dec 2014: oil prices declined by 40 percent.
- Median pass-through highlights (medians reported in source Table 2):
  - All countries gasoline average (Jan 2005–Dec 2014 average): 101
  - Advanced Economies gasoline median by subperiod: 138; 159; 146; 168; 121; Average 146
  - MENA gasoline median by subperiod: ..; 41; 1; 16; 0; Average 15
  - Sub-Saharan Africa gasoline median average: 94 (subperiods: 172; 91; 82; 85; 39)
- Interpretations:
  - Pass-through tends to increase with income level and is highest in liberalized-price regions.
  - In many developing countries gasoline pass-through > diesel > kerosene.
  - Pass-through during falling price spells (e.g., Jul 2014–Dec 2014) is generally lower than during rising price spells, indicating asymmetry.
  - Low pass-through in Jul–Dec 2014 compared to Jul 2008–Feb 2009 suggests governments may use declines to rebuild fiscal buffers or recoup revenue lost when prices were high.

### V. Dynamic analysis — impulse response functions (local projections)
- Empirical approach:
  - Univariate model of retail gasoline price and crude oil price (both in USD).
  - Impulse responses estimated via Jorda (2005) local projections.
  - Baseline horizons h = 0,..., H with H = 23 (24 months); p = 12 lags of retail and crude prices.
  - 95 percent confidence intervals via nonparametric block bootstrap with 1000 replications.
  - Sample period for dynamics: January 2000–December 2014.
- Baseline dynamic result (average country):
  - A one cent per liter increase in crude oil prices translates into a 1.2 cent increase in gasoline prices per liter at peak (a pass-through of 120 percent), six months after the shock.
  - Speed: one-to-one parity reached within first three months.
  - Decay: impact declines to less than 0.5 cent by the 10th month and remains persistent in medium term.
- Heterogeneity across country groups:
  - Advanced economies: peak pass-through up to 200 percent within six months; medium-term drop to about 50 percent.
  - Emerging Europe and Sub-Saharan Africa: more gradual adjustment, lower peak than advanced economies.
  - MENA, Developing Asia, Latin America: muted and slow responses; MENA peak barely above 50 percent.
  - Persistence of shock driven mainly by advanced economies and Sub-Saharan Africa.

### VI. Asymmetry of pass-through (positive vs negative price shocks)
- Method:
  - Split crude price changes into positive and negative parts (Crudet+ and Crudet−); estimate impulse responses to each separately.
  - Interested coefficient: π1 = α1 − β1 (difference between responses to positive and negative shocks).
- Aggregate results:
  - Evidence of asymmetry: retail gasoline prices respond faster and more strongly to positive oil price shocks than to negative shocks.
  - Peak difference at about 8 months: pass-through for positive shocks ~10 percentage points larger than for negative shocks.
  - Gap narrows as positive-shock pass-through phases out more rapidly.
- Disaggregated group results:
  - Advanced economies: peak asymmetry ~20 percentage points (positive > negative).
  - Low-income countries: peak asymmetry ~10 percentage points.
  - Emerging economies: peak asymmetry ~5 percentage points; asymmetric effect dies out within first seven months.
  - MENA: asymmetry nearly inexistent.
  - Latin America: magnitude comparable to Developing Asia but more persistent.
  - Sub-Saharan Africa: larger and more persistent asymmetry than other developing groups.
- Possible explanations:
  - Market structures: oligopolistic behavior, lack of competition, search costs.
  - Inventory/production adjustment lags.
  - Government policy: lower pass-through of price declines may be used opportunistically to raise net fuel taxes or rebuild fiscal buffers after oil price declines.

### VII. Fiscal cost of incomplete pass-through — end-of-year vs yearly average prices
- Method for subsidy cost estimation:
  - Benchmark price for net oil importers: international USD product price adjusted by $0.20 per liter (shipping and distribution/retailing cost).
  - Benchmark price for net oil exporters: international USD price.
  - Petroleum consumption from IEA; nominal GDP from International Financial Statistics.
  - Fiscal cost compared using yearly average retail prices vs retail prices as of December each year (2005–13).
- Main findings:
  - End-of-year retail prices often underestimate the fiscal cost of incomplete pass-through relative to estimates using yearly average prices.
  - Peak underestimation averaged 1.4 percent of GDP in 2008.
  - Diesel is the largest contributor to the discrepancy.
  - Underestimation is larger in years with large and volatile international oil price increases (correlation coefficient 0.30).
  - Regional disparities: underestimation in 2008 largest in MENA.
  - In rising-price years (e.g., 2007) end-of-year prices may overestimate fiscal cost relative to yearly average.
- Policy implication:
  - High-frequency (intra-year) retail price data are critical to accurately measure fiscal costs of fuel subsidies and to track subsidy reform progress.

### VIII. Conclusion — key implications
- Dataset value:
  - New monthly retail fuel price database (162 countries, four products) enables dynamic analysis of pass-through and refines fiscal subsidy estimates.
- Empirical conclusions:
  - Static analysis confirms prior findings: pass-through lowest in MENA, highest in advanced economies and emerging Europe; gasoline pass-through often larger than diesel and kerosene in developing economies.
  - Dynamic analysis: average peak pass-through of crude to retail gasoline is 120 percent (1.2 cents per 1 cent crude increase) at six months, with speed and persistence varying by region.
  - Asymmetry confirmed: retail prices respond faster and stronger to positive shocks than to negative shocks.
  - Fiscal cost estimates based on end-of-year prices can substantially understate true fiscal costs—example peak underestimation 1.4 percent of GDP in 2008.
- Policy relevance:
  - Governments using low international prices to rebuild fiscal buffers can lead to muted pass-through during price declines.
  - Accurate monitoring of subsidies and pass-through requires high-frequency retail price data to avoid biased fiscal cost estimates.

*Source: _wp16254 - References (PDF), canonical URL: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16254.pdf*

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

### _wp16254 - References ............................................................................................................. 27

### References
- References ................................................................................................................................................................. 27

### Tables
- 1. Average Retail Fuel Price by Country Groups, January–December, 2014 .......................................... 10
- 2. Median Pass-Through Coefficients by Country Groups and Over Time ............................................ 15

### Figures
- 1. Sample Composition by Country Groups...................................................................................................... 7
- 2. Number of Countries by Year, 1970–2014 .................................................................................................... 8
- 3. Trends in Average Retail Gasoline Prices by Country Groups, January 2000–2014 ........................ 11
- 4. Trends in International Oil Prices, 2005–2016 ............................................................................................ 13
- 5. Dynamic Responses of Retail Gasoline Price to Crude Oil Price Shock .............................................. 17
- 6. Dynamic Responses of Retail Gasoline Price to Crude Oil Price Shock, by Income and  
    Regional Groups ................................................................................................................................................. 19
- 7. Asymmetric Effects of Pass-Through of Crude Oil Price Shocks: Difference Between  
   Positive and Negative Price Shocks ............................................................................................................... 21
- 8. Asymmetric Effects of Pass-Through of Crude Oil Price Shocks: Difference Between  
    Positive and Negative Price Shocks .............................................................................................................. 22
- 9. Discrepancy Between  the Fiscal Cost of Incomplete Pass-Through at the  Average and  
    End-of-Year retail Fuel Prices, 2000-2013 ................................................................................................... 24
- 10. Discrepancy Between  the Fiscal Cost of Incomplete Pass-Through at the  Average and  
    End-of-Year Retail Fuel Prices in Selected Regions, 2000–2013 .......................................................... 24

### Appendixes
- 1. An Overview of Existing Databases on Retail Fuel Prices ....................................................................... 29

*Source: _wp16254 - References (PDF), canonical URL: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16254.pdf*

### 2. Composition of Country Groups (IMF World Economic Outlook Classification) ............................. 32

### 2. Composition of Country Groups (IMF World Economic Outlook Classification) ............................. 32

### I. Introduction — scope and dataset
- Research questions:
  - What is the impact of crude oil price changes on retail fuel prices?
  - How fast do crude oil price shocks feed through to domestic fuel prices?
  - Are responses to price declines symmetric to responses to price increases?
- Dataset compiled:
  - Monthly retail prices for four petroleum products: gasoline, diesel, kerosene, and LPG.
  - Coverage: 162 countries, monthly data for most countries starting from the early 2000s until December 2014; for some countries data go as far back as the 1970s.
  - Total of more than 65,000 monthly observations across products.
- Contributions:
  - Confirms static pass-through findings from prior literature.
  - Estimates impulse response functions of domestic retail gasoline prices using high-frequency data and Jorda (2005) local projections.
  - Differentiates positive versus negative crude oil price shocks to test asymmetry.
  - Demonstrates importance of within-year price changes for fiscal cost estimates of incomplete pass-through.

### II. The new global database on retail fuel prices — sources and methodology
- Primary sources:
  - Publicly posted data from country authorities (regulatory agencies, ministries, state-owned petroleum companies/refineries, national statistics agencies, central banks).
  - IMF staff-collected data for countries not publishing prices online (~40 percent of countries post prices online; IMF staff account for close to 40 percent of the sample; existing databases ~30 percent).
  - Remaining data from press coverage.
- Sample composition (based on 2014 monthly data):
  - Sub-Saharan Africa: 27 percent of sample
  - Advanced Economies: 21 percent
  - Developing Asia: ~14 percent
  - Latin America and the Caribbean: ~14 percent
  - Middle East, North Africa, and Pakistan (MENA): ~14 percent
  - Emerging Europe, Commonwealth of Independent States: remainder
- Data construction and adjustments:
  - End-user, tax-inclusive prices in local currency units (LCU): gasoline, diesel, kerosene in LCU per liter; LPG in LCU per kg (motor LPG per liter).
  - No extrapolation/interpolation; averages used to convert daily/weekly to monthly series.
  - When governments control prices and multiple changes occur in a month, month averages used, except price changes in last week are reported in following month.
  - Averaged across grades when multiple grades exist to limit breaks.
  - Harmonized units: 1 US gallon = 3.78 liters; 1 Imperial gallon = 4.54 liters. LPG unit price uses most common household canister ~12kg or 25 lbs.
  - Adjusted for currency redenomination and currency changes; fuels priced in USD converted at prevailing official exchange rate.
  - Where national series unavailable, capital city averages reported.
- Limitations noted:
  - LPG data coverage limited (12 percent of observations) due to market segmentation, quantity-differentiated pricing, limited reporting.
  - Kerosene coverage limited (19 percent of observations) as household use declines in some countries.
  - Database excludes discounts/rebates, and fuel sold below pump price for specific consumers; shortages and smuggling not captured when official data absent.

### III. Stylized facts on price levels and trends
- Average retail price levels (January–December, 2014, USD per liter except LPG in USD per kg):
  - World average: Gasoline 1.3; Diesel 1.2; Kerosene 0.9; LPG 1.2
  - Advanced Economies: Gasoline 1.9; Diesel 1.7; Kerosene 1.5; LPG ..
  - Emerging Europe: Gasoline 1.8; Diesel 1.8; Kerosene ..; LPG ..
  - Developing Asia: Gasoline 1.2; Diesel 1.1; Kerosene 1.0; LPG 1.5
  - Latin America and the Caribbean: Gasoline 1.1; Diesel 1.0; Kerosene 1.0; LPG 1.1
  - Middle East and North Africa: Gasoline 0.7; Diesel 0.6; Kerosene 0.5; LPG 0.2
  - Sub-Saharan Africa: Gasoline 1.4; Diesel 1.3; Kerosene 1.1; LPG 1.4
  - United States (memo): Gasoline 0.9; Diesel 1.0
- Key observations:
  - Advanced economies and emerging Europe have highest average retail fuel prices; United States is an outlier with lowest gasoline and diesel among advanced economies.
  - MENA countries exhibit lowest retail prices—major oil producers heavily subsidize domestic prices.
  - Sub-Saharan Africa has the highest average fuel prices among developing countries (more than twice MENA), reflecting high supply, transport, storage, and refinery costs.
  - Gasoline usually more expensive than diesel; differentials largely driven by taxation and policy objectives (e.g., lower diesel taxes to limit transport/food price impacts).
  - Wide cross-border price disparities fuel smuggling and complicate subsidy reform.

### IV. Static pass-through analysis — median coefficients by region and period
- Pass-through coefficient definition (expressed in percent):
  - PT_i = 100 * (P_t^i * e_t - P_{t-1}^i * e_{t-1}) / (P_t^* - P_{t-1}^*)
  - Where PT is pass-through in percent; P_t^i retail price in LCU; P_t^* crude oil price at t; e_t exchange rate (USD per unit of LCU); i indexes fuel product.
- Five international oil price subperiods identified (2005–2014):
  - Jan 2005–Dec 2006: gradual increase to peak July 2006.
  - Jan 2007–Jun 2008: dramatic rise, peak 2008 crisis.
  - Jul 2008–Feb 2009: sharp decline at global financial crisis onset.
  - Mar 2009–Jun 2014: rebound then relative stability.
  - Jul 2014–Dec 2014: oil prices declined by 40 percent.
- Median pass-through summaries (selected highlights; values are medians reported in source Table 2):
  - All countries gasoline average (Jan 2005–Dec 2014 average): 101
  - Advanced Economies gasoline median by subperiod: 138; 159; 146; 168; 121; Average 146
  - MENA gasoline median by subperiod (where reported): ..; 41; 1; 16; 0; Average 15
  - Sub-Saharan Africa gasoline median average: 94 (subperiods: 172; 91; 82; 85; 39)
  - Diesel and kerosene medians show similar regional patterns: highest pass-through in advanced economies and emerging Europe; lowest in MENA.
- Interpretations:
  - Pass-through tends to increase with income level and is highest in liberalized-price regions.
  - In many developing countries gasoline pass-through > diesel > kerosene, reflecting differentiated taxation and protection of products used by the poor.
  - Pass-through during falling price spells (e.g., Jul 2014–Dec 2014) is generally lower than during rising price spells, indicating asymmetry.
  - Low pass-through in Jul–Dec 2014 compared to Jul 2008–Feb 2009 suggests governments may use declines to rebuild fiscal buffers or recoup revenue lost when prices were high.

### V. Dynamic analysis — impulse response functions (local projections)
- Empirical approach:
  - Univariate model of retail gasoline price and crude oil price (both in USD), impulse responses estimated via Jorda (2005) local projections.
  - Baseline equation estimated for horizons h = 0,..., H with H = 23 (24 months); p = 12 lags of retail and crude prices.
  - 95 percent confidence intervals via nonparametric block bootstrap with 1000 replications.
  - Sample period for dynamics: January 2000–December 2014.
- Baseline dynamic result (average country):
  - A one cent per liter increase in crude oil prices translates into a 1.2 cent increase in gasoline prices per liter at peak (a pass-through of 120 percent), six months after the shock.
  - Speed: one-to-one parity reached within first three months.
  - Decay: impact declines to less than 0.5 cent by the 10th month and remains persistent in medium term.
- Heterogeneity across country groups (peak pass-through and dynamics):
  - Advanced economies: peak pass-through up to 200 percent within six months; medium-term drop to about 50 percent.
    - Example interpretation: one cent crude increase → two cent retail increase at peak.
  - Emerging Europe and Sub-Saharan Africa: more gradual adjustment, lower peak than advanced economies.
  - MENA, Developing Asia, Latin America: muted and slow responses; MENA peak barely above 50 percent.
  - Persistence of shock driven mainly by advanced economies and Sub-Saharan Africa.

### VI. Asymmetry of pass-through (positive vs negative price shocks)
- Method:
  - Split crude price changes into positive and negative parts (Crudet+ and Crudet−); estimate impulse responses to each separately.
  - Interested coefficient: π1 = α1 − β1 (difference between responses to positive and negative shocks).
- All-countries result:
  - Evidence of asymmetry: retail gasoline prices respond faster and more strongly to positive oil price shocks than to negative shocks.
  - Peak difference at about 8 months: pass-through for positive shocks ~10 percentage points larger than for negative shocks.
  - Gap narrows as positive-shock pass-through phases out more rapidly.
- Disaggregated group results:
  - Advanced economies: peak asymmetry ~20 percentage points (positive > negative).
  - Low-income countries: peak asymmetry ~10 percentage points.
  - Emerging economies: peak asymmetry ~5 percentage points; asymmetric effect dies out within first seven months.
  - MENA: asymmetry nearly inexistent (retail prices move little).
  - Latin America: magnitude comparable to Developing Asia but more persistent.
  - Sub-Saharan Africa: larger and more persistent asymmetry than other developing groups.
- Possible explanations:
  - Market structures: oligopolistic behavior, lack of competition, search costs.
  - Inventory/production adjustment lags.
  - Government policy: in developing countries, lower pass-through of price declines may be used opportunistically to raise net fuel taxes or rebuild fiscal buffers after oil price declines.

### VII. Fiscal cost of incomplete pass-through — end-of-year vs yearly average prices
- Method for subsidy cost estimation:
  - Benchmark price for net oil importers: international USD product price adjusted by $0.20 per liter (shipping and distribution/retailing cost).
  - Benchmark price for net oil exporters: international USD price.
  - Petroleum consumption from IEA; nominal GDP from International Financial Statistics.
  - Fiscal cost compared using yearly average retail prices vs retail prices as of December each year (2005–13).
- Main findings:
  - End-of-year retail prices often underestimate the fiscal cost of incomplete pass-through relative to estimates using yearly average prices.
  - Peak underestimation averaged 1.4 percent of GDP in 2008 (a year marked by oil price crisis).
  - Diesel is the largest contributor to the discrepancy (diesel typically heavily subsidized).
  - Underestimation is larger in years with large and volatile international oil price increases (correlation coefficient 0.30).
  - Regional disparities: underestimation in 2008 largest in MENA (because international prices peaked mid-year and collapsed by December, while subsidies remained high mid-year).
  - Conversely, in rising-price years (e.g., 2007) end-of-year prices may overestimate fiscal cost relative to yearly average.
- Policy implication:
  - High-frequency (intra-year) retail price data are critical to accurately measure fiscal costs of fuel subsidies and to track subsidy reform progress.

### VIII. Conclusion — key implications
- Dataset value:
  - New monthly retail fuel price database (162 countries, four products) enables dynamic analysis of pass-through and refines fiscal subsidy estimates.
- Empirical conclusions:
  - Static analysis confirms prior findings: pass-through lowest in MENA, highest in advanced economies and emerging Europe; gasoline pass-through often larger than diesel and kerosene in developing economies.
  - Dynamic analysis: average peak pass-through of crude to retail gasoline is 120 percent (1.2 cents per 1 cent crude increase) at six months, with speed and persistence varying by region.
  - Asymmetry confirmed: retail prices respond faster and stronger to positive shocks than to negative shocks.
  - Fiscal cost estimates based on end-of-year prices can substantially understate true fiscal costs—especially in volatile years (example peak underestimation 1.4 percent of GDP in 2008).
- Policy relevance:
  - Governments using low international prices to rebuild fiscal buffers can lead to muted pass-through during price declines.
  - Accurate monitoring of subsidies and pass-through requires high-frequency retail price data to avoid biased fiscal cost estimates.

*Source: Authors’ calculations and analysis contained in the IMF working paper chapter "2. Composition of Country Groups (IMF World Economic Outlook Classification)".*

### REFERENCES

### _wp16254 - REFERENCES

### References
- Arze del Granado, F. J., D. Coady and R. Gillingham, 2012, “The Unequal Benefits of Fuel Subsidies: A Review of Evidence for Developing Countries,” World Development, Vol. 40, Issue 11, pp. 2234–48.
- Bachmeier, L., and J. Griffin, 2003, “New Evidence on Asymmetric Gasoline Price Responses,” The Review of Economics and Statistics, Vol. 85, Issue 3, pp. 772–76.
- Baig, T., A. Mati, D. Coady, and J. Ntamatungiro, 2007, “Domestic Petroleum Product Prices and Subsidies: Recent Developments and Reform Strategies,” IMF Working Paper No. WP/07/71 (Washington: International Monetary Fund).
- Balke, N., S. Brown, and M. Yucel, 1998, “Crude Oil and Gasoline Prices: An Asymmetric Relationship?” Economic Review, Federal Reserve Bank of Dallas, pp. 2–11.
- Bloomberg, 2014, “Pain at the Pump: Gasoline Prices by Country”, Dataset, http://www.bloomberg.com/visual-data/gas-prices/
- Borenstein, S., C. Cameron and R. Gilbert, 1997, “Do Gasoline Prices Respond Asymmetrically to Crude Oil Price Changes?,” Quarterly Journal of Economics, Vol. 112, Issue 1, pp. 305–339.
- Clements, B., D. Coady, S. Fabrizio, S. Gupta, T. Alleyne, and C. Sdralevich, eds., 2013, Energy Subsidy Reform: Lessons and Implications (Washington: International Monetary Fund).
- Coady, D., R. Gillingham, R. Ossowski, J. Piotrowski, S. Tareq, and J. Tyson, 2010, “Petroleum Product Subsidies: Costly, Inequitable, and Rising,” IMF Staff Position Note No. 10/05 (Washington: International Monetary Fund).
- Coady D., I. Parry, L. Sears, and B. Shang, 2015, “How Large Are Global Energy Subsidies?” IMF Working Paper No. WP/15/105 (Washington: International Monetary Fund).
- Coady D., V. Flamini, and L. Sears, 2016, “The Unequal Benefits of Fuel Subsidies Revisited: Evidence for Developing Countries”, in Inequality and Fiscal Policy, Clements, B., R. de Mooij, S. Gupta, and M. Keen (eds) (Washington: International Monetary Fund).
- European Commission, 2015, “Weekly Oil Bulletin”, available at https://ec.europa.eu/energy/en/statistics/weekly-oil-bulletin
- GIZ (Gesellschaft für Internationale Zusammenarbeit), 2014, “International Fuel Prices 2012/2013: 8th Edition” available at http://www.giz.de/expertise/downloads/giz2014-en-international-fuel-prices-2013.pdf
- International Energy Agency (IAE), 2015, “End-Use Prices: Energy prices in US dollars,” IEA Energy Prices and Taxes Statistics (database).
- International Monetary Fund, 2015, “World Economic Outlook Update: Cross Currents,” (Washington).
- Johnson, R. N., 2002, “Search Costs, Lags and Prices at the Pump,” Review of Industrial Organization, Vol. 20, pp. 33-50.
- Jordà, O., 2005, “Estimation and Inference of Impulse Responses by Local Projections,” American Economic Review, 95(1), 2005, pp. 161-82.
- Kojima, M., 2009a, “Changes in End-User Petroleum Product Prices: A Comparison of 48 Countries,” Extractive Industries and Development Series No 2, (Washington: World Bank).
- Kojima, M., 2009b, “Government Response to Oil Price Volatility: Experience of 49 Developing Countries,” Extractive Industries and Development Series No. 10 (Washington: World Bank).
- Kojima, M., 2012, “Oil Price Risks and Pump Price Adjustments,” Policy Research Working Paper No. 6227 (Washington: World Bank).
- Kojima, M., 2013, “Petroleum Product Pricing and Complementary Policies: Experience of 65 Developing Countries since 2009,” Policy Research Working Paper No. 6396 (Washington: World Bank).
- Parry, I., D. Heine, S. Li, and E. Lis, 2014, Getting Energy Prices Right: From Principle to Practice, (Washington: international Monetary Fund).
- Peltzman, S., (2000), “Prices Rise Faster than They Fall,” Journal of Political Economy, No. 108, Issue 3, pp. 466-502.
- Radchenko, S., (2005), “Oil Price Volatility and the Asymmetric Response of Gasoline Prices to Oil Price Increases and Decreases", Energy Economics, Vol. 27, Issue 5, pp. 708-730.
- Ross, M., C. Hazlett, and P. Mahdavi, (2015) “The Politics of Petroleum Prices: A New Global Dataset,” Mimeo.

### Appendix 1. An Overview of Existing Databases on Retail Fuel Prices
- Main institutional sources: GIZ, European Commission (EC), International Energy Agency (IEA), UN Economic Commission for Latin America and the Caribbean (ECLAC), and Bloomberg.
- Limitations of available datasets:
  - GIZ: large country coverage and a fair time dimension, but low frequency (annual); limited product coverage (does not compile kerosene and LPG).
  - EC and ECLAC: higher frequency data but coverage tends to be restricted to certain countries in specific regions.
  - IEA and Bloomberg: coverage tends to be restricted to certain income groups or country sets.
  - Datasets from specific studies: may not cover a sufficiently long time horizon and can be limited in country coverage or data frequency.
- Importance of high-frequency data:
  - “This is crucial in light of the observed short-term volatility in fuel prices, and the policy responses to them.”
  - Footnote example: “For instance, high frequency data are needed to gauge the short-term response of consumer prices to changes in retail prices (see Abdallah and Kpodar, 2015).”
- Institutional database features (description based on the 2016 version of these databases):
  - GIZ database:
    - Publication: international fuel prices report published every other year.
    - 2012/13 report: overview of retail gasoline and diesel prices in about 170 countries based on a snapshot survey of prices conducted in November of every other year, starting in the 1990s.
    - Categorization using world crude oil price, retail prices in the United States and Luxembourg as benchmarks into four groups: high fuel subsidies, fuel subsidies, fuel taxation, and high fuel taxation.
    - Supplemented by country factsheets for about 130 countries with information on fuel price trends, fuel price policies, composition of prices, and assessment of pricing mechanism (ad hoc pricing or liberalized prices).
    - Footnote clarifications:
      - “Countries with high fuel subsidies are those where retail prices are below the price of crude oil on the world market. For countries in the fuel subsidies category, retail prices are above international crude oil prices, but below retail prices in the US. When retail prices are above that in the US, but below prices in Luxembourg (the lowest in the EU in November 2012), the country is classified in the fuel taxation category. Countries with high taxation are those where retail fuel prices are above the levels in Luxembourg.”
  - IEA database:
    - Covers 33 OECD member countries with quarterly data from 1978 through 2014 on end-consumer prices for various energy products, including gasoline and diesel.
    - Compiles prices for industries and provides data on import costs and taxes on energy products.
    - For a small subset of countries (France, Germany, Italy, Spain, UK, Japan, Canada and the United States), the IEA releases monthly data (starting from 2005) on gasoline, diesel, heating oil, and fuel oil prices and their applicable taxes.
    - Note: “The IEA used to compile data quarterly energy prices and taxes for selected non-OECD countries, but the publication has been discontinued.”
  - EC database:
    - Publishes weekly retail fuel prices and associated taxes with breakdown in different taxes (VAT, excise, other indirect taxes) since 2009 for the 28 countries of the European Union.
  - ECLAC database:
    - Covers 11 countries in Latin America.
    - Compiles annual information since 2001 on prices for various petroleum products (regular and premium gasoline, diesel, kerosene, fuel oil, LPG), composition of prices (import prices, taxes, gross margin), and provides information on government revenue from taxation of fuel products.
  - Bloomberg database:
    - Compiles quarterly gasoline prices for 61 countries, mainly advanced and emerging countries, since 2013.
    - Provides rankings by average price, affordability (measured by the average day’s wages needed to buy a gallon/liter of fuel), and the expenditure share of gasoline spending (portion of annual income spent on total gas purchases).
- Research-study datasets and findings:
  - Coady et al. (2010):
    - Compiled end-of-period retail fuel prices (gasoline, diesel, kerosene) from 2003 to 2008 for 186 countries to analyze pass-through and estimate magnitude and trends in fuel subsidies.
    - Found that petroleum subsidies had increased significantly during the period as many countries failed to fully pass through the sharp increase in international oil price in 2007 and early 2008.
    - Noted that tax subsidies accounted for the bulk of fuel subsidies, reflecting suboptimal taxation.
  - Coady, Flamini and Sears (2016): extended Coady et al. (2010) data to end-2014.
  - Clements et al. (2013): updated fuel subsidy estimates for 176 countries in 2011 and extended analysis to natural gas, coal, and electricity; discussed macroeconomic consequences and ingredients for successful subsidy reforms based on 22 country case studies.
  - Kojima (2013):
    - Analyzed petroleum product pricing in 65 developing countries based on a survey of end-user prices for gasoline, diesel, and kerosene in July 2012 and pass-through coefficient estimates in January 2009 to July 2012.
    - Documented price control mechanisms (retail price ceilings, retail or wholesale price control) and price adjustment mechanisms (ad hoc price adjustments, automatic pricing mechanisms, price smoothing, fully-liberalized prices).
    - Discussed mitigating responses: energy efficiency improvement, energy diversification, and efforts to lower fuel supply costs.
  - Ross, Hazlett and Mahdavi (2015): compiled monthly gasoline price data for 157 countries to assess the degree to which a country has market-based prices.
- Cross-references:
  - Baig et al. (2007) noted as a similar study for 2003–06 with a smaller sample (51 countries).
  - Kojima (2009a, 2009b) examined retail fuel prices and country responses to price volatility in smaller samples (48 countries and extended to January 2009 respectively).

### Appendix 2. Composition of Country Groups (IMF World Economic Outlook Classification)
- Advanced Economies:
  - Australia, Austria, Belgium, Canada, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hong Kong SAR, Iceland, Ireland, Israel, Italy, Japan, Korea, Luxembourg, Malta, Netherlands, New Zealand, Norway, Portugal, Singapore, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Taiwan Province of China, United Kingdom, and United States.
- Emerging Economies:
  - Algeria, Angola, Argentina, Armenia, Azerbaijan, Bahrain, Belarus, Botswana, Brazil, Brunei Darussalam, Bulgaria, Chile, China, Colombia, Costa Rica, Croatia, Dominican Republic, Ecuador, Egypt, El Salvador, Equatorial Guinea, Fiji, Gabon, Georgia, Guatemala, Hungary, India, Indonesia, Iran, Iraq, Jamaica, Jordan, Kazakhstan, Kuwait, Latvia, Lebanon, Libya, Lithuania, Malaysia, Mauritius, Mexico, Morocco, Namibia, Oman, Pakistan, Panama, Paraguay, Peru, Philippines, Poland, Qatar, Romania, Russia, Saudi Arabia, Serbia, Seychelles, South Africa, Sri Lanka, Swaziland, Syria, Thailand, Trinidad and Tobago, Tunisia, Turkey Turkmenistan, United Arab Emirates, Uruguay, and Venezuela.
- Low Income Developing Countries (LIDC):
  - Afghanistan, Bangladesh, Benin, Bolivia, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo, Republic of, Cote d'Ivoire, Democratic Republic of the Congo, Djibouti, Ethiopia, Gambia, The, Ghana, Grenada, Guinea, Guinea-Bissau, Haiti, Honduras, Kenya, Kiribati, Kyrgyz Republic, Lao P.D.R., Lesotho, Liberia, Madagascar, Malawi, Maldives, Mali, Mauritania, Moldova, Mongolia, Mozambique, Myanmar, Nepal, Nicaragua, Niger, Nigeria, Papua New Guinea, Rwanda, Samoa, Sao Tome and Principe, Senegal, Sierra Leone, Sudan, Tajikistan, Tanzania, Timor-Leste, Dem. Rep. of, Togo, Uganda, Uzbekistan, Vietnam, West Bank and Gaza, Yemen, Zambia, and Zimbabwe.
- Commonwealth of Independent States:
  - Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Russia, Turkmenistan, Kyrgyz Republic, Moldova, Tajikistan, and Uzbekistan.
- Developing Asia:
  - Brunei Darussalam, China, Fiji, India, Indonesia, Malaysia, Philippines, Sri Lanka, Thailand, Bangladesh, Kiribati, Lao P.D.R., Maldives, Mongolia, Myanmar, Nepal, Papua New Guinea, Samoa, Timor-Leste, Dem. Rep. of, and Vietnam.
- Latin America and the Caribbean:
  - Argentina, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Jamaica, Mexico, Panama, Paraguay, Peru, Trinidad and Tobago, Uruguay, Venezuela, Bolivia, Grenada, Haiti, Honduras, and Nicaragua.
- Middle East, North Africa, and Pakistan:
  - Algeria, Bahrain, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Pakistan, Qatar, Saudi Arabia, Syria, Tunisia, United Arab Emirates, Afghanistan, Djibouti, Mauritania, Sudan, West Bank and Gaza, and Yemen.
- Sub-Sahara Africa:
  - Angola, Botswana, Equatorial Guinea, Gabon, Mauritius, Namibia, Seychelles, South Africa, Swaziland, Benin, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo, Republic of, Cote d'Ivoire, Democratic Republic of the Congo, Ethiopia, Gambia, The, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mozambique, Niger, Nigeria, Rwanda, Sao Tome and Principe, Senegal, Sierra Leone, Tanzania, Togo, Uganda, Zambia, and Zimbabwe.

*Content adapted from _wp16254 - REFERENCES (IMF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16254.pdf_