## _wp13128

## Source details

**Canonical URL:** [_wp13128](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13128.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13128.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13128.pdf.json)

---

### Introduction
- Fuel subsidies are common in many developing and emerging economies, including India.
- Objectives of the paper:
  - Evaluate fiscal and welfare implications of fuel subsidy reform in India.
  - Identify issues for designing subsidy reforms.
  - Structure: literature review (Section II), magnitude and financing and required price increases to eliminate subsidies (Section III), impacts on inflation and household real incomes (Section IV), lessons and reform agenda (Section V), conclusions (Section VI).
- Recent policy actions noted:
  - June 2010: petrol pricing liberalized; intention to liberalize diesel prices announced.
  - 2012/13 budget: intention to limit all central subsidies (including fuels) to less than 2 percent of GDP in 2012/13, and reduce them to under 1.75 percent of GDP over three years.
  - January 2013: oil-marketing companies given greater flexibility in setting diesel prices; bulk users of diesel to pay unsubsidized prices.

### Background: Evidence and Literature
- Cross-country and India-specific findings:
  - On average, the top income quintile received 6 times more in subsidy benefits than the bottom quintile (Arze del Granado, Coady, and Gillingham, 2012).
  - Petrol subsidies: over 80 percent of total benefits accrue to the top two quintiles.
  - Diesel and LPG subsidies: respectively, 65 and 70 percent of benefits accrue to the top two quintiles.
  - Kerosene subsidies: benefits accrue more uniformly, but substantial leakage to higher income groups remains.
  - Estimated that increasing fuel prices by $0.25 per liter results, on average, in an increase in the cost of living (CPI) of around 6 percent; about half or more of this increase is due to indirect effects on prices of other goods.
- Selected India-specific studies:
  - Anand M. K. (2012): about a fifth of diesel sales are bulk sales; almost half of diesel sales are retail sales along the national highway network; concludes gradual adjustment to international prices preferable.
  - Lahoti, Suchitra and Goutam (2012): urbanization and income closely correlated with increased use of LPG; liberalizing LPG prices would likely fall less on the poor.
  - Bhanumurthy, Das and Bose (2012): full pass-through of higher global oil prices has negative short-term effects on growth and adds to inflation, but reduces current account deficit; reallocating budgetary savings to public capital expenditure results in higher growth.
  - Tapsoba (2012): reducing untargeted transfers and replacing them with capital expenditure or targeted subsidies raises growth relative to equivalent deficit reduction without reallocation.
  - Parikh and others (2012): a 10 percent increase in diesel prices decreases real incomes on average by around 0.45 of a percent, regressively distributed across income deciles; a 30 percent diesel-price shock shows short-term adverse impacts but after four years inflation under reform nearly 1.5 points lower and GDP 0.7 percentage points higher.

### Institutional and Committee Assessments
- Major committee recommendations:
  - Rangarajan Committee Report, 2006:
    - Use international (“trade parity”) prices as reference for petrol and diesel pricing.
    - Restrict subsidized kerosene to below poverty line (BPL) families.
    - Raise retail price of LPG; finance any remaining subsidies directly from the budget.
  - Parikh Committee Report, 2010:
    - Full liberalization of petrol and diesel prices at refinery gate and pump.
    - Target subsidized kerosene to BPL and raise price yearly according to growth in nominal agricultural GDP per capita; price kerosene sold outside PDS close to diesel; quantity-ration or replace subsidized LPG with direct cash transfers to BPL households and liberalize LPG prices.
  - Nilekani Task Force Interim Report, 2011:
    - Replace in-kind fuel and fertilizer subsidies by direct cash transfers using the Unique identification (UID) system to reduce leakage.
  - Kelkar Committee Report, 2012:
    - Roadmap including elimination of diesel subsidies over a two-year period followed by full price deregulation in 2014; elimination of LPG subsidies over three years; reduction of kerosene subsidies by one-third over same period.

### Magnitude and Financing of Fuel Subsidies in India
- Structural features:
  - India is the fourth largest global consumer of fuel products; demand increasing with high economic growth and rising incomes.
  - Two state-owned companies—Oil and Natural Gas Corporation Ltd (ONGC) and Oil India Ltd (OIL)—account for just over 75 percent of all extracted crude oil; ONGC alone accounts for nearly 65 percent of extracted crude.
  - Refining roughly equally divided between public and private refineries since most capacity increases from private sector since 2009.
  - Distribution: IOC, BPCL, and HPCL still dominate marketing side, controlling over 80 percent of the market.
- Under-pricing and subsidy measurement:
  - Retail prices of diesel, kerosene, and LPG are centrally administered; petrol prices have been liberalized.
  - Subsidy levels calculated as difference between international reference prices and regulated retail prices; international reference prices include international import prices (inclusive of international freight and insurance charges, and landing charges), product taxes, and transport and marketing margins.
- Estimated fuel subsidies (aggregate):
  - 0.6 percent of GDP in 2004/5.
  - 1.9 percent of GDP in 2007/8.
  - 1.9 percent of GDP in 2011/2.
- Diesel volume share:
  - Diesel accounts for about 60 percent of the total volume of the four fuel products consumed in the country.
- Financing modalities over time:
  - 2004/5: subsidies financed almost exclusively off-budget (OMCs absorbing losses, transfers from upstream oil companies).
  - Subsequent increases financed mainly by government issuing oil bonds to OMCs in lieu of direct budget transfers, and by increased financing from upstream companies and OMCs.
  - Upstream financing increased substantially in 2008/9; decreased with subsidy decline in 2009/10.
  - No oil bonds issued since 2009/10.
  - Sharp increase in subsidies thereafter financed mainly from the budget and by upstream companies.
  - Note: elimination of subsidies improves the budget only if off-budget financing is redirected to the budget.

### Formula and Measurement of Subsidies
- Appendix Table 1 describes government formula for “under-recoveries”.
- Subsidy calculations include “tax subsidies” arising when actual taxes are below target taxes.
- Fuel product customs duties collected by refineries are reported by government officials as being retained by refineries (not handed over to revenue authorities), operating like a protective tariff.

### Evolution of fuel product prices and subsidies (2004–2012)
- Petrol:
  - Domestic petrol retail prices liberalized in June 2010.
  - Prior to liberalization, domestic petrol prices tended to move with international prices with small subsidies except 2007–2008 when domestic prices barely changed while international prices surged.
- Diesel:
  - Retail diesel prices remained regulated; diesel subsidies increased sharply in 2007–2008, collapsed in late 2008 with international prices, rose until end-2010, then diverged and subsidies reemerged and increased.
  - Regulated retail diesel prices were increased by about Rs. 3.50 per liter in mid-September 2012.
  - January 2013 announcement: all bulk diesel sales at unsubsidized prices and OMCs have discretion to periodically increase diesel prices by small amounts.
- Kerosene and LPG:
  - Subsidies per liter and per 14.2 kg cylinder traditionally much higher than petrol/gasoline; peaked mid-2008, fell sharply, then increased continually to return to mid-2008 peak.
  - End-2012: government restricted households to 6 cylinders per year at subsidized LPG prices.
  - January 2013: quota increased to 9 cylinders.

### Distributional patterns of subsidy incidence
- Most benefit accrues to higher income groups who consume greater amounts of fuels.
- Per capita monthly expenditure on fuel products and share of fuel expenditures in total expenditures are substantially higher for higher income households:
  - Bottom two income deciles allocate around 1.6 percent of total expenditures to fuel consumption.
  - Top two income deciles allocate nearly 6 percent of total expenditures to fuel.
- Highest income decile spends more than twenty times as much as the lowest decile on fuel in per capita terms.
- Fuel composition differs by income:
  - Lowest income groups: almost all fuel expenditure on kerosene.
  - Highest income groups: petrol and LPG dominate.

### Methodology for price and welfare impact analysis
- Two channels for household welfare effects of higher domestic fuel prices:
  1. Direct impact: higher prices for fuels consumed for cooking, heating, lighting, and personal transport.
  2. Indirect impact: higher prices for other goods and services as higher fuel costs increase production costs and consumer prices.
- Direct impact calculation uses household survey budget shares:
  - Direct Impact = Budget Share X Percentage Increase in Fuel Price X 100
  - Interpreted as short-run impact or upper-bound long-run impact (assumes no substitution).
- Indirect impact estimated using the Coady and Newhouse (2006) price-shifting model:
  - Assumes increases in fuel production costs are fully passed forward to domestic prices of goods and services.
  - Uses India input-output table for 2003/04 (130 sectors).
  - Assumes zero demand price elasticity — interpret as short-term impact or upper-bound.
- Distributional classification:
  - Households allocated to welfare quintiles by consumption per capita.
  - Impact considered progressive if percentage welfare loss increases with total per capita household consumption.

### Welfare impact of eliminating subsidies (simulation based on end-October 2012 under-pricing)
- Average aggregate effect:
  - Eliminating fuel subsidies would decrease household real incomes (or equivalently, increase the CPI) by around 4 percent.
  - About three-quarters of the total impact comes from the direct impact on household fuel consumption (large increases in kerosene and LPG prices).
- Indirect impacts:
  - Indirect impact is similar across income groups in magnitude but differs in composition:
    - Lower income groups: indirect impact dominated by higher food prices.
    - Higher income groups: indirect impact dominated by higher non-food prices.
- Distributional outcome:
  - Total impact is progressive: percentage impact higher for higher income groups.
  - Exception: top income decile shows somewhat lower impact due to relative importance of petrol consumption and small petrol price increase.
- Wholesale Price Index (WPI) impact:
  - Impact on the WPI is much lower at 2.2 percent compared to the CPI.

### Absolute per capita impacts and compensation implications
- Average per capita welfare impact (monthly):
  - Bottom two income deciles: about Rs 20 per capita per month.
  - Top two income deciles: around Rs 120 per capita per month.
- Policy implication:
  - Because lower income groups receive a very small proportion of total fuel subsidies, it should be possible to fully protect lower income groups while generating substantial net fiscal savings.
- Note: simulations use 2009/10 National Socio-Economic Household Survey expenditure data and India input-output table for 2003/04; indirect impacts in simulations assume they arise only from the increase in the diesel price.

### Specific under-pricing estimates, November 2012 (summary from Table 1)
- Table format: Import Parity Price | Regulated Price | Difference | Percent of Regulated Price
  - Kerosene: 46.9 | 14.8 | 32.1 | 217.4
  - LPG: 9 | 11.5 | 4 | 10.5 | 501.0 | 122.1  (table line preserved from source as provided)
  - Diesel: 57.0 | 47.2 | 9.8 | 20.9
  - Gasoline: 72.7 | 68.1 | 4.7 | 6.8
- End-October 2012 under-pricing basis used for simulations.

### International experience with fuel subsidy reform — key design ingredients
- Elements that enhance successful and durable subsidy reform:
  - Communications campaign: emphasize fiscal cost, regressivity, crowding-out of priority spending.
  - Comprehensive reform plan: clear vision for petroleum sector, timeline for eliminating subsidies, transparent on-budget recording, and mitigation measures for lower-income households.
  - Gradual and sequenced reform: large price increases phased to allow energy-efficiency adjustments and lower inflationary impact; sequence increases toward fuels less important for the poor; adopt automatic pricing mechanisms with smoothing.
- Country examples:
  - Brazil: sequential deregulation (petrol 1996, LPG 1998, diesel 2001), withdrawal of LPG subsidies 2001 with targeted voucher program, Bolsa Familia consolidation 2003.
  - Philippines: energy price liberalization from 1996, automatic pricing mechanism, full liberalization by July 1998; scaled up cash transfers during 2007–2008.
  - Turkey: deregulation early 1990s; automatic fuel pricing mechanism adopted 1998, implemented by Energy Market Regulatory Authority since 2003; gradual fuel tax increases and scaling-up of targeted cash transfer programs.

### Fuel subsidy reform in India — overall findings
- Historical and distributional impacts:
  - Fuel subsidies increased from around 0.8 percent of GDP in 2009/10 to 1.9 percent of GDP in 2011/12.
  - Eliminating subsidies will have a substantial negative impact on real incomes of households, estimated to range from 4 percent for the lowest income groups to 5 percent for higher income groups.
  - About three-quarters of the impact arises from substantially higher prices for kerosene and LPG.
- Policy recommendations from commissions:
  - Use international fuel prices as appropriate reference.
  - Eliminate diesel subsidies over the short-term followed by full diesel price liberalization.
  - Gradually remove kerosene and LPG subsidies.
  - Replace in-kind subsidies with targeted cash transfers.

### Diesel subsidy reform — findings and recommendations
- Required price adjustment:
  - To remove existing diesel subsidies, regulated diesel prices would need to be increased by Rs. 10 per liter, from Rs. 47 to Rs. 57, i.e., a 20 percent increase.
- Implementation timing:
  - If OMCs increase prices by Rs. 0.5 per month (as recently allowed), the diesel subsidy could be completely removed by mid-2014 (alternative passage: by the end of 2014 if increases of Rs. 0.5 per month are applied).
- Trade-offs and risks:
  - Gradual increases reduce inflationary and income impacts but prolong political risk of reform reversal and reduce near-term savings if international diesel prices rise.
  - Full liberalization passes through sharp international price increases quickly, potentially causing public resistance; an automatic pricing mechanism with smoothing can limit magnitude of single price changes while ensuring full pass-through over the medium term.
  - Smoothing reduces domestic price volatility but increases fiscal volatility.

### LPG subsidy reform — findings and recommendations
- Rationing history:
  - From September 2012, government imposed annual limit of six 14.2 kg cylinders at subsidized prices; January 2013 quota increased to 9 cylinders.
- Price and required adjustments:
  - Current subsidy on LPG is around Rs. 500 per cylinder.
  - Eliminating this subsidy would require a 217 percent increase in price from the current price of Rs. 410 per cylinder.
- Reform path:
  - Eliminate existing subsidy over a two-year period.
  - Adopt an automatic LPG pricing mechanism with a smoothing component.
  - Once market LPG prices reach unsubsidized levels for consumption beyond rationed quantities, consider transforming universal rationing into a targeted system focusing on lower-income households and eventually converting in-kind subsidy into a cash transfer.

### Kerosene subsidy reform — findings and recommendations
- Scope and required adjustment:
  - International kerosene prices increased by around 150 percent since early 2009, while subsidized PDS kerosene domestic prices hardly changed.
  - Per liter subsidy increased almost threefold, from around Rs. 10 in early 2009 to Rs. 30 in October 2012.
  - Removing the subsidy requires nearly a 220 percent increase in the price of kerosene.
- Channels to reduce subsidies:
  - Implement better beneficiary registration and monitoring to reduce illegal leakage (estimated that 38 percent of PDS kerosene does not reach intended beneficiaries).
  - Rely on self-selection where higher-income households must appear in person to receive subsidized kerosene.
  - Limit kerosene subsidy to BPL households by removing entitlement from APL households (Rangarajan 2006; Parikh 2010).
  - Consider an automatic kerosene pricing mechanism with smoothing; consider eventual replacement of in-kind kerosene subsidies with cash transfers.
- Institutional opportunity:
  - Nationwide rollout of UID/Aadhaar is promising for reducing leakage; a conservative estimate found a real internal rate of return of around 53 percent to adoption of Aadhaar for social program registration.

### Cash transfers and targeting
- Rationale and design:
  - Cash transfers equal to unit subsidies on kerosene and LPG times existing rationed quantities; could be inflation-indexed.
  - Cash transfers remove government involvement in fuel distribution and let households decide expenditures.
- Potential savings from better targeting:
  - Eliminate non-poor households from receipt of transfers; savings depend on poverty line used and on inclusion of poor households currently excluded.
- Implementation advice:
  - Use pilots to test design, implementation, beneficiary response, and build political support.

### Conclusions and reform sequencing for India
- Priorities:
  - Priority to eliminate diesel subsidies in the short term since diesel accounts for the largest share of total fuel subsidy.
  - If OMCs increase diesel prices by Rs. 0.5 per month as recently allowed, the subsidy could be eliminated by the end of 2014 (alternate passage: by mid-2014).
  - Follow diesel elimination with full price liberalization and consider an automatic pricing mechanism to smooth adjustments and avoid large single price changes.
- LPG and kerosene sequencing:
  - Use quantity rations for subsidized LPG to protect low-income households in the short term; adopt automatic pricing with smoothing; later transform universal rationing into targeted safety nets and eventually to cash transfers.
  - For kerosene, develop better beneficiary registration and monitoring (leveraging UID/Aadhaar), pursue better targeting (withdraw entitlements from APL households), adopt automatic pricing with smoothing, and consider replacing in-kind subsidies with cash transfers.
- Targeting systems:
  - UID will help eliminate duplicate and ghost beneficiaries but is not a substitute for an effective targeting framework; current BPL/APL identification needs reform.
  - Pilot programs recommended to refine design and implementation before nationwide scaling-up.

*Source: _wp13128 - References*

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

### _wp13128 - References

### Introduction
- Fuel subsidies are common in many developing and emerging economies, including India.
- Subsidies aim to shield consumers (especially poor households) from high and volatile fuel costs for lighting, cooking, and transportation.
- Subsidies are characterized as inefficient and inequitable (IMF, 2010, 2013), with effects including:
  - Encouraging overconsumption of fuel.
  - Delaying adoption of energy-efficient technologies.
  - Crowding out high-priority public spending (physical infrastructure, education, health, social protection).
- Most benefits of fuel subsidies accrue to higher income groups who consume more fuel (Arze and others, 2012).
- Policy actions noted:
  - June 2010: petrol pricing liberalized; intention to liberalize diesel prices announced.
  - 2012/13 budget: intention to limit all central subsidies (including fuels) to less than 2 percent of GDP in 2012/13, and reduce them to under 1.75 percent of GDP over three years.
  - January 2013: oil-marketing companies given greater flexibility in setting diesel prices; bulk users of diesel to pay unsubsidized prices.
- Paper objectives:
  - Evaluate fiscal and welfare implications of fuel subsidy reform in India.
  - Identify issues for designing subsidy reforms.
  - Structure: literature review (Section II), magnitude and financing and required price increases to eliminate subsidies (Section III), impacts on inflation and household real incomes (Section IV), lessons and reform agenda (Section V), conclusions (Section VI).

### Background: Evidence and Literature
- Cross-country review (Arze del Granado, Coady, and Gillingham, 2012) of 20 developing countries (9 Africa, 4 Central and South America, 5 Asia and the Pacific, 2 Middle East) with India included:
  - On average, the top income quintile received 6 times more in subsidy benefits than the bottom quintile.
  - Petrol subsidies: over 80 percent of total benefits accrue to the top two quintiles.
  - Diesel and LPG subsidies: respectively, 65 and 70 percent of benefits accrue to the top two quintiles.
  - Kerosene subsidies: benefits accrue more uniformly, but substantial leakage to higher income groups remains.
  - Estimated that increasing fuel prices by $0.25 per liter results, on average, in an increase in the cost of living (CPI) of around 6 percent; about half or more of this increase is due to indirect effects on prices of other goods.
- India-specific studies:
  - Anand M. K. (2012):
    - About a fifth of diesel sales are bulk sales (largely for transportation and power generation).
    - Almost half of diesel sales are retail sales along the national highway network (much for long-haul freight).
    - Concludes gradual adjustment to international prices preferable to reduce adjustment costs.
  - Lahoti, Suchitra and Goutam (2012):
    - Using India Human Development Survey data, urbanization and income closely correlated with increased use of LPG.
    - Liberalizing LPG prices would likely fall less on the poor than on higher-income households.
  - Bhanumurthy, Das and Bose (2012):
    - Macroeconomic model allowing for monetary and fiscal responses to oil price shocks.
    - Full pass-through of higher global oil prices (vs preventing pass-through and raising subsidy bill) has negative short-term effects on growth and adds to inflation, but reduces current account deficit.
    - Reallocating budgetary savings from raised fuel prices to public capital expenditures results in higher growth.
  - Tapsoba (2012):
    - Using a DSGE model, finds that reducing untargeted transfers (universal fuel subsidies) and replacing them with capital expenditure or a mix of capital expenditure and better-targeted subsidies raises growth relative to equivalent deficit reduction without reallocation.
  - Parikh and others (2012):
    - VAR approach comparing a 30 percent increase in diesel prices to a “no-reform” baseline (2011–2015).
    - Short-term adverse impact on GDP and inflation, quickly reversed: after four years, inflation under reform nearly 1.5 points lower and GDP 0.7 percentage points higher.
    - A 10 percent increase in diesel prices decreases real incomes on average by around 0.45 of a percent, regressively distributed across income deciles.

### Institutional and Committee Assessments
- Major committee reports and recommendations:
  - Rangarajan Committee Report, 2006:
    - Recommended using international (“trade parity”) prices as a reference for petrol and diesel pricing.
    - Recommended subsidized kerosene be restricted to below poverty line (BPL) families.
    - Recommended retail price of LPG be raised, with any remaining subsidies financed directly from the budget.
  - Parikh Committee Report, 2010:
    - Recommended full liberalization of petrol and diesel prices at refinery gate and pump.
    - Recommended: (i) subsidized kerosene sold through PDS be targeted to BPL families and its price raised yearly according to growth in nominal agricultural GDP per capita; (ii) kerosene sold outside PDS be priced close to diesel to eliminate diversion incentives; (iii) subsidized LPG be quantity-rationed or replaced by direct cash transfers to BPL households with LPG prices fully liberalized.
  - Nilekani Task Force Interim Report, 2011:
    - Proposed replacement of in-kind fuel and fertilizer subsidies by direct cash transfers using the Unique identification (UID) system to reduce fiscal cost by eliminating leakage (duplication, ghost beneficiaries).
    - Set out an action plan for pilot reforms.
  - Kelkar Committee Report, 2012:
    - Roadmap for fiscal consolidation including timeline for fuel subsidy reduction.
    - Recommended elimination of diesel subsidies over a two-year period followed by full price deregulation in 2014.
    - Recommended elimination of LPG subsidies over three years.
    - Recommended reduction of kerosene subsidies by one-third over the same period.

### Magnitude and Financing of Fuel Subsidies in India
- Structural features of the petroleum sector:
  - India is the fourth largest global consumer of fuel products; demand increasing with high economic growth and rising incomes.
  - Increased domestic production has met much demand growth, but stagnant domestic crude production has required increased crude oil imports by refineries.
  - Exports of fuel products have also been increasing due to increased refining capacity.
  - Government heavily involved upstream and downstream through state-owned enterprises.
  - Reliance Industries Ltd (RIL) is the only private sector company present at all levels.
- Sector composition:
  - Crude oil extraction:
    - Two state-owned companies—Oil and Natural Gas Corporation Ltd (ONGC) and Oil India Ltd (OIL)—account for just over 75 percent of all extracted crude oil.
    - ONGC alone accounts for nearly 65 percent of extracted crude.
  - Refining:
    - Main public sector refiners: Indian Oil Corp Ltd (IOC), Bharat Petroleum Corp Ltd. (BPCL), Hindustan Petroleum Corp Ltd (HPCL), Chennai Petroleum Corp Ltd (CPCL).
    - With most refining capacity increases coming from private sector since 2009, refining is now roughly equally divided between public and private refineries.
  - Distribution:
    - Prior to mid-2000s, downstream distribution controlled largely by IOC, BPCL, HPCL.
    - Mid-2000s reforms authorized other companies (including ONGC; GAIL; Mangalore Refinery and Petrochemicals Ltd (MRPL); Numaligarh Refinery Ltd (NRL); RIL; Essar Oil Limited (EOL); and Shell India Pvt. Ltd (SIMPL)) to distribute fuel products.
    - IOC, BPCL, and HPCL still dominate marketing side, controlling over 80 percent of the market, because private sector finds it unprofitable to distribute subsidized products.
- Under-pricing and subsidy measurement:
  - Retail prices of diesel, kerosene, and LPG are centrally administered, producing large consumer price subsidies for some products.
  - Petrol prices have been liberalized (discussed for comparison).
  - Subsidy levels are calculated as the difference between international reference prices and regulated retail prices.
  - International reference prices are computed as the sum of international import prices (inclusive of international freight and insurance charges, as well as landing charges), product taxes (customs tariffs, central government excises, and state sales taxes), and transport and marketing margins.

*Source: _wp13128 - References*

### Appendix Table 1 describes the

### _wp13128 - Appendix Table 1 describes the

### Formula and measurement of subsidies
- Appendix Table 1 describes the detailed formula used by the government to calculate subsidies (referred to as “under-recoveries”).
- Subsidy calculations include “tax subsidies” that arise when actual taxes are below target taxes.
- Fuel product customs duties collected by refineries are reported by government officials as being retained by refineries (not handed over to revenue authorities), operating like a protective tariff.

### Evolution of fuel product prices and subsidies (2004–2012)
- Domestic petrol retail prices were liberalized in June 2010. Prior to liberalization, domestic petrol prices tended to move in line with international prices, with generally small subsidies except for 2007–2008 when domestic prices barely changed while international prices surged.
- Diesel retail prices remained regulated through the period; diesel subsidies increased sharply in 2007–2008, collapsed with international prices in late 2008, rose with international prices until end-2010, then diverged and subsidies reemerged and increased.
  - Regulated retail diesel prices were increased by about Rs. 3.50 per liter in mid-September 2012.
  - January 2013 government announcement: all bulk diesel sales would take place at unsubsidized prices and OMCs have discretion to periodically increase diesel prices by small amounts.
- Kerosene and LPG subsidies per liter and per 14.2 kg cylinder respectively:
  - Traditionally much higher than petrol/gasoline.
  - Peaked with international prices in mid-2008, fell sharply, then increased continually over the last three years to recently return to mid-2008 peak.
  - End-2012: government restricted households to 6 cylinders per year at subsidized LPG prices.
  - January 2013: quota increased to 9 cylinders.

### Distributional patterns of subsidy incidence
- Most benefit from price subsidies accrues to higher income groups, who consume greater amounts of fuels.
- Per capita monthly expenditure on fuel products and the share of fuel expenditures in total expenditures are substantially higher for higher income households.
  - Bottom two income deciles allocate around 1.6 percent of total expenditures to fuel consumption.
  - Top two income deciles allocate nearly 6 percent of total expenditures to fuel.
- Highest income decile spends more than twenty times as much as the lowest decile on fuel in per capita terms.
- Fuel composition differs by income:
  - Lowest income groups: almost all fuel expenditure on kerosene.
  - Highest income groups: petrol and LPG dominate.

### Magnitude and financing of fuel subsidies
- Estimated fuel subsidies:
  - 0.6 percent of GDP in 2004/5.
  - 1.9 percent of GDP in 2007/8.
  - 1.9 percent of GDP in 2011/2.
- Diesel accounts for about 60 percent of the total volume of the four fuel products consumed in the country.
- Financing modalities over time:
  - 2004/5: subsidies financed almost exclusively off-budget (OMCs absorbing losses, transfers from upstream oil companies).
  - Subsequent increase financed mainly by government issuing oil bonds to OMCs in lieu of direct budget transfers, and by increased financing from upstream companies and OMCs.
  - Upstream financing increased substantially in 2008/9; decreased with subsidy decline in 2009/10.
  - No oil bonds issued since 2009/10.
  - Sharp increase in subsidies thereafter financed mainly from the budget and by upstream companies.
- Note on off-budget financing: elimination of subsidies improves the budget only if off-budget financing is redirected to the budget.

### Methodology for price and welfare impact analysis
- Two channels for household welfare effects of higher domestic fuel prices:
  1. Direct impact: higher prices for fuels consumed for cooking, heating, lighting, and personal transport.
  2. Indirect impact: higher prices for other goods and services as higher fuel costs increase production costs and consumer prices.
- Direct impact calculation uses household survey budget shares:
  - Direct Impact = Budget Share X Percentage Increase in Fuel Price X 100
  - Interpreted as short-run impact or upper-bound long-run impact (assumes no substitution).
- Indirect impact estimated using the Coady and Newhouse (2006) price-shifting model:
  - Assumes increases in fuel production costs are fully passed forward to domestic prices of goods and services.
  - Requires input-output table describing production cost structure; India input-output table for 2003/04 (130 sectors) used.
  - Assumes zero demand price elasticity — interpret as short-term impact or upper-bound.
- Distributional classification:
  - Households allocated to welfare quintiles by consumption per capita.
  - Impact considered progressive if percentage welfare loss increases with total per capita household consumption.

### Welfare impact of eliminating subsidies (simulation based on end-October 2012 under-pricing)
- Average aggregate effect:
  - Eliminating fuel subsidies would decrease household real incomes (or equivalently, increase the CPI) by around 4 percent.
  - About three-quarters of the total impact comes from the direct impact on household fuel consumption (large increases in kerosene and LPG prices).
- Indirect impacts:
  - Indirect impact is similar across income groups in magnitude but differs in composition:
    - Lower income groups: indirect impact dominated by higher food prices.
    - Higher income groups: indirect impact dominated by higher non-food prices.
- Distributional outcome:
  - Total impact is progressive: percentage impact higher for higher income groups.
  - Exception: top income decile shows somewhat lower impact due to relative importance of petrol consumption and small petrol price increase.
- Wholesale Price Index (WPI) impact:
  - Impact on the WPI is much lower at 2.2 percent compared to the CPI.

### Absolute per capita impacts and compensation implications
- Average per capita welfare impact (monthly):
  - Bottom two income deciles: about Rs 20 per capita per month.
  - Top two income deciles: around Rs 120 per capita per month.
- Policy implication:
  - Because lower income groups receive a very small proportion of total fuel subsidies, it should be possible to fully protect lower income groups while generating substantial net fiscal savings.
- Note: simulations and results use 2009/10 National Socio-Economic Household Survey expenditure data and India input-output table for 2003/04; indirect impacts in simulations assume arise only from the increase in the diesel price.

### Specific under-pricing estimates, November 2012 (Table 1, in rupees per liter)
- Table 1 columns: Import Parity Price | Regulated Price | Difference | Percent of Regulated Price
- Kerosene: 46.9 | 14.8 | 32.1 | 217.4
- LPG: 9 | 11.5 | 4 | 10.5 | 501.0 | 122.1
  - (Table line in source: "LPG911.5410.5501.0122.1" — preserved exactly in structure above as individual values where extractable.)
- Diesel: 57.0 | 47.2 | 9.8 | 20.9
- Gasoline: 72.7 | 68.1 | 4.7 | 6.8

*Italic: Source: IMF Staff estimates and analysis (content unit: _wp13128 - Appendix Table 1 describes the).*

### 0.2 of a percent of GDP (Figure 9). This compares with a gross fiscal savings from subsidy

### _wp13128 - 0.2 of a percent of GDP (Figure 9). This compares with a gross fiscal savings from subsidy

### International experience with fuel subsidy reform
- Key design ingredients that enhance successful and durable subsidy reform:
  - Communications campaign: emphasize that fuel subsidies are fiscally costly, regressive (accrue mostly to higher income groups), and crowd-out higher priority public spending.
  - Comprehensive reform plan: identify a vision for a more efficient petroleum product sector, a clear timeline for eliminating subsidies, transparently record total cost of fuel subsidies on-budget, and specify measures to mitigate adverse impacts on consumers, especially lower-income households.
  - Gradual and sequenced reform: implement large price increases gradually to allow energy-efficiency adjustments and reduce inflationary impact; sequence initial increases toward fuels less important for lower-income households (e.g., petrol) and more gradually for fuels important to the poor (e.g., kerosene); adopt automatic pricing mechanisms to smooth adjustments and limit magnitude of single price changes.
- Country examples illustrating these ingredients:
  - Brazil: sequential deregulation—petrol pricing deregulated in 1996, LPG in 1998, diesel in 2001; withdrawal of LPG subsidies in 2001 followed by a targeted LPG voucher program and a nationwide targeted cash transfer program consolidated under Bolsa Familia in 2003.
  - The Philippines: energy price liberalization from 1996 with advance planning and consensus building; adopted an automatic pricing mechanism and fully liberalized fuel pricing by July 1998; launched and scaled up a cash transfer program during 2007–2008 international price shocks.
  - Turkey: energy sector deregulation in the early 1990s; automatic fuel pricing mechanism adopted in 1998 and implemented by the independent Energy Market Regulatory Authority since 2003; gradual fuel tax increases and scaling-up of targeted cash transfer programs.

### Fuel subsidy reform in India — overall findings
- Historical and distributional impacts:
  - Fuel subsidies increased from around 0.8 percent of GDP in 2009/10 to 1.9 percent of GDP in 2011/12.
  - Eliminating subsidies will have a substantial negative impact on real incomes of households, estimated to range from 4 percent for the lowest income groups to 5 percent for higher income groups.
  - About three-quarters of the impact arises from substantially higher prices for kerosene and LPG.
- Policy recommendations from commissions:
  - Use international fuel prices as the appropriate reference for setting prices.
  - Eliminate diesel subsidies over the short-term followed by full diesel price liberalization.
  - Gradually remove kerosene and LPG subsidies.
  - Replace in-kind subsidies with targeted cash transfers.

### Diesel subsidy reform — findings and recommendations
- Fiscal and price adjustments:
  - To completely remove existing diesel subsidies, regulated diesel prices would need to be increased by Rs. 10 per liter, from Rs. 47 to Rs. 57, i.e., a 20 percent increase.
  - If oil marketing companies increase prices by Rs. 0.5 per month (as recently allowed), the diesel subsidy could be completely removed by mid-2014 (alternative passages state by the end of 2014 if increases of Rs. 0.5 per month are applied).
- Rationale and risks:
  - Diesel subsidy elimination addresses the distortion where petrol prices are around 45 percent higher than diesel prices, which has led to substitution toward diesel automobiles and larger fiscal costs.
  - Gradual increases reduce inflationary and income impacts but prolong political risk of reform reversal and reduce near-term budget savings if international diesel prices rise.
  - Full liberalization passes through sharp international price increases quickly, potentially causing public resistance; smoothing via an automatic fuel pricing mechanism can limit the magnitude of any single price change while ensuring full pass-through over the medium term.
- Trade-offs:
  - Smoothing reduces domestic price volatility but increases fiscal volatility.
  - Avoiding reintroduction of subsidies may require not fully passing decreases in international prices through to consumers.

### LPG subsidy reform — findings and recommendations
- Recent rationing actions and their effects:
  - From September 2012, government imposed an annual quantity limit of six 14.2 kg cylinders of subsidized LPG per household; in January 2013 this limit was increased to 9 cylinders, likely offsetting most savings from the earlier reform.
- Price and required adjustments:
  - Current subsidy on LPG is around Rs. 500 per cylinder.
  - Eliminating this subsidy would require a 217 percent increase in price from the current price of Rs. 410 per cylinder.
- Reform path:
  - Eliminate existing subsidy over a two-year period (somewhat faster than the three-year period recommended by the Kelkar report).
  - Adopt an automatic LPG pricing mechanism with a smoothing component.
  - Once market LPG prices reach unsubsidized levels for consumption beyond rationed quantities, consider transforming universal rationing into a targeted system focusing on lower-income households and eventually converting in-kind subsidy into a cash transfer (recommended by the 2010 Parikh Report and the 2011 Nilekani Report).

### Kerosene subsidy reform — findings and recommendations
- Evolution and scale of subsidy:
  - International kerosene prices increased by around 150 percent since early 2009, while subsidized PDS kerosene domestic prices hardly changed.
  - Per liter subsidy on kerosene increased almost threefold, from around Rs. 10 in early 2009 to Rs. 30 in October 2012.
  - Removing the subsidy requires nearly a 220 percent increase in the price of kerosene.
- Channels to reduce subsidies:
  - Implement a more effective beneficiary registration and monitoring system to reduce illegal leakage from duplicate and ghost ration cards (estimated that 38 percent of PDS kerosene does not reach intended beneficiaries).
  - Rely on self-selection where higher-income households must appear in person to receive subsidized kerosene.
  - Limit kerosene subsidy to BPL households by removing entitlement from APL households (as recommended by the 2006 Rangarajan Report and the 2010 Parikh Report).
  - Consider an automatic kerosene pricing mechanism with smoothing to avoid large one-off domestic price increases when international prices spike.
  - Consider eventual replacement of in-kind kerosene subsidies with cash transfers.
- Institutional opportunity:
  - Ongoing nationwide rollout of India’s Unique Identification (UID/Aadhaar) system — biometric identification of all Indians over the next three years — is promising for reducing leakage; a conservative estimate found a real internal rate of return of around 53 percent to adoption of Aadhaar for social program registration.

### Cash transfers and targeting
- Rationale for transforming in-kind subsidies into cash:
  - The level of cash transfers should equal unit subsidies on kerosene and LPG times existing rationed quantities and could be inflation-indexed to protect households from future international price shocks.
  - Cash transfers would remove government involvement in fuel distribution and let households decide expenditures; government would ensure adequate fuel supply access.
- Potential additional savings from better targeting:
  - Eliminate non-poor households from receipt of transfers; better targeting would require more effective approaches than current BPL/APL identification, which has substantial undercoverage and leakage.
  - Magnitude of savings depends on the poverty line used for targeting and on inclusion of genuinely poor households currently excluded.
- Implementation advice:
  - Use pilot schemes: effectively designed, implemented, and evaluated pilots can inform program attractiveness to beneficiaries, refine design before scaling up, and build political support.
  - Learning from pilots already implemented in some Indian districts is crucial, though design and performance information was not available.

### Conclusions and reform sequencing for India
- Priorities and sequencing:
  - Priority should be given to eliminating diesel subsidies in the short term, since diesel accounts for the largest share of total fuel subsidy.
  - If oil marketing companies increase diesel prices by Rs. 0.5 per month as recently allowed, the subsidy could be eliminated by the end of 2014 (alternate passage: by mid-2014).
  - Follow diesel elimination with full price liberalization and consider an automatic pricing mechanism to smooth adjustments and avoid large single price changes while ensuring pass-through over the medium term.
- LPG and kerosene require more gradual and comprehensive reforms because of larger required price increases and greater importance for poor households:
  - Use quantity rations for subsidized LPG to protect low-income households in the short term; adopt automatic pricing with smoothing; later transform universal rationing into targeted safety nets and eventually to cash transfers.
  - For kerosene, first develop better beneficiary registration and monitoring (leveraging UID/Aadhaar), then pursue better targeting (e.g., withdraw entitlements from APL households), adopt automatic pricing with smoothing, and consider replacing in-kind subsidies with cash transfers.
- Targeting systems:
  - Developing a better targeting system is key to medium-term subsidy reduction.
  - UID will help eliminate duplicate and ghost beneficiaries but is not a substitute for an effective targeting framework; current BPL/APL identification needs reform to reduce undercoverage of poor households and leakage to non-poor households.
  - Pilot programs are recommended to refine design and implementation before nationwide scaling-up.

*Source: IMF Staff estimates and analysis as presented in the provided content.*

### References

### _wp13128 - References

### Key References Cited
- Anand, M. K. (2012), “Diesel Pricing in India: Entangled in Policy Maze,” Working Paper No. 2012-108, National Institute of Public Finance and Policy. Available at: www.nipfp.org.in/newweb/sites/default/files/WP_2012_108.pdf
- Arze del Granado, J., D. Coady and R. Gillingham (2012), “The Unequal Benefits of Fuel Subsidies: A Review of Evidence for Developing Countries”, World Development, 40(11), pp2234-48, Elsevier.
- Bhanumurthy, N.R., S. Das, and S. Bose (2012). “Oil Price Shock, Pass-Through Policy, and its Impact on India.” National Institute of Public Finance and Policy Working Paper No. 2012-99.
- Coady, D., and D. Newhouse, 2006, “Evaluating the distribution of the real income effects of increases in petroleum product prices in Ghana”, in A. Coudouel, A. Dani and S. Paternostro, eds., Analyzing the distributional impacts of reforms: Operational experience in implementing poverty and social impact analysis, World Bank, Washington DC, 2006.
- Fiszbein, A., and N. Schady, 2009, “Conditional Cash Transfers: Reducing Present and Future Poverty,” Policy Research Report (Washington: World Bank).
- International Monetary Fund, 2013, Energy Subsidy Reform: Lessons and Implications, forthcoming, Washington DC.
- Jha, S., and B. Ramaswami, 2011, “The Percolation of Public expenditure: Food Subsidies and the Poor in India and the Philippines”, Background Paper for 2011 India Policy Forum.
- Kelkar Committee Report, 2012, Report of the Committee on Roadmap fpr Fiscal Consolidation, Government of India.
- Lahoti, R., J.Y. Suchitra, and P. Goutam (2012). “Subsidies for Whom: The Case of LPG in India,” Economic and Political Weekly. Vol. XLVII, no. 44, November 3, 2012.
- Lang, K., and P. Wooders, 2012, “India’s Fuel Subsidies: Policy Recommendations for Reform”, Global Subsidies Initiative, International Institute for Sustainable Development.
- Nilekani, S. N., 2011, Interim Report of the Task Force on Direct Transfers of Subsidies on Kerosene, LPG and Fertilizer, Government of India.
- Parikh Committee Report, 2010, Report of the Expert Group on a Viable and Sustainable System of Pricing of Petroleum Products, Government of India.
- Parikh, K., J. Parikh, P. Ghosh, R. Panda, and A. Kaur, 2012, “Taming Diesel Subsidy to Curtail Inflation and Foster Economic Growth”, Integrated Research and Action for Development, October, India.
- Rangarajan C, Parikh KS, Chaudhuri S, Lahiri A, Dholakia B, Srinivasan M, Tripathi S., 2006, Report of the Committee on Pricing and Taxation of Petroleum Products, Technical Report, Government of India. Available at: http://petroleum.nic.in/Report1.pdf
- Tapsoba, Jules (2012). “Strategies for Fiscal Consolidation in India,” forthcoming IMF Working Paper.
- The Energy Resources Institute (TERI), 2012, “Fossil Fuel Subsidy Reform in India: Cash Transfers for PDS Kerosene and Domestic LPG”. Available at: http://www.iisd.org/publications/pub.aspx?pno=1656
- World Bank, 2011, Social Protection for a Changing India (Volume I), Washington DC.

### Appendix Figure 1 — Key players in India oil industry
- Ministry of Petroleum & Natural Gas
- Upstream / Exploration & Production: ONGC, OIL, Private Co. (Cairn, RIL)
- Downstream / Refining & Marketing: HPCL, BPCL, IOCL, CPCL, RIL
- Distribution: GAIL India, RIL
- Source noted: Ministry of Petroleum & Natural Gas

### Appendix Table 1 — Price Structure of Petroleum Products as of November 1, 2012 (Rs/liter-cylinder)
- Column headers: Diesel | PDS Kerosene | LPG (14.2 kg)
- C&F price: 43.46 | 43.27 | 71.82
- Import charges (Insurance, port, ...): 0.4 | 0.29 | 6.31
- Customs duty: 1.13 | -- | --
- Import parity price: 44.99 | -- | --
- Export parity price: 42.9 | -- | --
- Trade parity price 1/: 44.57 | 43.56 | 78.12
  - Note: The trade parity price for diesel is an average of import (80%) and export prices (20%).
- Premium recovered: 0.04 | -- | --
- Inland freight and delivery: 0.85 | 0.76 | 8.31
- Marketing cost of OMCs: 0.67 | 0.41 | 12.58
- Marketing margin of OMCs: 0.72 | 0.33 | 6.74
- Bottling charges: -- | -- | 38.68
- Total desired price: 46.85 | 45.06 | 74.44
- Under-recovery of OMCs: 9.84 | 31.34 | 78.45
- Subsidies: -0.82 | 22.58 | --
- Price charged to dealers: 37.01 | 12.96 | 73.41
- Excise duty: 3.56 | -- | --
- Dealer commission: 1.09 | 1.13 | 7.25
- VAT 2/: 5.49 | 0.7 | -
  - Note: The VAT rate is 12.5 percent for diesel.
- Retail selling price: 47.15 | 14.79 | 410.5
- Memo:
  - Total taxes: 10.18 | 0.70 | -
  - Total margins: 3.37 | 2.63 | 133.56
  - Required price increase (Rs): 9.84 | 32.12 | 501.03
  - Price increase (%): 20.92 | 17.41 | 22.1
- Source: Indian authorities.

### Appendix Figure 2 — Composition of Indirect Welfare Impact
- Chart categories: Transport | Non-food | Food
- Labels indicate percent share of total indirect impact (visual scale 0% to 100%)
- Source: IMF Staff estimates based on the Indian 2009/10 Household Socio-Economic Survey

### Appendix A — Description of data
- Household Survey Data:
  - National Sample Survey (NSS), sixty-sixth round, July 2009-June 2010.
  - Collected household consumer expenditures and employment/unemployment data.
  - Used to generate average household monthly per capita expenditure (MPCE), distribution over households and individuals, and breakdown over commodities.
  - Survey spans nearly 13,000 rural and urban sampling units and more than 100,000 households.
  - MPCE estimates considered reasonably robust at national, urban/rural, and large-state levels; inform apex planning, resource allocation, CPI construction, and assessments of living standards, poverty, inequality, and inclusiveness.
- Input-Output Data:
  - Central Statistical Office (CSO) Input Output Transactions Tables (IOTT) published for 1973–74, 1978–79, 1983–84, 1989–90, 1993–94, 1998–99, 2003–04, and 2006–07.
  - 2003–04 tables at 130-sector classification; IOTT record flows of commodities at factor cost among sectors.
  - Sources: National Accounts Statistics for primary and tertiary sectors; Annual Survey of Industries (ASI) for formal manufacturing; NSSO 2000–01 unorganized manufacturing survey for informal sector.
  - Paper uses the “absorbtion matrix” from 2003–04, transformed into the Leontief coefficient matrix by dividing input costs for each industry by total gross output; used to estimate impact of fuel subsidy reform on prices.

### Appendix B — Methodology for estimating indirect price effects
- Overview:
  - Indirect welfare effect of higher fuel prices arises because households consume goods and services that use fuel as inputs.
  - Methodology modeled after Coady and Newhouse (2006).
  - Requires a model of price-shifting behavior to identify pass-through of higher fuel costs to consumer prices.
- Commodity grouping:
  - (i) Cost-Push Sectors: higher input costs are passed on fully to output prices (loosely non-traded).
  - (ii) Controlled Sectors: output prices are controlled by government; burden of higher costs depends on government adjustments to controlled prices.
  - Aggregate commodity categories are assumed composed of proportions α (cost-push) and β (controlled), with 0≤(α, β)≤1 and α+β=1.
- Production technology and notation:
  - Input-output coefficient matrix A with typical element a_ij denoting cost of input i in producing one unit of output j.
  - Price changes interpreted as percentage changes with unit user price normalization.
- Controlled sectors:
  - Producer prices determined by pricing controls and assumed equal to user (consumer) prices so that:
    - p~ = q~  (equation as in source)
  - Formula for price changes:
    - Δp~ = Δq~  (equation as in source)
  - Right-hand side specified exogenously in the reform package.
- Cost-push sectors:
  - Relationship between user and producer prices:
    - q_c = p_c + t_c  (equation as in source)
      - where q_c is price paid by users, p_c is price received by producers, t_c is sales or excise taxes.
  - Producer prices determined as:
    - p_c = p_c(w, q)  (equation as in source; written in source as ),(wqpp cc = )
  - For these sectors, cost increases are assumed fully pushed forward onto user prices with factor payments fixed. Assuming taxes do not change:
    - Δq_c = Δp_c  (equation as in source)
  - Using the input-output coefficient matrix and fixed factor prices, change in producer prices derived as:
    - Ap...~.. β α Δ + Δ = Δ p_c  (equation represented in source as ApAqp cc ..~.. β α Δ + Δ = Δ)
  - Substituting from prior relation yields:
    - ApApp cc ..~.. β α Δ + Δ = Δ
    - so that:
    - VApp c ...~ β Δ = Δ  (equation as in source)
      - where V = (I - α·A)^-1 and I is nxn identity matrix; typical element v_ij captures combined direct and indirect use of cost-push sector i to produce one unit of cost-push sector j.
  - Change in sector aggregate prices:
    - Δq = Δq~ β α + Δ  (represented in source as qqq c ~.. Δ + Δ = Δ β α)
- Implementation notes:
  - All price-change vectors are nx1 row vectors where n is number of commodity groups.
  - (α, β) treated as nxn diagonal matrices in the algebra above.
  - Leontief/inverse matrix V used to capture direct and indirect inter-sectoral linkages and pass-through of fuel cost increases to consumer prices.

*Source: _wp13128 - References (source PDF content as provided).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13128.pdf_