## wp18174

## Source details

**Canonical URL:** [wp18174](https://www.imf.org/-/media/files/publications/wp/2018/wp18174.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18174.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18174.pdf.json)

---

### Major motivations for a UBI
- Two common arguments motivating adoption of a UBI:
  - A UBI can be a more effective way of supporting low-income households when existing income support programs are inefficient and the source of these inefficiencies (such as administrative constraints) cannot be easily addressed over the short term.
  - A UBI can help generate public and political support for implementing structural reforms (for example, increasing energy prices) by mitigating the adverse impact of these reforms on households, especially low-income and middle-income households.

### Scope, data, and design assumptions
- Analysis anchored in 2011 using the Indian 2011–12 National Sample Survey (NSS).
- Focus on fiscally neutral reforms to highlight trade-offs between policy objectives and abstract from potential crowding out of other public expenditures.
- UBI design in simulations:
  - Every individual receives an unconditional uniform cash transfer.
  - The common transfer level is set to fully exhaust the fiscal gains from subsidy reforms.

### Public Distribution System (PDS): background and institutional features
- Origin and evolution:
  - Dates back to the 1960s; initially targeted urban consumers via ration cards for fixed quantities of food and fuel (kerosene) at subsidized prices through Fair Price Shops (FPSs).
  - 1992: Revamped Public Distribution System (RPDS) launched to strengthen coverage in remote areas.
  - 1997: Targeted Public Distribution System (TPDS) replaced PDS; introduced APL (above-poverty-line) and BPL (below-poverty-line) cards.
  - 2001: "Antyodaya Anna Yojana" (AAY) added to reach the poorest of the poor with larger benefits.
  - 2013: National Food Security Act (NFSA) transformed TPDS toward a rights-based approach.
- NFSA entitlements and coverage targets:
  - Cover up to 75 percent of the rural population and up to 50 percent of the urban population.
  - Every eligible beneficiary entitled to receive five kilograms of food grain at a highly subsidized price, with higher entitlements for AAY families.
- Operational responsibilities:
  - Central Government (including the Food Corporation of India) is responsible for procurement, storage, transportation and bulk allocation of wheat, rice, sugar and kerosene to State Governments; sets quantities, quotas, and minimum prices.
  - State Governments are responsible for actual procurement, identification of eligible families, issuing ration cards, setting ration prices, and supervision of FPSs.

### PDS implementation variation and systemic inefficiencies
- State-level divergence in PDS implementation prior to 2013:
  - "90 percent of households consume rice from the PDS in Tamil Nadu" where the state enforces a near-universal PDS.
  - "around 45 percent of households consume PDS rice in Maharashtra" where the state enforces stricter eligibility.
  - State-level implicit income transfers can vary substantially: "around 220 rupees/month/household in Tamil Nadu" (state PDS grain was free in 2011) versus "around 60 rupees/month/household in Maharashtra."
- Documented systemic inefficiencies and leakages:
  - "36 percent of total PDS allocation never reaches final beneficiaries" because of “out-of-system” leakages (Ministry of Finance, Government of India, 2017a).
  - Example for kerosene: "41 percent gap between the total PDS subsidized kerosene allocation by the Central Government and actual household consumption" as captured by the 2011–12 NSS (Ministry of Finance, Government of India, 2016).
- Coverage and targeting shortfalls:
  - "approximately 20 percent of households in each of the bottom two income quintiles do not receive any benefits" (NSS 2011–12).
  - "the richest 40 percent of households receiving 35 percent of total PDS subsidies."
  - Because of under-coverage and leakage, "most income deciles receive similar shares of total PDS benefits," an outcome comparable to a UBI program.

### Analysis of substituting a UBI for the PDS — coverage, targeting, and distributional impacts
- A UBI would "outperform the 2011 TPDS program in terms of coverage of lower income groups" because it covers all households by design.
- Improved coverage under a UBI would cause "a slight deterioration in benefit targeting and generosity at the bottom" and "leave benefit progressivity virtually unchanged."
- Distributional impacts:
  - "On average, 50 percent of households in the bottom four income deciles would face a 6 percent welfare loss, while the other 50 percent would gain 3 percent."
  - Households with the largest losses are those with the largest PDS benefits by design.
- Potential mitigation via efficiency savings:
  - Estimated "36 percent of PDS total spending never reaches the intended households" (Ministry of Finance, Government of India, 2017a): "out of every 100 Rupees spent on the program, only 64 reaches households."
  - Recycling these savings into the UBI: the 36 percent leakage implies a required increase in the uniform transfer of "55 percent" (i.e., "36 divided by 64").
  - Effects of recycling efficiency gains:
    - Reduces the number of losers in the bottom four income deciles "to below 40 percent," bringing down the average share of losers in the bottom four income deciles "from about one half to a third."
    - Reinforces average gains for other households in the bottom four deciles "from 3 to 5 percent."
- Targeting by excluding top deciles or fiscal measures:
  - "Reallocating UBI transfers going to households in the top three income deciles, on top of the efficiency gains, would result in a 42 percent increase in the per capita level of UBI for the rest of the population and reduce the proportion of losers in the bottom four deciles by 28 percent (from an average of 50 to 22 percent)."
  - Alternative: increase income tax rates at the top of the income distribution.
- Characteristics of likely losers and possible complementary measures:
  - Losing households in the bottom four deciles receive "around 45 percent higher" PDS subsidies than other beneficiary households.
  - Higher probability of being a loser in specific states: Delhi, Manipur, Gujarat, Daman and Diu, and Nagaland.
  - Urban and large households with a greater share of youth and elderly have higher probability of losing.
  - Proposed complementary programs: targeted conditional cash transfer (CCT) programs linking eligibility to investments in education, health, and nutrition of children.

### Administrative and political implementation considerations; lessons from LPG reform
- Administrative challenges:
  - Cash transfers likely simpler to administer than the PDS, but require "ensuring an effective network for transferring money directly to households across the country."
  - Need to account for "increased take-up costs for households in remote areas without bank access, and transition costs (e.g., disruption, learning about a new system, changes in consumption behaviors, etc.)" in communication and complementary measures.
- Aadhaar and DBT as enablers:
  - Improved identification technology (Aadhaar) and direct benefit transfer (DBT) systems could help realize efficiency savings and eliminate ghost beneficiaries.
- LPG (PAHAL/DBTL) pilot and rollout:
  - Reform replaced subsidized LPG cylinder distribution with a cash transfer equal to the price difference between market and subsidized prices, paid to bank accounts.
  - "By 2016–7 the share of LPG subsidies allocated through the DBTL program was reported to be 58 percent."
  - Policy to encourage well-off households to forgo subsidies ("give-it-up" movement) created initial fiscal space, though some consumers switch back as unsubsidized LPG prices increase.
  - DBTL highlights differences between commodity-specific reforms (need to identify both consumers and quantities) and a universal cash scheme.

### Energy pricing, subsidy incidence, and UBI substitution for energy subsidies
- Efficient energy pricing concept:
  - Consumers should face a price reflecting: supply cost, a standard consumption tax (revenue-raising), and a Pigouvian tax to internalize externalities.
  - Estimated required price increases to reach efficient energy prices in India (Coady and Hanedar, 2016):
    - "gasoline (67 percent), diesel (69 percent), kerosene (10 percent), LPG (67 percent) and coal (455 percent)."
- Incidence and regressivity of energy subsidies (based on 2011 consumption patterns):
  - "On average, household subsidy benefits are equivalent to around 10 percent of their total expenditures," with distribution:
    - "11.5 percent of total expenditures in top three income deciles" on average.
    - "9 percent in bottom three income deciles" on average.
  - Composition and distributional patterns:
    - Indirect benefits from low coal prices: "relatively neutrally distributed across deciles."
    - Indirect benefits from low electricity prices: "regressive."
    - Indirect benefits from low diesel prices: "progressive."
    - "Around one quarter of total subsidy benefits come through the direct effect and are highly regressive."
- Redistribution potential of replacing energy subsidies with a UBI:
  - Under current subsidies: bottom four income deciles receive "17 percent of total energy subsidies"; top four income deciles receive "69 percent."
  - Replacing these subsidies with a UBI in a budget-neutral way would yield:
    - Bottom three income deciles would receive "30 percent of UBI transfers" versus "only 9 percent of energy subsidies" under the current system.
    - For bottom three income deciles, a UBI would represent "25 percent of their total expenditure," versus "around 9 percent" under current energy subsidies.
  - Gains and losses:
    - "Over 90 percent of households in each of the bottom four income deciles gain from the switch to a UBI."
    - A small share of households lose, and their losses are sizeable: "On average, 2 percent of households in the bottom three income deciles would incur a substantial average welfare loss of 11 percent."
    - These losing households are very energy intensive; consumption adjustments and complementary policies to promote efficient energy use can reduce welfare losses.

### Recent energy subsidy reforms and policy context
- Recent policy moves:
  - "Gasoline prices have been liberalized since 2010."
  - Since 2013 diesel prices were increased gradually until reaching international parity and "have been fully liberalized since January 2015."
  - Since June 2017, gasoline and diesel prices at the pump are "changed daily according to a revised automatic pricing formula."
  - Excise duties on petrol and diesel have been "increased nine times since November 2014."
  - Proposal to bring petrol and diesel under GST that was introduced in July 2017.
- LPG subsidy transition:
  - Government committed in 2017 "to gradually phase out LPG subsidies by April 2018 for domestic consumers, by increasing the cost of LPG cylinders by four rupees/month" (Ministry of Finance, Government of India 2017b).
  - All households are entitled to a DBTL cash transfer "corresponding to the price difference between the subsidized and unsubsidized LPG cylinder prices, up to twelve cylinders a year."
- Use of JAM (Jan Dhan–Aadhaar–Mobile) and digitalization:
  - Indian authorities assessed using digitalization and increased financial inclusion to substitute cash transfers for food and energy subsidies; priority to subsidies "with high levels of leakages" and "a high degree of central government control."
- Robustness of distributional conclusions:
  - Distributional results "would still hold true even if the fiscal space freed through terminating the PDS program and energy subsidies would be smaller."

### Implementation constraints and political economy considerations
- Centralized cash transfer to individuals requires identification, banking coverage, and effective access to cash after transfer.
- Since 2014, Indian authorities have launched large-scale initiatives commonly referred to as the “JAM trinity”:
  - Aadhaar biometric ID system covering nearly 99 percent of the population aged 18 and over in 2017.
  - Jan Dhan scheme helping households open a bank account, reaching an average banking coverage of 46 percent across Indian States, lower in rural states.
  - Mobile penetration – average penetration lower than 56 percent only in 2 states.
- Additional implementation and political challenges:
  - There may be several policy priorities (for example, to increase public spending in health or education, or to reduce the fiscal deficit) competing for financing. The desirability and level of a UBI in such a context of scarce resources has to be assessed against other governmental priorities and the potential for greater revenue mobilization.
  - Many vested interests in the status quo can make it difficult to reform long-lasting and well-known programs.
  - Direct cash transfers may be difficult to implement and the risk of excluding poor households remains, even under universal schemes.
  - Phasing in a UBI through the gradual introduction of categorical cash transfers such as social pensions and child benefits could be an efficient way to transition from subsidies to (broad) cash transfers (Khera 2016 and Drèze 2017).

### Summary and conclusions (high-level findings)
- Two reform contexts analyzed (both designed to be budget neutral):
  - Replacing the Public Distribution System (PDS) with a UBI (PDS provides price subsidies for wheat, rice, sugar and kerosene consumption).
  - Introducing a UBI as part of a structural reform program centered around increasing energy prices to efficient levels that reflect the true social cost of energy consumption.
- Findings: replacing PDS with a UBI
  - Replacement would help address under-coverage of low income households under the PDS.
  - Trade-offs and distributional effects:
    - The gain in coverage would come at the expense of a slight increase in leakage of benefits to higher income groups.
    - A sizeable percentage of existing PDS beneficiaries would lose from the reform, including many low-income households.
    - The number of losers and the magnitude of their losses could be reduced by:
      - Recycling efficiency gains from avoiding “out-of-system” losses throughout procurement, storage and distribution as a higher UBI benefit; or
      - Excluding higher income groups from the UBI and recycling the savings; and/or
      - Using other more targeted programs to protect poor households adversely impacted by the reform.
- Findings: replacing energy subsidies with a UBI
  - Replacing inefficient energy subsidies—raising domestic energy prices to efficient levels—would deliver:
    - Unambiguous distributional gains.
    - Strong incentives for improving energy efficiency with associated environmental and health gains.
  - Reasoning and distributional impacts:
    - Very high leakage of benefits under universal energy price subsidies means equal sharing of these subsidies under the UBI would deliver significant income gains for low-income households, financed by losses for the highest income households.
    - Welfare gains will be even higher to the extent that higher energy prices incentivize decreased wasteful energy use and improved energy efficiency.
    - Only a few low-income households lose from the reform, albeit some significantly; to the extent such losses reflect wasteful use of energy, higher energy prices provide a strong incentive to reduce waste.
    - Complementary programs can focus on facilitating energy-saving behaviors by consumers, especially low-income households.

### Data and methodology notes
- PDS and energy subsidy estimates and incidence analysis are based on data from the most recent publicly available National Sample Survey (2011–12 NSS) which:
  - Covers the whole of the Indian Union.
  - Records data on expenditure of more than 100,000 households sampled for the survey.
- For analysis:
  - Expenditure is used as a proxy for income.
  - Household expenditure is divided by the number of household members to obtain the welfare indicator (expenditure per capita).
  - Deciles are based on per capita household expenditure, with each decile representing 10 percent of the population.
  - Subsidy estimates assume that demand does not respond to price changes and therefore estimates should be interpreted as short-term welfare effects following price increases.
- Methodology for subsidy computations:
  - PDS Subsidy = PDS quantity x (market price – PDS price).
  - For energy: Direct Effect = Budget Share x Percentage Price Increase.
  - Indirect effects for non-energy goods reflect increases in diesel and coal prices estimated using the Coady and Newhouse (2006) model, which assumes increases in energy production costs are fully passed forward onto domestic output prices of goods and services.
  - Total energy subsidies are the sum of direct and indirect effects.

### Key quantitative incidence and budget-share statistics (NSS 2011–12 based)
- PDS and market budget shares for rice, wheat, sugar, and kerosene (Percent of total expenditure – all households; decile values shown):
  - Decile 1: Rice PDS 1.8%, Rice non-PDS 9.2%; Wheat PDS 0.8%, Wheat non-PDS 4.8%; Sugar PDS 0.4%, Sugar non-PDS 1.6%; Kerosene PDS 1.3%, Kerosene non-PDS 0.4%.
  - Decile 2: Rice PDS 1.2%, Rice non-PDS 8.8%; Wheat PDS 0.6%, Wheat non-PDS 4.3%; Sugar PDS 0.3%, Sugar non-PDS 1.7%; Kerosene PDS 1.0%, Kerosene non-PDS 0.4%.
  - Decile 3: Rice PDS 1.1%, Rice non-PDS 8.3%; Wheat PDS 0.5%, Wheat non-PDS 4.0%; Sugar PDS 0.2%, Sugar non-PDS 1.7%; Kerosene PDS 0.9%, Kerosene non-PDS 0.4%.
  - Decile 4: Rice PDS 0.9%, Rice non-PDS 7.7%; Wheat PDS 0.4%, Wheat non-PDS 4.0%; Sugar PDS 0.2%, Sugar non-PDS 1.7%; Kerosene PDS 0.8%, Kerosene non-PDS 0.4%.
  - Decile 5: Rice PDS 0.8%, Rice non-PDS 6.9%; Wheat PDS 0.4%, Wheat non-PDS 3.7%; Sugar PDS 0.2%, Sugar non-PDS 1.7%; Kerosene PDS 0.7%, Kerosene non-PDS 0.3%.
  - Decile 6: Rice PDS 0.7%, Rice non-PDS 6.5%; Wheat PDS 0.3%, Wheat non-PDS 3.4%; Sugar PDS 0.2%, Sugar non-PDS 1.6%; Kerosene PDS 0.6%, Kerosene non-PDS 0.3%.
  - Decile 7: Rice PDS 0.6%, Rice non-PDS 6.0%; Wheat PDS 0.2%, Wheat non-PDS 3.1%; Sugar PDS 0.1%, Sugar non-PDS 1.5%; Kerosene PDS 0.5%, Kerosene non-PDS 0.3%.
  - Decile 8: Rice PDS 0.5%, Rice non-PDS 5.3%; Wheat PDS 0.2%, Wheat non-PDS 2.8%; Sugar PDS 0.1%, Sugar non-PDS 1.4%; Kerosene PDS 0.4%, Kerosene non-PDS 0.3%.
  - Decile 9: Rice PDS 0.3%, Rice non-PDS 4.5%; Wheat PDS 0.1%, Wheat non-PDS 2.3%; Sugar PDS 0.1%, Sugar non-PDS 1.2%; Kerosene PDS 0.3%, Kerosene non-PDS 0.4%.
  - Decile 10: Rice PDS 0.1%, Rice non-PDS 2.8%; Wheat PDS 0.1%, Wheat non-PDS 1.4%; Sugar PDS 0.0%, Sugar non-PDS 0.7%; Kerosene PDS 0.1%, Kerosene non-PDS 0.3%.
  - Total (all deciles combined): Rice PDS 0.7%, Rice non-PDS 6.2%; Wheat PDS 0.3%, Wheat non-PDS 3.2%; Sugar PDS 0.2%, Sugar non-PDS 1.4%; Kerosene PDS 0.6%, Kerosene non-PDS 0.4%.

- Average household budget shares for energy products (Percent of total expenditure – all households; decile values shown):
  - Decile 1: electricity 1.9%, PDS kerosene 1.3%, kerosene 0.5%, coal 0.1%, LPG 0.4%, petrol 0.1%, diesel 0.0%.
  - Decile 2: electricity 2.0%, PDS kerosene 1.0%, kerosene 0.4%, coal 0.1%, LPG 0.6%, petrol 0.3%, diesel 0.0%.
  - Decile 3: electricity 2.1%, PDS kerosene 0.9%, kerosene 0.4%, coal 0.1%, LPG 0.7%, petrol 0.4%, diesel 0.0%.
  - Decile 4: electricity 2.3%, PDS kerosene 0.8%, kerosene 0.4%, coal 0.1%, LPG 0.9%, petrol 0.6%, diesel 0.0%.
  - Decile 5: electricity 2.4%, PDS kerosene 0.7%, kerosene 0.3%, coal 0.1%, LPG 1.1%, petrol 0.8%, diesel 0.0%.
  - Decile 6: electricity 2.5%, PDS kerosene 0.6%, kerosene 0.3%, coal 0.1%, LPG 1.3%, petrol 1.1%, diesel 0.0%.
  - Decile 7: electricity 2.7%, PDS kerosene 0.5%, kerosene 0.3%, coal 0.1%, LPG 1.5%, petrol 1.5%, diesel 0.1%.
  - Decile 8: electricity 2.9%, PDS kerosene 0.4%, kerosene 0.3%, coal 0.1%, LPG 1.9%, petrol 2.1%, diesel 0.1%.
  - Decile 9: electricity 3.0%, PDS kerosene 0.3%, kerosene 0.4%, coal 0.0%, LPG 2.1%, petrol 2.7%, diesel 0.1%.
  - Decile 10: electricity 3.0%, PDS kerosene 0.1%, kerosene 0.3%, coal 0.0%, LPG 1.7%, petrol 3.7%, diesel 0.3%.
  - Total (all deciles combined): electricity 2.5%, PDS kerosene 0.6%, kerosene 0.4%, coal 0.1%, LPG 1.3%, petrol 1.6%, diesel 0.1%.

*Source: wp18174 - introduction of a UBI in India that this paper does not address (IMF working paper, based on Indian 2011–12 NSS).*

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

### wp18174 - References

### Major motivations for a UBI (as discussed in the paper)
- Two common arguments motivating adoption of a UBI:
  - A UBI can be a more effective way of supporting low-income households when existing income support programs are inefficient and the source of these inefficiencies (such as administrative constraints) cannot be easily addressed over the short term.
  - A UBI can help generate public and political support for implementing structural reforms (for example, increasing energy prices) by mitigating the adverse impact of these reforms on households, especially low-income and middle-income households.

### Scope, data, and design assumptions
- Analysis anchored in 2011 using the Indian 2011–12 National Sample Survey (NSS).
- Focus on fiscally neutral reforms to highlight trade-offs between policy objectives and abstract from potential crowding out of other public expenditures.
- UBI design in simulations:
  - Every individual receives an unconditional uniform cash transfer.
  - The common transfer level is set to fully exhaust the fiscal gains from subsidy reforms.

### Public Distribution System (PDS): background and institutional features
- Origin and evolution:
  - Dates back to the 1960s; initially targeted urban consumers via ration cards for fixed quantities of food and fuel (kerosene) at subsidized prices through Fair Price Shops (FPSs).
  - 1992: Revamped Public Distribution System (RPDS) launched to strengthen coverage in remote areas.
  - 1997: Targeted Public Distribution System (TPDS) replaced PDS; introduced APL (above-poverty-line) and BPL (below-poverty-line) cards.
  - 2001: "Antyodaya Anna Yojana" (AAY) added to reach the poorest of the poor with larger benefits.
  - 2013: National Food Security Act (NFSA) transformed TPDS toward a rights-based approach.
- NFSA entitlements and coverage targets:
  - Cover up to 75 percent of the rural population and up to 50 percent of the urban population.
  - Every eligible beneficiary entitled to receive five kilograms of food grain at a highly subsidized price, with higher entitlements for AAY families.
- Operational responsibilities:
  - Central Government (including the Food Corporation of India) is responsible for procurement, storage, transportation and bulk allocation of wheat, rice, sugar and kerosene to State Governments; sets quantities, quotas, and minimum prices.
  - State Governments are responsible for actual procurement, identification of eligible families, issuing ration cards, setting ration prices, and supervision of FPSs.

### Policy questions and trade-offs analyzed
- The paper evaluates:
  - Whether adopting a UBI could address shortcomings of existing food and energy subsidies (incomplete coverage, leakage to the rich, operational inefficiencies, fraud and corruption).
  - How a UBI could mitigate adverse impacts on households from removing subsidies or raising prices as part of structural reforms (notably energy price reforms reflecting social costs such as domestic pollution, congestion and global warming).
- Emphasis on trade-offs that need careful management when replacing subsidies with a UBI, including administrative and political challenges in ensuring universal coverage.

### Related analytical focus (as presented in the paper)
- Replacement of PDS with UBI: assessment of coverage, targeting, and generosity across deciles (figures referenced).
- Substituting a UBI for energy subsidies and for the PDS: analysis of gains and losses from substitution, including scenarios with efficiency gains.
- Estimation and decomposition of PDS and energy subsidy effects at household level (appendix tables and figures referenced).

*Source: wp18174 - References (excerpt).

### introduction of the NFSA.

### introduction of the NFSA.

### PDS implementation variation and systemic inefficiencies
- State-level divergence in PDS implementation prior to 2013:
  - "90 percent of households consume rice from the PDS in Tamil Nadu" where the state enforces a near-universal PDS.
  - "around 45 percent of households consume PDS rice in Maharashtra" where the state enforces stricter eligibility.
  - State-level implicit income transfers can vary substantially: "around 220 rupees/month/household in Tamil Nadu" (state PDS grain was free in 2011) versus "around 60 rupees/month/household in Maharashtra."
- Documented systemic inefficiencies and leakages:
  - "36 percent of total PDS allocation never reaches final beneficiaries" because of “out-of-system” leakages (Ministry of Finance, Government of India, 2017a).
  - Example for kerosene: "41 percent gap between the total PDS subsidized kerosene allocation by the Central Government and actual household consumption" as captured by the 2011–12 NSS (Ministry of Finance, Government of India, 2016).
- Coverage and targeting shortfalls:
  - "approximately 20 percent of households in each of the bottom two income quintiles do not receive any benefits" (NSS 2011–12).
  - "the richest 40 percent of households receiving 35 percent of total PDS subsidies."
  - Because of under-coverage and leakage, "most income deciles receive similar shares of total PDS benefits," an outcome comparable to a UBI program.

### Analysis of substituting a UBI for the PDS
- Coverage, targeting, and progressivity comparisons (based on Indian 2011–12 NSS):
  - A UBI would "outperform the 2011 TPDS program in terms of coverage of lower income groups" because it covers all households by design.
  - Improved coverage under a UBI would cause "a slight deterioration in benefit targeting and generosity at the bottom" and "leave benefit progressivity virtually unchanged."
- Distributional impacts:
  - "On average, 50 percent of households in the bottom four income deciles would face a 6 percent welfare loss, while the other 50 percent would gain 3 percent."
  - Households with the largest losses are those with the largest PDS benefits by design.
- Potential mitigation via efficiency savings:
  - Estimated "36 percent of PDS total spending never reaches the intended households" (Ministry of Finance, Government of India, 2017a): "out of every 100 Rupees spent on the program, only 64 reaches households."
  - Recycling these savings into the UBI: the 36 percent leakage implies a required increase in the uniform transfer of "55 percent" (i.e., "36 divided by 64").
  - Effects of recycling efficiency gains:
    - Reduces the number of losers in the bottom four income deciles "to below 40 percent," bringing down the average share of losers in the bottom four income deciles "from about one half to a third."
    - Reinforces average gains for other households in the bottom four deciles "from 3 to 5 percent."
- Targeting by excluding top deciles or fiscal measures:
  - "Reallocating UBI transfers going to households in the top three income deciles, on top of the efficiency gains, would result in a 42 percent increase in the per capita level of UBI for the rest of the population and reduce the proportion of losers in the bottom four deciles by 28 percent (from an average of 50 to 22 percent)."
  - Alternative: increase income tax rates at the top of the income distribution.
- Characteristics of likely losers and possible complementary measures:
  - Losing households in the bottom four deciles receive "around 45 percent higher" PDS subsidies than other beneficiary households.
  - Higher probability of being a loser in specific states: Delhi, Manipur, Gujarat, Daman and Diu, and Nagaland.
  - Urban and large households with a greater share of youth and elderly have higher probability of losing.
  - Proposed complementary programs: targeted conditional cash transfer (CCT) programs linking eligibility to investments in education, health, and nutrition of children.

### Administrative and political implementation considerations; lessons from LPG reform
- Administrative challenges:
  - Cash transfers likely simpler to administer than the PDS, but require "ensuring an effective network for transferring money directly to households across the country."
  - Need to account for "increased take-up costs for households in remote areas without bank access, and transition costs (e.g., disruption, learning about a new system, changes in consumption behaviors, etc.)" in communication and complementary measures.
- Aadhaar and DBT as enablers:
  - Improved identification technology (Aadhaar) and direct benefit transfer (DBT) systems could help realize efficiency savings and eliminate ghost beneficiaries.
- LPG (PAHAL/DBTL) pilot and rollout:
  - Reform replaced subsidized LPG cylinder distribution with a cash transfer equal to the price difference between market and subsidized prices, paid to bank accounts.
  - "By 2016–7 the share of LPG subsidies allocated through the DBTL program was reported to be 58 percent."
  - Policy to encourage well-off households to forgo subsidies ("give-it-up" movement) created initial fiscal space, though some consumers switch back as unsubsidized LPG prices increase.
  - DBTL highlights differences between commodity-specific reforms (need to identify both consumers and quantities) and a universal cash scheme.

### Energy pricing, subsidy incidence, and UBI substitution for energy subsidies
- Efficient energy pricing concept:
  - Consumers should face a price reflecting: supply cost, a standard consumption tax (revenue-raising), and a Pigouvian tax to internalize externalities.
  - Estimated required price increases to reach efficient energy prices in India (Coady and Hanedar, 2016):
    - "gasoline (67 percent), diesel (69 percent), kerosene (10 percent), LPG (67 percent) and coal (455 percent)."
- Incidence and regressivity of energy subsidies (based on 2011 consumption patterns):
  - "On average, household subsidy benefits are equivalent to around 10 percent of their total expenditures," with distribution:
    - "11.5 percent of total expenditures in top three income deciles" on average.
    - "9 percent in bottom three income deciles" on average.
  - Composition and distributional patterns:
    - Indirect benefits from low coal prices: "relatively neutrally distributed across deciles."
    - Indirect benefits from low electricity prices: "regressive."
    - Indirect benefits from low diesel prices: "progressive."
    - "Around one quarter of total subsidy benefits come through the direct effect and are highly regressive."
- Redistribution potential of replacing energy subsidies with a UBI:
  - Under current subsidies: bottom four income deciles receive "17 percent of total energy subsidies"; top four income deciles receive "69 percent."
  - Replacing these subsidies with a UBI in a budget-neutral way would yield:
    - Bottom three income deciles would receive "30 percent of UBI transfers" versus "only 9 percent of energy subsidies" under the current system.
    - For bottom three income deciles, a UBI would represent "25 percent of their total expenditure," versus "around 9 percent" under current energy subsidies.
  - Gains and losses:
    - "Over 90 percent of households in each of the bottom four income deciles gain from the switch to a UBI."
    - A small share of households lose, and their losses are sizeable: "On average, 2 percent of households in the bottom three income deciles would incur a substantial average welfare loss of 11 percent."
    - These losing households are very energy intensive; consumption adjustments and complementary policies to promote efficient energy use can reduce welfare losses.

### Recent energy subsidy reforms and policy context
- Recent policy moves:
  - "Gasoline prices have been liberalized since 2010."
  - Since 2013 diesel prices were increased gradually until reaching international parity and "have been fully liberalized since January 2015."
  - Since June 2017, gasoline and diesel prices at the pump are "changed daily according to a revised automatic pricing formula."
  - Excise duties on petrol and diesel have been "increased nine times since November 2014."
  - Proposal to bring petrol and diesel under GST that was introduced in July 2017.
- LPG subsidy transition:
  - Government committed in 2017 "to gradually phase out LPG subsidies by April 2018 for domestic consumers, by increasing the cost of LPG cylinders by four rupees/month" (Ministry of Finance, Government of India 2017b).
  - All households are entitled to a DBTL cash transfer "corresponding to the price difference between the subsidized and unsubsidized LPG cylinder prices, up to twelve cylinders a year."
- Use of JAM (Jan Dhan–Aadhaar–Mobile) and digitalization:
  - Indian authorities assessed using digitalization and increased financial inclusion to substitute cash transfers for food and energy subsidies; priority to subsidies "with high levels of leakages" and "a high degree of central government control."
- Robustness of distributional conclusions:
  - Distributional results "would still hold true even if the fiscal space freed through terminating the PDS program and energy subsidies would be smaller."

*Source: Authors’ estimates based on Indian 2011–12 NSS.*

### introduction of a UBI in India that this paper does not address

### introduction of a UBI in India that this paper does not address

### Implementation constraints and challenges
- Centralized cash transfer to individuals requires identification, banking coverage, and effective access to cash after transfer.
- Since 2014, Indian authorities have launched large-scale initiatives commonly referred to as the “JAM trinity”:
  - Aadhaar biometric ID system covering nearly 99 percent of the population aged 18 and over in 2017.
  - Jan Dhan scheme helping households open a bank account, reaching an average banking coverage of 46 percent across Indian States, lower in rural states.
  - Mobile penetration – average penetration lower than 56 percent only in 2 states.
- Additional implementation and political challenges:
  - There may be several policy priorities (for example, to increase public spending in health or education, or to reduce the fiscal deficit) competing for financing. The desirability and level of a UBI in such a context of scarce resources has to be assessed against other governmental priorities and the potential for greater revenue mobilization.
  - Many vested interests in the status quo can make it difficult to reform long-lasting and well-known programs.
  - Direct cash transfers may be difficult to implement and the risk of excluding poor households remains, even under universal schemes.
  - Phasing in a UBI through the gradual introduction of categorical cash transfers such as social pensions and child benefits could be an efficient way to transition from subsidies to (broad) cash transfers (Khera 2016 and Drèze 2017).

### Summary and conclusions (high-level findings)
- The paper discusses two common arguments motivating adoption of a UBI:
  - (i) that it can be a more effective way of supporting low-income households when existing income support programs are inefficient; and
  - (ii) that it can play an important role in generating public and political support for the implementation of structural reforms in support of economic growth.
- Empirical analysis uses data for India to illustrate the trade-offs involved; discussion has broader relevance for countries considering similar reforms.
- Two reform contexts analyzed (both designed to be budget neutral):
  - Replacing the Public Distribution System (PDS) with a UBI (PDS provides price subsidies for wheat, rice, sugar and kerosene consumption).
  - Introducing a UBI as part of a structural reform program centered around increasing energy prices to efficient levels that reflect the true social cost of energy consumption.

### Findings: replacing PDS with a UBI
- Replacement would help address under-coverage of low income households under the PDS.
- Trade-offs and distributional effects:
  - The gain in coverage would come at the expense of a slight increase in leakage of benefits to higher income groups.
  - A sizeable percentage of existing PDS beneficiaries would lose from the reform, including many low-income households.
  - The number of losers and the magnitude of their losses could be reduced by:
    - Recycling efficiency gains from avoiding “out-of-system” losses throughout procurement, storage and distribution as a higher UBI benefit; or
    - Excluding higher income groups from the UBI and recycling the savings; and/or
    - Using other more targeted programs to protect poor households adversely impacted by the reform.

### Findings: replacing energy subsidies with a UBI
- Replacing inefficient energy subsidies—raising domestic energy prices to efficient levels—would deliver:
  - Unambiguous distributional gains.
  - Strong incentives for improving energy efficiency with associated environmental and health gains.
- Reasoning and distributional impacts:
  - Very high leakage of benefits under universal energy price subsidies means equal sharing of these subsidies under the UBI would deliver significant income gains for low-income households, financed by losses for the highest income households.
  - Welfare gains will be even higher to the extent that higher energy prices incentivize decreased wasteful energy use and improved energy efficiency.
  - Only a few low-income households lose from the reform, albeit some significantly; to the extent such losses reflect wasteful use of energy, higher energy prices provide a strong incentive to reduce waste.
  - Complementary programs can focus on facilitating energy-saving behaviors by consumers, especially low-income households.

### Data and methodology notes
- PDS and energy subsidy estimates and incidence analysis are based on data from the most recent publicly available National Sample Survey (2011–12 NSS) which:
  - Covers the whole of the Indian Union.
  - Records data on expenditure of more than 100,000 households sampled for the survey.
- For analysis:
  - Expenditure is used as a proxy for income.
  - Household expenditure is divided by the number of household members to obtain the welfare indicator (expenditure per capita).
  - Deciles are based on per capita household expenditure, with each decile representing 10 percent of the population.
  - Subsidy estimates assume that demand does not respond to price changes and therefore estimates should be interpreted as short-term welfare effects following price increases.

### Key quantitative incidence and budget-share statistics (NSS 2011–12 based)
- PDS and market budget shares for rice, wheat, sugar, and kerosene (Percent of total expenditure – all households; decile values shown):
  - Decile 1: Rice PDS 1.8%, Rice non-PDS 9.2%; Wheat PDS 0.8%, Wheat non-PDS 4.8%; Sugar PDS 0.4%, Sugar non-PDS 1.6%; Kerosene PDS 1.3%, Kerosene non-PDS 0.4%.
  - Decile 2: Rice PDS 1.2%, Rice non-PDS 8.8%; Wheat PDS 0.6%, Wheat non-PDS 4.3%; Sugar PDS 0.3%, Sugar non-PDS 1.7%; Kerosene PDS 1.0%, Kerosene non-PDS 0.4%.
  - Decile 3: Rice PDS 1.1%, Rice non-PDS 8.3%; Wheat PDS 0.5%, Wheat non-PDS 4.0%; Sugar PDS 0.2%, Sugar non-PDS 1.7%; Kerosene PDS 0.9%, Kerosene non-PDS 0.4%.
  - Decile 4: Rice PDS 0.9%, Rice non-PDS 7.7%; Wheat PDS 0.4%, Wheat non-PDS 4.0%; Sugar PDS 0.2%, Sugar non-PDS 1.7%; Kerosene PDS 0.8%, Kerosene non-PDS 0.4%.
  - Decile 5: Rice PDS 0.8%, Rice non-PDS 6.9%; Wheat PDS 0.4%, Wheat non-PDS 3.7%; Sugar PDS 0.2%, Sugar non-PDS 1.7%; Kerosene PDS 0.7%, Kerosene non-PDS 0.3%.
  - Decile 6: Rice PDS 0.7%, Rice non-PDS 6.5%; Wheat PDS 0.3%, Wheat non-PDS 3.4%; Sugar PDS 0.2%, Sugar non-PDS 1.6%; Kerosene PDS 0.6%, Kerosene non-PDS 0.3%.
  - Decile 7: Rice PDS 0.6%, Rice non-PDS 6.0%; Wheat PDS 0.2%, Wheat non-PDS 3.1%; Sugar PDS 0.1%, Sugar non-PDS 1.5%; Kerosene PDS 0.5%, Kerosene non-PDS 0.3%.
  - Decile 8: Rice PDS 0.5%, Rice non-PDS 5.3%; Wheat PDS 0.2%, Wheat non-PDS 2.8%; Sugar PDS 0.1%, Sugar non-PDS 1.4%; Kerosene PDS 0.4%, Kerosene non-PDS 0.3%.
  - Decile 9: Rice PDS 0.3%, Rice non-PDS 4.5%; Wheat PDS 0.1%, Wheat non-PDS 2.3%; Sugar PDS 0.1%, Sugar non-PDS 1.2%; Kerosene PDS 0.3%, Kerosene non-PDS 0.4%.
  - Decile 10: Rice PDS 0.1%, Rice non-PDS 2.8%; Wheat PDS 0.1%, Wheat non-PDS 1.4%; Sugar PDS 0.0%, Sugar non-PDS 0.7%; Kerosene PDS 0.1%, Kerosene non-PDS 0.3%.
  - Total (all deciles combined): Rice PDS 0.7%, Rice non-PDS 6.2%; Wheat PDS 0.3%, Wheat non-PDS 3.2%; Sugar PDS 0.2%, Sugar non-PDS 1.4%; Kerosene PDS 0.6%, Kerosene non-PDS 0.4%.

- Average household budget shares for energy products (Percent of total expenditure – all households; decile values shown):
  - Decile 1: electricity 1.9%, PDS kerosene 1.3%, kerosene 0.5%, coal 0.1%, LPG 0.4%, petrol 0.1%, diesel 0.0%.
  - Decile 2: electricity 2.0%, PDS kerosene 1.0%, kerosene 0.4%, coal 0.1%, LPG 0.6%, petrol 0.3%, diesel 0.0%.
  - Decile 3: electricity 2.1%, PDS kerosene 0.9%, kerosene 0.4%, coal 0.1%, LPG 0.7%, petrol 0.4%, diesel 0.0%.
  - Decile 4: electricity 2.3%, PDS kerosene 0.8%, kerosene 0.4%, coal 0.1%, LPG 0.9%, petrol 0.6%, diesel 0.0%.
  - Decile 5: electricity 2.4%, PDS kerosene 0.7%, kerosene 0.3%, coal 0.1%, LPG 1.1%, petrol 0.8%, diesel 0.0%.
  - Decile 6: electricity 2.5%, PDS kerosene 0.6%, kerosene 0.3%, coal 0.1%, LPG 1.3%, petrol 1.1%, diesel 0.0%.
  - Decile 7: electricity 2.7%, PDS kerosene 0.5%, kerosene 0.3%, coal 0.1%, LPG 1.5%, petrol 1.5%, diesel 0.1%.
  - Decile 8: electricity 2.9%, PDS kerosene 0.4%, kerosene 0.3%, coal 0.1%, LPG 1.9%, petrol 2.1%, diesel 0.1%.
  - Decile 9: electricity 3.0%, PDS kerosene 0.3%, kerosene 0.4%, coal 0.0%, LPG 2.1%, petrol 2.7%, diesel 0.1%.
  - Decile 10: electricity 3.0%, PDS kerosene 0.1%, kerosene 0.3%, coal 0.0%, LPG 1.7%, petrol 3.7%, diesel 0.3%.
  - Total (all deciles combined): electricity 2.5%, PDS kerosene 0.6%, kerosene 0.4%, coal 0.1%, LPG 1.3%, petrol 1.6%, diesel 0.1%.

- Methodology for subsidy computations:
  - PDS Subsidy = PDS quantity x (market price – PDS price).
  - For energy: Direct Effect = Budget Share x Percentage Price Increase.
  - Indirect effects for non-energy goods reflect increases in diesel and coal prices estimated using the Coady and Newhouse (2006) model, which assumes increases in energy production costs are fully passed forward onto domestic output prices of goods and services.
  - Total energy subsidies are the sum of direct and indirect effects.

*Source: wp18174 - introduction of a UBI in India that this paper does not address (IMF working paper, based on Indian 2011–12 NSS).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18174.pdf_