## Pandemic, Poverty, and Inequality: Evidence from India (wpiea2022069-print-pdf)

## Source details

**Canonical URL:** [Pandemic, Poverty, and Inequality: Evidence from India (wpiea2022069-print-pdf)](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022069-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022069-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022069-print-pdf.pdf.json)

---

### Context, motivation, and primary objective
- Pandemic raised concerns about increases in poverty and inequality and fiscal deficits from decline in tax revenues and higher expenditures.
- For the first time in several decades extreme poverty (PPP1.9 = $1.9 PPP 2011 dollars per person per day) increased worldwide in 2020; paper examines India’s contribution to that world increase.
- Primary objective: estimate poverty including in-kind subsidies and provide a consistent time-series of inequality and poverty (PPP$1.9 and PPP$3.2) for each year 2004-5 to 2020-21.

### Data challenges and baseline methods
- Traditional household surveys infeasible during pandemic; conventional method extrapolates most recent survey consumption by national accounts growth (PovcalNet approach).
- World Bank (Mahler et. al. 2021) estimate an increase of 97 million extreme poor worldwide in 2020 and a reduced 76 million in 2021.
- Country-level pandemic 2020 India estimates by others range 32 - 230 million additional poor.
- Authors construct synthetic annual series (agricultural year July-June) 2004-5 to 2020-21 using 2011-12 household distribution as benchmark and update per capita nominal consumption by either:
  - national PFCE growth (primary), or
  - nominal state domestic product per capita (alternative).
- Poverty lines updated with CPI rural and CPI urban by State; separate base distributions used for URP and MMRP.

### Recall-period methods, survey validity, and implications
- Two recall methods: Uniform Recall Period (URP) and Modified Mixed Recall Period (MMRP; official Indian method post-2011-12).
- Large divergences between URP and MMRP:
  - 2011-12: URP 21.8 %, MRP 18.5 %, MMRP 12.2 %
  - 2009-10: URP 33.8 %, MRP 29.4 %, MMRP 21.5 %
- Gap between URP and MMRP poverty estimates ~10-12 percentage points, corresponding to ~125 million difference at identical poverty line.
- NSS 2017-18 CES judged non-valid; Appendix II documents non-use of 2017-18 CES.
- Because S/NA (survey-to-national-accounts) consumption ratio stabilized post-2011-12 (~50–55 % depending on method), PFCE nominal consumption growth is a viable updating mechanism.
- Back-cast test to 2004-5: NSS URP consumption Rs. 699 pcpm vs PFCE back-cast Rs. 682 pcpm (2.4 % lower).

### Macro patterns, pass-through assumptions, and growth–poverty links
- Selected CAGR (in %) highlights (Table 1):
  - Nominal GDP: 2004-11 12.7; 2011-14 10.7; 2017-19 7.7; 2014-19 8.7
  - Real GDP: 2004-11 6.4; 2011-14 5.0; 2017-19 4.1; 2014-19 5.4
  - Nominal PFCE: 2004-11 12.4; 2011-14 11.8; 2017-19 9.1; 2014-19 9.5
  - PFCE deflated by PFCE Deflator (real per capita consumption): 2004-11 6.2; 2011-17 6.4; 2014-19 6.2
  - PFCE deflated by CPI (poverty line): 2004-11 4.0; 2011-17 4.7; 2014-19 5.9
  - Poverty Decline (URP) PPP$1.9: 2004-11 -14.1; 2011-17 -24.0; 2014-19 -32.9
- Pass-through assumption: nominal average household consumption growth set equal to average nominal PFCE growth (pass-through rate of unity assumed).
- Decline in CPI inflation contributed to faster poverty decline during 2014-19 despite lower GDP growth.

### Food subsidy (PDS) design, quantities, and rupee-equivalent estimation
- PDS has supplied subsidized cereals, sugar, and kerosene since 1978; 2013 Food Security Act (FSA) limited coverage to bottom 75 % rural and bottom 50 % urban.
- Pandemic 2020: ration doubled from 5 kg per month to 10 kg per recipient.
- NSS unit-level data (1999-00, 2004-5, 2011-12): average monthly per capita consumption of ~10 kg grain; 2011-12 consumption 9.73 kg (sd 0.51); aggregate share ~58 % rice, 42 % wheat.
- Off-take and observed consumption divergence (leakage):
  - Pre-2011 off-take ~3.1 kg transferred to “average” PDS recipient vs CES consumption ~1.0 kg pppm (2004-5).
  - 2011-12 NSS PDS consumption 2.04 kgs pppm; off-take 3.76 kg; > two-thirds leakage in 2004-5 estimated.
- Market prices and rupee-equivalent subsidy:
  - Pandemic market prices 2020-21: wheat Rs. 24.5/kg, rice Rs. 30.8/kg; weighted market price Rs. 28.1/kg.
  - Pre-pandemic nominal PDS prices: Rs. 2/kg (wheat) and Rs. 3/kg (rice) for 5 kg entitlement.
  - Rupee-equivalent subsidy = quantity obtained × (market price − subsidized price).
- Effective transfer rate assumed to increase to 86 percent in 2014-15 onward from pre-FSA 2011 level of 54 percent (projection); CMIE-based Bhattacharya & Sinha Roy (2021) All India average received during pandemic 84.6 percent (rural 89.1 percent, urban 77.3 percent).

### Empirical effects of food subsidies on poverty and inequality
- Monthly Subsidy (Rs) and Poverty line (Rs. per month) examples (Table 4):
  - Monthly Subsidy (Rs): 2004 = 4.0; 2011 = 23.8; 2014 = 72.8; 2017 = 88.6; 2019 = 119; 2020 = 192.7.
  - Poverty line (Rs. per month): 2004 = 480; 2011 = 865; 2014 = 1095; 2017 = 1246; 2019 = 1312; 2020 = 1399.
  - Subsidy to Poverty Line (%): 2004 = 0.8; 2011 = 2.8; 2014 = 6.6; 2017 = 7.1; 2019 = 9.1; 2020 = 13.8.
- Poverty and inequality adjusted for food transfers (Table 5 selected values):
  - PPP$1.9, Modified Mixed Recall (With transfers): 2004 = 31.9; 2011 = 10.8; 2014 = 5.1; 2017 = 1.9; 2019 = 0.8; 2020 = 0.9.
  - PPP$3.2, Modified Mixed Recall (With transfers): 2019 = 14.8; 2020 = 18.1.
  - Inequality Gini (Consumption Real): 2004 = 31.1; 2011 = 30.9; 2014 = 30.6; 2017 = 30.7; 2019 = 30.4; 2020 = 29.4.
  - 90th/10th percentile consumption ratio: 2004 = 4.1; 2011 = 4.0; 2014 = 3.9; 2017 = 3.9; 2019 = 3.9; 2020 = 3.7.
- Key empirical findings:
  - For PPP$1.9 extreme poverty, extreme poverty has stayed ≤ 1.1 percent of the population for the last three years including pandemic year; extreme poverty in 2020-21 was 0.8 % when transfers included.
  - Without any food subsidies, extreme poverty in the pandemic year would have increased by 1.05 % (from 1.43 to 2.48 %).
  - For PPP$3.2 (LMI) line, poverty increased from 14.8 % in 2019 to 18.1 % in 2020-21 (an increase of 3.3 percentage points, equivalent to 44 million for population of 1,340 million).
  - Doubling of entitlements in 2020 helped maintain extreme poverty at low levels for 2019-20 and 2020-21.
  - Using URP produces much larger pandemic-era increases (e.g., Kochhar (2020) 75 million), whereas MMRP with transfers indicates virtually no increase.

### Double counting, PFCE bias, and growth decomposition
- Double-counting concern: implicit cash transfer from rations equals quantity × (market price − ration price); because food consumption is inelastic, PFCE growth can under-estimate household consumption growth inclusive of food subsidies (a “negative double-counting” bias).
- Growth decomposition used: Y = (1 − w)*Ynf + w*Yf ⇒ Ynf = [Y − w*Yf]/(1 − w).
- Table 6 selected year entries showing PFCE growth and over-estimate of growth:
  - 2004: Market Price (Food) Rs/kg = 10.8; Share of food (%) = 7.8; PFCE (%) = 6.7; Food Growth = 31.7; Non-Food Growth = 4.6; Over-estimate = 2.1.
  - 2011: Market Price (Food) = 20.3; Share of food (%) = 6.2; PFCE (%) = 16; Food Growth = 6.3; Non-Food Growth = 16.6; Over-estimate = -0.6.
  - 2019: Market Price (Food) = 29.4; Share of food (%) = 3.9; PFCE (%) = 8.5; Food Growth = 1.4; Non-Food Growth = 8.8; Over-estimate = -0.3.
  - 2020: Market Price (Food) = 30.8; Share of food (%) = 4.2; PFCE (%) = -3.8; Food Growth = 4.8; Non-Food Growth = 4.1; Over-estimate = -7.9.
  - Period averages: 2005-11 Over-estimate = -0.2; 2012-19 Over-estimate = -0.3; 2004-20 Over-estimate = -0.1.
- Conclusion: double counting theoretically present but empirically absent or leading to under-estimate of growth when present.

### Inequality trends and determinants
- Real consumption inequality (Gini) declined to near its lowest level in forty years:
  - Gini was .284 in 1993/94 and .292 in 2020-21 (author-reported points).
- Reported real consumption inequality (URP Gini) historical:
  - 1983: 0.3272; 2004/5: 0.3324; 2009/10: 0.3389; 2011-12: 0.3392.
- Incorporation of food subsidies reduces measured inequality:
  - Example PFCE adjustment MMRP: 2019 Gini Without Transfers 31.4, With Transfers 30.4; 2020 Gini Without Transfers 31.5, With Transfers 29.4.
- Determinants of real inequality changes (base assumption: 2011-12 nominal distribution constant; real changes via):
  1. Differential price-level changes across states and urban/rural (poverty line updates).
  2. Differential evolution of average consumption across states via state domestic product growth (or identical adjustments under PFCE).
  3. Differential accrual of food subsidies due to eligibility/access (FSA eligibility bottom 75 % rural, bottom 50 % urban).

### Policy recommendations and welfare implications
- Given virtual elimination of extreme poverty at PPP$1.9, Government of India and World Bank should formally switch to the LMI country poverty line of PPP$3.2.
- World Bank should abandon the Uniform Recall Method and adopt the Modified Mixed Recall Method for India.
- Further research recommended to estimate the impact of the broader “welfare stack” (including LPG subsidy, electricity subsidy, asset transfers for toilets and affordable housing) on poverty incidence.
- Targeted in-kind transfers can achieve consumption impacts comparable to basic income at lower fiscal cost; targeted basic income or targeted in-kind transfers are more cost-effective than universal basic income.
- Pandemic fiscal response assessment:
  - India’s Gross Fiscal Deficit was 4.6 percent (with medium term target of 3 percent).
  - Three elements: rely on automatic stabilizers; target new expenditures/subsidies carefully; coordinate with central bank and provide monetary support and fiscal credit guarantees.
  - Evidence indicates targeted expansion of food subsidy (FSA + Aadhaar) and doubling entitlements in 2020 worked well to contain extreme poverty.

### Appendix II summary: credibility of 2017-18 CES and cross-checks
- 2017-18 CES was not released due to poor data quality; leaked reports analysed by multiple commentators.
- Key leaked-survey inferences labelled inconsistent with administrative and other survey data:
  - Reported average real rural consumption decline 8.8 % and urban increase 2 % between 2011-12 and 2017-18 ⇒ average decline 5.2 % over 6 years.
  - Reported large declines in Gini (rural 0.2872 → 0.2581; urban 0.3685 → 0.3298) not observed historically.
  - Reported extreme poverty increase from 12.2 % to ~17 % in 2017-18.
  - Implied nominal aggregate Rs. 2502 pcpm yields S/NA ratio 39.8 % vs 54.8 % in 2011-12.
- Multiple “smell tests” using administrative data (food production, medical and education expenditure trends, durable goods, mobility, employment earnings) find leaked 2017-18 CES an outlier; administrative data do not support the large declines in rural consumption or the extreme changes in inequality implied by the leaked survey.
- Authors justify non-use of 2017-18 CES and rely on PFCE-updated 2011-12 base distribution.

*Source: Section I, Section IV, Appendices II–III — wpiea2022069-print-pdf*

### Section I: Introduction ................................................................................................

### Section I: Introduction

### Context and motivation
- The pandemic raised concerns about increases in poverty and inequality globally; fiscal deficits increased because of decline in tax revenues and increased expenditures to combat pandemic effects.
- For the first time in several decades extreme poverty (those falling below the $1.9 PPP 2011 dollars per person per day; hereafter PPP1.9) in the world increased in the pandemic year 2020.
- The paper examines India’s contribution, if any, to the world increase in extreme poverty.

### Data challenges and conventional methods
- Traditional poverty analysis relies on household surveys, but pandemic constraints made thorough household surveys infeasible in recent years.
- The conventional method when no survey is available is to take the most recent survey data and update individual consumption (or personal) income by the corresponding growth rate observed in the national accounts; this extrapolation underpins PovcalNet estimates.
- Using this method, World Bank (Mahler et. al. 2021) estimate an increase of 97 million extreme poor worldwide in 2020 and a reduced 76 million in 2021 (that is, the number of extreme poor decreased by 21 million in the second pandemic year 2021).
- Country-level estimates for India derived by others are in the range of 32 - 230 million additional poor in 2020.

### Incorporating fiscal/in-kind interventions
- None of the cited poverty estimates incorporate the effect of fiscal interventions such as food subsidies and subsidized loans to the self-employed.
- The paper stresses that estimates of poverty that rely exclusively on household consumption expenditure from surveys will overestimate poverty rates unless the estimation method incorporates the effects of in-kind transfers.
- In-kind transfers reduce household consumption expenditure on items supplied free or at subsidized rates; adjustments for such transfers are necessary for reliable consumption and poverty estimates.
- In-kind food transfers have been an integral part of redistributive policy in India since the early 1980s.

### Contributions and methods highlighted
- One primary objective is estimation of poverty including in-kind subsidies.
- The paper provides a consistent time-series of inequality and poverty levels for both extreme poverty PPP$1.9 and low middle income poverty (PPP$3.2) for each year 2004-5 to 2020-21 (see Appendix I).
- The analysis uses state-level GDP and population data, administrative data on consumption items, several survey-based data (including NSS), and verification against national private final consumption expenditures (PFCE).
- Appendix II documents the non-validity of the NSS 2017-18 consumer expenditure survey and the authors’ non-use of it.

### Key findings reported in Section I
- Real inequality, as measured by the Gini coefficient, has declined to near its lowest level reached in the last forty years:
  - Gini was .284 in 1993/94 and .292 in 2020-21.
- Incorporation of food subsidies yields striking effects on poverty:
  - Extreme poverty has stayed below (or equal to) 1 % for the last three years.
  - In the pandemic year 2020-21 extreme poverty was 0.8 % of the population.
  - As early as 2016-17, extreme poverty had reached 2 %.
- According to the low middle income (LMI) poverty line of PPP$3.2 a day:
  - Poverty in India registered 14.8 % in the pre-pandemic year 2019-20.
- Historical comparisons:
  - In 2011-12, the official poverty level for the lower PPP$1.9 line was 12.2 %.

### Paper organization (plan)
- Section II surveys data, definitions, and methods for estimating poverty and inequality.
- Section III examines available poverty data using NSS surveys and World Bank methods and alternative estimates of growth and inflation for 2004-5 to 2020-21.
- Section IV outlines a method to incorporate in-kind food subsidies into poverty estimation, addressing possible double counting with PFCE adjustments.
- Section V concludes; Appendices document supplementary data, the non-use of the 2017-18 NSS survey, and comparisons with recent research findings.

*Source: Section I: Introduction — wpiea2022069-print-pdf*

### 12.6 and 10.8 %, respectively, with correspondingly large declines in the estimate of poverty (12.3

### wpiea2022069-print-pdf - 12.6 and 10.8 %, respectively, with correspondingly large declines in the estimate of poverty (12.3

### Recall methods, MMRP adoption, and impact on poverty estimates
- Two recall methods discussed: Uniform Recall Period (URP) and Modified Mixed Recall Period (MMRP; sometimes MRP referenced).
- In two consecutive two-year apart surveys (2009/10 drought year; 2011/12 normal rainfall year) the gap between the URP and the MMRP poverty estimate was around 10-12 percentage points (ppt).
- This gap corresponds to a difference in the estimate of poverty of around 125 million for an identical poverty line.
- The MMRP method is now the official Indian method of measuring poverty; the NSS decided to move “permanently” to the MMRP method and the 2017-18 CES survey gathered information exclusively on the MMRP method.
- World Bank continued use of URP for India is noted as outdated; results are presented for both methods with clear labeling that URP is no longer the official method.
- Specific extreme poverty measures reported for surveys:
  - 2011-12: URP 21.8 %, MRP 18.5%, MMRP 12.2 %
  - 2009-10: URP 33.8 %, MRP 29.4 %, MMRP 21.5 %

### Global practice and recall-period tradeoffs
- Longer recall periods:
  - Better at encompassing expenditure on infrequently purchased items.
  - Can lead to underreporting if respondents forget past purchases.
  - Despite lower average consumption, measured poverty might be lower under the longer recall period because it captures purchases of low-frequency items of poorer households.
- Short recall periods:
  - Mitigate underreporting but can lead to telescoping (reporting consumption outside reference period).
- Examples of country recall practices cited:
  - Jamaica: seven-day and 30-day used.
  - South Africa: weekly or monthly for food.
  - Kyrgyz Republic and Nicaragua: one-week recall.
  - Brazil: two-week recall.
  - India: separate recall periods for many non-food items; later adopted MMRP exclusively in 2017-18 CES.

### Survey capture divergence and consequences for updating methods
- Historical S/NA (survey-to-national-accounts) consumption ratio (survey capture) in India:
  - Averaged between 90-95 % for 1957-58 to 1977-78.
  - Declined to 62 % in 1993-94.
  - “Normalized” to around 50 % in 2004-5, 2007-8, 2009-10, and 2011-12 (URP method).
  - Around 55 % in 2009-10 and 2011-12 (MMRP method).
- Because of relative constancy in the S/NA ratio post-2011-12, national accounts nominal consumption growth (growth in Private Final Consumption Expenditure or PFCE) can be used to forecast consumption growth from the 2011-12 CES.
- Back-cast accuracy test for 2004-5:
  - NSS URP consumption level for 2004-5: Rs. 699 per capita per month.
  - Back-cast from 2011-12 using PFCE extrapolation: Rs. 682 per capita per month (just 2.4 % lower).

### Methods for estimating poverty in the absence of survey data
- Three data sets required:
  1. Base-year household distribution (2011-12 consumption distribution of 100,000+ households used separately for URP and MMRP).
  2. Updating procedure for per capita nominal consumption (two procedures used: national PFCE and nominal state domestic product per capita).
  3. Poverty line for each household across years (CPI rural and CPI urban for each State used to update State poverty lines; then merged with base household data by state and urban/rural status).
- Synthetic panel formed for agricultural years (July-June) 2004-5 to 2020-21 with 2011/12 household data as benchmark.
- Per capita consumption forecast forward/backward using national account consumption growth (PFCE) or state per capita GDP.
- Poverty estimated annually from resulting per-household nominal consumption distributions and residence-specific poverty lines.
- Inclusion of food subsidies: base forecast contains PFCE consumption at market prices; for households eligible for food subsidies, expenditure treated as per PFCE extrapolation plus equivalent cash transfer (Section IV incorporates role of Public Distribution System (PDS) subsidies).

### Data quality, 2017-18 NSS omission, and validation of PFCE updating
- Appendix II concludes the 2017-18 NSS consumer expenditure survey results did not correspond with most known data from diverse sources and therefore cannot be used for reliable poverty estimates (conclusion also reached by the government of India in 2019).
- Second conclusion: growth in national accounts expenditures over the last two decades conformed closely to growth from surveys and administrative data, supporting PFCE updating as reliable for generating poverty estimates.
- Meyer, Mok and Sullivan (2015) noted declining survey quality, unit non-response, item non-response, and measurement errors; administrative datasets provide alternative sources.

### Consumption inequality assumptions and determinants
- Base assumption: 2011-12 nominal distribution remains constant over time; changes in real inequality occur via three determinants:
  1. Differential price-level changes across states and urban/rural areas (poverty line updates).
  2. Differential evolution of average consumption across states via state domestic product growth (identical for all individuals when using PFCE updating).
  3. Differential accrual of food subsidies due to eligibility/access (2013 Food Security Act: bottom 75 % of rural and bottom 50 % of urban households eligible).
- Reported real consumption inequality (uniform recall period, Gini coefficient):
  - 1983: 0.3272
  - 2004/5: 0.3324
  - 2009/10: 0.3389
  - 2011-12: 0.3392
- Statement: real consumption inequality changed very little over 30 years (1983 to 2011-12) absent incorporation of food subsidies.

### Macro patterns of growth, inflation, and implications for poverty (Table 1 highlights)
- CAGR (in %) for selected periods (as presented in Table 1):
  - Nominal GDP: 2004-11 12.7, 2011-14 10.7, 2011-17 10.0, 2017-19 7.7, 2014-19 8.7
  - Real GDP: 2004-11 6.4, 2011-14 5.0, 2011-17 5.7, 2017-19 4.1, 2014-19 5.4
  - GDP deflator: 2004-11 6.3, 2011-14 5.7, 2011-17 4.3, 2017-19 3.6, 2014-19 3.3
  - Nominal PFCE: 2004-11 12.4, 2011-14 11.8, 2011-17 10.8, 2017-19 9.1, 2014-19 9.5
  - PFCE deflated by PFCE Deflator: 2004-11 6.2, 2011-14 6.1, 2011-17 6.4, 2017-19 5.5, 2014-19 6.2
  - PFCE deflated by CPI (poverty line): 2004-11 4.0, 2011-14 3.9, 2011-17 4.7, 2017-19 6.4, 2014-19 5.9
  - Inflation as indicated by PFCE Deflator: 2004-11 6.3, 2011-14 5.6, 2011-17 4.4, 2017-19 3.6, 2014-19 3.3
  - Inflation as indicated by CPI (poverty line): 2004-11 8.4, 2011-14 7.9, 2011-17 6.1, 2017-19 2.6, 2014-19 3.6
  - Poverty Decline (URP) PPP$1.9: 2004-11 -14.1, 2011-14 -16.6, 2011-17 -24.0, 2017-19 -35.2, 2014-19 -32.9
  - Poverty Decline (URP) PPP$3.2: 2004-11 -4.6, 2011-14 -7.1, 2011-17 -10.2, 2017-19 -22.4, 2014-19 -17.0
- Key inferences drawn:
  - GDP growth highest during 2004-11 (6.4 % per annum) and higher than 5.4 % per annum during 2014-19.
  - National accounts real per capita consumption growth (deflated by PFCE deflator) stayed relatively constant at 6.2 % per annum.
  - Real per capita consumption growth (deflated by CPI) was 5.9 % per annum for 2014-19, higher than the 4.0 % during 2004-11.
  - Decline in inflation contributed to faster poverty decline during 2014-19 despite lower GDP growth.
  - Pass-through assumption: nominal average household consumption growth set equal to average nominal PFCE growth (pass-through rate of unity assumed for nominal consumption growth); average pass-through for conversion of real per capita GDP growth to real per capita consumption growth cited as 0.67 in World Bank 2020 report.
  - Gap between CPI inflation and PFCE deflator inflation is a major source of divergence between survey-deflated consumption growth and national accounts growth.

### Poverty levels and trends 2004-2020 (traditional estimates, excluding food transfers; Table 2 highlights)
- Most important takeaway: pre-pandemic 2019 extreme poverty in India ranged between 1.4 % (official MMRP method, PFCE growth) and 5.4 % (outdated URP method, state domestic product growth).
- According to official MMRP method, poverty in pre-pandemic 2019 was 1.4 ppt, a decline of 10.8 ppt since 2011-12.
- Table 2 select figures (Poverty (in %) and Gini Estimates for India - Traditional, No Food Transfers):
  - PPP$1.9, Updates based on PFCE, Modified Mixed Recall:
    - 2004 32.7, 2011 12.2, 2014 7.4, 2017 2.9, 2019 1.4, 2020 2.5
  - PPP$1.9, Updates based on PFCE, Uniform Recall:
    - 2004 45.5, 2011 21.8, 2014 14.6, 2017 7.2, 2019 3.4, 2020 6.1
  - PPP$1.9, Updates based on SDP, Modified Mixed Recall:
    - 2004 37.1, 2011 12.2, 2014 9.4, 2017 4.2, 2019 2.2, 2020 4.1
  - PPP$1.9, Updates based on SDP, Uniform Recall:
    - 2004 49.7, 2011 21.8, 2014 17.8, 2017 9.4, 2019 5.4, 2020 8.8
  - PPP$3.2, Updates based on PFCE, Modified Mixed Recall:
    - 2004 73.8, 2011 53.6, 2014 43.3, 2017 29.0, 2019 18.5, 2020 26.5
  - PPP$3.2, Updates based on PFCE, Uniform Recall:
    - 2004 80.8, 2011 64.0, 2014 55.4, 2017 41.7, 2019 30.4, 2020 38.9
  - PPP$3.2, Updates based on SDP, Modified Mixed Recall:
    - 2004 76.8, 2011 53.6, 2014 47.6, 2017 33.1, 2019 23.3, 2020 31.0
  - PPP$3.2, Updates based on SDP, Uniform Recall:
    - 2004 82.5, 2011 64.0, 2014 58.9, 2017 45.3, 2019 34.6, 2020 43.0
- Additional notes:
  - World Bank estimates for 2014-15 (URP method, Newhouse-Vyas via a different imputation method) reported 14.6 %, identical to the authors' URP method based on PFCE growth rates.
  - Edochie et al. (2022) point estimate for head-count poverty in 2017-18: 10.4 %; the authors note “across a wide range of publicly available data sources, the paper finds no evidence of an increase in poverty between 2011/12 and 2017/18.”
  - The 2017-18 CES was exclusively MMRP data collection; poverty level in 2011-12 under MMRP was 12.2 % versus 21.8 % under URP.

*IMF WORKING PAPERS Pandemic, Poverty, and Inequality: Evidence from India*

### Section IV –Food Subsidy, Consumption Expenditures, and Inequality.

### Section IV –Food Subsidy, Consumption Expenditures, and Inequality

### Role and design of in-kind food support
- Direct measures (in-kind consumption support) are distinct from cash support and do not enter directly into survey estimates of consumption and therefore poverty.
- India’s Public Distribution System (PDS) has supplied subsidized cereals, sugar, and kerosene to ration card holders; major policy expansion in 1978.
- In 1985 Rajiv Gandhi concluded only 15 percent of funds meant for redistribution to the poor reached the poor.
- 2013 expansion of food subsidy limited coverage to the bottom 50 percent of urban and bottom 75 percent of rural population (Food Security Act).
- Pandemic 2020: food grain ration doubled from 5 kg per month to 10 kg per recipient.

### Consumption patterns and quantities
- Unit-level NSS data (1999-00, 2004-5, 2011-12) converge to an average monthly per capita consumption of 10 kg of grain.
- Total consumption of wheat and rice approximately constant at 10 kgs for 1993-94 to 2011-12; in 2011-12 consumption was 9.73 kg. per capita with standard deviation 0.51.
- Aggregate share decomposition: around 58 percent rice and 42 percent wheat.
- Table 3 (selected values from NSS unit-level data):
  - NSS 2004 monthly per capita consumption (Rs): Quintile 1 = 286, 2 = 408, 3 = 528, 4 = 725, 5 = 1550, All = 699.
  - NSS 2004 PDS transfers (Kg): Quintile 1 = 1.3, 2 = 1.2, 3 = 1.1, 4 = 0.9, 5 = 0.5, All = 1.
  - NSS 2004 Rice (Kg pppm): Quintiles = 6, 6.1, 6.2, 6, 5.3, All = 5.9; Wheat: 3.4, 4.1, 4.4, 4.6, 4.6, All = 4.2; Rice+Wheat: 9.4, 10.2, 10.6, 10.6, 9.9, All = 10.1.
  - NSS 2011 monthly per capita consumption (Rs): Quintile 1 = 734, 2 = 1053, 3 = 1368, 4 = 1878, 5 = 3982, All = 1803.
  - NSS 2011 PDS transfers (Kg): Quintile 1 = 2.8, 2 = 2.4, 3 = 2.2, 4 = 1.9, 5 = 1.1, All = 2.1.
  - NSS 2011 Rice (Kg pppm): Quintiles = 6, 5.8, 5.6, 5.4, 4.8, All = 5.5; Wheat: 4.1, 4.3, 4.2, 4.3, 4.1, All = 4.2; Rice+Wheat: 10.1, 10.1, 9.8, 9.7, 8.9, All = 9.7.

### Targeting effectiveness and leakage
- NSS CES (1993-94, 1999-00, 2004-5): PDS food quantity per person per month (pppm) consumed was in range 0.95 to 1.02 kgs.
- Government food off-take data indicated average of 3.1 kg transferred to the “average” PDS recipient (pre-2011).
- 2011-12 NSS PDS consumption: 2.04 kgs pppm (1.40 kg rice, 0.64 kg wheat); off-take indicated 3.76 kg delivered.
- Leakage estimate: more than two-thirds in 2004-5 (difference between off-take and observed consumption), interpreted as non-delivery/corruption.
- Top consumption quintile still received subsidized food: 0.5 kg in 2004-5 and over 1 kg in 2011-12.

### Estimating rupee-equivalent in-kind subsidy and methodology
- Pandemic year cereal market prices (2020-21): wheat Rs. 24.5/kg, rice Rs. 30.8/kg; weighted market food grain price Rs. 28.1/kg.
- Pre-pandemic nominal PDS prices: Rs. 2/kg (wheat) and Rs. 3/kg (rice) for 5 kg entitlement; pandemic additional 5 kg free.
- From 2014 onward (post-FSA) official allocation per person = 5 kgs pppm.
- Policy reduced eligible population from 1,280 million in 2013 to 870 million in 2014.
- Off-take increased transfer by 38 percent—from 3.9 kgs/month per person to 5.4 kgs pppm in 2014.
- Assumed effective transfer rate increased to 86 percent in 2014-15 onwards from pre-FSA 2011 level of 54 percent (projection based on trend); compares to CMIE-based Bhattacharya & Sinha Roy (2021) All India average 84.6 percent received during pandemic (rural 89.1 percent, urban 77.3 percent).
- Rupee-equivalent subsidy = quantity of food obtained × (market price − subsidized price); when household survey data unavailable, subsidized quantities estimated from consumption (inclusive of leakage) and market prices.

### Implications for poverty and measured impact of food subsidies
- Table 4 (selected values):
  - Off-take (lakh tons): Rice 2004 = 232.0, 2011 = 321.2, 2014 = 307.3, 2017 = 350.1, 2019 = 349.8, 2020 = 563.2.
  - Off-take (lakh tons): Wheat 2004 = 182.7, 2011 = 242.6, 2014 = 252.2, 2017 = 252.8, 2019 = 272.2, 2020 = 367.7.
  - Total off-take (lakh tons): 2004 = 414.7, 2011 = 563.8, 2014 = 559.5, 2017 = 602.8, 2019 = 622.0, 2020 = 930.9.
  - Monthly Subsidy (Rs): 2004 = 4.0, 2011 = 23.8, 2014 = 72.8, 2017 = 88.6, 2019 = 119, 2020 = 192.7.
  - Poverty line (Rs. per month): 2004 = 480, 2011 = 865, 2014 = 1095, 2017 = 1246, 2019 = 1312, 2020 = 1399.
  - Subsidy to Poverty Line (%): 2004 = 0.8, 2011 = 2.8, 2014 = 6.6, 2017 = 7.1, 2019 = 9.1, 2020 = 13.8.
- Table 5 (poverty and inequality adjusted for food transfers; selected values):
  - PPP$1.9 poverty line, Modified Mixed Recall: 2004 = 31.9, 2011 = 10.8, 2014 = 5.1, 2017 = 1.9, 2019 = 0.8, 2020 = 0.9.
  - PPP$1.9 poverty line, Uniform Recall: 2004 = 44.7, 2011 = 19.9, 2014 = 10.9, 2017 = 4.6, 2019 = 1.9, 2020 = 2.1.
  - PPP$3.2 poverty line, Modified Mixed Recall: 2004 = 73.5, 2011 = 52.2, 2014 = 39.7, 2017 = 25.2, 2019 = 14.8, 2020 = 18.1.
  - PPP$3.2 poverty line, Uniform Recall: 2004 = 80.1, 2011 = 62.9, 2014 = 52.0, 2017 = 37.7, 2019 = 25.5, 2020 = 29.9.
  - Inequality Gini (Consumption Real): 2004 = 31.1, 2011 = 30.9, 2014 = 30.6, 2017 = 30.7, 2019 = 30.4, 2020 = 29.4.
  - 90th/10th percentile consumption ratio: 2004 = 4.1, 2011 = 4.0, 2014 = 3.9, 2017 = 3.9, 2019 = 3.9, 2020 = 3.7.
- Key findings:
  - For the PPP$1.9 extreme poverty line, extreme poverty has been ≤ 1.1 percent of the population for the last three years (including pandemic year); virtually no increase in number of extreme poor in the pandemic year.
  - For PPP$3.2 poverty line, poverty increased by 3.3 percentage points between 2019 (14.8 percent) and 2020-21 (18.1 percent), an increase of 44 million for population of 1,340 million.
  - Doubling of entitlements in 2020 helped maintain extreme poverty at low levels for 2019-20 and 2020-21.

### Double counting, PFCE growth bias, and empirical results
- When household survey data available, implicit cash transfer from rations equals quantity of ration (Kgs) × (market price − ration price).
- Because food grain consumption is very inelastic, nominal PFCE growth can under-estimate household consumption growth inclusive of food subsidies (a “negative double-counting” bias).
- Growth decomposition formula used: Y = (1 − w)*Ynf + w*Yf, rearranged to Ynf = [Y − w*Yf]/(1 − w).
- Table 6 (selected values showing PFCE growth under-estimates Non-food growth):
  - 2004: Market Price (Food) (Rs Per kg) = 10.8; Share of food (wheat and rice) (%) = 7.8; PFCE (%) = 6.7; Food Growth = 31.7; Non-Food Growth = 4.6; Over-estimate of growth = 2.1.
  - 2011: Market Price (Food) = 20.3; Share of food (%) = 6.2; PFCE (%) = 16; Food Growth = 6.3; Non-Food Growth = 16.6; Over-estimate = -0.6.
  - 2014: Market Price (Food) = 28.5; Share of food (%) = 6.1; PFCE (%) = 10.5; Food Growth = 16.8; Non-Food Growth = 10.1; Over-estimate = 0.4.
  - 2017: Market Price (Food) = 26.1; Share of food (%) = 4.2; PFCE (%) = 8.9; Food Growth = -3.7; Non-Food Growth = 9.4; Over-estimate = -0.5.
  - 2019: Market Price (Food) = 29.4; Share of food (%) = 3.9; PFCE (%) = 8.5; Food Growth = 1.4; Non-Food Growth = 8.8; Over-estimate = -0.3.
  - 2020: Market Price (Food) = 30.8; Share of food (%) = 4.2; PFCE (%) = -3.8; Food Growth = 4.8; Non-Food Growth = 4.1; Over-estimate = -7.9.
  - Period averages:
    - 2005-11 Food share = 13.3; PFCE = 13.5; Over-estimate = -0.2.
    - 2012-19 Food share = 10.9; PFCE = 11.2; Over-estimate = -0.3.
    - 2004-20 Food share = 10.8; PFCE = 10.9; Over-estimate = -0.1.
- Conclusion: double counting theoretically present but empirically absent or results in an under-estimate of growth when present.

### Basic Income versus targeted transfers; fiscal and policy considerations
- Ghatak-Muralidharan recommended a transfer equal to Rs. 120 a month per citizen (2018-19 prices).
- FSA has provided transfers since 2014: Rs. 109 and Rs. 200 in 2020-21 (Table 4 context), reducing need for universal transfer.
- India effectively has a universal basic income for farmers: Rs. 6,000 a year for approximate family of five → Rs. 120 per person per month.
- Authors and referenced work argue targeted basic income or targeted in-kind transfers are more cost-effective than UBI.
- Virmani and Bhalla (2018) propose an Aadhaar-based net income transfer: NET = (Yb − x*Y) for Y < Yb, with taxes applied for Y > Yo and no transfers for Yo > Y > Yb; financing from reallocation of existing subsidies and transfers.
- Paper’s comparison: consumption impact of targeted food subsidy ≈ impact of basic income program but at lower cost.

### Pandemic fiscal response and assessment
- India’s Gross Fiscal Deficit was 4.6 percent (with medium term target of 3 percent).
- Three elements of pandemic relief:
  1. Rely on automatic stabilizers and reduce non-compliance penalties.
  2. Target new expenditures and subsidies carefully (iterative approach as information became available).
  3. Coordinate with central bank and provide monetary policy support; augment with fiscal credit guarantee program.
- Evidence indicates the second element (targeted expansion of food subsidy) worked well: Food Security Act (2013) and Aadhaar reduced leakage; doubling entitlements in 2020 maintained extreme poverty at low levels.

*Source: Section IV –Food Subsidy, Consumption Expenditures, and Inequality, wpiea2022069-print-pdf*

### 0.8 percent level. Without any food subsidies, extreme poverty in the pandemic year would have

### Pandemic, Poverty, and Inequality: Evidence from India

### Main findings on food subsidies and poverty incidence
- Food subsidies (PDS) were critical in preventing increases in extreme poverty during the pandemic; without any food subsidies, extreme poverty in the pandemic year would have increased by 1.05 % (from 1.43 to 2.48 %).
- Lower middle income (LMI) poverty with a PPP$3.2 poverty line would have increased by 8 % (18.5 % to 26.5 %) without food subsidies.
- The low level of extreme poverty is around 0.8 % in both 2019 (0.76 %) and 2020 (0.86 %) when transfers are included.
- Using the outdated Uniform Recall Period (URP), Kochhar (2020) concluded 75 million Indians were pushed into extreme poverty in 2020-21; by contrast, the Modified Mixed Recall Period (MMRP) indicates no increase with food transfers, and only an estimated 13.4 million increase (1 % of population) without transfers.
- The paper treats in-kind transfers as monetary-income-equivalent transfers on the grounds that households could sell subsidized food grains in the open market.

### Methodology and adjustment approach
- The paper adjusts 2011-12 household consumption expenditure levels to derive consumption distributions for 2011–2020 using estimates based on average per capita nominal PFCE growth.
- The analysis derives average rupee food subsidy transfer to each individual for each year 2004-5 to 2020-21 to avoid overestimating poverty when relying solely on reported consumption expenditures.
- Alternative adjustments are also presented using SDP (Gross State Domestic Product) growth rates to update the 2011-12 consumption distribution.

### Comparative measurement: MMRP versus URP
- NSS formally switched to the Modified Mixed Recall Method (MMRP) after the 2011-12 “experiment”; differences in definitions and survey consumption methods have produced large divergences in poverty assessment.
- Key illustrative numbers:
  - Table A1-2 (With Transfers, MMRP): 2019 extreme poverty 0.76 %, 2020 extreme poverty 0.86 %.
  - Table A1-1 (Without Transfers, MMRP): 2019 PPP$1.9 poverty 1.4 %, 2020 PPP$1.9 poverty 2.5 %; 2019 PPP$3.2 poverty 18.5 %, 2020 PPP$3.2 poverty 26.5 %.
  - SDP-based projections (Table A1-5, With Transfers, MMRP): 2019 PPP$1.9 poverty 1.30 %, 2020 PPP$1.9 poverty 1.42 %; 2019 PPP$3.2 poverty 19.0 %, 2020 PPP$3.2 poverty 22.8 %.

### Inequality and distributional effects
- Real inequality (Gini) with the type-2 (MMRP) consumption distribution shows:
  - Table A1-3 (PFCE adjustment): 2019 Gini Without Transfers 31.4, With Transfers 30.4; 2020 Gini Without Transfers 31.5, With Transfers 29.4.
  - Table A1-6 (SDP adjustment): 2019 Gini Without Transfers 32.4, With Transfers 31.3; 2020 Gini Without Transfers 32.4, With Transfers 30.3.
- Food subsidies have reduced poverty consistently since the enactment of the FSA in 2013 and improved targeting via Aadhar.

### Policy recommendations and welfare implications
- Given the virtual elimination of extreme poverty, the Government of India and World Bank should formally switch to the LMI country poverty line of PPP$3.2.
- The World Bank should abandon the Uniform Recall Method and adopt the Modified Mixed Recall Method for updating poverty estimates for India.
- Further research is warranted to estimate the impact of the broader “welfare stack” (including LPG subsidy, electricity subsidy, asset transfers for toilets and affordable housing) on poverty incidence.
- Temporary fiscal policy interventions can absorb a large part of the pandemic-induced temporary income shock; improved targeting combined with entitlement expansion during the pandemic absorbed part of the consumption shock.

### Additional technical notes and select quantitative points
- The paper documents PDS food subsidies’ average rupee transfers to individuals for 2004-5 to 2020-21 (derived and incorporated into poverty estimates).
- Table A1-7 highlights potential underestimation/over-estimation issues in PFCE growth due to presence of PDS wheat and rice, listing year-by-year values for Weighted Market Price (Rs Per kg), Share of Food (%), PFCE Growth (%), NSS Food Expenditures Growth Rate (%), and Over-estimate of growth (%).
- The analysis treats in-kind transfers as cash-equivalent transfers for welfare measurement purposes.

*Source: IMF Working Paper — Pandemic, Poverty, and Inequality: Evidence from India (excerpted content).*

### Appendix II: Data Supplement and Accuracy of the 2017 “Ill-Fated” Survey

### Appendix II: Data Supplement and Accuracy of the 2017 “Ill-Fated” Survey

### Absence of 2017-18 CES and leaked data
- The 2017-18 Consumption Expenditure Survey (CES) was not released due to poor data quality.
- Parts of a leaked survey report were analysed by Jha, Somesh (2019); Bhalla (2019); Bhalla and Bhasin (2019); and S. Subramaniam (2019, 2020).  
- The leaked source is described as: "Computations based on data in the 2011-12 NSO Report on Consumer Expenditure (68th Round) and Tables T3 and T4 of the 2017-18 NSO draft Report on Consumer Expenditure (75th Round)."
- The authors compile administrative datasets (BARC, Census Bureau, Ministry of Petroleum and Natural Gas, SIAM, Department of Telecommunications, Ministry of Railways, Ministry of Civil Aviation, Animal Husbandry Statistics, Ministry of Agriculture and Farmer’s Welfare) to compare with the 2017-18 CES.

### Key reported outcomes from the ill-fated IF2017-18 survey
- (i) "Average real rural incomes consumption declined by 8.8 % and real urban consumption increased by 2 % between 2011-12 and 2017-18. Given an urbanization ratio of approximately 33 %, this yields an average decline in real consumption of 5.2 % over 6 years."
- (ii) Inequality declines reported by Subramanian (2019): "Gini coefficient of inequality, from 0.2872 to 0.2581 in the rural areas, and from 0.3685 to 0.3298 in the urban areas."
- (iii) Such large declines in real consumption inequality had not been observed post-1982; typical deviations in the Gini are in a narrow 1 to 2 ppt range.
- (iv) "Extreme poverty increased from the NSS 2011-12 estimate of 12.2 % in 2011-12 to around 17 % in 2017-18."
- (v) Inferred nominal average consumption in 2017-18: "Rs. 1892 per capita per month (pcpm) in rural India and Rs. 3739 per capita per month in urban India leading to a nominal Rs. 2502 pcpm aggregate level (for the MMRP method)."  
  - This yields a S/NA ratio of 39.8 % for India; comparison: "the 2011-12 NSS survey/national accounts ratio of household consumption was 54.8 %."

### Methodology: "Imagine There is No Country" smell tests and cross-checks
- Central idea: compare IF2017-18 findings with administrative datasets, national accounts and other household surveys.

- Smell Test 1 – Growth in Food Production (Per-Capita Log Growth Rate in Food Production)
  - Table A2-1 annualized per-capita log growth rates (2004-11 and 2011-17):
    - Rice: 2.8 (2004-11), 0.6 (2011-17)
    - Wheat: 4 (2004-11), 0.4 (2011-17)
    - Coarse Cereals: 2.7, 1.4
    - Meat: 12.5, 5.1
    - Eggs: 4.9, 5.5
    - Milk: 4, 4.8
    - Oils: 2.3, 0.4
    - Fruits: 5.2, 3.5
    - Vegetables: 4.7, 3.3
    - Pulses: 3.2, 6.1
    - Rapeseed & Mustard: -2.6, 3.6
    - Groundnut (as oil): -0.2, 4.2
    - Sugar (cane): 5.4, 0.3
    - Spices: 5.1, 4.7
    - Tea: 2.1, 2.7
    - Tobacco: 3.8, 3.4
  - For about 42 % of food consumption items, the average (weighted) per capita growth in 2004-11 was 1.2 % versus 1 % in 2011-17.

- Smell Test 2 – Inferences from other NSS consumer surveys
  - Education (5.1 %) and health (6.2 %) account for a significant share of total expenditures.
  - National accounts per capita real annual growth rates between 2011-12 and 2017/18:
    - Medical care: 7.6 %
    - Education: 6.9 %
    - Joint average growth for these two items: 7.3 %
  - Comparison: average real PFCE growth rate: 4.5 % per annum.

- Smell Test 3 – Consumption of Non-Food items (Per Capita Log Growth Rate)
  - Table A2-2 entries:
    - PFCE: 4.6 (2004-11), 5.6 (2011-17)
    - Medical - National Accounts: -0.2 (2004-11), 7.6 (2011-17)
    - Education - National Accounts: 6.5, 6.9
    - Two wheelers: 11.6, 5.8
    - Cars: 12.5, 3.9
    - TV: 4.3, 7
    - Fuel Consumption: 3.8, 4.8
    - Electricity: (2011-17) 5.6
    - Airline Passengers: 31.1, 10.7
    - Railway Passengers: 5.5, -0.4
    - Mobile Users: (2011-17) 5.6
  - Combining food production and social consumption covers a large 53 % of total consumption. The average growth rate of 2.4 % for 2011-17 is 0.9 % per annum higher than 2004-11.

- Smell Test 4 – Consumption Growth, 2011-12 to 2014-15
  - Newhouse-Vyas (2018) method (non-identical surveys) reports:
    - Real urban consumption growth between 2011-12 and 2014-15: 3.3 %
    - Real rural consumption growth: 8.6 %
    - Average annual growth: 2.9 % per annum
  - Contrast: IF2017-18 reported cumulative growth of 2 % for urban and -8.8 % for rural between 2011-12 and 2017-18.

- Smell Test 5 – Pattern of real wage growth, 2017-18
  - Table A2-3: Per Capita Log Growth Rate in Consumption
    - PFCE - National Accounts: 4.3 (2004-11), 4.5 (2011-17)
    - HCES 2017-18: -5.2 (real per capita consumption)
    - PLFS 2017-18: -11.5
    - IHDS: 2.7 - 4.9
  - Table A2-4: NSS Employment Surveys - Real earnings per person month (Ages 15-64)
    - 2004: 4288 (All), 2567 (Female), 4882 (Male)
    - 2011: 6383, 4347, 6947
    - 2017: 7005, 5137, 7493
    - 2018: 7248, 5454, 7713
    - 2019: 7445, 5613, 8006
    - Growth Rate (%) 2011-17: 9.7 (All), 18.2 (Female), 7.9 (Male)
    - Growth Rate (%) 2011-19: 16.6 (All), 29.1 (Female), 15.2 (Male)
  - Wage inequality (real Gini) — Table A2-5:
    - 1999: 0.53
    - 2004: 0.52
    - 2011: 0.50
    - 2017: 0.43
    - 2018: 0.42
    - 2019: 0.43

- Smell Test 6 – Comparing consumption growth in CES and PLFS surveys
  - PLFS (one consumption question) indicates per capita consumption declines of 11.5 % between 2011 and 2017 (rural: 12.6 % decline; urban: 10.4 % decline).
  - Leaked CES shows rural decline of 8.8 % and urban increase of 2 % between 2011-12 and 2017-18.
  - Conclusion: consumption growth estimates from CES and PLFS for 2017-18 are not comparable.

### Overall assessment of IF2017-18 credibility
- Multiple checks (administrative data, national accounts, other surveys) consistently indicate IF2017-18 CES is an outlier in terms of data quality.
- Administrative data for 2017-18 supports the conclusion that the 2017-18 survey suffered from acute measurement problems, justifying non-release.
- The IF2017-18 CES implied unprecedented movements: large decline in rural consumption/incomes, large decline in inequality via "levelling down", increase in extreme poverty from 12.2 % to around 17 %, and an unusually low S/NA ratio of 39.8 % versus 54.8 % in 2011-12.

### Appendix III – Comparing our results with other findings (summary)
- Standard practice uses World Bank PPP $1.9 and PPP $3.2 poverty lines; PPP $1.9 corresponds to Rs. 865 per person per month in 2011-12 prices and matches India’s Tendulkar poverty line.
- Recent studies using CMIE CPHS (e.g., Gupta, Malani & Woda (2021); Basole et al. (2021)) did not adjust consumption for food transfers and used wage-based or wage-threshold poverty lines, producing higher poverty lines than PPP$1.9:
  - Basole et al. rural poverty line: Rs 2900 per person per month (Rs 94 per person per day); urban: Rs 3,344 per person per month (Rs 108 per person per day).
  - PPP$1.9 translates to Rs 28.5 per day.
- Criticisms of CMIE-based studies:
  - They often do not incorporate food transfers, biasing consumption downwards and poverty upwards.
  - Wages must be adjusted for number of earners and poverty lines adjusted to 2011 values; many studies do not perform these adjustments.
  - Some studies report estimates for stipulated periods rather than annual estimates on the agricultural year used by NSO CES.
- Kochhar (2021) using PovcalNet and national accounts estimates pandemic increased poverty by 75 million using PPP$1.9, but did not state that the consumption distribution used was the obsolete uniform recall method, nor provide rupee poverty lines for rural/urban areas or estimates for 2019.
- The authors’ historical estimates based on NSO 2011-12 base distribution and PPP $1.9 / PPP $3.2 align with World Bank estimates for preceding years and use the agricultural year for comparability.

*Source: Appendix II and Appendix III of wpiea2022069-print-pdf*

### References:

### References (Pandemic, Poverty, and Inequality: Evidence from India — Working Paper No. WP/2022/069)

### Major topical clusters in the cited literature
- Poverty measurement and survey methodology
  - Deaton, A., & Grosh, M. (1998). Designing household survey questionnaires for developing countries lessons from ten years of LSMS experience, chapter 17: Consumption (No. 218).
  - Deaton, A. (2003). Household surveys, consumption, and the measurement of poverty. Economic Systems Research, 15(2), 135-159.
  - Meyer, B. D., Mok, W. K., & Sullivan, J. X. (2015). Household surveys in crisis. Journal of Economic Perspectives, 29(4), 199-226.
  - Government of India (2005). Report of the Expert Group on Cross-validation study of estimates of private consumption expenditure available from household survey and national accounts, Sarvekshana, Vol.XXV(4) and XXVI(1),Issue No.88, 1–70.
  - Government of India (2008). Report of the Group for Examining Discrepancy in PFCE estimates from NSSO Central Statistical Organisation, Ministry of Statistics and Programme Implementation.
  - Government of India (2015). The Report of Prof. A. K. Adhikari Committee on PFCE. Central Statistics Office, Ministry of Statistics and Programme Implementation.
  - Newhouse, D. L., & Vyas, P. (2018). Nowcasting poverty in India for 2014-15: A survey to survey imputation approach (No. 6).
  - Mahler, D. G., Castaneda Aguilar, R. A., & Newhouse, D. (2021). Nowcasting Global Poverty.

- COVID-19 impacts on poverty, inequality, and welfare
  - Edochie, I. N., Freije-Rodriguez, S., Lakner, C., Moreno Herrera, L., Newhouse, D. L., Sinha Roy, S., & Yonzan, N. (2022). What do we Know about Poverty in India in 2017/18?.
  - Gupta, A., Malani, A., & Woda, B. (2021). Inequality in India Declined During COVID (No. w29597). National Bureau of Economic Research.
  - Mahler D. G., Yonzan N, Lakner C, Andres Castaneda Aguilar R, Wu H. Updated estimates of the impact of COVID-19 on global poverty: turning the corner on the pandemic in 2021? World Bank Blogs. Accessed September 28, 2021.
  - World Bank. 2020. Poverty and Shared Prosperity 2020: Reversals of Fortune. Washington, DC: World Bank.
  - Basole, A., Abraham, R., Lahoti, R., Kesar, S., Jha, M., Nath, P., ... & Narayanan, R. (2021). State of working India 2021: one year of Covid-19.
  - Virmani, A., & Bhasin, K. (2020). Growth Implications of Pandemic: Indian Economy (Vol. 7). Working paper no 2/2020, Foundation for Economic Growth and Welfare, New Delhi.
  - Hamilton, L., Roll, S., Despard, M., & Maag, E. (2021). Employment, Financial and Well-Being Effects of the 2021 Expanded Child Tax Credit: Wave 1 Executive Summary.

- India-focused analyses of poverty, growth, redistribution, and data interpretation
  - Bhalla, S. S. (2002). Imagine there's no country: Poverty, inequality, and growth in the era of globalization. Peterson Institute.
  - Bhalla, S. S. (2010). Raising the Standard – The War on Global Poverty. In Sudhir Anand, Paul Segal and Joseph Stiglitz (ed), Debates on the Measurement of Global Poverty, Oxford University Press, Oxford.
  - Bhalla, S. S. (2015). Food, Hunger, and Nutrition in India: A Case of Redistributive Failure.
  - Bhalla, S. S. (2017). The new wealth of nations. Simon and Schuster.
  - Bhalla, S. S., & Bhasin, K. (2019a). Towards a Targeted Basic Income Policy for India.
  - Bhalla, S. S., Bhasin, K. & Virmani, A. (2020). Poverty, Inequality, and Growth in India: 2011-2018. Presented at NCAER.
  - Ghatak, M., & Muralidharan, K. (2019). An inclusive growth dividend: Reframing the role of income transfers in India’s anti-poverty strategy.
  - Virmani, A. (2004). Accelerating Growth and Poverty Reduction: A Policy Framework for India's Development. Academic Foundation.
  - Virmani, A. (2005). Policy regimes, growth and poverty in India: Lessons of government failure and entrepreneurial success! (No. 170). Working Paper, ICRIER.

- Debates, commentary, and media perspectives on Indian data and welfare trends
  - Bhalla, S. S., (2019, July 27). It is time we recognised that survey data cannot be interpreted in the way it used to be. Indian Express.
  - Bhalla, S. S., & Bhasin, K. (2019b, June 29). Rethink poverty — and policy. Indian Express.
  - Bhalla, S. S., & Bhasin, K. (2019c, November 28). Opinion: Is the NSO’s consumption data for 2017-18 beyond salvation? Live Mint.
  - Bhalla, S. S., & Virmani, A. (2018, January 27). Smart policies for redistribution. Indian Express.
  - Felman, J., Sandefur, J., Subramanian, A., & Duggan, J. (2019). Is India’s Consumption Really Falling? Center for Global Development blog.
  - Himanshu. (2019, August 15). Opinion: What Happened to Poverty during the First Term of Modi? Live Mint.
  - Jha, S. (2019a, February 6). Unemployment rate at four-decade high of 6.1% in 2017-18: NSSO survey. Business Standard.
  - Jha, S. (2019b, November 15). Consumer spend sees first fall in 4 decades on weak rural demand: NSO data. Business Standard.
  - Rangarajan. C., & Mahendra Dev., S. (2019). Mind the statistics gap. Indian Express.
  - Subramanian, S. (2019, November). What is happening to rural welfare, poverty and inequality in India. In The India Forum (Vol. 29).

- Global poverty, inequality, and methodological advances
  - Lakner, C., Mahler, D. G., Negre, M., & Prydz, E. B. (2022). How much does reducing inequality matter for global poverty?. The Journal of Economic Inequality, 1-27.
  - Mahler, D. G., Yonzan N, Lakner C, Andres Castaneda Aguilar R, Wu H. Updated estimates of the impact of COVID-19 on global poverty: turning the corner on the pandemic in 2021? World Bank Blogs.
  - World Bank. (2018). Poverty and shared prosperity 2018: Piecing together the poverty puzzle.

### Notable specific data or findings cited within references
- Jha, S. (2019a) reports "Unemployment rate at four-decade high of 6.1% in 2017-18: NSSO survey."
- Gupta, A., Malani, A., & Woda, B. (2021) published as National Bureau of Economic Research working paper No. w29597: "Inequality in India Declined During COVID."
- Mahler et al. (World Bank Blog) entry was accessed on September 28, 2021.

### Implicit analytical and policy emphases drawn from the references
- Importance of survey design, cross-validation, and reconciling household survey consumption with national accounts for accurate poverty measurement.
- Considerable focus on the effects of the COVID-19 pandemic on poverty, inequality, employment, and welfare in India, including nowcasting approaches and immediate policy responses.
- Ongoing policy debates in India around targeted income transfers, basic income proposals, redistribution strategies, and the interpretation of official consumption and poverty statistics.
- Engagement with both academic and policy-oriented sources, as well as media commentary and working papers, indicating a multidisciplinary evidence base.

*Pandemic, Poverty, and Inequality: Evidence from India — Working Paper No. WP/2022/069*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022069-print-pdf.pdf_