NiTi Aayog ‘Poverty’ Stats: Serious Theoretical, Methodological, Empirical Questions 

The NITI Aayog along with the UNDP released a discussion paper on the findings of their Multidimensional Poverty Index, attempting to study the decline of poverty rates and the number of multidimensionally poor people in India across various time periods. 

There are serious theoretical, methodological, and empirical questions that remain yet to be settled on a subject that has severe policy, welfare, and state-ideological implications. Unfortunately, there is less of an intent or expressed willingness by the government, and those in NITI Aayog too, to engage with these issues. 

Rather, what we see is a coercive push -unfortunately by different institutional ways and means- to keep beating down any possible data towards a rhetorical modelled reality painted by those who are obsessed with the view that ‘poverty reduction in India’ is a consequential reality/happening in the last ten years (under the Modi government).  

Look closely…

But let’s look more closely at the Multidimensional Poverty Index created by NITI Aayog in alignment with the globally acclaimed Alkire Foster index methodology. 

The National Multidimensional Poverty Index or NMPI measures simultaneous deprivations across three equally weighted dimensions of health, education, and standard of living that are represented by 12 sustainable development goals-aligned indicators, according to NITI Aayog. 

These include three Health (nutrition, child and adolescent mortality, maternal health); two Education (years of schooling, school attendance); and seven Standard of Living indicators (cooking fuel, sanitation, drinking water, electricity, housing, assets, and bank accounts). 

As per the findings of the Index, multidimensional poverty (MPI) in India declined from 29.17% in 2013-14 to 11.28% of the population in 2022-23, with about 24.82 crore people moving out of this bracket in nine years to 2022-23. They also claim that Uttar Pradesh, Bihar and Madhya Pradesh registered the largest decline.

Given a paradoxical reflection of magical realism shaping much of the underlying data around growth and other developmental indicators and metrics on the Indian economy, and in particular terms to what we have seen in the quality (and politicisation) of India’s statistical infrastructure over the last 10 years, one may ask the question: Are these claims-as projected by the MPI findings, credible? 

Economist Santosh Mehrotra doesn’t think so (and on grounds of empirical reasoning one can agree with his observations below).

According to Mehrotra’s recent The Wire column, 

“There is no prima facie reason for assuming that the 7.9% per annum GDP growth rate would deliver similar results (as applicable for MPI) to a period when the GDP growth rate for the recent 9 years fell to 5.7% per year. As though that presumption was not incredible enough, the NITI Aayog paper goes further, drawing upon NFHS 5 data for 2019 and 2021 (please note, not 2019 to 2021 because the survey was stopped after data collection was stopped in 22 states due to COVID), to project beyond 2021 – to 2022 and 2023. In other words, yet another linear projection was made by the authors to extend their conclusions to two years beyond the end of COVID. 

“In other words, it was using data for non-COVID years to extend non-COVID rates of improvement after COVID, to 2022 and 2023. Thus the question is legitimate as to whether that assumption is justified and credible or not. The whole purpose of making the NMPI the poverty indicator for India, while consumption expenditure surveys were not done for eight years from 2014 to 2022, is part of a political strategy.”

Evidence for massive jump in poverty after COVID

As part of our InfoSphere edition on the Great Poverty Debate in India, we discussed earlier how the disproportionate impact of the pandemic and the pandemic-induced lockdowns (and other restrictions) affected poorer households and states. 

An analysis of household income data analyzed by APU revealed that the decline in incomes during the COVID-19 period was higher for lower-income percentiles and gradually decreased for higher percentiles. 

The bottom 10 percentiles experienced a significant 27 percent drop in incomes, while the decline was 23% for the 40th to 50th percentiles and 22% for the top 10 percentiles. Furthermore, there was a 15-percentage point increase in rural areas and nearly 20 percentage points in urban areas.

Although the income declines in urban areas were relatively higher compared to rural areas, the difference between poor and relatively well-off households may seem small in percentage points. However, it represents a significant decline in absolute terms, exacerbating the challenges faced by vulnerable populations.

Even for non-monetary deprivations (as measured by a vulnerable group’s access to education, healthcare, nutrition etc.), one can see a sharp rise in absolute and relative poverty measures. 

According to the Hunger Watch national survey of the Right to Food campaign, a crisis emerged in December 2021 – January 2022 due to declining incomes and severe food insecurity, especially among the economically vulnerable and marginalized sections of society:

• 80% of people reported some form of food insecurity, and 25% reported severe food insecurity, such as skipping meals, cutting back on food, running out of food, not eating throughout the day and going to bed hungry
• 41% of respondents said the nutritional value of their diet had worsened compared to  pre-pandemic times
• 67% of people could not afford cooking gas in the month before the survey, further reducing their ability to cook

Also, a report by Pew Research Centre highlights that around 75 million more people in India fell into poverty in year 2022 because of the pandemic-induced economic recession, compared with what it would have been without the outbreak. That number for India accounts for nearly 60% of the global increase in poverty in 2020. In that study, it defined the poor as people who live on $2 or less daily.

Source: Infosphere CNES (Data from Pew Research Centre) 

More recently, The State of Working India Study conducted by Azim Premji University offers valuable insights into the impact of COVID-19 on Indian households. 

The study reveals that from March to October 2020, households experienced a significant loss of income, with an average decline of 22% in cumulative income. However, the income loss was more pronounced among poorer households, resulting in a substantial increase in poverty rates across the country. The APU analysis primarily relies on monthly household income data from the CMIE-CPHS, along with data from the India Working Survey (IWS) and Azim Premji University Covid Livelihoods Phone Survey (CLIPS). 

CMIE-CPHS method refers to the Consumer Pyramids Households Survey (CPHS) which is a comprehensive and nationally representative survey conducted by the Centre for Monitoring the Indian Economy since 2014. 

Regardless of which method one uses, the larger issue underlying the methodological concern to poverty measurement in India is a crisis of poverty (and welfare) statistics in a rapidly politicising and compromised statistical infrastructure, making any critical scrutiny erroneously difficult.  

A crisis of poverty-statistics in Modi’s India 

Poverty estimates, as rightly argued by Himanshu, “Are not just an academic exercise but are crucial parameters to judge policy outcomes and the overall functioning of the government”. 

Our own Centre’s research team, while creating the Access Equality Index (AEI) in exploring nonmonetary forms of deprivations had discussed the wider implications of states across India being asymmetrically ranked on providing ‘access’ to basic social, economic services (from food, healthcare, education, job security, finance, legal recourse-to name a few). 

What’s unfortunately happening is that research in Indian social sciences, not just in economics, has broadly transitioned to ‘proving hypothesis I or II’ rather than seeing the ‘data for what it actually is’ and then making reasonable conclusions. 

In other words, data is used for theory-validation rather than the other way around. 

There is more to it…

Growth alone is a bad indicator of quality of life as it fails to tell us how deprived people are doing. Thinking of developmental goal in terms of utility has perhaps the only merit of looking at what processes do for people in letting them ‘be’ what ‘they want to be’.

Poverty, for worse, represents a state of powerlessness – lack of opportunity and possible upward-mobility for an identified group, one that often lives/positions itself on the bottom of consumption-income pyramid. 

Part of the reason there are conflicting estimates of poverty for the same period is the loss of reliable data and a yardstick to measure poverty and inequality after 2011-12. 

There is a poverty of statistics that crowds out any meaningful policy or academic discourse on the statistics of poverty – and much of other social policy.

Economists, dare I say, are increasingly toeing the government narrative when it suits them, and oft dismiss any critical insight that brings the government (or its own policy-methodology) to account.

There are government economists -along with those writing to align with the regime’s rhetoric and ideology who constantly argue that ‘proxy’ indicators on poverty reflect ‘a fall’ in its actual estimate, while others weighing their evidence against the ‘facts’ show contrary results (see here and here for a more detailed discussion of the author on this). It’s because of this reason that social policy debates in India-not just on poverty but so many other vital welfare issues-lack both, clarity and critical scrutiny.

More importantly, having closely worked on development indices, while creating the Access(In)Equality Index in 2021 and updating its input-output indicators more recently, one becomes aware of choice-issues related with the ‘form’, ‘structure’ and ‘design’ limitations of such hand-picked, crafted metrics and what they may directionally -if not empirically, imply. 

In a country where we haven’t seen the regular conduct and release of national consumption expenditure survey data (from which poverty measurement principally happened in the past) or the census, there is room for much debate on whether the Modi government is honest about its intent to show/reflect the ‘truth’ of poverty performance in India. 

Poverty and income distribution are issues for public discussion as much as they are instruments of governance and public policy for an economy which still has a substantial population that’s financially vulnerable even if not officially poor. 

Unless we take conscious measures, at the level of state-authority, to consciously depoliticise data and encourage independent critical scrutiny of existing (and new) methodologies of public data accounting/analysis, there is little hope on what we might do on both social and economic policy. The policy making framework and those behind social policy evaluation can do better in making the ‘discourse’ more analytically coherent than operate under constant cycles of confusion. 

Deepanshu Mohan is a professor of economics and director, the Centre for New Economics Studies, Jindal School of Liberal Arts and Humanities, OP Jindal Global University.

This essay’s empirical data and findings are drawn from the previous 2023 investigative research undertaken by Centre for New Economics Studies (CNES) InfoSphere team on poverty assessment in India using a pre-covid and post-covid difference in difference perspective. For details, please review InfoSphere’s work here for more details.

Team credits feature Amisha Singh, Aditi Desai as Senior Research Analysts with CNES and team leads of InfoSphere team.