Terms of Trade | The Great Indian Poverty Debate is at a crossroads
With each debate, it gets more complex. But with comparable data hard to come by, it's time to ask if asset delivery is more key to politics than income support
Last couple of weeks have seen a renewed interest in what is known as the great Indian poverty debate. In late March, reports appeared that a new paper by economists Arvind Panagariya and Vishal More — the former headed the NITI Aayog under the first Narendra Modi-led government — has found that reports of poverty increasing in the post-pandemic period in India were not true. Earlier this month, economists Jean Dreze and Anmol Somanchi — the former has been among the biggest supporters and defenders of anti-poverty programmes such as MGNREGS in India — published a detailed essay on the Ideas for India website questioning the efficacy of poverty estimation using data from the Centre for Monitoring Indian Economy’s (CMIE) Consumer Pyramid Household Surveys (CPHS).
The economists in the debate will take the road less travelled by, which may hold complex debates about changes in poverty and methods — but not have much of an impact on the busier road of mainstream politics. (Shutterstock)
For those uninitiated to this debate, a 2022 working paper by two World Bank economists, Sutirtha Sinha Roy and Roy van der Weide, had calculated poverty estimates for India using CPHS numbers and found that while poverty has been declining in India, its magnitude was much smaller than what was claimed in a 2022 IMF working paper by Surjit Bhalla, Arvind Virmani and Karan Bhasin (click here for a discussion on these two papers). Bhalla, at the time of writing this paper, was a Government of India nominated executive director at the IMF.
An absence of data, an abundance of estimates
Ironic as it is, despite having seen four new papers on poverty numbers in India, we still do not have an official number for the extent of poverty in India. The reason is simple. India’s poverty estimation is done using a Consumption Expenditure Survey (CES) — it is held every five years and the last one was held in 2011-12 — which is conducted by the National Sample Survey Organisation (NSSO). The government junked the 2017-18 CES and did not even publish its data which has led to non-availability of poverty estimates, and a host of other information, after 2011-12.
Because there is no CES to calculate poverty numbers for India, economists have been trying to figure out alternative ways of doing this. Bhalla et al, consistent with their former position that CES underestimates private consumption, have used Private Final Consumption Expenditure (PFCE) data to construct these estimates. Roy and Weide have used adjusted CPHS numbers as a substitute for CES data to construct their numbers. Dreze and Sonmanchi have critically examined the adjustments to CPHS which forms the basis of Roy and Weide’s calculations. Panagariya and More have used consumption data given in the Periodic Labour Force Survey (PLFS) to prepare their own poverty estimates.
The methods which have been applied by these economists are quite technical in nature. The World Bank paper and Dreaze and Sonmanchi’s critique of it, for example, requires understanding something called max-entropy method, which has been used to adjust CPHS data. It will take years, if not decades to get the skill set to understand these concepts. Simply speaking, such methodological issues are unlikely to be of interest or retain the attention of an interested reader who is not a quantitative methods geek. At the risk of sounding anti-intellectual, this author would like to argue that there is no point in following the ongoing poverty debate.
There are two reasons for such a conclusion. One of them is strictly academic.
The entire debate around poverty in India wants to make an informed argument about long-term trends in poverty. This is only possible if the methodology of calculating poverty remains the same through the reference period. This involves two things; the poverty line and the statistical base (in this case, the CES) for calculating the share of people below the poverty line.
There is a broad-based agreement across the left and the right with respect to the poverty line until the 2011-12 poverty numbers with the exception of Utsa Patnaik’s dissenting argument which holds that the recent poverty line threshold needs to be set at a much higher level to ensure a comparable level of calorie intake.
It is the quality and consistency of CES data which has been a bigger debate among the economists when it comes to poverty numbers in India. The first is the issue of divergence between CES and National Account Statistics (NAS) consumption estimates, which is where Bhalla et al have been the most vocal. On this count, an overwhelming consensus is in favour of not junking the use of CES numbers despite the growing divergence with NAS estimates.
The second is the debate around comparability of CES numbers due to tweaks in the kind of questions which are asked while recording consumption. This was the overarching theme in the poverty debate which first erupted in the early 2000s and actually led to a change in CES methodology between the 1999-2000 and 2004-05 CES rounds.
If Himanshu’s April 12 article published in The Indian Express is to be believed — he is among the most authoritative voices on the NSSO data in India — the ongoing CES round has once again changed the methodology of recording consumption in a way which will make the results, whenever they come, non-comparable to the previous poverty numbers. This will mean an end of long-term poverty comparisons in India, unless, of course, the government conducts a fresh CES which corrects for the change in methodology and conducts a new round to maintain non-comparability.
This brings us to the second reason, which is strictly political. If the current government’s track record in showing sensitivity (or lack of it) to concerns around quality and even availability of statistics is evaluated, it is very unlikely that a bunch of economists writing about tweak in CES methodology is going to lead to a change in what has been decided after a report with uncomfortable conclusions was junked. The government has not even conducted the decadal census which was due in 2021 and refuses to make any commitment about when it intends to conduct one. At the risk of sounding provocative, one can say that a change in CES methodology will require nothing short of change of government in the Centre. The number of economists who are likely to be aggrieved because of comparable CES data not being available cannot even influence the result in an assembly constituency.
What explains the lack of data collection?
Why is the current government not interested in publishing comparable poverty statistics? The lazy answer is that it is afraid that the numbers might show its record in poverty reduction in a bad light. To be sure, the government’s scrapping of the 2017-18 CES – they did show a fall in consumption levels – has given credibility to this argument.
However, what needs to be kept in mind while relying too much on a conspiracy theory argument about non-publication of poverty statistics is the fact that the Bhartiya Janata Party (BJP) under Modi has invested a lot of political and fiscal capital in building a narrative that it is pro-poor in nature. These efforts, to be sure, are very different compared to what the Congress made during its United Progressive Alliance (UPA) government.
While the UPA focused on anti-poverty programmes which were aimed at generating income/consumption — guaranteed employment under a demand driven MGNREGS and large expansion in food safety net under the NFSA are the biggest examples — Modi government has invested more in asset enhancement programmes such as provision of free toilets, gas cylinders, houses and now portable water. By their very nature, asset enhancement programmes do not give a direct boost to regular income of beneficiaries and therefore unlikely to lead to a major reduction in poverty numbers. This could have been the biggest motivation behind (what can only be described as) an internal sabotage of poverty statistics, as we knew them, by the current government.
Of course, the Modi government has been pragmatic enough to double down on income-support programmes during periods of acute crisis. Expansion in MGNREGS and PMGKY which offered free food grain during the pandemic are the biggest examples.
The political evidence, at least so far, suggests that there has been no major backlash by the poor against the Modi government or the BJP. In fact, recent research such as by Shikhar Singh, currently a post-doctoral fellow at University of Pennsylvania suggests that the BJP has made political gains due to delivery of material benefits such as LPG cylinders. As long as the BJP continues to gain from such policies, it is unlikely to lose power, and economists are unlikely to see a change in CES methodology to give comparable data with the past.
This brings up the key political question on the fate of the great Indian poverty debate. Has Indian politics entered a phase where asset delivery has emerged as a more effective route to secure political support than schemes which provided income support? If the answer to this question is yes, then the future of poverty debate in India is likely to become more and more technocratic and distant from the wider political economy narrative.
Unless the opposition manages to bring back a consumption/income support demand which proves to be a political game changer, we have arrived at a fork in the road moment in the great Indian poverty debate. The economists participating in the debate will take the road less taken. This road will see a lot of complex debates about changes in poverty, but they will not have much of an impact on the busier road of politics.
If one were to paraphrase one of Modi's famous one-liners, it is going to take a lot of hard work (by the opposition) to preserve the sanctity of Harvard (the academic debate over poverty statistics in India).
Every Friday, HT’s data and political economy editor, Roshan Kishore, combines his commitment to data and passion for qualitative analysis in a column for HT Premium, Terms of Trade. With a focus on one big number and one big issue, he will go behind the headlines to ask a question and address political economy issues and social puzzles facing contemporary India
E-Paper
