Do we need a new line to assess poverty based on CES for 2022-23?

At the $3.2 per capita per day poverty line, which the World Bank uses for lower-middle-income countries such as India, poverty rate declined from 53.6% during 2011-12 to 20.8% during 2022-23

Poverty estimates in India have mostly been shrouded in controversy. The recent one based on the household consumption expenditure survey for 2022-23 is no exception.

The SBI Research and economists Surjit Bhalla, along with Karan Bhasin, made poverty estimates based on the latest survey, using different parameters. 

The SBI Research, headed by Bank group chief economic advisor Soumya Kanti Ghosh, took the Suresh Tendulkar Committee recommendations for 2011-12 as the base and adjusted it for the inflation rate over the years and also for free food items among other goods and services provided by the government schemes. It then concluded that the poverty rate declined significantly during 2022-23 compared to that in 2011-12 (see chart).

On the other hand, Bhalla and Bhasin used the World Bank poverty lines to estimate the poverty rate on the basis of the survey. They found that on the $1.9 per capita per day poverty line on the purchasing power parity (PPP), the poverty rate declined from 12.2 per cent during 2011-12 to two per cent during 2022-23. It should be noted that the Bank later revised the poverty line to $2.15 on a PPP basis at 2017 prices, but the two economists took $1.9 as the poverty line because it was at 2011 prices, which closely corresponds to the Tendulkar estimates for 2011-12.

At the $3.2 per capita per day poverty line, which the World Bank uses for lower-middle-income countries such as India, the poverty rate declined from 53.6 per cent during 2011-12 to 20.8 per cent during 2022-23. 

The Tendulkar Committee, appointed in 2005, shifted away from the calorie consumption-based poverty estimation prevalent till then. It based its calculations on the consumption of food items like cereal, pulses, milk, edible oil, non-vegetarian items, vegetables, etc, as well as fuel, clothing, footwear, education, medical, entertainment, personal & toilet goods, etc. It based its poverty estimates on household consumption expenditure surveys for 2011-12.

However, its poverty line, which comes to around Rs 27.2 a day per person in rural areas and Rs 33.33 for urban areas for 2011-12, created controversy, with many experts and political parties questioning the methodology. 

In the wake of criticism, the government appointed another committee under the chairmanship of C Rangarajan in 2012. The committee used certain normative levels of adequate nourishment, clothing, house rent, conveyance, education and also the behavioural determination of non-food expenses. It also considered average requirements of calories, protein and fats based on ICMR (Indian Council of Medical Research) norms differentiated by age and gender.

Based on this methodology, the Rangarajan Committee estimated poverty much higher than that of the Tendulkar panel (see chart) for 2011-12.

The Rangarajan committee also used different methodologies wherein a household is considered poor if it is unable to save.

Is it then correct to gauge poverty by adding inflation and imputation factors (to account for free food grains and other items given by the government schemes) to the Tendulkar committee recommendations or the World Bank poverty lines to gauge poverty rates based on household consumption expenditure survey for 2022-23? Or should an expert panel be appointed to come out with new poverty lines in reference to the new consumption expenditure survey?

Meanwhile, the Niti Aayog estimated the multi-dimensional poverty rate for 2015-16 and 2019-21 based on indicators such as nutrition, child and adolescent mortality, maternal health, years of schooling, drinking water, housing, electricity, assets, bank account etc. It found that the poverty rate declined significantly to nearly 15 per cent during 2019-21 from nearly 25 per cent during 2015-16. 

Rangarajan tells Business Standard that appointing another committee by itself does not serve the purpose because the committee, headed by him, discussed multiple indicator approaches, too.

"But we preferred the consumption expenditure survey route better than the multiple indicator approach," he says.

He says the best thing is to look at both the approaches -- household consumption expenditure survey and multi-dimensional index -- to arrive at poverty figures. "Each has its own merits," he says.

However, he says one can look at international experience in this regard.

"After all, the World Bank came out with the poverty line of $2.15 per person per day on a purchasing power parity (PPP) basis. How did they get that number? They are applying that number across all countries," he wonders.

Pronab Sen, chairperson of the Standing Committee on Statistics, thinks that there should be a new poverty line because the latest consumption expenditure survey is a different exercise from earlier ones on various parameters.

"The sampling methodology has been changed a little bit. The questionnaire is different. Most of the basic things remain the same, but the number of items has gone up hugely," he points out.

Sen, a former chief statistician, says earlier surveys were based on one visit to each household, but the latest one is based on three visits.

"That makes it non-comparable. There is no guarantee that the same person will answer questions in all three visits. So you have to do it all over again. You have to set up a committee to go into the issues and decide the poverty line all over again," he says. 

He reminds that the Tendulkar poverty line, like earlier poverty lines, was derived from consumption expenditure surveys. Those were not taken out of thin air, he says, adding those were not like the World Bank poverty line -- $2.15 per capita per day on PPP basis for the whole world.

"Our poverty line is specific. Rural and urban poverty lines are different poverty lines," Sen explains.

When asked whether the new committee should take the Tendulkar panel report as the base report or the one by Rangarajan, Sen says it should base it on the former or else there would be serious issues of non-comparability.

In comparison with the multi-dimensional poverty rate, he explains that the multi-dimensional index measures assets but does not measure income.

"The consumption expenditure survey is the measure of income which is current, whereas assets are a history. For instance, getting a house is an asset built up over time," he elucidates.

But which one should we rely on to count the poor in the country?

To this, Sen says one should go by the English language.

"In the English language, the words poor and poverty are applied to income. Lack of education, lack of health, and lack of housing are indicators of poverty and not the measurement of poverty. I prefer to call them multi-dimensional deprivation index," he says. 

Former chief statistician TCA Anant says the concept of the poverty line was created by a Planning Commission-constituted group in 1977 based on specific calorie requirements. It was revised two to three times since then. Those revisions basically retained the basic structure of the original line.

There is enough evidence to suggest that the basic calorie norm is an inadequate poverty measure in the sense that it only looks at the consumption of cereals; he says adding the composition of food production has changed enormously since then.

It would be a better idea to say that "that poverty" is over and finish that debate, he sums up.

"There are a set of people who are hung up on that debate. Till you don't admit that it has been solved, you won't move ahead," he warns.

In fact, the Niti Aayog report on multi-dimensional poverty is an attempt at addressing that problem, Anant points out.

"By all means create a new poverty measure. Nobody says that we have become a rich country. We are still a poor country. Identify dimensions on which we are poor rather than going circles around debating how poor we are now than we were in 1977. Accept that progress has been made and move on," he says.

Soumya Kanti Ghosh, who authored the SBI research note on poverty estimates, says any new poverty line would show a lower rate than estimated by his research note.

"We have taken a higher inflation rate than what was given by NSSO over this period. NSSO has taken CAGR (compounded annual growth rate), and we have computed based on yearly inflation data. Our estimates are conservative," he emphasises.

Ghosh says there is no harm in updating the Tendulkar poverty line. " A committee may be set up to create a new poverty line after the Census exercise is over," he suggests.

He points out that the poverty rates on the basis of the consumption expenditure survey and multi-dimensional index complement each other. "They are not substitutes," he opines.