The great Indian poverty game-Sonalde Desai
Nowhere are the argumentative Indians more visible than in the cacophony surrounding poverty estimates. Poverty is declining; inequality is increasing; no one can live on Rs 28 a day; nine per cent of Indians are poor; 70 per cent of Indians are poor.
Poverty is too important to be used as ping-pong between optimists and pessimists on the Indian economy. I am deeply disillusioned to discover that there are no certainties in economic statistics, including the ones I am about to report below. But reasonable approximations of reality suggest the following.
First, incomes, as proxied for by consumption expenditure, have gone up in both urban and rural areas, even at the bottom of the income distribution. In constant prices, per capita monthly expenditure of the lowest 10 per cent of population in rural areas increased by Rs 55 between 2004-05 and 2009-10. However, those at the top gained the most, as can be seen by the divergence of the lines in the graph.
Second, our definitions of poverty are so outdated that these data tell us nothing about the number of poor in India. While the Tendulkar Committee report devoted considerable attention to tinkering around with urban-rural price differentials, it made no effort to evaluate the validity of the original poverty line, proposed in the report of a task force appointed by the Planning Commission in 1979. Such is the foundation on which all Indian poverty estimates are built — we have no idea of their validity. In 1979, the task force estimated, most households were receiving adequate calories in 1973-74 at the level of Rs 1.64 per person per day in rural areas and Rs 1.94 per person per day in urban areas. These estimates were based on the 28th round of the National Sample Survey (NSS), in 1973-74.
There are two serious problems with this estimate. One, as G C Manna, deputy director-general of the National Sample Survey Organisation (NSSO), has perceptively pointed out in an article in The Economic and Political Weekly, the 28th round sample contained just 23,348 households. Had the line been drawn using the prior round, which surveyed 125,000 households, then – using the same methodology – it would have been about 17 per cent higher.
The other problem is that the calorie-content norms used at that time – 2095 for urban residents and 2435 for rural residents – are totally outdated. The new norm recommended by the Food and Agriculture Organisation for India is about 1800 calories per day for both urban and rural residents.
Thus, a new poverty line is likely to be drawn at a very different level. There are reasons to believe it could be higher; there are reasons to believe it might be lower. We have no way of drawing any conclusions.
Third, the debates over poverty line serve no real purpose. In theory, it determines the level at which a family is defined as being “below the poverty line” (BPL), making it eligible for food subsidies as well as other welfare benefits. In practice, however, poverty status, defined by consumption expenditure, has only a weak relationship with ownership of a BPL card. NSS data from 2004-05 show that only 40 per cent of the poorest quintile of rural households had a BPL card, while 15 per cent of the richest quintile possessed a BPL card.
Moreover, when state and local governments provide BPL cards, they use very different criteria from those used by the Centre and have a vested interest in inflating the number of BPL cardholders. Thus, as 2004-05 NSS data indicate, rural residents in Gujarat and Karnataka were far more likely to have a BPL card than those in Bihar, something that is at odds with the comparative prevalence of poverty in these states. All the Centre is able to do is cap the number of BPL cards for which it will provide subsidies for the states using NSS poverty data — and even that is unlikely to hold firm given coalition politics.
At a time when a variety of urgent issues demand our attention, it makes little sense to indulge in this breast-beating. If we are looking at a poverty line to set up a scorecard for the nation, let us compare the incomes of individuals at various income deciles over time to see if any growth is taking place, and whether this growth is concentrated among the poorest or the richest. If we are searching for a litmus test to help us identify the deserving poor and shoo off the parasitical rich, there is nothing in this debate that is going to help us. We will have to come up with different standards and develop greater tolerance for uncertainty and leakages. In a nation led by an eminent economist, is some degree of sanity in a whirlpool of statistics too much to ask?
The writer is a senior fellow at the National Council of Applied Economic Research and a professor of sociology at the University of Maryland