Let's increase poverty

The 55th Round of the National Sample Survey (1999-2000) led to a lot of debate on poverty levels, mostly centred on the method of collecting data compared to earlier rounds. Poverty has many dimensions and income poverty is only one indicator.

However, it is the indicator commonly used, even though poverty ratios really capture expenditure or consumption poverty rather than income poverty.

Yet another the NSS large sample has collected data for 2004-05 and the results are expected in 2006. Many of these issues will be revisited then.

Three propositions should be evident. First, the growth rate has increased in post-reform India, with variations among states.

Second, since there is no evidence of expenditure distribution worsening, the trickle down effect should lead to poverty reduction.

Third, since income distributions are log normal, as the thick part of the distribution passes above the poverty line, decreases in head count ratios can be sharp.

Hence, no one should be surprised if NSS throws up a poverty ratio of around no more 20 per cent in 2006; extrapolated, India could have a poverty ratio of around 12 per cent in 2015.

However, poverty numbers are also a function of the level of development. As average income rises, so should the measurement bar. The US poverty line is not the same as the Indian.

In 1993, an expert group examined issues related to the poverty line, but did not question the core thrust of the Indian poverty line, enunciated by a study group in 1962.

The 1962 group had argued that the poverty line should be primarily based on food, and health and education should be excluded because these two would be provided by the state and were irrelevant for purposes of private consumption.

At its core, the Indian poverty line is still based on food for 80 per cent of the basket, with another 20 per cent thrown in for clothing and housing. But the proposition that health and education will be state-provided is now questionable.

Beginning especially with the 55th Round of the NSS, there is evidence of expenditure switching from food to non-food items, even for the poor.

In part, this may be the effect of a shift consequent to income growth. But it may also be due to increased recourse to private provisioning for both health and education.

Whatever the reason, the 1962 premise should now be discarded and food should account for a diminished share in the poverty line. Reflecting India's economic growth, there should be a new poverty line.

Calorie norms are used for food items, but their non-applicability is not a serious problem for non-food items. For instance, in the US, food costs are multiplied by a factor to obtain the poverty line, this factor is a function of the importance of food expenditure in the poor's consumption basket.

Nor should the construction of a time-series that goes back and permits inter-temporal comparisons be that much of a problem. In 1993, poverty ratios were recomputed backwards till 1983.

There is of course a political problem, because raising the poverty bar will mean that poverty ratios will immediately increase and everyone will not scrutinise differences in definitions. But any increase in poverty should gladden the hearts of the Left, which has a vested interest in high numbers.

That apart, there will be greater consistency with the 81.3 million ration cards issued by states for people allegedly below the poverty line!