You are here: Home » Opinion » Columns
Business Standard

Busting a few myths about poverty

While economic growth may reduce poverty, no systematic correlation between the pace of poverty decline and the pace of economic growth can be found in the data

Ajit K Ghose 

Ajit K Ghose

Recent publication of new numbers on by the has revived some old controversies. The numbers show that while the incidence of declined both between 1993-94 and 2004-05 and between 2004-05 and 2011-12, the pace of decline was much faster in the second period. This has led to a fresh airing of some old views. One is that the line, defined by an Expert Committee chaired by Professor in 2009, is just too low. The implicit argument is that the observed decline in looks impressive precisely because the line is so low and a higher line would have shown rather different trends. A second view that has got a fresh airing is that it is that reduces poverty; the more rapid the economic growth, the more rapid is decline. India’s own experience appears to provide supportive evidence; the growth of per capita income was much faster between 2004-05 and 2011-12, and so was the pace of decline. Whether or not the line is too low is not an issue that can be fruitfully debated. Arbitrariness is impossible to avoid in defining a line, which essentially expresses a set of social norms. Consensus on such norms is not to be expected. The Tendulkar Committee explicitly recognised this and abandoned the earlier practice of anchoring the line to a calorie norm. But its independently established minimum standards for various forms of expenditure were again ultimately arbitrary. The has now constituted another Committee of Experts, which will redefine the line in due course. But there is no reason to expect that the new line will be universally judged as appropriate. Redefining the line again and again is really a pointless exercise. It is much more useful to examine the sensitivity of trends to changes in the line. We can usefully study the alternative trends that emerge when we use different multiples of the current line — 1.25 times the line, 1.5 times the line, 1.75 times the line, and so on. In fact, the Planning Commission’s Press Note on Estimates, 2011-12 reports the results of an exercise of this kind but makes too little of this; it only points out that decline is faster in the second period irrespective of the line used. But the exercise also shows that the larger the multiple of the Tendulkar line, the slower is the pace of decline in both periods.

A possible explanation is that the lower the line, the larger is likely to be the percentage of the poor just below the line. It would be useful to systematically carry out such exercises in greater detail and analyse the results. We can then have a better understanding of what kind of is showing what trends. As for the view that it is that reduces poverty, it needs to be said straightaway that nobody has ever argued that does not reduce What has been contested is the idea that there is a linear relationship between the pace of and the pace of reduction. And a fresh look at India’s experience tells us why. If we look at the full set of estimates (see Table) available, the first thing we notice is that, at the all-India level, the decline in extreme was much faster during the period 2009-10 to 2011-12, when the pace of had slowed down, than during 2004-05 to 2009-10, when the pace of had been unprecedentedly rapid. The Planning Commission’s Press Note manages to hide this uncomfortable fact by focusing on the overall trend. It does mention that the whole reason for conducting a special survey in 2011-12 was that there was a drought in 2009-10. As such, it is likely that there was a temporary rise (or at least a slowdown in decline) in the incidence of extreme in that year. Hence, the Note seems to be implying, the trend observed for the entire period since 2004-05 is truer. As it happens, however, 2004-05 was also a drought year. Agricultural output declined by 1.6 per cent between 2003-04 and 2004-05, as it did by 1.4 per cent between 2008-09 and 2009-10. This means that the decline in the incidence of extreme is being underestimated for the period to 2004-05 (because the terminal year was not normal) and overestimated for the period from 2004-05 (because the initial year was not normal). As such, no firm conclusions about the effect of higher growth on can really be drawn from the comparison. A look at state-level estimates gives us more reasons for caution. In Andhra Pradesh, the incidence of apparently declined by 1.76 percentage points per annum between 2004-05 and 2009-10 and by 5.95 percentage points per annum between 2009-10 and 2011-12; yet the growth of per capita income was 7.9 per cent per annum in the first period and 7.4 per cent per annum in the second period. In Bihar, the incidence of declined by 0.55 percentage points per annum between 1993-94 and 2004-05, by 0.18 percentage points per annum between 2004-05 and 2009-10 and by as much as 9.90 percentage points per annum between 2009-10 and 2011-12; the growth of per capita income in the three periods was 2.2 per cent, 8.7 per cent and 12.7 per cent per annum respectively. More such cases could be cited, but two points are already clear: that the decline in extreme observed for the period since 2004-05 is vastly exaggerated, and that no systematic relation between the pace of decline and the pace of can be found. Where does all this leave us? We need to avoid sterile controversies and engage instead in serious analysis. There is nothing to be gained by constituting a Committee of Experts every few years to revise the line. There is much to be gained by studying the effects of changes in the line on time-trends in There is nothing to be gained by arguing about whether or not growth reduces There is much to be gained from serious analysis of the nature of the relationship between growth and This requires us to distinguish the effects of short-run fluctuations from those of growth. And the purpose of analysis is not to develop some simplistic view of how the rate of growth relates to the rate of decline. The purpose is to understand what kind of growth has what kind of effect on and why.

The author is, currently, Honorary Professor at the Institute for Human Development, New Delhi and, formerly, Senior Economist at the International Labour Office, Geneva

First Published: Sat, August 31 2013. 21:50 IST