‘Bihar’s data on consumption expenditure creates serious noise factor in poverty analysis’

Amitabh Kundu,Professor (Economics) at JNU speaks to Surabhi.

Written by Surabhi | Published: July 24, 2013 1:34:27 am

Amitabh Kundu,Professor (Economics) at Jawaharlal Nehru University and former member of National Statistical Commission,in an interview with Surabhi,says that even though poverty level have declined sharply,state level data presents anomalies that the NSC must look into. Excerpts:

Poverty estimates are compiled every five years. Why have the estimates been prepared for 2011-12 when the last figures were released for 2009-10?

Poverty: Read full report

2009-10 was an exceptionally bad year for agriculture,which brought down the overall growth. So the annual rate of decline in poverty during 2004-05 and 2009-10 was 1.5 percentage point only. With the high growth in national income since 2003,it was expected that poverty rate would have dropped sharply. The 12th Plan notes that the decline in poverty has been 2 percentage points annually between 2004-05 and 2011-12. The special survey with large sample size in 2011-12 was conducted to validate this. However,if the survey covered 2012-13,which saw a low growth in income,the poverty outcome would not have been that satisfying.

Is the decline in poverty sharper in rural areas than urban areas?

Data from 2011-12 shows that rural consumption expenditure grew by 9 per cent per annum during the preceding two-year period while the corresponding urban figure was 6 per cent only. So the decline in poverty level should be higher in rural than urban areas. Given the data related problems and the short time gap between the two rounds,one should not draw very strong conclusions about rural urban gaps. But poverty decline was higher in urban than rural areas during the five-year period before 2009-10,accentuating rural urban disparity significantly.

Has there really been such a dramatic drop in poverty levels or is there some statistical problem?

I have to refer to unit level or at least state level data. Rajasthan,for example,has shown a stable and high decline in poverty over the past two decades,which follows a trend. However,Bihar emerges as an enigma. Here,rural poverty levels can be estimated to have dropped by over 21 percentage points. This is as unbelievable as the data shows that Bihar’s poverty levels during 2004-05 and 2009-10 had gone up or remained stable in rural and urban areas. Bihar,which recorded the highest growth rate among states during this period,must have made an impact on consumption and poverty. But this was not captured by the NSS in 2009-10. So the growth rate of consumption in the subsequent two-year period works out to be phenomenally high. Interestingly,if you exclude Bihar from the poverty calculation at national level,the decline in rural poverty would come down by 2 percentage points and rural poverty decline would be less than that in urban areas. Bihar’s data on consumption expenditure creates a serious noise factor in the analysis. These anomalies increase the skepticism of the civil society in the reliability of the estimates. The NSC must probe into this problem.

Do you think the conclusions from consumption expenditure data go on well with the recent job trends?

There is a paradoxical situation. Employment rate has gone down by all four alternate measures employed by NSS. How can this coincide with a high growth in consumption and rapid decline in poverty? But the decline in employment is very high by usual principal status,particularly in rural areas. When we include subsidiary employment,the decline is not that sharp.

For all the latest News Archive News, download Indian Express App