India’s data problem is real, fixing quality a pressing priority
The issue of data quality in national surveys has reared its head again. A member of the Economic Advisory Council to the Prime Minister (EAC-PM) recently claimed national surveys like the National Sample Survey (NSS), National Family Health Survey (NFHS), and the Periodic Labour Force Survey (PLFS) are based on faulty sampling. The government believes the samplings are based on obsolete methodologies and do not represent the population accurately, and hence they underestimate India’s progress and development. Although the EAC-PM member questioning the ‘quality’ of government data is baffling, it also raises critical questions: What is the government doing to fix the problem (if at all there is one)? Is the data being questioned only because it does not concur with the government narrative on the economy?
The data collected by the National Sample Survey Office (NSSO) has been questioned by all sides of the political divide as well as by academicians, but for varied reasons. The incumbent government has been continuously at loggerheads with the NSSO because the former claims the data in the latter’s surveys—NSS and PLFS—are not reflective of the progress India has made recently vis-à-vis economic and social indicators. Others have criticised the data for underestimating poverty and distress because of its non-coverage of several politically sensitive areas. The recent debate over India’s actual ‘poverty’ rate further exposed India’s data problem. While pro-government economists passionately argued that the poverty rate in India has fallen, others have argued that poverty levels have remained stagnant over the last decade. If prominent economists cite the same data and arrive at inferences which are poles apart, it does not augur well for the nation’s policymaking.
The data quality gets even poorer at the state level. Agriculture and labour are two critical areas that come under the states’ statistical systems. Statisticians have suggested more frequent changes in sampling frames, base year, etc., to correctly capture the change in consumption and employment patterns. The quality of data is critical for the government to frame economic and social sector policies. And if there is an issue with the quality, it needs to be fixed. And while the intellectual honesty of those who politicise data can be countered with one argument or another, the quality of data cannot be compromised.