The Hindu Business Line : Data doubts

At a recent conference of States and Union Territories, Prime Minister Manmohan Singh rightly expressed concern over the reliability of India’s macroeconomic data. If figures as basic as Gross Domestic Product are way off the mark, it will lead to wrong investment decisions and policy choices. While the Prime Minister pointed to difficulties in data collection in agriculture, the problem extends to industry and services as well. Admittedly, it is no mean task to impr ove the quality of data in a heterogenous, fragmented economy, where the unorganised sector accounts for almost half the national product. The output of the unorganised sector is derived as the product of gross value added (GVA) per worker and the workforce in a particular sector, both of which are uncertain estimates.

However, data on the organised private sector are not reliable either. After industrial licensing was scrapped industries were no longer required to submit their output data to the government. Sample studies by the Reserve Bank of India on the finances of private and public limited companies are used to derive GVA per worker and sectoral output. But these surveys are not reliable, as their sample size is small and extrapolations are made on the basis of growth of paid-up capital when current data are unavailable. Besides, workforce size is hard to estimate for establishments employing 10-25 workers. The data for the public sector, based on budget documents and annual reports of enterprises, are perhaps the most reliable of all.

Service sector output figures, derived from Census workforce data and enterprise and employment surveys conducted by the National Sample Survey Organisation (NSSO), are considered inaccurate for a variety of reasons. If the sample size is small, it could distort results in sectors marked by the co-existence of small, medium and large players. The NSS survey design will need to be changed in some cases to capture this heterogeneity. While enterprise surveys are done about once in five years, output in the intervening period is calculated on the basis of questionable quantity and price indices. When the country’s entire output estimates are based on such approximations, it is a serious issue. Not surprisingly, there is a huge gap between the NSS’ monthly per capita consumption expenditure figure of about Rs 720 and the monthly per capita income estimate of about Rs 2,700 arrived at through the output route. The National Statistical Commission has a big mess to sort out. Its task would be that much easier if the Government were to push for financial inclusion of enterprises and individuals and improve tax governance.

Related Stories:
Evolve alternative methods to generate industrial data: PM