|
At the state statistics ministers conference, the prime minister questioned the quality of industrial and agricultural output data. This was the first statistics ministers conference. In June 2007, on Mahalanobis’ birth anniversary, we celebrated the first Statistics Day. That first-ever event also saw the PM giving a speech, where he had lamented the lack of credibility and transparency of India’s statistical system. It is good question to ask what the government did between June 2007 and August 2008 to improve data quality. A National Statistical Commission was established in 2006. But its presence has been, shall we say, statistically insignificant. Among developing countries, India’s statistical system isn’t bad. But India should be looking at different standards. There are time-lag issues in India data. There are problems of the kind flagged by the Rangarajan-chaired Statistical Commission’s report in 2001. Some relate to the unsatisfactory nature of data for unregistered/informal sectors. This affects agriculture, industry and services and leads to gaps between advance and final estimates. The PM mentioned lack of data-sharing across ministries and departments, a habit that negatively affects collective evaluation of public expenditure. His solutions include third party audit (using talent in universities, management and research institutes, NGOs). Using the Right to Informatuon Act, a very fine piece of legislation, may help in getting better data on specific programmes but isn’t much use when it comes to macro data.
Taxpayers, as the PM said, pay for information collection. What’s the best use of that money? T here is some evidence that the quality of government data has deteriorated since 1991, probably due to human resource and financial constraints. In the same period, the private sector has entered the data collection/processing/forecasting business. Private players do this more efficiently than the government does. However, this cannot be a full substitute to official statistics because only the government has the legislative backing to ask for mandatory disclosure; the power comes from the Collection of Statistics Act, 1953. So the best option is that the government outsource data collection and processing, while ensuring confidentiality. But there should be a caveat. Outsourcing should be to large organisations, not small and fragmented agencies. Therefore the official proposal on using postmen to collect data for a rural consumer price index (CPI) is a politically correct, methodologically horrendous idea.
|
