Saturday, March 04, 2006

Politics of statistics

This [NSSO consumer expenditure] survey was conducted between July 2004 and June 2005, i.e. it was completed more than eight months ago. One would have thought that in this age of computers and Infosys, the government would be able to process the survey information a bit faster than the days of hand-held calculators. In a parallel situation, the NDA government released preliminary results of the 1999-00 survey in September of 2000, just three months after the completion of the survey.

But the Congress is in a dilemma- if the poverty reduction according to the NSS data is "large", even its savvy public relations experts would find it difficult to take credit; the Congress-led UPA has been in power only since May 2004. Only if the results of the 2004-05 survey are "bad" i.e. reveal low poverty decline, will the UPA have an empirical (as opposed to ideological) basis for asserting the need for a Mrs Indira Gandhi- or Venezuela's Hugo Chavez-style populism. And the data, when released, are unlikely to support populist policies- hence, the release of the data post-Budget 2006.

From this this piece by Surjit Bhalla. The piece is filled with polemic (which can safely be ignored), with some speculations about what the results of NSSO survey are likely to be. We will just have to wait and see if what he says is true.