Opinion
-
Editorials
Grappling with estimates
THE BAD NEWS continues. Even if it is about the past, the revelation by the Central Statistical Organisation (CSO) that the gross domestic product in the year 2000-01 grew at four per cent does not reduce the gloom. That year was widely expected to register a healthy growth rate. When the year started, the CSO made an advance estimate of six per cent in January 2001, this was revised to 5.2 per cent in June 2001 (revised estimates) and now, the country is told that the growth rate could be only four per cent (quick estimates). This, however, does not necessarily mean that the bad news is over as the final figures could well throw up another grim piece of statistic. The data generation process of the CSO is quite well known. First comes the Advance Estimate, followed by the Revised, Quick, Provisional and Final estimates. In a way, the Organisation has only been true to its past as it constantly updates figures. The Economic Survey of 2000-01, for instance, gave revised CSO estimates of the GDP for as far back as 1994-95 and 1996-97.
It is easy to charge the Organisation with tailoring statistics, but before that is done it will be in order to recognise that there are some real computational difficulties. One reason is the continued inflow of revised information on industrial production, agricultural output and services. While this has been the practice of the past, upward revisions are normally accepted easier than downward ones. Another difficulty is the problem with analysing data from the services sector, which has steadily increased its share to the national income. Wide differences, for instance, have been reported in the Revised and Quick estimates. Under the broad heads of trade, financial services, community, social and personal services, the estimates have varied significantly. Some of the differences in the figures mentioned for the Revised and Quick estimates are pronounced. Financial services was expected to grow at 9.1 per cent in the Revised estimate, but the latest information places the growth at 2.9 per cent. The economic activities classified under trade, hotels, transport and communication in the Revised estimates were expected to grow at 6.9 per cent. In the Quick estimates this category has been re-defined into two groups: trade, hotels and restaurants (3.8 per cent) and transport, storage communication (8.2 per cent). Three areas of economic activity — manufacturing, electricity, gas & water supply and construction — had upward revisions.
Now that a comparison between the two sets of figures shows the broad reasons for the dip in the projections, there are important questions that should be addressed. Quite often, statistical bodies find themselves accused of jugglery to paint a rosy picture or to meet an immediate non-economic compulsion. If that were the case, the Economic Survey of 2000-01, presented by the National Democratic Alliance Government, need not have mentioned a better performance for the year 1996-97, when the country was under the rule of the United Front. However, this is not to say that all is well with the habit of re-casting growth rates. The exercises that measure economic performance will have to be upgraded, especially when it comes to making projections based on available trends and data. Making data-collection and analysis a more reliable and credible operation calls for a move away from the present dependence on Ministerial and Governmental resources. If there is one real purpose to be served by the sense of disillusionment caused by the CSO's latest revelations, it is the realisation that the formation of an independent statistical authority, as suggested by the Andhra Pradesh Governor, C. Rangarajan, can be delayed no longer.
Send this article to Friends by
E-Mail
Opinion
|
