Off the Mark

It is time for the government to consider getting out of forecasting and leaving it to the pros. Forecasting forms an integral part of decision making in any organisation, especially if it has a client base of 1.2 billion. The Indian government has not covered itself with glory in getting the forecasts wrong on a consistent basis. Let us look at inflation first. Even if we keep aside the fact that the overhyped figure has a very limited influence on our daily lives, it is interesting that the government even manages to estimate it incorrectly on a regular basis. Another crucial area where forecasting is the key is weather. ‘Met gets it wrong, monsoon delayed’ is an often encountered headline. Often weather forecasts take to obfuscation like “although IMD had prepared this year’s forecast over Kerala on the basis of improved statistical models applying six predictors, ground stations from the southern state have not reported favourably”. In a nation where the majority of the population stares at the sky to find out whether they will survive this season, predicting the monsoon incorrectly, failing to identify a possible drought can and have been very costly. And we all are aware of how the monsoon impacts farmers directly and consumers indirectly, in the form of food price inflation. Thus, a wrong met prediction leads to wrong food price inflation estimation… and the cycle goes on. We totally agree that in weather, tiny differences in input could quickly become overwhelming differences in output, something called the Butterfly Effect — the notion that a butterfly stirring the air today in Tokyo can transform storm systems next month in Manhattan. But isn’t that why the taxpaying common man is paying through his nose for the supercomputers and the ‘superbrains’? The irony is that almost every department has its own forecasting team of eminent statisticians and economists. The decisions that are taken by the government, which determines our fate, are solely based on these projections and they are at times way beyond the accepted level of statistical error of 5 per cent. The performance of our five-year plans is a clear indication of the potency, rather the lack of it, of our forecasting prowess. One may wonder how is it possible to have a five-year planning process along with a market economy, but let’s not get into that. Even if we look at the government’s annual budget estimates, the poor marksmanship is obvious at the end of the year — year after year. Business organisations, which have a big stake in getting accurate forecasting in areas of their interest, increasingly rely on private and paid sources rather than on free public data. The time has come for the government to realise that forecasting is a professional job, which is best outsourced to forecasting agencies. Collecting data in a vast country of great diversity and 1.21 billion people is not easy and government agencies such as the CSO, which collects national accounting and industrial production statistics, and RGI, responsible for conducting census of India every 10 years, are doing a fine job on that front. While data collection and population enumeration have to remain with the government departments and data made available in public domain, forecasting should be left to professional agencies that have to compete and be accountable to their paying clients, including the government, for the accuracy of their forecasts. Public policies and programmes will stand to gain the most if forecasting is on the mark.




Comments are closed.