During the last few years, a policy and media environment has been created that suggests there is a serious crisis in agriculture. That growth rates in agriculture have hovered around 2%, a large percentage of farmers (more than 40%) do not want to continue with farming, many of them are committing suicides, our food security is at threat, and so on. Perhaps, such an environment has encouraged the finance minister to announce a loan waiver for the farming community in his Budget speech which may cost the exchequer Rs60,000 crore or so.
Much of this perceived crisis in agriculture is often created by the “advance” estimates of agricultural growth released by the government early in the year, based on very preliminary information about area sown and some guesstimates about yields.
For example, an advance estimate of agricultural GDP (gross domestic product) growth for the year 2007-08 is given at 2.6%, way below the targeted growth rate of 4%. A similar “advance” estimate was given for the year 2006-07 at 2.7%. But how many people noted that in February 2008, the Central Statistical Organisation (CSO) released the “quick” estimate of agriculture growth rate for the year 2006-07 as 3.8%, revising the “advance” estimate of 2.7% by a whopping 40%?
This is not the first time that the deviation between “advance” and “quick” estimates has been so wide. For the year 2005-06, the deviation was even bigger, with an “advance” estimate of 2.3% of agricultural GDP growth getting revised to a “quick” estimate of 6%—an unbelievable 160% rise.
If this is the level of variation in our growth estimates for agriculture, there is something seriously wrong not as much with our agriculture, but perhaps with our statistical system. It raises concerns regarding the credibility of our growth estimates, and also questions the wisdom of relying on such poor “advance” estimates for taking any rational policy decision in time. This has often cost the nation heavily. For example, in case of sugar cane, the “first” advance estimate released by the ministry of agriculture was 283.4 million tonnes (mt), while the “final” is put at 355mt.
Based on the “advance” estimates, the government followed a policy of controls on sugar exports, but the reality was that it was a bumper crop which created a glut of sugar in the market. Later on, the government had to intervene to clear the glut by announcing sops to the industry for stocking and exports of sugar.
It is a highly inefficient policy, costing the exchequer heavily. The reason: poor “advance” estimates of sugar cane. This happens not only with sugar cane, but a similar story is repeated with cotton, and many other crops, even grains such as wheat and rice. It seems a case of utter failure in economic intelligence of agricultural produce estimates. Given this state of affairs, one is obviously provoked to question the reliability of these numbers as also the credibility of the method of estimation. The gap between advance, revised, quick and final estimates in agriculture is too large, logically unacceptable, and could render the entire policy vision upside down.
CSO has been coordinating most of the statistical activities in the country for a long time and has had a legacy of maintaining high statistical standards among the developing world. The significance of these numbers is huge as they form the basis of any policy decision and the future course of action in different sectors of the economy. While revisions are tenable, it is hard to digest the rationality of a swing from bust to boom. If one has to make any sense out of these numbers, there should be more clarity on the accompanying factors that steer such revisions. The statistical system for agriculture seems to be in the doldrums, be it the statistics relating to horticulture or livestock or commercial crops such as cotton or sugar cane.
The people and institutions collecting that information and revising it need to be given due incentives for better and accurate information, and also made accountable in case of poor performance. At least an attempt should be made to provide an accompanying note that adequately explains the wide deviations.
What is urgently needed is an overhauling in the centralized data monitoring and compilation system. The central agency need not just sit back and wait for any information from the state agencies, but be proactive and cross-check information from alternative sources closer to the ground realities.
The agriculture sector, to say the least, seems to be held hostage to this number game—and the policymakers and media alike have been crying wolf. The sector has been growing at a trend growth rate of 2.95% over the last two decades, since 1985, albeit with wide fluctuations. This may be below the targeted growth rate of 4% in the last two Plans, but it is not a crisis situation.
There may be some under-performance, which needs to be fixed, but before that we also need to fix our agricultural statistical system.
Ashok Gulati is director in Asia, and Kavery Ganguly a research analyst at the International Food Policy Research Institute. Comments are welcome at email@example.com