India has produced great statisticians, led by P.C. Mahalanobis. Their contributions to statistical sampling enabled speedy determination based on small samples, of a close approximation of gross domestic product, prices, prediction of election results and so on. Great applied economists have used statistics to gain very insightful information without counting all the numbers. All of us also tend to try and forecast future events on the basis of past statistics. Stock market players and investors are famous for speculating on future trends on the basis of past statistics and their understanding of the implications of events. Many mistakes are made because of wrong interpretation or using unrepresentative samples. Many vested interests also deliberately use statistics to mislead, confuse, and even lie. Politicians, shady entrepreneurs and investment advisors are some of them.
One example is represented by an editorial in The Economist about the latest report of the Intergovernmental Panel on Climate Change of the United Nations. Climate change forecasts are made by scientists who have studied many aspects that affect climate for years. Based on past observations over long periods, they forecast future trends. By nature, forecasts require assumptions that may be proved wrong. But factual observations cannot be ignored and have meaning for the future. Thus, global air temperatures did not rise as earlier forecasted, by 0.2 degrees per annum, for the decade since 1998. But the picture is different when a different base year is used. Basing the average on temperatures in 2000 shows that temperature rose by almost the forecast. The rise was higher than the rise in the earlier decade, the 1990s. It turns out that 1998 was an unusual year owing to the El Nino (an occasional warming of the Pacific Ocean), which boosted temperatures that year. Other such events affected some other forecasts. But the melting of the Arctic ice and the rise in sea levels is an observed fact, as are some other observations, and these have implications for climate change. However, anti-climate change groups (oil and mining companies among others) have used the more gradual warming to discredit IPCC warnings on climate change.
Politicians also selectively use statistics by changing the base year, quoting an exceptional year, calculating averages with and without exceptional years and so on to make a pre-conceived point. Thus, Narendra Modi claimed that the National Democratic Alliance government had delivered more GDP growth than the United Progressive Alliance government. He quoted a figure of 8.4 per cent GDP growth versus the recent estimates for 2012-13 of 4.5 to 5 per cent. P. Chidambaram, the finance minister, shot back that the average GDP growth over the NDA’s six-year rule was well below 8 per cent. The NDA showed 8.5 per cent only in its last year. But the UPA averaged over 8 per cent per year over its nine-year rule. Each is right in his figure. The conclusions are contradictory. Each politician picked the figure that supported his political argument.
Figures of GDP are not gospel truth; they are estimates and the estimates are corrected at least twice more before a ‘final’ figure is announced, usually in the third year. The ‘final’ figures in many cases have been very different from the earlier ‘quick’ estimates. But no one notices since the earlier figure has been repeated by so many that it is embedded in the memory.
India has a large ‘black’ economy. A considerable part of the rural economy is also outside the tax net and the markets. It has to be estimated periodically. The GDP figures, therefore, always underestimate the reality. The GDP estimates of both the actual amount and its growth rate do not catch the reality. But the media, politicians and most of us take these numbers as veritable truths. We tend to conclude that the GDP achieved in any year is because of the policies and actions of that year’s ruling government and offer credit or discredit.
For instance, the liberalization policies that Rajiv Gandhi’s government introduced in the later 1980s boosted the growth rates in the 1990s. His ‘broad banding’ of industrial licences so that a manufacturer could move into related lines without a new licence and his initiation of the rapid growth of telecommunications and information technology have expanded the economy, exports and India’s image.
P.V. Narasimha Rao’s government made dramatic economic policy changes. But these were built on Rajiv Gandhi’s policies. The years of NDA rule also benefited from the opening up of the economy under Narasimha Rao. The UPA-I government has since benefited from the infrastructure spending and social policies of Atal Bihari Vajpayee’s government. UPA-II splurged on social welfare expenditures to gain rural votes and damaged healthy macroeconomic parameters. Recent economic setbacks must be attributed to policies that started six years ago. A successor government in 2014 will have to clean up the macroeconomic mess left by UPA-II. One year is not representative of the effects of a current government’s policies. Not all the economic developments (good and bad) derive from only a current government’s policies.
We all know that a good part of government expenditures on infrastructure and social welfare is stolen, wasted, and does not reach whoever it is intended for. Rajiv Gandhi famously once said that only 17 per cent of government social welfare expenditure reaches those it is meant for — for example, the public distribution system, kerosene subsidy, now the rural employment guarantee scheme, the national health mission and so on. His figures were not confirmed by field investigations. But past studies by the National Council of Applied Economic Research and others have shown that about 40 per cent of cheap kerosene for the poor is diverted towards adulterating diesel for trucks. There are similar numbers for bogus ration cards, diverted ration foodgrains, and so on. Surveys of the Mahatma Gandhi National Rural Employment Guarantee Act in some states show that only 50 per cent or so of the money spent actually reached those it was meant for. But it is Rajiv Gandhi’s figure of 17 per cent of welfare expenditures reaching targets that is still widely believed.
We see the use of statistics to make preconceived political points in discussions on inflation. Politicians and the media in India measure inflation by the wholesale price index. This is released with a six-week gap. Only about 40 per cent of the items included in calculating the index consist of things which affect the consumer directly. The wholesale price index does not measure inflation for the consumer. Prices of industrial goods do not vary as sharply and frequently as do those of items bought by the consumer. The WPI may be relatively static when consumer prices are shooting up. Governments like to use it because it shows moderation in real inflation for the consumer. For the poor who buy in small quantities, prices are even higher.
Some weeks ago, an eminent economist said that there were 1,000 or so child deaths in India every day, because of inadequate food. This was contradicted by others. The figure of 1,000 is probably an exaggeration, but it helped the government in pushing for the food security bill. Statistics, true and false, most times based on many (unstated) assumptions and collection methods, influence us to believe in the effectiveness of one policy over another.
We should not give credence to statistics quoted by politicians, even when announced by the government. We must look closely at the statistics given to us. We must understand the underlying assumptions and methods of collection, or we would let the government and politicians mislead us about the state of our economy.