A debate has broken out in British Press about the quality of economic forecasting. When the Brexit referendum voted to leave by a majority of 52% to 48%, there was a shock.
A debate has broken out in British Press about the quality of economic forecasting. When the Brexit referendum voted to leave by a majority of 52% to 48%, there was a shock. The Sterling collapsed overnight. Dire forecasts were made about a recession. In the event, the British economy not only failed to have a recession but is performing surprisingly well. The estimates for GDP growth in the fourth quarter of 2016 have just been revised upwards from 0.6% to 0.7%. That may not seem much, but it comes after two quarters of growth of 0.5% which surprised many.
It is almost becoming fashionable for economists to doubt their ability to forecast. Not everywhere, of course.
The demonetisation exercise in India inspired a lot of forecasts even before the 50 days had passed since November 8. Manmohan Singh forecast a reduction of two percentage points in the growth rate. The maximum loss was put at 3.5%. Some people were confused between levels and growth rates. They said GDP would decline by 0.5%. Since, in the first two quarters of FY17, growth had been estimated at around 7%, if we put GDP at 100 for the beginning of the fiscal year, it would have been 103.5 by end of second quarter of the fiscal. For GDP to go down 0.5%, we would need GDP growth to be minus 4% in the next two quarters.
Of course, we do not know what the growth rate would be for the full FY17. Indeed, the estimates, whenever they come out, will be subject to revisions and corrections. It is unlikely that the question would be settled before sometime in late 2018. It will become a factor in the general election of 2019. Indeed, the early forecasts of the impact of demonetisation could have an impact on the elections going on at the moment. whether they are validated by data later on or not.
This illustrates the paradox of economic forecasting. Everyone knows, and often says, that economic forecasts are no good. Yet, they use them as political missiles. The Opposition will claim that demonetisation was disastrous and things are bad even now. The ruling party will talk about the political payoff of reduced corruption or depriving terrorists of fake currency. Either way, the truth will not be known for a while though the political impact will be felt by March 11.
Why can’t economists forecast? I would distinguish between routine forecasts in absence of large shocks such as Brexit vote or demonetisation. In such normal circumstances, today is the best estimate of what will happen tomorrow. In technical terms, first-order auto regression is the best forecast. It is when there is a shock when such may not be the case. It is difficult to identify when a change is a shock. It has to be unanticipated as the demonetisation announcement was. The Brexit vote on the other hand did not change everyday economic life at all. Various people said expectations about the future will change and investment may suffer. But, that was all about a year or two ahead.
Thus, after the referendum, the British economy chugged on as usual. Sterling depreciated almost overnight. That apart, no other change took place. So, today was a good predictor of tomorrow. With demonetisation, the shock was unanticipated and large. Even so, the final effect can be difficult to forecast. This is because while the amount of cash declined sharply from the second half of November onward and did not recover till mid-February, the effect will be different by type of goods. Those goods which are bought with high frequency—daily, such as vegetables—were not bought as much while cash was scarce. But since checks were still negotiable (plus debit cards and credit cards), large ticket items need not have suffered decline. This is because money being cash plus deposits, deposits increased while cash declined. If big ticket items were bought at the usual pace (except for jewellery and real estate) while small ticket items reduced, the overall impact on consumers expenditure may not be large.
You might also want to see this:
Economics has no easy answers. Even when we may have accurate data, forecasting the future is difficult, except in a smoothly-growing economy. But, of course, data are neither available nor accurate. There are still debates about how much old currency was returned and how much never came back. We may not know till June 2016. The government had thought that idle cash-hoards would be devalourised by the suddenness of the announcement. People laundered their money by buying gold or jewellery at above normal prices. Thus, they took a loss, but not of 100%, on their cash hoard. In absence of demonetisation, the transactions would have gone unrecorded or would have been recorded at the lower (false) white prices. If so, the final effect on consumers expenditure would be positive!
The physicist Max Planck said to the philosopher-mathematician Bertrand Russell that after talking to the economist Maynard Keynes, he had decided that economics was too difficult a subject. He would stick to physics.
The author is a prominent economist and Labour peer