Let the numbers speak

The discussion on the fall in employment after the November 8 demonetisation relies on anecdotal evidence. It is not backed by data.

Written by Bibek Debroy | Published: December 8, 2016 12:05 am
demonetisation, demonetisation black money, unemployment, unemployment ratio, unemployment rate india, job creation, currency demonetised, india jobs, india news, latest news For urban, the unemployment rate had declined to 6.89 per cent on October 31; it then rose to 7.05 per cent on November 30.

This is the season for quoting, and misquoting, John Maynard Keynes. “When the facts change, I change my mind. What do you do, sir?” This statement is attributed to Keynes, though there is no actual evidence Keynes ever said, or wrote, this. However, both Paul Samuelson and Joan Robinson have attributed this to Keynes. Statements have also been misattributed to Sherlock Holmes too. “Elementary, my dear Watson,” is the obvious example. But there is one statement clearly made by Sherlock Holmes. I am quoting from A Scandal in Bohemia”. There is a slightly different, and earlier, version of this statement in A Study in Scarlet. “It is a capital mistake to theorise before one has data, insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” Whether Keynes or Holmes, there is an emphasis on facts and data. Many problems with policymaking in India stem from the paucity of data. This is understandable for an economy where people are largely self-employed and work in the informal/unorganised, and even rural, sectors. These constraints are mentioned in the 2001 Report of National Statistical (Rangarajan) Commission and continue despite amendments to the Collection of Statistics Act (1953). Indeed, the present National Statistical Commission is examining ways to revamp India’s statistical system. In some places, there are holes in data. At other places, there are time-lags.

You can’t devise policy in 2016 on the basis of Census of 2001 or the National Sample Survey of 2011-2012. Sometimes, policy requires a census, surveys won’t suffice. The Socio Economic Caste Census is an instance of a census, which needs to be updated after 2011-2012, to identify beneficiaries of government programmes. Most assertions about MSMEs are based on a census conducted in 2006-07. In a way, calling it a census is unwarranted. It’s a census only for registered MSMEs. For unregistered MSMEs (which constitute an overwhelming bulk), the census actually presents projections based on a survey. There was a time when respectable surveys, particularly on consumption expenditure, income, access to public goods and services, were conducted only by the government — the NCAER being a bit of an exception. By “respectable”, I mean a survey with a proper sampling design and large samples, spread across all states — not one where one thrusts a microphone before 10 people in Delhi and extrapolates to the rest of the country.

To the best of my knowledge, there are two surveys that satisfy this respectability test. CMIE in collaboration with the BSE brings out two products. And there is PRICE (People Research of India’s Consumer Economy). PRICE did a survey in August 2016, interviewing more than 300,000 households, straddling 216 districts, 1,217 villages, 487 towns and 25 states. There was a survey in 2014. Therefore, when you look at the 2016 snapshot, you can find out improvements that have occurred in the two years. Since a panel of 12,000 households is constant between 2014 and 2016, that improvement scrutiny is even more robust. There is a wealth of data. Let me focus on a few. 98.7 per cent of the households have bank accounts (98.9 per cent urban, 98.6 per cent rural), a fact worth remembering when citing dated data on financial inclusion. (This doesn’t mean financial products have penetrated enough.) Between 2014 and 2016, income share of the bottom 20 per cent of the population has increased, while that of the top 20 per cent of the population has declined. Surplus income (defined as what is left of the income after routine expenditure is met) of the bottom 20 per cent has increased, while that of the top 20 per cent has declined. Between 2014 and 2016, the largest increase in income (three times the increase in national average) has been for urban labour, while the sharpest decline in income has been for rural labour employed in agriculture. Ninety-one per cent of households have electricity (a word of caution for those who quote census figures.) and 92 per cent of households own mobiles (also a caution for those who quote the census.), 57 per cent of households have LPG connections, while 55 per cent have tap water (for both these indicators, the census figures were less than half these numbers.)

For CMIE, the two products (available on the internet) relate to consumer sentiments and unemployment. There is again a wealth of data generated from 158,624 households. Take the unemployment rate, last updated on November 30. It’s given separately for rural and urban India, but let me cite all-India figures. That rate was 9.84 per cent on August 31, 8.97 per cent on September 30, 6.34 per cent on October 31 and 5.69 per cent on November 30. There is a tendency to ascribe everything to what happened on November 8 and also highlight rural distress, based on anecdotal accounts. The decline in rural unemployment rate mirrors the all-India trend. For urban, the unemployment rate had declined to 6.89 per cent on October 31; it then rose to 7.05 per cent on November 30. Till the end of November, this indicator doesn’t reveal a great loss in jobs, either rural or urban. In case one thinks unemployment rate is an imperfect indicator, let’s look at the consumer sentiments index. For rural India, the dip occurs from the end of August to the end of September and there is a steady upwards climb thereafter. There is no spike because of November 8. For urban India, the dip occurs from end-September to end-October, with a steady climb thereafter, with no spike because of November 8.

My intention, here, is not to debate the impact of November 8, but to point out that any debate must be supported with data, not anecdotes. CMIE’s unemployment product is a treasure trove of data, state-wise too. For instance, there seems to be a sharp increase in voluntary unemployment.

The writer is member, Niti Aayog. Views are personal.

For all the latest Opinion News, download Indian Express App

    Live Cricket Scores & Results