CMIE uses thin surveys to estimate unemployment and also links this with falling investments. But investments are rising and there are better sources of data on jobs.
By TV Mohandas Pai & Yash Baid
There have been several articles in the media in the past month raising the issue of ‘increasing unemployment’. These reports cite ‘data’ from blogs written by the think tank, CMIE (the Centre for Monitoring Indian Economy Private Ltd.). The CMIE has been putting out alarmist reports without adequate data from a thin survey to back their stance. Recently, the CMIE put out a report saying there has been a reduction of 11 million jobs in the year 2018.
Their analysis is based on a paper-thin sample survey conducted on just 1,40,000 (bit.ly/2QN2XlT) respondents to determine job trends for a nation of over 1.3 billion people. Further, the results from the survey are based on preliminary estimates two months ahead of schedule. They promise to complete the survey by February 2019 and a recent release says more or less the same! Looking at their methodology for these claims, we notice that their purported job loss calculations are based on falling investment proposals. Quoting Mahesh Vyas, the MD and CEO of CMIE from a recent article in The Hindu (bit.ly/2S2BXPZ), “new investment proposals had peaked at Rs 25 trillion in 2010-11. In 2017-18, these were down to Rs 11 trillion and, in 2018-19, these are unlikely to cross Rs 10 trillion. The impact of this fall in investments is visible in shrinking jobs”. Basing jobs on investment proposals is nothing but voodoo economics.
An investment proposal is a statement of intent and does not reflect investment on the ground. Proposals take 4-5 years to materialise and become tangible. It has no direct linkage with employment except for when actual investments flow in. Further, only about 35-40% of the investment proposals come into being. To say that investment proposals have come down and have suddenly impacted in a loss of jobs in just the past one year is an alarmist prediction. ‘Breaking news’ with such scant data to back their claims is a prime example of fearmongering.
If one looks at the recent CSO data in the first revised estimate of national income, consumption expenditure, saving and capital formation for 2017-18 (bit.ly/2EGRErm), we can see the actual investment in the economy. The yearly change in a nation’s Gross Fixed Capital Formation (GFCF), is a measure of investments in that country. At constant figures, one can see from the graphic that the GFCF has grown consistently from Rs 28.97 lakh crore in 2011-12 to Rs 41.36 lakh crore in 2017-18. There is no data to signify any reduction in 2018-19. So the idea that any ‘unemployment’ is due to reduction in investment between 2014-18 is bogus and not borne by data! The reduction between 2011-12 and 2013-14 is more due to a reduction in household investment in real estate than the private sector. This data also shows no reduction in private sector investment during this time except for a slight dip in 2016-17. There has been a discernible increase in government investment between 2014-18.
The last four years have seen a consistent uptick, with the change in 2017-18, in particular, being the most prominent. Even if you take investments at current prices, the GFCF at current prices has grown very well from 2014 to 2018, with a CAGR of 8.6%. The rate of GFCF to GDP, too, has increased from 30.8% to 31.4% in constant terms and from 28.2% to 28.6% in current terms in the past year. When investment to GDP is going up, there is no basis to say that employment can come down just because investment proposals are down.
Further, India has seen a constant accretion in its Gross Value Added (GVA) as can be seen in the graphic. For the years 2014-18, we witnessed a healthy CAGR of 7.5% in constant prices and 10.6% in current prices. The value addition in these years has been significant. Even if we credit a strong 5% of this growth in constant prices to increasing labour productivity per year, the remaining 2.5% growth in constant terms can be attributed to job growth. Hence, during 2014-18, this data shows an increase in total employment by 10%. On the base of a 502-million-strong labour force (World Bank), this 10% increase shows an increase in employment of 50 million (a lesser number if we use actual employed out of this). Analysts can assume any yearly productivity growth but the data will show strong employment growth contrary to the alarmist viewpoint.
When macroeconomic data like GVA, GFCF, social security (EPFO, PPF, NPS) data on transportation and from I-T on professionals is readily available, there must be strong questions raised against the need and sole use of thin surveys like the ones conducted by CMIE. The EPFO/ESI have over 10,00,000 employers registered, most of whom file monthly returns and pay contributions for around 190 industries across 29 states! This enormous quantum of data, every month, is more than any data from sample surveys and can be the base to make estimates. The CMIE reportedly has 500,000 respondents but this database has 10,00,000 employers and is more reliable. Today, the EPFO/ESI, after discounting duplication, shows 7.5 crore people getting a monthly salary. Governments, both Central and states and parastatals, employ another 2.5 crore, making a total of 10 crore getting a monthly salary with verifiable data. Increase in the monthly payroll, available through monthly reports of EPFO/ESI/NPS, show strong growth of around 70 lakh per year-verifiable data!
The Income Tax (I-T) data shows that 2.14 crore non-corporate entities filed returns in the fiscal year 2016-17 under the head ‘income from business/profession’. Assuming each employs 5 people on average, this could potentially show a stock of 10 crore employed. The MOSPI data on total stock of vehicles registered in India shows around 5 crore employed in this sector. All this shows verifiable data of 25 crore people employed. Even if we discount this further by 20%, we get a stock of 20 crore people employed from these databases. We can infer from the EPFO/ESI/NPS, vehicle sales and from data on professionals from the I-T database that, yearly, around 1/1.1 crore people are getting jobs, a far cry from the reduction in employment claimed by CMIE! Of course, this excludes the jobs in the agriculture sector!
India’s challenge is that around 43% of people depend on agriculture which is 15% of GDP, and shrinking, with their income growing at 2.9% a year, with low per capita income. They need to be moved to the industry and services sector where income is growing at 7.5-9%. The disparity is widening and only faster growth can bridge the income divide!
India has these rich verifiable data bases available with most giving monthly data. Sadly, MOSPI, too, depends on a thin sample of around 160,000 households for NSSO employment data surveys in a country of vast diversity with over 26 crore households, and the questions asked are based on the low growth economies of the 1980s and 1990s. When verifiable data is available, depending on such antiquated thin surveys does no justice to India. India needs a modern system of job growth estimation using verifiable databases. The government of India needs to take a policy decision to move from thin surveys to verifiable databases to enable regular issue of verifiable data. Our economists and commentators, too, should look at verifiable databases rather than pontificate on thin surveys that do not reflect the reality of a country which has grown at 8.7% a year in dollar-terms between 1991 and 2018!
(Pai is chairman of Aarin Capital Partners and Baid is head of research at 3one4 Capital)