A report in this paper (IE, November 1) on various employment estimates for 2017-18, based on the PLFS (periodic labour force survey) data, has received considerable attention. We have received several queries from policy makers regarding the differences in our employment estimates with studies by Laveesh Bhandari and Amaresh Dubey (BD), and Himanshu that are quoted in the report. In this article, we clarify the reasons for the differences and highlight the major findings of our study.
First, our estimates are different from that of BD because theirs are based only on the usual principal status (UPS) of employment. Our study is based on both usual principal and subsidiary status (UPSS) of employment. Since Himanshu’s estimates are also based on UPSS, our figures are not much different except for the 2017-18 PLFS figures. Himanshu’s and our estimates for employment in both 2004-05 and 2011-12 are almost the same (the negligible difference is because of the use of population multipliers that differ at second and third decimal points). The discrepancy with Himanshu’s 2017-18 employment figure is only because he probably used a lower population multiplier based on a projected total population of 1.33 billion.
But, while comparing Himanshu’s estimates with that of BD for 2017-18, we find that both their figures are the same (this is surprising since BD have ignored the subsidiary status of employment unlike Himanshu). We have explored why this has happened. BD have used a projected population of close to 1.36 billion based on their argument that the annual population growth of India was about 1.7 per cent between 2011-12 and 2017-18. This is a very high overestimation of India’s population growth rate during this period. The World Bank’s population projection (available at World Development Indicators) shows that annual population growth in India was only 1.2 per cent per annum in 2012 and 1.04 per cent in 2018, and the total population was about 1.34 billion in 2017-18. Our population projections (based on a monthly exponential growth rate) show that the total population in India was about 1.345 billion in 2017-18 (the same as the World Bank’s estimate).
The National Family Health Survey 2015-16 shows that there has been a phenomenal decline in the total fertility rate (the number of children born to women in the reproductive age group of 15-49), reaching a near replacement rate of 2.2 per woman, from 2.7 in 2005-06. Hence, the 1.06 per cent population growth rate we use is logical; and 1.7 per cent cannot be India’s current population growth rate. The use of the latter artificially raises the workforce number that BD came up with, 457 million in 2017-18. This is surprising since they had missed out on subsidiary workers altogether. Hence, they fail to find that the total jobs in India fell by nine million in six years, a first in India’s history.
Moreover, since the employment and unemployment figures are always estimated considering both the principal and subsidiary status of employment, BD’s employment estimate, based only on the principal status, is incomplete and misleading. The principal employment status is recorded for a relatively longer duration of time in the survey year (normally for a period of six months and more). It leaves out those who work for less than six months, and who must also be counted among those working. The subsidiary status of employment data (covering those working 30 days but less than 182 days), covers those not employed in the principal status, and must be added to workers in principal status. So, by ignoring the subsidiary employment, BD provide an incomplete and incorrect employment trend for India.
Hence, BD’s claim that manufacturing employment during 2011-12 and 2017-18 was “stagnant” is also misleading. In fact, based on the UPSS, we found that manufacturing employment actually declined by three million between 2011-2012 and 2017-2018. While labour intensive sub-sectors were mainly responsible for this decline, a few capital intensive manufacturing sub-sectors did register consistent growth of jobs during 2004-05 and 2017-2018. Although the quality of jobs is very poor in the services sector, it has played a major role in driving job growth, but only partly offsetting the fall in employment in agriculture, manufacturing, and slowing employment creation in the construction sector. In fact, because of their problematic methodology, BD fail to notice the actual fall in total employment (including self-employment of the Ola/Uber or Mudra variety, which is captured in a comprehensive sample of the unorganised and organised sector workers).
The share of regular and formal employment increased marginally due to growth of formal jobs in the private sector, but the share of informal jobs (without social security) within the government/public sector has increased. A dominant and rising share of jobs is still generated by micro and small units in the unorganised sectors without any formal or written job contract. In both government and private sectors, the number of contract jobs (with no written contract, or less than a year’s contract, or one to three-years contract) is on the rise post 2011-12. Although BD’s paper also reveals the same story, we differ from them on the absolute number of jobs. The rate of growth of jobs in the unorganised component is much higher than in the organised sector (those with more than 10 workers).
Moreover, BD’s research also provides a partial picture of the unemployment trends in India as they could not capture the upsurge in open unemployment among educated youth (which has more than doubled in a matter of six years), nor the recent rise in the disheartened youth labour force (those in the age group of 15-29 and not in educational institutions or the labour force). We found that the number of disheartened youth (having secondary and above level of general education, and with technical and vocational qualifications) is rising at an alarming rate.
With total jobs falling at a time when new entrants into the labour force are increasing at an accelerating pace, it is not surprising that real wages have not increased in both rural and urban areas. From now on, entrants into the labour force will increase at an accelerating pace, at least until 2030. Hence, a comprehensive employment policy combined with an industrial policy (to boost growth of medium and large firms) is necessary to address the transformation of agrarian workforce, boost real wages, and to ensure industrial development.
This article first appeared in the print edition on November 7, 2019 under the title ‘The jobs challenge’. Mehrotra is professor of economics and chairperson, Centre for Labour, JNU. Parida is assistant professor of economics, Central University of Punjab.