The Dark Clouds Of India’s Jobs Data, With One Silver Lining
Workers carry an iron pipe on their shoulders at a steel and iron market area in India. (Photographer: Prashanth Vishwanathan/Bloomberg)

The Dark Clouds Of India’s Jobs Data, With One Silver Lining

BloombergQuintOpinion

The Ministry of Statistics and Program Implementation recently released the results of the Periodic Labour Force Survey conducted by the National Sample Survey Organization for 2017-18 (July-June) and the remaining quarters until December 2018. The PLFS provides a variety of statistical indicators of India’s labour market on a quarterly basis for urban India and, on an annual basis, for rural India as well as rural and urban India combined.

An important feature of PLFS is the use of a rotating panel for urban areas, with each household being visited four consecutive quarters and no more. The real substantive change is that education in the PLFS, as opposed to monthly consumption expenditure in the old employment-unemployment surveys (EUS), is being used as the criterion for stratifying households within each urban block or rural village chosen for the survey. Because of this difference, the MoSPI is providing a cautionary note on comparing PLFS statistical indicators with the earlier EUS ones.

As explained in the PLFS documents, using education as the household stratification criterion is motivated by the importance of tracking labour-market dynamics within the various education categories in an “aspirational” society like India’s, with a rapidly expanding educated population. Besides, it should be obvious that education is more clearly identifiable than income, especially for urban informal and rural sector workers.

However, if the sampling is done right, the statistical indicators computed at the aggregate rural, urban or overall (rural plus urban) levels should not be different under the two approaches: namely, stratifying by education versus by consumption expenditure.
Workers eat at a wholesale grain market in Rewari, Haryana, on May 8, 2019. (Photographer: T. Narayan/Bloomberg)
Workers eat at a wholesale grain market in Rewari, Haryana, on May 8, 2019. (Photographer: T. Narayan/Bloomberg)

Data Consistency

Let me explain what I mean by sampling being done right. Firstly, the PLFS education categories or, alternatively, EUS expenditure categories have to be fully exhaustive (cover all possibilities). Secondly, the households chosen under each category have to be representative of the population within that category (within a particular block or village), with the sampling weights correctly telling us accurately what proportion of the population each household that was chosen for inclusion in the sample represents. Then the only source of difference between statistical indicators obtained under the two methodologies for a given year or quarter is a statistical discrepancy, which is natural to arise in a country India’s size and complexity. A simple, illustrative example is where one can divide students in a school into household income categories or, alternatively, parents’ education categories, and calculate their group average scores on a mathematics test. Whether one then calculates the overall school average from household income category averages or, alternatively, from parents’ education category averages, one should obtain the same overall mean, provided the person calculating accurately knows the proportion of students in each category. These proportions are analogous to sampling weights for PLFS or EUS.

If we look at the overall unemployment rate for India under the usual status, which is the proportion of the labour force (those working plus those looking for jobs) that is not employed based on the activity status of individuals over the past 365 days, the PLFS figure for the year 2017-18 was 6.1 percent, as compared to the EUS figure of 2.7 percent in 2011-12. The former is 126 percent larger than the latter. A gap so large between indicators generated by methods that are both trying to determine the actual proportion of the country’s labour force that is unemployed, in this case by the same definition of usual status, can only be very partially attributed to a statistical discrepancy.

It seems obvious then that there has been a significant increase in the unemployment rate.

For the purists who want to make comparisons only when the numbers are generated using the same methodology, the news again is not that good, since we can compare the 2017-18 PLFS current weekly status unemployment number with the same for the final three quarters of the calendar year 2018. While the former is 8.9 percent, the latter averages 9.7 percent. Clearly, unemployment has been increasing over the last year or so. These observations are consistent with the recent fall in GDP growth to 5.8 percent in the last quarter, which, according to some analysts, is driven by sluggish investment and a weak agricultural sector. Also, consistent with this, the unemployment rate calculated by the Center for Monitoring the Indian Economy has gone up from 5.74 percent to 7.17 percent over the last year.

Fall In Participation Rate

Moving to the labour force participation rate (LFPR), it was very low, at 37 percent under the usual status for 2017-18. While it is 55.5 percent for males, it is a dismal 17.5 percent for females. Actually, the overall male LFPR has been around 55 percent for many years in both rural and urban India, starting in 1993. The female LFPR in rural areas has constantly been declining from 33 percent in 1993 to 25 percent in 2011 and in 2018 to 18 percent. The urban female LFPR has remained fairly stable in the 15-18 percent range for the last 25 years. Researchers have found that a combination of better access to education for girls and rising household incomes, that keep women focused on household work, steers them away from low-paying jobs.

A lack of appropriate employment opportunities, and measurement problems, relating to working women often reporting their primary activity as household work, can explain low and falling female LFPR.

Also read: In Charts: Who Is Hurting Most From India’s High Unemployment?

Quality Of Jobs

The best news coming from the PLFS is the rise in the share of regular wage/salaried employees and the fall in the share of casual labour in overall employment. The share of regular workers in rural male employment grew from 10 percent in 2011-12 to 14 percent in 2017-18, while for rural females the analogous share increased from 5.6 percent to 10.5 percent. The rural casual labour share fell from 35.5 percent to 28.2 percent for males and 35.1 percent to 31.8 percent for females during the same period. The urban regular employment share went up from 43.4 percent to 45.7 percent for males and 42.8 percent to 52.1 percent for females.

Clearly, the quality of jobs has improved, and that aspect of employment is not being captured by changes in the unemployment rate and LFPR.

As average incomes increase and access to education improves, it is natural to see a smaller proportion of the population in the lowest age groups to participate in the labour force. We also expect people to be able to afford to wait longer to find jobs of their desired type and quality.

Thus, creating quality jobs, which keep pace with the growth in the working-age population, is a huge challenge for policymakers in India. Some recent labour reforms, mainly at the state level, along with rising incomes and better access to education, are likely to have improved average job quality. But the high and rising unemployment and the low and falling LFPR are still a cause for concern. The current government, at the start of its second term, needs to get serious about formulating policies aimed at expanding labour-intensive manufacturing as well as dealing with the problem of stagnant exports. Further labour reforms, careful thinking about land acquisition, and a rationalisation of the Goods and Services Tax are needed.

Devashish Mitra is Professor of Economics and Cramer Professor of Global Affairs, at The Maxwell School of Citizenship and Public Affairs, Syracuse University.

The views expressed here are those of the author and do not necessarily represent the views of BloombergQuint or its editorial team.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.