The Periodic Labour Force Survey (PLFS) is officially out, and with it, the controversy on job growth over the last 5 years has resurfaced. At the onset, I must state that the unit level data of the survey should have been released long ago. As far as the report is concerned, we’ve enough evidence to now know for sure that it suffers from serious inconsistencies.
For starters, let it be known that PLFS and the previous Employment-Unemployment Surveys have different sampling techniques. While the EUS uses income levels for stratification, PLFS used education attainment for stratification. Therefore, any comparison of unemployment rates for these two techniques is fundamentally flawed so any article that claims unemployment is at a 45-year high is junk and you’re better off ignoring it rather than wasting your time on it.
With the PLFS report finally out, commentators can now explore the inconsistencies within the report in far greater detail. It will also be possible now to use the unit level data and adjust for the change in the sampling framework to arrive at near comparable estimates of unemployment. One such inconsistency has been in the rate of urbanization as reported by PLFS which is lower than the one provided by the NSSO in 2011-12. This suggests that post-2011 there was a decline in urbanization, but rapid urbanization is visible across the country. Another issue with the PLFS is its underestimation of the population which clearly is yet another inconsistent finding. Therefore, there are multiple problems with the PLFS, and it simply does not pass the basic smell test!
The question that is often being asked is why the government released the report if it is incorrect- but then when the report was not being released everyone was criticising the government for withholding it. Fact remains that the report is definitely problematic irrespective of the fact that it has been now officially released. This release should be welcomed as now we have the entire report with us to identify what went wrong with the sampling framework and the survey design so that we can rectify it for the future. But would I base my understanding of India’s labour market solely on the PLFS? Absolutely not!
As far as the issue of job creation in the economy is concerned, the debate has been far too politicized in the recent past and it continues to be politicized even after the elections. Such politicization will not help address the issue at hand and it would only direct our attention to the inconsequential and never-ending debate on whether there is robust job creation in the economy or not. The answer to this question is likely to be highly subjective: if you were looking for a job and got one then you’d believe in robust job creation, if not then you’d believe that there are no jobs in the economy.
But a modern economy like India does require a time series statistic that can estimate unemployment (and hence job creation) on a regular basis. This was the initial motivation for the Periodic Labour Force Survey and perhaps, there were some errors that were made in the process but now would be the time to correct for them and reconstitute a similar survey with an improved sampling framework along with a questionnaire that aims at capturing the self-employed and flexible employment contract workers in the process of estimation.
Additional questions on whether an income was earned and if yes then from what source could be instrumental in capturing the source of livelihood for a respondent. It would be difficult to quantify job creation during the last 5 years in the absence of such data, but there is macroeconomic evidence to suggest that job creation has definitely improved between 2014-18 compared to 2009-13.
Karan Bhasin is a political economist by training and has diversified research interests in the field of economics. He tweets @karanbhasin95.