Making sense of employment statistics
On 1 July, the National Sample Survey Office (NSSO) started another round of employment-unemployment survey (EUS), part of its 75th round. This survey is a large one and is part of the quinquennial surveys on employment and unemployment. Usually, the quinquennial surveys are held after every five years but this one has been delayed by a year. Nonetheless, the results of this survey are eagerly awaited by all those who are interested in finding out what happened to employment between 2011-12 and 2017-18, particularly under this government. Unfortunately, the results are unlikely to be available until the end of next year or early 2019, just before the general elections.
The anticipation is also heightened because of the confusion that has been created about the robustness of employment data in the country. The NITI Aayog has set up a committee to suggest measures to improve the quality of employment statistics in the country. The report of the committee has not yet been submitted but media reports suggest that the committee is likely to recommend ways of improving the quality of statistics on informal employment and also ways of getting better data at shorter frequency.
Unfortunately, the suggestions of the committee will have no impact on the ongoing survey. Since a lot of work on improving the quality of employment statistics, including increasing frequency and coverage has been going on for almost a decade, it would have made sense to mainstream those efforts rather than start afresh.
While NSSO surveys are definitely the benchmark and the most trusted estimates, they have been under criticism because these surveys are not held every year and there is a time lag for release of estimates. These concerns have been voiced for quite some time now, and some of them were taken care of by NSSO when it restarted the annual EUS surveys after 2004-05. Between 2004 and 2011-12, six EUS surveys were conducted by NSSO. Given the need for high frequency data on employment, the Amitabh Kundu committee report on Periodic Labour Force Surveys (PLFS) was accepted in 2009 and pilots were conducted between 2011 and 2013. These surveys would give quarterly estimates of employment for the economy based on a rolling panel survey. The delay in rolling out PLFS and EUS should have attracted more attention from the NITI Aayog than revisiting the entire issue again.
While there is no information available from EUS of NSSO (which was delayed by this government), there have been estimates from comparable and credible sources on the extent of job losses in the economy. These have largely come from the labour bureau, which is an organ of the ministry of labour. The bureau has been conducting various surveys to collect information on trends in employment. It started a quarterly employment survey (QES) immediately after the financial crisis to capture the extent of job losses in specified sectors.
Based on these surveys, a total of 36.3 lakh jobs were created in selected sectors between September 2008 and September 2014. That is, 6.04 lakh jobs per year. On the other hand, total employment created between December 2014 and December 2015 has only been 1.35 lakh. Clearly, the rate of employment creation has declined by one-fifth during the present government’s tenure on a comparable basis.
The new series has expanded the coverage of the sectors by adding construction, which is an important sector for job creation. While these surveys cover only 20 million workers, the extent of employment covered by them is a good indicator of the changes in employment in the country since these cover the dynamic sectors including manufacturing and construction which are supposed to respond to economic growth. Given that agriculture has been losing workforce at more than 5 million workers every year, these numbers do not give any confidence that the economy is creating anywhere near the 20 million jobs that the government had promised.
The worsening of the employment situation is also confirmed by the annual surveys of the labour bureau started in 2009, the fifth round of which was in 2015-16. These surveys again are held nationwide and similar in concept to NSSO’s EUS. The last two surveys of 2014-15 and 2015-16 are improved surveys compared to the first three, with better geographic coverage and also larger sample size. The sample size of the last two surveys is more than 150,000 households compared to 100,000 households by the last two by NSSO.
Based on these surveys, the total workforce in the economy declined by 16 million between 1 March 2014 and 1 July 2015. While the survey data is clearly showing the worsening of the employment situation, data from industry such as the IT sector has corroborated the extent of decline in employment with more than 100,000 workers expected to lose jobs this year alone. But for those not satisfied by the survey data, the protests by farmers and the dominant agrarian communities of Jats, Patels and Marathas is evidence that the rural economy is not creating enough non-farm opportunities. Clearly, the numbers are showing what most economists have privately admitted—that the economy has now reached a job-loss growth phase.
The attempt by NITI Aayog to argue for weaknesses in the statistical system is nothing but obfuscation and an attempt to divert attention from the core issue of why it is not creating jobs. While the setting up of a committee to revisit the issue of employment estimates is understandable, what the NITI Aayog should have done is to push for starting the PLFS and for restarting the annual surveys. But it is not to argue that there is no need to revisit the issue of employment estimates available in the country. There are genuine problems with our employment estimates and much of it is not because the surveys are wrong but a reflection of the complexity of the labour market of a country which remains overwhelmingly informal even today. Despite that, there is no denying the fact that the economy has failed to create enough jobs.
What is needed is a better understanding of why the economy is not creating jobs. Now after three years, when the employment situation has become unmanageable, criticising employment statistics is nothing but an attempt to shoot the messenger. Given that the informal sector has borne the brunt of demonetisation and now the goods and services tax, a better understanding of the impact on employment potential of the informal sector would have helped policymakers. This would require conducting special surveys apart from the regular surveys. But in a country where counting of old demonetised notes takes forever, blaming the statistical system is a convenient excuse to sidetrack the core issue of job losses in the economy.
Himanshu is an associate professor at Jawaharlal Nehru University and visiting fellow at Centre de Sciences Humaines, New Delhi.