Hatchet job on the work front

Job creation has been a matter of intense debate and concern,  and rightly so. Some commentators and news articles have said that the quantum of new jobs created (employment generation) has dropped consistently. This is based on the Quarterly Report on Employment Scenario (QES),  released by the Labour Bureau, Ministry of Labour and Employment. QES data suggests that the country added about 1.1 lakh new jobs over the period of July-October 2016. In contrast, about 3.15 lakh new jobs were added in July-September 2011 and about 1.43 lakh in July-September 2013. Based on this, critics conclude India confronts a jobless growth scenario, apart from raising questions about growth numbers and the composition of growth.  Before deciding that, one must remember that there are serious limitations about employment data in general and the QES in particular. 

Broadly, there are three sources for assessing job-related data. The first is the Labour Bureau’s QES. This is based on surveys of industries/establishments. Two other sources include the employment-unemployment surveys undertaken by the National Sample Survey Office (NSSO) and the  Labour Bureau, respectively. These two surveys are household-based surveys where responses from large samples of individuals are collected for analysis.

Each of these sources has a different release frequency and study methodology. In addition, establishment-based surveys will always give an imperfect picture in a situation where a substantial amount of employment isn’t through employer-employee relationships and where there are large chunks of self-employment.  Which establishments are being covered?  Those that are registered in some fashion, and most enterprises aren’t.

QES is a quarterly release document, which indicates the employment situation within select sectors. It is prepared by surveying non-farm establishments which employ 10 or more workers. Further, these eight sectors include, manufacturing, construction, trade, transport, education, health, accommodation and restaurants and IT/ BPO. Note the following: Starting January 2009, 28 survey results were released till December 2015. The sample size of these surveys was around 2,000 establishments spread across 8 industrial sectors within 11 states. The limitation in sample size impacted the quality of data. 

To correct this, an expert group under the chairmanship of Prof SP Mukherjee was constituted. Based on the expert recommendations, the methodology was completely revamped. A new series was instituted, which increased the sample size to around 10,600 establishments spread across the country. The first survey based on the new series was released in September 2016, with estimated data on jobs as on April 1, 2016. 

Till date, three releases (quarterly basis) based on the new series have been done. The second release provided information on the employment situation as on July 1, 2016 (vis-a-vis the situation on April 1, 2016), the last release being March 2017 covering the same as on October 1, 2016 (vis-à-vis that on July 1, 2016).  

Given this, before deducing anything, bear in mind the following: a) Individuals/households are not surveyed. Only establishments, and that too with more than 10 employees. In that sense, the entire population of smaller establishments, informal economy players, individuals, etc, are not represented at all in QES; b) Major changes in study methodology make it difficult to compare and contrast the results of the new series with that of the old series. Hence, comparing July-September 2013 employment data with that of July-October 2016 is an apples-to-oranges comparison; and c) The most important aspect is that QES completely leaves out jobs in services and the agriculture sector. It only considers industry. 

We now turn to “Employment and Unemployment Situation” surveys undertaken by NSSO and Labour Bureau. NSSO usually undertakes such surveys every five years. It is done through primary interactions with a large sample of households spread across the entire country. The last such round was undertaken five-six years ago in July 2011-June 2012 and hence can’t be referred to for the current debate on job creation. Furthermore, since the last round, NSSO has not released any new survey results. The Labour Bureau also undertakes an “Employment-Unemployment Survey”. This survey examines broader key labour force trends such as unemployment rates, labour force participation rates, etc. But unlike QES, this is household-based and is undertaken annually. Till date, the Labour bureau has released five surveys. The last one, released in September 2016, covered the period 2015-16. That said, the Labour Bureau survey is yet to evolve institutionally and hence establishing its accuracy and credibility may take some time. 

So while the NSSO survey is ideal in terms of its methodology, reliability, coverage, rigour etc, we effectively face a situation where we do not have an updated, credible and holistic estimate of job growth. Therefore larger inferences on employment creation or the lack of it may not be yet warranted. 

The authors are economists with NITI Aayog. Views are personal.