The jobs question

Employment survey needs to be better, faster

The Indian Labour Bureau recently released its quarterly newsletter which showed a reduction in employment in the third quarter of 2015-16. Of the eight sectors covered, all but one saw a reduction in employment. In other words, the Bureau’s sample survey showed that employment went down in information technology (IT) and IT-enabled services, transport, jewellery, leather, metals, automobiles and the handloom and powerloom sectors. Only textiles and apparel showed positive growth during the quarter. This is a worrisome trend. However, the data set is from a quick survey undertaken of fewer than 2,000 establishments across the country. It does not cover many important industries, and more importantly excludes the informal sector. It is also dated, being released months after the data are collected, and conditions may have changed in the intervening period. Indeed, this exercise was intended, right from its very inception in the late 2000s, to be a quick survey with results that are essentially indicative. Even so it has been deteriorating over time. Since 2009, the coverage of establishments has fallen from 3,192 to 1,936 in the third quarter of the year 2015-16.

The larger issue is the obvious need for employment data. India needs an employment survey that includes the whole economy, including the unorganised sector, released within weeks after collection. This is not an easy task, but it is not impossible. Employment is by far the most important metric for policy, and arguably political imperatives are all driven by employment considerations. Whether it is monetary or fiscal policy or welfare measures, timely and quality data on employment are critical; their absence forces decision-making towards the domains of speculation and conjecture.

There is enough institutional and intellectual ability in the country for such data to be collected and made available. Further, data collection is not that expensive an exercise that India cannot afford it. It is fortunate, therefore, that the survey is reportedly being expanded to cover 18 sectors instead of just eight, and 10,000 establishments instead of 2,000. The release lag is also supposed to be reduced. If this new and improved survey is still not good enough, further adjustments must be made. Investing in employment data will enable more appropriate policy, fiscal decisions, and monetary policy. Several reports have been written by many committees arguing as such, and any progress would be welcome.