India Has No Robust System Of Measuring Job Creation And It’s Time It Got One

Be it the Gross Domestic Product (GDP) figures, the Index of Industrial Production (IIP), the Consumer Price Index (CPI) or even the employment numbers – producing reliable data on important economic indicators has always been a challenge in India. The debates surrounding the numbers on aforementioned indicators have particularly grown louder under the current regime. The government has responded to the growing criticisms by coming up with new indices improving upon the old ones as we saw with the new GDP, IIP, CPI, and WPI series. The focus has now shifted to employment-unemployment figures which have put the current government in poor light indicating that the economy has probably entered an era of jobless growth.

Those who disagree with such an assessment opine that the job indicators used to reach such conclusions are faulty and outdated. The government seems to be in agreement with this line of thinking. On 11 May, it appointed a Task Force for improving employment data under the chairmanship of Vice Chairman, NITI Aayog. The Task Force has put its report out for public comments. The report discusses in great detail the existing sources of data on employment in India, highlighting the current state of affairs and gaps in the data, looks at the practices in the United States and United Kingdom and provides recommendations for generating employment and unemployment data going forward.

It lists four potential sources of measuring employment and unemployment: household surveys, enterprise surveys, administrative data, and data from government schemes.

Household Surveys

Currently, two official household surveys – the Employment-Unemployment Survey (EUS) by the NSSO and the Annual Labour Force Survey by Ministry of Labour and Employment – and a population census collect employment statistics from households. The EUS is the most comprehensive of all but has serious drawbacks. It is conducted every five years and the results become available after a lag of more than a year. The limitation of Annual Labour Force Survey is that it is less comprehensive in extent than the NSSO survey and the data collection takes place during a part of the year, instead of being spread out over the entire year. Population census survey is held every 10 years and the final data is available after a lag of several years, which makes the whole exercise pointless from policy making point of view.

Based on these lacunas and after analysing the best practices in the US and the UK, the Task Force recommends three changes to improve household surveys: conduct the surveys on an annual basis, introduce a time-use survey and progressively introduce the use of technology that can speed up data collection and reduce the time lags between data collection and processing.

The EUS survey can be replaced by an Annual Household Survey which will provide annual estimates of labour force, employment, unemployment, industry structure of workforce, nature of employment and wages nationally and regionally on an annual basis. It will also have a quarterly module in urban areas, where households will be visited four times constituting a rolling panel of three quarters, thus making it easier for tracking the seasonal employment.

Apart from this, the report recommends instituting a new time-use survey to be conducted every three years. This will basically collect information on how people are allocating their time over a specified time period. It will be useful for:

a) determining the extent to which people engage in multiple occupations and the share of their time spent in performing productive but non-market activities;

b) determining time spent on leisure activities; and

c) assessing the reasons for shifts in labour participation rates and the effects of policy changes on the pattern of activities.

Enterprise Or Establishment Surveys

Unlike household surveys, there is no robust mechanism for collecting enterprise surveys because there is no one reliable sample frame for tracking all enterprises and one which is also updated frequently. Factories Act, for instance, leaves out all service sector establishments and even industrial establishments with less than 10 workers if using power and less than 20 workers if not using power. The Annual Survey Of Industries (ASI) is the only regular and frequent establishment survey, but the big drawback is that it uses the enterprises registered under the Factories Act.

Ideally, the Economic Census which covers the entire universe of non-agricultural establishments regardless of size or sector, should’ve served as the sample frame for conducting enterprise surveys but since 1977 when the first such survey was conducted, only five subsequent surveys have been conducted so far and those too at irregular intervals. Additionally, it doesn’t cover self-employed workers which form the majority of workforce in the country.

Other establishment surveys such as Unorganised Sector Surveys of Industries and Services by NSSO, Quarterly Employment Survey (QES) by Labour Bureau, Micro, Small and Medium Enterprises (MSME) Census etc also suffer from drawbacks like limited coverage, infrequent data collection, long time lags, outdated sample frames.

To rectify these lacunas, the Task Force has recommended an Annual Enterprise Survey using GSTN as the sample frame conducted with higher frequency, preferably every quarter. Since GSTN database has only those enterprises with turnover of more than Rs 20 lakh, the report recommends a separate annual survey of remaining small establishments.

Additionally, the Task Force has also recommended using GSTN across all legislation, ministries and departments as the universal establishment number to greatly simplify the process of data collection.

Administrative Datasets

India doesn’t use the data from administrative sources such as the Employees’ Provident Fund Organisation (EPFO), the Employees’ State Insurance Corporation (ESIC), the National Pension Scheme (NPS) etc for gathering information about labour market. This is probably because new entries into these datasets do not necessarily represent new jobs. However, information from these sources can be used to gauge the extent of formalisation of the workforce.

But nailing down the extent of formalisation has proved difficult. Many commentators have disputed the reports which put the proportion of informal workforce above 90 per cent. There is no official definition of ‘formal workers’ as of yet and the term ‘organised workers’ which is currently in use leaves out all workers who may have decent and steady jobs but either do not work in large enough enterprises or do not have written contracts. To remedy this, the report recommends that workers covered under The Employees’ State Insurance Act, 1948 (or other similar insurance), Employees’ Provident Funds and Miscellaneous Provision Act, 1952 (or other similar social security scheme) be counted as formal workers. Government and other public sector employees, workers having coverage under private insurance or pension schemes or provident funds and workers subject to tax deduction at source on their income through submission of Form 16 or similar income tax form should also be counted as a part of formal workforce.

Employment Generation Through Government Schemes

The Task Force calls for evaluating the employment impact of government schemes such as MUDRA loans, Housing For All, National Highway Construction, PMGSY, MGNREGA,

PMKVY, DDUGKY etc as these may be having significant impact on job creation especially in the current climate where private investment has not taken off and public expenditure is mainly driving the economic growth.

Apart from these, the Task Force has recommended creating a central facility from which all administrative data can be accessed. The report cites the examples of the Reserve Bank of India, the states of Andhra Pradesh and Gujarat, which are already working along these lines. Additionally, it calls for eliminating duplication in data gathering by various ministries and clearly mentioning the limitations of the estimates in their reports to avoid the confusion.

Finally, it emphasises on the importance of sound data collection and evidence-based policy making. It recommends bringing about serious institutional changes and providing necessary financial resources to revamp India’s statistical systems.

If implemented, these recommendations have the potential to transform the quality of employment-unemployment data collection in the country and in turn the output of government policies.