The problem with India’s jobs data is that they hide more than they reveal

Measures of employment are released sporadically and exclude a number of sectors, notably the informal economy that employs over 80% of Indians. While the Narendra Modi government has put in place a task force to address these gaps, there is still no clarity over how the new data will be computed or how often they will be published.

However, there are bigger concerns, too, according to a new study on India’s labour statistics by the McKinsey Global Institute, a unit of consulting firm McKinsey.

For instance, while the current data assessment by the Labour Bureau and the National Sample Survey Office does indicate broad trends, it doesn’t provide any insight into the quality of the country’s workforce. McKinsey notes that these labour statistics leave out information on independent work, as well as flexible or part-time jobs, even as they constitute a key part of today’s employment scenario.

Additionally, some quarterly numbers announced by the Labour Bureau are based on extremely small sample sizes. For example, the “Quarterly Report on Changes in Employment in Selected Sectors” is based on a sample of just 1,936 enterprises. As a result, McKinsey says, “Conclusions about aggregate national trends derived from them may not be accurate.”

Here are the six key limitations that McKinsey explains in its study:

Labour force participation

India’s current surveys measure the country’s labour force participation rate – people who are working and those who are willing to work and looking for jobs. But according to McKinsey, the participation rate by itself does not tell the full story. The rate is typically affected by a number of factors, including age, education, income, cultural attitudes, and job opportunities. So, a declining rate could just mean that young people are staying out of the workforce to study, for instance, and not necessarily a sign of deteriorating labour market dynamics.

“Similarly, declining female participation may actually be a sign of higher income security in some cases: the labour force participation rate of women is highest in low-income households that combat extreme poverty, and the first sign of entering the middle class is often for the woman of the household to withdraw from poor-quality work,” McKinsey said in the study.

Unemployment

While government surveys have shown that India’s unemployment rate has held steady at some 4% for the last four years, McKinsey notes that the assessment does not take into account the quality of work Indians do. For this, gainful employment is a better indicator, the study says, measuring “the quantity and type of work done by people already in employment, growth in labour productivity, higher earnings, and aspects of work quality such as safety, cleanliness, flexibility, income security, and intellectual challenge.”

“Rarely would a poor rural boy who had dropped out of school remain ‘unemployed’—he would typically be put to ‘work’ on his family’s small piece of land or would lend a hand at the local kirana shop owned by his uncle,” McKinsey explained. But though both these activities qualify as jobs, and would reflect in labour market surveys, the worker himself may not be gainfully employed.

Disguised underemployment

McKinsey also notes that India’s employment data only indicate whether a person has worked or not, without showing how much actual work was carried out. That means it’s difficult to gauge whether people are underemployed, i.e. working less than the full potential of their skills and abilities.

“A person who engages in 10 months of work is on par with one who has engaged in seven months of work during the 12-month reference period,” the study explained.

Social mobility

Labour assessments don’t ask the right questions about the social lives of workers.

“(The current) trends in the labour force participation rate and unemployment do not reflect social or economic mobility,” McKinsey said in the study. Surveys, it said, should ask individuals about their standard of living, hours of work, and overall satisfaction, so that the responses can be compared every year to track changes.

Sectoral employment

The study points out that new jobs in specific sectors, notably the most watched industries of manufacturing and services, are categorised based on the employer or company, not the actual job itself. As a result, if a manufacturer hires for a designing job, it would be captured as an addition to the manufacturing sector, instead of what it really is, which is an increase in service jobs. That could provide the government and policy-makers with misleading trends, McKinsey said.

Difficult interpretation

Finally, it’s the timing of the surveys that adds to the problems, making it impossible to formulate accurate, up-to-date, and comparable conclusions about the state of India’s employment.

“The annual labour market surveys involve a six-month data collection phase, in which respondents are required to describe their labour market status in the preceding year,” McKinsey notes. So, the data that are published in the 2015-16 annual employment survey – the latest data available – actually reflects what was happening in 2014-’15.

This article first appeared on Quartz.