Number theory: The difference in public and private estimates of employment in India

India does not have official high-frequency employment data, unlike countries such as the US, where statistics on employment, hours of work, overtime work, weekly and hourly income are available every month.

India does not have official high-frequency employment data, unlike countries such as the US, where statistics on employment, hours of work, overtime work, weekly and hourly income are available every month. Official employment data in India, which is published in the Periodic Labour Force Survey (PLFS) reports, are partial estimates of quarterly employment trends, which come with a lag of up to a year.

According to PLFS data, the all-India average labour income in 2018-19 was ₹11,225 per month.

Comprehensive employment estimates are available only on an annual basis, and come with a bigger lag. This is too big a delay for those who are interested in tracking the economy on a high-frequency basis. The need for high-frequency employment data has become even more important in the post-pandemic period.

It is this void in public statistics that has forced a number of analysts to rely on private sources of job numbers. The most popular among these are the employment estimates collected by the Centre for Monitoring Indian Economy (CMIE). CMIE’s employment estimates, which are based on its Consumer Pyramid Household Survey (CPHS) data, are available in various time frames, including on a weekly basis.

The reliance on CPHS to track India’s labour market has not gone unquestioned. Economists, including Jean Dreze, have argued that CPHS data excludes the poorest sections of the Indian economy and, therefore, might not convey the real picture. Is this criticism of CPHS data valid?

A working paper by Mrinalini Jha and Amit Basole, economists at the Azim Premji University, conducted a comprehensive analysis of PLFS and CPHS for 2018-19. The findings can be used to answer the question of compatibility between PLFS and CPHS numbers.

Here are three charts that summarise their findings.

CPHS overestimates the labour earnings for the pre-pandemic year

According to PLFS data, the all-India average labour income in 2018-19 was ₹11,225 per month. Using three different methods, the authors calculated the all-India income figures from CPHS to be ₹13,547, ₹14,531 and ₹17,354 per month. This is 19% to 55% higher than the monthly labour income from PLFS, conveying that CPHS likely overestimated labour earnings for the pre-pandemic year.

Monthly labour incomes in CPHS diverge from those in PLFS because of two reasons. One, CPHS fails to collect data from the less educated, poor and marginalized communities. Therefore, the average monthly figures are more likely to represent the labour earnings of the relatively better-off sections within the society.

Two, CPHS and PLFS collect the income and employment data differently. While PLFS collects the income data for individuals who report having jobs, CPHS collects income and employment data for each individual in two separate modules. These two modules do not have common points to relate with each other.

Considering this limitation, the authors merged these two modules to estimate labour incomes in CPHS. Therefore, it is likely that the monthly income figures diverged between CPHS and PLFS because of this difference in the data collection process.

Besides, these two surveys define employment differently. PLFS considers anyone to be employed if that person has earned income for at least one hour in the previous week. CPHS considers only those individuals with work on the day of the survey or a day before. This is one of the other reasons for the authors to use three different methods to calculate the labour income from CPHS.

The wage gap between men and women in higher in CMIE than PLFS

The wage gap between men and women is higher in CPHS than PLFS. PLFS data show that an average female worker earned 63% of the amount earned by a male worker in 2018-19. But, CPHS data show a much lower figure. An average female worker earned 47% of the amount earned by a male worker in 2018-19, as per CPHS.

This divergence was because women who do unpaid work in families – cooking, cleaning, taking care of the elderly and children -- are included in the data collection process in CPHS. While these women report being employed, they do not have any regular income to report during the survey. It is interesting to note that a fourth of the women workers in CPHS data have no incomes.

PLFS does not capture the proportion of women doing unpaid work in families. Therefore, if one excludes women doing unpaid work from CPHS, an average female worker was found to earn 55% of the amount earned by a male worker, which is much closer to the PLFS figure of 63% in 2018-19.

Both agree on urban-rural inequality, median and top incomes

Despite the differences, both CPHS and PLFS agree on three things. One, both showed similar extent of income inequality in rural and urban areas. The inequality in urban areas is higher than rural areas. In urban areas, the rich earned three times as much as the middle class in 2018-19. In rural areas, the rich earned two times .

Two, if the monthly income of an individual is ₹50,000 in 2018-19, then she belongs to the richest income class, that is top 5%, within India, state the authors.

Three, the median income of a worker was ₹10,000 a month or less in 2018-19. In January 2019, the expert committee on minimum wages, led by Anoop Satpathy, recommended a national minimum wage of ₹375 per day, or ₹9,375 per month for rural India, and ₹430 per day, or ₹10,750 per month for urban India. This means that the bottom 50% of the workforce earned less than the recommended minimum wage in the country.