The bitter truth of female labour participation
The labour participation rate (LPR) is the proportion of working-age people (which is people of 15 years or more) who are willing to work and are either actually working or are actively looking for work. In my opinion, the LPR is more important than the unemployment rate because it tells us how many people are willing to work.
If a very small proportion of people are willing to work then a low unemployment rate does not mean much. A high LPR directly contributes to growth and greater well-being.
If only adult men work, but adult women don’t, we are much worse off compared to a situation where both adult men and women work. A family where men and women work equally (and share household work also equally) is, in my opinion, an ideal modern family today.
In India, women’s participation in the labour force has been low.
This relatively lower female participation rate is not unique to India. It is a global phenomenon. Except Burundi and Mozambique, there is no country in the world where female labour participation is higher than male participation. Low female participation rates is a lot more pronounced in South Asian countries where it is 28 per cent compared to global rate of 67.5 per cent.
But CMIE’s Consumer Pyramids Household Survey (CPHS) for the period May-August 2018 shows the female labour participation rate to be a shocking 10.7 per cent! It is shocking to a point of being unbelievable. Although the numbers are not comparable at all, it looks like Indian female labour participation rate is better only than Yemen’s 6 per cent.
Last month, Rajiv Kumar, vice-chairman, NITI Aayog, publicly ticked off the entire survey because of this low female labour participation rate. Surjit Bhalla, member, Economic Advisory Council to the Prime Minister, has been repeatedly critical of this data similarly. Neither have studied the data sufficiently or correctly to justify such remarks. But, the onus of explaining this data and dispelling doubts rests on CMIE and I take up that task here.
We compare CMIE’s CPHS estimates with those from the official surveys of NSSO. The appropriate indicator to compare CMIE’s CPHS estimates of labour participation and unemployment with the NSSO’s estimates is the NSSO’s Current Daily Status estimates. Surjit Bhalla and his co-author Tirthtanmoy Das have completely misread CMIE’s surveys by wrongly interpreting them as comparable to NSSO’s Principal Status estimates. Their paper “All you wanted to know about jobs in India — but were afraid to ask” for the EAC-PM concluding that India added 12.8 million jobs in 2017 is therefore completely wrong. Bhalla and Das are also wrong in assuming arbitrarily that the CPHS’ female labour participation is unacceptably low.
According to the NSSO, rural female labour participation dropped from 19.7 per cent in 2009-10 to 18 per cent in 2011-12, averaging a drop of 85 basis points a year. According to CMIE’s CPHS, this dropped further to 15.75 per cent by early 2016 indicating an average fall of 56 basis points a year. So, comparing CPHS to NSSO implies a slower fall in rural female labour participation rate. Next, CPHS shows an increase in female labour participation rate in mid2016. So, there can be no complaint against CMIE for its estimate of rural female LPR till mid-2016.
In the case of urban female LPR, NSSO shows an increase from 12.9 per cent to 13.6 per cent between 2009-10 and 2011-12. CPHS shows a further increase to 15.8 per cent by early 2016. Again, CPHS estimates till mid-2016 show no unacceptable divergence from the direction and levels seen in the NSSO data.
Post mid-2016, the data shows a very sharp turn. I conjecture that the turn is caused by demonetisation. But, that is not important. What is important is what happened.
Female labour participation rates fell dramatically in both rural and urban regions after mid-2016. In rural India, it fell to 14.4 per cent in late 2016 and then to an average of 12 per cent in 2017 and around 11 per cent in most of 2018. In urban India, it fell from 16.4 per cent to 11 per cent in the first eight months of 2018.
CPHS is a panel survey. The same sample, the same execution machinery and the same estimation methods are used before and after 2016. So, if the data till 2016 are comparable with official data and therefore acceptable, then those after 2016 should also be acceptable.
The two government economists should now accept that the CPHS data is indeed credible and their misgivings are unfounded. The problems are in the real world, not in the data.