The COVID-19 hit economy is presenting a confusing picture. The government says that green shoots are appearing, while economic data gives mixed signals.
Data from different sources are giving a contrasting picture of the economy. For instance, some reports suggest that workers returning to the city are not finding work, while other reports talk of industry facing a labour shortage. Both are correct depending on which businesses one is talking of. The textile industry is perhaps running currently at 60% capacity, while FMCG is at 90% and the aviation sector at 25%.
The Centre for Monitoring Indian Economy (CMIE) and the ILO-ADB recently released their estimates on the state of job losses. The CMIE said that 18.9 million salaried jobs were lost during April 2020 to July 2020. But, employment in the informal sector increased by 8 million, compared to last year July.
The ILO report suggests that 4.1 million youth have lost jobs, especially from construction and agriculture. But these are mostly informal sector jobs.
So, one report suggests loss of jobs in the informal sector while the other portrays an increase in jobs. The CMIE data suggesting loss of salaried employment would also imply a large number of young losing employment so that the ILO numbers should have been larger.
The ILO report talks of the risk of a ‘lockdown generation’, implying a long lasting fall in employment and the economy. The CMIE also hints at a long term impact because it says that salaried jobs do not recover quickly.
Be that as it may, how accurate are these numbers? After all, employment data will give some sense of what is happening to production and incomes in the economy. It will reveal the impact of lockdown on production and how quickly the economy is recovering. The quarterly data released consists largely of limited data from the corporate sector and agriculture. It does not have the data for the unorganised non-agriculture sector. In effect it does not even fully represent the organised sector growth (or decline), a point made by this author since the time of demonetisation in 2016.
Also read: COVID-19 Impact: 41 Lakh Youth Lost Jobs in India, Says ILO-ADB Report
CMIE Survey: Impact of a shock
The estimates being given out by ILO and CMIE are based on quick surveys. The CMIE website presents its methodology and gives the sample size as 174,405 households consisting of 63,430 households from 3,965 villages and 110,975 households from urban areas. These households are surveyed over four months. 10,900 households are surveyed every week and 43,600 every month. So, CMIE is able to put out employment figures every week by projecting its results from its weekly sample.
A migrant worker wearing a protective face mask goes to work after few restrictions were lifted, during an extended nationwide lockdown to slow the spread of the coronavirus disease (COVID-19), in Kochi, India, June 1, 2020. Photo: Reuters/Sivaram V
The question arises how reliable would this weekly data be? When things are relatively stable this procedure may be alright. But in a fast changing situation as witnessed during lockdown and after it was eased in stages, the reliability would be less. The economy literally collapsed due to the huge shock administered to it. So estimation procedures that work during normal times are likely to fail when a shock is administered to the economy. A shock changes the various parameters of the economy.
For instance, the sample selected may no more be representative. As it is, the CMIE sample consists of 63.63% of households from urban areas while the share of such households in the population is about 30%. There must be good reasons for this choice. During normal times, this configuration remaining the same comparisons over time would be valid. But, not when the economy suffers a shock and the composition of population changes.
When people migrate in large numbers as they did from March to May, the population shares changed so that both the rural and the urban samples needed change but there is no way of doing that. So, the error in the survey could be quite large.
Next, those who stayed back in urban areas are the ones that most likely did not lose employment. The ones who lost employment are the ones who migrated but they would not be counted in the survey because they could not have been a part of the sample. So, an upward bias would occur in CMIE’s employment figures. With the exception of agriculture and some activities in open areas, most of the unorganised sector came to a halt during lockdown. Further in agriculture also there were persistent reports of lack of work during lockdown. This was on account of perishables being allowed to rot in the fields rather than being harvested. So, the unemployment figures would very likely be larger than what is depicted in the CMIE survey.
Distinction between work and employment
There is a bigger problem with data. Namely the definitional one – the distinction between working and being employed. During lockdown and afterwards many remained in employment but could not go to work or could not work from home. In such cases, counting employment would not give the complete picture of how much work was getting done. Various examples of this can be cited.
Also read: GDP Growth Rate Slump: Another Blow to Modi Govt’s Economic Growth Story
Shops, salons, theatres, hotels, restaurants, airlines, etc. did not retrench workers but work stopped. So, workers in these sectors would have reported themselves as employed but they were not working. Many of them did not receive a salary or had to take massive cut in salaries. Similarly, in manufacturing, factories largely closed down but did not fire workers who would have considered themselves as employed. Truckers deserted their trucks but would have been taken as employed.
Labourers wearing protective face masks load grocery items onto a supply truck at a wholesale market during the coronavirus disease (COVID-19) outbreak in Kolkata, July 24, 2020. Photo: Reuters
In the government sector work stopped over large swathes but government employees remained in employment. They received salaries but produced little. Technically they could be counted as producing since their contribution to the national income is measured as the salary they receive. Even after easing of lockdown many worked on alternate days. In the case of railways even now only about 800 passenger trains are running daily while 13,000 used to run in normal times. The staff is receiving salaries but work has greatly reduced.
Educational institutions were closed and classes were held via net. But teachers reported that many students could not attend classes. Many students reported that teachers were lecturing sporadically. Again, employment was not lost but work was greatly truncated. Temporary, ad hoc or guest lecturers or those working for private institutions hoped to continue and possibly reported themselves as employed but work stopped.
Self-employed like, an auto or an Ola driver possibly reported themselves as employed even if they could not ply their vehicles. Later when auto drivers started to sell vegetables using their autos as carts, they would have said that they are employed even if marginally so. Migrants who took up work under MGNREGS would be counted as employed but their regular work was gone. It is the employment under MGNREGS that has given a boost to rural employment figures. One needs to think about how to count this kind of work compared to what was being done earlier.
Daily wage labourers and homeless people wait to receive free food during an extended nationwide lockdown to slow the spread of the coronavirus disease (COVID-19) in New Delhi, India, April 28, 2020. Photo: Reuters/Danish Siddiqui/Files
In brief, three categories of issues are pointed to but there are more. First, those who reported as employed but did not work, second, those who worked intermittently and finally, those who got marginal employment in some other job. So, the employment figures from surveys designed before the shock was administered to the economy do not give the true picture of the economic decline persisting after the lockdown and quarterly data on GDP will not capture it. The economy is headed towards a new normal and not its pre pandemic state so new methods will have to be used to estimate employment and GDP.
Arun Kumar is Malcolm Adiseshiah Chair Professor, Institute of Social Sciences and author of forthcoming book on `COVID-19 and its Socio-economic Impact’. Penguin Random House.
