7.6 Index of Average Per Capita Household Consumption, by Worker Group

headed by informal and formal wage workers is significantly positive in April 2020 and insignificant in May 2020 (annex 7A, table 7A.7). These regressions suggest that, in rural areas, formal and informal workers were equally vulnerable to the early COVID-19 shock.

Evidence suggests that India’s flagship rural jobs guarantee program (MGNREGA) has cushioned the impact of COVID-19 in rural areas (Afridi, Mahajan, and Sangwan 2021). Given that there was a lockdown on MGNREGA activities until at least April 20, 2020, however, it is unlikely that MGNREGA alone explains why there was no differential job loss between formal and informal workers in rural areas in April 2020. Future research is needed to look into why formal jobs were as vulnerable as informal jobs to the early COVID-19 shock in rural areas.

THE IMPACT OF COVID ON CONSUMPTION

The CMIE survey also collects panel data on consumption, which the analysis used to estimate the impact of the COVID crisis on welfare. Figure 7.6 shows trends in

FIGURE 7.6 Index of Average Per Capita Household Consumption, by Worker Group

Source: based on CPhS data. Note: The figure shows mean per capita household consumption (natural logarithm) by worker category. For ease of comparability, all group means are normalized to 100 in February 2020, which is used as a base because it was the last prepandemic month. 95% standard errors are reported in shaded areas around the group means.

consumption per capita before, during, and after the onset of the COVID crisis by illustrating the same approach that was adopted in the case of income in figure 7.3. Specifically, total household consumption is divided by household size to calculate per capita individual consumption, and households are classified into three worker categories based on the employment status of the household head in December 2019. The log per capita consumption of each group in February 2020 is normalized to 100. The 95 percent confidence intervals of log per capita consumption are represented in the figure by the shaded bands.

As in the case of income per capita, consumption among all categories of workers was severely reduced in the months of maximum restriction linked to the pandemic. However, figure 7.6 also highlights an interesting puzzle: why do informal workers, who were exposed to larger job and income losses relative to formal workers, reduce their consumption to a lesser extent? This lower reduction is confirmed by estimations of equation 7.1 with the log of per capita consumption as the dependent variable. The results are presented in annex 7A, table 7A.8, and figure 7.7, which plots the interaction of time dummies with worker category indicators in table 7A.8, column 3.

The key message of figure 7.7 is that there is a positive, statistically significant consumption differential involving informal workers after March 2020 and a similar phenomenon (although not statistically significant) involving the self-employed. Without controlling for time-specific fixed effects, the reduction in consumption in April 2020 with respect to February 2020 was about 64 percent among formal workers (column 2 in annex table 7A.8), but it was only 45 percent among informal workers. Some of this differential is accounted for by industry-month and district-month fixed effects. The magnitude of the coefficient on Informal × April 2020 in annex 7A, table 7A.8, column 3, is 40 percent of the magnitude shown in column 1. However, even within the same industries, occupations, and districts and despite larger income losses, informal workers are affected by a smaller reduction in consumption relative to formal workers.

Three possible mechanisms may explain this puzzle: (1) forced savings, (2) differential price changes, and (3) differential adjustments to the crisis. First, the COVID economic crisis is not only an employment and income shock among households. It is also a goods and services supply shock. Some people, even if their incomes have not declined much, may be prevented from consuming certain items because some shops and services are closed or people wish to maintain voluntary social distancing. The data not only show that formal incomes are higher than informal incomes; they also show that the composition of the consumption of formal workers is quite different from that of their informal counterparts. Formal workers devote a much higher share of total consumption to nonfood items and to services, such as restaurants and other recreational services, that were unavailable during the lockdown. The unavailability of these services may have forced formal households to save a larger part of their income than they typically do.

Indeed, preliminary evidence on this forced saving effect may be seen in annex 7A, figure 7A.1. The figure plots the log of per capita income and consumption between July