COVID-19 Sends the Bill: Socially Disadvantaged Workers Suffer the Severest Losses in Earnings

*Tharcisio Leone*

## **Abstract**

This work uses a nationally representative household survey conducted by phone during the COVID-19 pandemic to estimate the short-term impacts of lockdown measures on employment and income in Brazil. In May 2020, 18 percent of the employed population (around 15.7 million workers) were temporarily absent from their jobs due to the lockdown policies while 56.6 percent of them were no longer earning an income from work. Similar figures were registered in June 2020. This decrease in employment has generated a fall of 18 percent in the average work income and an increase of 0.014 points in the Gini coefficient. The vulnerable among the population have been hit hardest by the pandemic: the average earnings of the lowest income decile decreased from BRL 389.07 to 0 while for the secondlowest a 70.2 percent reduction has been seen (from BRL 878.08 to BRL 262.06). Thanks to the implementation of the COVID-19 Emergency Aid, the Brazilian government has been able to reduce the losses in income for all social classes. Nevertheless, the average income of the first decile is 5 percent lower than the value pre-pandemic while for the second decile the equivalent figure is 15.2 percent.

**Keywords:** COVID-19, lockdown effects, income, employment, emergency aid, Brazil

#### **1. Introduction**

Since the outbreak of the novel coronavirus disease in China, COVID-19 has profoundly affected the daily routine of the great majority of the global population and plunged the world into a crisis of unprecedented scope [1]. At an early stage herein, the goal was to avoid the overburdening of the health system. Many countries have worked to "flatten the curve", taking such restrictive measures as travel bans, lockdowns, stay-at-home orders, and quarantines—some of them extremely stringent—to reduce the movement of persons, and, consequently, to slow down the spread of the virus [2].

As the pandemic unfolded, Brazil would become a global hotspot. At the time of this writing, the seventh-most populous nation on the planet was then the secondworst-affected country worldwide with more than 18 million confirmed cases and 510 thousand deaths due to COVID-19. Since the confirmation of the first coronavirus case in Brazil on February 26, 2020, the policy responses to combat the spread of the pandemic have been scattered and uncoordinated [3]. The federal and local

governments have found themselves in constant disagreement over the lockdown measures necessary to flatten the curve [4]. However, despite the opposition of President Jair Bolsonaro, all twenty-seven Brazilian states would implement between March 13 and 24, 2020, lockdown measures to reduce the circulation of persons and consequently the spread of the virus. In subsequent weeks, the municipalities followed suit, enacting additional legislation to regulate these stay-at-home orders.

Some empirical studies have already been able to confirm that these lockdown policies were successful in increasing social distancing during the pandemic. Leone [5], for example, used geolocation data from nearly sixty million smartphone users in Brazil to show that the population numbers socially distancing grew considerably after the implementation of the lockdown measures. While the share of stay-athome individuals in the pre-pandemic phase (January and February 2020) was close to 20 percent, this number increased to 50 percent in the first weeks after lockdown policies were introduced. Similar results were also reported for Italy [6], Sweden [7], and the United States [8–10].

It is therefore no surprise that the impacts of the pandemic go way beyond the mortality rate, and the government responses to it will certainly cause turmoil for the economy. Fernandes [11] estimates a decline of 10.4 percent in global gross domestic product under the scenario whereby the lockdowns last until the end of July. In contrast to many European countries who can mitigate at least parts of their lockdowns' economic disruption through welfare states, in developing ones the most vulnerable among the population have tended to be the biggest losers during the pandemic given the lack of social security coverage [12]. A World Bank study concluded that, at the global level, COVID-19 is pushing between 40 and 60 million into extreme poverty [13]. Sumner et al. [14] also used simulation models to quantify the potential short-term economic impact of the lockdown policies, highlighting that in some regions of the world the pandemic could result in poverty levels being reached similar to those recorded 30 years ago. Based on the worst-case scenario whereby the per capita income decreases 20 percent—the number of people living in poverty could increase by up to 580 million as compared to 2018.

Despite the valuable contribution of all these empirical simulations of the economic costs related to COVID-19, it is high time to abandon the forecasts and start to estimate the real socioeconomic consequences of the pandemic. This is exactly the main contribution of this chapter to the literature. This work will apply realtime measures of work activities and income levels during the pandemic in Brazil to quantify the short-term economic impacts of COVID-19-related social-distancing policies.

Consequently, this chapter contributes to the rapidly growing literature describing the economic impacts of the coronavirus crisis. While similar studies for developed countries have largely flourished over the past year [15–17], empirical evidence for developing countries is still rare in the literature. Therefore, this work will be the first to apply data from the PNAD COVID-19—a (national) representative household survey conducted by phone with 349,306 Brazilian residents during the pandemic—to estimate the negative impact of the lockdown on employment and income levels in Brazil. Overall, the findings suggest that the measures to flatten the curve have led to a reduction in employment and income—with more significant losses occurring for the most vulnerable parts of the Brazilian populace.

#### **2. Data and method**

This manuscript uses data from PNAD COVID-19, a recent nationally representative longitudinal survey conducted by the IBGE (Brazilian Institute of Geography

#### *COVID-19 Sends the Bill: Socially Disadvantaged Workers Suffer the Severest Losses… DOI: http://dx.doi.org/10.5772/intechopen.102030*

and Statistics) with 193,662 households (349,306 individuals). The intention behind it is to continuously produce information about the health status and labor-market characteristics of the domestic population during the pandemic (May–November 2020). Data collection was carried out remotely via telephone calls, and drew on the sample of the PNAD-Contínua (Continuous National Household Sample Survey).<sup>1</sup>

To provide information in real time about the pandemic, PNAD COVID-19 adopts a rotating panel scheme of interviews to produce weekly and monthly consolidated data. This means that every week one-quarter of the involved households are interviewed, and the ongoing main descriptive statistics are published immediately. Then, at the end of the month, the data for the whole sample (193,662 households) are consolidated and made available to the public in the form of microdata. Given the panel structure of the survey, the households can be correctly identified across all the months of PNAD COVID-19, and they can be also linked with the data of the Continuous PNAD—thereby providing the relevant information for the period pre-coronavirus.

From the PNAD COVID-19 sample, two main pieces of information related to personal income will be used in this study here. The first variable (C10) refers to the (normal) earnings before lockdown policies were implemented in March 2020, while the latter (C11a) investigates the same type of income during the pandemic itself. In addition, I use the information on the allocation of COVID-19 Emergency Aid (D0051) to estimate the effects of this social scheme on income distribution.

In a first step, the manuscript identifies the economically active population from the sample, which corresponds to employed and unemployed persons aged 15–64. Employed are those individuals who have worked at least one hour during the reference week or who were temporarily absent from their job; unemployed persons meanwhile are those who were not employed but made some specific active effort to find a job in the same reference week. PNAD COVID-19 shows that the lockdown policies have deeply affected job-seeking, since many people were dissuaded from leaving the home or have anticipated that companies would postpone their new hires. For this reason, this chapter assumes as unemployed also those persons who have not looked for a job but wanted to be working during the identified reference week. Finally, this manuscript differentiates between the population employed in the formal and informal labor markets. Informal are those workers with no employment contract registered via the Work and Social Security Card and self-employed persons who do not pay social security contributions (INSS).
