Social distancing policies are necessary from a public health perspective but can have negative effects on economic activity. Using a newly constructed dataset of sectoral dependence on the use and sale of intermediate goods, this column investigates whether social distancing policies can have negative spillover effects on sectors that are not directly targeted due to input-output linkages. It finds that firms that depend on the sale of intermediate goods to sectors affected by social distancing measures are more affected by the crisis.
The COVID-19 health crisis has prompted governments to take extraordinary measures to save lives, including lockdowns and social distancing measures. The combined effect of the spreading of the virus and these measures has resulted in an unprecedented sharp decline in economic activity, as affected sectors were essentially shut down (Guerrieri et al. 2020).
In a recent paper (Laeven 2020), I investigate whether the social distancing policies, while necessary from a public health perspective, can have negative spillover effects on sectors that are not targeted by the policies, thus further depressing the overall economy. Specifically, I study the role of firm input-output linkages and social distancing in the transmission of the COVID-19 shock to the valuation of US corporates. Economic theory suggests that the initial shock to affected sectors can spill over to unaffected sectors through input-output linkages (e.g. Long and Plosser 1983, Acemoglu et al. 2012). To the extent that firms in unaffected sectors rely on intermediate inputs and demand for products from firms in affected sectors, social distancing can disrupt the ability of firms in unaffected sectors to produce and sell goods. One would therefore expect that firms whose suppliers and customers are concentrated in industries and states that are more affected by the COVID-19 shock and related lockdown measures would experience larger stock price declines compared to otherwise similar firms.
There is indeed a large variation in stock price performance across sectors during the outbreak of the virus (Table 1). The hardest hit sector is the mining, oil and gas sector, which was adversely affected by the sharp drop in commodity prices as the global economy came to a halt. On average, stock prices declined by 53.4% in this sector. Next are the entertainment and the hotel and restaurant sectors, which declined by 45.2% on average. Both sectors were largely closed down because of lockdown measures.
Table 1 Sectoral stock returns during the COVID-19 outbreak
Notes: This table reports firm stock returns over the first three months of 2020, averaged at the two-digit NAICS sector level. Sector 55 (Management of Companies and Enterprises) and sector 92 (Public Administration) are excluded from the table because there are no publicly listed firms from these sectors in our dataset.
To assess the significance of this spillover channel through input-output linkages, I construct a new dataset of the sectoral dependence on the use and sale of intermediate goods, using input-output tables from the US Bureau of Economic Analysis, which I combine with information on lockdown and social distancing measures at the state and sectoral level, including information on each sector’s physical contact-intensity from Kóren and Petö (2020) based on data from O*NET and the designation of (non)essential industries by the US. I measure the initial COVID-19 shock using information on the number of reported COVID-19 cases and deaths in the state, and combine all this information with financial and stock price data from Compustat.
Table 2 reports the measures of sectoral contact-intensity and dependence on the use and sale of intermediate goods by affected sectors. On average, about 37.1% of the workers were negatively affected by social distancing, but the variation across industries is large, from a low of 13% in the apparel manufacturing industry, which is heavily automated, to a high of 91% in health and personal care stores. For the average industry, 2.7% of total production depends on the sale of products to industries adversely affected by social distancing, but this can be as high as 26.4% in the case of the automotive repair and maintenance industry, which depends heavily on the sale of products to the motor vehicle and parts dealers industry. Dependence on intermediate inputs from social distancing-affected sectors also varies across sectors and is generally higher than dependence on the sale of products to other sectors. For the average industry, 14.0% of output depend on the supply of intermediate inputs from industries affected by social distancing, but this can be as high as 27.2% in the case of the nonferrous metal production and processing industry. This industry depends heavily on the metal ore mining industry for its inputs, which is an industry with a social distancing value of 71%, and to a lesser extent on the electric power generation, transmission, and distribution industry, which has a social distancing value of 46%.
Table 2 Social distancing and input-output linkages
Notes: This table reports industry-level social distancing and input-output linkages variables at the 2-digit NAICS sector level. Sector indicates the names of the two-digit NAICS sector name. Distancing is the share of industry employment affected by social distancing from Kóren and Petö (2020). Affected-sold is the fraction of total production sold to industries affected by social distancing. Affected-intermediate is the fraction of total output consisting of intermediate products from industries affected by social distancing.
My analysis focuses on the first quarter of 2020. This period encompasses the initial stage of the corona crisis, including: the outbreak of the virus in Wuhan, China, reported to the World Health Organization (WHO) on 31 December 2019; the first confirmed case of local transmission in the US in late January; the declaration of a public health emergency by the Trump administration on 31 January 2020; the designation by WHO of COVID-19 as a global pandemic on 13 March 2020; the guidance on (non)essential critical industries issued by the US Department of Homeland Security on 19 March 2020; and a series of lockdowns and stay-at-home orders at the state level during the second half of March 2020. I focus on stock returns over this period, conditioned on pre-determined firm characteristics measured as of end of 2019. I include the social distancing and input-output linkages measures alongside standard controls of stock return regressions (including market-to-book ratio, market value, cash holdings, and leverage).
I find that the returns of firms that operate in sectors that are more sensitive to social distancing measures are more adversely affected by the crisis. Moreover, I find a role for input-output linkages in the sense that firms that depend on the sale of intermediate goods to sectors affected by social distancing measures are more affected by the crisis. Several tests are consistent with the view that bigger firms and firms with larger cash buffers are better able to withstand these shocks, consistent with the notion that scale and deep pockets can help to buffer liquidity shocks (e.g. Harford 1999).
Both the direct effects of social distancing and its indirect effects through input-output linkages appear to be important drivers of stock prices during the outbreak of the pandemic. Our estimates imply that a one standard deviation increase in the industry’s share of workers affected by social distancing (the direct effect) is associated with a decline of 2.7% in stock returns, while a one standard deviation increase in the fraction of output sold to social-distancing affected sectors (the indirect effect) is associated with a decline of 3.1% in stock returns. The indirect effect of social distancing from the sale of products to other firms is therefore estimated to be quantitatively at least as important as the direct effect from social distancing. This implies that lockdowns impose large negative externalities on firms through input-output linkages, even for firms that are not directly affected by social distancing measures.
Acemoglu, D, V M Carvalho, A Ozdaglar and A Tahbaz-Salehi (2012), “The Network Origins of Aggregate Fluctuations”, Econometrica 80: 1977–2016.
Guerrieri, V, G Lorenzoni, L Straub and I Werning (2020), “Macroeconomic Implications of COVID-19: Can Negative Supply Shocks Cause Demand Shortages?”, Mimeo, University of Chicago.
Harford, J (1999), “Corporate Cash Reserves and Acquisitions”, Journal of Finance 54: 1969–1997.
Kóren, M and R Petö (2020), “Business Disruptions from Social Distancing”, mimeo, Central European University, March.
Laeven, L (2020), “Pandemics, Intermediate Goods, and Corporate Valuation”, CEPR Discussion Paper No 15022.
Long, J B and C I Plosser (1983), “Real Business Cycles”, Journal of Political Economy 91: 39–69.