The economic carnage from the COVID-19 pandemic has unleashed a wave of corporate bankruptcies, most resolved either by reorganizing a company’s balance sheet or, more painfully, by liquidating its assets. The latter route can leave deeper scars on nearby businesses for years, find Harvard’s Shai Bernstein, Chicago Booth’s Emanuele Colonnelli, Columbia’s Xavier Giroud, and Brigham Young’s Benjamin Iverson.
Historically liquidations under Chapter 7 of the US Bankruptcy Code triggered 20 percent more job losses among neighboring companies in the five years following the bankruptcy filing than did reorganizations under Chapter 11, the researchers calculate.
The findings could inform policy makers as they wrestle with the potential snowballing effects of small businesses failing because of COVID-19. Since January 2020, about one-third of small businesses have closed, according to Opportunity Insights, a Harvard-based nonprofit tracking the economic effects of COVID-19.
The study results also challenge the popular theory of creative destruction, which holds that letting struggling companies fail ultimately benefits the economy, in part by attracting more-productive newcomers.
When a company liquidates, three-quarters of the local jobs lost can be directly tied to the bankruptcy, while the remaining quarter come from a lack of new jobs being created.
“Under creative destruction, we would expect higher employment following liquidation, or at least higher entry into the area, yet neither is supported by the data,” write Bernstein, Colonnelli, Giroud, and Iverson.
The researchers built a data set of 91,000 establishments that filed for bankruptcy between 1992 and 2005 using geolocation data from the Census Bureau’s Longitudinal Business Database and bankruptcy filings from LexisNexis. They focused specifically on the 40 percent of companies that started in Chapter 11 but later converted to Chapter 7, as this made it possible to compare the spillover effects of the two bankruptcy regimes.
But capturing these effects is tricky, since liquidations may be more prevalent in economically depressed and declining areas. To achieve a more directly comparable analysis, the researchers exploited the fact that jurisdictions assign bankruptcy judges at random, and judges interpret federal bankruptcy laws differently, with tougher judges tending to force Chapter 7 more often. “In a large data set, the characteristics of local economies assigned either a tough judge or a lenient one will be the same, with the only difference being that one block got dealt a liquidation and another didn’t,” says Colonnelli. “This allows us to claim that any difference in economic outcomes will be driven specifically by the effect of liquidation.” (The “block” Colonnelli is referring to is a census block, the smallest geographic area tracked by the Census Bureau, usually a rectangular grid about the size and shape of a city block or, in a less urban area, delineated by roads or other boundaries.)
The researchers also explored how the spillover effects of bankruptcy tend to propagate within a given census block. Liquidations, they find, wreaked broader havoc than reorganizations, mainly by reducing consumer traffic and hampering informal knowledge sharing between establishments. Larger liquidated companies (relative to the size of their blocks) spread more pain than smaller ones, and the spillover effects lessened over larger geographical areas, suggesting that a portion of dislocated workers found jobs in nearby regions.
The researchers note that liquidations caused larger job losses within blocks containing many small and young establishments, which tend to be more susceptible to business-cycle fluctuations. However, losses at small companies eventually spread to neighboring areas, driving down overall employment on the blocks before leveling out after the fifth year following a Chapter 7 filing, they find.