How has the COVID-19 crisis been affecting hourly workers at small businesses? I've been working with @joinhomebase along with Marianne Bertrand, Feng Lin @rothstein_jesse @mattunrath @CAPolicyLab @UChiPovertyLab @RustandyCenter @ChicagoBooth to find out [1/10].

@joinhomebase provides scheduling & time clock software to small businesses across the US, generating daily data on hourly employee hours, providing a high-frequency picture of employment & hours. We build on Homebase's work ( https://joinhomebase.com/blog/real-time-covid-19-data/?sscid=41k4_6y5bc&utm_source=affiliate&utm_medium=partner&utm_campaign=389818) & document four facts: [2/10]

Fact 1: Firms have dramatically reduced employee hours. Each sub-plot in this figure shows distribution of hours across firms, measuring relative to average hours per week. By Mar 22, over 40% of firms shut down entirely and most others have large hours reductions. [3/10]

Fact 2: Hours reductions vary by industry's ability to operate under stay-at-home orders. Hours reductions are largest in Beauty & Personal Care / Leisure & Entertainment, declining over 90%, and smallest in Home & Repair / Transportation (even those hours declined by 50%) [4/10]

Fact 3: Hours start falling earlier in states with stay-at-home orders, but fall sharply by March 16 in almost all states. By March 28, total hours declined over 70% in states with the earlier stay at home orders, about 15 percentage points more than states without them. [5/10]

Fact 4: Hours reductions are primarily explained by firm shutdowns and hours reductions, not layoffs. [6/10]

I wouldn't be a good academic if I didn't add some caveats. Notably, Homebase data is disproportionately composed of small firms in food service, retail, & other sectors that employ many hrly workers. Data exclude most salaried employees & firms who don't use Homebase. [7/10]

Also, what we refer to as a firm shutdown is a shutdown of Homebase measured emp. If firms employ workers that don't schedule their time using Homebase & some workers remain employed, some of the hours losses we attribute to shutdowns may be properly attributed to layoffs [8/10]

We will post regular updates of this analysis, as well as further analysis of Homebase and other labor market and small-business data, at http://ChicagoBooth.edu/PovertyLab/COVIDresearch (you can also check the page out for more empirical details on the facts above) [9/10]

Thanks to @joinhomebase and the great teams at @CAPolicyLab, @IRLEUCB,
@RustandyCenter, @UChiPovertyLab for excellent work in pulling this together. [10/10]