New Preprint with Daniel Prinz: How do #Shutdown policies affect unemployment during #COVID19? We find that limits on bars and restaurants caused modest increases in unemployment and discuss implications for relaxing restrictions. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3581254 #EconTwitter

To get high-frequency data on unemployment, we proxy for UI claims using @GoogleTrends data, borrowing from @paulgp and @aaronsojourner's awesome work predicting future UI claims with Gtrends data.

Using NPI timing from @KFF, we show that NPI announcement leads to an impulse response in Google searches for “file for unemployment.”

...but this response is a fraction of the overall increase in Google searches over the same time period. We estimate that bar/restaurant limits caused only 6% of UI claims between March 14 and 28, and 29% of UI claims in the Food and Accommodation industry.

Our results suggest that “direct pandemic effects” on demand and supply account for the majority share of job losses (even if the bars are open, most people aren’t going out). @Opentable reservations were already declining before closure orders (h/t @CharlesFLehman)

So these jobs come back if bars and restaurants open? We discuss why our estimates are likely an *upper bound* for the jobs that would return (due to re-hiring costs, severed employer-worker relations, and worker/firm expectations). See our paper for this and lots of robustness!

We benefited from a growing effort of economists like @Jabaluck, @IvanWerning, @paulgp, @aaronsojourner, @Claudia_Sahm, @jasonfurman, @arindube, and many others to contribute to the econ/health policy discussion around COVID,

and we are grateful for suggestions from @Samantha_Burn, @MonicaFarid9, Ed Glaeser, @lkatz42, Jim Stock, and especially @timothyjlayton.