Do Minimum-Wage Increases Accelerate Automation in Restaurants? A Difference-in-Differences Test Using the Automatable-Employment Share

A higher wage floor pushing restaurants toward self-service kiosks is a causal hypothesis, not an identity. No public data count kiosks, so we test the observable signature the hypothesis must leave: the share of restaurant employment held by cashiers and counter attendants, the two occupations whose core tasks a kiosk reproduces. We build that share for every state and year from 2011 to 2019 from the public-use files of the Current Population Survey. Treatment is coded from the Vaghul–Zipperer minimum-wage series: seventeen jurisdictions that raised their binding floor by at least five percent in 2014, 2015, or 2016, compared with twenty-one states that stayed at the federal $7.25 throughout. We estimate staggered-adoption effects with the Callaway–Sant'Anna estimator and a border design of 26 neighboring state pairs, with wild-bootstrap inference clustered by state. We find no detectable effect. Averaged over the four years after an increase, the automatable share moves by +0.02 percentage points (95 percent confidence interval −1.02 to +1.05) against a baseline of 11.3 percent. The border pairs give −0.50 (−1.70 to +0.71). The estimates rule out declines larger than about one percentage point, or roughly a tenth of the front-counter share, in the main design. A placebo, alternative control groups, weighting, a broader occupation set, employment levels, a dose-response analysis, and a Seattle case study all point the same way. Minimum-wage increases of the size states adopted in 2014–2016 did not measurably shift restaurant employment away from the occupations kiosks replace within four years.