Source: Data for this table come from the monthly sales records of Casas de Pollo Rey (CDPR) for the stores in the experimental sample from December 2020 to December 2022.

Note: This table presents the impacts of digital training on monthly store sales in United States dollars (USD). The balanced sample includes stores that remained open from December 2020 to December 2022, six months after the experimental follow-up survey. Controls in the differences-in-differences specifications include store fixed effects, time dummies, municipality-specific time trends, and time trends that depend on initial store sales. Controls in the analysis of covariance (ANCOVA) specifications include 12 rounds of baseline sales, dummies for missing sales for each baseline round, municipality-specific time trends, and time trends that depend on initial store sales. Standard errors are robust to heteroskedasticity of unknown form and are clustered at the store-owner level. *** Significant at 1 percent. ** Significant at 5 percent. * Significant at 10 percent.