The Public Health Effect of Mandatory and Voluntary Food Safety Measures: Generalized Synthetic Control Methods on Milk Pasteurization in the United States

This paper investigates the public health effects of mandatory and voluntary milk pasteurization adoptions in the United States. With generalized synthetic control methods (GSCM), we estimate the casual health impact of city pasteurization ordinances in 16 treated cities. Our results show the average treatment effect of city with pasteurization ordinances was obvious. Children diarrhea mortality rates on average in the treated group is 7.62 lower than their synthetic versions. And the number was only 1.04 in the control group. The treatment effect measured by GSCM is also different from the results measured by classical linear based Differences-in-Differences model. Next, we implement a FE analysis to examine the effect of voluntary pasteurization adoptions. We choose robust FE estimators to minimize the influence of observations of extreme values. Also, our robust panel estimations indicate the spread of pasteurization was significantly responsible for the improved health profile in US cities. In sum, we find pasteurization had its unique and significant contribution to public health in history