The implementation of social distancing policies is key to reducing the impact of the current COVID-19 pandemic . However, their effectiveness ultimately depends on human behavior . In the United States, compliance with social distancing policies has widely varied thus far during the pandemic . But what drives such variability? Through six open datasets, including actual human mobility, we estimated the association between mobility and the growth rate of COVID-19 cases across 3,107 U.S. counties, generalizing previous reports . In addition, data from the 2016 U.S. presidential election was used to measure how the association between mobility and COVID-19 growth rate differed based on voting patterns . A significant association between political leaning and the COVID-19 growth rate was measured . Our results demonstrate that political orientation may inform models predicting the impact of policies in reducing the spread of COVID-19.