OBJECTIVE: The COVID-19 pandemic presented an unprecedented worldwide challenge to public health and the economy. However, the effect varied in different population groups and locations. This study examines how the preexisting socioeconomic conditions, income, education, access to broadband internet, and healthy foods affect people's pandemic-related health-protective behavior and lead to this differential impact in the U.S. METHODS: This study utilizes a dynamic panel-data-based regression analysis on a nationwide dataset of U.S. counties. RESULTS: Higher income, education level, and broadband speed reduce the probability of violating recommended preventive measures, promote health awareness among individuals, and lower COVID-19 cases. In contrast, low access to healthy foods increases the number of cases. Additionally, our study reveals several interesting findings related to people's food habits, living conditions, and the connection between education and access to broadband internet that contribute to the differential impact of countywide COVID-19 cases. CONCLUSIONS: Insights from our data-driven analysis enable healthcare authorities to understand where resources should be allocated at the beginning of a pandemic to reduce disease spread and thus, build a greater resiliency for the next pandemic. Hence, our findings provide significant insights to government and healthcare officials to devise health awareness campaigns and develop advocacy measures and policies in conjunction with policymakers to contain future pandemics.