Data from: The association of county-level presidential election outcome and COVID-19 mortality in Colorado, 2020-2022
Data files
May 12, 2025 version files 7.24 KB
-
CovidrepositoryDF.csv
4.82 KB
-
README.md
2.42 KB
Abstract
Context:
The Coronavirus disease 2019 (COVID-19) pandemic occurred during a time of political tension in the United States. County-level political environment may have been influential in COVID-19 outcomes.
Objective:
This study examined the association between county-level political environment and age-adjusted COVID-19 mortality rates from 2020 to 2022.
Design & Setting:
Political environment was measured by the 2020 Presidential election results and compared with age-adjusted COVID-19 mortality rates by county in Colorado.
Main Outcome Measures:
Rate ratios (RR) and 95% confidence intervals (CI) were estimated using negative binomial regression incorporating a population offset term. Models adjusted for populational differences using the demographics percentile from Colorado’s EnviroScreen Environmental Justice Tool.
Results:
Age-adjusted county mortality rates ranged from 14.3 to 446.8.0 per 100,000. 2021 COVID-19 mortality rates were nearly twice as high in counties voting for Donald Trump compared to those voting for Joseph Biden (adjusted RR = 1.98, 95% CI: 1.59, 2.47). Results for 2020 and 2022 mortality models were also in the positive direction, though the confidence intervals crossed null values.
Conclusion:
These results build on a growing body of evidence that the political environment may have been influential for COVID-19 mortality, helping to understand the drivers of health outcomes. Implications for the public health system as we shift into the endemic period of COVID-19 include motivation for collaborative work to restore and rebuild trust among and between stakeholders and the community, as well as increase health education given its’ influence on both individual and community behaviors.
https://doi.org/10.5061/dryad.573n5tbfh
Exposure data derived from the Secretary of State of Colorado website for the 2016 and 2020 Presidential elections. Covariate data obtained from EnviroScreen. Outcome data obtained through a data request for COVID-19 mortality by county from Colorado Department of Health and Environment.
Description of the data and file structure
All data collected and added to a constructed data frame in Excel for ease of use. Mortality data with * indicated suppressed data cells to avoid identification of cases when 2 or fewer deaths occurred in a calendar year for each county.
- county = Colorado county observation
- population20 = 2020 county level population
- population21 = 2021 county level population
- population22 = 2022 county level population
- binary2016 = binary indicator variable for 2016 election*
- binary2020 = binary indicator variable for 2020 election*
- percTrump2016 = percent of county voting for Trump in 2016
- percClinton2016 = percent of county voting for Clinton in 2016
- percTrump2020 = percent of county voting for Trump in 2020
- percBiden2020 = percent of county voting for Biden in 2020
- demographics = CO EnviroScreen demographics percentile for each county
- rate2020 = age adjusted Covid-19 mortality rate per 100,000 in 2020
- rate2021 = age adjusted Covid-19 mortality rate per 100,000 in 2021
- rate2022 = age adjusted Covid-19 mortality rate per 100,000 in 2022
* For this dataset, the logic variables indicated exposed and unexposed groups. Specifically, the binary2016 and binary2020 variables describe the county-level Presidential election outcome for each year. 0 indicates that Hilary Clinton won in 2016 and Joseph Biden won in 2020, while any 1 for either variable indicates Donald Trump won that county in that year's election.
Sharing/Access information
Links to other publicly accessible locations of the data:
All exposures and covariate data was publicly available. Mortality outcome data obtained through a data request for Colorado Department of Public Health and Environment. Data was organized into an Excel file for ease of use and analyzed in R.