Vaccination is Local: COVID-19 Vaccination Rates Vary by County and Key Characteristics

Introduction

COVID-19 has disproportionately affected certain underserved and high-risk populations, including people of color, those with underlying health conditions, and those who are socioeconomically disadvantaged. Ensuring access to COVID-19 vaccines for these communities can help address the disparate health effects of the virus and achieve herd immunity. The Biden administration has identified vaccine equity as a priority, but states and local jurisdictions vary in how and the extent to which they prioritize equity. Given that vaccine roll-out in the U.S. is inherently local, understanding how vaccination rates vary at the local level is important for informing outreach efforts and addressing equity. Earlier CDC analysis found that, as of early March, counties with high social vulnerability had lower vaccination rates than counties with low social vulnerability.

This issue brief builds on that analysis by analyzing how vaccination rates to date vary by counties and identifying key county characteristics that are associated with higher or lower county vaccination rates. It is based on KFF analysis of data from Centers for Disease Control and Prevention (CDC) that shows the percent of the population fully vaccinated at the county-level; the data also provides this share for those ages 65 and older. The CDC data, while incomplete (see methods), provides comprehensive data to examine vaccine rollout at the county level. The results in this brief use data as of May 11, 2021 and include 77% of all counties (2,415) in the US. See Methods box for a more detailed discussion of the data, measures included in the analysis, and methodology.

Key Findings

There is wide variation in reported vaccination rates by county across the US (Figure 1). Across the 2,415 counties included in this analysis, the average county vaccination rate weighted by population is 33.6% for the total population. Nearly a third of all counties have fully vaccinated a third or more of the county population, while 1% of all counties have fully vaccinated 50% or more of the population.

There are mixed findings on how county vaccination rates align with national prioritization recommendations and state prioritization decisions. The CDC recommended that as long as vaccine supplies were limited, certain groups be prioritized for early vaccine access, including healthcare workers and long term care residents, as well as older adults and people with medical conditions that put them at greater risk of severe COVID-19 illness. Although all states have now expanded eligibility to everyone ages 16 or older, initially, they prioritized vaccinating older adults in addition to healthcare workers and long-term care residents. In part due to earlier prioritization of older adults as well as higher vaccine uptake at older ages, counties with higher shares of people ages 65 and older have higher vaccination rates (31.4%) than those with lower shares of people ages 65 and older (29.8%). In contrast, most states were slower in opening up eligibility to those with high-risk medical conditions and the list of qualifying conditions differed across states. We found that counties with higher shares of people with certain high-risk medical conditions have a lower average vaccination rate, compared to those with smaller shares (25.5% vs. 33.9%).

Additionally, county vaccination rates do not seem consistently correlated to COVID-19 impact. Importantly, counties with high scores for community transmission of COVID-19 have higher vaccination rates than counties with low community transmission scores. The CDC defines community transmission levels for counties based on cases per 100,000 population and shares of positive tests in the past week. Counties classified as having “high” community transmission levels (and thus have higher numbers of cases and positive tests) have an average vaccination rate of 30.5% compared to 27.2% for counties with “low” community transmission levels. Because community transmission levels reflect current cases and positive tests, people in counties with higher transmission levels may be more motivated to get vaccinated. It may also be the case that COVID-19 transmission is currently higher in counties that have higher vaccination rates due to other factors. At the same time, counties with high shares of cumulative cases and deaths per 100,000 population have lower vaccination rates than counties with low shares of cases and deaths. In the counties with the highest cumulative cases and deaths per 100,000 people, the average vaccination rates are 29.1% and 28.3%, respectively compared to 31.8% and 32.1%, respectively in counties with the lowest numbers of cases and deaths per 100,000 people.

Higher county uninsured rates and poverty rates are associated with lower vaccination rates. The average vaccination rate in counties with high uninsured rates is 25.8% compared to 33.9% in counties with lower uninsured rates. Higher poverty rates, which are associated with lack of insurance coverage and may contribute to other barriers to accessing vaccines, are also associated with lower county vaccination rates.

Consistent with CDC’s recent research, counties that rank high on the Social Vulnerability Index (SVI) (those with higher vulnerability) have lower vaccination rates than counties that rank lower on this index (25.8% vs. 32.5%). This finding is not surprising given that SVI aggregates measures of socioeconomic status, age, race/ethnicity, some of which we discuss above, along with additional measures of disability, housing type, and transportation access. Regardless, counties with higher vulnerability currently have lower vaccination rates than counties with lower vulnerability.

Metro counties have higher vaccination rates for the total population than non-metro counties, and vaccination rates are lower in counties that voted for Trump compared to those that voted for Biden. The share of the total population vaccinated in metro counties is 31.3% compared to 28.7% in non-metro counties. In addition, consistent with evidence of vaccine hesitancy among Republican voters, the average vaccination rate in counties that voted for Trump in the 2020 election is 28.5% compared to 35.0% in counties that voted for Biden.

Conclusion

This analysis of county-level vaccination rates provides further evidence of inequities in COVID-19 vaccination efforts to date. Counties with higher shares of people disproportionately affected by COVID-19, including people with high-risk medical conditions, and those living in poverty, have lower vaccination rates than counties with lower shares of these populations. Additionally, counties with higher social vulnerability have lower vaccination rates that counties with lower SVI. These data can be used to help direct continued outreach and vaccination efforts going forward. In addition, we also find that counties with higher levels of community transmission of COVID-19 have higher vaccination rates, suggesting that people in those areas are responding to the situation in their communities and efforts to accelerate vaccination and stave off rising cases in some areas appear to be working.

Methods
This issue brief analyzes data at the county level and draws from multiple sources.

Our main outcome of interest, vaccination rates by county, was collected from the Centers for Disease Control and Prevention’s (CDC) COVID-19 Integrated County View. The CDC data reports completed vaccination rates for total population and population over age 65. Data are not reported for Hawaii, New Mexico, Texas, and the smallest counties in Alaska and California. In addition, we exclude data for counties where less than 80% of vaccination records include county of residence, which eliminated data for Colorado, Georgia, Vermont, Virginia, and West Virginia. The analysis includes data for 2,415 counties, 77% of total counties (3,142) in the US.

The average population-weighted county vaccination rate for the total population is slightly lower than national estimates due to missing data from several states and counties as noted above.

We also use the CDC data for “community transmission level”, which classifies counties based on new cases per 100,000 people and the share of positive tests in the last 7 days.

Data to categorize counties by demographic characteristics of residents is pulled from the Census Bureau’s 2019 American Community Survey 5-Year Estimates by county. We use ACS data to categorize counties by residents’ age, race/ethnicity, poverty, and health coverage. Specifically, we calculate the share of the county population that is over age 65, people of color, Non-Hispanic Black, Hispanic, in a family with income below poverty, and is uninsured.

Data on the cases and deaths were pulled on May 11, 2021 from the Johns Hopkins University county data. To calculate cases/deaths per 100,000 population, totals for each county were pulled from the Census Bureau’s demographic data – using total population.

The 2020 Presidential Election results were pulled from a GitHub repository that compiled data from media sources including The Guardian, townhall.com, Fox News, Politico, and the New York Times. Alaska is excluded from this component of the analysis as the only data available is at the district-level and cannot be cross-walked onto counties.

Metro and non-metro classifications are based on the U.S. Department of Agriculture’s 2013 Rural-Urban Continuum Codes. Counties with codes 1 through 3 are classified as “metro” and 4 through 9 are classified as “non-metro.”

Data on the share of the population with underlying medical conditions that put them at higher risk for severe COVID-19 illness by county came from the CDC. The conditions included chronic kidney disease, COPD, heart disease, diagnosed diabetes, and obesity (BMI >= 30).

County Social Vulnerability Index (SVI) is from the CDC’s Agency for Toxic Substances and Disease Registry. SVI indicates a community’s vulnerability based on certain social conditions (i.e. socioeconomic status, household composition, language, etc.) that may affect the community in the event of a disaster.

To classify counties, we translate continuous measures into categorical outcomes, using the group definitions below:

  • For continuous measures of age, poverty rate, uninsured rate, high-risk medical conditions, SVI, and cases and deaths per 100,000, we classified counties below the 25th percentile of overall distribution of counties for each measure as “low,” counties above the 75th percentile as “high,” and all other counties as “medium.” Specific breaks for each variable are indicated in Figure 2.

Given ongoing concerns related to equitable access to the COVID-19 vaccine, assessing differences in vaccination rates by race/ethnicity at the county level would add to existing national and state level data. However, the results of our analysis do not have face validity when compared to data analyzed at the individual level showing that people with Hispanic ethnicity are vaccinated at lower rates than White people, leading us to conclude that there are confounding factors driving the results based on county-level racial and ethnic composition. Therefore, we do not include comparisons of vaccination rates based on county racial and ethnic composition.

KFF Headquarters: 185 Berry St., Suite 2000, San Francisco, CA 94107 | Phone 650-854-9400
Washington Offices and Barbara Jordan Conference Center: 1330 G Street, NW, Washington, DC 20005 | Phone 202-347-5270

www.kff.org | Email Alerts: kff.org/email | facebook.com/KFF | twitter.com/kff

The independent source for health policy research, polling, and news, KFF is a nonprofit organization based in San Francisco, California.