In addition to the dozens of statistics that County Health Rankings (CHR) use to determine the health outcomes and health factors rankings, they provide dozens of county level statistics that are not factored into the rankings such as: life expectancy, infant mortality and demographics. These statistics that can be just as illuminating as the ones included in the rankings. This post focuses on statistics significantly associated with COVID full vaccination rates. This is for Pearson coefficients greater than 0.634 or less than -0.634.

Motor Vehicle Mortality

The motor vehicle mortality rate was significantly associated with the vaccination rate (getting the first 2 shots). This association was negative, accounting for 63.8% of the variability with every increase by one in the motor vehicle mortality rate there is a predicted 1.67% decrease in the vaccination rate. Fayette and Bedford Counties were both outliers in this relationship. Fayette County has a high motor vehicle mortality rate and a high vaccination rate. Bedford County has a high motor vehicle mortality rate and the lowest vaccination rate in the 10 county region. These outliers cancel each other out in the regression model. Counties with higher case mortality rates from COVID appear lighter colored in the graph above. This could mean individuals who are more reckless with driving are more reckless with getting vaccinated.

Uninsured Rates

The percentage of the county having no health insurance is negatively associated with the full vaccination rate. This relationship is somewhat weaker than motor vehicle death rates accounting for 51.3% of the variability. For every 1% increase in the uninsured rate 5.4% decrease in the vaccination rate. The low percentage of the variability is due to 3 outliers: Bedford, Centre, and Fayette Counties. Bedford has a high uninsured rate and a low vaccination rate. Centre and Fayette have high uninsured rates and high vaccination rates.

Women and Men’s Earnings

Women’s median earnings were positively associated with vaccination rates accounting for 58.5% of the variability. The regression model states that for every $1,000 increase in women’s median earnings there is a 1.4% increase in the vaccination rates.

Men’s median earnings had a slightly weaker association with vaccination rates than Women’s accounting for 53.4% of the variability. The regression model states that for every $1,000 increase in women’s median earnings there is a 1.1% increase in the vaccination rates. The median household income and the gender pay gap was not associated with vaccination rates.

Child Care Center Rate

The rate of child care centers in the county was positively associated with the vaccination rates accounting for 55.6% of the variability. For every unit increase in child care centers in the county, there is a predicted 2.9% increase in the vaccination rate. Bedford County is a similar outlier as in the other charts.

% Broadband Access

The % with broadband internet access in the county is positively associated with vaccination rates accounting for 45% of the variability. For every 1% increase in broadband access there is a predicted 1.5% increase in vaccination rates. Bedford County had one of the lower rates than most at 77% along with the highest case mortality rate.

% Non Hispanic White

The % of the non-hispanic white population in the county is negatively associated with vaccination rates. This accounts for 51.4% of the variability. For every 1% increase in this percentage, there is a 1.6% decrease in the vaccination rate.

% Rural

The percentage of the rural population is negatively associated with vaccination rates. This accounts for 64.2% of the variability. This is the strongest association of all of these relationships with vaccination rates. For every 1% increase in the rural population there is a predicted 0.25% decrease in the vaccination rate.

These variables are all correlated with each other. The sample size is too small to try to tease these associations apart. The state does not make the data for all 67 counties in PA readily available. Next, I will look at the variables correlated with COVID Case Mortality.