Blinder-Oaxaca decomposition analysis of what psychographic factors explain the gap in vaccine aversion between the most and least averse subgroups based on selected indicators for the unvaccinated US adult population, Household Pulse Survey, 1 December 2021 to 7 February 2022
| Explanatory variables | Indicator (least vaccine-averse subgroup vs most vaccine-averse subgroup) | |||||||||
| Marital status (single vs married) | Mental illness symptoms (present vs absent) | Education level (<high school vs graduate degree) | Sexual orientation (bisexual vs undisclosed sexual orientation) | Gender (female vs undisclosed gender) | Health insurance (Medicaid vs VA/Tricare) | State (Maryland vs Alaska) | Race/ethnicity (black vs white) | Age, years (18–24 vs ≥65) | Annual household income, $ (<25 000 vs ≥200 000) | |
| Vaccine safety/effectiveness | ||||||||||
| Concern about possible side effects of a COVID-19 vaccine. | 0.0 | −0.1 | −2.7 | 3.7 | 3.0 | −0.4 | 0.7 | −0.1 | −0.6 | −1.8 |
| Not sure if a COVID-19 vaccine will protect me. | 0.3 | 0.9 | −1.4 | 0.2 | 0.2 | 3.0 | −1.2 | 0.3 | 0.7 | −0.8 |
| Plan to wait and see if it is safe and may get it later. | 20.2 | 23.9 | 3.5 | 20.8 | 14.3 | 30.6 | 27.6 | 10.0 | 43.6 | 11.1 |
| Any reason related to vaccine safety/effectiveness | 22.3 | 23.1 | 6.2 | 17.9 | 12.8 | 30.9 | 29.7 | 12.6 | 41.5 | 11.7 |
| Access challenges | ||||||||||
| My doctor has not recommended it. | 1.8 | 0.9 | 0.9 | 2.0 | 2.2 | 0.7 | 1.9 | 2.1 | 0.1 | 2.5 |
| Concern about the cost of a COVID-19 vaccine. | 0.4 | 0.2 | −0.6 | −0.4 | −3.9 | −0.1 | 2.8 | 0.0 | 0.1 | 0 |
| It’s hard for me to get a COVID-19 vaccine. | 2.5 | 1.9 | 0.7 | −1.6 | −6.2 | 6.7 | −1.1 | 1.7 | 2.3 | 2.1 |
| Any access reason | 4.7 | 2.9 | 1.0 | 1.4 | −6.5 | 7.2 | 3.2 | 3.9 | 2.3 | 5.3 |
| Social capital deficits | ||||||||||
| I don’t trust COVID-19 vaccines. | 6.9 | 3.3 | 8.5 | −10.7 | −0.4 | 8.8 | 19.7 | 12.2 | 0.7 | 8.5 |
| I don’t trust the government. | 6.3 | 3.3 | 12.3 | 11.4 | 21.1 | 20.8 | 30.6 | 15.8 | 2.6 | 20.1 |
| Any social capital deficits | 8.9 | 4.5 | 13.7 | 2.2 | 12.4 | 20.9 | 33.7 | 18.7 | 1.8 | 19.7 |
| Perceived susceptibility/seriousness | ||||||||||
| I don’t believe I need a COVID-19 vaccine. | 9.5 | 18.7 | 26.9 | 13.8 | 5.8 | 33.3 | 30.4 | 22.9 | −22.3 | 28.4 |
| I don’t think COVID-19 is that big of a threat. | 8.6 | 18.2 | 25.2 | 24.1 | 18.8 | 46.7 | 31.8 | 23.7 | −21.5 | 32.0 |
| Any indicator for perceived susceptibility/seriousness | 12.3 | 25.0 | 35.5 | 25.7 | 15.2 | 53.2 | 42.9 | 31.3 | −29.2 | 41.1 |
Decomposition analysis was performed for only those indicators with at least 10 percentage points difference between the most averse and least averse subgroups. Coefficients from Blinder-Oaxaca decomposition analysis can be positive or negative and all estimates above are in percentages. A positive value, for example, a coefficient of 20%, means that the observed gap in vaccine aversion between the two subgroups being compared would have been narrower by 20% had the most vaccine-averse group had the same level of the explanatory variable(s) as the least vaccine-averse group. Conversely, a negative value, for example, a coefficient of −20%, means that the observed gap in vaccine aversion between the two subgroups would have been even wider by 20% had the most averse group had the same level of the explanatory variable(s) as the least averse group.
Boldface indicates statistically significant explanatory effect (p<0.05).
VA, Veterans Affairs.