Ethnic and educational inequalities in COVID-19 vaccine hesitancy: cross-sectional analysis of the UK household longitudinal study

Background Vaccination is crucial to address the COVID-19 pandemic but inequalities in uptake may exacerbate existing health inequalities. We investigate the UK prevalence of COVID-19 vaccine hesitancy, identify which population subgroups are more likely to be vaccine hesitant, and report stated reasons for hesitancy.

Methods Nationally representative survey data from 12,035 participants were collected from 24th November to 1st December 2020 for wave 6 of the UK Household Longitudinal Study (‘Understanding Society’) COVID-19 web survey. Participants self-reported ethnicity, highest educational attainment, gender, age, how likely they would be to have a vaccine if offered and their main reason for hesitancy. Weighted cross-sectional analysis assessed the prevalence of vaccine hesitancy and logistic regression models estimated independent associations.

Results Overall vaccine hesitancy was low (18% unlikely/very unlikely). Vaccine hesitancy was higher in women (21.0% vs 14.7% in men), in younger age groups (26.5% in 16–24 year olds vs 4.5% in 75+) and in those with lower education levels (18.6% no qualifications vs 13.2% degree qualified). Vaccine hesitancy was high in Black (71.8%) and Pakistani/Bangladeshi (42.3%) ethnic groups. Odds ratios for vaccine hesitancy after adjustment for age and gender were 13.42 (95% CI:6.86, 26.24) in Black, 2.54 (95% CI:1.19, 5.44) in Pakistani/Bangladeshi groups, and 1.76 (95% CI:1.10, 2.82) for Other White (including Eastern European) ethnic groups (compared to White British/Irish). Vaccine hesitancy was not higher in all minority ethnic groups; for example, ORs were 1.11 (95% CI:0.64, 1.95) for Indian ethnicity and 0.67 (95% CI:0.24, 1.87) for Other Asian (including Chinese) ethnicity. Lower education was also related to vaccine hesitancy (no qualifications versus degree OR 3.54; 95% CI:2.06, 6.09) but ethnic differences largely remained when education was included in the model. For those who were vaccine hesitant the most common stated reason for hesitancy was concerns over unknown future effects (42.7%). However, when compared to the White British/Irish group, Black participants were more likely to state they ‘Don’t trust vaccines’ (29.2% vs 5.7%) and the Pakistani/Bangladeshi ethnic group more frequently cited worries about side-effects (35.4% vs 8.6%).

Conclusion Vaccine hesitancy is strongly associated with education and ethnicity, with marked ethnic heterogeneity. Black and Pakistani/Bangladeshi participants reported considerably greater vaccine hesitancy than White British/Irish ethnicity, but some minority ethnic groups did not. Educational inequalities did not account for ethnic differences. Vaccine programmes need to understand reasons for vaccine hesitancy within specific population sub-groups and take urgent action to improve uptake.