COVID-19 vaccination willingness in peri-urban Tanzanian communities: Towards contextualising and moving beyond the individual perspective.

With only 5.1% of the population fully vaccinated against COVID-19, Tanzania has one of the lowest vaccination rates in the world and after two years of changing policies regarding the disease, the country struggles to get its vaccination campaign on the rails. In this study, we identify the determinants of COVID-19 vaccination willingness in two villages of the Mvomero district in Eastern Tanzania. Based on survey data, we performed univariate analyses to assess differences in vaccination intention for various social groups, and built a four-dimensional multivariate ordered logistic regression model that comprises respondents' personal and socioeconomic characteristics, the channels through which they get their information, their attitudes and perceptions towards COVID-19, and their social network embeddedness. Only 37.0% of the respondents indicated that they would be willing to get vaccinated against COVID-19. Vaccination willingness differed significantly according to gender, age, educational attainment and religion; with men, the elderly, people with post-secondary education and Catholics and Muslims more likely to accept a vaccine. Predictors of vaccination willingness were gender, age, social media and informal contacts as information sources, perceived effectiveness of the vaccine and of alternative medicine, fear of side effects, a general dislike of vaccines, and the proportion of vaccinated people and the highest value of trust in international organisations in one's network. Although people's attitudes and perceptions have the largest share of the explanatory value, our model shows that all four of our model's building blocks were imperative in explaining vaccination willingness. Therefore, our paper presents a compelling case for the inclusion of respondents' social embeddedness as a common dimension for exploratory models of vaccination willingness.