Effect of retirement on COVID-19 vaccination in Europe: a quasi-experimental study

This study investigates how the transition to retirement influenced the uptake of COVID-19 vaccination in Europe. We employed a quasi-experimental regression discontinuity design to assess the impact of retirement on COVID-19 vaccination, accounting for unobserved confounding and controlling for factors such as subjective health status, age, vaccination for other diseases, medication use, cancer diagnosis, and internet use. We used individual-level data from the Corona Surveys of the Survey of Health, Ageing, and Retirement in Europe (SHARE), conducted in 2020 and 2021. Our outcomes were binary indicators capturing whether the respondents were either vaccinated or expressed the intention to get vaccinated. The independent variable of interest was retirement status, which is instrumented using age eligibility for pension benefits as discontinuous policy threshold in a fuzzy regression discontinuity design. Estimates indicated that retirement was associated with a decline in COVID-19 vaccination uptake. Specifically, retirees were less inclined to get vaccinated compared to individuals of similar age who had not yet retired. The estimates remained robust to different model specifications and falsification tests. Changes in the frequency of social contacts following retirement could offer a plausible explanation for the observed relationship. Given that retirees may have fewer social interactions and, as a result, reduced exposure to potential health risks, they might have viewed social isolation as a potential substitute for COVID-19 vaccination.