Motivation: Equality of Access Need Not Imply Equity in Outcomes The COVID-19 pandemic has increased inequality across a variety of social, economic, and psychological dimensions through disruptions to education, economic activity, and employment (Bitler et al. 2020, Chetty et al. 2020, Halloran et al. 2021, Agostinelli et al. 2022). Work by our team (Alsan et al. 2021, Gilstrap et al. 2022) and others has shown that Black and Hispanic communities have been disproportionately affected, exacerbating already severe racial and ethnic inequalities (Perry et al. 2021, Cornelissen et al. 2022). The development of effective vaccines against COVID-19 held the promise of ameliorating the pandemics devastating effects on vulnerable and historically disadvantaged groups by empowering individuals to again live normal lives. The vaccines are free to all by law (Federal Register 2020) and are therefore an important government benefit with potential for profound impacts on social lives and economic livelihoods. Yet with COVID-19 vaccines initially in scarce supply, governments were forced to ration the distribution of this benefit as is common in many other public programs ranging from heating subsidies (Administration for Children and Families 2020) to housing vouchers (Chicago Housing Authority 2015) to Medicaid expansions (Finkelstein et al. 2012). In distributing any benefit, governments consider two important goals. One is equality: providing equal access to everyone. Another is equity: targeting using criteria such as social vulnerability, race, ethnicity, and age to equalize outcomes such as vaccination rates rather than access. While randomization can achieve equality in access to invitations, equity in outcomes is much more difficult to attain. Even if a disadvantaged group clearly has worse outcomes and therefore should be targeted, there is an empirical question of how intensively they should be targeted to attain equity in outcomes. In the case of COVID-19 vaccination, the answer depends not only on differences in outcomes but also on differences in takeup. We aim to overcome a major barrier to understanding the tradeoff between equality and equity in the distribution of vaccines and other health care to older adults using a rigorous research design with strong scientific premise based on an unusual natural experiment generated by randomized COVID-19 vaccine invitations issued by Michigan Medicine, a large health system with a national reach. Initial invitations prioritized equality in vaccine access through randomization. Approximately 14,000 of 113,000 eligible patients aged 65 plus received randomized invitations in January 2021 before vaccine supply unexpectedly dwindled. When supply recovered after initial invitees had sufficient time to be fully vaccinated, Michigan Medicine prioritized equity in outcomes by targeting more randomized invitations to individuals who met certain criteria by social vulnerability, race, and ethnicity. In later invitations to adults younger than age 65, Michigan Medicine targeted invitations using measures of health. We are constructing novel data on randomized vaccine invitations linked to electronic health records, state-level vaccination and death registries, and pre-vaccination survey data on social and economic behaviors. We have been approved by the Michigan Medicine IRB to make our data publicly available for further research. Our detailed data, novel context, and interdisciplinary expertise in health economics, econometrics, epidemiology, vaccine hesitancy, and medicine will enable us to inform new policies to address inequities.
