Association Between Statewide COVID-19 Lottery Announcements and Vaccinations

In the United States, the COVID-19 vaccination rate slowed from a peak of 3.6 million vaccinations per day during the week of April 5, 2021, to fewer than 2 million vaccinations per day by the week of May 3, 2021. To boost vaccine uptake, 19 states announced large cash lotteries by July 1, 2021, that were tied to COVID-19 vaccination.

For instance, on May 12, 2021, Ohio announced Vax-a-Million, a set of weekly $1 million drawings to be held over 5 weeks for Ohio residents 18 years or older who had received at least 1 COVID-19 vaccine dose. One recent study failed to find an association between the Ohio drawings and increased vaccinations.¹ In this case-control study, we assessed if announcements of cash drawings in 19 states were associated with increased vaccine uptake by comparing vaccination trends in states that announced drawings with states that did not using a difference-in-differences framework.

We used daily state-level COVID-19 vaccination data from the Johns Hopkins University Vaccine Tracker between April 28 and July 1, 2021, a period during which vaccine appointments were widely and rapidly available.² These data were combined with announcement dates from public news reports. We also collected COVID-19 case counts from The New York Times and restaurant and bar foot traffic from SafeGraph Inc.

We used a difference-in-differences framework for analysis, which compared daily reported COVID-19 vaccinations administered per 1000 population before and after the announcement in the 19 states that announced a lottery with those in states that never announced, while controlling for covariates. Details of the model can be found in eMethods in the Supplement. We also estimated event-study models, which interacted the difference-in-differences estimator with binary indicators for the number of days prior to or after the announcement. These estimates allowed us to assess how the association between announcements and vaccinations changed over time.

Analyses of secondary, deidentified data are exempt from institutional review board review, as determined by the University of Colorado. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.

There were 37.2 million first doses of COVID-19 vaccine administered in the United States between April 28 and July 1, 2021, including 19.2 million in states that announced cash drawings. Ordinary least squares estimates of the difference-in-differences model are summarized in the Table. Estimates of the association between an announcement and vaccination rates were very small in magnitude and statistically indistinguishable from zero. For example, for first doses, the adjusted regression estimate was −0.06 (95% CI, −0.43 to 0.30) daily vaccinations per 1000 population.

This finding was reinforced by the event-time estimates plotted in the Figure.³ No statistically significant association was detected between a cash-drawing announcement and the number of vaccinations before or after the announcement date, a period that included announcements of lottery winners for most lottery states. This finding was evident using a conventional event-time model (Figure, A) and a novel technique that better accounts for staggered announcements (Figure, B).

With the upper bound of the 95% CI for the first-dose estimate of 0.30 daily vaccinations per 1000 population, we can rule out fairly small associations between lottery announcements and vaccinations. If the cash-drawing announcements had been associated with 70% of US adults receiving a first vaccine dose by July 4 (President Biden’s stated goal), we would have expected an estimate of approximately 1.22 daily vaccinations per 1000 population.

Results of this case-control study may reflect several factors. Lottery-style drawings may be less effective than incentives that pay with certainty. Another possibility is that drawings were not an informative vaccine promotional strategy and that more complete messaging on vaccination would have been far more effective.^4,5 Also, individuals who are hesitant to receive COVID-19 vaccinations may be influenced by vaccine misinformation. This study is subject to limitations intrinsic to all analyses with nonexperimental data, and the potential for confounding factors remains. Moreover, these findings do not necessarily generalize to incentives for other vaccines.⁶

Accepted for Publication: August 18, 2021.

Published: October 15, 2021. doi:10.1001/jamahealthforum.2021.3117

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Dave D et al. JAMA Health Forum.

Corresponding Author: Andrew I. Friedson, PhD, Department of Economics, College of Liberal Arts and Sciences, University of Colorado Denver, PO Box 173364, Campus Box 181, Denver, CO 80217 (andrew.friedson@ucdenver.edu).

Author Contributions: Dr Sabia had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: All authors.

Acquisition, analysis, or interpretation of data: Dave, Friedson, Sabia.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: Dave, Friedson, Sabia.

Statistical analysis: Dave, Hansen, Sabia.

Obtained funding: Sabia.

Supervision: Dave, Friedson, Sabia.

Conflict of Interest Disclosures: Dr Sabia received research support for data analysis from the Center for Health Economics & Policy Studies at San Diego State University, which has received grant support from the Troesh Family Foundation and the Charles Koch Foundation. No other disclosures were reported.

Additional Contributions: We thank Kyutaro Matsuzawa, MA, and Samuel Safford, MA, who were compensated as research assistants at the Center for Health Economics & Policy Studies at San Diego State University.