The potential for vaccines to interrupt U.S. transmission of Covid-19 depends not only on technical efficiency in vaccine distribution, but also on the willingness of a large proportion of the public to be vaccinated. Though there has been strong demand for the relatively small amount of vaccine available initially, maintaining interest in vaccination is a longer-term challenge.
To understand public attitudes toward taking a Covid-19 vaccine and the factors likely to affect willingness to do so going forward, we examined 39 nationally representative, randomized polls with publicly available tabulations that were conducted between August 2020 and February 2021 (see Supplementary Appendix, available at NEJM.org). Our framework provides a perspective different from that of much of the media reporting on individual polls and informs our recommendations for outreach efforts to encourage vaccine uptake — efforts in which we believe physicians can play an important role.
Polls suggest that much of the U.S. public is currently undecided about whether to take a Covid-19 vaccine. This point is often overlooked, since interpreters of several recent polls have predicted that a majority will get vaccinated.1 But these interpretations are missing an important detail: reporting of poll results often involves collapsing various categories of responses. For example, reporting combines people who say they will “definitely” get vaccinated (40%) with those who say they “probably” will (23%) (see table). Polls using other wording reveal similar patterns, with substantial proportions of respondents indicating that they are “somewhat likely” to take a vaccine, for example. But evidence from political polling suggests that people who say they will “probably” or are “somewhat likely” to take an action do not always do so.2 Moreover, many people, when offered the choice, say they are “not sure” or that they will “wait until it has been available for a while to see how it is working for other people.” Thus, though there is potential for a majority of Americans to take Covid-19 vaccines, many people are apparently still making up their minds. Information they receive in the next few months could determine their decisions. To help motivate the public, we believe information should account for three central features of public opinion evident in these polls.
First, although the broader literature on Covid-19 vaccine adoption in the United States indicates that perceived effectiveness is an important motivator, poll findings warn that public perceptions of effectiveness may not align with expert views. In particular, the public may expect that an effective Covid-19 vaccine means not only strong protection against the virus, but also immediate changes in their daily experience. Indeed, when asked about reasons they would take the vaccine, many people said it would allow them “to go back to normal activities like work or school” (52%) or “feel safe around other people” (75%). Such perceptions may conflict with expert warnings that getting a vaccine — however effective it may be in preventing infection — does not warrant a return to normal.
Second, and also in keeping with other research on Covid vaccine uptake in the United States, polling shows that safety is a key consideration in individual decision making. Safety-related issues are consistently the top reasons for hesitancy regarding Covid-19 vaccination cited in polls, with 71% of respondents in one recent poll, for example, citing concerns about side effects. However, despite scholarship emphasizing the role of trust in institutions to provide relevant information, polls suggest that sources of technical information about safety are not greatly trusted. Specifically, there is limited trust in the media or pharmaceutical companies to provide Covid-19 vaccine information: as few as 16% and 20% of respondents, respectively, say they have “a great deal/quite a bit” of trust in these organizations to provide such information. The public also has only moderate trust in information provided by the Food and Drug Administration. More people, however, trust their own doctor or even health professionals as a group. This finding is consistent with polling about other health information, which reveals that doctors are frequently more trusted or viewed more positively than other sources, including elected officials, government agencies, and medical scientists.
Third, willingness to be vaccinated varies among U.S. communities. Black adults are less likely than White adults to say they will take a Covid-19 vaccine. In a recent poll, 31% of Black adults as compared with 46% of White adults said they would “definitely” get vaccinated. In the context of what is known about well-justified distrust among Black adults owing to historical and current racism in the U.S. biomedical enterprise,3 polling provides insights about how this distrust plays out in relation to Covid-19. For example, Black adults are less confident than White adults that vaccines have been properly tested for safety (67% vs. 76%), and less confident that medical scientists act in the public’s best interest (33% vs. 43%).
For different reasons, though also related to distrust, Americans who identify as Republicans are less likely than Democrats to say they will get vaccinated. A quarter of Republicans (26%), as compared with half of Democrats (52%), say they “definitely” will. This finding reflects a more recent polarization in the United States that affects responses to nearly every facet of contemporary policy.4 Regarding Covid-19, polls show that Republicans have less confidence than Democrats do that medical scientists will act in the public’s best interest (36% vs. 54%), and less trust in every source of vaccine information polls have asked about, with the exception of former President Donald Trump. Republicans thus have had little trust in President Joe Biden regarding such information (23% vs. 93%).
Having explored multiple polls, we believe that there is great potential for public willingness to receive Covid-19 vaccines but that effective public education and outreach are needed to maximize the proportion of the population that will do so quickly. We also believe that clinical physicians, rather than pharmaceutical companies, political leaders, or even medical scientists, should be at the fore of education and outreach strategies. Featuring clinicians in messaging is particularly important given that many people will not see their own physician when making vaccination decisions: current vaccine policy and cold-chain logistics mean that people will largely be attending mass-vaccination clinics. To reach communities that are less trusting of vaccine efforts, outreach should be led by, or should meaningfully incorporate, physicians reflecting the diversity of the relevant communities. Practically, this recommendation means that Black physicians, including those affiliated with historically Black medical institutions, should have a key voice. Similarly, physicians from well-respected medical institutions in Republican-leaning states should be incorporated into efforts in those states.
Assuming that the vaccines continue to have strong safety and efficacy profiles, vaccination programs can emphasize that message. Clinicians can also provide information about mitigating risks for patients with allergies or chronic conditions, which may affect people’s concerns about safety. At the same time, the media, including social media, may damage perceptions of vaccine safety by reporting frequently on the small number of people who experience serious side effects.5 It is therefore important to highlight for reporters as well as the public the relative rarity of these side effects.
Finally, for many people who want a vaccine, the motivation to “return to normal” is paramount. Although we must explain that precautions such as mask wearing will still be needed in many circumstances until vaccine adoption is widespread, leaders who wish to motivate vaccine adoption can balance such warnings with positive messages about how vaccination can facilitate a return to activities such as seeing friends and family or working in an office. These messages are especially important in relation to vaccines that require two doses, to avert drop-off before the second dose.
Such broad-strokes guidance will need to be adapted to state and local contexts, given variation in cultural contours even within particular demographic groups. Where possible, approaches should leverage additional polling and qualitative data comparing subgroups such as younger and older Black adults. Moreover, wherever possible, approaches should be customized to the group that is undecided within a particular population. Nonetheless, we believe that these broad lessons from polling can guide the large-scale messaging efforts needed at the national level to help motivate a large fraction of the public to receive Covid-19 vaccines.