How do we get to yes?
Which is to say: How can public health experts and policymakers convince the persistently recalcitrant fraction of people who haven’t gotten vaccinated against Covid-19? It’s the key to ending the pandemic, and a brutal irony in the United States, one of the few countries with the money and wherewithal to have enough of the drug for every citizen. Yet still, something like 25 percent of US adults and teens haven’t done it. Or maybe more relevantly, between 15 and 30 percent of US residents tell pollsters and researchers that they will never get one—nope, nuh-uh, no way.
More than 700,000 Americans and more than 5 million people around the world have died from the disease. Vaccines against it reduce illness, deaths, and transmission. It’d be very good to get unvaccinated people to become vaccinated people.
One theory says: If someone could just find the exact combination of words, that might bring people around. Another idea is that vaccination should be mandatory. Carrots versus sticks, nudges versus shoves.
The literature on finding those perfect conceptual frames is a years-old canon. (Oh, you like hesitating about vaccines? Check out this T-shirt from resisting the human papillomavirus vaccine mandate.) But Covid is new, and the pandemic has been fast-moving and protean; that has made it frustratingly hard to assemble a coherent, effective public health campaign in the US, where resistance to vaccination (and other measures, like masks) has taken many forms.
New data from the polling company Civis made public this week has some insights here. Researchers there asked more than 5,000 unvaccinated people to look at eight different ways of talking about vaccination—“messages,” in the parlance of political communication—and then asked them if, after hearing those messages, they were more or less likely to go do the thing. People got randomized into groups that heard different approaches to talking about vaccines, and there was a control group of people who didn’t see any of the messages being tested. The messages were stuff like “It protects kids” and “You’ll be able to do things that require proof of vaccination,” all the way through “Vaccines are very safe” and even the deathbed words of someone who didn’t get vaccinated and got the disease.
The answer? It was complicated. Protecting kids made people about 6 percent more likely to say they now planned to get vaccinated. The safety and deathbed messages actually caused a guilt-induced backlash—people said they were less likely than they were before the test. Each message had varying effects depending on the recipients’ demographic backgrounds and political affiliations—a patriotic, do-it-for-America message worked only on Latinx respondents, but a message of personal decisionmaking only moved conservatives. (The biggest caveat, of course, is that people only reported their intention to get vaccinated. The study didn’t count actual needles in shoulders.)
That’s all in keeping with previous research. In a study of vaccine hesitancy in the United Kingdom in late 2020—before Covid vaccines were available—researchers at Oxford University asked 5,114 people how willing they’d be to get vaccinated, and why or why not. Of the ones who said they wouldn’t, the most common reason was that they didn’t believe in its “collective importance.” They either didn’t think there was a dangerous pandemic going on (belief in conspiracy theories correlates with vaccine resistance) or they didn’t care that a vaccine would help other people.
A few months later—early 2021—the same researchers did another study, a randomized trial of 18,855 people. This time, they measured the effectiveness of 10 different kinds of messages on people’s stated willingness to get vaccinated. (The control group just heard the basic UK National Health System spiel: That the vaccines are safe, effective, and protect against Covid.) Telling people that vaccines help prevent them from transmitting the virus to others, that it was safe even though it had a speedy development process, or that the pandemic is really bad and everyone getting vaccinated would help end it sooner—none of that worked. The only thing that moved the needle was emphasizing the personal benefits of getting the shot: It makes it less likely that you will get sick. Less likely that you will die.
Another survey, this one of a representative sample of 3,048 US residents last summer, similarly compared changes in intent-to-vaccinate after hearing language about vaccine safety, private benefit, social benefit, and economic benefit—sometimes in combination with each other. (A single basic message about rigorous testing of the vaccines in trials functioned as the control.) The biggest boost by far was in the private-benefit group; as the journal article about the research noted, even though vaccine-resistant people often cite concerns about safety and side effects, in practice the this-is-safe message did only slightly better than the control.
The Civis researchers found much the same thing. “We have tested that kind of framing in the past. With wearing a mask, for instance, we found that personal benefit was far more effective to talk about than protection of others,” says Crystal Son, director of health care analytics at Civis. “When we tested messages talking about the importance of mask use, one of them was: ‘Wear it for grandma.’ That one was zero percent effective, and had a high probability of negative backlash.”
Son’s team found more subtle effects in their most recent vaccine research. In similar Civis research a year ago, telling people that Covid vaccines might keep children from getting sick didn’t seem to change anyone’s mind; now, with most kids back in school and vaccine approval for kids looking more imminent, it did.
But in trial after trial, the only reliable way to convince vaccine-resistant folks to get vaccinated is to tell them it’ll benefit them personally.
So I’m a schmuck, basically. Since the beginning of the pandemic, I’ve been writing about how disasters actually make people work together, and how cooperation is the only way out of all this. In just about every story I wrote about vaccines once they became available, I’d add some boilerplate about how they were safe and effective, and that while vaccines would keep each of us safer, the best reason to get them was that they helped each of us protect us all. What a waste of pixels.
Here’s the weird part: People resisted entreaties, messages, even prizes. But when the hard-assed mandates started to slam down—get vaccinated or lose your job—the volume of complaints didn’t change, but the volume of vaccinations did. They worked. Lots of those people who said they’d never, nope, nuh-uh get vaccinated? Poll numbers showed percentages of “never” in the teens and “only if you make me” in the single digits. After mandates, those conditions flipped—it turns out the “nevers” were the much smaller group.
So what happened? Was there a different, successful message somehow embedded in the mandates? (Beyond, you know, “Do it.”) Or were the never-gonnas … lying?
To an armchair statistician, the numbers suggest it’s the second thing. Faced with the possibility of losing their jobs, the holdouts did not hold out. At United Airlines, just about 300 out of 67,000 US employees lost their jobs rather than get vaccinated. Of 26,500 people working for UCHealth in Colorado, just 100 said no. In California, where a mandate applies to all health care workers, 97 percent of Kaiser Permanente’s 216,000 employees got it; the company suspended 2,000 who didn’t. San Francisco has 35,000 city employees; just about 2,000 remain unvaccinated. Of New York City’s 150,000 public school employees, 99 percent of principals, 96 percent of teachers, and 94 percent of staff got it. New York state has more than 650,000 hospital and nursing home workers, and by the time the mandate there went into effect, 92 percent had gotten at least one dose—up 10 points from a week before. Tyson Foods went from 50 percent of its 120,000 US employees vaccinated in early August, when it announced a mandate, to 91 percent by the end of September. A long but incomplete list of companies with mandates includes Google, Facebook, Twitter, Goldman Sachs, Walgreens, Disney, and Morgan Stanley. Conde Nast, WIRED’s parent company, has one too.
Arguably, the message in all those mandates is something like: “Our organizational values include safety and care for others, and you are part of the organization, soooooo …” Or alternatively, maybe the effective message is: “You will lose your livelihood if you don’t get vaccinated.”
The nudge-versus-shove dichotomy is probably a false one. It’s about differences within the vaccine-resistant group, which turns out to be not a monolith at all. And it’s about the multiple ways that public health campaigns reach multiple audiences.
One way to think about why the personal benefit message works is to consider politics. In the same way that Covid denialism and anti-vaccine, anti-mask sentiments correlate pretty well to conservative political beliefs, those conservative beliefs correlate well to distrust in authority and an insensitivity to the idea of collective benefit and communitarian values. Political identification shows up as a factor again and again in this research. As the pollsters at the Kaiser Family Foundation have found, people who identify as Democrats are way more likely to be vaccinated, and believe that the recent Delta variant-fueled surge in Covid numbers is the fault of unvaccinated people. People who identify as Republicans are way less likely to be vaccinated, and think the Delta surge is the fault of immigrants and tourists. In the typology of the American Communities Project, liberal, younger cities, and closer-in suburbs tend to have vaccination rates of 60 percent; conservative and whiter evangelical hubs and working-class country areas are more like 35 percent.
Changes in the pandemic change people’s opinions. According to KFF, 56 percent of unvaccinated workers say they’d choose the alternative to vaccination built into the federal version of the mandate, which is getting a weekly Covid test; 12 percent say they’ll take the vaccine. The Delta surge pushed vaccination uptake of at least a single dose in the US from 67 percent to 72 percent, and the number of people making the (let’s face it, preposterous at this point) argument of “wait and see” to just 7 percent. But KFF still shows 4 percent saying they’ll only get vaccinated if required, and 12 percent saying they definitely won’t. That’s the opposite of what seems to actually happen when mandates hit. (Though technically if you get vaccinated because otherwise you’ll lose your job, that counts as a personal benefit, in a way.)
So it’s mandates for everyone, right? Maybe not. “Nudges are more about allowing authorities to direct the population in a positive direction by providing options,” says Mark Donald C. Reñosa, an epidemiology and biostatistics researcher at the Research Institute for Tropical Medicine in Manila and co-lead author of a recent review of nudgey vaccine interventions. “Shoves tend to be complicated, oftentimes use more resources, are expensive, and carry greater risk of failure.” Even though polls show most people in the US generally favor mandates, they can also (as some who oppose them have warned) crystallize opposition.
The difficult slalom, as I’ve written, is that mandates only work if they’re done right—less a shove, more a hug. They have to come with time off work to account for possible short-lived side effects, with education campaigns, with easy access to vaccination sites, with clear explanations. “Mandates work well if they’re not draconian,” says Saad Omer, director of the Yale Institute for Global Health and a vaccine mandates expert. “If you turn down the pressure valve altogether, a large chunk of people find ways to avoid the mandate. But the balance of convenience should be in favor of vaccination and away from non-vaccination.”
Timing matters too. None of these mandates would have worked nearly as well, Omer says, if they’d gone into effect this January, right alongside the arrival of the vaccines themselves. Public health interventions need time and encouragement to grow into becoming social norms and habits. Once all that gets in place, if there are still people who just don’t wanna, then the mandates do their work. “You don’t get from 30 percent to 80 percent with a mandate,” he says. “You go from 70 percent to 90 percent.”
In other words, the buffet of options has to be plates full of lots of different kinds of carrots, and very well curated sticks. That’s why all the different methods of communication, all the different types of messages aimed at different kinds of people, have to continue even in the face of resistance. “That’s the basic responsibility of government—to communicate effectively around a public health intervention, even if it’s also a mandate,” Omer says. “There was communication around seat belt effectiveness. We never stopped communicating until it became a social norm and a habit, not just ‘It’s the law.’” Those people who said they’d never get a shot and then turned around and got one weren't being deceptive. They were just, as Omer puts it, poor approximators of their future behavior. They needed the right inducement and the space to get there.
The job isn’t over yet. Omer warns that these halcyon days of mandate compliance could crumble again in the next few months, if approvals for a kids’ Covid vaccine lead to school districts adding it to the list required for enrollment, as the state of California already has. “Not everybody has to work with a large employer, but everyone has to send their kids to school,” Omer says. “I am not sure that political leaders recognize what’s coming. I still think if it’s done appropriately, I think it’s still the right thing to do, but they better be prepared.” (Especially if the resistance isn’t the garden-variety anti-vaccine sentiment familiar from the Before Times but is, as The Washington Post has reported, an astroturfed proxy war funded by the usual conservative superpowers, including the Koch, DeVos, and Walton families.) The political divide will be there; the language for supporting the mandates has to be there too. The road to yes, it seems, is long and winding.