Why Paying People to Get Vaccinated Doesn’t Work

I’m going to start this week with a thought experiment.

I’m going to give you 1 million dollars. Here’s the catch: You can’t keep any of it. You need to use it to get as many people vaccinated against COVID-19 as possible. How would you do it?

The sky’s the limit here. You could pay Jalen Hurts a million dollars to do an Instagram story about getting vaccinated. Or you could buy billboards all around your local area, encouraging people to get the shot. Or you could send direct mail to people’s houses. Or robocalls! It’s up to you.

Some of you, brainstorming about this, might decide to just cut out the middleman. If I have a million dollars to get people vaccinated, maybe I’ll just pay people to get vaccinated. I could offer 100,000 people $10 each or 10,000 people $100 each.

This is the central insight behind financial incentives in public health. We spend a lot of money creating public service announcements to get people to eat healthy and exercise. We create websites full of accurate, reliable information about the benefits of quitting smoking or reducing alcohol intake. But instead of paying for an ad to convince people to stop smoking or something, what if we just pay the person to stop smoking?

Direct financial incentives are a fascinating intervention for a variety of behaviors we want to encourage that require a little extra motivation. And they’ve been tried, with some significant successes, across a variety of behaviors — from quitting smoking to taking HIV medications. But would it actually work for vaccination? Turns out, probably not.

Here’s the situation. It’s November 2021, about a year after the COVID vaccines came out, and you are German. More specifically, you live in Ravensburg, a charming city in the south of the country.

One day, a flyer comes to your door, signed by the mayor, giving you information about seven upcoming free COVID vaccination events. Are you going to go?

An alternative version is the same flyer, but this time it comes with a barcode. Go to a vaccination event and bring the barcode, and after your shot, you’ll get €40. Are you going to go now?

Final version. The same flyer, but this time, one person in your house gets the barcode but you don’t. You just get the information about the vaccination event — no promise of €40. But you know that some people are getting €40. Are you going to go now?

This thought experiment illustrates the concept of “spillover” in public health interventions. It is the effect of an intervention directed at one group of people on the people the intervention was not specifically directed toward. The problem is that a lot of studies only look at the targeted population, not the entire population, and failing to consider the big picture could lead to some incorrect results.

This phenomenon is ably described and evaluated in this study, «Financial Incentives for COVID-19 Vaccination,» appearing in JAMA Network Open, by John Ternovski and colleagues. I blatantly stole the thought experiment I just gave you from their study design. But they actually did it, so I can show you the results.

First, a bit more detail on how the study worked. The entire city of Ravensburg was randomized at the household level, so you have intervention houses and control houses.

One adult in each house was picked as the “representative” adult in the house. In control houses, each adult was sent the flyer that provided only information about the upcoming vaccination events. In intervention houses, the representative adults got the flyer with the bonus payment, but every other adult got just the informational flyer.

This setup allows three really interesting comparisons.

We can use the “representative” adults from each house to see how effective the bonus payment would be on increasing vaccine uptake among the targeted group. This is the so-called “direct” effect. I’ll just point out that most research studies would stop here. They would just say, OK, people who got the flyer with the promise of money were more likely to be vaccinated, without considering the other people around them.