Imagine that a thousand people—randomly selected from the U.S. population—had unprotected sex yesterday. How many of them will eventually die from contracting HIV from that single sexual encounter?

Now, imagine a different thousand people. These people will drive from Detroit to Chicago tomorrow—about 300 miles. How many will die on the trip as a result of a car crash?

Which of those two numbers is bigger?

If you’re anything like the participants in a new study led by Terri D. Conley of the University of Michigan, the HIV estimate should be bigger—a lot bigger. In fact, the average guess for the HIV case was a little over 71 people per thousand, while the average guess for the car-crash scenario was about 4 people per thousand.

In other words, participants thought that you are roughly 17 times more likely to die from HIV contracted from a single unprotected sexual encounter than you are to die from a car crash on a 300-mile trip.

But here’s the deal: Those estimates aren’t just wrong, they’re completely backward.

According to statistics from the U.S. Centers for Disease Control and Prevention and the United States National Highway Traffic Safety Administration, you are actually 20 times more likely to die from the car trip than from HIV contracted during an act of unprotected sex.

Why were the participants’ estimates so far off?

Conley and her colleagues think the answer has to do with stigma: Risky behavior related to sex is judged more harshly than comparable (or even objectively worse) health risks, when you control for the relevant differences between the behaviors.

“It seems that as a culture we have decided that sex is something dangerous and to be feared,” Conley told me in an interview. That’s why, she argues, U.S. parents try to “micromanage” their children’s sexuality, “with the danger of STIs [Sexually Transmitted Infections] being a large part of that.”

At the same time, “parents are excited about kids getting their driver's licenses, and do not regularly forbid their child from driving … they know there are risks but assume the kids must learn to manage those risks.”

She thinks this approach should be applied to sex as well.

Of course, there could also be a moralistic aspect here—a kind of hangover from America’s Puritan founding. I raised this possibility with Shaun Miller, a philosopher at Marquette University who focuses on love and sexuality. “I'm not sure if it relates to our Puritan values,” he told me, “but I do think the stigma is a proxy for moral judgment. Sexuality has always had to do with one's moral character, and so if one has an STI, it suggests that one's character is ‘infected’ as well.”

To test this idea that sex-related risks are more stigmatized than other types of risk, Conley and her colleagues ran a follow-up study. In the study, they wanted to control for some of the differences between driving cars and having sex—two activities that both carry risk, sure, but which are different in other ways.

If these differences could somehow explain the weird estimates that participants gave in the first study—without having anything to do with sex-related stigma, specifically—it would undermine Conley’s theory.

Conley and her team designed a test that would compare “apples to apples”—two cases where a health threat was transmitted through sex, but only one of which was an actual STI.

They gave a collection of 12 vignettes to a large number of participants—one vignette per person. All of the vignettes told the same basic story: Someone transmits a disease to someone else during a casual sexual encounter, without knowing that they had something to transmit. There were two diseases: either chlamydia, a common STI that rarely causes serious health problems (and that can be completely cured with a course of antibiotics), or H1N1—commonly known as the swine flu—which can be seriously bad for your health or even kill you.

The main thing they manipulated between the different vignettes was the severity of the outcome caused by the disease. A “mild” outcome was described as getting sick enough to have to see the doctor, and then take a week’s worth of medicine. A “moderate” outcome was the same, except that you had to go to the emergency room first. A “serious” outcome was getting hospitalized and nearly dying. And a “fatal” outcome was, well, dying.

The last two conditions only applied to H1N1, because chlamydia rarely gets that bad.

Once the participants read their vignette, they had to say what they thought about the person who transmitted the disease. The participants would rate the person on how risky and how selfish their behavior was, as well as how dirty, bad, and immoral, and dumb they were for doing what they did.

The results were surprising. Participants who read the story about someone unknowingly transmitting chlamydia—with a “mild” outcome—judged that person more harshly than participants who read about the swine-flu case where the other person actually died!

Even Conley didn’t expect to see this. “Why would there be so much culpability surrounding a ‘sex disease’ but not a non-sexual disease transmitted through sex?” she said.

It’s a good question. Unjustified stigma about STIs—Conley’s preferred explanation—could be one answer. But there’s another possible answer as well, and it’s one that points to a potential weakness in the methodology of this second study.

There’s an important difference between chlamydia and swine flu in terms of how you can prevent them from being transmitted, and it has to do with condoms. Using a condom will dramatically reduce your chances of transmitting an STI like chlamydia, but it would have no effect on transmitting the swine flu. This is because swine flu isn’t passed on through genital contact, but rather through the respiratory system (so you could get it through kissing, or coughing).

So participants who were given the “chlamydia” vignette might have reasoned something like this. “If the person in this story had made sure that condoms were being used—which is the responsible thing to do in a casual sexual encounter—then the STI would very likely not have been transmitted. But it was transmitted. So the person was probably not using condoms. I’m going to rate this person harshly now, because I disapprove of this irresponsible behavior.”

Similarly, as the philosopher and cognitive scientist Jonathan LaTourelle of Arizona State University pointed out to me, “people might think that if you have chlamydia there is at least some probability you have it because of some prior sexual behavior that they disapprove of as well.”

In the swine-flu case, the same kind of judgment just couldn’t apply. That’s because even if safe-sex strategies were being employed, the virus would transmit exactly the same.

To their credit, Conley and her colleagues acknowledged this limitation in their paper, earning praise from other researchers I talked to. But limitations aside, Conley’s team thinks their study has important implications for public health. The main one, in their view, is that the stigma surrounding STIs needs to be drastically reduced. Otherwise, they fear, it could backfire, leading to more STI-transmission, not less.

“The basic research on stigma is quite clear on one issue,” Conley and her colleagues write in the paper. “Stigmatizing behaviors does not prevent unhealthy activities from occurring. For example, the more individuals experience stigma associated with their weight, the less likely they are to lose weight.”

So, they conclude, “we have every reason to suspect that stigmatizing STIs will [likewise] be associated with poorer sexual-health outcomes.”

They give two examples to illustrate this risk. One: If someone thinks they might have an STI but worries that their doctor will stigmatize them, they might be less likely to seek medical treatment. And two: If someone thinks their potential sexual partner will judge them for having an STI, then they’ll be less likely to bring it up.

But it might not be that simple. Stigmatizing some behaviors (like overeating) doesn’t seem to reduce them, but what about other behaviors—like smoking? There is some evidence, though it is contested, that increasing stigma around smoking actually has been pretty effective in reducing the number of smokers over time. When it comes to stigmatization, then, the question is whether risky sex is more like smoking, or more like overeating.

As the scientific cliché has it, “more research is needed.”