The logic seems so simple: more condoms, less disease, fewer teen pregnancies. That was the rationale behind a major policy push across the country in the early 1990s: Facing the then-fatal threat of HIV/AIDS, districts from Colorado to California to New York introduced condom-distribution programs in schools. According to a study published this month, over the course of just a few years, 22 districts in 12 states implemented this kind of program, affecting roughly 484 schools. In about two-thirds of the schools, kids had to go through mandatory counseling to get the condoms.

Condom access didn’t decrease the rate of teen births, though. It seems, in fact, to have increased it. The researchers estimate that these programs were responsible for roughly two additional births per 1,000 teens. And this rate was significantly higher when students could get condoms without any counseling. The effects would have been even greater if condoms were freely available to the entire U.S. high-school-age population, the researchers suggest—in that case, they projected, there would have been an additional five births per 1,000 teens.

The findings seem to pose a challenge to groups that advocate these kinds of school-based contraception-distribution programs. But they’re also troublesome for those who oppose sex education in schools. Easy access to condoms created worse outcomes for students, not better. But it seems that lack of access to information may have hurt them even more.

The debate over contraception distribution in schools often focuses on whether districts should only teach abstinence. In the schools included in this study, “counseling often included the message  that abstinence is the safest method of protection against STIs,” wrote the researchers, Kasey Buckles and Daniel Hungerman of Notre Dame, in an email. But the programs weren’t necessarily geared toward abstinence—“in some cases, a counselor might be required to show how a condom is put on, and in other cases a counselor might be forbidden from doing that.” Opponents worry that this kind of show-and-tell, paired with easy access to condoms, encourages kids to have sex or engage in risky sexual behavior. The researchers speculated that the ’90s-era condom programs might have done exactly that.

But a lot has changed over the last couple of decades, suggesting that alternative policies might have different outcomes. Since 1990, teen birth rates have fallen dramatically, declining from roughly 62 births per 1,000 girls aged 15 to 19 to roughly 24 among the same group in 2014. Teen abortions have also decreased significantly: In 2011, roughly 14 out of every 1,000 adolescent girls had an abortion, compared to 44 out of 1,000 in 1988. Teens are also waiting longer to have sex, and when they do, they’re doing it more safely.

The way young women use birth control is also changing. According to the CDC, condom use decreased by about half among women aged 15 to 19 between 1995 and 2010, while reliance on hormonal birth control has gone up, especially among women under 30. While women in their late 20s and 30s are more likely than teens to get IUDs, the greatest decrease in condom use as the only method of birth control was among teens.

These trends have likely been driven by a number of factors, including new public-health policies. Recently, a number of programs have shifted their emphasis away from condoms and toward hormonal birth control. While condoms are still the most common kind of contraception used by teens, they have higher failure rates for preventing pregnancy than hormonal birth control. As the researchers in this study point out, girls also have less control over condom use than they do over something like the pill; they can make a long-term decision to get on birth control, rather than deciding whether or not to use a condom when they’re about to have sex.

That’s why some states and localities are pushing for wider distribution of other kinds of birth control, including long-acting reversible contraceptives, or LARCs. An initial study of one such program—Colorado’s Family Planning Initiative, founded in 2009—suggests it has reduced teen birth rates in the state by about 5 percent.

Then again, as the researchers point out, some public-health organizations, including the American Academy of Pediatrics, have recently advocated condom-distribution programs in schools. Because of this, “many school districts (including Boston and Chicago) have recently considered the role of condoms in schools,” Buckles and Hungerman wrote in an email. In some areas of the country, alternative programs, including those that provide students with free IUDs, have had trouble catching on. This is especially true in conservative states like Texas, as my colleague Olga Khazan reported earlier this month.

For all of the positive trends involving teen sex, pregnancy, and contraception use in the United States, studies suggest that fewer teens are getting formal sex education in school now than in the past. While the Notre Dame study is a reminder that it’s difficult to predict the outcomes of public-health policies, it’s also a warning: lack of education, along with the specifics of how health policies are constructed, can be an important factor in teen birth rates.

“Our work suggests that if you get this type of intervention wrong, there could be unintended consequences,” Buckles and Hungerman wrote in an email. “That is a message that will probably never lose relevance!”