Birth control—or our skill at using it—is getting better.

A new analysis from the Guttmacher Institute found significant improvements in the efficacy rates of all reversible contraceptive methods. Considering that 45 percent of all pregnancies in the United States are unintended, this is great news for those who don’t want to conceive. Unfortunately, it also comes at a time when the Trump administration and congressional Republicans are working to roll back progress made on both contraceptive access and sex education.

Guttmacher researchers analyzed 2006-2010 data from the National Survey of Family Growth (NSFG), a national representative survey of individuals 15-to-44 conducted by the Centers for Disease Control and Prevention (CDC) on a regular basis. People who participate in the survey keep a monthly contraceptive calendar, in which they detail their contraceptive use and pregnancies for four years. The researchers also used data from the Guttmacher Institute’s Abortion Patient Survey in order to better capture contraceptive failures that ended in abortion.

Using this data, the researchers calculated rates that represent the probability of failure for each method during the first 12 months of use. The birth control pill now has a failure rate of 7 percent (down from 8 percent in 2002), and the male condom has a failure rate of 13 percent (down the most from 17 percent in 2002). Withdrawal is the only method with a failure rate that had not changed: Its failure rate remains at 20 percent. The researchers did not give a change rate for long-acting reversible contraceptive (LARC) methods, such as intrauterine devices (IUDs) and implants, which remain the most effective with a failure rate of 1 percent.

The likelihood of failure for all methods declined from 12.4 percent in 2002 to 10.3 percent between 2006 and 2010.

In order to truly understand these numbers, it is important to remember that all contraceptive methods actually have two failure rates. The first is the “perfect use rate,” also known as the “method failure rate.” Calculated during clinical trials of methods, this rate shows how well a method of contraceptive can work if it is being used consistently and correctly.

The rates calculated for this study, by contrast, are “typical use rates” or “user failure rates,” which examine how well a method works under real-life conditions. These rates include those who are using the method incorrectly or inconsistently, and even include “outright nonuse among individuals who report using.”

This is an important distinction when deciding how well a contraceptive method works. If, in your NSFG contraceptive calendar, you report being a birth control pill user during March even though in truth you forgot to pick up your pills until the last week of the month, you are counted as a pill user. And, if you get pregnant in March because you missed so many pills, that is counted as a failure of the pill when the “typical use rate” is calculated.

It doesn’t seem fair to blame a pill that wasn’t taken any more than it’s fair to blame a condom left in a drawer, but it does help people understand how much effort each method might require of them. LARC methods have the lowest failure rate in large part because the user doesn’t have to do anything to be successful. Once the method is put in place, it will work for three to ten years with no effort from them. So, if a method has a low perfect-use failure rate and a high typical-use failure rate, it’s reasonable to assume it takes more commitment or effort on the part of the user to be successful.

As a sex educator, I think it’s exciting that user failure rates have dropped, though I often prefer to look at it in the reverse so as to concentrate on success: In other words, a method that has a 7 percent failure rate is 93 percent effective. Though some clinical trials may have been done on the perfect-use efficacy of contraceptive methods, this particular study of typical-use rate shows that we users are doing a better job. The fact that condom efficacy has improved is particularly heartening because it remains the only contraceptive method that also prevents sexually transmitted infections.

The current study can’t tell us why or how we got better at using birth control, but there are a number of plausible explanations. It’s possible that access improved; if it is easier to obtain and afford a method, we can have it on hand without missing a day or a sex act. It’s also possible that between sex education and widespread internet access, more people are getting the information they need about these methods and how to use them correctly. Finally, it’s possible that people are more motivated to avoid pregnancy and therefore to use contraception consistently and correctly.

Any one of these would be a positive development and it seems most likely that all three factors played a role. Certainly, there is no one government policy that could be pointed to as the origin of improved contraceptive use, especially because the data used in this study—which dates back as far as 2003—span both the George W. Bush and Barack Obama administrations, which had very different views on contraception.

Still, I fear that the policies the Trump administration will unveil in the months and years to come may reverse some of the progress we’ve made.

Together Trump and congressional Republicans seem poised to get rid of the Affordable Care Act, and, in particular, the provision that requires insurance companies provide some birth control at no cost to consumers. As Rewire Vice President of Law and the Courts Jessica Mason Pieklo pointed out in a recent Rewire article, U.S. Department of Health and Human Services Secretary Tom Price doesn’t believe that the provision is necessary, and “Trump’s cabinet is packed tip to tail with contraception foes opposed to the ACA generally and health insurance coverage for birth control specifically.”

This is distressing, especially in light of initial evidence that the birth control provision is helping women access and use contraception. Other than withdrawal and condoms, most birth control methods require prescriptions and either a monthly cost (such as for the pill) or a high up-front cost (such as for an IUD). If women lose access to insurance, it will become harder for them to afford the initial doctor visit as well the methods themselves.

As for information about contraception, Rewire’s analysis of sex education under Trump suggests we may lose ground there as well. We may possibly revert to government-funded abstinence-only-until-marriage programs, which tend to discuss contraception only in terms of failure rates. Unlike comprehensive sex education programs, which have been shown to increase the likelihood young people use contraception when they do become sexually active, abstinence-only programs have never been proven to change young people’s sexual risky behavior.

In addition, those of us who were working in sex education during George W. Bush’s years in the White House had a serious case of déjà vu when the U.S. Environmental Protection Agency (EPA) was asked to remove climate change documents from its website. In 2003, the CDC was pressured to take down its fact sheet that promoted condoms. The fact sheet that replaced it was longer, more confusing, and included new language promoting abstinence and monogamy. It’s not hard to imagine efforts to purge pro-contraception information from the CDC or other government websites in the future. Without this information, people may be less aware of their contraceptive options and where to find them.

As for motivation to avoid pregnancy, many factors likely play a role. One’s own outlook for the future probably has a lot do with it, so it is very unclear whether this will change in either direction given the current political climate.

While we wait and see what the Trump administration will do with the ACA, birth control, and sex education, we should remember that whether contraceptives work is largely up to each of us (or at least those of us who are in a relationship in which unintended pregnancy could become an issue). The methods have been invented. They work exceptionally well at preventing unintended pregnancy, provided that they’re available to us and we use them consistently and correctly. Now it’s up to us to use them.

Congratulations to us all for these improvements. Let’s keep up the good work and the pressure on elected officials so they don’t get in the way of our recent successes.