Contraceptive Access and Declining Teen Birth Rates: A Connection Still Unproven

Recent headlines have been celebrating the findings of a CDC study, which shows that birth control use among sexually active teens is becoming more common. Some reporters have made a connection between this study and other reports indicating a lower rate of teen births, even reporting that the birth rate is lower because of increased contraceptive use. But do any studies actually show that birth control is the reason for falling teen pregnancy rates?

While it may, at first, seem logical that birth control access is behind this decline in births to teen mothers, the reality isn’t that simple. It appears that many factors unrelated to contraception have contributed to the decline in teen birth rates, as other recent studies indicate. Journalists misuse science when they suggest that the extant research settles this complex debate.

Teen Birth, Poverty, and Birth Control

While teen births in the United States are declining overall, they are not decreasing to the same degree everywhere. Regional differences in poverty and in education have long been known to be closely tied to variations in teen birth rates. However, a recent study authored by Raid Amin and colleagues upset conventional wisdom with its finding that poverty and education cannot explain all variation in teen birth rates.

Some news articles have interpreted the study to indicate that increased access to contraception is needed, but the researchers did not look at contraceptive use, nor do their findings support the conclusion that birth control is the answer.

What did the study find? The study’s authors first identified the metropolitan areas with the highest teen birth rates between 2006 and 2012. In absolute numbers, the highest rate was seen in San Antonio. The second highest was in Denver. Since they expected poverty and educational level to be the factors explaining teen births, they then compared girls at the same income level and with the same education in each community. When adjusted for poverty, the top two areas were different cities in the same two states: Fort Worth, TX and Grand Junction, CO. When they adjusted for educational level (high school diploma), Texas disappeared completely from the top ten; Denver took over the number one spot, with Washington, DC, coming in second.

Amin and coauthors conclude, based on this information—and perhaps to no one’s surprise—that understanding why some teens get pregnant and give birth is more complicated than simply being poorer or less educated. In their discussion section, they then go on to speculate about what other factors may be involved.

Is Birth Control the Linchpin?

In a paragraph that was widely quoted by the media, Amin et al. do speculate that one factor affecting decline in teen pregnancy might be access to contraception. They state: “It is an important observation that recent observed declines in the teenage birth rate have been largely attributed to an increase in contraception use rather than a change in sexual practices of adolescents.”

It is unclear upon what information this speculation—the only part of the paper quoted by many media reports—is based. It is not the authors’ own data from their study. Rather, the authors reference two other studies promoting the link to contraception access, yet that reference is, again, only to the discussion sections, not the data, of those two papers. One of these papers studied the incidence of teen pregnancy, and states in its discussion section: “Our analysis did not address factors that might explain the decline between 2008 and 2011, but several possible factors should be considered.” This paper then goes on to discuss factors that might or might not contribute to the decline, among which are an increased desire for pregnancy related to an improving economy (meaning fewer pregnancies are classified as unintended), and increased use or efficacy of contraceptives, among other possible causes.

In other words, the speculation of Amin et al. is a reference to the speculation of other authors. Neither Amin et al. nor the studies they reference have data relating contraceptive use, sex education, or contraceptive availability to teen birth rates (or to teen pregnancy rates), nor are there data on teen sexual activity (which is “presumed to be” unchanged, but this was not studied).

The studies themselves, then, do not even purport to offer evidence that access to birth control decreases teen pregnancy rates, but that didn’t stop the media from drawing their own conclusions. When their study was released, media reports latched on to the speculative discussions and selectively highlighted some of the data. Several stories indicated that if the highest teen birth rate in the country is in a city in Texas, this must be because Texas does not allow teens to obtain birth control on their own (unless they are married).

Not So Fast

While the highest teen birth rate was in a city in Texas, the second highest, in absolute numbers, was in Denver. Colorado has easy access to birth control for all teens, offering contraception to minors without parental notification. Beginning in 2008, the state has also aggressively promoted long-acting reversible contraceptives (LARCs), even offering them free or at very low cost to teens. Attempts have been made to link this wide availability of contraception in Colorado to the reported decline in its statewide teen birth rate, but given nationwide falls in teen births, the fact that only a fraction of Colorado teens received LARCs, and declining rates of teen sex—among other factors—the data do not bear this connection out. The purported connection can’t explain why a city with easy access to birth control has the second highest teen birth rate, following a city with markedly less access.

The remainder of the “top ten” locations for high birth rates in Amin et al. were scattered throughout the country and do not seem to correlate with ready availability of birth control. Again, this is not to fault the research, for its object was not to establish such a connection; if it were, the authors would have collected the relevant data. It is rather to fault those who have hastily used the research to defend conclusions that it cannot support.

Needed Context: Abortion Rates and Intention

Ultimately, the data of Amin et al. are limited in their ability to tell the whole story about teen births. In the first place, the researchers looked at birth rates—not pregnancy rates. They did not collect any data on abortions. The authors quote a study showing that abortion rates in the setting of an unplanned pregnancy have not changed over time, even though unintended pregnancies have declined, but the referenced paper has national, not regional, data. This does not rule out the possibility that there is regional variation in abortion rates, which are probably affected by differences in culture and religious beliefs, which of course differ across the country.

We might reasonably suppose, for example, that the same cultural milieu that results in abstinence-only education in Texas might also mean a pregnant teen is less likely to choose abortion; Amin et al. provide no data to suggest otherwise. A higher birth rate in some metropolitan area may not mean there were more teen pregnancies; it might just mean that fewer pregnant teens chose abortion.

Another piece of missing data is the girls’ intentions: how many of them wanted to get pregnant? One study referenced by Amin et al. found that, of adolescent mothers studied, more than 20 percent of them planned to get pregnant at that time or sooner, and another 65 percent wanted to have a baby but had not planned to do so right away, leaving only 15 percent of the pregnancies unwanted. (These data came as no surprise to me. I remember, as a medical student working in a gynecology clinic, asking young, sexually active women if they wanted to have a baby. Many of them just smiled, and some admitted to me that yes, actually, they would like that.)

Interestingly, the recently released CDC study on contraceptive use reported similar findings. The CDC found that those teens who wanted to avoid pregnancy were more likely either to abstain from sex or to use contraceptives, whereas those teens who chose to have “unprotected” sex were less likely to report that they wanted to avoid pregnancy.

What we see in both these studies, then, implies that the news reports claiming that “better birth control access” is the solution to the teen birth rate have it backwards: the data indicate that many teens choose to not use contraceptives because they are not trying to avoid pregnancy.

What We Do Know: Not Much

So, what can we conclude about what causes higher teen birth rates? We know that poverty and educational level affect teen birth rates but do not provide a complete explanation of regional variation. We know that there are some areas of the country with more teen mothers, and that some of these areas have abstinence-only education and lower levels of access to contraception for minors, but others do not.

We do not have information about regional variation in teen pregnancy rates, abortion rates, or differences in desire to have children. As Amin and colleagues conclude, we know that there must be other factors, but we do not know what they are.

All of this confirms that the relationship between sex education, availability of contraception, and teen pregnancy rates is complex and controversial. One recent study from the UK actually found, to the investigators’ surprise, that expenditure cuts on teen pregnancy services actually reduced teen pregnancy rates. The authors found (lower) educational levels and (higher) alcohol consumption to be predictors of teen pregnancy, and, like Amin et al., concluded that more research is needed.

The declining teen birth rate calls for further study. To distill such a complicated issue into a flashy headline relating teen pregnancy to any individual factor might generate attention and publicity, but it’s an egregious oversimplification. Worse yet, to ignore the actual results of scientific research by twisting a single hypothetical statement from a paper’s discussion section into a political argument is to weaponize science and undermine its concern for the truth.