by Stan E. Weed
For the first time in recent decades, national statistics have provided some encouraging news about teen pregnancy rates. They are dropping! It is not a huge drop, and we cannot yet say whether these lower rates will be sustained (or better yet, whether the decline will continue). But at this point, the data clearly demonstrate a reduction in teen pregnancy rates.
Furthermore, this reduction in pregnancy is reflected in a corresponding drop in teen birthrates and in teen abortion rates. The federal, state and private agencies which have worked at this issue for some time are of course delighted – as they should be. Teen pregnancy affects lives in significantly negative ways.
What is not so encouraging is the rush to ascribe or claim credit for these results based on some evidence that is available. Such prominent agencies as the U.S. Department of Health and Human Services, the Office of Population Affairs at HHS, the National Adolescent Reproductive Health Partnership, the Centers for Disease Control and the Planned Parenthood Federation of America (PPF) have all interpreted the national data to mean that contraceptive policies and practices account for the drop in teen pregnancies. The following quotes illustrate the popular and prevailing explanation for the recent trends:
"The positive change in teens sexual behavior is a result of ... increased condom usage."
Advocates for Youth press release, September 18, 1998
"(HHS) Secretary Shalala said the dramatic increase in contraceptive use at first intercourse ... may be responsible for the leveling off and recent decline of teenage birth rate."
HHS press release, May 1, 1997
"The dramatic decline in unplanned pregnancy has occurred to a large extent as a result of higher contraceptive prevalence and use of more effective methods."
Planned Parenthood press release, January 17, 1998
"While the proportion of adolescent females who have experienced sexual intercourse has increased over time, their likelihood of pregnancy has decreased. Increases in contraceptive use by adolescent females contribute to this change."
Office of Population Affairs, cited on HHS web site, November 9, 1998
Some agencies attribute the drop in teen pregnancy to both an increase in contraceptive use and a decrease in sexual activity rates among teens. All of them, however, want to credit the use of contraceptives for at least part of the decline. What is the evidence that leads so many groups to the same conclusion? Is the data really that clear, that compelling? If it is clear, then how do we explain this sudden success after years of failure?
What is the Basis for the Claims Made by PPF and Others?
The sources listed above essentially claimed that the drop in teen pregnancy rates is associated with a corresponding increase in contraceptive use – more specifically, an increase in condom use. Setting aside for a moment the problem of correlation versus causation, let's look more closely at the relationship between contraceptive use and teen pregnancy rates. Figure 2 illustrates the data pertaining to condom use at last intercourse for males and females. It does in fact indicate an increase. That change, however, does not tell the full story. In any valid research endeavor, one always looks for alternative explanations beyond the first impression. Figure 3 illustrates that the use of oral contraceptives has declined during that same time period. This trend of declining use of the more effective oral contraceptives more than offsets the increase in condom use. The net effect in terms of contraceptive use (both methods combined) is minus 12.2%. Even adding in other forms of birth control, such as Norplant or Depo-Provera, does not change this pattern. How then can one attribute decreased pregnancies to increased contraceptive use when in fact the total contraceptive use has declined? If there is no correlation, as illustrated here, between contraceptive use and teen pregnancy, there will certainly be no causal relation.
A more thorough analysis of the data reveals another striking element. What appears to be a decline in teen pregnancy rates as shown in Figure 1 is somewhat misleading because it combines the married and unmarried teens. And for the unmarried teens, it also combines the sexually active teens with the not sexually active teens. These combinations confound the data to the point that they become uninterpretable, let alone amenable to causal claims such as
"The dramatic decline in unplanned pregnancy has occurred to a large extent as a result of higher contraceptive prevalence and use of more effective methods"
(Planned Parenthood press release, January 17, 1998)
By separating the married from the unmarried, and the sexually active unmarried from the abstinent unmarried, we are able to obtain a more plausible explanation based on the current evidence. A more complete explanation of this examination is available in the recent publication by the Consortium of State Physicians Resource Councils entitled "The Declines in Adolescent Pregnancy, Birth and Abortion Rates in the 1990s. What Factors are Responsible" (Jones, Toffler, Bell, et al. 1999). In summary, the estimated non-marital birthrate per 1,000 sexually active females 15 to 19 has gone from 85.2 to 111.8 (between 1988 and 1995) – an increase of 31.2%. And this increase occurred during the highly acclaimed increase in condom use (and commensurate but less touted decrease in pill use). So, for those unmarried teens who are sexually active, birth rates are increasing rather than decreasing! This is not the chorus we are hearing sung by various federal, state and media agencies.
A Better Explanation
How then do we explain the drop in overall teen pregnancy rates, if the change in contraceptive use fails to account for that drop? The other trend in national statistics which has caught the attention of even the popular media is the change in sexual activity rates. For the first time in recent decades, the trend of increasing numbers of teens engaging in premarital sex has reversed. Data from similar time periods (1988 to 1995) from the National Survey of Family Growth, the National Survey of Adolescent Males and the Youth Risk Behavior Survey all report a clear and consistent trend. More adolescents are abstaining from sexual activity in recent years, and there are now more teens (15-19) who are not having sex as compared with those who are. Figure 4 illustrates this trend. Figure 4 data comes from the National Survey of Family Growth (for females) and the National Survey of Adolescent Males.
The data from the Youth Risk Behavior Survey also shows an overall decline in sexual activity rates, but that data indicates more change is due to a decrease for males than for females. In any event, there is a decline in the percentage of 15-19 year old adolescents who have ever had sex. We also see an overall decline in those who have had sex in the past three months, and in the number of partners they have ever had. What once appeared to be an inevitable increase in sexual activity rates each year has now changed direction toward more abstinent behavior. If we are looking for a plausible explanation to account for decreasing teen pregnancy rates, the change in sexual activity is certainly more viable than the argument being made that the drop in teen pregnancies is a function of increased condom use.
The shift towards abstinent behavior is likely the result of multiple factors and forces operating simultaneously, including awareness and concern about AIDS and other STDs. Not to be ignored in this shift, however, is the large increase in the number of teens exposed each year to programs that promote abstinence as their central message. These programs have multiplied dramatically and account for a twelve-fold increase since 1986 in the number of teens exposed to a clear and direct message each year about sexual abstinence. I am not aware of any other factor that might account for the shift towards abstinent behavior. Certainly the media has become more, rather than less, prevalent and provocative with its messages regarding premarital and extra marital sex.
So where did the commentators go wrong in their logic and in their interpretation of the data? It seems that four flaws account for their leap of logic, any one of which would be fatal.
They made the typical mistake of combining married and unmarried teens when referring to teen pregnancy.
They made the mistake of combining sexually active, virginal and not-sexually-active teens.
They failed to distinguish between condom use, oral contraceptives and other methods.
They drew a causal link based on correlational data, and the correlational data was flawed to begin with – as per a, b, and c above.
It would also appear that the popular mantra of the policy makers about needing to provide more condoms to kids because "we all know they are going to be sexually active" is directly challenged by the national data. Given the trends we see in multiple sources of national data, sexual activity is neither inevitable nor irreversible. Second, the sexually active unmarried adolescents who are supposedly using more condoms, more regularly, more competently, are not showing a decline in pregnancy rates. Third, the effort to make condoms more readily available through distribution programs (based on the simple assumption that what kids need to become more regular condom users is greater, non-embarrassing access) has recently been tested in the Seattle Public schools. This social experiment demonstrated the fallacy of that approach. The results indicated that making condoms available to students did not increase condom use. Among students who had engaged in sex during the preceding 3 months, the percentage who used a condom the last time they had sex actually declined from 57% to 51% among the Seattle students, and the decrease was much greater among students in schools that had clinics (and distributed many more condoms) than among students in schools without clinics.
Abstinence as a Viable Strategy?
The number of abstinence-centered sexuality education programs has increased dramatically in recent years. The recent welfare reform act (Title V) has generated new interest in abstinence education and made new program funds available through block grants to states. Many existing programs have evolved, become more sophisticated, more theory driven, more intensive and of longer duration. And as pointed out earlier, more and more teens are involved as program participants each year. While this expansion and evolution of abstinence programs is occurring, efforts to measure impact and effectiveness have lagged behind. Very little funding has been available for extensive evaluation efforts, and those programs which are evaluated often change components or elements in order to respond to what has been learned in the evaluation. This results in a moving target for evaluation efforts. All of this puts the field of abstinence education in a state of flux at present. That is not a criticism, but merely a comment on the current condition.
In spite of this recent expansion of programs, the evaluation funding limitations and the moving target of evolving programs, there are some things we know about the characteristics of effective programs. Our work in the field over the past ten years has led to several conclusions — most of which are empirically based, while others are probably in the "educated guess" category. First, the old paradigm of "sex education" was based on the premise that young people lack information (physiological, biological, reproductive). It was believed that increasing the quantity and quality of that information will lead to better sexual decision making and fewer negative consequences. A multitude of studies have now lain to rest (or put a stake in the heart of) that old assumption. Information, per se, is not a central or potent factor in driving behavior. Surprisingly, there are still programs that cling to the information–decision making paradigm. Application of this premise will not have a significant or meaningful effect on sexual behavior. Doing more of what does not work, and doing it at younger ages, will not reduce the pregnancy rates of teens.
Understanding the mediating factors that do influence sexual behavior has been a central feature of our own work. Based on data collected over several years from 35,000 teens in 23 different studies and samples, we have identified several factors that explain or predict sexual risk taking behavior. This has obvious implications for program development or selection, because it identifies characteristics of effective programs. Our research demonstrates that there are a number of factors operating simultaneously which account for sexual behavior and its consequences. For example, sexual values, peer sexual environment, future orientation and related risk behaviors are all factors that contribute directly and significantly to the behavior of interest. It is also evident that some factors are more directly linked to the targeted behavior, and have a much stronger connection than do other factors. Sexual values – the teen's conviction about what is good and bad, right and wrong, important and not important – has been the most consistent and potent predictor of all the variables we have measured. Sexual behavior is also influenced by the extent to which the students described their immediate peer environment as sexually permissive, or felt he or she would experience opportunities or face pressures from a partner to have sexual intercourse. Programs that take these multiple factors into account in an effective way are likely to have a significant influence on teen sexual behavior and all of its consequences.
Other program features are also emerging as important contributors to significant outcomes. Those which engage the parents as partners, and strengthen the degree of "connection" between parents and youth will have greater success at transmitting those important values mentioned above. Programs which move beyond a "quick fix" mentality and offer intensive exposure over an extended period of time, with "booster" shots to reinforce the experience will be more effective than the one time or short course exposures. And those programs which give a direct and clear message about abstinent behavior do better than those which waffle on the message.
The important news in all of this is that the move towards abstinence is growing, it appears to be having an effect on adolescents as reflected in the national data, and we are learning more about the nature and dynamics of sexual risk taking behavior which will lead to more effective interventions. Moving forward and building on this success will sustain the positive trends in teen pregnancy reduction, and lower birth and abortion rates – certainly a common hope across a broad spectrum of ideological positions. Those who have long embraced abstinence on moral and religious grounds can find comfort in this recent data that indicates it is possible to reach adolescents with a positive message of hope, respect, future and self control. Even for those who approach this problem from simply a health perspective, the news is encouraging and the program implications are significant. The self-indulgent and immediate gratification messages so prominent in the media do not have to be the dominant and overpowering lessons of life. There is a better message, and teens are listening.
Stan Weed, Ph.D. (in Social Psychology) is a widely published researcher and former professor. He has been a partner/co-founder of the Institute for Research and Evaluation (Salt Lake City) since 1985, focusing primarily on teen pregnancy, AIDS prevention and character education curricula.