Barbara Devaney
Senior Fellow 
Mathematica Policy Research 
101 Morgan Lane 
Plainsboro, NJ  08536 

Does abstinence work? Well, of course, abstinence works.  Teenagers who are abstinent hardly ever acquire a sexually transmitted disease, they never get pregnant, they never drop out of school because of pregnancy, and they will not be a single welfare mom while a teenager. 

But the question that I want to address today is not whether abstinence works, but whether abstinence programs work.  And that question is a lot more difficult to answer.  In fact, I argue that we don't have an answer to that question. 

Let's take a closer look at the situation.  With age at menarche currently around 12, most teenagers have a long time before marriage in which to become sexually active.  Some teenagers will stay abstinent during this time, for a variety of personal reasons, no matter what kind of program or intervention is aimed at them.  Others will become sexually active, no matter what kind of program or intervention is aimed at them.  But those in the middle of the two extremes--the vast majority of teenagers--are influenced by many, many different factors.  The programs and interventions they receive can have an influence on them. 

Over the past twenty years or more, a variety of programs have been designed and implemented to influence teenagers' decisions about their sexual activity.  Our focus here is abstinence programs, which promote a strong message of sexual abstinence before marriage.  In the abstinence category, some programs promote abstinence and provide information on contraception.  (These programs are called Abstinence-plus.)  Other programs promote only abstinence and do not discuss contraception, unless it is to emphasize high rates of contraceptive failure.  These abstinence-only programs are clearly of the most policy interest today. 

Unfortunately, the development of strong and promising abstinence-based programs have not been matched by good, solid research on program effectiveness. We don't know whether these programs succeed in reducing teenage sexual activity.  Why don't we know?  To answer this, let's first consider what we need to know to decide whether a program is effective.  At a minimum, I think we need answers to the following two basic questions: 

• If you provide an abstinence program to a group of teenagers or pre-teens, how many of them remain abstinent during their teenage years?

• How does this number compare with the number who would have been abstinent without the program?

• Teenagers move or leave the program.  Some leave because they become sexually active.  It's hard to find the original group year after year, so often you are left with a very select group of teens who do not move or leave the program.  They may be precisely the group mentioned earlier--the ones who would be likely to be abstinent anyway.  They don't tell us anything about the teens who left.  To find out about the ones who left, we need to follow them over time and that can be difficult and very expensive.

• Researchers try to compare teenagers volunteering for a program with teenagers who do not volunteer.  Again, the volunteers may be the group of teenagers who are likely to remain abstinent anyway, and those who do not volunteer may be the ones who are likely to have sex anyway.  Neither group tells us anything about what kind of influence the abstinence program has on the vast majority of students who are in between these two extremes.

• Teenagers who have had sex, especially those who are in a program that delivers a strong abstinence message, are reluctant to admit it.

• Teenagers hit their peak years of sexual activity starting at ages 15 and older.  Yet, most abstinence programs start at young ages, grades 6th, 7th, or 8th, to influence pre-teens before they become sexually active.  So you have to follow the pre-teens for many years to find out what happens to their sexual activity and pregnancy rates.  This is very expensive.  Most evaluations don't do it and most don't go on long enough to provide the information needed.

So what do we know about the effectiveness of abstinence programs?  We know that abstinence programs produce a short-term increase in participating teens' pro-abstinent attitudes and values.  Some programs lead to increases in parent-teen communication about abstinence;  others apparently do not.  We know that some abstinence-plus programs appear to reduce teenage sexual activity, at least when we look at what teens are doing after short follow-up periods.  But we do not know whether abstinence-only programs lead to reduced sexual activity.  And we do not know whether abstinence programs in general reduce teen pregnancy. 

This rather negative summary does not mean that abstinence programs are failures.  It does mean that evaluations of the programs have been too limited in design and scope to answer the two basic questions about program effectiveness.  In particular, the evaluations of abstinence-only programs have suffered from follow-up periods that were not long enough, from sample sizes that were too small, and from difficulties in determining what would have happened without the program in question. 

Despite the lack of conclusive evidence on the effects of abstinence programs specifically, several reviews have identified the characteristics of effective programs that attempt to prevent teenage risk-taking behavior.  These reviews suggest that the following traits characterize effective teenage risk prevention programs.  Effective prevention programs: 

• Focus on a small number of specific behavioral goals. 

• Are based on a sound theoretical framework.

• Include basic information on risks

• Reinforce clear and appropriate values

• Use small group discussion and role-playing to teach teenagers resistance skills

• Are sufficiently long and intense

• Target students at relatively young ages

Nearly all of the abstinence-only programs reviewed in the literature share many of these characteristics.  In particular, abstinence-only programs focus attention on a small number of behavioral goals--that is, reducing teenage sexual activity, include basic information on the risks of sexual activity, and target young, pre-teen students.  However, only a few were sufficiently long and intense, lasting at least three months and requiring more than one hour of time per week. 

However, several abstinence-only programs that meet most of the criteria, especially those that are sufficiently long and intensive, have currently not been fully evaluated.  (Best Friends meets more of the criteria for success than any other abstinence-only program).  At this point in time, most evaluations of abstinence-only programs have themselves been short in duration and they have been of programs that were of relatively short duration and intensity. 

Should we care about whether abstinence programs work?  Yes, absolutely.  Abstinence is the current policy emphasis in teenage pregnancy prevention.  A considerable amount of federal dollars are being used to support the abstinence message. It is imperative that we know whether these programs reduce sexual activity and teenage pregnancy to levels lower than they would be without such programs. 

Furthermore, abstinence programs are not implemented in a vacuum.  Abstinence programs are likely to replace existing or previous programs that may provide education about contraceptive use or even provide access to contraceptives.  If such programs reduce unprotected intercourse among sexually active teenagers, and if these programs are replaced by abstinence programs, we need to know what happens to teens who do not remain abstinent. 

Finally, a wide variety of abstinence programs are available.  Some are school-based, some community-based.  Some target just teenage girls, while others target both girls and boys.  Some start as early as 5th grade, while others target older teens.  Evaluations tell us not only whether abstinence in general is effective, but also which programs are most effective in which circumstances.  Knowing which programs work and in what context is therefore important both for our decisions on whether to continue specific programs and whether to replicate them elsewhere.

Return to Abstinence Education Agenda