I was in Catholic community center today for a sporting event when a brightly colored poster on a bulletin board caught my eye. The picture was of a parachutist falling gently toward the earth and the caption read “Would he do it if he knew his gear would fail to protect him from injury or death 1 out of 5 times? Why trust a condom to save your life?”

I have no idea where the people who produced the poster got their statistics on the effectiveness of condom use but I have to wonder if “failed to protect him” perhaps means that he prefers the freedom of jumping without a chute at all or waits until he is five feet from the ground before pulling the cord. Or maybe he decides to lather the chute up with oil and that causes it to deteriorate during his decent. Maybe his pack is too big and slips off his shoulders as he pulls away from the plane. I have no idea what the author intended by the caption. I just know that if he is using a condom properly it isn’t going to fail to protect him one out of five times – it is improper use that causes failure. 

Now if he jumped with just a condom on, that would be a bad idea.

The methodology for measuring the effectiveness of condom use is a challenge given how difficult it is to control for behavior that cannot be directly observed. In several studies where one partner has AIDs and condoms are used consistently they have been shown to be 100% effective in protecting the uninfected partner. Improper, or inconsistent, usage though, creates to the possibility of transmission of disease.

The problem is then that no matter how effective a protective method is when used properly, if people are careless then sex becomes risky even if their intention is to practice safe sex. 

There is a psychological barrier that is crossed the first time a person has sex. Given this, crossing that barrier is akin to paying a fixed cost that once it is paid does not have to be paid again. If adolescents avoid having sex initially to avoid paying this fixed cost then once that line has been crossed it is more difficult to argue for abstinence.  And we observe this in the data; once a person starts having sex it is much more likely that they will have sex in the future than it is for an individual who has never had sex at all. 

So if schools and health offices make contraceptives available and that availability lowers the expected cost of crossing the line between virginity and sexual activity then the availability of contraceptives could increase sexuality in teens in both the short-run and in the long-run. Distributing condoms to teens may therefore lower teen pregnancy rates in the short run, since there will always be teens that are having sex regardless, but could also increase pregnancy rates in the long run given that condom use is sloppy.

Using the National Longitudinal Survey of Youth that collected information from thousands of American Youth in 1997, a new working paper conducts policy simulations that test this theory that contraceptive availability will increases pregnancy rates in the long-run. * 

And they find that it does. In the simulation a hypothetical reduction in contraceptive availability to adolescents, age 14, results in an almost 8% increase in the rate of unprotected sex in the year the policy comes into effect relative to when contraceptives were available. Three years later, when the hypothetical teens are 16, the increase in unprotected sex is less than 4%. 

The same reduction in contraceptive availability also increases the rate of no sex at all with a 3% increase immediately following the policy and an over 5% increase three years later.

Despite the increase in unprotected sex, and as a result of the reduction in sexual activity, pregnancy rates in the simulation drop in all three years of the policy. 

So this sounds like a policy recommendation to reduce the availability of contraceptive to teens – and that is exactly what the paper I am reviewing advocates. But do I think that we should pull contraceptives from the schools?

No, I don’t. The problem is this: Since when does three years count as “the long-run”, especially when we are talking about human sexuality? The teens in this policy simulation are extremely young and therefore are bound to be the most elastic when it comes to decisions about sexuality, i.e., they are the most responsive to small changes in the price. This policy simulation may say something about their behavior but says nothing about how older teens will respond to the lack of contraceptives in their early years – regardless of whether or not they are sexually active.

I would personally like to see this same policy simulation conducted not over three years but over six, or more. My guess is that what you will observe is that increased access to contraceptives early in their teens increases both a willingness to use them, and improved skills at doing so, later on.   

My father used to say, if you are going to give a person a gun you can expect them to shoot it (and he was talking about contraceptives, just in case you were wondering, and yes he did say this when I was a teenager). He was a smart man, but I think on this point we will have to disagree. 

* Arcidiacono , Peter, Ahmed Khwaja and Lijing Ouyang (2011). “Habit Persistence and Teen Sex: Could Increased Access to Contraception have Unintended Consequences for Teen Pregnancies?” Working Paper.