Update: Yvain, who unlike me knows what he's talking about, takes a somewhat more nuanced position and I infest his combox.
This is basically a response to a comment that got so long I made it a post.
In my last post I recommended looking for similarities between both sides of ideological polarizations to ameliorate our ideological biases. As an example, I mentioned I don't think either side of sex ed works. In response to that Leo asked:
What do you mean by "neither kind of sex-ed (abstinence only or comprehensive) shows any signs of effectiveness"? I'd certainly expect education about the morning-after pill to reduce unwanted pregnancies, and about safe toys to reduce ER admissions.
I was thinking of what we normally mean by sex ed: There are lessons in school that will include some information and some behavior advice. Then reactionaries like me hope the kids have less sex, modernists hope they contracept more and both sides hope for less pregnancies and STDs.
On that the evidence is pretty clear: The lesson content doesn't matter, there are simply no statistically significant differences between kids getting the different options available. See here and here. In other words the sex ed doesn't work. I don't think that is in any way surprising. Nowadays nobody growing up in the western world is dependent on school for information on where babies come from, and it actually would be surprising if horny dumb kids heeded their teachers moralizing, particularly since we already know that drug scaring programs don't work either.
To be fair, both sides of the American sex ed wars can point to better effects for much more intense community based programs like mentoring or situationally adapted one-on-one counseling. Basically, if you give lots of individual attention to a kid from a broken home, that kid is likely to take your advice seriously. But that's not what we normally mean by sex ed, it's basically proof that social work works.
The two effects Leo proposes are actually variations I hadn't considered before, so I did a bit of a literature search.
On the morning-after pill there seems to be a scientific controversy whether preemptively dispensing it for possible future use with some counseling leads to more usage and whether it displaces other contraceptives. Different studies come down on different sides of both questions. My intuition says both effects probably are real but small and that seems consistent with the data, but other conclusions are also possible. Anyway the effects are small in usage and too small to be observable in actual pregnancies at realistic sample sizes. Also, just teaching about it without handing out the actual pill doesn't seem to work.
I didn't find anything on sex-toy education. Probably it would be prohibitively hard to get a sufficiently large sample on that question, because both of the possibly related attributes (receiving sex-toy education and presenting at an ER because of having used unsuitable tools on ones genitals) would be rare. I certainly didn't receive any sex-toy-ed, despite my German sex ed being essentially a long advertisement for (contraceptive) sex. Abstinence wasn't presented even as an option, much less a preferable one, and there were tips like preemptively masturbating to avoid erections at swimming lessons, so there clearly was no Christian bias like you might suspect in some parts of the United States. Still, safe toys didn't come up, probably because nobody thought them a relevant question. On the other side, accidental sexual self-mutilations occasionally make the news, but my understanding is that they are newsworthy because they are rare freak accidents. Also, my suspicion would be that teen-agers using dangerous stimulation tools do know that there are safer toys but just don't have them. A correlation between two rare attributes is very hard to measure, so I think there probably is no science on this. Anyway, if your supposed effect exists, it would have a very minor effect on public health. You certainly could get much more, ahem, bang for the buck by spending the lesson time on pro-vaccination propaganda.
So, to sum it up, you can always think of special cases, but on the whole neither kind of what we normally think of as sex ed has the desired effects.