Last Friday, Texas Governor Rick Perry issued an executive order requiring that "[t]he Health and Human Services Executive Commissioner shall adopt rules that mandate the age appropriate vaccination of all female children for HPV prior to admission to the sixth grade," though allowing parents to refuse on conscientious grounds. (The relevant state statute, Health & Safety Code § 161.004(d), provides that "A child is exempt from an immunization required by this section if ... a parent, managing conservator, or guardian states that the immunization is being declined for reasons of conscience, including a religious belief.")

The governor's action has led to a good deal of objection, from both social conservative and libertarian circles. I thought I'd pass along a few tentative thoughts of my own. Because the aggregate of the thoughts is long, I'll break it up into a few posts; in this one I'll just set out the key facts:

HPV (Human Papillomavirus) apparently causes about 70% of cervical cancers; there are about 10,000 new cases of cervical cancer in the U.S. per year, and 3700 deaths, which means that HPV kills about 2500 American women each year. HPV is sexually transmitted; the vaccine, the CDC reports, has been tested on 9-to-26-year-olds and appears to be nearly 100% effective if administered before a woman acquires HPV. It has not, however, been tested on women over 26, and thus is not recommended for them.

I can't speak to how safe the vaccine is, though I have no reason to doubt the CDC's view that it is safe. If it is indeed too risky, that would of course be a plausible objection to it, but I'll set that aside for now. Rather, in coming posts, I'll discuss whether the vaccine is likely cost-effective; whether it might be counterproductive because it will lead to more risky sex and thus more diseases that stem from it; whether HPV is different enough from other diseases that government-mandated or even government-strongly-suggested HPV vaccination is improper; and how we might think about truly mandatory vaccination (which this doesn't seem to be) more generally.

Not everything that improves health, or even saves lives, is worth mandating or even strongly pressuring (even if your philosophy is social welfare maximization; I'll speak later about how those with a more libertarian bent might view this). Saving a few lives at the costs of billions of dollars may end up not being cost-effective, especially given that the expense could reduce people's ability to spend on much more cost-effective health improvements.

But my very rough back-of-the-envelope calculation suggests that the HPV vaccination is pretty cost-effective. The CDC reports that the retail price of the HPV vaccine is about $360; let's assume this ends up being roughly the cost, setting aside bulk discounts, extra labor costs for administering the vaccine and the like. About 2500 American women die of HPV each year, which means that the lifetime risk for a typical American woman is roughly 2500 x 75 / 150,000,000 = one in a thousand.

Given that most estimates of value of life saved I've seen run in the $5 million to $10 million range, avoiding a 1/1000 risk of premature death for $360 sounds pretty cost-effective. Of course a more full analysis would have to include many other factors: It should consider years of life lost rather than just deaths. It should consider the possibility that the people who die of cervical cancer are also the ones least likely to get the immunizations; cervical cancer can largely be avoided through regular pap smears, so it stands to reason that women who get it tend to be ones who are least likely to have good health insurance coverage, and they may be the ones who are most likely not to get immunized when they are girls. It should also consider, as a factor countervailing to this one, the possibility that herd immunity provided by very broad immunization will benefit even those who aren't immunized. And it should doubtless consider a bunch of other things, too.

Still, given that $360,000 is way under $5 to $10 million, it seems safe to say that include all the other factors will still yield a judgment that the HPV immunization is likely cost-effective.

Some argue that vaccinating girls for HPV would give them a false sense of security, and will thus lead to young teens' having more sex. I'm pretty skeptical about that.

Those girls who think about the risks of sex (and who knew in the first instance that sex can cause cervical cancer) will still realize that sex can spread AIDS — which kills roughly 2300-2500 American women per year, which causes over 5000 more per year to fall ill, and which strikes me as a much more publicly focused on risk than is HPV. They'll also recognize that sex can cause pregnancy; and those who blithely have sex without paying much attention to these remaining risks likely would have blithely ignored the risk of cervical cancer, too. "I wasn't going to have sex, because I was afraid of getting cervical cancer or AIDS, but now that I know that I'll only be at risk of AIDS, I'll go ahead and do it" strikes me as an implausible reaction.

In principle, I'm open to arguments that legal requirements that seem to increase safety will lead to riskier behavior that will eliminate much of the safety benefit. That's certainly a danger that sound policy analysis should take into account. But in this particular situation, it just doesn't seem very plausible.

I've heard some argue that HPV is different, and a less proper candidate for a government-mandated (or even government-strongly-pressured) immunization, because HPV is acquired through risky, and usually personally chosen, behavior. This, though, strikes me as mistaken.

Though having multiple sexual partners increases one's risk of getting HPV, all it takes is one sexual partner. Nearly every woman will have sex at some point in her life. Even if she is a virgin bride, she can get HPV from her husband on her wedding night. True, if she marries a virgin, and her husband never cheats on her, then she's not at risk (setting aside the possibility of rape). But even the most moral behavior on her part, under any definition of morality short of lifelong abstinence from sex (including marital sex), won't protect her. The vaccine, on the other hand, likely will protect her.

There's also a very different kind of risky behavior argument: I'm told that regular pap smears, and the medical procedures used when the pap smears show a dangerous result, are very reliable in preventing even HPV-infected women (about 50% of the population, I have read) from developing cervical cancer. In a sense, then, actually getting cervical cancer may be said to be the woman's "fault" not because of her sexual behavior but because of her medical laxness. (That sounds harsh, but I take it that we do say that in some measure, though not without sympathy, in other contexts: If someone dies of untreated pneumonia -- consider, for instance, Jim Henson -- we might think that this death was in some measure his fault.)

Yet that seems to me to be a not very good argument against immunization. Many women don't get pap smears because they're fairly poor. Even those that could easily get them but don't seem to deserve some protection. And even for those who get them and thus don't get cervical cancer, the treatment used to avoid death from cervical cancer is expensive, unpleasant, and emotionally distressing -- and it can lead to infertility.

Of course, all this then raises the broader libertarian objection: Why should some people be forced to be immunized (assuming the Texas law mandated immunization, which this one seems not to, given its broad exemption for parents with conscientious or religious objections) in order to protect others? I'll turn to that in the next post.

Given Texas's conscientious objector exemption, the vaccine seems not to be really mandatory. (I say "seems" because I'm not positive how the exemption, which is on its face very broad, is applied in practice.) But what if it were mandatory? Should we oppose that on broadly libertarian grounds? When I say "we," I'm referring not just to hard-line libertarians (of which I'm not one) but also to those who have general Millian "free to do what I please so long as I don't hurt others" sympathies, even if those sympathies can sometimes be trumped by other concerns.

My tentative sense is that immunizations against communicative diseases are often quite proper, even as a libertarian matter. I say "tentative" because I'm sure others have thought about the subject in much more depth than I have, and perhaps they can prove me mistaken. But let me quickly lay out my thinking.

It is a sad fact of biology that we can spread communicable diseases without any conscious decision on our parts, even without knowing that we are infected. Any time we do this, we are indirectly causing harm to someone else. Say Alan has sex with Betty, who then has sex with Carl, who then has sex with Denise; say Alan is infected with HPV, and each sexual act would (absent immunization) spread HPV; and say Betty isn't immunized against HPV. Betty's failure to get immunized would lead to her unwittingly spreading the virus, which ends up hurting Denise. She hasn't intentionally harmed Denise, but she has harmed her -- you might categorize the harm as negligent (in that it flows from negligent failure to get immunized) or not, but it is indeed the infliction of harm.

Now it's true that the harm also flowed from Denise's voluntary decision to have sex with Carl. But, as I noted in an earlier post, it's hard to see why this should excuse the harm caused by Betty, any more than Denise's voluntary decision to get on the road excuses the harm that someone imposes on Denise by crashing into her with a car (or, if you prefer, that Betty imposes on Denise by crashing into Carl's car, which then crashes into Denise's).

Even if you think that some people's having many sexual partners should affect the analysis, remember that HPV can be spread even among people who are about as sexually constrained as can be expected. The Alan-Betty-Carl-Denise connection can happen even if Betty was a virgin when she married Alan; if she then didn't have sex with Carl until she married him (assume Alan had died, or had left Betty); and if Denise was a virgin when she married Carl (again, assume Betty had died, or had left Carl). This very scenario might be rare -- but lots of other scenarios in which people had led fairly safe lives, but find themselves getting HPV, are also quite plausible. And more broadly, even if people are leading somewhat riskier lives than this, participating in spreading a disease to them may still be quite rightly seen as harming them, despite their own role in choosing risky behavior.

Of course, if HPV immunization were 100% reliable, and 100% available, then this analysis wouldn't apply with quite the same strength: Presumably any person who remains at risk of HPV infection would be at risk because of her own refusal to get the vaccine. Yet while the immunization is supposed to be extremely reliable for 9-to-26-year-olds, it hasn't been tested on over-26-year-olds, and thus isn't recommended for them. Moreover, some people won't get the vaccine, possibly because they can't afford it. ($360 isn't chopped liver for many, especially for people who aren't in America.) Even an "assumption of risk" presumptive libertarian may reasonably conclude, I think, that refusing to get immunized is wrongful behavior, because it may lead to one's becoming a vehicle for transmitting a dangerous and sometimes deadly disease to third parties, and thus harming those third parties (in a way that an "assumption of risk" argument would not excuse).

Finally, recall that the question here is whether to immunize girls who are under 18, girls who may well get infected before 18 if the immunization is delayed until then. Even if it was just their own health on the line, and not the health of others whom they might indirectly infect, we could rightly say that they don't have the maturity to refuse this protection, and that their parents shouldn't be entitled to refuse this protection on their behalf. But since the question of what kinds of modest physical risks parents should be free to have their children run is thorny, contested, and old hat enough that I'm not sure I can add much to the subject, I thought I'd focus primarily on how parents' refusal to immunize their daughters may hurt others, and not just the daughters themselves.

So a brief summary: There may well be practical problems with truly mandatory immunization, and it may well be that herd immunity would mean that 90% immunization is good enough to reduce the risk to a level that doesn't merit regulation. There may of course also be practical objections to immunization if the immunization seems unduly risky (a question I set aside in the first post in this chain). But as a moral matter of individual liberty, it seems to me that there's little support for a claimed freedom from getting immunized -- and especially a claimed freedom from getting your underage children immunized. A requirement that people not allow their bodies to be media for unwitting transmission of deadly diseases strikes me as quite compatible with a generally libertarian perspective on the world.

Megan McArdle (Asymmetrical Information) posts about HPV immunization — she thinks it's good even if it's mandatory (which the Texas version apparently isn't) — and draws, among others, this comment:

HPV, like HIV and unlike Polio, Tetanus, Pertussis, etc is 100% preventable by behavioral change.

Let me stress it again: It is 100% preventable (setting aside rape and congenital transmission) by behavioral change — lifelong sexual abstinence. It is not preventable by only having sex within marriage, since your husband or husbands (even the most religious don't mind remarriage after a spouse's death) might have been infected with HPV when you married them, or might cheat on you after marriage, get infected, and reinfect you.

Having fewer sexual partners will decrease your chances of getting infected, but they won't decrease them by 100%, and the chances will remain pretty substantial, I'd wager, given that "At least 50% of sexually active people will get HPV at some time in their lives." Having one sexual partner and making sure that he was a virgin at the time decreases the chances further (if he doesn't cheat on you), but, you know, sometimes the person you fall in love with — even the person you think God may have meant for you, if you think about things that way — happens to have done some wrong things in the past. If "behavioral change" means "dumping the one person you've fallen in love with and want to spend the rest of your life with because they've had sex with some other people" (or even "because they fall within the 50% of the public that has been infected with HPV"), then it might be good to make that clear (though again even that won't give you 100% immunity unless you are 100% confident that your spouse won't cheat on you).

And of course on top of all this, we know that teenagers, even well-brought-up teenagers, aren't always the best at behavioral change. However moral you might be, and however moral you are bringing up your daughter, are you 100% sure that she's going to take that supposed 100% prevention approach? If your religion teaches that sin, and giving in to temptation, is part of human nature — and if you think our society has turned into a moral cesspool that constantly bombards young people with praise of sexual immorality — then why do you think that even well-intentioned, well-taught teenagers are going to be able to completely resist temptation and sin?

Look, if we're talking assumption of risk and "you could have prevented it yourself" in the context of skydiving or bungee-jumping, I can understand that. But if "100% preventable" means living an entirely asexual life, and "almost 100% preventable" means dumping the person you're in love with because he isn't a virgin (and maybe is even part of the 50% of the public that's HPV-infected), then we're way out of the range of normal assumption-of-risk talk. And when you add to that the fact that failure to immunize turns the unimmunized into unwitting but still dangerous vehicles of transmission of deadly disease — when a woman dies of HPV-related cervical cancer, the unimmunized people who helped spread the HPV to the woman helped cause her death — then the case for "never mind immunizing, people should prevent HPV themselves because it's 100% preventable by behavioral change" becomes very weak indeed.

A 1999 article reports:

Objective: To determine whether patients with genital warts carry human papillomavirus (HPV) DNA on their fingers.

Methods: 14 men and eight women with genital warts had cytobrush samples taken from genital lesions, finger tips, and tips of finger nails. Samples were examined for the presence of HPV DNA by the polymerase chain reaction.

Results: HPV DNA was detected in all female genital samples and in 13/14 male genital samples.

HPV DNA was detected in the finger brush samples of three women and nine men. The same HPV type was identified in genital and hand samples in one woman and five men.

Conclusion: This study has identified hand carriage of genital HPV types in patients with genital warts. Although sexual intercourse is considered the usual mode of transmitting genital HPV infection, our findings raise the possibility of transmission by finger-genital contact.

(Sex Transm Inf 1999;75:317–319)

I don't know whether the study has since been confirmed or undermined, though some casual searches suggested this possibility is being taken seriously, but the magnitude of the risk is unknown. If you know more than I do about this, please post something on the subject in the comments.

Likewise, while my casual assumption is that if hand-genital transmission is possible, hand-hand transmission followed by hand-genital transmission by the recipient (i.e., a woman shakes an HPV carrier's hand, and then transfers the HPV to her hand when touching her own genitals), I'm not sure whether that's right. Of course, even if it's not, this means that an HPV-positive parent could transmit the virus to an infant or a toddler when wiping, washing, or examining the child. (Mother-to-newborn transmission of HPV is also possible, though my vague sense from a few snippets I've seen is that it's not seen as a major likely source of serious disease.) In any case, given the recent debates about HPV immunization, I'd be happy to hear more about this from those who know.

The world seems to be an even ickier place than I had thought.