My journey began one blustery day in October of 2007 when I came upon an article in the New York Times. This article completely shook my perspective. It didn't change my belief that abortion was murder or my desire to save the lives of unborn babies. Instead, it simply completely overhauled my tactical focus and made me realize that the current efforts of the pro-life movement are extremely backwards.
The first thing I learned from that New York Times article shocked me: it turns out that banning abortion does not actually affect the abortion rate.
The source for the article was the Guttmacher Institute, which has been turning out reports on abortion and birth control for years. It therefore proved rather hard to try sort through the pile of papers to track down little details like how they calibrated their surveys and what their systematic errors were. In fact I never found those little details. So I will have to try to do it for them.
The first thing to realize is that when abortion is banned, you can only estimate the rate instead of counting hospital statistics. The latter is subject to typographical errors and similar slippages, but the former has to be estimated from other factors. Those connections to other factors are not known exactly, and the degree to which you don't know these represents a systematic uncertainty. For example, the rate of maternal injuries after a particular type of abortion is apt to vary from place to place, and if the injury is known to be associated with abortion and is concealable, you may not even have accurate statistics on that. Some other practice, such as FGM, may increase or decrease injury rates. So you can estimate that from thus-and-such an injury rate there must have been X many abortions, but that has to be accompanied by an estimate of your uncertainty.
The second thing to realize is that when you want to make such estimates, you should try as hard as possible to calibrate your estimators, preferably with a known rate but in lieu of that, cross-calibrate with as many other estimators as you can. Without that you have no confidence that your calibration is correct. As seen below, it isn't clear how they calibrated their estimates.
This is their report on the incidence of abortion in Nigeria.
Methods: Experienced physicians conducted interviews at a nationally representative sample of 672 health facilities in Nigeria that were considered potential providers of abortions or treatment for abortion complications. The data were used to estimate the number of abortions and to describe the provision of abortion-related services.
This isn’t enlightening. Nor are the references to UN reports, which are typically tables of numbers without sources.
Tracking down references (such as this one used to justify an estimate that half of non-physician abortions result in complications requiring physician assistance) turns up stomach-churning euphemisms like "menstrual regulation by vacuum aspiration." It also turns up "Regional variations in the estimated abortion rate are high" in the Philippines. Likewise in Nigeria, the rate they estimate for abortions varies by about a factor of 4 from the more Western South and the Muslim North.
This variation can be interpreted two ways: the abortion rate depended strongly on the culture, or their methods have a systematic error of up to a factor of 4.
Mexico is a candidate for partial calibration. Guttmacher estimates 533,000 abortions in 1990 and 875,000 in 2006, or about 33 per thousand women. Johnston references the number of abortions in Mexico during this time (again using some of those unreferenced UN sources), and ends with a number of about 13000, or rate of 2.2 per thousand! (even if you double that (I don't know his method either) that’s still quite a factor away from Guttmacher!). Then the laws changed to allow abortions under some circumstances in some areas, and the measured rate was about 40,000/year. This does not measure exactly the same thing, but it does not fill me with confidence in Guttmacher’s numbers, which would predict 70,000 in Mexico City without any magnet traffic from districts with more restrictions.
From the Mexico numbers I cannot get a clean estimate for the systematic error, but there’s a factor of 15 between what was able to be measured and what they estimated. Another group quotes higher numbers than the Johnston's numbers but lower than Guttmacher's. If I assume that Mexico City did not attract anybody from neighboring areas for abortions (which is a whale of an assumption), then Guttmacher's numbers are 75% too high. (If it did attract from outside, their numbers are even farther wrong.)
In consequence, the chart on the Economist's site referencing Guttmacher's numbers is misleading at best.
In the USA, the abortion rate shows several interesting features. The Guttmacher estimates are always higher than CDC measurements by at least 20%, so in countries where abortion is legal we can estimate their systematic uncertainty at 20%. The other relevant feature is that over 7 years from 1973 to 1980, the number of abortions went up linearly starting at 615,000 until it more than doubled at 1,298,000. The population grew too, but not at that rate, so we are looking at a change in demand for abortions: it grew when the procedure became legal.
Making abortion legal changed the culture as well as the law.
The executive summary is that their claim that abortion laws do not affect abortion rates is not supported by their results. I am not claiming that they are getting their numbers out of thin air, but I am claiming that they have not represented their results in ways that would pass muster in hard science.
Therefore I am confident that the claim accepted by the author of the original post
I was shocked to find that the countries with the lowest abortion rates are the ones where abortion is most legal and available, and the countries with the highest abortion rates are generally the ones where the practice is illegal. It's true.is not proven, and almost certainly not true.
Suppose the Guttmacher numbers for Nigeria were only 75% wrong, and not a factor of 4. Then their very numbers contradict their claim: two different cultures have wildly different abortion rates in the same country with the same laws. In the US changing the law created demand.
Then the author takes up birth control vs abortion, using another Guttmacher study looking at Europe, presumably on the theory that this means they don't have to correct for different cultures. One major problem with this is that the culture is exactly the critical point at issue. A minor problem is that the East and Western cultures have some serious points of difference. And since the same countries that restrict birth control also ban abortion, and thus the Guttmacher estimates for abortion rate cannot be relied on (as described above), they cannot reliably compare countries with easy abortion and those that ban it to study the effects of birth control availability. The only appropriate way to do this is to study it using the uncertainty estimates that they leave out.
The author riffs on this finding thusly:
The cause of abortions is unwanted pregnancies. If you get rid of unwanted pregnancies the number of people who seek abortions will drop like a rock. Simply banning abortion leaves women stuck with unwanted pregnancies. Banning abortion doesn't make those pregnancies wanted. Many women in a situation like that will be willing to do anything to end that pregnancy,
It is hard to know where to begin with this. To start with, "wanted" isn't a simple binary state. The "many women" is a variable fraction which will change depending on the culture those women are embedded in. Unwanted pregnancies come from sex (in this country mostly unmarried), and the rate of unmarried sex likely to result in unwanted pregnancies will depend on things like the cultural attitude towards birth control. The rate of unmarried sex is "elastic" in the economic sense. (I'm not addressing married birth control here.)
The author also claims that "I realized that the only world in which opposing birth control made any sense was one in which the goal was to control women’s sex lives. After all, birth control allows women to have sex without having to face the "consequences" of sex."
That isn't correct either. Whether you believe their doctrine or not, Catholics (and not that long ago almost all other Christian groups) consider sex without openness to life to be disordered, and that breaking the connection between sex and babies is to reject God's gift. This was true for both men and women, so the author's invidious limitation of restrictions to women is also false. Do not bother trying to cite that 98% statistic: it is a misrepresentation of a misinterpretation, playing "telephone" with a dubious study.
Given that an important part of sex is mental, it shouldn't surprise anyone that sex with birth control is going to be different than sex without. If nothing else, there is a sense of adventure associated with the latter. Setting aside the claims for spiritual differences, there are real psychological differences.
The first generation of the Pill used very high hormone doses, and IIRC it did in fact prevent implantation and did act as an abortifacient. Current generations use smaller doses, and it is plausible that their abortifacient component is smaller. UIDs, on the other hand, seem to be so toxic that sperm, ova, and any resulting embryos don’t survive—it is abortifacient.
The section on how many spontaneous abortions there are, and how hypocritical pro-life groups are when they don’t worry about that, is also pointless. You can entirely eliminate spontaneous abortions by sterilizing women, and the logic of the section suggests that pro-lifers should conclude that is a good thing. No.
The author says there are so many spontaneous abortions that consistent pro-lifers should hold fundraisers for research to "save the zygotes." This is a straw man, of course: First because anybody who knows any biology at all knows it is impractical, and second because there is a world of difference between dying from a defect and murder. Purpose matters. Many people go to hospice to die, but we don't want nurses putting cyanide in their IVs.
She notes a study that makes a prediction about abortion rates:
According to Dr. Jeffery Peipert, the study's lead author, abortion rates can be expected to decline significantly—perhaps up to 75 percent—when contraceptives are made available to women free of charge. Declaring himself "very surprised" at the results, Peipert requested expedient publication of the study, noting its relevance to the upcoming election.
I flat out don't believe that contraceptives are so expensive that 75% of the abortions would go away if contraceptives were free. The key phrase here is "relevance to the upcoming election." This smells of buried assumptions (such as effectiveness in the lab vs how people actually use things).
Mea maxima culpa. As pointed out by bs king below, the study dealt with implants, and not the usual pills or condoms. I was fed up with the Guttmacher misdirection and annoyed by the red flag of political utility, but that is no excuse for not examining the study. Implants are more idiot-proof, and if they were pushed more strongly and paid for by others more young women would doubtless use them, and fewer would get pregnant. That could well reduce the abortion rate, and it is quite reasonable for Libby Anne to cite this as a desirable possibility.
In a different post the author claims that studies have proven that abstinence-only sex education fails.
Who shall I believe, the study compilers or my lying eyes? I'm old enough to have seen the changes for myself. The "sexual revolution" meant that many more people began having sex outside marriage, and the VD rates prove it. The culture used to deprecate extra-marital sex. People did it anyway; it caused problems. Now the culture encourages it. Many more people indulge, and the problems are vast.
So what do parents try to do to change the culture? The first step is to teach their own children at home, of course, but the second is to try to make sure the schools don't condone extra-marital sex. This is what the abstinence-only classes are trying to do. Since school boards are supposed to be under local control, you'd think this was possible, but apparently it isn't supposed to be allowed.
Some jibber that "You can’t turn the clock back." If the culture has taken the wrong turn (and wrt sexual attitudes that seems glaringly obvious(*)), you have to try to change direction. Parents, faced with a popular culture largely out of their control, try to change what they can. If they succeed with the first steps, perhaps we can have changes in the entertainments, and then changes in behavior. We do not expect perfection, but improvement. We know it can be better.
Do you doubt that cultural changes matter? Look at Guttmacher's own notes on Nigeria. The Northern culture has a much smaller abortion rate than the Southern. The abortion rate in the US rose with time, as it became culturally acceptable. The VD rate skyrocketed when people starting believing they were entitled to have sex. Culture matters.
(*) The fraction of children without fathers is skyrocketing, the fraction of couples who can commit long enough to raise children is low, the fraction of ever-adolescent men is catching attention, a sexual economy that exploits women (give it away when you're young, and when you're older the men want the young ladies instead), the barrenness, the perpetual pressing of boundaries because the ordinary does not satisfy: all point to something deeply unhinged. Little seems to encourage prudence or self-control or courage.
3 comments:
Thank you for being honorable enough to do the work on this. I tend to just think Guttmacher = deceitful and read no further.
Libby Anne has come to the conclusion she has been advertised into, but believes it is one she has thought through logically. That was the purpose. My reasoning goes the other way: if the data has to be so systematically altered, isn't that an admission that something is amiss?
I suppose that cuts both ways as well. I certainly have know evangelicals who try to advertise people into their prolife belief rather than convince them logically, though some elements of logic remain present. It's a hard discussion to have.
Please see your e-mail.
A few thoughts....
I like your take on the legality/abortion rate point. It is bizarre to me that one would make the assertion that the legality of a behavior has no bearing on how often people engage in that behavior.
Of course it does.
As you point out, the desire for pregnancy is not a binary state. In fact in the oft quoted statistic "half of all pregnancies are unplanned" there are actually 3 categories: planned, unplanned, and mistimed. Presumably at least some of the "mistimers" get abortions. If abortion were illegal, I would guess most of them would not (presuming mistimed means something along the lines of "with a good committed partner that I wanted babies with eventually").
Anyway, abortion is (for most women) a complex choice. To say legality doesn't factor in at all is silly. I know it's anecdotal, but all of the friends I have who have had abortions would likely have not had them had it been illegal. All of them agonized over it, and I believe it would have tipped them in the direction of not doing it if they'd had to seek out something in a back alley.
That being said, I have to quibble with you on one point...and that's about the Peipert study.
In that study, the free contraceptives included implantable hormonal birth control and IUDs. Both of those run around $1000 to get, but last for several years. I actually think $1000 up front is prohibitive for most young women if not covered by insurance. These also are the most effective and lowest maintenance forms of birth control, meaning even the most irresponsible woman can't screw it up once it's in. I think providing women with free (near) foolproof birth control that they would not otherwise have would actually be quite likely to reduce abortion rates. Even presuming that the failure rate was maxed out (about 7 per 1000 women) and that every single one of those resulting pregnancies was aborted, that's still lower than your average inner city abortion rate now.
I would very much like to see that strategy tested on a larger scale actually, especially with an eye towards STD rates and marriage/out of wedlock birth rates.
I do think the pro-life movement has gotten a little bogged down in the same few strategies, that as far as I can tell haven't made much of a difference in my lifetime. I think exploring a few new tactics couldn't hurt.
Post a Comment