No doubt quite a few of my readers have now heard about this study, deliberately released just now with the intent of influencing the election (!), that purports to show that free contraception reduces abortions.
I've entitled my post "My present thoughts," because I anticipate that there will be more analysis on this subject from others as time goes on. However, I have gotten hold of the paper itself (thanks to Serge at Life Training Institute) as well as this Powerpoint PDF from the study author.
Before getting into criticisms of the methodology and conclusions of the study, I want to ponder something that I haven't seen discussed elsewhere. The celebratory article on the study from NBC says this:
The results were so dramatic, in fact, that Peipert asked the journal of Obstetrics & Gynecology to publish the study before the Nov. 6 presidential election, knowing that the Affordable Care Act, and its reproductive health provisions, are major issues in the campaign.
“It just has so many implications for our society,” he told NBC News.
Since when are professional scientific journals openly and proudly attempting to influence political elections by their decisions about publication? Isn't there something a tad unprofessional about that? More than a tad? The study's author, Jeffrey Peipert, tells a news organization unabashedly that he expressly asked the journal, Obstetrics and Gynecology, to hasten the publication of his article in order to influence this fall's presidential election. Frankly, if I ever tried such a thing on a philosophy journal editor, even an editor who had already accepted an article of mine, I hope (and still believe) I'd receive a sharp rebuke. (Not that my philosophy articles have political implications anyway; the scenario is hypothetical.) Such a request should be taken as an insult to the professionalism of the editor. Regardless of whether the article was accepted independently of political considerations, the timing of its release should be decided on the basis of academic and professional considerations, including time for possible revisions and the courtesy owed to other authors whose articles were submitted and accepted longer ago. The utterly unashamed announcement that author and editor colluded to time the article to influence the U.S. presidential election should leave a bad taste in everyone's mouth, especially in the mouths of scientists and scholars, and should even cast a small amount of doubt on the objectivity of the review process itself.
Now, on to content issues.
There are some major methodological questions about and problems with this study, and in my opinion they should have influenced its publication negatively. At a minimum, the journal should have required the article's revision, as a condition of publication, to clarify some things that are right now quite unclear and to include more raw data and more explicit calculations for public evaluation. But the problems go deeper than that.
The abortion data for study participants may have been estimated rather than collected directly.
As a matter of fact, the article says that the abortion rate for study participants was estimated:
We also estimated abortion rates in our metropolitan area and in our cohort and compared these rates with U.S. and regional abortion rates. (Peipert et. al. "No-Cost Contraception and unintended Pregnancies," Obstetrics and Gynecology, December, 2012, p. 2.) [Side note by me: "December, 2012"? That's what the PDF says. Maybe in later years we're all supposed to forget that it was deliberately released in October to influence the 2012 election?]
Any reviewer worth his salt should have read that sentence, and the absence, there or elsewhere in the paper, of an explicit statement of how such an estimate took place, and said, "What?" At a minimum, the authors should have been made to answer the following questions: Was that a misstatement, or do you really mean to say that you did not collect raw data on abortions among study participants but rather estimated abortion rates in your cohort? If so, why? You were interviewing participants repeatedly throughout the study. Why estimate the data? If that isn't what you meant, then change the sentence. If that is what you meant, you need to explain precisely how you estimated abortion rates in your study group. (Another possibility is that the reference to estimating the abortion rate in the cohort refers to the standardization/weighting of the data, which I will discuss below. But this is an implausible interpretation. Estimating an actual rate of abortion in a group is one thing. Standardizing data from the group is another. It doesn't seem that this is what the sentence refers to, but if that is what the sentence means, again, it is radically unclear.)
Obviously, if the actual abortion rate among the participants was merely estimated, while we are not told how it was estimated, the article's statistical value is highly questionable from the outset.
The abortion rate, however collected or estimated, was altered by a weighting factor, and the raw data are not given.
However the researchers obtained their rates of abortions for study participants in the first place, they are quite explicit that they then renormed the data in an attempt to take into account the group's unrepresentative nature in terms of race and age.
(The acronym CHOICE refers to the study group. Cute, huh?)
Because the CHOICE cohort represents a higher-risk population (median age of 25 years and 50% black) than the general population, we standardized the CHOICE abortion rate to the age and racial (black and white) distribution of females who reside in the St. Louis region using data from the 2010 U.S. Census (direct standardization). We compared the CHOICE standardized rate with the St. Louis regional rate...(p. 3)
Translation: Because our study group was unrepresentative of the St. Louis female population by race and age, we expected to get more abortions among our study group than we would have gotten if the group had been representative demographically. So the abortion numbers we're giving you aren't the real abortion numbers, even if we did collect the real abortion numbers to begin with. They are numbers which have been massaged by a weighting factor to give the number of abortions we believe would have occurred if more of the study participants had been white and over 25 years of age.
The raw data are given nowhere in the article. Neither, ipso facto, is the precise calculation showing the weighting factor and the derivation of the stated "abortion rates" among study participants from the initial data.
Let me be clear: I am not saying that using weighting factors is per se an unethical thing to do. Nor am I saying that the weighting factors in this case were designed with malice aforethought to get a low abortion number for political ends. Maybe the number of abortions they report in the paper is a fair indication of the number of abortions per thousand that there would have been in their study sample if it had had a different demographic makeup. Weighting factors of this kind are often applied in statistical contexts. However, they can be quite contentious, and social scientists should also supply the raw data and tell precisely how the numbers they want to use were derived.
It would perhaps be uncharitable to say that a publication of study results that gives results derived by standardization without providing this information is statistically worthless. But only a little uncharitable.
Again, this is the kind of thing that the peer review process should have caught. A reviewer should have asked that the article be revised to include fuller information.
Implants and IUDs were aggressively up-sold to study participants.
This point is less geeky than the first two but just as big a deal, if not bigger. From the news reports one would have gathered that the study participants were simply offered their choice of "a range" (the exact range is an interesting question I'll discuss later) of contraceptive devices or methods, all free-free-free, and that it was simply the removal of cost that induced such a surprisingly high number to choose implants and IUDs. Conclusion: Ah, many women out there are just yearning for contraceptive implants and IUDs if only, tsk-tsk, they didn't cost so much. Were the "cost barrier" removed, all else (including the alleged reduced abortion rate) would follow as the night the day. A typical statement is this one, from the Washington Examiner story:
When price wasn't an issue, women flocked to the most effective contraceptives — the implanted options, which typically cost hundreds of dollars up-front to insert.
"As a society, we want to reduce unintended pregnancies and abortion rates. This study has demonstrated that having access to no-cost contraception helps us get to that goal," said Alina Salganicoff, director of women's health policy at the Kaiser Family Foundation.
This impression is quite misleading. One of the study's express objectives was
...to promote the use of the most effective contraceptive methods (IUDs and implants) (p. 2)
This objective was promoted actively:
Contraceptive counseling included all reversible methods but emphasized the superior effectiveness of LARC methods (IUDs and implants). (p. 1)
All participants were read a brief script informing them of the effectiveness and safety of LARC methods at initial contact and completed an in-depth, evidence-based contraceptive counseling session at enrollment. (p. 2)
In addition, the project provided education to promote the use of the most effective contraceptive methods, IUDS, and implants in an effort to alter population outcomes. (p. 5)
In other words, cost was by no means the only factor in the participants' decision to choose IUDs and implants--essentially, to sterilize themselves for a fairly long period of time quite effectively. Rather, those methods were expressly and deliberately pushed as the best methods, with researchers making a quite unabashed attempt to "alter outcomes" by this up-selling method.
The article implies that dramatic reductions in unintended pregnancies and abortions can be expected nationwide if only the study methods are generalized as a matter of policy, and it explicitly likens these methods to the Obamacare mandate for free contraception:
The Institute of Medicine recently recommended that eight primary preventive health services for women be covered without cost to patients under the Patient Protection and Affordable Care Act of 2010. Among these eight services, the Institute of Medicine recommended "a fuller range of contraceptive education, counseling, methods, and services so that women can better avoid unwanted pregnancies and space their pregnancies to promote optimal birth outcomes." As a result, all FDA-approved contraceptive methods would be covered without cost. The Contraceptive CHOICE Project essentially simulated this recommendation in our region for reversible contraceptive methods. (p. 5)
Are we to believe that it is part of the Obamacare contraceptive project that all women, or even all women seeking contraception, will be read a special script up-selling IUDs and long-term implants? As far as I know, this is not part of the present plan, though perhaps we shouldn't give the Obamacrats any bright ideas. And how would one induce all women (all women of childbearing age? all women of childbearing age who were not presently trying to get pregnant?) to listen to such a script, anyway?
It seems highly probable that the deliberate pushing to participants of these specific methods, rather than their mere provision at no cost, was in no small measure responsible for any subsequent reduction in unintended pregnancy rates that did occur.
Unless we are willing to envisage a program under which American women generally receive a high-pressure sales pitch for having an IUD or implant inserted, we shouldn't be too quick to imagine any straightforward generalization of this study's methods via Obamacare's contraception diktat.
The study participants were unrepresentative in other specific ways, and these ways were features, not bugs, of the study.
--All study participants had to be willing to use contraceptive devices and/or hormones of some kind. Yes, this may sound like a too-obvious point, but there are women in the world, even women who are sexually active and would prefer not to get pregnant, who don't want to use contraceptive devices, hormones, or implants. Some prefer just to "take their chances." Others use NFP. Others, with their male partners, might count on the method of conception-avoidance used by Onan in Holy Writ, on which I won't expand further just here.
--All study participants who were already using contraception had to be willing to change their method of contraception. This is stated expressly on slide 15 of this presentation. It's unclear exactly how this willingness was measured. Does this mean that they actually had to change their method or only to say that they were willing in theory? What changes were acceptable? For example, if they said they were willing to "try a new contraceptive method" but then, after counseling, chose to have an implant removed and to switch to contraceptive pills, which were not one of the methods the study authors were especially pushing, would this have taken them out of the study? None of this is clarified in the paper itself. But I think we can say pretty definitely that it is not in general characteristic even of all sexually active American women that they are "willing to try a new contraceptive method." So the study was unrepresentative in this way, and this aspect of the study was a fundamental part of its design. Since it was designed to promote the use of IUDs and implants, which participants might not have been using before, women who were definitely satisfied with their method of birth control and were not "willing to try a new contraceptive method" were deliberately excluded.
--It looks as though options offered at no cost to study participants may have excluded over-the-counter items such as spermicides, condoms, and sponges. This is not absolutely clear, which is one reason I did not include it among the "biggies." In some places the paper says things like, "All participants received the reversible contraceptive method of their choice at no cost" (p. 1) and "Participants were offered all U.S. Food and Drug Administration (FDA)-approved contraceptive methods and could choose any method." (p. 2)
That seems pretty unequivocal. Yet there is some reason to suspect that this only means methods that require a prescription, not simpler, non-hormonal, over-the-counter contraceptives. This Washington Times article from March of this year, for example, describes the study like this:
Called the Contraceptive CHOICE Project, the study is providing those options [IUDs and implants] and a range of others for free. Participants also can choose from birth control pills, a monthly patch, a monthly vaginal ring and a once-every-three-months shot.
That list of options which participants could receive free corresponds almost exactly to the list of "what do women choose" on slide 21 in the Powerpoint presentation. The only difference is the inclusion on the slide of the diaphragm, chosen by less than .1 of participants. (A diaphragm also requires medical involvement and a prescription.) The complete absence of non-prescription methods from the methods actually chosen is especially suspicious considering the known high popularity of condom use among contraception-minded couples. (See, for example, the charts on p. 7 and p. 27 of this document.)
If in fact study participants could not choose a non-prescription method of contraception as their designated form of contraception to receive for free, this makes the study sample all the more unrepresentative. And if that was the case, it fits extremely well with the study's overall emphasis on getting away from less effective, more compliance-requiring methods of contraception.
All of these major issues of non-representativeness in the study group are relevant to the politicized and over-the-top conclusions of the study. The article actually states,
In fact, based on our calculations in Table 3, changes in contraceptive policy simulating the Contraceptive CHOICE Project would prevent as many as 62-78% of abortions performed annually in the United States. (p. 6)
How, exactly, are policy changes going to simulate on a nationwide scale the highly artificial situation created in this study? Besides the matter of subjecting everyone to sales pitches for IUDs and implants, mentioned above, how would policy by itself be able to bring about a national situation in which all sexually active women who do not wish to become pregnant are
a) willing to use contraception
b) willing to change their method of contraception, and
c) willing to use a prescription-type method of contraception?
The promotion of liberal policy approaches to sexual issues for "reducing abortions" ignores the funding of abortions and the opposition to abortion restrictions.
This point concerns not the study per se but the use which the study authors and other liberals wish to make of its results. Since the study authors are quite blatant about the desire to influence people, including pro-lifers, to endorse President Obama's policies, we should ask ourselves what other liberal policies are making a difference in the world right now that are relevant to abortion statistics. We could start by talking about the instantaneous Obama revocation of the Mexico City Policy, but hey, let's stick to domestic issues only. So far, using threats of loss of federal money, outright lawsuits against states, and direct grants to Planned Parenthood when states stopped funding PP, the Obama administration has foiled or attempted to foil state attempts to defund PP in
The study authors actually admit that lessened government funding for abortion might be relevant to a downturn in abortion statistics:
[T]he analysis comparing repeat abortion in the St. Louis region with that in Kansas City and nonmetropolitan Missouri is essentially an ecological study. There may be several factors that affect the rates of repeat abortions, such as the economic recession, federal changes in Title X funding for family planning, and Missouri state laws that limit access to abortion. (p. 6, emphasis added)
Then, of course, there are those state laws limiting access to abortion. It should go without saying that the Democrat position on such laws is opposition of the strongest kind.
This study is clearly supposed to have a political take-home message: If you're pro-life, you now have an excuse for supporting Obama with his contraception mandate, and you can point to "science" as having "shown" that Obama's policies will reduce the numbers of abortions, maybe even dramatically. If someone is so taken in by this as not to make the obvious retort concerning Obama's and the Democrats' overall policies on abortion, he probably was looking for an excuse to be taken in. But for the record, let's not allow ourselves to forget the manic support of this administration for unfettered access to abortion and the fact that such a mania is scarcely calculated to decrease the number of abortions in the country.
I won't go quite so far as to say that this study is another example of how to lie with statistics. I will, however, say that it is an example of sloppy publication, probably sloppy because of political motivation, and seriously over-hyped results, results that are not, pace the study authors, readily generalized to the population at large nor a simple model for public policy. (And that is, of course, even waiving all the obvious moral objections to the public policies in question.)
There is a kind of casual and unnerving blindness-tending-to-totalitarianism in the study authors that causes them not to bat an eyelash at the prospect of a propaganda campaign, nationwide, to induce all sexually active women who do not wish to become pregnant to use IUDs and implants. (In fact, they do seem to envisage such a campaign. See slide 28 here which, in describing their study as a model, says that these methods should be "FIRST LINE.")
Then there is the sloppiness that apparently prevents their wondering whether women generally, who have not signed up for a free contraception study, stated that they are willing to use contraception, to choose from one of the methods offered, and to "try a new method," are going to be greatly influenced by the campaign they envisage (if it is not backed up by actual coercion). They seem to assume that all or most women secretly want what they, the elites, want them to want and are just looking to have costs taken on by someone else to make those desires come to pass. This is hardly a known scientific fact. Indeed, it is in the general neighborhood of what the study was supposed to support, making their argument, if not precisely circular, something near to a closed curve in space. But the study authors don't seem to realize this.
Ideologically committed leftism is blinding. It is blinding not only morally but also scientifically, which is why initial skepticism about such studies is always warranted until they have been investigated. We live in a world in which the "other side," especially the conservative side, is always asked to respond to any new claim instantaneously. If one does think of a pertinent instantaneous response that, one can be sure, would not be undermined by further investigation, there is nothing wrong with making it. But better still, if possible, to take some more time, get as close to the actual data as you can, and make your responses after further thought. That is what I've tried to do here, and I am grateful for the copy of the paper and the other links permitting me to do so.