Over the course of the last month, I've spent significant time updating Abort73's international abortion statistics. In addition to maintaining pages for all of the major English-speaking countries, Abort73 also has a page dedicated to worldwide abortion statistics. Unfortunately, worldwide abortion statistics don't actually exist, forcing us to rely upon estimates instead.

Though a few different organizations report on the incidence of abortion worldwide, there is only one source for global abortion totals—the Guttmacher Institute. Both the United Nations (UN) and the World Health Organization (WHO) republish Guttmacher's estimates as reliable and authoritative. But are they? That's the question I recently set out to answer. 

The Guttmacher Institute, it's fair to say, has a conflict of interests. By their own admission, Guttmacher is a "research and policy organization committed to advancing [abortion] rights in the United States and globally." In other words, the Guttmacher Institute has a clear and active agenda that influences everything they do—which is why their publications go well beyond the mere reporting of facts. Guttmacher also interprets those facts in an ongoing effort to normalize abortion around the globe. But what if there are no facts to interpret? Would they be willing to create their own?

According to Guttmacher's latest estimates, more than 56 million abortions occur each year around the world. The source for this number is a 2016 study authored by Guttmacher's Principal Research Scientist, Gilda Sedgh, and published in The Lancet. The study's Background Summary concludes with a dual declaration of purpose. First, the study aims at estimating the percentage of pregnancies that end in abortion. Second, the study seeks to answer this question: Does the legal status of abortion have any bearing on the rate of abortion? 

If that question strikes you as absurdly self-evident, I know how you feel. Does the legality of abortion have any influence on a woman's willingness to have an abortion?! Yes, it has a MASSIVE influence—and both history and common sense bear this out. Nevertheless, the abortion industry is determined to convince the world that criminalizing abortion has no bearing on the frequency of abortion—a proposition that is largely propped up on medical studies that are too dense for most people to critically analyze. 

If you hadn't noticed, 56.3 million abortions is a HUGE number. The United States makes up approximately 4.3% of the global population and has one of the highest abortion rates in the world. And yet, if the entire world aborted at the same rate as the United States, there would be "only" 21.5 million abortions per year. That's barely a third of Guttmacher's global estimate. Is that right? How could one of the largest nations in the world, with one of the highest abortion rates in the world, account for such a small percentage of abortions in the world?

Without any further ado, let's take a look at some actual numbers. To make their global abortion tabulation, Guttmacher gathered data for 184 countries over a 25-year period (1990-2014). If all 184 countries had published official abortion statistics for all 25 years, there would have been 4,600 country-years of data to analyze (184 x 25). In reality, there were only 962 country-years of data available, and 527 were deemed incomplete. The remaining 435 country-years represent only 9.5% of the theoretic whole. Though 26 additional country-years were mined from published and unpublished abortion surveys, Guttmacher raised all 26 of these tallies by 55%—on the assertion that women underreport their abortions by 30-80%. Finally, 81 more country-years of statistics were obtained from "nationally representative studies," but an undisclosed number of these studies used "an indirect approach to estimate abortion incidence in developing countries." 

All told, there wasn't a single year of public or private abortion data available for fully half of the 184 countries considered, and only 74 of the remaining countries had existing abortion data for more than one year. For the most recent reporting period (2010-2014), only 33% of the countries in the world had even a single year of abortion data available—which is why most of Guttmacher's abortion statistics come from Europe and North America. Only 11% of Latin American countries had any abortion data available between 2010 and 2014, compared to 19% of African countries and 32% of Asian countries. 

In other words, of the 56.3 million global abortions Guttmacher estimates annually, no more than 10% can be officially accounted for and only 23% have any basis in empirical evidence. The other 77% are assumed based upon a complex statistical model that estimates regional abortion totals according to the number of women of reproductive age, the assumed percentage of those women who fall into various reproductive categories, and the assumed frequency of abortion for each category.

In the study’s own words:

Our estimates have several limitations. Information about abortion incidence in the developing world is scarce. The quantity and precision of data in developing regions are reflected in the wide uncertainty intervals around estimates for these regions... Because the evidence base of abortion rates is scarce, we present abortion incidence for 5-year periods rather than for each year, and we present rates for subregions rather than for countries. 

If you're counting at home, somewhere around three-quarters of Guttmacher's global abortion tally is built on at least three levels of uncertainty. First, researchers must estimate how many women in a given region are married or cohabiting. Second, they must estimate the distribution of these women across four reproductive categories (those without need of birth control, those using birth control, those whose birth control fails, and those not using birth control). Finally, they must estimate what percentage of women in each of these categories is likely to obtain an abortion. Because of “data constraints,” the authors simply assume a fixed abortion percentage for all unmarried, non-cohabiting women. Though the study notes that abortion incidence in each subgroup varies with "coital frequency, fecundity, the strength of motivation to avoid carrying an unintended pregnancy to term, and a woman's ability to act on her preferences,” they chose not to account for these variances in their model.

The question we're left with then is this: In the absence of measurable data, how in the world are their regional abortion estimates made? The study answers this way:

Past estimates of global abortion incidence relied on available abortion data and on qualitative assessments of exchangeability to make inference to countries and territories lacking data. After the last global estimates were made, country-specific estimates of the level of contraceptive use and unmet need for contraception among married women, and the proportions of women who are married, were published. These estimates, and their association with abortion incidence, make possible a systematic, model-based approach to estimating abortion incidence.

So, in the absence of statistical data, the authors look to neighboring countries to fill in the gaps—assuming that abortion rates will be fairly constant across regions. Once these estimates are published, they are then used as the basis for all succeeding estimates. You see the danger, of course. If those first estimates happen to be wrong, then everything down the pipeline will be wrong too. That's the risk they take by allowing each successive study to take as its foundation the guesswork of the study before. On the one hand, the authors assure us of the reliability of past studies; on the other hand, they say things like this:

The current estimates are also higher than previously published estimates for Southern and Northern Africa and Western Asia. We expect that the previously published estimates were lower than the model-based estimates largely because prior estimates made conservative assumptions and likely insufficient adjustments for underreporting… The estimates of abortion incidence for 1990–2014 are intended to override previously published estimates for 1995, 2003, and 2008. 

Said differently: our prior estimates are the statistical basis for 75% of our current estimates, but our prior estimates were probably wrong, so we're replacing them with our current estimates, that are predominantly based on the old estimates. Make sense?! It's all rather humorous, except for the sinister impact of their conclusions. Without any empirical evidence, The Lancet study estimates that 8.3 million abortions occur annually in Africa and 6.5 million occur in Latin America. Why should these numbers raise a red flag? Because abortion is broadly illegal in both regions, and these totals would require regional abortion rates that are among the highest in the world.

So far as I can tell, the only way to account for such bloated abortion numbers in Africa and Latin America is to assume a constant abortion rate that is unaffected by the legality of abortion. By all appearances, the authors simply assumed from the beginning what they were ostensibly trying to figure out. After all, for fully 70% of the countries in Africa and 60% of the countries in Latin America, Guttmacher couldn't find a single piece of abortion data for any of the 25 years in question. That doesn't leave much room for "qualitative assessments of exchangeability."

Though I took issue with the study's purported objective at the outset, I was far more incredulous by the time I'd read all the way through. It isn't much of an exaggeration to summarize the study like this: This study seeks to determine whether the legality of abortion has any relationship to the frequency of abortion. To answer that question, we've assumed that the legality of abortion has no relationship to the frequency of abortion. Not surprisingly, Guttmacher found the conclusion it assumed. In the words of the report: "We did not observe an association between the abortion rates for 2010–14 and the grounds under which abortion is legally allowed." 

What they did "observe" is even more outlandish. According to their probability models, women who live in countries where abortion is illegal are more likely to have an abortion than women who live in countries where abortion is legal. Yes, you read that correctly; welcome to backwards world. Guttmacher even had the audacity to publish an infographic asserting this nonsensical claim. At best, their assessment is wildly speculative. At worst, it is intentionally deceptive. Either way, it helps sustain a familiar narrative—one that goes like this: women will have abortions whether they're legal or not. Some will, certainly, but the vast majority will not.

What we see in all of the countries where actual data exists is a dramatic increase in abortion frequency in the years immediately following its legalization. We also see that even minor restrictions to the circumstances under which abortion is permissible can significantly reduce the rate of abortion. Waiting periods, mandated counseling sessions, parental notification laws—all of these requirements reduce a woman's likelihood of going through with an abortion. If relatively insignificant obstacles to abortion demonstrably reduce its frequency, just imagine what legal prohibition would do. 

Common sense notwithstanding, virtually no one questions the numbers put forward by Guttmacher, let alone their conclusions. Even Christians who purportedly oppose abortion are apt to read these numbers and naively conclude that laws don't matter—or that they do more harm than good. Do not be deceived. Guttmacher has a horse in this race. Guttmacher knows that if abortion is already occurring in countries where it is against the law, at rates that are indistinguishable from countries where it is protected by law, then all of the pragmatic arguments for outlawing abortion cease to exist. 

That's why we must always take Guttmacher's global abortion numbers with a grain of salt. Here's a good rule of thumb. Whenever the abortion industry suggests ways to reduce the frequency of abortion, take it under advisement, but don’t take it too seriously. Their lobbyists have no interest in reducing the frequency of abortion. And when the abortion industry tells you that women would be more likely to have an abortion if abortion were outlawed, do your best not to laugh—or scream.