Getting religion and statistics

makemydataA couple of years ago, there was a flurry of stories about a study on abortion rates worldwide. Published in the Lancet based on research by the Guttmacher Institute, the stories said that abortion rates were the same throughout the world, regardless of legality. Yet if you looked at the data, you saw some pretty significant variations. North America reported 33 abortions for every 100 live births while Africa reported 17 abortions for every 100 live births. An accompanying graph showed variations of 12 abortions per 100 live births on up to 105 abortions per 100 live births. And, most importantly, the abortion rate is the number of abortions per 1,000 women. The numbers above are ratios.

Well, Guttmacher is out with another study -- I notice that some of the authors of the Lancet study are doing this one as well -- on abortion rates. Again, that's the number of abortions per 1,000 women. Here's the BBC's take:

Restricting the availability of legal abortion does not appear to reduce the number of women trying to end unwanted pregnancies, a major report suggests.

The Guttmacher Institute's survey found abortion occurs at roughly equal rates in regions where it is legal and regions where it is highly restricted.

You will probably not be surprised to find that the accompanying graph doesn't really bear this out. Just, for example, the graph shows Latin America and the Caribbean with a rate of 31 abortions per 1,000 and Oceania with 18 abortions per 1,000 women. If you go to the actual report, you'll see that the abortion rate for Eastern Europe in 2003 is 48 abortions per 1,000 women compared to 12 abortions per 1,000 women for Western Europe. One region having quadruple the abortion rate of another doesn't read the same way to me as it does to most reporters, I guess.

Now, I read through the report and the authors are honest about dealing with some significant limitations in data collection. All of the numbers, in fact, are estimates and not actual. But the significance of the report is that it is a longitudinal study. It compares abortion rates from 1995 with those from 2003 and attempts to show that restricting abortion doesn't reduce the number of women trying to end unwanted pregnancies.

Unfortunately the BBC report doesn't demonstrate this finding in its story at all. Which is a shame since that's the point of the study. But moving beyond that, it would be nice to have a report that got a bit more balance in any case. Take, for instance:

Even the UK, which has a relatively high rate, fares well in comparison to the US, where the number of abortions is among the highest in the developed world. The institute says this rate is in part explained by inconsistencies in insurance coverage of contraceptive supplies.

Now, remember that the point of the study is to show how abortion rates change as restrictions are increased or lessened. The above paragraph doesn't speak to that issue. (To be fair, the report itself doesn't deal much with that issue in these regions either.) It also doesn't explain that the United States has more liberal abortion laws than the UK. Of course, the abortion rate declined in the U.S. at the same time that some states enacted some restrictions on the practice.

Perhaps most egregiously, the paragraph above alleges that the report says something it never did. I quickly read the report and never came across such a claim about the U.S. abortion rate being explained by insurance coverage. I did a word search for insurance and while there is discussion of insurance in various countries, the claim above is not made. And all the mentions of inconsistency deal with contraceptive use, not insurance coverage of same. I looked at the press release that Guttmacher sent out and it didn't make that claim either.

I know this is the week for inventing quotes out of whole cloth, but really. Reporting shouldn't be this difficult. And on that note, while the study itself frequently mentions the role of religion in influencing societal reaction to abortion law, you'll note that it's not mentioned here by the BBC.

Again, though, the real problem with the story is that it takes the claims of the Guttmacher Institute as the Gospel truth. And while Guttmacher is a respectable research outlet, it's unabashed in its advocacy work as well. If you simply read the report, you'll see that it's all about advocating for liberalized abortion laws. This isn't surprising, considering that Guttmacher was founded as the research arm of Planned Parenthood and they are very open about this aspect of their work. However, there are plenty of academics, including those of differing political persuasions, who disagree with their findings. Reporters have a basic duty to include them as well.

Please respect our Commenting Policy