Michael J. New: Finding Flaws in an Abortion Rates Study and the Mexico City Policy

Anti-abortion protesters take part in the National March for Life on Parliament Hill in Ottawa, May 9, 2013. | (Photo: Reuters/Chris Wattie)

One significant pro-life accomplishment thus far during the Trump administration has been the decision to reinstate and expand the Mexico City policy, which prevents U.S. foreign aid from funding organizations that perform or promote abortions. It has been in place during the administration of every GOP president since Ronald Reagan. Under Trump, the policy now includes not only family planning funds but also funding for global-health assistance.

Unsurprisingly, supporters of legal abortion have been vigorous in their opposition to this policy, aggressively promoting studies that claim to show that the Mexico City policy has in fact increased abortion rates overseas. A 2011 study published by the World Health Organization, and more recently a 2019 study by Rutgers University professor Yana van der Muelen Rodgers both have received a significant coverage.

This June, a new study on this topic appeared in the British Medical Journal The Lancet Global Health, authored by a team of researchers at Stanford University, has been receiving a considerable amount of attention. This study is somewhat different than the other two, analyzing more years of data. It is more rigorous analytically, and it appeared in a prestigious academic journal, but even so, it contains many of the same methodological shortcomings as the previous studies.

For one thing, it is important to note that overseas abortion data are often incomplete. Like other studies assessing the effect of the Mexico City policy, this one suffers from missing many data points. The study analyzed abortion data from 26 African countries between the years of 1995 and 2014. The authors report no abortion data for 170 of the 520 state-year pairs, meaning that nearly one-third of the potential data points are missing. Also, the authors obtained abortion data for all 20 years from only six of the 26 countries covered in the study, meaning that more than three-quarters of the countries in the dataset have missing data points.

Click here to read more.

SOURCE: Christian Post, Michael J. New