Abortion is an issue that stirs up deeply felt passions and seems to offer little basis for compromise. But there is one thing that both sides of the debate agree on — fewer abortions are better. The pro-life side opposes abortion in principle, while pro-choice advocates generally hold that preventing unwanted pregnancies is preferable to terminating them.
That shared outlook could provide common ground on one of the most important federal initiatives concerning abortion — the Mexico City Policy. This executive order, announced in 1984 by the Reagan administration at the United Nations International Conference on Population and Development, requires all foreign nongovernmental organizations that get U.S. family planning assistance to certify they will not perform abortions or provide counseling about the procedure.
The Trump administration has greatly expanded the policy to condition almost all U.S. global health aid on compliance with these restrictions, including HIV, malaria, and maternal and child health programs. The U.S. spent more than $7 billion on international health assistance in 2017.
We recently published research (Brooks, Bendavid, and Miller 2019) indicating that, rather than reducing abortion, the Mexico City Policy has had the unintended effect of significantly raising abortion rates in a set of African countries that rely heavily on U.S. family planning and reproductive health aid. Many of these abortions are likely to have been performed unsafely, endangering the health of women who had them, previous studies of abortion indicate (Grimes, et al 2006).
Why would abortions go up under a policy that bans promoting the procedure? We found strong evidence that, when the policy was in force from 2001 to 2009, contraceptive use dropped in countries more reliant on U.S. aid and was accompanied by increases in pregnancies and abortions.
That is not surprising. International organizations that offer abortion counseling, such as the International Planned Parenthood Federation, are also major suppliers of contraceptives and family planning information. Their decision to continue offering abortion information made them ineligible for U.S. funding, which forced them to scale back a broad range of family planning services. Our research suggests that cutbacks such as these are driving the higher pregnancy and abortion rates we find.
Groups opposed to abortion and those that believe it is a woman’s right to choose do not agree on much, but they may share the view that the jump in abortion rates associated with the Mexico City Policy is a significant problem.
Only a few studies have rigorously examined the Mexico City Policy’s impact (Bendavid, Avila, and Miller 2011; Jones 2015). Our research stands out for the long period of time and large number of countries we examined. In addition, we took advantage of a peculiar feature of the Mexico City Policy — it has alternately been imposed and lifted by successive Republican and Democratic administrations. We looked at trends in contraception use, pregnancy, and abortion in 26 sub-Saharan African countries from 1995 to 2014 during the presidencies of Bill Clinton, who rescinded the policy; George W. Bush, who reinstated it in January 2001; and Barack Obama, who overturned it again in January 2009.
As a result, the organizations that were the most important providers of family planning and reproductive health services in the countries we studied lost U.S. funding when George W. Bush was in the White House, a gap other donors did not fill. Funding resumed under President Obama. This stop-and-start pattern serves as a natural experiment testing what happens to abortion when the policy is in place and when it is no longer in effect.
Our main data source for abortions and pregnancies was Demographic and Health Surveys (DHS), funded by the U.S. Agency for International Development. Information on contraception came from the United Nations Population Division’s World Contraceptive Use dataset. We obtained data on U.S. family planning and reproductive health aid by country and year from the Organization for Economic Cooperation and Development and also used the World Bank’s World Development Indicator for data on a range of economic and demographic variables. In total, we examined data on pregnancies and abortions for nearly 750,000 women.
In our statistical analysis, we separated countries into those with relatively high or low exposure to the Mexico City Policy, which was a measure of how much countries relied on U.S. family planning assistance, when the Mexico City Policy’s funding restrictions were not in place. We then compared rates of contraception use, pregnancy, and abortion in high- and low-exposure countries when the policy was in effect and when it was not.
The results were striking. In the highly exposed countries, contraceptive use was 14 percent lower, while pregnancies and abortions were, respectively, 12 percent and 40 percent higher when the policy was in force during the George W. Bush administration compared with the rates during the Clinton and Obama presidencies. Simply put, when the Mexico City Policy was active, there was less contraception and more pregnancy and abortion. When the policy was voided and family planning aid restored, those trends reversed.
An obvious question is whether something besides the Mexico City Policy might explain the abortion ups and downs that we found. Our investigation was not a randomized trial in which variables can be carefully controlled. In a statistical study like ours, there is always a chance that some unobserved factor may drive the results.
Nevertheless, we are confident that our interpretation of the Mexico City Policy’s effects — that abortion and pregnancy rates varied depending on access to contraception — is correct. Any confounding factor would have to vary systematically both with the sharp off-again, on-again pattern of the policy and with the degree to which family planning programs in each African country benefits from U.S. assistance — an unlikely scenario.
There could also be natural concerns about abortion reporting. Consider two related potential issues. First, abortions are notoriously difficult to measure, largely because many are illicit and women are uncomfortable reporting these. In particular, the DHS survey data is widely considered to underreport abortions. In addition, it can be hard to distinguish induced abortions from spontaneous abortions in the DHS data.
However, as with any other potential confounder, underreporting would have to vary with the sharp off- again, on-again pattern of the policy and also with the degree to which family planning programs in each African country benefit from U.S. assistance. We conducted extensive simulations to determine the extent to which underreporting of abortion could have influenced our findings and ultimately concluded that although underreporting is present, it did not meaningfully bias our study’s results.
Second, the Mexico City Policy could have made women more reluctant than usual to report if it altered in some fashion the legal or cultural environment for abortion. Because our abortion data is reported by individual women rather than governments or private organizations, this seems less likely. But importantly, if there were selective incentives for abortion reporting under the policy, the most likely scenario would be relatively less reporting of abortions when the policy was active and in countries benefiting more from U.S. aid. If present, this would therefore lead us to underestimate the unintended consequences of the Mexico City Policy that we find — or the true extent of the rise in abortions when the policy was in effect would be greater than our analysis suggests.
Our research did not look at the effects on women’s health of the changes in contraceptive use, pregnancies, and abortions under the Mexico City Policy, but it is reasonable to believe that maternal deaths and injuries have risen, perhaps significantly. Abortion incurs risk, and more abortions are likely to drive up mortality rates. Beyond that, to the extent that international organizations are forced to cut back family planning services, more abortions are probably performed under unsafe conditions, putting women in greater danger.
This study, unprecedented in scope, strengthens the case that making family planning services and contraception more widely available is an effective way to reduce abortion. From that perspective, the Mexico City Policy does exactly the opposite of what is needed to prevent the unwanted pregnancies that prompt many women to seek abortions.
In this regard, it seems possible that everyone could agree that a policy designed to curb abortion but winds up increasing it is a failure.
 Benin, Burkina Faso, Burundi, Comoros, Eswatini, Ethiopia, The Gambia, Ghana, Guinea, Kenya, Liberia, Lesotho, Madagascar, Malawi, Mali, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, Tanzania, Uganda, Zambia, Zimbabwe
 We focused on modern methods of contraception, defined as female and male sterilization, oral hormonal pills, the intra-uterine device, the male condom, injectables, implantables, vaginal barrier methods, the female condom, and emergency contraception.
Bendavid, E., Avila, P., Miller, G. United States aid policy and induced abortion in sub-Saharan Africa. Bull World Health Organ 2011; 89: 873–80.
Grimes, David A., et al. Unsafe abortion: the Preventable Pandemic. The Lancet 2006; 368.9550: 1908-1919.
Jones, Kelly M. Contraceptive Supply and Fertility Outcomes: Evidence from Ghana. Economic Development and Cultural Change 2015; 64(1): 31–69.