Over the last 25 years, the USA has swung backwards and forwards on the Mexico City policy, also known as the "global gag rule", depending on the party in power. The policy is intended to reduce the number of abortions, not in America but in places poor enough to receive American aid. Opponents describe the policy as cruel and harmful to women, while supporters believe it's a totemic protection of the unborn. Medical scientists have shown it has the exact opposite effect to its intent.

First introduced by Ronald Reagan in 1984, the Mexico City policy not only bans American aid going to the performance of abortions but prevents federal funding going to other activities run by agencies that perform abortions. Even organizations that tell pregnant women how to get abortions elsewhere are defunded.

The rule put health care providers in a bind. Abortions and counseling are usually a small portion of what they do, yet most chose to lose their largest source of funding rather than buckle to what they saw as bullying.

Bill Clinton removed the rule within days of his inauguration. It has since been reinstated or scrapped every eight years as one of the first acts of the new president, depending on their party. Although it has been shown to have a negative effect on women's health, attempts to measure its impact on abortion numbers have often been limited in scope.

Dr Eran Bendavid of Stanford University obtained surveys of three-quarters of a million women in 26 countries over the Clinton-Bush-Obama changes. In the Lancet Global Health, Bendavid shows that countries heavily affected by the rule had a 40 percent increase in abortions while it was in force relative to those where the major funding sources were elsewhere. It's the correlation between US influence and additional abortions that makes the connection far more convincing than if there had just been a spike in African abortions while Bush was in office.

The explanation appears to be that by defunding major distributors of contraceptives, the Policy increased pregnancies by 12 percent in highly affected countries. Even with abortions being harder to obtain, a larger pool of women seeking them caused the increase Bendavid identified. Bendavid and co-authors note that during this period, the abortions performed were likely far less safe, probably causing many extra deaths.

The authors acknowledge several possible sources of error in their methods. However, they note that most of these, if real, would have caused them to underestimate the effects of the rule, suggesting the true number of extra abortions was probably even higher.