Thread by @secularprolife

🧵 1/ A quick example of the problems with confusing correlation and causation, from this 2020 study: "Unintended pregnancy and abortion by income, region, and the legal status of abortion: estimates from a comprehensive model for 1990-2019" thelancet.com/journals/langl…

2/ We're considering three factors here:

1. Unintended pregnancy (UIP) rates
2. Abortion laws
3. Abortion rates

Higher UIP drive up abortion rates. Stricter abortion laws drive down abortion rates.

3/ So we would expect countries with high UIP and lax abortion laws to have the highest abortion rates and countries with low UIP and strict abortion laws to have the lowest abortion rates.

4/ But often countries with high UIP have strict abortion laws and countries with low UIP have lax abortion laws. These combinations have a mitigating effect on one another and mask the relationships between UIP, abortion law, and abortion rates.

5/ This mitigating effect means countries with strict abortion laws and countries with lax abortion laws have similar abortion rates. People then incorrectly conclude the law doesn't really affect abortion rates.

6/ But the mitigating effect also means countries with high UIP (73 per 1000 women) and countries with low UIP (58 per 1000 women) have similar abortion rates (36 vs 40 abortions per 1000 women). Yet no one concludes unintended pregnancy rates don't affect abortion rates.

7/ Also worth noting is % of unintended pregnancies aborted. In countries where abortion is restricted, about 50% of UIP are aborted; in countries where abortion is broadly legal, it's about 70%. Both high numbers, to be sure, but a 20 point difference is huge.

8/ Abortion rights proponent and researcher Diana Greene Foster discussed this issue (based on similar prior data) in her column "Stop Saying That Making Abortion Illegal Won't Stop People From Having Them" rewirenewsgroup.com/article/2018/1…

9/ The 2020 study in the OP also found that global UIP decreased 18% from 1990-2019. If abortion rates are affected primarily by UIP, rather than the law, we'd expect global abortion rates to decrease too. Instead they remained about the same. Why?

10/ There are many factors in play, but abortion law is sure to be a major one. The Center for Reproductive Rights issued a report summarizing international abortion law from 1994-2014 which found "a global trend toward abortion law liberalization." reproductiverights.org/sites/crr.civi…

11/ If global UIP decreased, but global abortion laws on aggregate got more lax, it wouldn't be surprising that the abortion rate didn't change much. Again, UIP and the law can have mitigating effects on each other.

12/ Just as correlation isn't always causation, lack of correlation isn't always lack of causation. If we look at global abortion rates only in the context of abortion law, we don't see a correlation. But if we control for UIP, it appears strict laws decrease abortions.