Correlation is not causation

correlationisnotcausationI meant to highlight this good example of skepticism last week by USA Today's Cathy Grossman. It all comes out of an interesting study showing that states with more Catholics tend to favor expanded rights for homosexuals. Grossman highlights the finding and discusses a bit of it in context:

The main thrust of the study was to examine whether there is "pro-gay bias in policy making" (the authors conclude no) or a tyranny of local majorities "in which anti-gay majorities trump minority rights" (the authors again say no).

For adoption, marriage, and civil unions, conservative state majorities can win out. But for hate crimes, health benefits, housing protection, and job protection, there is no tyranny of the majority blocking minority rights. Indeed, here, the majority seems to favor these civil rights protections.

Mark Silk, who heads the Leonard E. Greenberg Center for the Study of Religion in Public Life at Trinity College and blogs at Spiritual Politics analyzed the study and finds that six of eights states with majority populations supporting same-sex marriage are the states with the highest proportion of Catholics:

Silk tosses a dig, "way to go, Bishops!" but is that a fair shot?

The bishops have campaigned long, loudly and clearly against same-sex marriage but the Catholic Church also offers a pervasive message of social justice, an umbrella many liberal Catholics stand under when they argue for marriage equality or life issues such as abortion, contraception and end-of-life decisions.

While I understand how Catholics would follow church teaching on social justice when it comes to abortion, contraception and end-of-life care, I'm not sure I quite understand how liberal Catholics argue for same-sex marriage under the umbrella of Catholic social teaching on social justice issues. I'm not disagreeing with what is written -- just wondering how it's done since church teaching is rather clear on marriage being a heterosexual institution. That's probably for another story.

But more importantly, religion journalists and scholars need to remember that correlation is not causation. Just because two variables exist side by side (high proportion of Catholics in a given population and high support for same-sex marriage) does not mean that one causes the other. It makes no more sense to say that support for same-sex marriage leads people to be Catholic than to say that being Catholic leads people to support same-sex marriage. Either of those things could be true, of course, but there's nothing in this particular study to indicate that.

Now, it's still an interesting statistic and I would encourage social scientists and journalists to dig further. We just have to avoid logical fallacies while doing so.

Art is the cover of a Soul Coughing album that includes a song with the headline above.


Please respect our Commenting Policy