“[O]ne of the strongest factors predicting divorce rates (per 1000 married couples) is the concentration of conservative or evangelical Protestants in that county,” the researchers explain. Religiously conservative states Alabama and Arkansas have the second and third highest divorce rates in the U.S., while religiously liberal New Jersey and Massachusetts have two of the lowest. Full graph below shows the regional correlation:

Describing their findings as a “puzzling paradox,” the researchers explained that the higher divorce rate among religious conservatives is tied to early marriage and early childbearing — factors already known to contribute to strained marriages and divorce. “Starting families earlier tends to stop young adults from pursuing more education and depresses their wages, putting more strain on marriages,” Glass stated.

Full piece here.

Also:

In America, religiosity and conservatism are generally associated with opposition to non-traditional sexual behavior, but prominent political scandals and recent research suggest a paradoxical private attraction to sexual content on the political and religious right. We examined associations between state-level religiosity/conservatism and anonymized interest in searching for sexual content online using Google Trends (which calculates within-state search volumes for search terms). Across two separate years, and controlling for demographic variables, we observed moderate-to-large positive associations between: (1) greater proportions of state-level religiosity and general web searching for sexual content and (2) greater proportions of state-level conservatism and image-specific searching for sex. These findings were interpreted in terms of the paradoxical hypothesis that a greater preponderance of right-leaning ideologies is associated with greater preoccupation with sexual content in private internet activity. Alternative explanations (e.g., that opposition to non-traditional sex in right-leaning states leads liberals to rely on private internet sexual activity) are discussed, as are limitations to inference posed by aggregate data more generally.

So by the transitive property...