Sometimes social science is like a snipe hunt, just stumbling through the dark with a sack in your hand yelling "Here, snipe!"
Is X a good idea?
First, you must state a hypothesis why it is or is not a good idea. No points for an opinion that it's icky or that it's wonderful.
Then you have to hunt up a database that might test your hypothesis.
Then you use multivariate analysis to attempt to eliminate confounding variables. The broader the hypothesis, the more statistical noise.
It can take a long, long time to produce real evidence, but somebody ought to do it. The snipe hunters below undertook to test hypotheses from both sides of the political debate over marriage equality. Looks like they had about the same luck most snipe hunters have.
Note that this hunt for real effects has little to do with fairness. Those of us who say bans on gay marriage are plain vanilla sex discrimination need nothing beyond a grasp of logic and an understanding of fairness no more sophisticated than found in kindergarten.
Sometimes fair policies have unfortunate collateral effects.
You open the professions to women without upgrading the pay and status of K-12 teachers and you get K-12 teachers who are simply not as smart as the women who were forced into K-12 teaching for lack of other opportunities.
You pass open housing laws and you get the black middle class bailing from the inner cities for the non-racist reason that underpins some "white flight." If you can afford it, you want your kids raised with better schools and less crime.
You don't reinstate sex discrimination in employment or race discrimination in housing.
For similar reasons, I don't think you pine for the days of sex discrimination in marriage laws if the snipe hunters ever find a practical down side. You find ways to deal with the down side in the interest of being fair.
The abstract below the fleur-de-kos was copped from an academic file sharing website. The authors are economists at Emory University, where you could find them if you have a professional interest.