Steel man one: controlling the social costs of mate competition. Struggling to find a wife? Start an affair with a married woman, persuade her to divorce her husband and marry you. If that is permitted then the happy husband has to guard against it
all the time. That is a cost. By the exchange doesn't create more women; the gain to one man is exactly cancelled by the loss to the other. Socially it is net loss. Complications: the gain to the woman on finding a more suitable husband, the loss to the woman when the second man tires of her and divorces her and her first husband doesn't want her back. The distress of the children of the first marriage.
Exercises for the reader: (1)Gender swap the previous paragraph. (2)Add in the effects of alimony, child custody, and child support that we are willing to talk about. (3)Add in the dark side of child custody and support with women making false allegation to win custody and the money that comes with it.
Steel man two: Some people are so attractive that they have a choice of who to marry. Homo Economicus studies divorce law, and the effect of divorce incentives on mate choice is immediate. Normal people are influenced indirectly, by the marriages and divorces of parents, aunts, uncles, the parents of their friends, movies and novels about family life. Perhaps the eventual effect that filters through to normal people is the same as what Homo Economicus calculates so promptly, but with a lag of twenty to forty years.
Does a woman chose a dangerously sexy man, or a boring good provider? She doesn't ask the questions: what if he beats me? Trades me in for a younger model? Dies in a fight or motorbike accident? Bores me to despair? But all those scenarios play out in the romantic fiction she reads, and she knows which are the fantasy ones, and which are the ones that play out under real life divorce law.
If divorce is forbidden, people make safe choices. If divorce is permitted people take risks, and society gets to pick up the pieces. The usual argument in favour of divorce to end violent, unhappy marriages treats the rate of violent, unhappy marriages as exogenous. A rival argument is that the rate of violent, unhappy marriages is endogenous. The causal chain runs: permit divorce, more risky marriages, more violently unhappy marriages. Which leads to a tricky calculation: more bad marriages, but divorce ends a bad marriage, so many of the bad marriages as less bad because briefer. (complication: some of the safe-choice marriages turn out to be pretty bloody miserable)
That is not the correct calculation. There is inter-generational transmission of trauma; the children of bad marriages that break down, grow up to have bad marriages that break down. That adds another twenty to forty years of lag on the effects of changes to divorce laws filtering through.
Back in 1700 the considerations above would have been a little scandalous, because that way of thinking is atheistic. Say it out loud and people would think that you had read too much Thomas Hobbes. But they had a curious notion of providence, with faith in God being bolstered by the sense that He had handed down moral laws that worked out for the best. I think that they had their own ways of combining Theism and social-realism.