The intended comparison is presumably not be to all marriages, but to marriages which are strained enough for divorce to become a question. I don’t know what go

The intended comparison is presumably not be to all marriages, but to marriages which are strained enough for divorce to become a question. I don’t know what good data on that would look like. Perhaps ‘Outcomes for children of parents who went to marriage counselling’ Even then, couples that get counselling are probably a healthier subset of all marriages than would make for an ideal reference.