How come when studies look at the impact of divorce on children they compare the outcomes of children of divorced parents with the population as a whole? Surely they should compare the outcomes of children of divorce with the outcomes of children of unhappily married parents?
Or maybe I'm missing something and there is such a study and I've missed it? Always possible 
My parents are divorced and for me parental unhappiness is a massive burden far beyond the simple logistics of who lives where.