There's an interesting post at the economics blog Marginal Revolution entitled What's Wrong With Polygamy? It includes a link to this paper in which three economists develop a model of the marriage market. The paper "examines why developed countries are monogamous while rich men throughout history have tended to produce polygyny (multiple wives)." When it comes to the evolution of social norms, I much prefer anthropologists, who are obsessed with cultural details and take them seriously, to economists, who ignore them. Thus we read, for example: "A key assumption of the model is that high quality men and women are more efficient in producing higher quality children, which generates a comparative advantage for high quality parents in raising higher quality children." That's culture to an economist: people are high-quality or low-quality.
Anyway, the interesting comments at Marginal Revolution deserve futher discussion: Polygamy does not contradict the idea of quality children, relative to available alternatives: the kids get papa's good genes and full-time attention from mama. If "good genes" are the only input papa gets, that's hardly high-quality parenting. And the real-world example I'm familiar with via reading, LDS polygamy in 19th- and 20th-century Utah, generally led to poverty as well as absentee fathers, hardly conducive to producing "high-quality children."
MR continues: Keep in mind if this is worse on average than other options, women won't want the deal. That assumes young women of marriageable age are free to choose, which is often not the case (although it is an assumption of the economic models of marriage, it seems). In patriarchal systems, patriarchs make important decisions, including who gets the young marriageable women or girls. If Papa thinks little Susie ought to marry Uncle Burt, the fact that little Susie doesn't want the deal might not carry much weight.
Next, MR states: If there is a social cost from polygamy, it more likely stems from the young men who cannot find wives and resort to violence and risky behavior. Okay, but let's not forget the young women. In a utilitarian maximization model, it seems like decreased happiness (or utility) of young women is also a social cost.
Finally, this: Polygamy ends when children cease to be a net economic asset. As society progresses and urbanizes, there are cheaper ways of having sex with multiple women, if that is one's goal. When I do this -- equating "polygamy" with a way of "having sex with multiple women" -- conservative Mormons get shrill very quickly. If MR is suggesting that polygamy as discussed in the paper or in general is (despite the preceeding economic context) simply a way to have sex with multiple women, then it seems the "marriage model" needs to incorporate sex drives, choices, and constraints. For example, one might hypothesize that power leads to greater sexual appetite, then in societies where polygamy is discretionary those higher up the social or hierarchical ladder would be more likely to enter into polygamy as a means of "having sex with multiple women." At least that's one way of incorporating MR's suggestion into the "marriage model."