Disclaimer. Don't rely on these old notes in lieu of reading the literature, but they can jog your memory. As a grad student long ago, my peers and I collaborated to write and exchange summaries of political science research. I posted them to a wiki-style website. "Wikisum" is now dead but archived here. I cannot vouch for these notes' accuracy, nor can I even say who wrote them. If you have more recent summaries to add to this collection, send them my way I guess. Sorry for the ads; they cover the costs of keeping this online.
Roeder. Power dividing as an alternative to ethnic power-sharing. In Sustainable Peace: Democracy and power After Civil Wars, eds. Roeder and Rothchild. Ithaca: Cornell University Press [forthcoming].
An excellent critique of (Lijphart's) power sharing on theoretical and empirical grounds. Points out many potential problems of power sharing as an institutional mechanism (x) to prevent escalation of ethnic conflict (y). Roeder sees three forms of democracy, not just two (like Lijphart does): Westminster majoritarianism, Consociational supermajoritarianism, and power-dividing multiple majoritarianism. He argues that power-dividing has several theoretical advantages over power-sharing in divided societies--it is likely to be a more stable institutional arrangement, it will reduce incentives to focus on ethnic concerns, it will help other issues become salient, and so on. See my notes in the margins for a quick summary. Interestingly, Phil outlines a middle ground in the "efficiency" vs "representative" debate. Usually, we think of Westminster systems as efficient and consociational systems as representative. In Phil's system (basically the American system), less contentious subjects are handled efficiently (i.e. delegated to an agency for regulation by a committee using majority rule) and more important subjects are handled representatively (with consensus by at least three majorities: Senate, House, President).
Ethnofederalism is bad (p 16), but federalism can be good if it creates new cross-cutting cleavages, as in Inda (p 20).
Even in cases where power sharing appears to work (and in most cases it has failed), it works only because power sharing is only a small part of a broader power dividing regime (12, 18).
The statistical test is very nice. For a quick summary, see Figure 3.1 (page 33). Note that there are two dependent variables (different measures of the same thing); one is dashed and the other is solid. Power dividing institutions have a substantively strong effect on reducing the occurrence of ethnic violence and escalation.
He ends with a nice normative argument (35) showing that, in this case, what is pragmatic (power dividing) is also strongly in accord with liberal theory [unlike power sharing/consociationalism, which has been accused of being less than liberal].
This draws heavily on Madison and the US experience.
Research on similar subjects