My launching off point is the view that the extreme polarization of the political parties is a product of long-term historical transformations; thus, there is not a lot of room for optimism that this polarization can be meaningful changed by various institutional reforms that have been proposed (including by me, eg, here). Empirical studies cast increasing doubt on whether the conventional reforms proposed would make a dent in the polarization of our parties. Thus, I suggest considering a different tack.
Here is an excerpt:
My suggestion is that, if we are looking for solutions, we should re-define the problem of effective governance in our era as one of political fragmentation rather than one of political polarization. By fragmentation, I mean the external diffusion of political power away from the political parties as a whole and the internal diffusion of power away from the party leadership to individual party members and officeholders. It is political fragmentation that makes it that much more difficult, in a political world that rests on polarized parties, for party leaders nonetheless to engage in the kinds of negotiations, compromises, and pragmatic deal-making that enable government to function effectively, at least in areas of broad consensus that government must act in some way (budgets, debt-ceiling increases). And because of political fragmentation, party leaders in all our political institutions have less capacity to play this kind of leadership role than in many previous eras. When political fragmentation that makes it that much harder for party leaders to command their parties is added to highly polarized parties, the mix is highly toxic to the capacity of our political institutions to function effectively.