Building off yesterday’s skepticism that either party will become “merely regional” or whatever over the long-term, let me say that I think a lot of people engage in poor prognostication because the experience of Democratic Party domination from 1933-1968 is misleading. A 35 year period in which the Republicans had the White House for only eight years and did even worse in congress creates the impression that that sort of thing might happen again at some point. This, in turn, tends to generate a desire to write books like Building Red America, One Party Nation, or (on the other side) The Emerging Democratic Majority predicting that we’ll emerge from the rough muddle that’s existed since 1969 with decisive control for one side or another.
On a theoretical level, however, political parties are self-interested institutions in the business of winning elections. Meanwhile, there are only two of them. We ought, therefore, expect to see a lot of minimum winning coalitions, a lot of median-voter theorem behavior, out-of-power parties repositioning to fit demographic and ideological trends, etc. and all this should produce rough long-term parity. Before Civil Rights, on this view, the United States didn’t have a genuine two party political system, and that opened up the possibility of dominance. You can see something similar, perhaps, in contemporary Canada where the Liberals can lose (as they did at the last election) but the Conservatives can’t really win a majority, because a share of Liberal losses wind up benefitting the Bloc Québécois rather than the alternative national party.
The real long-term political question, on this view, isn’t about partisan dominance, but about the terms of equilibrium — how do both parties shift around to maintain rough parity and what are the implications of those shifts for policy.