There’s been a pretty strong reaction by some bloggers on the Internet today (see here and here) because state legislators in Pennsylvania are considering legislation to change Pennsylvania’s electoral college votes from winner-take-all to a congressional district system: one electoral college vote for each district won, plus two votes for winning the state at-large (this conforms to the Constitutional structure — total representatives + Senators — of the state EC vote distribution). The partisan implications are obvious; whichever presidential candidate was likely to win the state under the at-large system will come out a loser, as at least some proportion of the congressional districts will give their vote to the other candidate. This is likely to hurt the Democrats, as they have won Pennsylvania in recent elections and, at any rate, Pennsylvania’s 20 electoral votes (back then 21) are part of the math of most plausible Obama maps.
I’m not going to rehash all of the usual arguments about the value of the winner-take-all vs. CD system or the consequences of changing from one system to another nation-wide. You can read about them in this decent political science piece if you like. I will, however, make three points that I have not seen mentioned:
1) The winner-take-all direct vote system used in almost all states is the product of two strong historical forces. The Constitution is very clear that each state may choose its electors in any way the state legislature directs. This includes the winner-take-all direct votes system, the congressional district direct vote system, a division by proportional statewide vote, a vote of the state legislature, a hand-picking by the governor, a caucus system, a vote by state legislative districts, or anything else you can dream up. It’s not even out of the realm of possibility that the system could contradict other portions of the Constitution. If a state chose to let only people over 40 participate in a direct vote system, it’s not at all clear to me that the federal courts would intervene (it might very well be struck down in state court). The Constitution clearly leaves the decision up to the state legislature.
So why then do all but Nebraska and Maine use the winner-take-all direct vote system? For two reasons. First, the spread of universal suffrage and mass democracy in the early 19th century put pressure on the states to shift the vote from the legislature (which all states had been using as the selection methods in the first two elections) to the people. This trend accelerated in the 1820’s, and by 1828 only Delaware and South Carolina were still picking electors in the state legislature. (South Carolina did not switch to direct vote until after returning from secession). Second, using a winner-take-all system maximizes the leverage of the state. Candidates need not pay much attention to a state in which the best they can do is win, for example, 6 of 10 electoral votes and the worst they can do is win 4 of 10 electoral votes. In effect, such a state has reduced its electoral votes to 2, and becomes proportionally less important, which in theory means candidates will be less likely to bend policy promises toward the state’s electorate, and definitely means local and state politicians will have fewer chances to appear on stage with the national candidates. Thus individual states, jealous of their political clout within the federal system, had strong incentives to make the electoral votes winner-take-all.
2) I suspect we will see (a few) more congressional district systems in the future. Once upon a time, individual states were less diverse. That is, the interests of the people of the state were narrower and more uniform. This is a lot less true now than it was 100 or 150 years ago. In fact, the sorting of the political parties by ideology has accelerated the trend such that conservatives in state X are much more like conservatives in state Y than they are like liberals in state X. Consequently, it’s not obvious that partisans in any state still have reason to maximize the state leverage in the federal system by using a winner-take-all electoral college system. In effect, national partisan concerns are now far more dominant than state or local concerns, partisan or otherwise. While 150 years ago, the national parties were loose confederations of state parties that had disparate underlying constituencies and policy goals, almost the exact opposite is true now: the state parties are, using the broadest brush, administrative units of a functionally nationalized party system. This overstates the case a bit, I know, but the logic still holds: the national parties have no interest in winner-take-all systems in states they won’t win, and without independent and diverse state parties to constrain things, it is likely that we’ll see more tinkering with electoral college systems in the future.
3) That said, I don’t expect Pennsylvania to change it’s system. Nor do I see a lot of states going to a congressional district system in the near future. While the overarching conditions for maintaining the winner-take-all system are crumbling, the necessary political context to trigger a conversion to a congressional district system is rare. The only time a legislature would theoretically consider a shift would be when either:
(A) they honestly think a congressional district system is better on the merits; or
(B) when they see political advantage in doing so.
Or both. But the problem is that choice A is unrealistic; institutional change in political systems is rarely driven by normative concerns, instead usually occurring as the byproduct of decisions that maximize political interests. But the conditions that make choice B hold are not common: you need a situation in which the state population is likely to vote for candidate A, but the entire state government is controlled by the party of candidate B, and the party of candidate B believes they will not be badly punished by the voters for supporting such a change.
Right now, the second condition holds in Pennsylvania: the GOP controls the state government. The working assumption is that the first condition also holds — that President Obama is likely to win Pennsylvania. That, however, is not a slam dunk in 2012; the GOP could easily end up costing itself electoral votes by shifting to a congressional district system (although, it should be noted, if President Obama cannot win Pennsylvania, it is unlikely to be the difference in the election). And the third condition is the most dubious: in a world where President Obama wins Pennsylvania in 2012, I think it highly probable that at least some proportion of GOP state legislators (and perhaps some Pennsylvania GOP House Members) will pay an electoral price for supporting a change to the system. Which could make some legislators gun-shy about supporting the change. Therefore, I think it highly unlikely that such a change occurs this cycle.
This is not to say that such changes are not possible. But my guess is that the factors mentioned above will attenuate most impulses to change the systems.