Laura points out a good article from Rich Bolen, a Constitutional law scholar and a member of the Federalist Society, rebutting arguments against NPV, such as the basis of my skepticism, the impact on smaller states in presidential elections:
Objection: Won't small states be ignored if we eliminate winner-take-all rules and adopt the National Popular Vote?
Response: No, because even in states that are reliably red or blue a bigger margin of victory or smaller margin of defeat can offset losses in other states. Candidates will work hard in states they will definitely win so they can get out more of their supporters to influence the total number of votes cast for them nationally. In fact, in strongly supportive states, candidates can get a better return for their campaign investment.
During our meeting, we spent quite a bit of time on this question, and I'm still not entirely convinced it will have the intended effect — but I'm more open to the point than I was previously. States like Utah, South and North Dakota, and Oklahoma are usually so overwhelmingly Republican that candidates don't bother spending much time or effort there; the same is true for Democratic states like Vermont, Connecticut, and so on. Their winner-take-all EC policy means that if a candidate is more or less guaranteed to win it all without lifting a finger, then they won't — because the margin of overall victory doesn't matter. In an NPV system, every vote would count, and both parties will be forced to compete in smaller states to either maximize or minimize the eventual margin of victory — and that means every state will get attention.
Also, the NPV system would only "activate" under specific circumstances. If the popular vote winner doesn't get the EC victory, then each state would have to fulfill their requirement under the interstate compact to allocate their EC votes to the popular-vote winner, and only if enough states have joined the compact to get to 270 votes. (Interstate compacts are perfectly legal, by the way, when they involve only powers properly held by the states — and allocation of Electoral College votes is completely a state's decision.) But the possibility that an election will hinge on these outcomes will force candidates to campaign in all 50 states, NPV advocates argue.
I'm still at least somewhat skeptical. Adopting NPV essentially means that all presidential elections are popular-vote contests, which does make the EC less relevant while retaining its form and Constitutional finality. I'm not sure at all that the NPV will actually get candidates to spend time in smaller states, especially Democrats, who will use their GOTV votes in urban areas to build huge popular-vote leads in New York, California, and other coastal states in a cost-efficient manner that may not be replicable in exurbs or rural areas. However, the NPV advocates rightly note that the current system means fewer down-ticket resources applied in those areas in both parties because of the perceived lack of need for those votes now in national elections.
It's worth debating, but given the relative lack of crises in the last 134 years, it will probably be difficult to move states into accepting such a change.
