Rick Hasen has a good column about about the national GOP effort to get Republican-controlled legislatures in states which vote for Democrats in presidential elections to switch to apportioning electoral votes by Congressional districts. Mainly I agree with his bottom line: it's unlikely to go anywhere. In particular, he focuses, correctly, on some of the ways in which the incentives for the legislators and governors involved run against the incentives of national presidential candidates.
Hasen is certainly correct that there's nothing illegal or unconstitutional about the scheme. I don't really think he's right that Democrats have been overly hyping the thing, though; even if it's both legal and unlikely to happen, it's still pretty outrageous.
But it's worth going over again just how tricky the math of this scheme is.
Remember, while a national electoral-vote-by-Congressional-district plan would strongly favor Republicans, there's virtually no chance it could happen. So we're talking state-by-state "reform," which in practical terms means states with partisan control. Not only that, it means states with Republican partisan control which support Democrats in presidential elections. There aren't many of those!
Indeed, the 2010 GOP landslide sandwiched by two good Democratic years set up the rare situation in which there actually are several states which sort of fit. Which leads to columns noting how if all of them -- FL, WI, VA, PA, MI, and OH -- went for it, Romney could have defeated Obama.
But here's the thing. Outside of all the other reasons that it's unlikely, the math of doing it in a few selected states gets weird quickly.
That's because we don't know, in advance, exactly how the next election is going to turn out; nor do we know the exact rank of the states (that is, from most to least Republican).
The thing is that splitting the electoral votes is double-edged: it helps Republicans if Democrats win the state, but it helps Democrats if Republicans win it. That's true even if there is a strongly Republican bias; even if Republicans gain a lot more if Democrats win than Democrats do if Republicans win (say, a state with 20 electoral votes in which the districts will likely produce a 12-6 GOP edge in a close race), it's still a real problem for Republicans. That's the complexity of the situation that, say, Micah Cohen really overlooks.
See, the more Republican the state, the more Republicans would risk giving away electoral votes. But if they only do it in relatively safe Democratic states -- some subset of the six states mentioned above -- then the total electoral vote haul is less likely to make a difference.
And, again, we don't know the rank-order in advance. Could Ohio go Republican while Virginia and Florida go for the Democrats? Sure. And if Ohio is the only state that buys in and then goes Republican in an election which would have yielded a very slim GOP electoral vote edge, it could easily turn instead into a Democratic win. It's even worse...in order to do this the GOP needs to act now, before the risk that midterm elections make it impossible to do in some states -- but that also opens up the risk of a big Democratic win in one or more state followed by a flip back, thus potentially leaving the scheme in a "wrong" combination of states (such as only the most Republican-leaning one).
Not to mention that, as I said in the earlier piece, the people who actually have to support this -- elected Republicans -- are very likely to take the very fact of their election as a reason to be believe that their state is trending Republican, and therefore passing the measure would be counterproductive! When it comes to their own elections, I'm perfectly willing to believe that politicians are paranoid and would want the largest possible electoral advantage in their district; that's why bipartisan, incumbent-protection gerrymanders are more common than partisan, seat-maximizing gerrymanders.
But the point here is that even if state Republicans were perfectly willing to ignore their own incentives and instead do whatever the national party believed was best, it still would be extremely difficult to game out the proper combination of states in advance. If they could do the entire nation, then it would be easy. But since that can't be done, what remains just isn't very promising.