There's a slight inconsistency to Republican efforts to reform the electoral vote. For years, party activists have tried (with varying levels of effort) to dismantle the winner-takes-all system of electoral vote distribution in states like California, for the obvious reason that California is a big state that consistently hands its electoral votes to Democratic candidates.

In Nebraska, there's a new push in the opposite direction, as the New York Times reported over the weekend. Nebraska does precisely what many (Southern) Californian Republicans would like to see: distributes some of its electoral votes to the winner of each congressional district. In 2008, that meant that Barack Obama won an electoral vote in the state, thanks to those hippies in Omaha. So state Republicans want a winner-takes-all system. Politics is complicated.

What's interesting is that if every state in the union switched to a system that divvied up electoral votes based on the presidential results in each congressional district, the outcome of five of the last six elections would have been the same. And the one that would have been different isn't, as you might suspect, the hyper-close contest of 2000. It's the far-less-close race of 2012.

Several months ago, we pulled data on the presidential vote in each congressional district for a number of recent elections in an effort to show the decline in split-ticket voting; that is, the decrease in voters that chose a Republican House representative and a Democratic president, or vice versa. The graph below shows the 1998 and 2014 House results compared to the presidential vote two years prior. The closer to the yellow line, the stronger the link between the margin for a House candidate of one party and the presidential candidate of the same party. In 2014, that link was much tighter.

We'll come back to this. First, here are the results of each presidential election by congressional district since 1992. We've mapped each. Beneath the map is a pie chart showing the actual electoral vote, the number of House districts won by each presidential candidate (not the House seat itself), and the adjusted electoral vote total if votes were divvied up by the House district winner.

An important note on that. Nebraska has three House districts and five electoral votes, so it has two winner-take-all electoral votes and allocates the other three by House district (One electoral vote for each House seat.) The data below simply divide up the electoral votes by House winner, meaning, for example, that California's House districts each got 1.09 electoral votes (55 electoral votes, 53 House districts). And so on. This is why the electoral vote count in the adjusted version doesn't add to 538. It's rounding. Okay. On with it.

Notice that only in 2012 would the results of the election have been different.

Why? The same reason that Democrats aren't going to retake the House any time soon: A majority of House districts strongly favor Republican candidates. There are a lot of reasons for that, that we've explored any number of times before, including geographic sorting and gerrymandered district lines. And, of course, increased partisanship, as in that first graph.

In other words, shifting to an all-congressional-district-based electoral voting system wouldn't necessarily better reflect the will of the population. After all, Democrats (famously) got more votes in House races in 2012, but still lost the House badly. Of course, the existing electoral vote system doesn't always reflect the popular vote either. (See: Gore, Al.)

So what's the best system? Coin tosses, probably.