I find it usually pretty difficult to disagree with the estimable Jonathan Bernstein, but here I think he is just well and truly wrong, and in a big way:

Indeed, somewhere north of 80% of all voters are basically reliable partisans in presidential elections. And you know what? That means that no one is interested in your vote, even if you live in the closest state in a desperately close contest. In other words, electoral college or no, your vote isn’t going to be the deciding one.

Yes, in the current system, you’re going to be totally ignored if you live in Texas, and massively pandered to if you live in Florida. But not really: you, the non-swing voter, are going to be totally ignored regardless of where you live. So why do you care whether it’s the swing voters in a handful of states or the swing voters in whatever areas get the best bang for the advertising buck?

And that’s before you get to the basic math of the situation: from the perspective of an individual voter in a nation of 300 million and more, no, your vote isn’t going to make a difference in a national election.

Now, yes, there are some policy implications of the electoral college and those matter, and fair enough. And yes, political is still to some extent geographical, so you’re more likely to meaningfully participate in ways other than voting if you live in Ohio than in New York. And perhaps that’s enough to make reform a good (if not an urgent) idea. Perhaps. But there really is something screwy in the assumptions of the anti-EC crowd about who “counts” and who doesn’t in different types of election systems.

This is hugely misguided, and here’s why – in 2008 there was a huge turnout gap between different states:

California – 61.2%

Iowa – 69.7%

New York – 58.3%

North Carolina – 66.1%

Ohio – 66.6%

Pennsylvania- 63.7%

Texas – 54.4%

Virginia – 67.5%

I’m sure we could get Nate Silver to do a regression analysis, but the pattern is still pretty clear to the naked eye – there is a major turnout gap between the states which were the focus of campaigning and those that were not.

The real point here is that in a purely popular national election, the entirely calculus is flipped. You don’t merely shift to targeting swing voters in non-swing states – the entire game becomes much more about turnout. Let’s say for argument’s sake that Democratic-leaning voters are less likely to turn out. So if you could increase turnout in New York, say, you would probably be gaining a lot of votes. What if Obama could campaign in New York – could he boost turnout to the level it was in, say, Virginia or North Carolina? That would mean more than a million more voters!

Basically, there’s a ton of voters that are basically ignored every four years due to the electoral college. And people who are ignored are less likely to vote. And that doesn’t mean only campaigning in non-swing states of your “color” – if you had a national popular election Obama would have just as much incentive to campaign in Texas as New York, presuming there are just as many Democratically-inclined voters in Texas who no-show as there are in New York. Maybe there are even more!

Anyway, the real injustice of the electoral college is that the entire theory that underlines it – that the key actors in choosing the President of the United States should be the arbitrary institution of the “state” as opposed to the popular will on a one-citizen one-vote basis – is backwards and unjust. I’m surprised this isn’t more obvious.