Ramesh Ponnuru observes (with a caveat) that while Republican congressional candidates lost significant support from libertarian-leaning voters since the last midterms, they actually gained relative to 2004. From this, we’re meant to infer that even if libertarians are up for grabs, they’re hard to appeal to, since they’re just congenitally truculent folks who will always be moving in the opposite direction from the broader electorate.

This seems a bit hasty. I think a more plausible explanation would be that libertarians were probably more apt to turn against the war earlier than the modal voter, and so you had higher left-libertarian turnout in 2004 among people voting against Bush. More generally, I’d expect that in areas like the West, you’d find plenty of libertarians who are happier with the kind of less socially conservative Republicans who win office there than with the Republican Party generally. So even absent the war, I’d think whether it’s a presidential election year would make a difference. It seems to make more sense to compare midterms with midterms.