Kirby and Boaz establish that there are a lot of voters who can be described, with varying degrees of looseness, as libertarian, and these voters are not as a group reliable partisans of either party. Of course, we already knew these things. They also provide evidence suggesting that 73 percent of libertarians voted for Republicans in House elections in 2000, 70 percent in 2002, 53 percent in 2004, 54 percent in 2006, and 66 percent in 2008. That is, libertarians moved in nearly perfect opposition to the public at large, which was swinging toward the Republicans from 2000 through 2004 and against them from then through 2008. The same pattern holds in the presidential races. Bush gets 72 percent in 2000 and 59 percent in 2004 while McCain gets 71 percent in 2008.

Perhaps the lesson should be that parties seeking political success should run as far as possible from these voters? This is not of course the lesson Kirby and Boaz wish to draw. Their claim is that the libertarian bloc is a leading-edge bloc: Its move away from Republicans presaged the public’s, and the public may again be following its subsequent move away from the Democrats. They may be right to interpret the data this way, but they do not offer any particular reason for thinking so.

Boaz and Kirby note that Gallup has sometimes divided Americans into conservatives, populists (socially conservative and economically conservative), liberals, and libertarians. Conservatives have generally been the largest bloc. In 2009 the numbers were 31, 19, 18, and 23, respectively.

So let me be the master of the obvious here. Any Republican strategy for victory has to involve getting most of the conservatives, some of the populists, and some of the libertarians; the equivalent Democratic strategy building on most liberals requires a bigger share of the populist-plus-libertarian vote than the Republicans have to get to win. But how either party should appeal to the libertarians without scaring too many of the populists, or vice-versa, is the crucial question.