Writers in both National Review and the New Republic have dismissed David Kirby’s and my warning that Republicans are losing libertarian voters by noting that President Bush’s percentage of the vote went up in 2004 even though he lost libertarian votes. Thus, Ramesh Ponnuru and Jonathan Chait say, losing libertarian votes is no problem for the Republicans.

In National Review, Ponnuru writes:

The electorate as a whole swung toward Bush during those years: He increased his percentage of the overall vote from 48 to 51. Libertarians swung one way; the remaining 85 percent of the electorate swung the other way, and swung far enough to overwhelm the libertarians.

In the New Republic Chait agrees:

Boaz and Kirby …stress that President Bush’s share of the libertarian vote dropped precipitously between 2000 and 2004. But, during that time, Bush’s total share of the vote rose by almost 3 percent.

It’s true enough that Bush increased his percentage of the total vote even as libertarians were swinging away from him. But Chait and Ponnuru would have us believe that Bush succeeded because his policies alienated libertarians and appealed to a larger group of non-libertarian voters. But what policies would those be? Did he achieve re-election on the strength of the war in Iraq? His massive over-spending and prescription drug entitlement? His support for the gay marriage amendment? Not likely. (For a discussion of state marriage amendments and the 2004 vote, see here.)

Indeed, the large question about 2004 is why a president with a strong economy won only 51 percent of the vote, 6 points behind what economic models of presidential elections predicted. The biggest answer is the war in Iraq, which was increasingly unpopular by November 2004 and which likely turned off both libertarians and other independent and centrist voters.

Meanwhile, along with the economy, what accounted for Bush’s gains from 2000 to 2004?

It’s terrorism, stupid. The most important number in the 2004 exit polls was this: 58 percent of respondents said they trusted Bush to handle terrorism, while only 40 percent trusted Kerry. You can’t win a post‑9/​11 election if only 40 percent of voters trust you to protect them against terrorists; people may not have been happy with the war in Iraq, but many of them thought terrorism was the bigger issue. Indeed, our study found that libertarian-leaning voters who cited “terrorism” as the most important issue in 2004 voted heavily for Bush, while those who cited some other issue gave a majority of their votes to Kerry.

And of course, our post-election 2006 data found that libertarians again gave Democrats a larger share of their votes than they had historically done. And this time it did cost the Republicans. Independents–many of them libertarian-minded–turned sharply away from Republican candidates. Disgruntled libertarians probably cost the Republicans congressional seats in New Hampshire, Montana, Arizona, and Colorado, Nevada, and Iowa, and possibly also in Florida, Ohio, and Pennsylvania.

If Republicans can’t win New Hampshire and the Mountain West, they can’t win a national majority. And they can’t win those states without libertarian votes. This may be good news for Democrat Chait. But Ponnuru should worry about it.