THERE WERE SIGHS of relief in Republican circles last week when the third quarter's economic growth rate was announced as 7.2 percent. But if the central political assumption of the Bush administration is true--that we are in the midst of a world war that is far from over--the relief may prove premature, if not irrelevant.

American voters behave very differently in wartime elections than they do in peacetime elections. The Democrats found that out in November 2002. In that election, by the normal criteria of economic data and domestic issue debate, Democrats had every reason to hope for at least modest gains in the House and Senate. But all year, polls indicated instead a status quo election, and toward the end of the campaign Democrats found themselves on the defensive on the war-related issue of homeland security. The result--historically surprising GOP gains, including the recapture of the Senate--suggested an electorate focused on war, not the sluggish economy or other domestic issues that mildly favored the Democrats.

The course of the Democratic nomination struggle this year suggests that the war-centered mood of the electorate hasn't changed. A candidate mainly associated with opposition to the Bush administration's conduct of the war, Howard Dean, has vaulted from the back of the pack to frontrunner status. The only other contender seemingly able to excite elements of the Democratic electorate, retired general Wesley Clark, has mounted a candidacy that will almost certainly live or die based on his ability to make sense of the war. If voters did not believe us to be in wartime, it's hard to imagine any real interest in a candidate who not only has minimal political experience, but who apparently entered the race having given little thought to the simplest questions of why he is running and what he believes.

For President Bush and the Republicans, strong economic growth would nicely complement a continued GOP advantage on war policy. But if the electorate sees Bush losing his grasp of the war, economic strength will not prevent political setbacks. In 1966 and 1968, Democrats were presiding over the sixth and eighth years, respectively, of the strongest economic expansion in U.S. history. But in those same years, a marked loss in confidence in President Johnson's conduct of the Vietnam war coincided with a 47-seat loss in the U.S. House in 1966 and with a decline in the Democratic presidential vote from 61 percent to 43 percent between 1964 and 1968.

If election history is any guide, the rules are simple. If America sees itself at war, war-related issues trump domestic issues when the two are in conflict. If a role reversal takes place--that is, Republicans gain an advantage on domestic issues while Democrats gain the upper hand in the war debate--Bush and his party will be net losers.

One of Bush's chief advantages since 9/11 is that his view of the war has seemed to coincide roughly with that of the electorate. He has seen the enemy as protean, resourceful, and crossing the usual sectarian, regional, and ideological barriers. He also sees the enemy as implacable and irredeemably evil. He believes we must put pressure on the terrorists and their rogue-state allies and facilitators all over the world before they come after us to inflict another 9/11, or worse.