The 2004 presidential election: the emergence of a permanent majority?

The 2004 election results are best viewed from the perspective of incumbent-versus-challenger evaluations and of retrospective and prospective evaluations. Incumbent President George W. Bush survived Senator John F. Kerry's challenge mainly because he convinced a small majority of voters that he had done a good enough job and that the prospects for positive government performance were sufficiently good that he should be reelected. This produced, according to the major exit poll, summary evaluations by the electorate that were narrowly favorable to Bush and narrowly negative for Kerry. These evaluations were closely reflected in the division of the vote.

The most distinctive feature of the presidential election of 2004 was its similarity to the election of 2000. The change in the popular vote was small, as was the change in the pattern of outcomes in the states and the resulting electoral vote. Once Kerry had effectively secured the Democratic nomination for president, he and Bush were locked in a close contest. Throughout the spring and early summer, Kerry trailed Bush by one or two percentage points in the polls. (1) He even held a very small lead over Bush from July until about the middle of August. Bush got a healthy convention "bounce" to a lead of three or four percentage points, but Kerry's stronger performance in the debates narrowed Bush's lead to an average across the various polls of 50.9 percent for Bush to 49.1 percent for Kerry on 31 October, very close to the final result. To be sure, Bush picked up about two points in the major-party vote from his first to his second election, but that is very little gain for an incumbent successfully seeking reelection.

Not only were the last two presidential elections very close, but they also shared a great many other similarities. Perhaps most obviously, only three of the fifty states switched sides in the two contests, and these involved small changes in vote percentages that tipped two states (Iowa and New Mexico) that narrowly voted Democratic in 2000 to voting Republican in 2004 and one state (New Hampshire) that narrowly voted Republican in 2000 to voting Democratic in 2004. Bush's electoral vote increased from 271 to 286, and eight of the fifteen votes he gained came from switches among these three states. The remaining seven-vote gain resulted from a net increase of House seats among the twenty-nine states that he won in both 2000 and 2004.

Still, the major story about the 2004 elections is its narrow Republican victories. There was very little change in the House vote. The Republicans picked up four additional seats, but these gains resulted from a partisan redistricting plan in Texas that was implemented in 2003. The Republicans also picked up four Senate seats, giving them a secure, but not filibuster-proof, majority. Most Senators won reelection handily, but the GOP picked up five open Democratic seats in the South, giving them a large majority in that region. These victories came in states that Bush won handily as well, so it is possible that, in this precise and specific way, Bush had presidential coattails.

In addition, most groups voted in 2004 much as they did in 2000. The most notable exceptions were Hispanics, who supported Kerry less strongly than they did Al Gore in 2000, and a shift of women to Bush that reduced the "gender gap" in 2004. The candidates campaigned in much the same way as well. They traveled to, spent advertising money in, and focused campaign organizational efforts on a small number of "battleground" states-- almost exactly the same battleground states in 2004 as in 2000. This meant that the residents of most states saw little of the presidential contest in either election. Except for the presidential and vice presidential debates, they did not see the candidates, hear their speeches, or see or hear their ads (what became known as the "air war" in the battleground states); they scarcely witnessed even much of the national campaigns' organizing of registration and turnout drives and other lower-profile aspects of campaigning (campaign flyers, telephoning, radio and other local advertising, and so on) that make up the "ground war."

Despite the similarities, important differences, particularly in electoral context, can be found between the two years. The impact of September 11, of two "hot" wars, in Afghanistan and in Iraq, and of the war on terror at home and abroad meant that no candidate could long ignore peace and security. Incumbency had switched between the parties. And although incumbent Vice President Gore could run a campaign that was surprisingly distinct from the Clinton administration, Bush could not have run as anything other than the incumbent president. Gore distanced himself from Bill Clinton and the Clinton-Lewinsky scandal, but as a result, he also distanced himself from a nearly eight-year economic boom, although one that was beginning to end. Bush could not distance himself from the economy, because presidents are held accountable for its performance. At the start of the year, the Democrats were happy to remind the public of the President's accountability. As the year wore on, Bush's claim that the economy had turned the corner was reflected in the hard indicators of economic performance and, then, increasingly in the eyes of the public. To be sure, he was not riding a boom, as was Clinton in 1996 or as was Ronald Reagan in 1984, but the voting public was, on balance, slightly positive about the economy-maybe slightly, but still just enough for Bush to avoid a downfall like his father's twelve years earlier.

These differences in electoral context--foreign affairs and the economy--lend themselves to incumbent-versus-challenger assessments and, as will be discussed below, to retrospective and prospective evaluations. These contrast with the common interpretations of the 2004 presidential election, at least those offered by pundits in the first months after the voting. Pundits made much of the "red state--blue state" distinction, that is, cleavages between Republicans and Democrats, between conservatives and liberals, to discuss the growing polarization that has characterized congressional politics and national party politics in general and that has, therefore, linked Republicans with conservatism and Democrats with liberalism. To be sure, the two major candidates were, at one and the same time, an incumbent facing a challenger and a Republican opposing a Democrat. It is sometimes difficult to distinguish whether a difference in the vote is tapping more of a distinction between incumbent and challenger or more of a Republican--Democrat difference. However, there are good reasons to suspect that the public chose, this time, between Bush and Kerry more in terms of their evaluations of how well the incumbent performed and how well they thought the challenger might do than in terms of party and ideology. Although both were important components of voters' decisions, we think the former outweighed the latter.

REGIONAL VOTING PATTERNS

As noted above, the 2004 presidential election results were very similar to those of the 2000 election. There was a shift toward Bush of 2.8 percent of the total vote. These gains resulted, in part, from the reduced share of the vote for third-party or independent candidates. With 99.0 percent of the vote going to the major-party candidates, compared with 96.2 percent in 2000, Bush was able to score gains in forty-eight of the fifty states and in the District of Columbia. But Kerry's share of the total vote also increased over Gore's in twenty-six states and DC.

The similarity between the 2000 and 2004 results raises the question of voting patterns over a longer time span. Sixteen states with 135 electoral votes have voted Republican in all seven presidential elections between 1980 and 2004. These include six of the Southern States (by which we mean the eleven states of the old Confederacy), all four Prairie States, three of the eight Mountain States, and Alaska, Indiana, and Oklahoma. During this quarter of a century, only Minnesota and DC, with thirteen electoral votes, have remained equally loyal to the Democrats. Indeed, only eight states and DC, with ninety-two electoral votes, voted Democratic in all five elections between 1988 and 2004. (2)

Neither party has an electoral vote base adequate to ensuring victory, and in the wake of short-term forces, either party can win. Although the Republicans have been advantaged since 1980, in 1992, Clinton was able to defeat George H.W. Bush by a 370- to 168-vote margin, winning such "red" states as Arkansas, Colorado, Louisiana, Nevada, and Tennessee. In 1996, Clinton defeated Bob Dole by a 379- to 159-vote margin, winning Arkansas, Louisiana, Nevada, and Tennessee. He even won Arizona, the only state to have voted Republican in all eleven elections between 1952 and 1992.

The Republicans now rely heavily upon the South, and the transformation of the South from a solidly Democratic region to a Republican region is the most dramatic change in postwar American politics. The Republican strength in this region is impressive, especially given Democratic dominance in the South from the end of Reconstruction through the end of World War II.

But Republican dominance in the South displays a dramatically different pattern in 2004 than in the days of the solid Democratic South, when the Democrats usually won most Southern States by substantial margins. Figure 1 presents a map showing Bush's margin of victory over Kerry in the thirty-one states that Bush won. Figure 2 shows Kerry's margin over Bush in the nineteen states (plus DC) that Kerry carried.

[FIGURES 1-2 OMITTED]

As Figure 1 shows, Bush's victory in the South was strong, but he was strongest in the Prairie States. He also did well in the Mountain West, although he carried New Mexico and Nevada by narrow margins. Figure 2 contrasts markedly with Figure 1. Although Bush won fourteen of his thirty-one state victories by over twenty percentage points of the total vote, Kerry carried only three of his nineteen state victories by this margin, as well as DC. Although he carried all the Mid-Atlantic States, he won with a fifteen- to twenty-point margin only in New York. And of the four Midwestern States bordering the Great Lakes that he won, his margin in three of those four was less than five percentage points. Like Democrats in the last four elections, he carried all three of the Pacific Coast states, but only in California did he approach a ten-point margin.

Obviously, there is considerable variation among the states, but we would caution analysts not to exaggerate its extent. For example, one can construct a measure of regional voting by taking the percentage of major-party voters in the South who voted Democratic and subtracting the percentage of major-party voters outside the South who voted Democratic. In 1944, the South was …