Pundits seem to be converging on a new conventional wisdom: that the drawn-out and extraordinarily competitive Democratic presidential primary race between Barack Obama and Hillary Rodham Clinton has cleaved the party in two. Many voters insist that they will not support any Democratic candidate in the general election except their original favorite, according to exit polls, and that has caused party elders to fret about whether the eventual nominee will be able to unify the party and defeat presumptive GOP nominee John McCain.
McCain may also have problems unifying his party in the fall, especially attracting conservative Republicans. In some recent primaries, as many as a quarter of Republican voters supported Mike Huckabee, who quit the race in March, or Ron Paul, who has stopped campaigning. GOP leaders worry that these voters might not turn out for the more moderate McCain.
Both parties can rest easy. Despite ugly battles and policy differences that sometimes seem intractable, the reality is that presidential campaigns tend to unify each party behind its nominee. Political scientists call this phenomenon the "reinforcement effect." It was described in 1940 in the first major study of a presidential campaign. The study's authors -- Paul Lazarsfeld, Bernard Berelson and Hazel Gaudet -- noted that voters tended to "join the fold to which they belong," with Democrats gravitating to Franklin D. Roosevelt and Republicans to Wendell Willkie. These voters were not blindly following whichever shepherds their parties nominated, the study concluded. Rather, their partisan loyalties reflected their underlying values, and the parties' nominees solidified their support by emphasizing these same values as the campaign unfolded.
The same thing happened in the 2000 presidential campaign, in which both Al Gore and George W. Bush emerged from contested primaries to be their parties' nominees. Early on, Democrats and Republicans appeared less than fully enthusiastic about their candidates. For instance, in June 2000, only 71% of likely Democratic voters said they would vote for Gore in the general election, according to the National Annenberg Election Study, a survey conducted by the Annenberg Public Policy Center at the University of Pennsylvania. The rest said they would either vote for Bush or another candidate, or were undecided. Among Republicans, only 82% said they planned to vote for Bush. But voters in both parties overcame or set aside their early doubts as the campaign unfolded. According to the Voter News Service's election-day exit poll, 86% of Democrats voted for Gore and 91% of Republicans for Bush. Most partisans rejoined the fold.
The chances are high that they will again this fall. Today's partisans are actually more loyal to their parties than they were in June 2000. According to a Pew Center for the People and the Press survey taken in late April, 77% of Democrats supported Obama over McCain, and 81% picked Clinton over McCain. As for the GOP, 85% of Republicans backed McCain against either Democrat.
And evidence suggests that the "reinforcement effect" is already underway. After Obama's big win in the North Carolina primary and better-than-expected showing in Indiana, a Gallup poll found increased support for him among most categories of Democrats.
But what about those Clinton supporters who say they won't vote for Obama in November? And what about those Republicans who still like Huckabee more than McCain? The existence of these voters certainly suggests that support for the parties' likely nominees could be less than overwhelming, but responses to exit polls conducted in primaries do not accurately predict how people will vote in November. In the 2004 New Hampshire primary, for instance, only 62% of Democrats who didn't vote for John Kerry said they had a favorable impression of him. By election day, however, ambivalence within the party seemed to vanish: 89% of Democrats who voted chose Kerry.
What about those white working-class voters Obama has had trouble attracting? Will they rejoin the Democratic fold in November?
They probably will because class differences have not divided Democrats in recent elections. For instance, in the 2004 election, 86% of white Democrats without a college degree voted for Kerry, as did 92% of those with a college degree. White Democratic voters who made less than $50,000 a year were just as loyal to their presidential candidate as those who earned more, according to the National Election Pool's exit poll. Democrats were largely unified across class boundaries despite Republican attempts to portray Kerry as an effete cosmopolitan out of touch with "real" Americans.
But if Obama, an African American, wins the nomination, as is expected, race could make the 2008 election different from previous presidential contests. There are certainly some white Democrats who won't vote for a black for president. An imperfect indicator is those Democratic primary voters who supported Clinton and said race was a factor in their decision. In the Kentucky primary exit poll, this group constituted 17% of all Democratic voters.
Nevertheless, in the voting booth, partisan loyalties may prove more powerful than racial prejudice. Benjamin Highton, a political science professor at UC Davis, studied 357 contested House races in the 1996 and 1998 elections. He found that white voters were no more or less likely to support black candidates than white ones. Prejudice against blacks still exists in many forms, but it does not guarantee that large numbers of Democratic voters will abandon an African American nominee for his white Republican opponent.
The reinforcement effect is especially likely to be strong this year, for two reasons. One is that partisan loyalties are increasingly salient to voters. Over the last 50 years, the proportion of Americans who identify with or lean toward either the Democratic Party or the Republican Party has held steady or slightly increased, according to the American National Election Study, the longest-running academic survey of political attitudes. Currently, 90% of the electorate identifies with one or the other major party. And during this same period, partisans have become increasingly loyal to their party's candidates, as in 2000 and 2004.
The second reason is that the Democratic and Republican parties have become more polarized, especially during the Bush administration. Democrats and Republicans are more divided in their views of President Bush and the Iraq war than they have been for any other president or war, according to data compiled by Gary Jacobson, a political science professor at UC San Diego. For instance, in a May Rasmussen poll, 70% of Republicans, but only 11% of Democrats, approved of the president. Given this gulf, most Democrats and Republicans are unlikely to find the opposing candidate attractive, especially after that candidate is criticized by their party's nominee for months on end.
The reinforcement effect, of course, doesn't guarantee victory for either party. In a close election, even a little party disloyalty can bring down a nominee. But history suggests that the discord we read about in the two parties today will fade, and Democrats and Republicans will coalesce around their candidates as Nov. 4 approaches.