The Democratic primary once had an obvious and inevitable outcome, but now the process seems interminable. It is not delay or indecision, but division, and it could have devastating effect in November. Recent polls and history show this to be a real possibility.
ABC News/Washington Post recently polled Democratic-leaning adults and asked Clinton and Obama supporters whether they’d vote for their primary opponent in the general election. Thirty-six percent of Clinton supporters said they probably or definitely would not vote for Obama in November, while 38 percent of Obama supporters said the same of Clinton. An April 21 Quinnipiac University poll found similar results among Pennsylvania voters.
This might be a temporary symptom of a heated race, but the threats of defection are not idle. In the last 100 years, six elections have shown that the party faithful can make decisions that change presidential races. These movements take three forms.
The most focal, a real split, occurs when a party leader bolts and takes partisans with him. In 1912, Teddy Roosevelt lost the Republican nomination but then ran as an independent. He won more votes than Taft, his hand-picked successor, but propelled the Democrat Wilson to the presidency. In 1968, Wallace abandoned the Democrats to run as an independent. Wallace’s 10 million popular and 46 electoral votes helped Nixon win a narrow victory over Humphrey.
The second type, an effective split, happens when a third-party candidate disproportionately draws voters from one party. Ross Perot’s 1992 run attracted 20 million votes and benefited Bill Clinton, who won the presidency with the smallest portion of the popular vote — 43 percent — since Wilson in 1912. The 2000 election crammed a similar outcome into a smaller package: Nader’s quixotic third-party run polled just 3 million votes nationally but proved crucial in Florida, the election’s decisive state.
The final kind of voter movement — where a large group of voters simply fails to visit the polls — is a distinct possibility this year. Historically, two examples of suppressed turnout stand out. In 1988, then Vice President Bush totaled 5.6 million votes fewer than Reagan had in 1984. Dukakis polled 4.2 million votes more than Mondale had in 1988, but that leaves 1.4 million votes that vanished between elections.
The 1996 disappearance was far more pronounced, though it occurred within a third-party candidate’s totals. Perot garnered 12 million fewer votes than in 1992. While Clinton gained 2.5 million votes more than in 1992 and Dole got 100,000 votes more than Bush’s total. Thus over 9 million voters disappeared — or rather failed to reappear — between the two elections. The influx of 1992 became the reflux of 1996 and a lesson that rapidly energized voters may be as quickly unmotivated.
In four of these cases, these vote movements fundamentally altered the political landscape. Wilson’s 1912 victory ended a Republican stranglehold on the presidency — they had won in all but two of the previous 13 elections. Similarly, the 1968 election broke the Democrats’ presidential dominance — they lost only two of the previous nine elections, but won only three of the following ten. In 1992, Bush appeared unbeatable following victory in the Persian Gulf War and Republicans had lost only one of the previous six elections. In 2000, Gore could campaign on “peace and prosperity” and gain more popular votes, but he still lost the election.
Significant voter shifts within parties are not uncommon. And three current factors make such a shift possible, and potentially important, this November. The electorate is more volatile now; of the six cited examples of meaningful voter shifts, four occurred in the last 20 years. Contemporary voters demonstrate a marked willingness to “shift or sit.”
Second, the contest is close already. The 1988 and 1996 races, where apathy reduced turnout, were not close and voter absence was not decisive. In today’s race, a statistical dead heat, reduced turnout could be quite significant.
Finally, some Democratic poll respondents said they would not only refuse to support their primary opponent — they would actually support McCain. This doubles the effect of a defection. The partisan who sits, takes a vote away from a party. The voter who shifts to another candidate takes a vote away from a party and gives it to another.
An interesting nominating season could pale in comparison to the general election’s turmoil. McCain has already pulled off an amazing victory to secure the nomination. Another becomes ever more within reach the longer factors that could shift Democratic voters remain in play.
— J. T. Young served in the Department of Treasury and the Office of Management and Budget from 2001-2004 and as a congressional staff member from 1987-2000.