Site Presented By
Saturday, Jul 11, 2009
A look at party loyalties shows that Kerry has a much better chance than Bush of persuading undecided voters.
Jul 21, 2004 | There are many reasons for the Democrats to be hopeful heading into Boston next week, but the most important of these may be that the Bush campaign has maximized its potential and trails in the polls. There is a boundary to the limits of any political coalition, and the Bush-Cheney campaign is near the edge of its electoral reach.
The Bush campaign has mobilized its core base of conservative white male Republicans very effectively. Now what? Now is when Karl Rove wishes he were Mary Beth Cahill, John Kerry's campaign manager. From nearly every angle that the Bush strategists peer, the turf they view for expanding their coalition is decidedly less friendly than the landscape enjoyed by Team Kerry.
Let me begin with a notion that is very plausible in today's red-state/blue-state, 50-50 nation: This election will be determined primarily by party loyalty. Even two-thirds of the voters who call themselves independents lean toward one of the major parties, leaving 10 percent or fewer voters who are truly independent of party loyalties.
Over time, voters with differing intensities of partisan loyalty tend to vote with varying levels of consistency in favor of their preferred parties. Examining those patterns of partisan loyalty across a span of elections permits one to estimate the likelihood of a Democratic or Republican vote for any such class of partisan identifiers, including true independents. The academic pedigree of this model lies in the work of prominent political scientist Philip Converse of the University of Michigan during the 1960s; its current major revisions and applications belong to the University of Missouri's John Petrocik. My model shares the basic concept but departs methodologically from its prior calculation.
For the nation as a whole, this model of expected votes, premised on the assumptions above, predicts a virtual tie vote in the 2004 presidential election: 50 percent for the Democratic candidate and 50 percent for the Republican candidate. The model basically relies on partisan benchmarks: It assumes that the partisan predispositions of voters will not be dramatically disrupted by seismic shifts on major policy issues, the nomination of a particularly odious major-party candidate, or catastrophic external events that turn the world topsy-turvy.
How does this model suggest that the Bush-Cheney campaign is running out of welcoming, available votes? It does so by estimating the expected Republican vote from key voter groups and measuring President Bush's performance in the most recent polls among those groups. The difference between the expected Republican vote and Bush's observed vote tells a great deal about how well Bush is doing and what his prospects are.
Consider the breakdown of Bush's support by party identification presented in the most recent Newsweek poll: 90 percent among Republicans, 10 percent among Democrats, and 34 percent among independents. The partisan benchmark model estimates the expected Republican vote as 91 percent among Republicans, 12 percent among Democrats, and 49 percent among independents.
Bush nearly matches expectations among Democrats (at 10 percent, only two points below where he should be) and Republicans (at 90 percent, only one point below expectations). But with only 34 percent support among independents, Bush is running 15 points below the objective for a Republican hoping to capture 50 percent of the national vote. Bush's task is thus to make huge inroads among an amorphous group of voters, most of whom do not align with his party.
The same poll shows Kerry with 83 percent of the Democratic vote, 6 percent of the Republican vote, and 53 percent of the independent vote. These numbers suggest that Kerry is running reasonably well -- two points better than expected among independents, three points worse than expected among Republicans, and five points worse than expected among Democrats. The latter figure should give his campaign pause, however, as Kerry should be doing better among Democrats. And it also highlights the value of a meaningful baseline. While 83 percent sounds pretty good, the baseline value informs us that Kerry is not doing as well as he should be within his own party.
But the baseline measure also provides insight about comparative strategic hurdles. Kerry is running behind expectations among the one group that ought to be, and likely will be, most receptive to him as the campaign unfolds. So Kerry needs to make small gains among friendly voters, while Bush needs to make huge gains among relatively unfriendly voters.
A similar picture appears when the focus shifts to ideological orientations, as in Zogby's latest national poll.
Kerry nearly matches expectations among moderates and is running six points better than expected among liberals and five points worse among conservatives. By contrast, Bush reaches expectations among conservatives, with 76 percent of their votes. His shortcomings rest with moderates (six points below expectations) and liberals (11 points below expectations). A typical Republican should expect to do no better than Bush is doing among conservatives.
The votes Bush needs are to be found among moderates and liberals -- hardly an auspicious prospect for him. It is possible, but unlikely, that Bush will amass enough additional votes from conservatives to make up for this deficit.
