Diverging trends in national and state polls are adding another level of uncertainty to the closest presidential race in decades--and inspiring speculation that Al Gore might lose the popular vote yet still win an electoral college majority that places him in the White House.
Over the last several days, national polls have largely converged, providing Republican George W. Bush a small but steady lead of 1 to 5 percentage points. Yet the latest surveys in critical battleground states still show Gore running well enough in most of them--particularly Michigan, Pennsylvania and Florida--to remain within reach of the 270 electoral votes needed for victory.
The disparity between the national numbers and the state results reflects an intriguing dynamic in the race: Bush is generally further ahead in states where he leads (especially in the South and Mountain West) than Gore is in the states where he leads. That pattern enlarges Bush's margin in polls measuring the popular vote, even as polls in the key battleground states show a closer race or give Gore an edge.
Most analysts believe that in the end the closest states will tip toward the winner of the popular vote, as they usually do. But the split between the national and state numbers opens the outside possibility that Gore could lose the popular vote yet still squeeze out an electoral college majority with narrow victories in several states, especially the pivotal trio of Pennsylvania, Michigan and Florida.
"There's a 30% chance that Gore loses the popular vote and wins the electoral vote," Democratic strategist James Carville said.
In a paper released Thursday, two Columbia University political scientists even calculate that Gore could win an electoral college majority as long as he stays within 2.2 percentage points of Bush in the popular vote.
"Gore would be the favorite as long as he doesn't lose by more than that amount," co-author Robert Erikson said. "The reason is that most of the battleground states are a little bit more pro-Gore than the nation overall."
More Than a Century Since Last Split Verdict
There hasn't been such a split verdict since 1888--when Democratic President Cleveland won the popular vote but lost the electoral college decisively to Republican Benjamin Harrison. Similar split outcomes occurred in the 1824 and 1876 presidential elections.
But today it's unknown how Americans--many of whom are only dimly aware of the electoral college's workings--would react to seeing the loser of the popular vote occupy the Oval Office.
Thomas E. Mann, a political scientist at the nonpartisan Brookings Institution think tank in Washington, predicts that, while the public would accept the results, there would be an immediate demand to replace the electoral college with a direct national vote.
"We live in such a plebiscitary world, a world very different from 1888, that I think you simply can't sustain" the current procedure, Mann said.
USC law professor Erwin Chemerinsky said no legal basis could be found to challenge an election in which one candidate won the popular vote and the other won the electoral college.
"The Constitution says that it is the electoral college that determines who is the president," Chemerinsky said. "It leaves no doubt."
While concurring in that assessment, one prominent Republican lawyer said a lawsuit might still be filed as part of an effort to pressure the actual electoral college "electors" to switch their vote to the popular-vote winner.
Nothing in the Constitution requires the electors, who meet in December, to vote for the winner of their state. About half the states have legal penalties for electors who disregard the results in their state, but no one has ever been prosecuted for doing so, according to a recent analysis on the political Web site Voter.com.
Electors Usually Are Party Loyalists
In fact, seven times in the last 52 years "faithless electors" have deviated from their state results to back another candidate. In 1976, for instance, one elector from a state carried by President Ford voted for Ronald Reagan. If Bush won the popular vote but lost the electoral college, Republicans might launch a campaign urging individual electors to switch, some believe.
Such a campaign would face long odds, though, since the electors are chosen by their state parties, usually from their most loyal workers and supporters.
Bush aides discount the possibility of a split verdict, largely because they say their polling shows no disparity between the national and state trends. In that polling, Bush holds leads equal to or greater than his national advantage in most of the key battleground states, they say.
Karl Rove, Bush's chief strategist and a political history buff, flatly dismisses the possibility of an 1888 replay. Cleveland won a 90,000-vote majority largely because he amassed huge margins in Southern states where the Republican Party still wasn't competing in the aftermath of the Civil War, Rove notes. Harrison won most of the key Northern states by much smaller margins.
Though Bush is comfortably ahead in most of the South, Rove notes, in most places he isn't approaching the 35- to 50-percentage-point margins Cleveland ran up throughout the South. That means Bush's current lead in the national polls reflects more diversified regional strength than Cleveland displayed.
"It is highly unlikely that the model of 1888 is going to be repeated," Rove said.
Tracking Candidates' Strengths by Region
Yet, to a lesser degree, the same basic phenomenon of that year is present today. Bush has big leads in almost all of the states in the Republican "L" that stretches down from the Mountain West across the South. In Texas alone, Bush's lead may exceed 30 percentage points. Overall, the latest CNN/USA Today/Gallup tracking poll shows Bush leading by 19 percentage points in the South, by far the biggest margin either candidate holds in a region.
By contrast, Gore's lead in most states outside the Northeast has shrunk to single digits. For instance, the latest polls in California and Illinois--where Gore once led comfortably--now show him holding just 7-point advantages in both.
Bush's success at reducing Gore's margin in these states has helped the Texan pad his lead in the national polls. Indeed the desire to affect the national figures, which influence news coverage, is one reason Republicans have spent so much money on campaign ads in Illinois and California, even as they acknowledge they are unlikely to win either.
The bottom line is a lopsided battlefield reminiscent of Cleveland and Harrison. "Where Bush leads he leads big, where Gore leads he leads small," independent pollster John Zogby said. As that disparity factors into the national polls, Bush has opened his lead, which ranges from 4 percentage points in the Gallup survey to 3 points in Zogby's and 1 point in a CBS/New York Times poll.
But a related finding by the Gallup poll vividly illustrates Gore's stronger position in some battleground states. Gallup calculates the results from voters in seven states: Ohio, Wisconsin, Michigan, Missouri, Pennsylvania, New Jersey and Illinois. In those states, Gore led Bush by 4 points Wednesday.
Assuming Gore holds the Democratic base states--particularly New York, California, New Jersey and Illinois--and Bush takes the reliable Republican states as well as Ohio and Missouri (where he leads), the states that are most likely to decide the election are Michigan, Pennsylvania and Florida. Most analysts believe the candidate who wins two of those three becomes the favorite for victory. If either wins three, he would become a prohibitive favorite.
The most recent polls in Michigan show Gore leading, by margins of 4 to 10 percentage points.
In both Pennsylvania and Florida, the public polls diverge. The Los Angeles Times Poll earlier this week showed Bush leading narrowly in both. But three public surveys released Thursday gave Gore a 4- to 5-point lead in Florida (though Republicans are more optimistic in their surveys).
In Pennsylvania, Zogby's tracking poll and an American Research Group survey released Thursday gave Gore leads of 4 and 6 percentage points, respectively.
The list of other states too close to call includes Washington, Oregon, Iowa, Wisconsin, Tennessee, Nevada, New Mexico, Minnesota, New Hampshire and Maine.
Calculating the Chances
In recent elections, almost all of the closely contested states ultimately have tipped toward the winner of the popular vote. But heading into the campaign's last weekend, these states show no consistent pattern.
Gore has been gaining ground in Tennessee, Minnesota and New Mexico, while Bush appears to be advancing in Maine, Oregon and perhaps Washington.
In the end, any possibility of a split verdict depends on a razor-thin popular vote total, most experts agree.
Political scientist Erikson, like most analysts, said that if Bush maintains the 3- to 4-point lead he enjoys in most national polls, "he is going to win and he is going to run the table of the [close] states."
Only if Gore can reduce Bush's overall popular vote lead to no more than 2 percentage points would the vice president have a realistic chance of capturing an electoral college majority, Erikson predicts.
Which means that, even to win a split decision, Gore over the next few days probably has to win the final round of this bruising title match.