If economic discontent had a face in this election season, it might look something like Kenneth Tyner's. But just because the 30-year-old is suiting up these days for a new job at a Flint, Mich., McDonald's, just because he hasn't held a job longer than seven months in the past four years, doesn't mean Tyner is seething against President Bush.
"He has no responsibility as to how I screw up my life," said Tyner, who, along with his working wife, is trying to raise two children on an annual income somewhere south of $35,000. "I support the president because I like what he's done in Iraq."
For economy watchers, it has been a confounding election season. Economists who have correlated economic statistics and past elections say the president should be waltzing to victory Tuesday, with as much as 57 percent of the national vote. In the hard-hit battleground of the upper Midwest, pollsters say, Democratic challenger Sen. John F. Kerry (Mass.) should have locked up states such as Michigan by now.
Instead, the outcome is anybody's guess.
"Whenever one uses econometric [statistical] work to make predictions, you have to assume the future is going to be like the past," said Ray C. Fair, an economist at Yale University, whose election model predicts Bush should win 57.48 percent of the two-party vote. "Foreign policy has never trumped economics in the past, but this may be the year."
In Michigan, where layoffs and work furloughs still dominate the headlines, Kerry's lead remains surprisingly precarious, said Ed Sarpolus, an independent pollster. That, he said, is in large part because the Bush campaign -- with the assistance of Kerry -- has been able to keep the focus on Iraq and terrorism. Bush has actually expanded his 2000 margin among white men like Tyner, despite the state's job losses.
"The advantage is still Kerry's," Sarpolus said. "But if it flips and Bush wins, it's because the last message is Iraq and the war on terrorism, and [Michigan voters] didn't hear the counter message from Kerry on health care and jobs."
Beyond Michigan, the razor-thin margins in some of the most fiercely contested states seem to defy those states' economic fates. According to recent Labor Department data, Ohio lagged behind only Michigan in job losses over the past year, shedding 17,900 payroll positions. Yet some polls indicate the presidential contest in the Buckeye State is at a virtual tie.
Minnesota, which Vice President Al Gore won in 2000 by 58,607 votes, appears deadlocked -- despite the 20,000 jobs lost since the president took office. Iowa, which Gore won by a mere 4,144 votes, has since lost 25,900 payroll positions, six times Gore's margin of victory. Yet Bush has led Kerry in several recent Iowa polls.
On the other hand, Florida led all states in job gains over the past 12 months, with 125,500 new positions. Yet it also is too close to call.
The failure of the economy to play decisively in either candidate's favor may be due to the mixed signals of the economy itself, pollsters and economists say. If Bill Clinton's slogan in 1992 was, "It's the economy, stupid," Kerry's and Bush's might be, "It's a stupid economy," or at least a murky one.
"In order for Kerry to get real traction, he needs consistently bad news on the war and on the economy," said Ed Goeas, a Republican pollster. "For Bush to get real traction he needs the opposite: consistently positive news on the war and the economy. Really, neither candidate has gotten that."
The Commerce Department is expected to announce today that the economy grew at a strong clip in July, August and September, the second straight quarter of significant expansion. But that growth is facing a gusty headwind: oil prices over $50 a barrel and average gasoline prices topping $2 a gallon.
The nation's payrolls have expanded for 14 straight months and 1.9 million jobs, allowing Bush to say at campaign rallies, "The economy is strong and getting stronger." But that job growth has barely kept up with population increases and has been insufficient for the president to overcome Kerry's charge that Bush will be the first president since Herbert Hoover to preside over a net job loss.
The unemployment rate stands at 5.4 percent, below the post-World War II average and exactly where it stood when Clinton won reelection in 1996. But that statistic has been aided by workers who have given up finding a job and left the workforce. The labor force participation rate is 65.9 percent, compared with the 67.2 percent rate that Bush inherited in January 2001 and the 67 percent rate of November 1996.
If the labor force was more in line with historical trends, the unemployment rate would be between 6.5 and 7 percent, said Nariman Behravesh, chief economist at the Massachusetts-based economic forecasting firm Global Insight Inc.
Nonetheless, the economy should be a large plus for Bush's reelection bid, according to three different election models. Looking at income growth, unemployment, the power of incumbency and other factors, Global Insight's model forecasts Bush garnering 55.7 percent of the two-party vote. Fair's model considers economic growth, the inflation rate and the number of quarters with unambiguously "good news." Under that model, Bush should do even better.
Macroeconomic Advisers LLC, a forecasting firm in St. Louis, takes yet another tack, looking at income growth in the six to nine months before the election, the inflation rate and housing starts. Again, Bush should be running away with the election.
Housing looms large because it is a proxy for economic well-being, said Joel Prakken, an economist at the firm. People would not be buying homes or building new ones if they feared for their economic futures. Just this week, the Commerce Department and the National Association of Realtors reported unexpected surges in new and existing home sales, to near-record levels.
"The unemployment rate is not an important variable," Prakken said, despite all the energy politicians put into labor trends. "The truth of the matter is, 5 percent of the country is usually unemployed, but several hundred million people are employed."
Yet something has gone wrong this year, the modelers readily concede. Fair suggested that the historical correlation between economic growth and job growth may be breaking down, leading voters to take little solace in a steadily expanding gross domestic product.
Another factor may be income growth, Behravesh said. The models use the total growth of after-tax incomes in the economy, but if that growth is concentrated at the top of the income scale, the forecasts will not pick up discontent among middle- and lower-income voters.
The problem, Prakken suggested, is not with the economic measurements, he said. It is with trying to separate the economy from other issues.
"The models don't know anything about insurgents lining up Iraqi army recruits and executing them," he said. "They don't know anything about  tons of missing explosives. I think the economy is the most important thing most of the time, but I do not expect the kind of romp for Bush that the economic models predict. It's much too messy this time around."
Tyner has more than enough reason to grouse. In down-on-its-heels Flint, he has cast around for four years, from McDonald's to temporary agencies to installing cable television and back to McDonald's. The longest job he held since Bush came to office was a seven-month stint with a lawn service. On Monday morning, he scraped together $10 to put gasoline into the tank of his old Chevy pickup truck. By that afternoon, the tank was empty.
But his wife has held down her job in an accounting office of USA Today, and he has always found some work, no matter how meager the pay.
"My wife and I, despite the economy, gas prices and yadda, yadda, we are doing okay for ourselves," he said, shrugging. "I can understand that the economy is in the pooper, but, no matter how bad things get with this economy, I have always found a job. You just have to get off your butt."