A little-known sector of the election forecasting universe – models based on reams of economic data — proved more accurate in this presidential election than traditional opinion polls from well-known pollsters like John Zogby. Mr. Zogby grabbed attention when he predicted a victory for Senator Kerry in May. Nine of Mr. Zogby’s polling competitors had Mr. Bush and Senator Kerry in a virtual tie, often with a 2% or greater margin of error.

In contrast, four economics-based models had President Bush taking at least 53% of the vote, with several predicting landslide like totals of 56% and 58%. The final popular vote tally was 51%-48%.

Mr. Zogby’s spokeswoman did not return a call seeking comment. A statement posted on Zogby International’s Web site said, “We feel strongly that our pre-election polls were accurate on virtually every state. Our predictions on many of the key battleground states like Ohio and Florida were within the margin of error. I thought we captured a trend, but apparently that result didn’t materialize.”

Economics-based models do not include measures of public sentiment on policy issues. Instead, they are built on a combination of published and forecast data on national and regional unemployment, gross domestic product growth, and inflation. One modeler, Global Insights senior economist Nigel Gault, said, “If you took the last four year’s worth of economic data into account, there was little chance that the voters would reject an incumbent when GDP was growing at a 3.5% rate.”

Global Insights’ model predicted between a 55% and 56% final popular vote tally, Mr. Gault said. He said the difference between his model and the final tally represented “unquantifiable issues,” such as voter dissatisfaction with health-care costs and the Iraq War.

Mr. Gault said opinion polls often fail to take into account the massive advantage an incumbent president has over a challenger, which is magnified when the incumbent is presiding over a period of historically low unemployment and inflation. “Job growth was much lower for President Bush than President Clinton, but there has clearly been job growth,” he said. He added that incumbents who lost – Presidents Hoover, Ford, Carter, and George H.W. Bush – all presided over periods of GDP decline.

The ability to understand the difference between voter concern and voter anger over an issue like employment is another reason that economy-based models do better than opinion polls, said Economy.com’s Robert Dye. He said that people are always concerned about job safety, but the things that have affected it the most since 2001-outsourcing, the tech bubble collapse and the corporate governance scandals – are things that President Bush couldn’t do much about.

Mr. Dye said that the 5.5% unemployment rate is a figure that would get most recent presidents handily reelected. Economy.com’s model predicted a Republican victory, with Mr. Bush taking 53.2% of the popular vote.

Another forecasting approach for the presidential election that beat the polls and consistently called President Bush as the winner were online gaming and futures markets.

The Iowa Electronic Market, a project of the University of Iowa’s School of Business, offered traders the chance to speculate cash on the presidential election’s winner and the final percentage of popular vote received. Since the political conventions this summer, the IEM’s traders had forecast President Bush’s re-election, with the president’s popular vote lead eventually narrowing to 2.7 points early Tuesday.

Winning futures trades on the IEM – where traders have called every presidential election winner since 1988, although traders predicted then Governor Bush would carry the popular vote in 2000 – pay out the difference between $1 and the cost of the future. Late on November 1, the IEM was pricing President Bush’s futures at 51.2 and Senator Kerry’s at 48.5. A future is a contract that allows the holder to bet on the price move of an asset or commodity.

Online gaming sites such asTrade-Sports.com- many based in Great Britain – had been offering an even rosier re-election scenario for President Bush, who held an almost 11% lead going into Tuesday morning. The site’s bettors had wagered more than $5 million dollars that President Bush would be re-elected.

But a rumor leaked that Senator Kerry’s exit polls were exceptionally strong and wiped out that lead, sending the Senator ahead in the betting on the popular vote totals and numerous state races, including Florida and Ohio.