Trade Resources Industry Views Statistical Modeling Helped Some Prognosticators Predict Outcome of U.S. Elections

Statistical Modeling Helped Some Prognosticators Predict Outcome of U.S. Elections

Computerworld - Statistical modeling techniques that retailers and manufacturers use to find and target customers helped some prognosticators predict the outcome of this month's U.S. elections with stunning accuracy.

Because of his connection to The New York Times, blogger Nate Silver may be the best-known quantitative analyst to accurately predict the election results, but many others also used statistical models and got similar results. The spot-on forecasts have focused unprecedented attention on quants, as quantitative analysts are known, and their ability to predict future events and trends.

As far back as June, Drew Linzer, an assistant professor of political science at Emory University, predicted in his blog, Votamatic, that Barack Obama would win re-election with at least 52% of the popular vote and 332 Electoral College votes. In the end, Obama took 51% of the popular vote and 332 electoral votes. 

Already an Insider? Sign in

Computerworld - Statistical modeling techniques that retailers and manufacturers use to find and target customers helped some prognosticators predict the outcome of this month's U.S. elections with stunning accuracy.

Because of his connection to The New York Times, blogger Nate Silver may be the best-known quantitative analyst to accurately predict the election results, but many others also used statistical models and got similar results. The spot-on forecasts have focused unprecedented attention on quants, as quantitative analysts are known, and their ability to predict future events and trends.

As far back as June, Drew Linzer, an assistant professor of political science at Emory University, predicted in his blog, Votamatic, that Barack Obama would win re-election with at least 52% of the popular vote and 332 Electoral College votes. In the end, Obama took 51% of the popular vote and 332 electoral votes.

Linzer, like other quants who accurately predicted election results, made his forecasts by aggregating state-level poll data with economic indicators and data from previous polls. He started by constructing a baseline forecast for each state using the Time-For-Change statistical model developed years earlier by Emory colleague Alan Abramowitz.

Time-For-Change predicts the incumbent party candidate's national vote share by looking at factors such as the president's approval rating in June, the percentage change in gross domestic product in the first two quarters of the year, and the number of years the incumbent party has held the presidency.

Poll data is thrown into the mix as Election Day nears. "The basic idea is that on Election Day, or in the weeks leading to Election Day, polls are the best indicator," he said.

Despite minor fluctuations in support levels for the candidates, the data always showed Obama winning, he said.

"I never saw it as being a close race," Linzer said. "When I started producing my forecast in late May, the historical model that I was using showed that Obama would get about 52% of the major party vote."

David Rothschild, chief economist at Microsoft and developer of the model used by Yahoo's The Signal blog, which also accurately predicted the outcome of the presidential race, called the forecast "a triumph of science over punditry."

Back in February, before Mitt Romney had secured the Republican nomination, The Signal had a baseline forecast predicting an Obama win.

Rothschild said his model creates a baseline by evaluating historical data, state-level economic indicators and factors like the president's approval rating and the advantages of incumbency.

"For most of the election cycle, we had Obama at around 303 [Electoral College votes]," Rothschild said.

Ultimately, the accuracy of the polls made all the difference, said Josh Putnam, a visiting professor of political science at Davidson College and author of FHQ, another blog that early on predicted a 332-206 Obama electoral vote victory. If the polls had been wrong, the forecasts would have been wrong as well, he said.

Putnam didn't use statistical models; he simply aggregated state-level poll data to arrive at his forecasts.

"It was not very complicated," he said. "My forecasts were based simply on a weighted average of poll data."

This version of this story was originally published in Computerworld's print edition. It was adapted from an article that appeared earlier on Computerworld.com.

Source: http://www.computerworld.com/s/article/9233753/Statistical_modeling_yields_accurate_election_forecasts
Contribute Copyright Policy
Statistical Modeling Yields Accurate Election Forecasts