Computer Model Says Polls Wrong, Romney Will Win
Mitt Romney will win – and win big – according to a computer model that has successfully predicted the last eight presidential elections.
Romney will win every “swing” state and will wallop President Barack Obama 320-218 in the electoral college vote, according to the forecast by two professors running the University of Colorado computer model, as reported by the London Daily Mail: The computer model projection has correctly predicted every presidential election since 1980.
Most current polls give Obama an edge of 283 to 255 for Romney in the electoral votes, whereby “winner takes all” of the delegates in any state, regardless of the size of the popular vote.
The computer model was constructed by Profs. Kenneth Bickers and Michael Berry, who base their predictions on economic data, income per capita, and unemployment figures for each state. The computer model prediction could change as data changes. The current forecast is based on June data and will be updated after September data is collected.
Obama enters the election race, which officially begins after each candidate wins the nomination at the party conventions, with a jobless rate of 8.3 percent. There are no signs of its declining significantly, and it even could rise slightly. Moreover, the “real” rate is above 15 percent when including millions of Americans who simply have given up looking for work.
Prof. Berry told the London newspaper that if voters perceive the economy as having improved since Obama was elected in 2008, he could win a second term. If not, Romney will win several states that polls show as doubtful or leaning towards the president.
Meanwhile, the popular polls show a virtual dead-heat, while most Americans, asked to predict the outcome of the November vote, think that President Obama will be re-elected.
One negative factor for Obama could be a low turnout of Democrats, whom polls show as favoring him but without much enthusiasm, while Republican voters are fervent in their dislike of the president.