Last night Republicans witnessed presidential election results far from the victory they had been promised. Many never saw it coming.
In the days leading up to the Election, conservative pundits clung to the notion polls were incorrect. Wishful supporters of the GOP claimed public opinion polls were oversampling Democrats; this was the reason President Obama maintained a small but persistent advantage, they insisted.
Right-leaning coverage also did little to forewarn mistaken Republican voters. Pundits claimed the presidential race was destined to be a victory for Mitt Romney.
“Nobody knows anything,” wrote Peggy Noonan for the Wall Street Journal, in an effort to dismiss the general idea that examining data, applying reason to polls, and attempting to determine who would cast votes could provide guidance to the elections’ results. Noonan was positive independents were “breaking for Romney” on the simple basis of intuition.
On Monday, Fox News contributor Dick Morris told viewers to expect a Romney landslide, the reason being majority of polls had oversampled Democrats. By “unskewing” poll results, Obama’s narrow lead would instantly vanish.
As November 6 drew closer, it became apparent that the overwhelming belief amongst Republicans was Romney would win big in most swing states, largely because popular conservative media outlets told them so.
At Election Day’s end, Nate Silver’s FiveThirtyEight blog’s prediction of Obama’s re-election and many others like it, were correct. Romney did turn North Carolina red, and ended up with a win in Indiana. But Obama, by the slimmest margins, took away Colorado, Virginia, Ohio, and Wisconsin— a startlingly different result than the huge win they assumed Romney would capture in each state.
The reality is, most polls were not skewed and were not, in fact, oversampling Democrats. The electorate is now composed of much more Democrats, especially in swing states.
According to a Gallup poll of party affiliation, 32 percent of Americans identified as Democrats. The poll, conducted September 24-27, finds the percentage of Americans identifying as Republicans at 28 percent and 38 percent considering themselves Independents. From the end of summer 2012 to September 27, on average, only 28 percent identified as Republicans and 33 percent considered themselves to be Democrats. 38 percent of Americans were reported as being an Independent.
Poll results regarding the presidential election were reached by seeking out a demographically representative sample– blacks, whites, young, old, lower-income, upper-income, and from differing regions of a state. It is then, after sifting through demographics do pollsters question party affiliation. Respondent pools are not based on specific numbers of Republicans, Democrats, and Independents.
Pollsters such as Nate Silver set aside assumptions of how independents would vote and how undecided voters were leaning to examine hard data from a variety of public opinion polls. Our own election predictions here at PollHeadlines forecasted Obama winning re-election with 290 electoral-votes. Although we did not predict Virginia correctly (and perhaps Florida pending results), by using public opinion polls, we were also able to predict Obama’s win.
What Republicans can take away from this election is a wholly different approach to viewing polls. Rather than hyping a win in the hopes the crowd will follow the assumed winner and going off intuition, it may just be all in poll numbers after all.
