The number of voters who haven鈥檛 made up their mind whether to vote for President Obama or Mitt Romney is so low that there now is only a small and ever dwindling chance they might influence the outcome of the presidential election, said a Vanderbilt University political scientist.
鈥淚t鈥檚 not impossible that something that happens in the last weeks of the election season or during the debates could influence undecided voters in one direction,鈥 said at Vanderbilt and the leader of the .
鈥淏ut that (swing vote) sliver is so small these days that it will be very hard to do,鈥 he said. 鈥淭he fundamentals of this election seem pretty much in place and Romney needs to get control of the narrative to move this contest.鈥
YouGov uses a representative sample of 600 Americans with an over-sample of 200 independent voters to rate campaign ads. YouGov most recently evaluated dueling Obama and Romney advertisements, each blaming the other for the loss of American manufacturing jobs to China.
Everyone who participated in the polls was shown the ads before taking the polls, meaning that the effects should be more potent than on most voters who see the ads on their own. That鈥檚 because there is a lot of competing information and the power of an ad dissipates over time.
Even with this kind of strong test, the results show little movement.
鈥淔or the representative sample of Americans who saw the Obama attack ad on Romney, the president held a 5.6 percentage point lead,鈥 Geer said. 鈥淔or Americans who saw the Romney attack ad on Obama, the president held a 5.4 percentage point lead.
鈥淚n other words, these ads did not change the dynamics of the race at all.鈥
Obama鈥檚 attack ads did sway a few of the handful of swing voters his way, while Romney鈥檚 ad did not move the dials among these voters,鈥 Geer said.
鈥淧eople have opinions about Obama and ads cannot alter them much,鈥 Geer said. 鈥淏ut Romney is less well known and the ads may be able to have some effect among this sliver of the electorate.鈥
Results and both China ads are posted on a website for .