In the presidential election next week, it’s all coming down to a few states, Ohio chief among them. According to HuffPost pollster aggregation figures, Obama has the edge in the “birthplace of aviation” state, averaging 48.5% compared to Romney’s 46%.
But according to the latest YouGov/Xbox poll, Ohio has slightly more Romney supporters than Obama supporters.
That said, while slightly more respondents said they intend to vote for Romney, a majority of the same respondents said they expect Obama to win. And here’s the twist: According to economist David Rothschild, “when the intention (i.e., support) and expectation of respondents in a poll point in opposite directions, the expectation is correct over 75 percent of the time.”
YouGov/Xbox is the poll Microsoft and YouGov have been running through Xbox LIVE. To vote, you answer questions in a preliminary questionnaire that surveys your gender, age group, geographical location (by state) and general voting preferences (Democrat, Republican, Independent, etc.) before answering the poll question(s) on tap.
I didn’t participate in the Monday, Oct. 29 poll that generated the above results, but to give you a sense of how this works, today’s poll questions include stuff like: “Is your opinion of Barack Obama favorable or unfavorable?”, “What is your opinion of Mitt Romney?” and “How serious a problem do you think global warming is?”
In Monday’s YouGov/Xbox Poll, Rothschild says 12,479 U.S. respondents were interviewed, 642 of those indicating they live in Ohio. Of the latter group, 48.3% indicated support for Obama and 51.7% for Romney. But 51.1% of the same Ohio-based respondents said they expected Obama would win (as opposed to 48.9% for Romney).
But wait, shouldn’t intention matter more than expectation in an election? It sounds like common sense, right? Not necessarily. According to an academic study on voter expectations co-conducted by Rothschild and University of Michigan Business and Public Policy professor Justin Wolfers, “polls of voter expectations consistently yield more accurate forecasts than polls of voter intentions.”
One possible reason for this: Expectation questions incorporate more information, says Rothschild.
Every person possesses a batch of both public and private information about the election, and when a pollster asks them about their intention, the pollster is extracting only a small portion of that information; the pollster is just getting the respondent’s support on that day. The expectation question captures information about the respondents social network (e.g., who the respondent thinks her friends and family will vote for) and more public information (e.g., what the respondent is seeing in the news), along with the respondents’ intention to vote and support for one or the other candidate.
It’s an intriguing idea, one that’s especially interesting considering this “expectation” survey is being conducted through such an unconventional medium (the Xbox 360).