Mark Blumenthal’s National Journal column is a good overview of the potential poll spoilers in this election. In summary, the following things could cause things to go awry in polling: 1) underrepresentation of cell-phone only households in polling; 2) confounds in “likely voter” models due to unprecedented turnout; and 3) respondents failing to tell the truth in surveys due to social discomfort over not supporting a black candidate.
Items #1 and #2 seem to favor Obama. He’s crushing the younger demographic, and according to Marc:
“Since 2004, the cell-phone-only population as a percentage of all adults has more than tripled and now includes nearly a third of adults between 18 and 24.”
That’s a huge number, and it’s been way underrepresented in some polls. Same for the turnout … higher turnout favors democrats, especially this cycle, due to the unusually drawn-out primary campaign. Obama registered tons of folks in late-voting primary states that otherwise might have been somewhat ignored.
Item #3 certainly cuts the opposite way, but Marc points this out:
“Over the last 10 years, according to a paper [PDF] by Harvard post-doctoral fellow Daniel Hopkins that studied 133 statewide races between 1989 and 2006, the apparent polling bias in such races largely disappeared.”
Remember, this doesn’t mean that racial bias in elections has disappeared. It just means that people are less likely to lie about it. The so-called “Bradley Effect” doesn’t mean that folks won’t vote for a black dude … it means they’ll tell a pollster that they will, then do the opposite in the voting booth. Most of the folks who won’t vote for Obama in November because of his race are saying, “McCain” when someone asks them who they’re voting for.
-Education Dude
No comments:
Post a Comment