Right Angle Blog catches a piece on the Channel 3 website Friday bragging about their polling on the primary. Well, WKYC's got a right to brag; their election-eve SurveyUSA poll really, truly nailed the final results. But I'm still very curious about the likely voter sample, which was heavily skewed toward voters who've gone to college, compared to the city's adult population as a whole.
Since SurveyUSA's predictions were so accurate, it's reasonable to infer that their sampling was accurate too. Both their samples of self-identified likely primary voters (just before the election and a week earlier) had the same education breakdown: 15% had been to grad school, 17% had four-year college degrees, 28-29% had some college experience, and just 40% had never attended college. As a sample of Cleveland voters this is remarkable, since in the 2000 Census only 35% of college-age city residents had attended college at all, only 15% had degrees of any kind, and two-thirds said they'd never gone past high school.
Has college education become the major factor defining "likely voters" in low-turnout Cleveland elections? Is the educated minority of city residents taking over the role once played by senior citizens -- the role of core voters who will always show up, no matter what?
Well, here's an interesting chart comparing two sets of percentages -- the percentage of people age 25 and older in each ward who've attended college (Census 2000), and the percentage you get by dividing each ward's votes in the very high-turnout November 2004 election by its votes in the very low-turnout election last Tuesday.
With a few striking exceptions, the two percentages follow similar patterns from ward to ward. A higher percentage of college-educated residents tends to coincide with a higher percentage of November voters who also turned out last week.
At this point you may be thinking "Well, du-u-u-h. Of course college-educated voters are more likely to show up at the polls than less educated voters! Isn't that true everywhere?"
That's one reasonable way to look at it. But here's another possibility:
Most people go to the polls only when they have a reason to go -- something or someone they specifically want to vote for, or against. Last November a historically high number of Clevelanders of all education and income levels, more than 170,000, turned out to vote for Kerry against Bush. (Kerry won every single city precinct.) Last Tuesday, only 30% of that number -- apparently the most educated 30% -- found the mayor's race similarly compelling.
Could this be because none of the candidates were making a serious effort to talk to the less educated 70%... the poor and blue-collar majority of the city?
More about this tomorrow.