Sunday, March 2, 2008

Columbus Dispatch Poll Less Disturbing at First Glance

The Columbus Dispatch's recent poll looks very bad for Obama, as it has Hillary Clinton winning 56-40, which, when combined with the so-called Bradley effect, would be giving her more than the margin she needs to actually win more pledged delegates.

But a closer look at the poll makes things look far better for Barack Obama.

First of all, they only poll registered Democrats, when Independents are allowed to vote as well (and since the GOP contest is done, will be voting far more heavily in the Democratic primary; Obama tends to do far better among Independent voters).

Second of all, the crosstabs for the registered Democrats seem to be underemphasizing Obama's strong demographics.

First of all, they expect the black percentage of the vote among registered Democrats to be only 12%, the same as the black percentage in the overall population, in a state where, thanks to the Democratic efforts in 2004 (and 2006, to a lesser extent), black voters are far more likely to be registered to vote than they would be in many other states because Democrat, with effectiveness, as the 2004 Ohio loss was due to factors other than failure to get out the African-American vote (although there almost certainly was some disenfrancisement with the moving of the voting machines out of urban precincts and various other tactics, this didn't come into play in terms of voters being registered in most cases, so AA voters are being called by the campaigns).

Second of all, Ohio has one of the most liberal voting laws for those who've committed felonies, second only to Maine and Vermont (who allow even those in prison to vote), allowing anybody not actually in prison/jail to vote, so that only 2.64% of African-American voters are disenfranchised (African-Americans are far more likely to be arrested, prosecuted and convicted of a felony for similar drug-related crimes than whites; this is the major difference).

Moreover, their numbers involve rather laughable age estimates (they expect 18-24 year olds to be 1% of turnout and voters under 35 to be 7% of turnout, and voters under 45 to be 17% of turnout).

The Wisconsin exit polls had 9% 18-24, and 38% under 45.

No comments: