Ohio Polls and PPP

Two new polls are out today regarding the presidential race in Ohio. Rasmussen reports the race tied at 45% each for Obama and Romney. Public Policy Polling (PPP) reports Obama is ahead by 48% to 45%. Since I have noticed that PPP polls always overstate the lead for Obama when compared to other polling organizations, these results are not surprising. Almost certainly, the Rasmussen poll is correct while the PPP poll is skewed, but I wanted to see if I could tell how PPP got its slant into the results. It was not hard to find.

The racial breakdown in Ohio includes 12% of the population which is African American. Normally, whites vote in a higher percentage than blacks both across the country and in Ohio. In 2008, 85% of the Ohio electorate was white; the remainder was African American, Asian or Hispanic. The polling in 2008 showed that blacks voted in much higher numbers than usual because Obama was the first black nominee of a major party. The polling this year has shown that much of that enthusiasm has disappeared; blacks still support Obama, but they are not likely to vote in such a great percentage as last time. At the same time, enthusiasm among Republicans has risen dramatically since 2008. As a result, at a minimum one would expect the Ohio electorate to be at least 85% white or perhaps higher.

The sample polled by PPP for its latest Ohio poll is 82% white, 13% black and 5% Hispanic or Asian. Think about that. The PPP sample has more blacks in it than the Ohio population as a whole, and blacks vote at a lower rate than whites. This is a critical difference since the black vote goes over 95% to Obama and the white vote goes strongly to Romney. A sample that was just 3% more white than the one from PPP would mean a pickup of about 1.75% for Romney and a reduction of the same amount for Obama. Suddenly the 48-45 lead for Obama found by PPP becomes a 47 to 46 lead for Romney.

I have examined about sixty different polls in many different states. No matter whether PPP is polling registered voters or likely voters, its polls always show Obama with a substantially bigger lead than the polls from the other pollsters. What seemed at first to be just interesting has become scandalous. Somehow, PPP produces slanted results for its polls. PPP had better explain how this can be, or it should no longer be taken seriously.