Obama's walrus has faulty plan...what else is new?
The Folly of David Axelrod’s Turnout Model
Much of this blog is spent arguing — persuasively I hope — why the polling data and get out the vote activities favor Mitt Romney this election cycle. But if President Obama wins re-election where could all of this analysis have erred?
Possibly the final remaining question that will determine the election outcome is whether Obama campaign manager David Axelrod is correct on the racial composition of the electorate being 72% White or whether is he fooling both himself and his acolytes in the media? President Obama regularly polls with national support among Whites as low as between 36 – 38%. If Axelrod is correct then Barack Obama has a good chance to eek out a close re-election. But if Axelrod is wrong on the electorate composition and Whites make up closer to 75% of voters then not only will Barack Obama lose this election, he could lose it badly with a final tally in the area of 338 to 200 electoral votes.
The entire Obama campaign is predicated on a voting public with a racial composition that in my opinion is highly unlikely to appear in 2012. The Obama campaign has revealed their assumption that White voters will only comprise 72% of the national voting public this year. State percentages will vary but those models are not revealed publicly. The Census Bureau’s Current Population Survey (CPS) is the final arbiter on racial make-up of an election. For months following an election, they pour through every precinct and get an accurate read on election statistics before releasing their results which unsurprisingly tend to differ from election night exit polls.
The racial breakdown in 2008 election according to the Census Bureau was White 76.3%, Black 12.1%, Hispanic 7.4%, Asian 2.5%, Other 1.7%.
The CNN exit polls which are still used regularly in news stories reported: White: 74%, Blacks: 13%, Hispanics 9%, Asians 2%, Other 3%.
When you look at the differences, Whites are underrepresented by 2.3%, Blacks overrepresented by 0.9% and Hispanics over-represented by 1.6%. In the last election Whites voted Republican (55 – 43) while both Blacks (95 to 4) and Hispanics (67 to 31) voted Democrat. When national polls reflect the CNN racial make-up they are over-sampling Democrat groups and under-sampling Republican groups.
As outlined previously, 1.7 million White voters (who voted in 2004) did not vote in 2008. This means from a racial composition stand-point the 2008 demographic breakdown is over-generous to the non-White groups and somewhat of an anomaly due to the missing White demographic who voted in 2004 but sat out last time. This is not an argument against the decreasing percentage of White voters in election, because the overall increase in the non-White voting population is very real. The problem with Team Obama’s assumptions, however, is they decrease the White vote -4.3pp to 72% this election cycle when there is no evidence to support such a steep decline.
Before 2008, the average decrease in the White percent of the popular vote was -1.4 percentage points (pp) since 1988. This was not solely because of decreased interest from White voters in Presidential elections but due to the increasing number of non-Whites both in the population and participating in Presidential elections. The rate of change from election to election was a decrease of -0.4pp in the White composition of the voting public and most of that coming between 1992 and 1996 in another election when white Republicans were less than enthusiastic with their nominee. Something dramatic, however, happened in 2008. The number of whites as a percentage of the voting population dropped -2.9pp to 76.3% from 79.2% (Pew Research, April 30, 2009). |
