Tuesday, October 6, 2009

Blacks and the 2009 election

According to the latest census numbers New Jersey's population is about 15% black.

And if the turnout on November 3rd reflects that Jon Corzine can probably start planning for his second term.

Here's the thing though: even last year with Barack Obama on the ballot exit polls showed turnout from African Americans lagging their share of the population at 12%.

Now most of the pollsters I could find sample composition by race for are expecting something even lower than that. Democracy Corps has it at 10%, we have it at 9%, and Monmouth has it at only 8%. Research 2000, at 12%, is the only company I could find projecting it at the same level as last year.

I usually think of 'worst case scenario' for black turnout as being 80% of the state's black population, i.e. a state with a 15% black population would have a 12% black electorate. But given that's where New Jersey was last year we don't have much choice but to rethink that assumption.

It's a pretty similar story in Virginia. The population there is 20% black and so was the electorate last year. We've had it in the 16-18% range on all of our polls. Insider Advantage had it at 18% on their recent poll, Research 2000 had it at 16%, and SurveyUSA had it at 17% last week and then 15% this week. We'll be watching African American poll response closely the next four weeks...I think it's possible it could go even as low as 13 or 14%...or as high as 18 or 19%. It's just really hard to say what that engagement level's going to be when election day gets here.

Barack Obama will probably show up to campaign for these folks and no doubt folks will come to see him...but will that get them excited enough to show up and vote for Jon Corzine and Creigh Deeds? Hard to say.


Anonymous said...

Does anyone know what percentage of that 15% is of voting age? Maybe since the state's overall voting age population would decrease as a percentage of total population the 15% would hold steady.

Jayant Reddy said...

Tom, I think one possible mistake you're making, one I'm surprised you didn't consider, is assuming the exit polling is right.

The reason I'm surprised is you guys at times past have been at the forefront of calling out provably erroneous exit poll results, such as the 2004 North Carolina black vote share which exit polls pegged at 26%, but actual state data pegged at 19.6%. And more recently you guys pointed out that South Carolina's exit polling, too, had to be wrong in one of the past two Presidential elections because it purported black vote share DECLINING from 2004 to 2008, despite Obama on the ballot in 2008.

That being the case, I think you need to consider that maybe last year's New Jersey exit poll was just wrong and understated black vote share. I just looked at the 2004 exit poll on CNN.com, and it showed the 2004 black vote share at 14% in New Jersey. You say in your blog post that the 2008 exit poll shows the black vote at just 12%, which I confirmed just now again at CNN.com.

So the black vote DROPPED in NJ with Obama on the ballot? I don't think so. I'm guessing the 2008 exit poll you cite erroneously understated black turnout in the state. It's possible that the 2004 exit poll actually overstated black turnout as it did in NC, but I still doubt black vote share was as low as 12% in 2008.

That said, the 2006 NJ Senate exit poll showed black vote share at just 10%, and that may, indeed, be closer to reality next month. The total white vote in 2006 was 77% of the total, compared to, according to what again I think are erroneous exit poll results, 70% in 2004 and 73% in 2008.

I think it's hard to say, too, what black vote share will be in the age of Obama. One hypothesis to be tested is whether Obama's Presidency raises black interest in other elections without him on the ballot, above levels in the pre-Obama America. Probably not, but I still hope......

Web Statistics