Beat the Pollsters (NC: Obama wins 54% to 46%, IN: Clinton wins 53% to 47%)

Given that Democratic primary has been more about demographics than policy, I'm going to use past exit polls and demographic data to predict what will happen tomorrow. I suspect that my simple analysis could easily beat the large polling companies.

Let's start with North Carolina:

NC is 22% African American. In the rest of the south, the percentage of the Democratic electorate that is African American tends to be between 135%-196% greater than their overall share of the population. The average is 166%. If the trend holds I would suspect African-Americans will make up 36.5% of the electorate tomorrow.

State	AA% of Population	AA% of Electorate	Percent Increase
Georgia	30	51	170%
Virginia	20	30	150%
Louisiana	32	48	150%
Alabama	26	51	196%
Mississippi	37	50	135%
Tennessee	17	29	171%
South Carolina	29	55	190%

It has been reported that 40% of all early voting has been by African American. I read that the Obama campaign has made early voting a large part of their campaign efforts so I suspect the overall AA percentage of the Democratic electorate will be slightly less than 40%.

North Carolina is better educated and wealthier than Mississippi and Alabama but slightly less educated and poorer than Georgia. Obama got 26% of the white vote in Mississippi and Alabama but received 43% of the white vote in Georgia. Given the current tone of the campaign and the election trends, I suspect Obama should receive about 34% of the white vote in North Carolina. So here is my breakdown of how North Carolina will look tomorrow:

Vote	Overall	Clinton	Obama
Africian American	36%	10%	90%
White (and Others)	63%	66%	34%
Overall	100%	46%	54%

Headline: Obama Wins 54% to 46%

Next let's look at Indiana:

Indiana should be very similar to Pennsylvania and Ohio, which Clinton won by 10 points. All three are rust belt states and in all three Hillary Clinton has been endorsed by the most prominent statewide Democratic office holder.

Indiana is less educated than Ohio and Pennsylvania and has few African Americans. This is good for Hillary Clinton. The only positive demographic figure for Obama is that Indiana has fewer citizens over 65. I'm assuming these slight differences will balance out. Given all this, Hillary Clinton should win Indiana by about 10 points.

The wild card element is how much the Chicago's media market overlap into Northern Indiana should help Obama. I think the effect will be noticeable but not game changing. While Indianans should as a result "know" Obama better, Obama should be very recognizable figure to most voters after this long primary battle. I would be surprised if the proximity factor helped Obama pick up more than an extra 2-4% of the vote.

Headline: Clinton Wins Indiana 53% to 47%

Barack Obama and the Whitest States in America

I've recently been examining the demographics of the states which have already voted, looking for any interesting patterns. I was not surprised to discover that Barack Obama has not lost a single state where over 18% of the population is African American. Given his strong support among African Americans, it would be nearly impossible for him to lose a contest where they make up a significant percent of the electorate.

What I did find interesting is that Obama has not lost a single state where 1% or less of the population is African American. Surprisingly, he has won Maine, Vermont, Idaho, Wyoming, North Dakota, and Utah; all have populations which are 1% or less African American. Although, Barack Obama did lose New Hampshire, where slightly more than 1% of the population is African American.

A large African-American community or near lack of one are both good indicators of where Obama is most competitive. If the patterns continue, Obama should be expected to not only do well in North Carolina, which is 21.7% African-American, but also be strong in nearly all-white Montana and North Dakota.

*all numbers from US Census Bureau

Forget Pennsylvania, Look to May 6th

Hillary Clinton's three victories last night ensure that the Democratic primary will continue. Both the Clinton campaign and the media as a whole seem to be ignoring Wyoming (Saturday, March 8th) and Mississippi (Tuesday, March 11th). The fact that everyone assumes it is impossible for Clinton to win in either of these states somehow makes them not newsworthy. That said, all the focus now seems to be on Pennsylvania.

This is good news for Hillary Clinton. History, political establishment, demographics, and primary rules all favor Clinton. Clinton has long vacationed in the state. She has the endorsement of the governor. The state's population is solidly white, blue collar workers, and it is a closed primary. Hillary Clinton does need to win Pennsylvania, and all indicators suggest that she will.

If the Obama campaign is smart it will steal a page right out of Clinton's playbook: redefining the battleground. Obama should embrace the idea that the Democratic primary is far from over. This is now a long fight, and there will be no knock-out blows. Make it all about winning states, votes, and delegates. The Obama campaign should fully admit that they do not expect to win Pennsylvania, just like Clinton does not expect to win Mississippi. They should try to shift the talk from Pennsylvania on April 22nd to Indiana and North Carolina on May 6th.

Focusing attention on May 6th shouldn't be hard. North Carolina and Indiana are the 10th and 15th largest states, respectively. They have a total of 187 pledged delegates up for grabs. With over 29% of the remaining unselected 611 pledged delegates, May 6th is now the biggest day left on the calendar. North Carolina's demographics should heavily favor Obama, and Indiana's proximity to Illinois should be very helpful. If the Obama campaign is smart, we will hear "wait until May 6th" from Obama as often as we heard "wait until March 4th" from Clinton.