Monday, October 20, 2008

Opposing Forces Give Uncertainty To Final Weeks

Machine vs Bradley
I started to mull the possible impacts of two unseen forces that may/may not emerge in the final two weeks. The first-which we have heard plenty about in this election-is the tacit effect of racism in voter decision. I went back and researched what discrepancies occurred between polls and election results in a number of African-American races.

It seems that prior to 1996, the swing was about 3.1 points. From 1996-2006, the discrepancy was insignificant. Many look to the New Hampshire primaries this January, where Obama led by 7-8 points the day before, only to lose by 3 in the actual vote. Some pundits suggest that the Bradley Effect is simply a result of faulty polling. I tend to agree. But for the sake of argument, let us consider what would happen if the Bradley Effect did occur.

Let's take a look at some battleground poll averages from RealClearPolitics:

With the current polling numbers, Obama is poised to win the electoral college 364 to 174. Let's apply the 3.1 point swing to every battleground state and see what happens. This would flip the following states for McCain: North Carolina (15) and Missouri (11). This changes the EC to 338-200. Let's see what happens even if the swing is 5 points... McCain now wins Florida (27), Ohio (20), and Nevada (5) as well. This puts us at a 286-252 Obama win. So EVEN if the Bradley Effect swings the results FIVE whole points, Obama will still win by Bush-Kerry margins.

The second silent force that the polls may fail to account for is the power of the ground game. As you can see on the map above, Obama has 18 campaign offices in Indiana alone. Now let us consider the comparison of field offices and the early voting laws to see if we can quantify a possible effect.

If you notice, Florida is the only state in our Bradley group where McCain has more offices. Consider also the available funds that each campaign has until election day. Obama will spend an estimated $200 Million in the final month of the campaign. McCain, on the other hand, will spend roughly $50 Million during that time. Combining the cash values with the ground game, where some of the money will surely be spent on Get Out The Vote, we find that in Ohio, Nevada, and North Carolina, the "machine" advantage helps Obama. In Missouri, the advantage of field offices is somewhat dampened by the lack of early voting. Nonetheless, there will still be a measurable effect.

Though there is no way to quantify this impact, let us estimate that the ground game advantage will lend a 3 point swing. Combined with our earlier figures from the Bradley Effect, this gives Florida (27) to McCain and the remaining four states to Obama. In the end, the Electoral College ends up at 337-201 Obama. Thus, any swing that we might see from the Bradley Effect, even an exaggerated 5 points, will be countered or softened by the converse effect of the ground game/money advantage.


Bill Sanchez said...

But what about mass voter purges, voter fraud, and machine error? There's plenty of things that can go wrong in this election. Let's not get cocky

Sebastian Lederer said...

shouldn't a correct poll already include the better ground game advantage for Obama. Isn't that how they poll people, based on likely voters. So if there is an expectancy, that Obama has a better ground game, one should expect that there are more likely voters from his favored demographics... We'll see. And I agree, let's not get cocky...