The odor of a Poll's cross tabs
I wanted to take some time to look at that “poll” claiming to show that Obama leads Romney among Small Business Owners as it just did not smell right.
Well… The odor is indeed strong and as I always say about polls I need to see the cross-tabs.
Be interesting to see if there is some push back on the data of this poll since it is such and outlier compared to a similar polls of this same segment over the past few months.
First lets look at polling from Gallup that showed Republicans (and leaners) at 49% to Democrats (and leaners) of 42%, it is reasonable to assume that the Democratic advantage by party affiliation in 2012 will be somewhere around non-existent for the general population.
Scott Rasmussen, has also noted. “Over the last 20 years, the number of people who in a presidential election have considered themselves Republicans has been between 32 percent and 37 percent. The number of people who have been Democrats has stayed between 35 and 39 percent in that time frame. Yes, there are changes, but there are not massive swings.”
The poll being touted on the left surveyed 6,145 small business owners and was conducted by George Washington University School of Political Management and Thumbtack.com. Their sample of likely voters claims,
Rep = 29%
Dem = 32%
Ind = 39%
Looking at a survey by a Small Business Association of members and non-members you can see the cause for my questioning the validity of this GWU Poll
Here is the breakdown of Party Registration of Small Business Owners for the following years.
Rep = 54%
Dem = 30%
Ind = 16%
Rep = 48%
Dem = 33%
Ind = 20%
Rep = 54%
Dem = 16%
Ind = 19%
They also asked the question of how “Regularly” they Vote:
Rep = 95%
Dem = 86%
For Comparison they also included Fortune Mag/Zogby Poll from a 2007 Survey and they indicate that Zogby has not followed up that poll with new data.
Rep = 37%
Dem = 35%
Ind = 26%
Given the Party Affiliation reported from Prior polls of this demographic and the results of others Polls of this same group (I have not looked at the cross tabs of those other polls) does it seem reasonable that the Obama who could fill a stadium in 2008 and can’t fill one at his own convention in 2012 has engendered the kind of enthusiasm among Small Business Owners to get them to change their party affiliation by double from 2011? Or better yet if looking at 2008 which showed Dems at a similar ratio of 33% to the GWU survey of 32% that ALL of the swing has been from Rep to Ind?
What would be the result of the GWU Poll if the weighting based on Party Affiliation was more in line with other samples of this same demographic and did not so heavily discount the affiliation of Republicans?