5/30/2009

No relationship between the Chrysler dealers who lost their dealerships and them being Republicans

My son Maxim put together this information. It is pretty clear to me that there is no relationship here and that conservatives would do well to stay away from this claim.

Methodology

50 dealerships were picked at random from the list of 789 closing dealerships, and another 50 from the list of 2392 dealerships that are to remain open. I used a random number generator to determine which dealerships were selected.

The two groups of dealerships happen to be spread through the country similarly. In our sample of 50 closing dealerships, 60% came from states that Obama won. In the sample of 50 open dealerships, 62% came from states Obama won. The average amount Obama won by in these states was also very similar for both groups (closing dealerships were in states where Obama got an average of 51.36% of the vote, whereas open dealerships were in ones where he got 49.84% of the vote.)

Data on contributions are from Opensecrets.org and are for the 2008 election cycle.

Results

Among closing dealerships, 6 out of 50 donors (12%) donated to Republicans. 4 out of 50 donated to Democrats (8%)

Among open dealerships, 7 out of 50 (14%) donated to Republicans. 5 out of 50 donated to Democrats (10%)

The average size of donations was nearly identical for both parties, in both samples.

While there is no evidence of a systematic political bias, that does not rule out the possibility that there were a few politically motivated decisions regarding the biggest donors, as some are suggesting might be the case.

Our data do indicate one major factor in determining whether a dealership was closed or not: The size of the dealership, measured by the number of product lines carried (the 4 lines are Chrysler, Jeep, Dodge, and Dodge Truck.) The average store that will be closed in our sample carries just 2.5 of those product lines, whereas the average store that will stay open carries 3.64.

Labels: , ,

4 Comments:

Blogger Conant said...

How representative a sample is 50?

5/30/2009 6:26 AM  
Blogger John Lott said...

Given the size of the overall sample, looking at 100 cases (50+50) is fine. If the means were at least a little different, I would look at more, but these two samples look very similar.

5/30/2009 1:13 PM  
Blogger Nelson said...

I don't understand why only 100 dealerships were sampled, when we supposedly have data on all 789. From what I've read on the emerging research, the dealerships being closed overwhelmingly gave to Republican interests, to the tune of 90%, and only a very small amount donated to Obama's campaign. I thought that was data on the whole, not just a sample. Maybe you can clear this up for me?

6/02/2009 9:31 AM  
Blogger GunGeek said...

It sounds like they just looked at those dealerships where the dealership itself (the corporate entity) donated. I can't imagine that only 10% of the owners (stereotypically being rich old men) of those dealerships made donations.

I'm sure it would take a lot longer to figure out to whom the owner(s) donated, but that's what would have to be done before any kind of conclusion could be made.

6/08/2009 11:24 AM  

Post a Comment

<< Home