Dealergate and Statistics
Yet another statistical analysis of the Chrysler dealership closings has been conducted, although this one appears to be both much more thorough (albeit preliminary) and concentrated on the correct data (my emphasis):
To start with, we pulled raw donor data from The Center for Responsive Politics / OpenSecrets.org for the 2008 election cycle and extracted ~865 megabytes of 2008 individual contribution (“IC”) cycle table entries.
… this particular output is the widest available dataset on contributions. We matched this data against two Chrysler dealer lists:
First, Docket #797 “Document #3” “Schedule of Designated Domestic Dealer Agreements and Cure Costs Related Thereto” (a list of dealers expected to survive).
Second, the famous “Exhibit A” document of dealers to be closed.
We ran binary logistic regressions across the variables. The results are interesting but the most dramatic was saved dealers v. donations by candidate and/or party.
The results of the analysis suggest that donors to Hillary Clinton in the recent presidential race received some preferential treatment. That does not mean that anyone has proven anything, nor that the statistical analysis makes any sort of unassailable case. It merely raises a concern that, given the probabilities, Clinton donors appear to have survived the dealership closings surprisingly well.
This puzzled us. Why would there be an
significantnoticeable (we have rightly been called out for using significant here) and highly positive correlation between dealer survival and Clinton donors? Granted, that P-Value (0.125) isn’t enough to reject the null hypothesis at 95% confidence intervals (our null hypothesis being that the effect is due to random chance), but a 12.5% chance of a Type I error in rejecting a null hypothesis (false rejection of a true hypothesis) is at least eyebrow raising. Most statistians would not call this a “find” as 95% confidence intervals are the gold standard for this sort of work. Nevertheless, it seems clear that something is going on here. Specifically, the somewhat low probability that the Clinton data showing higher survivability of Clinton donors could result just from pure chance. But why not better significance with any of the other variables? Why this stand out?
Then we got to thinking. Steven Rattner, the Car Czar, is married to Maureen White, one-time national finance chairman of the Democratic National Committee. What does Maureen do now? From her website:
Maureen White is currently Chairman of the Board of Overseers of The International Rescue Committee (IRC), a member of the North American Advisory Board for the London School of Economics, and a National Finance Chair of the Hillary Clinton for President Campaign. (emphasis ours)
That website looks dated, but you get the idea.
Again, we want to point out that our findings are preliminary and subject to change. But whatever the result, the Administration has made themselves very vulnerable by taking charge of the dealership closing decisions.
I’m still not sure if there’s anything to the allegations, but there seems to be more than enough anomalies to warrant some questions being asked of the Obama administration. It should be noted that the theory regarding potential shenanigans has morphed from Obama creating a Republican hit list with the closings, to Obama benefiting Democrat donors by allowing their dealerships to survive (and thrive), to Obama’s “Car Czar” rewarding donors to his wife’s favorite political candidate (Hillary Clinton). When the theory moves that much, often it’s a sign that one is fishing for a villain. And despite the evidence amassed in this case showing that an unusual number of Democrat donors are set to prosper from the closing decisions, that may be the case here.
However this all turns out, one thing is certain: by involving itself so deeply in the fate of Chrysler (and GM), the Obama administration invited scrutiny concerning its decision-making processes. Furthermore, in being so opaque about how the government is picking winners and losers (not to mention that it is making these decisions at all), the Obama administration has left itself open to attacks of favoritism. That has nothing to do with Obama or partisanship in particular, but with the fact that unaccountable power rightfully raises fears and suspicions of favoritism. If Chrysler had been left to fend for itself in bankruptcy, none of these questions would have been raised.
The government arrogated to itself tremendous amounts of power over what would normally be private business decisions. In the process, the Obama administration blatantly used its power and influence to reward a favored constituent group (the UAW). Now that statistical evidence suggests more favoritism may have been in play, it’s a little late to cry “conspiracy theory.” Instead, the Obama administration should start opening the books and answering questions.