Since our last update of the blogger investigation into the Chrysler dealer closing fiasco now known as Dealergate, another significant benchmark has been met. All of the information from the .pdf list of Chrysler dealers remaining open has been put into spreadsheet format. With both lists now in this form, researchers are finally able to analyze the MOPAR dealer data and find significant patterns buried deep within it.
The number crunchers at Zero Hedge subjected the “safe” dealer data to quantitative analysis and made the first finding: There is a “noticeable and highly positive correlation between dealer survival and Clinton donors.” Doug Ross translates ZH’s math-speak into plain English:
Using raw donor data from OpenSecrets.org (865 megabytes of 2008 individual contribution records) and the list of Chrysler dealerships (both closed and open lists), their preliminary conclusions align with the anecdotal evidence. After a major data cleansing effort, they matched the information against dealer status (safe or nuked); the party affiliation of the contribution recipient; the presidential candidate to whom the dealer contributed (if applicable); the amount of the contribution; and the ZIP code of the dealership.
But one thing bothered Doug about ZH’s inference. The confidence level of their methodology was 87%, which while high, doesn’t quite approach the 95% to 97% levels that he calls the “gold standard” confidence interval. Doug did some investigation of his own and discovered the reason:
The database of dealership owners lists only a single party in each record. But many dealerships are owned by multiple parties. And, in some cases, those parties are billionaires who do not want their names listed as primary owners in these types of documents.
In other words, Zero Hedges greatly understated the already high correlation between dealer donations to the Clinton campaign and their status as a franchise which was allowed to remain in the Chrysler dealer network.
An illustration of how many Democrat donors are masked by the data can been seen in the case of RLJ-McLarty-Landers, a partnership which owns a number of Chrysler dealerships which will remain open while other MOPAR dealers in RLJ’s markets were shut down. Neither former Clinton White House chief of staff Mack McLarty nor Robert Johnson, big-time Democrat campaigners and donors, have names which appear as owners in the database ZH used. Those dealerships are listed as owned by the third partner of the business, Steve Landers. Therefore, none of the group’s six dealerships were counted in ZH’s results.
Speaking of understatements, we earlier reported that by the strangest of coincidences, Steven Rattner, who runs President Obama’s auto task force, is married to Maureen White, was once a fundraising chair for the Democratic National Committee. As it turns out, she was also National Finance Chair of the Hillary Clinton for President Campaign. Go figure…
Thomas Lamb has also been going over the Chrysler dealer data with a fine tooth comb. The aspect of the closings which piqued Thomas’ interest was minority-owned dealerships. It was widely reported that about a hundred minority dealers would be closed, but in fact only 38 of the 170 minority-owned MOPAR deaerships got the axe, among them a number of franchises owned by Hispanics, some of whom donated to or otherwise supported Republicans:
If the trend continues on the closures of African American dealerships versus Hispanic ones, then you are looking at a civil rights lawsuit.
Affirmative action for African American owned dealerships and not for Hispanic owned dealerships does not work out too well.
On Friday, Doug reported on the Left’s reaction to some of the facts uncovered earlier last week regarding Dealergate. Their defense of how Chrysler-Obama-Fiat-UAW Motors is handling its dealers has been weak, ineffective and based on poor methodology, which is to say no methodology at all. It seems to have amounted to a few Google searches, and:
“They somehow forgot to ask the key question: what was the ratio of GOP-to-Democrat donations for the dealerships that were closed by Obama and his minions?”
Doug checked that one out, too:
The ratio of donations for dealers that were shuttered was 42 to 1, GOP to Democrat.
Joey Smith, who found the first clue to the correlation between Clinton donors and Chrysler dealers remaining open, has also begun documenting the influence President Obama’s auto task force has had on Chrysler dealer closings:
“… the Task Force made Chrysler cut some of its dealerships in exchange for some more money from the government.
The final article includes a quote from GM CEO Fritz Henderson saying that GM has been under pressure (presumably from the Task Force) to make faster and deeper cuts.’”
With the General Motors bankruptcy announcement coming today, there will be even more work for the growing army of bloggers who are investigating what the in-the-tank-for-Obama media refuses to even casually look into. GM, of course, didn’t make the critical mistake made by Chrysler. The company has refused to release its lists of open and closed dealerships.
In true horror story fashion, we leave you with some really scary food for thought. With a hat tip to Eric Erickson, may I point out that the expert President Obama has entrusted with the task of overseeing the dismantling of GM for the White House is just 31 years old. He is not a car guy, having never set foot in an automobile factory before being given responsibility for GM. Nor is he a business school graduate. In fact, he’s a law school dropout named Brian Deese. And though he’s not an economist, he was the top economic policy staffer for — cue the soundtrack from the Bud Light “Real Men of Genius” ads — the Hillary Clinton presidential campaign. Just another one of those uncanny coincidences you find when you dig into the dirt from which sprang the federal government’s intrusion into the auto industry.
Update: Zero Hedge acknowledges what Doug and I have taken pains to point out:
“We are only counting majority owners in our donor matches. It’s entirely possible that other influential donors who are minority owners, or former owners or otherwise connected in ways we cannot see are not being counted.”
See related post at Libertarian Republican.
- JP
13% p-value
froid Monday, June 1st at 8:46AM EDT (link)The Clinton data correlation have a 13% p-value (from this, the 87% confidence was derived, but I’m not comfortable using that term.)
As somewhat of a statistician, I thought I’d try to explain what that means.
Suppose you took the list of all Chrysler dealers, but instead of using the actual information on who was closing or not, you assigned them randomly, then you looked for correlations in the data. Since you assigned them randomly, you know that there are none. However, random chance might make a relationship appear anyway. Now, suppose you did this over and over.
What a p-value of 13% means is that, if you did this over and over randomly, 13% of the time, random chance would show a Clinton relationship at least as strong as the one that was found in the real data. Thus, there’s a 1 in 8 chance that the real data is like that; the relationship is random.
That’s not all, though. Remember, going into the analysis, nobody was saying “Clinton donors were spared”, the hypotheses were “Republican donors were targeted” and “Democratic donors were spared”. Those were tested too, but showed no correlation it all. But think for a moment. If the Clinton point and the Republican point had no correlation, but the general Democratic donation had a positive correlation at a p-value of 13%, we’d be reading the same article. Similarly, if the McCain or Republican correlations were negative, we’d be right here.
That’s four possible results that could have been treated as interesting. If they were independent, the chance that one of them was randomly at 87% significance would be about 40%. That is to say, if we did the random draw experiment, 4 in 10 times, we’d see one of the four show a result like that. Now, the different donation variables are not independent. Obviously, knowing someone donated to a Democratic candidate tells you a lot about the chances that they donated to Clinton. Thus, the math gets an awful lot harder to figure out the overall chance that the correlation could be randomly simulated. In fact, I can’t do it without the data itself.
But consider. If none of those four values matched, maybe someone would run 2004 fundraising next? Would a correlation with Kerry donations make indicate something underhanded was going on? Maybe the Obama administration is punishing Bush supporters? What about looking at lifetime donations? The more possible correlations that one finds interesting, the tighter the standard on the significance one must apply to each one.
This case is one I’ve been following closely, because the hubris associated with the act of co-opting a private business and forcing politically motivated business practices on them in this fashion is breathtaking. But an extraordinary claim like that needs solid evidence, and the present statistical analysis is not it.
–
froid
Wow, you know...
evanm Monday, June 1st at 8:54AM EDT (link)Froid, your comment has to be the most thoughtful critique of the Dealergate methodology I’ve read to date. And trust me, I read them all.
I want Josh and Doug and everyone involved to keep digging, because I really do think there’s a story here. However, the story isn’t goig to be statistical unless we factor in location, profit-margins, etc. That’s why I feel, at this point, statistics were the wrong avenue to demonstrate a bias in these closings. We should be focusing on the mechanism. We can’t count on the Democrats to always be so biased that we can demonstrate it statistically with 95% confidence. They’re just not competent enough.
At this point, I need to stop trumpeting the story, because the goalposts keep moving. First it was Democrats vs. Republican donors, now it’s Clinton donors? Most of the posts I read today didn’t even mention that there was no correlation between Democratic donations and closed dealerships. That destroys the credibility of those articles for me, because that’s what I thought Dealergate was… yesterday.
Anyhow, here’s to hoping something comes out of this- and I don’t mean to diminish anyone’s hard work. Just my thoughts.
a misrepresentation of p-value
Streiff Monday, June 1st at 9:19AM EDT (link)p-value can only be used when assuming randomness. If the underlying data aren’t random, which is pretty much the discussion here, p-value cannot, by the definition of the statistical test itself, be used to evaluate the data.
For instance, we have two cards face down. An ace and a king. We try 20 iterations of me picking the cards. I pick the ace 15 times and the king 5 times. By p-value, this is chance.
Same experiment. Though now the cards are marked. I pick the ace 15 times and the king 5 because I want to continue fleecing the credulous and pseudo-statisticians.
Statistically the same outcome,. but because the underlying data were not random it was a test that was misapplied and has no relevance.
“A man does what he can and endures what he must.”
So mean
Neil Stevens Monday, June 1st at 9:46AM EDT (link)You’re wrecking his whole cargo cult operation there. If he can’t pretend he’s being rigorous with statistics, what’s he left with to make himself feel scientific?
Want to run for conservatives? Give.
There Is No Crisis
Be careful using the word "mean" in a discussion about prob & stats
bk Monday, June 1st at 9:55AM EDT (link)Not sure...
evanm Monday, June 1st at 10:01AM EDT (link)I’m no stats nut, but is this actually true?
I thought, in stats, p-value was a representation of the likelihood that a given data set would be produced by a random (i.e. unweighted) process? In the examples above, the p-value would be extremely low for both experiments, indicating a strong likelihood of biased sampling. Not a certainty, but that’s the point, right?
Just looked it up
Neil Stevens Monday, June 1st at 10:11AM EDT (link)Everything I have available to me, in definining the p-value, assumes that the observations are random.
However we *have* the entire set of dealerships that Obama closed, as well as the entire set of dealerships.
There’s no random variable there.
Want to run for conservatives? Give.
There Is No Crisis
True.
evanm Monday, June 1st at 10:17AM EDT (link)But think about the examples above: If you flip the card 20 times, you know all 20 outcomes, not a random sampling of them. The question we’re asking, and the question p-value purports to answer, is whether those outcomes were due to chance, or a relationship.
If the flips were close to 10 for each card, the p-value would be high, i.e. there is a strong probability that the outcome was due to chance.
If the flips weren’t (e.g. 16 to 4,) the p-value would be low, i.e. there is a low probability that the outcome was due to chance.
True, if you’re going to use a sample, it has to be random. But that doesn’t mean you have to use a sample.
Your first paragraph is not quite correct
Neil Stevens Monday, June 1st at 10:27AM EDT (link)The p value tells you if there’s a relationship. It doesn’t tell you there’s no randomness.
Want to run for conservatives? Give.
There Is No Crisis
Right.
evanm Monday, June 1st at 10:29AM EDT (link)nt
no, not correct
Streiff Monday, June 1st at 10:38AM EDT (link)p-value can’t tell you “radomness” if the events aren’t random.
In the card example, getting 16-20, if the cards weren’t marked, right might indicate that you were a good guesser or clairvoyant. If the cards are marked, i.e., the events aren’t random, then p-value tells you nothing.
“A man does what he can and endures what he must.”
what if there is 0% chance
Streiff Monday, June 1st at 10:15AM EDT (link)that an action happened by chance?
Actually we know that none of these closings were chance, they were each decided upon through a deliberative process of some sort. Viewing this series of discrete actions as nothing more than a coin toss is bizarre to say the least.
It is the difference between you getting hit by a meteorite and me bouncing a chunk of concrete off your head because I don’t like you.
“A man does what he can and endures what he must.”
Reference
benjjneb Monday, June 1st at 1:43PM EDT (link)This page does a decent job explaining statistical significance.
In short, to claim statistical significance you must reject the null hypothesis that your sample is uncorrelated with your putative explanatory variable. P-value is the probability the null hypothesis would explain your sample; Froid is not misusing this.
I apparently...
evanm Monday, June 1st at 1:50PM EDT (link)…need to scrub my vocabulary of “chance” and “randomness” and replace them with “null hypothesis.”
You guys are begging the question
Neil Stevens Monday, June 1st at 1:53PM EDT (link)You’re asking us to presume your argument: That Obama isn’t a corrupt weasel, and that the choices of dealerships to be confiscated by the state were unrelated to paying tribute to The Leader’s political party.
You then use that assumption of randomness… to demonstrate randomness.
No sale.
Want to run for conservatives? Give.
There Is No Crisis
Sorry, but ...
froid Monday, June 1st at 2:08PM EDT (link)that isn’t a political position, that’s just how statistics works. Mathematically speaking, you always assume the null hypothesis and then attempt to disprove it by showing how improbable it is.
Want to show that smoking is correlated to cancer? Assume that it doesn’t, and then work to disprove your assumption by showing that your data are not likely due to chance.
Want to show that SAT scores help with college admissions? Assume that they don’t, then work to disprove your assumption.
Want to show that political contributions affected dealship closings? Assume they don’t, then work to disprove the assumption.
Statistics just doesn’t have good tools to prove relationships between variables. All of the tools disprove the null hypothesis, which in this case is that the political donations of the dealer had no effect on the chance that they would be closed.
See below
Neil Stevens Monday, June 1st at 3:08PM EDT (link)The burden of proof is on you to demonstrate Obama’s good faith here, such that these statistics are even relevant.
Want to run for conservatives? Give.
There Is No Crisis
Burden of Proof
evanm Monday, June 1st at 3:21PM EDT (link)I don’t see where that comes from. I understand that we want Obama/Chrysler to prove there was no wrongdoing in selecting dealers, but we’re not going to do that by blindly pretending we have fantastic statistical evidence. We’re going to do it by getting better evidence.
What?
seandparnell Monday, June 1st at 5:36PM EDT (link)Umm, I’m sorry, but there are certain people speculating or even accusing the Obama administration of engaging in particularly shady activities. There’s no legal standard that MUST be followed here (this isn’t a court, after all), but as a general rule I like to see the “burden of proof” on the people alleging a certain thing to be true.
Frankly, if you make an assertion and are then unable to provide evidence that it is true, I’m not going to bother wasting my time trying to disprove it - your own inability to support your contention is evidence in and of itself. “I think it’s true” doesn’t get very far with me.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
And we are obligated to care what gets very far with you why?
janis Monday, June 1st at 5:50PM EDT (link)You’re the one who showed up pooh-poohing the notion that Obama is, as the inestimable Neil Stevens declared, a “corrupt weasel.” And then you tsk, tsked at us rubes for believing that this shining example of the Chicago Machine at it’s finest would for even a moment consider such shady behavior.
Given all that’s happened since The Won showed up on the scene, I think it’s probably more fair that YOU have to show the burden of proof that he’s the innocent lamb that you wish to consider him.
Me? I’ll have mine with mint jelly and a side of roasted new potatoes, thank you very much.
You aren't under any obligation
seandparnell Tuesday, June 2nd at 9:58AM EDT (link)And you seem confused - I’m not pooh-poohing the idea that “Obama is… a corrupt weasel.” I’m casting doubt on the notion that he is guilty of this specific thing that he is accused of.
Look, you can adopt whatever standard you like. “Obama’s a bad man, so he’s guilty until proven innocent of everything he’s accused of” is pretty stupid in my book, but hey, whatever you want is fine.
It’s just worth remembering that there was something referred to here as “Bush Derangment Syndrome,” wherein every action by Bush was automaticaly assumed by certain individuals on the left to be part of the grand master plan to allow Halliburton to put poison in the wells and sell your grandmother to the Arabs. I thought those charges fairly ludicrous, just as I do these.
Obama Derangment Syndrome is not an improvement over Bush Derangement Syndrome.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
So...Did You And Larry Sabato Discuss....
rcov092 Tuesday, June 2nd at 11:05AM EDT (link)how this was Obama derangement syndrome. Sabato is a shaill for the Democrat party, since he is on the board of CCP I will assume a fair amount of influence from him over the organization.
“Not One Red Dime for the NRSC or NRCC till they stop trying to elect liberals”
Join the RedState Strike Force
Suffice it to say...
seandparnell Tuesday, June 2nd at 3:41PM EDT (link)…that your assumptions aren’t going to get you very far. You might want to check the others on the board of acadmic advisors, you’ll se a mix. Then you might want to check our board of directors, who actually, you know, set policy, hire me, guide the organization, etc.
Yeah, that Ed Crane, he’s a real Dem shill…
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
Let's get some sanity
Brad Smith Friday, June 5th at 9:55PM EDT (link)Look, I started the Center for Competitive Politics and I hired Sean. You think I’m an Obama shill? Better do your google search.
Larry Sabato is not on the Board of CCP, he is on the Board of Academic Advisors. If you’ve read what Sabato has said and written over the years on campaign finance, you’ll see that he’s strongly deregulation. And, in fact, if you really read Sabato more generally, you’d know he’s not a Dem shill. In any case, suffice it to say that I asked Larry to be on the Board of Academic Advisors, and suffice it to say that while we value his expertise on the issues that concern us, he has no role in the management of the organization. That’s not what academic advisors do.
The point of Sean’s posts, as I read them - is that we need to be careful of making accusation without good evidence, and careful of assuming that allegations are proven without better evidence. Your post about Sabato having great influence at CCP indicates that indeed you don’t understand that.
Brad Smith
Professor of Law
Capital University Law School
Capital University website
Center for Competitive Politics website
<font color="Beige">**cough**</font>Global Warming<font color="Beige">**cough** </font> <font color="OldLace"> --nt--</font>
6eorge Jetson Monday, June 1st at 9:24PM EDT (link)Wait...
evanm Monday, June 1st at 2:11PM EDT (link)You then use that assumption of randomness… to demonstrate randomness.
Wait, you mean “null hypothesis,” right?
/confused
No
Neil Stevens Monday, June 1st at 3:07PM EDT (link)The math you guys are referring to deals with random variables.
None of it is valid if none of this was random.
Our whole point is that a political process doesn’t generate random outcomes in spots like this.
Want to run for conservatives? Give.
There Is No Crisis
No, it doesn't.
evanm Monday, June 1st at 3:16PM EDT (link)Neil, I actually agree with your last sentence: “Our whole point is that a political process doesn’t generate random outcomes in spots like this.” But that was not the point of the analysis done by ZH, or Doug Ross, or anybody analyzing the statistics of this thing. My original comment on this was that it’s a mistake to make the argument centered on statistics, because it doesn’t matter whether there is a “statistically significant relationship” between dealer closings and donations if the donations were relied on to begin with.
That doesn’t change the fact that froid’s original comment was correct, and that he was not mistaken in interpreting the statistics. We don’t get to redefine things like “p-values” and Baye’s theorum just because the outcomes don’t match the solution we desire. The fact is, I think there could actually be a pretty good statistical case that the Democrats were engaging in pay-to-play over dealerships- but as froid and others have pointed out, this ain’t it.
Excuse me?
Neil Stevens Monday, June 1st at 3:40PM EDT (link)I challenge you to demonstrate where streiff or I have redefined anything here.
That or retract your accusation and apologize for it.
Who’s we anyway? Who are you?
Want to run for conservatives? Give.
There Is No Crisis
No offense was intended, Neil.
evanm Monday, June 1st at 3:50PM EDT (link)In that context, we = everybody.
And no, I wasn’t attempting to accuse you of redefining anything. My point was that the math is valid, even if none of this was random, to prove to skeptics that it wasn’t random. That’s what this whole exercise in statistics is, isn’t it?
To the extent that I implied that you (or anybody, for that matter) were deliberately misleading, I certainly retract and apologize. It was not my intention.
And, since you asked, I’m Evan. I’m not sure what you want to know beyond that.
Y es, the math is valid
Neil Stevens Monday, June 1st at 4:10PM EDT (link)I never said that the cold math is invalid.
But it just has no meaning unless its assumptions are met.
Whether those assumption are met is the point of contention, Evan.
Want to run for conservatives? Give.
There Is No Crisis
Exacty.
evanm Monday, June 1st at 4:25PM EDT (link)And my argument was that the assumptions, as I understand them (which is admittedly little), had been met. Otherwise, I doubt Doug and ZH (and, in reporting, Josh) would have made much a deal of the 87%+ confidence level.
I’m not going to beleaguer the point, however. I can see I should be treading lighter.
Random
benjjneb Monday, June 1st at 5:45PM EDT (link)Neil, you are being misled by the admittedly confusing use of the word ‘random’ in discussions of the null hypothesis used to establish statistical significance.
What is meant is not that the outcome is truly ‘random’ in the sense of being drawn from a hat. It is that it is ‘random with respect to’ the explanatory variable under consideration. More specifically, knowledge of the explanatory variable does not increase your likelihood of predicting the outcome.
As an example, the likelihood of a baseball player hitting a home run in a particular at bat is not ‘random’. It clearly depends on their past performance and factors like the pitcher and ball park. However, it is ‘random with respect to’ whether or not their uniform number is even or odd. This independence can be demonstrated by use of a ‘random’ null hypothesis.
This is the sense in which consideration of p-values determined from a null hypothesis are valid and informative.
Not a good comparison
Neil Stevens Monday, June 1st at 5:46PM EDT (link)Baseball at-bats are not determined in advance.
The burden, again, is on you to show that there was no pre-determined outcome here with the car dealers.
You do *not* get to determine the terms of debate.
Want to run for conservatives? Give.
There Is No Crisis
One more try for me...
evanm Monday, June 1st at 6:33PM EDT (link)True, baseball at bats are not determined in advance. However, in terms of using the data available to determine whether there is a relationship between political donations and dealerships closed, the null hypothesis is that there is no relationship. If the p-value of the data gives a high null hypothesis, then the chances are high that the null hypothesis explains the result.
Neil is correct that, given a high p-value, we cannot say that political donations didn’t determine dealership closings. However, we can say that this data set tells us nothing about the relationship between political donations and dealership closings.
That, at least, is the point I have been trying to make. It’s not that the Administration didn’t choose which dealerships to close based on owner; I suspect it’s likely that they did; it’s just that this data doesn’t say much of anything to that effect.
And even that is a separate issue from who the burden of proof should be on, which, as Neil keeps point out, is in every case the public officials in charge of making the decision.
Neil: Is this any better, at all?
Not really
Neil Stevens Monday, June 1st at 6:56PM EDT (link)It’s like you’re doing everything you can not to cast any suspicion at all on The Leader.
Want to run for conservatives? Give.
There Is No Crisis
I know you have to be on the lookout for trolls and all...
evanm Monday, June 1st at 7:54PM EDT (link)However, I really have been supportive of the Dealergate research that Doug Ross has been doing, and Josh has been reporting on.
Then let's just drop it here. (nt)
Neil Stevens Monday, June 1st at 7:58PM EDT (link)Want to run for conservatives? Give.
There Is No Crisis
If the "sample" is the whole population then there is no null hypothesis
LJ "Beaglescout" Miller Monday, June 1st at 9:06PM EDT (link)You are misrepresenting the data.
“Each of us has a natural right, from God, to defend his person, his liberty, and his property.”
Nobody's moving goalposts.
Josh Painter Monday, June 1st at 9:27AM EDT (link)We’re simply reporting findings as they are brought to our attention. Differenct researchers are interested in different aspects of the data.
And it has not been demonstrated that there’s no correlation between political donations and whether dealerships remained open or were closed. That remains to be seen. We only have one preliminary study done since the open dealer list was plugged into a spreadsheet, and the researchers only counted majority owners in their donor matches. They have admitted that they “might be undercounting for this reason.” Doug labored to point that out, as did I in my diary. Yet neither one of you bothered to acknowledge that in your comments.
- JP
“An armed society is a polite society” - Robert A. Heinlein, “Beyond This Horizon” (1942)
Josh, I get it...
evanm Monday, June 1st at 9:46AM EDT (link)I’m with you 100%. Like I said, I’m convinced there was impropriety here. However, froid has a point- if we test a bucket of different lists (okay, 4 different ones), we’re bound to find something close-to-statistically significant. That doesn’t mean that’s where the misconduct is.
I personally think all of these initial runs will be mooted if we can compare them to dealer profits/losses or revenues, because it appears from the anecdotal evidence that failing dealerships are being saved at the expense of successful ones. But will we ever get that data? In the meantime, we’ve given the wingnuts a big “Not statistically significant” headline to trumpet.
As for ignoring your labor to point out undercounting Democratic dealerships: mea culpa. However, I don’t see anywhere your post (I know you don’t write or speak for Doug) mentioning the p-value, or confidence of a correlation for, Democratic dealerships in general. When I read through all of these stories this morning, I kept asking myself “Clinton, great, but what about democratic dealerships in general?” Doug had the info in a table, but didn’t discuss it. Again, not accusing you of anything short of doing yeoman’s work on Dealergate, but I want you to understand where I, and probably many other conservative readers, are coming from.
"statistically significant"
Streiff Monday, June 1st at 9:56AM EDT (link)you realize that “statistically significant” refers to “statistical significance”, right? And that statistics is the study of “probablility,” right? And that statistics can’t demonstrate malign intent, right?
The only way the issue will ever be definitively proven is if the process for pulling franchises is made transparent on a case by case basis. Until then, it is incumbent upon the people making the decisions to demonstrate that they were made for business reasons, not to hammer the other side’s political donors.
“A man does what he can and endures what he must.”
And given the "transparency" we've seen from Obama so far
bk Monday, June 1st at 10:02AM EDT (link)I can hazard a guess that the probability that “the process for pulling franchises is made transparent” is 0, or maybe ε, but that’s close enough to 0 for me.
Umm...
evanm Monday, June 1st at 10:03AM EDT (link)Everything you just wrote is precisely the point I was trying to make. So… yes?
actually, if you'll read your comments
Streiff Monday, June 1st at 10:10AM EDT (link)you’ll see that you’re continually whining about the lack of “statistical significance.”
If you’re going to troll, for heaven’s sake, be consistent.
“A man does what he can and endures what he must.”
No.
evanm Monday, June 1st at 10:23AM EDT (link)Read them again.
We can’t count on the Democrats to always be so biased that we can demonstrate it statistically with 95% confidence. They’re just not competent enough.
In other words, I read nothing into the lack of a statistically significant result. That’s why we shouldn’t have been focusing on one in the first place.
However, froid has a point- if we test a bucket of different lists (okay, 4 different ones), we’re bound to find something close-to-statistically significant. That doesn’t mean that’s where the misconduct is.
Again, me saying statistical significance doesn’t matter.
And actually… those are the only two times I’ve used the phrase. So I’d say me “whining about the lack of ’statistical significance’” is a fairly major mischaracterization of my trolling at this point.
unfortunately, I did read your comments
Streiff Monday, June 1st at 10:34AM EDT (link)and I’d really recommend you doing the same hitting “post comment”.
like this little gem
Are you saying that you’re only trolling a little bit?
“A man does what he can and endures what he must.”
Streiff...
evanm Monday, June 1st at 10:40AM EDT (link)I would love to respond, but my reading comprehension skills are not up to the task.
I have seriously zero idea what you’re talking about.
Other than “my trolling,” which, I think it was pretty painfully obvious, was supposed to be sarcasm.
...
froid Monday, June 1st at 10:20AM EDT (link)In a nutshell, my comment theme was “The present data analysis does not yet sufficiently support the hypothesis.” I hope that is what you got out of it, because that is what I tried to put in it.
To be clear, I’m not accusing anyone of moving goalposts in a deliberately deceptive fashion. The behavior I’m talking about is perfectly natural and understandable, and is reasonably described as “simply reporting findings as they are brought to our attention”. I’ve done it myself in data analysis projects. However, intent does not change what it means statistically. The more correlations you look for, or might have considered looking for, in a dataset, the stricter you must be with the significance of each correlation treated singly.
I didn’t mention that there might be undercounting due to unknown unlisted partners with unknown donation habits because that isn’t part of the present data analysis, and has no hard data supporting it at all, just one example. There might also be overcounting due to the unknown donation habits of unknown unlisted partners with unknown donation habits in every case but that one. We just don’t know.
But, since you want to talk about it…
One of the reasons that I was initially captivated by the possibility in this story was that the conjectured activity was so easy and simple to do. They had a list of dealers; they had a list of donors. One guy could have matched them up in an underhanded way for his suggested closing/keeping list in an evening. Once we start talking about owners not in the listing however, the plausibility of the scenario gets much shakier to me. Now people have to research the dealerships to make sure they are properly sorting by the politics of the unlisted partners, and the whole thing becomes a much more elaborate and complex scheme. Still doable, but not by one guy in an evening, and it rings less true to me.
This is not to say that it is impossible, but I would need decent evidence. One threshold would be sufficient for me to believe it, but it would take a more stringent threshold for me to want to go to bat with it. Coming back around to this comment, what I am saying is, I don’t have that evidence yet.
What I was trying to do is explain why I believe that the conclusion that you refer to in your post (the sentence you put in bold), was not yet sufficiently supported by the data analysis you cited, at least in its implications.
And lastly, you say above “And it has not been demonstrated that there’s no correlation between political donations and whether dealerships remained open or were closed”. I have bad news. It never will be. Statistics can, strictly speaking, never prove the null hypothesis. It can only fail to disprove it. This is fundamental to the mathematical assumptions on which statistics is built. Has it failed to disprove it well enough? Possibly, possibly not. That depends on how clean the data are, and how much more can be added to the set. That’s a question I can’t answer, because I haven’t seen the data, and I don’t know how easy it is to get a comprehensive list of the names of the unlisted partners.
I just want to make one thing clear once more. I am not assuming that these allegations are untrue. I really really want to know if they are true. I just want to know correctly.
I should probably have taken back...
evanm Monday, June 1st at 10:26AM EDT (link)I think I was the first to use “moving the goalposts.”
That was in haste. My bad.
are you seriously contending
Streiff Monday, June 1st at 10:31AM EDT (link)that Chrysler doesn’t know the name of partners in their franchises? Or the White House would not have access to this?
I really hope that you know the info Chrysler has on franchisees is a little bit more detailed than looking at a /dba name and trying to figure out who is behind it.
“A man does what he can and endures what he must.”
No
froid Monday, June 1st at 10:49AM EDT (link)I’m not. What I’m contending (and I haven’t thought through in much detail) is that adding in multiple partners, only one of whom is listed in the public records of the closing, increases the complexity of the presumed conspiracy. And any increase in the complexity of a conspiracy decreases my tendency to believe it.
Of course Chrysler has the information. Could the person implementing the cheat get it? I don’t know.
However, then they have to make sure that the public list shows the “right” owner, such that their chicanery is hidden. How do they achieve this?
It’s not impossible, really. It might not even be that hard. But it is more complicated, and less elegant, therefore I am less inclined to think it might be true.
However, that’s just my gut. I don’t trust my gut very much, which is why I am far more interested in the data.
I don't understand your point
Streiff Monday, June 1st at 11:01AM EDT (link)We know that the White House was involved in the closing and used confidential Chrysler sales/performance data, allegedly, to close the least profitable franchises. That is not a matter of conjecture.
I fail to see how this “increases the complexity” when it actually means that the complexity was reduced because the White House had all the information Chrysler had.
“A man does what he can and endures what he must.”
...
froid Monday, June 1st at 11:43AM EDT (link)My point is not very exciting, not very interesting, and not particularly relevant to the main argument. I only brought it up originally because it seemed like I had been called out on not addressing the issue and only talking about the math.
When I first heard that this whole thing was a possibility, I imagined that the most likely way it would have happened, if it happened, is as follows.
Someone in a position of influence on the automotive task force decides, either on their own or because of a whispered suggestion from higher up, that closing dealerships is an opportunity to score political points/leverage payback. They quietly and privately put together a list of major donors on both sides to keep handy during discussions. Whenever one of those dealerships comes up, this person leans the conversation toward keep if the dealer donated to Democrats, and leans close if they donated to Republicans.
As the amount of information this person needs to have had access to increases, and the number of decisions and procedures they need to have influenced to their own ends goes up, the scenario becomes less likely to me.
This is not the only way it could have happened, of course. It could have been open knowledge in the room that the plan was to use the decisions politically. That scenario just doesn’t resonate with me personally, as well. Not because I believe the people in the room are righteous and virtuous people who would never do such a thing; they’re politicians, or they were assigned by politician. It doesn’t sit with me because conspiracies like that break down under their own weight and complexity. It’s a lot harder to find 6 or 8 people you trust absolutely than 1 or 2.
It still could have happened that way. Or the first way. It just seems less likely to me. I’m not trying to convince everyone that I’m right about this theory, I’m just explaining what my thinking was.
On the other point, what the statistics have thus far shown, I am trying to convince everyone else, and that’s what I actually care about. This whole other thing was essentially just an off-the-cuff aside about my first reaction to the added quirk of unlisted partners in the dealership data, and was never carefully reasoned to withstand careful argument.
–
froid
a stunning lack of perspective
Streiff Monday, June 1st at 11:59AM EDT (link)If you’ve ever been in the room when a selection committee picks a winner in a contract competition you’ve seen this (or if you’ve seen a grant to an academic researcher evaluated). To deny that it happens and infer that we are ruled by philosopher-kings is simply an insult to our intelligence.
Contracts and grants are awarded every day by the US government based on the political clout of the company/university involved. It doesn’t mean that they aren’t competent, and AFAIK no one is asserting that non-viable franchises have been left alone regardless of politics, it only means that people want to please the boss, or, if they can’t do that they’d prefer to not antagonize the boss.
You’re trying to allege a “conspiracy”, it seems, as a way of denying that a group of political appointees reviewing which franchises to cancel, might actually use politics as part of their decision process.
And I’m still more than a little unclear what the “statistics have thus far shown” since no one, AFAIK, has alleged that the decisions were arrived at on the basis of chance and measuring any deviation from chance, under those circumstances, seems a lot like mathematical onanism.
“A man does what he can and endures what he must.”
I haven't been in the room
froid Monday, June 1st at 2:35PM EDT (link)To your first point:
I haven’t been in the room, and I don’t know what goes on in decision-making like that. I suppose I do believe that if these people have the indecency to make choices for such crass reasons, they have at least the decency to know its wrong and try to hide it. Perhaps I am assuming too much. Maybe it wouldn’t take a conspiracy of people who understood they were being underhanded, and they would just treat it like another day at work. I’d like to believe that isn’t true, but I could be wrong.
Back to the stats, let me try to explain why I keep referring to random chance.
The null hypothesis for this case says that our explanatory variables (the owners donation habits) are unrelated to our data (whether the dealership closed or not). Obviously, something was used to make those decisions, and they were not random. But in the universe where we just know those two facts about each dealership, they are apparently random according to the null hypothesis.
The basic principle behind disproving the null hypothesis (and therefore showing that there is a relationship between donation behavior and the keep/close decision) is to show that randomness isn’t even apparent.
The way you do that is to say “What if, instead of being a result of dozens of factors that we don’t have documented here, the decisions were in fact truly random? Made by a monkey with a typewriter? If that were the case, would I be surprised to see a result like the one I’m looking at now?” If the answer to that question is yes, and the results could have plausibly been randomly assigned, then you can be pretty confident that the actual assignment wasn’t done by using the relationship.
That is what statisticians measure. How likely random chance (which is shorthand for “things that we don’t know about, don’t relate to what we don’t know about, and are effectively chance to us) is to produce the results, by comparing them to what actual random chance would do. In some cases, the comparison is mathematical; the chances are figured out by formula. In other cases, it is empirical; you literally randomly assign the outcomes to the explanatory variables, run your statistical models, and repeat. It is this second process that I would like to do with this data.
G'bye
Neil Stevens Monday, June 1st at 6:07PM EDT (link)I’ve noticed you basically have one mission here: To shill for The Leader on this issue.
I hope they’re paying you or that’s just truly pathetic, to devote your life like some fascist wanna-be.
Want to run for conservatives? Give.
There Is No Crisis
I wonder how come he never said this in 2007
Jack_Savage Monday, June 1st at 6:12PM EDT (link)“It doesn’t sit with me because conspiracies like that break down under their own weight and complexity. It’s a lot harder to find 6 or 8 people you trust absolutely than 1 or 2.”
But they believe that the government of the United States, including its Congress, was duped into a war to benefit Haliburton.
The only thing breaking down are Known Facts….
Excellent work
seandparnell Monday, June 1st at 10:43AM EDT (link)Thanks for your explainations. I was one of the first of the “skeptics” on this issue, although I’m probably more persuaded than you that there just isn’t much here (see my blog post “Obama almost certainly not using ‘enemies list’ to close Republican auto dealers’ at: http://www.campaignfreedom.org/blog/detail/obama-almost-certainly-not-using-enemies-list-to-close-republican-auto-dealers).
Look, auto dealers are overhwelmingly Republican, and if you’re closing auto dealers then you’re closing lots and lots of Republican dealers. Please note that this is a different statement than “the Obama administration/auto task force/Chrysler management did a good job of picking which dealerships to close” or “dealerships should have been closed.” Those are separate arguments, and I make no claim one way or another on those.
But as yet there remains no reason to believe that political payback played any role in closing dealerships, unless you want to count “paranoia” as a reason.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
Your level of confidence in Obama's motives is commendable.
janis Monday, June 1st at 11:07AM EDT (link)I can’t for the life of me figure out where you come up with it though. The man uses political payback all the time, and, by the way, it isn’t paranoia if they really ARE out to get you.
Up to you
Neil Stevens Monday, June 1st at 11:23AM EDT (link)If you’re really not interested in making Obama answer for his actions, that’s up to you.
So if you really have a hard time believing anything political could come from a man who is a product of the Chicago machine and who’s played politics with sharp elbows since his first primary election, that’s you.
We see differently.
Want to run for conservatives? Give.
There Is No Crisis
ah, no
seandparnell Monday, June 1st at 5:47PM EDT (link)One does not have to believe President Obama is not motivated by politics or that the ‘chicago machine’ ways haven’t rubbed off on him to doubt the credibility of this theory. He’s been accused (or rather, his administration has been) nefarious activities, specifically using dealer closings to exact political retribution. The truth or falseness of this accusation does not rest on whether Obama is from the ‘chicago machine’ or is known to play hardball. It rests on whether this specific charge is true.
Is Obama a bad president who often acts in politically calculated ways to reward friends and influence opponents? I’m not saying (here) one way or another, because I’m speaking on behalf of my organization (hence the title, org, and web address under by name). What I am saying is that there is not yet sufficient evidence to support tise charge. Some people aren’t comfortable believing their political opponents aren’t guilty of everything they’re accused of, I prefer to have a little more open mind on this sort of thing.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
Why let him off the hook?
Neil Stevens Monday, June 1st at 5:49PM EDT (link)Why can’t he simply demonstrate otherwise?
Why can’t the President answer for himself?
Why does he have all these people who mysteriously come here trying to sweep it under the rug?
Want to run for conservatives? Give.
There Is No Crisis
hate to be in your court
seandparnell Tuesday, June 2nd at 10:12AM EDT (link)Sorry, Neil, but I participate in public policy discussions all the time. I don’t allow people to make simple assertions without evidence to back them up (rather, I don’t allow myself to give much respect to people who don’t offer evidence to back them up), and when they do offer evidence I go after it - hard - to see if it’s true or not.
Most of the “evidence” I’ve seen offered here are basically theories - “Obama’s a bad man, he’s from the Chicago machine, so this must be true, it’s up to him to prove it isn’t.” That doesn’t get very far with me, and frankly is just “Obama Derangement Syndrome,” not much different than “Bush Derangement Syndrome.”
And I don’t know who is “mysteriously” coming here to “sweep it under the rug.” You’ve got a couple of people saying they’re skeptical, and providing their reasons. I really haven’t seen much refutation/challenge to the reasons we’ve provided for being skeptical, just “Obama bad, story must be true.” Wow, I’m really not impressed.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
a word of advice
Streiff Tuesday, June 2nd at 10:22AM EDT (link)You’ve made a rather complete ass of yourself. The next step will be to send you back to post full time for the benefit of the tens of readers who come to your website.
For someone who allegedly “participate[s] in public policy discussions all the time,” and doesn’t “allow people to make simple assertions without evidence to back them up” you have shown a singular inability to comprehend what is being talked about and to date you have yet to provide anything to this but a series of unsupported assertions, as well as some vigorous fellating of the Administration.
Obviously your sense of shame is somewhat atrophied because otherwise you would have disappeared by now considering the number of self-beclownings you’ve performed so I must act in its stead.
Your next post that doesn’t contain some stunning analytical insight will be your last.
“A man does what he can and endures what he must.”
Sean Parnell: Let me clear it up for you.
Martin Knight Tuesday, June 2nd at 12:00PM EDT (link)What we’re seeing is that dealerships with excellent sales records are being closed while competing dealerships with less impressive numbers are being closed. Now we’re discovering that the bulk of the people being closed are Republican donors while the bulk of the people *not* being closed are Democratic donors.
At first glance, this bears investigation, wouldn’t you agree? Imagine if this was being done under the auspices of a Republican Administration and Democratic donors with excellent sales records were being closed while their less praise-worthy Republican competitors were being shut down? Would the Press be on it?
Now, if you want to know the reason why everyone is obviously thinking that you’re an obtuse idiot, it is the fact that you’ve offered no evidence that you’ve looked through any of the information gathered. The only thing coming from your direction so far is an airy fairy dismissal of the entire thing (i.e. “move along … nothing to see here”) while repeating Nate Silver’s wholly inadequate talking point as if it were the gospel truth.
That’s it. Thanks.
To me, “consensus” seems to be the process of abandoning all beliefs, principles, values and policies. So it is something in which no one believes and to which no one objects … There are still people in my party who believe in “consensus” politics. I regard them as Quislings, as traitors … I mean it.
- Margaret Thatcher
NOTE: “consensus” = “Bipartisanship™”/”Centrism™”
Earning the B team
benjjneb Tuesday, June 2nd at 3:42PM EDT (link)I think we can see the root of this argument in your following sentence:
This statement is directly refuted by the Zero Hedge article being presented in the main post. They consider exactly the question of whether Republican (or Democratic) donations affect the likelihood of a dealership being closed, and find that it has no effect.
I don’t think that is an irrelevant point, and raising it is not “airy fairy dismissal”. Accuracy is valuable, as is well stated in this old EPU diary. Framing a debate and advocacy is important, but correctness has its place as well.
Thanks for getting back to the question
seandparnell Tuesday, June 2nd at 4:22PM EDT (link)I did offer analysis on why this probably wasn’t true (the assertion that Obama was using the GOP dealer list to close political “enemies” dealerships): http://www.campaignfreedom.org/blog/detail/obama-almost-certainly-not-using-enemies-list-to-close-republican-auto-dealers
To recap:
Point 1: Car dealers are overwhelmingly Republican. Thus, if you’re going to be closing dealerships, by definition you’re going to be closing an awful lot of GOP dealerships. This is not different than noting that if you’re hiring for an NBA basketball team, you’re hiring an awful lot of tall people.
Zero Hedge, linked to at the top of this diary, provides the following numbers in her table: Dem dealers, 244, GOP dealers, 514. Gee, that sounds an awful lot like the 2:1 ratio I refer to in my post.
Point 2, admittedly less powerful than the above point, minority owned dealers took the same hit as non-minority owned dealers on a proportional basis. If the Obama folks were using political criteria to close down their opponents, they probably (but not necessarily) would have taken a lesser hit.
And I have looked at the data. I’m not saying it’s not worth looking into. I’m saying it’s foolish to make the accusation before you have any real evidence, only theories. It’s also worth noting that we are looking at a grand total of 53 Clinton donor dealerships out of a total of roughly 3,000 dealerships. Does that suggest to you that a the movement of a few numbers one way or the other could change the results? If 25% of all dealers were supposed to be closed, that means on a randomly assigned basis you could expect 13.25 Clinton dealerships would be closed. What if only 8 were? Wow, only 15%, meaning a 40% ‘advantage’ to Clinton dealers!
When small number shifts make big changes to the results, you’ve got problems saying anything other than “there may be something here, or there may not be.” Which is somewhat short of “Obama’s guilty.”
Also, define “bulk of” and “democratic donors.” Because ‘bulk of” says half or more to me, which really isn’t possible when you look at the numbers. And “democratic donors” - looking at the zero hedge data, are you excluding Obama and other democratic donors from your definition of democratic donors, and only counting Clinton donors as democratic donors? Because what zero hedge’s data show (if it’s accurate) is that Dem and Obama donors don’t seem to have been treated any differently, only possibly Clinton donors.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
You know, if your mind's open enough, everything just falls out. n/t
janis Monday, June 1st at 6:00PM EDT (link)_
Sure
seandparnell Tuesday, June 2nd at 10:16AM EDT (link)I often say there’s a difference between being open-minded and empty-minded - same point, different way of saying it.
But I loathe knee-jerk responses, and I despise theories masquarading as evidence. Theory is what you think happened or how you think something happened (or will happen - let’s not get hung up on wordplay). Evidence is what supports the theory.
What I see here is lots of theory turned into accusation, without much in the way of supporting evidence. Yeah, I’m open minded enough not to get fooled into adopting every conspiracy theory that comes along.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
seandparnell, do you know the difference between a conspriacy theory and investigative journalism?? Apparently not....nt
Aaron Gardner Tuesday, June 2nd at 10:27AM EDT (link)Aaron’s Archive
conform and celebrate diversity….or else!!!
Investigative journalism?
seandparnell Tuesday, June 2nd at 4:26PM EDT (link)Investigative journalism would be having suspicions, looking into it to see if it’s true or not,then presenting what you found along with appropriate conclusions - they’re guilty, they’re innocent, it’s hard to say.
Here, Obama’s guilt seems (for most) to be assumed, and the demand here is that he prove it isn’t true. I’m not a journalist, but that doesn’t sound like investigative journalism to me.
That said, investigate away. I’ve called Obama on enough stuff, but I’ve usally waited until I knew what the facts were.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
shill much?
Streiff Monday, June 1st at 11:36AM EDT (link)Your post was simply a dog’s breakfast of assertions without an attempt at analysis.
Calling people paranoid simply because they’ve done something you haven’t bothered to do, to wit, look at the facts, doesn’t make you look very smart or honest.
“A man does what he can and endures what he must.”
Shill?
seandparnell Monday, June 1st at 5:50PM EDT (link)Please.
An assertion has been made regarding Obama and auto dealers. It’s either true or false. Eventually, someone may come up with evidence that it’s true. Right now, all I’m seeing are a lot of hysterical observartions. Car dealers are overwhelmingly Republican - given that, care to explain to me how it would be possible NOT to shut down an awful lot of Republican dealerships as part of a process that involves shutting down an awful lot of car dealers?
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
I could have said buttwipe or lickspittle
Streiff Monday, June 1st at 5:55PM EDT (link)with equal accuracy.
But shill will suffice for now.
“A man does what he can and endures what he must.”
That was priceless, Streiff!
janis Monday, June 1st at 6:01PM EDT (link)It’s good to giggle helplessly on a Monday. Thanks.
right under the "leave a comment" box
seandparnell Tuesday, June 2nd at 10:18AM EDT (link)is a statement:
“Be respectful, or be banned. No Profanity.”
Worth considering, although given the prevailing winds here I’m more likely to get banned for daring not to instantly assume Obama is guilty just becuase somebody’s got a theory of his latest nefarious doings.
Sean Parnell
President
Center for Competitive Politics
http://www.campaignfreedom.org
We are respectful to those who are respectful
janis Tuesday, June 2nd at 11:41AM EDT (link)to us. Which includes the notion of not thinking you are always right and the smartest man in the room in every discussion. We’ve had several of those and they are no longer here.
May your stay at RedState be as productive for you as it was for them.
I am going with shill
Jack_Savage Monday, June 1st at 6:02PM EDT (link)Although the other two work as well.
I absolutely love the standard of evidence that he requires, though. Maybe we should roll with the “fake but accurate” template of the left - it sure would save us a lot of analysis and research…
The problem with that, Jack Savage, is that
janis Monday, June 1st at 6:10PM EDT (link)if we fake something, it would probably make Obama look BETTER than the real stuff appears to. What I still haven’t seen anyone explain is why Chrysler had to shut down ANY dealerships. Well, other than that the White House made them do it in exchange for more money in the bailout. Is it still being called a bailout or do we have some other nifty new name for stealing our money these days?
You are right
Jack_Savage Monday, June 1st at 6:15PM EDT (link)We would be far too benevolent if we were to fake something. The real thing is too sickening.
I wonder if anyone is questioning whether Obama is a socialist now?
Yeah, I think Russia is. Pravda called what's going on
janis Monday, June 1st at 6:23PM EDT (link)“marxism”, so I guess we’ve passed socialist and are going full-bore Commie.
Thanks, but...
froid Monday, June 1st at 11:46AM EDT (link)… I haven’t done any work yet. All I’ve done is read about and talk about the work someone else has done, and wish I could do it too. I’m not at all sure that I won’t find anything if I do. Once the dataset is made public, if it is, then I’ll have done something worthwhile, I hope.
–
gnfnrf
Not ready to start taunting my Democrat friends
NotSoBlueStater Monday, June 1st at 9:43AM EDT (link)Here’s the digestible sound bite needed here, if it exists:
Percent of all dealerships owned by Republican donors = X
Percent of all dealerships owned by Democrat donor = Y
Percent of dealerships to be closed owned by Republican donors = Z
Percent of dealerships to be closed owned by Democrat donors = Q
I could run with that.
There are some amazing-sounding stats here, but I’m only somewhat convinced that it mean anything because I haven’t seen the simple numbers.
-
The Conservative creed has never offered a life of ease without effort. Democracy is not for such people. Self-government is for those men and women who have learned to govern themselves. - Margaret Thatcher
Number of Republican owned GM/Chrysler dealerships
izoneguy Monday, June 1st at 10:13AM EDT (link)Ready to switch to Toyota, Honda, Nissan, BMW, Mercedes, etc…..
is what Obama should really worry about. Obama is poisoning the well that is the dealership network. Without dealerships it will be hard to support a car company that can only sell to government agencies, and other cities. Of course not every GM/Chrysler can fold or will want to fold but good luck in getting new dealers interested. I guess those big democratic donors might be interested in an Obamanation dealership. GM & Chryslers new motto: Democratically built cars for democrats, nothing conservative about our Outos.
“When the government fears the people, there is liberty. When the people fear the government, there is tyranny.”
Thomas Jefferson
I heard an ad today touting the new car that GM
janis Monday, June 1st at 10:43AM EDT (link)will be building soon, that nasty little lawnmower thing with a windshield. The ad said it would do 90 mph. That was a horrifying mental image. People are going to die in droves in those things. Can we mandate that all legislators who supported this bailout travesty with the attendant bankruptcies have to drive these little death-traps?
Me, I’m sticking with my used Fords for as long as we can get parts to fix them with. I bet a lot of other people will do the same.
Saving a few gallons of gas per week is not worth the safety of my family
izoneguy Monday, June 1st at 10:56AM EDT (link)“When the government fears the people, there is liberty. When the people fear the government, there is tyranny.”
Thomas Jefferson
No izoneguy, Smart car means 'no people'
penguin2 Monday, June 1st at 11:32AM EDT (link)We’ll all end up dead. Maybe that’s their plan. Would save on the health care costs.
My thought, make them drive the same cars they want to sell us. There ought to be a law…
Isn’t that the ugliest car you’ve ever seen?
Resistance to tyrants is obedience to God.
Benjamin Franklin
Ah, penguin2, the image of the presidential motorcade
janis Monday, June 1st at 11:38AM EDT (link)featuring a string of these little go-carts–that’s worthy of a giggle or two. Would look like the circus came to town and brought lots of little clown cars. How very apt for this administration.
Thanks for the laugh, what a visual! (nt)
penguin2 Monday, June 1st at 11:43AM EDT (link)Resistance to tyrants is obedience to God.
Benjamin Franklin
As I peruse the articles relating to this bankruptcy...
$peciallist Monday, June 1st at 11:18AM EDT (link)…I’m having a hard time finding out how many jobs are going to Vanish…
I wonder why?….I know…the Old Media is Dead….
Fox News...
evanm Monday, June 1st at 11:25AM EDT (link)This morning Fox News was reporting 20,000. I suspect that doesn’t include the previous layoffs.
General Motors said Monday that it would close 14 plants
$peciallist Monday, June 1st at 11:37AM EDT (link)… including seven in Michigan, as part of its restructuring in bankruptcy.
It will be Way more than 20,000……way to go, Barry
Michigan's not a swing state.
Moe Lane Monday, June 1st at 11:40AM EDT (link)It’ll give its EVs to Obama in 2012, so what does he care?
Check out my new blog at http://moelane.com/.
http://twitter.com/moelane
My (blogging-related) wish list.
You are right
izoneguy Monday, June 1st at 11:47AM EDT (link)Michigan will go for Obama no matter what…
I am sure we will see plenty of democrats flocking around the Arlington, TX plant for the next 18 months. Texas did not go for Obama. I fear an all out assault is planned for Texas.
I don’t think all the closed Texas GM.Chrysler dealerships will let Obama off the hook that easily.
“When the government fears the people, there is liberty. When the people fear the government, there is tyranny.”
Thomas Jefferson
On behalf of the closed dealers
Josh Painter Monday, June 1st at 1:17PM EDT (link)NADA has retained two law firms. One to represent closed MOPAR dealers and another for closed GM dealers.
But the courts won’t lift a judicial finger to stop the forclosures, both of which are abuses of bankruptcy laws.
http://www.heritage.org/research/economy/tst052209a.cfm.
- JP
“An armed society is a polite society” - Robert A. Heinlein, “Beyond This Horizon” (1942)
Have the Folks in Michigan ever seen a Pitchfork?
$peciallist Monday, June 1st at 11:53AM EDT (link)attention Michigan voters…google search *pitchfork*
Go to Home depot and buy one….they are located right by the Torches..
speciallist, there's probably a union rule about who
janis Monday, June 1st at 12:02PM EDT (link)can operate the things. See your shop steward for the Pitchfork Operator trainees and Torch Lighting journeymen. Carrying the torches requires trained operators as well. Failure to comply with union rules will result in stiff fines and namecalling.
Wow, that many in Michigan?
izoneguy Monday, June 1st at 11:41AM EDT (link)I would have thought they would only close plants in Redstates?
Arlington, TX has a GM plant. They build the Escalade & Tahoe.
Let’s see if they get the ax.
“When the government fears the people, there is liberty. When the people fear the government, there is tyranny.”
Thomas Jefferson
The Saturn plant in Tennessee is going to be "idled", although
janis Monday, June 1st at 11:43AM EDT (link)I haven’t seen yet for how long. That’s going to impact a lot of suppliers and local businesses that depend on the Saturn workers for their income. So the damage will not be limited to just the plants that close.
No, Journalism is dead.
Josh Painter Monday, June 1st at 1:10PM EDT (link)The old media are, unfortunately, still breathing, even if it is starting to sound like a death ratlle.
- JP
“An armed society is a polite society” - Robert A. Heinlein, “Beyond This Horizon” (1942)
Silver Lining
erp Monday, June 1st at 11:38AM EDT (link)So our 12 year old Chrysler Concorde in perfect shape will be a legacy for a grandchildren. Thanks Big O.
erp
I'm waiting for the punitive measures that will prevent us from
janis Monday, June 1st at 11:45AM EDT (link)continuing to drive older cars. There’s no doubt that they have something cooked up to force us into the go carts.
This would all be resolved if Pres. Obama lived up to his cal for transparency in govt...
Aaron Gardner Monday, June 1st at 11:42AM EDT (link)And since we, as taxpayers, now own a significant portion of Chrysler we should be allowed to see the process involved in deciding which dealers were closed.
Can’t we file a FOIA to find out what is really going on?
Aaron’s Archive
conform and celebrate diversity….or else!!!
What's the big deal?
Karina Monday, June 1st at 4:55PM EDT (link)Obama’s got it covered. The car czar is married to Hillary’s Campaign Finance Chair and the guy in charge of GM was the top economic advisor for Hillary. They had a great campaign with NO money problems…at all..except for…well…Vive La France! We’re in big trouble.
Stop confusing me with the facts, I’m making up my own imagination. ~My grandmother who voted for Obama
Deeper Analysis Needed
MSU_Charles Monday, June 1st at 12:57PM EDT (link)My minor in grad school was stats and I still use them when publishing articles, today. My comment on this is not to get too caught up in the fact of the 0.87 P-Value (even though that is a weak value), but to not put too much stock in the correlation. Those of us that use these stats on a regular basis normally use correlations as lower level measures prior to conducting more in-depth analyses. My suggestion for those guys with the data is to run a logistic analysis including other dealer specific variables (e.g. profitability, location, sales volume, dealership seniority, etc.). That would control for more of the entire situation and give us a better understanding.
In addition, I would suggest not just simply examining which dealers donated to each party (i.e. a yes or no dummy variable) but examining the amount of money each contributed. I suspect that when comparing those Repub donating dealers that were closed to Repub donating dealers that were not, those that were not closed donated less money indicating less powerful or less involved political activists for Repubs. On the other hand, I suspect the opposite for the Dem donating dealers. Those that donate more were probably spared. Just a thought.
I also agree with above posts
MSU_Charles Monday, June 1st at 12:59PM EDT (link)on general reporting with simple stats, such as the mean. As Dick Morris commonly states, those with longer arguments usually lose the issue.
However, the deep anlaysis is still important.