March 28, 2007
From the desk of Jane Galt:All politics is local
Gack. This terrible, terrible study on Justice department investigations is all over my favourite liberal blogs. Over at Free Exchange, we discuss why it's so awful:
But it seems to me that there is another, at least equally parsimonious explanation: local officials in cities are, as far as I know, overwhelmingly disproportionately Democrats. Cities are also much more likely to be targeted by corruption investigations, for two reasons: they offer more opportunities for corruption, because they provide more services, and officials are much more removed from the local population; and they offer opportunities for bigger thefts. Proportionately, stealing $7,000 from the Phelps, New York town beautification fund may be as big a blow as stealing $5 million from the New York City sewer system, but only the latter is likely to trigger a federal investigation. Also, small towns or counties have fewer officials, which means fewer people in on any corruption, which means fewer whistle-blowers to take down a conspiracy.This thesis would also explain why there is no variation at the national and state levels; statewide offices offer sufficient scope for corruption in any state that any illicit activity is likely to bring Justice swooping down.
And indeed, when I look at the list of local investigations, they seem to be disproportionately concentrated in urban areas. To decide that this is a plot of some kind, I would have to compare the results from the Bush justice department to the Clinton justice department, an obvious check that the authors inexplicably decline to make. Instead, they calculate the chi-square as if Republican and Democratic politicians were randomly (i.e. basically evenly) distributed throughout the population. I'd declare this study not worth the paper it's written on, if only it weren't in electronic format.
At Matt Yglesias', commenter Brett Bellmore points out that it may even worse than that; according to him, the study didn't look at actual incidents, but only media accounts. Needless to say, the majority of media accounts of federal corruption investigations are going to come from major daily papers. And where do you find major daily papers? In big cities. And who dominates the political scene in big cities?
Is this really all you can expect from a communications professor? Someone needs to take their computers away until they promise to take Statistics 101 and actually listen this time, instead of just staring blankly at the pretty shapes on the board.
Update More here.
Posted by Jane Galt at March 28, 2007 6:00 PM | TrackBack | Technorati inbound links"); ?>When I first read accounts of the study, and noted that no comparable figures were given for what occurred in the previous administration, I concluded that the entire thing was likely ridiculous.
Posted by: Will Allen on March 28, 2007 6:18 PMNothing happens for free. Where is the money flow, the power game that publicised this action against these particular people? Someone hoisted this story into the limelight, someone is keeping the scrutiny in the public eye. If we as a nation are being led to further someone's agenda, who is using this story, and what do they hope to gain?
From the attorney refusing to prosecute a porn dealer selling the exact same merchandise as a porn distributor being managed by federal courts, to the appearance of integrity of the others, nothing is really new here. Shakeups, political skullduggery, character assassinations, career lynchings -- dirty tricks, intimidation, and oh-so-polite brutality have been with us since before the nation began. Who is using this story to divert attention from their own warts, or using the national media and public to damage which opponent?
This doesn't take a conspiracy theory. The story is compelling, but the story has been spread too high and wide not to imply someone pushing, for a specific purpose. I want to know that purpose, in order to appreciate what needs to be done and in what order.
Posted by: Brad K. on March 28, 2007 11:49 PMUniversity of Missouri- Saint Louis...UMSL
Maybe they should call Communications PhD's - Propaganda PhD's.
UMSL. The Missouri state system is not like the California system folks. UMSL is not the UCLA of Missouri. It's more like LACC of Missouri.
Posted by: Richard Pointer on March 29, 2007 3:15 AMSo someone ignored a selection bias and we'll have to wait for another study to find its quantitative magnitude. Welcome to academia.
A professor once told me that he would never trust estimates until at least ten pieces of independent research reach the same conclusion. Good advice.
The real problem isn't the study, but that most journalists know nothing about statistics (or anything on which they're reporting). The Economist is a notable exception, which is why I subscribe and advocate that all of my friends do the same. But I think it's really the case that journalistic ignorance is more to blame than a researcher who simply didn't control for everything.
Posted by: Michael on March 29, 2007 3:35 AM
Measuring the rate of indictments is no way to measure this either; the Rovian tactic is to get the local law to proclaim some sort of investigation of the political opponent. An indictment is unlikely to be the end product of such a politically motivated (and likely flimsy) investigation.
You'd want to look for well-publicized investigations happening right around election time. The very aim of such an investigation is the publicity, therefore it seems reasonable to look in newspaper databases.
By the way, wouldn't comparison to the previous administration be implicitly stating that the the Clinton Justice Department was "fair" and the Bush Justice Department is "questionable"?
I suppose the debate has already accepted that the Bush Justice Department is highly questionable, and moved on from there.
I think that it implies that the previous admin was just as questionable. Who could know what fair is?
Posted by: aaron on March 29, 2007 6:31 AMWill's (and Jane's) point is correct, unless we know how this compares to previous administrations it is difficult to say whether the anti-Democratic bias is intentional or systematic.
Posted by: wallster on March 29, 2007 9:07 AM"The real problem isn't the study, but that most journalists know nothing about statistics (or anything on which they're reporting). "
I agree. A researcher can only study what they can observe. Examining available, imperfect data is a perfectly legitimate step in preperation for deciding if it is worthwhile to make the investment in time and money to equip oneself with the means to obtain better data.
I haven't bothered reading the reports (wake me when they do a thorough study), but the authors can't be blamed if the press draws eroneous conclusions about it. If the Republican party hadn't made corruption an official party policy, if dishonesty were not the default position of most Republicans, stories like these would have no legs.
Posted by: Njorl on March 29, 2007 9:17 AMhttp://www.epluribusmedia.org/columns/2007/20070212_political_profiling.html
We compare political profiling to racial profiling by presenting the results (January 2001 through December 2006) of the U.S. Attorneys' federal investigation and/or indictment of 375 elected officials. The distribution of party affiliation of the sample is compared to the available normative data (50% Dem, 41% GOP, and 9% Ind.).
This 50% vs 40% ratio compares to the point that seven times as many democrats were investigated as republicans.
It does not look like your explanation conforms to the data.
Right, as I said, they're acting as if the Democrats and Republicans are randomly distributed throughout the population, which they're obviously not. Democrats are, afaict, concentrated in urban areas, as are the newspapers from which they're pulling their accounts of the investigations. Republicans are disproportionately represented in rural areas which are less likely to attract attention because any crimes will be much smaller (involving fewer people and less money); there are likely to be fewer (or no) major newspaper stories about any investigation (The Finger Lakes Times does not have a crack team of investigative reporters with connections in the US Attorney's office); and small local offices may offer much less scope for crime to begin with than a big city office. The sourcing is garbage, the methodology is garbage, and the result is hence garbage as well. This is Stats 101 stuff here.
Posted by: Jane Galt on March 29, 2007 9:50 AMAlso, don't forget the "looking under the streetlamp" effect - big cities are generally where the US Attorney's offices themselves are.
Posted by: Crank on March 29, 2007 10:21 AMEither these guys deliberately constructed a study in a very poor fashion (it really wouldn't have been much more work to come up with something substantially more meaningful), or they have no idea of how to do the work. Which is the worse thing to say about "researchers"?
Posted by: Will Allen on March 29, 2007 10:24 AMThey didn't even "measure" indictments. In fact, they didn't really measure anything. They googled for "politician" and "investigation" and then sorted the stories and called it "data," without ever establishing a baseline or checking their "data" for relevance.
The two are "professors emeritus." For those of you unfamiliar with the ways of academia, that means "already retired and not drawing a real check, and thus cannot be fired or detenured for whoring their credentials."
As former communications profs, they may well be utterly lacking in quantitative methodogical skills, but they're wll aware of logical fallacies, and use pretty much the entire volume of same in their so-called "study."
Do go read the take-downs in the "More here" link.
Posted by: Tully on March 29, 2007 10:35 AMyour objections don't make any sense.
first, if big city papers cover big issues then having a Federal Prosecutor located in your city investigating politicians for corruption is big news. it is big news even if the prosecuter is investigating corruption in small outlying towns without newspapers willing or able to report same. it is the action of the federal prosecutor that is newsworthy, not the location of the target.
second, federal investigations are undertaken without publicity in order to protect the innocent. that these investigations were leaked to the press in the first place is a very significant consideration in evaluating the imprtance of this study.
third, if the Clinton appointed prosecutors investigated republicans seven times more often than they investigated democrats we would know this by now. anyone with lexus-nexus could duplicate the study for the Clinton years and use that as ammunition against this study. that no conservative whacko has made a similar claim against Clinton is conclusive proof that they cannot make such a claim.
fourth, your pretence to object on grounds of methodology is pathetic. you would make a good flat earther or a libertarian.
Posted by: ken on March 29, 2007 11:54 AMLMAO
"fourth, your pretence to object on grounds of methodology is pathetic. you would make a good flat earther or a libertarian"
You'd make a good Shields or Cragan! :-)
Garbage in, garbage out. And they hand-picked their garbage to tune the smell.
Posted by: Tully on March 29, 2007 12:13 PMAgree with Jane's take on the bad methodology of the study and her opinion of journalists knowledge of statistics, except on one inferred assumption
All that is 'bad' only if you assume that said communications was actually making a good faith attempt to 'get it right' per Jane's definition, and said journalists were also. Is that a good assumption? Maybe they did 'get it right' per their definition.
Posted by: j mct on March 29, 2007 12:23 PMtully, you don't make any sense.
the facts are presented in the study.
you seem to be saying, without offering any proof, that other facts exists which were not included in this study. and that, if included, would bring the frequency of democratic/republican investigations in line with the norm.
if you have any such facts it us up to you to offer them.
until then this study stands up pretty well.
your reliance on dark matter argument is typical of people with little knowledge but lots of sophistry, or of libertarians.
Posted by: ken on March 29, 2007 12:31 PMif the Clinton appointed prosecutors investigated republicans seven times more often than they investigated democrats we would know this by now.
You have a serious need for a clue here. The point of comparing with Clinton was not to show that Clinton was "just as bad." It was to see if perhaps Clinton's USA's had also investigated Dems significantly more often that Reps. If so, then it means something other than political animus is at work.
in line with the norm.
What norm? The point is, we don't know what the "normal" rate of corruption investigations is, so we can't compare this result to it.
Posted by: Rob Lyman on March 29, 2007 12:37 PMKen,
If you'd bothered to follow the links provided in the post, you would see that the Stubborn Facts post lists several easily discovered investigations of Republicans that were not mentioned in the study. That is, the study manifestly undercounted investigations of Republicans. Given that we know the study's count to be inaccurate, how can we give credence to the analysis?
Posted by: TWL on March 29, 2007 1:13 PMRob Lyman, a whole new category of clueless is needed for you.
first, if Janet Reno's justice department investigated democrats seven times more frequently than they did republicans that would have been all over the Rush Limbaugh show, Fox News would have done a weeks worth of specials, the WSJ editorial page would have used acres of newsprint pointing it out, and Isokoff would have written a book. didn't happen.
secondly, if the justice department was anything but even handed in it's handling of corruption charges some crazed wacko consevative would already be in possesion of that information. there is nothing new to be uncovered from the Clinton era.
Posted by: ken on March 29, 2007 1:18 PMOkay Ken, say that they didn't. I can think of a reason why not, which doesn't mean that's it true though, that differ from yours and is perfectly consonant with Jane's point, and is also, perfectly plausible.
Posted by: j mct on March 29, 2007 1:36 PMKen, I don't get it. Apparently it took an "academic study" and accompanying press release to bring to light that the Bush DOJ was investigating more Ds than Rs.
But you argue if Clinton had done the same thing, no study or whatever would have been necessary, because Rush Limbaugh could have what, felt it in his bones?
I guess that explains why the authors of this study didn't bother to examine Clinton--they knew they could rely on Rush's super-duper trick knee to tell them all they needed to know about Reno's DOJ.
Seriously, you can't count on crazed wacko conservatives to establish your baseline for you, you need to actually do the work yourself. And if you fail to, that either makes you a lousy researcher or a dishonest one.
Posted by: Rob Lyman on March 29, 2007 1:58 PMRob Lyman, your point is .... that the Clinton administration was under investigated?
I'm laughing at you Rob.
do you really think that something like skewing the justice department investigations to target republicans could have been gotten away with by Janet Reno? Scafie somehow missed this?
and conversely, if the press reported seven times more stories of democrats under corruption investigation than republicans we wouldn't have heard about it from Fox News or Rush Limbaugh?
i don't know about you but i still get e-mails from your friends listing all the murders Clinton is responsible for.
if democrats were seven times more corrupt than republicans, and this information came from leaks from Clinton appointed prosecutors, you would be crowing about that right now. instead, you got nothing.
Posted by: ken on March 29, 2007 2:19 PMdo you really think that something like skewing the justice department investigations to target republicans could have been gotten away with by Janet Reno?
I never said anything of the sort.
if the press reported seven times more stories of democrats under corruption investigation than republicans we wouldn't have heard about it from Fox News or Rush Limbaugh?
Not unless somebody actually went back, and you know, counted the press accounts. Which apparently has been done for Bush but not Clinton.
you got nothing.
You, on the other hand, have 1) difficulty with reading comprehension, and 2) bald assertions that things must be the way you think they are, because. Just because.
I'm not sure which of us comes out the winner in that fight.
Posted by: Rob Lyman on March 29, 2007 2:31 PMKen,
I felt no need to do the study you outlined. Everyone knows Democrats are seven times more corrupt than Republicans. I for one think the present study is beyond reproach.
Posted by: Rush Limbaugh on March 29, 2007 2:32 PMken,
The point is that it would be nice to know what fraction of Clinton-era USA investigations & prosecutions of politicians were of Democrats, because we can assume that Clinton didin't have it in for Democrats, and then we would have a sort of benchmark against which we could compare Bush Jr. If the Clinton-era investigations were also mostly of Democrats, we can fairly conclude that there's no particular partisan slant now. If Clinton-era investigations have a markedly different political profile, then it begins to look as though political considerations play a role — how big a role would take a lot of other tests to find out.
But at the moment we don't know anything about the Clinton-era investigations/prosecutions, because the study doesn't bother with anything before the current administration. Why not?
Posted by: Michelle Dulak Thomson on March 29, 2007 2:41 PMMichelle,
ken's point is that no such investigation is necessary, because if Clinton had similar numbers, I would be crowing about it, and since I'm not, that proves that Clinton didn't have similar numbers.
Personally, I'm inclined to think that if Clinton was so remarkably evenhanded, ken would be crowing about it, and providing links.
But given that ken also claims to get emails from my friends--a claim I find, to say the least, dubious--you might discount some of his other claims, and award the point to me.
Posted by: Rob Lyman on March 29, 2007 2:50 PMMichelle Dulak Thomson,
you can go all the way back to the Truman era if you want. i guarantee you will not find another justice department, until the current one under conservatives, that skewed it's investigations to target a political party. do a *study*. you won't find anything new.
up until the present administration the justice department could be counted on to be evenhanded in both practice and in appearance.
Rob Lyman, if you really believe that the Clinton administion has been under investigated then by all means go ahead and run a lexis nexus search.
we both know what you will find. you would rather pretend otherwise but even you must know how weak your position is - unless you are a libertarian, they don't know a thing.
Posted by: ken on March 29, 2007 3:06 PMken,
Why don't you do the study since you are the one guaranteeing that one would find no other justice department that "skewed" its administration of justice.
Rob is correct, your argument cuts equally both ways. It is interesting, don't you think, that the authors of the present study chose not to do the comparison, or if they did, they found it inconvenient to report the result. Would not their case for bias be far stronger if they had, or are they just stupid?
Posted by: Yancey Ward on March 29, 2007 3:16 PMYancy, my point is seemingly lost on people who never read anything other than conservative or libertarian produced propoganda.
just a modest study of American history would confirm that the justice department has traditionally been above partisan politics in the implementation of its duties. that is the norm. i point this out only because it seems that liberatarians are not familiar with this fact.
that this information is not included in study under question is no argument against the study.
and trying to claim ignorance of this fact is just plain ignorant. there is a large body of knowledge out their regarding past administrations, including the Clinton administration.
some people here think that the Clinton administration was under investigated. but go ahead, check it out. you won't find anything like what the current administration has done.
Posted by: ken on March 29, 2007 3:47 PMken you're not related to J.Edgar Hoover, are you?
"....Justice Department counted on to be evenhanded in both practice and appearance."
HA!
Posted by: Will Allen on March 29, 2007 3:57 PMKen,
You seemingly keep missing the point. I and most of the others have taken no position on whether or not the Bush Administration's administration of the Justice Department is biased, we are just arguing that the study in question proves nothing in regards to the question. Firstly, a number of clear errors in data mining have been uncovered, including the fact that the present study is undercounting the numbers of Republicans being investigated/indicted by the present administration; and, secondly, we are questioning why the present study did no comparison to the previous administration's record. Such a comparison would serve to address the questions of whether the results of today are actually out-of-line due to bias, or are the result of the heterogeneous distribution of politicians-a very plausible explanation for skewed results. We are questioning your assertion that previous adminstrations were more balanced- an assertion that has been supported by no data as yet-just your appeal to us to learn history. My guess is that, within a week or two, we will have some data from someone who has taken the time to try to do a comparison; then we can revisit the matter.
Posted by: Yancey Ward on March 29, 2007 4:10 PMJeez, Ken. Let me take a stab at this...
"first, if big city papers cover big issues then having a Federal Prosecutor located in your city investigating politicians for corruption is big news."
Ok, There are 4 claims here, you reference one.
1) Big Cities are closer to the Prosecutors, so there's likely to be an easier find.
2) There are more Politicians in Big Cities, so more chances of getting caught.
3) Small town politicians are'nt "big names" so they don't make stories as often.
4) Small towns have small money, and a $2500 scam isn't front-page news.
But that's ok, I'm sure you considered all of this before simply dismissing this out of hand.
"second, federal investigations are undertaken without publicity in order to protect the innocent."
Yep, and some of those investigations were of civilians who had been working with politicians. The Politicans was enver investigated, charged, or even suspected of a crime, but they got on the list. Why? They were in a Political story of an investigation. Why was it a story? Because the investigation was local news of the city the newspaper was located in... odd, that would overcount Demcorats due to city based investiagtions, wouldn't it?
"third, if the Clinton appointed prosecutors investigated republicans seven times more often than they investigated democrats we would know this by now."
Are you stupid, or inattentive? It's likely that Clinto Appointed Prosecutors investigated DEMOCRATS more often than Republicans. Which would exactly match the data of this story. Nobody is making the claim you've debunked here and it would be a confusing claim to make. If the focus is on Cities, and Dems are the major politicians in cities, then Everyone would have more investigations of Democrats in cities. That was the claim, but you missed it entirely.
"fourth, your pretence to object on grounds of methodology is pathetic."
Although, not nearly as pathetic as your understanding of the objections.
Posted by: Gekkobear on March 29, 2007 4:15 PMThe Democratic Party has always been up to its elbows in corruption; it's in the nature of big city machine politics, on which Democrats have held a lock since the days of Tammany Hall through Daley's Chicago ("vote early and often") right up to the present day.
Look at the list of big city mayors below, as a proxy for the politics of the city. I'm not saying that the worthies below are corrupt, but the chances of a big city corruption investigation hitting a Democrat are, well, much more than even money, just on a statistical basis.
(Please pardon the funky formatting - I couldn't get the entries to space properly otherwise.)
City-----------Mayor-----------------Party
Atlanta-------Shirley Franklin-------Democrat
Baltimore-------Sheila Dixon-------Democrat
Boston-------Thomas Menino-------Democrat
Chicago-------Richard M. Daley-------Democrat
Denver-------John Hickenlooper-------Democrat
Detroit-------Kwame Kilpatrick-------Democrat
Houston-------Bill White-------Democrat
Los Angeles-------Antonio Villaraigosa-------Democrat
Minneapolis-------R. T. Rybak-------Democrat
New Orleans-------Ray Nagin-------Democrat
New York-------Michael Bloomberg-------Republican
Newark-------Cory Booker-------Democrat
Oakland-------Ron Dellums-------Democrat
Philadelphia-------John F. Street-------Democrat
Pittsburgh-------Luke Ravenstahl-------Democrat
Richmond-------Douglas Wilder-------Democrat
Salt Lake City-------Rocky Anderson-------Democrat
San Antonio-------Phil Hardberger-------Democrat
San Diego-------Jerry Sanders-------Republican
San Francisco-------Gavin Newsom-------Democrat
San Jose-------Chuck Reed-------Democrat
Seattle-------Gregory J. Nickels-------Democrat
St. Louis-------Francis G. Slay-------Democrat
Washington, DC-------Adrian Fenty-------Democrat
For those who didn't follow the link at the bottom of the post, here's a rebuttal to the concept that this "study" was done with anything other than partisan political posturing in mind. From the horse"s mouth:
http://stubbornfacts.us/politics/partisanship/more_debunking_of_the_political_profiling_study
"Submitted by Donald C. Shields, Ph.D. (not verified) on Sun, 03/18/2007 - 5:35pm.
Below is our reply to the Smerconish diatribe that the Philadelphia Inquirer has yet to print.
STUDY OF U.S. ATTORNEYS’ POLITICAL PROFILING IS SWIFT-BOATED
We are writing in response to a special Sunday Review news report by Michael Smerconish, “Head Strong / Poor Way to Defend Fumo,” that appeared in the Philadelphia Inquirer, February 18, 2007. We have no association with Mr. Fumo, Mr. Sprague, or Mr. Smerconish, but our national study of 375 politicians, who were investigated by U.S. attorneys over the last six years, found that seven of ten investigated politicians were Democratic elected officials or candidates. This disparity constituted political profiling on the part of the Ashcroft/Gonzales appointees. Mr. Smerconish chose to swift-boat us and our research in the tradition of the attacks made on Senator Kerry and Congressman Murtha’s war records in 2004, not to mention Rush Limbaugh’s ridiculing Michael J. Fox’s Parkinson tremors."
Using this language proves to me, at least, that Dr.(!) Shields is not acting as an unbiased, dispassionate observer, but rather as a partisan political operator who published this study as yet another "gotcha" operation aimed at an administration that he obviously feels is worthy of nothing but contempt.
you must know how weak your position is - unless you are a libertarian, they don't know a thing.
I conclude that you must be a libertarian, since you don't seem to understand the weakness of your position.
But seriously, to suggest that our lovely and tolerant hostess, or, say, Eugene Volokh "doesn't know a thing" says a lot about you.
Posted by: Rob Lyman on March 29, 2007 8:39 PMRob,
libertarians may be nice as people but intellectually they are children trapped in the 'me' stage of development.
they can survive only in a well structured liberal environment that indulges their fantacies of self relience while protecting them from the consequences of their desires.
Posted by: ken on March 29, 2007 9:40 PMFolks, please forgive our dear Ken, he was dropped as a child...seven or eight times. Two of those incidents may have invovled deliberate accelleration, but the legal council for his former nanny refuses to comment on the matter.
To prevent further damage, a large portion of his skull was replaced with a protective metal plate, resulting in this. Hence you will get nothing at all like logical consistency out of him, and will only frustrate yourself in the trying. Suggest merely smiling and nodding when he says something, and then talking around him.
Posted by: Logical Reasoning Fairy on March 30, 2007 3:32 AMken,
No one is reading this thread at this point, I suppose, but, well, whatever . . .
up until the present administration the justice department could be counted on to be evenhanded in both practice and in appearance.
Ummm. Someone up above mentioned J. Edgar Hoover. 'Nuff said, I think.
Look, you still don't get our point, apparently. The reason to look at what the Clinton DoJ did is to see whether the current preponderance of Democrats under investigation is a product of the Bush Administration, or just an artifact of who has the wherewithall to be corrupt in so big a way that the Feds might get involved, or a bit of both.
My own guess, like Yancey's, is that Clinton's DoJ prosecuted Democrats vs. Republicans in pretty much the same ratio Bush Jr.'s did — because you don't get corruption on a nationally-noticeable level unless there's lots of money to disburse, and there is lots of it in major cities, and major cities are (see helpful list up-thread) run almost entirely by Democrats.
But I would like to see the numbers. I hope that Yancey is right, and some kind soul will do the research that the authors of this study didn't bother to.
Posted by: Michelle Dulak Thomson on March 30, 2007 2:47 PMI didn't take a scientific sample, but it makes me laugh to read so many people commenting on this issue who don't understand the point of the article. I had a semester of statistics in high school. Almost every time I read about a "study" in the newspaper the writer fails to give an outline of the methodology. If I had a dollar for every study about failing school performance, for example, that never mentions the number of non-English speaking students in the classroom, I would be rich. If their analysis of a statistical study doesn't impress a reader with my weak credentials, what good is it?
Posted by: tyree on March 31, 2007 2:15 PMPeople who think we're even to the point of having to check whether the Clinton administration "also" investigated Democrats more heavily than Republicans aren't paying attention: This study doesn't prove that there are more investigations of Democrats than Republicans. At most, it proves that there are more news accounts of investigations of Democrats than Republicans.
Since when have the news media ever claimed to be even trying for a representative sample?
Ah, let's see, which party is it that keeps telling us that we need less federalism, more federal intervention, more block grants, more community development grants, more funding for NCLB, and so on because poor cities and poor states simply can't afford all of those services? And because, of course, "states rights" is code for discrimination? Which party is that, exactly?
Posted by: Eric H on March 31, 2007 2:48 PMHmm...what major city is majority white and run by White Republicans that the government is ignoring? I can't think of one. n
Posted by: Brian G on March 31, 2007 4:02 PMAs a bit of background on the possible political proclivities of the professors in question - the "study" was published at 'e pluribus media', a site founded by some Daily Kos regulars troubled by the Jeff Gannon debacle a few years ago.
Posted by: Tom Maguire on March 31, 2007 6:46 PMOr, if the D Kos connection does not persuade you, this link might persuade you that Cragan and Shields are not playing with a full deck.
Posted by: Tom Maguire on March 31, 2007 6:52 PMken, you may want to go back and read your own "dark matter" post. It didn't apply very well in your example, but it has applied to mosts of your posts since then.
The study claimed the Bush DoJ was investigating 7 times as many D's as R's. That 7-1 ration is already dumped because of missing data that has already been submitter. But maybe it's still some ratio that is unfair, in appearance at least. The OP gives several possibilities why appearances may deceive in this matter. You have addressed none of them, other than to offer "well, if there were any real info, we would know it already." I don't find that persuasive in the least.
njorl's argument is even weaker: "well, maybe they're not guilty, but it's their own fault for having always been guilty before." Rather circular and impressionistic, I think.
Posted by: Assistant Village Idiot on March 31, 2007 10:03 PMIn the quote by the study's authors they claim 7 out of 10 not 7 to 1. Where does the 7 to 1 meme come from?
I'd fisk out Ken's intentional sins against data and methodlogy, but I see others have already done so.
Garbage in, garbage out. The study is bad because it not properly done in the first place. It's partisan polemic and propaganda dressed in a Halloween costume. The academics involved DO understand methodology and logical fallacies--and they used most of the logical fallacies available.
Bottom line is that the study is so intentionally flawed that it tells us absolutely nothing about anything, save what it tells us about the authors.
Posted by: Tully on April 2, 2007 11:48 AM'njorl's argument is even weaker: "well, maybe they're not guilty, but it's their own fault for having always been guilty before." Rather circular and impressionistic, I think.'
So you think that reputation should be dismissed?
We know that the Bush Justice department behaves unethically. Ethical people who raise red flags get discreet inquiries that look for evidence of genuine wrongdoing before things get publicized. Unethical people get the red flags themselves publicized. If the red flags are innocuous, they have no one to blame but themselves. You are only entitled to a presumption of innocence in a court of law.
It is similar to the Kathleen Willey business. Her accusations were publicized because Clinton destroyed his reputation. I don't blame the reporters who publicized Willey's false story.
Posted by: Njorl on April 3, 2007 11:16 AM