Interview with Dr. Donald Shields

Roxy Caraway
Original Publication Date

Since February 18th 2007 when ePluribus Media published the ongoing study The Political Profiling of Elected Democratic Officials, it has been widely referenced. Senators Dick Durbin (D-IL) and Patrick Leahy (D-NH) quoted the study's findings on the Senate floor; Paul Krugman has summarized its findings; and Dr. Shields was interviewed by Stephen Colbert. 

Roxy Caraway of ePluribus Media interviewed Dr. Shields to discuss all the attention and find out how the on-going study is progressing.

ePluribus Media: Dr. Shields, thank you for taking the time for this interview. We understand there have been many questions about the "basis" for your study. Let me start out with an easy question. What do you mean by investigated?

Dr. Donald Shields: When we say investigated, we mean that the popular press (newspapers or television stations) ran a story naming a particular elected official or candidate as being under investigation by the Department of Justice (DOJ) under the direction of a U.S. Attorney-led grand jury or U.S. Attorney's Office.

ePMedia: Investigation -- isn't that rather ephemeral? What about charges, dismissals, dropped, closed, pled guilty, and convictions?

DS: The focus of the study was the reporting (and thus the leaking) of federal investigations from supposedly secret grand jury subpoenas, federal agent inquiries, and grand jury testimony. Those press reports are not ephemeral at all. That's why we give the approximate date of each story (or indictment, if any) -- so that readers can go search it out for themselves and confirm the investigation.

ePMedia: So, the study tracks the leaks of "investigation." Why is that newsworthy?

DS: As it turns out, our study of press reports of federal investigations has become a terribly important study. It provides powerful sign evidence of the extensiveness of the politicization of the U.S. Attorneys, that isolated anecdotal evidence could never have provided alone.

ePMedia: Can you explain that?

DS: The quantitative study was "predictive" as it raised the issue that given such a disparity in investigations, there might well be a national policy of politicization.

What is now coming out of Washington provides concurrent validity of this longitudinal quantitative U.S. Attorney study. The disparity in investigations highlighted by the quantitative study predicted that evidence would be found in the emails, testimony, and whistle-blowing accounts of the Bush-Ashcroft-Gonzales DOJ indicating that a policy of politicization existed.

ePMedia: I have heard of some Democrats/Republicans/Independents not identified in your sample; doesn't that mean your sample is in error, or you didn't get everyone, or your sample suffers from selection bias?

DS: Selection bias is what the data proves the U.S. Attorneys offices engaged in ... the DOJ was investigating Democratic elected officials and candidates at a far greater rate than Republican and Independent candidates vis-a-vis a comparison to the elected Republican presence in the actual population of elected officials. It's interesting that Spokane's Republican Mayor West was one that we missed. It's also interesting that he was cleared of any federal charges by the U.S. Attorney. That is a courtesy it appears Democratic elected officials seldom received. Usually, the U.S. Attorney replies that "our office can neither confirm nor deny" that the official was under investigation.

ePMedia: What specific procedure(s) did you use to uncover the names?

DS: This study included the names found in a spectrum of media press reports -- from print, television and internet search engines -- beginning in the Spring of 2004.

ePMedia: I know you missed some Republicans, like Alaska's Senator Ted Stevens; didn't you miss others?

DS: Just to keep the record straight, we also missed some Democrats like former Democratic Mayor James Hayes of Fairbanks, AK, and an Independent or two like Detroit's former Councilwoman, Sheila Cockrel. Fortunately, citizen contributors -- Republicans, Democrats, Independents, and even news reporters -- have been sending in new names from all over the country. From Alaska to Alabama and California to Connecticut. We thank each of them for their assistance.

ePMedia: Some of our readers want to know how they can be assured you are reporting all of the Republicans and Independents you come across.

DS: Readers are free to offer their own analyses. One such reader in New Jersey reported 154 cases of public corruption in that fair state and found that the U.S. Attorney Christie had investigated 75% Democrats. That number is consistent with our nation-wide finding.

As we continue our study into 2007, it appears the new cases maintain the percentages reported in the February release of the data (as the theory of large numbers would predict), just as those February numbers reflected the earlier April 2005 interim reports of the data to the Southern States Communication Association in Baton, Rouge, LA (based on 229 cases), and the November 2005 report to the National Communication Association in Boston, MA (based on 267) cases.

What these four studies show across time is the consistency of the percentages of investigated Democratic, Republican, and Independent elected officials, not only with the various reports of the data that we've given, but with independent studies by other citizen researchers.

ePMedia: What about Congressman Foley (R-FL); he was not on your list?

DS: Correct! That's because from the Fall into Winter of 2006, Foley was under investigation by the U.S. House Ethics Committee and the State of Florida, not by any Office of the U.S. Attorney. If the US Attorney's office opens an investigation -- and leaks that they have done so, then we will include it in our study.

ePMedia: Why study press reports of federal investigations of elected officials when everyone knows that what is important is case numbers given after indictments (charges), that can be tracked via Lexis-Nexis?

DS: The DOJ is required by law to report that study every year (although they tend to run a couple of years behind for publication). The DOJ does it for all public corruption cases (without separating elected officials and candidates) and without giving those candidates' party affiliations.

We do not claim our study to be a definitive study of all Federal investigations -- only a study of those publicly reported investigations. Any investigation not leaked and which did not lead to indictment would not have had the political impact or communication valence that we sought to highlight.

A true study of federal investigations would require that the DOJ cough up the information not just on charges, convictions, and sentences related to political corruption, but the actual number of cases investigated involving elected officials and candidates. The politicization of the U.S. Attorneys' Offices is too serious a matter -- to both political parties -- to allow padding of conviction rates by basing them only on indictments as opposed to investigations.

ePMedia: Do you mean that the politicization of the U.S. Attorneys Office is creating an aura of corruption around elected Democratic officials and Democratic candidates?

DS: Yes, exactly. A hundred years ago if a crime was discovered, the U.S. Attorneys' Offices sought to investigate who did it. Now, they rely on electronic surveillance of phones, computers, bank transactions, and so forth (often under the Patriot II Act with absolute abandon). With electronic surveillance technologies, they can surveil without any prior suspicion of a crime. We could call it "Catch and Release."

ePMedia: Can you elaborate on the harms of "Catch and Release"?

DS: Democrats are harmed in much the same way as a "catch and release" fish turns belly-up when returned to the water. That's why the DOJ needs to report the number of investigations, as opposed to indictments.

ePMedia: So, you're saying that reports of investigations are harmful -- even if there has been no crime?

DS: Yes, absolutely. The very fact that an "investigation" was reported in the media is damaging. That's why the DOJ needs to report the number of investigations of elected officials and candidates regardless of whether they led to charges. Most important, they need to report the political party affiliation of those elected officials.

ePMedia: But the U.S. Attorneys often say that their offices are staffed by professionals and they never know the political party affiliation of whom they are investigating?

DS: In this information age, it is simply not credible for U.S. Attorneys to claim they do not know the political affiliation of those elected officials and candidates they investigate. Certainly, Kenneth Starr knew what political party elected President Bill Clinton, and Patrick Fitzgerald knew that the Republican Bush Administration employed Scooter Libby. Well, the same is true of any U.S. Attorney in any of the 93 districts. As the ol' vaudeville and radio comedian Will Rogers used to say, for the government to claim such a thing is just "dust in the eyes of the people."

ePMedia: Again, why is it you didn't use the DOJ's categories of indictments, pleas, and convictions, rather than press reports of investigations?

DS: To have used the DOJ's and U.S. Attorneys' categories seems to us an overly narrow, very legalistic, worldview. As communication professors, we chose to go in another direction. We leave it to someone with a more legal mind to conduct that study, and indeed we would again ask the DOJ to provide their data based on separating elected officials and candidates from other cases of public employees engaging in corrupt acts. That would be a great start toward transparency.

ePMedia: What's it like receiving your own fair share of criticisms?

DS: Although we thrive on the give-and take of honest criticism, we would ask of those who would criticize this study to limit any critique to the study we actually did, not some study we didn't do!

And, of course, don't overlook the point that the major reason we provided the Appendices was so that those who truly want to know the outcome of any given case can simply Google and study the news reports around the dates given and form their own opinion.

ePMedia: Doesn't your study demonstrate that Democratic elected officials and candidates are just more corrupt than Republicans and Independents?

DS: No it doesn't. What some critics seem to have overlooked is the all important Appendix B. That appendix shows that at the both state-wide and national (U.S. Senate and Congress) level, there is no significant difference in the ratio of U.S. Attorney investigations between Democratic and Republican officials, vis-a-vis their presence in the actual elected population. The rate of investigation is the same as the officials' party affiliation in the population at-large. Thus, one clearly would be misinterpreting our full study's data to argue that either Democrats or Republicans are more corrupt.

ePMedia: Well, if Democratic office holders aren't more corrupt, intrinsically, then how are we to interpret your findings?

DS: The findings reported in Appendix B illustrate the point: If a given political party is not more corrupt at the state-wide and national level, why would one or the other party be more corrupt at the local level? The answer is that that particular statistical finding points to a disproportionate investigation of Democratic candidates and elected officials by the Bush Administration's Offices of the U.S. Attorneys.

ePMedia: Is there any evidence that would concurrently validate your study's findings?

DS: Well, recall I noted the New Jersey study above. And, of course, in weeks since ePluribus Media's publication of the quantitative political profiling study, the accuracy of the study's predictive capacity has been enhanced. By this I mean that if the study is correct that at the local level there has been a disproportionate investigation of Democratic elected officials, then it would be predictable that within the Justice Department there would be a policy of encouraging political selectivity in investigation and prosecution of Democrats in ways calculated to influence the electoral process. This appears to have been demonstrated by Bush Administration's memos, e-mails and witness and whistle-blower testimony since the December 2006 Gonzales 8 U.S. Attorney firings.

The quantitative political profiling study's general accuracy is supported by its correct prediction that there is a policy of "punishment and reward" for U.S. Attorneys by the Gonzales-led DOJ based on "political performance." Nor can it be denied -- except perhaps by the most zealous of Administration apologists -- that this cancerous national policy has a local metastasis. For example, Western Missouri's Bradley J. Schlozman, an early Patriot II "interim" assigned straight from the DOJ to Kansas City, investigated and indicted three local Democratic elected officials within a space of three days early in 2007, two of whom were candidates in Kansas City's February and March 2007 City Election. This followed his actions only days before the November 2006 election in the hotly contested Talent-McCaskill race that determined control of the U.S. Senate, where he had filed the same type of "election fraud" cases that the Bush-fired U.S. Attorney Iglesias refused to file in New Mexico.

ePMedia: Many of your critics say your sample is seriously flawed. Is that possible?

DS: I think it is important to note that there is no claim by the DOJ or the critics that the study's presented cases are significantly wrong; nor any DOJ or critics attempt to offer the real numbers (if they are different from ours). Indeed, the DOJ claims both that they do not keep such statistics and that they've investigated lots of Republicans.

ePMedia: Are we to accept the DOJ's explanation?

DS: The DOJ asserts the presumption of not being political. We used to believe that. Now, however, we offer evidence that they are very much political (only in 1 in 10,000 replications could those disproportions have occurred by chance), and in so doing, we overcome the presumption of innocence on the part of the DOJ. They must now offer evidence if they hope to carry the burden of rebuttal.

ePMedia: Aren't there more Democratic elected officials and candidates because the investigations have been in the big cities where Democrats tend to control?

DS: To the extent that this is true, it proves our point: The Bush-Ashcroft-Gonzales DOJ has selectively investigated Democratic elected officeholders and candidates. But, in a specific counter to your question, one is hard pressed to find any of the 93 U.S. Attorneys Offices that do not include large sections of suburban, small town, and rural areas. Suburban, small town, and rural areas are where this fallacious argument would say Republicans live.

ePMedia: So the Bush DOJ may well have been ignoring areas within their jurisdictions where there are Republican officials all the while concentrating on areas where there are Democratic elected officials.

DS: Well, we're not quite ready to concede that all urban areas just elect Democrats, but essentially, our data would support that view.

Moreover, as one connects the dots from the Gonzales 8, the White-out 3, and the Gonzales Baker's Dozen 13 or more, one can see that there are emails, witness testimony, whistle-blower testimony, and other anecdotal evidence that demonstrate the Bush DOJ's systematic plan to do just what our study predicted.

Also, don't overlook the fact that lots of small towns and suburban and rural counties are represented in our data, from Cherokee County, AL to Logan Co., WV, from Benton, TX to Pine Bluff, AR. Democrats appear to be the primary recipients of leaked federal investigations in those places, too!

ePMedia: Many critics have asserted that your study is discredited, a sham, a study that fails the requirements of scientific rigor. What is your answer to their criticism and is it valid?

DS: Every freshmen statistics student knows the stock arguments against any study. But, asserting a stock argument does not mean that it is true. What makes a study stand up to the stock arguments is its intrinsic reliability and validity. Of the 375 names, I know of none that have been presented without the backing of a news report of an investigation. Critics have offered less than a handful of names claimed to be misidentified as to political party. That's an enviable reliability correlation for any scientific study. Similarly, to claim we missed a name here or there in the absence of an actual scientific replication of our study is nothing more than assertion against overwhelming statistical evidence. For our critics to claim they can't replicate the study (Google any name we provided) seems both remiss and negligent.

ePMedia: It is our understanding that Dr. Cragan has declared he cannot continue the project. Can you tell us a little more about that and what is behind his decision?

DS: Due to new time and career commitment concerns, my long-time research colleague and best friend, Professor John Cragan, has asked to bow out as a contributor to this on-going study. I will miss him, but I, and this research project, will persevere.

ePMedia: And how is the study going? I understand you have new updated data that reflects an ever stable finding?

DS: Yes, I have uncovered a number of newly confirmed cases (thanks also to the input of diligent reader responses) for the period of the study (2001-2006) that will be going into the updated and expanded tables. As well, there are already a number of new 2007 cases. The new report will reflect more than 500 cases of press reports of the U.S. Attorneys' Offices investigating elected officials since January of 2001. I am anticipating the release of the newly confirmed data soon.

ePMedia: Thank you for taking this time to explain your on going study.  We look forward to checking in with you again as you continue to research this phenomenon.

Editors note:One of the subjects in the preliminary data presented in the February 2007 opinion editorial "The Political Profiling of Elected Democratic Officials" is Kathryn Shields, the sister of Dr. Shields. Dr. Shields began compiling the information contained in the study in 2003. The study was first released in 2004 and then was presented twice in 2005 before being published in February by ePluribus Media. Because it is a longitudinal study of the Bush Justice Department, it won't be ready for final publication until 2008.

ePluribus Media Researchers, Contributors & Fact Checkers: Cho, standingup, Aaron Barlow, AvaHome, Jenn718