STATS ARTICLES 2009
Dorothy Parker Meets The Marlboro Man:
The Milwaukee Journal Sentinel’s BPA Conspiracy Theory
Dr. S. Robert Lichter, Professor of Communication
Director, STATS and CMPA, George Mason University
It’s not clear which part of the Milwaukee Journal Sentinel's article on STATS needs the most improvement – the reporting, the ethics, or the logic. It’s practically a seminar on how to use rhetorical devices to mask holes in an argument, including ad hominem, red herrings, straw men, begging the question, poisoning the well, selective quotation, false context, and even a faulty syllogism!
Introduction: Just Trying to Be Helpful
STATS tried to help the Milwaukee Journal Sentinel improve its flawed coverage of the chemical BPA. God knows we tried.
In the course of his lengthy article, STATS.org editor Trevor Butterworth solicited input from the lead authors of The European Food Safety Authority’s 2006 risk assessment, the current EFSA BPA panel, NSF International’s risk assessment on BPA and the chairman of the Center for Evaluation of Risks to Human Reproduction’s expert panel on BPA, all of whom provided specific and detailed scientific criticism of the way the Journal Sentinel portrayed the evidence on BPA in its “chemical fallout” series.
He also drew their attention to the published NSF and EFSA risk assessments, as well as other risk assessments and specific studies from around the world, including those by Norway’s Scientific Committee on Food Safety and Neurotoxicity, Denmark’s Environment Protection Agency, Germany’s Federal Institute for Risk Assessment, France’s Food Safety Agency, Food Standards Australia and New Zealand, and Japan's Research Center for Chemical Risk Management, all of which contradicted what Journal Sentinel was reporting.
To help them correct their errors, he carefully spelled out the National Toxicology Program’s 2001 evaluation of the statistical methods required to conduct reliable research on BPA for risk assessment purposes, none of which the Journal Sentinel took into account when establishing what counted as reliable research.
He warned them about their over-reliance on a few environmental activist groups to interpret the science on BPA, and referred them to a recent George Mason University survey of members of the Society of Toxicology, which showed that toxicologists overwhelmingly saw activist groups as overstating chemical risk.
Finally, he made repeated attempts to interview Journal Sentinel reporter Suzanne Rust and Watchdog editor Mark Katches, to let them know about the methodological problems in their piece. Alas, his efforts went unrequited.
As Dorothy Parker Once Said…
Now, the Journal Sentinel has published its response. Alas, the lesson it holds is best expressed in Dorothy Parker’s famous pun – you can lead a horticulture, but you can’t make her think.
Instead of discussing any of the substantive issues, the Journal Sentinel instinctively retreated to the “shut up, he explained” school of reporting, which holds that journalists always operate in the public interest, so any media criticism can be dismissed as expressing special interests. (Out of 22 email inquiries the Journal Sentinel sent to STATS, only one dealt with the substance of STATS’ critique. The rest were mainly of the “who are you and where does your money come from” variety.)
Since these science reporters clearly were not interested in the science at issue, they decided to tar STATS as part of an industry cabal determined to maximize profits by foisting a dangerous chemical on a helpless public. There’s only one problem with this rhetorical strategy – STATS accepts no money from industry sources. So what to do, what to do?
Well, the Journal Sentinel decided to do the next best thing – to assert that STATS is a mere creature of another nonprofit research organization, the Center for Media and Public Affairs, which once did accept a grant from an industry source – not the chemical industry, mind you, but even better – the evil tobacco industry:
“STATS claims to be independent and nonpartisan. But a review of its financial reports shows it is a branch of the Center for Media and Public Affairs. That group was paid by the tobacco industry to monitor news stories about the dangers of tobacco.”
This is a curious phrasing. STATS is not “a branch” of CMPA; they are independent 501(c) (3) nonprofit research organizations, as the annual 990 IRS filings for the two organizations clearly show. They have distinct but related missions, as outlined on their websites. One might just as well (and just as inaccurately) call CMPA “a branch” of STATS, or call both “branches” of George Mason University, with which they are affiliated. The homepage of the STATS website clearly identifies it as a “sister organization” of CMPA. The Journal Sentinel elsewhere calls CMPA “the parent organization” - a sleight of hand used to impugn STATS’ independence by claiming that, even though it does not accept money from industry, it is merely “a branch” of an organization that does.
Enter the Marlboro Man
Of course, the logic of asserting that STATS is a tool of the chemical industry because its sister organization once did a study for a tobacco company is not exactly Euclidian. Rather, it follows the more circuitous route of ad hominem argumentation. This explains why an article putatively about STATS quickly morphs into an attack on CMPA, which continues for several paragraphs. This is a pretty ballsy maneuver, since CMPA’s research output includes a dozen books, scores of academic journal articles and reference works, and hundreds of major media citations, and its lead researchers were described as “America’s preeminent news analysts” in the Annals of the American Academy of Political and Social Science.
But none of this interests the Journal Sentinel. What interests them is the shattering discovery that CMPA once took money from Philip Morris! The source of this shocking news is a draft of a 1994 internal corporate memo that proposed using a CMPA content analysis to advance the company's public relations goals. And the sordid truth is, in 1994 Philip Morris did commission CMPA to conduct a content analysis of “the topical focus, sources and tone of presentation of tobacco stories” in the national media over a period of several months. (Spoiler alert: The coverage was negative.) The Journal Sentinel amusingly describes this as an assignment “to pick apart stories critical of smoking.” For a short introduction on what a content analysis actually does, we refer readers to the “Research Methods” page on the CMPA website.
If the Journal Sentinel had been watching more closely while they were rummaging among 15 year old tobacco industry files (or bothered to look on the CMPA website), they might have noticed that Philip Morris funded another CMPA study – a survey of public attitudes toward the media, which we designed and the Louis Harris polling organization (now Harris Interactive) administered. Our write-up of the findings appeared in a 1997 CMPA “Media Monitor,” which prominently displays the fact that the funding came from Philip Morris
In fact, far from hiding this association, we were pleased to join a company that includes the American Civil Liberties Union, Lincoln Center, the NAACP, and the National Organization for Women, all of which have put Philip Morris money to use in socially beneficial activities.
(Incidentally, contrary to the Journal Sentinel article, Philip Morris did not ask CMPA to “track and report on two or three case studies,” Philip Morris did not contract with CMPA “at least twice during the 1990s to monitor media coverage of tobacco,” and the entire memo, of whose existence we were unaware, is falsely identified by the Journal Sentinel as “Lichter’s proposal to the tobacco company.” We told the Journal Sentinel that all these statements were false; they included them anyway, without providing any evidence to back them up, as none exists. As a communications scholar I find this troubling. But as a civil libertarian, I don’t believe in filing libel suits; anyhow, I’d prefer to get back to pursuing the Journal Sentinel’s red herring before it gets lost at sea.)
Thank You for Not Smoking Gun
And now for CMPA’s own investigative bombshell: “A review of its financial reports” reveals the shocking fact that the Journal Sentinel itself has long had a policy of accepting some tobacco ads, thereby reaping financial benefits from a product that sickens and kills millions of people, which continues to this very day!
OK, our investigation actually consisted of asking publisher Betsy Brenner about it. We’re inclined to accept Ms. Brenner’s assurance that “I do not think it affected any aspect of the paper’s news coverage,” not only because she seems an honorable and forthcoming person, but also because of the empirical evidence, i.e., the paper’s critical coverage of tobacco.
And that raises an interesting point with regard to CMPA’s own “coverage” of tobacco. The Journal Sentinel insinuates that because CMPA accepted money from a tobacco company, we somehow assisted the industry’s devious efforts to make the media landscape safer for smoking.
As it happens, however, we were actively involved in publishing and publicizing research on the environmental causes of cancer, beginning in 1993 (the year before Philip Morris commissioned a content analysis) and continuing over the next several years.
CMPA’s research showed that cancer researchers not only “placed tobacco in a league of its own among cancer agents,” as I wrote, but that they believed the media actually understated the cancer risks associated with tobacco. I testified before Congress on these findings in July 1993, and a condensed version appeared in CMPA's “Media Monitor” newsletter that November.
The research eventually culminated in the 1999 Yale University Press book that I co-authored on the environmental causes of cancer (Environmental Cancer – A Political Disease?), from which the above quotation is taken. (STATS has also repeatedly noted the dangers of smoking, because – unlike BPA – the scientific evidence for the harm it causes is robust. See sidebar).
So the Journal Sentinel's criticism comes down to the following: STATS is not independent and nonpartisan, because it is really part of CMPA (which is false). CMPA once took money from a tobacco company (just as the Journal Sentinel still does). And this taints CMPA's reputation because we... continued to research and publicize the evils of tobacco before, during, and after receiving this money.
And if you’ll believe that, have I got an Alar story for you.
Organizations that commission content analyses generally have a grievance with the media, which they hope to document through the use of social scientific methods. We always tell such groups the same thing – we will find out what's out there, for better or worse.
If the data support their argument, that's great; if not, that's too bad. If they want to argue that the coverage is unfair or unwarranted, that’s a case they have to make.
For example, the same year we studied media coverage of tobacco, we studied television's coverage of Hispanics for the National Council of La Raza, which used the data to begin a dialogue with television producers over their portrayal of Hispanic characters.
CMPA has also studied, among others, media coverage of the Catholic Church for the Catholic League for Religious and Civil Rights, anti-Semitism on talk radio for the American Jewish Committee, news coverage of food and nutrition for the International Food Information Council, and news and entertainment portrayals of government workers for the Council for Excellence in Government.
After reviewing our files at some length, however, I do think the tobacco study was the only one we conducted while engaged in other research whose findings condemned the product of the company that commissioned it.
The Journal Sentinel Cries Wolff
After this lengthy aside, the Journal Sentinel returns to the ostensible focus of the article. Unfortunately, their segue is inauspicious: “The STATS analysis of media coverage of BPA echoed the approach used in the tobacco analysis.” In fact it did not. CMPA performed a social scientific content analysis. STATS conducted a journalistic inquiry based on a close reading of Journal Sentinel’s articles. The different methodologies reflect the different missions of the two organizations and illustrate why one is not a “branch” of the other.
But let that pass, and consider the substance of the Journal Sentinel’s specific complaint: “Trevor Butterworth, editor of STATS, has become BPA's fiercest advocate. He combs the Internet for stories that raise concern about the chemical, even on the most obscure blogs, and he chastises those who claim BPA can be harmful.”
That sounds awfully, well, consuming. However, the record shows that if any party in this dispute is obsessed with BPA, it is not Butterworth. Since his first major piece on BPA appeared in February 2007, he has written over 70 articles and almost 100 blog posts for STATS on the media coverage of such diverse topics as antidepressants, food safety, the risks of lawn mowing, and whether "pixie dust" can regrow human tissue [we kid you not], in addition to contributions to the Huffington Post, Financial Times, Bookforum, and elsewhere. The Journal Sentinel, on the other hand, published some 40 articles on BPA in 2008.
Immediately after presenting Butterworth as a Javert-like nemesis of BPA’s critics, the Journal Sentinel takes its best shot in depicting him as a stealth p.r. operative of the chemical industry, as follows:
“Butterworth offers this advice on a journalism Web site:
‘Forget conventional PR! If some bratty journalist gives you a whack, whack back with obscene, jaw dropping disproportion: knee him in the groin, pull what's left of his hair out, tell him he writes in clichés, and misuses the semicolon, and stomp on his iPhone! A hack is like a bully, and charming a bully is a bit like reasoning with a psychopath or . . . writing a novel on Twitter. For the tough cases, go Dada. Defending the brand means exacting respect and that will come from fear not charm.’ ”
Now, anyone reading this “advice” is likely to think it's pretty strong stuff, maybe even over the top. More importantly, they’re likely to assume that it’s directed toward beleaguered flacks of the chemical industry. So they might be surprised to learn that (1) it’s intentionally over the top; and (2) it’s directed toward beleaguered flacks at the New York Times!
Yes, the Journal Sentinel transposed a humorous comment Butterworth posted on Romenesko about the New York Times’ problems with sniping from media critics, in order to make it appear that he was giving take-no-prisoners advice to the chemical industry. Specifically, he was poking fun at Times flack Catherine Mathis, who told the New York Observer that she had gone on a “charm offensive” to defang dyspeptic Vanity Fair critic Michael Wolff. (Wolff characteristically responded that the offensive had been so small he hadn’t noticed it, prompting Butterworth’s tongue-in-cheek suggestions as to how the Times might adopt a more, um, aggressive tack.)
Didn’t think that kind of sleight of hand happened anymore at “respectable” newspapers, did you? It’s all here in the accompanying sidebar. Read it ‘n’ weep for the Milwaukee reading public.
Journalism on Hormones
But wait! There’s more! Having tried to link STATS to industry by taking material out of context, the Journal Sentinel used a similar technique – this time that of selective quotation – to make it appear that STATS falsely claimed to receive funding from the eminently respectable Hormone Foundation of the Endocrine Society.
Here is the passage from the Journal Sentinel article:
Butterworth listed the Hormone Foundation, an affiliate of the Endocrine Society, as a contributor to STATS on its Web site. But when the Journal Sentinel asked the Hormone Foundation about the apparent support, Paula Correa, the foundation's director, denied it and sent Butterworth this e-mail:
"I have checked your site and saw that the Society/Foundation is listed as supporters of that study. This is absolutely false and I ask that this information be taken down from the STATS Web site immediately."
Now here is the full text of the email from Paula Correa:
“I just got a call from the Milwaukee Sentinel asking me about an apparent conflict of interest between STATS and the Endocrine Society on BPA. They said that you told them that the Hormone Foundation/Endocrine Society has paid STATS $70,000 to support that study [i.e. Butterworth’s BPA article]. I told them that you were incorrect and that the Foundation hired STATS to do a study on a different topic that is yet to be published.
I have checked your site and saw that the Society/Foundation is listed as supporters of that study. This is absolutely false and I ask that this information be taken down from the STATS website immediately.”
In fact Ms. Correa was a victim of a rhetorical technique called “poisoning the well.” The Journal Sentinel falsely asserted to her that STATS claimed the Hormone Foundation/Endocrine Society had funded Butterworth’s study (and then hid this fact from readers when they selectively published part of her response). Then she looked at a STATS website posting which detailed our responses to MSJ’s questions about our funding sources, and saw what the Journal Sentinel reporter falsely told her was there, rather than what was actually there.
In actual fact, as the posting from the STATS site shows, the Society/Foundation was listed as providing funds to STATS, but not for Butterworth’s study. And since this material came straight out our emailed responses to their queries, the Journal Sentinel reporters knew this. But note the article’s careful phrasing, which makes it appear that Ms. Correa was denying that her foundation gave any money to STATS: “Butterworth listed the Hormone Foundation… as a contributor to STATS on its Web site. But … the foundation's director denied it…”
Ms. Correa’s actual concern was that her organization not be perceived as making a donation to STATS, but rather as contracting with STATS for the performance of specific services. At her request, we further clarified the funding relationship by noting that the money was given “for work related to a planned conference on the Women's Health Initiative.” (See sidebar)
The Journal Sentinel also uses the endocrine connection to briefly engage the George Mason survey of scientific opinion on chemical risk (which STATS co-sponsored) as follows: “The survey has been criticized for focusing on toxicologists, as opposed to endocrinologists, who study hormones and their effects on the human body.”
Actually, we haven’t yet encountered this criticism. But we’d love to have a serious discussion about why a survey of scientific opinion on BPA shouldn’t focus on toxicologists, whose field of study is the adverse effects of chemical, physical, and biological agents. Alas, serious discussion still seems outside the Journal Sentinel’s range.
Paging Professor Plato
Finally, and appropriately enough for an article built around rhetorical fallacies, the Journal Sentinel’s argument closes with the flourish of a faulty syllogism. This begins with the paper taking one last shot at the special-interest bogeyman by noting that STATS received funding from a conservative foundation:
“The Sarah Scaife Foundation reported giving STATS $100,000 in 2007, an amount that equaled all of STATS' assets ... The Scaife Foundation funds a number of organizations that promote public policy against regulation, including the Heritage Foundation, the American Enterprise Institute and the Cato Institute.”
Oh my God! STATS was funded by the Scaife Foundation! That’s almost as bad as the chemical industry! Or the tobacco industry! Booga booga! It’s true that 2007 was a tough year for STATS – the only grant came from the Scaife Foundation. As we noted above, however, during the following year, when Butterworth did most of his work on the BPA article, STATS received income totaling $210,000 from several sources, none of them related to Scaife.
And the faulty syllogism? Well, we wondered what the Journal Sentinel article was referring to when it claimed at the outset, “a review of its finances and its Web site shows that STATS is funded by public policy organizations that promote deregulation.” And now we know:
1. The Scaife Foundation funded STATS.
2. “The Scaife Foundation funds a number of organizations that promote public policy against regulation…”
3. Therefore, “STATS is funded by public policy organizations that promote deregulation.”
We know times are tough at newspapers these days, but might we suggest that the Journal Sentinel would benefit from adding a logic editor? (Repeat after me: All men are mortal; Socrates is a man…)
In fact, looking back over the MJS article, it’s not clear which is most in need of improvement – the reporting, the ethics, or the logic. It’s practically a seminar in the use of rhetorical devices to mask holes in an argument, what with the use of ad hominem, straw men, begging the question, poisoning the well, selective quotation, false context, etc. The faulty syllogism that opens and closes the article provides appropriate bookends for this farrago of fallacies.
Coda: The Public’s Interests
Oh, by the way, did I mention the results of CMPA’s Harris Poll on public attitudes toward journalism? Not too encouraging, alas. Barely half (51 percent) said the news media generally get the facts right, 79 percent said journalists should do more stories that point out the errors and distortions in their reporting, 85 percent wanted local news councils to investigate complaints about reporting and issue corrections, 70 percent thought journalists should be required to obtain a license to practice their profession, and 70 percent actually believe that the courts should impose fines on the media for biased or inaccurate reporting.
Not exactly a vote of confidence, I’m afraid. But journalists shouldn’t worry too much about what the public really thinks of its self-appointed tribunes. After all, the poll was funded by a tobacco company.