We can have facts without thinking but we cannot have thinking without facts. — John Dewey


A big Bloomberg News piece on insurance has unleashed the wrath of that industry, entangled New York’s Deadline Club in an awards dispute, and now pits journalist against journalist.

Matthew J. Winkler, Bloomberg’s editor in chief, accused the journalism club of “unethical” conduct in its most recent awards contest for allowing an insurance industry trade group to compromise the “integrity” of the club’s judging.

In a three-page letter to Tim Paradis, an Associated Press reporter and the club’s president, Winkler said that in its jury deliberations it accepted industry allegations that a Bloomberg Markets magazine expose, “The Insurance Hoax,” contained factual errors without talking to Bloomberg. Winkler said the club eventually backpedaled and admitted that the piece didn’t contain errors after all but found other reasons to deny the story an award.

“We don’t know why ‘The Insurance Hoax’ failed to win any of the four awards for which it was a finalist,” Winkler wrote. “We do know the process by which the stories were judged was irregular, opaque and unethical.”

The text of Winkler’s letter, obtained by The Audit, is available here.


The issue goes far beyond a journalism contest and pulls the curtain back on an aggressive public relations campaign by the insurance industry and others against Bloomberg News and a bitter argument over what is and isn’t an “error.”

The campaign has been led by the Insurance Information Institute and its president Robert Hartwig, who in a widely circulated letter to Bloomberg Markets’ editor Ronald Henkoff calls the story “malicious,” “biased, “inaccurate,” “intellectually shabby,” and more. Significantly, he contends the series contains several factual errors, ranging from the misuse of a key industry ratio to blown arithmetic, that are so important the series couldn’t have been written without them.

In an email to me, Hartwig goes so far as to call the series the Million Little Pieces of insurance journalism.

“It made a big splash when it came out but when subjected to fact checking it simply didn’t pass muster,” Hartwig says. “The authors and editors at Bloomberg should be embarrassed because the piece is a disgrace to the Bloomberg name.”

Bloomberg says it stands by its story and, after meeting with Hartwig, declined to run a correction. A spokeswoman adds that Bloomberg is still waiting to hear from the Deadline Club.

Paradis, the club’s president, told me the club plans to write back to Winkler. Meanwhile, he defended its awards process and offered what amounts to a qualified endorsement of the facts in the stories.

“We received complaints about the story and therefore had no choice but to examine it,” Paradis said. “We did so in good faith and gave both sides fair treatment. The results speak for themselves as the story was named as a finalist in several categories. If an entry doesn’t merit being a finalist, it’s not included. We had several categories for which no entrants were recognized.”

I’ll cut to the conclusion: a review by The Audit found no significant factual errors and no errors at all involving the insurance industry. The III’s allegations are unfounded. Details are below.

As a side note, Winkler’s allegation of an “unethical” and “compromised” awards process also appears to lack support. While the club seems initially, and erroneously, to have accepted the industry’s charges of fact problems, it backed off those conclusions after talking to Bloomberg.

To me, though, the issue of whether the story contained fact errors is central. With errors, the series’ credibility is undermined and the focus shifts to the journalists and their methods. With no errors, the series stands as a compilation of facts that portray an industry out of control and ripe for reform.

But it goes even deeper than that: everyone’s entitled to his own opinion but not to his own facts. Anti-Bloomberg forces say the story has errors; Bloomberg denies it; the Deadline Club doesn’t say one way or another.

But without agreement on facts, journalism itself becomes impossible.

The backdrop for all this is a bitter struggle over the public image of the insurance industry and growing (though still isolated) calls on Capitol Hill for its reform. Critics charge the industry with exploiting a dysfunctional regulatory system to cheat policyholders at every turn. The industry says it is in fact suffocating under regulatory burdens and is unfairly vilified by critics who either don’t understand the business or have agendas of their own.

These competing black-and-white portrayals of the industry has only grown starker since Hurricane Katrina, in August 2005, an event that triggered a torrent of lawsuits alleging wholesale abuse of policyholders even as insurers reaped (then) record profits.

Outsiders are often taken aback at the vitriolic nature of insurance debates. There are a few reasons for this.

First, as insurers know well, an allegation of claims-handling misconduct represents more than a simple contract dispute. It’s illegal. To knowingly deny or underpay a legitimate claim is the equivalent of a bank refusing a withdrawal: it better have a good reason. Hence, even a single allegation is a problem.

Second, insurance is a backwater in the mainstream business press, and its actors are thus unused to serious, arms-length scrutiny. The reasons for this are many, but as this post reminds me, insurance is a difficult beat that requires immersion in an ocean of obscure ratios and facts and long, grinding arguments over their meaning with an adroit research operation headed by Hartwig, a Ph.D. Suffice it to say that from a career point of view, most business reporters find the beat unrewarding.

Third, despite this seeming surfeit of information, insurance lacks a useful metric for—of all things—claims-handling performance. While deeply ironic to anyone who has ever hoisted a phonebook-sized volume of, say, Best’s Aggregates and Averages, the industry fact bible, it is nonetheless true that no aggregate figures exist to document the amount policyholders actually demanded from insurers, a key figure that could then be compared to how much insurers chose to pay. There is no “insurer payout ratio,” which I like to call the “Starkman ratio.” Imagine a discussion about a mutual fund without knowing its past performance.

As a result, arguments over insurer claims performance tend to take on almost a theological tone. News organizations point to harrowing individual cases, statistics that show insurers pay little in claims as percentage of premiums, (as low as 55 percent or 65 percent depending on who’s counting) and the industry’s wild profitability as proof that insurers routinely take advantage of policyholders in their moment of need.

Insurers say the sob-story anecdotes are either anomalous or bogus altogether, that the use of certain ratios out of context is irresponsible and misleading, that policyholders are more likely to game the system than insurers, and that critics do not understand or fail to note the industry’s (supposedly) wildly cyclical nature.

Into this howling maw dropped “The Insurance Hoax.” The two-part, 11,000-word series, written by David Dietz and Darrell Preston, compiles court records, whistleblower accounts, internal documents won in discovery, and financial figures to describe profiteering on a vast scale, driven in part by a claims-handling system engineered by consultants to lowball legitimate claims or deny them outright.

The series (I refer to a series but the main disputes center around the first day’s story on insurer claims conduct; the second day deals with regulatory failings) won recognition in other journalism awards contests as well as high praise from me).

The series’s strength, in my view, is precisely its remorseless assembly of a factual record. It includes testimony from former adjusters (“It’s tough to look at people and know you’re lying”) engineers (“we were working for insurance companies and they wanted certain results”), policyholders (“If they defrauded me, how many more are they going to defraud”), and cases in which juries found insurers had acted in deliberate bad faith. The series makes use of blockbuster discovery material, particularly a sequence of slides prepared by McKinsey & Co. for Allstate Insurance Co. in the early 1990s that characterized claims as a “zero-sum game” and recommended insurers adopt a combative posture toward customers who wanted more than insurers offered.

But where I saw strength, critics saw weakness. They believed the story lacked balance and unfairly used anecdotes, even if true, to tar an entire industry.

Sam Friedman, a Deadline Club member and editor-in-chief of National Underwriter, a trade publication, in an online column called the series a “hatchet job” and asked readers to write to the Deadline Club to deny Bloomberg a prize.

In one of his posts, Friedman said he anticipated being dismissed as an “industry shill” but that he found the piece so one-sided and misleading that even its nomination was a travesty.

In an email to me, Friedman says that if he influenced the club, “so be it.”

This particular article, “The Insurance Hoax,” was a hatchet job, plain and simple. Not only did it have numerous factual errors, but the publication did not grant industry representatives a fair opportunity to refute the central assertions in the article, nor did they publish any corrections or even a letter offering the other side, to my knowledge.

In addition, the article painted the entire industry with a single brush, which I feel is unfair, considering that the overwhelming majority of Katrina claims were, in fact, paid, and that the industry laid out tens of billions to rebuild these flood-prone areas—often when the cause, wind or flood, was not clear.

He also takes issue with my findings that Bloomberg did not make factual errors:

Wow, so as long as something is “factual,” it is “accurate” in your view, even if taken totally out of context? Two plus two does equal four, but if there are six elements, and one arbitrarily decides to ignore the other two, the story is fine??? Very odd logic, sir.

The fight shifted venues earlier this year when the Deadline Club named “Hoax” as a finalist in four categories, sparking objections from Friedman, Hartwig, and others who pointed to what they saw as the series’s flaws and called on the club to reconsider. The club then wrote Bloomberg.

According to Winkler’s letter, Paradis on May 5 emailed a four-page memo to Bloomberg with nine “allegations of inaccuracy” and said the club’s research “highlighted parts of the article with incorrect, inexplicable or questionable material.” The issue at this point was whether the stories should be disqualified from the contest altogether.

On May 9, Paradis and other club officials reviewed the stories in a conference call with Bloomberg editors.

In the end, the club took a middle path: it denied the series an award but left it as a finalist. Paradis declined to discussion how the club made its awards judgments but emphasized that the series was not removed as a finalist.

Still, someone has to settle this “error” question. So, ladies and gentlemen, if I could ask you to kindly step behind CJR’s specially reinforced, lead-lined blast barrier: The Audit is going in.

In his August 2007 letter to Bloomberg Markets’ Henkoff, Hartwig lays out several alleged errors, including one that is indeed a flat mistake: Bloomberg said the storm killed “more than 16,000” people when the figure was closer to 1,600.

Bloomberg agreed and corrected it in the text (though it doesn’t note the correction online).

The rest are actually disagreements, and I’ll handle the closest calls:

Alleged error: Bloomberg said insurers paid out only 55 percent of premiums to policyholders in 2006; Hartwig says the correct number is 65 percent.

The Audit explains: Bloomberg is relying on a figure known as the “loss ratio,” which is the amount actually paid to policyholders excluding various expenses directly associated with the claim, including legal fees, the cost of adjusting and investigating the claim, etc.

Hartwig says it is irresponsible and misleading to exclude these expenses, which also benefit policyholders. I say that Bloomberg language was precise: it was characterizing the amount paid to policyholders. It has a more than reasonable basis for its position (and is not alone in making it), and did not make an error.

Alleged error: Bloomberg said insurer profits since 1994 increased by “an annual average of 46 percent.” Hartwig says the correct figure is 15.9 percent.

The Audit explains: Hartwig believes the appropriate measure is compounded annual growth. Bloomberg took the percentage increase from one year to the next and averaged them all. This emphasizes the large jumps industry profits took from one year to the next.

Bloomberg’s method is unusual, but allowable. It’s certainly not a factual error.


Alleged error: Hartwig says Bloomberg “incorrectly insinuates that states have no prosecutorial power” over insurers.

The Audit explains: What Bloomberg said was that state insurance departments have no prosecutorial powers and that prior to Katrina no state or federal prosecutor had started a criminal investigation of insurer claims handling. The first statement is categorically true, and Hartwig’s letter offers no refutation of the second.

Alleged error: Bloomberg says insurers spent $98 million on lobbying. Hartwig says that figure is an exaggeration because it includes life and health insurers when the story dealt with property/casualty insurers.

The Audit explains: I think this one speaks for itself. Bloomberg never said otherwise. It’s not an error.

Alleged error: Bloomberg said the federal flood program “helped the industry increase profits by 25 percent in 2005, the year of Katrina.” Hartwig says the finances of private insurers are “entirely independent” of the flood program and that what the flood program did or didn’t do could not have an impact on insurers’ profits.

The Audit explains: While Hartwig is right that the industry’s and the government’s books are separate, private insurers are actually intimately involved in the flood program. Private insurers adjust federal flood claims for the government and were charged with the task of assigning the cause of Katrina damage to either wind (meaning insurers paid) or flood (meaning the government paid). Litigation and congressional testimony supports Bloomberg’s assertion that insurers on at least some occasions attributed losses caused by wind to flood damage, and thus avoided payment at the government’s expense.

Bloomberg’s phrasing was clumsy, however, in that it could lead to the conclusion that the flood program helped boost insurer profits by 25 percent when the sentence merely meant that profits grew 25 percent and the flood program helped. It is beyond question that this is true.

Audit’s bottom line: Bloomberg could have saved itself some headaches by using figures that were less tendentious: if it said insurers paid out 65 percent, instead of 55 percent, is that something insurers should be proud of? Should Bloomberg have said profits increased “six fold,” rather than an average of 46 percent a year in a given period?

It could have but didn’t have to.

Another point: facts are important, but arguments over the number of angels dancing on the head of a pin can serve as an unhelpful distraction from the need to address what are obviously real problems.

Bloomberg’s story had a point of view: that the insurance claims system is broken to the point that underpayment is now routine, indeed systemic. It used facts to support its point.

To me, that’s a strength. But, of course, that’s just my opinion.

If you'd like to get email from CJR writers and editors, add your email address to our newsletter roll and we'll be in touch.

Dean Starkman Dean Starkman runs The Audit, CJR's business section, and is the author of The Watchdog That Didn't Bark: The Financial Crisis and the Disappearance of Investigative Journalism (Columbia University Press, January 2014). Follow Dean on Twitter: @deanstarkman.