‘How Bermuda rigs rates’ (October 25)

The following is an analysis that is part of a graduate thesis. Learn more.

Despite its headline, St. John did not discuss in this story very much “rigging” of rates by Bermuda, or really much about that country in general. Instead her focus was on the reinsurance companies located there, and how their rate-setting practices affected Florida insurance companies and in turn Florida homeowners. More specifically, the issue addressed here concerned the years after Hurricane Katrina: How did reinsurers react and what were the effects of those reactions?

Reinsurers are companies that provide insurance to insurance companies designed to pay out in the event of a catastrophe requiring more funds than the insurance company has. St. John concluded that Floridians “took [a] hit” (3) from reinsurers in three primary ways.

First, that reinsurers “reduced the storm coverage they were willing to give Florida,” or in some cases “refused to write policies for months, convinced they could extract an even higher price from insurers that neared collapse” (5). Second, that the reduction in supply caused insurers to “still [cobble] together hurricane protection in August and September, during the peak of danger, and [pay] three times the January rate” (6). Third, that “the cost was paid by Florida property owners” (7). The first and third conclusions were descriptive claims about a previous state of the world; the second claim contained the descriptive claim about the way the world was, and also a descriptive conclusion regarding causation.

What evidence did St. John offer to support her first claim regarding reinsurers’ reduction or purposeful limiting of available coverage? One part of the claim was better justified than the other. The only evidence offered to demonstrate that insurers “reduced the storm coverage” they offered was St. John’s reference to “reinsurance contracts and comments by executives” that showed that reinsurers reduced the capital they offered to Floridians “even when they had money in the bank and board approval to use it” (24). This was essentially a restatement of the original claim with an oblique citation to “contracts and comments,” but readers were not told anything about the contracts or comments. It was left to readers to accept the claim on St. John’s authority.

Given the relatively weak evidence available for the first half of the claim, it was surprising to see very strong evidence made available for the second half, that reinsurers refused to write policies so as to squeeze insurers. St. John quoted several conference calls from within reinsurance corporations, with executives telling analysts that “we held back capacity” or that insurance companies would eventually return as “distressed buyers” (27–37). Admittedly, it wasn’t clear how St. John accessed these conference calls. But assuming that she accurately reported their substance, then readers had plenty of evidence to use in accepting the claim. St. John claimed that “some” reinsurers “purposefully refused to write policies,” and some — not necessarily all or “many” — had been shown to have done so.

On to the second claim, which came in two parts. The first part was that insurers were “still cobbling together hurricane protection in August and September” (6). The closest St. John came to offering evidence in support of her claim came in paragraph 39, where she wrote that regulators “began a watchlist of insurers without full coverage at the start of hurricane season”:

Industry sources said five insurers were put under temporary supervision. Records obtained by the Herald-Tribune show at least one, United Property and Casualty, was still short in mid-September and operating under a regulatory consent order, even as it sought a state loan to expand.

This paragraph fell short of demonstrating the conclusion. St. John began by discussing insurers “cobbling together hurricane protection,” but this paragraph showed evidence only of companies without protection, not of companies struggling to obtain protection. (To be sure, neither situation was desirable for citizens, the state, or insurers.)

The second part of the second claim depended on the first part, which had not been demonstrated. But assuming arguendo that the first part of the claim was true, St. John then went on to claim that when insurers were putting together coverage in August and September, they were “paying three times the January rate” (6).

This part of the claim was not demonstrated, either. Instead, it was simply repeated, in paragraph 36: “By July, Florida’s cost to reinsure against the biggest hurricanes had tripled.” As it provided no additional evidence or citations, this paragraph served only to restate the conclusion, and could not be used to support the conclusion without also committing the Begging the Question fallacy (Damer, p. 55). A highly particular reader, though, might have noted that paragraph 36 does not relate to the same time period as did the original claim — July, rather than August and September — and included a qualifier missing from that claim (reinsurance “against the biggest hurricanes,” as opposed to all hurricanes).

The third conclusion, that “the cost [of reinsurance rate increases] was paid by Florida property owners” (7) came with some evidence, but the evidence bordered on becoming an instance of providing correlation without causation:

The average home premium increased 80 percent. Residents near the coast saw increases of 300 percent. More than 300,000 Florida families lost their private coverage, forced to find a new company or join Citizens, the state-run insurer of last resort. (41)

This paragraph, though lacking a source, did appear to show that Florida residents suffered during the hurricane season in question. But no evidence was provided to suggest why reinsurance costs, or insurer struggles, were the cause of the increase in “average home premium” or the total loss of private insurance for “more than 300,000 Florida families.” Admittedly, however, it is difficult to think of another potential cause that would account for the change. It is perhaps best to acknowledge that many factors affected insurance costs for homeowners, but that the cost of reinsurance for insurers was one of them, and that it was likely that the trends St. John discussed earlier affected insurance costs for homeowners. The third conclusion, then, was fairly well supported.

Leave a comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s