The main job of any reporter is to be accurate with the facts, and when we make mistakes do our best to correct and not repeat them. However, when there is a difference of opinion about the accuracy of a story we have two options–ignore the complaint, or do a follow-up. In this case, it’s the latter.
Back in March, Anna Olsen, a researcher, and Keith Porter, associate research professor in civil, environmental and architectural engineering–both from the University of Colorado in Boulder–gave a presentation about their work on the economics of demand surge during a conference on hurricane risk at Princeton University, sponsored by Willis.
The conference brought together members of the Willis Research Network, a group of 18 universities that study risk.
In my March 20 story on the NU Online News Service–”Demand Surge Not Driven By Economic Demand”–I reported that Ms. Olsen’s study of three major hurricanes and two major earthquakes, and the subsequent demand surge for insurers, had less to do with labor and materials and more to do with the way claims were settled.
The story reported her saying that demand surge–an increase in labor and material costs after a storm–was more likely caused by the combination of a lack of insurer resources to settle claims, as well as insurers settling claims for more than their limits. She was reported to have said there were not enough adjusters to settle claims, and that poorly trained adjusters were settling for more than they should have.
She was also reported to have said that insurers paid out higher claims to sophisticated customers who were willing to fight the carriers.
“When I saw the first story that came out, it wasn’t exactly what had been presented,” said Ms. Olsen, who was interviewed along with Mr. Porter on June 5. “Both Keith and I felt that it got everything that was said, but it didn’t get what the point of the presentation was.”
It was not until the two started to receive inquiries from individuals in the industry that they turned to Willis, which contacted NU. That prompted the follow-up interview.
Mr. Porter explained that while their research did not find major volatility in labor and material cost increases above the national average, there were other sources of information they did not look at it.
“It doesn’t prove that demand surge doesn’t come from material prices and labor prices. It simply shows that those two sources don’t support the hypothesis, but other sources might,” he said.
Both Ms. Olsen and Mr. Porter contend the suggestion that demand surge may be contributable to poor claims adjusting was based on conversations and literature on the subject, and not their research. “It is a hypothesis worth pursuing, but we have not reached conclusions one way or the other about it,” said Mr. Porter.
The point of their research is not to prove the existence of demand surge but to explain why it happens and its significant contributors, said Ms. Olsen. From that, added Mr. Porter, the hope is that they will be able to produce a predictive model for insurers to know how much demand surge to expect in a verifiable way. “That’s the goal of the research, but we’ll have to wait until all the work is done to say that was the case or not,” Ms. Olsen noted. “We’re trying to prove which ones hold water and which ones prove to be a good explanation.”
Close to a month after NU‘s story appeared, a group of consumer watchdogs used the article in a letter to the nation’s insurance commissioners as more evidence that demand surge, as a component in premium pricing, should be more closely examined. Birny Birnbaum, executive director for the Center for Economic Justice, said those costs unnecessarily add 20-to-30 percent to property insurance premiums. “Demand surge has been a relatively unexamined component of catastrophe models,” he added.
When asked about this, Mr. Porter said that is not the case. The models have been thoroughly examined, especially by Florida’s Department of Insurance, but they remain the proprietary property of the modelers and are not made public.
It is their hope, he noted, that their research will get to the “granular level” of what causes demand surge, and that such knowledge would contribute to the larger understanding for underwriters writing risks.
When informed of Ms. Olsen’s comments about demand surge from the NU interview, Mr. Birnbaum said in an e-mail that the opinion of the group of consumer watchdogs remained unchanged–that there are problems with these models. “A few major assumptions–including assumptions about demand surge–can dramatically impact the model output,” he wrote, adding:
o “Such major assumptions require public oversight.
o “Demand surge is particularly susceptible to manipulation by insurer claim settlement and government policies.
o “Demand surge can and should be managed with proper government policy.
o “A public model is necessary for the public to have confidence in the modeling used for homeowner insurance rates.”
“To this last point I will add that one cause of the financial market crisis was an overreliance on financial risk models–models which failed in extreme circumstances,” said Mr. Birnbaum. He added that it seemed odd that the research is “not ready to criticize demand surge in cat models, but is ready to endorse it.”
We no doubt have not heard the last word in this ongoing debate.
To respond to Associate Editor Mark E. Ruquet’s column, e-mail him a email@example.com