Because of disasters such as the tsunami in Japan, tornadoes in Joplin, Missouri and Tuscaloosa, Alabama, flooding along the Mississippi and Missouri Rivers and the Yangtze River in China, the concept of “risk”  has recently received a significant amount of attention.

Munich Reinsurance (an insurance company for insurers) recently reported that worldwide losses for the first half of 2011 have already surpassed the total losses for the historic high set in 2005 when Hurricane Katrina hit New Orleans.  This has caused insurance companies to re-evaluate the future negative effects of climate change. Moreover, Congress in the debt ceiling debates has questioned the role the federal government plays in providing disaster relief, with the St. Louis Beacon reporting that with “minor” catastrophes are on the rise there is little federal assistance available to cover major catastrophes.

As we have previously noted, Congress is debating the major reforms to the National Flood Insurance Program.  The NFIP is over $18 billion in debt which will likely increase after the total losses for 2011 flood  season and upcoming hurricane season are tallied.  A report to be released later this month by the Federal Emergency Management Agency indicates that the nation’s floodplains are expected to increase by 40 to 45% in the next 90 years, which will put stress on the already indebted program.

These trends have caused Sen. Durbin (D-Il.) to ask the following: should the federal government use probabilistic catastrophe modelling or “cat modelling” when assessing risk of severe weather created by our changing climate? In addition to historical date of past severe weather events, cat modelling uses prospective date generated from recent climate change studies to model the probability of future events.  Since the mid-1990′s, the insurance industry has increased its use of cat modelling  to evaluate insurance portfolios in risk-prone regions of the country.   To be sure, cat modelling has been criticized as deterministic, based on uncertain science, and providing cover for insurance companies to raise premium rates to unjustifiably high levels.  But until now it has not been considered for use by the federal government and such use could significantly impact communities around the country as well as business investment decisions.

For communities affected by the NFIP, the use of cat modelling is troubling.  First, insurance companies increased use of cat modelling to hedge against the risk of floods will result a rise in premium rates.  Both the House and Senate bills phase-out the federal subsidy for insurable structures by increasing premium rates at a rate of 20% per year up to the full actuarial rate.  Without reforms allowing for the regulation of cat models, insurance companies would be free to overcharge premiums to hedge against the risk of a catastrophic flood.  It is estimated that catastrophe modelers overshot actual insured losses for 2006 through 2010 by as much as $53 billion.

More importantly, if FEMA were authorized to use cat modelling to map flood zones such a determination would be very difficult to appeal.  The basis for appealing a flood rate insurance map (“FIRM”) is the FIRM’s technical accuracy.  FIRM challenges tend to be difficult to win on administrative appeal and federal courts are limited to examining the technical merits of the FIRM and often give FEMA wide discretion absent a technical error.  Given the uncertainty of predicting climate change, any appeal would essentially come down to whether the studies and cat models relied upon by FEMA were a reasonable assessment of risk – an assessment subject to the high bar set by the ”arbitrary and capricious” standard under Section 706 of the Administrative Procedures Act.  Without guidance from Congress, cat models would reshape current flood zones and potentially subject new urban areas to flood zone mapping.

The collective result will without a doubt affect business decisions to locate, expand, lease, or otherwise investment in affected communities, as evidenced by the experience of Metro East communities in St. Louis.  From a policy perspective, it is not unreasonable for the effect of cat modelling to transfer the risk of a natural disaster back onto property owners choosing to own property in risk-prone areas.  However, use of cat modelling should be tempered to avoid potentially devastating economic consequences for affected communities.