Impact Scenarios from Hurricanes Past
By Rick McDonald
Katrina packed quite a wallop. The megastorm will likely prove the most expensive hurricane in U.S. history in both nominal and real (inflation-adjusted) terms. Official estimates for insurance claims are approaching $25 billion.
Given the general rule of thumb that total damage is double insured losses, this estimate would top damage estimates from Hurricane Andrew in 1992, and fall in line with the combined impact of all four hurricanes that hit in August and September, 2004. Here is a rogue's gallery of the 10 most damaging U.S. storms, upon which Katrina sits comfortably atop:
|Hurricane||Category||Date||Damage (Bil. $)||Infl. Adj. Damage (Bil. $)|
|Katrina (LA, MS, AL)||4||8/05||$50.0E||$50.0E|
|Andrew (FL, LA)||5||8/92||$26.5||$34.1|
|Ivan (AL, FL)||3||9/04||$14.2||$14.5|
|Hugo (SC, GA, NC, VA)||4||9/89||$7.0||$10.0|
|Georges (FL, MS, AL)||2||9/98||$5.0||$5.8|
|Floyd (Mid-Atl., NE)||2||9/99||$4.5||$5.1|
|Fran (NC, SC, Mid-Atl.)||3||9/96||$3.2||$3.8|
|Opal (FL, AL)||3||10/95||$3.0||$3.6|
We at Action Economics expect economic disruptions on the regional economy from the hurricane to be less than what occurred over the two-month period of August and September, 2004, when four Category 3-4 hurricanes hit, as the net days of disruption are now less. But Katrina should provide a bigger disruption than Hurricane Andrew alone.
It also may be the case that the impact on individual monthly economic releases will be less than with Hurricane Andrew in 1992, as the late occurrence in the month could leave the net impact divided between two months. But the concentration of the U.S. energy industry in the region may prompt larger overall effects on the national economy than was seen with the storms in Florida and will create many effects that should prove peculiar to this hurricane.
Here's a detailed review of the impact of the largest hurricanes on various U.S. economic indicators:
Retail sales generally see a small dip and subsequent rebound following a hurricane, although the pattern is erratic. The impact on services is even less discernible. Emergency stockpiling prior to the hurricane, and subsequent replacement spending thereafter, tends to mute the net negative impact that many people intuitively assume.
Payrolls generally fall short of the underlying growth trend in the hurricane month, but then rebound in the months afterward. Interestingly, the workweek tends not to be affected. Katrina's impact will be limited by its coming well after the August payroll survey, but well before the September survey. Weekly initial jobless claims, however, can be depressed following a hurricane, as reporting centers are closed, and individuals focus on more pressing needs in the immediate aftermath of the disaster.
Industrial production historically is the most heavily affected report, as power outages and distribution delays affect utility, mining, and even manufacturing production.
Housing permits tends to show a more dramatic impact than starts. Conventional wisdom holds that permits tend to be more insulated from weather distortions, but the permit data is less abstract -- as the Census Bureau polls government permit agencies directly. With government agencies often closed during hurricanes and subject to the bureaucracy, the permit process has historically seen more of a distortion than starts. Construction spending has not historically seen much of an effect. Plus, Katrina hit an area with a relatively small construction sector.
New home sales tends to see a larger distortion than existing home sales. But this may be more a reflection of Florida being a high-growth state with a highly developed new-construction industry. Katrina's high-impact areas should prove to have less of an affect on new home sales.
The trade deficit tends to widen for hurricanes across the Southeast overall, due to the region's bias toward Latin America exports over imports during hurricane months. But Katrina's high-impact area is a critical oil-import region, which will lead to imports being depressed on the month -- especially given refinery concerns that would create bottlenecks for unloading oil supertankers.
Inflation data tend to be biased slightly higher in the immediate aftermath of the hurricane.
The U.S. GDP and current-account data should show limited impact from insurance payments for damages, though corporate profits will take a huge hit similar to the third quarter of 2004 -- $160 billion by our estimates.
Personal income components are also distored from the "capital consumption" impact of insurance payments, but the distortions are offsetting and so should have little impact on the total personal income figure.
Overall, any aberrant surprise to the downside in upcoming monthly data should viewed with a skeptical eye. From a broader perspective, an event of this magnitude clearly suggests downside risk for third-quarter GDP relative to what otherwise was the case. Our storm-adjusted third-quarter GDP forecast of 3.8% reflects what probably amounts to a 0.5% subtraction to growth from the impact of Katrina on underlying source data due primarily to utility and other production disruptions, which is in line with the "triple hurricane impact" seen in the third quarter of last year.
This estimate assumes no extraordinary impact on the petroleum industry that some fear will cause be so prolonged as to drag well into the fourth quarter. And that's a big assumption. The jury is still out on this factor until the energy industry is able to fully assess the damage -- and determine how long the likely fuel shortages will last.
MacDonald is global director of investment research and analysis for Action Economics