Case study

We estimate the HVI for a sample of southeastern counties in order to demonstrate the estimation process and method to interpret the results. This sample of counties provides at least one county from each of the eight states; we chose two counties from three states (Texas, Florida, and North Carolina) that have large coastlines and a high probability of hurricane strikes. The 11 counties are: Cameron (TX), Galveston (TX), Orleans (LA), Harrison (MS), Mobile (AL), Miami-Dade (FL), Bay (FL), Chatham (GA), Charleston (Sc), New Hanover (NC), and Dare (NC). The indicator definitions, and minimum and maximum values, are listed in Table 3. There is significant variation between the indicators for the

1 We do not include indicators for non-residential construction but assume that it would be proportional to residential construction.

sample counties. Most of these coastal areas have experienced rapid population growth and development growth in recent decades despite the risk of hurricane damage.

Table 3. Hurricane Vulnerability Index Indicators

We use values for population and housing data from the 2000 U.S. Census to estimate the four demographic HVI indicators. We use the coastal vulnerability index from Thieler and Hammer-Klose (1999) for a measure of the physical shoreline characteristics that increase the susceptibility to hurricane damage. We use the Insurance Services Office (ISO) rating on building code enforcement. The ISO's Building Code Effectiveness Grading Schedule evaluates the effectiveness of a municipality's enforcement of building codes. For the climatological probability, we use U.S. Landfalling Hurricane Probability Project estimations of the likelihood that a hurricane will strike a particular county.

The HVI value for the 11 sample counties is presented in Table 4. A higher HVI value indicates a county that where the risk of hurricane damage is greater. The maximum and minimum risk possible would be indicated by a value of 10 and 0, respectively. The most and least vulnerable county to hurricane damage from our sample is Miami (5.52) and Chatham (0.11), respectively, based on the HVI values. Examining the individual indicators can explain much of the ranking for a particular county. For example, of the 11 sample counties, Miami is the most at risk for three of the indicators - population, number of houses, and probability of a strike - and among the most risky for the other four indicators. Chatham County is the least at risk for two variables - probability and vulnerability - and near the lowest risk for two variables - population and the number of houses.

Indicator

Minimum

Maximum Risk

Resident Population

Number of Housing Units

Median value of owner-occupied housing

Climatological probability of hurricane strike (percent)

Average Building Code Enforcement Grade

Percentage of homes built after 1990

Vulnerability to sea-level rise

2253362 852278 137200 100 10 100 4

County

HVI Value

Miami-Dade, FL Dare, NC Charleston, SC New Hanover, NC Bay, FL Harrison, MS Galveston, TX Orleans, LA Cameron, TX Chatham, GA Mobile, AL

0.69

The hurricane return period, probability of a strike, and hurricane damage cost for each sample county is listed in Table 5. Comparing the HVI value to the historical hurricane damage cost provides some validation of the HVI values. The county with the most and least cumulative damage is Miami-Dade and Chatham, respectively. The two counties have the highest and lowest HVI, respectively. Between 1851 and 2006, the greatest number of major hurricane strikes occurred in Florida (37), and the fewest strikes occurred in Georgia (3) (Blake et al., 2007). Coastal areas that are predicted to have a higher probability of a hurricane strike -Miami-Dade, Charleston, Cameron, and Dare. - have the highest HVIs, generally. Although Cameron County has the third highest strike probability, its HVI is ranked sixth. This is partially the result of Cameron having the lowest housing values of the sample. Examining individual indicators can indicate which factors contribute the most to risk for a particular county, and therefore which factors should receive the most attention. For example, Bay County's HVI is slightly higher than Chatham County's HVI. The principle difference between the indicators for the two counties is that Bay County has a much higher vulnerability indicator than Chatham. This difference may account for some of the higher historical damage costs for Bay County. The higher vulnerability for Bay County would suggest that hurricane mitigation efforts and resources should be focused on creating less vulnerable communities in Bay County. Building hurricane-resistant housing or protecting shorelines would be productive activities.

Return Period Hurricane Strike Hurricane Damage

Counties

Hurricane

Probability1

Costs(1950-2

(years)

(climatological)

(2008 $)

Cameron, TX

25

2.7

47,301,684

Galveston, TX

18

1.9

869,886,903

Orleans, LA

19

1.7

988,065,664

Harrison, MS

18

1.9

1,322,042,826

Mobile, AL

23

1.9

622,166,867

Bay, FL

17

0.5

164,097,087

Miami-Dade, FL

9

5.0

3,523,941,209

Chatham, GA

34

0.4

3,596,486

Charleston, SC

15

2.9

878,583,973

New Hanover, NC

16

1.1

399,547,814

Dare, NC

11

2.4

88,484,892

Probability of intense hurricane striking the county

Sources: Blake et al., 2007; U.S. Landfalling Hurricane Probability Project; Hazards & Vulnerability Research Institute

Table 5. Hurricane Statistics for Sample Counties

Probability of intense hurricane striking the county

Sources: Blake et al., 2007; U.S. Landfalling Hurricane Probability Project; Hazards & Vulnerability Research Institute

Table 5. Hurricane Statistics for Sample Counties

The HVI can be helpful in creating equitable rates for property insurance. With the recent frequency of storms and hurricanes, insurance companies and state agencies have reassessed their insurance policies in coastal markets. Insurance companies began changing their coastal policy writing practices shortly after 1992's Hurricane Andrew, which struck south Florida and caused an estimated $32 billion in property damages. Insurers immediately attempted to limit coverage and raise rates where coverage was provided in areas subject to hurricane impacts. Additional, heavy losses in the years following Andrew have caused coastal insurance for wind and hail to become much more expensive and even difficult to obtain in many areas as insurance companies attempt to reduce risk. For example, following the 2004 hurricane season, premiums doubled for windstorm insurance in many parts of the Florida with owners of 1,500 square foot homes facing premiums of $10,000 for wind damage alone with total insurance costs of $13,000 and deductibles of up to $18,000 (Mortgage News Daily 2007).

As insurance companies have reduced coverage and raised rates, consumers have become increasingly concerned and groups such as the real estate industry have called for increased government intervention. State insurance agencies, which provide oversight of insurance rates, can use the HVI in order to more accurately regulate insurance companies. For example, Miami-Dade's HVI value of 5.52 is much higher than Charleston County's of 1.34, which is the second highest HVI. This indicates that the risk of damage, and therefore insurance rates should be much higher in Miami-Dade County. However, examining individual indicators is important in order to assess rates correctly. Dare County's HVI, for example, is ranked third; this is due in part to the modest population and housing totals. Other indicators, such as hurricane probability and vulnerability are very high for Dare, which would suggest that rates should be adjusted accordingly.

In addition, the HVI can be used to inform residents about areas of greater risk. Property owners moving to counties with high HVI values can be informed of the potential risk. Although we provide an illustration of the HVI calculation using a sample of coastal counties, expanding the analysis to include all counties from the southeastern U.S. or adding other indicators would follow this methodology.

0 0

Post a comment