Historical validation of enhanced U.S. demand surge
Historical catastrophe events provide empirical evidence of demand surge and demonstrate how reconstruction timelines, labor market disruption, and construction cost inflation influence post-event losses.
Verisk's enhanced demand surge is grounded in empirical observations of reconstruction costs following major natural catastrophes in the United States. Historically, construction labor shortages and delays in reconstruction can cause losses to increase after certain large events. The industry view of demand surge in the U.S. has been defined by hurricane seasons in 2004 (Charley, Frances, and Jeanne), 2005 (Katrina), and 2017 (Harvey and Irma). There have been other notable examples of events with demand surge, but few others have caused the significant increase in construction costs seen from these events.
These case studies examine three hurricane seasons—2004, 2005, and 2017—that have demonstrated demand surge in the United States. The table below shows the enhanced demand surge framework results compared to historical demand surge observations. Historical observations are based on Verisk 360Value Residential Reconstruction Cost Analysis data for the impacted state. Historical ranges (minimum and maximum) are based on Verisk 360Value Residential Reconstruction Cost Analysis data for impacted local areas. For each scenario, modeled demand surge is compared to available historical data. In comparing modeled demand surge results to historical events, it is important to consider economic conditions at the time of event. To compare modeled results, results are presented using both long-term average cost inflation trends for post-event construction cost inflation and cost inflation trends at the time of the event.
| Event | State | Demand Surge framework | Demand Surge framework (w/inflation at time of event) | Observed demand surge (actual) |
|---|---|---|---|---|
| Hurricanes Charley, Frances, Ivan, and Jeanne (2004) | FL | 10.3% | 12.4% |
17.9% (8.3%-27.5%) |
| Hurricane Katrina (2005) | LA | 13.0% | 11.3% |
6.6% (5.8%-11.4%) |
| Hurricane Harvey (2017) | TX | 7.6% | 5.7% |
2.7% (0.9%-5.3%) |
| Hurricane Irma (2017) | FL | 9.0% | 8.9% |
4.2% (1.4%-5.5%) |
Source: Verisk. Note: The new demand surge framework results using long-term average cost inflation trends from 2010 to 2023 for post-event construction cost inflation. Demand Surge Model (w/inflation at time of event) uses cost inflation trends at the time of the event using 10-year centered-rolling averages (i.e., 1999-2009 for 2024, 2000-2010 for 2005, and 2012-2022 for 2017). Cost inflation values are based on Verisk 360Value Residential Reconstruction Cost Analysis data for the impacted state. Actual observed demand surge use Verisk 360Value Residential Reconstruction Cost Analysis data for the impacted state at the time of event and assumed claim settlement time. Historical observed ranges (minimum and maximum) are based on Verisk 360Value Residential Reconstruction Cost Analysis data for impacted cost areas.
Case study: 2004 hurricane season
The 2004 hurricane season is notable for four named storms making landfall in Florida within a six-week period: Hurricane Charley in August, followed by Hurricanes Frances, Ivan, and Jeanne in September. In Florida, it was estimated that one in five residential structures sustained some level of damage during the season (Guy Carpenter, 2014). The 2004 hurricane season highlights how multiple storms in a single season can contribute to demand surge.
The clustering of events in Florida created increased demand for reconstruction resources, delaying recovery and rebuilding for many. For example, an extensive survey of impacted households conducted by the Bureau of Economic and Business Research at the University of Florida found that repairs had been completed for only 35 percent of damaged homes nine months after the storms. At the time, 44 percent of damaged properties still had repairs underway, 16 percent were planning but had not yet started repairs, and 5 percent were not planning to make repairs at all (Smith & McCarty, 2006). This reconstruction profile, characterized by high demand for repairs, created conditions for significant labor market disruption and prolonged reconstruction and insurance claim settlement timelines.
Modeled demand surge for the 2004 season in Florida is 10.3 percent using long-term assumptions for post-event construction cost inflation and 12.4 percent using cost inflation trends at the time of the event. Modeled results are consistent with the observed demand surge of 17.9 percent in Florida, on average, and localized demand surge ranging between 8.3 and 27.5 percent. Modeled demand surge is within the range of historical observations. These results demonstrate the model's ability to capture demand surge for clustered events that can occur in a hurricane season.
It is also important to consider economic conditions at the time of event. The 2004 hurricane season coincided with a strong housing construction market in Florida with already high levels of construction cost inflation. A previous retrospective of demand surge following the 2004 hurricane season found that construction permits per worker were at elevated levels in Florida prior to the 2004 hurricane season. The high level of construction activity suggests that the construction labor market had limited excess capacity to absorb the sudden surge in repair demand. In response, construction wages rose sharply following the 2004 hurricane season, while construction material prices continued to follow national trends (AIR Worldwide, 2020). This analysis highlights that demand surge can be significantly driven by local construction labor market disruptions at the time of the event.