Housing Recovery on the Gulf Coast

Housing Recovery on the Gulf Coast

A house damaged by hurricane Katrina in 2005.

Recent research from HUD’s Office of Policy Development and Research (PD&R) examines how $19.7 billion in Community Development Block Grant Disaster Recovery grants were used to rebuild Louisiana, Mississippi, and Texas following the devastation of hurricanes Katrina and Rita in 2005. PD&R has published two reports from this research as well as a summary of findings. The first phase of the research involved windshield observations of significantly damaged homes and neighborhoods in the three states. The second phase surveyed owners of badly damaged properties about their experience rebuilding their properties and their lives. In the words of Assistant Secretary Raphael Bostic, “The federal government plays an important role in response to, and recovery from, natural disasters. HUD hopes that this research helps to inform that role and to help disaster-affected communities recover faster and more fully.”

The Phase I windshield observations of the most affected blocks, conducted in early 2010, showed that approximately 75 percent of all affected properties in the most damaged areas were in good condition, approximately 15 percent contained a residential structure needing substantial repair, and slightly more than 10 percent of the sites were vacant because the house formerly on the site had been destroyed by the storm or subsequently demolished. The areas with the highest percentage of vacant properties were the Lower Ninth Ward of New Orleans and adjoining St. Bernard Parish in Louisiana, and Hancock County, Biloxi, and Waveland/Bay St. Louis in Mississippi. Although Texas had a low percentage of vacant properties, about one-fourth of the existing structures needed substantial repair.

One important finding from the Phase II property owner survey was that having insurance coverage before the hurricanes was a strong predictor of rebuilding. This finding emphasizes the importance of homeowner’s insurance and flood insurance, particularly for households in disaster-prone areas. Affected households who moved reported higher neighborhood satisfaction than those who stayed, suggesting that in some cases relocation may serve affected households better than rebuilding. In addition, neighborhoods suffering from widespread, significant flood damage are less likely to be successfully rebuilt, suggesting that those neighborhoods either should not be rebuilt or will require substantially more resources per unit to rebuild.

Phase II also examines the different models of homeowner assistance offered by the three states. Texas implemented a traditional home repair and rehabilitation model, in which housing assistance was tied directly to the home for the exclusive purpose of rebuilding and the state hired and managed the contractors. This program served roughly 1,000 households. Mississippi and Louisiana, having sustained more widespread damage, had approximately 28,000 and 129,000 households, respectively, in need of assistance. To speed the recovery effort, these states received a waiver allowing them to provide assistance to individual homeowners in the form of a compensation grant, which did not necessarily have to be used for rebuilding. Homeowners could hire their own contractors and begin construction immediately. The average length of time from the start of construction to the end of construction was only 8.1 months in Texas compared with 10 months in Mississippi and 13.9 months in Louisiana. It seems that Texas—with a relatively small program and abundant contractor capacity—was able to carry out its program effectively once the program was developed. However, when measured from the date of the storms to the end of construction, rebuilding in Texas took considerably longer than it did in Louisiana and Mississippi despite the smaller size of the Texas program. This finding suggests that the compensation model did enable Louisiana and Mississippi to begin rebuilding sooner than they would have with the standard rehabilitation model.

Finally, the study finds strengths and weaknesses in both the rehabilitation model and the compensation model. The compensation model does appear to have been successful in Mississippi and Louisiana. As of 2011, rebuilding rates are comparable across the three states even though the Mississippi and Louisiana programs served far more households than did the Texas program. The rehabilitation model, on the other hand, resulted in more complete application of funds to rebuilding; 16 percent of households in Mississippi and 6 percent of households in Louisiana did not use any of their grant funds for rebuilding. Although the rehabilitation model may be difficult to scale up, some evidence suggests that it is an appropriate strategy when rebuilding is the priority and adequate capacity exists among building contractors. The rehabilitation model also enables the government to supervise how rebuilding occurs. The organization managing the program can ensure compliance with building standards and incorporate other improvements (such as energy-efficiency upgrades). A homeowner-driven rebuilding model runs the risk that assisted households may not pursue these important goals. Ultimately, however, the most appropriate rebuilding strategy depends on the scope of the damage and the community’s plan for recovery.