Quantifying the Benefits of Home Buyouts for Mitigating Flood Damages

Abstract

With increasing costs of floods across the globe, interest has grown in using managed retreat strategies, such as home buyouts, for flood mitigation. For years, national-level estimates of the benefits of flood mitigation investments in the United States have been estimated as being on the order of 4:1 (USD) or more recently 6:1 (USD), according to the Federal Emergency management Agency and others. Localized evaluation of the true costs and benefits for communities is limited, however, and primarily centered around avoiding direct structural damages. Home buyouts, where individuals’ properties are purchased and removed from risk while reducing impervious areas, are only one of many mitigation options captured in the national estimates. This paper presents a scenario analysis to estimate a localized return on investment of a home buyout program for Nashville-Davidson County, Tennessee in the United States that justifies investment in such programs. It considers the cost of buyouts across the community over several decades. It also compares those costs to benefits such as changes in stormwater attenuation, damages avoided, and exposed populations for a specific extreme flood event. Overall, at a localized scale, the study demonstrates that the return on investment is comparable to national averages, and that proactive implementation significantly increases the value of benefits relative to costs. The approach has potential applicability to other areas in the United States and beyond. It can be used to make the case for investment in such mitigation programs utilizing similarly available data.

Introduction

Globally, estimated annual damages due to riverine flooding events total tens to hundreds of billions of dollars (USD). These damages will likely continue to increase under a changing climate. Ward et al. (2017) estimated the current expected annual damages from river floods as ∼$94B USD given existing flood protection standards. This current level of flood damage will likely increase up to twenty times by 2080 (Winsemius et al., 2016) in response to climate effects and increased socioeconomic development. Recognition of the growing threat of flood damages has led to calls for increased flood damage mitigation efforts, particularly those focused on prevention.

Mitigating efforts to prevent river flood damage include strategies such as the construction of levees, dykes, and dams. Other efforts include building raising and flood-proofing, installing stormwater runoff storage tanks, green infrastructure development, urban development regulations, and managed retreat. Often, the benefit-cost ratio associated with such flood mitigation efforts in the United States (U.S.) are claimed as 4:1 (USD), or more recently, 6:1 (i.e., one dollar invested in mitigation leads to six dollars of benefit as reduced damages, etc.) (National Institute of Building Sciences (NIBS, 2017). This is a national average, however, based upon a 75-year horizon and multiple flood mitigation strategies. These estimated benefits may not necessarily apply at the local level, especially in urban areas experiencing gentrification and rising housing costs, or for individual mitigation strategies.

Among the options for flood mitigation, managed retreat has gained increasing popularity and attention as a strategy for reducing flood risk in urban areas over the past few decades (Freudenberg et al., 2016). Managed retreat strategies have appeared across the globe since the 1980s, with most known cases occurring in the past decade (Hino et al., 2017; Greiving et al., 2018). Managed retreat programs range from large, national-level government-driven, mandatory programs—such as the prohibition on rebuilding residential structures near areas flooded during Typhoon Haiyan in 2013 implemented in the Philippines (Hino et al., 2017)—to voluntary relocation of entire small communities following a flash flood in 2011 (Sipe and Vella, 2014). Voluntary home buyout programs (buyouts) targeted towards individual property owners in high flood-risk areas are also possible.

In the United States, managed retreat typically occurs via voluntary buyouts funded by a combination of local and federal funds. The federal funding comes from the U.S. Federal Emergency Management Agency (FEMA) or the U.S. Department of Housing and Urban Development’s (US HUD) Community Development Block Grant Program (CDBG). The increased interest in buyouts in the U.S. (and around the globe) is in part due to the growing recognition of problems frequently attached to gray infrastructure (e.g. levees) mitigation. These problems include risk transference and increased development in floodplains due to a false sense of security (also known as the levee paradox), which are not associated with buyout activities (Palmer et al., 2015). By acquiring and removing repetitive loss structures and converting them to publicly-owned open space, buyouts stop future flood loss and discourage further development in floodplains (Zavar, 2015; Conrad et al., 1998). Furthermore, the conversion of buildings to open space along flood-prone streams and rivers offers opportunities for creation of natural recreational areas and parks. These natural areas benefit local residents and build the capacity of floodplains to naturally improve water quality, attenuate flooding, and provide habitat for wildlife.

The FEMA - Hazard Mitigation Grant Program (HMGP) provides a primary source of federal funding for buyouts. The program allocates most of this funding to communities applying to HMGP only after a significant disaster event with a Presidential Disaster Declaration (Conrad et al., 1998). HMGP funds cover 75% of the costs of flood mitigation projects, the state and local governments sharing the remainder of the costs (Zavar, 2015). Given that FEMA-funded buyouts must undergo a benefit-cost analysis before approval to demonstrate that the benefits will outweigh the costs, assuming that buyouts are cost effective is logical. The capital cost of buyout programs is so high, however, that without greater evidence of cost effectiveness, particularly observation-based evidence, many communities may likely choose not to pursue buyout opportunities. Prior studies on buyout programs have reported long-term benefit-cost ratios ranging from 5:1 (Rose et al., 2007) to 2:1 (Conrad et al., 1998; Kousky & Walls, 2014) to less than 0.5:1 for some CDBG funded buyouts (Tate et al., 2016) based on modeled expectations of direct flood damages avoided. However, limited observation-based evidence exists of the effectiveness of buyout programs at reducing direct flood damages and very few studies address secondary costs avoided. In fact, to date, most benefit-cost analyses related to flood mitigation investments focused centrally on quantifying structural losses and damages avoided in comparison to capital costs of the mitigation option (White, 2011; Highfield and Brody, 2013). In addition, though often discussed, to our knowledge, the benefits of proactive implementation of buyouts is still un-quantified.

Limited research exists on quantifying benefits of buyouts beyond loss avoidance. When considering secondary losses related to direct property damage, Rose et al. (2007) examined multiple FEMA mitigation grants related to flooding and found an average benefit-cost ratio of approximately 5:1 USD. The Rose et al. (2007) study included environmental damage, and death and injury losses using flood models to estimate flood depths at participating properties. While this study addressed secondary costs of flood hazards, the property damages were not based on observed values. Secondary benefits were also largely utilized generalized values found within the literature. This same study suggested that estimates of flood depths are among the greatest sources of uncertainty in benefit-cost analyses of flood hazard mitigation, exceeded only in strength by discount rates (Rose et al., 2007). In a related study, Whitehead and Rose (2009) recommended investigations of environmental benefits using local data when possible, as variations from general averages found in the literature are likely. In a more recent evaluation of property acquisition through HGMP and CDBG programs in Cedar Rapids, Iowa, after flooding in 2008, Tate and colleagues (2016) calculated the benefit-cost ratios for individual parcels. These ratios were based on expected structure and contents losses associated with different flood return periods, which were modeled using a coupled 1D/2D hydrodynamic model. They found that more than 60% of the properties that participated in one of these acquisition programs had a benefit-cost ratio of more than one over a 30-year lifetime based solely on expected structure and contents losses. While this study applied an advanced flood modeling approach, which may compensate for a lack of observed flooding data, it did not account for indirect and secondary costs associated with flooding.

In this study, we therefore address the following research questions:

• How does proactive implementation impact calculated benefit-cost ratios for buyout programs?

• To what extent does inclusion of secondary benefits alter benefit-cost projections for buyouts?

Using a test study area, to answer these questions, we examine the costs and benefits of a buyout program in terms of both direct damages avoided as well as secondary benefits, such as municipal stormwater management costs avoided and population protected. We use a unique dataset of observed flood depths from a recent severe flooding event, a household survey of flooded residents, building footprints, and tax-parcel data to calculate flood damages at high spatial resolution. This information provides a greater degree of certainty than is typically available when employing most tools for estimating damage. We employ retrospective counterfactual scenarios to compare the costs of the flood event on the community under the following considerations: (1) the costs of having not implemented the buyout program, (2) the costs of the flood if expansion of the buyout program had been proactive instead of reactive, and (3) the costs of the flood if the buyout program had proactively acquired all identified repetitive loss properties. These counterfactual scenarios allow characterization of the realized benefits of the buyout program for the study area. They also indicate the extent to which proactive implementation of buyout programs increases/decreases their cost effectiveness.

In May 2010, central Tennessee, U.S. experienced a severe precipitation event where more than 13 inches (33 cm) of rain reportedly fell in the Nashville region over 32 hours, more than twice the historic record (National Oceanic and Atmospheric Administration (NOAA), National Weather Service, 2011), exceeding expectations for a 1,000 year, 48 -h rainfall event (National Oceanic and Atmospheric Administration (NOAA), 2018). At least eleven fatalities occurred in the area due to flash flooding of streams and tributaries of the Cumberland River, more than 150 water rescues were conducted and more than 11,000 buildings were damaged at an initial estimated cost of about $2B USD (National Oceanic and Atmospheric Administration (NOAA), National Weather Service, 2011).

Following the flood, the City of Nashville in Davidson County, Tennessee, U.S. took steps to improve emergency response and mitigate flood impacts. Fig. 1 shows the location of Davidson County in the U.S. The Nashville Office of Emergency Management put in place improved emergency communication plans and Nashville Metro Water Services (MWS) prepared a Unified Flood Preparedness Plan (UFPP) that assessed the effectiveness of different flood response and mitigation strategies including modifications to water and wastewater treatment plants and construction of flood walls (Nashville Metro Water Services (MWS, 2013). In addition, MWS utilized U.S. Federal Emergency Management Agency (FEMA) - Hazard Mitigation Grant Program (HMGP) funding to significantly expand the small home buyout program that had been in place since the late 1970s.

The MWS buyout program has been active since 1977. However, funding for buyouts remained fairly limited prior to the 2010 flood. Fifty-six properties were acquired between the years 1977 and 2009. In contrast, 246 properties were acquired between 2010 and 2016 following the flood. As of 2016, the residents of an additional 48 properties targeted by the buyout declined to participate and continue to reside in their repetitive flood loss homes. MWS has identified another 89 properties on their high priority list yet currently lacks sufficient funding to acquire these properties. In Nashville, homes are prioritized for buyout based upon on FEMA’s repetitive loss list, extent of damages, and also the willingness of homeowners to participate in the program (some eligible homeowners refuse buyout offers). The total cost of acquisition of all buyout properties purchased prior to the flood in 2010 was ∼$9.6 M USD while the cost of acquiring all properties purchased between 2010 and 2016 was ∼$34.2 M USD. While the average cost of acquisition was ∼$145,000 USD the costs ranged from approximately $11,000 USD to $480,000 USD. Fig. 2 displays the locations of all acquired properties in the Nashville area and the inundation boundaries of the 2010 flood.

For the Nashville community, floodplain managers are primarily focused on removing residents from harm's way. Within the community most observed injuries and fatalities related to flooding were due to people remaining at or attempting to leave their property during a flood event. The preference for demolition over structure elevation stems from multiple considerations including the potential compromise of electrical components and mold contamination in a previously flooded home. Furthermore, Nashville has experienced significant growth and development in the past 20 years and there are few options for relocation of the older, single-level homes to smaller lots. Another concern with raising homes is that people will likely be victims of flooding again in the future and that elevated homes may create a false sense of security similar to the levee effect.