Der Artikel untersucht die Auswirkungen der Freibord-Politik auf Hausbesitzer in verschiedenen Einkommensgruppen, wobei Jefferson Parish als Fallstudie dient. Sie unterstreicht die unverhältnismäßigen Auswirkungen von Überschwemmungen auf einkommensschwächere Gemeinschaften und die Notwendigkeit eines gerechten Hochwasserrisikomanagements. Die Studie misst anhand diskriminierender Analysen die Anfälligkeit und den Nutzen der Umsetzung des Freihandelsabkommens und zeigt erhebliche Unterschiede bei der wirtschaftlichen Belastung und dem langfristigen Nutzen auf verschiedenen Einkommensniveaus auf. Die Ergebnisse unterstreichen die Bedeutung maßgeschneiderter Maßnahmen, um gefährdete Gemeinschaften zu unterstützen und eine faire Verteilung der Vorteile aus dem Hochwasserschutz zu gewährleisten.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
The National Flood Insurance Program (NFIP) encourages the adoption of at least a 1ft. freeboard against flooding. Though freeboard is an effective flood mitigation method, the strategy involves additional investment that may cause economic liabilities. Moreover, such strategies are often generalized and do not consider optimal requirements at the micro-level. To this end, this study identifies the factors of economic vulnerability and proposes a framework for investigating the implications of such policy on different income groups. The discriminant analysis method was used to empirically quantify vulnerability and identify factors that are significant contributors to vulnerability. In addition, a vulnerability-benefit analysis was conducted by quantifying the benefits in terms of savings. The framework has been elucidated with a case study conducted for a group of cities in Jefferson parish with a poverty rate of 20% or more. It was found that though economic vulnerability increased with a decrease in income level, the benefits associated with the use of freeboard were almost equally distributed across all levels of vulnerability. The framework developed in this study is expected to provide helpful insight regarding the impact of different policies formulated for mitigating and managing flood disasters on various population groups at the micro-level.
Hinweise
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
In the U.S., flooding is the second deadliest natural disaster (OCWW 2020). In 2017 alone, inland floods resulted in the loss of 25 lives and 3 billion dollars in property damage (UCS 2018). Previous studies have shown that lower-income communities suffer disproportionately from flood disasters than higher-income communities. A survey conducted after Hurricane Katrina revealed that a higher percentage of lower-income communities resided in areas flooded by the event (Logan 2006). Flood mitigation tools and strategies are typically categorized into two broad groups: structural mitigation and non-structural mitigation. While structural strategies involve structural alternatives such as engineered floodwalls/seawalls, floodgates, and levees, non-structural strategies include elevated structures, property buyouts, and permanent relocation (Cigler 2017). While structural approaches involve large financial investment and long-term planning by the government, which allows for proper benefit–cost analysis (BCA) before implementation, non-structural approaches are generally in the form of recommendations for precaution at the individual level that often leads to economic burden (Brody et al. 2009).
Freeboard is a non-structural strategy that was first nationally recommended by the International Residential Code (IRC 2014) to provide additional protection against flooding for single-family homes. Freeboard elevates homes above the base flood elevation (BFE), reducing the chances of flooding. The use of freeboard to increase the height of the base level of the building is one of the most common preventive measures adopted by private homeowners for flood protection in new building constructions. According to the Federal Emergency Management Agency (FEMA), freeboards tend to compensate for the many unknown factors that could contribute to flood heights greater than the height calculated for a selected size flood and floodway conditions, such as wave action, bridge openings, and the hydrological effect of urbanization of the watershed. The use of freeboard also allows for reduced insurance rates (FEMA 2020). Several studies have been conducted to assess and analyze its effects and benefits in preventing flood loss (Wu et al. 2011; Brandimarte and Baldassarre 2012). In addition, FEMA has developed a Community Rating System (CRS) to encourage local governments to implement freeboard during construction in flood zones (Highfield and Brody 2013; Zahran et al. 2009; Highfield and Brody 2017). However, the cost involved in using freeboard for home construction could be a considerable burden on lower economic groups due to their limited financial resources and pre-existing conditions (Miljkovic and Miljkovic 2014).
Anzeige
The benefit–cost analysis (BCA) of freeboard implementation may vary significantly across income groups if socio-economic standing along with geographical context is taken into consideration in the analysis. For instance, wealthier communities might afford higher costs for more substantial flood protection measures, and they might also perceive higher benefits due to the greater value of their properties and infrastructure. Conversely, lower-income communities might struggle to bear the upfront costs of such implementations, even if they stand to benefit from reduced flood risk. This disparity suggests that a standardized, one-size-fits-all guideline for freeboard implementation might not be equitable or effective. Additionally, the absence of a comprehensive benefit–cost analysis (BCA) complicates the determination of optimal freeboard levels. Without this analysis, it becomes challenging to quantify the benefits accurately, which can result in sub-optimal decisions regarding freeboard implementation (Gnan et al. 2022a, 2024). Thus, standardized freeboard guidelines across a larger geographical area can lead to significant distributional concerns, given that the financial burdens and perceived benefits are unequally distributed among different income groups. Consequently, policy guidelines should consider these economic disparities to ensure fair, effective, and inclusive flood risk management across diverse communities.
Freeboard is not required by NFIP standards, but communities are encouraged to adopt at least a one-foot freeboard to account for the one-foot rise built into the concept of designating a floodway and the encroachment requirements where floodways have not been designated. Based on this suggestive measure, local governments have been designing new mandates and ordinances to implement freeboard into construction practice as an effective strategy for mitigating flood losses. One such example is Jefferson Parish in the state of Louisiana in the U.S., which passed an ordinance to mandate freeboard guidelines for home construction (Al Assi et al. 2023). Jefferson Parish lies on the coastal floodplain of the Gulf of Mexico and is prone to extreme flooding as most of the parish is below sea level, with 56% of its surface covered by water bodies (Adaptation Clearinghouse 2022). Thus, it has both AE and VE flood zones that are considered high risk. In addition, the parish has significant income inequalities, with an inequality ratio of 16.57 in January 2022 (Trading Economic 2024). However, the freeboard mandate only takes into account the flood zone and location of individual houses (inside or outside the levee-protected area) when determining the height of the freeboard required.
To this end, this study aims to deduce a suitable methodology that may be implemented for benefit–cost analysis of freeboard policy considering the socio-economic context at the micro-level, thereby providing better insight into distributional inequalities. The study's objective is to empirically measure the vulnerability of different communities and the benefits obtained from adapting to the new policy. A vulnerability-benefit analysis with respect to different income levels in the community can help optimize freeboard requirements at the micro-level.
The main research question that the study explored is that given the new policy as mentioned above is mandated in Jefferson parish:
1.
How will it impact the economic vulnerability of different income levels?
2.
The distribution of long-term benefits of such policy among the different income levels; are the benefits the same across all income levels, or do they vary?
Anzeige
The study identified and listed several socio-economic factors from the extensive literature review that may influence the economic vulnerability of different income groups, affecting their ability to adapt to the new policy through freeboard. A discriminant analysis method was used to identify the significant contributors of vulnerability among the identified factors and empirically measure the vulnerability of different income groups towards adapting to the new policy. The benefit obtained from adapting the new policy was measured in terms of projected savings obtained from reduced flood insurance costs and reduced annual flood loss. A comparison of vulnerability to benefit was made to analyze the disparity in implications of the policy on different income groups.
Climate change and socio-economic development exacerbate flood risk, necessitating flood risk management (FRM) strategies that are context-specific, accounting for socio-economic and cultural factors to be effective, feasible, and sustainable. A study by Klijn et. al., supports that acknowledging these differences can better rationalize current FRM policies and practices, as evidenced by Deltares projects which highlight significant variations in global FRM approaches (Klijn et. al., 2021). Another study conducted in New York highlights the urgency for equitable climate adaptation policies as climate change impacts intensify, revealing that traditional benefit–cost analysis often neglects distributional inequalities. Income-based adjustments can help shift investment toward poorer and disadvantaged areas, though this approach involves trade-offs in overall economic benefits (Lockwood et al. 2024). Additionally, previous research underscores the disproportionate impact of climate change on vulnerable and marginalized populations, emphasizing that without proactive policies, climate change will exacerbate existing socioeconomic disparities (Shonkoff et al. 2011). This highlights the importance of integrating equity considerations into climate adaptation strategies, such as freeboard implementation, to ensure that benefits and burdens are distributed fairly across diverse communities.
The framework developed in this study is expected to benefit local, state, and federal governments by providing a decision-making tool that can identify the vulnerability of different communities when implementing new measures for improving the long-term resiliency of communities against natural disasters. The insights obtained from this study can help agencies develop appropriate assistance and financial schemes to enable more equitable distribution of the benefits of freeboard policies, making them more inclusive rather than profitable and catering only to certain sections of society.
2 Methodology
While various studies have examined the distributional inequalities of climate adaptation policies and mitigation strategies due to socio-economic and cultural differences, this study presents a new approach by specifically exploring the distributional impact of freeboard implementation. Unlike previous research, which is often generalized across multiple adaptation strategies, our study uniquely focuses on the equity of economic burden and benefits for different income groups within the context of freeboard policy. This methodological contribution aims to measure how the economic burdens and benefits of freeboard implementation are distributed among diverse income groups, addressing a gap in the existing literature. To the best of our knowledge, no other studies have utilized this specific approach to examine the implications of freeboard implemetation on the equity, making this research a pioneering effort in this area. The proposed framework using discriminant analysis allows for the empirical measurement of vulnerability toward adapting to different disaster mitigation strategies among different sections of the community. It applies to various categories of vulnerability factors for natural disasters and their corresponding mitigation strategies. In this study, the application of the framework is demonstrated by measuring the vulnerability of different income groups towards adapting to the new freeboard policy in Jefferson Parish as a flood mitigation strategy implemented at the individual level. One of the contributions of this study is listing the index factors that may influence vulnerability towards adapting to the new freeboard policy. The proposed framework then helps to provide empirical evidence for identifying these index factors that are significant contributors to the disparity between different income groups towards adapting to the new policy. The study also measures the benefits of the new policy for each population unit. Thereby, the study conducts a vulnerability-to-benefit analysis to identify the implications of the new freeboard policy on different income groups. Figure 1 shows the overarching method used in this study. The method for empirically measuring vulnerability using a discriminant analysis was initially proposed by Lam et al. (2016) and later modified in another study by Deria et al. (2020).
Fig. 1
Overview of the methodology proposed in this study
×
2.1 Data
This study identified socio-demographic, geographic, and infrastructure-related factors that affect economic vulnerability and thereby affect the ability of different economic groups in the community to adapt to the policy by use of freeboard. These factors listed are based on a combination of facts, logical assumptions, and preconceptions based on data trends. Most of these factors have been previously identified as factors of vulnerability in natural disaster studies (Balica 2007; Kuhlicke et al. 2011; Müller et al. 2011; Behanzin et al. 2016; Rana and Routray 2016). Each factor is elaborated in detail in the following sections. The factors were calculated with respect to the total population count in the population unit. This study was conducted at the census tract level, where each tract was considered a population unit. Among the factors considered in this study, some factors increase the vulnerability or economic burden in adapting to the policy, while others reduce vulnerability and increase the probability of readily adapting to the policy. Those factors that increase the vulnerability are termed incremental factors, whereas those that decrease vulnerability are termed decremental factors in this study.
2.1.1 Incremental factors
Poverty (P)
Poverty exists when people lack the means to satisfy their basic needs, making it a significant economic factor that can nullify a household's ability to invest in disaster mitigation strategies for long-term resiliency. This study considers the percentage of the population in poverty in each population unit, as determined by the U.S. Census Bureau (TABLE ID: S1701). Data for this factor was collected from the U.S. Census Bureau (USCB, 2019) and calculated as the percentage of individuals with income below the poverty level.
Unemployment (U)
Unemployment refers to the situation where a person of employable age cannot find a job, reducing people's ability to afford risk management practices (Balica 2007; Kuhlicke et al. 2011). This study considers the percentage of the population aged 16 and over who are unemployed in each population unit. Data were collected from the U.S. Census Bureau (TABLE ID: B23025) (USCB, 2019).
Dependent population
Children (C) and the elderly (E) increase a household's economic vulnerability due to their physical fragility and financial dependence, limiting the ability to invest in risk management practices (Müller et al. 2011). This study considers these two sub-factors as separate indexes. The percentage of the population under 18 years is regarded as the factor 'Children,' while the percentage of the population aged 65 and above is considered the factor 'Elderly.' Data were collected from the U.S. Census Bureau (TABLE ID: B01001) (USCB, 2019).
Household size (H)
An increase in household size raises vulnerability by dispersing resources among more members, thereby decreasing per capita expenditure (Behanzin et al. 2016). Data for average household size were sourced from the U.S. Census Bureau (Table ID: B25010) (USCB, 2019).
Mortgage (M)
A household with a mortgage may heighten the vulnerability of investing in additional mitigation strategies due to the existing financial burden (Rana and Routray 2016). Data on the percentage of households with a mortgage were obtained from the U.S. Census Bureau (Table ID: B25081) (USCB, 2019).
Ethnicity (R)
Blacks, Latinos, and other minority races experience greater financial hardships compared to Whites, increasing their vulnerability when adapting to new policies requiring additional investment (Lam et al. 2016). Data on the percentage of non-White population in each unit were collected from the U.S. Census Bureau (Table ID: B02001) (USCB, 2019).
Cost of construction with freeboard (CC)
The cost of construction with freeboard significantly contributes to vulnerability in adapting to new policies, affecting certain income groups more than others. Higher freeboard requirements directly increase additional construction costs, thus escalating financial burdens. The building reconstruction cost (BC, in USD) is calculated as the product of the building's livable area (A) and unit replacement cost (\({C}_{R}\) in USD) in Eq. 1.
$$BC=A\times {C}_{R}$$
(1)
The cost of freeboard installation (\({CC}_{i}\)) is estimated as a percentage of the construction cost at Base Flood Elevation (BFE). The percentage of increase in construction cost associated with freeboard (\(F)\) provided in Appendix 1 is multiplied by the B.C. at BFE (\({BC}_{BFE}\)) to obtain the cost of freeboard installation for each building (\({CC}_{i}\)), as shown in Eq. 2. The average cost of freeboard installation for the entire population unit was considered for this factor.
$${CC}_{i}=F\times {BC}_{BFE}$$
(2)
Building value (BV)
The market value of a building is contingent upon various factors, including the prevailing housing market conditions, geographical location, and additional amenities. Buildings equipped with freeboard, which offer enhanced flood protection, are anticipated to command higher market prices. This increase in value can impose greater financial burdens, such as elevated mortgage costs, thereby heightening vulnerability. The data for this factor was collected from the U.S. Census Bureau (TABLE ID: B25077) (USCB, 2019).
2.1.2 Decremental factors
Flood zone (FZ)
Living in flood-prone areas increases the likelihood of experiencing flooding events and consequently raises annual flood losses. Therefore, residents in such areas are more inclined to support the adoption of the new freeboard policy to enhance flood protection and mitigate flood losses. Data on the percentage of residential buildings in moderate to high-risk flood zones were sourced from the building inventory maintained by the Jefferson Parish Department of Floodplain Management & Hazard Mitigation (Jefferson Parish Geoportal 2021).
Average annual loss (AAL)
The Average Annual Loss (AAL) for each residential unit is computed using depth-damage functions (DDFs) provided by the U.S. Army Corps of Engineers (USACE; 2000), which relate flood depth to relative loss based on building construction costs. Higher annual flood losses for residential buildings are associated with a greater likelihood of owners adopting the new freeboard policy to mitigate future losses. The mean AAL for all residential buildings in the population unit was calculated following methods detailed in Gnan et al. (2022a, 2022b, 2024).
Insurance (I)
Homeowners in moderate to high-risk flood zones without flood insurance may exhibit reluctance or lack awareness of flood event consequences, potentially affecting their willingness to adopt the new freeboard policy. Conversely, homes with existing insurance policies may be motivated to implement freeboard to qualify for discounted premiums provided by FEMA. In this study, separate indexes were established:
1.
IAE: Percentage of households in flood zone AE with active flood insurance policies.
2.
IX: Percentage of households in flood zone X with active flood insurance policies.
The data for this factor was collected from OpenFEMA Data Sets (Dataset Name: FEMA NFIP Redacted Policies) (OpenFEMA Data Sets, 2022).
2.1.3 Standardization of factors
To conduct discriminant analysis effectively, it is crucial to standardize the data to ensure that variables are comparable and contribute equally to the analysis (Lakshmanan, 2019). Standardization involves transforming the variables to have a mean of zero and a standard deviation of one, which removes differences in the scales of measurement without altering the relationships between variables. This ensures robustness and accuracy in interpreting how different variables interact and contribute to the outcomes being studied. It also mitigates biases and inconsistencies that may arise from variations in data scales, magnitudes, or units. Once the data is standardized, each variable is expressed as a percentage relative to its total value within the study area, as shown in Eq. 3. This approach, known as weighted mean calculation, accounts for the influence of each variable proportionally to the total population of the study area (U.S. Census Bureau, TABLE ID: 2010: DEC Redistricting Data (PL 94–171)). Equations 4 and 5 show the equations used for standardizing the incremental and decremental factors, respectively.
$$Weighted \;Mean\; of \;factor= \left({\sum }_{i=1}^{n}Value \;of \;Factor \;in \;i*Population \;in \;i\right)/ Total\; population\; in \;study \;area$$
(3)
where,
n=total no of census tracts
$${\text{IF}}_{\text{i}}=\left(\left(\text{Value of Factor in i}-\text{Weighted mean of a factor in the study area}\right)/\left(\text{Weighted mean of the factor in the study area}\right)\right)*100$$
(4)
$${\text{DF}}_{\text{i}}= \left(\left(\text{Weighted mean of a factor in the study area}-\text{Value of factor in i}\right)/ \left(\text{Weighted mean of the factor in the study area}\right)\right)*100$$
(5)
where,
IF
index factors listed as incremental factors.
DF
index factors listed as decremental factors.
i
population units such as a state, county, census tract, block group, or blocks. We used the census tract for this study.
2.2 Vulnerability analysis with respect to median income
This study employed discriminant analysis to identify discriminating or grouping variables from a pre-selected list of predictor variables. These predictor variables, also referred to as index factors, include incremental and decremental factors, as outlined in Section 2.1. The choice of index factors can vary across studies depending on the targeted natural disasters and specific research objectives. In this study, we restricted the index factors to those that influence economic vulnerability in adapting to the new freeboard policy in Jefferson Parish.
The vulnerability score measures the relative economic vulnerability of households regarding the implementation of the freeboard policy, accounting for both the cost of adoption and the likelihood of take-up. This comprehensive index incorporates multiple socio-economic factors, providing a holistic understanding of vulnerability that extends beyond mere income and savings. Factors such as Average Annual Loss (AAL) from flood events and insurance take-up are included to represent the direct financial impacts that could deter freeboard adoption. By integrating these factors, we aim to capture the full spectrum of economic burden and the likelihood of policy implementation, creating a comprehensive measure that provides deeper insights into the actual challenges faced by households. Combining the costs of implementation and the probability of take-up into a single measure allows for a better understanding of the intertwined nature of economic burden and implementation feasibility. This approach ensures that our index accurately reflects both financial strain and adoption feasibility, offering a explicit understanding of household vulnerabilities.
2.2.1 Correlation test and wilks lambda test
In discriminant analysis, akin to MANOVA, an important assumption is that predictor variables should not be highly correlated to maintain predictive power (Carlson 2017). A correlation test was conducted among predictor variables to ensure correlations were below 0.7 (Statistics Solution 2021). Variables exceeding this threshold were assessed using a canonical structure matrix to eliminate those with lower discriminatory power. The significance of the discriminant model was assessed using Wilk's Lambda test, where a smaller value indicates stronger discriminatory ability (SPSS Statistics 2021). A significance level of 0.05 was employed for the study.
2.2.2 Test of equality for variable selection
The equality of group means test assessed each variable's significance in the discriminant model. Variables with a significance level greater than 0.2 were considered to contribute minimally to the model and were eliminated (SPSS Statistics 2021). The remaining variables proceeded to the next step of the analysis.
2.2.3 Calculation of vulnerability score
The income range and cost of living in Jefferson Parish were considered to determine the income groups. This classification is unbiased and independent of other factors considered in our study. As per the Pew Research Center that conducts extensive surveys and research on demographic trends, household income for middle-income groups ranges between two-thirds to double the median household income after adjusting for household size (PRC 2018). The median income of Jefferson Parish for a household size of 3 is $73,936. Table 1 shows the income range for each of the income groups calculated based on the median income of Jefferson parish (U.S. Census Bureau (TABLE ID: B19013)). Each census tract or population unit was categorized into one of the income groups, using these income ranges in Table 1.
Table 1
Income ranges of different income groups (Hubbard 2020)
Income Group
Median household income range
Low Income
Less than $49,290
Middle Income
$49,290—$147,892
High Income
More than $147,892
In this study, the SPSS software was used for conducting the discriminant analysis. The unstandardized canonical discriminant function coefficient for each parameter obtained from the discriminant analysis was used for calculating the vulnerability score for each census tract using Eq. 6.
$$V= \sum\nolimits_{i=1}^{n}wi \;Fk\; ,\; k=1\; to \;m$$
(6)
where,
V
Vulnerability Score,
w
unstandardized canonical discriminant function coefficient obtained from the discriminant analysis,
F
Standardized value of the factor for the particular parish
n
Total number of non-correlated and significant factors
m
Total number of census tracts
The vulnerability score is an empirical measurement of the vulnerability of a population group that could pose a barrier towards undertaking any new measure towards improving resiliency, which in this case is adapting to the new policy for flood mitigation in Jefferson Parish through the use of freeboard. The vulnerability score was calculated using Eq. 4 for each population unit. A scatter plot was developed to display the vulnerability score values corresponding to the population unit's adjusted median income. The adjusted median income was obtained by extrapolating the median income for an average household size of 3 members for each census tract. A line of best fit was obtained to express best the relationship between vulnerability and median income for the population units considered in the study.
2.3 Benefits analysis with respect to median income
To investigate the effectiveness of a mitigation strategy, a vulnerability vs. benefit study was conducted to analyze whether the vulnerability associated with adapting to the new policy in Jefferson Parish is proportional to the benefits obtained through freeboard. As previously mentioned, the benefits obtained from using freeboard were calculated in the form of savings obtained from two sources: 1) Reduced annual losses from flood damages and 2) Reduced insurance premium rate.
2.3.1 Reduced insurance premium
For homes in special flood hazard areas (i.e. A, AE, V, and VE flood zone), NFIP coverage is mandatory (U.S. Senate 2011). Flood insurance premiums vary based on the property's location, the flood zone, first-floor elevation, building characteristics, and the BFE. The higher the elevation compared to BFE, the less likely the home will flood and lower the premium. To assess the potential benefits of adding freeboard, the estimated reduction in annual flood premiums is added to its expected average annual flood loss. Although a decrease in flood premiums with elevation increase is not typically addressed in such studies, its inclusion allows for a more comprehensive evaluation of benefits associated with adding freeboard (FEMA 2008).
Annual premium at BFE, BFE + 1, BFE + 2, BFE + 3, and BFE + 4 scenarios is calculated based on Appendix 2. In this study, we considered the best-case scenario out of the five scenarios: BFE, BFE + 1, BFE + 2, BFE + 3, and BFE + 4 for each residential building in the population unit based on total annual savings (TS). The saving from insurance premiums was calculated as shown in Eq. 7. The average savings from the annual premium for all residential units in a population unit was calculated for this factor.
Using freeboard to increase the height of the residential building above the designated BFE helps substantially reduce annual flood losses. The average annual loss for the building was calculated per the corresponding best case (\({AAL}_{Freeboard}\)) out of the five scenarios that maximized the total annual savings (TS). The saving from average annual loss (SAAL) was calculated as shown in Eq. 8. The mean of reduced average annual loss for all residential units in a population unit was calculated for this factor.
The benefit from adapting to the new freeboard policy was obtained in terms of total savings, as shown in Eq. 9.
$${TS}_{i}= {SI}_{i}+ {SAAL}_{i}$$
(9)
where,
TS
Total annual savings.
SI
Savings from reduced premium.
SAAL
Savings from a reduced average annual loss.
I
population units.
The total savings was plotted against the adjusted median income for each population unit to analyze whether benefits varied with the income level. A scatter plot and a line of best fit were obtained to express the relationship between Benefits and Median income for the population units considered in the study.
2.4 Vulnerability to benefit analysis
The higher income groups can be expected to have a higher flood loss in terms of dollar value due to the possession of higher wealth. It can also be expected that higher income groups may pay a higher insurance premium due to the higher building value of their residential units (bigger homes in terms of sq. ft.). Since the savings are dependent on reduced annual flood loss and reduced insurance premiums, savings will be higher for the higher-income groups. Thus, a direct comparison of savings vs. vulnerability may not yield conclusive answers. Furthermore, the value of money changes with the economic standing. For example, dollar x may have a different value for two different people, where one person's income is twice that of the other. Therefore, to get a more accurate comparison of benefits vs. vulnerability, the benefit for each population unit was calculated as a percentage of the median income of the population unit, as shown in Eq. 10.
The benefits were then plotted against the vulnerability score for each population unit to understand the relationship between vulnerability and benefit from freeboard. The benefit-to-vulnerability analysis can be a substitute for a cost–benefit analysis to understand and quantify the feasibility of adapting to the new policy through freeboard. Such a framework can also be used by government agencies when planning the implementation of new risk management strategies in the future.
3 Case study
Jefferson Parish, in the state of Louisiana, has a significantly high poverty rate. It is a part of the New Orleans-Metairie metropolitan area that has a history of flooding events. It is subjected to flooding from two sources: (1) tropical storm/hurricane induced storm surge from the Gulf of Mexico and (2) urban pluvial flooding from heavy rainfall related to tropical storms, hurricanes, and unusual rain events. Thus, a considerable amount of the population in the parish is expected to suffer from heavy annual losses from flood damages. Since the purpose of this study is to analyze the benefits of using freeboard for the lower-income group, only those cities of Jefferson parish that had a high poverty rate were chosen. The data from the year 2019 was used to conduct this study. The cities with a poverty level of 20% or higher were selected for this study. As such, the following cities were taken into consideration:
1.
Terrytown
2.
Bridge City
3.
Waggaman
4.
Westwego
5.
Marrero
6.
Avondale
7.
Gretna
8.
Harvey
9.
Woodmere
The study was conducted at the census-tract level to obtain a substantial amount of standardized data suitable for conducting the discriminant analysis. Also, the census tract-level data could be easily obtained with an acceptable margin of error and very few missing data points, which could be easily imputed. A total of 42 data points were initially considered. However, four census tracts had no data available and, therefore, were eliminated from the study. A total of 38 data points were finally considered for the study. The income level of each census tract was determined using the median household income after adjusting for an average household size of 3 members. Using these income ranges, each census tract was categorized into one of the income groups as per Table 2. Only the low-income and middle-income census tracts have been considered for this study, as none of the census tracts in the chosen study area fell under the high-income group. Figure 2 below shows the location of the block groups selected for the study on the geographical map.
Table 2
Count of data points in each income level for each city selected for the study
City
Poverty Rate
Total
Low
Medium
High
Inconclusive Data
Terry town
28.6%
7
3
2
0
2
Bridge City
26.7%
1
1
0
0
N/A
Waggaman
26.6%
2
1
1
0
N/A
Westwego
24.8%
2
2
0
0
N/A
Marrero
24.7%
10
4
6
0
N/A
Avondale
21.8%
2
2
0
0
N/A
Woodmere
21.8%
3
2
1
0
N/A
Harvey
20.2%
7
3
2
0
2
Gretna
20%
8
1
7
0
N/A
Total
42
19
19
0
4
Fig. 2
Map showing the location of selected higher and lower-income census tracts in the Jefferson Paris
×
This study is sensitive to building area as several cost-related factors, such as AAL, BV, and CC, vary with the area of the built structure. Therefore, only single story-homes with a ground coverage range of 500 sq. ft. – 2000 sq. ft were considered for this study. Since over 99% of the residential buildings in the study area were located in flood zone X or flood zone AE, only these two zones were considered. In addition, this study was conducted only for owner-occupied households. The Jefferson Parish ordinance has mandated the following freeboard guidelines for home construction under the Code of Ordinance, Chapter 14 (Jefferson Parish 2022).
1.
BFE + 1 ft of freeboard for new home constructions and substantial home renovations in flood zone AE located inside the levee-protected area.
2.
BFE + 1 ft of freeboard for new home constructions and substantial home renovations in flood zone VE located outside the levee-protected area.
3.
BFE + 2 ft of freeboard for new home constructions and substantial home renovations in flood zone AE located outside the levee-protected area.
4 Results and discussions
4.1 Vulnerability analysis with respect to median income
The data for each factor, along with required calculations and standardization for each, were conducted as per Section 2.1. The Pearson correlation test found a high positive correlation between AAL, CC, and FZ. The detailed results of the correlation test are provided in Appendix 3. A geographical area in a high-risk flood zone can be expected to have a higher flood level and, therefore, a higher annual loss. A higher flood level also means a greater height for optimum freeboard and, thereby, a higher value for the construction cost. Therefore, this correlation was expected. As per the structural canonical matrix, it was found that FZ had a higher importance in the model as compared to AAL and CC. Therefore, FZ was retained, whereas AAL and CC were eliminated from the model. The discriminant analysis was then conducted with the remaining 11 predictor variables in the study. The results of Wilk’s Lambda Test are given in Appendix 4. The significance level of the model was less than 0.05. Therefore, the discriminant model was deemed significant for the study. The test of equality of group means as shown in Appendix 5 deemed the following variables as significant, whereas the rest of the variable were eliminated from the study.
1.
Poverty (P),
2.
Household Size (H),
3.
Children (C),
4.
Elderly (E),
5.
Ethnicity (R), and
6.
Building value (BV)
The unstandardized canonical discriminant function coefficients obtained from the discriminant analysis were considered as weights of each significant parameter for calculating the vulnerability score. Thus, using Eq. (6), the vulnerability score for each census tract was calculated as shown in Eq. 11:
A scatter plot was obtained for Vulnerability score vs. median income to understand the relation between the vulnerability associated with freeboard installation and the different income groups. A best-fit curve was then overlaid to understand the trend depicted in Fig. 3.
Fig. 3
Scatter plot showing the comparison of Vulnerability Score with Median Income
×
As is evident from Fig. 3, vulnerability decreases with increasing median income. In other words, census tracts in the lower-income category have greater liabilities and barriers to investing in freeboard as a mitigation strategy to reduce flood loss. Thus, a correlation can be drawn here that the lower-income population faces more significant barriers and liabilities toward freeboard implementation. These findings align with existing literature emphasizing the disproportionate impact of climate change and adaptation measures on economically disadvantaged populations (Shonkoff et al. 2011), thereby establishing that the current freeboard policy in Jefferson Parish has a similar impact.
4.2 Benefit analysis with respect to median income
A scatter plot was obtained for each census tract's Annual Savings vs. Adjusted Median Income to understand the benefits of investing in freeboard and the different income groups. A best-fit curve was then overlaid to understand the trend depicted in Fig. 4.
Fig. 4
Scatter plot showing the comparison of Annual Savings with Median Income
×
The scatter plot in Fig. 4 shows that annual savings from freeboard implementation do not correlate strongly with median income. This horizontal trendline suggests that other factors, such as AAL, flood zone status, and building value, significantly influence savings. Therefore, while freeboard implementation can provide substantial economic benefits, these benefits are not uniformly distributed across different income groups.
4.3 Vulnerability to benefit analysis
The total savings was also plotted against the vulnerability score for each census tract, as shown in Fig. 5.
Fig. 5
Scatter plot showing the relation of total annual savings with vulnerability
×
As evident from Fig. 5, census tracts with lower vulnerability scores realize greater absolute savings from freeboard implementation. However, this direct comparison does not account for the varying financial contexts of different income groups. As the higher income groups possess the higher value of financial assets, the savings will also be higher as compared to lower income groups who possess financial assets of lesser value. Therefore, comparing savings in terms of monetary value does not provide a true reflection of how it may impact different income groups. To address this, we recalculated savings as a percentage of median income and plotted these against vulnerability scores, as shown in Fig. 6.
Fig. 6
Plot showing vulnerability vs. benefits obtained
×
As illustrated in Fig. 6, there is an increasing trendline for vulnerability versus benefits, indicating that census tracts with higher vulnerability scores experience comparatively greater savings when freeboard is implemented. This suggests that the economic benefits of freeboard are more pronounced in areas with higher vulnerability, aligning with the increased need for mitigation strategies in these regions.
If the current practices and costs associated with freeboard construction are considered, the implementation of freeboard as a flood mitigation strategy proves to be more advantageous for residents in high-risk flood zones with significant flood loss potential. These residents stand to gain more substantial economic benefits compared to those in areas with lower flood risks and nominal losses. This finding underscores the critical role of targeted adaptation measures in maximizing the efficacy of flood risk management policies. However, the affordability of freeboard remains a crucial factor influencing the extent of benefits realized. Households in lower-income brackets often face significant financial barriers to implementing such mitigation strategies despite the potential long-term savings. Consequently, there is a pressing need for policy interventions to support these vulnerable communities.
To enhance the adoption rates and equitable distribution of benefits, the following measures are recommended:
Assistance Programs: Government and non-governmental organizations should establish assistance programs that provide financial aid or subsidies to lower-income households for freeboard implementation. These programs can significantly reduce the upfront costs, making it more feasible for economically disadvantaged groups to adopt freeboard measures.
Special Schemes: Tailored schemes, such as low-interest loans or grants specifically designed for flood mitigation efforts, can provide the necessary financial support for vulnerable populations. These schemes should be designed to prioritize high-vulnerability areas to ensure that resources are allocated where they are most needed.
Discount Rates: Offering discounted rates for freeboard construction to households in high-vulnerability zones can incentivize adoption and help balance the economic burden. This approach can be particularly effective when combined with public awareness campaigns that educate communities about the long-term benefits and cost savings associated with freeboard.
By implementing these targeted financial strategies, policymakers can alleviate the economic barriers faced by vulnerable communities, thereby enhancing the overall effectiveness and equity of flood risk management practices. This approach not only promotes greater resilience against flooding but also ensures that the benefits of mitigation strategies are distributed fairly across diverse socio-economic groups. These findings contribute to the broader discourse on climate adaptation and socio-economic equity, highlighting the importance of integrating vulnerability considerations into policy frameworks. While this study is conducted at the census tract level for a specific area, the methodology is robust and adaptable for application at the individual household level, provided that detailed data is available. This adaptability ensures that our approach can be tailored to various contexts, enhancing its relevance and applicability across different regions and populations.
5 Future works and limitations
The authors acknowledge that the list of index factors identified as predictor variables in this study may not be all-encompassing. More variables may be added to this list in future works. There are several limitations of the study, mainly related to data collection and the selection of parameters for initial consideration. These limitations are as follows:
Since data in this study is collected from various sources, maintaining consistency across all parameters could be an issue.
The margin of error was not considered while collecting the data for each parameter.
The parameters are not constant and may differ with geographical location, population type, and measure implemented even for the same natural disaster.
This study only considered the direct physical economic losses. It does not consider pre-existing conditions or vulnerability due to indirect means such as disruption, displacement, or relocation.
Furthermore, the lower to the middle-income category in the American housing market generally tend to buy an existing home due to the lower cost. Only a tiny percentage of the population buys brand-new houses. The price is based on the home appraisal, which considers the location and price of other homes in the vicinity rather than the actual construction cost. In such a scenario, the freeboard cost might affect the builder's profit margin but does not affect the homeowners. In this study, we assumed that there would be policies mandating a substantial increase in the base elevation of buildings above BFE at some point in time. In such a scenario, we can expect that there will be a rise in the price of homes in every locality due to the extra cost associated with elevating the building above BFE, such as by using freeboard. Secondly, people in the U.S. can buy land and build their own homes after procuring a building permit from the city or county administration. However, this method is not commonly practiced. However, for this research, we assumed that homeowners are key participants or stakeholders in the construction process and, therefore, have a choice of whether to use freeboard depending on their economic capabilities.
6 Summary and conclusion
With the increase in the number of flood events in the U.S., the federal and state governments have been facing massive yearly losses due to flooding, and lower-income communities tend to suffer disproportionately. Though several researchers have come up with solutions and incentive programs that may help to reduce losses and chances of flooding, especially for low-income households, standardizing these solutions is necessary to enable their appropriate and optimized implementation. This study aimed to develop a framework to help identify underlying causes of economic disparity and empirically measure the vulnerability involved in freeboard implementation using a discriminant analysis method. It was found in this study that:
The discriminant analysis method can help to identify underlying causes of disparity.
Using the coefficients obtained from the discriminant analysis, it is possible to empirically calculate the level of economic risk or vulnerability involved for different income groups.
The economic risks or barriers to implementing freeboard increase with a decrease in income.
However, the benefits in terms of annual savings have an almost equal distribution across all levels of vulnerability.
From this study, it is convincing that there is a need to promote the use of freeboard amongst low-income communities and low-cost housing. The lower-income communities tend to bear higher flood loss, and the construction cost involved in implementing freeboard could be an additional economic burden. In this regard, they may not be capable of readily adapting to the new policy in Jefferson parish through the use of freeboard. Therefore, there is a need for the federal, state, and local governments to incentivize the use of freeboard for the low-income group to encourage and enable them to invest in construction with freeboard. Lowering the price of freeboard may also allow the communities to use freeboard over the recommended optimized level to further reduce or completely nullify losses due to flood damage and reduce insurance premium rates to a bare minimum.
Declarations
Competing interests
The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix 1: Upper Limit of Mean Cost of Construction Increase (%), by Freeboard (FEMA 2008)
Freeboard (ft)
A, AE and X-Zone
1
2.3%
2
4.5%
3
6.8%
4
9.1%
This study, conducted by Louisiana State University, analyzed various scenarios from no freeboard to four feet above Base Flood Elevation (BFE), aiming to optimize freeboard usage for flood loss reduction at minimal investment (Gnan et al. 2022a, 2024; FEMA 2008).
Appendix 2: Calculation for Annual Insurance Premium
For each flood zone, Flood Insurance Rate Map (FIRM) assigns the 100-year flood as BFE, and rates are estimated by comparing the building's elevation to BFE. This research calculates premiums using the rate tables published in Appendix J (Rate Tables) of the NFIP (2021) Flood Insurance Manual's post-firm construction rates for a single-family residence for multiple elevation levels. Basic rates for building and content are applied to every $100 of the basic building and content coverage limits, and separate additional rates for building and content are used for every $100 additional coverage. For single-family homes, $60,000 is the basic building coverage, and $25,000 is the basic content coverage, with maximum limits of $250,000 for building and $100,000 for content. NFIP requires a minimum deductible of $1,250 for building and content if the coverage exceeds $100,000 (NFIP 2021); therefore, $1,250 was chosen as a conservative value.
For each building, the total building basic insurance premium (\({G}_{{b}_{B}}\)) is the basic coverage limit (\({P}_{{b}_{B}}\)) for the building multiplied by its basic rate (\({R}_{{b}_{B}}\)). The total additional insurance premium for the building \({(G}_{{a}_{B}})\) is the additional coverage amount (\({P}_{{a}_{B}}\)) multiplied by the building's additional rate (\({R}_{{a}_{B}}\)), as shown in equation (12).
Total contents basic insurance premium (\({G}_{{b}_{Ct}}\)) is the basic coverage limit (\({P}_{{L}_{Ct}}\)) for contents multiplied by its basic rate \({R}_{{b}_{Ct}}\), whereas the total additional insurance premium for contents (\({G}_{{a}_{Ct}}\)) is the additional coverage amount (\({P}_{{a}_{ct}}\)) multiplied by contents additional rate (\({R}_{{a}_{Ct}}\)) as shown in equation (13).
\({G}_{{b}_{B}}\) is added to \({G}_{{a}_{B}}\) and the \({G}_{{b}_{Ct}}\) is added to \({G}_{{a}_{Ct}}\) to calculate the principal premium (\({P}_{PL}\)) as shown in equation (14).
\({P}_{PL}\) is multiplied by the deductible factor \(d\) (NFIP 2021) for the chosen deductible to obtain the deducted premium (\({P}_{d}\)), as shown in equation 15.
$${P}_{d}={P}_{PL} \times d$$
(15)
According to (NFIP 2021), the annual premium is calculated as shown in Eq. 11. The \({P}_{d}\) is added to the Increased Cost of Compliance (ICC) premium, then subtracted by the Community Rating System (CRS) discount. The Reserve Fund Assessment (RFA) percentage is added to the total premium after calculating the ICC premium and CRS premium discount. The Homeowner Flood Insurance Affordability Act of 2014 (HFIAA) surcharge and federal policy fee (FPF) are added to determine the total annual premium (P, in USD), as shown in equation 16.
