Mandatory climate risk disclosure, housing prices, and credit supply
Xi Chen
Review of Accounting Studies · 2026-07-09
CC BY 4.0 · https://creativecommons.org/licenses/by/4.0/
Source: https://link.springer.com/article/10.1007/s11142-026-09974-z

Abstract

This paper examines how climate risk transparency influences house prices and credit supply. I contend that inadequate property-level climate risk disclosures induce homebuyers to demand a risk aversion discount on house prices, creating a potential market for lemons. Employing a stacked difference-in-differences design, I find that flood risk disclosure laws, by enhancing dwelling-specific flood risk transparency, raise house prices on average by 7.6%. Results strengthen in states with stricter disclosure requirements. Exploiting within-state heterogeneity and controlling for housing market trends, I find that the effects persist and intensify in regions with higher aggregate exposure to flood risk, greater information frictions, and more attention to climate risks. Conversely, the effectiveness of flood risk disclosure laws attenuates in regions where households are less concerned about climate change. Additionally, less sophisticated lenders extend credit to financially constrained borrowers in response to the laws but experience lower profitability.

Introduction

This paper examines how mandatory, property-specific flood risk disclosures affect house prices and credit supply. Coastal flooding is among the most economically damaging consequences of climate change (Buchanan et al. 2017 ). By 2040, an estimated 30-cm sea-level rise could threaten at least one million people (Hauer et al. 2016 ) and coastal properties valued at nearly $1 trillion dollars (Rao 2017 ). While studies document house price discounts in flood-prone areas (e.g., Bernstein et al. 2019 ), they typically do not examine the information channel through which flood risk is reflected in housing markets.

Information friction surrounding individual properties’ flood exposure remains substantial because the primary public information source, the Federal Emergency Management Agency’s flood zone maps under the National Flood Insurance Program, provides flood risk information only at an aggregate level. This simple location-based classification fails to capture property-specific characteristics and thus obscures the actual risk faced by homeowners (Palmer and Turrentine 2022 ). As Scata ( 2025 ) notes, “Inadequate or non-existent [property-level] disclosures are putting unsuspecting buyers of previously flooded homes at substantially more risk of paying out of pocket for unexpected flood damages.” Second, in many states, sellers are not required to disclose key dwelling-level information, such as whether a property is covered by private flood insurance. Flood insurance information can convey forward-looking risk exposure at the property level and can materially affect buyers’ decisions. Darlington ( 2018 ) writes: “Hiding the flood insurance information from consumers can lead them to buy a severe repetitive loss property without knowing it.”

When asymmetric information exists about a dwelling’s idiosyncratic exposure to flood risk (Bakkensen and Barrage 2022 ; Ouazad and Kahn 2022 ), buyers, who bear added uncertainty, may demand compensation in the form of lower average selling prices—a so-called lemons discount (Frankel et al. 2025 ; Cho et al. 2024 ; Akerlof 1970 ). To address these information frictions, 15 coastal states enacted flood risk disclosure laws between 1993 and 2017. These laws require a more comprehensive and standardized approach to flood risk disclosure in real estate transactions. The laws go well beyond federal measures and the map-based proxies by mandating detailed, property-specific disclosures that buyers and lenders cannot readily obtain on their own. Under the laws, before a contract is signed, sellers must provide, to prospective buyers, a standardized disclosure form containing property-level information, such as prior flood damages, flood frequency, flood insurance coverage, disaster-aid receipts, and remediation efforts. Importantly, the disclosures include not only past flood damages but also the status of private flood insurance policies, which conveys information about future risk exposure. Flood risk disclosure laws further specify sellers’ legal liability for any errors, omissions, or inaccuracies in their disclosures. Noncompliance may invalidate the sales contract and expose sellers to monetary penalties, including damages of up to three times the actual loss as well as buyers’ legal costs (National Association of Realtors 2019 ). Overall, flood risk disclosure laws aim to make flood risk information available at the dwelling level, enhancing housing market transparency.

I hypothesize that flood risk disclosure laws can inform prospective buyers about property-specific flood risk and consequently reduce information frictions and the lemons discount, thus raising housing prices, on average, in the market. Because lenders evaluate credit risk based on collateral values, reduced uncertainty about housing prices also lowers credit risk and facilitates mortgage lending. Studies provide both theoretical and empirical support for this argument. Di Maggio et al. ( 2017 ) and Christiano et al. ( 2014 ) show that heightened uncertainty depresses housing prices and credit supply, while Nanda and Ross ( 2012 ) demonstrate that asymmetric information leads buyers to demand price discounts. Related evidence from other markets further supports this explanation. Cho et al. ( 2024 ) show that improved information reliability increases selling prices in used car markets, and Sinha ( 2022 ) finds that corporate fracking disclosure laws expand mortgage lending in nearby areas.

At the same time, flood risk disclosure laws may not fully eliminate risk-related price discounts if existing location-based information is already sufficient, if sellers do not comply, or if buyers and lenders are unaware of or do not use the mandated disclosures. Compliance can be costly for sellers, as they must prepare and verify the information. Large institutional lenders may rely less on flood risk disclosure laws than small lenders do because they have sophisticated risk models in place (Nguyen et al. 2022 ). And less experienced market participants may neglect climate risk information due to limited attention to or disbelief about climate change (Baldauf et al. 2020 ). Overall, whether the laws influence the house prices and credit supply is an open empirical question.

To examine the issue, I combine residential real estate prices from the Federal Housing Finance Agency, mortgage loan data from the Home Mortgage Disclosure Act, aggregate flood zone classifications from the National Oceanic and Atmospheric Administration, media outlet information from the Federal Communications Commission, and disaster declarations from the Federal Emergency Management Agency. These data allow me to construct a census tract-level panel from 1990 to 2017. I supplement these data with demographic characteristics from the U.S. Census Bureau and bank-level information from Call Reports. Using a stacked difference-in-differences design (Cengiz et al. 2019 ; Baker et al. 2022 ), I find that house prices increase by approximately 7.6% following the adoption of flood risk disclosure laws, compared to properties without such laws, consistent with lower information asymmetry driving a reduction in the risk premium. This magnitude is economically meaningful, as studies show that improvements in market transparency on property conditions and past transactions raise house prices by 6.4% (Nanda and Ross 2012 ; Eerola and Lyytikäinen 2015 ) and mortgage lending by 9.9% (Di Maggio et al. 2017 ).

I conduct robustness tests to validate these findings. Specifically, I include state-year fixed effects and estimate triple-difference-in-difference specifications to control for broader market trends in each state. I also perform parallel trends analyses using multiple F-tests, placebo adoption dates, and a contiguous border-county design to mitigate concerns about omitted confounders. I find no evidence of spillover or anticipation effects. Together, these results suggest that the likelihood of an unspecified correlated omitted variable driving the findings is low. For example, if other contemporaneous state policies were to explain the observed price increases, they would need to disproportionately affect high-flood-exposure regions.

Having established the baseline results, I next present three categories of cross-sectional analyses. First, I examine cross-state heterogeneity by assessing whether the effects vary with the strength of state-level disclosure requirements. Using both a ranking from the Natural Resources Defense Council and a self-constructed disclosure index based on nine legal dimensions, I show that the positive effects of flood risk disclosure laws strengthen in states with more stringent disclosure requirements. Second, exploiting within-state variations, I document that the laws’ effects strengthen in markets where buyers demand larger price discounts to compensate for uncertainty, such as areas with greater information friction (due to fewer media outlets) or climate risk attention (due to more salient natural disasters in the neighborhood). Third, I show that the results attenuate in regions where households are less concerned about climate change. This finding suggests that, while disclosure-based regulation reduces price discounts by promoting detailed and credible information, its effectiveness may be limited in markets characterized by weak beliefs in the issues being disclosed.

I next examine whether the flood risk disclosure laws-induced improvement in collateral transparency affects mortgage market outcomes by increasing lenders’ willingness to supply credit. By reducing uncertainty about property-specific flood risk and collateral values, flood risk disclosure laws improve lenders’ assessment of mortgage risk and facilitate loan origination. This expansion in credit supply, especially among financially constrained borrowers, can further amplify housing price responses (Favara and Imbs 2015 ). Consistent with this mechanism, I find that following the adoption of these laws, lenders are less likely to reject mortgage applications from buyers facing housing cost burdens. The lenders also offer larger loans and higher loan-to-value ratios to these borrowers. Moreover, the post-adoption increase in lending strengthens among smaller and less profitable banks, suggesting that lender sophistication and economic incentives affect the extent to which the mandated flood risk information is incorporated into credit decisions. Taken together, these findings support the view that flood risk disclosure laws make flood risk information more accessible to less experienced lenders, which reduces their uncertainty about collateral values and credit risk and thus supports housing price increases. Nevertheless, I find that the expansion in lending has unintended consequences for banks’ profitability, indicating greater risk-taking at the expense of returns (Kang et al. 2021 ).

As discussed above, these analyses rely on the assumption that sellers comply with flood risk disclosure laws and provide standardized disclosures that buyers and lenders are aware of and use. Although this assumption is plausible, market participants may rely on other information sources. To address this concern, I survey 100 mortgage brokers, real estate agents, and recent homebuyers in the United States using Qualtrics. The majority of respondents (77%) report high compliance with the laws, and 100% of respondents indicate that disclosures are provided using standardized forms. Approximately 70% believe that the laws help buyers make better-informed decisions, and most support expanding similar disclosure requirements nationwide. Importantly, 86% report that the disclosure forms contain information not otherwise publicly available. Overall, this evidence supports the view that flood risk disclosure laws increase transparency and that sellers comply with them.

This study contributes to the literature on the real effects of mandatory disclosure and responds to Leuz and Wysocki’s ( 2016 ) call to study disclosure mandates outside traditional corporate settings, where causal effects are often difficult to identify. I examine a novel nonshareholder disclosure regulation designed to protect consumers, in which the behaviors of sellers, buyers, and lenders, unlike in corporate disclosure settings (e.g., Greenstone et al. 2006 ), are directly observable. Given rising climate-related damages and growing policy interest in transparency-based solutions (Carattini et al. 2022 ), my evidence yields timely and policy-relevant insights into how mandated property-level climate risk transparency affects house prices and credit supply.

The study also adds to the growing literature on climate risk disclosure (Frankel et al. 2025 ; Christensen et al. 2021 ), which has focused on non-U.S. settings (Downar et al. 2021 ; Fiechter et al. 2022 ; Krueger et al. 2024 ). Footnote 1 In the United States, there is no universal, industry-spanning federal mandate for climate-risk reporting, and the SEC’s proposed standardized climate disclosure rule has been paused due to legal concerns. Footnote 2 State-level flood risk disclosure laws therefore represent one of the few mandatory climate-related reporting regimes in place. These laws inform investment decisions made by millions of households whose housing and financing choices significantly affect the real economy. Footnote 3 By first integrating both the timing and the comprehensive content of multiple U.S. flood risk disclosure mandates, I provide new evidence that mandatory climate-related disclosures increase housing prices and bank lending by reducing dwelling-level information asymmetry. Footnote 4 Taken together, the results suggest that mandatory disclosures, rather than policies that require costly real actions, can serve as an effective starting point for improving investment decisions and market transparency (Bolton et al. 2021 ). However, the findings also highlight that the effectiveness of disclosure regulation depends on not only information availability but also market participants’ beliefs. In the climate context, expanding the information alone may be insufficient when participants do not accept or fail to process the revealed risks.

Finally, this study has implications for banking, as it contributes to the growing literature on the collateral channel of mandatory disclosures (Sinha 2022 ; Kang et al. 2021 ). While research shows that improved reporting can enhance lending outcomes (Kang et al. 2021 ) and that disaster-related policies can affect banks’ risk taking (Lu and Nikolaev 2024 ), I find that flood risk disclosure law-induced credit expansion has unintended consequences for bank profitability. These results point to a trade-off associated with disclosure mandates, as improved transparency can facilitate credit growth but also encourage bank risk taking.

State flood risk disclosure legislation

2.1 Background of U.S. flood risk disclosure legislation

In the absence of flood risk disclosure laws, information frictions surrounding individual properties’ flood exposure are high. The main information source, the Federal Emergency Management Agency’s flood zone map under the National Flood Insurance Program, has at least four drawbacks for buyers. First, the flood zone map provides information only at the aggregate level. Palmer and Turrentine ( 2022 ) demonstrate that this simple location-based information obscures the actual risk faced by homeowners and fails to reflect property-specific characteristics. In other words, the flood map indicates the community's overall exposure but does not specify the risk for individual homes within that community. Moreover, buyers cannot search past flood damage claims using a home’s address (Kusisto 2019 ), and the agency only shares flood claim history with current homeowners upon written request (Darlington 2018 ). The Privacy Act of 1974 also prohibits local governments from sharing this information publicly. Second, sellers need not disclose certain key information, such as whether they have private flood insurance. Flood-insurance information reflects the future risk exposure of individual properties and plays a key role in buyer decision-making. Footnote 5 In addition, sellers have no liability to buyers regarding post-purchase flood damages, and buyers cannot sue sellers for nondisclosure of environmental attributes (Nanda and Ross 2012 ). Third, only properties in Special Flood Hazard Areas within participating communities are required to disclose their flood zone location information. The mapping process for these areas is often seen as political, and a large proportion of areas that are exposed to flood risk are not required to disclose (Nguyen et al. 2022 ). Footnote 6 Lastly, the map is not always accurate and is difficult to interpret and compare across properties (Bakkensen and Barrage 2022 ; Palmer and Turrentine 2022 ). Matthew Eby, founder of the climate-risk-assessment company First Street, has criticized the information provided by the agency, saying, “These are areas where there was either no analysis done, or where analysis was done but the region was determined to have no special flood hazards …. We know the maps are fundamentally wrong” (Kusisto 2019 ).

Brokers and agents do little to eliminate the buyers’ information disadvantage (Garmaise and Moskowitz 2004 ). Brokers earn a percentage of the sale price, so they have an incentive to push buyers to pay more. And, in many jurisdictions, the buyer’s broker legally represents the seller, not the buyer. Appraisers and inspectors offer little help as well. Because they work quickly and cheaply, they do not always gather detailed data on hidden attributes related to flooding. They tend to satisfy regulatory requirements but otherwise do little to inform buyers. Even a thorough inspection reveals only a property’s physical condition, not whether it has received federal disaster aid or been covered by private flood insurance. In contrast, the sellers’ understanding of local risk and property-specific exposure is much deeper than a low-cost appraisal can capture.

In the housing market, sellers’ information advantage over buyers is often amplified when it comes to environmental attributes (Pope 2008 ). Potential homebuyers are likely to face high search and attention costs for learning the risks specific to each home and often make their home-buying decisions within a short period. Footnote 7 In addition, the lender may have a financial incentive not to provide the buyer with flood information early in a transaction: if the value of completing the transaction exceeds the cost of default risk, then providing timely flood information may discourage a potential buyer and impose additional costs on the lender. In contrast, sellers (current homeowners) are likely to be better informed about their home’s flood risk exposure, due to floods they have observed near their homes. Their informational advantages pertain not only to past damages but also to flood insurance and remediation efforts.

2.2 Flood risk disclosure legislation requirements and compliance

To address the information frictions described above, 15 coastal states introduced laws requiring more comprehensive and rigorous disclosure of property-level flood risk in real estate transactions from 1993 to 2017 (Table 1 Panel A). Most of the regulated states are in coastal areas with significant flood hazard potential. Figure 1 Panel A shows the number of mandates for flood risk disclosures by state and the distributions of the contents of these disclosures. Flood risk disclosure laws go far beyond federal measures and the map-based proxies by mandating detailed, property-specific disclosures that buyers and banks cannot readily obtain. A flood risk disclosure law typically requires sellers to complete a standard form that includes questions about factors such as a home or property’s flood history, flood insurance policies, hidden structural damages, flooding claim history, and past disaster-related aid (Federal Emergency Management Agency 2022 ). Appendix B shows, as an example, the disclosure form in Louisiana, which must be given to prospective buyers prior to the contract being signed (Federal Deposit Insurance Corporation 2019 ; Federal Emergency Management Agency 2022 ). Beyond these general requirements, states may also mandate more detailed disclosures. For instance, some states require disclosure of measures and steps to remedy flood problems (e.g., Mississippi and Louisiana), while others request information about drainage (e.g., District of Columbia) or a property’s dam inundation. California also grants a three-day rescission period during which buyers who have signed a contract can terminate a transfer if the disclosure is not made.

Under flood risk disclosure laws, sellers must disclose not only past flood damages but also the status of private insurance policies with insight about future risk exposure. Sellers cannot choose their method of disclosure, as a standard format is stipulated. More importantly, these laws articulate the sellers’ liability arising from errors, omissions, and inaccuracies in the disclosures (Troy and Romm 2004 ). Failure to comply invalidates the contract of sale, so sellers have an incentive to disclose the required information early in the process. Sellers can be sued and held liable to buyers for monetary penalties (National Association of Realtors 2019 ) of up to three times the actual damages, and the court may additionally require the seller to pay a buyer’s costs and legal fees.

To validate whether lenders truly care about flood risk–related disclosure, I examined their public disclosures and plotted, in Fig. 1 Panel B, the number of financial institutions that disclose flood risk in their annual reports each calendar year. The trend indicates that flood risk has become an increasingly important risk type for creditors over time. To further illuminate compliance with flood risk disclosure laws, I surveyed 100 mortgage brokers, real estate agents, and recent homebuyers in the United States. The respondents had balanced demographic backgrounds (gender, age, Hispanic origin, ethnicity, and region) and were given a 20-day response window. The survey was created using the software Qualtrics, and its questions can be broadly classified into four categories: background information, current flood risk disclosure practices, legal protection and enforcement, and market impact and consequences of noncompliance. The final survey instrument consisted of 15 questions. Table 2 reports the key results. Footnote 8 The majority of respondents (77%) report high compliance with flood risk disclosure laws, and 100% of respondents indicate that disclosures are provided using standardized forms. Approximately 70% believe that flood risk disclosure laws help buyers make better-informed decisions, and most support expanding similar disclosure requirements nationwide. Importantly, 86% report that the disclosure forms contain new information not otherwise publicly available. Among real estate agents and mortgage brokers, 57% report difficulty obtaining flood risk information prior to transactions, 71% cite lack of information availability as the primary barrier before flood risk disclosure laws are implemented, and 43% find flood risk information difficult to interpret, consistent with substantial information asymmetry. Moreover, 86% report experiencing transactions that were delayed or terminated due to flood disclosure issues. Overall, the survey strongly supports the view that flood risk disclosure laws enhance transparency and that sellers comply with them.

2.3 Predictors of U.S. flood risk disclosure legislation

If states introduce flood risk disclosure legislation in times of local economic stagnation, this could threaten my identification strategy. To examine whether economic, political, and fiscal factors explain the introduction of flood risk disclosure laws, I use a predictive regression. In my main tests, I intentionally focus on the timing as opposed to the passage of the laws (i.e., when, not if), and I believe that this timing is plausibly exogenous. Following Granja ( 2018 ), I estimate a Weibull proportional hazards model to explain the timing of flood risk disclosure laws across states as a function of local economic and political factors. Time to FRDLs is the expected time to the flood risk disclosure laws. Disasters is the average number of natural disasters that could have caused flooding (e.g., coastal storms, floods, hurricanes, typhoons) and that affected a given state in the past three years. Controls are a vector of state-level socioeconomic variables. Table 1 Panel B shows that most lagged state characteristics do not significantly predict the introduction of flood risk disclosure laws. This lack of predictability is consistent with the presence of considerable friction in the passage and implementation of this sort of legislation.

Although Disasters is positively significant, it indicates that natural disasters are associated with delays in the timing of the adoption of flood risk disclosure laws. The model results suggest that states are more likely to delay the laws when flood risk is more salient. However, the effects I consider occur several years later. In other words, the disasters occur several years before the housing market periods I consider, by which time their economic effect has been purged. Thus, the delays are unlikely to affect my conclusions. Moreover, I obtain consistent results for a sample without any natural flooding disasters (Table OA3).

Empirical predictions

Given the information frictions discussed in Sect. 2.1 , there is a pooling equilibrium between high- and low-exposure properties in the market before the enactment of these laws. Potential homebuyers cannot differentiate the cross-sectional variations, so they price based on the worst-case scenario (Gilboa and Schmeidler 1989 ); that is, they may request a risk aversion premium, resulting in market-level discounted housing prices. Nanda and Ross ( 2012 ) suggest that the presence of asymmetric information typically causes a decrease in the average selling price of homes, as buyers request a premium to compensate for the uncertainty. By providing more complete, accurate, and timely property-level information to prospective buyers, flood risk disclosure laws separate dwellings based on their idiosyncratic risk. This decrease in information frictions should reduce both the risk aversion premium and the adverse selection in the housing markets covered by the laws. Studies (e.g., Botosan 1997 ; Nanda and Ross 2012 ) have shown that reducing uncertainty via higher transparency increases average asset values. Thus, the decrease in information frictions should raise prices on average.

In the absence of flood risk disclosure laws, real estate lenders may insufficiently incorporate the idiosyncratic (cross-sectional) risks of flooding for each property due to (1) the lack of precise, cheaper information and (2) the availability of misinformation (Cleetus 2019 ). The adverse selection and high uncertainty result in a market for lemons in climate risk, where lenders protect themselves by pricing climate risk into their credit supply (Nguyen et al. 2022 ; Ouazad and Kahn 2022 ). Additionally, lenders’ inability to identify low flood-risk properties weakens the sellers’ incentives to take measures that mitigate the risk of damage (Jin and Leslie 2003 ). By decreasing information frictions at the dwelling level, flood risk disclosure laws reduce lenders’ uncertainty about the value of housing collateral, lowering credit risk and facilitating mortgage lending. The literature (Di Maggio et al. 2017 ; Christiano et al. 2014 ) provides both theoretical and empirical support for the idea that high uncertainty leads to lower housing prices and a reduced credit supply (a risk premium). Sinha ( 2022 ) shows that laws improving the corporate disclosure of pollution risk around oil and gas facilities increase mortgage lending nearby.

However, the prediction that flood risk disclosure laws reduce the lemon discount in the housing market assumes that sellers comply and that buyers and lenders use these mandated disclosures. Sellers might not comply if Federal Emergency Management Agency’s location-based information is sufficient, as they incur additional costs in collecting, compiling, and verifying the information. Institutional lenders, which process many mortgage applications daily and have advanced risk-assessment systems (Nguyen et al. 2022 ), may be less likely to rely on flood risk disclosure laws. Unsophisticated agents might also neglect to incorporate climate risk information due to limited attention (Baldauf et al. 2020 ).

Data and descriptive statistics

4.1 Data and sample selection

Residential real estate prices come from the Federal Housing Finance Agency (Chaney et al. 2012 ; Favara and Imbs 2015 ). Footnote 9 The agency provides a home price index, which broadly measures the movement of single-family home prices in the United States. Because of the breadth of its sample, the index provides more information than other house price indices (Chaney et al. 2012 ). It has been available at the census tract level since 1975, and I use the census tract-level real estate prices as a proxy for the value of traded properties. I obtain local housing supply elasticities data, which are available for 95 metropolitan statistical areas (MSAs), from Saiz ( 2010 ).

The mortgage data come from a comprehensive sample of mortgage applications and originations collected by the Federal Financial Institutions Examinations Council under the provisions of the Home Mortgage Disclosure Act. Under these provisions, the vast majority of mortgage lenders are required to report (Dagher and Kazimov 2015 ). The data, which are available at the loan application level, record information on the year of application (data are available on an annual frequency), the amount of the loan, the lender’s decision, the characteristics of the applicant (including income, race, and gender), and the location of the property, down to the census tract level. The data became available in the early 1990 s, and my sample spans from 1990 to 2017.

My original sample is 500 million observations. After I restrict the sample to institutions that report to either the Office of the Comptroller of the Currency, the Federal Reserve System, or the Federal Deposit Insurance Corporation, the number of observations is reduced to 211 million. I further restrict the sample to mortgage applications made for owner-occupied home purchases of conventional single-family properties (Dagher and Kazimov 2015 ). To minimize noise, I follow Dell’Ariccia et al. ( 2012 ) and Dagher and Kazimov ( 2015 ) and drop applications below $25,000 and above $1 million, as well as those missing income and census-tract data. I aggregate the data up to the census tract level and obtain demographic attributes, such as population, race, gender, and house units from the U.S. Census Bureau (Mian and Sufi 2009 ). Following Cengiz et al. ( 2019 ) and Baker et al. ( 2022 ), I employ a stacked difference-in-difference research design for all specifications. My primary stacked sample includes 286,866 census tract-year observations.

I collect information on the implementation years and the contents of the disclosure regulations from state law codes and the Natural Resources Defense Council. I identify the aggregate census tract-level exposure to flood risk within my sample using the National Oceanic and Atmospheric Administration sea level rise calculator (Marcy et al. 2011 ; Bernstein et al. 2019 ). It provides detailed sea level rise shapefiles that describe the latitudes and longitudes that will be inundated following a one- to six-foot increase in average global ocean level. I define District Exposure as an indicator equal to one if six feet or less of sea level rise would put the census tract area underwater (Marcy et al. 2011 ; Bernstein et al. 2019 ). I use the geographic mapping software ArcGIS Pro to assess the exposure level of each census tract within a coastal state that has flood risk disclosure mandates. Figure 2 Panel A provides the map. The exposed census tracts are concentrated in the Gulf region, Washington State, and along the Eastern Seaboard. Panel B displays a more precise representation of district exposure in California and Louisiana.

4.2 Descriptive statistics

I define the variables in Appendix A . In particular, Housing prices is the natural log of one plus the home price index. Loan size is the natural log of one plus the average amount of mortgage loans originated (in thousands of dollars) in census tract i’s year t (Favara and Imbs 2015 ). Loan-to-value is the natural log of one plus loan value divided by house price index instrumented by local housing supply elasticity (Saiz 2010 ). I control for time-variant county-level and census tract-level demographic attributes, including population, education, ethnicity, applicant income, gender, the number of housing units, and households.

Table 3 Panel A presents the descriptive statistics for the main variables for my sample. Of the census tract-year observations, 51% (49%) are after (before) the implementation of flood risk disclosure laws. The mean Housing prices is 5.342, the average Loan size is 5.395, and the average Loan-to-value ratio is 0.692. The gender distribution within a census tract is balanced. About 16.1% of mortgages are in districts exposed to high flood risk. These statistics are all broadly consistent with those reported in the literature (Nguyen et al. 2022 ; Sastry 2022 ). The mean and median values of the different variables are reasonably close. Panel B of Table 3 displays univariate analyses before and after the flood risk disclosure laws, comparing the High Exposure and Low Exposure districts. It indicates price protection (lemon discount) (median values: 5.090 versus 5.378) in high-exposure districts, with the change in premium being greater than in low-exposure districts (0.288 versus 0.221). Footnote 10 In addition, there is an aggregate exposure discount (5.090 versus 5.149) before the flood risk disclosure laws.

Research design

I in-difference approach to examine how flood risk disclosure laws affect housing prices and mortgages at the census tract level. My primary analyses follow the literature (Cengiz et al. 2019 ; Baker et al. 2022 ) and use the following specification:

(1a) is my baseline model, where my variable of interest is FRDL , an indicator that equals one if census tract i’s year t is after state j has enacted a flood risk disclosure law and zero otherwise. I include census tract-cohort fixed effects and year-cohort fixed effects. β 1 essentially captures a stacked difference-in-difference estimator. I cluster standard errors by state-cohort of census tracts because FRDL is a state-level variable (Klasa et al. 2018 ; Barrios 2021 ). To control for geographic heterogeneity, I add census tract-level and county-level time-varying demographic attributes, such as population, education, ethnicity, applicants’ income, housing units, and households.

In areas with high aggregate flood exposure, investors face greater uncertainty regarding property-specific flood risk. Investors in low-exposure regions should have a smaller flood-related information disadvantage and therefore respond less to mandated disclosures. I thus expect the price effects to be larger in markets with greater aggregate flood exposure. To control for state-level housing market trends and time-varying state characteristics, I estimate model (1b) by including state-year-cohort fixed effects, absorbing FRDL. The variation arises from both the timing of the laws and time-varying heterogeneity in district-level flood-risk exposure. I essentially compare average changes in housing prices and mortgages across markets with high versus low exposure within the same state at the same time.

Results discussions

6.1 Flood risk disclosure laws and housing prices

Table 4 reports my baseline results, with both models estimated using a stacked regression design. Both columns include census tract-cohort fixed effects. Column 1 includes year-cohort fixed effects, while column 2 adds state-year-cohort fixed effects to control for housing market trends. The state-year-cohort fixed effects absorb the main effects of flood risk disclosure laws. Footnote 11 The coefficient for FRDL in column 1 is positively significant at the 1% level, suggesting a lower risk premium associated with information frictions after the laws. A positive and significant coefficient of FRDL × District Exposure suggests that property-level disclosures remain informative, even in areas where aggregate flood risk is known to be high, consistent with flood risk disclosure laws adding incremental dwelling-specific information beyond location-based measures.

In terms of economic magnitude, the regulations alone increase the affected census tracts’ house price index by close to 7.6%. To put this into perspective, Di Maggio et al. ( 2017 ) find that a one standard-deviation decrease in local economic uncertainty leads to a 9.9% rise in real estate transactions and leverage, illustrating the strong link between uncertainty and market activity. Additionally, the housing literature (e.g., Nanda and Ross 2012 ; Eerola and Lyytikäinen 2015 ) indicates that improved information in the housing market about property conditions and past transactions can reduce risk premiums by approximately 6.4%. Overall, these findings suggest that my estimates are economically meaningful and align with the broader understanding of how uncertainty influences housing markets.

In the housing market, the price signal serves as the most direct indicator of market information asymmetry (Nanda and Ross 2012 ), which differs from in the equity market. There are two main reasons for this. First, an average household completes only a few property transactions over its lifetime. Each price quotation must communicate information, making the price signal immediately responsive even when trading is sparse. In contrast, equity markets trade continuously, so liquidity consistently reflects how quickly information is incorporated into prices. Second, the value of houses that is used as collateral, being directly connected to credit risk, is more critical for banks’ lending decisions than the volume of house transactions (Nguyen et al. 2022 ; Kang et al. 2021 ).

Although the above patterns are well established, I conduct additional tests to illuminate the liquidity effects. First, using Zillow’s house sales data, as shown in Table OA5, I find that flooding risk disclosures reduce uncertainty for real estate market participants (buyers), which in turn leads to increased sales volumes. Footnote 12 Second, I follow Cho et al. ( 2024 ) and compute county-level housing price standard deviations. As shown in Table OA6, both univariate analyses and full-sample regressions confirm that flood risk disclosure laws lead to a significant increase in housing price variances. In other words, once disclosure separates high‐ from low‐risk homes, transaction prices spread out more, even as average values rise. I refrain from making strong welfare claims for flood risk disclosure laws and simply note that they raise both mean prices and price dispersion, leaving any net welfare impact an open question.

6.2 Assumptions and robustness tests

In this section, I test several key assumptions of the difference-in-differences design. First, my setting exploits the staggered adoption of flood risk disclosure laws and uses the time-varying attribute of the treatment to identify a counterfactual distribution—a common approach used in empirical work in economics. However, this approach has specific weight issues and involves the use of previously treated units as controls. I address these issues by using a stacked regression design throughout the paper (Cengiz et al. 2019 ).

Second, I mitigate the influence of unobservable state time-varying factors that correlate with the timing of flood risk disclosure laws’ adoption and the housing market. To control for the general housing market trend, I employ a triple difference-in-differences design with cohort-state-year fixed effects. In Table 4 Panel A column 2, FRDL × District Exposure is statistically significant at the 1% level, indicating that the increase in house prices induced by flood risk disclosure laws strengthens when the local market is more exposed to flood risk uncertainty. These results show that my findings are not caused by state time trends and are consistent with the notion that separating the pooling equilibrium has a greater effect in areas where the aggregate flood risk is known to be high (e.g., in coastal districts).

Third, I assume that the average outcomes among the treated and comparison populations would have followed parallel trends in the absence of the treatment. To better understand the dynamics of my findings, I create a series of event-year indicator variables, FRDLs , that are indexed from t-3 to t + 3 , with t = 0 being the year when a flood risk disclosure law takes effect. The variables take the value of one if a law takes effect within t-i and zero otherwise. I find that the coefficients on t-3 , t-2 , and t-1 are all statistically insignificant, have inconsistent signs or have a small magnitude. (The p-value of an F-test of joint significance is 0.697.) In contrast, the coefficients on t , t + 1 , t + 2 , and t + 3 are all consistently positive and statistically significant. (The p-value of an F-test of joint significance is 0.00.) Figure 3 , which plots these patterns, shows no evidence of a differential pre-trend. Importantly, this graphic further suggests that the reported effects are not limited to the short run. Thus, the documented effects are unlikely to be explained by the macroeconomic factors and potential reversals when house prices rebound from temporary drops.

Fourth, to alleviate concerns about potential anticipation effects of law adoption, I examine whether there is any discernible spillover to neighboring states. The stable unit treatment value assumption requires that the response of a particular unit depends only on the treatment to which the unit was assigned, not the treatments of others around it. But one might expect that, when a neighboring state enacts a flood risk disclosure law, firms in bordering states begin to respond in anticipation of similar legislation being adopted in their own states. To test this, I construct a dummy variable, FRDL_neighbor , which is set to one in a focal state’s census tracts after a neighboring state enacts flood risk disclosure laws. The results in Table 4 Panel B column 1 indicate no significant spillover effects to the focal states’ census tracts when neighboring states pass similar laws. My results remain robust after dropping neighboring counties (Table 4 Panel B column 2). Additionally, I exclude observations from the year preceding the law’s adoption to account for potential anticipation effects (Table 4 Panel B column 3). However, note that, if real estate prices increase in an area for exogenous reasons, the prices in surrounding areas are also likely to be affected by a shift in demand. As such, I do not expect the laws’ effects to be limited to the areas directly affected by flood risk, and I examine all of the districts in states covered by the disclosure laws. Footnote 13

Fifth, to further alleviate concerns about unspecified omitted variables, I use a series of machine-generated random placebo state adoption dates that do not overlap with the actual adoption dates (Cornaggia et al. 2015 ). As discussed in Sect. 2.2 , to strengthen identification, I intentionally focus on the laws’ timing rather than their passage in my main tests (i.e., when, not if). I expect that the placebo dates, unlike the actual events, will not affect house prices. Indeed, Table 4 Panel B column 4 suggests that the placebo law adoption has no significant effects on house prices. In an additional robustness test, I expand my sample to include both the variation in timing and the adoption status. Table 4 Panel B column 5 shows the results, including those in non-adoption coastal states. My findings remain consistent.

Finally, to further alleviate the concern that my results are driven by unobservable local factors in the same tract-year as my treatment, I follow Dube et al. ( 2010 ) and compare the outcomes in adjacent counties in different states that are differently affected by the regulation change. The assumption is that the counties, being adjacent, are very similar in observable aspects. More importantly, they may share unobservable aspects that are difficult to control for. Indeed, research (Huang 2008 ; Fan 2025 ) indicates that contiguous border counties offer significantly better control groups than randomly selected counties or counties chosen using propensity score matching. Specifically, I estimate the impact of flood risk disclosure laws on house prices within pairs of adjacent counties on opposite sides of state borders. I identify these pairs using the U.S. Census Bureau County Adjacent file and then merge this dataset with my main sample, resulting in 81,740 census tract-year observations. Table 5 shows that, in both models, flood risk disclosure laws significantly raise house prices in the counties subject to them. This supports that my main results are not driven by unobserved local macroeconomic conditions that coincide with my treatment.

6.3 Endogeneity discussions and limitations

The main endogeneity concern about this paper is whether unobservable state time-varying factors influence both the timing of flood risk disclosure law adoption and housing market outcomes. To mitigate this concern, I implemented the various robustness tests above (e.g., a triple difference-in-difference, a contiguous border–county analysis). Nevertheless, to characterize the likelihood that a correlated omitted variable explains the relationship between housing prices and flood risk disclosure laws, I follow the procedure described by Cinelli and Hazlett ( 2020 ). I find that the effect of a correlated omitted variable would have to be implausibly large to lower the laws’ estimated effect to zero. Specifically, Table OA7 in the online appendix shows that FRDL remains significant, even if I allow for the presence of an unspecified correlated omitted variable that is three times as strong as GDP. The t-value of 13.86 exceeds the adjusted critical t-value of 13.26, and only unobserved confounders (orthogonal to the covariates) that explain more than 2.89% of the residual variance of both the treatment and the outcome are strong enough to bring the point estimate of FRDL to 0. GDP, in comparison, explains only 0.12% and 0.13% of the residual variance of the treatment and outcome, respectively. This result, along with the other procedures I conducted, suggests that the likelihood of an unspecified correlated omitted variable explaining my results is low. A hypothetical omitted variable would have to be both unconditionally and conditionally correlated with flood risk disclosure laws to affect my results. Essentially, if states passed laws that increased housing prices at the same time as they initiated flood risk disclosure laws, then the other laws would also need to disproportionately affect counties with high flood-risk exposure to explain my results.

Despite the robustness checks discussed above, several caveats temper my causal claims. First, I cannot rule out that states enact flood-risk disclosure laws for non-exogenous reasons. Second, other contemporaneous regulations could confound the results (although this is unlikely, as they would have to systematically overlap with the flood risk disclosure laws’ timelines across different states). Third, the border-county design relies on the assumption that adjacent counties have similar regulatory environments, which may not always hold.

6.4 Cross-sectional results

6.4.1 Cross-state heterogeneity in disclosure quality requirements

Studies suggest that the design of regulations can significantly impact their effectiveness and externalities (e.g., Meyer and Rowan 1977 ; Fung et al. 2007 ). This suggests that different states’ flood risk disclosure laws, which vary in stringency and specificity, will have differential impacts on the quality of disclosure. To test this, I consider two measures of stringency. First, I construct NRDC Quality , a ranked variable from 0 to 4 (representing grades A to F, with A implying the highest disclosure quality), to measure the quality of flood risk disclosure laws, based on grading conducted by the Natural Resources Defense Council 2019 . Footnote 14 The organization grades the quality of states’ flood risk disclosure laws on a scale of A to F. States with a grade of F have no statutory or regulatory requirements for seller disclosure of a property’s flood risks or past flood damages to potential buyers. Building on the methodology, I also create a self-constructed law index that measures the quality of disclosure laws based on the quantity of dimensions the laws address. FRD Quantity is constructed based on nine flood risk disclosure laws disclosure dimensions: flood damage to land, flood damage to property, flood insurance, flood zone classification, flooding problems ever, drainage ever, federal disaster aid, additional form, and measures to remedy the problem. I assign each of the above dimensions a value of one or zero and sum the nine indicator variables. Table 6 presents the results. The findings suggest that investors react more strongly to flood risk disclosure laws when they require higher-quality and more multidimensional disclosure about specific properties. This suggests that, if federal regulators attempt to design a national flood risk disclosure law, they should carefully consider the effects of heterogeneous flood risk disclosure laws in different states.

6.4.2 Within-state heterogeneity in information frictions

In contrast to Sect. 6.4.1 ’s focus on cross-state differences in statutory design, the next analysis uses within-state variation to illuminate the mechanism that allows me to include cohort-state-year fixed effects. In particular, I test whether flood risk disclosure laws matter more where local information frictions rise and alternative information sources weaken. As discussed in Sect. 2.1 , information frictions arise when homebuyers face difficulties in acquiring, processing, or verifying property-specific flood-risk information, leading to greater uncertainty and price discounts. To capture variation in local information frictions, I construct a county-level measure of local media presence using television and radio station listings data from the Federal Communications Commission. Research has relied on the commission’s broadcast license databases to identify local television and radio outlets (e.g., Berry and Waldfogel 2001 ). For each county-year, I count the total number of active media outlets (e.g., television and radio stations). The literature uses media presence to proxy for the strength of information dissemination and the efficiency of local information environments (e.g., Engelberg and Parsons 2011 ). Areas with fewer media outlets are likely to have higher information frictions.

I define LowMedia as an indicator equal to one if the total number of media outlets in county i in year t is below the sample median and zero otherwise. Table 7 column 1 reports the results. I find that, in the absence of flood risk disclosure laws, house prices in low-media counties exhibit significantly larger buyer price discounts: the coefficient on LowMedia is negative and statistically significant at the 5% level. This result is consistent with buyers demanding compensation for higher uncertainty when the local information environment weakens. I then examine whether flood risk disclosure laws mitigate these information frictions by interacting with LowMedia and the FRDL indicator. The coefficient on FRDL × LowMedia is positive and statistically significant at the 1% level, indicating that the price effects of flood risk disclosure laws strengthen in regions with fewer media outlets. Economically, this suggests that the laws substitute for weak local information environments by providing detailed property-specific flood-risk disclosures directly from sellers. Importantly, the sum of the coefficients on LowMedia and FRDL × LowMedia (P-value is 0.617) is not statistically different from zero, implying that flood risk disclosure laws reduce the price discount associated with high information frictions. Together, these results indicate that the laws operate through an information channel to improve price efficiency, with greater improvements in markets where alternative information sources are limited.

6.4.3 Within-state heterogeneity in climate change attention

Beyond the formal information environment captured by media presence, local attention to flood risk may also vary with the salience of nearby climate-related disasters, even when the focal properties are not directly affected. Research shows that exposure to disasters in neighboring regions heightens awareness of climate risk and influences economic behavior through attention and salience rather than physical damage alone (Bernstein et al. 2019 ; Baldauf et al. 2020 ; Engle et al. 2020 ; Ilhan 2020 ). Disasters may increase the relevance of flood risk information and amplify the role of disclosure in pricing decisions.

To capture variation in flood-risk attention driven by nearby disasters, I identify major flood-related disaster declarations in neighboring counties using the Federal Emergency Management Agency Presidential Disaster Declaration database. I create NeighborDisaster , an indicator equal to one if at least one adjacent county experienced a major flood-related disaster during year t and zero otherwise. Importantly, this measure captures heightened attention to flood risk without any confounding effects from direct local damage. The results in Table 7 column 2 show that the interaction term FRDL × NeighborDisaster is positive and statistically significant at the 1% level, indicating that the price effects of flood risk disclosure laws strengthen when disasters have recently occurred in nearby regions. Thus, even in the absence of direct exposure, buyers’ heightened attention to flood risk appears to increase their responsiveness to property-specific flood disclosures, leading to larger price adjustments when such information becomes available. Overall, these results suggest that attention triggered by nearby disasters reinforces the role of disclosure in mitigating information frictions and shaping how climate-related risks are capitalized into housing prices.

6.4.4 Within-state heterogeneity in climate change beliefs

The effectiveness of flood risk disclosure laws may depend on local beliefs about climate change and flood risk. Studies show that market participants’ climate risk beliefs influence real estate prices and financial decisions (Bernstein et al. 2019 ; Baldauf et al. 2020 ; Engle et al. 2020 ). Households that underestimate or dismiss climate-related risks may discount or ignore the information in mandated disclosures, making the disclosures less effective.

To examine this possibility, I analyze heterogeneity in the flood risk disclosure laws’ effects using a time-varying, county-level measure of climate change concern from the Yale Climate Opinion Maps (Murfin and Spiegel 2020 ). This measure captures the share of residents who do not believe that global climate change is occurring and reflects local skepticism toward climate risks. Disbelief is the percentage of residents who are not at all worried about global climate change in county c in year t. The results in Table 7 column 3 show a significantly negative coefficient on the interaction term FRDL × Disbelief , indicating that the positive price effects of flood risk disclosure laws weaken in counties where disbelief in climate change prevails. These findings imply that simply expanding the information set through disclosure mandates does not fully resolve information frictions when market participants believe less in the underlying risk. Taken together, the results indicate that beliefs play an important role in shaping how flood risk disclosures are reflected in housing prices. Even when disclosures provide standardized and credible information, their effectiveness depends on whether buyers perceive flood and climate risks as relevant. This finding highlights an important limitation of disclosure-based regulation in the presence of heterogeneous or weak beliefs.

6.5 Credit expansion

The cross-sectional evidence above supports an information channel. I next examine whether the flood risk disclosure law-related improvement in collateral transparency affects mortgage market outcomes by changing lenders’ willingness to supply credit, which can amplify housing price responses. By reducing uncertainty about flood risk and collateral values, mandated disclosures could improve lenders’ assessments of mortgage risk, facilitating loan origination and generating a positive feedback loop between credit supply and housing prices. Consistent with this mechanism, Favara and Imbs ( 2015 ) show that expansions in mortgage credit significantly increase housing demand and prices. To examine whether flood risk disclosure laws influence lenders’ decision-making, I analyze how reductions in collateral uncertainty affect mortgage lending. If the flood risk disclosure law-induced price effects arise solely from increased buyer demand due to lower information frictions, then the effects should be concentrated among financially unconstrained buyers. In contrast, if the price effects arise because lenders respond to improved collateral transparency by expanding credit, then the effects should strengthen among financially constrained households, which benefit more from enhanced lending. To test this, I obtain county-level housing cost burden data from the Harvard Joint Center for Housing Studies. Pre Burden is defined as the average share of households facing housing cost burdens in county c prior to the enactment of the disclosure law.

Housing prices in the United States exhibit substantial geographic heterogeneity, driven by variation in housing supply elasticities, land-use regulations, and geographic constraints (Gyourko and Saiz 2006 ; Saiz 2010 ). As discussed in Sect. 6.3 , these factors may correlate with both the adoption of flood risk disclosure laws and housing market outcomes, potentially threatening identification. To address this concern, I follow Chaney et al. ( 2012 ) and Favara and Imbs ( 2015 ) and instrument census tract-level housing prices using local housing supply elasticities. I then use the instrumented prices to compute census tract-level loan-to-value ratios, which serve as a measure of credit supply. Loan-to-value ratios are widely used indicators of mortgage risk in the United States (Bian et al. 2018 ), and lenders adjust them to manage collateral risk, particularly in flood-prone areas (Sastry 2022 ).

Table 8 Panel A presents stacked regression estimates in which I interact FRDL with Pre Burden and examine outcomes including housing prices, loan size, loan-to-value ratios, and loan denials. All specifications include cohort-state-year fixed effects to control for time-varying housing market conditions and concurrent regulatory changes. The FRDL indicator is absorbed by cohort-state-year fixed effects, and the Pre Burden indicator is absorbed by cohort-census tract fixed effects in this section of analyses. The coefficients on FRDL × Pre Burden are positive and statistically significant at the 1% level across columns (1) through (3). These results indicate that, following the adoption of flood risk disclosure laws, lenders extend larger loans and allow borrowers to finance a greater share of property values, particularly among financially constrained households. Consistent with Sastry ( 2022 ), the findings suggest that reducing uncertainty about collateral values mitigates adverse selection and encourages lenders to increase loan-to-value ratios. I also observe a significant decline in mortgage denials among constrained buyers (Table 8 Panel A column 4)—further evidence that lenders adjust their lending behavior in response to improved information. These results imply that the disclosure laws-induced reduction in collateral uncertainty disproportionately benefits borrowers facing financial constraints, leading to credit expansion and amplifying the increase in housing prices. This evidence supports a credit supply mechanism and complements the main findings.

Next, it would be interesting to know which characteristics of lenders contribute to frictions in the housing market. I therefore examine whether certain bank characteristics affect flood risk disclosure laws’ impact on the credit supply. Since banks have different exposures to and abilities to incorporate flood risk information before the laws’ enactments, I expect heterogeneity across lenders. In particular, I focus on banks’ size and profitability. The impact of banks’ sophistication on their use of flood risk disclosure laws is unclear, since larger ones have more resources and more sophisticated models to price climate change risk (Stiroh 2020 ) but smaller ones have superior local knowledge (Garmaise and Moskowitz 2004 ; Lim and Nguyen 2021 ). I construct two measures: Pre Size , the bank’s mean amount of real estate loans scaled by total assets before the disclosure law’s enactment, and Pre NI , the bank’s mean net interest income scaled by total operating income before the disclosure law’s enactment. The results in Table 8 Panel B indicate that flood risk disclosure laws’ effects on lenders strengthen when the lenders are smaller and less profitable before the laws’ enactment. The findings suggest that the impact of the disclosure regulation depends on investors’ sophistication and the costs of acquiring the flood risk information. In general, big lenders with more expertise benefit less from the mandated disclosures.

Finally, if flood risk disclosure laws, by reducing uncertainty about collateral values, encourage banks to expand lending, this may also change banks’ risk-taking incentives. I hence test whether the credit expansion associated with the laws comes with trade-offs for banks’ profitability. Following Nguyen et al. ( 2022 ), I collect banks’ balance sheets and financial performance information from Call Reports. Since Home Mortgage Disclosure Act data do not track loan performance post origination (Davis et al. 2023 ; Saadi 2020 ), I use two methods to tackle this issue. First, I use the rate spread variable recorded in the database to measure the mortgage risk and implied expected default rate. Second, I check whether institutions that offer high loan-to-value ratios to financially constrained buyers are associated with high default rates (Table OA8). Footnote 15 In Table 8 Panel C, I find that the high-default loans are associated with lower profitability after the adoption of flood risk disclosure laws, as the coefficients of the triple interaction term FRDL × Pre Burden × Expected Default are negatively significant at 5% in column 1 and 10% in column 2. This finding aligns with those of Di Maggio et al. ( 2017 ), who notes that a decrease in uncertainty shifts lending from safer borrowers to financially constrained ones, resulting in increased risk-taking incentives. These loans may increase default rates and negatively affect banks’ balance sheets. Footnote 16

6.6 Additional analyses

6.6.1 Flood insurance

Lenders may have flood risk information under the National Flood Insurance Program since the program requires transfer disclosure and flood insurance purchases for houses in Special Flood Hazard Areas. This suggests that, for those houses, lenders might be less likely to rely on the information in flood risk disclosure laws. In reality, however, many mortgage companies do not undertake the required flood determinations until they sell to federally regulated institutions (Nguyen et al. 2022 ). Even with the insurance requirement, many property owners purchase policies at the time of financing but drop them later, as lenders often fail to ensure that insurance is maintained (Pope 2008 ). The literature also suggests that it is difficult for lenders to sanction or penalize violations of flood insurance purchases (Nguyen et al. 2022 ). At the same time, a large proportion of areas exposed to flood risk are not included in Special Flood Hazard Areas and thus are not covered by flood insurance requirements (Pralle 2019 ). All of this raises the question of whether the disclosure regulation’s effect depends on whether flood insurance is mandatory. To test this, I follow Ouazad and Kahn ( 2022 ) and use Special Flood Hazard Areas to identify where flood insurance is required. I create the indicator SFHA , which is equal to one if flood insurance for agency-backed mortgages in the areas is required and zero otherwise. Footnote 17 Table OA9 shows that the coefficients of FRDL × SFHA are not statistically significant, suggesting that flood insurance requirements do not change the influence of flood risk disclosure laws. In other words, national flood insurance does not appear to substitute for the effects of mandatory disclosures.

6.6.2 Securitization

I next investigate the effects of passing flood risks to government sponsored enterprises, such as Fannie Mae and Freddie Mac (Nguyen et al. 2022 ; Ouazad and Kahn 2022 ). My premise is that lenders are more likely to rely on seller-disclosed information if the risk cannot be sold to government sponsored enterprises, and that the information premium should vary depending on whether the loan is eligible for securitization. When loans do not meet securitization requirements, lenders, who tend to be risk averse, directly bear the costs of default. I therefore expect the positive effects of flood risk disclosure laws to strengthen for non-government sponsored enterprise loans. To test this prediction, I use county-level, time-varying thresholds on loan limits set by Fannie Mae and Freddie Mac and define jumbo loans as those with a loan amount greater than the county-level conforming loan limit. Consistent with my credit expansion explanation, the results (Table OA10) indicate that flood risk disclosure laws’ housing price effects do indeed strengthen for government sponsored enterprise-ineligible loans and that lenders also tend to charge lower interest rates for those loans following the adoption of flood risk disclosure laws.

6.6.3 Alternative explanations

In this section, I attempt to address alternative explanations. The first concern relates to federal disaster support. The alternative explanation is that the observed house-price and loan-to-value increases stem from uneven federal aid distribution, which lowers credit risk for wealthier families. However, my results hold even when I restrict the sample to properties that did not receive any federal disaster aid and that lie outside the National Flood Insurance Program (Table OA11). This suggests that unequal aid distribution alone cannot drive the observed price and loan-to-value increases.

A second concern is that banks’ diversification laws may impact my findings. To address this, I utilize my bank loan-tract-year panel data and construct a diversification measure by counting the number of distinct census tracts in which each bank originates loans. I re-estimate my baseline regressions while controlling for this diversification and then split the sample into more diversified and less diversified banks. While geographic diversification slightly attenuates the FRDL coefficient for the more diversified banks, FRDL remains both positive and highly significant across both subsamples (Table OA12). This pattern shows that, although diversification partially moderates the effect, it cannot fully explain my findings. FRDL continues to exert a robust influence on lending behaviors, regardless of banks’ geographic footprints.

I conduct additional tests to address a third concern: politically driven lending. First, to see whether early-2000s political-lending initiatives account for the observed effects, I restrict the sample to loans originated before 2001. This analysis leaves my core flood risk disclosure law results largely robust (Table OA13). Second, political pressure would lead to a larger effect on government sponsored enterprise-eligible loans, which small banks find safer and easier to sell. Section 6.6.2 shows that flood risk disclosure laws’ price effects are somewhat more pronounced for ineligible loans. Moreover, I find that banks reduce lending to financially constrained borrowers once the flood risk materializes and information improves (Basu et al. 2022 ) (Table OA14). This counters the political-lending hypothesis, under which banks would be expected to expand credit to minority or low-income households.

Conclusion

This paper examines the effects of mandatory climate risk disclosure on housing prices and mortgage supply. Using stacked difference-in-differences and continuous border designs, I find that house prices increase following the adoption of flood risk disclosure laws, consistent with a risk premium decline due to reduced information asymmetry. The effect strengthens in markets with strict disclosure requirements, more information frictions, and greater attention to climate risks. Households’ weak beliefs in climate change attenuate the effectiveness of the laws. Importantly, less sophisticated lenders react to the laws by rejecting fewer loans and issuing larger loans with higher loan-to-value ratios to cost-burdened buyers. These lenders, however, see lower profitability. My study responds to the call for research on disclosure mandates outside traditional corporate settings (Leuz and Wysocki 2016 ) and opens the discussion on the costs and benefits of climate risk disclosure mandates (Christensen et al. 2021 ; Downar et al. 2021 ). By reducing information uncertainty, flood risk disclosure laws facilitate price efficiency and encourage credit expansion to disadvantaged borrowers (Di Maggio et al. 2017 ). However, this expansion comes at the expense of lower profits, and lenders may selectively tighten credit when they anticipate a high likelihood of disaster-related defaults, which could unintentionally worsen inequalities in credit access among vulnerable populations.

I acknowledge that applying results from the housing markets to the financial markets to the financial markets raises external validity concerns, despite the presence of similar fundamental forces operating in both markets (e.g., information asymmetry and moral hazard). Similar concerns are noted by Cho et al. ( 2024 ), who use the used-car market to understand the role of information credibility at the expense of generalizability. I also acknowledge that I cannot fully disentangle demand effects from supply effects. Instead, I highlight the existence of both and suggest that they are driven by a common information channel.

Notes

Two notable exceptions are the work of Christensen et al. ( 2017 ) and Tomar ( 2023 ). However, these studies target heavily polluting, highly regulated industries.

https://www.reuters.com/legal/government/us-appeals-court-hits-pause-challenges-sec-climate-rule-2025-09-12/

See Kielty et al. ( 2023 ), Sinha ( 2022 ), Fan ( 2025 ), and Kang et al. ( 2021 ).

Troy and Romm ( 2004 ) focus only on one state. I consider multiple states and compile a comprehensive dataset coding both the laws’ timing and disclosure content. The time-varying nature of these events allows me to isolate the effects of macroeconomic trends and offers a more robust setting in which to test the role of transparency in climate risk within the housing market.

Unlike other latent characteristics, flooding risk can be measured and cannot be fully insured, especially since the value of time and emotional costs related to recovery cannot be insured (Turnbull et al. 2013 ). The information on flooding insurance under flood risk disclosure laws is particularly relevant for potential buyers.

The market penetration rate of flood insurance outside National Flood Insurance Program is only about 1%.

They have problems learning the information and drop the offer since they face intense pressure from the other involved parties (real estate agents, sellers, mortgage lenders, etc.), who have financial incentives to ensure that the transaction is completed soon (Pope 2008 ).

I report the full length of survey questions in Online Appendix Table OA2 . Some additional anecdotal evidence is in Appendix C .

Previously known as the Office of Federal Housing Enterprise Oversight.

High Exposure district increases from 5.090 to 5.378, and Low Exposure district rises from 5.149 to 5.370.

To assess the incremental explanatory power of the flood risk disclosure laws, I follow Breuer and DeHaan ( 2024 ) and Jennings et al. ( 2024 ) and report both the adjusted R 2 and the within R 2 , which isolates the portion of the variation explained by my treatment, net of fixed effects. Flood risk disclosure laws plus my control variables account for 9% of the remaining variation. By contrast, Jennings et al. ( 2024 ) find that state noncompete laws and controls explain only 4% beyond firm and year fixed effects. These comparisons confirm that flood risk disclosure laws contribute meaningfully to the explained variance in my setting.

Zillow house sales data cover a shorter period (from 2008) and fewer geographic areas (only MSAs). For a balanced sample, I include sales data from only three states (i.e., those where the flood risk disclosure law mandate was in place after 2008), and they are available only at the MSA level. These factors significantly reduce my sample size.

My analysis focuses exclusively on coastal flooding because it contributes the most to banks’ physical climate risk and remains hard to predict, which heightens coastal states’ information uncertainty relative to that of noncoastal ones (Nguyen et al. 2022 ).

This is a nonprofit international environmental organization, and I report the details of its methodology in Appendix D . I complement its measure using a self-constructed index.

Following Saadi 2020 , I collect Freddie Mac’s single-family mortgage performance data to create a census tract-level default rate based on loan delinquency (based on mortgages at least 90 days overdue). These data track mortgage performance but are limited to high-quality Government Sponsored Enterprises loans. They are the only publicly available data tracking mortgage performance. Because they only cover a few high-quality loans, I do not use them in my main analyses (Adelino et al. 2016 ).

Both Pre Burden and Expected Default measures are continuous variables; hence, I run a three-way interaction. The results should be interpreted with caution, as the interest rate data are available only from 2004. I focus on regulatory events after 2004 to ensure balanced pre- and post-periods.

Flood Zones A, AO, A1-A30, AE, A99, AH, AR, AR/A, AR/AE, AR/AH, AR/AO, AR/A1-A30, V1-V30, VE or V. See https://www.fema.gov/glossary/special-flood-hazard-area-sfha .

https://grist.org/article/your-future-home-could-be-in-a-flood-zone-and-no-ones-required-to-tell-you

https://www.nrdc.org/bio/joel-scata/undisclosed-flood-damage-financially-soaks-home-buyers

https://www.nrdc.org/bio/joel-scata/flooding-can-put-unsuspecting-home-buyers-financially-underwater

References

Adelino, M., A. Schoar, and F. Severino. 2016. Loan originations and defaults in the mortgage crisis: The role of the middle class. The Review of Financial Studies 29:1635–1670.

Article Google Scholar

Akerlof, G. A. 1970. The market for lemons: Quality uncertainty and the market mechanism. The Quarterly Journal of Economics 84:488–500.

Article Google Scholar

Baker, A. C., D. F. Larcker, and C. C. Y. Wang. 2022. How much should we trust staggered difference-in-differences estimates? Journal of Financial Economics 144:370–395.

Article Google Scholar

Bakkensen, L. A., and L. Barrage. 2022. Going underwater? Flood risk belief heterogeneity and coastal home price dynamics. Review of Financial Studies 35:3666–3709.

Article Google Scholar

Baldauf, M., L. Garlappi, and C. Yannelis. 2020. Does climate change affect real estate prices? Only if you believe in it. The Review of Financial Studies 33:1256–1295.

Article Google Scholar

Barrios, J. M. 2021. Staggeringly problematic: A primer on staggered DiD for accounting researchers. Working paper. https://ssrn.com/abstract=3794859 .

Basu, S., J. Vitanza, W. Wang, and X. R. Zhu. 2022. Walking the walk? Bank ESG disclosures and home mortgage lending. Review of Accounting Studies 27:779–821.

Article Google Scholar

Bernstein, A., M. T. Gustafson, and R. Lewis. 2019. Disaster on the horizon: The price effect of sea level rise. Journal of Financial Economics 134:253–272.

Article Google Scholar

Berry, S. T., and J. Waldfogel. 2001. Do mergers increase product variety? Evidence from radio broadcasting. The Quarterly Journal of Economics 116:1009–1025.

Article Google Scholar

Bian, X., Z. Lin, and Y. Liu. 2018. House price, loan-to-value ratio and credit risk. Journal of Banking & Finance 92: 1–12.

Article Google Scholar

Bolton, P., S. J. Reichelstein, M. T. Kacperczyk, C. Leuz, G. Ormazabal, and D. Schoenmaker. 2021. Mandatory corporate carbon disclosures and the path to net zero. Management and Business Review 1:21–28.

Article Google Scholar

Botosan, C. A. 1997. Disclosure level and the cost of equity capital. The Accounting Review 72:323–349.

Article Google Scholar

Breuer, M., and E. D. DeHaan. 2024. Using and interpreting fixed effects models. Journal of Accounting Research 62:1183–1226.

Article Google Scholar

Buchanan, M. K., M. Oppenheimer, and R. E. Kopp. 2017. Amplification of flood frequencies with local sea level rise and emerging flood regimes. Environmental Research Letters 12 : 064009.

Article Google Scholar

Carattini, S., E. Hertwich, G. Melkadze, and J. G. Shrader. 2022. Mandatory disclosure is key to address climate risks. Science 378:352–354.

Article Google Scholar

Cengiz, D., A. Dube, A. Lindner, and B. Zipperer. 2019. The effect of minimum wages on low-wage jobs. The Quarterly Journal of Economics 134:1405–1454.

Article Google Scholar

Chaney, T., D. Sraer, and D. Thesmar. 2012. The collateral channel: How real estate shocks affect corporate investment. American Economic Review 102:2381–2409.

Article Google Scholar

Cho, C., R. Frankel, and X. Martin. 2024. Information reliability and market outcomes. Management Science 70:6255–6279.

Christensen, H.B., E. Floyd, L.Y. Liu, and M. Maffett. 2017. The real effects of mandated information on social responsibility in financial reports: Evidence from mine-safety records. Journal of Accounting and Economics 64: 284–304.

Article Google Scholar

Christensen, H. B., L. Hail, and C. Leuz. 2021. Mandatory CSR and sustainability reporting: Economic analysis and literature review. Review of Accounting Studies 26:1176–1248.

Article Google Scholar

Christiano, L. J., R. Motto, and M. Rostagno. 2014. Risk shocks. American Economic Review 104:27–65.

Article Google Scholar

Cinelli, C., and C. Hazlett. 2020. Making sense of sensitivity: Extending omitted variable bias. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 82:39–67.

Article Google Scholar

Cleetus, R. 2019. The truth about flood risk can worsen American inequality. https://www.wealthmanagement.com/real-estate/the-truth-about-flood-risk-can-worsen-american-inequality . Accessed 2 Apr 2026.

Cornaggia, J., Y. Mao, X. Tian, and B. Wolfe. 2015. Does banking competition affect innovation? Journal of Financial Economics 115:189–209.

Article Google Scholar

Dagher, J., and K. Kazimov. 2015. Banks’ liability structure and mortgage lending during the financial crisis. Journal of Financial Economics 116:565–582.

Article Google Scholar

Darlington, A. 2018. Little-known federal law keeps buyers from finding out if a home routinely floods. Post and Courier. https://www.postandcourier.com/news/little-known-federal-law-keeps-buyers-from-finding-out-if/article_9e348564-9b13-11e8-934c-6ffa499e136b.html . Accessed 22 Apr 2025.

Davis, M. A., W. D. Larson, S. D. Oliner, and B. R. Smith. 2023. A quarter century of mortgage risk. Review of Finance 27:581–618.

Article Google Scholar

Di Maggio, M., A. Kermani, R. Ramcharan, and E.G. Yu. 2017. Household credit and local economic uncertainty. FRB of Philadelphia Working Paper No. 17–21. https://ideas.repec.org/p/fip/fedpwp/17-21.html . Accessed 18 July 2025.

Downar, B., J. Ernstberger, S. Reichelstein, S. Schwenen, and A. Zaklan. 2021. The impact of carbon disclosure mandates on emissions and financial operating performance. Review of Accounting Studies 26:1137–1175.

Article Google Scholar

Dube, A., T. W. Lester, and M. Reich. 2010. Minimum wage effects across state borders: Estimates using contiguous counties. The Review of Economics and Statistics 92:945–964.

Article Google Scholar

Eerola, E., and T. Lyytikäinen. 2015. On the role of public price information in housing markets. Regional Science and Urban Economics 53:74–84.

Article Google Scholar

Engelberg, J. E., and C. A. Parsons. 2011. The causal impact of media in financial markets. Journal of Finance 66:67–97.

Article Google Scholar

Engle, R. F., S. Giglio, B. Kelly, H. Lee, and J. Stroebel. 2020. Hedging climate change news. The Review of Financial Studies 33:1184–1216.

Article Google Scholar

Fan, G. H. 2025. Economic consequences of public pension accounting regulation changes: Evidence from housing markets and local economies. The Accounting Review 100:249–275.

Article Google Scholar

Favara, G., and J. Imbs. 2015. Credit supply and the price of housing. American Economic Review 105:958–992.

Article Google Scholar

Federal Deposit Insurance Corporation. 2019. Consumer Compliance Examination Manual. https://www.fdic.gov/consumer-compliance-examination-manual . Accessed 14 May 2023.

Federal Emergency Management Agency. 2022. Flood risk disclosure: Model state requirements for disclosing flood risk during real estate transactions. https://www.fema.gov/sites/default/files/documents/fema_state-flood-risk-disclosure-best-practices_07142022.pdf . Accessed 17 Apr 2025.

Fiechter, P., J. Hitz, and N. Lehmann. 2022. Real effects of a widespread CSR reporting mandate: Evidence from the European Union’s CSR Directive. Journal of Accounting Research 60:1499–1549.

Article Google Scholar

Frankel, R., S.P. Kothari, and A. Raghunandan. 2025. The economics of ESG disclosure regulation. Review of Accounting Studies . https://doi.org/10.1007/s11142-025-09900-9 .

Article Google Scholar

Fung, A., M. Graham, and D. Weil. 2007. Full disclosure: The perils and promise of transparency . Cambridge: Cambridge University Press.

Book Google Scholar

Garmaise, M. J., and T. J. Moskowitz. 2004. Confronting information asymmetries: Evidence from real estate markets. The Review of Financial Studies 17:405–437.

Article Google Scholar

Gilboa, I., and D. Schmeidler. 1989. Maxmin expected utility with non-unique prior. Journal of Mathematical Economics 18:141–153.

Article Google Scholar

Granja, J. 2018. Disclosure regulation in the commercial banking industry: Lessons from the national banking era. Journal of Accounting Research 56:173–216.

Article Google Scholar

Greenstone, M., P. Oyer, and A. Vissing-Jorgensen. 2006. Mandated disclosure, stock returns, and the 1964 Securities Acts amendments. The Quarterly Journal of Economics 121:399–460.

Article Google Scholar

Gyourko, J., and A. Saiz. 2006. Construction costs and the supply of housing structure. Journal of Regional Science 46:661–680.

Article Google Scholar

Hauer, M. E., J. M. Evans, and D. R. Mishra. 2016. Millions projected to be at risk from sea-level rise in the continental United States. Nature Climate Change 6:691–695.

Article Google Scholar

Huang, R. R. 2008. Evaluating the real effect of bank branching deregulation: Comparing contiguous counties across U.S. state borders. Journal of Financial Economics 87:678–705.

Article Google Scholar

Igan, D., and L.U. Laeven. 2012. Credit booms and lending standards: Evidence from the subprime mortgage market. Journal of Money, Credit and Banking 44: 367–384.

Article Google Scholar

Ilhan, E. 2020. Sea level rise and portfolio choice. Working paper. https://ssrn.com/abstract=3743046 . Accessed 30 Dec 2024.

Jennings, J., J. M. Kim, J. Lee, and D. Taylor. 2024. Measurement error, fixed effects, and false positives in accounting research. Review of Accounting Studies 29:959–995.

Article Google Scholar

Jin, G. Z., and P. Leslie. 2003. The effect of information on product quality: Evidence from restaurant hygiene grade cards. The Quarterly Journal of Economics 118:409–451.

Article Google Scholar

Kang, J. K., M. Loumioti, and R. Wittenberg-Moerman. 2021. The harmonization of lending standards within banks through mandated loan-level transparency. Journal of Accounting and Economics 72 : 101386.

Article Google Scholar

Kielty, P. D., K. P. Wang, and D. L. Weng. 2023. Simplifying complex disclosures: Evidence from disclosure regulation in the mortgage markets. The Accounting Review 98:191–216.

Article Google Scholar

Klasa, S., H. Ortiz-Molina, M. Serfling, and S. Srinivasan. 2018. Protection of trade secrets and capital structure decisions. Journal of Financial Economics 128:266–286.

Article Google Scholar

Krueger, P., Z. Sautner, D. Y. Tang, and R. Zhong. 2024. The effects of mandatory ESG disclosure around the world. Journal of Accounting Research 62:1795–1847.

Article Google Scholar

Kusisto, L. 2019. Is your home at risk of flooding? The data is hard to find. Wall Street Journal. https://www.wsj.com/articles/is-your-home-at-risk-of-flooding-the-data-is-hard-to-find-11560418204 . Accessed 14 May 2025.

Leuz, C., and P. Wysocki. 2016. The economics of disclosure and financial reporting regulation: Evidence and suggestions for future research. Journal of Accounting Research 54:525–622.

Article Google Scholar

Lim, I., and D. D. Nguyen. 2021. Hometown lending. Journal of Financial and Quantitative Analysis 56:2894–2933.

Article Google Scholar

Lu, Y., and V.V. Nikolaev. 2024. The impact of climate hazards on banks’ long-run performance. Working paper. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4569935 . Accessed 2 June 2025.

Marcy, D., W. Brooks, K. Draganov, B. Hadley, C. Haynes, N. Herold, J. McCombs, et al. 2011. New mapping tool and techniques for visualizing sea level rise and coastal flooding impacts. In Solutions to Coastal Disasters 2011:474–490.

Meyer, J. W., and B. Rowan. 1977. Institutionalized organizations: Formal structure as myth and ceremony. American Journal of Sociology 83:340–363.

Article Google Scholar

Mian, A., and A. Sufi. 2009. The consequences of mortgage credit expansion: Evidence from the U.S. mortgage default crisis. The Quarterly Journal of Economics 124:1449–1496.

Article Google Scholar

Murfin, J., and M. Spiegel. 2020. Is the risk of sea level rise capitalized in residential real estate? The Review of Financial Studies 33:1217–1255.

Article Google Scholar

Nanda, A., and S. L. Ross. 2012. The impact of property condition disclosure laws on housing prices: Evidence from an event study using propensity scores. The Journal of Real Estate Finance and Economics 45:88–109.

Article Google Scholar

National Association of Realtors. 2019. State flood hazard disclosures survey . National Association of Realtors.

Natural Resources Defense Council. 2019. How states stack up on flood disclosure. https://www.nrdc.org/flood-disclosure-map . Accessed 13 Oct 2022.

Nguyen, D. D., S. Ongena, S. Qi, and V. Sila. 2022. Climate change risk and the cost of mortgage credit. Review of Finance 26:1509–1549.

Article Google Scholar

Ouazad, A., and M.E. Kahn. 2022. RETRACTED: Mortgage finance and climate change: Securitization dynamics in the aftermath of natural disasters. The Review of Financial Studies 35: 3617–3665.

Article Google Scholar

Palmer, A., and J. Turrentine. 2022. It’s time to fix our water-logged National Flood Insurance Program. Natural Resources Defense Council. https://www.nrdc.org/stories/time-fix-water-logged-national-flood-insurance-program . Accessed 22 April 2025.

Pope, J. C. 2008. Do seller disclosures affect property values? Buyer information and the hedonic model. Land Economics 84:551–572.

Pralle, S. 2019. Drawing lines: FEMA and the politics of mapping flood zones. Climatic Change 152:227–237.

Article Google Scholar

Rao, K. 2017. Climate change and housing: Will a rising tide sink all homes? Zillow Research.

Saadi, V. 2020. Role of the Community Reinvestment Act in mortgage supply and the U.S. housing boom. The Review of Financial Studies 33: 5288–5332.

Article Google Scholar

Saiz, A. 2010. The geographic determinants of housing supply. The Quarterly Journal of Economics 125:1253–1296.

Article Google Scholar

Sastry, P. 2022. Who bears flood risk? Evidence from mortgage markets in Florida. Working paper. https://ssrn.com/abstract=4306291 .

Scata, J. 2025. Flooding can put unsuspecting home buyers financially underwater. Natural Resources Defense Council. https://www.nrdc.org/bio/joel-scata/flooding-can-put-unsuspecting-home-buyers-financially-underwater . Accessed 25 Apr 2025.

Sinha, K. 2022. Fracking disclosure, collateral value, and the mortgage market. The Accounting Review 97:427–454.

Article Google Scholar

Stiroh, K.J. 2020. Climate change and risk management in bank supervision. Speech, March 4, 2020. https://www.newyorkfed.org/newsevents/speeches/2020/sti200304 . Accessed 1 Sept 2025.

Tomar, S. 2023. Greenhouse gas disclosure and emissions benchmarking. Journal of Accounting Research 61 (2): 451–492.

Troy, A., and J. Romm. 2004. Assessing the price effects of flood hazard disclosure under the California natural hazard disclosure law (AB 1195). Journal of Environmental Planning and Management 47:137–162.

Article Google Scholar

Turnbull, G.K., Zahirovic-Herbert, V. and Mothorpe, C. 2013. “Flooding and liquidity on the bayou: The capitalization of flood risk into house value and ease‐of‐sale.” Real Estate Economics 41 (1): 103–129.

Wooldridge, J.M. 2021. Two-way fixed effects, the two-way Mundlak regression, and difference-in-differences estimators. Working paper. https://ssrn.com/abstract=3906345 .

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 63 KB) (download DOCX )

Rights: © 2026 The Author(s)
Rights and permissions 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/ . Reprints and permissions
Edition changes: Licensed narrative text edition. Original headings, paragraphs and references are preserved with normalized whitespace and layout. Figures, tables, display equations and separately licensed third-party material are not reproduced; use the linked original full text for those elements and definitive pagination. This is an attributed republication, not newly conducted research.