Sea Level Rise and Home Prices: Evidence from Long Island

Abstract

Global sea level rise is a known consequence of climate change. As predictions of sea level rise have grown in magnitude and certainty, coastal real estate assets face an increasing climate risk. I use a complete data set of repeated home sales from Long Island in New York State to estimate the appreciation discount caused by the threat of sea level rise. The repeat sale methodology allows for time-invariant, unobserved property characteristics to be controlled for. Between 2000 and 2017, I find that residential properties that were exposed to future sea level rise experienced an annual price appreciation rate of roughly 1% point below unexposed properties. I provide numerous robustness checks to confirm this result. I also find evidence of demand spillovers by estimating an appreciation premium for properties that are near the coast but are relatively safe from sea level rise.

Appendices

Appendix A

In Table 5, I provide results where I limit the sample to properties within 2 km of the coast. In Fig. 9, I provide an elevation map where I indicate which areas of Long Island are within 2 km of the coast. On the south coast, much of this area is at very low elevations, whereas the north coast contains steeper terrain that extends to higher elevations

Fig. 9

Elevation map of the study area. Areas of Nassau and Suffolk counties that are more than 2 km from the coast are shown in black

Appendix B

To generate the sample of properties used in analysis, I drop a number of observations to focus the sample on observations that are most relevant to the research question and fit with the repeat sales methodology. In this Appendix, I provide results using alternative samples to test the robustness of results. Overall, I find results are not generally sensitive to the specific decisions made regarding which observations are dropped.

I limit the sample to only properties that sold multiple times in order to include fixed effects at the property level. If properties that sell frequently are not representative of overall trends in the market the regression results will also not be representative of the overall market. In Table 8, I show results among properties that sold exactly twice, rather than at least two times. If the bias in the repeat sale sample is correlated with sale frequency, this subsample should be more representative of the overall market. I find results on the limited sample are robust and estimates are very close to the full repeat sale sample analysis (Table 2).

The analysis of this paper is focused on the housing market. I exclude other property types, particularly commercial and agriculturally zoned land. These non-housing property types make up 4.4% of transactions. In Table 9, I repeat the main analysis but include these other property types. I find results are very similar to the main estimates of the paper.

Table 8 Effect of sea level exposure on price appreciation, only properties that sold exactly twice

Table 9 Effect of sea level exposure on price appreciation, all property types

Appendix C

Table 10 provides split sample results where the effect of exposure is estimated separately for properties on the North and South sides of the study area. I split the sample area with a line that runs equidistant from the north and south coasts.

I find that the significant negative price effect of exposure is driven by properties on the south coast. In fact, I find a marginally significant positive price effect for exposure on the north coast (column 1). Because hurricane activity approaches from the south, the north coast is largely protected from storm surge events. The marginally positive price effect could be evidence of demand substitution towards coastal properties that are considered to be relatively low risk

Table 10 Effect of sea level exposure on price appreciation, north coast vs south coast

Appendix D

In Table 11, I provide a horse-race regression where I include the elevation based at risk definition simultaneously with the FEMA flood zone definition. Across the three regressions, I find that all six coefficients are negative, though not all are statistically significant. The result suggests that conditional on elevation, homes within FEMA defined flood zones appreciated less quickly. Also, the result suggests that conditional on being in a flood zone, lower elevation properties suffered a greater price penalty. The two risk indicators are highly correlated, potentially leading to multicollinearity issues. Therefore, the results should be interpreted with caution

Table 11 Effect of sea level exposure on price appreciation, impact of elevation vs flood zone