Project Description and Objectives of Research:
There has been much disagreement about how to best measure the benefits of cleanup of hazardous waste sites. In 1980, Congress mandated the EPA to clean up abandoned hazardous waste sites. The Comprehensive Environmental Response, Compensation and Liability Act (CERCLA) required the EPA to establish criteria to prioritize sites based on risks to health, environment, and welfare. Welfare was interpreted to mean impacts associated with health and the environment, not economic and social impacts (Greenberg and Schneider, 1995). Consequently, the real effect of hazardous waste sites on property values has often been neglected in cost-benefit analyses. Incorporating losses in property values in the analyses may yield a different conclusion about the effectiveness of remedial actions.
Previous academic studies have attempted to measure benefits from the cleanup of hazardous waste by showing that residential property values become lower as the distance to a hazardous waste site decreases. Extending this argument, if the hazardous waste site is removed, then the discount for being in a location that is close to a former hazardous waste site should be recovered. After environmental contamination is completely cleaned up, ceteris paribus, one would expect residential property values to regain their lost values. The benefits of cleanup are then the difference between what property values would be if the hazardous waste site never existed and what property values are with the hazardous waste site. We argue that this reasoning is faulty because of hysteresis or path dependence. If stigma effects from a site exist, the stigma associated with a hazardous waste site leads to irreversible losses in property values. Past studies have ignored the effects of stigma and, therefore, may have overvalued the benefits of cleanup of hazardous waste sites. The hypothesis to be tested is whether there is a stigma effect from environmental damage on residential property values. If the stigma from environmental damage is significant, then the framework developed in this study can be used to analyze issues of environmental justice.
This project augments the existing literature on the environmental contamination and residential property values because property values by analyzing such values over a long period of time. Specifically, existing studies have not looked at property values after cleanup has been completed which we proposed to do in this study.
Summary of Findings:
We developed a theoretical model that includes both path dependence and expectations in the determination of property values. This research makes an important contribution to the tipping (or residential succession) literature by showing that, with external economies and adjustment costs, both path dependence and expectations can play a role in determining whether there will be a stigma equilibrium. The model was estimated empirically using hedonic price techniques. The pooled data set covered the period 1979 to 1995 and included more than 200,000 observations. The time period included observations before, during, and after cleanup of a contaminated site. We found that only houses located in a very close proximity to the hazardous waste site were stigmatized.
A variety of estimation methods were used in this study. The first approach was to estimate a distance model. We found that the coefficient on distance from smelter starts out positive, then turns negative after cleanup. Standing alone, one might conclude from this result that there is no stigma. However, in the first post-cleanup period (1987-90), there were no sales within one mile of the smelter.
We hypothesized the price gradient on distance from the hazardous waste site is unlikely to be continuous. We used two approaches to deal with the possible discontinuity of the price gradient. The first approach is using a linear spline function in place of the distance variable in the hedonic price equation. The second approach is to replace the distance variable with discrete distance dummy variables. We conclude from the results of both approaches that the effect of the smelter diminishes rapidly with distance. The smelter?s sphere of influence is no greater than one mile. There is stigma for houses within one mile of the smelter.
A repeat sales subset of the data set which consists of houses that were sold more than once was identified, and a repeat sales model was estimated separately for the smelter area and a control area. The hypothesis tested that the rates of return are the same across areas. The repeat sales data set was also used to estimate separate hedonic regressions for houses that were sold both before and after cleanup. The coefficients on the year dummy variables in the years after additional contamination concerns about the RSR smelter site arose (1991-1995) are more negative for the smelter area than the control area. In the period during and initially after cleanup (1981-1990), the return is better in the smelter area. This allows for a non-confounded analysis because the structural characteristics of the house do not usually change over time. However, there may be selection bias because the most stigmatized houses may not have been sold after cleanup.
The dynamic effects of the smelter were analyzed by estimating its effect on housing appreciation rates. Using the repeat sales data set, appreciation rates were calculated for each of event-driven time periods. Appreciation rates multiplied by 100 were used as the dependent variable in a regression with housing, neighborhood, and environmental attributes used as the independent variables. In the period in which identification and cleanup occurred, which was also a period of intense media coverage (1981-1986), a location that is farther away from the smelter had a positive and significant effect on the appreciation rate of the house. In the first post-cleanup period (1987-1990), the houses with locations that are in close proximity to the smelter experienced a significantly higher appreciation rate than houses located farther away. Finally, in the period of additional concern about the smelter area (1991-1995), a location that is farther away from the smelter had a positive but insignificant effect on the appreciation rate of the house.
In order to compare the smelter area with a control area, models with an indicator variable, which is equal to one when the distance from the smelter is less than a specified number of miles and zero otherwise was estimated. This model was estimated with the smelter area equal to a circle with a radius of one mile and a radius of four miles. There is a discount for being within one mile of the smelter in each period, which can be interpreted that a stigma exists on properties within one mile.
The effect of media coverage was also analyzed in this study. The bulk of the coverage occurred in the period in which identification of the site and cleanup occurred (1981-1986). The results indicate that the estimated coefficient on the media variable in this time period was negative and significant for properties sold within four miles of the RSR site, while the estimated media coefficient was positive and significant for properties sold greater than four miles away from the site. Media coverage again increased in the period of new concern after cleanup (1991-1995). The media variable coefficient was again negative and significant for properties sold within the smelter area, while it was positive but insignificant for properties sold greater than four miles from the smelter.
We also estimated a dynamic discrete time model in order to analyze the evolution of perceived risk around a hazardous waste site and its effect on property values. Perceived risk is different from scientifically assessed risk because it can be manipulated. We found that media coverage and high prior risk perception increase current perceived risk. Increased perceived risk surrounding the hazardous waste site, in turn, lowers property values. This research is very innovative because it uses a revealed preference approach to estimating perceived risk, while other attempts to estimate perceived risk rely on survey data. Using Generalized Maximum Entropy (GME) estimation techniques, we found that perceived risk, weighted by distance, has the expected negative relationship with housing price. The coefficient on lagged perceived risk is positive and less than one, so that perceived risk does not explode. Finally, the media coefficient is positive. Media coverage increases perceived risk.
The theoretical model shows that stigma can be caused by both path dependence and uncertainty, rather than uncertainty alone, which is the accepted explanation in the environmental economics literature. The theoretical model also shows that stigma is not the only outcome after contamination. Recovery is also possible. In our examination of Dallas County, Texas, and the RSR lead smelter site, we found that stigma exists close to the site, but it dissipates rapidly with distance. Before the identification of the site by the EPA, there was already a discount for a buying a house with a smelter location. During the period of smelter identification and cleanup, there was a period of high media coverage, and the discount increased. In the first four years after a court ruled that cleanup was complete, no houses were sold within one mile of the site. Finally, in the subsequent five years, houses were sold within one mile of the site but a discount remained. Media coverage of the site caused property values to decrease in the smelter and increase outside the smelter area.
Papers and Manuscripts:
McCluskey, Jill J. and Gordon C. Rausser, 1998. "Estimation of Perceived Risk and Its Effect on Property Values," Working Paper, Department of Agricultural Economics, Washington State University.
McCluskey, Jill J. and Gordon C. Rausser, 1998. "Stigmatized Asset Value: Is it Temporary or Permanent?" Working Paper, Department of Agricultural and Resource Economics, University of California, Berkeley.
McCluskey, Jill J. and Gordon C. Rausser. "Causes and Compensation for Environmental Inequality," In preparation.
"Stigmatized Asset Value: Is it Temporary or Permanent," Presented at: University of California, Berkeley--Real Estate Ph.D. seminar at the Haas School of Business, November 1997; Pennsylvania State University--Department of Energy, Environmental, and Mineral Economics, January 1998; Department of Agricultural and Resource Economics--North Carolina State University, January 1998; School of Public Policy and Management--The Ohio State University, May 1998.
"Stigma of Environmental Damage on Residential Property Values," Selected Paper presented at the 1997 American Agricultural Economics Association meetings, Toronto.