1 May 30, 2003 Timewise and Directional Heterogeneity in Distance Profiles reflecting Superfund Taint: the Dynamics of Neighborhood Change by Trudy Ann Cameron and Graham Crawford University of Oregon ABSTRACT Certain sociodemographic groups often seem to be relatively more concentrated near environmental hazards than in the surrounding community. However, snapshot statistical analyses cannot reveal how residential mobility for these different groups reacts to public perceptions of environmental hazards. Panel data for census tracts for sixteen different Superfund localities allow us to examine how ethnicities, the age distribution and family structure vary over time with proximity to these major environmental disamenities. For any particular group, a distance profile with a slope that decreases over time suggests the group may have been “coming to the nuisance.” We find many statistically significant time patterns in distance profiles. However, there appears to be no way to generalize the mobility patterns for different groups in the face of evolving environmental hazards. This heterogeneity may account for the difficulty other researchers have experienced in identifying such systematic effects. Our secondary theme concerns directional heterogeneity in externalities generated by point sources of pollution. In the context of an airborne pollutant, we explain how to let distance effects vary systematically with direction. Failure to allow for directional heterogeneity can obscure otherwise statistically significant distance effects. The “downwind” direction can be estimated, or the true downwind direction can be imposed upon the model using information about prevailing winds in the area. If appropriate, the downwind direction can be allowed to vary seasonally. We find frequent evidence of statistically significant directional effects in our sixteen data sets. Acknowledgments: This research was funded in part by the US Environmental Protection Agency (CR 824393-01) through a cooperative agreement with Cornell University (29067-5808, PI: William D. Schulze) and a subcontract to the University of California at Los Angeles. Support from research funds accompanying the R.F. Mikesell Chair in Environmental and Resource Economics at the University of Oregon are also gratefully acknowledged.