In this paper, we attempt to contribute to the growing environmental justice debate by exploring environmental equity in the forty-nine counties of Ontario, Canada. We use multiple regression analysis to address a central research question: what variables predict the location of pollution emissions? Data were extracted from the 1993 National Pollutant Release Inventory and the 1991 Census of Canada to assess relationships among socioeconomic class variables, industrial and land-use variables, and pollution emissions. The results agree with the findings of recent US studies. Manufacturing employment, urbanization variables, dwelling value, and household income were all significantly related to pollution emissions. These relationships took the same direction as in most of the US studies. In total, the four variables account for about 63% of the variation in pollution emissions (adjusted R-2 = 0.626, p < 0.0001). Contrary to a hypothesis of environmental inequity, the household income variable displayed a positive relationship with pollution emissions. A conceptual model of the development process is formulated to explain the positive relationship between pollution emissions and income levels. We conclude the paper with suggestions for future research.
