What influences the pattern of pollutant releases? An investigation of firms’ siting and households’ sorting decisions in Ontario, Canada

Abstract

The relationship between the pattern of pollutant releases and neighborhood characteristics has been in the center of environmental justice debates for the last decades. One part of the discourse revolves around the question if the siting decision of polluting firms or the sorting decision of households is the explanatory factor for the pattern of pollutant releases. The aim of this study is to contribute to the discussion by investigating the potential correlation between polluting facilities and socioeconomic characteristics of neighborhoods in Ontario, Canada, using data from the Canadian census and National Pollutant Release Inventory. The analysis shows that when the results are considered from the perspectives of the siting decision of firms and sorting decision of households’ hypotheses, the results weigh in favor of the sorting decision of households.

Keywords:

• local pollution
• firms siting
• households sorting

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Goods-producing industries contribute to more than 22% of GDP in Ontario. The contributing sectors to goods-producing industries in the order of magnitude are: construction 23.95%, transportation equipment 14.11%, utilities 8.67%, food, beverage, and tobacco 8.62%, primary and fabricated metals 8.57%, chemical and petroleum products 5.83%, electrical and electronic products 3.68%, agriculture 3.96%, and other primary and manufacturing industries contribute 6.88 and 16.32%, respectively (Statistics Canada 2015).

2. The elimination of the mandatory-long-form census and its replacement with the voluntary National Household Survey in 2011 initiated an ongoing debate about the reliability of the data collected. In particular, studies show that there are systematic differences between the non-respondents and respondents in similar voluntary surveys (Green and Milligan 2010; Brochu, Morin, and Billette 2014), where people at the extremes of the socioeconomic scale are unlikely to fill out voluntary surveys (Collier 2010). Consequently, there are major concerns about the representativeness of the 2011 survey because of the selection bias inherent in the voluntary survey (Green and Milligan 2010).

3. The indicator variable ‘Census Metropolitan or Agglomeration Area’ is used as a proxy for urban areas. A Census Metropolitan or Agglomeration Area is one or more adjacent municipalities centered around a core of at least 50,000 or 10,000 people, respectively. In addition, a Census Metropolitan Area must have a total population of at least 100,000 people (Statistics Canada 2012).

4. The sub-samples of the Greater Toronto Area, and urban versus rural areas in Ontario are focused in particular.