Determinants of community response to stormwater pollution in an urbanized watershed.

Abstract

Increasing federal regulation of stormwater has been accompanied by decreasing amounts of federal aid available for its mitigation. Left to fend for themselves, communities in urbanized watersheds with diverse demographic, environmental, and spatial characteristics are likely to exhibit different levels of response to stormwater pollution. The purpose of this research is to measure community response to stormwater pollution and to identify the factors which influence the variations in response. A general response model is developed to identify the factors influencing variation in community response. Basic constructs of the model use environmental hazard theory, integrated with analysis of community demographic and political forces affecting local government decision processes. Community stormwater response is measured by an index comprised of twelve response indicators. Selection criteria for the response indicators include their effectiveness, variability, and general representation of the spectrum of stormwater response activity. Observable activities are scored progressively based upon their relative ecosystem impacts. Weighted and unweighted indicator scores are computed, with a survey of stormwater experts providing the basis for the weights. The study area is the urbanized River Rouge watershed in southeast Michigan containing forty-eight communities. Data sources include interviews with public officials, staff, and public documents. Results indicate a wide variation in response levels among the selected communities, with the finding that specific planning activities, such as ordinances, may act as catalysts for additional stormwater implementation and monitoring activities. Regression analysis identifies four significant predictors of response at the watershed scale: the mean age of householder over 25, the total tax rate, community type (city or township), and the amount of green staff (presence of recycling coordinators, foresters, or environmental planners). This mix of predictors demonstrates the applicability of hazard theory to stormwater response, as community stormwater response is influenced by the capacity and economic means to effect change--two central constructs of hazard theory. Locally, a case study of two communities suggests that site-specific visual and temporal characteristics of the stormwater hazard influence community involvement and affect the level of stormwater response. Weighted indicators produce a regression model with slightly stronger statistical and interpretive characteristics than unweighted indicators.