1.Background

The siting and operation of a petrochemical plant often face widespread public opposition due to its potential environmental health risk. The public perceived risk associated with its synthetic chemical emissions and products commonly translates into a generalized stress or a fear of illness. Numerous countries have implemented risk communication programmes from ‘top-down’ to alleviate public opposition. Thus, this study aims to improve the knowledge about the effects of risk communication on petrochemical industrial risk perception through mapping the geographical patterns of local perceived risk among the potentially affected population and examining their determinants from 'bottom-up'.

2.Methods and data

Combining a social psychology theory with a Risk Perception Mapping approach, we develop a Social Risk Communication and Perception Model of Petrochemical Indutry Complex (SRCPMPIC) to explain public risk attitudes towards petrochemical industry and their determinants. We apply a method integrated a geographical information system based spatial statistics technique with a spatial regression analysis into testing the SRCPMPIC. The data were collected by a survey for the residents of two municipalities in Central Taiwan. One of the municipality owns an established petrochemical plant, while the other is still under planning and construction. The survey was conducted through a face-to-face and door-to-door interview, which was designed by integrating focus group meetings with pre-test surveys. Then, using an ethnographic sampling approach, 452 respondents were used in the analysis.

3.Results

This study compares the extent of which the residents’ perceived risk is spatially autocorrelated within and between communities, and explaining why clustering of specific risk attitudes occur in certain locations. Results show that the spatial patterns of residents’ perceived risk across the two municipalities are not dispersed randomly, but instead spatially correlated in certain regions. Using both a spatial-lag and a spatial-error regression analysis, findings show that social distrust, pollution experiences, proximity to plants, farmers and women have a significant and positive relation to perceived levels of risk. Especially, risk communication is a strong contributor to mitigate perceived risk. Thus, policy lines can focus more on communicating petrochemical risk to the residents, as well as improving the environment for encouraging dialogue and taking particularly affected people’s welfare into consideration.