Risk Information and Perceived Environmental Risk in Communities Hosting Hazardous Facilities: A Survey Experiment

The intervention consisted of providing respondents with factual, site-specific information about the hazardous facility located in their municipality. Respondents were randomly assigned either to receive this information immediately before the risk-assessment module or to a control condition that received no additional information.

Respondents within each of the ten study municipalities were randomly assigned to an information treatment or a control condition. Of the 514 respondents, 258 were assigned to the information condition and 256 to the control condition, with allocation balanced across the ten study sites.

Respondents assigned to the treatment condition received a site-specific information block immediately before answering questions on perceived environmental and health risks and support for risk-management measures. The control group proceeded directly to these questions without receiving the additional information.

The treatment provided factual information about the facility located in the respondent’s municipality. The information followed a common structure describing the facility’s principal activity, relevant hazardous substances or hazards, possible accident scenarios and their potential consequences. For industrial facilities, information on the extent of the emergency planning zone was included where applicable. The format and approximate length of the information were kept similar across sites, while the substantive content was adapted to the characteristics and risks of each facility.

The information therefore differed across study sites. Some texts contained relatively reassuring information, such as limited expected environmental impacts or the absence of anticipated evacuation of residents, whereas others described hazardous substances, potential accident scenarios or impacts potentially extending to nearby residential areas. Information provided for landfill sites covered risks including landfill gas, fire, explosion and wind-borne waste.

All variables used to characterize respondents and examine treatment-effect heterogeneity were measured before the intervention. Outcome measures were administered immediately after the treatment.

2025-11-24

2025-12-28

Overall perceived environmental risk attributable to the facility.
Perceived risks related to specific environmental, health, and safety dimensions.
Support for regulatory, risk-management, and compensatory measures related to the facility.

Overall perceived environmental risk is measured on a four-point scale ranging from very low to very high risk.

Specific risk perceptions are measured separately for the release of hazardous substances, explosion or fire, air pollution, water pollution, soil pollution, long-term health consequences for residents, long-term health consequences for the respondent’s household, noise, odor, and personal sense of safety. Each item is measured on a four-point scale.

Support for policy measures is measured separately for independent inspections, public information on risks, protective zoning, mandatory emergency exercises, an operator contribution to a municipal fund for safety and the environment, more stringent environmental standards with sanctions, and restrictions on the facility’s operations. Each item is measured on a four-point scale.

The study uses a randomized survey experiment conducted among adult residents of ten Czech municipalities hosting industrial facilities or landfills subject to heightened environmental or safety regulation. Within each municipality, respondents were randomly assigned either to receive factual, site-specific information about the facility before answering the outcome questions or to a control condition that received no additional information. Outcomes were measured immediately after the intervention.

The experiment was embedded in a face-to-face computer-assisted survey conducted in ten Czech municipalities. Respondents were randomly assigned within municipality to either an information treatment or a control condition. Of the 514 respondents, 258 were assigned to the treatment group and 256 to the control group, with allocation balanced across the ten study sites.

All pre-treatment characteristics, including environmental attitudes, self-reported proximity to the facility, frequency of contact with its surroundings, prior awareness of the facility’s activities, perceived informedness about risks, and perceived advantages and disadvantages of the facility, were measured before treatment assignment was revealed to the respondent.

Respondents assigned to the treatment group then received a factual, site-specific information block describing the facility’s activities, relevant hazards, possible accident scenarios and potential consequences. Respondents in the control group proceeded directly to the outcome module without receiving this additional information. Outcome measures were administered immediately after the intervention.

Computer-generated randomization implemented automatically within the survey software. Respondents were assigned according to a pre-specified allocation sequence generated by the software.

Individual respondent, stratified by study site (municipality).

No

500 individual respondents

500 adult respondents

250 respondents in the information treatment condition and 250 respondents in the control condition.