Citizens' legitimacy judgments on multi-stakeholder governance models. Corporate political activity and social trust as antecedent variables
The dataset was generated by the fieldwork for an empirical study consisted of a large-scale survey of residents, through street interviews, in the three localities (El Morell, La Pobla de Mafumet and Perafort-Puigdelfí) closest to a petrochemical complex in Spain. These localities received the greatest impact of the positive and negative externalities generated by the activity of the firms in the analyzed complex. The research was done during the first fortnight of September 2018. It consisted of closed questions with items measured on a five-point Likert scale (1, lowest agreement and 5, highest agreement). A total of 390 responses were obtained using simple random sampling. Residents living in the three localities were randomly invited to answer the questionnaire. The total number of inhabitants of the localities surveyed was 8,807 according to figures obtained from the Spanish Institute of Statistics (2018), which for a 95% confidence level represents a sample error of ±5% (p=q=0.5). The research sample comprised 149 (38.20%) respondents from La Pobla de Mafumet, 146 from El Morell (37.44%), and the smallest group, 95 residents, from Perafort-Puigdelfí (24.36%). The total sample of interviewees in each of the three localities was significantly representative of the population universe in their locality. The model variables are: citizens' perceptions of financial incentives (corporate political activity, CPA), citizens' trust in companies, citizens' trust in public authorities, and legitimacy that citizens confer on a multi-stakeholder governance model (political corporate social responsibility).
Steps to reproduce
Structural equation modeling (SEM) was used to empirically validate the model via the EQS 6.2 statistical software package utilizing the maximum likelihood estimation method. The modeling consists of two stages: a confirmatory factor analysis test and a structural model or causal test.