Spatio-temporal patterns in citizen feedback: Comparison of postal code areas in the City of Turku

Abstract

This thesis aims to increase knowledge on area-level differences in providing feedback to the local government. Citizens' direct feedback to city officials about public utilities, services, and other issues is a form of democratic participation that is often overlooked. However, participation in co-production activities has previously been shown to be unevenly distributed among population groups. Previously the provision of feedback to the local government has been studied in the contexts of American cities, but to my knowledge no such study has yet been conducted in the Finnish context with a Finnish feedback dataset.

Area-level differences in giving feedback about snow removal is studied using area-level covariates on residents' linguistic makeup, income, education level, owner-occupancy and age. The statistical datasets forming the covariates and meteorological datasets used for data filtering are retrieved using open source software and open methods. The analyses were conducted using the Bayesian INLA framework for spatio-temporal model building.

Results show that feedback provision across different postal code areas is spatially and temporally structured. The results from the models also indicate that areas with higher education and higher proportion of residents who speak other than Finnish or Swedish have higher rates of feedback provision, whereas the effect of distance from city centre is negative for all feedback and positive for feedback provision related to snow ploughs. Further studies are needed to explain individual level variation in feedback provision.