Recent years have seen an upsurge in the use of citizen satisfaction evaluations of governmental services in both academia and government itself. Yet, there remains inconsistency in our understanding of what satisfaction evaluations actually represent. We have recently witnessed a move away from performance-based models to cognitive-implicit models of citizen satisfaction (Andersen & Larsen 2016; Van Ryzin 2004). Yet, the inherent political nature of local government, its institutions, and services remains largely ignored. Indeed, municipal services are not delivered in a political vacuum. They are connected to local governments and political principals via a chain of functional responsibility. This paper addresses this shortcoming by testing a theory of citizen satisfaction that accounts for the political nature of municipal services.
In line with theories of partisan motivated reasoning (Baekgaard et al. 2017; James & Van Ryzin 2016; Jilke 2017), we argue that citizen satisfaction evaluations are partisan colored. By this, we mean that partisans are more likely to be satisfied with municipal services when the responsibility for these services is vested in political principals from the same political spectrum. Thus, the resulting partisan bias may hinder unbiased evaluations of local services. However, drawing upon the literature on the clarity of responsibility of political actors (Hobolt, Tilley & Banducci 2013; Rudolph 2006), we argue that the partisan bias in citizen satisfaction comes only into being when the line of responsibility between political principals and service delivering institutions is clear. In contrast, when this line of accountability is dispersed across multiple actors, citizens become less likely to use their partisan lenses when evaluating the quality of municipal services.
To test the predictions of a political theory of citizen satisfaction, we conduct a two-wave panel survey of voters in municipal elections. The survey will be conducted 2 weeks before and 6 weeks after the Danish local elections, which will be held on November 21 2017. We will employ a Difference-in-Difference estimator to identify the effects of partisan match between voters and elected municipal leaders to assess whether changes in shared partisanship with municipal leaders affects changes in citizen satisfaction evaluations of local services. In other words, we expect that a change in partisan match would either increase or decrease satisfaction with local services. In addition to the quasi-experimental design outlined above, we will randomly allocate voters into three experimental conditions. Using a between-subjects design, participants will be randomly allocated to three conditions containing experimental-condition-specific, but realistic, information prior to answering satisfaction items in both waves. Condition 1 will include a direct mentioning of the functional responsibility of the mayor for these services (i.e., a high clarity of responsibility). Condition 2 will complement this information with the mentioning of all public bodies who are involved in the delivery of these services, thereby blurring responsibility across multiple actors (i.e., a high institutional dispersion). The third condition serves as a control group, where no such information is provided.