Association for Public Policy Analysis & Management

In recent years, government agencies have undergone substantial reforms aiming to make the provision of public services more business-like. This includes treating service users as customers rather than citizens. However, the unintended behavioral consequences of the way how government agencies treat service users has not been addressed to date, neither by scholars nor by government practitioners.

The goal of this study is to test whether a reform-induced shift of people’s role identities towards public service delivery affects their willingness to coproduce public goods. This expectation is derived from identity labeling theory which assumes that addressing public service users as citizens (or customers) will lead them to act in accordance to the traits assigned to this identity (relative to not addressing them with their respective identities). This means that self-identified citizens (in a civic-republican sense) who will be addressed as such, will be more likely to contribute to the provision of local public goods – in line with citizenship duties inherent to this identity. Self-identified customers, however, will be less likely to do so, which is in line with market-based individual norms of free-riding in the provision of public goods. In this sense, treating service users as customers may crowd-out their willingness to contribute to the provision of public services.

We test these expectations using a naturalistic survey experiment among demographically representative sample New Jersey residents. A total of 1,000 randomly selected residents will be contacted via phone and asked to assess their self-perceived role identity toward a government agency. Based on their answers, respondents are grouped into three survey arms (#1: self-identified citizens; #2: self-identified customers; #3: others) where they are asked whether they would provide their email address to volunteer with the New Jersey Office of Volunteerism. This is a much used measure of people’s revealed citizenship behavior. Simply asking them whether they would volunteer might be prone to social desirability bias, with respondents over-reporting their willingness. Asking respondents to provide their email address, however, is an observable behavior which comes much closer to their actual citizenship behavior.

In the study arms for self-identified citizens and customers, people are randomly addressed with their respective identity (i.e., the experimental manipulation) so that the effect of labeling them with their respective role identity can be compared to a control group where this has not been mentioned. Arm #3 does only provide the email request, and serves as a baseline comparison group; it also helps us to assess the average citizenship behavior for respondents in the population of NJ residents and whether perceiving and communicating with service users in a certain way does increase their willingness to coproduce public goods.

This study has important implications in how changes in the way government agencies address and perceive their clients affect people’s subsequent willingness to coproduce important public services which are typically provided in partnership with public agencies.