Encouraging Citizen Coproduction: Does Communication Style and Message Sender Make any Difference?

The outcomes of public services, such as education, health care, and safety, are improved considerably when citizens contribute with input to the public service production; that is when citizens coproduce. Public organizations... [ view full abstract ]

The outcomes of public services, such as education, health care, and safety, are improved considerably when citizens contribute with input to the public service production; that is when citizens coproduce. Public organizations are therefore seeking different ways to enhance citizen input. A commonly used strategy to increase citizen coproduction is to provide citizens with information that encourages citizens to coproduce more. Yet empirical studies estimating the effect of such strategies to enhance citizen coproduction remain scarce. Surprisingly, one of few exceptions suggests that a simple information strategy consisting of providing parents with information material in the form of a small leaflet may have negative effects on parents’ motivation to coproduce (Jakobsen & Thomsen, 2015). Drawing on insights from psychology on human’s motivational responses to external events we argue that information strategies by public organizations may to varying degrees be experienced as controlling by the recipients depending on the communication style employed – and thereby potentially crowd out citizens’ motivation to coproduce. Moreover, citizens may have different confidence in the message sender depending on whether the sender is a distant contact or not. This may also have implications for citizens’ experience of the information strategy as controlling and thus their motivation to coproduce. We test these arguments in the area of education focusing on parents’ coproduction of educational services in the form of reading with their children. First, we employ a field experiment (n>2,100) to examine the effect of providing information materials to parents that is communicated in either a more or a less controlling style. Second, we employ a survey experiment (n>800) in the same setting to test the effect of the message sender, that is, does it makes any difference whether the sender of information material is a distant contact (the municipality) or not (child’s teacher).