Citizen Coproduction

Abstract

Citizen coproduction, that is, citizens’ input to the provision of public services, holds great potential to improve services provided to citizens. It is therefore important to understand why some citizens are more likely to coproduce than others. Citizens’ skills and knowledge to coproduce are argued to be crucial for their contribution to coproduction, but research on this topic is sparse. Building on coproduction theory supplemented with theoretical insights from social psychology theory, the main contribution of this study is to develop theoretical arguments that describe how self-efficacy perception may moderate the influence of knowledge of how to coproduce on citizen coproduction undertaken by individual citizens. A large-N study in the field of education is used to examine this relation.

Keywords

coproduction knowledge and self-efficacy

Introduction

In the last decade, it has become increasingly common for public managers and government organizations to involve citizens in the provision of public services (Brandsen, Pestoff, & Verschuere, 2012). For example, when patients do physical training after surgery performed by public employees, or when parents help their children with homework, they contribute to the provision of public services that they or their relatives consume. In other words, citizens are coproducers of public services (for similar examples of this type of coproduction, see Jakobsen & Andersen, 2013; Pestoff, 2012). Several studies have shown that citizen participation in coproduction of public services holds great potential to improve services provided to citizens. Benefits include greater quality and quantity of the services provided and more efficient service provision (Brudney, 1983; Jakobsen & Andersen, 2013; Parks et al., 1981; Percy, 1983; Vamstad, 2012). It is therefore important to understand why some citizens are more likely to coproduce than others.

Scholars have argued that the ability and motivation to coproduce are important for citizens’ participation in coproduction of public services (Alford, 2002, 2009; Jakobsen, 2013; Porter, 2012; van Eijk & Steen, 2014). The ability to coproduce has been defined as having the knowledge, skills, and materials necessary to coproduce (Alford, 2009; Brudney, 1983; Jakobsen, 2013; Percy, 1984; Sharp, 1980), and empirical research has shown that citizens’ knowledge and skills to coproduce are important for their contribution to coproduction (Alford, 2002, 2009). Moreover, a recent study found that citizens’ self-efficacy perception to coproduce, that is, their own judgment of their competences to coproduce is related to their contribution to coproduction (Parrado, Van Ryzin, Bovaird, & Löffler, 2013). Although both knowledge and self-efficacy perception seem to be important for citizen coproduction, we have a limited understanding of the circumstances under which these factors are important. Prior research in the coproduction literature has devoted no attention to whether self-efficacy perception and knowledge interact in their influence on citizen coproduction. Such interactive relations may have been overlooked because most prior research relies on case studies and detecting moderator effects often requires large-N studies. However, understanding questions of this sort is highly relevant in terms of guiding local governments on how to increase citizen coproduction.

The main contribution of this study is to develop theoretical arguments that describe how self-efficacy perception may moderate the influence of knowledge of how to coproduce on citizen coproduction undertaken by individual citizens. In developing these arguments, the study builds on coproduction theory supplemented with theoretical insights from social psychology theory (see, for example, Bandura, 1986, 1993). Specifically, the study argues, first, that both knowledge of how to coproduce and self-efficacy perception are positively related to citizen coproduction level, and, second, that the relationship between knowledge of how to coproduce and citizen coproduction level is likely to vary by different strengths of self-efficacy perception. These theoretical arguments are tested using a large-N study on parents’ coproduction of educational services, which focuses on parents’ involvement in developing their children’s reading skills. The empirical results show that parents’ self-efficacy perception is significantly and strongly associated with their level of coproduction of educational services. Moreover, the results confirm that the relationship between knowledge of how to coproduce and parents’ level of coproduction is dependent on the strength of their self-efficacy perception. These results imply that government organizations that attempt to increase parental input to the provision of educational services should not only lift constraints on their knowledge of how to coproduce but also ensure that parents perceive themselves as highly efficacious with regard to coproducing educational services.

The study proceeds as follows. First, the study develops the theoretical arguments concerning how citizens’ knowledge of how to coproduce and self-efficacy perception may interact in their influence on citizen coproduction level. Next, the research design and case are presented and discussed, followed by the empirical results. Finally, the study discusses the results and implications for future research.

Theoretical Background

Various definitions of coproduction have been proposed over time since the development of the coproduction theory in the late 1970s. Parks et al. (1981) defined coproduction as a mixing of the productive efforts, or input, of public service agents and citizens to the provision of public services. The mixing of input may occur through coordinated efforts in the same production process or through independent, but related efforts (Parks et al., 1981). Later, this definition has been expanded by Bovaird (2007) to include volunteers and community groups as coproducers. In line with most research, this study draws on the definition of coproduction outlined by Parks et al. according to which both public service agents and citizens provide input to the provision of public services. However, citizen may contribute to and benefit from coproduction in different ways. This can be illustrated by a typology of coproduction developed by Bovaird, Van Ryzin, Loeffler, and Parrado (2015), who used two criteria to categorize different types of coproduction (see Table 1). The typology only deals with how individuals outside the government agency may contribute to and benefit from coproduction, and input from public service agents is thus implicitly assumed. The first criterion concerns whether input to coproduction by individuals outside the government agency is individually or collectively provided. The second criterion concerns whether the benefit from coproduction in terms of the service received is enjoyed by the individual directly involved or a larger group of individuals who are not necessarily contributing.

Table 1.

Different types of coproduction.

Input to coproduction (from individuals outside the government agency)	Benefits from coproduction (for individuals outside the government agency)
	Individually enjoyed	Collectively enjoyed
Individually provided	A	C
	Private individual coproduction	Philanthropic individual coproduction
Collectively provided	B	D
	Private collective coproduction	Philanthropic collective coproduction

Source. Adapted from Bovaird, Van Ryzin, Loeffler, and Parrado (2015).

Note. The typology does not focus on the input from public services agents.

This study focuses on the type of coproduction outlined in Box A in Table 1. On this basis, coproduction refers to the mixture of input from both public service agents and individual citizens to the provision of a public service that primarily benefits the individual citizen involved or their relatives. Examples of citizen coproduction, as conceptualized in this study, are when parents in addition to teachers provide input to their children’s education by helping with homework or reading to them, or when patients with diabetes do exercises to avoid health problems (for similar examples, see Jakobsen, 2013; Jakobsen & Andersen, 2013; Pestoff, 2012). In the first example, the main beneficiary and the participant are not the same person, while this is the case in the latter example. Although this study focuses on a type of coproduction that mainly entails the creation of private value to an individual citizen or their relatives, it often also entails the creation of public value because of potential positive effects on, for example, service outcomes (Alford, 2009). The theoretical arguments and examples presented in the remainder of this section apply to the type of citizen coproduction defined above. The implication of focusing on one type of citizen coproduction for the generalizability of the results is discussed in the conclusion.

While a number of studies of coproduction have focused on inducing citizens to engage in coproduction, and thus perceive coproduction as a relationship between public service agents and citizens that needs to be designed (Alford, 2002, 2009; Brudney, 1983; Ostrom, 1996), recent studies have argued that you cannot have service delivery without coproduction; in other words, coproduction is an essential and core component of all service delivery (Osborne & Strokosch, 2013; Radnor, Osborne, Kinder, & Mutton, 2014). This implies that instead of focusing on whether citizens engage in coproduction or not, the essential question is how to ensure high levels of citizen coproduction. For instance, while public education in general is coproduced some parents may only provide a minimum of input whereas others involve themselves wholeheartedly . Following this latter conceptualization of citizen coproduction, the remainder of this section seeks to explain variations in citizen coproduction level. In general, this question has only received modest empirical attention among coproduction scholars. Several coproduction scholars have paid theoretical attention to factors such as the ability and motivation to coproduce (Alford, 2002, 2009; Brudney, 1983; Jakobsen, 2013; Percy, 1984; Porter, 2012; Rosentraub & Sharp, 1981; van Eijk & Steen, 2014). The ability of citizens to coproduce has been defined as having the knowledge, skills, and materials necessary to coproduce (Alford, 2009; Brudney, 1983; Jakobsen, 2013; Percy, 1984; Sharp, 1980). It is argued that if citizens know how to coproduce and understand that their input matters, they are more likely to coproduce (Jakobsen & Andersen, 2013; Thomsen & Jakobsen, 2015). An example is that parents who understand the value of reading with their children are more likely to coproduce educational services to their children. Moreover, parents who know how to facilitate the development of their children’s reading skills are more likely to do so and their input is likely to be of higher quality. Similarly, patients who know the right exercises and how to perform them are more likely to coproduce their own health services and provide input of higher quality. Despite the theoretical attention devoted to the importance of the ability to coproduce, there is only little empirical research on this topic. Prior empirical research has mainly relied on case studies and focused on exemplifying how knowledge and skills to coproduce or lack thereof may matter for citizen coproduction (see Alford, 2002, 2009). This brings us to the first hypothesis to be tested:

Hypothesis 1: The individual citizen’s knowledge of how to coproduce influences his or her level of coproduction positively.

Another factor that might explain variations in citizen coproduction level, but has received less theoretical attention, is citizens’ self-efficacy perception to coproduce, that is, their own judgment of their competences to coproduce. A recent study by Parrado and colleagues (2013), which relied on a large-N study, showed that especially perceived self-efficacy, measured by whether citizens feel they can make a difference, is an important predictor in explaining variation in citizen coproduction across countries and services areas (Parrado et al., 2013). While coproduction scholars have underlined the influence of citizens’ knowledge and perceived self-efficacy to explain variations in citizen coproduction level, no empirical studies have taken both factors into account. Before moving to this part, the theoretical argument for why citizens’ self-efficacy perception is important for citizen coproduction level is elaborated. The concept of self-efficacy has its origin in social psychology theory where Bandura argued that it is not only the possession of objective competences such as knowledge and skills that determines what activities a person becomes involved in and how successful the performance is. Another important factor is a person’s judgment of his or her competence to perform a given action (Bandura, 1986). The theoretical reasoning behind perceived self-efficacy as important for becoming an active citizen is, according to Bandura, that it determines, in a specific context, which activities to become involved in, the effort expended, and persistence in the face of obstacles. The stronger the self-efficacy perception, the more likely a person is to become involved in an activity and the greater and more persistent the efforts are (Bandura, 1986). Applied to citizen coproduction, as conceptualized in this study, this implies that parents who feel competent to help their children become good at reading are more likely to coproduce educational services to their children than parents who doubt their own competences. Moreover, in the case of obstacles highly efficacious parents are more likely to be persistent in their efforts than less efficacious parents. Similarly, patients who feel competent to make a difference in their own rehabilitation are more likely to coproduce their own health services than patients who doubt their competences. This brings us to the second hypothesis:

Hypothesis 2: The individual self-efficacy perception of a citizen influences his or her level of coproduction positively.

I will now discuss how citizens’ perceived self-efficacy and knowledge of how to coproduce may interact in their influence on citizen coproduction level. In developing these arguments, I draw on social psychology theory. Specifically, the social psychology theory has argued in a number of studies that persons with the same skills and knowledge may perform a task differently depending on their self-efficacy perception (Bandura, 1986, 1993). Accordingly, the relation between knowledge and action may be contingent on the strength of a person’s self-efficacy perception. However, the social psychology theory does not explicitly state how the influence of knowledge and skills may vary by the strength of the self-efficacy perception. Applying these theoretical insights to citizen coproduction, as conceptualized in this study, we would not necessarily expect that citizens with high perceived self-efficacy and a high level of knowledge of how to coproduce are more likely to coproduce than citizens with a high self-efficacy perception, but less knowledge. The reason is that high self-efficacy perception fosters active involvement along with persistent efforts in case of obstacles. For instance, parents who feel competent to help their children become good at reading and know how to facilitate the development of their children’s reading skills through shared reading are not necessarily likely to read more frequently with their children than highly efficacious parents who do not possess this type of knowledge. Likewise, patients who feel competent to make a difference in their own rehabilitation and know which exercises to perform after surgery are not necessarily more likely to provide more input to their own health services than highly efficacious patients who do not possess this type of knowledge. In contrast, less efficacious citizens are more likely to give up quickly if they face obstacles, and we might expect less efficacious citizens with greater knowledge of how to coproduce to face fewer obstacles than less efficacious citizens with less knowledge. The implication is that knowledge of how to coproduce should have a stronger influence on citizens’ level of coproduction among citizens with low perceived self-efficacy than among citizens with high perceived self- efficacy. Thus, the third hypothesis to be tested is as follows:

Hypothesis 3: The influence of a citizen’s knowledge of how to coproduce on his or her level of coproduction varies by the strength of the self-efficacy perception.

Research Design and Data

To test the hypotheses, I use data from a cross-sectional study on coproduction of educational services, which focuses on parents’ involvement in the development of their children’s reading skills during primary education. Parental participation in coproducing educational services is a highly relevant case for two reasons. First, when parents, for example, read with their children they are coproducing educational services to their children together with public employees, that is, the teachers. Second, educational service is an enduring service, which parents and children are locked into for long periods, and they often cannot rely on exit (Pestoff, 2012). This creates favorable conditions for involving parents in this type of coproduction activity because they gain influence on the services their children consume. Therefore, if the theoretical arguments presented in the previous section can be supported by empirical data, it is expected to be the case in this service area.

The data used in this study are survey data on parental coproduction behavior combined with register data on parental characteristics. The survey data were collected in September and October 2013 among parents of children enrolled in Grades 2 and 3 from 28 primary schools in Aarhus Municipality, Denmark. All survey questions were translated to English, Somali, Arabic, and Turkish to increase the response rate among parents with an immigrant background. The study population consisted of 2,895 families and parents were asked to answer the survey questionnaire either by mail or via a web survey. If email addresses were available for both parents, the link to the web survey was sent to the child’s mother and father. In the few cases where both parents answered the survey, the survey completed first was chosen. If no email addresses were available, a letter with a questionnaire was sent to the child’s home address. The response rate for the survey questions included in this study was 52%. The register data on parental characteristics were obtained through Statistics Denmark and linked to the survey data via the Danish Civil Registration System. The analysis is based on a final sample of 1,419 parents of children nested within 1,386 families. For these parents, there is information on coproduction, knowledge of how to coproduce, self-efficacy perception, and background characteristics.¹

The dependent variable for this study, citizen coproduction level, is measured by asking parents how frequently they read with their child. A similar measure has been used in previous studies of parents’ involvement in coproducing educational services (see, for example, Jakobsen, 2013; Thomsen & Jakobsen, 2015). The first independent variable, knowledge of how to coproduce, was measured by an item capturing parental knowledge of how to read effectively with children. According to prior research, dialogic reading is a very effective method for parents to facilitate their children’s language and reading development. Dialogic reading techniques encompass that parents (a) talk to their children and (b) ask them questions about the book while reading together (Whitehurst et al., 1988). Specifically, parents were asked about what kind of activities they believe improve their child’s language and reading the most, and were offered five possible answers. Parents were given the value 1 if they chose the two activities acknowledged as the most effective in facilitating a child’s language and reading development according to the dialogic reading literature, whereas parents who chose other activities or did not know were given the value 0. A similar item capturing parental knowledge of how to coproduce effectively with their children has been used in one prior study on parental coproduction of educational services (Jakobsen, 2013). Of course, many different items could be applied to measure citizens’ knowledge of how to coproduce, and the item applied here certainly does not capture all aspects of parental knowledge of how to coproduce educational services. However, a large advantage of this item is that it captures the type of knowledge often provided by government initiatives to lift constraints on the ability to coproduce (Jakobsen, 2013; Jakobsen & Andersen, 2013). The second independent variable, individual self-efficacy perception to coproduce, was measured by a question asking parents to rate on a scale from 0 to 10 how certain they were that they could help their child become good at reading. Thus, compared with the knowledge item, which is an objective measure, this item is subjective and captures how parents judge their competence to help their child become good a reading. The more efficacious, the higher the number they would choose. The item was phrased the same way as prior parental self-efficacy questions developed by Bandura (see Bandura, 2006). The exact wording and coding of all survey items is outlined in Appendix A.

Prior to collecting the survey data, all survey items were pilot tested among 30 families with children enrolled in Grades 2 and 3 to ensure that the questions and response categories were clear and that there were variations in the answers. Based on the pilot test, the response categories for the measure on citizen coproduction were further specified to prevent leeway in the interpretation of the response categories.² Although the survey questions were pilot tested, a concern in relation to the question about how often parents read with their child is that it may be subject to social desirability, and in particular among highly educated parents (Hofferth, 2006). I try to mitigate this problem by including a control variable for the respondent’s educational attainment, which is further described below. However, this approach may not be enough to address the problem with social desirability, which is discussed further in the concluding section.

Finally, the study controls for a number of child and family characteristics that may influence parental coproduction behavior and knowledge of how to coproduce or perceived self-efficacy. These are register data from Statistics Denmark, and mainly from 2012. Variables for family characteristics include information on the respondents’ educational attainment, gross income, employment status, age (years), gender, origin, and number of children in the household. Prior studies of coproduction have shown that especially education and income levels are positively associated with citizen coproduction level and the ability to coproduce (Rosentraub & Sharp, 1981; Warren, Rosentraub, & Harlow, 1984). The study controls for employment status, age, and gender of the parent, as studies of parental time spent with children indicate that employed, older parents as well as mothers spend more time on educational activities with their children than unemployed, younger parents and fathers (Bonke & Esping-Andersen, 2011; Guryan, Hurst, & Kearney, 2008). Moreover, the study controls for number of children in the household, as studies of parental time with children have shown that the number of children is negatively related to parental time spent with children (Craig & Mullan, 2012). Last, the study controls for the following child characteristics: gender, age, grade level (2nd or 3rd grade), and whether the child lives with both parents or a single parent. These variables are included as they are also expected to be related to parental time with children. For instance, prior studies have shown that parents spend more time with younger than with older children (Bonke, 2009; Stafford & Yeung, 2005). Descriptive statistics on all variables are shown in Table B1 in Appendix B. All analyses presented in the following section are conducted using ordered or binary logistic regression combined with clustered robust standard error at the family level. As the sign and significance test of interaction effects may be misleading when we use logistic regression (Ai & Norton, 2003; Greene, 2010), all interaction models have also been estimated using ordinary least squares (OLS) regression. An OLS regression yields results similar to those presented in the following section.

Empirical Results

The two hypotheses, that citizens’ knowledge of how to coproduce and perceived self-efficacy influence their level of coproduction positively, are tested in Models 1 to 4 in Table 2. The reported coefficients are logits. Looking at the results for the knowledge item, we only identify a significant relationship between knowledge of how to coproduce and parental input in the bivariate analysis in Model 1. When we include self-efficacy and family characteristics in Model 4, the relationship becomes insignificant. In contrast, the relationship between perceived self-efficacy and parental input is positive and highly significant at the .001 level in Models 2 to 4. Thus, the empirical results only lend support to the second hypothesis concerning perceived self-efficacy being positively related to citizens’ level of coproduction.

Table 2.

Ordered logistic regression analysis for the association between knowledge, perceived self-efficacy, and parental input.

	Model 1	Model 2	Model 3	Model 4	Model 5
Knowledge, high	0.20^† (0.10)		0.14 (0.10)	0.00 (0.11)	2.06** (0.68)
Self-efficacy		0.28*** (0.04)	0.28*** (0.04)	0.25*** (0.04)	0.37*** (0.05)
Knowledge × Self-Efficacy					−0.24** (0.08)
Education (Reference: Lower secondary)
Upper secondary education				0.17 (0.24)	0.12 (0.24)
Higher education, short-cycle				0.27 (0.30)	0.22 (0.30)
Higher education, medium-cycle				0.39 (0.25)	0.35 (0.25)
Higher education, long-cycle				0.30 (0.26)	0.29 (0.26)
Employment (Reference: Employed)
Unemployed				0.28 (0.36)	0.23 (0.36)
Student/retired				0.07 (0.19)	0.06 (0.19)
Income (10,000 DDK)				0.00 (0.00)	0.00 (0.00)
Number of children				−0.12* (0.05)	−0.14** (0.05)
Age of parent				0.01 (0.01)	0.01 (0.01)
Non-Western origin, parent				−0.51* (0.20)	−0.52* (0.20)
Responding parent, mother				0.15 (0.13)	0.14 (0.13)
Boy				0.26* (0.10)	0.27** (0.10)
Age of child				−0.18^† (0.10)	−0.19^† (0.10)
Second grade				0.67*** (0.14)	0.66*** (0.14)
Living with (Reference: Parent in new relationship)
Both parents				0.52^† (0.27)	0.52^† (0.27)
Single parent				0.46 (0.30)	0.46 (0.30)
Pesudo-R²	.00	.02	.02	.05	.05
N	1,419	1,419	1,419	1,419	1,419

Note. The reported coefficients are logits. Dependent variable: every day, 4-5 times per week, 2-3 times per week, once a week, once a month, and never. Standard errors in parentheses.

†

p < .10. *p < .05. **p < .01. ***p < .001 (two-tailed tests).

To get an idea about the strength of the association between self-efficacy perception and parental input to coproduction, predicted probabilities for different strengths of self-efficacy perception were calculated for an average parent. This means that all continuous variables are held at their mean, whereas ordinal variables are held at their median and nominal variables are held at their mode. Calculations show that the predicted probability of answering “every day” is .32 for parents with the lowest perceived self-efficacy (10th percentile) and .50 for parents with the highest perceived self-efficacy (90th percentile), which is a fairly strong difference (significant at the .05 level).³ Thus, the association between self-efficacy and parental input is not only significant but also fairly strong.

The third hypothesis stating that the relationship between knowledge of how to coproduce and citizens’ level of coproduction is dependent on the strength of self-efficacy perception is tested by including an interaction term between knowledge of how to coproduce and self-efficacy perception in Model 5 in Table 2. The interaction term in Model 5 is negative and significant at the .01 level, which suggests that the association between knowledge of how to coproduce and parental input to coproduction decreases with the strength of perceived self-efficacy. To convey a better understanding of this moderator effect, Figure 1 presents differences in the predicted probabilities of providing a high level of input (answering: “every day” or “4-5 times per week”) between highly and less knowledgeable parents at different strengths of perceived self-efficacy. The predicted probabilities are calculated for an average parent.

Figure 1.

Interaction between knowledge and perceived self-efficacy in the prediction of parental input.^a

Whereas the results in Table 2 were based on an ordered logit model, the predicted probabilities in Figure 1 are based on a binary logit model with a binary dependent variable, which captures whether a parent provides either a low or a high level of input.⁴ Figure 1 shows that among parents with a low self-efficacy score, the knowledgeable parents are significantly more likely than less knowledgeable parents to provide a high level of input. Moreover, this difference is much larger for parents with a self-efficacy score of 5 than parents with a score of 7. With regard to the strength of the association, Figure 1 shows that the relation between knowledge and parental input for parents with a low self-efficacy score is fairly strong. For example, the difference in the predicted probability between highly and less knowledgeable parents, but a self-efficacy score of 5, is .27. Finally, additional results (not presented) reveal that parents with the highest level of perceived self-efficacy, whether they are knowledgeable or not, are the most likely to provide a high level of input (corresponding to a predicted probability between .82 and .89). As mentioned, the predicted probabilities presented in Figure 1 are based on an average parent, which means that all continuous variables are held at their mean, ordinal variables are held at their median and nominal variables are held at their mode. As a result, if the values for the control variables are changed, Figure 1 is likely to depict another graph. However, robustness checks using an average marginal model (AME), holding all control variables at their mean, show similar results to those in Figure 1. Thus, the empirical results lend support to the third hypothesis stating that the relationship between knowledge of how to coproduce and citizens’ level of coproduction is dependent on the strength of their self-efficacy perception.

In summary, the empirical findings show that perceived self-efficacy, in contrast to knowledge of how to coproduce, is significantly and strongly associated with parental input to coproduction. Moreover, the analysis confirms that the relationship between knowledge of how to coproduce and parental input to coproduction is dependent on the strength of self-efficacy perception. Specifically, knowledge of how to coproduce was more strongly related to parental input among less efficacious parents than among highly efficacious parents. Additional analyses (not presented) show that the results presented in Table 2 are robust when two other measures, also asking parents to judge their competence in relation to facilitating their children’s language and reading development, are applied.⁵ Moreover, the results are similar to those in Table 2, when parental input to coproduction is measured by how frequently parents help their child with homework.

Discussion and Conclusion

The question of why some citizens are more likely to coproduce has received modest empirical attention among scholars in the coproduction literature. The few studies that have addressed this topic have theoretically argued that factors such as the ability and motivation to coproduce are important for whether citizens participate in coproduction of public services (Alford, 2009; Brudney, 1983; Parks et al., 1981; Percy, 1983; Rosentraub & Sharp, 1981; van Eijk & Steen, 2014; Warren et al., 1984). The main contribution of this study was to develop theoretical arguments that describe how self-efficacy perception may moderate the influence of knowledge of how to coproduce on citizen coproduction undertaken by individual citizens. It was argued, first, that both knowledge of how to coproduce and self-efficacy perception are positively related to citizen coproduction level, and, second, that the relationship between knowledge of how to coproduce and citizen coproduction level is likely to vary by different strengths of self-efficacy perception. These relationships were tested using data from a large-N study on coproduction of educational services, which focused on parents’ involvement in the development of their children’s reading skills.

The empirical analysis supports the hypothesis that perceived self-efficacy is positively associated with citizen coproduction level. Next, the empirical analysis found empirical support for the hypothesis that the relationship between knowledge of how to coproduce and citizen coproduction level varies by different strengths of self-efficacy perception. Specifically, knowledge of how to coproduce was found to be more strongly related to parental input to coproduction among less efficacious parents than among highly efficacious parents. In fact, an additional analysis for an average parent with high perceived self-efficacy showed no substantial relation between knowledge of how to read effectively with children and parental input to coproduction. A likely explanation, as argued by Bandura (1986), is that highly efficacious persons are very likely to engage in activities and be persistent in the face of obstacles. In contrast, for an average parent with low perceived self-efficacy, the relation between knowledge of how to read effectively with children and parental input to coproduction was fairly strong. A likely explanation is that less efficacious persons are more likely to give up if they face obstacles (see Bandura, 1986).

There are three limitations to the empirical results that need to be discussed. First, as the hypotheses were tested in the field of education, the results are mainly generalizable to this service area. However, the theoretical arguments and examples are not restricted to this area. To better understand how knowledge of how to coproduce and self-efficacy perception are related to citizen coproduction level and may interact in their influence, it is highly relevant for future studies to replicate the results in other service areas. Second, the theoretical arguments presented in this study apply to the type of citizen coproduction that involves input from an individual citizen to the provision of public services that benefit the individual involved or their relatives. Thus, the empirical results may not be generalizable to other types of coproduction among parents with children enrolled in primary education (cf. Bovaird et al., 2015). As far as explaining variation in citizen input to collective types of coproduction, one could imagine that collective self-efficacy perception, that is, a group of citizens’ shared judgment of its competences may be more important than individual self-efficacy perception. Questions of this sort should be discussed theoretically and examined empirically in future studies. Third, the empirical analysis relied on cross-sectional data where the independent and dependent variables were measured simultaneously, which may bias the empirical results. For instance, if some parents tend to overstate both their input to coproduction and perceived self-efficacy, social desirability may be a confounding variable and the estimated association between these two variables will be upward-biased (Jakobsen & Jensen, 2015). The empirical results therefore cannot be interpreted as causal effects, and future research is encouraged to use panel data or field experiments that are better able to handle this potential bias.

Nevertheless, this study can still be regarded as an important contribution to the coproduction literature, as it is the first study to provide theoretical insight into understanding how knowledge of how to coproduce and self-efficacy perception may interact in their influence on coproduction undertaken by individual citizens. The implication of the empirical results is that government organizations that attempt to increase parental input to the provision of educational services that primarily benefit their children should not only pay attention to lifting constraints on their knowledge of how to coproduce. It is also highly relevant to ensure that parents perceive themselves as highly efficacious with regard to coproducing a given educational service (see also Parrado et al., 2013). While this study showed that parents who are highly efficacious tend to coproduce very frequently, whether they are knowledgeable or not, it remains to be examined how these factors relate to and interact in their influence on the quality of citizen coproduction. One could, for example, argue that parents who both possess knowledge and are highly efficacious will provide the most stimulating reading environment. If these arguments can be empirically supported, it underlines that government organizations need to develop public initiatives that target both knowledge of how to coproduce and self-efficacy perception. However, questions of this sort remain to be examined in future studies.

Footnotes

Appendix A

Appendix B

Acknowledgements

I would like to thank Simon Calmar Andersen, Søren Serritzlew, and four anonymous reviewers for their helpful comments on previous versions of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research for this project was partly funded by TrygFonden’s Centre for Child Research.

Notes

Author Biography

Mette Kjærgaard Thomsen is an assistant professor at the Department of Political Science, Aarhus University. Her research interests include coproduction of public services, the effect of government initiatives on citizen coproduction and parental time investment in children.

References

Norton

E. C.

(2003). Interaction terms in logit and probit models. Economics Letters, 80, 123-129.

Alford

(2002). Why do public-sector clients coproduce? Toward a contingency theory. Administration & Society, 34, 32-56.

Alford

(2009). Engaging public sector clients: From service-delivery to co-production. Basingstoke, UK: Palgrave Macmillan.

Bandura

(1986). Self-efficacy. In Bandura

(Ed.), Social foundations of thought and action: A social cognitive theory (pp. 390-453). Englewood Cliffs, NJ: Prentice Hall.

Bandura

(1993). Perceived self-efficacy in cognitive development and functioning. Educational Psychologist, 28, 117-148.

Bandura

(2006). Guide for constructing self-efficacy scales. In Pajares

Urdan

(Eds.), Self-efficacy beliefs of adolescents (pp. 307-337). Greenwich, CT: Information Age Publishing.

Bonke

(2009). Forældres brug af tid og penge på deres børn [Parental use of time and money on their children]. Copenhagen, Denmark: Rockwool Fondens Forskningsenhed, Syddansk Universitetsforlag.

Bonke

Esping-Andersen

(2011). Family investments in children: Productivities, preferences, and parental child care. European Sociological Review, 27, 43-55.

Bovaird

(2007). Beyond engagement and participation: User and community coproduction of public services. Public Administration Review, 67, 846-860.

10.

Bovaird

Van Ryzin

G. G.

Loeffler

Parrado

(2015). Activating citizens to participate in collective co-production of public services. Journal of Social Policy, 44, 1-25.

11.

Brandsen

Pestoff

Verschuere

(2012). Co-production as a maturing concept. In Pestoff

Brandsen

Verschuere

(Eds.), New public governance, the third sector, and co-production (pp. 1-9). New York, NY: Routledge.

12.

Brudney

J. L.

(1983). The evaluation of coproduction programs. Policy Studies Journal, 12, 376-385.

13.

Cameron

A. C.

Trivedi

P. K.

(2010). Microeconometrics using stata, revised edition. College Station, TX: Stata Press.

14.

Craig

Mullan

(2012). Lone and partnered mothers’ childcare time within context in four countries. European Sociological Review, 28, 512-526.

15.

Greene

(2010). Testing hypotheses about interaction terms in nonlinear models. Economics Letters, 107, 291-296.

16.

Guryan

Hurst

Kearney

(2008). Parental education and parental time with children. The Journal of Economic Perspectives, 22(3), 23-46.

17.

Hofferth

(2006). Response bias in a popular indicator of reading to children. Sociological Methodology, 36, 301-315.

18.

Jakobsen

(2013). Can government initiatives increase citizen coproduction? Results of a randomized field experiment. Journal of Public Administration Research and Theory, 23, 27-54.

19.

Jakobsen

Andersen

S. C.

(2013). Coproduction and equity in public service delivery. Public Administration Review, 73, 704-713.

20.

Jakobsen

Jensen

(2015). Common method bias in public management studies. International Public Management Journal, 18, 3-30.

21.

Osborne

S. P.

Strokosch

(2013). It takes two to tango? Understanding the co-production of public services by integrating the services management and public administration perspective. British Journal of Management, 24, S31-S47.

22.

Ostrom

(1996). Crossing the great divide: Coproduction, synergy, and development. World Development, 24, 1073-1087.

23.

Parks

R. B.

Baker

Kiser

Oakerson

Ostrom

. . . Wilson

(1981). Consumers as co-producers of public-services: Some economic and institutional considerations. Policy Studies Journal, 9, 1001-1011.

24.

Parrado

Van Ryzin

G. G.

Bovaird

Löffler

(2013). Correlates of co-production: Evidence from a five-nation survey of citizens. International Public Management Journal, 16, 85-112.

25.

Percy

S. L.

(1983). Citizen coproduction: Prospects for improving service delivery. Journal of Urban Affairs, 5, 203-210.

26.

Percy

S. L.

(1984). Citizen participation in the coproduction of urban services. Urban Affairs Review, 19, 431-446.

27.

Pestoff

(2012). Co-production and third sector social services in Europe. In Pestoff

Brandsen

Verschuere

(Eds.), New public governance, the third sector, and co-production (pp. 13-34). New York, NY: Routledge.

28.

Porter

D. O.

(2012). Co-production and network structures in public education. In Pestoff

Brandsen

Verschuere

(Eds.), New public governance, the third sector, and co-production (pp. 145-168). New York, NY: Routledge.

29.

Radnor

Osborne

S. P.

Kinder

Mutton

(2014). Operationalizing co-production in public services delivery: The contribution of service blueprinting. Public Management Review, 16, 402-423.

30.

Rosentraub

M. S.

Sharp

E. B.

(1981). Consumers as producers of social services: Coproduction and the level of social services. Southern Review of Public Administration, 4, 502-539.

31.

Sharp

E. B.

(1980). Toward a new understanding of urban services and citizen participation: The coproduction concept. The American Review of Public Administration, 14, 105-118.

32.

Stafford

Yeung

J. W.

(2005). The distribution of children’s developmental resources. In Hamermesh

D. S.

Pfann

G. A.

(Eds.), The economics of time use (pp. 289-313). Amsterdam, The Netherlands: Elsevier.

33.

Thomsen

M. K.

Jakobsen

(2015). Influencing citizen coproduction by sending encouragement and advice: A field experiment. International Public Management Journal, 18, 286-303.

34.

Vamstad

(2012). Co-production and service quality: A new perspective for the Swedish welfare state. In Pestoff

Brandsen

Verschuere

(Eds.), New public governance, the third sector, and co-production (pp. 297-316). New York, NY: Routledge.

35.

van Eijk

C. J. A.

Steen

T. P. S.

(2014). Why people co-produce: Analysing citizens’ perceptions on co-planning engagement in health care services. Public Management Review, 16, 358-382.

36.

Warren

Rosentraub

M. S.

Harlow

K. S.

(1984). Coproduction, equity, and the distribution of safety. Urban Affairs Review, 19, 447-464.

37.

Whitehurst

G. J.

Falco

F. L.

Lonigan

C. J.

Fischel

J. E.

DeBaryshe

Valdez-Menchaca

M. C.

Caulfield

(1988). Accelerating language development through picture book reading. Developmental Psychology, 24, 552-559.