The clear model as a predictor of candidacy for council membership

ABSTRACT

This paper tests to what extent the dimensionof the CLEAR model can be used as reliable predictors of an individual’s decision to pursue council membership. By using the CLEAR framework as an extension of known frameworks such as self-determination theory and public service motivation that only focus on motivation, we develop a more comprehensive view of an individual’s decision to run for candidacy. Our research shows that people do not distinguish between the ‘can do’ and ‘like to’ dimensions. Rather, these two dimensions must be considered as three sub-dimensions in our research coined as amotivation, intrinsic motivation, and extrinsic motivation. Furthermore, in explaining why people tend to decide to run for council membership, only amotivation, intrinsic motivation and social support (enabled dimension) appeared to be reliable predictors. To further our analysis, we have analysed how these different dimensions differ across different societal groups.

KEYWORDS:

• CLEAR framework
• council membership

1. Introduction

Fewer and fewer citizens seem to be interested in a position as a council member in local councils (Kjaer 2007). This is also the case in the Netherlands, where the interest in council membership is declining and local political parties are having trouble finding suitable candidates (Denters 2012). Fewer people are members of political parties, and political parties therefore have fewer options to recruit competent candidates (Boogers, De Jong, and Voerman 2018). In continental Europe, this trend has been apparent for a long time now (Aars and Offerdal 1998) and can be understood as part of the broader decline of interest in participating in democratic and political structures (Koestner et al. 1996; Kjaer 2007).

Perhaps surprisingly, what determines people’s decision to apply for council membership has received little academic attention in continental Europe (see for notable exceptions Pedersen, Pedersen, and Bhatti 2018; Pedersen 2014). When authors do focus on (potential) council members, studies involve the considerations of current council members (Ritz 2015) or focus on the considerations for recruitment (Aars and Offerdal 1998; Broockman 2014). Studies about motivations for a public cause are mainly focused on the motivations of civil servants (i.e., public service motivation literature, e.g., Perry 1996; Vandenabeele 2008). Therefore, what considerations people have that constitute the decision to run for council membership is largely unknown. Furthermore, we do not understand how to increase an individual’s willingness to apply to be a council member. We argue that if we need to make our city councils (more) equal representations of our society, it is imperative to increase our understanding of these considerations among people stemming from different societal backgrounds In order to fill this gap in the literature, in this research, we test to what extent the CLEAR-model as developed by Lowndes, Pratchett, and Stoker (2006) offers a useful framework to explain what constitutes people’s decision to run for candidacy and we examine how the CLEAR dimensions vary across people stemming from different societal backgrounds. As the CLEAR-model has only been marginally empirically and quantitatively tested, and as many of the dimensions of the CLEAR model correspond to personal motivation, we first need to understand how the CLEAR-model relates to frameworks of personal motivation i.e., self-determination theory – SDT and public service motivation – PSM. In doing so, this study follows the example of Fox and Lawless (2005), who argued that ‘in order to understand fully the decision dynamics involved in moving from “eligible potential candidate” to “actual office holder,” it is necessary to step back and assess nascent ambition – or the inclination to consider a candidacy’ (Fox and Lawless 2005, 644).

In order to do so, we conducted a panel survey of 1,650 Dutch individuals – a representative sample of the Dutch population. The questionnaire used in the survey combined several dimensions from SDT and PSM and the CLEAR framework as developed by Lowndes, Pratchett, and Stoker (2006)

This paper makes four contributions: 1) we extend our understanding of what determines an individual’s decisions to consider council membership as a possible option, 2) we show how different determinants have different effects across different societal groups, 3) we show to what extent well-known theoretical frameworks such as SDT and PSM have value in predicting the chance that an individual will apply, and 4) we test the CLEAR framework on a large quantitative scale and its value to explain what determines people’s willingness to run for council membership.

The remainder of this paper is structured as follows: in the next section, we elaborate on how the different theoretical frameworks of SDT, PSM, and CLEAR relate to each other, and we present an integrated theoretical framework to test in our studies. In Section 3, we assess the research methods and strategies used to accomplish our research objective. In Section 4, we present the results of our analysis by showing that amotivation, intrinsic motivation and social support appeared to be reliable predictors of people’s decision to apply for council membership. Furthermore, we show that these determinants differ across different societal groups. Section 5 elaborates on the implications of these results for our understanding of what determines people’s decision to apply for council membership. The last section (6) offers a conclusion and a reflection on the validity of our research methods.

2. Theoretical framework

Running for a public office is considered fundamentally different from acquiring a position in a private sector context. In the Public Service Motivation literature, it is argued that the motivation of public servants refers to an ‘altruistic motivation to serve the interests of a community of people, a state, a nation, or humanity’ (Rainey and Steinbauer 1999 , 23). Although PSM refers to motivation of public servants, motivation for council membership is often defined along similar lines (Pedersen 2014; Ritz 2015), as well as aiming for re-election (Ritz 2015) and pro-social motivation (Benabou and Tirole, 2006). However, what constitutes one’s decision to run for council membership is likely to be formed by more than just personal motivations. Elements that can be derived from other branches of literature such as the idea of contributing to a larger cause (i.e., self-efficacy Madsen 1987); social capital (Coleman 1988); or whether someone fits the profile according to his/herself (see for instance Ladam, Harden, and Windett 2018 who examined whether women inspire other women to pursue political careers) may be highly influential to that decision as well. In order to understand this decision we need to go beyond a focus on internal and external motivation and pay attention to other aspects as well.

The CLEAR model as developed by Lowndes, Pratchett, and Stoker (2006) offers a starting point of such a comprehensive framework. CLEAR stands for people participating when they can, when they like, when they are enabled by a social and institutional infrastructure, when they are asked to participate, and when they have the feeling that collaborative partners are responsive to their input (Lowndes, Pratchett, and Stoker 2006, 281). This model was initially developed to understand why citizens participate in their communities, including both motivational aspects (e.g., ‘can do’ and ‘like to’) as well as more social aspects, such as their surrounding network. Adopting the CLEAR framework to explain considerations for council membership requires translation of its main elements to the context of political participation. It necessitates broadening of the framework in order to do justice to the extended literature on motivations to contribute to the public good. We therefore integrate insights from SDT and PSM into the CLEAR framework. In this section, for each dimension of the CLEAR model we elaborate on how it can be defined and reformulated to be suitable for our research and how the dimension can be further refined with aspects of SDT and PSM.

Can do

The ‘can do’ dimension refers to the question of whether people believe they possess the appropriate skills and resources to participate (ibid; 286). This is closely related to the ‘resource-based view of political participation,’ which indicates that the availability of sufficient time, money, and civic skills is crucial for engagement in political activities (Fox and Lawless 2005). Lowndes, Pratchett, and Stoker (2006) have mentioned that participatory skills, such as speaking in public or having the capacity to organise events or gather many people, are usually found among people with higher social-economic status, who are better educated and employed. In SDT, this dimension of ‘can do’ corresponds to what Ryan and Deci label introjected regulation (2000a, 2000b). Classically, this refers to the urge to demonstrate one’s ability (or avoid failure) in order to maintain feelings of worth. Gagne et al. (2015) conceptualised this dimension by focusing on the extent to which self-esteem, pride, and the feeling of shame or inability explain why they put effort into their job. In this research, we use this can do dimension to get more insight into people’s considerations for council membership.

Like to

Lowndes, Pratchett, and Stoker (2006) have argued that the ‘like-to’ dimension – the feeling of being part of a community – determines whether an individual is willing to participate. However, we argue that the answer to the question ‘Would you like to participate?’ is based on a far wider spectrum of motivations than just the feeling of belonging to a certain community, as Lowndes, Pratchett, and Stoker (2006) have argued. Again, looking at SDT, a range of dimensions and concepts may help us to understand why people like to participate or, in Ryan and Deci’s (2000a) terms, do their work.

First of all, amotivation. Amotivation refers to the state of lacking the willingness to act. This can refer to not valuing an activity or not expecting it to yield a beneficial outcome (Ryan and Deci 2000a , 72). Intrinsic motivation refers to doing something because it is inherently interesting or enjoyable (ibid). Ryan and Deci have preserved the term ‘intrinsic motivation’ exclusively for the joy of doing one’s work, because the work itself is enjoyable. Most aspects of self-determined motivation do not refer to intrinsic motivation but rather to how individuals internalise external motivators (Ryan and Deci 2000b). Extrinsic motivation refers to an activity that generates a separate outcome rather than the activity itself. In the research of Fox and Lawless (2005), this is referred to as the competitive traits that people can have, namely career ambitions or material ambitions.

As Figure 1 depicts, there is great variation in extrinsic motivators. For example, an individual doing his or her homework because the individual believes it is necessary for his or her future career or because the individual’s parents tell him or her to do it are both extrinsic motivators. However, the former refers to a personal choice (synthesis of goal congruence) while the latter refers to compliance with an external regulator (salience of extrinsic rewards or punishments). Four types of extrinsic motivation form the middle of this continuum, namely external regulation, introjected regulation, identified regulation, and integrated regulation (see Figure 1). Externally regulated refers to behaviours that are performed to satisfy an external demand or reward (such as the homework to earn the parents’ favour). We have already briefly discussed introjected regulation within the ‘can do’ dimension. Here we can add that although the motivation is intrinsic, the behaviour has an external locus of causality. This means that the reason the individual feels motivated is outside of the individual (e.g., pressure from a relative). Regulation through identification is a conscious valuing of a behavioural goal so that the effort is accepted as personally important. It is different from intrinsic motivation in the sense that it is not the activity itself that is perceived as desirable, but rather the outcome the activity generates (e.g., donating money to a charity). Integrated regulation refers to the complete assimilation of identified regulations. This means that the activity is closer to what one might call ‘vocation.’ The activity in itself may be unenjoyable (e.g., cleaning a toilet), but since it prevents infant mortality in third world countries, people are more willing to do the job. This also means that these activities are in line with one’s values and needs, but they remain extrinsic, because the focus remains on an external outcome rather than on the inherent joy of doing them.

Figure 1. Self-determination theory framework (Ryan and Deci 2000a).

Since we are interested in people’s willingness to participate in the public domain, with the possible association with vocation and mission it is also useful to pay attention to PSM within this dimension. PSM is rooted in the idea that the motivation of public officials is different from that of people working in the private sector, or as Perry (1996) put it, ‘an individual’s predisposition to respond to motives grounded primarily or uniquely in public institutions and organisations’ (p. 6). Vandenabeele (2007) specified Perry’s definition in ‘beliefs, values and attitudes that go beyond self-interest and organisational interest, that concern the interest of a larger political entity’ (p. 547). In this sense, being a council member could arguably be considered a form of PSM. Perry’s (1996) initial framework consisted of four dimensions, namely attraction to policy making, commitment to public interest, compassion, and self-sacrifice. However, Vandenabeele (2008) has convincingly argued that Perry’s original model does not fit the conception of PSM in a European context (i.e., the values of the original four-dimension model do not approach the threshold values for assessing fit; p. 1094). Therefore, for a European context Vandenabeele ‘translated’ Perry’s model into a three-dimension model in which the dimensions public interest and civic duty and self-sacrifice formed one dimension (PSM I) and the dimensions compassion (PSM II) and attraction to policy making and politics (PSM III) were retained. Ritz (2015) applied a reformulated version of the PSM framework to study the motivations of current council members in Switzerland. Although interesting, this applied version is not of much use for our research on understanding why potential members might be motivated to run for council membership since it focuses on how council members enjoy their current job and position.

Connecting SDT to PSM, we argue that conceptually PSM needs to be distinguished from intrinsic motivation as conceptualised by Ryan and Deci (2000a) since intrinsic motivation refers to the pure joy of performing a certain task. PSM is about pro-social behaviour and assigning meaning and purpose to a certain action. Therefore, PSM can be considered a form of internalised extrinsic motivation (Wilkesmann and Schmid 2014, 9) – that is, doing good in the service of other people and society. This corresponds to the fourth dimension of SDT: integrated regulation.

In this research, given the much more elaborate state of the literature on what motivates people to do work in general (SDT) and to perform public tasks and duties in particular (PSM), this research applies these frameworks within the context of the ‘like to’ dimension of the CLEAR model.

Enabled to

The ‘enabled to’ dimension involves the idea that people participate if they are supported by groups or organisations (Lowndes, Pratchett, and Stoker 2006, 288). From earlier studies we know that if people live in a social context where there is political interest, they are much more likely to be interested in politics (Fox and Lawless 2005). Social groups and organisations can act as amplifiers of an individual’s willingness. Without such a supportive social context, the persistence of a person to keep on participating fully depends on his or her own motivation. Thus, the existence of social networks and the types of social networks and civic infrastructures may be conditional for people to participate (ibid). This corresponds to our understanding of social capital. Social capital refers to the idea that social relations can facilitate the production of economic or noneconomic goods (Paxton, 2002; 256). In addition, social capital resides not in individuals but in relations between individuals (Coleman 1988). Social capital requires (1) objective associations among individuals and (2) associations of a particular type: reciprocal, trusting, and involving positive emotion (Paxton, 2002; 256). It is important to acknowledge that there is a succinct relation between social capital and collective action, but this collective action can result in citizens’ initiatives to organise social care as well as protests, petitions, and riots. Therefore, it is not only interesting to examine whether social capital correlates to candidacy for council membership but also what kind of social capital. Therefore, we apply Paxton’s (2002) taxonomy of social capital – membership of trade unions, religious groups, sports and recreation, politics, arts and culture, social work, youth work, health, animal rights, community action, environment, human rights, and peace – to this research.

Asked to

The ‘asked to’ dimension refers to the idea that people need to be mobilised in order to participate (Lowndes, Pratchett, and Stoker 2006, 288). They are engaged more easily when they are invited to do so. In fact, as the literature shows, being invited is one of the main reasons people tend to participate politically (Verba, Schlozman, and Brady 1995). Lowndes, Pratchett, and Stoker (2006) added to this that it is even more powerful if people in power are the ones who invite citizens to participate. In the literature on council membership, the importance of being invited as a condition for actual candidacy is also acknowledged (Broockman 2014; Bowman and Boynton 1966). In addition to the work of Lowndes, Pratchett, and Stoker (2006), we argue that previous political activities, such as being a member of a political party (Boogers, De Jong, and Voerman 2018) and having a history of political activity within a certain political party (Hooghe, Stolle, and Stouthuysen 2004) may correlate with being invited (or not) to participate in a council membership race.

Responded to

The ‘responded to’ dimension refers to the idea that for people to participate they have to believe that their involvement makes a difference (Lowndes, Pratchett, and Stoker 2006, 289). They have to believe that they are listened to, which is different from agreed with, and that their view has been taken into account. In SDT literature, this dimension is partly captured within the amotivation dimension, where people were asked whether their amotivation stems from the fact that they think their work is pointless or a waste of time (Gagne et al. 2015). More positively formulated, this corresponds to the concept of self-efficacy. In terms of political self-efficacy, this refers to ‘the feeling that individual political action does have, or can have, an impact upon the political process … the feeling that political and social change is possible, and that the individual citizen can play a part in bringing about this change’ (Campbell, Gurin, and Miller 1954, 187, quoted in Madsen 1987, 572). In line with the conceptualisation of Parrado et al. (2013), in this research we examine to what extent people have the feeling that as council members they can contribute to different policy domains, such as public safety, health, environmental issues, or education.

In our next section, we elaborate on how we have operationalised and tested our framework.

3. Research methods

Data

Data were collected in November 2019 by conducting a panel survey of 1,650 Dutch individuals. To collect data, we made use of the Longitudinal Internet Studies for the Social sciences (LISS) panel administered by CentERdata (Tilburg University, the Netherlands). In total, 1,335 panel members completed the survey, resulting in a response rate of 80.9%. Approximately 45% of the respondents are male, and the age group with the most respondents is 65 years and older (39.9%). The predominant educational level is lower professional education. The data are reasonably representative of the Dutch population with a small overrepresentation of less educated people.

Measures

Dependent variable

Our dependent variable candidacy was measured using one question: ‘Suppose you were asked to be a candidate for a political party in the upcoming municipal elections, what would be your answer?’ Respondents could answer ‘no’ (= 0) or ‘yes’ (= 1).

Independent variables

To the best of our knowledge, there is no measurement scale to measure the CLEAR framework. Therefore, we relied on other existing measures to measure the separate CLEAR dimensions. Gagne et al. (2015) developed the Multidimensional Work Motivation Scale (MWMS) to measure work motivation in line with SDT. Focusing on this scale taught us that this scale is congruent with the ‘can do’ and ‘like to’ dimensions as described by Lowndes, Pratchett, and Stoker (2006).

Can do was measured using three items: 1) ‘I don’t consider council membership because I don’t have the right skills,’ 2) ‘I consider council membership because I have to prove to myself that I can do it,’ and 3) ‘I consider council membership because it makes me feel proud of myself.’

Like to was measured using the distinction in different types of motivation as described in SDT. Amotivation was measured on a 5-point Likert scale using three items. An example item is ‘I don’t consider council membership because I really feel that I am wasting my time.’ Extrinsic regulation was measured on a 5-point Likert scale using six items. An example item is ‘I consider council membership because others will respect me more.’ Introjected regulation was measured on a 5-point Likert scale using four items. An example item is ‘I consider council membership because it challenges me.’ Identified regulation was measured on a 5-point Likert scale using seven items. Here we added four items from Ritz (2015) on PSM and the questions of Gagne et al. (2015) to our questionnaire. Following Ritz (2015), we made a distinction between self-interested and other-interested motives. An example item of self-interested PSM is ‘I consider council membership because I am very interested in politics,’ and an example item of other-directed PSM is ‘I consider council membership because I consider public service as my duty.’ Finally, intrinsic motivation was measured on a 5-point Likert scale using four items. An example item is ‘I consider council membership because I like being a councilor.’

Explanatory factor analyses were performed on the original survey items. Based on this, we made a few modifications. As it turned out, our respondents did not recognise the dimensions can do and like to as two separate dimensions. Furthermore, the sub-dimensions as developed by Ryan and Deci (2000a) were not recognised as such. Respondents recognised amotivation as a separate sub-dimension (α = .75). However, questions belonging to external regulation and introjected regulation were considered one dimension. We reformulated these two subdimensions into the dimension extrinsic motivation (α = .94). Finally, our respondents demonstrated that they considered the initial subdimension of intrinsic motivations (considering council membership because the work itself is enjoyable) and subdimensions identification and integration as one dimension. We have reformulated this dimension as intrinsic motivation (α = .97). Hence, we reformulated the dimensions can do and like to into amotivation, extrinsic motivation, and intrinsic motivation. Subsequently, we performed a confirmatory factor analysis in IBM SPSS AMOS version 26 to test the factor structure. Each indicator significantly loaded onto the appropriate factor and all loadings were above .40 (range from .60 to .91). The fit values for this model were .96 (TLI), .97 (CFI) and .058 (RMSEA). Based on these fit indices, one can conclude that this model is a good fit (Byrne 2001).

Enabled to was measured using four questions: ‘If you were a candidate for councilor, to what extent do you think you would experience support from: 1) your household members, 2) your family, 3) your friends/acquaintances, 4) your employer.’ Respondents could answer on a 5-point scale ranging from 1 (no support) to 5 (very much support) (α = .92). In addition, we asked respondents if they have a relative working as a council member. Respondents could answer ‘no’ (= 0) or ‘yes’ (= 1). Furthermore, we asked respondents whether they participate in one or more of the following communities: labour union; religious organisation; sport or recreation; welfare; music, culture, and art; animal rights organisation; human rights organisation; and environmental organisation. Respondents could answer ‘no’ (= 0) or ‘yes’ (= 1).

Asked to was measured using one question: ‘Have you ever been invited to become a council member?’ Respondents could answer ‘no’ (= 0) or ‘yes’ (= 1).

Responded to was measured using 11 items: ‘How much influence do you think you can have as a council member in the field of: 1) safety, 2) education, 3) healthcare, 4) sustainability/environment, 5) economy, 6) culture and sport, 7) infrastructure, 8) spatial planning and public housing, 9) integration, 10) catering industry, 11) tourism and recreation.’ Respondents could answer on a 5-point scale ranging from 1 (no influence) to 5 (very much influence) (α = .94).

Control variables

We included gender (0 = male, 1 = female), age (six categories: 15–24, 25–34, 35–44, 45–54, 55–64, 65 and older), educational level (six categories: elementary school, high school – low, high school – high, lower professional education, middle professional education, higher professional education), income (continuous variable gross income), housing situation (dummy variables: single, living together with no children, living together with children, single with children), and primary occupation (dummy variables: job or school, unemployed, retired).

4. Results

In this section, we assess the results of our analysis. We begin this section by showing the means, standard deviations, and correlations of the study variables (Table 1). Subsequently, we present to what extent the dimensions of the reformulated CLEAR model (as the result of our factor analysis) can be considered reliable predictors of an individual’s consideration for candidacy. Finally, we present how these dimensions may have different effects across different groups within society. We distinguish among differences in sex, age, education level, (gross) income, housing situation, and primary occupation.

Table 1. Correlation matrix ***Correlation is significant at the 0.00 level (2-tailed); ** Correlation is significant at the 0.01 level (2-tailed); * Correlation is significant at the 0.05 level (2-tailed).

	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Candidacy	0.09	0.29	1.00
2. Sex	0.55	0.50	-,073**	1.00
3.age	57.11	17.24	-,115**	-8.00	1.00
4.Level of education in CBS-cat	3.23	1.49	0.05	-,081**	-,280**	1.00
5.income	2228.18	1645.89	0.00	-,356**	-0.01	,394**	1.00
6. Single	0.29	0.45	-0.02	0.00	,139**	-,070*	,061*	1.00
7. Living together, no children	0.41	0.49	-0.01	-,066*	,337**	-,073**	-,079**	-,530**	1.00
8. Living together, with children	0.25	0.43	0.00	0.02	-,448**	,157**	0.05	-,366**	-,482**	1.00
9. Single with children	0.05	0.22	0.05	,110**	-,158**	0.00	-0.05	-,149**	-,196**	-,135**	1.00
10. Job or school	0.47	0.50	,075**	-0.04	-,663**	,300**	,294**	-,152**	-,242**	,377**	,113**	1.00
11. Unemployed	0.21	0.40	-0.02	,186**	,074**	-,176**	-,329**	0.03	0.01	-0.04	0.01	-,482**	1.00
12. Retired	0.32	0.47	-,065*	-,119**	,644**	-,168**	-0.03	,136**	,256**	-,368**	-,130**	-,652**	-,351**	1.00
13. Amotivation	2.82	1.12	-,141**	0.05	0.04	-,097**	-,108**	-0.01	0.04	-0.03	-0.02	-0.05	0.02	0.04	1.00
14. Extrinsic motivation	1.66	0.76	,337**	-0.02	-,106**	0.03	-0.04	0.02	-,082**	,066*	0.02	0.02	0.01	-0.03	0.00
15. Intrinsic motivation	2.00	1.02	,414**	-,075**	-0.02	,130**	,073**	0.01	-0.02	0.02	-0.02	-0.02	-0.02	0.04	-,114**
16. Relative working as council member	0.05	0.23	,108**	-0.04	0.04	,055*	,061*	-,060*	,101**	-0.04	-0.02	-0.03	-0.01	0.03	-0.05
17. Participating in labor union	0.08	0.28	,055*	-,168**	0.05	0.05	,115**	0.00	0.02	0.01	-,057*	0.03	-,081**	0.04	0.00
18. Participating in religious organization	0.09	0.28	-0.01	-0.01	,100**	0.01	-0.03	-0.03	0.03	0.01	-0.03	-,086**	0.03	,067*	-0.04
19.participating in sport or recreation	0.23	0.42	0.00	-,079**	-0.01	,139**	,104**	-0.05	0.03	0.03	-0.05	0.04	-,106**	0.05	,072**
20. Participating in welfare	0.04	0.19	0.02	0.04	0.02	0.04	-0.01	,056*	0.01	-0.04	-0.05	-,054*	,082**	-0.01	-0.01
21. Participating in music, culture or art	0.07	0.26	0.01	0.01	,058*	0.05	,057*	-0.01	,064*	-0.05	-0.02	-0.05	-0.03	,075**	-0.01
22. Participating in animal rights organization	0.01	0.09	,057*	0.03	-0.02	0.01	-0.02	-0.02	0.02	-0.01	0.02	0.00	0.04	-0.03	-0.01
23. Participating in human rights organization	0.01	0.10	-0.03	0.05	-0.01	0.03	0.02	0.02	0.01	-0.02	-0.02	-0.01	-0.02	0.02	-0.02
24. Participating in environmental organization	0.02	0.14	0.01	-0.04	,072**	,071**	0.05	0.00	,064*	-,072*	-0.01	-0.05	-0.05	,094**	-0.01
25. Social support	2.61	4.65	,257**	-,094*	-,468**	,231**	,125**	-,125**	-,091*	,186**	0.01	,366**	-,098*	-,354**	-,179**
26. Working for political party	0.01	0.08	0.04	-0.02	0.00	0.02	0.01	0.03	-0.03	0.02	-0.02	0.00	0.03	-0.02	-0.04
27. Ever invited to become councilor	0.03	0.18	0.04	-,106**	,096**	,063*	,082**	-0.01	,068*	-0.04	-0.04	-,076**	-0.02	,102**	-,068*
28. Influence	2.38	7.83	,129**	-,111**	-,070*	,084**	,088**	-0.01	-0.04	,071*	-0.01	,093**	-0.05	-0.05	-,094**
	14	15	16	17	18	19	20	21	22	23	24	25	26	27
1. Candidacy
2. Sex
3.age
4.Level of education in CBS-cat
5.income
6. Single
7. Living together, no children
8. Living together, with children
9. Single with children
10. Job or school
11. Unemployed
12. Retired
13. Amotivation
14. Extrinsic motivation	1.00
15. Intrinsic motivation	,803**	1.00
16. Relative working as council member	0.05	,116**	1.00
17. Participating in labor union	0.03	,064*	-0.02	1.00
18. Participating in religious organization	-0.01	0.04	,115**	,061*	1.00
19.participating in sport or recreation	-0.03	0.00	0.03	,058*	0.03	1.00
20. Participating in welfare	0.00	0.03	0.02	-0.02	,133**	0.01	1.00
21. Participating in music, culture or art	-0.01	0.05	,124**	-0.03	,107**	,089**	,079**	1.00
22. Participating in animal rights organization	0.04	0.05	0.05	0.00	0.00	0.03	-0.02	0.01	1.00
23. Participating in human rights organization	0.04	,059*	-0.02	0.02	-0.01	0.03	0.02	0.03	,072**	1.00
24. Participating in environmental organization	0.02	0.04	0.01	0.01	-0.03	0.04	,082**	,082**	0.05	,194**	1.00
25. Social support	,233**	,278**	,107*	0.02	-0.05	0.04	-0.01	0.02	0.05	-0.02	-0.06	1.00
26. Working for political party	0.02	,106**	,102**	-0.02	,105**	0.04	0.03	0.01	-0.01	,081**	,118**	0.04	1.00
27. Ever invited to become councilor	0.05	,145**	,213**	0.05	0.03	0.05	0.05	0.04	,075**	0.02	,062*	0.01	,238**	1.00
28. Influence	,140**	,176**	0.01	0.02	,057*	-0.03	0.02	0.03	,096**	-0.02	0.02	,217**	0.03	0.05

Table 1 shows that approximately 9% of our sample indicated that they were willing to apply for candidacy. When focusing on how our respondents on average evaluated the dimensions of our (adapted) CLEAR model, we see that on average (all on a 5-point scale except for asked to, which has a 2-point scale), amotivation resulted in a mean of 2.80, extrinsic motivation 1.66, intrinsic motivation 2.00, enabled to 2.61, asked to 0.03, and responded to 2.38. When correlating these dimensions to our dependent variable (candidacy), we see that all dimensions are significantly correlated except for the asked to dimension. This dimension correlates significantly with numerous other variables such as sex, age, household, and the amotivation and intrinsic motivation dimensions of our model. Concerning the other dimensions of our model, amotivation had a correlation of −0.448 (p < 0.01), extrinsic motivation 0.337 (p < 0.01), intrinsic motivation 0.414 (p < 0.01), enabled to 0.257 (p < 0.01), asked to 0.042, and responded to 0.129 (p < 0.01).

CLEAR as a predictor for candidacy

Table 2 displays to what extent the dimensions of the CLEAR model can be considered reliable predictors for people’s decision to apply for candidacy. The Goodness of Fit tests presented at the bottom of the table show good results to interpret these results. Moreover, the Classification Table (see Table 3) shows that the overall fit of the model yielded 92.2% correct classifications what can be considered as a good result. In addition, the criteria for classification accuracy is satisfied as the overall case classifications was more than 25% higher than the proportion by chance rate (54.3%).

Table 2. Regression analysis.

	B	S.E.	Wald	Df	Sig.	Exp(B)	95% C.I.for EXP(B)
Predicting variable							Lower	Upper
Sex (1 = female)	−0,471	0,436	1,163	1	0,281	0,625	0,266	1,469
Age	−0,015	0,016	0,898	1	0,343	0,985	0,954	1,017
Level of Education (CBS Category)	−0,170	0,156	1,186	1	0,276	0,843	0,621	1,146
Income	0,000	0,000	2,623	1	0,105	1,000	1,000	1,000
Living together no children (ref.cat = single)	−0,061	0,573	0,011	1	0,915	0,941	0,306	2,891
Living together with children (ref.cat = single)	−1,182	0,599	3,898	1	0,048	0,307	0,095	0,991
Single with children (ref.cat = single)	−1,134	1,326	0,731	1	0,392	0,322	0,024	4,326
Unemployed (ref.cat = job or school)	0,128	0,641	0,040	1	0,842	1,136	0,324	3,987
Retired (ref.cat = job or school)	−1,870	0,899	4,323	1	0,038	0,154	0,026	0,898
Relative working as council member	0,734	0,675	1,183	1	0,277	2,083	0,555	7,817
Participation in labour union	0,553	0,577	0,916	1	0,339	1,738	0,560	5,389
Participation in religious organisation	−0,119	0,653	0,033	1	0,856	0,888	0,247	3,194
Participation in sport or recreation	−0,210	0,504	0,174	1	0,677	0,810	0,302	2,177
Participation in welfare	0,066	1,112	0,004	1	0,952	1,069	0,121	9,445
Participation in music, culture, art	0,717	0,764	0,881	1	0,348	2,048	0,458	9,149
Amotivation	−0,716	0,276	6,727	1	0,009	0,489	0,285	0,840
Extrinsic motivation	−0,023	0,354	0,004	1	0,948	0,977	0,488	1,955
Intrinsic motivation	1,710	0,334	26,241	1	0,000	5,530	2,874	10,639
Enabled (social support)	0,153	0,052	8,747	1	0,003	1,166	1,053	1,290
Asked (ever invited to become council member)	−0,817	0,897	0,829	1	0,362	0,442	0,076	2,564
Responsive (Influence)	−0,006	0,026	0,048	1	0,827	0,994	0,944	1,047
Hosmer and Lemeshow test	Chi-Square 9,422			8	0,308
Nagelkerke R Square					0,546
N	477

Note: n is lower than the original sample due to missing cases of independent variables. Three variables (participation in animal rights organisation, human rights organisation and environmental organisation) were deleted from the analyses due to skewness.

Table 3. Classification Table.

Observed			Predicted
			Candidacy		Percentage Correct
			No	Yes
	Candidacy	No	413	8	98,1
		Yes	29	27	48,2
	Overall Percentage				92,2

a. The cut value is ,500

Based on our analysis we can observe that from our model, only amotivation, intrinsic motivation and the level of social support (enabled) had a significant effect on the likelihood that an individual would apply for candidacy (amotivation B = −0,716 p = 0.009; intrinsic motivation B = 1.710, p = 0.000; enabled B = 0.153, p = 0.003). Extrinsic motivation did not have a significant effect. Our analysis also shows that the level of assumed impact (responsive) is not a significant predictor of an individual’s decision to run for council member. Furthermore, we can observe that the effect of asked to is not significant, showing that being invited does not reliably predict an individual’s decision to apply. In sum, it appears that the decision to run for office depends on whether the individual considers the work of council members enjoyable, finds it important to contribute to a common cause and/or enjoys enough social support to embark on such a mission.

Finally, we can see that the category ‘living together with children’ (B = −1.182, p = 0.048) and ‘retired’ (B = −1.870, p = 0.038) appeared to be reliable predictors of an individual’s application. This indicates that most societal group characteristics – such as gender, age, level of education, or being employed or not – are by themselves not reliable predictors of an individual’s decision to run for council membership when combined in one model with the dimensions of the CLEAR model.

This analysis shows that most societal group characteristics do not significantly explain an individual’s decision to apply for council membership. In contrast, the dimensions amotivation, intrinsic motivation and social support (enabled) of the (adapted) CLEAR model proved to be significant predictors. In the next section, we go one step further by analysing how the significant dimensions differed between different societal groups. Doing so enables us to estimate whether, for instance, incentives intended to increase an individual’s intrinsic motivation would lead to an increased likelihood to run for office among a certain type of people.

How CLEAR dimensions differ across different societal groups

Table 4 shows how the significant dimensions in our model differed across the different societal groups. We distinguish in this research differences in sex, age, education level, housing situation, primary occupation, and income. We elaborate on the significant differences below Table 4.

Table 4. Differences in CLEAR dimensions among societal groups.

		Amotivation			Intrinsic motivation			Enabled
		Mean	F	Sig	Mean	F	Sig	Mean	F	Sig
Sex	Male	2.76	0.71	0.40	2.08	1.14	0.29	2.72	2.64	0.11
	Female	2.87			1.93			2.51
Age	15–24	2.92	1.69	0.14	2.14	1.55	0.17	3.32	32.44	0.00***
	25–34	2.75			2.11			3.42
	35–44	2.71			2.02			2.91
	45–54	2.77			1.86			2.64
	55–64	2.73			1.93			2.53
	> 65	2.92			2.03			1.75
Education level	Elementary school	2.87	3.52	0.00***	1.95	7.73	0.00***	2.24	7.79	0.00***
	High school – low	2.97			1.82			2.22
	High school – high	2.70			2.18			2.84
	Lower professional education	2.78			1.99			2.7
	Middle professional education	2.68			2.11			2.83
	Higher professional education	2.61			2.42			3.04
Housing situation	Single	2.82	0.08	0.49	2.01	0.42	0.73	2.32	7.55	0.00***
	Living together with no children	2.88			1.96			2.48
	Living together with children	2.77			2.02			2.89
	Single with children	2.74			1.91			2.68
Primary occupation	Unemployed	2.86	1.86	0.16	1.95	1.12	0.33	2.29	50.23	0.00***
	Job or school	2.76			1.97			2.88
	Retired	2.89			2.05			1.73
Income	0–1,500	2.80	3.74	0.01**	1.92	2.79	0.04**	2.54	0.88	0.48
	1,501–2,500	2.90			2.01			2.61
	2,501–4,000	2.78			1.93			2.80
	> 4,001	2.43			2.28			2.67

***ANOVA is significant at the 0.00 level (2-tailed); ** ANOVA is significant at the 0.01 level (2-tailed); * ANOVA is significant at the 0.05 level (2-tailed)

Amotivation

Concerning amotivation, our analysis showed that there were only significant differences between people with different levels of education and different levels of income. Our analysis showed that people with a low level of professional education had significantly more amotivation (were significantly less motivated) towards council membership (mean of 2.97 on a 5-point scale) compared to the people with the two highest levels of education (means of 2.68 and 2.61). With regard to the level of income, people with the highest level of income (+€4,001 per month) had significantly lower amotivation (were more motivated, mean of 2.43) than people with the two lowest income levels (means of 2.79 and 2.90).

Intrinsic motivation between different societal groups

Focusing on intrinsic motivation, we found only significant differences between groups of people with different education levels and different income levels. It appears that people with higher education levels are significantly more intrinsically motivated (highest level mean of 2.42 on a 5-point scale) than people with lower education (high school – low mean 1.99). People with sequential education levels (e.g., middle education level and higher education level) did not differ significantly from each other. Therefore, we can conclude that intrinsic motivation is the highest among people with the highest levels of education. We found a similar though less explicit significant difference between people with different levels of income. Here we found that people with the highest level of income (+€4,001) had significantly more intrinsic motivation (mean 2.28) than people with the lowest level of income (€0–1,500, mean 1.91).

Experienced support between different societal groups (enabled)

Our analysis of the enabled dimension shows that there are significant differences between people with 1) different ages, 2) different education levels, 3) different housing situations, and 4) different primary occupations. It is perhaps peculiar that we found that the older people get, the less social support for candidacy they experience from their environment. The youngest people in our sample (15–24 years) estimated their social support on average at 3.32 (on a 5-point scale), while the eldest people (+65) estimated this on average at 1.75. When focusing on the difference between education levels, our analysis generated a similar result as our analysis of the intrinsic motivation dimension. Here we also see that people with higher education levels experience significantly greater social support than people with lower levels of education (higher professional education has a mean of 3.04, high school – low has a mean of 2.22), while people in sequential education categories do not differ from each other significantly. Perhaps a surprising result is that people with different housing situations experience significant differences in social support. Here we see that families (people living together with children) experience significantly more social support (mean 2.89) than singles (mean 2.32) or people living together without children (mean 2.48). There appeared to be no significant differences with the single parents group. The highest F-score (50.23) was found in our analysis of differences between people with different primary occupations. People with a job and students experience the most social support (mean 2.88), while unemployed people have an average mean of 2.29. Retired people experience significantly lower social support (mean 1.73). We can conclude that according to our respondents, council membership is most attractive to people with families who are relatively young and well educated.

5. Discussion

Based upon our analysis, the image of people interested in a position of council member is not very surprising. People with higher socio-economic status are not only more motivated to run for council membership, they also have a social network that supports them in such a position. Especially the latter is an important finding. As formally almost everyone older than 18 in the Netherlands can run for council membership (Minstry of Interior 2021), officially there should be no reason why council members should not be equal representations of society. However, skills like debating, the ability to comprehend loads of information, having a network in society and the ability to check on the mayor and aldermen (Politiek adviseur 2021) indicate that it is not a job for everyone. Hence, the ambition for a more representative representation of municipality councils is a quest for breaking through vicious circles. Our analysis show that people with lower socio-economic status more inclined to have high levels of amotivation towards council membership. But as long as the social structures are in place that supports such an adverse attitude towards governmental institutions it is unlikely that if there is an individual from lower socio-economic background who has the ambition to apply for candidacy that he or she will receive the necessary support to do so. We elaborate on the implications of our findings in the remainder of this section.

What constitutes an individual’s decision to apply for council membership ?

When we focused on what constitutes motivation, we saw that people made no distinction between the ‘can do’ dimension and the ‘like to’ dimension. Based on our factor analysis, we conclude that when considering what constitutes motivation, it seems to make more sense to distinguish between amotivation, intrinsic motivation, and extrinsic motivation. As such, although theoretically it seems to be logical to distinguish between whether people can and whether people want to (Lowndes, Pratchett, and Stoker 2006) in the decision to run for council membership, in reality, these two aspects overlap.

In terms of the other aspects of the model, our study shows that social support (enabled) is conditional for an individual’s decision to consider candidacy. However, the taxonomy as presented by Paxton (2002) did not prove to generate significant differences among the different categories.

Furthermore, we saw that being asked to run for office is not a reliable predictor of an individual’s decision to do so. This conclusion is contrary to the conclusions of many other authors, who have concluded the opposite (see Bowman and Boynton 1966 as one of the earliest; see also Meadowcroft 2001; Broockman 2014). Perhaps this finding can be explained against the background of declining interest in political participation in general. Since being invited can only be considered a motivational stimuli if the recipient regards the invitation as something positive (e.g., an honour to belong to a certain group), the declining legitimacy of representative institutions possibly means that people consider it less honourable to be associated with a political party or representative body.

Finally, our analysis also shows that the level of assumed impact (responsive) is not a significant predictor of an individual’s decision to run for council member. This is peculiar given that intrinsic motivation has such a strong positive effect on this decision. It seems that our respondents consider making an effort to contribute to a common good more important than actually having impact. This is an addition to the work of Fox and Lawless (2005), who concluded that efficacy is a critical predictor for nascent political ambitions. However, they gauged efficacy by an individual’s self-perceived qualifications (which corresponds to how we defined the can do dimension). Here, efficacy seems to adhere more to how an individual can contribute to a larger cause. This is also reflected in our study. Our results add to this idea that for people who are motivated by intrinsic factors, contributing is more important than the result.

Differences across societal groups

We conclude that it is valuable to pay attention to the stratified character of democratic and political participation. Models like the CLEAR framework do not take into account differences in income, education, social status, and the like. Our analysis showed some significant differences in how the various dimensions of the CLEAR model are differently considered across different societal groups. For example, based on our analysis, we can expect that when incentives are introduced in particular to address people’s intrinsic motivation, predominantly highly educated people (high school – high and middle and higher levels of professional education) and (although to a lesser extent) people with high incomes will respond to the invitation. However, although this group seemed to be most interested in the position of council member (they also had the lowest level of amotivation), we can question whether an increase in incentives will lead to more candidates from this pool. As Meadowcroft (2001) has stated, it is especially this target group who are hesitant to participate because of ‘the personal and career costs of the time required for public service as a local authority member, coupled with the declining prestige and status of council service’ (Meadowcroft 2001 , 22).

Furthermore, our findings reveal possible biases with regard to the kind of people invited to run. The primary target groups of political parties do not seem to be the groups with the most potential (in terms of motivation), as can be deduced from the fact that the people who indicated that they have been asked usually belong to a favourable group for political parties (Broockman 2014). However, this excludes numerous other potentially receptive groups. Our study shows that women, younger people, people with lower incomes, and people with lower levels of education are significantly less frequently asked. This is troubling when one aims to have representative councils. However, the problem not only lies with the bias of political recruitment systems; additionally, people with lower socio-economic status seemed to be less convinced of the benefits (Meadowcroft 2001) of being a council member.

6. Conclusion

To explain what constitutes citizens’ willingness to run for office, we tested an adapted model of the CLEAR framework. In doing so, our paper served four purposes: 1) extend our understanding of what determines an individual’s decisions to consider council membership as a possible option, 2) show how different determinants have different effects across people belonging to different societal groups, 3) show to what extent well-known theoretical frameworks such as SDT and PSM have value in predicting the chance that an individual will apply, and 4) test the CLEAR framework on a large quantitative scale and its value to explain what determines people’s willingness to run for council membership.

Our study showed that although the CLEAR model offers us building blocks to understand an individual’s decision to run for office, it does require some reformulation (i.e., subdividing ‘can do’ and ‘like to’ into amotivation, intrinsic motivation, and extrinsic motivation). After reformulation we were able to apply the CLEAR model in a large N setting, including insights from other well-known motivational frameworks (SDT and PSM). In doing so, we extended our understanding of what contemplations people have to make the decision to run for council membership. On the one hand this generated confirmation of insights that were to be expected (i.e., intrinsic motivation is much more important than extrinsic motivation). On the other hand our analysis showed that the extent in which people feel supported by their network also determines their decision to apply. Our analysis also showed that people are not considering council membership for any kind of external aspect (i.e., rewards, but also not because of the possible influence or the effect that they can generate as council member) but that they consider it merely as their civic duty (being a council member is a virtue in itself). Our analysis on how these dimensions differ across different societal groups also confirmed that council membership is something that is much more considered among people with higher socio-economic status than among people with a lower socio-economic status. But also here, we see that not only people with lower socio-economic status are less interested, they also do not have the network to support them if they are interested in such a position. Therefore, the search for a more representative municipality council is also a broader question about how to restore faith in governmental institutions across different societal groups.

Of course, we must acknowledge the limitations of this study. First, we were able to detect significant differences across different societal groups. However, this research is less equipped to formulate theoretical insights about why these differences occurred. Especially if we want to go one step further and formulate incentives for different kind of groups, we need to understand the underlying explanatory mechanisms. In addition, because of privacy issues, we were not able to ask people about their ethnic and/or religious background. Therefore, we were unable to conclude how the dimensions of our model may differ across societal groups with a potentially crucial difference. Second, since our model explains only marginally what constitutes an individual’s choice to run for office (total variance explained by the model), there are numerous other possible factors that explain this choice better than the CLEAR model does. Most likely, people’s decision to run for council member is heavily depending on local politics and local policy issue. In our analysis we were not able to include this dimension. Therefore, we have to be reluctant in drawing harsh conclusions about our understanding of what determines an individual’s decision to apply for council membership. Third, the strength of this study (analysing the considerations of people who are willing to politically participate) is also its weakness: to establish a complete understanding of what influences people’s willingness to run for office, we also have to analyse the motivations of those who have agreed to do so.

These limitations also offer us a few directions for future research. As our research suggests significant differences between people stemming from different societal backgrounds. The key-question is how trust can be restored and if that may lead to a more representative municipality council? Second, as mentioned above, we were able in this research to extent our understanding of one’s consideration to run for candidacy. But as only 28% of our model explains this decision, there is still a lot to discover. Hopefully future research will generate a more comprehensive image of what determines such a decision. Last, we propose to extent this research specific to the group of people that applied for candidacy (note, this is not the same as the people that actually became council member) and test whether the CLEAR dimension has a different turn out with these people.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

William Voorberg

William Voorberg is an assistant professor at the department of Public Administration and Sociology of the Erasmus University. His work focusses on applying design methods (both as a problem-solving strategy and a research approach) to contemporary challenges, involving issues obstructing circular economy, restoration and conservation of biodiversity, and climate change. Also, his work aims to understand how citizen participation in many forms can be understood e.g. how co-creation and co-production may be useful strategies to collaborate with citizens, what motivations people have to participate and how we can address the outcomes of such participation processes.

Brenda Vermeeren

Brenda Vermeeren is an Associate Professor at the Department of Public Administration and Sociology, Erasmus University Rotterdam and a senior advisor at ICTU-InternetSpiegel (a programme of the Dutch Ministry of Interior and Kingdom Relations). Her research interests include the relationship between HRM and performance in the public sector and the role of line managers and employee wellbeing within this relationship.

Arwin Van Buuren

Arwin van Buuren is a Professor of Public Administration at the Erasmus University Rotterdam. He is alos the Academic Director of the Erasmus Governance Design Studio and  GovLab010, a knowledge hub between the city of Rotterdam and the Erasmus University. His research is about the interrelationships between design, collaboration and innovation focusing on participative and invitational forms of governance.