Does Stewardship Theory Provide a Viable Alternative to Control-Fixated Performance Management?

: Stewardship theory provides an interesting alternative to agency theory, which in the recent New Public Management era supported the introduction of rigorous performance management systems based on generalized mistrust in and control of public employees. However, we lack empirical validation of the feasibility and positive outcomes of the new forms of trust-based management recommended by stewardship theory. As such, there are few examples of alternative ways of boosting the motivation of public employees that can serve as beacons for public service organizations (PSOs) eager to ﬁnd new ways of motivating their sta ﬀ to create public value for the users of public services and society as a whole. This article aims to remedy this problem by exploring a seemingly successful empirical case of trust-based management to see whether the core principles of stewardship theory apply and how new management practices may inﬂuence the motivation and well-being of the employees, the perceived satisfaction and involvement of the users, and overall organizational performance, including cost e ﬃ ciency.


Desperately Seeking Ways to Boost Intrinsic Task Motivation and Public Service Motivation
As knowledge and evidence of both the motivational inefficiency and detrimental effects of control-and incentive-based performance management on intrinsic task motivation (ITM) and public service motivation (PSM) accumulate (Jacobsen et al. 2014;Perry et al. 2009), we need to explore and test alternative ways of governing frontline personnel in public service organizations (PSOs). Stewardship theory (Schillemans 2013) provides an interesting and welcome alternative to agency theory, which in the era of New Public Management (NPM) has prompted the introduction of a new performance management system based on generalized mistrust in and control over public employees (Hood 1991;Pollitt and Bouckaert 2011;Van Dooren et al. 2015). However, we lack the empirical validation of the feasibility and outcomes of the new forms of trust-based governance recommended by stewardship theory. As such, there are few examples of alternative ways of boosting public employee motivation that can serve as beacons for PSOs desperately seeking new ways of motivating their personnel to create public value for the users and society in general. This article aims to remedy this problem by exploring a seemingly successful empirical case of trust-based management to see whether the core principles of stewardship theory apply and how new management practices possibly influence the motivation and well-being of the employees, the perceived satisfaction and involvement of the users, and overall organizational performance (including cost efficiency). 1 There has been growing interest in how to motivate public employees to do good work and produce value for the users of public services and society in general. Hence, instead of uncritically subscribing to the Weberian assumption that all public employees are well-motivated and loyal servants, narrowly concerned with the pursuit of the public interest, researchers and practitioners have taken a more realistic stance and started examining the actual levels and forms of employee motivation and how it can be impacted by different forms of public management and governance. Motivation is defined as the amount of energy that an employee is prepared to invest in accomplishing a particular task. The research on employee motivation, commonly distinguishes between intrinsic task motivation, which arises from the value of the work for the individual employee, and extrinsic motivation, which arises from the desire to obtain outcomes that are apart from the work itself and typically derived from managerial assessments of employee performance (Amabile 1993;Benabou and Tirole 2003;Locke and Schattke 2019;Sansone and Harackiewicz 2000). New research insists that public employees are possibly also motivated by an altruistic endeavor to produce services that help citizens to attain a better life and to produce outcomes that are valuable for society in general (Breaugh et al. 2018;Hondeghem and Perry 2009, p. 6). Some see this additional public service motivation as an internalized type of extrinsic motivation, because it is oriented towards producing outcomes of benefit for others beyond oneself (Vandenabeele 2007). Hence, public employees are driven by a mixed bag of motivations:

I.
Extrinsic motivation, which is based on self-interest and drives action based on the desire to receive performance-related rewards and to avoid punishment (Sansone and Harackiewicz 2000). Hence, employees are assumed to respond to positive and negative reinforcements, such as praise, merit wages, reproach, and threats of termination.

II.
Intrinsic task motivation is hedonistic and refers to the empowering feeling enjoyed by employees who use their knowledge and skills to competently and efficiently accomplish a task they deem to be meaningful (Conger and Kanungo 1998;Franco et al. 2002;Thomas and Velthouse 1990). ITM is thus likely driving a surgeon carrying out a heart operation, a teacher helping children to read and write, and a professional member of the navy navigating a destroyer on patrol in domestic waters. III.
Public service motivation is defined by as 'an individual's orientation to delivering service to people with the purpose of doing good for others and society' (Hondeghem and Perry 2009, p. 6). It is an altruistic, pro-social motivation to serve the public interest out of obligation and loyalty, to help specific citizens and users by delivering services based on expertise and compassion, or to help to create and maintain an efficient and effective public sector (Andersen et al. 2011). Hence, PSM may equally drive a social worker to work hard to help an unemployed person to find gainful employment or an executive manager in the Ministry of Finance who is eagerly fighting unwanted budget transgressions that risk undermining public spending policies.
Understanding the different work-related motivations of public employees is important due to the labor-intensiveness of public service delivery (Franco et al. 2002), but there is good reason to assume that it is more important to enhance the ITM and PSM than the extrinsic, incentive-based motivation of public employees. While the latter fails to produce the expected positive effects on PSO performance ), the former tends to enhance work engagement, job satisfaction, and organizational commitment (Pandey and Stazyk 2008) together with public performance and the production of desirable outcomes Naff and Crum 1999). Moreover, on a more fundamental 1 While the focus of our study is on management practices that aim ensure that a given set of tasks are carried out and goals and objectives are achieved, we shall like to note that these practices are predicated on particular forms of governance such as responsible autonomy (Fairtlough 2007) often associated with the New Public Governance (Mills et al. 2019) as well as new forms of leadership such as distributed and horizontal leadership (Bolden 2011) or servant leadership (Reinke 2004). level, ITM and PSM are autonomous forms of motivation, as the individual understands the value of their work-related efforts. This stands in contrast to the extrinsic motivation, where the sense of volition is higher in ITM and PSM because the employees have internalized the value of their efforts, which is a good thing if this value is central to the organization (Vandenabeele 2007). Fortunately, the conditions in the public sector for enhancing ITM and PSM are favorable, as there tends to be greater emphasis on professionalism and public value creation (Moore 2000;Perry and Wise 1990) and less focus on monetary rewards (merit pay) and punishment (contract termination) than in the private sector (Burgess and Ratto 2003;Rainey and Bozeman 2000).
Employee motivation is influenced by cultural factors as well as organizational factors pertaining to the different ways that PSOs are governed and managed (Franco et al. 2002;Pandey and Stazyk 2008). Unfortunately, the last three to four decades of NPM reforms have aimed to strengthen extrinsic carrot-and-stick motivation at the expense of ITM and PSM (Van Dooren et al. 2015). NPM draws inspiration from the corporate sector and organizational economics in recommending marketized service production based on privatization, outsourcing, and the commercialization of the remaining public sector based on the import of performance management tools from private firms. While the introduction of quasi-markets draws on neoclassical market economics, the introduction of performance management is founded on agency theory (Jensen and Meckling 1976;Lane 2005), which assumes that information asymmetries between the public principals and their administrative agents allow the latter to act opportunistically and pursue their narrow self-interests for less work and more pay at the expense of public performance. The cure for the self-serving and sub-optimizing behavior of public employees is performance management involving a combination of management by objectives, performance measurement, transparent reporting, and transactional leadership based on conditional rewards and punishments (Barber 2008;Van Dooren et al. 2015). Performance management tends to focus on financial incentives that may crowd out ITM and PSM by making public employees focus on receiving tangible rewards in exchange for good performance and perceiving these rewards as more important that the more intangible rewards gained from competently solving a meaningful task or doing something good for specific users, society, or the public sector (Franco et al. 2002;Stazyk 2009). New research shows that the crowding out effect of performance management tools tends to be conditioned on the employees' perception of these tools as being controlling and based on mistrust of their competences and intentions (Jacobsen et al. 2014). This all goes to show how the excessive use of control-fixated performance management risks creating a workforce of cynics who only work to receive financial rewards and not because they enjoy what they are doing and think it has value for others (Moynihan 2010).
If the crowding out effect of performance management was the only negative impact, it would be enough to counterbalance the incentive-based transactional leadership with a transformative leadership aiming to 'crowd in' ITM and PSM by seeking 'to develop, share, and sustain a vision intended to encourage employees to transcend their own self-interest and achieve organizational goals' (Jacobsen and Andersen 2015, p. 832). However, it is well-established in the literature that attempts at solving the principal-agency problem by means of performance management also lead to goal-displacement and sub-optimization (Bohte and Meier 2000), enhance standardization and hamper innovation (Moynihan 2013), spur gaming and window dressing that seeks to produce apparent rather than real results (Hood 2006), and increase transaction costs, thereby taking away resources from service delivery (Musso and Weare 2020). Considering the many unintended negative effects of performance management, of which the crowding out of ITM and PSM is our main concern, it is easy to understand why PSOs are desperately looking for alternatives to control-based performance management.
Trust-based management models aiming to empower public employees through a combination of distributive and horizontal leadership provide an interesting alternative to control-based performance management, although they tend to complement rather than replace the latter (Nyhan 2000;Vallentin and Thygesen 2017). What makes the new trust-based governance models interesting is that they abandon the principal-agency model, which is based on individual utility maximization, and adopt a new stewardship model that assumes a considerable goal congruency between public leaders and public employees. This assumption completely changes the control-trust nexus and lays the foundation for a new type of trust-based management that may help to boost ITM and PSM. 2 The remainder of this article aims to explore the empirical feasibility and results of the introduction of a trust-based management model in PSOs. It begins by explaining the origins and core assumptions of stewardship theory and comparing it with agency theory. It then introduces the empirical case that we will scrutinize, justifies the case selection, and accounts for the qualitative methods applied in the case study. The first part of the empirical analysis aims to test whether the core management principles associated with stewardship theory apply in the empirical case and how they combine with the management principles associated with agency theory. The second part examines the outcomes associated with introducing trust-based management aimed at empowering employees and enhancing their ITM and PSM. The discussion reflects on the theoretical and empirical contributions, and the conclusion calls for the further study of trust-based management models founded on stewardship theory.

Stewardship Theory and Agency Theory Compared
New ideas in public administration and management are often imported (with considerable delay) from the private sector, which is perceived as a major source of legitimacy due to its perceived efficiency and innovativeness. Agency theory and stewardship theory are no exception. The shift from the former to the latter emerged in corporate governance theory in the early 1990s (Davis et al. 1997;Donaldson and Davis 1991;Muth and Donaldson 1998) and came to the public sector more than a decade later (Dicke 2002;Schillemans 2008;Van Slyke 2007).
In their seminal article, Donaldson and Davis (1991) describe how corporate governance has been influenced by agency theory, which holds that shareholder interests will only be safeguarded in firms where the CEO is not also the board chair and where the board has created proper incentives and control mechanisms that curtail the CEO's ability to act opportunistically in ways that make shareholders incur a loss, compared to a situation where they had exercised direct control over the corporation. Rejecting the underlying 'economic man' model, Donaldson and Davis proceed to argue that far from being a self-interested shirker, the CEO will typically want to be a good steward of the corporate assets, obtain intrinsic satisfaction from doing a good job, and carry out unrewarding tasks out of some sense of duty. The corporate performance of the CEOs does not depend on incentive and control schemes but rather on the creation of facilitative and empowering structures that include having the CEO to chair the board of directors, thus concentrating power and authority in one person.
There have been many attempts at empirically validating either agency theory or stewardship theory as the allegedly 'best way' to conduct governance. Although recent results of empirical studies do produce partial support for the positive impact of the stewardship model, the findings do not give unequivocal answers to the question about the relative superiority of agency or stewardship theory (Daily and Dalton 1994;Davis et al. 1997;Donaldson and Davis 1991;Fox and Hamilton 1994;Schillemans 2013). Hence, research tends to stress the contingency of the performance of agency and stewardship models on different cultural, contextual, and situational factors, and it increasingly focuses on the impact of different combinations of the two models. Hence, stewardship should not necessarily be seen as a complete substitute for agency-based governance, but rather as an alternative model based on different motivational assumptions that may modify, balance, or transform existing practices of agency-based governance (Caers et al. 2006).
In line with this, Schillemans observes that the stewardship theory advanced by Donaldson and Davis (1991) helps us to move beyond the clash of interests between equally self-serving principals 2 While the kind of trust-based horizontal governance relations in networks, organizations and communities area inherent to stewardship theory are a cornerstone of civilization and may be traced back to tribal and clan-based societies (Stout and Love 2018), the scientific conceptualization of stewardship and the attempt to draw out its consequences for management are a recent endeavor. and agents in the public sector and to chase away the fear that contract-bound agents-be they public employees, public agencies, or private service deliverers-may exploit the principals' limited information time, resources, and knowledge to shirk and do less to provide public solutions and services than they are supposed to do. Assuming, with stewardship theory, that subordinate agents in the public sector are not 'individualistic, opportunistic and self-serving', but rather 'collectivistic, pro-organizational and trustworthy' (Davis et al. 1997, p. 20) may help us not to waste public money on costly systems of control-based performance management (McCubbins et al. 1987) and avoid alienating and demotivating public employees who become subject to close and frequent monitoring that contains the stigma of punishment (Meier and Krause 2003).
Stewardship theory criticizes both the critical diagnosis of shirking advanced by agency theory (see also Pierre and Peters 2017) and the control-based cure offered by performance management, and it simultaneously provides the foundation for a new, trust-based management model that nurtures the intrinsic and pro-social motivations of public employees. The argument is not that subordinate agents are exclusively motivated by professional aspirations and pro-social goals but rather that a governance model that supports intrinsic and pro-social motivation and empowers the agents enables the principals to tap into the otherwise hidden resources, ideas, and energies of their agents. Hence, despite the structurally defined power asymmetries, principals and agents may actually share the same goals and play for the same team, thereby creating benefits for public leaders, employees, users, and society in general (Schillemans 2008, pp. 544-45). As such, it is reasonable to expect nurses to be motivated by improving public health, elderly care assistants to be driven by the ambition to provide a dignified life for the elderly, and social workers to strive to enhance the employability of the unemployed and to help them find a job. What stewardship theory suggests is that we draw on the consequences of this insight for public management.
The best way to capture the novelty, basic features, and consequences for public governance of stewardship theory is to compare it head to head with agency theory. To that end, Table 1 below compares agency theory with stewardship theory with regard to how they perceive the motivation of agents/stewards, the goals of public managers and employees, the role of public managers, the types of public management and leadership, and the role and organization of power. Table 1. Comparison of agency and stewardship theory (adapted from Schillemans (2008) and inspired by Hernandez (2012)).

Motivation of agents/stewards
Extrinsic motivation based on self-interest Intrinsic motivation based on self-realization and pro-social orientation

Goals of public managers and employees
Conflicting-principals and agents have diverging interests Shared-there is considerable overlap

Role of public managers
Reduce risk by enhancing control through strict performance management based on objectives, performance indicators, evaluation, and corrective measures Absorb risk by enhancing trust through the promotion of dialogue, empowerment, and joint learning

Type of public management and leadership
Transactional, top-down Transformative, distributive, horizontal

The role and organization of power
High 'power distance' allows powerful formal leaders to regulate and discipline less powerful employees Low power distance facilitates knowledge-sharing between empowered peers, coordination based on shared outcome responsibility, and self-governance within mutually agreed boundaries Two relatively coherent stories seem to emerge from the head to head comparison of agency theory and stewardship theory. Agency theory assumes that agents are motivated by individual utility maximization and envisages a conflict between principals and agents that the former is bound to lose, unless they develop a strict system of performance management and exercise top-down transactional leadership based on regulatory and disciplinary power. In contrast, stewardship theory assumes that public employees are motivated by carrying out meaningful professional tasks that create value for others and that the goals of public leaders and the employees who are stewarding them are basically the same, or will tend to be the same, if the leaders and managers: (a) engage in a trust-based, learning-focused dialogue with the employees about visions, goals, and tools (servant leadership, see Reinke 2004, and transformational leadership, see Bass and Riggio 2006); (b) empower them to engage in bounded self-regulation and invite them to exercise team-related leadership tasks (horizontal and shared leadership; see Pearce et al. 2008;Carson et al. 2007); and (c) involve them in making strategic leadership decisions on behalf of the entire organization (distributed leadership, see Bolden 2011 andGronn 2002).
In empirical reality, these contrasted stories and practices of public governance tend to be combined. However, a recent study of Dutch quasi-autonomous agencies conducts a survey experiment that shows that the perception of the actual use of elements from stewardship theory is lower than what the respondents would regard as optimal (Schillemans and Bjurstrøm 2019). This might be explained by the continued predominance of NPM among public administrators, and it begs the question of what can be gained by shifting the balance more in the direction of trust-based management supported by stewardship theory and New Public Governance (Mills et al. 2019).
Trust-based management encourages leaders and managers to share power, information, and resources with employees. And as Guy Peters (1996) explained long ago, we might expect that 'more empowered workers should be willing to work harder, share more ideas with management, and treat their clients more humanely since they are themselves being treated better' (p. 51). This expectation resonates well with empirical studies of staff empowerment in private service firms (see Fernandez and Moldogaziev 2011) that show that trust-based power-sharing improves effectiveness and productivity (Kirkman and Rosen 1999;Lawler et al. 1992); promotes innovativeness (Spreitzer 1995); and increases employee job satisfaction (Kirkman and Rosen 1999), organizational commitment (Guthrie 2001;Kirkman and Rosen 1999;Lawler et al. 1992Lawler et al. , 1995, and job involvement (Coye and Belohlav 1995). There is a dearth of studies of the impact of employee empowerment in PSOs and very few positive examples of what a gravitational shift towards trust-based governance may accomplish. This study seeks to change that by exploring the form and impact of trust-based governance in a public organization.

Methodology
While most of the recent research applications of stewardship theory in the public sector have studied the relationship between contract-governed projects and agencies, this study examines the relationship between public leaders and employees in a single public service organization. We have chosen to study an extreme case in which trust-based management and empowered self-regulation are deliberately and wholeheartedly pursued in order to assess the feasibility of public governance based on stewardship theory and explore if and how it delivers on the positive expectations.
The selected case is the Job and Activity Center (JAC) in Gentofte Municipality, Denmark. JAC provides empowering activities, education, and job training for adults with special needs. The target group consists of citizens who cannot readily gain a foothold in the labor market due to physical, psychological, and/or social challenges. It serves about 400 adults from Gentofte and other municipalities in the Copenhagen Capital Region together with outreach and outplacement to private firms. It has about 100 employees, including a group of seven leaders. The executive leader came to JAC in 2012 and initiated a gradual shift towards a trust-based management model that has gained considerable renown and won several awards. The leader frequently describes JAC as a '90% leaderless organization', but it seems more correct to say that leadership is exercised in new ways rather than being reduced in any way. One key condition for the new and somewhat experimental management model is that JAC is situated in a municipality that is not only affluent and well-managed, but also known to be highly innovative. Moreover, the Danish public sector is characterized by a low structuration of work, a reluctant implementation of NPM, and close and informal management-employee relations. Hence, JAC has developed and thrived under favorable conditions. Selecting an extreme case such as JAC renders it impossible to generalize the results to a larger population of public service organizations. However, the purposeful selection of JAC makes it possible to explore the positive and negative impact of a public management model aimed at replacing control-based with trust-based management, despite some bureaucratic and performance-related controls being deliberately preserved.
We decided to conduct a qualitative case study due to our interest in exploring the complex interaction between public leaders and managers and their employees, how this interaction is shaped by the recent introduction of a new trust-based management model, and how this model is described and assessed by the situated actors. In order to understand JAC as an organization, to acquaint ourselves with the people and practices, and to see how JAC has been discussed in public, we studied the many texts and documents available on the JAC website (https://jacinfo.gentofte.dk/), which contains information about the staff, organization, mission, target group, internal documentation systems, and the user-board. The website also contains press clippings and references to publications about JAC. As such, it provides a rich resource for document studies. Other important documents include the Best Place to Work survey and the Social Auditing Report that are both conducted by external agencies. These preliminary document studies were supplemented by qualitative interviews with two centrally placed leaders, two shop stewards and two other JAC employees, and the administrative director at the City Hall who is responsible for employment policy. We interviewed the executive leader and a middle manager to understand the background, design, and operation of the new governance model in theory and practice. We interviewed two shop stewards who had interacted with the leadership group and represented the group of employees, about whom they had good and broad knowledge. We also interviewed two ordinary employees who could give us further information about what it was like to work in an organization like JAC. Finally, we interviewed the administrative director for employment policy to get an outside-in look at JAC. The interviews were all carried out on-site, which provided an opportunity just to hang around, to talk to the users, and to get a glimpse of everyday life at JAC and the city hall. The interviews were semi-structured in order to get clear and distinct answers to our questions about the governance model and its impact on leadership practices, staff motivation, work practices, and empowered participatory governance, while still providing opportunity for the informants to tell their own stories and voice any possible concern or criticism. The interviews lasted about one hour each and were recorded with the individual informant's consent and subsequently transcribed and coded using the dimensions in Table 1 supplemented with codes pertaining to positive and negative experiences, problems, results, and achievements. We cross-coded the transcribed interviews and other textual material to enhance reliability and to use the qualitative data material to inform our analysis and to document key observations (for an overview of the data sources, see Appendix A).
We encountered one potential problem in relation to data collection: The informants were selected and approached with help from the executive leader of JAC, who has a vested interest in the JAC experiment being presented in a positive light, which might have biased the selection process. However, the informants were clearly informed about the confidential and anonymous use of data and we found that the informants spoke quite openly and did not present an overly rosy picture of the new governance model; at least not a picture that deviated from the official reports and evaluations (see later).

Governance, Leadership, and Empowerment in JAC
This section first describes the executive leader's attempt to transform the JAC management model and then compares the new management model with the key aspects of the agency and stewardship models summarized in Table 1 above.
When the new executive leader of JAC took office in 2012, the organization was based on hierarchical and compartmentalized steering; it had a negative, 'gossipy' culture; while an atrocious financial situation had necessitated cutbacks and layoffs. The new leader exploited this crisis to set a new, positive agenda that encouraged the remaining staff members to stand together, to 'give more than they take', and to become actively involved in developing and governing JAC. This was assumed to help to enhance the quality of services and to recruit more paying municipal customers in the long run. To show that the new executive leader was prepared to walk the walk as well as talk the talk, she formed two task forces with employee participation, which were made responsible for strategic development and creating a new documentation tool, respectively. In the gradual emergence of a new trust-based management model focused on staff empowerment, three things proved important: careful selection of a new, visionary, and unified leadership team; supportive involvement of the trade-union and shop stewards; and clear communication with all staff members about the kind of commitment and involvement that was expected from them if they wanted to work for JAC. Nevertheless, the first two years were somewhat conflict-ridden, until the new governance model finally received general acceptance. The new model was based on a joint vision according to which all citizens should be moved from A to B in terms of skills, competences, and life quality and that this movement should be clearly documented. A more recent, updated vision emphasizes the contributions to public value creation made by the citizens themselves.
The new, trust-based governance model in JAC involves the employees in all decisions except confidential personnel cases. Let us take a closer look at the new management model to see how it compares to the stipulations and recommendations of agency and stewardship theory. Starting with the former, there are clear signs that ideas and practices pertaining to agency theory can still be found at JAC. First, extrinsic motivation is found amongst the employees who admit that they appreciate receiving praise for successful team efforts and pecuniary rewards, such as bonuses for extraordinary achievement. However, it should be noted that both public praise and pecuniary rewards are based on visions and values that the employees have been involved in defining and that the entire team rather than its individual members are subject to public praise.
Second, there are conflicting interests, as not all JAC employees are interested in being empowered and actively involved in collaborative workplace activities. Some prefer to be told what to do, and then do their chores and go home. However, this is less a management-employee conflict and more within the group of employees who view the new governance model differently. Over time, the conflict has almost disappeared due to staff turn-over, which has changed the composition of the employees so that almost everyone now supports the new focus of staff empowerment and participation in collaborative workplace governance. In this assessment, we rely heavily on the accounts of the two shop stewards who are well acquainted with the opinions of the other employees.
Third, all staff interaction with users must be monitored and documented, but the applied documentation tool is not only co-created by the leaders and employees, it is also primarily an internal work tool helping individual staff members and their team to assess the problems and challenges of the users and to track and document progress and results. Hence, it is not used to assess individual employee performance. JAC has mechanisms for detecting the team-based budget transgressions, but such problems are seen as triggers of collective learning processes rather than mischief calling for disciplinary action.
Fourth, transactional leadership is practiced, as successful interventions are praised and bonuses are paid on top of wages. However, it should be noted that praise is given in plenary meetings and based on self-reported stories about exceptional results obtained by an entire team and that such bonuses are paid out on the basis of collective bargaining at the team level. Hence, the leadership is rather de-centered on the reward side. In contrast, the leaders assume a central responsibility for punishment. Punishment is used in relation to employees who either in theory or practice fail to commit to the basic JAC vision and values. Such employees first receive a corrective talk with a manager and are fired if the value conflict proves unbridgeable.
Finally, there are concrete examples of the executive leader having explicitly referred to her leadership prerogative in professional disputes with employees and having insisted on users being given one more chance or special, favorable treatment. Moreover, the many layoffs for which the executive leader was responsible in the first years were reported by several employees to have created a high power distance vis-à-vis the staff, which took much time 'and many cups of coffee' to overcome.
Turning to the analysis of ideas and practices that are implicitly founded on stewardship theory, the interviews make clear that the trust-based management ideas are full of positive energy and indicate exactly what JAC as an organization aspires to do in order to develop and consolidate. Starting with the employee motivation, ITM and PSM seem to be strong among the employees we interviewed and heard about. New employees are recruited based on their ITM and PSM. Job ads clearly signal that applicants must be driven by an interest in using their professional skills to create development, learning, and value for the users. Moreover, celebrating small and large successes is an integral aspect of the organizational culture, such as when a task is carried out particularly well, a goal is reached, and citizens are moved from A to B. In the Best Place to Work survey, 99% of the employees agree that successes are celebrated. One informant explains how: 'The celebrations help to give us a common language and team spirit. They also provide recognition of professional competences and allow us to take pride in what we do'. Another informant reports that: 'In the beginning, we were a little like "we're just doing our job . . . ". But now we've learned to stand tall and take pride in doing a good job'. A third informant claims: 'It's the many small ways that we're recognized for what we do-that's why we like our job'. The Best Place to Work survey reveals that 98% of the employees consider their work to be meaningful. A fourth employee explicitly refers to her job motivation: 'I'm not motivated by my wage. What motivates me is to feel that we can make a difference. That really gives me a kick. Also, we have room to develop ourselves, and there are few limitations on how we go about our work as long as we help to improve the lives of our users'. Hence, task-related self-realization and PSM seem to go hand in hand. In support of this, the Best Place to Work survey shows that 99% of the organization's employees appreciate how their work contributes to society.
Considering next the goals and interests of leaders and employees, both parties agree that the management and employees share a high degree of goal congruence. One of the leaders claims that: 'Basically, we have a shared understanding of the goals'. The other leader concurs: 'What's important is that we have the same basic values'. She continues: 'This is partly ensured by hiring people with the right values'. One of the shop stewards contends: 'If there are two conflicting parties that clash, there tend to be a lot of tactical and political games. But here we have a lot of trust and a high degree of openness about what we think and what we see happening in the organization. We have the same goals. That's why we collaborate with the leadership and have so few conflicts'. The executive leader confirms this account and reports that she works very closely with the shop stewards if there are problems with an employee. The shop stewards immediately report back to her if they have spoken with the trade union about problems with an employee. 'We've created a "we" rather than an "us" and "them"', she concludes.
The role of the leaders and managers at JAC is not first and foremost to instruct and control the employees-far from it-Instead, they see themselves as promoters of mutual trust. They aim to facilitate dialogue with and amongst the employees at all levels, empowering the employees by organizing self-governing teams and stimulating joint learning. Decision-making processes are made to be transparent and inclusive. According to one employee: 'The result is that we trust each other. We have a large degree of trust between leaders and employees' (Agertoft 2019). One leader explains that if he is approached by an employee who reports a problem requiring the attention of the leadership, the standard response is to facilitate dialogue between the parties involved and ensure that everybody is heard in the process. He explains: 'It's no use if I step in and try to solve the problem. They won't learn anything from that. It's better to create a process that helps them to manage and solve the problem themselves'. Such facilitative leadership results in a lot of meetings. One of the employees reports that the leaders motivate the staff to come to meetings by emphasizing how they are essential for joint and professional learning. 'This convinced most employees that they should come to the meetings, and now we're increasingly allowed to determine the content of the meetings', the employee tells us. Other employees concur with this view. The leaders seem to have succeeded in promoting trust and staff empowerment. In a survey from 2019, 100% of the employees respond positively to the question 'The leaders trust employees to do their work well without looking over their shoulder'.
The JAC leaders aim to exercise transformational leadership, but not in the traditional sense of developing and presenting a pre-designed vision and mission that the employees are expected to internalize and pursue. Rather, the employees are invited to co-create the goals and values of the organization, which instils a sense of ownership of the overall vision. The leadership group regularly stages joint discussions of strategic management issues, but they always 'set out the corner flags of the playing field' based on the jointly formulated goals and values, the legal requirements, and the budgetary constraints. Hence, the dialogue between leaders and employees always take place within an overall framework that defines the basic conditions on which the joint discussions build. The leaders are keen to remind the staff of these 'basic conditions', which are in part defined by the employees themselves but also subject to future revision through joint discussions. The Best Place to Work survey indicates that 98% of the employees agree that the organizational leaders clearly specify their expectations to joint efforts.
Based on this particular kind of transformational leadership, the JAC leadership aims to mobilize the employees' ideas, competences, resources, and energies through the exercise of distributed leadership and servant leadership. The informants talk at length about this, and their descriptions are confirmed by the recent Social Auditing Report (Socialtilsyn Hovedstaden 2019), according to which the staff experience a high degree of involvement in and influence on strategic and organizational matters. Hence, decisions that are normally considered a leadership prerogative are delegated either to self-regulating teams in which the employees exercise horizontal leadership of their peers or to the joint tasks forces and plenary meetings at the organizational level, where leaders and staff discuss a broad range of important issues. The daily operations at JAC are carried out by self-governing teams that are responsible for distributing tasks, determining whether specific users fit their program, selecting and developing professional methods, and taking financial decisions, including the use of subs within the allocated budget. The teams are also responsible for recruiting new employees. A leader from the leadership team will participate in the selection process, but the process is driven by the team members, who work together with the new employee. An informant reports that the result is that new employees feel 'chosen'. The teams are also responsible for allocating bonuses, and the process has proven to run surprisingly smoothly, with an exceedingly high level of satisfaction with the outcome. The Best Place to Work survey shows that 77% of the employees are satisfied with the wages they receive for the work they are doing. Finally, if a team encounters problems that it is unable to solve on its own, it can request assistance from an external team leader, who offers to coach and advise the team.
While the delegation of leadership responsibilities to self-regulating teams in which leadership is exercised in a horizontal manner through peer-to-peer interaction is relatively common, the massive involvement of all employees in the central strategic leadership decisions, which we find in JAC, remains relatively rare. Task forces combining leaders and selected employees were responsible for developing the overall organizational strategy in two rounds over the first five years, and a new, third round is under preparation. The new assessment and documentation tool was also developed by a joint task force. Moreover, the entire staff body was involved in a process leading to cuts. The leaders openly informed the staff at a plenary meeting about the poor financial situation, and the leadership received many ideas and suggestions that helped turn the budget around. One of the employees we interviewed described how: 'We felt we had real influence in the budget process-where we had to make cuts. Not all of our ideas and suggestions were followed, but they were all listened to and taken seriously'. Finally, the organizational structure has been changed through a joint process whereby the employees ended up designing two alternative organizational diagrams that were put to a vote. Indeed, staff involvement is carried to the extreme as all of the leadership group meetings have been opened up for all employees, except for discussion of confidential personnel issues.
More and more strategic and organizational issues are being subject to joint discussions in task forces or plenary meetings, which tends to make the mandatory works council superfluous. Employees reported feeling silly to have formal meetings in the works council to confirm decisions that have already been taken by leaders and employees in task forces and plenary meetings. The executive leader and shop stewards are therefore now working on a new model wherein the plenary staff meetings are regarded as types of work council meetings.
Staff involvement in self-governing work teams and discussions of organizational development mean that there are many meetings. However, several of the employees tell us that the level of involvement and active engagement may vary as long as the employees are committed to improving the lives of the users. 'It's more the approach to your work that matters than your involvement in all aspects of the organizational development', we were told.
The final aspect of the management model concerns the role and organization of power in JAC. The executive leader is convinced that it is better to lead by empowering and trusting the employees than by controlling them (Agertoft 2019). However, 'trust-based leadership is not the same as absence of leadership', she claims. The JAC leaders are visible, and they are keen to get out of the office and participate in dialogue meetings throughout the organization. Employees and shop stewards agree that the leaders in general and the executive leader in particular are very approachable. There is always an open door, and you can always come in for a coffee and discuss what is on your mind. This helps to minimize the leader-employee power distance, and distance is not necessary, since the leadership places a high premium on input from the employees, joint coordination, and shared responsibility for goal achievement.
In sum, the JAC management model is clearly, albeit unknowingly, informed by the stipulations of the new stewardship theory. Nevertheless, as described in the beginning of this section, some elements remain from the old management model that draws on classical agency theory. Management and leadership are primarily based on a soft power that frames and facilitates involvement in the collaborative processes on both the operational and strategic levels. However, this soft power is ultimately backed by a hard power (see Nye 2008), although this has shrunk to the occasional use of corrective measures and firing of employees who, after several conversations, do not fit the unique JAC work culture. Hence, in JAC, public management based on stewardship theory has relegated management based on agency theory to a somewhat marginal position.

Outcomes
Does it matter which management model is applied in a PSO such as JAC? Does it affect the organizational outcomes produced by JAC? Apparently, it does. In 2018 and 2019, JAC received the 'best public workplace' award in Denmark, and it was awarded the title 'best public workplace in Europe' in 2019 and 2020 by the European organization Best Place to Work. This result was achieved based on three outcomes. First, the motivation and well-being of employees receives top scores. All of the responding employees in the Best Place to Work survey say that they experience a higher well-being and work engagement, and 100% of them answer that JAC is a 'fun place to work'. In the last two years, fewer than 2% of the employees have left JAC. One employee explains as follows: 'The trust and recognition that we're met with as JAC employees is fantastic. Three months after I was hired, I was involved in making a new strategy for our day services. The faith in what we can do as employees gives us a lot of satisfaction and involvement with our work'. Echoes of this overly positive evaluation of employee empowerment is found in the accounts of the other employees as well.
Second, the perceived user satisfaction, user empowerment, and the overall quality of the work delivered by the JAC employees have all increased. We do not know of any survey of user satisfaction, but the external assessment of the quality of work conducted by the Social Auditing Agency gives JAC a 4.8 score out of 5. The Social Auditing Report also concludes that the citizens served by JAC appear trusting and satisfied, and they feel that they receive high-quality services that help them develop their full potential (Socialtilsyn Hovedstaden 2019). As a concrete indicator of the high trust and satisfaction among the users, recent statistics show that the use of force has fallen by 64% in the period 2014-18. Moreover, the general experience is that the new management model has strengthened the employees' work with the users of their services and empowered the latter. One of the employees with whom we spoke reports of how 'the users' needs and demands are met differently now, because we have more surplus energy and empowerment. The users feel the satisfaction we have from working together as a team, and we pass that on to them'. One of the leaders confirms that 'many of the employees experience that they can really help and "be something" for the users. This is something I see much more today than beforehand'. Another employee explains how the trust-based empowerment of the employees has strengthened their professionalism, from which the users benefit: 'We constantly look at the JAC users to assess what's best for them. What we offer in our team might not be the best. We can proudly say we're good at what we do, but it might not be the best for a particular user'. This observation is corroborated by another employee: 'We've had three open spots on our team for new users, but turned down four because they won't benefit from what we can offer. In that sense, our work has become more serious and professional'. Lastly yet importantly, there seems to be a spill-over in the sense that trust in and empowerment of the employees result in more trust in and empowerment of the users. One of the employees explains this as follows: 'That we have more autonomy also means that we give the users more influence, albeit with due respect for their different levels of functional competence. Some of them have a good sense of what different activities cost and what they want, and they're given influence on what we do and how we do it. Others are merely given a choice between different activities'. Hence, employee empowerment appears to also foster user empowerment contingent on the cognitive capacity of the users.
Third, the overall organizational performance, including cost efficiency, has improved. Not only has input from the employees helped JAC to deliver the same or better service for less. In the period 2013-18, sick leave fell by 30%, which saves JAC and the municipality considerable resources. Many of the informants refer to how there are fewer persons on sick leave, one mentioning how: 'We pride ourselves on being professional. So you don't call in sick just because you don't feel like going to work. If you do that, you're unfair to your colleagues-who should do the work you're supposed to do? We really feel that we make a difference'. Increased presenteeism resulting from attempts to boost ITM and PSM finds support in recent research based on Danish panel data (Jensen et al. 2019). Still, one might suspect that the strong sense of obligation to colleagues, workplace, and organizational values may become a source of stress. However, fewer than 1% of the JAC employees have been on sick leave due to stress-related factors. This number should be understood in the context that working in the field of disability and psychiatry is generally known to be emotionally strenuous.
The bright side of the JAC management model, which implicitly leans heavily on stewardship theory, might be countered by considering the dark side of trust-based management. Hence, empowered employees might work at a slower pace, as they spend time consulting their team members and using their professional skills to meet the particular needs of service users. Their behavior may also be less consistent and uniform than that of employees operating in bureaucratic organizations based on hierarchy, rule-following, and control. Finally, the costs of selecting, training, and involving employees may incur a loss to the organization Lawler 1992, 1995). While we are repeatedly told that the leadership group's preparation of staff involvement in strategic management decisions is time-consuming and demanding, there are no complaints about the other dark issues, despite encouragement of the informants to speak their minds and to speak openly about positive as well as negative experiences. This might be because slower, more individualized service provision is seen as an integral component of a stronger professionalism and because employee involvement in empowered, collaborative governance on both the team and organizational levels creates better solutions with broad ownership, which enhances efficiency in the day-to-day operations-at least the informants suggest as much.
However, there is one undeniably dark side of the JAC story, which relates to the initial employee turnover (Seibert et al. 2011). The mismatch between the new trust-based management model and a large group of staff members who were uninterested in taking more responsibility and participating in self-regulated goal attainment resulted in the lay-off of 20% of the original employees. The JAC leaders report that immediately after the organizational turnaround, 'there were some employees who said: "This isn't for me. I need a leader who can take the decisions and tell me what to do"'. Although extensive attempts were made to develop the competences and motivation of the employees and to encourage them to engage in the new empower participatory governance model, some still find it to be 'too much'. As one of the leaders puts it: 'Some employees quit their job, telling us that this isn't a way for them to work-with all this influence, participation, and self-regulation'. Hence, offers of being trusted are not always received positively and may even be perceived as a kind of poisoned chalice (Bentzen 2019;Skinner et al. 2014). The executive leader is well aware that JAC is a workplace for a particular type of employees, which begs the question: What do we do with all of the employees who do not want to invest as much in their job as JAC employees? If all public workplaces became like JAC, where would the 20% of the employees who quit or were laid-off go?

Discussion
PSOs that do not want to live with the many problems associated with the control-fixated performance management regime that in recent decades has been recommended by NPM and legitimized by classical agency theory may want to seek inspiration in the assumptions proffered by the new stewardship theory that has been imported to public administration theory from corporate governance theory. However, there is not a lot of evidence that a management model relying on trust-based empowerment of employees may work in PSOs. Our analysis of JAC in Gentofte aims to remedy this problem by providing a relatively thick description of a relatively successful case of trust-based management that may serve as a beacon and inspire others to test out new, trust-based governance models.
Our main findings are that trust-based management inspired by stewardship theory provides a feasible governance model for local PSOs seeking an alternative to governance models steeped in bureaucratic control and performance management and that most of the positive expectations have been fulfilled, which has created a triple-win in the sense that the new public management model seems to be better for the employees, better for the users, and better for the public sector. Although the new forms of governance and management at JAC are not modeled after the principles of stewardship (which are unknown to the JAC personnel), the analysis shows how all of the five core aspects of stewardship theory can be identified in the JAC management model. Equally importantly, the new management model has proven to be relatively easy to develop, implement and consolidate, although not without some stress at the managerial level. However, we should bear in mind that JAC has benefitted from favorable conditions and that it has been spurred by a highly committed public leader who has invested extensive time, energy, and prestige in the project. The absence of these conditions may seriously impede replication.
The analysis of the outcomes clearly demonstrates the attractiveness of a shift towards the ideas and practices inherent in stewardship theory and trust-based management. Employee motivation and well-being is enhanced, the users seem to receive better service, and the service delivery costs are reduced. These results have helped JAC to earn several workplace awards, which has in turn conferred beacon status on the institution. Danish ministries, other municipalities, and public administration and work-life researchers have been lining up to visit and study JAC. Hence, diffusion of the trust-based management model to other Danish PSOs is most likely and these organizations may be able to obtain some of the same benefits as JAC even if they do not commit to a full scale implementation.
The positive results obtained through the trust-based management of self-regulating teams of empowered employees is confirmed in case studies of the innovative use of autonomous nursing teams in the independent Dutch homecare provider Buurtzorg in the Netherlands, which has also earned itself high employee and user satisfaction while managing to cut costs (Drennan et al. 2018;Gray et al. 2015). Still, the advantage of JAC as a beacon for the public sector is that JAC is a traditional local PSO, whereas Buurtzorg is a new independent service provider that does not have to follow the rules typically found in public bureaucracies. Moreover, compared with Buurtzorg, JAC complements the self-regulating teams with employee involvement in strategic management at the organizational level. As such, JAC sets a new standard for trust-based management in PSOs.
Next to the apparent feasibility and positive outcomes of the trust-based management model, the most significant empirical finding is that the predominance of ideas and practices inherent in stewardship theory seems to be predicated on retaining some elements of governance and management founded on agency theory. Hence, as noted above: the new soft power associated with trust, empowerment, and collaborative governance is complemented by elements of hard power that are necessary, for example, to trim personnel in order to better match the employees and the organizational values. This finding confirms existing research (Caers et al. 2006;Schillemans 2013), which has stressed that it is not so much a matter of replacing one governance model with another, but rather a question of finding the right combination; ultimately, it is the mix that matters. This finding resonances well with new research on hybridity that emphasizes the competition and co-existence of different governance paradigms (Torfing et al. 2020) and different modes of governance (Koppenjan et al. 2019).
While the main focus of this article has been empirical, these empirical observations may also give rise to a theoretical contribution; hence, it is interesting to see how transformational leadership, focusing on the formation and communication of visions and values, is not only used to motivate individual employees, but also to frame collaborative interaction and strategic dialogue between leaders and employees. It is also interesting how the visions and values that are communicated are a product of joint discussions between leaders and led. These observations challenge the traditional notion of transformational leadership (Bass and Riggio 2006) that tends to stress the central role of charismatic leaders in setting the course of the ship. The idea that value-based narratives may be jointly formulated and then serve to frame horizontal interaction processes is a part of new theories of metagovernance (Torfing and Sørensen 2009), and future research may explore how transformational leadership theory can be broadened through an exchange with theories of metagovernance. However, both theories of transformational leadership and metagovernance have been blind to the idea that the discursive framing of interaction relies on input from those who are led or metagoverned.
It is not possible to generalize the empirical results from a study of a purposively selected singular case that has grown in highly fertile soil to a wider population. However, the statistical generalization of results was never the purpose of this article, which instead has sought to explore and document the feasibility of applying principles inherent to stewardship theory to PSOs and to assess the outcomes in order to provide an antidote to the prevalence of control-fixated performance management, which often ends up crowding out the intrinsic motivation of public employees. JAC may indeed provide a source of inspiration for other PSOs that might want to adopt some of the key principles and practices associated with stewardship theory while adapting them to their specific context. The lesson they can learn from JAC is that trust-based management that involve employees in self-regulating teams and strategic discussions may help to achieve a triple-win in terms of improved job engagement, user satisfaction, and cost efficiency.

Conclusions
As we have seen, stewardship theory developed in the private sector, where it challenged the notion of an inherent conflict of interest between principals and agents and replaced it with the path-breaking (yet plausible) idea of a high degree of goal congruence. As we have seen, this new basic assumption spurs the development of a trust-based management model, which is of great relevance for service-producing public sector organizations that are seeking new ways to boost the intrinsic task motivation and public service motivation of their employees in order to cope with the cross-fire between rising citizen expectations and scarce public resources.
Aiming to make up for the dearth of studies that apply stewardship theory to PSOs and focus on the relationship between managers and employees, this study has conducted a qualitative study of the Job and Activity Center (JAC) in Gentofte that has aimed to swim against the stream and implement a new, trust-based management model that empowers employees by involving them in autonomous teams and strategic co-management. The study concludes that the new governance model is aligned with the core principles of stewardship theory but retains some elements emanating from agency theory. The trust-based management model has been relatively simple to implement and has produced a range of desirable outcomes, including enhanced staff motivation, better services, and cost containment, which has earned JAC several workplace awards. The drawback of the model is that it depends on a good organization-employee match, which has resulted in the loss of a large number of employees. As such, one of the explanations of how ITM and PSM have been boosted is that the employees who are short on these forms of motivation have been fired, replaced by new employees with the right work ethic. This observation tends to suggest that, in a paradoxical way, the JAC governance model succeeds in using soft power conditional on the use of hard power. As observed in previous research, trust and control need not be opposites, but may function as parallel strategies that may supplement and balance each other in handling risk (Edelenbos and Eshuis 2012; van Thiel and Yesilkagit 2011;Weibel and Six 2013).
Further research is needed to justify the claim that PSOs will benefit from a gravitational shift from agency to stewardship theory. Comparative case studies will enable us to explore the impact of contextual conditions for the trust-based governance of PSOs to thrive and produce desirable outcomes. The qualitative comparative analysis of medium n cases of trust-based governance may help us to identify the competing constellations of factors, such as institutional designs and forms of leadership, which generate successful outcomes. Large n studies of PSOs may draw on multi-level regression analysis to study the causal impact of institutional designs drawing on agency and stewardship on employee motivation, service quality, and cost containment. We clearly need such studies, since we have many studies of the shortcomings of management models based on agency theory and seem to lack good, well-documented alternatives.

Conflicts of Interest:
The authors declare no conflict of interest.

Key documents (14)
• External supervision rapport of JAC 2020 (1) • Information folder about JAC 2019 (1) • Rapport from Best Place to Work 2019 and 2020 (2) • Articles about JAC in the press (4) • News from the website of JAC (4) • Job postings announced at the website og JAC (2)