Interactive Political Leadership in Theory and Practice: How Elected Politicians May Benefit from Co-Creating Public Value Outcomes

This paper argues that elected politicians may strengthen their political leadership role by initiating, orchestrating and engaging in the co-creation of public value outcomes. The collaborative turn in public value theory shows how public managers may mobilize the knowledge, ideas and resources of users, citizens and organized stakeholders, but it has so far neglected the role of elected politicians who tend to be reduced to a legitimizing sounding board for public managers aiming to advance public value creation in collaboration with a plethora of public and private actors. This paper seeks to compensate this benign neglect by advancing a new notion of ‘interactive political leadership’. This new construct aims to conceptualize the way that elected politicians may develop new and better policy solutions through a problem-focused interaction with relevant and affected actors from the economy and civil society, including users, volunteers, citizens and other lay actors. The theoretical argument about the development of interactive political leadership, which takes us beyond the traditional forms of sovereign political leadership that perceives politicians as ‘elected kings’, is illustrated by empirical examples drawn from local, national and supranation levels of government.


Introduction
The aim of this paper is to show that elected politicians may benefit from initiating, orchestrating and engaging in the co-creation of public value outcomes. The collaborative turn in public value theory emphasizes the role of public managers and their employees in co-producing and co-creating public services and service systems (Alford 2010;Bryson et al. 2017;Crosby et al. 2017). However, it has so far neglected the role of elected politicians who seem to be reduced to a legitimizing sounding board for public managers aiming to advance public value creation in collaboration with a plethora of public and private actors. This paper seeks to compensate this benign neglect by advancing a new notion of 'interactive political leadership'. This new construct aims to conceptualize the way that elected politicians may develop new and better policy solutions through a problem-focused interaction with relevant and affected actors from the economy and civil society (Ansell and Torfing 2017). The theoretical argument about the development of interactive political leadership, which takes us beyond the traditional forms of sovereign political leadership, is illustrated by empirical examples drawn from the local, national and supranation levels of government. The illustrative examples add flesh and blood to the theoretical concept of interactive political leadership and allows us to reflect on its conditions. The paper aims to contribute to empirically informed theory development that can help bridge the gulf between political leadership theory and collaborative theories of public value creation.
The paper is based on the claim that the public value perspective (Moore 1995(Moore , 2000Meynhardt 2009; Benington and Moore 2011;Bozeman and Johnson 2015) is a game changer for public managers in particular and the public sector as a whole, as it transforms the way that public organizations deliver, improve and innovate public services, and the way that public solutions to complex problems are designed and implemented. Hence, when public managers focus broadly on the production of public value outcomes rather than more narrowly on the administration of particular programs and the delivery of specific services, it becomes clear that a broad range of public and private actors, including service users, citizens and civil society actors, may contribute to defining and producing public value for specific groups and society at large (Stoker 2006a;Alford 2010;Page et al. 2015;Bryson et al. 2017;Crosby et al. 2017). Hence, the public value perspective encourages public managers to embrace co-creation, defined as the process through which two or more public and private actors collaborate in defining problems and designing and implementing solutions .
While there is a growing focus on how public managers and frontline personnel co-produce public services with service users, citizens and voluntary 'third-sector' organizations (Brandsen and Pestoff 2006;Osborne and Strokosch 2013;Brandsen et al. 2018), there has only been scant attention paid to the role played by elected politicians in public value creation, beyond their role as 'authorizers' of the public value pursued by entrepreneurial public managers (Moore 1995). In order to fill this gap, we propose that politicians may improve and qualify their political leadership by initiating, orchestrating and participating in the co-creation of public value outcomes. Politics is essentially about articulating and allocating values for a group, population or society (Easton 1965), and drawing on Tucker (1995) and Nye (2008), we define political leadership as the performance of five related tasks: (1) the discursive formation of the political community that they aim to lead; (2) the identification of the societal problems and challenges that call for public action; (3) the development of innovative, feasible and robust policy solutions; (4) the mobilization of widespread support for the implementation of these solutions; and (5) the highlighting of the public value that results from public policy-making. Elected politicians can improve the performance of these political leadership tasks through co-creation of public value outcomes with relevant and affected actors from the public, private and third sector. Hence, rather than merely endorsing the public value propositions advanced by public managers, elected politicians may themselves engage actively in public value creation by seeking to mobilize a broad range of public and private actors that can provide the inputs they need to formulate and implement new and better solutions to wicked and unruly problems. Such an endeavor is likely to increase input and output legitimacy and thus enhance trust in government.
The structure of the paper is as follows: The first section reviews the literature on public value, arguing that the public value perspective is a game changer that highlights the importance of co-creation as a core tool of public administration and governance. The second section looks at how elected politicians and their exercise of political leadership may benefit from initiation, orchestrating and engaging in the co-creation of public value. This discussion prompts the definition of the concept of interactive political leadership. The third section scrutinizes the concept of interactive political leadership, compares it with other concepts of political leadership, draws out its institutional and democratic implications, and reflects on its scope conditions. The fourth section provides and compares a number of empirical examples of interactive political leadership in order to explore its practical relevance and further elucidate its scope conditions. The conclusion summarizes the key findings and suggests some future research avenues.

Public Value as a Game Changer
When the public sector expanded in the postwar era, it was organized as a Weberian and Wilsonian bureaucracy with a sharp distinction between politics and administration, hierarchical control, a horizontal division of labor, impartial rule-based decision-making, and meritocratic recruitment (Wilson 1887;Weber 1946). Despite the equity and efficiency gains derived from rule-governing, centralized resource allocation and horizontal specialization, the public sector became subject to fierce criticisms from neoliberal public choice and principal-agency theorists, who accused the public sector of being rigid and inefficient and of delivering poor and costly services (Downs 1967;Niskanen 1971;Myerson 1982).
Neoliberal critics and advocates of New Public Management reforms compared the ossified and wasteful public sector to the flexible and efficient private sector (Osborne and Gaebler 1992). The political attack on the public sector-epitomized by President Reagan's famous statement that 'government is not the solution, but the problem'-created an enduring inferiority complex within the public sector. Whereas the private sector was praised for its ability to use competition and innovation to produce private value that was validated by large numbers of consumers and subsequently appropriated as profits by shareholders, public bureaucracies were depicted as an unimaginative and unproductive parasite feeding off the value created by the private sector (Benington and Moore 2011, pp. 7-8). Moreover, it was readily asserted that private contractors would be far better at producing low-cost and high-quality services in response to the needs of private users, and that the implementation of a strict performance management system was necessary to secure goal attainment and reduce slack in public organizations (Hood 1991). Finally, public norms and values aiming to ensure the rules of law, equity, transparency and accountability were seen as red-tape rules preventing efficient problem solving and service delivery (Osborne and Gaebler 1992).
Against this background, it is difficult to overestimate the positive impact that the discovery and development of the public value perspective has had on the public sector in general and public managers in particular. First, the public sector is described as a unique type of organization with a distinct revenue source (taxation) and a distinct value form (public value) (Moore 1995(Moore , 2000. The public sector does not produce private value for a small group of shareholders. Rather, it produces politically mandated public value for specific groups of citizens and society at large. Second, public managers are no longer portrayed as unimaginative and wasteful squanderers of value extracted from the private sector, but as curious explorers and change agents aiming to discover and define what would be valuable for the public sector to do in order to meet social needs and solve societal problems. Like managers in the private sector, public managers are inventive, well-intended and mission driven. As such, the public value perspective helps public managers to restore their self-worth as public managers (Rhodes and Wanna 2007, p. 407). Finally, the public value produced by public service organizations is not supposed to be validated merely by individual consumers operating in the new service markets based on the free choice of service providers, but supposedly subjected to democratic political debate between public managers, elected politicians and relevant stakeholders (Moore 1995).
Despite the fact that the public value perspective emerged in the heyday of New Public Management reforms and readily acknowledged that the public sector was underperforming and could learn from the strategic management thinking from the private sector (Benington and Moore 2011, pp. 7-9; see also Ongaro and Ferlie 2015), it offered a way out of the above-mentioned public inferiority complex. It rescues the sense of public purpose and community orientation in government institutions, nurtures the idea that public managers are closer to being mission-driven knights than self-interested knaves (Le Grand 2003), and insists that political and democratic debate trumps markets when it comes to endorsing public value propositions.
The core concept of public value is defined in terms of the social purpose of the public sector. It is measured in substantive and symbolic rather than fiscal terms and embedded in laws, regulations and mission statements that are politically authorized and guide the daily operations of public organizations (Moore 2000). Rather than reflecting the professional norms of the public employees, or the nature of the concrete tasks carried out by public organizations, the concept of public value captures the positive impact that public interventions may have on societal problems and social needs. It is a multi-dimensional construct that is shaped by the collectively expressed and politically mediated preferences of public managers, elected politicians and selectively involved users, citizens and private stakeholders (O'Flynn 2007, p. 358). The inherent tension between the contribution of public sector officials to the identification and definition of the public good and the influence of users, citizens and civil society organizations is well captured by the definition of public value as something that adds value to the public sphere and is valued by citizens (Benington and Moore 2011).
According to (Moore 1995(Moore , 2000, the identification and pursuit of public value is the central activity of public managers. Public value gives purpose and direction to the daily operations of public organizations. Public managers play a crucial role in formulating value propositions that guide public organizations, but they are not free to judge any organizational purpose as valuable to the public. Hence, in order to legitimize their public value proposition and obtain access to financial resources and delegated power, they must consult elected politicians, interest groups and citizens in their authorizing environment to garner support for the social purpose they want to pursue. When their public value proposition has been authorized through a process of political and democratic deliberation, public managers must focus their attention on the operational capacity of their organization in order to make sure that it commands the necessary finance, technology, skills and competences and communication structure to successfully pursue the public value set out in its mission statement and imbedded in laws and regulations. Together, the relations between public value, the authorizing environment and organizational capacity make up the 'strategic triangle' that helps public managers to understand the strategic challenges at stake and reflect on the complex choices that they face. At its core, Moore's public value perspective is essentially managerial. Public managers are not only Moore's target audience, but also the primary vehicle in the production of public value. Whereas public managers must seek upward authorization of their public value propositions through a dialogue with politicians, experts and private stakeholders that oversee the activities of public organizations, the initiative to define and propose what counts as public value lies with the public managers. Moreover, although the value chain perspective emphasizes users' and citizens' co-production of the services they receive or the regulations they are subjected to, it is the downward and inward management of their own organization that enables public managers to pursue their public value propositions (Moore 2000, p. 198). Hence, while the know-how and capability that is needed to achieve public value outcomes sometimes lies outside the organization led by the public managers, the organizational capacity of their own organization is usually sufficient to ensure the production of the planned results. Exploiting the organizational assets of their own organization, and occasionally supplementing these assets with input from the external environment is, together with the formulation and authorization of public value propositions, the key task of public managers.
While the emphasis on the distinct contribution of the public sector to the production of public value constitutes a neo-Weberian rebellion against the efficiency-focused New Public Management paradigm (Pollitt and Bouckaert 2004), the celebration of the proactive role of public managers brings the public value perspective in line with the managerialism of New Public Management (Hood 1991). As such, New Public Management calls for strategic management based on goal steering and performance management in order to curb the opportunistic behavior of public employees. Public managers are subjected to high-powered incentives schemes that penalize managerial failure and reward success. Hence, New Public Management seeks not only to 'let managers manage', but also to 'make managers manage' (Norman 2001).
The managerial approach to public value creation has been criticized by Rhodes and Wanna (2007) and re-articulated by Stoker (2006a). Rhodes and Wanna (2007) claim that Moore's public value perspective might fit the pluralist political system in the US, but that it is incompatible with the Westminster Model, in which there is a hierarchical relationship between elected officials and non-elected civil servants who do not enjoy much freedom to explore and define the public value outcomes. As such, Rhodes and Wanna criticize the alleged supremacy of public managers for exaggerating their autonomy, entrepreneurship and ability to do politics, and for overlooking the crucial role of elected politicians, political parties and ministers for setting the course for the public sector and defining the kind of public value to be pursued by public service organizations. Alford (2008) comes to Moore's rescue by arguing that the authorizing environment, which among others, comprises elected politicians, places a legitimate limit on the autonomy of public managers to define what counts as public value.
He also problematizes the sharp line of demarcation between politics and administration that seems to inform the criticisms voiced by Rhodes and Wanna. Stoker (2006a) takes a different and less confrontational approach than Rhodes and Wanna when evaluating Moore's public value perspective. In a re-appropriating gesture, Stoker views the public value perspective as a new management style that supports the recent emergence of networked governance. Public organizations aim to create public value outcomes, but there is a need to recognize a wide range of stakeholders, including citizens, interest organizations and civil society actors, when it comes to defining and achieving public value outcomes. There is whole range of private actors that can support public organizations' pursuit of public value outcomes by providing support, knowledge and resources, and new technologies that make citizen engagement easier, more flexible and less time consuming. In short, the public value perspective and the growing appreciation of networked forms of governance would appear to work hand in glove. Although this conclusion is reached without direct criticism of Moore's initial formulation of the public value perspective, it decenters the role of public managers by urging them to govern in and through the construction of networks of public and private actors. This view was later echoed by O'Flynn (2007).
Recent elaborations of this line of argument (Alford 2010;Bryson 2011;Page et al. 2015;Bryson et al. 2017;Crosby et al. 2017) support the claim that public value is a game changer that transforms the modus operandi of the public sector. Far from being the privilege of public managers acting as guardians of the public interest, public value production involves a broad range of public and private actors that can all contribute valuable experiences, ideas, forms of knowledge and resources to the fulfillment of unmet social needs and the creation of innovative solutions to wicked problems. Frequently, private actors have important assets and resources that can be mobilized in the process of defining and achieving public value. Users have first-hand experience with problems and challenges; they are experts in their own lives and can help evaluate new solutions. Civil society organizations have creative ideas, alternative ways of doing things, and scores of volunteers. Private firms have access to capital and new technology, and experience developing business models. Mobilizing all of these assets in the creation of public value outcomes recasts the public sector from being a primarily legal authority regulating and disciplining citizens and a service provider aiming to provide satisfactory services to its customers, to being an arena for the co-creation of public value outcomes. The implications of this insight are wide-ranging, as it calls for a more distributive, horizontal and integrative leadership (Torfing 2016), the construction of physical and digital platforms that support the emergence of collaborative arenas (Ansell and Gash 2017), and new forms of evaluation that supports co-creation of innovative public value outcomes (Patton 2010). (Moore 1995(Moore , 2000 does recognize the role that private actors play in authorizing public value propositions and co-producing services, and Benington and Moore also talk about the emergence of a 'new pattern of co-creation' (Benington and Moore 2011, p. 15). However, on closer inspection, the 'co-creation' of public value outcomes with relevant and affected actors is reduced to the 'co-production' of predetermined public services in a dyadic relation between service providers and service users. The reduction of co-creation to co-production is particularly clear when we are told that 'clients play a crucial role in co-production' and that the success of government agencies depends on 'millions of individuals accepting the obligation and doing the duty' (Benington and Moore 2011, p. 269). This is still far from a situation in which a broad range of public and private actors, including citizens, civil society organizations and private firms, collaborate in order to define and pursue public value outcomes (Nabatchi 2012).
The idea of an intrinsic link between public value production and co-creation based on cross-boundary collaboration between multiple actors is not new. The Ostroms and colleagues gradually came to the same conclusion. Early on, the Ostroms insisted that the public sector produces public goods rather than private goods and claimed that it was unlikely that a single integrated bureaucracy was the best provider of public goods (Ostrom and Ostrom 1971). Instead, we should opt for multi-organizational arrangements. A few years later, Eleanor Ostrom reported empirical findings demonstrating that the performance of local police is better when the police co-produce services with local citizens and community actors (Ostrom 1973;Ostrom and Whitaker 1973). In a later work, she reported on studies of sanitation problems and primary education, which show that the co-production of public goods and services through contract-based collaboration between public and private actors is crucial for achieving public value outcomes in developing countries (Ostrom 1996). While this leaves us with a sense of co-production as an exceptional add-on to public bureaucracies in some areas and some countries, Osborne et al. (2016) claim that co-production is inherent to the public sector due to its service-dominant logic, which means that users play a pivotal role in the integrated processes through which public services are produced and delivered. Finally, Osborne and Strokosch (2013) supplement the original focus on 'consumer co-production' that aims to empower users in discrete processes of service delivery with a new focus on 'participatory co-production' in which users participate in strategic planning and service design and on 'enhanced co-production' that involves relevant and affected actors in developing new and innovative public service solutions. From here, there is only a short step to embracing the idea that co-creation aims to include a broad range of actors with relevant assets in the production of public value outcomes .
The concept of co-creation builds on the assertion that public value production will benefit from the mobilization of resources that are not available within public organizations but are widely distributed among a broad range of relevant and affected actors (Ostrom 1996(Ostrom , p. 1079. Hence, while the New Public Management has attempted to deal with the cross pressure between citizens' rising expectations to public problem-solving and service provision and the scarcity of public resources due to fiscal constraints by aiming to use existing resources in the most efficient way possible (Radnor and Osborne 2013), co-creation aims to mobilize new and different resources and ideas to make ends meet and spur public innovation (Torfing 2019). Hence, while there have always been examples of co-production and co-creation, the new thing is that governments at different levels have begun to embrace co-creation as a core governance tool that spurs public value creation through collaborative networking.

Co-Creation of Public Value Outcomes as a Tool for Elected Politicians
So far, the co-creation of public value outcomes has mainly been an administrative endeavor. Public service organizations are short of resources and many societal problems are too complex for public administrators to solve singlehandedly. Moreover, the public sector is often incapable of influencing the behavior of citizens, firms and organizations and secure desired outcomes without involving intermediary organizations and associations, or the social actors themselves. As a result, public professionals increasingly invite users, community groups and private non-profit or for-profit organizations to participate in collaborative governance processes (Warren 2009). Administratively initiated co-creation is primarily motivated by the search for effective problem solving, but it also carries a democratic potential that resides in the effort to align what governments do and what citizens want.
To reap the fruits of co-creation in full, we need to realize that co-creation of public value outcomes is not a tool reserved for administrators only. It may also provide a useful tool for elected politicians who tend to find themselves squeezed between the urgent need for political leadership of policy innovation in the face of wicked problems, and the imminent risk of being side-tracked and losing influence on public policy making. On the one hand, elected politicians are being called upon to find new and better solutions to the pressing problems confronting Western societies. On the other hand, many politicians suffer from the development of tunnel vision as a result of their hard work in sector-specific standing committees, from being decoupled from policy-making processes that are frequently dominated by executive administrators and policy advisors and from political insulation from citizens and stakeholders who participate in governance networks that are facilitated and controlled by the administration (Kjaer and Opstrup 2016). Frequently, politicians who are a part of government are involved in drafting legislation and forging political compromises in relation to key political decisions. However, there is a strong tendency for politicians elected for parliamentary assemblies at various levels of government to be locked into time-consuming political committee work, technocratic policymaking processes, forums for administrative oversight and mediatized political blame-games, while simultaneous being locked out of key policy-formulating networks and attempts to metagovern such networks (Koppenjan et al. 2009;Ansell and Torfing 2017).
An additional set of constraints further limits the ability of elected politicians to exercise political leadership. First, the political community is fragmented and polarized along material and post-materialist lines that make it difficult, if not impossible, to create a tendentially unified demos (Dalton 2018). Second, the mass mediatization of politics favors quick solutions and personalized conflicts over substantial political debate aiming to create a robust agreement about long-term solutions, thus making it difficult to sustain a public dialogue about how to solve pressing problems (Klijn and Korthagen 2017). Third, political power is distributed among a growing number of levels, sectors and actors, thus limiting the sovereign power of elected politicians (Crosby and Bryson 2005). All of this tends to make it an uphill struggle for elected politicians to deal effectively with complex societal problems such as the refugee crisis, economic slumps, climate change, soaring health expenditure and increasing drug abuse. The stakes are high since the failure of elected politicians to convince the electorate that they are doing a good job in tackling pressing societal problems leads to declining trust, anti-politics and the rise of populist movements that may undermine the institutions of liberal democracy (Stoker 2006b).
In sharp contrast to this bleak scenario, there is a growing number of signs that politicians are eager to solicit input to public policy-making and to reform their political working conditions to allow for more systematic and sustained dialogue with citizens and stakeholders that can enrich their understanding of societal problems, stimulate innovation and enhance democratic legitimacy (Lukensmeyer and Brigham 2005). As such, they are not only governing for the people, but with the people (Neblo et al. 2018), and thus endeavoring to ease the tension between political leadership and self-governance that lies at the hard of liberal democracy (Kane and Patapan 2012).
A study based on interviews with leading politicians at the national and state level in five Anglophone countries reveals that political leaders increasingly respond to their need for information, knowledge and ideas by seeking input from external societal actors (Lees-Marshment 2015). What seems to be emerging from the individual and somewhat sporadic contacts initiated by elected politicians is a new model of 'deliberative political leadership' through which politicians solicit inputs from actors inside and outside government, evaluate the quality of those inputs, and integrate new ideas into their own deliberations about how to solve key policy problems (Lees-Marshment 2015).
An Australian study analyzed the contact patterns of public officials seeking to enhance public innovation. The study shows that local politicians, and in particular mayors, were keen to contact non-government actors to obtain strategic and political advice and input to creative problem solving (Considine et al. 2009). However, their networks are not always as large as those of the public administrators.
In much the same vein, a recent study of Dutch mayors (Karsten and Hendriks 2017) shows that a majority saw themselves as 'bridging-and-bonding mayors' rather than 'get-it-done mayors', and that they cared more about process values such as procedural fairness, transparency and responsiveness than about substantive values. Their principal role is to bring different groups of people together and unite them through a democratic process of policy deliberation. The same picture is found in other consensus democracies where local politicians aim to involve local citizens and stakeholders in policy deliberations (Hall et al. 2009;Piattoni 2011;Van Der Heiden and Krummenacher 2011).
However, the picture is not uniform. Nabatchi and Farrar's studies of the interaction between legislators and citizens in the US suggests that thick participation based on two-way communication and facilitated deliberation is in decline. Thin participation based on one-way communication, such as websites, newsletters and direct mailings, or limited two-way communication, such as opinion surveys, polls and social media, is preferred by elected politicians because it is easier to organize, control and scale. So-called 'tele town-halls' are frequently used, but the quality of the interaction is poor (Nabatchi and Farrar 2011).
Interactive political leadership can be enhanced by the development of new institutional designs (Sirianni 2009;Smith 2009;Nabatchi 2012). A broad collection of studies assembled by Gastil and Levine (2005) indicates that institutional reforms aiming to link representative democracy with more participatory and deliberative forms of democracy were not only initiated by civil society organizations and nongovernmental institutions. Hence, governments have created a broad variety of community forums, deliberative polls, planning cells, citizen juries and participatory budgeting processes. The methods for selecting the participants may vary, but the shared ambition is to provide tools for eliciting, appreciating and utilizing different inputs to arrive at collective decisions.
Efforts by governments to enhance interaction with citizens and relevant stakeholders are being supplemented by similar efforts undertaken by national parliaments. As such, a cross-national study conducted by the Inter-Parliamentary Union (IPU) found that parliaments around the world are busy reforming their institutional working conditions in pursuit of effective representation (Beetham 2006). These institutional reforms do not only consist of attempts to become more inclusive, better at communicating with the public, and more effective in exercising oversight of the executive branch of government; parliaments also seem willing to experiment with new ways of engaging with the public, including civil society, and enabling social actors to contribute to the legislative process.
Elected politicians' engagement in the co-creation of public value outcomes paves the way for a new kind of interactive political leadership defined as a dedicated and systematic effort by elected politicians to engage a broad range of relevant and affected actors in problem-focused interaction processes that result in an authoritative and legitimate definition and production of public value outcomes.
While the concept of interactive political leadership breaks the classical notion of sovereign political leadership, the empirical novelty of the phenomenon to which it refers is disputable. Hence, it could be argued that politicians have always involved other actors when developing new policy solutions. In liberal democracies, political leaders are forced to interact regularly with their followers in order to properly represent them. Moreover, it is implicit in the Westminster Model that political leaders interact with and receive strategic policy advice from their administrative aides in terms of executive public managers and special advisors. Finally, both pluralism and corporatism attest to the fact that elected politicians meet with lobbyists and interest group representatives on a regular basis to discuss new legislation. In this limited sense, we agree that interaction with other actors is by no means new. What is new, however, is the deliberate, systematic and orchestrated effort by elected politicians to improve their political leadership by engaging in collaborative processes that facilitate the co-creation of public value outcomes (Sørensen 2019).
Interactive political leadership creates a space for political attempts to manage differences between manifold actors aiming to contribute to the definition and production of public value. This space defies description in terms of the classical chain of government according to which voters delegate power to and subsequently control the representative assembly that, in turn, delegates power to and controls the government that finally delegates responsibility for policy implementation to the administration, which is controlled through auditing and performance management. Relations between relevant and affected policy actors form a complex network rather than a linear chain, and its actors are held together by their mutual dependency vis-à-vis the problem or challenge at hand rather than by hierarchical relations between principals and their agents (Kickert et al. 1997). While the exercise of interactive political leadership engages elected politicians in networked processes of creative problem solving in which political power is decentered and distributed, they retain their relative autonomy vis-à-vis the network, which is a condition for acting as leaders of the political community. Hence, politicians still have the formal powers to set the agenda and initiate policy making, to decide who should be involved, when and how, and to reject, amend or approve the outcomes of collaborative policy interaction. These formal powers over the policy-making process are combined with the more informal power they acquire from working with other actors in collaborative settings, and which may enhance their capacity to solve problems in new and innovative ways. As Follett (1941) remarked many years ago, political leaders exercise not only 'power over', but also 'power with'. This early insight is exactly what the concept of interactive political leadership aims to capture. Machiavelli (1961) provided the first modern account of political leadership when he linked political leadership to grapping power and achieving glory ('entering History') rather than to virtuous behavior and the pursuit of the common good (Ongaro 2017). Stable political leadership, he argued, requires the exercise of power both over friends and enemies, and in the excise of power, political leaders must combine strategic and tactical maneuvering with brute force, including violence. Much later, Nye (2008) insisted that the hard power recommended by Machiavelli is not the best way for political leaders to recruit followers in modern liberal democracies. Political leaders who have a real societal impact often make use of soft power based on charisma and persuasive communication and aim to help the political community to formulate and realize shared objectives by creating a joint sense of situational awareness and mobilizing active followers behind collective strategies. Soft power can sometimes be combined with elements of hard power (carrot and stick). Nye (2008) describes this combination as 'smart power'. While Nye shift the balance from hard to soft power, both Tucker (1995) and Keohane (2014) insist that although leadership involves the exercise of power, it cannot be reduced to the exercise of power since not all powerful persons are leaders in the sense that they enable societal transformation. Leaders may use different forms of power as a resource. However, the essence of leadership does not lie in being powerful, but rather in getting things done by giving voice to visions, creating and implementing strategies, and solving pressing problems (Keohane 2010). This understanding of political leadership underlies Tucker's functional definition of political leadership, which aims to capture the leadership tasks that all political leaders must perform in order to lead. As such, Tucker (1995) famously asserts that political leaders play a key role in defining the problems confronting the political community, designing solutions and securing support for their implementation.

The Concept of Interactive Political Leadership
While Tucker (1995) goes against the traditional association of political leadership with the exercise of political power, Nye (2008) takes issue with the idea that political leaders lead a predefined political community consisting of passive and obedient followers. Thus, he insists that political leaders play an active role in constructing the political community and that they engage in an ongoing dialogue with a broad range of active, independent and critical followers in order to learn from them and secure their continued support. Drawing together these insights, we define political leadership in terms of five key functions performed by political leaders: (1) the discursive formation of the political community that they aim to lead; (2) the identification and definition of problems and challenges that call for public action; (3) the development of innovative, feasible and robust policy solutions; (4) the mobilization of widespread support for the implementation of these solutions; and (5) the highlighting of the public value outcomes that result from public policy-making.
Political leaders produce public value by diagnosing problems and challenges confronting a particular political community, giving direction to and devising solutions, securing political and popular support, and communicating the outcomes. In so doing, they set goals, adjudicate conflicts, take a stand, mandate action and deploy incentives. However, they also listen to proposals and petitions, give voice to unmet needs and demands, seek council and mobilize resources (Keohane 2010, p. 26). Following this line of argument, we claim that co-creation supports and promotes all of the above-mentioned political leadership functions and enables politicians to produce public value. Hence, collaborative interaction with citizens, community leaders, private businesses and NGOs assists elected politicians in creating a multifaceted story about the community that they are leading by weaving together a plurality of voices. It also helps to diagnose problems and challenges by bringing together multiple perspectives, experiences and observations. It enables the development of innovative policy solutions through processes of brainstorming, mutual learning and experimentation that aim to take society in a particular direction. It builds joint ownership over new and bold solutions through more or less inclusive participation. Finally, it enables political leaders to enhance the visibility of the outputs and outcomes of policy processes, thus demonstrating their capacity for creative problem solving. The positive impact of multi-actor collaboration on the exercise of political leadership cannot, however, conceal the risk of interest capture in collaborative arenas that can make political leaders look weak and defeated (McFarland 1987). However, this problem can be mitigated, or perhaps even avoided, through careful institutional design that ensures broad-based participation as well as political autonomy and discretion in the final decision-making (Smith 2009).
Compared to the kind of authorizing political leadership found in Moore's public value theory, our concept of interactive political leadership insists that the role of elected politicians and office holders is not merely to endorse public value proposals crafted by well-intended public managers. Like Rhodes and Wanna (2007), we maintain the primacy of politics and interpret this primacy as meaning that elected politicians should play an active role in diagnosing societal problems and designing public value solutions in and through collaborative interaction with public managers as well as a plethora of relevant actors from the economy and civil society. In this perspective, political leaders are neither authorizers nor sovereign decisionmakers, but responsive and responsible leaders who aim to govern with and through the people.
The democratic implications of interactive political leadership are not only to be found in the potentially positive effects on the level of citizen engagement, democratic legitimacy and trust in government. Interactive political leadership also contributes to a rethinking of key aspects of liberal democracy by transforming the traditional understanding of 'civic engagement', 'democratic mandate' and 'democratic accountability', and replacing current forms of 'resistance democracy' with a new kind of 'interactive democracy'.
First, the traditional understanding that 'civic engagement' is a good thing in itself because it gives affected citizens political influence over policy outcomes, replaced with a new understanding of how such engagement can help to strengthen and democratize political leadership by providing the input that elected politicians need to define and solve pressing societal problems (Lees-Marshment 2015). Interaction with different kinds of citizens and stakeholders can help elected politicians to lead the local community.
Second, the classical conception that voters give elected politicians a 'democratic mandate' to govern on their behalf is replaced with a new recognition of the need for elected politicians to govern in and through an ongoing dialogue with the multiple groups and interests that they claim to represent (Rosanvallon 2011). Hence, the democratic mandate is redefined as an obligation to interact with citizens between elections.
Third, the conventional understanding of 'democratic accountability' as a question of voters holding elected politicians to account on election day is replaced by a new, horizontal form of accountability where all those involved in politically initiated processes of collaborative policy innovation evaluate their separate and joint efforts to enhance public value production and produce accessible narrative accounts of their deeds in order to facilitate public scrutiny (Michels and Meijer 2008;March and Olsen 1995).
Finally, current forms of 'counter-democracy' (Rosanvallon and Goldhammer 2008) through which citizens have increasingly assumed the right to criticize, oppose and reject political decisions and public solutions in and through public hearings and consultations, participation in user boards and electronic citizen panels, the exercise of free choice of service providers etc. are being replaced by a new type of 'interactive democracy' (Rosanvallon 2011) in which citizens and local stakeholders are expected to contribute to and be co-responsible for the design and implementation of new policy solutions in return for their ability to influence policy outcomes.
Both the positive impact that collaborative interaction may have on political leadership and the wider implications of interactive political leadership on democratic governance in societies marked by the surge in networked governance (Torfing et al. 2012) and the proliferation of assertive citizens (Dalton and Welzel 2014) beg the question of how to create institutional frameworks that facilitate the exercise of interactive political leadership. We agree with Ostrom (Ostrom 1996(Ostrom , p. 1080) that the design of institutional arrangements that bring together public and private actors in the pursuit of public value is a daunting task. However, new research suggests that the public sector should play an active role in creating platforms for collaborative governance (Ansell and Gash 2017). Platforms are institutional frameworks that comprise rules, procedures, templates, resources, knowledge and communication systems that facilitate the creation, transformation and multiplication of collaborative arenas (Janssen and Estevez 2013). They are digital or physical meeting places that use goals and visions to attract a particular set of actors, sponsor collaborative interaction to leverage ideas and resources, stimulate mutual learning to enhance innovation, and provide political and administrative support to the implementation of new co-created solutions. While a platform provides a relatively permanent opportunity structure within which temporary arenas of co-creation and interactive political leadership may emerge, operate and adapt, it can neither determine the precise form and content of these arenas nor guarantee their successful contribution to public value production. Nevertheless, deliberate attempts by elected politicians and their administrative aides to build platforms for collaborative interaction between elected politicians and lay actors will generate new opportunities for strengthening the political leadership of the former and the democratic participation of the latter (Sørensen and Torfing 2019a). The institutional design of platforms for interactive policy making will tend to supplement the institutions of representative democracy with democratic institutions based on participation and deliberation, thus stimulating the formation of a new type of hybrid democracy (Sørensen and Torfing 2019b).
The formation of platforms and arenas for interactive political leadership may help to enhance the input and output legitimacy of public governance. However, there are several limiting factors that may discourage elected politicians to go down this road. First, elected politicians may not want to share their political power and responsibility with non-elected actors that may be conceived as 'unaccountable' or depicted as 'self-interested pressure groups'. Second, their focus may be on interparty rivalry and the creation of short-term solutions and quick wins that help to ensure their re-election rather than on improving the quality of policymaking through interaction with relevant and affected actors. Finally, ideological purism and populist belief in strong personal leadership may prevent political leaders from engaging in collaborative processes aiming to spur innovation based on dialogue and mutual learning (Ansell and Torfing 2017).

Public Value Creation through Interactive Political Leadership
The institutional design of platforms and arenas for interactive political leadership may help to connect citizens and politicians, enhance public value production and renew our troubled democracies. However, we need to see whether this argument resonates with practical experiences. To do so, let us look at some illustrative examples of interactive political leadership in order to gauge what it might look like in practice, and to judge whether it offers a viable future for political leadership reforms. Since the size of the political community seems to matter when it comes to connecting politicians with citizens and relevant stakeholders, we will consider empirical cases drawn from different levels of government.
The first case is from local government. In 2015, after several years of successful experimentation with citizen participation and continuous discussions about how to strengthen local political leadership, the Danish Municipality of Gentofte decided to reform the functioning and operation of the City Council. According to the new model (Sørensen and Torfing 2015), the City Council continues to be the ultimate forum for authoritative decision-making about budgets, policies, service standards and new initiatives to improve local welfare. However, to qualify its political decision-making, the City Council formed a number of Task Committees in which elected politicians and local citizens and stakeholders work together to solve pressing problems confronting the municipality. The Task Committees typically comprise five politicians, ten citizens and three to four administrators who serve as resource persons and process facilitators. The Task Committees meet regularly on fixed dates over a period of 3-9 months. In order to find time to work in the new Task Committees, the role and function of the Standing Committees are transformed so that they only meet four times per year and give priority to the strategic monitoring of their policy area over administrative case processing.
The Task Committees are regulated by a mandate drafted by the City Council. The mandate defines the policy problem in broad terms, describes its background and the existing policies, and specifies the type of delivery (a vision, strategy, action plan, policy proposal or initiative) that the Task Committee is expected to submit to the City Council, which will then discuss, amend and endorse it at an ordinary meeting. As a part of its mandate, the City Council defines ten so-called competence profiles of relevant and affected citizens that the politicians want include in the Task Committees. Since the elected councilors already represent the people, they are not looking for a representative panel of citizens, but rather for different voices that can bring new experiences, ideas and resources to the table and help to co-create new and innovative solutions to wicked and unruly problems. The competence profiles are broadly advertised on websites and posters and in local newspapers. Interested citizens can sign up to participate in the Task Committees on the municipal webpage. When the deadline has expired, the politicians match interested citizens with the predetermined competence profile, and are then invited to join the Task Committees. Local stakeholder organizations may also participate in the Task Committees, but only as individuals with a special competence or perspective, and not as representatives of their respective organizations.
One of the first Task Committees was charged with developing a new youth policy. Ten local youth coming from secondary schools and different types of high schools, some with experience from student councils and voluntary work, together with a few drop-outs and former drug abusers, worked with five politicians over 10 months to create a new policy. Already at the second and third meeting, it dawned on the local councilors that the problems that the local youth were experiencing were different from what they had imagined and were also quite similar across different youth groups and segments. Equipped with a new understanding of the problem, they developed a rather innovative youth policy that was short and easy to read. Two high school girls presented the new policy to the City Council, which adopted the new policy after serious political discussion. Subsequently, the two girls toured schools and youth clubs to present the new youth policy, which was already beginning to be implemented even before the administration had drawn up an implementation plan. The evaluation shows that both the politicians and the local citizens participating in the Task Committees learned a lot about the policy problem and contributed to developing new ideas and solutions. Several of the youth participating in the Task Committee on youth policy reported that they had started to think about whether to become involved in local politics and perhaps run for office (Sørensen and Torfing 2016). Some problems remain since the new model is very time-consuming, and it is unclear how citizens can contribute to the implementation of new policies and monitor the progress of new initiatives.
The second case is from national-level policy making in the UK. While it is common for political committees in national parliaments to receive letters and deputations, there are rarely any proactive attempts on the part of such committees to interact and collaborate with external, societal actors in the design of new policies. An interesting exception is the select committees in the British parliament that are regulated by the House of Commons Standing Order no. 152 from 2002. In 2007, the House of Commons Home Affairs Select Committee organized a parliamentary enquiry into the issue of 'honor killings'. Assisted by a small group of administrative staff, the 15 members of the select committee set out to explore a complex set of questions about domestic violence, forced marriage and 'honor'-based violence against women in order to define the problem and develop a proper policy solution (House of Commons 2008). Honor killings appeared on the national political agenda in several European countries at the beginning of the new millennium, and in most countries are perceived as a culturally based type of murder that mainly occurs in Islamic and Middle Eastern communities (Ercan 2014). In the European context, honor killings are allegedly part of the development of 'parallel societies' that allow culturally defined groups of immigrants to live by rules that are in conflict with the values and lifestyle of the majority society.
In its search for a solution to this mediatized problem, the select committee instigated an enquiry that consisted of seminars with relevant stakeholders, visits to women's organizations and refuges, the soliciting of oral evidence from expert witnesses including survivors of domestic violence and their relatives, and an online consultation process that lasted for six weeks and resulted in more than 240 postings. Although the enquiry was based on a relatively traditional 'hearing' and 'consultation' format, there were many opportunities for the politicians to engage in a well-informed and learning-based dialogue with the relevant and affected actors.
The enquiry resulted in a new understanding of the problem, which the select committee ended up describing as 'so-called "honor" killings' (House of Commons 2008, p. 8). Instead of following the mainstream culture-based framing of 'honor' killings as an ethnic aberration, the select committee adopted a gender-based framing that regards 'honor' killings as an extreme form of violence against women that should be treated as such in criminal law (Ercan 2015). As a result of this innovative problem definition, the select committee concluded that there was no need for new legislation, as the laws against violence against women were deemed perfectly capable of handling this particular form of violence.
The third case is drawn from supranational policy making in the European Union (EU). The progressive expansion of the European Union enhances the distance between the political decision makers in the European Commission and the European Parliament, and European populations. To counteract the ensuing democratic deficit, there have been an increasing number of attempts to involve citizens and organized stakeholder groups in EU policy making (Boucher 2009). The European Commission has created an elaborate system for consultation that invites different societal actors to provide feedback on EU policy proposals. Gradually, the range of participants has been expanded from economic experts and business actors, via the involvement of social partners, to the inclusion of civil society organizations and the wider public. The vocabulary has also shifted from 'consultation', via 'partnership', to 'participation' (Quittcat and Finke 2008). The directly elected Members of the European Parliament (MEP) have also stepped up their efforts to interact with European populations in relation to EU policy making. The European Citizens' Consultation (ECC) in 2009 is an ambitious and pioneering example of this.
The question posed to European populations at the initiation of the ECC was: what can the EU do to shape our social and economic future in a globalized world? While the general character of this question seems to have made it difficult for MEPs to translate the recommendations into new policies, the question was well suited to stimulate broad-based participation and open debate, which may inform and shape future policy agendas. As a part of the ECC, 1635 randomly selected citizens from 27 countries participated in national consultations that were prepared and supported by websites and online debate forums and concluded with a debate between citizens and a panel of politicians consisting of national members of the EP and candidates in the upcoming elections. Participation by the wider public was facilitated during the interim online phase leading up to the European Citizens' Summit, at which the 270 national recommendations were discussed and reduced to 15 proposals that were further debated with political and administrative leaders from key EU institutions.
The research-based evaluation (Leyenaar and Niemöller 2010) concludes that participation was based on fair conditions and equal opportunities, that the participants had a solid information base and were well qualified for the joint deliberations, that rules and procedures regarding the participation and deliberation were openly communicated, and that the time, money and energy invested in the participation were commensurate with the results. While media coverage was disappointing, a large majority of the participants claimed that they had learned a lot about the EU and that their support for EU institutions had increased. However, they had rather low expectations about their chances of influencing policy making. The participating EU parliamentarians thought that the ECC provided a useful forum for interaction with citizens, helping to strengthen political representation. However, they thought that the advisory function of the ECC was limited. In sum, the ECC seems to be designed to enhance popular support for the EU rather than to turn citizens into co-creators of public value outcomes.
In all three cases, elected politicians sought to solicit policy input from citizens and relevant stakeholders in order to improve policy making. Citizens and other societal actors were invited to debate policy problems and policy solutions with elected politicians and thus to co-create public value outcomes. However, it should be noted that the process begins and ends at the political level. Hence, the primacy of politics is maintained while the political leadership of the elected political representatives is simultaneously enhanced by the active involvement of, and interaction with, citizens and stakeholders.
The example of interactive political leadership from the local government level appears to involve more policy deliberation than the examples from the national and supranational levels, which tend to have a more traditional 'hearing' and 'consultation' format. The obvious explanation for this is the greater proximity between citizens and elected politicians at the local level. Another difference is that the impact of the policy interaction seems to be higher at the local and national level that at the supranational. This is probably due to the stronger problem focus of the local and national policy interactions, which means that the elected politicians were more eager to learn from the collaborative interaction with societal actors. Despite these differences, the three examples all attest to the fact that interactive political leadership is an emerging and seemingly viable practice at different levels of government.

Conclusions and Ways Ahead
This paper has argued that public value is a game changer that not only serves to highlight the specific contribution of the public sector, but also points to the plethora of public and private actors that can contribute to the production of public value outcomes. While there has been growing scholarly interest in how public managers and employees co-create public value through networks, partnerships and other collaborative forms of governance, this paper has argued that co-creation also constitutes a potent tool for elected politicians seeking to enhance their political leadership by engaging in participatory policy deliberations that enable them to learn more about policy problems, obtain inspiration for the design of new and better solutions, and create joint ownership over new and bold solutions. In order to capture this new type of political leadership, we have coined and defined the concept of 'interactive political leadership'. We have also discussed how interactive political leadership recasts core features of democracy, and we have provided illustrative examples of what this might look like at different levels of government.
Our research on interactive political leadership aims to create a rapprochement between theories of collaborative governance and political leadership theory by arguing that collaborative forms of governance and the co-creation of public value may help to remedy the problems that elected politicians are currently facing. Since this is a new field of study, we need more research to consolidate the concept of interactive political leadership, both theoretically and empirically. In particular, we need new research on how interactive political leadership can be supported and enhanced through the construction of institutional platforms (Ansell and Gash 2017), political forms of metagovernance (Sørensen and Torfing 2017) and the development of a new division of labor between politicians and public managers which enables elected politicians to participate in public policy-making at an early stage when they can influence the definition of policy problems and the design of policy solutions, and involve other relevant and affected actors in the co-creation of public value outcomes.
We also need more research on how interactive political leadership is hampered by the traditional image of politicians as sovereign decision makers steeped in a mixture of paternalism and political competition that precludes the involvement of external actors (Ansell and Torfing 2017). Another inhibitor that deserves scholarly attention is the media logic that tends to favor speedy solutions, personal dramas and political conflicts, thus undermining attempts to engage a plethora of public and private actors in lengthy policy processes aimed at fostering joint solutions based on policy deliberation and political compromise formation (Klijn 2014). Finally, we need to explore whether the political culture in majoritarian winners-take-all democracies, as opposed to consensual democracies, (Lijphart 2012) impedes the development of interactive political leadership.
Our hope is that future studies of the conditions for the emergence of interactive political leadership and its potential to produce public value outcomes will enable us to renew the way that we think about and practice political leadership in the 21st century. Engaged scholarship can help us to find and develop new forms of political leadership that are more attuned to the rise of interactive governance and assertive citizens.