Polyphonic Supervision Meta-governance in Denmark in Denmark ... This article will show how Luhmann’s concept of functional ... governance’ (Jessop, 2000; Jessop, 2003; Jessop,

■ Research Paper

Polyphonic Supervisioni—Meta-governance in Denmark

Anders la Cour* and Holger HøjlundCopenhagen Business School, Frederiksberg, Denmark

This article will show how Luhmann’s concept of functional differentiation makes itpossible to analyse how the government draws various stakeholders into governableterrain through the use of different functional systems. The article takes its outset in arecent reform aimed at a renewal of the authority relations between the municipalitiesand the non-governmental housing associations in Denmark. In doing so, the articleprovides an example of how the government uses meta-governance to define severalinteractive arenas for the various stakeholders. The article concludes thatmeta-governance means neither that the government retains its old power at the centreof society nor that it is hollowed out. Instead, we have to realize that the government’scapacity as an organizer relies on many other forms of media beyond power. Copyright© 2017 John Wiley & Sons, Ltd.

Keywords systems theory; meta-governance; supervision; partnerships

INTRODUCTION

According to a growing literature, new kinds ofwicked problems are currently challenging thedominant ideology of new public management(NPM). While NPM assumes that thegovernment can control the delivery of goodsand services, new problems have arisen, whichchallenge the idea that the government has theauthority to define both the nature of the chal-lenge and the kinds of services to be delivered.

These kinds of problems are called ‘wicked prob-lems’, because the definition of and solutions tothe problems are uncertain and controversial(Head &Alford, 2015; Termeer et al., 2015). Ratherthan acting on the basis of its autonomous power,the government has to invite relevant stake-holders into processes of negotiation about howto define the problems and to develop commoninitiatives to respond to them. Wicked problemsrequire in this sense a high degree of flexibility,reflexivity and stakeholder involvement acrosspublic, private and voluntary sectors (Durant &Legge, 2006; van Buuren et al., 2012).

However, because these types of collaborationsseldom establish themselves on their own—and ifthey do, they often appear messy, unpredictableand full of conflicts—a growing literature is

iParts of this article were previously published in Cour, Anders la &Andersen, Niels Åkerstrøm (2016) Metagovernance as Strategic Supervi-sion. In Public Performance and Management Review, 39(4): 905–925.

*Correspondence to: Anders la Cour, Copenhagen Business School,Porcelænshaven 18a, 2000 Frederiksberg, Denmark.E-mail: [email protected]

Copyright © 2017 John Wiley & Sons, Ltd.

Systems Research and Behavioral ScienceSyst. Res (2017)Published online in Wiley Online Library(wileyonlinelibrary.com) DOI: 10.1002/sres.2449

pointing to the fact that governments are increas-ingly involved in organizing how the differentstakeholders may coordinate their collaborationin order to develop a common understanding ofthe complication because it is all about invitingdiverse public, private and voluntary actors toparticipate in planning, development and promo-tion. In the governance literature, a government,who goes into the establishment of arenas of col-laboration, is said to be engaged in ‘meta-governance’ (Jessop, 2000; Jessop, 2003; Jessop,2004; Provan & Kenis, 2008; Sørensen &Torfing,2009; Torfing et al., 2012; Klijn &Koppenjan, 2015).

From the critical literature, it has been statedthat meta-governance has undermined the state’sability to exercise control over socio-politicalprocesses and developments. The state has beenapproached as being fragmented, eroded or evenhollowed out. A state that governs through meta-governance rather than through traditional formsof regulation has been seen as weakened withregard to its regulatory power (Rhodes, 1994,2000; Millward & Provan, 1993; Stoker, 2006).

While the governance literature has developedimportant insights into how governments havelost faith in central planning and instead engagein establishing themselves as meta-governors,we still need a better understanding of howgovernments make use of different strategies inorder to invite various stakeholders intogovernable terrain. Such an understanding willchallenge the idea that the government has lostpower and instead emphasize how it regainspower through subtle forms of supervision withregard to how the different stakeholders collabo-rate in order to meet the challenges of ‘wickedproblems’.

This article proposes to analyse thegovernment’s role in facilitating new arenas ofcollaborative governance through Luhmann’sconcept of functional differentiation, in order tobetter understand how meta-governance trans-forms—rather than diminishes—the power ofgovernment. We argue that Luhmann’s theoryof the functionally differentiated society offers apossibility for new and subtle analysis of howthe government retains its organizing powerthrough new ways of governing the self-governance of other organizations.

The line of the article’s argument proceedsthrough the following four steps: the first partprovides a short review of the literatureregarding the current research field of meta-governance. This section will argue that thisresearch field lacks a understanding of howgovernments strategically supervise the stake-holders they try to meta-govern. The second steppresents the systems theoretical concept of thefunctionally differentiated society. This sectionwill show how the modern functionallydifferentiated society provides the governmentwith a multifunctional horizon for how meta-governance can supervise the self-governance ofrelevant stakeholders. Third, the article willprovide a case of a reform of the Danish housingassociations as a prime example of meta-governance as polyphonic supervision. Thearticle concludes by discussing how the differentdesigns established by functional differentiationcreate a need for coordination among variousmedia and designs. In other words, Luhmann’stheory of functional differentiation makes itpossible to develop an eye for the multiplicityof perspectives involved in modern forms ofmeta-governance. Or to put it in more preciseterms, paraphrasing Helmut Wilke, well-knownsystems theoretician, we show how meta-governance represents the government’s attemptto develop a ‘functionally differentiated guidanceprogramme’ for the self-steering of the involvedstakeholder (Wilke, 1993, p. 293).

META-GOVERNANCE

The concept of meta-governance has challengedthe traditional picture of the government as thecentral locus of power. It has been debated howthe state’s ability to govern has changed duringthe last four or five decades. Robichau (2011)argues that the discussion has split into astate-centric and society-centric perspective. ‘Thestate-centric perspective maintains that the stateretains its power as the chief actor and centre ofsociety, while the society-centric position con-tends that the state is being hollowed out,decentred, and thus, is progressively relyingupon non-state actors to fulfil its duties’ (p. 117).

RESEARCH PAPER Syst. Res

Copyright © 2017 John Wiley & Sons, Ltd. Syst. Res (2017)DOI: 10.1002/sres.2449

Anders la Cour and Holger Højlund

Because of the state-centric approach, meta-governance redefines the role of the state. In or-der to maintain its regulatory powers, the statechanges dramatically in response to the changingcontext within which practices of governancetake place (Jessop, 2004; Börzel & Risse, 2010;Fawcett & Daugbjerg, 2012). This position is chal-lenged by the society-centric perspective, whichviews meta-governance as an example of howthe transition from government to governancecauses the state to diminish rather than tochange. Due to this perspective, the state ishollowed out (Rhodes, 1994, 2000; Millward &Provan, 1993; Stoker, 2006), is fragmented(Majone, 1994, 1997; Levi-Faur, 2005) or appearsto function as a shadow (Scharpf, 1994;Whitehead, 2003; Sørensen & Torfing, 2009;Poulsen, 2009; Damgaard & Torfing, 2010). Thelatter metaphor indicates some kind of ritualisticsurvival of government or survival fromgovernment in order to step in, if the process ofinteractive governance fails to come up withfeasible solutions to existing challenges (Torfinget al., 2012).Robichau’s (2011) review of the governance

literature argues that much research is neededif we are to gain a more precise picture onmeta-governance. The assumption thatmeta-governance represents the state’s attemptto ‘govern better than less’ (Wallington et al.,2008: 3) should be put to empiricalinvestigation. Or as Kjær (2004) argues, ‘in all,governance does not take place without govern-ment, and governance theory should leave therole of the state open to empirical investigationrather than simply assume that the role isdeclining’ (204). The same argument aboutempirical openness concerning the state’s rolehas been put forward with strong evidencefrom the perspective of Nordic and Europeanexperiences by Vallentin (2013) and Vallentinand Murillo (2010).Taking its outset in systems theory, this article

investigates how government makes use ofdifferent functional media in order to includevarious actors in the act of governing. New andpreviously neglected forms of co-optation areseen. We show how a new form of governingappears that does not represent a neutral

facilitation of interactive governance amongvarious actors but enhances the structural andcommunicative power of the state bydiscursively framing the stakeholders’ identitiesand possible observations of each other, therebyshaping the stakeholders’ capacities as politicalactors.

This article is not the first to introduceLuhmann’s notion of functional differentiationto discussions of public governance. Previousstudies have used the idea of functionaldifferentiation to challenge some of the principaldistinctions that hitherto have guided thegovernance field. Andersen (2005) haschallenged the distinction between public/private that has guided the governance literatureand emphasized that this distinction lacks theability to reflect the complexity of a functionallydifferentiated society. Esmark (2009, 2010)follows the same line of argument when hechallenges the distinction between hierarchy,market and networks used in governanceliterature as an exhaustive mapping of basicmodes of steering available in public governance.Esmark shows how existing modes ofgovernance exceed such tripartite distinctionsand instead make use of steering possibilitiesinherent in the different available functionalsystems, which result from the development ofmodern societies.

Others use Niklas Luhmann’s work to describecollaborative networks as independent systemsin their own right. Kooiman (1993, 2000), forexample, perceives such networks as indepen-dent interaction systems. La Cour and Højlund(2011, 2013) follow the same line of argument,when they construe collaborative networks asthird-order systems and describe how theydevelop into an independent systemic order.Andersen and Loftager (2014) use systems theoryto describe governance networks as meta-organi-zations—organizations of organizations—andshow how these special kinds of social systemsbring together the observations of differentfunctional systems and enable intensive andreciprocal irritability. Similarly, Göran andBrunsson (2008, 2011) have proposed the notionof meta-organization to describe organizationsthat consist of already organized interests. Their

Syst. Res RESEARCH PAPER


Polyphonic Supervision

meta-analytical approach combines aninstitutional with a systemic perspective. Finally,Van Assche and Verschraegen (2008) followanother path, showing in their work howLuhmann’s concept of functional differentiationrepresents both a challenge and possibilitiesfor the government’s attempt to coordinatedifferent organizations’ activities within a func-tionally differentiated society. We will take ouroutset in Assche and Verschraegen’s work inorder to show how meta-governance repre-sents a specific perspective on collaboration.Our argument is that meta-governance repre-sents a specific type of planning that attemptsto influence the way the different stakeholdersestablish themselves as partners for each other.In systems theoretical terms, meta-governancerepresents the government’s attempt to spana multifunctional horizon for its ability todiscursively frame stakeholders’ identities andpossible relationships with one another. Indoing so, new functional media come to thefore that may enhance our knowledge of howgovernments try to govern through meta-governance. In this way, systems theory addsvalue to the existing governance literature, inparticular with respect to developing a betterunderstanding of how meta-governance estab-lishes certain couplings of government andgovernance.

Luhmann’s theory of social systems shares thegovernance literature’s view on the self-regulation of stakeholders. Modern policy-making is not, as NPM saw it, a question ofinstructing each stakeholder on what to do;instead, the government becomes a question ofensuring that the relevant stakeholders assumeresponsibility for the development of local goalsand the initiatives needed to achieve them.Systems theory and the governance literaturethus both contest the basic notion of governanceas primarily about hierarchical commands,determination or prohibitions. Instead, they bothemphasize that ‘to govern’ means to govern theself-governance of others by designating struc-tures of possible stakeholder action that compelstakeholders to make use of their freedom inparticular ways. In other words, thegovernment’s only course of action is to discover

indirect modes of steering that provide otherwiseself-steering systems with the necessary capacityto steer themselves in the directions that thegovernment desires (Wilke, 1993). In thefollowing, by introducing Luhmann’s theory ofthe functionally differentiated society, we willshow how meta-governance represents thegovernment’s attempt to steer the self-steeringof stakeholders in a distinct way, namely, bycreating the foundational premises for how thedifferent stakeholders might create themselvesas partners for one another.

According to systems theory, stakeholdersbelong to the category of organizations.Organizations are social systems, whichestablish the distinction between membershipand non-membership as their primaryprinciple for creating a boundary betweenthemselves and their environment. In thissense, individual human beings are onlyobserved as relevant stakeholders if they areperceived as members of specific organizationsand thereby represent more than just them-selves in the collaboration.

Meta-governance refrains from attempts tosteer by means of defining ‘facts’ beforehand,such as defining how the problems at handshould be conceived and how they should beaddressed in meta-governance; the definitionsof such questions are left open to the stake-holder collaboration, and the governmentseeks to supervise how stakeholders might es-tablish themselves as partners for one another.In other words, the government’s supervisionis not about presenting the ‘facts’ of the‘wicked problems’ that stakeholders face.Instead, meta-governance becomes an exampleof how the government tries to influence howthe different stakeholders might conceive oneanother as mutual partners, because theirability to orient their actions and findsolutions to the ‘wicked problems’ dependson how they conceive one another as partners.In system theoretical terms, meta-governancecan therefore be seen as an attempt to influ-ence and coordinate the interaction betweenself-steering systems by offering ways forthem to conceive one another as mutualpartners.




FUNCTIONAL DIFFERENTIATION

Our point of departure is Niklas Luhmann’stheory of modern society as functionally differen-tiated. Our line of argument is that throughouthistory, differing functional systems have playeda dominant role (Luhmann, 2013, p. 87). A defin-ing characteristic of these systems is that theyevolve around their own way of perceiving theworld and the fulfilment of a specific function(for a pedagogical introduction, see Tsivacou,2005, in this journal). In their work, Roth andSchutz (2015) have provided a canon ofLuhmann’s 10 functional systems. In the follow-ing, we will focus only on four of these and adda fifth (the system of love), because they seemto dominate the specific case of the Danish re-form of the social housing associations, whichwe will introduce later on. The first one is the eco-nomic system, which forms the medium ofmoney and observes everything through the lensof having/not having in order to deal with scar-city. The political system forms the medium ofpower through the distinction govern/governedin order to produce collectively binding deci-sions. The legal functional system forms the me-dium of law through the distinction betweenthe legal and non-legal in order to be able tojudge. The pedagogical system uses the idea ofthe child as a medium formed through the codeof better/worse in order to socialize. And evenlove has a functional system, which forms themedium passion through the code of loved/notloved, which deals with intimacy (Luhmann,2012). Andersen and Born (2008) in particularhave outlined the analytical advantage ofobserving the communication of love as itappears in different organizational settings.Luhmann has dedicated a volume to each of thefunctional systems he has detected in hishistorical descriptions of modern society asfunctionally differentiated. In addition to theones mentioned previously, these are science,religion, sport, mass media, health and art.Common to all these systems is that they have

developed their own semantic and are closedaround their own specific way of observing theworld. In economic communication, everythingcan be seen through the lens of money, but this

also marks the limit of its observation; it cannotalso convict people or make collectively bindingdecisions. The binary codes of functional systemsare extremely effective for guiding observations.From the point of view of a function system,anything can be reduced to an either or question.Everything can be observed through its specificcoding. However, at the same time, this repre-sents the system’s limitation, because the selec-tive codes remain blind towards issues thatcannot be translated into one of the two sides ofthe binary code. In other words, functionalsystems are operatively closed systems; theyevolve around their own specific code, and theonly thing they are capable of is producing con-tinued communication within the same distinc-tions. Several critics have used this theoreticalframework to show how organizations can makeuse of more than one media in their communica-tion by oscillating between the different possibleobservations of the world made available to themby different functional systems (Andersen &Born, 2000; Andersen, 2003; Højlund & la Cour,2015; Højlund, 2009). Moreover, we havewitnessed the appearance of a new critical andsystemic management journal, Tamara, whichpublishes critical organizations studies guidedby this systemic approach [for an example, seeRoth (2014)]. Our approach is in many wayssimilar to this analytical strategy, because weseek to analyse the state as a multifunctionalorganization.

The article’s main argument is that the variousfunctional systems provide very different condi-tions for meta-governance’s attempts to structurethe partnerships of its stakeholders. The media ofpower, money, law, pedagogy and passion pro-duce rather different conditions for discursivelyframing the stakeholders’ identities and possiblerelationships with one another. Depending onthe specific media offered to them by the supervi-sion as a lens through which to observe eachother, the partners will come to observe eachother rather differently.

Within the medium of power, the supervisingentity tries to configure the partnership as away to empower the participating organizationsand institutions. Who has the power to do what,and how can the partnership be used as a way for




the different institutions to discover themselvesas someone with authority to decide and to takeresponsibility for creating themselves as someonewith the power to do so? This is especially impor-tant if a partnership includes institutions that arenot used to participating in traditional decisionprocesses, that is, the game of power. The statecan try to configure the partnership as a collabo-rative arrangement that challenges traditionalpower relationships between govern andgoverned. In such a case, the partners need tobe mindful of when the partnership is being usedto challenge traditional forms of power relation-ships, which are observed as an obstacle to thedevelopment of the partnership, but also of howthe partnership constructs new forms of powerrelationships, where partners are assigned newforms of authority with respect to decision-making.

Within the medium of law, the state tries tosupervise the partnership as a collaborativenetwork between agents with a legal obligationtowards each other. This can be in regard to areform that makes it a legal obligation for part-ners to plan meetings with each other, to informthe public about their decisions and to havecertain roles to fulfil in order to respect the part-nership as a legal constitution. Partnership isseen as right, and the other partners are observedas legal subjects, where the actors are evaluatedin light of the rules that are developed aroundthe partnership: are they fulfilling their legalobligations? This kind of supervision is aboutmaking the duties of the different partners clearto everybody.

Within the medium of money, the state tries tosupervise the partnership as between entitiesstriving to profit individually from the partner-ship. Thus, the partners will consider initiativesthat appear to be profitable in the situation. Part-nerships formed in this code therefore encouragethe partners to make calculation on the basis of acost/benefit logic with respect to what they maygain from this investment vis-à-vis a differentone. In this sense, money plays a double function:it is used as an incentive for the stakeholders toconsider what the partnership has to offer,namely, economic benefits, but it also functionsas a way for the partners to put pressure on each

other: do we benefit enough from the efforts weput into this partnership? These rational calcula-tions work to guide the self-governance of thestakeholders to ensure that the partnershipfunctions as an effective tool and achieves thegreatest possible utility for the involved actors.The guiding principle is ‘we must have valuefor the money’.

To supervise through the code of pedagogymeans that responsibility for creating oneself asa partner for the other is assigned to theinstitution itself. What is special about this partic-ular mode is that the institution is expected toobserve itself through the code of pedagogy,reproduce the institution as incomplete and turnitself into a project focused on creating itself aspartner for the other. The goal of pedagogicalsupervision is for the partnership to become aproject of self-development as a way to increasethe institution’s relevance as partner. When thestate’s supervision is formed in the pedagogicalcode, it forms the medium of ‘the child’(Luhmann, 1993). In this perspective, institutionsare perceived as mouldable, and the partnershipis symbolized as a learning endeavour, whereengaged institutions are expected to learn fromone another. In the observation of the pedagogi-cal code, institutions do not represent fixedentities but are expected to evolve through thepartnership.

To supervise the partnership in the medium oflove is not to establish the power relation amongthe partners or to explicate their duties towardsone another. It is also not about letting thepartners discover their own responsibility forself-development, as is the case within the peda-gogical code. Instead, it is about defining thepartnership as a matter of the partners takinginitiatives that are not motivated by ‘what is init for me’ but by initiatives that show respectand responsibility for the partnership as a whole.Here, the partners do not take on responsibilitybecause they have the authority to do so, becauseit is their legal duty or because it contributes totheir own self-development. Instead, the statetries to facilitate the partnership as a collabora-tion, where the different partners are responsibleabove and beyond their legal duties. They areexpected to develop engagement that




transgresses these expectations and to show inde-pendent initiative, which indicates that they‘love’ the partnership. Here, the partnership isconstituted as a way for the partners to guidethemselves in accordance with their ideas aboutthe needs of the other partners. It becomes aquestion of unselfishly ‘giving oneself’ to thepartnership. What is important is that the part-ners orient themselves towards the world of theother in order to find meaning that transgressestheir own way of seeing the world and replacesit with the world of another system (Luhmann,2012, p. 175). But this also works the other wayaround, which means that the partners are alsoexpected to enable ‘the other’ to give back (seeAndersen & Born, 2008, for further discussionsabout love as a communication medium).The examples of supervision through different

media are summed up here:The systems theoretical conceptualization of

the functionally differentiated society allows usto develop a more precise observation of howthe state seeks to meta-govern. By means ofwhich rationality does the state seek to governthe self-governance of the participating actors?When more than one code is possible, the super-vision may be suspected of indeterminacy orhidden agendas. This is a risk that accompaniessupervision as a medium for observation andself-observation. Supervision is by definition notvery conclusive. Multiple codes support thepicture of inconclusiveness. However, at thesame time, polyphonic coding produces potentialflexibility. The supervision may make itselfrelevant to different kinds of partnerships. Asobservational modus, supervision will producedifferent observations. Some observations willrun counter to each other, while others will bemutually supportive.Accordingly, we argue that meta-governance

represents a ‘functionally differentiated guidanceprogramme’ (Wilke, 1993, p. 293), which invitesthe relevant stakeholders to create themselves aspartners for one another in different and flexibleways. Luhmann’s notion of the functionally dif-ferentiated society makes it possible to analysemore specifically the ways in which meta-governance seeks to influence and facilitate theself-governance of others. Before providing an

example of how meta-governance makes use ofseveral media simultaneously, we will present asection on our methodology.

DATA AND METHODS

The sources of empirical evidence for this articleemploy document analysis. This method relieson policy documents, where the data are basedon a close examination of a variety of policydocuments (Freeman & Maybin, 2011, p. 155).Policy documents tend to refer to other policydocuments. We assume a coherence or a stabilereference structure, when such network of tex-tual relations appears in which all documents re-fer to each other; here, we address thedocuments as a semantic reservoir (Luhmann,1998, p. 19) from where we can abstract a kindof generalized vocabulary of the policy system.Taken together, the policy documents representa stabilized, self-reflexive and self-descriptivevocabulary of the given political area (Luhmann,2000, p. 219).

In the present semantic archive of the Danishreform of the housing associations, we include agovernment white paper on the reform(Ministry of Welfare, 2008). The white paperpresents the government’s policy prior to theintroduction of subsequent legislation. We alsoinclude the official guidelines on how to realizethe reform in practice from the ministry,municipalities and national association ofhousing associations (Ministry of Interior andSocial Affairs, 2009; Ministry of Social Affairs,2010; Copenhagen Housing Association, 2011;National Association of Housing Associationsand National Association of Municipalities,2009a, 2009b, 2010). Three documents are fromgovernment departments (Ministry of Welfare,2008; Ministry of Interior and Social Affairs,2009; Social Affairs, 2009). Two documents arefrom the National Association of HousingAssociations and National Association ofMunicipalities (2009a, 2009b, 2010). The lastdocument is from the largest housingassociation in Denmark, Copenhagen HousingAssociation (2011).




We conduct a close examination of a total of sixdocuments. These documents create their ownnetwork of textual relations. In our reading ofthe six documents, we follow a reading strategybuilt upon a set of methodological assumptionsderived from systems theory, which lay out twoequally important research questions: ‘what isthe case’ and ‘what lies behind’ (Luhmann,1993). The first question is ‘what is the case’.How does meta-governance represent a multi-horizontal guidance programme for the self-perception of the involved stakeholders? Andwhat lies behind? In other words, how is this aresult of a society that has developed into a func-tionally differentiated society, where the politicalsystem no longer exists at the centre of societybut has to rely on the capacity of other systemsto steer themselves in accordance with the goalsof the political system? The following case is builtup around these two research questions.

POLYPHONIC SUPERVISION1

The Danish government launched in 2010 asteering reform in order to improve the collabora-tion between municipalities and housing associa-tions in Denmark. The aim of the reform was tocreate a new governance framework in order tobreak down the traditional forms of governmentmediated by legal regulation and contracts.Instead of government from hierarchy, a frame-work of equality in positions was proposed tofoster dialogue-based interaction between thetwo stakeholders. The reform text was clear: themunicipalities would give up their role asrule-observing bureaucrats and instead createthemselves as equal partners who could developgoals, challenges and initiatives together with thehousing associations on an equal footing. Thehousing associations would no longer act as mereadministrators of the law but in addition developinto active policymakers. The ambition of thereform was that the institutionalization of inter-organizational cooperation would create effective

initiatives than the traditional hierarchical formof government, hitherto had been able to pro-duce. After all, the new forms of governancehad the ability to bring together resources acrossthe municipalities and the housing associations(National Association of Housing Associations& National Association of Municipalities, 2009a,2009b).

Having a closer look into the reform, thegovernance ideas seem unquestionable. Thereform turns the traditional hierarchical relation-ship between the municipalities and the localhousing associations into a collaborative partner-ship. The future challenges of the associations areaddressed as challenges related to activities thatare related to community building. Instead of for-mulating the municipalities with a controllingfunction in relation to the housing associations,the two actors are expected to establishthemselves as equal partners engaged in mutualdialogues about how, together, they may developinitiatives for handling future challenges (Minis-try of Social Affairs). The reform represents anexplicit and formal strategy for stipulating andinstitutionalizing formal interaction betweenpublic and non-public actors. In this sense, thereform represents yet another example of meta-governance in practice. But the reform also exem-plifies how the state makes use of several func-tional media in order to organize the conditionsof the partnership.

What were the dominating semantics of thehousing reform? In the core of the reform wassome law changes. Law was an important com-municative media, although the reform under-stood itself as representing a new directionaway from the traditional practices, where themunicipalities’main task was to provide superin-tendents and ensure that the housing associa-tions managed themselves according to the law(National Association of Housing Associations& National Association of Municipalities, 2009a,2009b). The housing associations still had tomanage themselves in accordance with the law,but this was not included in the reform agenda.Instead, the law was being used to obligate thepartners to establish formal meetings with eachother (at least one bilateral meeting each yearbetween the municipalities and each of the

1 Parts of case description also appear in Cour, Anders la & Andersen,Niels Åkerstrøm (2016). Metagovernance as Strategic Supervision. InPublic Performance and Management Review, 39(4): 905–925.




housing association) and to commit the munici-palities to sending a questionnaire to the housingassociations, which they had to fill out and whichcould be used as a starting point for the discus-sions at the meetings. The municipalities werealso obligated according to the law to make asummary of the meeting available to the public(Ministry of Social Affairs, 2010; NationalAssociation of Housing Associations & NationalAssociation of Municipalities, 2011b). Themedium of law cannot force the different part-ners to make collective decisions with each other,but it can focus on the mutual obligation to meetthe expectations of the law to have regular andinformed meetings with each other.From the very beginning, the reform legiti-

mized itself as a way to make the administrationof the housing associations more cost effective byallowing the municipalities and the housingassociations to reduce spending on administra-tive matters. But the reform also emphasized thatthe partnership would make it possible todevelop solutions that would be more effectivethan existing ones, also with regard to the useof economic resources (Ministry of Interior &Social Affairs, 2009). Thus, the reform seeks toinstal a certain kind of reflexivity among thestakeholders, where future options are evaluatedfrom the perspective of a potential return, whichis calculated in relation to the given investment.Accordingly, the initiatives taken can beperceived as an investment, which is intended,over time, to create increased value for theinvolved actors.One of the many intended outcomes of the

reform was for the housing associations todiscover themselves as independent organiza-tions with the authority to partake in a politicaldecision-making process. They were expected todevelop their own political stance concerning allpolicy areas in relation to the overall develop-ment of the municipality that they belonged to(National Association of Housing Associations& National Association of Municipalities,2011a). However, the reform also challenged themunicipalities, because the civil servants, whorepresented the municipalities at the many meet-ings, had to take on a more proactive role thanmerely one of observing bureaucrats. The reform

required the civil servants to become innovatorsof new political initiatives and to provide sus-tainable policy solutions (Poulsen, 2009, p. 118).

One of the dominant semantics of the reformwas the focus on development. From the verybeginning, the partnership was constituted as ajoint project for development, for example, howto extend collaboration. The reform language ismarked by a sense of potentiality with regard tothe development of the partnership and thepossibilities for the partners’ self-development(National Association of Housing Associations& National Association of Municipalities, 2009a,2009b, 2011a, 2011b; Ministry of social affairs,2010). The state wants to supervise the partner-ship as a tool for learning. In the code of peda-gogy, the institution can evaluate the other withwhom it finds itself in a partnership: how is theother institution’s capacity for self-developmentprogressing? Or it may evaluate the partnership:how is the partnership contributing to the self-development of the institutions? Or theinstitution: how is the institution using thepartnership as a way to self-develop? Thepedagogical code constitutes all three entities assomething unfinished, as something in need ofconstant development.From the perspective ofthe code of love, the reform invites the stake-holders to constitute themselves as contributorsto the common objectives and furthermore toallow others to constitute themselves in the sameway:

… to create effective development and goodresults requires that the partners similarlyplace demands on each other. Each partnerneeds to create initiatives and build the frame-work for their implementation and subse-quent evaluation (National Association ofHousing Associations & National Associationof Municipalities, 2009a, 2009b, p. 4).

Meta-governance seeks to bring a ‘shared we’to the fore by establishing a goal for the partnersto gather around. This goal is communicated as aflexible signifier for the partners to put a lot ofcommunality and shared feelings into. A centralfocus of the reform is the actors’ self-creation asrelevant partners for each other. Meta-governance seeks specifically to establish the




actors’ freedom to commit to the assumedresponsibility for the partnership. It representsan attempt to formulate mutual obligationsconcerning the self-creation of the individualpartners as responsible for, and relevant to, thepartnership. Through the co-optation of the com-mon goal and the values of communality, thepartners are expected to make the partnershipcentral to their own decisions (Figure 1).

Thus, in the image of meta-governance, thepartnership becomes rather complex; it calls onthe partners to establish themselves as legalentities with legal obligations towards each other.Not submitting the questionnaires, or fillingthem out, not extending invitations and notshowing up to the meetings are seen as againstthe law. It is also expected of the partners thatthey take on political responsibilities, byestablishing themselves as political authoritiesvis-à-vis each other with respect to developingnew policy agendas within the partnership. Butthe partnership needs to also be motivated bythe benefits that it offers for the time and effortput towards it. Moreover, it asks the partners to

assume an almost experimental attitude towardseach other, constantly exploring ways in which tolearn from the experiences gained throughout theprocess, and how this may contribute to thedevelopment of the partnership as such and totheir own self-development. And finally, thepartners are expected, through their actions, totune themselves to the world of the other in orderto develop a mutual eye for what most benefitstheir relationship as a whole, despite more indi-vidually motivated behaviours (Figure 2).

As shown, meta-governance seeks to organizethe conditions of the partnership within morethan one semantics and colour by more thanone code. This kind of polyphony can serve asan opening towards more flexible strategies,although it might at times challenge the identityof the organizations entering a partnership.Organizations marked by a strong homogeneousidentity will feel challenged by polyphonicsemantics. Figure 3 illustrates how different me-dia construct such challenges in different ways.

For a summary, each code tries to establish thepartnership from the perspective of its own

Figure 1 Supervision formed by media




Figure 2 Supervision forms

Figure 3 Polyphonic challenges




exclusive way of observing the world. Whatmakes a difference, what counts as a coordina-tion problem, a challenge or a solution is definedby the code through which the partnership isviewed. Our argument is that if you observehow meta-governance seeks to organize theconditions for self-governance through theestablishment of different kinds of partnershipbetween the relevant actors, many different codesappear. And none of them can be said to havehegemonic status. In order for the partnershipnot to collapse, meta-governance requires forthe partners to have the ability to constantlyswitch codes in accordance with what is goingon in the collaboration. In this way, functionaldifferentiation produces potential flexibility. Butbecause the different media that the stateprovides for the stakeholders’ self-governanceproduce different observations, there is also therisk that some of them will run counter to oneanother.

Establishing the partnership within themedium of economy, where the collaboration isall about creating benefits for the individualstakeholder, can easily create conflict with themedium of love, which produces expectationsabout devoting oneself to what is best for thepartnership. The code of love is preoccupied withhow the partnership can initiate the binding ofthe stakeholders in a relationship that does notmark difference but unity. In other words, it triesto establish a ‘we’. This goal also runs counter tothe pedagogical code, which requires the oppo-site of the love code’s demand for absorptionand asks stakeholders to engage in reflectionand a strict self-awareness about how the part-nership can contribute to their self-development.In the present context, we are unable to furtherdevelop these ideas, but the basic assumption isthat if the organizations can constitutethemselves as partners under the umbrella ofthe same coding, they are more likely to be ableto coordinate their interest, develop commoninitiatives and reach community goals, than ifthe opposite is the case. Polyphonic meta-governance, by contrast, requires the continuouschoice among, reproduction and replacement ofvarious codes, pushing and pulling the partner-ship in different directions. The shifting

perspectives, which the codes make possible,can serve as important and innovative vehiclesfor the development of the partnership. But theshifting codes may also produce disorder andunpredictability, because the stakeholders cannever be sure when the partnership invokes adifferent code.

CONCLUSION

This article has drawn attention to the ways inwhich Luhmann’s theory of functional differenti-ation can expand our knowledge about how gov-ernment works through meta-governance as away of organizing the conditions for self-organization among various stakeholders. Seenthrough the lens of systems theory, meta-governance represents a specific form of steering,where governments aim to coordinate differentways for stakeholders to engage in collaborationwith one another, without defining what a givenproblem or its solution looks like. Instead, meta-governance represents an attempt to developdifferent discursive frameworks that function asstrategic supervision of how the relevant stake-holders should construct themselves as relevantactors to participate in the governance arena.Using Niklas Luhmann’s theory of functionaldifferentiation, the article shows how this hasbeen done through different functional mediathat pave the way for a variety of ways in whichto couple relevant stakeholders. System theorymakes it thereby possible to analyse howmeta-governance is changing the patterns ofpossible interaction between stakeholders, whichinvolve that collaboration takes place in afragmented system of governance in whichvarious function systems offer their own logicsand their own expectations towards the pro-cesses of collaboration.

In the presented case, the government seeks tometa-govern through the use of five differentcommunication media: politics, law, economy,pedagogy and love. Some of these representnew and hitherto neglected forms of media forthe self-governance of the involved stakeholders.In this way, meta-governance represents a poly-phonic steering ambition, in the sense that it




provides openness with regard to the plurality ofcodes and semantics that may guide the develop-ment of the partnership. Therefore, it becomesimportant for the state that the partners use thepartnership as a kind of scanning mechanismfor possible codes and semantics, as a way torecognize the dispersal of capacities andresources relevant for developing the partner-ship. In this sense, polyphonic supervisionproduces anarchic conditions for the self-governing of relevant stakeholders. In order tobe strategic about their choices, however, it isimportant for stakeholders to see the scanningprocesses as strategic, in the sense that the choiceof semantic coding is related to matters of iden-tity and future reflexivity, because the choice ofcode will have a constitutive effect on the part-nership and the partners.In this article, we have only outlined a very

broad idea about how the theory of thefunctionally differentiated society can improvethe investigation of how meta-governance linksgovernment and governance. As a result, we willargue that the question is not whether or notmeta-governance undermines the power ofgovernment but rather how meta-governancemakes new steering media relevant for thegovernment’s ability to influence the self-governance of its stakeholders.This macro-sociological approach should be

supplemented by concrete empirical studies ofhow governments make use of different mediawhen organizing the conditions for the self-governance of stakeholders. In relation to suchempirical studies, the conceptualization of meta-governance as polyphonic supervision producestwo important research questions for future stud-ies. First, it encourages researchers not to losesight of the integral role that governments playin the different arrangements for the self-governance of others. Second, it opens uptowards further empirical investigations andanalysis of the kinds of media involved in gov-ernmental efforts to influence the self-governanceof other organizations.With inspiration from systems theory, meta-

governance has been constructed as an exampleof polyphonic supervision of how stakeholdersmight create themselves as partners for one

another. The case has been a governmentalreform that tries to bring stakeholders togetheracross public and civil sectors in different collab-orative networks. But polyphonic supervision isnot reserved for governments or other politicalorganizations in their attempt to govern the self-governance of others. In fact, it becomes an openempirical question of which kinds of organiza-tions make use of polyphonic supervision. In thissense, systems theory seems once again totransgress the limited scope of meta-governance.

REFERENCES

Andersen NÅ. 2003. Polyphonic organizations. InAutopoietic Organization Theory, Hernes T, Bakken T(eds.). CBS: Copenhagen; 151–182.

Andersen NÅ, Born AW. 2000. Complexity andchange: two “semantic tricks” in the triumphantoscillating organization. Systemic Practice and ActionResearch 13: 297–328.

Andersen NÅ, Born AW. 2008. The employee in thesign of love. Culture and Organization 14(4): 325–343.

Andersen SC. 2005. How to improve the outcome ofstate welfare services. Governance in a systems-theoretical perspective. Public Administration 83(4):891–907.

Andersen SC, Loftager J. 2014. Deliberative democraticgovernance. Administrative Theory & Praxis 36(4):510–529.

Börzel TA, Risse T. 2010. Governance without a state:can it work? Regulation and Governance 4(2): 113–134.

Copenhagen Housing Association. 2011. Hanbog omalmene boliger - spilleregler for bestyrelser, beboere ogansatte i den almene boligsektor. Litotryk Københavna/s: Copenhagen.

la Cour A, Højlund H. 2011. The emergence of a third-order system in the Danish welfare sector. In TheThird Sector, Hull R, Gibbon J, Branzei O, Haugh H(eds.) . Emerald Group Publishing Limited(Dialogues in Critical Management Studies: Bingley;87–111 Nr. 1.Bingley (red.)

la Cour A, Højlund H. 2013. Organizations,institutions and semantics: systems theory meetsinstitutionalism, pp. 185–202 i Anders La Cour &Andreas Philippopoulos-Mihalopoulos, LuhmannObserved: Radical Theoretical Encounters. Basingstoke:Palgrave Macmillan.

la Cour A, Andersen N. 2016. Metagovernance as Stra-tegic Supervision. Public Performance & ManagementReview 39: 905–925.

Damgaard B, Torfing J. 2010. Network governance ofactive employment policy: the Danish experience.Journal of European Social Policy 20(3): 249–262.




Durant RF, Legge JS. 2006. “Wicked problems,” publicpolicy, and administrative theory lessons from theGM food regulatory arena. Administration & Society38(3): 309–334.

Esmark A. 2009. The functional differentiation ofgovernance: public governance beyond hierarchy, mar-ket and networks. Public Administration 87(2): 351–370.

Esmark A. 2010. Systems theory. The SAGE Handbook ofGovernance. SAGE: London.

Fawcett P, Daugbjerg C. 2012. Explaining governanceoutcomes: epistemology, network governance andpolicy network analysis. Political Studies Review 10:195–207.

Freeman R, Maybin JO. 2011. Documents, practicesand policy. Evidence & Policy: A Journal of Research,Debate and Practice 7(2): 155–170.

Göran A, Brunsson N. 2008. Meta-organisations.Edward Elgar Publishing: Northampton.

Göran A, Brunsson N. 2011. Organization outsideorganizations: the significance of partial organiza-tion. Organization 18: 83–104.

Head BW, Alford J. 2015. Wicked problems implica-tions for public policy and management. Administra-tion & Society 47(6): 711–739.

Højlund H, la Cour A. 2015. Polycontextuality and thebody. In Systems Theory and the Sociology of Health andIllness. Observing Health Care, Knudsen WF/M (ed.).Routledge: London; 43–60.

Højlund H. 2009. Hybrid inclusion—the new consum-erism of Danish welfare services. Journal of EuropeanSocial Policy 19(5): 421–431.

Jessop B. 2000. Governance failure. In The New Politicsof Local British Governance, Stoker G (ed.). Macmillan:Basingstoke.

Jessop B. 2003. Governance and meta-governance: onreflexivity, requisite variety and requisite irony. InGovernance as Social and Political Communication, BangHP (ed.). Manchester University Press: Manches.ter.

Jessop B. 2004. Multi-level governance and multi-levelmetagovernance. Multi-Level Governance 1(9): 49–75.

Klijn HE, Koppenjan K. 2015. Governance Networks inthe Public Sector. Routledge: London.

Kjær AM. 2004. Governance. Polity Press: Cambridge.Kooiman J. 1993. Modern Governance. Sage: London.Kooiman J. 2000. Societal governance: levels, modes

and orders of social-political interaction. In DebatingGovernance: Authority, Steering, and Democracy, PierreJ (ed.). Oxford University: Oxford.

Levi-Faur D. 2005. The rise of regulatory capitalism:the global diffusion of a new order. The Annals ofthe American Academy of Political and Social Science598(1): 200–217.

Luhmann N. 1993. “Was ist der Fall?” und “Was stecktdahinter?” Die zwei Soziologien und dieGesellschaftstheorie. Zeitschrift für Soziologie 22(4):245–260.

Luhmann N. 1998. Geesellschaftsstruktur und Semantik 1.Suhrkamp: Frankfurt am Main.

Luhmann N. 2000. Die Politik der Gesellschaft.Suhrkamp: Frankfurt am Main.

Luhmann N. 2012. Love as Passion. Polity Press:Cambridge.

Luhmann N. 2013. Theory of Society, vol. 2. StanfordUniversity Press: Stanford.

Majone G. 1994. The rise of the regulatory state inEurope. West European Politics 17(3): 77–101.

Majone G. 1997. From the positive to the regulatorystate: causes and consequences of changes in themode of governance. Journal of Public Policy 17(2):139–167.

Millward HB, Provan K. 1993. The hollow state:private provision of public services. In Public Policyfor Democracy, Ingham H, Smith SR (eds.). BrookingsInstitution: Washington; 222–237.

Ministry of Interior & Social Affairs. 2009. Den almeneboligsektors dokumentation – Rapport fra arbejdsgruppenvedrørende almene boligorganisationers dokumentation.Silkeborg bogtryk: Silkeborg.

Ministry of Social Affairs. 2010. Vejledning omstyringsdialog. Silkeborg bogtryk: Silkeborg.

Ministry of welfare. 2008. Den almene boligsektorsstyring. Silkeborg Tryk: Silkeborg.

National Association of Housing Associations &National Association of Municipalities. 2009a.Styringsreformen – det fremtidige samarbejdemellem boligorganisationer og kommuner.

National Association of Housing Associations &National Association of Municipalities. 2009b.Styringsreformen – det fremtidige samarbejde mellemboligorganisationer og kommuner. KommunernesLandsforening & Boligselskabernes Landsforening:København.

National Association of Housing Associations &National Association of Municipalities. 2010.Styringsdialog – Et værktøj til samarbejde mellemkommuner og boligorganisationer. KommunernesLandsforening & Boligselskabernes Landsforening:København.

National Association of Housing Associations &National Association of Municipalities. 2011a.Styringsdialog – Et værktøj til samarbejde mellemkommuner og boligorganisationer. Litotryk Københavna/s: København.

National Association of Housing Associations &National Association of Municipalities. 2011b.Håndbog om almene boliger – Spilleregler for bestyrelser,beboere og ansatte i den almene boligsektor. LitotrykKøbenhavn A/S: København.

Provan K, Kenis P. 2008. Journal of Public AdministrationResearch and Theory 18(2): 229–252.

Poulsen B. 2009. Competing traditions of governanceand dilemmas of administrative accountability: thecase ofDenmark.Public Administration 87(1): 117–131.

Rhodes RAW. 1994. The hollowing out of the state: thechanging nature of the public service in Britain. Thepolitical Quarterly 65(2): 138–151.




Rhodes RAW. 2000. Governance and public adminis-tration. In Debating Governance, Pierre J (ed.). OxfordUniversity Press: Oxford; 54–90.

Roth S. 2014. The multifunctional organization: twocases for a critical update for research programs inmanagement and organization. Tamara: Journal forCritical Organization Inquiry 12(3): 37–54.

Roth S, Schutz A. 2015. Ten systems: toward a canon offunction systems. Cybernetics & Human Knowing 22(4): 11–31.

Robichau RW. 2011. The mosaic of governance: creat-ing a picture with definitions, theories and debates.Policy Studies Journal 39(1): 113–131.

Scharpf FW. 1994. Games real actors could play:positive and negative coordination in embeddednegotiations. Journal of Theoretical Politics 6(1):27–53.

Stoker G. 2006. Public value management: a new nar-rative for networked governance? American Reviewof Public Administration 36: 41–57.

SørensenE,Torfing J. 2009.Makinggovernancenetworkeffective and democratic through metagovernance.Public Administration 87(2): 234–258.

Termeer CJ, Dewulf A, Breeman G, Stiller SJ. 2015. Gov-ernance capabilities for dealing wisely with wickedproblems. Administration & Society 47(6): 680–710.

Torfing J, Peters BG, Pierre J, Sørensen E. 2012. Interac-tive Governance. Oxford University Press: Oxford.

Tsivacou I. 2005. The ideal of autonomy from the view-point of functional differentiation/integration of

society. System Research and Behavioral Science 22:509–524.

Vallentin S. 2013. Governmentalities of CSR: DanishGovernment Policy as a Reflection of Political Difference.Springer: Journal of Business Ethics.

Vallentin S, Murillo D. 2010. Government,governance, and collaborative social responsibility.In The Collaborative Enterprise. Creating Values for aSustainable World. Frontiers of Business Ethics 9.Forlaget Peter Lang: Oxford; 209–228 .i A. Tencati,L. Zsolnai

Van Assche K, Verschraegen G. 2008. The Limits ofPlanning: Niklas Luhmann’s Systems Theory andthe Analysis of Planning and Planning ambitions.Planning Theory 7(3): 263–283.

van Buuren A, van Boons F, Teisman G. 2012.Collaborative problem solving in a complexgovernance system: Amsterdam airport Schipholand the challenge to break path dependency. SystemsResearch and Behavioral Science 29: 116–130.

Wallington T, Lawrence G, Loechel B. 2008. Reflectionson the legitimacy of regional environmental gover-nance: lessons from Australia’s experiment in natu-ral resource management. Journal of EnvironmentalPolicy & Planning 10(1): 1–30.

Whitehead M. 2003. “In the shadow of hierarchy”:meta-governance, policy reform and urban regener-ation in the west Midland. Area 35(1): 6–14.

Wilke H. 1993. Systemtheorie. UTB für wissenschaft:Stuttgart.


