Making sense of governance

Abstract

Governance is a term which is widely, but not always precisely, used, and this article seeks to clarify what the term is being used to mean. In particular, it is concerned with whether it denotes a particular mode of government, or whether is a broad category encompassing all modes of government. It focuses on the arguments about political practice on which the original claims about governance were based, and the evidence that there has been a change is political practice which demands a new label. It concludes with a discussion of the way that accounts of government are used in the practice of governing, and the incentives that this gives both participants and observers to adopt the warm but fuzzy term ‘governance’.

What are we talking about?

There is no doubting the appeal of the term ‘governance’ in contemporary discussion of governing, but it can be asked if the widespread use of the term has been accompanied by fuzziness about its meaning. Does it, for instance, denote a particular way of governing, or does it cover all modes of governing — in which case, do we gain anything by using the term? This article seeks to open up discussion about governance as a concept, both for observers and for practitioners: about the meaning being given to it, about its relationship to practice, and about the extent to which it adds to our ability to make sense of the process of governing. It poses a counterfactual — an alternative theorisation of the same evidence — and explores the use of multiple accounts in the analysis and practice of governing, and the significance of this for the theorising of organisational form, and concludes with a review of governance as a construct. To facilitate this discussion, the argument is set out as a series of propositions.

1 Governance has been widely adopted as a concept, but without much clarity about its meaning (or perhaps because of this)

Fifteen years ago governance was relatively unknown as a concept, but now it gets more mentions on Google Scholar than ‘government’, ‘politics’ or ‘democracy’. (This not true of Google itself, suggesting that the term has much more appeal to academics than to the public at large.) For a term that has become so widely used, there has been remarkably little concern about its meaning. In Rhodes’ pioneering work Understanding Governance (1997), he identifies six different meanings given to the term, and suggests that it ‘really has no meaning’, but passes quickly and without explanation to a seventh, his own ‘stipulative definition’, without much concern to demonstrate why this is superior to the others. Van Kersbergen and Van Waarden (2004) identify nine different meanings, and Offe (2008, 2009) asks if the term is an ‘empty signifier’.

Rather than searching for ‘the meaning’ of such a variously used word, it is perhaps more fruitful to investigate the contexts in which ‘governance’ came to be used. But here we need first to take note of Offe's distinction between using governance as a way of distinguishing one mode of governing from another (Gegenbegriff), or using it as a way of encompassing all modes of governing (Oberbegriff). As we shall see, the term began as a distinguishing marker (Gegenbegriff) but is now commonly used as a comprehensive Oberbegriff. The increasingly authoritative Wikipedia (quoting Weiss & Thakur, 2007) defines ‘global governance’ in such all-encompassing terms as

the complex of formal and informal institutions, mechanisms, relationships, and processes between and among states, markets, citizens and organisations, both inter- and non-governmental, through which collective interests on the global plane are articulated, rights and obligations are established, and differences are mediated.

Although it is possible to find scattered historical examples of the term being used with a range of meanings or no particular meaning (e.g. Wilson's, 1976 memoir The Governance of Britain), the spectacular adoption of the term in the 1990s can perhaps be traced to two streams of discourse. One was among practitioners, marked by the World Bank's (1989) diagnosis that the lack of development in Africa was due to a ‘crisis of governance’. It defined governance as

“… the exercise of political power to manage a nation's affairs”. Good governance included some or all of the following features: an efficient public service; an independent judicial system and legal framework to enforce contracts; the accountable administration of public funds; an independent public auditor, responsible to a representative legislature; respect for the law and human rights at all levels of government; a pluralistic institutional structure, and a free press (Leftwich, 1993: 610).

Here, the bank was launching a political reform agenda, but as it is prohibited by its Articles of Agreement from involvement in political matters, it called this agenda ‘good governance’, and this usage became widespread among reformers. After all, who could be in favour of ‘bad governance’?

The other discourse about governance, by social scientists, came by a different route. The most significant voices were perhaps the international relations scholars who noted the construction of regimes of rule which did not rely on claims of the sovereignty of states, and wrote about ‘governing without government’ (Rosenau & Czempiel, 1992). This was a significant departure from the Westphalian assumption that governing is accomplished by the exercise of the authority of states, and was criticised for its imprecision (see Smouts, 1998), but it was congruent with observations of examples of cross-national governing such as Antarctica, where there are no universally recognised claims to sovereignty, or the management of epidemic disease, where one consequence of the SARS epidemic was the strengthening of the power of the WHO to involve itself in health crises without invitation from the national government. The term was applied to the analysis of government by Rhodes, building on earlier work about linkage in local government in the UK. He argued that the UK was moving from ‘government’ to ‘governance’, a situation in which

The state becomes a collection of interorganizational networks made up of governmental and societal actors with no sovereign actor able to steer or regulate (1997: 57).

So government is rule by direction; governance is rule by self-organising networks. Here, governance is clearly a distinguishing term (Gergenbegriff), used to clarify a claimed historical transition.

The new term attracted a good deal of attention, and became widely used by both scholars and practitioners (see, e.g. Jose, 2007) It had strong appeal in Western Europe, and had a particular attraction in the EU, where the quest to establish a single European market led to the very evident reconstruction of the regimes of rule in wide areas of social and economic life. The ‘rules of the game’ which had developed in different countries in various spheres of social and economic life were excavated, challenged, and superseded by new European rules. This was accomplished by an extended process of negotiation which was marked by complexity of identities and interests, information asymmetry, and limited scope for resolving contests by authoritative decision: it was a very public mode of regime-making, and seemed to exemplify the claim that rule by governments was being replaced by negotiation within networks, bringing the social science and practitioner perspectives together. This perception overlooked the way in which the now-superseded patterns of rule had been laid down; as Bismarck is said to have remarked, ‘Laws are like sausages: if you want to enjoy them, you should not see them being made’.

As Trofing and Sorendsen (this issue) note, ‘there is considerable confusion and disagreement about the definition of governance’. The term is applied both to particular forms of rule and the whole process of rule. There has been a range of Gegenbegriffe, coupled with a tendency to attempt to draw them all into a single Oberbegriff. Torfing and Sorendsen see governance as governing through networks, but Howlett and Ramesh (this issue) see it as ‘the mode of government coordination of societal actors exercised by state actors’ and Bell and Hindmoor (2009) argue that ‘governments typically use four different modes of governance’, markets, contracts, partnerships and inculcating self-discipline’: in both cases, ‘governance’ is what governments do. Outside academia, governance is commonly used in ways which have little to do with negotiation among stakeholders. In the Australian Public Service for instance, ‘governance’ is about managerial control, and ‘corporate governance’ in business is a sort of a constitutional law of the firm. The Wikipedia entry on ‘governance’, which includes hierarchy, markets and networks as forms of governance, also adds corporate governance, global governance, project governance and information technology governance. The question is what the use of the term add to our capacity to analyse the process of governing, or whether, as Offe asks, it has become an ‘empty signifier’ (Laclau, 1997).

2 This does not necessarily detract from its analytical utility

The fact that there are a number of different usages of a term in use is not in itself a problem, but it directs our attention to what the term is being used to signify, and in what way it strengthens analysis. The assertion that governance denoted a shift in practice, and that we were moving to an era of ‘governing without government’ through ‘self-organising networks’ was soon subjected to empirical challenge by researchers who found that the networks were not self-organising and government was not diminishing in significance (e.g. Bache, 2000; Johansson & Borell, 1999), and Marinetto (2003) questioned whether something new was being identified. This criticism did not appear to have much impact on the popularity of the term, since it was widely agreed that something was happening to government and we needed a word for it. By 2003, Rhodes was saying ‘We use governance as our preferred shorthand phrase for encapsulating the changing form and role of the state in advanced industrial societies and a key facet of these changes is public sector reform.’ (Bevir, Rhodes, & Weller, 2003: 13–14) although its application to ‘advanced industrial societies was qualified by the caution that “Governance” as applied to British government will mean something different to “governance” in France’. It was now much more of an Oberbegriff than a Gegenbegriff. The term also seemed to become less of an analytical construct than a label for the actors’ accounts of their own experience: ‘We sought to document elite constructions of governance in general and public sector reform in particular. … We focus on elite beliefs … we are not providing our account of the problems’ (ibid. 14), and this theme was developed in a further collection of ‘governance stories’ (Bevir and Rhodes, 2006).

Rather than defining all the practices of governing as governance, it might be more fruitful to take governance seriously as an analytic construct which identifies some specific social phenomena and see what it contributes to our understanding of observed practice — that it, to retain it as a Gegenbegriff. Stoker (1998) identifies the key aspects of governance as involving:

•	institutions and actors from within government but also beyond it;
•	a blurring of roles and responsibilities;
•	power dependence in relationships between institutions;
•	autonomous self-organising networks;
•	governing by the use of new techniques to ‘steer and guide’ rather than command.

This gives us a workable perspective — a Gegenbegriff. As Stoker says, this perspective ‘deliberately selects various trends and developments for our attention. Its value is to be judged by how good or bad the selection has been.’ (Stoker, 1998: 18, 26). We can be more precise about what it is that is being identified and how the explanation generated in this perspective compares with alternative explanations. We might also ask what it is about the term that makes it so appealing in such a diversity of ways.

3 Governing is much too important (and too difficult) to be left to the government (as we have always known)

The governance narrative embodies the assumption that (a) non-government participants have a significant role in the process of governing, and (b) this has happened recently. It is uncommon for these assumptions to be spelled out in any detail, or the basis for them established with much analytic rigour and empirical evidence. It is obvious that non-government is involved in governing, and this is clearly at variance with the account of governing as the work of the government. It follows that there has been a change: once, governing was done by the government, now, it is done through a process of negotiation involving non-government participants as well.

This reflects Foucault's observation (1986: 88–9) that in our political and social analysis, we have still not cut off the king's head: we have been accustomed to the use of a way of talking about governing which presents it as the work of a superior sentient being called ‘the government’. This is exemplified and reinforced by the discourses, practices and symbols of both practitioners and observers. For instance, it might argued that ‘the government’ needs to act to curb alcohol-fuelled violence. Academic writing accepts and reinforces this assumption. ‘Public policy is how politicians make a difference’, say Althaus, Bridgman, and Davis (2007: 5), and it is accomplished through cabinet decisions. Academics appraise the achievements of ‘the Howard government’ or ‘the Blair government’. This is reflects the modernist liberal democratic narrative of government, which elaborates a functional role for ‘the government’ and identifies this with a loosely defined condensation of political and bureaucratic authority. This narrative explains and validates both the practices of the political and bureaucratic leaders, and the contribution of social science to these practices. Governments (or their advisers) are said to recognise problems, identify ways of responding to them, evaluate consequences of these options, and arrive at decisions — a technology of government as episodic instrumentalism (‘intervention’). To paraphrase Voltaire, if ‘the government’ did not exist, it would be necessary to invent it.

That ‘non-government’ is an integral part of the process of governing has long been recognised. A.F. Davies opened his Australian Democracy

It is as though there were a political gateway through which all issues pass. Disputed from the moment they are in sight of it — and more hotly as they approach — they pass (if they pass) through, and drop out of controversy for a time. Managing the procession are certain ‘gatekeepers’ — not just the Cabinet of the day, but bureaucrats, journalists, association heads and independent specialists camped permanently around each source of problems. To talk of a political process is to recognise some hint of pattern in the way in many different fields the controversial is transformed into the routine (Davies, 1964: 3).

Here, nearly fifty years ago, Davies was anticipating by fifteen years the concept of a ‘policy community’ (Richardson & Jordan, 1979), and he prefaced these opening lines with a long quotation from an American political science classic, now over a century old: A.F. Bentley's The Process of Government (Bentley, 1908). Thirty years after Bentley, Griffith was advocating the study of ‘the ‘whirlpools’ of special social interest and problems’ (Griffith, 1939: 182–183) Similarly, Loveday's study on industry policy in Australia (Loveday, 1982) showed that the process of governing was not contained in neat constitutional envelopes, but was accomplished through a complex of organised relationships and interdependencies. So we cannot make sense of it in terms of an account of the separation of political, administrative and business participants, and the centrality of authoritative decision: they are all part of the one process, which cannot be understood in terms of the constitutional formulations of responsible government. Moreover, Loveday referred to similar arguments by political scientists in Britain (e.g. Self, 1972) and North America (Doern & Wilson, 1974; Heclo, 1972; Peters, 1978). He also referred to the emerging argument about ‘corporatism’, already becoming a ‘growth industry’ (Panitch, 1980), which was the first in a succession of metaphors recognising the interactive and negotiative nature of the process of government, including ‘policy community’, ‘issue network’, and ‘advocacy coalition’.

In other words, the key elements of the governance narrative — actors from inside and outside government, the blurring of roles and responsibilities, interdependence, and the relative unimportance of command — had already been recognised by political scientists before governance was coined as an analytical construct. But when these phenomena were identified as ‘governance’, they were seen as evidence of a change in practice: before, there was a clear separation between government and governed, and decisions were imposed by state authority, but now, there is organisational inter-linkage and outcomes emerge from negotiation. But these conclusions usually do not flow from a systematic comparison of the governing of some field of practice now as compared to some point in the past, but can simply be asserted because they are congruent with what ‘everyone knows’ — or rather, what social scientists assume: that is, that governing is the work of government, and that any finding that it is not must represent a significant change. For the practitioner it can be presented as a bold innovation (see Metze, 2006), and for the academic observer, an important discovery which can generate publications.

In this context, there is less inclination to subject to critical scrutiny the claim that government has only recently become interactive and negotiative, or the corollary that this has become necessary because of the increasingly complex nature of modern society (Jordan, 1990). As we have noted, there is little interest in the sort of historical comparison which might validate the claims about changing practice. Nor is there much interest in exploring the possibility that there might be different presentations of the process of governing, that they might be used for different purposes, and that the model of systematic problem-solving by official policy-makers might operate to validate the outcome rather than to describe the process by which it was reached. Comparing empirical accounts of present practice with official accounts of past practice is not conclusive evidence of a change in practice, particularly when there is reason to suppose (as suggested by the Davies and Loveday studies mentioned) that past practice was more interactive and negotiative than was suggested by official accounts.

4 Governing is accomplished through complex organisational forms (but some people want to tell a simpler story)

As Foucault noted, it is not easy to escape the assumption that governing is done by the government, given that the constant expansion, subdivision and reorganisation of the government workforce is described in terms of its functional concern for matters to be governed. This tendency is amplified by the propensity of organisational hierarchs to set or demand indicators of achievement by which the contribution of organisational units to governing can be judged. In this perspective, the assumption is that the governing is accomplished by a unit within government, and that the place of other bodies, whether within government or outside it, is problematic. In this government-focused perspective, even ‘multi-level government’ is seen as innovative. ‘Consultation’ with non-government bodies is acceptable, but the making and application of rules is the work of government. Involvement with these other bodies in a less-controlled way is seen as calling for noting and labelling — e.g. as ‘engagement with stakeholders’, ‘co-production’, or as ‘governance’. Boxelaar, Paine, and Beilin (2006) offer an illuminating example of the tension between a policy approach based on mobilising non-government actors (in this case, farmers) with a control-oriented management concern with ‘SMART (Specific, Measurable, Accountable, Realistic, Timebound) outputs/deliverables’. In this case, the extension agents saw the task as constituting modes of self-rule among farmers; the departmental management saw it as the delivery (by officials) of services to customers: it was important to demonstrate that it was the department and its staff that were doing the governing.

Rather than assuming that governing is a process in which ‘officers of the state attempt to rule’ (Goodin, Rein, & Moran, 2006: 3), it might be more promising (and certainly less pre-determined) to start with the field of practice, and to look for the ways in which the participants seek to order practice, the extent to which this involves recognising or constituting or deferring to explicit agencies of order, and the relationship between these agencies and the machinery of government. Rather than assuming that (for instance) the existence of a fire service is the result of the government recognising a problem and designing the most appropriate agency to deal with it (cf. Hill & Hupe, 2002: 190), we might ask how people have responded to the risk of fire, noting the activities of householders, neighbours, community groups, insurance companies and local authorities, and asking how this became an issue for ‘the government’. In this context, we might have a clearer idea of the nature of state bureaucratic involvement in the management of fire risk, and its relation to the continuing activities of householders, insurance companies and community organisations. We would not be surprised to find (for instance) that a body established to fight fires and described as a statutory authority of the New South Wales government is staffed by 700 government employees and 70,000 unpaid volunteers and that half its budget comes from insurance companies, nor would we be driven to find a new term to describe this apparently anomalous form of governing. We would recognise that actors in a variety of institutional settings had, over time, constructed and modified an array of practices for dealing with the risk of fire, which had then been ‘enacted’ through the organisational forms of government. In a wide range of areas of social practice, such as land use, the care of the aged, and the consumption of alcohol, we can see governing being accomplished by a continuing interplay of official and non-official forms, underpinned by the ‘collective puzzling’ (Heclo, 1974) through which the concerns of policy are created. What is new is not the involvement of ‘non-government’ in the state; it is the recognition — by officials and by academic and other observers — of the governing being done by non-government and its incorporation in the organisation chart (see Gaudin, 1998; for another illustration of this process, see Colebatch, 2002.)

Seen in this perspective, the phenomena which have been identified as the characteristics of governance may have been parts of the process of governing for some time, and not a late-20th century innovation in response to complexity. It could be that what is changing is not so much the practice of government but the public accounts that are given of this practice. In the second half of the 20th century, the discourses of governing in Anglophone liberal democracies tended to focus on bureaucratic state activity, and the other elements of governing received less attention. While (as we have seen) the place of non-bureaucratic activity in governing was recognised, it was not well theorised, being largely seen as a footnote to (perhaps a problem for) an account of governing as a process of organised public contention giving rise to bureaucratic state activity. One prominent Australian political scientist advised me sagely that discussion of these questions was really ‘organisational sociology’ and political scientists should ‘stick to their last’ (i.e. focus on voting, parties, issues and points of public contention) — an interesting illustration of the role of disciplinary gatekeepers in paradigm maintenance.

5 And multiple accounts are mobilised in constituting and explaining governing

This discussion of ‘official’ and ‘non-official’ elements of governing directs our attention to the way that the activity of governing is suffused by accounts of governing, that different accounts are in play at different times, with a good deal of overlap between accounts and ambiguity in their use. We can also see that the mobilisation of accounts is part of the practice of governing, and that what constitutes a ‘good account’ of the governing of some field of practice may depend on the context and the time. Seeing the governing of harvest practices as autonomous self-management will ‘make sense’ to farmers, but senior officials may make sense of it as the delivery of government services to customers. In this connection, we have to recognise that the overt actions of government may be expressing rule rather than creating it, that it is the interplay among active participates that generates an outcome which can then be ‘enacted’ (Weick, 1979) by the formal and visible practices of government. This raises questions about the relationship between different accounts of governing, and between these accounts and the practices of the participants. For this reason, we now need to examine the nature and significance of the different accounts of governing that are in use — and in this context, we will be confronted with the nature and significance of academic accounts of governing.

The use of multiple accounts in the explanation of governing is in fact normal practice, both for academic observers and for practitioners. People commonly make binary distinctions between accounts in use: e.g. Howard's practitioner informants noting that ‘the [policy cycle] model is really about theory, not practice’ (Howard, 2005: 10). Other distinctions include between formal and informal, or formal and effective (Bailey, 1969/2001), between normative and empirical, or sacred and profane (Colebatch & Degeling, 1986). The implication is that both accounts have their uses.

Graham Allison took this further in his analysis of the Cuban missile crisis (Allison, 1971), arguing that it was necessary to look at the same situation through different lenses to fully understand it, seeing government as being at the same time a rational actor, an array of disconnected specialisations, and an arena of rivalry. Each of these accounts, be argued, explained part of the outcome; none of them explained all of it. Tenbensel (2006), drawing on Flyvbjerg and Aristotle, argued that policy practitioners draw on multiple sources of knowledge — episteme (systematic knowledge from study), techne (knowledge from experience), and phronesis (ethical–practical judgement) — and to be effective in policy work, practitioners have to be able to deploy the appropriate knowledge in any specific context.

On a broader canvas, the ‘governmentality’ approach (e.g. Dean & Hindess, 1998; Dean, 1999; Rose & Miller, 1992; Rose, 1999) has shown the importance of understanding governing not only in terms of specific actions and structures, but also on the shared meanings and contexts within which these actions ‘make sense’, and the recognition of the persons and processes which constitute these contexts. But there are multiple forms of sense-making in play, and actors tend to recognise this, and the tension which it causes, and seek outcomes which make sense in different meaning-frames (see, for instance, Howes, 2008; Morrison, 2007). Meanings are fluid and context-related, and are developed and constituted in use (see McLaverty & Halpin, 2008).

For these reasons, I have previously argued (Colebatch, 2006a, 2006b, 2009) for an analysis which recognises three overlapping and interacting accounts:

•	authoritative choice, focusing on the structures and practices of legitimate authority and cast in terms of coherence and instrumentality;
•	structured interaction, focusing on the interplay between different organised participants inside and outside government, whose concerns are recognised and institutionalised in different ways;
•	social construction, focusing on the broader and long-running framing of what is to be governed, what is valid knowledge, what are shared values, and who can speak with authority — the ‘collective puzzling’ identified by Heclo (1974).

All of these accounts are relevant in creating and maintaining the ‘structured commitment of important resources’ (Schaffer, 1977) that makes policy, and I would argue that taking this approach to governing helps us to recognise the complex way in which it is constituted. Bureaucratic agencies assume the existence of other organisational forms, such as industry associations, schools and families, and seek to enlist them in their projects of governing. Stakeholders negotiate with one another ‘in the shadow of hierarchy’, aware of the need to seek official endorsement of negotiated outcomes, and of the possibility that rules can be imposed if consensus cannot be achieved. And rule-making takes place over matters that need to be ruled — that is, have been socially constructed as policy concerns. Crenson's landmark study (1971) showed how air pollution could be regarded as a policy concern in one steel-making town, but not in a comparable town which was a ‘company town’ where the one steel plant was the major employer in town; there, the pollution was regarded as normal. More recently, after Al Gore won the Nobel Peace prize for his work promoting public awareness of climate change, there was a move to draft him back into politics as the Democratic presidential candidate, which he resisted; one observer commented that Gore:

He's also come to believe that even a US president is powerless to act on climate change unless public opinion has moved, that acting as a teacher and advocate can have a greater political impact (Freedland, 2007).

These distinct narratives of governing reflect the different structural forces inherent in the constitution of policy, and interact with one another in the constitution of policy. And the accounts of governing form part of the governing, as the apparent outcomes of continuing negotiations over socially constructed concerns are ‘enacted’ through the forms of state authority. This is a structural dimension of governing, not something that has been forced on hitherto-competent bureaucratic states by increasing social complexity.

And in a broader sense, the use of these accounts of governing — by practitioners, commentators and academic writers — is part of the governing. It is primarily the observers of governing — academic researchers and writers, journalists and other commentators — who create the account of authoritative problem-solving, locating official acts in a narrative of coherence, intention and improvement. This account is then mobilised by practitioners, who articulate their activity in terms of ‘problems’, seek expert knowledge about the problems, consult with relevant stakeholders, invoke the appropriate authority, and announce the ‘decision’ — and in this way, ‘perform’ policy.

This calls for practitioners to be sensitive to context, and to frame their activity appropriately. But while this ‘tolerance for ambiguity’ becomes a valuable skill for practitioners (Tenbensel, 2006), it presents problems in academic discourse, which prefers universal, systemic explanations of practice. And the academic study of policy has emerged from a social science traditionally oriented to state bureaucratic activity, and sharpened by the response to Lasswell's call (1951) for a ‘policy science’, which gave rise to graduate programmes in ‘policy analysis’ which would equip the expert analyst to ‘advise the Prince’. It was assumed that governing was accomplished by the state structure, and what was needed was the knowledge before the event, and evaluation after it, which would enable the optimal choices to be made by ‘the Prince’. When the experience of governing was otherwise, this was taken as evidence of deviance: as ‘regulatory capture’ or ‘implementation failure’, and of the need for clear separation of decision and programme administration, more precise specification of the task, and more regulation of process.

But at the same time, there has been the increasing recognition (in some Western liberal democratic polities) of the non-official component in governing — in discourse, in practice, and in organisational forms. What were once stigmatised as ‘vested interests’ are now legitimated as ‘stakeholders’ and incorporated into state structures and practices. The governance discourse saw this as the result of changing practice: before, there had been separation of government and non-government, and rule by official direction; now, there is collaboration and negotiation. But there has been little empirical evidence of this change, and it is clear that the involvement of official and non-official practice in the accomplishment of governing is of long standing. But the narrative of authoritative choice, with its stress on order, specificity and transparency, remains strong, so practices of social construction and stakeholder negotiation have to be ‘enacted’ through discourses of ‘governance’.

6 Governance could be a useful analytic construct, but is likely to be more popular as an empty signifier

It appears that through a process of ‘creeping indeterminacy’ (Gaudin, 1998), ‘governance’ has come to convey no particular meaning. At best, it is an Oberbegriff, replacing terms like ‘government’ or ‘the political process’ without improving on them in any way. For Howlett and Ramesh (this issue), governance is an empty construct: any mode of governing is ‘governance’, whether ‘hierarchical’, ‘non-hierarchical’, ‘alternative’ or ‘market’. For them (and many other scholars), the term does not differentiate any political order from any other. At worst, it has become a tabula obscura, as it were, on which users can inscribe any meaning that suits their purpose. And we have only been concerned with the use of the term in academic social science; in ordinary discourse, governance is used in a wide variety of ways, few of which bear any resemblance to Rhodes (1997) vision of governing through self-organising networks. There is no intellectual obstacle to retaining it as a workable analytic construct, a Gegenbegriff, along the lines defined by Stoker ten years ago, but the dilution of the term into an Oberbegriff, and the distraction of other usages like ‘good governance’ and ‘corporate governance’ makes it difficult to use the term to clarify the argument.

If governance were to be used as a Gegenbegriff for the analysis of contemporary practice, it would be clear that it is both conceptually under-developed (see Offe, 2008, 2009) and lacks empirical support. The confident assertions of the 1990s about the emergence of new forms of rule would have to be heavily qualified. There have certainly been organisational changes, but it is not clear that governing is now the work of self-organising networks that are independent of the government. It would be recognised that the evidence provides little support for the proposition that the involvement of non-government in governing is a late 20th century development, or for the assertion that this form of governing is a functionalist necessity, forced upon us by the complexity of modern life.

In this context, we might look for the counterfactual, the alternative Gegenbegriff, and this paper has argued that this would focus on the social construction of rule, the way in which social actors (including government officials and political leaders) order their practice, and draw upon the discourses and practices of government (and other realms of practice) as part of this ordering (see Colebatch, 2002). We would recognise that governing involves many hands, is grounded in interaction rather than direction, and is a continuing process marked by indeterminacy and ambiguity. Official organisational forms and practices are mobilised as part of the governing, and are depicted as its source, but the enactment of governing is not the same as the process by which it is accomplished (as Bagheot, 1963 [1867] and Montesquieu, 1914 [1782] showed).

This would mean shifting the focus from ‘government as an actor’ or even ‘government actors’ to ‘action which creates governing’, and the place of official forms and practices within this action. This is not quite the same thing as investigating the characteristics of ‘collaborative and deliberative forms of governance’, but it is consistent with the approach of the Commission on Global Governance, composed of eminent public figures, which reported in 1995. It saw the defining characteristics of governance as:

-	it is neither a rule system nor an activity but a process;
-	it is founded on accommodation rather than domination;
-	it involves public and private actors at the same time;
-	it is not a formal institution but relies on continual interaction;

(see Commission on Global Governance, 1995; Smouts 1998).

This resonates with the usages of writers like Stoker (1998) and Rhodes (1997), but it is not focused on official institutions, nor is it an attempt to characterise a system; it is a way of talking about the practice of governing. In this respect, it represents a challenge to the disciplinary norms of academic life, since it is not focused on the phenomena which are traditionally seen as ‘political’, and suggests that their role in the process may be less important than is claimed. It recognises the significance of the institutional recognition of non-official actors in governing, but does not insist that their presence is a recent and functional reaction to increasing social complexity. But we may learn more from looking at ‘governance’ not as the next (or perhaps current) big thing in political science, but as a way of locating the concerns of political scientists within a broader framework for theorising social action.