Inclusive and Exclusive Social Preferences: A Deweyan Framework to Explain Governance Heterogeneity

This paper wishes to problematize the foundations of production governance and offer an analytical perspective on the interrelation between agents’ preferences, strategic choice, and the public sphere (defined by impacts of choices on ‘publics’ who do not have an input in strategic choice, and by contextual conditions). The value is in the idea of preferences being social in nature and in the application both to the internal stakeholders of the organisation and its impacts on people outside. Using the concept of ‘strategic failure’ we suggest that social preferences reflected in deliberative social praxis can reduce false beliefs and increase individual wellbeing. From this approach, the paper offers a taxonomy of production organisations, based on social preferences about two variables: (i) the governance form (i.e. ownership and control rights) (ii) other strategic decisions that characterize the management of a company at a more operational level, once its fundamental legal form has been chosen. Each dimension (governance and strategic decisions processes) is then categorised alongside two basic preferences: towards inclusion or exclusion of ‘publics’ that have no substantial access to decision power about these variables. Our framework explains governance heterogeneity by contrasting exclusive and inclusive social preferences in cooperatives, social enterprises, as well as traditional corporations. A discussion of the evolution of social preferences and organizational forms is addressed through examples and regional experiences.


Introduction
We can observe instances in which radical innovations in governance and decision-making processes have been introduced by innovators as highly reasoned and structured replies to the experienced failures of production organisations to meet wider societal needs. 'Creative responses', using Schumpeter's wording (Schumpeter 1947), were searched by the founder of worker cooperatives such as father Jose Maria Arizmendiarreta, creator of Mondragon in the Bask countries. This is perhaps the most followed and celebrated example, but certainly not the only one. Employee buyouts were pioneered, in the 1920s UK, by John Spedan Lewis in the retail sector and more recently, in the 1980s, David Erdal led the transition to employee ownership of the family paper mill Tullis Russell (Erdal 2011). The complex constitutional settings that innovators elaborated expressed preferences about aims and processes that were in stark opposition with the corporate governance and work policies of the 1920s and 1980s. The role played by individual choice, nonetheless, does not rule out the relevance of the context. In this sense, the entrepreneur's choice can be considered as a highly reasoned reply to historical contextual conditions, facilitated or obstacled by the institutional network and social relations in which they are embedded, whilst at the same time remaining central to the introduction of governance innovations and their diffusion (Granovetter 1992;North 2005). In line with socioeconomic approaches, we therefore view the entrepreneur's choice of governance and subsequent strategies as the expression of the preferences of a socially embedded individual (Granovetter 1992).
Leading from these considerations, the paper wishes to problematize the foundations of production governance and offer an analytical perspective that unbundles the interrelationship between agents' preferences, strategic choice and the public sphere (here defined by impacts of choices on 'publics' who do not have an input in strategic choice, and by contextual conditions). Specifically, by redeveloping a foundational perspective on the meaning of the public sphere, the paper aims at clarifying the potential of different preferences to meet societal needs. From this approach, the paper offers a taxonomy of production organizations, based on social preferences about two variables: (i) the governance form (i.e. ownership and control rights) (ii) other strategic decisions that characterize the management of a company at a more operational level, once its fundamental legal form has been chosen. Each dimension (governance and strategic decisions) is then categorised alongside two basic orientations: towards inclusion or exclusion of 'publics' that have no substantial access to decision power about these variables.
Analytically, the private and the public (or socially embedded) dimensions of individual action have been traditionally kept separate. In On Liberty, Mill seeks the philosophical basis for protecting individuality from the authority of society. The latter, for Mill, identifies the meaning of 'public', which should not interfere '…when a person's conduct affects the interests of no persons besides himself…' (Mill 1859/1869. The interconnections between private actions and public impacts have been however explored in the analysis of market failures. 1 More recently, behavioural theory has explicitly accounted for the interaction of preference formation with contextual conditions such as economic, social, political and cultural institutions (Bowles 1998). Within organisations, individual preferences have been argued to respond to incentive systems and to the experiences activated by interactions with co-workers and managers (Ben-Ner and Ellman 2012).
Consistently, the public dimension, in the interpretation of this paper, is not the arena of governmental policies, 2 as in Stigler (1971) and Posner (1974), or central planning, as in Hayek (1944). Rather, we consider the public dimension as (1) the objective environment represented by the social, economic, cultural institutions which affect the formation of particular tendencies in the way individuals act (prior to action); (2) the known and unknown variable wave of influences that radiates from each individual choice (following action). What we aim at stressing more explicitly, in comparison with established theories, is that preferences and related choices are not, by their very nature, purely private, not least in their antecedents and consequences. Rather, following Dewey (1922Dewey ( , 1927, we openly recognise that, not some, but each private choice must include a public dimension: 3 Breathing is an affair of the air as truly as of the lungs; digesting an affair of food as truly as of tissues of stomach … There are specific good reasons for the usual attribution of acts to the person they immediately proceed. But to convert this special reference into a belief of exclusive ownership is as misleading as to suppose that breathing and digesting are complete within the human body. (Dewey 1922, p. 24). 1 The idea of externalities has been widely acknowledge in economics. As we explain later in the paper, however, our approach differ from market failure, supporting the necessity of developing complex coordination mechanisms aimed at discovering complex connections through enquiry and deliberation, beyond the price mechanism. This can be justified because external effects may need to be discovered but also, and more crucially, because it is from the Footnote 1 continued choice of processes and praxis that wider social consequences derive. Moreover, deliberative processes originate social outcomes that are at least partially intended and governed, whilst the price mechanism generates externalities, which are not part of the objective function of the decision-maker (at last formally). 2 According to standard views in economics, the State is viewed as acting for the public interests against market failures or, as the Chicago school suggests, as the maker of regulatory policies which are nonetheless captured by specific industries for their own private interest. These perspectives and debate are reviewed by Chang (1997). 3 In Sandel's view, the temptation to decontextualize choice from its context has seduced Rawls (1971) who, whilst seeking a construct for achieving just choices, had to cut bridges with individual identity and experience (Sandel 1982;Quinn et al. 1997). This is however a problematic argument that would deserve a wider debate. In Rawl's defence, the pre-commitment to the creation of an unbiased normative framework can be considered as a necessary condition for the development of the type of democratic interaction envisaged for example in the pragmatist approach.
A Deweyan approach, in this sense, underpins also the economic contributions mentioned above, for which particular patterns in the choice of processes and aims are not to be attributed solely to the individual dispositions of the decision-maker (e.g. the entrepreneur, the worker, the consumer), but also to a contextual component defined by the habits, norms and established practices which underpin choice and the attainment of outcomes. Our working hypothesis, in line with behavioural theories, is that contextual conditions concur in the definition of individual dispositions. The other side of the coin would be that individual dispositions can affect existing institutions, socio-economic aims, processes and outcomes. We explain that enquiry-based processes are a pre-condition to make sense of the complexity of such interconnections, looking for solutions that reduce the failure of production organisation to meet each and every need across society. In this work, enquiry-as a way of thinking-explains also the nature of social preferences (since assessment and consequent changes in preferences are based on such praxis) and organisational forms.
A further clarification is also needed on the idea of consequences, or outcomes. Because the nature of the decision process is the outcome of 'an act of choice' (Sen 2002, p. 159), we regard the choice of process jointly with the generation of outcomes since, as Sen points out, particular processes are preferred in view of their anticipated ability to achieve certain outcomes or avoid undesired ones. 'Comprehensive outcomes', in Sen, include both the choice of process and their expected 'culminating' outcomes (ibid.). We say 'expected' since whilst the actor may have a particular aim in mind, the actual result may be different. Here contextual elements matter in determining the final outcome (Dewey 1922).
With an emphasis on the interconnectedness of individual choices and contextual components, we consider all preferences underpinning production choices as 'social', or having a public dimension, whether inclusive or exclusive of the effects on others and society. We then argue that in the current economic environment, strategic choices do not reflect, as a norm, dispositions towards the inclusion of the public dimension of choice, therefore preventing production choices to achieve collectively beneficial ends (Cowling and Sugden 1998a). Following these considerations, we present a framework to discriminate among business types and provide possible explanations for the emergence and persistence of exclusive rather than inclusive preferences in the choice of organisational forms and processes. In particular, we reason on what elements can be expected to lead to changes in the nature of social preferences amongst economic actors, reinforcing, in our conclusions, the role of individual dispositions as well as the meaningfulness of institutions and policy action in supporting and empowering the expression of inclusive social preferences. Non-systematic evidence based on specific examples is used to illustrate our arguments.

The Public Dimension of Preferences and Outcomes
To explain the ambiguities that can originate when overlooking the interconnections between individual action, socio-economic institutions, and public consequences consider the conceptualisation put forward in Ben-Ner and Putterman (1998): Self-regarding preferences concern the individual's own consumption and other outcomes, other-regarding preferences concern the consumption and outcomes of others, and process-regarding preferences concern the manner in which the individual in question and others behave, including the ways in which they attain outcomes of interest. We shall refer to process-regarding preferences mainly as values, but sometimes also as codes of behaviour, mores, ethics, and by other terms, depending mostly on the context (Ben-Ner and Putterman 1998, p. 7) A pragmatist perspective may help to comprehend how this conceptualisation may overlap at a number of crossroads. As private actions have public bearings, it follows that also self-regarding preferences underpinning private actions have a public impact: they regard the sphere of others. Applied to a standard ultimatum game for example, this means that the 'private' choice of the proposer engenders a public sphere, whatever the degree of fairness of the proposer's decision, as the proposer's choice impacts on the recipient's welfare and sets her reaction, which, in turn, affects the proposer's welfare. A pragmatist perspective suggests that the public dimension would be present even if the decision did not account for others' welfare, or that there is a broader public and societal dimension of consequences which transcends the individual perspective. 4 The idea of procedural preferences, in parallel, reflects how the agent wants to achieve a certain intended aim. They can therefore be understood in terms of the agent's assessment of the rights and duties to be entailed by the process attached to an outcome of interest . It follows that a process-based perspective always entails a view on aims and outcomes, since each process will be designed in view of opening up a certain set of opportunities, and avoid unwanted consequences (ibid.). Thus, the largest the distance between desired and existing processes, the lowest the individual wellbeing will be. As an illustration, suppose there are three potential alternative processes X, Y, Z and that the preferred outcome O can only be reached by processes X and Z. The agent prefers X to Z whenever X is available. Process Y instead can achieve outcome C. If in the agent's state of affairs only process Y is available, outcome O is not an opportunity. The actual outcome will not meet the agent's preference, meaning that her needs or desires will not be fulfilled.
Moreover, like Dewey in philosophy,  and Hirschman (1982) in economics, a number of scholars in organisational psychology and industrial democracy have reinforced the view that processes, like other outcomes, represent something from which individuals can receive fulfilment (Guthrie 2001;Spreitzer 1995). Within organisations, deliberative processes and employee participation in decision-making, in particular, have been shown to be a prerequisite for the development of high quality communication, information sharing and trust inside organizations (Ostrom 1990;Deci and Ryan 1990), thus contributing to reinforce workers perception of meaning, competence, self-determination and impact (Messersmith et al. 2011;Spreitzer 1995). It follows that individual wellbeing is not the exclusive result of attaining a preferred outcome, but derives also from elements of the psychological contract between the individual and its organisation, or from the enjoyment attached to the experience. For example, suppose agent A's actions at work are strictly directed and monitored by her line manager. Despite the high wage, in absolute and relative terms, she is dissatisfied. Contrary to the work practices currently in use at A's workplace, she greatly values autonomous thinking and critical engagement with colleagues before decisions of interest are taken. A's current work context and practices are therefore in contrast with her self-fulfilment.
From this example we can also appreciate the relation between process-outcomes and culminating outcomes. Besides being dissatisfied, A's critical thinking is frustrated by excessive direction and control. As a consequence, new ideas are scarce and problem solving is not effective (cf. Ostroff (1992) for an account of the relation between involvement, satisfaction and performance). The processes chosen by the firm are, as a result, highly incompatible with innovation, thus lowering the quality of services (and the wellbeing) offered to users. Workers and users' interests, in this example, are disregarded by the firm's organisational processes. This leads us to a further point, for which preferences on processes do have, like preferences on culminating outcomes, a social or public dimension, which implies that procedural preferences too regard others.

Enquiry and Deliberation
Besides the impacts of choices on society at large, Dewey talks about the existence of a plurality of 'publics' rather than 'the public' as a monolithic entity (Long 1990;Branston et al. 2006). In this sense specific publics are generated by each action, and each agent is part of one or more publics (Dewey 1927). An appreciation of outcomes, for Dewey, comes from the discovery of such complex interactions. One important element of knowing about the larger set of needs, views and implications of so-called private choice is that it strengthens assessment (Dewey 1917;Buchanan and Vanberg 1991). In economic terms, this means that new knowledge can affect what individuals believe and value, as well as their preferences (Witt 2003). The problem of beliefs, specifically, may be also understood by considering the limitations of inductive and deductive reasoning, for which inductive knowledge is subject to the limits of biased (or positional) observations (Popper 1959;Sen 1993), whilst deductive knowledge may suffer from the use of incorrect assumptions (Lakatos 1975). 5 The pragmatist approach builds on the desirability of enquiry, rather than (albeit surely not inconsistently) on the centrality of altruism and reciprocity (as in Fehr and Gächter 2000;Fehr and Schmidt 2001). Specifically, since the public dimension is only partially known prior to experience, enquiry-based thinking would represent a foundational element of all aspects of human experience, hence underpinning the constant assessment of needs, preferences, processes and outcomes. Decision-making processes, from this angle, would very much resemble the scientific construction of knowledge, which is by its very nature inter-subjective and evolutionary.
Facing the problem of knowledge coordination Hayek (1945), for example, argues in favour of the price mechanisms, whilst Dewey (1927), as mentioned, suggests the desirability of deliberative practices based on enquiry. The two mechanisms differ in the type of assumptions and outcomes. The market mechanism aggregates knowledge through price information, building on the existence of different but complementary interests of buyers and suppliers. Differently, deliberation contemplates a variety of perspectives and interests that may or may not be compatible or complementary. It requires, therefore, more complex rules of interaction, formal and informal. This approach differs from market coordination also in the way it accounts for social consequences stemming from socalled private action. The idea of positive and negative externalities deriving from private market choices recognises that prices account only for some of the effects of decisions, whilst spilling over on other agents who do not directly participate (Coase 1960). In this case the economic agent either ignores or does not care for impacts on others (unless different property rights or incentives are designed), and yet such consequences are part of the scenario. In the theory of externalities what are called selfish preferences can be described as situations in which the agent knows about the externality but decides to exclude such positive or negative effects from his/her decisions. Albeit selfish preferences can also generate positive externalities, when effects are negative this 'exclusive preference' causes most social dilemmas (Ostrom 1990). In the same way, partial knowledge and bounded rationality can prevent actors from recognising the externality problem. From a pragmatist perspective, however, something more fundamental than knowing about the externality is involved. In order to account for the public dimension (defined in terms of complex interactions with the context), processes and practices need to be designed and developed with the aim of enhancing learning, i.e. uncovering impacts, avoid undesired ones by cooperatively searching for possible solutions. This involves the definition of coordination mechanisms that are more complex than the market, such as processes centred on deliberation and shared decision-making amongst publics. It follows that decision-makers' preferences are expressed, in the first place, deontologically, i.e. in the definition of processes and praxis (rights and duties) from which wider social impacts derive. Moreover, deliberative processes originate social outcomes that are at least partially intended and governed, whilst the price mechanism generates externalities, which are not considered part of the objective function of the decision-maker (at last formally).
From a pragmatist perspective, the nature of preferences underpinning the choice of economic processes and relations can be assessed by looking at the extent to which these foster enquiry, cooperation and trust. Enquiry, as envisaged by Dewey, is in fact understood as a way of thinking that can eradicate partial understanding or false believes from our courses of action by considering each and every existing and future perspective critically. It underpins the experience of non-isolated individuals who are able and enabled to use their 'creative intelligence' to assess and change social institutions, as well as their own preferences and related outcomes (including processes). The argumentation goes as far as to indicate deliberation as the preferred coordination mechanism (Dewey 1927). Deliberative decision-making processes are defined as pluralistic, in the sense that the aims of participants may diverge, whilst still retaining a common will to find a deliberative shared solution to problems. To this end, deliberation supports open communication based on the quality of argument, on the explanation of meanings and experience (regardless of the medium used to express it, Young 2000) rather than on power or information asymmetries (Habermas 1984). In other words, the fact of having a particular aim in mind is not a sufficient reason for suggesting it to others, unless the agent finds a good reason or argument to support it, and for others to agree. Deliberation brings new knowledge in the decision process and this contributes to cast individual preferences. The shared process however is not seen as in contrast with autonomy. Likewise, the pluralism of deliberation improves agents' motivation to implement decisions, as well as agents' fulfilment in achieving results that are aligned with intended outcomes (Cohen 1989, p. 34). The efficacy of deliberative practices, in this sense, needs scrutiny, with the aim of assessing whether deliberative capacities exist, if diverse communication modalities are integrated and potentially conflicting interests accommodated, thus minimising failure to meet public needs and creating the conditions for individual fulfilment.

Inclusive and Exclusive Social Preferences
It follows that the first problem for the decision-maker becomes assessing the desirability of the process per se, its relation to the desired outcomes. Since enquiry asks for multiple perspectives to be equally considered, not only knowledge but also decision-making power needs sharing. Therefore, enquiry, as a way of thinking, favours the choice of inclusive processes, as for example those entailing shared deliberation. We call inclusive social preferences those that underpin the choice of inclusive processoutcomes.
If enquiry, as a way of thinking, can justify social preferences for inclusive processes, on the contrary the lack of enquiry conditions and attitudes (e.g. incentives that favour the exploitation of information and power asymmetries) further develops exclusive attitudes reinforcing the choice of exclusive process-outcomes. We call exclusive social preferences those that underpin the choice of exclusive process-outcomes. Social preferences for exclusive processes encompass the public dimension to the extent that they marginalise the interests of the publics affected (others) or the interests of society at large (the common good), therefore encumbering the needs and wellbeing of the excluded (culminating outcome).
Albeit inclusive preferences tend to adhere with inclusive process-outcomes and vice versa exclusive preferences with exclusive process-outcomes, processes and preferences are not equivalent. Inclusive preferences can be expressed, for example, in exclusive contexts. Likewise inclusive processes may host exclusive behaviours and fail to deliver culminating outcomes as envisaged (Ben-Ner and Ellman 2013). This may happen if actors express exclusive social preferences within an inclusive framework. Managers or workers may shrink due for example to inconsistent motivations, information asymmetries or lack of appropriate monitoring mechanisms (cf. Ben-Ner and Ellman 2013; Grimalda and Sacconi 2005 for experimental results). We have suggested that, because powered by enquiry, inclusive processes can reduce the distance between the culminating outcomes of the decision-making process and what publics deem as desirable, thus furthering fulfilment. Still there is no guarantee that this outcome will be achieved since it depends on a combination of contextual conditions and individual attitudes.
One outcome of engaging with the process is the refinement of social preferences. A movement towards inclusive preferences, for example, can be prompted by the failure of the conventional for-profit enterprise to respond to societal needs, as the experience of several co-operators and social enterprises shows (Borzaga et al. 2011). Alternatively, a failure of self-managed firms to deliver member benefits, paired by poor enquiry and recognition of the issues can reinforce a movement towards exclusive preferences amongst the members who may opt to exit the cooperative or transform it into a conventional firm. A more detailed discussion about the evolution of preferences is in the last part of the paper.

A Taxonomy of Production Organisations
Reflecting on the nature of free trade, Cowling and Sugden (1998a, p. 349) have referred to the impacts of exclusion from strategic decision-making processes as strategic failure or 'the failure of an economy's system or process of strategic decision-making to yield the most appropriate outcomes for the society served by that economy' due to strategic decisions in production being concentrated in corporate hierarchies and made by a restricted group of managers or stockholders. Differently from the type of intelligence that serves some interests even at the detriment of others, we have argued that true enquiry requires a cooperative discovery process, it entails knowledge creation but also shared access to knowledge and decisionmaking. Choices based on inclusive preferences reach beyond the individual actor's sphere not only because they impact on others and society at large, but also by means of processes based on engagement, shared decision-making and learning (the 'positive freedom' aspect) (Berlin 1958;Joas 1996;Offe 2011;Sacconi 2011).
Because of the observation of strategic failure across economies due to exclusive preferences and related choices in production organisation, we are raising a question on how production can move towards a more inclusive reality, so as to reduce failure to meet societal needs (Sacchetti and Sugden 2011). We have argued for the inherent public and social dimension of each and every preference expressed through individual decisions and have focused in particular on the need to establish conditions that encourage a habit of enquiry, pluralism and cooperation, as those are, tendentially, not pivotal in conventional production governance settings. We have considered preferences on aims, processoutcomes as temporary (because subject to enquiry) and comprehensive (because interrelated). The next step is to use these ideas to identify a framework which can support the assessment of the social preferences and choices expressed in production.
We consider two procedural aspects: the choice of governance form (as ownership, rights and duties) and the choice of decision-making processes (as the praxis of collective decision-making). The choice of governance and decision-making processes are the outcome of the decisionmaker's social preferences, within a particular institutional context. Culminating outcomes (firm's impacts) can be then associated with process-outcomes. Therefore, we suggest considering the choice of governance and of internal decision-making practices as a mediator between the decision-maker preferences and public outcomes. We use in particular social preferences regarding governance choices and other strategic decision-making practices 6 as an indication of the decision-makers pre-commitment towards enquiry.
Processes that reflect inclusive social preferences would be designed so that situations can be problematized, and not just regarding a restricted group's private concerns. The aim would be to define rights, duties and practices that allow the search and inclusion of the publics and their multiple perspectives, as well as considerations of the wider common good. Close to this ideal are, for example, organisations created with the core aim of providing welfare, cultural or environmental services through multistakeholder governance (Tortia 2010). Conversely, exclusive social preferences would not, as a norm, support the inclusion of other perspectives and interests in the process, rather than those of the decision-makers themselves. On this extreme we find for example traditional equity-based corporations with no or limited strategy towards stakeholder involvement.
If we bring together social preferences regarding formal governance structures with those about other decision-making practices, we obtain the following hypothetical combinations (Fig. 1).
The combination of social preferences regarding governance and those regarding strategy-making highlights situations of homogeneous processes, as in cells one and four. Cell's two and three present combinations of heterogeneous processes. In cell one, the initial inclusive social preferences supporting the choice of governance, exemplified, for example, by membership in self-managed firms, are consistently carried forward to include the strategies towards other publics, operating within (e.g. volunteers, salaried workers) and outside the organisation (e.g. suppliers or other actors in the civil society, such as users, costumers, the public administration, or other interested actors depending on the mission). Here are cooperatives that specifically produce an economic and a social surplus without following the profit-maximisation rule (Valentinov 2008). As Borzaga et al. (2011) emphasise, in cooperatives cost minimisation is not the one priority and, as long as the organisation is sustainable, the surplus takes also a social and psychological connotation. This is often the case for particular forms of social enterprises characterised by both mutualistic nature and multi-stakeholder governance. The crucial difference with socially responsible conventional business (cell three) is that conventional business fundamentally retains an exclusive governance structure centred on investor interests, even in the presence of corporate social responsibility. Differently alternative business forms such as self-managed firms with social aims have embedded, in principle, ideas of shared decision-making power and multi-stakeholder benefit in their aims and governance structure.
In cell four, we find quite the opposite, with a consistent persistence of exclusive social preferences, both in the initial choice of governance and in the strategic decisionmaking approach.
In cell two we find organisations, such as cooperative firms, which set up processes to include at least one major stakeholder, which sets its objectives in an exclusive way. This is typically not the investor but the weakest stakeholder, i.e. the stakeholder that would incur the greatest loss if it were not the owner of the organisation. Inclusive preferences, however, do not extend beyond membership. Albeit founded on democratic governance principles, these cooperatives are mainly accountable to their members and do not implement particular practices for the inclusion of other types of interests. An exclusive focus on membership would be consistent with the neo-institutionalist analysis of cooperatives, which grounds the emergence of cooperative governance in the need for particular publics (such as workers, consumers, users, producers) to minimise transaction costs when market failure is present (Hansmann 2000).
In cell three we may find conventional investor-owned firms engaging in genuine strategies for the search and inclusion of publics. We can position here also social enterprises and non-profit organisations in general (such as private foundations) with a board of managers that is strongly driven by the initial social mission, but with no membership.

Evolution
Our taxonomy depicts four representative situations. It is a static picture of ideal-typical features of organisations at any given time. But how do firms move from one cell to the other? Or, what elements can be expected to lead to the development of more or less inclusive patterns of behaviour amongst decision-makers? The contradiction that we are left to explain is why, despite the fact that inclusive social dispositions improve understanding and validity of choices, the reality of production organisation is widely characterised by exclusion. There must be, then, a cumulative cycle which perpetuates one type of approach. The problem does not lay perhaps in the absence of subjective dispositions towards enquiry and inclusion, but in the institutional bias which does not favour the expression of such preferences and enquiry-based processes more generally. Evolutionary theory has profusely made the point that institutional and organisational diversity and trajectories can be explained as path-dependence from specific historical accidents and choices (David 1985;Arthur 1994;North 1990). The relevance of contextual influences has been pointed out also by behavioural theory, which stresses that, at the macro level, the formation of socially inclusive habits is not immediate but associated with complex courses of value transmission through socialisation (Ben-Ner and Putterman 1998). Evolutionary economists, in parallel, have offered a number of perspectives on the persistence of established patterns. Witt (2003), in particular, has argued that limited knowledge and, consequently, bounded rationality applies to preference formation in a world where incomplete perspectives impact on what agents value and aim at. This argument which, for us, has a clear Deweyan flavour, reinforces the frictions surrounding change on the one hand and the relevance of deliberation on the other. Deliberative practices can support learning and a change in beliefs, following which individuals will move attention to new sets of values and related meansends [whilst still leaving other desires and behavioural frames 'ignored or neglected', (Witt 2003, p. 80)].
For example, strengthened by their resilience to the cyclicality of economies and to complexity, the self-managed organisations initiated by innovators inspired, to some extent, wider social recognition and diffusion by means of imitation. Still, these represent minority solutions to the production governance problem. Witt (2003) suggests that widespread changes across economies would require communication across the community to attract agents' attention. In particular, he argues that a crucial condition is that a 'critical mass' of communicating agents and groups is reached, so that the new set of values can spread across the community. Communication and agreement on a novel set of values (as embedded in a new norm for example) will cast mutual expectations on behaviour and stimulate agents' conformity with the new set of values (Grimalda and Sacconi 2005;Sacconi and Faillo 2010). Also, some degree of proximity in the perception of values, ends and means amongst networks of decision-makers is relevant to start the deliberative process (Sacchetti and Sugden 2009). The latter can be facilitated by policy action and agenda setting (Witt 2003). These conditions are important to break path-dependence and institutional inertia and help overcoming situations that can reinforce false believes, limiting or slowing down the opportunities for change, even when more socially or economically efficient alternatives are available or when individuals show different social preferences vis à vis those embedded in existing governance structures (cf. Mahoney 2000). Other accounts reinforce that preference change can be affected by the ability of specific publics to articulate their perspectives on reality (Dewey 1927), from the costs of participation, community size and actors' distance (Dixit 2009), or from disappointment and fear accumulated from prior interactions (Hirschman 1982;Meier andDurrer 1992 cited in Witt 2003).
Consistently with macro-approaches to change within organisations motivational theory explains that individuals can gradually internalise contextual interests, values and rules. These internalised rules of behaviour concur to the formation of agents' sense of who they are, so that their behaviour is sensed as autonomous and self-determined (Gagné and Deci 2005 p. 335). Thus, driven by social institutions and established ways of organising production, social preferences towards process-outcomes become central to the agent's identity. He or she would be likely to act in ways that are more or less consistent with enquiry and with including or excluding others more generally. Depending on these contextual conditions, the individual actor could, to different extents, come to appreciate the importance of multiple interests and of investing in deliberative decision-making processes. 7 More specifically, economic theory has explained preference change in the organisation as reactions to the nature of rewards and punishments, for example in the form of financial incentives (cf. Bowles and Polanía Reyes 2009 for a review). Incentives and processes in particular signal what the incentive provider values in terms of behaviours and outcomes (ibid.). For example, Frey (1997) argues that monitoring counteracts individual autonomy and selfdetermination, with the result of lowering individuals' trust and virtuosity (Ben-Ner and Ellman 2012). Ben-Ner and Ellman (2012), however, argue that preference change happens gradually, rather than as instantaneous feedback to processes, 'mediated through aversive interactions with work colleagues and bosses' (ibid., p. 405). In particular, the authors explain durable preference change with the emotions triggered by perceived inequity in the workplace. Placed in a context where selfish behaviour is rewarded, the altruist who experiences frustration can then decide to leave or to conform and stay. In the latter case, individual values and critical enquiry abilities are durably compressed and conformity increased.

Policy Action and Conformity with Inclusive Social Preferences
To illustrate some of our points, consider policy incentives towards alternative governance forms. These are provided through legal frameworks, local development policies, and more generally by means of deliberative skills through education and training [which, especially in business education, is also subject to the strategic choices of higher education organisations (Sugden 2013)]. 8 Awareness of alternatives and critical mass, markedly in the presence of habits of thought and organisational inertia, can be fostered by policies that channel the commitment of production organisations towards particular sets of strategies. Here commitment entails the possibility of stringent adherence to collectively defined rules, which may require contractual solutions or radical governance changes (Sacconi 2011). 9 This approach to the emergence of preferences for particular choices is supported by Grimalda and Sacconi's theory of preference formation for which preferences result from the joint consideration of different descriptions of states of affair (Grimalda and Sacconi 2005). If the description regards culminating consequences, preferences that are inclusive or exclusive of the interests of others are relevant. Differently, when descriptions are presented in deontological terms, as situations where preferences show whether the agent is capable of choices that are consistent with particular shared principles, then psychological preferences for conformity are relevant (Grimalda and Sacconi 2005;Sacconi and Faillo 2010). The way reality is described or presented (e.g. through policy, or through a 'social contract') does matter in triggering preferences, and in particular in prompting reciprocity in complying with principles, forming, through social interaction, beliefs resulting in degrees of mutually expected conformance.
As an exemplification of gradual movements from cell four towards cell three by means of pre-commitment to a set of principles, consider the recent introduction, in the UK, of the community benefit clause in public procurement (CBC). CBCs essentially require contractors to deliver social value added to communities. These clauses are generally meant to maximise local social welfare generated by public demand, for example for infrastructures or specific services (such as employability services). Specifically, policies at local and regional level in the UK have identified the production of value added with respect to employment, training and urban regeneration. The criteria set by public administrations aim at delivering wider social benefits than those associated exclusively with the provision of a particular good or service. For example, CBCs may require, directly or indirectly, contracting out activities to social enterprises. In this way, the conventional business firm commits (at least within the remit of the procurement contract) to the implementation of some inclusive social strategies, clearly encouraged by the institutional framework defined by CBCs. Moreover, when a conventional for-profit company commits to the production of community benefits, stakeholders' expectations towards the inclusion of wider public interests may change beyond the remit of the initial commitment contract and become a permanent feature of the aims and processes of organisations . Following renewed stakeholder expectations and learning generated through engagement with social enterprises, organisations may further adjust their governance and/or strategies, conforming to shared expectations (Grimalda and Sacconi 2005). Specifically the evolution of social preferences towards inclusion would occur when conventional businesses and social enterprises enter a reciprocal learning process which may prompt a change of strategic aims and related processes beyond and consistently with the remit of CBCs.

Institutions Matter: Historical Circumstances and Prevailing Values
Historical trajectories stemming from past choices have been argued to influence the evolution of institutions and socio-economic organisations and explain institutional 8 Academia has also been argued to have a specific policy role in selecting and weighting beliefs. Within economics and business, in particular, the discipline has historically exerted strong influence on economic policies as well as in shaping the nature of businesses and their strategies (Currie et al. 2010;Fleckenstein 1997). 9 The perspective is different from stakeholder theory, where the inclusion of stakeholder interests is typically presented in the context of win-win situations that emerge spontaneously and despite a conventional governance structure.
Inclusive and Exclusive Social Preferences 481 variety across regions and countries (North 1990;Salamon and Anheier 1998). To illustrate, consider that albeit representing in general the expression of a niche business culture, since the end of WWII, 10 in the southern part of Europe cooperation has considerably grown, whilst in England their presence remained weak. Countries like Spain and Italy, for example, have had a long-standing and stable tradition in self-management. 11 In Italy, as Zamagni (2006) observes, ideas of human dignity, fairness and solidarity across a variety of political orientations (liberal, socialist and catholic) ensured support across local administrations and contributed (together with other elements, such as the solidarity and ties amongst cooperatives formally coordinated through federations and consortia) to the continuity over time of the cooperative business form. This process was supported by the recognition in the 1948 Italian Constitution of the role of cooperatives, 12 and in the implementation of this principle through consistent legislation and fiscal incentives (Zamagni 2006). Traditionally operating in agriculture and credit, cooperatives extended also to the provision of social services, finding a suitable terrain in those Southern European Countries that where lacking sufficient provision (Borzaga 2004;Borzaga et al. 2011;Defourny and Nyssens 2010). The establishment of social cooperative enterprises is an interesting case, since it builds on previous critical mass of cooperative values embedded in the existing framework that defined cooperative firms and on other complementary institutional arrangements, including the existence of a cooperative credit sector. Still, this organisational typology required some degree of institutional innovation. The first specific law on social enterprises appeared in Italy in 1991, 13 together with a supporting regional and national system composed of intermediate associations, academic research and education, professional training and data collection on social cooperative enterprises. 14 The emergence of a network of supporting institutions and initiatives suggest, therefore, that the development and diffusion of the values of social enterprises, their aims and form of governance, coexisting with traditional ones, have been a viable but demanding challenge, which required individual initiative together with a co-evolution of habits, legal framework, production structure and supportive complementary institutions (Nelson 1994;Amable 2000;Boyer 2005).
The family nature of local capitalism and the prevalence of small and medium enterprises have also been argued to have left more space for the development of alternative business forms if compared to systems dominated by large equity-based corporations, such as the UK (Zamagni 2006;Everett and Minkler 1993). Here, in the late 1970s, a neoliberal approach to policy and economic choices, paired by the economic weaknesses, strategic mistakes, and member opportunism in collectively-managed organisations brought to the privatisations and demutualisation of most of the existing building societies as well as of other mutuals (Birchall 2001). 15 Mismanagement at firm level and demutualisation policy clearly illustrate movements from the expression, at least in principle, of inclusive social preferences towards the membership (as in cells 1 and 2), to conventional business forms (as in cells 3 and 4). More generally, mismanagement signals a weakening of inclusive social preferences in managers (assuming that such preferences existed) and their displacement by exclusive preferences, alongside the effects on members' motivation and preferences, as suggested by Ben-Ner and Ellman (2012).

Conclusions and Implications
The persistence of strategic failure, i.e. the misalignment between preferences reflecting the private interests of 10 In Italy, after 1924, during fascism, and until the end of the war, all civic and economic associations had been forbidden by law, thus putting a halt to the diffusion of cooperatives. 11 Over the last 30 years, in Italy cooperation entered a clear growing pattern. In 2001 cooperative firms represented 1.2 % of firms counting for about 6 % of the total employment (ISTAT 2008). Using national census data Zamagni (2006) observes that during 1990-2000 the overall occupation grew by 60.1 % within cooperatives, contributing to one-fourth of the overall occupational growth for the decade. 12 Article 45 states: ''The Republic recognises the social function of co-operation of a mutualistic, non-speculative nature. The law promotes and encourages co-operation through appropriate means and ensures its character and purposes through adequate controls…''. 13 In the Trentino region, where cooperation has a long-standing tradition, national legislation was anticipated by a regional law in 1989. 14 In 1994, Issan, an international research and policy network on cooperative and social enterprises later named Euricse, was created in Footnote 14 continued collaboration with the cooperatives federation, the representative association for commerce and tourism and the Faculty of Economics at the University of Trento, in the Trentino Region. Membership was later extended to ensure the development of the initiative and gain international visibility. The institutional recognition of cooperative models was strengthened further in 1997 when the Third Sector National Forum was officially instituted at the national level and recognised by the government as representative of the sector's interests, and in 1999 when sectorial data started to be collected in periodic census by the national statistical institute, ISTAT. 15 Differently from the UK, in some countries demutualisation is not an option. If it were, as the UK case shows, opportunistic behaviours of members or managers would be incentivised. In fact, because cooperatives accumulate indivisible reserves over time, selling an established cooperative permits members to appropriate all the value accumulated by previous members, placing the continuity of cooperative firms in jeopardy (Tortia 2007). restricted groups and those of publics and society at large, has provided socio-economic relevance to our study. We have argued that a reduction of strategic failure can occur by reinstating enquiry and inclusion in choices about production governance and processes. This would require cooperation rather than mere coordination, as in terms of shared access to decision-making and use of deliberation. These dispositions, at least in principle, are likely to be present in the governance settings of self-managed organisations (but also, to a more limited extent, in conventional firms, albeit confined to specific responsible practices). More generally the rules that define governance and strategic decision-making processes express the decisionmakers' pre-commitment towards enquiry. Some arrangements will acknowledge a variety of interests and perspectives into the initial constitutional process, such as organisations with multi-stakeholder governance; others will focus on some interests in particular. Our taxonomy identifies and classifies production organisations with respect to their potential to generate strategic failure or, in other words, by the degree of exclusion of publics and social good from comprehensive outcomes.
The role of policy and regulation, therefore, is not understood as constraining, but in fact as enabling particular types of behaviours and impacts which would otherwise be marginalised because of prevailing interests. Policy may not and probably should not, try to change habits directly. Rather it may change them indirectly 'by modifying conditions, by an intelligent selecting and weighting of the objects which engage attention and which influence the fulfilment of desires' (Dewey 1922, p. 26). In these respects, the challenge for the decision-makers at firm and policy level appears to be one of endowing individuals and their organisations with a variety of tools, including those that promote engagement in deliberative cooperative processes. Through deliberation, the knowledge of contexts, courses of action, and their effects is improved and used to critically assess production aims and means. In line with Offe (2011), Sacconi (2011, Cowling and Sugden (1998b) and Sacchetti and Sugden (2011) this requires that decision-making power is shared across social actors, deliberately acting to give voice to multiple publics and reduce strategic failure.
As part of its potential, the social preferences framework can be used to assess the degree of inclusion of publics and wider social values into economic decisions. It may represent also a viable explanatory model to assess the consequences of policy action, in terms of its potential in generating movement and variations across strategic choice categories. Moreover, the inclusive/exclusive nature of preferences reflected in production systems can be related to other socio-economic development measures to test which production systems are associated with higher levels of individual wellbeing (Erdal 2011;Wilkinson and Pickett 2010). Beyond private firms, the framework could be also applied to governmental organisations. Decision-makers in the public government arena develop different views of the world and adopt, not less than others, diverse behaviours with respect to the inclusion of publics and social interests. For example, the framework can help clarify aspects of social preferences as reflected in the analysis of the aims and outcomes of industrial policy (Cowling and Tomlinson 2011;Chang 1997), social policy, in the processes characterising regulatory arenas (Hancher and Moran 1989), and more generally in practices of problem solving in public policy dilemmas (Avio 2002).