Labour Constitutions and Market Logics

Abstract

The article evaluates labour law’s strategies of coping with the pressure put on its project of realizing justice by a hegemony of economic perspectives on labour markets. Its consequences for a methodology of labour law are set out by critically engaging with recent proposals made by Simon Deakin and Ruth Dukes. It is argued that a socio-historical perspective on the role of legal models in actually shaping labour relations can enrich the concept of a ‘labour constitution’.

Labour Law and the Hegemony of Economics

The increased influence of economic perspectives in the political discourse regarding labour relations has put pressure on labour law’s project of realizing justice. Labour law as a specific legal domain was coined in 1911 by Hugo Sinzheimer who argued for opening up state law to the issues of organizing labour relations, leaving a large part of the regulation to the organized parties of industry themselves. In this way, Sinzheimer tried to protect democracy from the potentially disruptive effects of industrial conflicts (Hepple, 1986: 6–9). Gradually political elites in Western Europe came to be convinced of the potential contribution of labour law to the legitimacy of nation states. In the aftermaths of both World Wars, significant steps were made towards legislation that codified a ‘legal order of labour’.

Since the 1980s, however, economic theories that emphasize ‘market’ aspects of labour relations have been embraced by politics, partly at the expense of the principles of this legal order. Politicians have welcomed economics’ claim to allow for universal, ostensibly ‘value-free’, solutions to controversial matters. The growing influence of economics has urged labour law scholars to redefine the position of their discipline in relation to this body of knowledge that was increasingly ousting them from policy-making processes.

Facing this growing hegemony of economics, labour law scholars have felt the need to develop coping strategies. In this article, I comment on these strategies in light of the question, which new methodologies would be needed to allow labour law to keep performing its duty of securing justice in labour relations. Part 2 recapitulates the problems involved in imposing a market concept on labour relations. Labour lawyers, trying to adjust their perspective to that of economics, should take due account of criticisms made of it. In the third part, I take a methodology recently proposed by Simon Deakin as an example of a ‘reformist’ strategy, and I use my criticism of it to unfold a broader perspective on the role that law plays in ‘economic’ interactions. I argue that law is ‘constitutive’ of labour relations in a much broader sense than is often assumed. In the fourth part, I argue on the basis of a foray into the history of labour relations that the concept of the ‘labour constitution’ can usefully be applied to premodern relations. New methodologies, I conclude, ought to recognize the changeability of practices and of the legal concepts that partly constitute them.

Labour Law’s Strategies

Attempts by labour law scholars to address the current challenge to labour law posed by dominant neo-liberal theories and policies testify to three main strategies. First, on the ‘conservative’ side, some argue that labour law’s context has changed, but that its values and goals, in particular that of compensating for the inequality of bargaining power in the employment relation, stand as firmly as before, so that there is no need to ‘re-invent’ labour law. According to Manfred Weiss, labour law is grounded upon four insights, already ‘brilliantly analysed by Hugo Sinzheimer’: ‘labour is not a commodity’; the basic problem of the employment relation is personal dependency; this dependency may threaten human dignity that therefore needs protection; and, finally, this protection ought to cover all the risks in an employee’s life. Changes in the conditions of production may require some adaptations to new circumstances, but there is no reason to revise the paradigm of labour law, defined in terms of a professional duty, or vocation, to protect workers (Weiss, 2011: 44, 56).

The ‘reformist’ strategy, second, is to make labour law adopt labour market policy goals and adjust its perspective. ‘Law of the labour market’ scholars, as Ruth Dukes calls them, started redefining labour law in terms of its contribution to the functioning of labour markets, moving it closer to declared policy objectives. In contradiction of neo-liberal economics’ claims, the argument is made that ‘regulation complements, rather than obstructs, the institutions of private law in providing a framework for exchange relations’ (Deakin, 2009: 29; Dukes, 2014: 9). The state should intervene to ensure the preconditions for a well-functioning market (Deakin and Wilkinson, 2004: 283–4). This may include promotion of particular social rights (to healthcare, education, equal treatment) (Dukes, 2014: 113). Protective rules are characterized as a means of combating social exclusion. Even collective labour rights are seen to improve the use of productive resources, by creating a balance of power in the workplace (Deakin and Wilkinson, 2004: 348–9; Dukes, 2014: 114–5). What marks the change in methodology here is, according to Dukes ‘the acceptance of the objectives of social inclusion and competitiveness, and the consequent adoption of market functioning or market efficiency as the rationalizing or organizing principle of the field’ (Dukes, 2014: 109).

A third strategy argues for the constitutive role of labour law in labour relations. Dukes has criticized the ‘reformists’ for the seemingly unpolitical character of their move towards functionalizing labour law. An economics discourse conveys the suggestion that all of us would benefit from an increased ‘efficiency’ of labour markets. However, Sinzheimer already noticed that in a liberalized private sphere property reveals itself as a basis of domination. He therefore urged government to bring production under democratic control. To Dukes this is not a worn-out goal; what is needed is to ‘bring politics back in’, to revive the constitutive role of labour law. In the 1920s, Sinzheimer could still put his trust in the nation state to act as guarantor of such a legal regime ¹ , but whom could we trust to act as such today?

Market and Labour: A Problematic Nexus

The ‘reformist’ strategy tends to follow economics both in naturalizing markets and in accepting that labour is a commodity that can be traded in ‘labour markets’. Both propositions are highly problematic.

‘In the beginning, there were markets’ – that was how Oliver Williamson formulated the revelation of economics, with a wink to Genesis (Williamson, 1983: 20). Giving priority to markets by endowing them with a ‘natural’ status was an 18th-century development. Naturalizing markets was an important way for economic theories to try to distance themselves from substantive matters of value and politics. ² Up until the 18th century, however, the fairness of transactions was an integral element of the market constitution. The conditions of market interactions, as defined by the market constitution, were considered to provide the basis of fair transactions, and the market was a locus of véridiction (truth-telling) (Foucault, 2008: 32, 36–7). ‘Fairness’ remained for a long time a separate criterion by which the validity of transactions could be legally rejected. Only the liberal strategy of separating a ‘private’ economic sphere, and the belief that a ‘natural’ equilibrium would result from freely transacting in this space, put an end to this constitutional tradition.

This strategy may have its advantages, but it is not harmless. Economics is hailed by some for its level of abstraction and its eminent universalizing potential and criticized by others for its ‘self-inflicted “veil of ignorance”’ and its colonization of social life (Zafirovski and Levine, 1999: 312). Douglas and Ney (1998: 74) have characterized its presumption of homo oeconomicus as a feedback system: a certain institutional form, adopted and favoured, generates its own version of practicable knowledge and this knowledge then generates the proofs and legitimations that protect it and, with their users, a vested interest in defending the current form and a suspicion of alternatives. In this sense, an economic perspective partly creates its own realities – it is even said to take ‘at least some (not all) of the classic functions of religious symbolisms as devices to cope with existential uncertainties’ (Deutschmann, 2015: 9). Rather than being value-neutral, the merit of economics would be to render evaluations invisible (Bourdieu, 2005; Latour, 2014: chapter 16). Such accounts of the character of economic knowledge ought to be reason for labour law scholars to be critical of the current hegemony of economics and to reject ways of presenting ‘market relations’ as something ‘natural’ that would, by that quality, take precedence over law.

The ‘spontaneity’ that economic theory tends to attribute to market functioning is hard to reconcile with historical accounts of the development of market trading. From a historical perspective, there is much more evidence for the political constitution of markets built upon legal devices (Fligstein, 2001). Historians have reconstructed how medieval rulers, interested in furthering trade, instituted and facilitated local markets. Markets were conceived of by participants as human devices for realizing fairness, long before they were ‘naturalized’ by economics in the 18th century (Howell, 2010: 276–8).

The second problematic proposition, which treats labour power as a commodity, can be said to have been prepared and facilitated by Enlightenment legal theory, in particular by the legal construction of ‘hiring oneself out’ as a transfer of a ‘particle of freedom’ treated as property (Pufendorf, Grotius). To Enlightenment theorists, an appeal to ‘contract’ in particular had the negative tenor of distancing oneself from hierarchical models of labour relations (corvée, guilds). In a next step, the formal autonomy and equality that was accorded to citizens had however to be reconciled with the subordinate position in which some of them had to be kept in labour relations. The articulation of a discourse of citizenship with that of the individual ‘ownership’ of labour power, including the ‘freedom’ to legally transfer it by contract to another for a limited period, turned out to be the ‘trick’ that was believed to solve the problem. ³ However, to treat labour power as something that one could own and might trade without alienating oneself, retained an unpleasant whiff of fictitiousness.

And what of the market that this ‘fictitious commodity’ (Polanyi) should be traded upon? Definitions of ‘market’ notably do not abound; a very general one would be: ‘a social structure for the exchange of rights in which offers are evaluated and priced, and compete with one another’ (Aspers, 2011: 4). Callon (2007: 140–4) regards the following aspects of markets to be essential: (a) an allocative decentralization of decision-making, (b) in the hands of actors defined as calculating agents, (c) who solve their conflicts of interest in transactions by way of equivalences measured in prices. Mainstream economists tend to use functional criteria, based on what markets are supposed to do. But what are the phenomena to which this functioning is being ascribed? To identify them additional criteria are required, two of which are particularly relevant here: the voluntarity of exchange (or the possibility of ‘exit’) and its specificity: ‘a substantive specification of both parts of the transaction’ (Rosenbaum, 2000: 468–9). Applied to labour contracts, each of these criteria is problematic. To ‘sellers’ of labour power, ‘the opportunity costs of not accepting an exchange offer’ as well as those of ‘exit’ may be too high. And contracts are ‘incomplete’ due to the impossibility of specifying ex ante what workers will have to do. This incompleteness is compensated by assigning the employer a position of authority that is actually irreconcilable with formal contractual equality.

Apart from these deficiencies of labour contracting as a market phenomenon, labour has been qualified as a ‘fictitious commodity’, at least since Polanyi (1944): labour is not ‘produced’, as other commodities are, to be traded in the market; it cannot be kept in stock and is, if unused, immediately wasted; it cannot be separated from its owner or, positively formulated: it is inseparably linked to the life of the worker (Polanyi, 1944; Sinzheimer, 1927).

Callon (2007) analyses markets and politics in terms of co-construction and considers disentanglements and framings (as to goods, agencies, price-setting mechanisms) to be necessary conditions for the functioning of markets – conditions that are usually only tacitly presupposed by economic theory. ‘Market framing constitutes powerful mechanisms of exclusion’. For instance: in buying a shirt, I can be assured that its Bengali producers will not sue me for damages suffered as a consequence of their working conditions. Each economic act or transaction occurs against, incorporates, and also re-performs a particular complex of institutional arrangements, rules and connections.

Framing implies the drawing of boundaries that are typically being legalized and it is legal devices that become a constitutive element of social interactions that are qualified as economic transactions (Knegt, 2008). For instance, the way in which ‘private’ and ‘public’ have been separated since the 18th century, and the associated allocation of powers to citizens, workers and capital owners have been constitutive of relations at the workplace.

Looked at from the ‘flipside’ of framing, we might say that markets constitute institutional devices for creating equivalences, between ‘commodities’ as well as, in certain respects, between traders. Metaphorically one could say that law constitutes the ‘Operating System’ on the basis of which the application programme ‘market transacting’ can be run. Coase (1992: 717–8) has criticized his fellow-economists for assuming that what is being traded on markets is ‘goods’: it is not goods that are traded but rights to perform certain actions upon them. Beckert (2009: 246) emphasizes that markets are not ‘natural’ or ‘spontaneous’ but are ‘highly presuppositional arenas of social interaction’. And as they have distributive consequences, market devices are themselves ‘essentially contested’. Even motivations for economic action are partly themselves the result of institutional variations (Zafirovski and Levine, 1999: 323–4).

A Strategy of Broadening the ‘Market’ Concept?

Might a possible ‘reformist’ strategy for labour law lie with interpreting ‘markets’ less restrictively than orthodox economists tend to do? ⁴ If ‘efficient functioning’ could be argued to encompass the sustainability of labour markets (i.e. not generating conditions that undermine their own existence), new requirements would announce themselves, partly to be met by labour law regulations. In addition to the regulatory requirements proposed by Deakin and Wilkinson (2004), ‘sustainability’ as an endogenous requirement of market functioning could be argued to require rules against the exhaustion of resources: pertaining to health and safety, training, income compensation in case of sickness, and facilities regarding reproductive activities such as the raising of children. In that case, rules of labour law would, to some extent, not need to be framed any more in terms of an external ethics of human relations. They could be argued to flow from internal requirements of the labour market itself. This strategy could potentially have the advantage of putting an end to worn-out (‘paternalistic’) legitimations of labour law in terms of the ‘protection’ of workers. However, not all of labour law can be reconstructed and re-legitimized in this way. Tucker (2010: 131–6) has criticized Deakin and Wilkinson for being overly optimistic about the happy coincidence of social rights and market-supportive functions, ignoring the unchanged character of conflictual relations in a capitalist economy. It would be impossible to explain the existence of, for instance, disability arrangements from a market logic. The approach set out in this section could thus only be valuable as a limited strategy of thinking along with economics in order to find ways to penetrate into its relatively closed bulwark.

There are thus several reasons for labour law to be careful in resisting the adoption of economic perspectives on labour relations. Labour law’s recent shift to economic methods of analysis is insufficiently grounded, Dukes has argued: ‘the market is as much a social construction as it is an economic or legal one’. Already by uncritically adopting the current concept of ‘market’, we run the risk of implicitly narrowing our view to an economic perspective on human interactions.

Is Labour Law Only External Regulation of Labour Markets?

Deakin (2014) has recently set out the contours of a research agenda to be developed from a new concept of the relations between markets and legal systems. He sets himself the task of reflecting upon the question ‘how the notion of the market is constructed through legal discourse’. Law is regarded as a reflexive realm where a conceptualization of interactions takes place. Starting from a systems–theoretical perspective, he considers legal concepts as ‘terms used by the legal system to describe [social] institutions’, as incomplete expressions of social practice. These expressions can be used ‘to disclose understandings of the nature, scope and limits of labour market relations’ (Deakin, 2014: 141, 149). He criticizes traditions of legal thought that claim legal terms to be describing a reality that abstracts from social practice; legal terms are descriptive ‘approximations of phenomena which exist in a realm of interactions which take place, by definition, beyond the law’ (Deakin, 2014: 141–2). ‘[L]egal scholarship should be actively contributing to the formulation of a better (in the sense of more empirically grounded) set of models of the market’ in order to inform policy choices. Even if legal models may tell us little about ‘external reality’, the way that they change over time – essentially adjusting themselves functionally in a process of selection to what reaches them from this external world – may be revealing (Deakin, 2014: 142–3, 149).

Law, conceived as a system, is functionally related to social practices and primarily adjusts itself to changes in these practices. ‘The rules of labour law operate in a functional relationship to the labour market phenomena they purport to regulate’ (Deakin, 2014: 149). The labour market does have a normative foundation, but this is not to be identified with law itself. The normative foundation consists of ‘shared understandings of what the employment relationship entails in terms of reciprocal obligations, which are articulated at the level of the legal system’ (Deakin, 2014: 151). Deakin distances himself from mainstream economics’ treatment of labour law rules merely as external interference with a supposedly spontaneous private ordering. Such rules are rather to be seen as endogenous responses to market failures that compensate for the absence, in practice, of the general self-equilibrating mechanism presumed in economics. The labour market ‘cannot effectively operate in the absence of the “fairness norms” and values of solidarity to which labour law gives expression’. We should ‘understand labour law as a mode of market governance which provides the basis for sustainable economic development’ (Deakin, 2014: 155, 159).

Deakin’s proposal may be welcomed in that it attempts to reverse the opposition of normativity to a quasi-natural market that liberalism has operated. But does his proposal offer adequate tools to develop a more comprehensive theory of the relation between law and practises of labour relations? I would argue that in this respect it is defective in several ways, and the more so as we try to grasp relations that are not part of modern market economies. Four interrelated criticisms are worked out below: (1) the proposal builds upon a conception of ‘(labour) law’ that is too restrictive; it defines the relation between law and social practices in a way that underplays (2) the socially contested as well as (3) the constitutive character of legal concepts; and (4) it underestimates the role that law plays in daily practices. These criticisms allow me to set out, by contrast, a different methodological strategy.

Is Law Identical with ‘Legal Systems’ at State Level?

Deakin’s concept of ‘law’ is restricted to a codified system of state regulation and case law of the judiciary, at least at state level. The ‘widely-held beliefs concerning limits to self-interest’ that constitute the ‘normative base’ of labour markets are treated as ‘conventional understandings’ that frame (rational) behaviour but are denied legal character (Deakin, 2014: 154; Deakin et al., 2017: 188–9). In other word, Deakin et al. recognize the importance of custom, but state that it has to be distinguished from law that is ‘primarily constituted by the state’. Significantly Deakin et al. (2017: 190) base this statement merely on indirect argumentation; the question of the criteria of demarcation is not systematically posed.

It is unnecessary to go into lengthy conceptual debates in order to point out that this conception curtails the scope of research into the relation between law and work relations, in particular those preceding the modern era. The legal regime of premodern guilds, for instance, can be argued to meet common requirements of ‘law’ (a system of general rules, partly codified; institutionalized procedures for enforcement and conflict resolution; reflection upon the validity and consequences of rules; conscious adaptation of rules to changing circumstances) but would be denied such a qualification (no state law, no separate judiciary) in Deakin’s conception. While it is of course necessary to draw boundaries, a conception of law that easily excludes from its scope regimes that on good grounds believed themselves to be ‘legal’, and that unjustifiably favours a ‘modern sense’ of law that only became dominant in the 19th century, cannot be a fruitful methodological starting point for a more comprehensive treatment of the relation between law and work relations.

The same problem is reiterated in the demarcation of labour law as ‘specific to work relations in modern market economies’. Labour law is actually coupled to the concept of ‘employment’ and its specific trade-off between subordination and protection (Deakin, 2014: 149–50), as if the relation between them would only have been a problem of modernity. If we start from a less restrictive concept of labour law, as comprising forms of regulation of working relations that included institutionalized enforcement according to contemporary conceptions of law, we may avoid excluding from our scope centuries of legal regulation of labour, for instance, as it has been operative in premodern European cities.

A ‘Functional’ Relation Between Law and Practices?

A second drawback of Deakin’s approach concerns the way the relation between law and social practices is defined. In essence, this relation is defined as ‘functional’, and its development as one of evolutionary adjustment (Deakin, 2014: 149, 158). Legal concepts have a capacity to effect real-world consequences, provided that they can be made sufficiently to ‘fit’ social relations, and in this way to contribute to the stabilization of these relations. What remains unclear, however, is how this process of ‘adjustment’ is to be conceptualized. Deakin’s account of the coevolution of law ‘and the wider economic and political structures in which those labour law regimes are embedded’ (Deakin, 2014: 142) suggests a process of incremental change. However, Brunkhorst (2014) has argued that in processes of social evolution legal concepts also figure in rapid or revolutionary change, generating normative constraints. Different from elements of evolutionary processes in the non-human biosphere, concepts are themselves products of social processes of knowledge acquisition as well as of political contestation. Therefore, a careful account would be required both of the potentially conflictual process of the selection and introduction of consequential legal concepts and of the methods of their stabilization. If, for instance, the employment contract has been an invention functional to the effectiveness of labour markets, how are we to account for the selection of this legal institution as well as for its consequent design, compared to alternative options?

Third, Deakin’s conception of legal concepts as ‘incomplete expressions’ of social practices (Deakin, 2014: 142) tends to underplay the role of law in constituting these practices. Legal models are more than incomplete descriptive representations of practices, they are models for practices. They are characterized by ‘a distinctive way of imagining the real’ (Geertz, 1983) that does relate to ‘real’ practices and can be studied in its practical consequences, but can neither be grasped completely in terms of ‘description’ nor subjected to an empirical test (vs. Deakin, 2014: 143). For medieval guilds, for instance, the model of ‘religious brotherhood’ had important consequences for working relations, largely irrespective of whether members actually acted in ‘brotherly’ ways or not. Brotherhood founded the legal–constitutional base upon which the workers in the guild can be said to have been ‘living their law’ (Brand, 1990).

Law as an Element of Practice

In Deakin’s approach, law as a ‘legal system’ is reflexively distanced from the ‘real life’ and from practices that are deemed to be inspired rather by customs and conventions. This is a conception of law that suits modern systems of ‘positive law’ much better than the law of premodern or non-Western social configurations and tends to blur our vision of the latter. There is an important stream of sociological research into the constitutive contribution of legal concepts to the day-to-day functioning of social practices. It has investigated the ways in which our ‘knowledgeability’ in social action is invested with legal notions and concepts, albeit partly in the guise of routines that are only dimly reminiscent of the legal innovations from which they stem (a.o. Giddens, 1984; Hunt, 1993; Sarat and Kearns, 1993; Silbey, 2005). For a more comprehensive view of the relation between law and work relations, the latter approach is more promising than a restrictive view that in fact takes the historically contingent figure of modern positive law as a benchmark for all legality.

Recognition of the implication of law in social practices requires a more complicated model of their relation that I can here only sketch briefly. The temporal dimension of social life urges participants to reflect upon and to try to stabilize social relations by drawing boundaries, practically and discursively allotting powers and responsibility for future actions to collective or individual participants. In this way they happen to create, ‘as the mediation of practical activities’, a normative dimension that is both part and parcel of the practices that comprise their relation and at the same time ‘a distinctive way of imagining the real’ (Geertz, 1983: 184). The ‘reflexive elaboration’ of the frames of meaning, connected with this normativity, ‘is characteristically imbalanced in relation to the possession of power’ (Giddens, 1976: 113). Boundaries and concepts may be contested, but, in the meantime, provide for a relatively fixed order that facilitates coordinated action. The framework provides for the accountability, not only to others but also to themselves, of the actions of those involved. To the extent that those who act ‘improperly’ face collectively recognized, institutionalized sanctions, this normative order may rightly be qualified as a ‘legal’ one. ⁵

It is not unusual for law to be undervalued in analyses of social practices. This is, first, due to law’s tendency, at the stage of social interactions, to hide behind the curtains as soon as normativity has been effectively translated into routine, only to reappear if practices on stage are being contested. ⁶ In this regard, my comparison with an operating system is certainly apt. Legal concepts are normalized and incorporated in daily routines in such a way that their normative – and therefore contestable – character goes unnoticed (Silbey, 2005: 330–2). Second, this tendency of law is also – unintendedly but unfortunately – obscured by methodological choices of sociology. The neo-positivist turn has curtailed sociology’s capacity to deal adequately with the normative dimension of social action, by restricting researchers methodologically to the study of observable behaviour. The resulting representation in terms of ‘expectations’ and of ‘social control’ is a two-dimensional representation of a three-dimensional social space. The methodological position that thus deserves to be argued for is that the role of ‘law’ in social relations can be adequately weighed only if we include the functioning of legal concepts within relations that, at first sight, might seem to be ‘free of law’ (Hunt, 1993: 225).

The ‘Constitution of Labour Law’ Strategy

This more sophisticated view of the role that law plays in social practices invites us to take a closer look at the third strategy that I distinguished above, the aspiration to ‘bring politics back in’ by reintroducing a constitutional mission of labour law.

The way Sinzheimer applied the notion of ‘labour constitution’ to labour relations, considered as the power relations of industrial production, was explicitly in analogy to what had been realized, by way of ‘constitutional law’, with respect to power relations in the ‘public’ sphere. In Sinzheimer’s view property in the ‘private’ sphere had, under capitalist modes of production, turned out to be the basis of the fierce, unrestricted domination of owners over expropriated workers. The legal conception of ‘employment’ as a transfer of property ended up creating relations of domination that were incompatible with the principles of freedom and equality that had been the basis of creating this ‘private’ sphere in the first place. This was a reason to introduce ‘public’ devices to counter these hegemonic effects, in a way comparable to what had already been done in the public sphere. In that way, a community would be created in which the structurally opposed interests of capital and labour would be pacified by a legally guaranteed, institutional structure of decision-making procedures, that would thus constitute all those involved as dignified, human, and to a certain extent substantively ‘free’ participants (Dukes, 2014: 212–3; Sinzheimer, 1927).

In Dukes’ updating of this approach, the labour constitution would not imply that relations should conform to any predefined particular pattern; but at the same time, its normative content is characterized as ‘the introduction of democratic principles and modes of action’ (Dukes, 2014: 196, 217). This is, to a large extent, how Sinzheimer and Kahn-Freund saw a solution to the problem of domination, but the question has to be raised whether these are timeless, or to some extent in particular early 20th-century remedies? Could it be fruitful to start by defining the problem at a more general level: as one of a fair, constructive, as well as efficient allocation of powers and responsibilities in a domain characterized by structurally different, and partly contradictory interests of participants? In order to come closer to an answer to these questions, I propose to broaden the perspective on labour constitutions by arguing that the legal basis of some types of premodern labour relations in Europe might also rightfully be characterized as ‘labour constitutions’.

A Historical Perspective on the ‘Labour Constitution’

Dukes’ perspective allows for a plurality of labour constitutions, though predominantly in what we now consider to be the typically ‘legal’ form of rules laid down in written documents (Dukes, 2014: 216). I will extend the scope of the concept by pointing out that what premodern contemporaries considered to be their ‘law’ (i.e. the constitution of their labour relations) need not have, and had often purposively not, been written down. I argue that this broader historical perspective reveals that in Western Europe a ‘constitutional’ approach to what we now call ‘labour relations’ has for almost a millennium been a rather self-evident element of daily life to many contemporaries. To substantiate this claim within the inevitable space constraints of this article, I briefly sketch just one premodern order of working relations that of German guilds before the middle of the 18th century.

‘May God protect the Empire’ was the greeting that accompanied the entrance of late-medieval wandering German journeymen amidst their peers into the regular inn of their trade (Brand, 1990: 110–1). It was an expression of their confidence in the Empire’s protecting the structure of guilds and journeymen’s associations across the different towns between which they travelled. To medieval and premodern associations related to work – both guilds and journeymen’s associations – the jurisdiction over its own matters was a crucial element of the constitutional order of the trade, as it was generally for those aspiring to power in those days. Both ‘honour’ and ‘freedom’ were less personal matters of individual craftsmen than qualifications that could only be achieved through membership in the association(s) of the trade. Membership in the trade was required for being allowed to practise it; therefore, exclusion was a much-dreaded sanction. The effectiveness of the trade’s jurisdiction rested upon the fact that towns were usually relatively closed political unities (in the continent, though less so in Britain), and that town governments adopted and thereby reinforced the guild’s rules, granting them effect towards outsiders. The members, united by their solemn oaths (‘coniuratio’) lived the law of the guild; it surrounded them with every step they took, it was orally transmitted as an element of the practices of their craft and mobilized to ground the claim to their status within the city. These associations performed important public functions, for instance, implementing quality requirements, but the ‘mystery’ of the trade, including its internal decision-making, was usually concealed from outsiders. They observed their law, holding ‘with an almost Byzantine formal rigour’ to ritual forms and formulas: the guild’s shrine, the fixed ritual of salutation, ritually drinking to each other and yearly common meals were not a way to ‘express commitment’ but were forms directly connected to the content of what it was to be a ‘brotherhood’ of craftsmen (Brand, 1990: 77–85; Dilcher, 1996: 10).

Beginning in the 14th century, the journeymen’s own internal jurisdiction began to be contested, in particular with respect to their control over the labour supply. Such internal conflicts were typically settled by city governments within the framework of a common notion of the constitutional importance of the guild structure. Participants in guilds could link up with the medieval static conception of ‘good old law’, according to which any change was an infringement of traditional freedoms and privileges. It was on that basis that the journeymen, mentioned at the start of this section, appealed to the Emperor as the God-given sovereign with the holy duty of maintaining the traditional law (Brand, 1990: 73, 394). For a considerable time, the guild constitution was able to stand up to challenges, such as the rise of merchants who tried to extend their control over the (textile) production process by way of putting-out systems, and the rise of consolidating territorial states trying to extend their control over the economy by imposing civil forms of law that undermined the corporative order.

During a large part of the last millennium, labour relations in European cities – in some, at least, from the 13th century – were ordered by a constitution operating as the foundation of a tripartite system of (a) masters, (b) journeymen and apprentices, and (c) city governments that predominantly kept each other in check. City governments promoted the self-regulation of associations, played an intermediary and mediating role in case of conflicts and tried to uphold, if necessary, the common interests of the city. The balance was always temporary, subject to changing power configurations, but struggles proceeded within a commonly recognized framework of communal life in the city. Labour relations to a large extent found their constitution in local conventions, experienced and held up as the ‘living law’ of workers’ collectives, in particular in the cities. Public regulation was limited and often merely covered the entry into and exit out of labour relations, ‘health and safety’ issues, and sometimes wage tariffs.

As city economies lost their relative coherence, due to economic changes and the rise of larger territorial entities, the corporate constitutions gave way to civil law regimes of liberal rule. In the 19th century, liberal state policies repressed local in favour of national laws and, on the continent, imposed a private law regime on labour relations. In times of gradual industrialization during the third quarter of the 19th century, contractual principles of this regime were challenged by industrialists’ revival of the issue, now as a legal one, of the subordination of workers. In Britain, ‘master and servant’ relations were strengthened, on the continent subordination was gradually incorporated legally into the ‘labour contract’. A counter-tendency of collectivization on labour’s side, in combination with political developments such as an extended franchise, resulted in national laws that began to embody a historical compromise of accepting the subordination of workers in exchange for guarantees as to the risks of wage labour, and, finally, in the employment contract as the legal form of this compromise. Beginning in the 1970s, this compromise was undermined by employers and governments embracing neo-liberal economic perspectives that inspired a tendency away from protections, back to the ‘normal’ enterprise of civil law contracts. The increasing amount of interdependencies, in which humans are nowadays involved, seems to promote an individual self-experience, and thus a stronger individualization, reducing the potential of traditional collective means of striving for a fair constitution of labour relations.

New Methodologies

The question raised at the outset of this article was that of which methodologies might be required for a study of labour law developments that does not begin by functionalizing law in relation to labour markets. The criticism of ‘reformist’ strategies and the arguments for an alternative perspective on the role of law in social practices, that I developed in the previous sections, resulted in a proposal that extends the ‘constitutional’ approach on the basis of a broader concept of ‘law’, so as to take due account of the historical variability of structuring labour relations.

Dukes has rightly reasserted ‘the “intensely political” nature of the question how markets are constituted’ (Dukes, 2014: 207), at a time when its political nature is being hidden by orthodox economics and, in the wake of this, by politicians who have embraced its concepts. Referring to the constitution as it had been proposed by Sinzheimer, she asks: ‘how might the idea be abstracted sufficiently from the particularities of the context in which it was developed so as to render it applicable to current conditions, while holding on still to the normative principles at its core?’ (Dukes, 2014: 194). To what extent would it be possible to define such a core that may be saved from historical contingency?

A broader historical inventory of the normative structuring of labour relations in cities located at a kind of ‘central axis’ in Europe, I argue, suggests that starting from the 13th century (Italian cities, Cologne) a tripartite configuration of (a) masters, (b) journeymen and apprentices, and (c) government, and a concomitant effort to reach a legally guaranteed balance between their interests, was a persistent element of the constitution of labour relations in these cities. As from the second half of the 19th century, this configuration replicated itself to a certain extent at national level. At the European level, however, as Dukes (2014: chapter 6) has shown, it failed to do so.

Such a configuration may, at a very general level, suggest as a basis of a ‘labour constitution’ an institutional structure that recognizes this configuration and attributes powers and responsibilities in such a way that a set of both efficient and fair relations can develop. One way to realize such a balancing structure might be to revitalize ideas which have been developed before the depoliticizing turn to neo-liberalism. Dukes (2014: 107) rightly states that opponents have not convincingly argued why the previously shared goals of labour law would now be irrelevant. It does not follow, however, that we should be restricted to searching for remedies in labour law’s past and, for instance, to placing our stake on the reinstatement of trade unions to a position of strength. That is one possibility, but not a necessary operationalization of what constitutes an efficient and fair distribution of powers and responsibilities. Guaranteeing fairness by way of mechanisms of the past, such as the collective membership of unions, is possibly challenged as a strategy by a number of recent social developments. I mention only a few: the changed structure of productive organizations, the socio-technical innovations of ICT or the increased use of supervisory bodies in regulatory domains. This comment suggests no definite conclusion about the operative possibilities of ‘old’ solutions; it is rather a methodological plea to use a constitutional perspective to look more systematically at requirements in light of current relations in the world of production.

By way of the selective retelling of part of labour law’s history above, I argue for a methodology that would include a more fundamental reflection on the relations between actual interdependencies and their ‘constitution’, and that would account for the contingencies of both ‘labour relations’ and the normative models that go into constituting them. These relations could be developed in several ways, provided that their mutually constitutive relation, the ‘duality of structure’ (Giddens, 1976; Sewell, 2005), is kept alive and not debased into a one-sided account, for instance, of ‘de-facto idealism’. ⁷ I would argue that a convincing case can be made here for the concept of ‘elective affinities’. ⁸ An example, based on the account of the previous section, may illustrate this point. When medieval guilds, as associations of competing craftsmen, felt a need to formalize their bond and had to choose among a contemporary repertoire of available models, they all selected the religiously inspired model of ‘brotherhood’. The values involved in this model implied the equality of the craftsmen and the power to consensually make their own rules. If we take account of interdependencies in the social space of the medieval city and of available normative models, we may conclude they could have hardly made any other choice. Having been made, the choice had definite consequences for their social and economic practices, but, on the other hand, did not necessarily imply that the values inherent in the model would be realized in practice. From an adequately dual perspective on social practices, it is no contradiction that, even while the values of the model kept being valued until the late 19th century, actual inequalities between workers only increased.

Looking to the future instead of the past, we may use this approach to ask questions about future constitutions of labour relations. If the prediction should come true that in a few decades robots will have replaced half of human manpower, what will be the consequences for (the normative structuring of) labour relations? Will the performance of robots be qualified as ‘labour’ or, in reverse, will human labour be conceptualized more akin to robotized power, for instance, as the return of an investment made plus an additional flow of energy to be supplied? Will models other than that of the exchange between a ‘transfer of labour power’ and a wage be mobilized?

Conclusions

Labour relations and the normative models that partly constitute them are subject to persistent changes. The conceptualization of the labour contract on the basis of a ‘transfer of property’ dates back hardly more than two centuries and the ‘permanent contract’, nowadays so easily used as a yardstick to bemoan all kinds of precarization, was probably unknown, still, to our grandparents. The problematic character of ‘labour’ as a (fictitious) commodity is fortified by recent historical-comparative research that concludes that those societies in world history that did commodify labour show a consistent pattern of short economic success, followed by growing social inequalities that result in their collapse (Van Bavel, 2016). Labour law methodology therefore ought to reckon with the possibility of not just piecemeal but fundamental change.

‘Reformist’ strategies attempt to reconcile labour law with currently hegemonic economic perspectives, and thereby adapt it to the restrictions that attach to mainstream economics. In this article, I have criticized the restriction of the concept of ‘law’ to (written) rules and argued instead for the recognition of law as co-constitutive of practice. Rather than viewing normative models such as the labour contract as ‘representations’ of actual relations, we ought to be attentive to their performative powers, in the sense that they partly create their own conditions of success, comparable to the way in which economic perspectives tend to turn humans into the utility-maximizing individuals that they were supposed to be in the first place. Law’s tendency to ‘hide behind the curtains’ should be an invitation to fling them open and uncover once more the normative landscape.

I have argued that a new methodology of labour law that recognizes its constitutive role in labour relations may be enriched by a historical perspective. Looking at labour relations in Western European cities since the 13th century, what strikes me is the balance between ‘parties’ involved that has for a considerable time been rather well preserved in the city economies of premodern times and has to some extent been restored at a national level during part of the 20th century. If we look for a basic formal model of a ‘labour constitution’, the institutional accomplishment of a persistent balancing effort in the tripartite involvement of (1) masters or employers, (2) journeymen, apprentices or workers, and (3) governments might make a good candidate. The realization of ‘democracy’, put forward in Dukes’ approach, would then be but one way of concretizing this constitution. Our methodological task would be to improve our insight into the institutional conditions under which such a balancing effort may be effective.

Accomplishing this task can be brought nearer by recognizing a certain selective affinity between types of interdependency that actors find themselves in, and the values that they tend to base the normative model of their organization upon. It is functionally of fundamental importance to a social unit that participants have values to appeal to, even if these values can be said to be insufficiently realized. These values and legal models cannot be immunized against historical change (vs. Dukes, 2014: 196), they are rather themselves the product of long-term developments in human interdependency, and the notion of selective affinity may give us a handle to investigate their relations. Both the role of legal models in shaping actual relations and the contribution that a concept of ‘elective affinity’ can make to a theory of the constitution of labour relations invite further consideration in light of new methodologies of labour law.