Where Does Solidarity End?

Tr@nsit Online

Richard Hyman considers the future of Solidarity. are sceptics correct to think that it has already been irrevocably destroyed by an epoch of individualism? Or can a redefined and reinvented concept of solidarity rescue it for future generations?

For some sceptics, solidarity has already ended: we are in an epoch of individualism, when concern for the welfare of others has no place in the rational-choice egoism of purely economic actors. As a corollary, trade unionism itself is obsolete. Several decades ago Mancur Olson, in The Logic of Collective Action, identified a paradox: that the rational individualist worker would avoid the costs of union membership, since s/he would benefit from the achievements of collective bargaining even if not a member (the ‘free-rider’ principle); but if all workers followed such instincts, there would be no trade unionism and all would be worse off. Olson’s solution to this paradox was unconvincing; and the ‘rational choice’ approach to collective action is itself flawed. In reality, most workers have always joined trade unions because they already see themselves as part of a community and do not identify themselves in purely egoistic terms. Is this no longer the case?

The question may be posed rather differently: not whether solidarity is at an end, but whether old models and conceptions are exhausted and thus whether solidarity must be redefined and re-invented. ‘Solidarity can only be rescued if it is placed on a new foundation,’ is the argument of Philippe Van Parijs. Likewise, Rainer Zoll has argued the need for a reassessment and redefinition of our understanding of solidarity. This is also the position which I wish to develop below.

The Many Meanings of Solidarity

The principle of solidarity has a long history, and like most evaluatively charged concepts its meanings have changed over time and its interpretations remain varied. For example, both socialists and catholics commonly appeal to the ideal of solidarity, but what they understand by this is in many respects very different. It is not my intention here to explore the etymology and historical evolution of the concept, but it is important to highlight some of its contrasting implications.

One conception presupposes common identity, the possession of characteristics which mark individuals as members of a group – the nation, the tribe, the religious sect (or perhaps also devotion to a particular football club or pop group) – with a collective loyalty and a clear sense of difference from those outside its ranks. Sometimes the homogeneity of the group may be reinforced by rituals, uniforms, an arcane vocabulary. The common obligations deriving from uniformity constitute what Durkheim defined as ‘mechanical solidarity’: externally imposed commitments which allow little or no space for choice or reflection. In the past, perhaps, this model of solidarity played a significant part in the collectivism of labour movements.

A second type of solidarity, at times overlapping with the first, is based on awareness of common interests which are best pursued collectively. This is the classic rationale for trade unionism: workers as a whole are victims of oppression and exploitation, individually weak as employees, consumers or citizens; but unity is strength. The foundation of effective labour movements depended on what, elsewhere, I have described as ‘solidarity as a mobilising myth’. It was by emphasising the commonality of interests that union organisers sought to persuade workers that ‘an injury to one is an injury to all’. And because interests are shaped by subjective perception as well as objective situation, belief could create its own reality. ‘Solidarity forever’ became factual, to the extent that the heroic myths actually defined workers’ understanding of their own circumstances.

A third understanding of solidarity involves mutuality despite difference. This may be based on a sense of interdependence, generating what might be called a second-order community of interest in sustaining a set of social relationships in which all are positively implicated. It was in this sense that Durkheim wrote that the elaborate division of labour within modern societies created the basis (or at least the potential) for an organic solidarity, more nuanced and more flexible than the rigid uniformity of earlier social structures. A different way of comprehending such mutuality is as an expression of the obligations of humanity. No man [or woman, we would add today] is an island, wrote John Donne. From this perspective, there is an obligation on the strong to support the weak – either on the pragmatic rationale that the roles might on some occasion be reversed, or through a more diffuse recognition of the human condition. This third approach may turn solidarity into a synonym for charity, implying pitying support for passive victims. This is far removed from the socialist view of solidarity as active and collective; but can aspects of this approach be used to inform and enrich the solidarity of labour movements?

Opposition, Class Oppression and Diversity

To explore this question we should note an important distinction, between what may be termed solidarity with and solidarity against. In the insurgent origins of most labour movements, workers’ unity was the basis for resistance to the oppressor. By contrast, conceptions of solidarity-as-charity often avoid any reference to conflicting interests and collective mobilisation and struggle. This is also true, we may note, of the conventional usage of solidarity in the contemporary politics of the European Union: in this discourse, it is necessary to combat social exclusion, but the latter is regarded as the product of impersonal social forces rather than of deliberate human agency.

Class opposition was historically an important foundation of trade union conceptions of solidarity. Yet the understanding of class was always somewhat problematic, and has become increasingly so. Unions were traditionally organised on the basis of particular constituencies: distinct occupations, specific employers, individual sectors of the economy. In uniting one group of workers, unions could divide them from others. To some extent, the ideal of class solidarity can be seen as an effort to overcome such divisions. And typically, if class was able to unite it was not so much through common interests in the present as through a vision of the future: ‘trade unionism often involved the articulation of an alternative social order comprising a wider class solidarity’. This is consistent with Alain Touraine’s theory of class: workers’ collective consciousness involved a sense of common identity which was conditioned by perceptions of antagonism to an external threat (typically, the employer); but what was necessary to give this true class significance was a totalising understanding of the context of this opposition, and linked to this a perception of a different form of economy and society. But if the traditional class utopias have lost their credibility, is class solidarity still possible?

More prosaically, it was easiest to identify trade unionism with class solidarity when the boundaries of class seemed relatively clear-cut. The typical union members of the past were manual workers in factory, mine or mill, on the docks or the railways, confronting employers whose own impressive class solidarity underlined the need for an equally effective proletarian counterforce. Now that manual workers in the traditional sense are a minority of the labour force, and in some countries also of trade union members, collective identities have become far more diffuse, and employee interests far more differentiated. Concurrently, with the expansion of public employment, the idea of the employer as oppressor has lost some of its force. Of course sociologists may argue that class analysis, with modifications perhaps, can encompass the situations of white-collar workers or public employees; but subjectively the heterogeneity of contemporary employment means that the collective interests of particular employee groups are often likely to be viewed in opposition to those of other workers rather than in parallel.

Class as the motor of solidarity is also challenged by the significance of other bases of interests and identities. One obvious example is gender. Trade unions have been forced to come to terms, often with some difficulty, with the fact that traditional conceptions of solidarity were essentially a masculine construct, while the universalism of the rhetoric of class typically expressed the circumstances of men and neglected those of women. The collective bargaining agenda normally presupposed a male, full-time employee; the social policy agenda very often presupposed a male ‘breadwinner’ with a dependent housewife. As unions have come slowly – more slowly indeed in some countries than in others – to recognise the more complex interdependence between waged employment and domestic labour, and the manifold ways in which gendered inequalities pervade the work environment, so many of the simple dimensions of old ideas of class collectivism have lost their coherence. More specifically, solidarity within and between gender groups involve very different issues, which in neither case can be reduced to old formulae of class solidarity.

Just as class relations are cross-cut by those of gender, so they are also permeated by the increasing ethnic diversity of contemporary societies. For ethnic minority workers, problems of class oppression are often reinforced by those of racism and xenophobia. In many countries, such workers are underrepresented in union membership, and far more so in positions of leadership. In the past, trade union job protections have often been associated, whether or not intentionally, with discriminatory practices. For victims of xenophobia and racial discrimination, solidarity with those in a similar position may be the first priority. Again, a challenge for trade unions is how far they can accommodate such concerns – especially when their own members may be beneficiaries and indeed perpetrators of discriminatory practices.

To take this logic further, workers possess multiple identities. The stereotype of the traditional proletarian status emphasised a common work situation, an integrated and homogeneous local community, and a limited repertoire of shared cultural and social pursuits. Though exaggerated, this stereotype did identify a core of historical reality, particularly in the single-industry manual working-class milieux in which ‘modern’ mass trade unionism had its strongest roots. By contrast, in contemporary society the spatial location and social organisation of work, residence, consumption and sociability have become highly differentiated. Today the typical employee may live a considerable distance from fellow-workers, possess a largely ‘privatised’ domestic life or a circle of friends unconnected with work, and pursue cultural or recreational interests quite different from those of other employees in the same workplace. This disjuncture between work and community (or indeed the destruction of community in much of its traditional meaning) entails the loss of many of the localised networks which strengthened the supports of union membership (and in some cases made the local union almost a ‘total institution’). At the same time, partly in consequence, a person’s employment is today less often than in the past the dominating signifier of individual identity. Even if the importance of ‘lifestyle politics’ is sometimes exaggerated, such aspects of identity as age, sexual orientation, leisure activities increasingly compete with occupational status in shaping workers’ perceptions of their interests. This poses obvious difficulties for traditional understandings of the solidarity of labour.

The Limits of Mechanical Solidarity

A century ago, in their classic analysis of trade union functions, Sidney and Beatrice Webb saw unions’ central purpose as establishing a ‘common rule’ governing the employment conditions of all members of each employee group. Mass labour movements often embraced this principle in a form more rigid than the Webbs themselves envisaged: as suggested earlier, collective regulation in industrial relations commonly showed considerable resemblance to Durkheim’s concept of mechanical solidarity. This was problematic in three key respects.

First, it presupposed a standardisation of regulation: a ‘one-size-fits-all’ model of employment conditions. Paradoxically, there was often an elective affinity between trade union rule-making and the standardisation imposed on workers by ‘Fordist’ employers. While employers insisted that workers were ‘not paid to think’, unions were suspicious of the notion that individual workers should exercise choice over their employment conditions. Understandably so, since the whole basis of trade unionism was that individual bargaining between worker and employer would be on management’s terms, and that the conditions of all employees could be advanced only by eliminating the scope for each to bargain separately. Yet the principle of standardisation often failed to recognise any differentiation of issues. Unions correctly identified some essential common rules – minimum rates of pay, maximum hours of work – which were essential if workers were not to engage in a competitive undercutting of conditions. On some issues, however, individual choice need not necessarily undermine general safeguards; but traditional union approaches to collective bargaining have been sceptical at best towards ideas of flexible regulation which allow scope for personal preferences. I return to this issue below.

Second, to some extent as a corollary, mass trade unionism was normally based on a hierarchical control which mirrored that of the employer, with a centralised determination of policy and insistence on disciplined observance of authoritative decisions. Such a model of solidarity could be justified in terms of the needs of an effective fighting organisation: ‘when a unit goes into battle it is impossible for each individual to follow their own judgment, everyone must be subordinated to a predetermined plan,’ wrote Hermann Jochade, president of the International Transportworkers’ Federation a century ago. More prosaically, from the Webbs onwards it has been common to argue that what they dismissed as ‘primitive democracy’ is incompatible with administrative efficiency and negotiating expertise: like business organisations, unions themselves should embrace the principles of scientific management.

Third, the solidarity of interest representation has always been selective. It is possible to identify four main types of issue of concern to unions. The first constitutes the traditional core agenda of ‘bread-and-butter’ collective bargaining over wages and other conditions of employment. The second relates more to procedure, status and opportunity: rights limiting employers’ arbitrary authority and underwriting employment protection, ‘fair’ mechanisms for promotion and career advancement, training opportunities and so on; and the regulation of production, the allocation of work and the determination of workloads. The third addresses the role of the state: the constitution of the social wage (hence concern with social welfare provision and taxation policy), the politico-legal framework of trade union organisation and action, the macroeconomic policies which shape the circumstances of the labour market. Finally there is an agenda not directly linked to the worker’s status as employee but addressing other facets of personal and social existence: war and peace, the environment, the sphere of consumption, the institutions and facilities of the local community. While all these varied themes have figured on the trade union agenda at different times, however, there have often been strong pressures to avoid questions which may prove internally controversial and to highlight those on which unions can deliver results through negotiation with employers or with governments. Commonly this has reinforced unions’ role as bureaucratic bargaining agents at the expense of their potential as social movements.

In my view, mechanical solidarity is an idea and an orientation whose time has passed. It has locked trade unions in many countries into a language and a mode of action which no longer attract and in many cases repel those whose interests unions wish to represent. There is an urgent need for new understandings of solidarity.

Towards New Solidarities

Union membership is almost universally in decline: the main cross-national difference is in the speed, not the fact of this decline. In the main, the composition of union membership reflects the labour force of half a century ago: predominantly male, heavily weighted towards manual workers in ‘old’ industries or in a shrinking public sector, working full-time on a ‘normal’ employment contract. It is also an ageing group: in most countries, the median age of a union member is well over 40, in some cases over 50. By the year 2020, in other words, the majority of today’s trade unionists will have retired. Can a new, and in many ways very different, generation be won to collective organisation in their place?

In many respects, the malaise afflicting labour movements in much of the world today stems from the exhaustion of their old model of collectivism. In its traditional form, ‘solidarity’ was a slogan which easily matched a conception of a working class which was not recognised as differentiated by gender, skill, ethnicity or other significant characteristics. How do we understand the idea of solidarity if the old notion of an undifferentiated proletariat is abandoned? The starting point must be that solidarity involves the perception of commonalities which extend, but do not abolish, consciousness of distinct and particularistic interests. Workers’ organisations themselves acquire cohesion and effectiveness by shaping the ways in which members and supporters identify their own circumstances and frame their own grievances and aspirations. This process can be double-edged. Most labour movements in the nineteenth and twentieth century expressed as general interests of the working class what were often the specific interests of its strongest sections. To cling to former models of solidarity means increasingly to suppress the more differentiated concerns of women workers, members of minority ethnic groups, those with marginal occupational status, new entrants to an increasingly unfriendly labour market, and so on.

A solidarity which can appeal to those outside the traditional ranks of trade unionism must accommodate, indeed welcome diversity. Hence it is more appropriate to speak of solidarities in the plural rather than in the singular. Solidarities must be multi-faceted, first because the labour market circumstances of different groups of workers are varied (and in some respects in competition); second because extra-work identities (which nevertheless affect how employees regard themselves as workers) are increasingly differentiated. In his thought-provoking book Beyond Individualism, Michael Piore has suggested a new understanding of trade unions, first as ‘communities of action’ (organisations that provide a context for individual self-realisation), and second as ‘borderland institutions’ (which bridge the ideals and perspectives of distinct social or cultural groups). Both themes imply abandoning the principles of mechanical solidarity. In other words, regulation of employment must be flexible, not uniform; common action must be the outcome of debate and discussion, not hierarchically imposed; and the suppressed alternatives of traditional collective action must be rediscovered.

A concrete example may be helpful. For more than a decade, the world of work has been permeated by risk and insecurity. The demand by employers – and increasingly, governments – for ‘flexibility’ means reversing many of the gains won by labour movements over more than a century, which ensured that workers could not be treated simply as disposable commodities. There can be no doubt that, among workers, this is a source of disorientation and resentment (though the outcome can often be to find scapegoats among even more vulnerable sections of the community). Yet the discourse of flexibility also connects with the attractions of individual autonomy and choice. Can trade unions bridge the borderland between insecurity and autonomy, giving positive force to what Ulrich beck has called ‘solidarity from anxiety’?

This raises difficult issues. It is right to be suspicious of the idea of ‘soft law’ (or the more recent analogue, the ‘open method of coordination’): its implication is a process of regulation which does not genuinely regulate. But consider some specific questions for contemporary trade unionism. For example, should overtime working be prohibited altogether; rewarded by extra payments; or compensated by time off at the individual’s own discretion? Should part-time employment be resisted, or should there be scope for variable hours of work – if employees themselves can agree or refuse, and can obtain the same employment rights as full-time employees?

The challenge for trade unions today is to develop forms of regulation which define a solid framework within which workers can exercise genuine choice. This means determining which rules should be primary and universal, which are secondary and discretionary. It means redefining flexibility, resisting flexibility as precariousness and vulnerability but embracing flexibility as the right to choose within an overall framework of protection. This is an expression of a genuinely organic solidarity: a combination of universal shelter and individual opportunity.

To survive, the principle of solidarity must be redefined and reinvented. Without this, does trade unionism have a future?


Cambridge: Harvard UP, 1965.

Refonder la solidarité, Paris: Editions du cerf, 1996.

Was ist Solidarität heute?, Frankfurt: Suhrkamp, 2000.

‘Imagined Solidarities: Can Trade Unions Resist Globalization?’ in Peter Leisink, ed., Globalization and Labour Relations, Cheltenham: Edward Elgar, 1999.

Andrew J. Richards, ‘The Crisis of Union Representation’, in Guy Van Guyes, Hans De Witte and Patrick Pasture, eds, Can Class Still Unite?, Aldershot: Ashgate, 2001.

La conscience ouvrière, Paris: Seuil, 1966.

Industrial Democracy, London: Longmans, 1897.

Cambridge: Harvard UP, 1995.

Risikogesellschaft, Frankfurt: Suhrkamp, 1986.


Tr@nsit online, Nr. 24/2002
Copyright © 2002 by the author & Transit – Europäische Revue.
All rights reserved. This work may be used, with this header included, for noncommercial purposes. No copies of this work may be distributed electronically, in whole or in part, without written permission from Transit.