https://cpes.org.uk/om/files/original/260906c0b442cbab30f010fb3d324883.pdf 693803222b6d47110617e6a144ccf2d3 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource International Contributions to Labour Studies 3 Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Text Any textual data included in the document International Contributions to Labour Studies (1993) 3, 67-90 Workplace leaders and labour organisation: limits on the mobilisation and representation of workers Robert Fishman and Carol Mershon* Despite the institutional guarantees and rights afforded by contemporary democratic societies, workers continue to encounter characteristic difficulties in securing representation, building organisation, and engaging in joint action. Analysts of labour often cite divisions within the working class as stubborn obstacles to collective mobilisation. Of equal consequence, although much less frequently noted by scholars, is the relative shortage of political skills-one might even say initiativewithin the working class. 1 This leadership shortage is especially severe at the workplace level, the foundation for working class organisation. We argue that the ways in which the spaces for workplace labour leadership are filled-or remain vacant-strongly shape the possibilities for the representation of workers and their collective mobilisation.2 Our purpose in this essay is two-fold. First, quite practically, we advocate more research centred on workplace leaders. We do so here by delineating their contribution to the broader profile of labour movement activity. (With the term 'workplace leaders', we refer to those who initiate, co-ordinate or represent collective worker action, individuals who usually hold formal positions of institutional responsibility. Under this label we do not include those 'activists' whose efforts are significant but who essentially support labour activities led by others.) Second, we make a more conceptually oriented case for how labour movements and collective worker action should be understood-independently of the operational issue of concrete research strategies. We show that plant-level leaders play essential roles in labour movements. We contend that leaders within workplaces should form a vital object of study for analysts of worker representation and collective action. Even when researchers *Department of Sociology and Kellogg Institute, University of Notre Dame, and Department of Government, University of Virginia. 1Among the few contemporaty social scientists to note the problem of political skills within the working class is Seymour Martin Lipset who discusses the phenomenon both in his 1960 and in Lipset, Trow and Coleman 1956. 2 0ur treatment of the leadership space of the union movement draws in part on the work of J. Samuel Valenzuela, who develops the concept of the organisational space of the labour movement and the filling of that space by one or more leadership groups. Valenzuela's work, however, focuses on the national level, leaving unexplored the crucial arena of the workplace. See Valenzuela (1981) and his doctoral dissertation, Columbia University. 1052-9187/93/010067 +24 $08.00/00 © 1993 Academic Press Limited 68 R. Fishman and C. Mershon devote primary attention to other equally important dimensions or aspects of labour's experience, their conceptual map of the world oflabour should incorporate the plant-level leaders and the role the leaders occupy. The workplace leadershipcentred understanding we advance here elaborates the conceptual underpinnings of our earlier empirical research on labour and politics in Italy and Spain, where we examined plant-level leaders in order to illuminate the broad challenges, accomplishments, and dilemmas of the labour movement. 1 In addition to considering the workplace leaders' powerful impact on the emergence and direction of collective worker action, we discuss the many constraints and impediments that reduce the leaders' autonomy and at times even threaten to preclude joint worker action. We do not underrate the significance of other actors and other foci in the study of labour. The workplace leaders whose role and efforts we emphasise are by no means capable of creating whatever organised following or joint action they might wish to initiate. But workplace leaders are crucial and problematic components of labour movements. By focusing on their contributions and on the difficulties they face, we hope to clarify the limits on the representation of workers' interests as well as the prospects for collective action. Indeed, the large, macro-level challenges confronting labour are manifested in part through their influence on the recruitment, the resources, and the actions of workplace leaderships. 1. The empirical reference points Before developing our argument about the distinctive importance of plant-level leaders, we turn to a brief discussion of the two cases we use most often to illustrate our reasoning. Our chief empirical reference points are Spain and Italy, the countries where we have done our own field work, although we draw on research on unionism in other national settings. In various respects, the Spanish and Italian labour movements are similar. In both Spain and Italy, multiple union confederations have historic ties to competing political parties and the party left is divided as well, with relatively strong communist parties (now post-communist formations). In both countries, workplace union leaders enjoy significant decisional autonomy and serve on elective structures chosen by all workers in the plant. It is common in both cases for the competing unions to be simultaneously present in the same workplace; representation is not restricted to the most widely supported labour organisation. Despite these similarities, the way that the spaces of workplace leadership are filled varies between the Spanish and Italian union movements. In Spain the central mechanism for representation and leadership at the plant level is the works committee (comite de empresa), an elective body representing all the workers in the firm regardless of whether they are union members or non-members.2 The competing union confederations, along with local or company unions and any interested independents, present candidates for the works committee and all workers in the company choose among them. The works committee is then charged with the negotiation of contracts with the employer, with leading any strike, and 1See Mershon, 1986, 1989, 1990A; Fishman, 1990. 2Among those scholars currently working on Spanish works committees are Modesto Escobar, Kerstin Hamann, Felipe Pimentel, and Lynne Wozniak. Workplace leaders and labour organisation 69 with the broader representation of the workers in the firm. Local union sections also exist in many firms, but their importance has usually been limited (despite the preference of one of the national union confederations for their development and recent efforts to strengthen them). A union must be able to advance candidates and secure their election to the works committee if it is to contribute to the representation of workers within a firm. In Italy the principal mechanism for representing workers and defending their collective interests before employers is the factory council (consiglio de fabbrica), an institution invented in the unprecedented wave of worker mobilisation that peaked in the Hot Autumn of 1969. The Italian factory councils represent all workers in a firm, union members and non-members alike, as do the Spanish works committees. Unlike the Spanish committee, however, the factory council is the joint base structure of the multiple national union confederations. That is, despite the tripartite division of Italian labour outside the workplace into politically distinct competing confederations, inside the workplace a single body-the 'unitary' factory council, to borrow Italian unionists' term-is recognised and acts as the unions' shopfloor agent. In 1972, at the same time that the Italian confederations created the CGIL-CISL-UIL Federation, they officially designated the factory council as their base structure. The CGIL-CISL-UIL Federation never laid down and rarely discussed formal, standard rules for the factory councils' electoral mechanisms and operations. (Prevailing practice has stressed egalitarianism, direct democracy, and co-operation among confederations.) Thus, whereas the Spanish committees are elected according to procedures and a mandate established by law, the Italian councils are elected by less institutionalised practices established in the course of collective worker action, and the councils' role, but not their procedure of election, is recognised by law. 1 The Italian councils, unusual in their origins and institutional design, are characterised by a high degree of decisional autonomy. Moreover, as we emphasise below, their emergence as part of a wave ofworker mobilisation helped to produce (above all for the councils' first decade of existence) a large pool ofleaders and activists to serve on the councils and support their work. Even so, the Italian unions face the same challenge as their Spanish counterparts: to recruit, encourage, and develop a plant-level leadership. For without workplace leaders capable of winning election to the factory council the Italian union confederations lack a direct channel for an immediate role in the representation of workers in the enterprise.2 'The Workers' Statute of 1970 (Article 19) guaranteed the union right to a shopfloor presence, using the generic term 'firm-level representative organ' (rappresentanza sindacale aziendale) and leaving the precise structure of such an organ to the unions and 'workers' initiative'. In 1972, the three confederations explicitly defined the council of delegates as 'the union base structure [istanza] with powers of negotiation in workplaces'. See Smuraglia (1975), 86-88,231-233,284-287. 2It is worth noting that in the late 1980s the Italian confederations began to reconsider the factory councils' institutional design, moving toward establishing greater formality in council procedures. Proposals were made-and in places implemented-for the creation of union-specific shopfloor bodies alongside the unitary factory councils. Moreover, the confederations aired proposals for organisms called 'workplace councils of union representation' (CARS or consigli aziendali di rappresentanza sindacale), which would have replaced the factory councils; and then in March 1991 the CGIL, CISL, and UIL signed an agreement to establish plant-level RSU (unitary union representatives, or rappresentanze sindacali unitarie). The agreement on the RSU, like the plans for CARS, has generated much discussion and little action; and the factory councils largely remain in place. See, for example, Nuova Rassegna Sindacale, November 16, 1992. 70 R. Fishman and C. Mershon We contend that crucial differences having to do precisely with workplace leaders help to account for a striking disparity between these two cases: Italian unions are far more successful in recruiting members than are Spanish unions. Throughout the 1980s Italian unions maintained relatively high levels of affiliation (unionisation stood at 49·8% of the dependent work force in 1980 and 41·7% in 1990), whereas Spanish unions enrolled only about 13% of the labour force in 1982, 5 years after the return of democratic freedoms in 1977.1 This extremely low figure reflects in part the disinterest of many plant-level leaders in persuading workers to formally join unions. As Fishman's work has emphasised, the workplace leaders' understanding of union action-based in large measure on the experience under the Franco dictatorship of mobilisational success despite the organisational weakness of the opposition labour movement--encourages them to represent and attempt to mobilise all workers without regard to formal membership.2 Recent research has confirmed that many workplace leaders do little to solicit formal affiliation although in some unusual work settings the efforts of the local leadership produce relatively high membership. 3 Thus, the Spanish model of high levels of collective mobilisation and extremely low levels of membership is incomprehensible without reference to the expectations and behaviour of plant-level leaders. In contrast, Italian workplace leaders have actively and successfully contributed to union recruitment. The nature of the factory councils as the within-plant organ of the union confederations has helped to foster the Italian workplace leaders' efforts to encourage formal membership.4 Although our analysis focuses on two cases that are fairly similar institutionally, the clear thrust of the argument is to emphasise common problems of broad applicability in the study of labour. Despite the evident explanatory relevance of important cross-national differences in institutions, we wish to underscore here the actors who fill or leave vacant institutionally defined positions. We contend that plant-levelleaders should not be taken for granted. Their actions cannot be deduced from features of institutional design or from the policies of the confederations to which they belong. The sensibilities and the efforts of actors-as well as the simple fact of their presence or absence-powerfully shape behaviour and outcomes within institutions. The study of plant-level leaders makes complementary the analysis of individual action and of institutional constraint. Thus throughout the essay we suggest more or less generalisable implications of our discussion and in the conclusion we further explore the broader significance of the case we build. 'Data on unionisation in Italy are from Romagnoli and Della Rocca (1982: 92) and Pirani et al. (1991: 52); on unionisation in Spain, see Fishman (1990, eh. 6). 2See Fishman (1990, chs 4, 6, forthcoming). 3This is one of the findings in the excellent study, Iriso Napal 1992. 4Along with the efforts of workplace leaders in Italy, we should note that institutional incentives generate relatively high unionisation rates in some sectors. Legislation in 1973 provided that union members may handle all procedures related to social security pensions in union offices, which accounts for a steep increase in the enrolment of pensioners from the mid-1970s to the early 1990s. In agriculture, 1973 legislation stipulated automatic union enrolment for workers carrying out filing procedures for the SCAU [Servizio contributi agricoli unijicatt] in union offices; and in 1990 union membership in agriculture stood at 84.3% of dependent workers (Pirani et al., 1991: 53). Workplace leaders and labour organisation 71 2. The methodological approach: avoiding characteristic pitfalls in the study of collective worker action By centering our understanding of collective worker action and our research activity on the workplace leadership role we seek to avoid several dangers posed by alternative methodological approaches. We may identify these dangers as excessive and undifferentiated methodological individualism, which can amount to rank-andfile determinism; excessive and unwarranted aggregation or macro-level determinism; and the misidentification of what is observed. Let us briefly consider these three pitfalls. Excessive and undifferentiated methodological individualism involves the attempt to explain all collective worker actions or outcomes by exclusive reference to the individual-level attitudes and characteristics of all workers concerned. This approach has to some extent been encouraged by the development of the sample survey and the possibilities for analysis it affords. The approach is by no means restricted, however, to practitioners of survey research; and, on the other hand, it is not inevitable in studies based on survey data. A telling example of individual-level determinism is found in Mark Van de Vall (1970). Van de Vall notes the great increase in union membership among US Federal civil service employees in the early 1960s, and attributes the positive outcome for unionism to a decline in the status of civil service work. This decline, he reasons, left the Federal employees more inclined to accept unionisation in a country where collective worker organisations had long been largely identified with blue-collar workers. A far more compelling explanation for the same outcome is provided by Hugh Clegg (1976). Clegg contends that the surge in civil service unionisation resulted from President Kennedy's extension of collective bargaining rights and recognition to Federal employees. Many times, as in this case, decisions or actions taken by collectivities, institutions or macro-level actors provide the crucial impetus for the observed behaviour of individuals (by offering incentives to individuals, shaping expectations, structuring opportunities for choice). We do not deny that the attitudes and characteristics of individual workers are among the determinants of collective outcomes. The pitfall is the assumption that all or most of the causally relevant phenomena can be observed at the level of the individual. The converse of excessive methodological individualism is excessive or unqualified aggregation-the misplaced assumption that collective actors actually and reliably behave just as their top-level leaders and official histories intend or proclaim. This tendency is readily found in scholarly work, including much of the highest quality. In identifying this pitfall we do not mean to argue against the reliance on the statements and views oftop-levelleaders and official bodies. Rather, we emphasise that the views elaborated and expressed cannot be assumed-without evidence-to represent or determine broad patterns of collective experience. The final pitfall is the misidentification of what is observed, an all-too-common occurrence when collective worker action is observed and interpreted with little or no understanding of the layering of involvement and, more specifically, of the role of critical minorities. When students of labour fail to appreciate the distinctive roles of workplace leaders, activists, informal leaders and other critical minorities, any 72 R. Fishman and C. Mershon insurgent collective mobilisation tends to be seen as an instance of generalised rank-and-file action-of working-class initiative from the base. Perhaps the most common scenario is the boisterous assembly or even union congress that rejects an agreement negotiated by higher-ranking leaders. Such instances of resistance to accommodation are regularly interpreted as signs of rank-and-file militancy even if the large majority of workers remain disconnected from the expression of opposition. A more specifically faulty perception occurs when students of labour take at face value every 'waving of the flag of union democracy', neglecting to note that the demand for union democratisation is frequently aired by subordinate leadership groups that enjoy no more-perhaps even less-genuine rank-and-file support than does the dominant union leadership. 1 We do not mean to argue that rank-and-file workers are irrelevant, that they are always inactive or quiescent, or that they form an undifferentiated mass. We insist instead that actions of informal leaders, insurgent alternative leaders, or even workplace-level official leaders are frequently mistakenly viewed as representative of generalised worker sentiments at the base. Thus through an appreciation of the indispensability of leaderships-and a recognition of the obstacles they face-we are better able to interpret what we observe.2 3. The indispensability of workplace leaderships How and why do workplace leaderships play essential roles within labour movements? We examine, in turn, the inadequate supply of workplace leaders, the leaders' distinctiveness, their autonomy in different national contexts, and the two levels of working-class life they bridge. 3.1. Indispensability and shortage: the difficulty offilling the workplace leadership role Without some union leadership at the plant level the possibilities for collective worker action are very limited. Workplace organisation depends on the presence of plant-level leaders. Thus leaders are necessary if unions are to engage in activity inside workplaces.3 And yet many workplaces---even some where the workers might in principle be quite favourably disposed toward collective action-suffer from an absence of leaders of any stripe whatsoever. This fundamental constraint on the representation of workers and collective worker action is a focal point of much of the analysis that follows. Union strength and mobilisational capacity at the national 10ne scholar wryly observes that in the Italian union movement 'whoever waves the flag of union democracy is always the one who is in the minority at the moment' (Guido Romagnoli, quoted in Lorenzini 1982: 90). On the instrumental demand for union democratization see Fishman (1990, eh. 2); Mershon (1990B); Regalia (1984A); and Sabel (1981). 21n emphasising the role of workplace leaders, we do not wish to suggest a simplistic view of union officials outside the workplace. Nor do we imply that rank-and-file workers constimte a unified, homogenous entity. In a work of this scope, however, our focus on workplace leaders precludes extended discussion of national leaders and the rank and file. Workplace leaders have received much less scholarly attention than have top-level leaders and the rank and file, and we hope to redress this neglect. For a judicious treatment of the complex divisions within unions, see Zeitlin (1989). 30n this point, see Fishman (1990:, 45-46). Workplace leaders and labour organisation 73 level rest, to a significant degree, on labour's organisational presence within plants. 1 Without such a presence, some workers may still join unions and unions may attempt to co-ordinate mobilisations exclusively from territorial or sectoral headquarters. But the labour movement cannot attain broad strength-the capacity to mobilise workers, a credible claim to speak for them, effectiveness in collective bargaining and in securing the implementation of agreements reached-without a reasonably extensive workplace presence. Of course, militant forms of worker protest may emerge in opposition to official union leaderships and, in rare instances, even in local contexts where no formal union presence exists. The most militant forms of worker opposition to practices of labour accommodation, however, tend to appear in those plants that have a history of fairly strong union action. This was the case in France's May 1968 and in Italy's Hot Autumn of 1969.2 Informal leaders often come forward where unionism is well established, and in many instances supplement, rather than contest, the activity of official leaders. But in workplaces with no union presence informal leaders are exceptions, not the norm. Thus even where informal leaders challenge the policies carried out by official union leaders, the informal leaders themselves tend to be products of union life and collective worker action. 3 And yet despite the indispensability of plant-level leadership for formal union activity to take place within the firm-and under normal circumstances for any collective worker action at all to ensue-the supply of leaders within the workplace is typically insufficient from the standpoint of the national labour movement and its broad objectives. In numerous firms no union leadership at all is present, and in many others the available leaders represent only a truncated segment of the larger spectrum of union alternatives found at the national level. In this and other ways, even in firms with some labour representation or union structures present, the leadership shortage shapes the type of union life that is possible. Data from the Spanish case underscore the magnitude of the leadership shortage in that national context. After the 1980 plant-level union elections in Spain, it was estimated that 52-57% of the workers eligible to vote were employed in firms where elections for works committees had actually occurred.4 Nearly half of the eligible workers, then, never had the opportunity to choose union representatives; in many plants, no workplace leaders came forward to present themselves and thus to initiate elections, leaving the firms with no organised union presence. Moreover, the Spanish labour confederations have frequently encountered difficulty in penetrating 1For discussions of different panerns of organisation in the firm, see Carew (1976) and Ludevid Anglada (1979). 2 0n the Italian case, see Mershon (1986, 1990B) and Regalia (1980). On the French case, see Mann (1973, eh. 6) and the sources cited therein. In some plants where significant episodes of informal opposition have occurred-one might think of Italy's FIAT, for example--union organisation inside the plant may be somewhat weaker than one would anticipate on the basis of the size and importance of the factory in question. Nevertheless, even in FIAT, linkages to extra-factory union functionaries and within-plant union efforts maintained a level of union activity far greater than that found in numerous small plants with no union present. 'Informal opinion leaders play an important role in the strongly union British factory studied in Batstone, Boraston and Frenkel (1978). However, many of those leaders with no elective position such as shop steward are in fact former stewards. 4This conclusion was reached by UGT in its official study of the 1980 elections (UGT, 1981). 74 R. Fishman and C. Mershon even those firms that do have works committees. In Fishman's 1981 survey of Spanish works committee chairs, no confederation succeeded in advancing candidates in every firm with a works committee, and the smaller confederations penetrated very few workplaces. Comisiones Obreras, the union confederation with the broadest presence at the time, was able to field candidates in 86·4% of the sample's firms. (For obvious reasons, the sample was drawn from firms where works committee elections had been held and therefore where at least some union leadership was present. Thus the magnitude of the leadership shortage is considerably greater than these data imply.) The UGT, the Socialist union confederation and in 1981 the second largest labour organisation, presented candidates in just 72·8% of the firms with committees, leaving over a quarter without the presence of the Socialist union option. Despite the great emphasis in the national media at the time on the political competition between these two confederations, workers could not actually choose between them in over one quarter of the firms studied. The leadership supply problem thus prevented a highly visible cleavage from reaching many workers inside their workplaces. The shortage of workplace leadership was far more severe for the smaller confederations. The Union Sindical Obrera was able to present candidates in only 11·4% of the firms surveyed, and the two Maoist confederations reached, respectively, 7·7% and 6·5% of the workplaces (Fishman, 1990, table 2.1). Even though, in some respects, these smaller unions appeared to mount a significant challenge to the larger confederations (as manifested in their vociferous opposition to the wage restraint of the Moncloa Pacts), in fact the leadership shortage confined them to a small minority of workplaces. The Italian evidence on the supply ofworkplace leadership is somewhat less clear. We have treated the existence of competing candidate slates as an indicator of the presence or absence of a workplace leadership of the various Spanish confederations. In Italian firms, formal lists of union candidates are rare, and the confederations do not engage in institutionalised competition in factory council elections. Factory council delegates are elected in constituencies of small work units. Workers generally meet in departmental assemblies to nominate colleagues in their work unit to the post of council delegate, and vote immediately thereafter with blank ballots or by a show of hands. 1 Although workers tend to know the confederal affiliation of candidates, the candidates do not stand as official representatives of a confederation. And yet, despite our lack of strictly equivalent measures, other sorts of evidence suggest that the problem of leadership supply constrains workplace union action in Italy as in Spain. Slightly over a third of all wage-earners in Italy (36·3% in 1977) elect a factory council. 2 Some Italian workers are represented in other ways: In 1In other ways, as well, the Italian factory councils display greater informality than do the Spanish works committees. The works committees elect a secretary or chairman; the councils generally lack institutionalised individual leadership. (Many large councils elect collective leadership in the form of an executive committee). The Spanish works committees, indeed, resemble the Italian internal commissions, which were supplanted by the factory councils. The distinctive traits of the factory councils-their emphases on egalitarianism, direct democracy, and interunion co-operation-reflect the councils' moment of invention, as we argue below. 2This figure is calculated from the most recent detailed 'census' of councils, that recorded in Coi (1979: 209). Workplace leaders and labour organisation 75 agriculture, public and private services, and the South, and in small firms of all sectors and regions, union representatives (either named by union functionaries or elected by union members) and union sections are a common alternative mode of representation. However, there are (unknown numbers of) other Italian workers with the legal right to union representation who remain without any such representation because of the absence of workplace leaders. 1 These workers, it is safe to assume, are concentrated in small and medium-sized firms, where a total absence of leaders may simply reflect the restricted size of the pool of available workers. Moreover, even where factory councils exist, telling evidence of a leadership shortage appears. In many factories, one or more council constituencies (often work units of white-collar workers) have no workers willing to serve as a factory council delegate.2 The result may be a vacancy impossible to fill; the successful drafting of a worker who is persuaded to represent fellow workers on the council; or the failure of an attempt at conscription, the rapid resignation of a draftee, followed-or not followed-by a by-election in the constituency. Where a resignation is not followed by a by-election, the position remains vacant, a clear demonstration that all present within the work unit are unwilling to take on the leadership role. The workplace leaders who do emerge in constituencies that would otherwise be left unrepresented (or 'uncovered' in delegates' jargon) may enjoy more autonomy than is often assumed given the lack of any alternative leadership. This helps to explain why some councils are able to remain in office for a number of years without any renewal of their electoral legitimacy.3 Thus the shortage of leadership limits the range of choices effectively open to workers. The shape of union representation inside a factory depends to a significant degree on the presence or absence of workplace activists willing to undertake the leadership role. We do not mean to argue that political choices are always non-existent at the workplace level. Cleavages visible at the national level in the labour movement do reach many workplaces, but this penetration never extends to all workplaces. In some work settings a full range of alternatives is effectively provided to workers; but in many other settings the range is narrowed, for some alternatives may be totally absent and others only weakly represented.4 Thus the shortage of workplace leaders 1Still other Italian workers, those in extremely small workplaces (under sixteen employees in commerce and industry and under six employees in agriculture), were not covered by the Workers' Statute until 1990 and thus lacked the legal right to union representation. Workers in the submerged economy, of course, are not unionised. 2Mershon also found that although the councils' design stipulates the representation of work units, in some factories the council observed an informal practice of encouraging the representation oflocal union minorities-and council leaders complained of the difficulty of convincing members of the minority (usually the UIL, the smallest of the three confederations) to serve on the council. 3The most widely known example is FIAT Mirafiori in Turin, but, in light of the spectacular ups and downs of unionism at FIAT, it may not be the best one. It is likely that the prolongation of councils' tenure between elections became more common in the 1980s, as union defeats and internal divisions discouraged workers from taking on the delegate role. For additional evidence on difficulties in filling delegate positions, see the important research of Regalia (1984B), especially pp. 96-103. 4We have emphasised that options may be completely absent from the workplace due to the leadership shortage problem. The supply of leaders is not a dichotomous variable, however, and the viability of union alternatives is a matter of degree. Leaderships and alternatives may be 'weakly' or 'strongly' present, depending on such factors as the levels of political skill they display, their length of experience, their unions' numerical strength, and their positions in the social networks of the workplace. This point deserves more attention than we have been able to give it here. 76 R. Fishman and C. Mershon limits the extent to which political or strategic options penetrate the firm and provides some autonomy-an influential role often without organised local opposition-to those workplace leaders who do emerge. Nevertheless, as we discuss below, numerous factors limit the autonomy of workplace leaders. 3.2. The distinctiveness of the plant-level leaders and the institutional basis for their autonomy There would be little need to focus on workplace leaders if their attitudes automatically, predictably mirrored the preferences and opinions of the rank and file or if the leaders loyally and unfailingly endorsed union policies established at higher levels. The evidence shows, however, that many workplace leaders hold attitudes that set them apart from the rank and file as well as from union officials outside the workplace. The leaders' distinctiveness, we may reasonably argue, results partly from recruitment processes: both self-selection and choices made by the rank and file may often elevate relatively radical workers to the position of workplace leaders. Once the leaders are selected, their experiences form another source of distinctiveness. The leaders are charged with fulfilling specific responsibilities and at times face-or at least have reason to fear-repression or negative sanctions from employers. Let us first compare workplace leaders with the workers they represent. Students of the labour movement have advanced sharply divergent arguments about the relationship between the views of leaders and of the rank and file, arguments that cover all imaginable possibilities: that leaders are more radical, more conservative, or typical of rank-and-file workers, or that the relationship between leadership and rank-and-file attitudes varies significantly from one time or place to another. 1 We consider the last thesis-of variation across contexts-to be the most convincing and the most consistent with the available evidence. It is quite likely, in any case, that the most prevalent pattern is the first: that in most contexts plant-level leaders are more radical than the majority of the rank and file. The Spanish workplace leaders interviewed in 1981 were clearly more radical than the rank and file. The leaders were much more likely than ordinary workers to perceive the firm as a fundamental arena of conflict between the employer and workers rather than as a team (Fishman, 1990, table 2.2). Moreover, the Spanish leaders were well aware of this contrast in perspectives: approximately 60% of the sampled leaders judged the workers in their own firm to be more moderate than the union representatives there, while almost all of the remaining leaders believed that the workers held roughly the same position as their representatives. Fewer than 5% believed the workers in their firm to be more radical than the union representatives (Fishman, 1990, table 2.3). The Italian evidence is perhaps somewhat less direct but broadly parallels the Spanish findings. Over 80% of the factory council leaders in Mershon's 1982 study perceived a conflict ofinterests between workers and employers; in contrast, roughly 1For example, a classic statement of the view that the organisational position of labour leaders moves them in a conservative direction remains Michels (1962). Two analysts who suggest that the labour leadership is often more radical than the working class as a whole are Lipset (1960, eh. 12) and Stephens (1979: 86). For further discussions with references to works advancing opposing views on this question, see Fishman (1990, eh. 2). Workplace leaders and labour organisation 77 half of the rank-and-file workers questioned in several large-scale sample surveys in Italy saw an essential harmony of interests between workers and employers.1 Even though the council leaders held more conflictual attitudes than did workers in general, many leaders denied the existence of precisely this attitudinal divergence. Over half of the Italian leaders regarded the workers and the council in their plant as equally conflictual. Under a third of the leaders deemed workers in their plant less conflictual than council representatives there. This difference between the Italian and Spanish plant-level leaders in their characterizations of workers' attitudes might in part reflect a greater genuine attitudinal similarity between leaders and workers in Italy, where the solidity of union organisation and working class political culture has been greater than in Spain. Even more important, the Italian leaders' insistence on similarities between workers and councils conforms to the leaders' conception of their role, a conception rooted in the founding moment of the factory councils. The factory council is the clearest institutional expression of the ideals communicated in the Hot Autumn: egalitarianism, class unity, and direct democracy. We explore below founding moments and their impact on the behaviour of workplace leaders. It is worth emphasising here that, despite some differences between Italy and Spain, in both cases the problem of leadership supply and the recruitment processes tend to yield leaders with a distribution of attitudes somewhat different from the working class as a whole. This conclusion, we should note, does not undermine the democratic legitimacy of the plant-level leaders. The workers choose representatives to lead and co-ordinate collective worker action, a task more taxing and complex than the mere mechanical reproduction of mass-level attitudes. Where workers select leaders more radical than themselves, they do so, presumably, either because they regard such leaders as best qualified to organise joint action or because more moderate leaders are not available. Moreover, collective worker action normally requires broad participation by the rank and file if it is to succeed; the leaders' ability to effectively pursue radical mobilisation is therefore constrained by the predispositions of workers. Thus the greater radicalism of the workplace leaders, when and where it is found, does not diminish the legitimacy of working-class representatives inside firms. The attitudes ofworkplace leaders frequently differ, as well, from those ofleaders higher up in the union hierarchy. And workplace leaders at times oppose or reinterpret or even ignore official union policy. Thus, as many students of unionism have noted, plant-level unionism may follow a dynamic quite unlike the one envisioned in official policies formulated at the national level. 2 By focusing specifically on plant-level leaders, it is possible to make this broad observation more concrete. In Spain, plant-level leaders frequently disagreed with the official position of their confederation on the institutional mechanism-works 1Mershon (1990A, table 7). We replicated questions on class radicalism across leaders and workers in each country, not across leaders in Spain and Italy. 2Examples of interesting work that offer related perspectives include: on Italy, Locke (1990) and Regalia (1984B); on Germany and other cases, Alien (1990); on Sweden, Korpi (1978), especially eh. 7, and (1981). 78 R. Fishrnan and C. Mershon committees or union confederations-appropriate for collective bargaining (Fishman, 1990, table 2.8). On issues that elicit widespread political enthusiasm or polarise convictions such differences with official policy might be more widely expected; differences are indeed found on highly charged political issues. That differences between the workplace leaders and their confederations also appear on an issue of organisational policy underlines the autonomy of leaders inside the firm in their perceptions and preferences. The Italian factory council leaders interviewed by Mershon defined the union movement's broad purpose in bargaining somewhat differently than did union officials outside the workplace (1986, table 8.1). Almost two-thirds of the council leaders spoke of extensive dissatisfaction among union cadres and militants, and frequently attributed this discontent to the national leaderships' lack of attention to cadres and militants. A number ofleaders also criticised the austerity policy adopted by the confederations in 1978.1 Thus workplace leaders may at times favour a course of action suggested neither by a reading of mass-level attitudes nor by official confederal policy. But the extent to which the leaders are actually able to act upon their inclinations-in other words, the dimensions and degree of their decisional autonomy-depends on several factors, including the institutional mechanisms for workplace representation. In both Italy and Spain workplace representatives are empowered to negotiate firm-level contracts with employers, although they may be constrained by agreements reached and mobilisational plans formulated at higher levels. Even where agreements or mobilisational strategies are decided further up the union hierarchy, the workplace leaders may fail to contribute actively to their implementation. In addition, in both Spain and Italy the plant-levelleaders receive their legitimacy from the workers who elected them and their pay from the firm employing them. Their de facto institutional autonomy is, therefore, perhaps even greater than official documents might suggest. The greatest weapon at the disposal of the confederations may be the threat of expulsion of unionists who violate discipline. This ultimate sanction is unusual in Italy and Spain, however, perhaps because of the leadership shortage as much as any other consideration.2 3. 3. The autonomy of workplace leaderships in different national contexts Despite some important contrasts in institutional design between the two countries, in both Spain and Italy the opportunity for workplace leaders to differ with higherups is considerable. To what extent do workplace leaderships in other settings share this opportunity? We address this question by highlighting a distinction and presenting a sketch of extreme conditions. We have discussed two types of workplace leadership autonomy in the cases of Spain and Italy: what might be called 'attitudinal autonomy', and decisional autonomy. The first consists of the workplace leaders' affirmation of opinions or 1Mershon (1986), pp. 85-286, 289-92. We should note that the workplace leaders had company in their criticism of austerity, for vertical coalitions within the union movement opposed the policy. See, among others, Golden (1988). 2In the exceptional case of Gij6n, the largest city in Asturias, the local leaders of Comisiones Obreras were expelled from the confederation because of their left-wing policies; they established a new left union in Gijon, which continues to exist. For an interesting study of this, see Vega Garcia (1989). Workplace leaders and labour organisation 79 ideals different from those of leaders at upper levels of the union hierarchy. The second refers to decision-making capacity and behaviour, to the possibilities for undertaking action independent of-and perhaps in opposition to-extra-factory union structures. Let us develop this distinction in considering the role of workplace leaders in a variety of national contexts. Imagine, for a moment, a country with no shortage of workplace leaders. Suppose further that this country's readily available set of plant-levelleaders is homogeneous in outlook and that it holds attitudes identical to the (again homogeneous) attitudes of upper-level union leaders. Of course, no country fits this description. Still, we should contemplate the implications of such an utter lack of attitudinal autonomy on the part of workplace leaderships. Under this extreme condition, the plant-levelleaders' decision-making spheres and capacities would be of no special interest. In such a context, regardless of the formal degree of union centralisation or decentralisation, the workplace leaders' choices would prove to be quite predictable and their decisional autonomy quite restricted in practice. Putting aside the extreme hypothetical case of a complete absence of attitudinal autonomy, the workplace leadership role is, admittedly, of greater significance in some national settings than in others, but it is never without importance. In countries where union centralisation is greater and workplace institutions are more functionally limited than in Spain and Italy, workplace leaders' decisional autonomy is by definition less extensive. And yet imperfect information and the need to respond to specific problems leave some latitude in workplace practice even in highly centralised union movements. 1 It is highly likely that the decisional autonomy of plant-level leaders depends both on the severity of the leadership shortage and on the salience of divisions within labour. The more pronounced the shortage of workplace leaders, the more national leaderships are forced to rely on a heterogeneous pool of representatives and the greater their incentives to tolerate local deviations from national policies. Where multiple, politically distinct union confederations organise workers, the room for plant-level differences is likely to be greater than where a single confederation exists. With a plurality of confederations, workplace leaders may borrow and recombine elements of strategy from various organised sources, and may find justifications for resistance to one national decision in another confederation's position. Extra-factory union officials, for their part, are constantly aware of the competition to represent workers and of the failure of this competition to penetrate numerous workplaces. Competition and scarcity together induce tolerance for plant-level departures from the national policy line. 3.4. The 'two languages' of the workplace leaders The role of the plant-level leaders is vital to the labour movement in other respects as well. Not only do the leaders' efforts establish the organisational presence of the unions in the firm, but they also link two very different levels of working-class life: the workplace with its diverse, highly specific, and frequently rather mundane 'On variations across workplaces in the Swedish case, for example, see Korpi (1978). For a concise comparison of the powers of plant-level bodies in Britain, Germany, the Netherlands, Norway and Sweden, see Lange (1984, table 6). 80 R. Fishrnan and C. Mershon concerns; and the national labour movement with its broad strategies and objectives. A co-ordinated national labour strategy of any sort-whether one of mobilisation or accommodation-requires the active participation of the workplace leaders if it is to reach the rank and file. In fact, the workplace leaders must generally be able to speak two different languages: the at times abstract political discourse of the national labour movement as well as the more direct and concrete language of the workplace. Although some rank-and-file activists may use these two 'languages' interchangeably, weaving them together in a single statement, as often as not these languages remain quite distinct. Plant-level leaders are typically able to adopt or drop one language or the other, depending on the sphere of activity in which they frame their appeals. To do their work well, the plant-levelleaders must maintain a firm grasp of the problems and concerns ofthe workers they represent. In our research we have found evidence, as well, of their fine appreciation of complex questions of political strategy and objectives. When asked about the appropriateness of a general strike in response to each of several different situations, the Spanish workplace leaders distinguished sharply among those situations on the basis of political criteria (Fishman, 1990, table 2.8). The Italian council leaders also displayed a careful calibration of response to several hypothetical political circumstances. These political sensibilities are likely to find expression in the work of the leaders: the autonomy of the plant-levelleaders allows them to rely to a considerable degree on their own political judgment as they attempt to shape collective worker action in the firm. Nevertheless, the leaders encounter numerous constraints and limits as they carry out their work, and it is to these that we now turn. 4. Limits on plant-level leader initiative Several basic constraints on the workplace leaders are readily apparent. The support of rank-and-file workers-although it need not be unanimous-is vital if leader initiatives are to succeed; mobilisation depends on active participation. The employers as well pose a constant limit to the leaders' efforts. In the extreme (and not uncommon) case, employer repression or negative sanctions against unionists continue to take place in national settings where such sanctions are formally illegal. Even where employers do not resort to repression, their response to union demands clearly helps to determine the outcome of labour conflict. The state, as well, is a frequent actor in labour relations, not only in instances of peak-level bargaining but also in setting and interpreting the legal basis for union activity at the plant level. In a broader sense, the political environment may provide opportunities for, or constraints on, union action. Rather than analysing here these fundamental factors, which have tended to occupy the attention of many macro-analysts, we shall instead focus on several less obvious limits on the action of the workplace leaders. 4.1. Foundational experiences The behaviour of the plant-level leaders, and even the inclination of individuals to fill the leadership role, may be strongly influenced by 'foundational experiences'-instances of collective worker action capable of generating fairly Workplace leaders and labour organisation 81 durable expectations, objectives, personal commitments, and even institutions. Powerful collective experiences such as the struggle against Francoist repression in Spain and the Hot Autumn of 1969 in Italy redefine the boundaries and repertoires of collective worker action. 1 These experiences create large pools of new leaders and activists, broadly shared memories of collective struggle, and new understandings of the range of possible allies, opponents, strategies, and challenges. They may also lead to the establishment of new organisations (for example, Comisiones Obreras in Spain) and new institutions (the factory councils in Italy). Students of the labour movement with no specific sensitivity to the role of plant-levelleaders may be aware of the importance of foundational experiences such as the Hot Autumn. We wish to stress that foundational moments can continue to shape collective worker action for years or even decades through their enduring impact on the beliefs and behaviour of the many workplace leaders formed or brought forward by those experiences and through the institutions or organisations that set the framework for collective action. 2 In both Italy and Spain we found that a disproportionately large subset of the leaders we interviewed had entered the workplace as employees shortly before or during crucial foundational experiences: the Hot Autumn in Italy and in Spain the surge of labour conflict in the late Franco years (Mershon, 1986, eh. 4; Fishman, 1990, eh. 3). These instances of collective struggle produced large pools of new workplace leaders (especially among young workers who had recently joined the work force), thereby alleviating the leadership supply problem without eliminating it entirely. An additional reason why foundational experiences create many new workplace leaders is that they may pull existing plant-level leaders out of the firm to higher positions in newly formed or expanded organisations. In such cases the 'vacancies' opened make it possible for new leaders to arise at the workplace level. 3 The large group of workplace leaders formed in the course of powerful collective experiences appears to act differently than do other leaders. In Spain, the young leaders socialised during the surge of protest in the final years of the Franco regime and the first key years afterwards are inclined, by an overwhelming margin, to carry out their union activity within Comisiones Obreras, the Communist-oriented union confederation that emerged in opposition to the Franco regime (Fishman, 1990, table 3.2). Older leaders are much more likely than the young to join the more moderate Socialist confederation. In Italy, the council leaders formed during the 'The notion of changing repertoires of collective worker action is developed in the work of Charles Tilly. See, for example, Tilly, Tilly and Tilly (1975). 2 0ther social scientists have developed similar themes about decisive historical moments. In their classic work, Upset and Rokkan (1967) stressed the importance of foundational experiences for party systems. The widespread mobilisation entailed in struggles for enfranchisement, they argued, channelled subsequent relations between representatives and represented. Linz (1978) has discussed the inauguration of a democratic regime as a critical moment, one capable of strucmring the subsequent alternatives open to elites and citizens. Tarrow (1989) analyses the cycle of mass mobilisation in Italy as one that evenmally enriched Italian democracy. Collier (1986) uses the term 'founding moment' to refer not to a phase of mass mobilisation itself but to regime changes that in some way respond to the demands for popular incorporation voiced in mobilisation and reform movements. The new political senlements, she emphasises, carry a logic that influences state-labour relations for years after they are reached. This research has been extended in Collier and Collier (1991). 3ln Italy, union apparamses grew rapidly after the Hot Aummn. See Coi (1979). This reasoning obviously follows the work of Harrison White on vacancy chains. 82 R. Fishman and C. Mershon Hot Autumn display distinctive patterns of partisan affiliation: leaders who were young delegates from 1968 to 1972 were very likely to have been members of a political party at the time or to have joined since then; leaders who were young union members, but not delegates, during the Hot Autumn were relatively unlikely to be partisans in 1982; and among party members only, the Hot Autumn cohort, more frequently than older leaders, enrolled in their party only after some union experience. 1 Different foundational experiences can provide the basis for variations among leaders in conceptions of their role, definitions of the purpose and goals of their union organisations, assumptions about the boundaries of possible strategies, and understandings of the proper activity of representative bodies.2 The Italian Hot Autumn expressed and diffused the ideals of confederal unity, egalitarianism, and direct democracy. These values are embodied in the factory councils' institutional design, but they also find an echo in the beliefs of council leaders. The Italian workplace leaders are much more likely than their Spanish counterparts to stress unity among the unions within the plant, and they are far more reluctant than the Spaniards to acknowledge the greater moderation of the rank and file. 3 Among the distinctive marks carried by many of the Spanish leaders from their founding moment in the late Franco years is the relative disinterest shown by many in pressing workers within their firm to formally join a union and pay dues. During the Franco period a remarkably high level of opposition protest was possible with relatively little permanent clandestine organisation. This tradition helps to explain the extremely low level of union membership in Spain today, even in many factories where union representatives are elected and strikes take place.4 Thus in both Spain and Italy many leaders-and the overall shape of the available set of leaders-are strongly influenced by foundational experiences. Powerful collective experiences help to create new possibilities for collective worker action, which live on, in part, through workplace leaders and their institutions. These founding moments may, at the same time, close off other possibilities, such as high union membership in Spain. In either case, these experiences shape the workplace leaders' actions and beliefs. Thus the enduring legacy of the foundational experiences cannot be fully understood without some attention to the plant-level leaders, individuals for whom these decisive moments remain a crucial point of reference 1See Mershon (1990A). These data, we should keep in mind, refer only to a subset of the unionists who witnessed the Hot Autumn: those council leaders holding office in 1982. A study of the Italian Communist Party has yielded similar findings: the phase of mobilisation that peaked in the Hot Autumn pulled forward a distinctive generation of activists and leaders in the PCI, and members ofthe generation at different institutional locations had different experiences during the cycle of conflict. See Hellman (1988, pp. 60-61, 131-35, 181-90. 21ndeed it is possible to argue that, for some leaders, role conceptions and related beliefs are forged in part by foundational experiences, and that these beliefs, in turn, guide the leaders' behaviour. For other leaders, not participants in founding moments, social bonds and institutional demands exert strong influences on beliefs. 3 See Mershon data discussed above and Mershon (1986), eh. 3. 4 See Fishman (1990), chs 4, 6. Some Spanish factories, of course, do have high levels of union membership. The point is that some workplace leaders, influenced by the experience of the opposition period, see little need to encourage union membership in order to initiate collective worker action and represent workers. Workplace leaders and labour organisation 83 and inspiration long after the society-wide impact seems to have been absorbed and faded. 4. 2. Allies, adversaries and indifference: the limits posed by the political context and climate of opinion Plant-level leaders in divergent settings-even if they share similar outlooks-will tend to act somewhat differently. Among the factors that shape the possibilities and constraints faced by the leaders, one must include the political environment at the national, regional and local levels. 1 Students of unionism have long noted the importance of 'external' climates of opinion.2 Political climates or contexts afford workplace leaders a broad sense of possible and appropriate demands or collective efforts. Political contexts differ, as well, in the relative strengths of political actors, the resources held by social actors (including their collective memories and the reach and density of their networks), and the prevailing patterns of alignment among political and social actors. Thus the actions of the workplace leaders will be affected by the allies available to them, the adversaries they confront and the indifference they encounter, even at times from former allies. The ties (and at times tensions) between unions and political parties constitute a common theme in labour movement activity across otherwise very different national political settings. Unions, of course, rely on parties for access to political power, a crucial issue for labour given the importance of legal regulation of union behaviour and governmental decisions on economic policy. Thus the political party system at the national, regional, and local levels-and even at the plant level in countries such as Italy with capillary party organisation-helps to define the opportunities available to the workplace leaders. The leaders may be influenced not only by the traditional stances toward labour of the relevant political parties and by the balance of power among the parties, but also by shifts in the parties' perspectives on labour. 3 The preference of many social democratic parties for nationwide arrangements between labour and capital has been widely discussed by social scientists. Until recently scholars have devoted less attention to the growing disinterest in the working class exhibited by many parties formerly-and still formally-allied to labour. Among the best current examples is the Socialist union confederation in Spain, the Union General de Trabajadores (UGT), which is routinely at odds with the government of the historically allied Partido Socialista Obrero Espaiiol (PSOE). More often than they would like, today's plant-level leaders find themselves without strong allies in the political party world. It could be argued that left parties have distanced themselves from manual workers-and from unions, the organisations they view as most clearly speaking for manual workers-as a response to changes in the class structure of advanced industrial countries. Industrial manual workers have declined in number or at least 'On the importance of local political alignments and subculmres in shaping union behaviour, see Golden (1988), Mershon (1986), eh. 12, and Locke (1990). 2 For a good discussion of the importance of the political climate of opinion, see Lipset, Trow and Coleman (1956), especially eh. 16. 3The fact that many workplace union leaders are themselves party members or activists may heighten their perceptions of the shifts. It would be interesting to investigate whether those leaders involved in parties are more likely than other leaders to find shifts reasonable and acceptable or, on the other hand, more likely to find the shifts troubling and difficult to accept. 84 R. Fishman and C. Mershon relative weight throughout Western Europe and North America; and left parties, impelled by electoral concerns and perceiving difficulties in balancing the demands of manual workers and others, have opted to de-emphasise their historic ties to industrial workers. 1 However one understands the causes of the parties' behaviour, this does not diminish the implications for plant-levelleaders-and for labour more broadly-of the fraying alliance between unions and left parties. A long and distinguished theoretical tradition has focused on ties between the union movement and intellectuals. Writers of widely different political and theoretical schools such as Gramsci, Perlman, Michels, Schumpeter, and Aron have underscored the importance of ties between intellectuals and the working class, especially working class organisations. Intellectuals have been viewed as contributors to the strategic vision of labour, to labour's organisational development, and to its self-confidence or morale. Intellectuals have given labour much encouragement and on occasion skills and theoretical orientations. They have furnished labour with a series of links to the world outside the working class and have helped to legitimate and draw attention to the actions of labour. Intellectuals, according to some analysts, have even endowed the working class with 'myths'. Italy and Spain number among those countries where bonds between intellectuals and the working class have been quite salient. A visionary leftist might well have complained in the 1970s that the links between the intellectual community and the working class, certainly in Spain and perhaps in Italy, did not suffice as a basis for working class hegemony. Nevertheless, in the 1970s, the collective benefits for the labour movement derived from these links-and the direct ties to the intellectual world enjoyed by many workplace leaders-were quite real. At a minimum, those Left intellectuals who wrote about the working class, spoke in union halls, invited unionists to speak in university settings, and so forth, promoted a sense of the importance of the labour movement among many workers and others outside the working class. Their efforts improved the morale of labour and reinforced its links to the larger political community. These bonds may well have indirectly aided the process of union leadership recruitment and they may have encouraged labour to stake out a position on comprehensive national questions. The ability of plant-level leaders to employ enthusiastically an abstract language of global political objectives is also due in part to the historic ties between intellectual and working class communities. After the 1970s, in many countries, including Italy and Spain, the bonds between intellectuals and the working class seem to have entered into crisis; and in some settings an already existing crisis has become more acute. The ties between the two groups have eroded significantly, so that some observers even refer to a 'divorce' between the working class and Left intellectuals, now more interested in other social forces. 2 Local union leaders have lamented the intellectuals' loss of interest and the growing tendency for erstwhile allies to view the unions as an 'obstacle'.3 10n left parties' search for support among manual and non-manual workers, see Przeworski and Sprague (1986). See also Lemke and Marks (1992). 20n the erosion of ties between the two groups in Spain, see Di Rosa (1986). 3Fishman's interview with local labour official, Cornella, Barcelona province, June 1985. Workplace leaders and labour organisation 85 Increasingly, the plant-level leaders, whose work was never easy, find themselves without the allies and the 'myths' on which they have traditionally depended. This loss of allies and supporting 'myths' in the intellectual and political party communities may limit the types of activity workplace leaders are inclined to undertake, but it is difficult to draw firm conclusions without research focused specifically on this issue. In any case, the political context in the broadest sense, which we understand to include political parties as well as intellectuals, has clearly shifted against labour in many national settings. Although the old founding moments live on in the commitments of leaders and activists, the constraints on these actors have changed-particularly with respect to the configuration of allies, adversaries and indifference they encounter. The change in the larger political environment does not alter the basic tasks performed by the workplace leaders or the indispensability of their role, but its new configuration does impinge on their activity and limit their success. 4.3. Social ties within the workplace: resources available to the leaders The social ties available to the leaders within the workplace itself provide the plant-levelleaders with another set of opportunities and constraints. All workplace leaders must rely on social ties to mobilise workers and co-ordinate action, but the scope of union-centred networks varies greatly from one plant to another. 1 Moreover, social ties cannot simply be willed into existence, no matter how dedicated a leader may be. Everything else being equal, leaders in smaller plants are more likely to reach the entire work force directly or indirectly through their networks; but everything else is not equal. The leadership supply problem leaves many small plants with no leadership at all, while many others must make do with relatively disinterested leaders or even 'leaders by default'. Leaders by defaultthose who win positions of representation simply because they are uncontested and not because of any genuine social support or base within the workplace-are more likely to be found in relatively small workplaces. As Mershon's research in Italy has suggested, these leaders by default have a more limited ability than do other leaders to mobilise worker participation and to secure concessions from the employer.2 Where leaders are 'disconnected' from the rank and file, their ability to mobilise workers and win concessions is sharply curtailed. This isolation ofleaders by default is in no sense a function of their ideological stance or even of their level of commitment. Rather it is a question of their position-and the position of union structures-in the social networks of the workplace. In a plant with competing leaderships no single leader is likely to fully dominate the workplace through social bonds, for alternative networks may be available to workers. But in such a competitive setting, a leader must enjoy a solid social base in order to rise as a representative. Only in a firm without competition among leaderships can representatives emerge by default, with restricted social resources or support and with highly limited opportunities for effective action. In some non-competitive settings, however, leaders may enjoy a great ability to mobilise worker participation: where 1An excellent study of workplace unionism that stresses the importance of social networks in the work of shop stewards is Batstone, Boraston, and Frenkel (1978). 2This analysis builds on Mershon's experience and findings; see (1986, pp. 121-24, 250-53, 394-95). 86 R. Fishman and C. Mershon collective efforts or social bonds place an uncontested leader within a social network of great reach in the workplace, the leader's capacity for effective action is considerable. 5. Conclusion 5.1. Recapitulation of the argument Unions acquire a concrete, immediate presence in workers' lives through the efforts of leaders at the workplace. Both the external political context and the internal social context help to establish the opportunities and constraints for the workplace leaders' activity. The leaders depend on-and draw a sense ofpurpose from-bonds of interest and identity among workers as well as alliances between workers and other groups. The world the leaders encounter, both in the firm and in the wider community, helps to establish the foundation for the success or failure of their initiatives. As we have shown, leaders are absent from numerous workplaces and in other work settings the available leaders represent only a portion of the spectrum of union alternatives with national visibility. Yet without some union leadership at the workplace level, organised labour is severely handicapped in its attempts to mobilise workers, to discern, mediate, and defend workers' concerns, and to assess and implement collective agreements. The workplace leaders form an essential layer of union organisation, distinct both from workers and from officials higher up in the union hierarchy, and, at the same time, they build a bridge between specific, prosaic problems at the workplace and the comprehensive, long-term aims of the entire union movement. Patterns of political and social life that extend well beyond the workplace-as experienced by the plant-level leaders in their efforts on behalf of labourpowerfully shape the possibilities for workers to attain victories in their collective struggles. Changes in those larger patterns may not depend on the commitments, or even the efforts, of the leaders. But in our view the impact of those broader changes can best be understood when we direct our attention to the way they impinge on the workplace leaders without whom no co-ordinated collective worker action can take place. 5. 2. Implications of the analysis If the argument we have developed is valid we should be able to demonstrate clear implications of our analysis. With this in mind we now briefly consider two very different types of implications: suggested lines of research and more broadly applicable interpretive lessons. 5.2.1. Research questions. Although we do not contend that research on plant-level leaders is of any greater importance than the study of other aspects of labour movement activity, we do claim that some significant questions can only be resolved through enquiry focused on those filling the workplace leadership role. We identify here a few of the basic questions that can be explored through studies of the sort we propose. Workplace leaders and labour organisation 87 One set of questions concerns the supply of workplace leadership. First, how successful are unions-and sectors of opinion within unions-in overcoming the fundamental challenge of leadership shortage? To what extent do the publicly visible union alternatives actually have available to them plant-levelleaders present in workplaces throughout the economy? To restate the point: Which publicly visible union options are significantly hampered in their development by a lack of leaders inside plants? Second, is union success in meeting the leadership supply problem a simple reflection of a broader ability of the union to appeal successfully to workers or is it (as we suspect to be the case normally) an independent question? We would expect to find significant differences between unions or informal sectors of labour opinion and among national or local settings in the extent to which the leadership supply problem is resolved, or on the other hand remains unmet, thus forming a major impediment to union development. Only once these questions have been posed will it be possible to ask as well: To what degree does success or failure in confronting the workplace leadership supply problem influence other dimensions of union success or failure? In a similar fashion the other issues we identify can generate specific researchable questions. To what extent do the attitudes ofworkplace leaders differ from those of the rank and file and the union higher-ups? To what degree do the plant-level leaders implement national union policy or transform it? How active and effective are the leaders in attempts to strengthen workplace union organisation and encourage membership? What is the position of the leaders in the social relations of the workplace? To what extent are presumed macro-level influences on labour actually experienced by workplace leaders, and how do such experiences correlate with plant-to-plant variations in union behaviour? These and other questions are suggested by our emphasis on plant-level leaders. 5. 2. 2. Conceptual lessons. We do not wish to suggest that all research on labour take up the agenda we outline. Indeed, one of the strengths of the field of labour studies is the variety of research agendas pursued by scholars of unionism. But in our view two broad elements of our argument could be usefully incorporated into the reasoning of those students of labour who follow other strategies of enquiry. First, we insist on the distinctiveness and potential autonomy of differing levels of union activity: the top-level leaders with nationwide visibility; the functionaries charged with co-ordinating local union activity; the plant-level leaders required to establish a formal union presence in the firm; the workplace activists who may provide much of the liveliness and colour of union life but who are virtually impossible to find in the absence of a plant-levelleadership willing to establish formal union life; and the rank and file whose heterogeneous views cannot be extrapolated from observations made at other levels. Second, we emphasise the inherently problematic nature of the plant-level leadership and the work it does. The existence of the workplace leaders cannot be taken for granted. Moreover, we cannot assume that specific national strategies will be effectively carried into the firm by workplace leaders. These two essential points have the potential to reorient significant interpretations of labour's collective efforts. 88 R. Fishman and C. Mershon Let us take up the example of an important recent contribution by Wolfgang Streeck. In a highly stimulating and forcefully reasoned rejoinder to Claus Offe and Helmut Wiesenthal (1980), Streeck (1990) contends that workers organise in their collective defence more easily than capital; employers, Streeck argues, actually face the more daunting barriers to self-organisation. As support for this assertion Streeck cites cross-national data showing that in most national settings business forms a greater number of separate associations whereas labour forms a smaller, perhaps even a much smaller, number of organisations. Thus labour organisation is more unified than the self-association of capital. This pattern is taken as proof that labour's organisational capacities are greater than are those of capital; the objective of associational unity across sectors is typically more nearly attained by labour than capital. Our argument provides a clear basis for a sharply different interpretation of the cross-national finding. Once we understand labour organisation to be highly constrained by a leadership shortage inside the workplace we can appreciate that an associational configuration-the presence of a greater or lesser number of competing unions-is the product in part of the weakness of some potential organisational competitors. In practice, national settings have at most a few nationwide confederations competing with one another. The outcome reflects in part the inability of those with alternative viewpoints to find a labour leadership in a sufficiently large number of workplace settings to gain nationwide credibility. It is highly likely that some genuine sectors of opinion within the working class-one thinks, for example, of conservative workers-find no distinctive organisational expression precisely because there are virtually no workplace leaders willing to organise an autonomous organisation resting on their beliefs. And some relatively visible and ideologically distinctive unions, such as Spain's Maoist confederations (in the first years of the new democracy), fail to establish themselves as serious national competitors partly because their workplace presence is highly limited given their tremendous leadership shortage. To reiterate: Organisational success does not flow automatically from atiitudinal predispositions among workers. Nor is success a simple consequence of the strengths of the victors. Typically it reflects in part the weakness of potential competitors. The shape of labour organisation is highly constrained by a workplace leadership shortage, which limits the number of viable associational players. Thus labour's superior attainment of associational unity reflects in large measure the enormous challenges and difficulties faced by labour organisation. In our view, the data presented by Streeck fit rather well the spirit of Offe and Wiesenthal's claim: that workers face distinctive difficulties in organising their collective defence. Many other conceptual lessons could be drawn but we do not intend to provide here an inventory of implications. Any such list would run the risk of being at once too long and too short. Rather, we propose a general sensibility to the workplace leadership in the hope that this sensibility may bring students of labour to draw new conclusions and lessons, many more than we ourselves could enumerate. 1 1For helpful comments, we would like to thank Peter Bearman, Gerry Berk, Steve Hellman, John Kautsky, the late Barbara Salert, John Sprague, Samuel Valenzuela, and two anonymous referees. We also thank Juan Linz for strongly influencing and encouraging the research that underpins this analysis. Workplace leaders and labour organisation 89 References Alien, C. 1990. Trade unions. Worker participation, and flexibility: linking the micro to the macro, Comparative Politics, April Batstone, B., Boraston, I. and Frenkel, S. 1978. Shop Stewards in Action, Oxford, Basil Blackwell Carew, A. 1976. Democracy and Government in European Trade Unions, London, George Alien and Unwin Clegg, H. 1976. Trade Unionism under Collective Bargaining, Cambridge, Cambridge University Press Coi, S. 1979. Sindacati in Italia: iscritti, apparati, finanziamento, Il Mulino, March-April Collier, R. 1986. Historical founding moments in state-labor relations, Brazil, Chile, Mexico, and Venezuela. Paper presented at the World Congress of the International Sociological Association, New Delhi, India Collier, R. and Collier, D. 1991. Shaping the Political Arena: Critical Junctures, the Labor Movement, and Regime Dynamics in Latin America, Princeton, Princeton University Press Di Rosa, P. 1986. The disenchantment of Spanish intellectuals: reflections on the relation between intellectuals and workers in contemporary Spain, Social Studies Honors Thesis, Harvard University Fishman, R. 1990. Working Class Organization and the Return to Democracy in Spain, Ithaca, Cornell University Press Fishman, R. The union movement in democratic Spain, in Valenzuela, J., (ed.), Labor Movements in Transitions to Democracy, Notre Dame, Indiana, University of Notre Dame Press Golden, M. 1988. Labor Divided, Ithaca, Cornell University Press Hellman, S. 1988. Italian Communism in Transition: The Rise and Fall of the Historic Compromise in Turin, 1975-1980, Oxford, Oxford University Press Iriso Napal, P. 1992. Sistemas de negociaci6n colectiva y acci6n sindical: sindicatos y trabajadores en la empresa, Doctoral Dissertation, Universidad Complutense de Madrid Korpi, W. 1978. The Working Class in Welfare Capitalism: Work, Unions, and Politics in Sweden, London, Routledge and Kegan Paul Korpi, W. 1981. Unofficial strikes in Sweden, British Journal of Industrial Relations, March Lange, P. 1984. Union Democracy and Liberal Corporatism: Exit, Voice and Wage Regulation in Postwar Europe, Western Societies Program Occasional Paper No. 16, Center for International Studies, Cornell University Lemke, C. and Marks, G. 1992. The Crisis of Socialism in Europe, Durham, N. C. and London, Duke University Press Linz, J. 1978. Crisis, Breakdown and Reequilibration. Vol. 1 of The Breakdown of Democratic Regimes, in Linz, J. and Stepan, A., (eds), Baltimore and London, Johns Hopkins University Press Lipset, S. 1960. Political Man. Garden City, New York, Doubleday Lipset, S. and Rokkan, S. 1967. Cleavage structures, party systems and voter alignments, in Party Systems and Voter Alignments, New York, Free Press Lipset, S., Trow, M. and Coleman, J. 1956. Union Democracy. New York, Free Press Locke, R. 1990. The resurgence of the local union: industrial restructuring and industrial relations in Italy, Politics and Society, vol. 18, no. 3 Lorenzini, L. 1982. Sindacato senza democrazia, Azimut, September-October Ludevid Anglada, M. 1979. El sindicato dentro de la empresa, Barcelona, Editorial Blume Mann, M. 1973. Consciousness and Action among the Western Working Class, London, Macmillan Mershon, C. 1986. The micropolitics of union action: industrial conflict in Italian factories, Doctoral Dissertation, Yale University Mershon, C. 1989. Between workers and union: factory councils in Italy, Comparative Politics, January 90 R. Fishman and C. Mershon Mershon, C. 1990A. Generazioni di leader sindacali in fabbrica: L'eredita' dell'autunno caldo, Polis, August Mershon, C. 1990B. Relationships among union actors after the Hot Autumn, Labour, Spring Michels, R. 1962. Political Parties: A Sociological Study of Oligarchical Tendencies in Modern Democracy, New York, Free Press Offe, C. and Wiesenthal, H. 1980. Two logics of collective action: theoretical notes on social class and organizational form, in Political Power and Social Theory, Vol. 1. Greenwich, Connecticut, JAl Press Pirani, S., et al. 1991. Gli anni 80 del sindacato, Nuova Rassegna Sindacale, September 30 Przeworski, A. and Sprague, J. 1986. Paper Stones: A History of Electoral Socialism, Chicago, University of Chicago Press Regalia, I. 1980. Delegati e consigli di fabbrica nelle ricerche degli anni '70, Annali delta Fondazione Luigi Einaudi, Vol. XIII, Torino, Einaudi Regalia, I. 1984A. Democrazia e sindacato: preliminari per una riftessione critica, Il Mulino, November-December Regalia, I. 1984B. Eletti e abbandonati. Modelli e stili di rappresentanza in fabbrica, Bologna, 11 Mulino Romagnoli, G. and Della Rocca, G. 1982. 11 sindacato, in Cella, G. and Treu, T. (ed.), Relazioni industriali, and Treu, T. Bologna, 11 Mulino Sabel, C. 1981. The internal politics of trade unions, in Berger, S. (ed.), Organizing Interests in Western Europe, New York, Cambridge University Press Smuraglia, C. 1975. Enciclopedia dei diritti dei lavoratori, Milan, Teti Stephens, J. 1979. The Transition from Capitalism to Socialism, London, Macmillan Streeck, W. 1990. Interest heterogeneity and organizing capacity: two logics of collective action, Working paper, Centro de Estudios Avanzados en Ciencias Sociales, Instituto Juan March, Madrid. (Also published as Imprenditori e sindacati: eterogeneita' degli interessi e capacita' organizzativa, Stato e mercato, April, 1991 Tarrow, S. 1989. Democracy and Disorder: Protest and Politics in Italy, 1965-1975, Oxford, Oxford University Press Tilly, C., Tilly, L. and Tilly, R. 1975. The Rebellious Century, Cambridge, MA, Harvard University Press UGT. 1981. Elecciones Sindicales de 1980: un Primer am:ilisis. Madrid, UGT Valenzuela, J. 1981. Uno schema teorico per l'analisi della formazione del movimento operaio, Stato e Mercato, vol. 3 Van de Vall, M. 1970. Labor Organization, Cambridge, Cambridge University Press Vega Garcia, R. 1989. El movimiento obrero gijones en la transicion: el fenomeno de la corriente de izquierda. MS., Department of Contemporary History, University of Oviedo Zeitlin, J. 1989. 'Rank and filism' in British labour history: a critique, International Review of Social History, vol. 34.  Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Workplace leaders and labour organisation: limits on the mobilisation and representation of workers Creator An entity primarily responsible for making the resource Robert Fishman Carol Mershon Source A related resource from which the described resource is derived International Contributions to Labour Studies 3, pages 67-90 Date A point or period of time associated with an event in the lifecycle of the resource 1993