Union effectiveness: In search of the Holy Grail

Introduction

Over the last four decades, union power and influence to protect and advance members’ interests have declined in most developed economies. The effect of this decline has been ever more keenly felt since the onset of financial crisis and economic recession in the global north. Consequently, the practical issue of union effectiveness has become ever more pressing and salient. The outcome of this decline within and outside the workplace has been that members’ living standards have fallen or stagnated while members have become ever more subject to the whims of employers and outcomes of increasingly deregulated labour, capital and commodity markets. Yet at present and for many years previously, the study of union effectiveness has been bedevilled by deficiencies, making it much more difficult for unions and their members to obtain the Holy Grail of union effectiveness.

The inadequacy of the academic contribution to understanding union effectiveness takes a number of forms. First, there has been lack of sustained and detailed attention per se, even a ‘dearth’ (Pyman, 2002), with it being clear that a working normative assumption operates whereby definitions of measures and outcomes of union effectiveness are seen as being rather self-evident, even unproblematic (see Burchielli, 2004: 337). Examples of those unspoken and implicit measures and outcomes are union density, bargaining coverage or wage settlements, making these partial studies at lower level units of analysis. To illustrate further, Sullivan (2010) suggested that a ‘density bias’ has impeded the identification of strategies for attaining union renewal, and by implication, union effectiveness. And, although hundreds of union wage effect studies have been published, there are apparently no more than 20 published studies that explicitly study union effectiveness itself. Many of these are relatively recent, and are concerned with the subset of organizing effectiveness. Second, where union effectiveness studies have been carried out, they have often lacked rigour and completeness. This we attribute to their insufficient conceptualization of (union) effectiveness. Consequently, scholars have typically concentrated narrowly upon collective bargaining or organizing effectiveness or (internal) organizational effectiveness. Measures have captured aspects of local or national union activity but seldom the totality of national unions and union movements (and especially so between and across nations and economies). While at the operational level a narrower focus may be appropriate, our objection here is mainly in regard to the conceptual level. Third, where studies have been carried out on the processes by which unions undergo renewal and revitalization, union effectiveness has been a surprising omission. The aforementioned deficiencies are heightened when the voluminous prognoses for union revival, renewal and revitalization (like social movement unionism and ‘organizing model’ unionism) are also taken into account.

Overall, relatively little advance has been made in theorizing union effectiveness since the contributions of Hammer and Wazeter (1993) and Fiorito et al. (1993). Therefore, collectively the import of these deficiencies is that the ‘bigger picture’ of union effectiveness is somewhat neglected. Put another way, the ‘wood’ cannot be seen for the ‘trees’, for the means (like union density) and ends (like greater worker control or higher wages) have no obvious and rounded connection. We believe that a grounded conceptualization of union effectiveness can provide the necessary connection. In order to navigate through the dialectic of agency and environment, the flow(s) of intention, action, process and outcome in relationship to union effectiveness must be identified and theorized. Our thrust, then, is to identify the conceptual and theoretical maladies and put forward a union effectiveness framework for addressing and resolving them based upon a more complete conceptualization. It is hoped that the value in this exercise is not merely academic, for unions as actors require clarity of thought to guide their actions and labour-orientated scholars have a role to play in helping that clarification of thought.

We begin by characterizing unions in order to establish the flow of intention, action, process and outcome by starting with the ‘big picture’ of unions under capitalism. We should acknowledge that our main focus is upon unions operating within (rather than across) nations, and more specifically unions within advanced western democracies and economies of the global north. Although much of this analysis may apply to specialized global union federations, unions at transnational levels such as the EU, or to general international federations and attempts to establish other forms of inter- and transnational worker solidarity, some points may not and there is clearly much that is relevant to such bodies that we do not consider. First, we outline a multifaceted conceptualization of union effectiveness before assessing and critiquing the literature and then returning to more fully elaborate our conceptualization and framework. As we proceed, we state the various propositions derived from the discussion.

Unions as worker organizations

Unions are of, by and for workers. ¹ This defines their character, means, purpose and interests. These features are all linked together by the goals of defending and improving members’ living standards (primarily through their paid employment) and members’ control over their working lives and conditions (primarily centred upon the experience of work in the workplace). Consequently, the first order of analysis for union effectiveness necessarily concerns outcomes, while the second order concerns the resources and processes by which to achieve outcomes. The resources comprise members’ activism, employed staff and financial resources as well as members’ collective power inside and outside the workplace. Alliances with other unions, community, religious or other groups may provide additional resources. The processes comprise collective bargaining, political action and mutual aid. In this logic, power is primarily a means to an end, and power itself neither specifies chosen outcomes nor guarantees effectiveness in gaining those outcomes. This is because the application and exercise of power – as we term it levers of power – must be specific to the situation and circumstance of the goal being sought and its temporal and spatial context.

The important rider to this logic is that acquisition and generation of power (and allied resources) is often an implicit goal for unions given that it is the critical modus operandi (see Lévesque and Murray, 2013: 779). Within unions as collective organizations, the generation of the ‘power to’ function – in regard to their members – is the best guarantor of generating the ‘power over’ function (vis-a-vis employers, government, etc.). Equally as important at this stage in establishing a framework for beginning to conceptualize and theorize union effectiveness is recognizing that the exercise and application of power – in terms of strategy and tactics – heavily influences whether, how and to what extent goals are attained. It is easy to envisage how an un-powerful union is also an ineffective union. It is less easy to imagine a prima facie powerful union that is also ineffective but such a union is, nevertheless, quite possible because the application and exercise of power is a critical variable in explaining whether union effectiveness is attained. Moreover, there is a seemingly inescapable logic that particular levers of power are best suited for use in specific arenas to attain certain goals. For example, generally, economic levers of power are best exercised in the economic arena for the attainment of economic goals. Meanwhile, political levers of power are best suited to and exercised in the political arena for the attainment of political goals. Matters are not always so clear-cut, of course. As Sturmthal (1973) noted, unions may resort to political means for economic ends when the state or the economic situation blocks union exercise of economic leverage.

To this needs to be added the context and situation represented by the system of capitalism in and under which unions exist. Unions are a reflection of the imbalance in power between capital and labour as well as representing an attempt to redress this in regard to the prosecution of workers’ interests. The power imbalance is most easily understood in terms of what flows from capital owning and controlling the means of production, distribution and exchange, namely, a superiority of economic, political, ideological and organizational resources. This sense of being the weaker party is reified by unions being ‘secondary’ (Offe and Wisenthal, 1985) and ‘intermediary’ organizations (Muller-Jentsch, 1985). Yet it is important not to ignore that unions are, nonetheless, active, conscious and independent collective agents which have special features by virtue of being voluntary and democratic organizations, based upon membership participation and whose primary resource is the collective activity of their members in the workplace (collective bargaining) and in public spaces (lobbying, political protest). As Clark (2009: 9–11) suggested, volunteer activism is critical to unions and the one factor determining effectiveness that unions can most influence.

As a result of the dependency upon voluntary membership participation (for legitimacy and mobilization), it may be expected that unions will be somewhat disadvantaged in achieving effectiveness relative to other organizations like businesses. Although conceptually we can speak of how well unions and businesses achieve their respective goals, we avoid comparing effectiveness between unions and businesses due to their widely differing goals. Unions lack most of the coercive and incentive power over volunteers that employers hold over staff, even while holding similar power over their own staff. Unions also may be somewhat constrained relative to other organizations in exercising their incentive and coercive powers over staff due to ideological and public relations considerations, e.g. it would be embarrassing for a union to criticize employers for relying on incentive pay schemes while using such schemes for its own staff. On the other hand, it may also be expected that with the legitimacy gained through democratic processes and commitment to social justice, unions may be advantaged in some regards relative to organizations such as businesses. Yet, as Chaison et al. (1993) noted, union legitimacy may have declined for several reasons, including autocratic leadership perceptions. Thus: ‘nonunion workers commonly believe that unions and their officers have an agenda that is not drawn from the needs and desires of the membership’ (Chaison et al., 1993: 151). More recently, the overall decline in political and civic participation by citizens has further compounded the problems for unions generating resources from their members as activists.

These aforementioned components constitute the foundation for our critique of the salient literature and the laying out of a definitional framework of union effectiveness. The preceding suggests four major propositions about union effectiveness that are to varying degrees definitional:

It is not enough to ask whether unions win improvements for members or for employees generally. Even though members may at times be willing to forgo having a say in their unions for the sake of material benefits, norms about unions in western democracies demand democratic unions. Hence, we are saying that even though members may regard an autocratic leadership and the union representation it provides as ‘effective’, we cannot consider such a union to be effective because it is not socially acceptable in societies that expect democratic unions. As an extreme analogy, we would be hesitant to call ‘effective’ a union that achieved its goals through the use (or threat) of violence, blackmail and intimidation.

Unions that make more efficient use of their resources (internal functional efficiency) are able to exert greater external influence on employers, governments and others from a given resource base. Along with Proposition 1, this proposition recognizes that unions simultaneously face administrative imperatives to be a ‘town meeting’ and representative imperatives to be an effective fighting force (‘army’) (Muste, 1928).

Certain factors, such as membership decline, may lead unions to refocus on community issues or consumer service provision. These may in some instances be useful supplements to a focus on work-related issues. There is a potential ‘tipping point’, however, where unions’ focus essentially ceases to be sufficiently work-related to warrant considering the entities to be (labour) unions. It becomes something else and may be effective or ineffective but it can no longer be considered a union.

Members provide the resources to empower their unions most obviously in terms of providing dues (subscriptions) that hire staff and compensate officers for time spent on union work. A second form of collectivization is the volunteer efforts of lay members by which much union work is conducted. Recent studies have focused solely on the resources provided by volunteers as per the ‘organizing model’. While this is an understandable reaction to an to overreliance on ‘servicing’ whereby volunteerism was eschewed in favour of officer and staff expertise, the more fundamental issue Proposition 4 stresses is that collectivization of member resources can take two basic forms, financial and human.

Union effectiveness

The first four propositions above define union effectiveness to some extent, but union effectiveness is not easily defined, and a more elaborate (albeit still rudimentary outline) is needed to guide our critique of the union effectiveness literature. Union effectiveness is a ‘highly complex, highly dynamic concept’ (Clark, 2009: 5). With that caution noted, Clark (2009: 5–8) proceeded to define it in terms of: (1) ability to organize new members; (2) collective bargaining outcomes; and (3) impact on society, quickly adding ‘the extent of democracy in a union’ as ‘an additional consideration that some believe is a measure of union effectiveness’. We define union effectiveness in broader terms, namely, of unions successfully applying appropriate levers of power to gain their purposeful goals. This is then an issue of performance set against desired objectives (with the rider of democratic participation). By appropriate, we mean applying the salient amount and type of power in the apposite arena. Consequently, union effectiveness in terms of both procedural and substantive outcomes involves creating deliberate and usable leverage with employers as well as state authorities in pursuit of objectives. Thus, this definition of union effectiveness is not merely about defining institutional effectiveness (i.e. a union as organization) and neither does it merely concern itself with collective bargaining or the union operating as a political actor. Consequently, the aspects of union effectiveness deployed in assessing the extant literature are:

This is a more highly specified definition than that offered by Hammer and Wazeter (1993: 303–305) because it steadfastly connects means with ends. For the moment, we are not concerned with the internal democratic processes by which union goals are decided so long as they do not become incoherent and contentious, thus leading to the dissipation of union resources required to achieve union goals (cf. Hammer and Wazeter, 1993: 304). In other words, we take union goals as a given and focus upon their attainment. That said, we acknowledge that more militant goals will require greater membership mobilization and leverage (Kelly, 1996) but we are not concerned that, in the current period, there is a mismatch between means and ends whereby desired maximalist ends (as per union goals) are vastly out of kilter with the reduced resource means. This discussion of union effectiveness suggests:

Recalling that internal union democracy is vital, our focus now shifts to a more concrete perspective in that union effectiveness on the operational level requires the union to exert external influence. The exercise of certain levers (e.g. economic, industrial, legal, political) is more effective in appropriate arenas (e.g. economic, industrial, legal, political).

Employers, politicians and governments do not heed ‘memberless’ unions, succinctly conveying why many unions heavily focus on recruiting and organizing in an era of steady union decline. Recalling Proposition 4, unions usually need membership to provide resources (although, one alternative is state funding under populist corporatist regimes like those of Peron in Argentina). As a practical matter, attrition through retirement, job leaving, etc. requires that unions continually act to acquire new resource providers. As noted elsewhere, organizing (and recruiting and retaining) could be considered a supra-strategy for unions in that it is essential for influence in any arena.

Literature review

We adopt a ‘Goal-System’ framework in critiquing previous work. In it, goal definition is the fundamental starting point for any study of assessing union effectiveness, no matter its particular aspects and subsets. This framework builds on earlier work by Fiorito et al. (1991, 1993), which in turn draws heavily on previous work by Bok and Dunlop (1970) on unions and by organizational theorists on effectiveness (e.g. Cameron and Whetten, 1983), among others. Thus, ‘[e]ffectiveness [concerns] … the degree to which goals are attained’ (Fiorito et al., 1993: 113). Goals – in terms of the constituency of interests they are intended to serve – are either sectional or class-based (albeit it should be acknowledged that intention and outcome may diverge). Union goals are essentially of economic, social and political natures, and some can also be achieved by linked but, nonetheless, separate organizations of workers such as friendly or mutual aid societies, cooperatives, or labour parties (Fiorito et al., 1993: 112–115). While this is uncontroversial, what is far more so is identifying the specific measures by which to judge union effectiveness. In studying union strategic choice in New Zealand, Boxall and Haynes (1997: 571) suggested:

Regardless whether or not workers are instrumentally and/or ideologically motivated, … unions are effective when they meet the expectations of their members in respect of equitable outcomes in these areas [better pay and conditions, protection against arbitrary management decisions, etc.] … Meeting these ends … depends on effective management of the primary or critical means of a union: its mode of engagement with employers.

This is one of the few explicit definitions offered. But it is too workplace – and economic – based (as Hammer and Wazeter [1993: 302] would no doubt concur), albeit a caveat on the importance of the state was subsequently entered into (Boxall and Haynes, 1997: 572). Although of use, we find other contributions similarly incomplete, such as those definitions offered by Hickey et al. (2010), whereby effectiveness in terms of outcomes is based upon membership participation, membership growth and union influence; by Rose and Chaison (1990), who in analysing organizing effectiveness, focused upon certification wins rate, number of workers covered and gaining a first contract (but not the quality of it); and by Bryson (2005), who stipulated union effectiveness comprises organizational (accountability to members) and bargaining (delivering improvements in work and working conditions) effectiveness.

In examining ‘the extent to which union strategies are associated with a union’s organisational effectiveness and membership decline’, Gahan and Bell (1999) did not put forward a concept of effectiveness (union or union organizational) or a framework for judging it. Instead, they used dependent variables to represent effectiveness (Gahan and Bell, 1999: 17–20), apparently regarding their validity as self-evident. One was a three-item scale for union effectiveness based on convention delegate responses to survey items (e.g. ‘This union is in a stronger position than most other unions’). The second was a measure based on government data on industry-level changes in union density. They used several delegate survey-based indicators of union strategies as predictors. Gahan and Bell stressed their findings for a ‘membership orientation’ strategy variable, the only predictor yielding clearly consistent results for the two dependent variables. Membership orientation had positive implications for delegate ratings of relative effectiveness and for industry-based density trends. In other words, ‘membership orientation’ was the only predictor that was positively related to perceived effectiveness by delegates and positively related to a measure based on government-based data on density.

Despite some criticism (e.g. Frege, 2002), the ‘Goal-System’ framework (Fiorito et al., 1993: 119–122) concerning national union effectiveness remains the most developed to date because it sets out a comprehensive framework of core penultimate goal criteria for evaluating union effectiveness. Fiorito et al. specified bargaining, politics, self-help, organizing, member solidarity and resource acquisition, and that these were complementary. Frege (2002) suggested that a focus on the process of influencing management needs to be added to these components. Similarly, Pyman (2002: 6) proffered that union effectiveness is ‘manifest in six general dimensions: … improved terms and conditions of employment; [membership] growth; internal democracy; organisational stability; activism and social movement unionism … [these being] typically expressed as union strategies’. Yet while suggesting ‘dimensions of effectiveness will invariably differ according to each individual trade union and across situational contexts’, she believed they represented a ‘core set of dimensions’ (Pyman, 2002: 6). However, no criteria or contextual framework was put forward for measuring and evaluating effectiveness, with the dimensions comprising various large and ill-defined components (of which culture, power and ideology were found in most).

By contrast, Burchielli (2004: 339–342), in research on Australian unions, outlined a union effectiveness typology with three – representative, administrative and ideological – elements of effectiveness. However, the components were not as distinct as claimed and lacked adequate internal specification for they pertained overwhelmingly to themes rather than criteria or goals. Representative effectiveness comprised recruitment, responsiveness to members and achieving key union goals; administrative effectiveness comprised structure and strategy, innovative practices, goal clarity, leadership and staff accountability; and ideological effectiveness comprised social values, ideological cohesion, developing active membership, union commitment and politicizing the environment. Very few of the components in themselves lead to effectiveness other than, for example, achieving union goals, suggesting that component parts and processes were missing. Moreover, there was no ordering of the components and no specification of their interrelationship. Nonetheless, Burchielli’s conceptualization was superior to most others. For example, Black et al.’s (1997) study of a single union did not offer a definition of effectiveness even though the concept was a focus of their article. They did, nonetheless, suggest size was no guarantor of effectiveness and evaluated a small union as being effective for the reasons of maintaining membership and a financial surplus as well as meeting membership expectations and needs and fostering a sense of membership-union identity (see also Sayce et al. [2006] on high levels of women’s participation in this union).

Within the literature, different writers make useful contributions. Yet they conflate the means and ends to the extent that ultimate goals and their attainment are not satisfactorily considered and exclude considerations of extra-workplace and political goals such as attaining political change through social democratic parties and their operation within the institutions of liberal democracy. Thus, extant definitions of union effectiveness lack both breadth and depth.

Returning to the ‘Goal-System’ framework, bargaining, politics, self-help, organizing, member solidarity and resource acquisition were criteria of penultimate goals, with ultimate goals being far less well specified (see Fiorito et al., 1993: 120), begging three key questions. First, what are the measures for evaluating attainment of the penultimate goals? Second, what are the ultimate goals? Third, how is attaining penultimate goals related to attaining ultimate goals? In regard to the latter, is it a simple mechanical relationship of gaining all penultimate goals that leads to the attaining of all ultimate goals? Are there intermediate mediators or moderators, primarily in terms of strategy and tactics? And finally, does innovation or the attaining of resources, for example, simply presuppose success in gaining an ultimate goal? Fiorito et al. (1997: 6) only briefly addressed the issue of the third question when arguing the six aforementioned components ‘were considered to be individual strategic elements, which when combined, would encompass union performance as a whole’.

More generally, this initial discussion brings forth a number of wider issues. First and foremost, are these measures (as per goals) to be specified by the unions themselves or by those external to them? (This is a different issue from which particular parties – political or otherwise and internal or external to the union – have influence in setting the union’s goals.) In most social science research, it is unquestioned that constructs can be informed by practitioners, lending them face validity, but constructs are defined and validated by scholars. Yet we see in the aforementioned research that scholars fairly often rely upon practitioner definitions (or provide vague definitions) as well as practitioner judgements. Are there then, in effect, subjective and objective measures? Will (or can) they converge? Subjects’ knowledge of nuanced circumstances can lead them to provide more meaningful albeit less objective assessments of union effectiveness. The same ‘bottom line’ percentage for membership growth may mean very different things in two sharply contrasting sectors. Subjective assessments potentially ‘net out’ differing circumstances. Although a general expectation of convergence seems reasonable, there are understandable reasons for divergence that are not simply measurement error. If measures are to be subjective, who are the subjects that give rise to them? Prior research used subjective views from employees in unionized workplaces (Bryson, 2005), union staff (Burchielli, 2004), convention delegates (Gahan and Bell, 1999) and national officers and staff (Fiorito et al., 1995), to name just a few possibilities. Second, should it be assumed that all unions have the same overall purpose but that this is expressed through different specific goals? The issues of sectionality, class and society loom large here (see, for example, Hyman, 2001). Third, and while accepting that some form of organizational survival is vital (even if not in its current form as a result of merger or amalgamation), is there to be an ordering of goals and, if so, what should this be?

Specification and ordering of goals do vary within unions and along vertical and horizontal axes. The ‘Goal-System’ framework deploys the concept of the ‘dominant coalition’ to seek to understand how the dominant group prevails and, thus, sets the goals. Even if the goals are not directly contested by other groups, they may prioritize and interpret them differently leading to variance (Fiorito et al., 1993: 118). Among these groups are members, activists, lay officials, officers (junior, senior) and internal political groupings and factions. Their goals may vary, as unions’ goals may do, in the extent of the ambition embodied in the goals – for example, moderate versus militant, or reformist versus revolutionary. In regard of goal ordering, temporal and processual aspects can be thought of – the former concerning short, medium and long-term goals, the latter ultimate and penultimate goals, and where some are operational (like efficiency, capacity building or resource acquisition) and have an instrumental use rather than being strategic and value-driven. For example, the goal of attaining ‘voice’ should not be seen as an ultimate goal – even if it is an important one for obtaining procedural justice – because it is more important what the outcomes of the ‘speaking to’ and ‘being listened to’ are, namely, whether desired actions are then taken by the employer or state. Similarly, a healthy state of union democracy is viewed as being desirable in itself but it is held to be more significant as an instrumental resource of legitimacy for the purposes of collective mobilization. The point is that whatever the decided goals, clarity is needed in whose goals are or are not being evaluated when assessing union effectiveness. Studies which were explicit on this score include Fiorito et al. (1997), focusing in this instance upon national union leader perception, Frege’s (2002) use of local leaders’ perceptions of influence, Burchielli’s (2004) respondents being union field officers and Bryson’s (2005) survey of workers and members.

Following from this, issues of scale and unit of organization are significant for smaller and larger components of union organization should not necessarily be assumed to have the same goals, dynamics and interests. Indeed, they may diverge and conflict. National unions’ concerns are further from the workplace so that their constituency of interest is normally the aggregation of workplaces their members work in (rather than the individual workplace as for the workplace union) as well as the national union as an organization and the national politics than impinge upon it. A national union–local union conflict in the American meatpacking industry famously illustrated this in the 1980s, becoming the centre piece of a popular documentary film, American Dream (Kopple et al., 1990). Moreover, the preponderance of employed union officers is necessarily higher in national unions than in workplace unions so that they are guided, notwithstanding the democratic process, by these officers’ views. The same is true for geographical regions and branches of unions vis-a-vis workplace unions. Part of this difference speaks to the conflicting and complementary rationalities of unions as organizations (as per their administrative or bureaucratic function) and unions as the members (as per their representative function). But whatever their derivation, measures of union effectiveness must be situated on an appropriate size scale. Thus, it is legitimate to consider whether Hammer and Wazeter’s (1993) study of teachers’ locals in one state provides lessons for national unions. Moreover, the implications of teacher unionism, with elements of professional control, also beg the question of whether the research site can generate valuable lessons for other public sector unions. The same point can be made about the study on organizing effectiveness by Badigannavar and Kelly (2005), and a similar query can be raised for Frege’s (2002) study of workplace unionism in Hungary and Slovenia. The wider point here is that unions’ goals may differ in that some may seek professional control as a form of job control while others with such ‘craftist’ mentality may seek legal regulation of their jobs instead. By contrast, the ‘Goal-System’ framework (e.g. Fiorito et al., 1993, 1995, 1997) on national unions and national leaderships is explicitly situated in the particular national context. Yet, even here, reflection on whether lessons are generalizable to other scales (inter- or transnational, subnational) is required. So, a clearer sense of what lessons are general or particular to those scales of union organization outside the research site is important. Indeed, in the case of the conflicting and complementary rationalities, this should also relate to the outcomes of the interaction of different scales of union organization over space and time. On a more practical level, the deepening process of capitalist globalization of the means of production, distribution and exchange requires issues of scale and interrelationship of constituent parts (as per the slogan, ‘Think global, act local’) to be more prominent in union effectiveness studies.

Quite understandably recent union effectiveness studies have been dominated by assessments of the effectiveness of ‘union organizing’. Over 10 years ago, Fiorito et al. (2001: 263) noted: ‘The focus on organizing effectiveness is partly a product of the contemporary crisis in union membership and density, and partly a product of views of organizing as a “supra-” or “meta-strategy” given its necessity to achieve virtually all union goals.’ It may seem churlish to criticize studies of union organizing for what they do not do, especially as assessments must necessarily be made manageable by placing limits upon their scope. However, given three points, criticism is not unreasonable. First, as Fiorito et al. (2001: 263) suggested, union organizing has been viewed by many as a panacea for resolving many aspects of union weakness, and union weakness and ineffectiveness overall. Second, union organizing quintessentially represents a particular means towards gaining an end, and a rather partial one at that. Third, the effectiveness of union organizing cannot be properly assessed on its own without also assessing items such as effectiveness of unions in terms of organizational effectiveness. Consequently, many union organizing effectiveness studies are rendered less than complete for they assess the means of union organizing using limited criteria. Charlwood (2004), for example, examined the factors that influence ‘consolidatory organising and recruitment effectiveness’ in Britain using the criteria deployed of recruiting and retaining members without regard for the bargaining outcomes (especially material ones). The authority cited for this was Fiorito et al. (1995: 614) even though they stated ‘a full evaluation of organizing effectiveness must consider how organizing efforts contribute to bargaining, political, and self-help objectives’. Elsewhere, Simms and Holgate (2010) ask ‘what is organising for?’ but do not answer their own question, not least because they did not provide a framework for analysing data in order to proffer an evaluation. Burchielli and Bartam (2009) examined what helps organizing work without examining does organizing work (in terms of union effectiveness). In contrast, Badigannavar and Kelly (2005) examined bargaining outcomes and achieving representational rights of two organizing campaigns in Britain (but without explicitly defining a concept of effectiveness). A similar omission appears in their related article on union effectiveness and union revitalization through organizing and partnership approaches (Badigannavar and Kelly, 2011).

Overall, the worth of assessments of union organizing effectiveness is limited because, while associations between outcomes and other factors are found, they less clearly establish direction of causation. For example, in examining the relationship between density and union effectiveness, Rose and Chaison (1996) showed there were associations where union effectiveness was defined in terms of organizing, bargaining and political outcomes (labour law, distribution of wealth) but they did not set out the dynamic, osmotic nature of the relationship of cause and consequence between the two. The same criticism can be levelled at Hammer and Wazeter (1993) over union mentality, participation and leadership. This type of deficiency is particularly important because some may regard the gaining of legitimacy, bargaining rights, new and additional members and resources for union action as goals (operational, penultimate, etc.) while others may regard them merely as facilitators and not goals as such. The latter have a point to make because while legitimacy and resources are, inter alia, necessary for union effectiveness, they are far from sufficient to achieve it. Moreover, even when all the necessary and sufficient parts of the equation are present, the whole cannot become greater than the sum of the parts until the issue of the appropriate application and exercise of power in terms of strategy and tactics is addressed in practice (and subsequently analysed). Fiorito et al.’s (2002) study of developments in innovation and use of information technology in relation to union and organizing effectiveness should be seen in this light. This remains true even when it is recognized that there are many factors which have been studied (like social identity, issue framing, etc.) with regard to subsets of union effectiveness and which influence the ability to mobilize members.

The issue of union effectiveness causes or determinants has been studied. Sometimes, such study has overlapped with the study of components of union effectiveness (see Hammer and Wazeter, 1993). Fiorito et al.’s (1993) set of determinants revolved around the degree of innovation; bureaucratic structuring and the locus of decision-making (i.e. centralization); union democracy; union strategies; and environmental influences (consisting of employment growth and employer opposition to unions). Frege (2002) categorized her causal factors in terms of economic, institutional, organizational resources, while Lévesque and Murray (2010) offered a more (union) internalized framework of four power resources and four strategic capabilities (see below). And, given our emphasis on the significance of strategy to generate and apply power, Boxall and Haynes’s (1997) examination of union strategic choice is significant for union effectiveness. They argued that effectiveness is not synonymous with strategic planning but rather with the enactment of strategy albeit in a rather circular manner: ‘Unions become effective only if they conceive and enact more effective strategies’ (Boxall and Haynes, 1997: 569). However, they did not then go on to study strategy and tactics per se in their examination of union behaviour approximating to their four models. Although complex, union strategy can be understood and operationalized (see Gall, 2011).

Even with Lévesque and Murray (2010), the same absence of union effectiveness is notable. They provided a conceptual and theoretical-based analysis of resources and strategic capabilities that constitute union renewal yet proffered no definition of union effectiveness (and, thus, nor obviously the criteria and measures by which to judge it). As argued before, union renewal is primarily a means to an end, and not an end in itself. It is a penultimate, not an ultimate, goal. The omission is notable because Lévesque and Murray (2010) took their starting point as union power and cited its loss as being at the core of union decline. It was because of this starting point and power being primarily a penultimate, rather than, ultimate goal that led to the omission. However, omission of union effectiveness is also predicated on the absence of a consideration of the terms of the application of power in regard of place and space, on the one hand, and strategy, tactics and organizational intelligence on the other (cf. Lévesque and Murray, 2010: 334, 336, 345). The ‘how’, ‘when’ and ‘where’ of the application of power were left unaddressed. Appropriate and sufficient application of power – in quantitative and qualitative terms as effective leverage – over external bodies to gain goals and objectives is fundamental to understanding not only successful union renewal strategies but, therefore, also union effectiveness.

Towards an integrated conceptual framework

We now return to laying out a framework for defining and measuring union effectiveness. Figure 1 provides a schematic representation of our ‘Goal-System’ framework of union effectiveness. We believe it advances understanding by helping to focus upon more identifiable ultimate measures; by ordering items in a way that distinguishes between the creation of penultimate (means) and ultimate (ends) goals; and by recognizing the totality and interconnectedness of gaining the panoply of union goals and the processes by which they are arrived at. Thus, whether a union is effective can be judged on a number of measures which, when put together and seen in the round, the whole becomes greater than the sum of the parts. This represents a form of totality which has hitherto been missing. Thus, two associated propositions are:

OPEN IN VIEWER

Figure 1.

Schematic Goals-Systems model of union effectiveness.

We briefly describe the model’s main features illustrated in Figure 1 and then elaborate upon certain points before providing a more detailed consideration of industrial leverage (Figure 2). First, Figure 1 suggests that environmental influences play a pervasive and, often decisive, role. The environment shapes aspects of agency such as aspirations and goals, appropriate choices for actions, and the influence of levers of power (where environments can be regarded as the accumulated outcome of agency, namely, decisions and actions). These endogenous variables (goals, actions and levers) along with the environment shape outcomes pertinent to the various dimensions of union effectiveness, which is indicated by a potential host of measures reflecting the breadth of aspiration encompassed in the notion of ‘advancing and defending worker interests’. Although depicting an overall sequential flow of influences (i.e. from environment to goals to actions to levers to outcomes), this is a dominant but certainly not the only direction of flow among these variables. The environment, for example, is not fixed for all time and is potentially altered by union, employer and state actions. One might also consider feedback effects, wherein, for example, evaluation may indicate that actions did not yield intended effects. This leads to organizational learning and attempts to apply actions more precisely in future opportunities (see also Hyman, 2007). Moreover, there are exceptions to the dominant causal flow posited. But this is not to say ‘everything is connected to everything else with indeterminate causality’. Despite feedback effects and exceptions such as those noted, there is nonetheless a dominant causal flow. This could be more formally stated through ancillary propositions – e.g. environmental factors influence goals – but this seems unnecessary detail at this time. A further proposition that follows is:

OPEN IN VIEWER

Figure 2.

Industrial leverage as a function of agency and environment.

Given our contextualization of unions as of, by and for workers, our starting point remains a simple but broad definition of union effectiveness concerning the successful application of the appropriate levers of power to gain purposeful goals. The advancement and defence of members’ economic, political and social interests is the definition of what unions are ‘about’ and ‘for’. The places of ideology and law are taken to be contained within these three categories of interest. Effectiveness pertains to an evaluation of the how, when, why and where – and if – of unions achieving this ultimate goal, and relates to both procedural (of ‘voice’ and ‘justice’) and substantive (of remuneration, etc.) outcomes (and their interrelationship). This then allows for different unions to act in different ways as their circumstances and strategies suggest but without losing sight of an overarching means of evaluating any or all union organizations’ achievements or failures in regard to effectiveness.

By levers, we mean the collective tools and resources needed to achieve effectiveness, where a lever represents the generation and application of power in a particular circumstance. By ‘appropriate’ levers of power, we mean applying the salient amount and type of power as a result of the choice of strategy and tactics in the apposite arena so that deliberate and usable leverage is created. Thus, unions need to align the extent of their ambition with the degree of the right kind of power so that they do not take a sledgehammer to crack a nut and vice versa. Specifically, this means understanding that, for example, some strikes are primarily of an economic or political nature (but seldom both) because of the employer concerned (capitalist, state), that the behaviour of capitalists is still influenced by the state so that the state may be targeted to effect change among capitalists, and states must be targeted to advance and protect social wages. Alternatively, the ‘leverage strategy’ practised by some Australian, British and US unions does not just require that the salient supply chains and commercial relationships are pinpointed but that the points of fragility and sensitivity within these are identified and appropriate levers – with finely tuned strategy and tactics – are exercised on them. But whatever levers are concerned, they are nearly always and everywhere necessarily collective ones based upon mobilization of members, their supporters and their resources. It is possible to imagine union power that is not based on mobilization or only very remotely so based. A leader’s persuasive skills at the bargaining table can be a source of power. A lawsuit or any action that relies heavily on union financial resources may be only tenuously linked to mobilization, but, ultimately, the decision to spend union resources is a decision about collectivizing individual resources. Thus, even a lawsuit is an example of the deployment of a collective resource (as well as likely having implications for other members and workers). It is from members, their supporters and their resources that the forcefulness of a union’s ideological, political, organizational and financial resources is derived. This reinforces Proposition 4, suggesting a refinement along the lines of ‘Union supporters’ commitment to collective action through financial and volunteer contributions is positively related to union leverage in various forms’ and which is closely linked to Proposition 9 (see below).

The arenas of operation comprise product, labour and capital markets in terms of the economic arena and the various organs and processes of public policy decision-making and legal enactment for the political arena. Consequently, the primary levers in aforementioned markets comprise, inter alia, disruption to the production, distribution and exchange of goods/services through the withdrawal of labour, restriction of output, etc., and damage to brand, reputation and investment opportunities (see also Batstone [1988] on the aspect of labour scarcity related to disruptive capacity). Underlying all is the disruption of the orderly provision of goods/services. For the private sector, this disruption implies infringement upon employers’ ability to create profit by imposing costs upon them. That said, leverage is greater in the industrial rather than economic arena because directly creating disruption to employer interest within product and capital markets is harder – disruption to the labour market is easier and has a spillover effect into product markets. But there is no proverbial Chinese Wall between the spatial bounds of arenas, for sufficient economic leverage can also create political leverage when governments feel compelled to intervene in disputes. The case of a (private sector) transport strike is an example. Similarly, the ideological arena heavily colours what is permissible and achievable within the political arena. Nonetheless, there are also some distinct separations whereby, for example, public sector workers have little economic leverage but more political leverage. Thus, a strike by public sector workers primarily seeks to exercise political rather than economic leverage (even when it involves disruption to revenue generation like tax collection).

In discussing economic or industrial leverage, Clark (2009: 5) noted the ‘highly dynamic’ nature of union effectiveness. This is salient for associated concepts such as leverage. This dynamism is brought forcefully into focus in strike leverage. While strikes are a source of leverage, Freeman (2007: 128–140) noted the ‘Great Doubling’ of effective labour supply resulting from rapid Chinese, Indian and former Soviet Union integration into an increasingly global labour market was arguably a ‘game changer’. Although strikes already may have been declining sources of leverage for other reasons such as employer dependence upon workers being reduced due to technological change, ² this rapid and dramatic labour market integration as part of wider neoliberal globalization has arguably transformed the strike option for workers from an often reliable source of leverage to one that is currently rarely practical for many workers in advanced industrial societies. Ackers (2013) suggested an entirely different argument for diminishing strikes, namely, that in some settings more integrated and ‘just-in-time’ (JIT) supply chains have made strikes too disruptive (i.e. too ‘deadly’) to be used. Rather, we contend that with JIT and lean production unions have performed poorly at identifying points of fragility and sensitivity within operating systems (see Gall, 2013).

Leverage can be exercised in direct and indirect ways and on vertical and horizontal planes within a single arena. (Albeit that it is recognized that the once seemingly simple bilateral relationship of the employer and worker is now considerably more complex as the increased use of self-employment, subcontracting and the like – along with continual churn in organizational reconfiguration – have made the employment relationship far more multifarious; see Marchington et al., 2005.) Industrial leverage (i.e. some form of collective mobilization like industrial action) is exerted more directly upon the employer (as per a bilateral relationship and in the workplace or at the enterprise level) while political leverage may see pressure applied directly upon the government over its policy as well as on the government as a third party to pressure the targeted employer. Leverage within one organization can be used against another (that is, a buyer or supplier). The utility of developing leverage – through public protests and brand damage – against the clients that determine the contracts and not the contractors themselves was highlighted in the ‘Cleanstart’ campaign to improve wages and conditions in contract cleaning in Australia and New Zealand (Crosby, 2009). Industrial leverage – in terms of its immediacy and magnitude of impact – is influenced by a range of factors like perishability as well as susceptibility of production of goods/service to disruption (like JIT systems) and availabilities of stock and alternative suppliers. Union ability to use these to create leverage depends upon other factors of social relations like union density and consciousness in terms of mobilizing capacity. However, despite the prevalence of JIT and lean production regimes, manufacturing workers seldom use this opportunity to press their case for resolving grievances. This may be explained by the use of ‘no-strike’ deals and the job insecurity experienced in a sector which has seen contraction and offshoring (see Gall, 2013). Indeed, the more effective tool of workplace occupation, rather than striking, against redundancy and closure – by raising their costs through sequestrating employer assets and control from inside the workplace – has been seldom used in the current recession (Gall, 2010). When confronting offshoring and outsourcing in ICT, again, workers do not use the employer’s dependence upon their cooperation (to train the new staff, to secure data and operating systems) to ensure a smooth transition of the service as a highly delimited window of opportunity for bargaining. These examples suggest the choice of the tactics to resist can be crucial – assuming a union has the ability first to mobilize.

We have been at particular pains to state that the issue of strategies and tactics, thus choosing attendant windows of opportunity in terms of the how, when and where to act (i.e. to apply levers of power), is vital to understanding union effectiveness. This is because the conception of particular circumstances when power is generated and applied as a lever has been sorely missing from considerations of union effectiveness. We believe the issue of strategy and tactics goes a long way to explaining why some unions are successful and others are not (and how successful some are). Issues of resource generation or securing penultimate goals are, thus, held to be necessary but not sufficient to explain success (or otherwise). In sum, the preceding suggests a broad but important proposition about union effectiveness (which follows on from Proposition 5):

Although there are some who think a particular lever such as the strike is a union’s ‘ultimate weapon’, there are clearly times and places, as we have described, where alternate tactics may be more apt and effective. In some public sector jurisdictions, for example, harsh laws against striking create potentially high strike costs to unions and their members. The ‘benefit/cost ratio’ for the union is very unfavourable. Publicity tactics to illuminate employer mismanagement or other failings may be less disruptive in the short term, but the benefit/cost ratio far better.

In the light of this, we now outline a model of industrial leverage (see Figure 2, where it is assumed that density and loyalty are sufficiently high to stymie normal operations being maintained by existing non-union staff and temporary replacement workers are not easily to hand). Industrial leverage is defined as ability to inflict costs and damage upon a bargaining opponent for purposeful gains – or to credibly threaten to do so either consciously or through the bargaining opponent’s perception. With an employer, those goods or services being targeted to create leverage are essentially either economic/private or political/public goods. This obliges unions to analyse the weak points in the supply chain they seek to target. The key components (with internal dimensions) for understanding whether leverage is generated concern:

Figure 2 suggests that unions must be cognizant that leverage is generated across this range of components, hence the interrelationship of the components. Thus, power and influence in one may help generate power and influence in others in an overall, cumulative manner. Using the dimensions as variables would enable evaluation of the worth of, inter alia, administrative or production staff, JIT and non-JIT systems, being up or downstream in a supply chain, and peak and non-peak times of demand. Although leverage results from operating across a number of the above variables, any and all are founded upon member activism and financial resources derived from their dues. As before, the ability to create leverage is still subject to the particular strategies and tactics used. Industrial leverage is believed to be the central form of leverage where other forms like political pressure and reputational damage but play a support role to this sharper and more malleable form.

Conclusion

While proffered some 20 years ago, Fiorito et al. (1993: 113) and Hammer and Wazeter’s (1993: 303) observations that union effectiveness was conceptually underdeveloped still remain broadly correct. In setting out the dimensions of union effectiveness, as well as the criteria by which to judge achievement in each, we have avoided providing excessively tight definitions because we want to establish the first principles, which can then begin to be operationalized in the research setting with the aid of a number of propositions (rather than hypotheses). From the propositions, hypotheses may be developed. We suggest that operationalization must be carried out (a) in a holistic (and not ‘pick and mix’) manner and (b) through largely quantitative research methods so that something akin to a universal scorecard can be created for each unit of union organization and that this is transferable across and between unions. An essential part of this is to measure the extent of union effectiveness as well as the efficiency in gaining effectiveness such that some robust statistical analysis can be developed. That said, we recognize our framework has to be held against temporal and paradigmatic dimensions for the practice of unions vis-a-vis union effectiveness varies considerably in different IR systems and in different epochs. (Indeed, it can also vary by industrial sector.)

Deploying a broader, more rounded conceptualization of union effectiveness – underpinned by that of leverage and then supplemented by exercise of that leverage in terms of strategy and tactics – helps get to the heart of the matter of theorizing the building of successful unions in an environment of asymmetrical power relationships within a zero-sum or mixed game where interests between capital and labour are not wholly compatible. This involves something of a critical synergy of two relatively disconnected literatures on union effectiveness and union power in order to provide the necessary springboard for constructing this more robust and rigorous theorization. (There are other important conceptual threads such as that on union legitimacy [for example, see Chaison et al., 1993] which might also be fruitfully incorporated into a discourse on union effectiveness so as to create a more robust analysis, but we leave that issue for further theorizing.) Thus, we worked through our critique of the existing literature in order to construct the basis for providing a totality for theorizing union effectiveness. The product of this then is a schematization of union effectiveness that is broad and deep, dynamic and more internally specified in terms of its subcomponents and their interrelationship. This does not provide for any ‘roadmaps’ or ‘best practice’ for unions or activists but it does signal that praxis – a priori theoretically informed practice – is essential for unions to be able to counter the power of capital and the state. This marks out our contribution. ³ Our keenness to rectify the identified maladies results from recognition of not only the power imbalance between capital and labour but also acknowledgement that goal setting and goal achievement are influenced by an external environment which represents the power, ideology and interests of capital. We believe we have also provided a useful basis from which to productively contemplate how these national dimensions may be transferred and replicated onto the global (international, transnational) economic stage on which unions must increasingly now seek to be effective within.