Trade union membership and power in comparative perspective

Abstract

The trade union movement around the world remains in the throes of a prolonged and deep decline, whether measured by membership and density, bargaining power in relation to employers or political influence over the ubiquitous neoliberal narrative that underpins the policies of many governments. Decline has not been arrested or reversed by the many strategic initiatives undertaken in recent years such as organising campaigns or coalition building, although it is possible that the state of the unions would be even more parlous if these initiatives had not been pursued. Against this bleak backcloth, there are some positive signs: unions representing specific occupations, such as school teachers, nurses and airline pilots, have retained high levels of density; and union confederations in many parts of Europe have launched successful general strikes against unpopular government reforms to pensions and welfare benefits. Unions need to position themselves as agencies that can help deal with the growing problems of wage stagnation, low wages, income inequality and insufficient economic demand. That in turn requires a coherent challenge to the dominant neoliberal narrative.

Introduction

Since the early 1990s, a growing number of trade union movements have responded to membership decline by adopting a variety of innovative approaches to rebuild their membership, diversify their representative structures and increase their bargaining power. Organising, coalition building and international campaigns are just some of the activities undertaken (Frege and Kelly, 2004; Gumbrell et al., 2013). Yet, despite the substantial resources poured into these activities, the decline of aggregate union membership has continued seemingly without hindrance. For some European trade union movements, membership density has been less important than collective bargaining coverage, but that too is under pressure. In this article, I first describe the economic and political context within which unions now operate. Section ‘Varieties of capitalism and the health of the trade union movement’ briefly introduces the Varieties of Capitalism (VoC) approach as well as a conceptual framework for gauging the health of a trade union movement (and of its individual affiliates). Section ‘Evidence on the state of the union movement around the world’ reviews the evidence on the various dimensions of trade unionism; section ‘Union revitalisation’ considers the opportunities for trade union revitalisation, and the final section concludes.

Global economy and neoliberal hegemony

Stiglitz (2002) notes,

The closer integration of the countries and peoples of the world … brought about by the enormous reduction of costs of transportation and communication, and the breaking down of artificial barriers to the flows of goods, services, capital, knowledge, and (to a lesser extent) people across borders. (p. 9)

If we leave aside the reference to ‘artificial’ barriers, this definition of globalisation is otherwise uncontentious. It captures the extraordinary transformation of the opportunities available to capitalist firms arising out of the deregulation of national and international trade and production, ushered in by the ascendancy of neoliberal ideas and policies. The neoliberal agenda centred on the systematic dismantling of institutions and organisations deemed inimical to product market competition and labour market ‘freedom’ – in other words, trade unions and collective bargaining. The opening up of the Chinese economy to foreign direct investment from the early 1980s and the collapse of the ‘communist regimes’ of Eastern Europe from 1989 provided further opportunities for mobile, and increasingly, multinational capital, to flee from the high-wage, unionised economies of Western Europe, North America and Australasia. The wholesale privatisation of state industrial corporations, and the opening up of hitherto public services to private capital, further extended and deepened the spread of profit-making capitalist firms.

However, it would be wrong to depict the period since the 1980s primarily as one of capital flight from the Northern to the Southern hemisphere. For example, the bulk of the world’s foreign direct investment is by multinational corporations (MNCs) from the advanced capitalist countries (ACCs) into other advanced capitalist economies (Dicken, 2011: 26–31). It is also important to note that a substantial proportion of domestic capital is highly immobile: retail stores, care homes and leisure centres, for example, cannot be relocated to Asia in response to upward pressure on wages (Glyn, 2006). Nonetheless, a growing percentage of Northern hemisphere workers has been exposed to international product market competition and to the threat of capital flight – from Germany into Eastern Europe or from the USA to Mexico, for example – while others have become more exposed to the equally powerful threat of privatisation. Within the European Union, the accession of Poland and other East European economies has also facilitated increased labour migration and labour market competition.

Globalisation has sometimes been depicted by its proponents as a process without a subject, a developmental tendency that simply cannot be resisted (see Hay and Rosamond, 2002 for a critique). As Sklair (2001) has argued, it is better understood as a political project orchestrated by various elements of an emerging international capitalist class, comprising the heads of key governments, the executives of major corporations and members of international economic organisations (e.g. the International Monetary Fund (IMF) and Organisation for Economic Co-operation and Development (OECD) and international think tanks. Social democratic parties as well as conservative and liberal parties have been implicated in the augmentation of capital’s power through the deregulation of labour and product markets and privatisation (Callaghan, 2000). The power and mobility of capital and its national and international supporters have created acute problems for trade unions in many parts of the world. In sectors of the economy exposed to international competition, particularly manufacturing and some financial services, the threat of competitive pressures and of capital flight has been used to demand wage and productivity concessions from unionised workers. In the USA, similar threats have been used in around 50% of union organising campaigns to deter workers from voting union (Bronfenbrenner and Hickey, 2004). The combined result is not only downward pressure on wages and conditions but increased difficulty for trade unions in demonstrating they have the power to ‘make a difference’ to terms and conditions of employment. The perceived ineffectiveness of unions arguably hinders both retention of existing members and the recruitment of new members. Within the Southern hemisphere, the process of ‘regime shopping’, in which foreign MNCs play off one host government against another in their search for tax and regulatory concessions, is an equally potent source of downward pressure on the wages and conditions of Asian workers.

VoC and the health of the trade union movement

The ‘Varieties of Capitalism’ literature (Amable, 2003; Hall and Soskice, 2001) identifies firms and financial markets as the core actors within capitalist economies, with less prominent roles for both the state and organised labour (Howell, 2005: 27). At its core, the theory combines rational choice assumptions with institutional analysis and a functionalist equilibrium model (Streeck, 2009: 18). Institutions enter the analysis because they are seen as helping firms solve ‘coordination problems’ in five spheres: employment relations, vocational training and education, corporate governance, inter-firm relations, and workforce motivation. Institutions in these different spheres are interconnected to produce ‘institutional complementarities’ (Hall and Soskice, 2001: 9). These interlocking and mutually reinforcing institutions are thought to co-vary systematically, yielding at least two major ideal-typical capitalist models – liberal market economies (LMEs) and coordinated market economies (CMEs); later research has added more variants such as Mixed Market Economies (MMEs) and social democratic models (Amable, 2003; Molina and Rhodes, 2007).

One reading of VoC theory is that institutions possess a significant degree of resilience even in the face of strong market pressures so institutional change is likely to be incremental, revealing a strong degree of ‘path dependency’, preserving the differences between LMEs and CMEs (Kelly, 2011: 64). A different reading suggests that institutions can be ‘reworked’ in order to alter or subvert their original functions so that the differences between national economies become less clear-cut over time (Thelen, 2009). A more critical appraisal by Baccaro and Howell (2011) suggests that the institutional differences between VoC are being undermined by global market competition leading to common processes of declining union membership, bargaining decentralisation, weaker employment protection laws and growing income inequality. Allied to this critique is a renewed emphasis on the role of the state as a critical actor in the transformation of employment relations. These issues will be considered again as we review the empirical evidence on union decline, but first, we must consider how best to conceptualise and measure the state of the trade union movement.

In many LMEs, such as Australia, the UK and the USA, the health of the trade union movement is often discussed in terms of trade union density, the proportion of eligible workers who belong to a union.¹ The implicit assumption is that a high level of density is a prerequisite for bargaining effectiveness and for a high level of collective bargaining coverage. In fact, many West European trade union movements negotiate wages and conditions for the majority of the workforce within their country despite low union density. In Austria, for example, bargaining coverage is around 80%, but union density in 2013 was just 27%. In France, unions have relied far more on their capacity to mobilise workers and citizens in protests and strikes against government than on membership density, which in 2013 stood at just 8%. Thus, we need to analyse the state of a union movement along a set of dimensions rather than assigning analytical priority to union density. The relative significance of different dimensions is likely to be context-specific: bolstering union density may be critical in the USA but far less important in France. Table 1 (based on Behrens et al., 2004) depicts four main dimensions of union health as well as a number of sub-dimensions.

Table 1. Measures of trade union performance.

Source: Behrens et al. (2004).

Trade unions typically count their membership in absolute numbers, but this can be misleading: Irish trade union membership increased by almost 50% between 1988 and 2006 (from 440,890 to 653,597), but the even more dramatic expansion of the labour force meant that union density plunged from 55% to 39% (Roche, 2008: 18–19). Equally important is the composition of the union movement and the degree to which it reflects the shifting composition of the labour force. German trade union density was highly stable throughout the 1980s and early 1990s at approximately 33%, but this reassuring figure concealed a significant under-representation of the growing female workforce that was becoming predominant in the weakly unionised private service sector.

If membership is a key resource for unions, equally important are the workplace representatives who form the public face of the union for many employees. The skills and commitment of these mostly unpaid volunteers play a key role in shaping membership involvement in the union and willingness to remain in membership. The finances of the union largely determine the number and types of officials it can hire to assist with negotiations, casework and campaigning. From time to time, unions will need to threaten sanctions against government or employers, and in these situations, it is not so much the size of its membership that counts as its willingness to engage in collective action (that is, the capacity of the union to mobilise its membership). Finally, under this heading, a union must have ‘strategic capacity’, the ability to monitor and respond to continuing and sometimes dramatic changes in its external and internal environments (Weil, 1997).

Union bargaining power needs to be analysed in relation both to employers and to governments, as there is no theoretical reason to assume the two are necessarily correlated. Bargaining power is normally manifest in the process and outcomes of negotiations but the distinction, in practice, between negotiation and consultation is sometimes difficult to establish. The interaction between a strong employer and a weak union may increasingly resemble consultation rather than meaningful negotiation. Following Lukes (2005), union power should be measured not only by the outcomes of issues that are subject to negotiation, but by the range of issues that comprises the agreed bargaining agenda. Finally, under this heading of economic bargaining power, we should attend both to the micro-level of union activity – bargaining outcomes within the enterprise and sector – and to the macro-level – the distribution of national income between labour and capital. Political power can be captured in similar vein by the range of issues on which unions are able to negotiate with government as well as the outcomes of those negotiations. Political power can also be measured indirectly through a range of institutional variables such as union–party linkages and union influence over party policymaking.

Evidence on the state of the union movement around the world

Union membership and density

Table 2 shows trade union density figures for 32 countries between 1980 and 2012/2013 (the latest year for which reliable data are available), and we can see that density has fallen in all but three of them, the exceptions being Belgium, Finland and Spain. One interpretation of this evidence challenges the significance of national institutions by emphasising the common experience of decline in many different VoC. The Belgian and Finnish cases could be analysed as exceptions reflecting the operation of the Ghent system in which union membership is either required or is beneficial in accessing unemployment benefit; the positive impact of this ‘selective incentive’ on union density is well established (Western, 1997). Spanish density is fractionally higher than in 1980 but is still less than 20% of the workforce.

Table 2. Trade union density (%) and density change in 32 countries 1980–2012/2013.

Source: Organisation for Economic Co-operation and Development (OECD, 2015).

LME: liberal market economy; CME: coordinated market economy.

^a 1993.

^b 1995.

^c 2009.

On the other hand, the rates of decline vary between the different ‘varieties of capitalism’, being highest in the LMEs and lowest in the social democratic regimes of Scandinavia. The latter are particularly interesting because the small economies of Northern Europe have long been exposed to global market competition owing to the limited size of their domestic markets: Sweden’s population is a little over 10 million, and Denmark, Finland and Norway each has a population of approximately five million. The high level of union density in these four countries suggests there is no necessary incompatibility between strong trade unions and strong, globally competitive firms. It also suggests that national employment relations institutions can be sustained by state and other actors in the face of liberalising pressures. Eastern European figures are slightly misleading because the very high figures for the early 1990s represent the legacy of compulsory communist trade unionism. Thereafter, however, the highly deregulated regimes in this part of Europe conform to the liberal market trajectory with a near-universal and rapid collapse of union membership. The Australian and British cases are interesting because although density levels were almost identical in 1980 at around 50%, the decline in Australia has been even more dramatic than in Britain: current Australian density is just one-third of its 1980 level compared to one-half for Britain.

Beneath the decline in aggregate density, however, the picture is more complex. In the UK, many occupational unions representing professional workers have maintained or expanded their membership. This is true of unions organising school teachers, health workers, airline pilots, train drivers, musicians and actors, and the same is true in Australia. Arguably, the success of these unions is due in part to their ability to express and promote a strong sense of occupational identity. In contrast, it is the large, multi-occupation, multi-industry unions that account for most of the membership decline over the past 30years in the UK, the USA and much of Western Europe. However, many of the stable or growing unions organise in the public sector, and as a consequence, many countries now display a growing concentration of union members in public services. In the UK, approximately 53% of trade union members in 1995 worked in public sector organisations, defined as those ‘owned, funded or run by central or local government’. By 2010, that figure had climbed to 62% because of public sector employment expansion under the 1997–2010 Labour governments, the continued contraction of private manufacturing industry and the expansion of the largely non-union, private service sector (Achur, 2011). Despite government cutbacks to the public sector since 2010, the long-run trend is likely to resume because of the continuing decline in union membership density in both private manufacturing and services.

Similar trends are at work elsewhere. In the USA, public sector union membership comprised 42.3% of the total in 1995 but 51.9% in 2010; in Ireland, the figure rose from 38.3% (2001) to 54.2% (2009) and in New Zealand from 28.4% (1988) to 54.6% (2008) and comparable trends can be found in Denmark, France, Germany and Spain. Australia and the Netherlands seem to have bucked these trends with public sector union membership stable at around half the total stock of union members between 2000 and 2010 (Visser, 2013). The compression of trade unionism into the public sector is problematic because it allows critics to portray the union movement as a ‘privileged’ and ‘special interest’ group with terms and conditions of employment often superior to those in the private sector. Looked at another way, public–private differentials in pay and pensions are testimony to the effectiveness of trade unionism but also lend themselves to an anti-union discourse about ‘privilege’, as evidenced in the assault on public sector collective bargaining in Wisconsin, USA. Theoretically, the growing public–private split in union density could be said to corroborate the VoC emphasis on the key role of the employer in deciding whether to recognise unions for collective bargaining. However, it also points to the equally critical role of the state in maintaining trade unionism and collective bargaining; the significance of this role has been underplayed in much of the VoC literature. In relation to trade unionism, this is a serious oversight because many of the most dramatic, rapid declines in national trade union density have occurred in the wake of anti-union legislation introduced by Conservative governments, for example, the UK in the 1980s and New Zealand in the 1990s or following the dismantling of national institutions such as the arbitration machinery in Australia and New Zealand. A more subtle and under-researched effect of state action can be seen in the extensive legislation on individual employment rights, particularly in Western Europe, which may erode the incentive for employees to join unions (Estlund, 2013).

Union organising

In the face of the haemorrhaging of union membership over the past 20years, a number of union movements, especially in the LMEs of Australia, Britain and America, have adopted variants of an ‘organising approach’ to arrest and reverse membership decline. There are several theoretical models underpinning the organising literature, of which mobilisation theory is one (Gahan and Pekarek, 2012; Kelly, 1998). In essence, the theory states that the collectivisation of a set of individuals requires them to adopt a particular set of beliefs: their conditions of employment are in some way unjust or unfair; fellow workers share their sense of grievance; the employer is either the cause of their employment problems or is responsible for their alleviation; union action against the employer will be effective and at minimal cost; and these beliefs are combined into a coherent narrative or ‘collective action frame’ that explains their predicament and legitimates their protest.

Many union organising campaigns, both in the UK and the USA, have sought to mobilise workers around a sense of injustice, to secure majority support for trade recognition in a secret ballot and then to establish a bargaining relationship with the employer leading to collective agreements. British Trades Union Congress (TUC) interest in organising developed quickly in the mid-1990s, culminating in the launch of its Organising Academy in 1998, one year before the new Labour government passed a union recognition law, obliging an employer to negotiate with a union once it had shown majority support among the relevant section of the workforce in a secret ballot. Between 1995 and 2010, when the Labour government was defeated, union organising in the UK is estimated to have produced collective agreements covering 948,330 employees (Gall, 2012). It is not known exactly how many of the workers covered by these agreements joined unions (and remained union members), but since recognition for bargaining only follows a majority ballot vote, then it must be at least 475,000. While some organising campaigns targeted large workplaces, the average establishment size brought under union coverage was just 261 employees (a figure slightly higher than the US figure of 192: Bronfenbrenner and Hickey, 2004). Nevertheless, union density in the UK continued to decline throughout this period, even under the relatively favourable conditions of a Labour government, legislative support for organising and relatively low unemployment (Simms et al., 2013). Union organising has also failed to halt the long decline in union density in the USA and Australia.

There are several factors implicated in the very modest impact of union organising drives. One is the problem of perceived lack of union effectiveness because survey data on non-union members have shown repeatedly that this belief is strongly correlated with their unwillingness to vote for or to join a trade union (cf. Clark, 2009: 34–35 for the USA; Peetz, 1998: 37 on Australia). Managers’ denial of access to union organisers and active discouragement of employee unionisation may also engender the view that trade union presence would fail to make a difference to terms and conditions of employment. The UK 2011 Workplace Employee Relations Survey of several thousand establishments showed that in the private service sector, only 9% of human resource (HR) managers were in favour of trade unions (Van Wanrooy et al., 2013). The service sector is important because it now provides employment for the overwhelming majority of new labour market entrants. Recent research on how young workers in UK hospitality, retail and social care deal with workplace problems shows that only 6% of the sample approached a trade union – a reflection of the absence of unions from the vast majority of such workplaces in the UK (Tailby and Pollert, 2011).

Resources and institutional vitality

Reliable comparative data on key union resources, such as workplace representatives and finance, are difficult to obtain. Yet, cross-national evidence from 12 European countries shows that the capacity of unions to support members at the workplace is one of the best predictors of their willingness to retain union membership (Waddington, 2014). In many European workplaces, this support will be provided by the union workplace representative. British data from the 2011 Workplace Employment Relations survey show that while the majority of manufacturing workplaces have at least one union representative on site, this is true for only half the unionised workplaces in the private services sector and less than half in public services (Van Wanrooy et al., 2013: 58–59).

In contrast, one could argue that many union movements around the world have displayed a significant degree of institutional vitality. There is no easy way to measure such a concept, but several indicators point towards positive results: a growing number of union movements have adopted strategic plans, and although such plans can be poorly formulated, developed in a top-down fashion and inadequately implemented, their existence suggests some recognition, however limited, of a crisis of union membership and influence. It is also clear that many unions and union confederations are becoming more open to initiatives such as coalition building hitherto regarded as somewhat peripheral to their main ‘repertoires of contention’ and to their traditional focus on negotiations with employers (Frege and Kelly, 2004; Gumbrell-McCormick and Hyman, 2013).

Economic bargaining power

Collective bargaining coverage, the proportion of employees whose pay is determined directly by collective bargaining, has proved far more resilient than union density as Figure 1 shows. In the EU14 (the EU15 minus the UK) mean, unweighted coverage was fairly stable at approximately 80% between 1980 and 2011 (the most recent year for which we have reliable comparative data). In the LMEs of the UK and the USA, in contrast, collective bargaining coverage is closely correlated with union density, so both have declined together.

Figure 1. Collective bargaining coverage (%) EU and USA, 1980–2011.

Source: Calculated from Visser (2013).

Employees covered by collective (wage) bargaining agreements as a percentage of all wage and salary earners in employment with the right to bargain (EU14 = Austria, Belgium, Denmark, Finland, Germany (from 1990), Greece, Italy, Ireland (from 2002), Luxembourg, Netherlands, Portugal, Spain, Sweden. For missing year data, closest year was substituted).

However, in Europe’s largest economy – the German ‘coordinated market economy’ – bargaining coverage has been steadily declining because of a growing reluctance by employers to be governed by collective agreements. In 1995, industry-wide agreements covered a little over 70% of employees in the Western states, but by 2010, the figure had fallen to 56% (and was less than 40% in the Eastern states) (Behrens, 2013: 216). The erosion of bargaining is particularly pronounced in the private services sector and that is worrying because of the key role of service firms in employment growth. The decline of bargaining coverage could represent a process of liberalisation, as Baccaro and Howell (2011) have argued, or it could signify a more complex process of dualisation, in which distinct forms of regulation emerge and co-exist in different segments of the economy (Thelen, 2009). If we dig a little deeper into the content of collective bargaining, there is further worrying evidence about the range of issues subject to joint regulation. Data from the UK, which may not be representative of Western Europe because of its liberal market character, show that the scope of collective bargaining in the private sector has been declining over time (Table 3).

Table 3. Scope of collective bargaining, UK private sector workplaces with recognised trade unions, 2004–2011 (%).

Source: Van Wanrooy et al. (2013: 81).

The evidence suggests that issues once subject to union–management negotiation are increasingly regulated by consultation, in which management retains the final say, or by unilateral employer decision without any consultation (Van Wanrooy et al., 2013).

If we turn to the outcomes of collective bargaining, the evidence on the stability of West European bargaining coverage appears in a different light. At micro-level, it is still the case that unionised workers enjoy significantly better terms and conditions of employment compared to their non-union counterparts. Hourly wage rates, sickness and maternity pay and annual leave days are all higher where unions negotiate with employers (Van Wanrooy et al., 2013: 90). However, if we turn to the macro-level and consider the increasingly topical issue of economic inequality, then trade union economic bargaining power looks far less impressive. One way of capturing bargaining effects is through the division of national income between wages and salaries on the one hand, and profit, interest and rent on the other. The data provide a very approximate guide to the class distribution of income between labour and capital, although the raw data and its interpretation are both problematic. Wage and salary data include the salaries of the chief executive officers (CEOs) of large private corporations, who might otherwise be designated as capitalists; the incomes of the self-employed need to be adjusted so they appear primarily as wages not profits; and adjustments can also be made to capture the welfare payments and tax credits that supplement the incomes of low paid workers (Stockhammer, 2013). What the data reveal is a quite remarkable and long-run shift in the distribution of national income from labour to capital (Figure 2).

Figure 2. Wage shares in national income 1970–2010.

Source: Calculated from European Commission–Economic and Financial Affairs (AMECO) (2015), using methodology of Stockhammer (2013: 1).

ADV: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, Spain, Sweden, UK, and USA.

Series used: Adjusted wage share: total economy as percentage of gross domestic product (GDP) at current prices.

There are some variations between countries and uneven shifts over time, but the overall trend is clear and appears to embrace both liberal and coordinate market economies: the share of national income accruing to labour, through wages and salaries, has been steadily declining across the advanced capitalist world since the late 1970s. The actual magnitude of this shift is even greater than the data suggest because the salary incomes of top capitalists are included as part of labour’s wage share. More insight into the distribution of national income emerges from recent data on the incomes of the top 1% and top 0.1% of income earners in the UK and the USA, groups that are heavily concentrated in financial and business services. Between 1978 and 2008, the highest paid 1% of the British labour force (approximately 250,000 people) saw their share of total wage income treble, from 5% of all income to 15% (Bell and Van Reenen, 2014).

The wage share data provide one possible explanation for the stability of bargaining coverage in the CMEs of Western Europe. In effect, unions appear to have traded wage moderation and productivity improvements for the preservation of industry-wide collective bargaining.

Political power

The limits on the ability of trade unions to extract higher wages from employers in a globalising economy pose potentially severe threats to the capacity of unions to maintain, let alone extend, their membership and influence. One response by unions has emerged with particular force in Southern Europe where interventionist governments in Greece, Italy, Portugal and Spain have sought to overcome the weaknesses of poorly organised employer confederations through far-reaching social and labour reforms. Both Social Democratic and Conservative governments have played key roles in retrenching welfare systems, deregulating labour markets – to ease dismissal and redundancy – and increasing the costs to employees of state and occupational pensions. Although union density was never high in any of these countries and, with the exception of Spain, has fallen steadily since the 1980s (see Table 2), union confederations in Southern Europe have shown a tremendous capacity to mobilise people in strikes and protests well beyond the ranks of their own membership. Such mobilising capacity is even more remarkable in France, where union density is the lowest in the advanced capitalist world at just 8%. There has been an upsurge of general strikes in Western Europe since 1980, directed against government reform proposals, and this is a particularly interesting phenomenon given the dramatic decline in the incidence of economic strikes against employers (Figure 3) (Hamann et al., 2013b). It suggests that the collapse in strike activity, especially in manufacturing industry, may significantly understate the continuing ability of unions to mobilise members and citizens in anti-government protests and campaigns. The evidence also suggests that the recent general strikes cannot be dismissed as futile protests, designed simply to vent frustration over government policy because around 40% have yielded significant concessions from governments, especially on welfare and pension reforms (Hamann et al., 2013a).

Figure 3. General and economic strikes, EU15 plus Norway, 1980–2014.

Source: Hamann, Johnston and Kelly dataset.

However, there is little evidence that even successful general strikes have generated a net, positive impact on aggregate trade union density (see Table 2). It may be that strike effects are apparent only for the unions most heavily involved, or it may be that any membership gains from strikes have been cancelled out by losses due to employment contraction.

One arena in which there do appear to have been general strike effects in some countries is that of trade union–political party affiliations. Across a wide range of countries, both formal and informal ties between trade unions and social democratic political parties have loosened considerably in recent years (Burgess, 2004). In three Scandinavian countries, the unions ended their almost century long practice of collective affiliation of union memberships to the social democratic parties, replacing it with individual affiliation and thereby signalling a much greater degree of autonomy from the party. This happened in 1990 in Sweden, 1995 in Denmark and 1996 in Norway. In Spain, the 1988 general strike by both socialist and communist confederations against the socialist government led one year later to the formal separation of the socialist confederation Unión General de Trabajadores (UGT) and the socialist party Partido Socialista Obrero Español (PSOE) (Hamann, 2012). The Spanish case encapsulates many of the processes that have led to the partial erosion of labour movement structures. Social democratic parties have moved to the right, particularly on economic and welfare policy, and to varying degrees have embraced elements of neoliberal thinking, such as the value of privatisation, deregulation of labour markets and the need to construct business-friendly policies on taxation and wages. Trade unions meantime have continued to represent constituencies that have borne the brunt of these policies and have continued to oppose them through protests and general strikes. The weakening of union ties to social democratic parties has been accompanied by the emergence of parties to the left of social democracy that have proved willing to challenge neoliberalism (again, to varying degrees) and have become potential allies of the trade union movement. In Germany, a number of local and regional unions now openly support the left party Die Linke, and leftist parties in other countries have also begun to erode the social democratic vote share: Syriza in Greece, the Socialist People’s Party in Denmark, Podemos and United Left in Spain, Left Bloc in Portugal and the Socialist Party in the Netherlands. These developments raise important questions about the value and limits of the VoC approach in dealing with the dynamics of party political and electoral systems.

Union revitalisation

The trade union movement in the advanced capitalist world is both smaller and weaker in 2015 than at the end of the world strike wave of 1968–1974. This position has been reached despite a host of strategic innovations, in organising, in the reconstruction of trade unions to advance the interests of women and ethnic minority workers and through mobilisations, protests and general strikes against government austerity policies. It is true that unions remain strong in the public sectors of many countries, and that professional workers with a strong sense of occupational identity continue to be highly organised. Yet generally, the pattern is one of widespread decline in membership, power and influence.

At the same time, however, the world’s major capitalist economies display a number of features and tendencies which could provide the foundations for union revitalisation centred on a ‘narrative’ of injustice and exclusion. First, average earnings have declined or stagnated for many workers in many countries, particularly since 2008, and this development acts as a brake on the growth of effective economic demand (Gregg et al., 2014). That said, it is always possible that the American and European banks which were bailed out and rejuvenated after the 2008 crisis will once more seek to fuel the consumption of goods, services and housing through the reckless expansion of credit. With many workers on low and stagnating incomes, the dramatic increase in reward packages at the top end of the distribution has led to a substantial rise in income inequality (Bell and Van Reenen, 2014). Earnings stagnation and growing income inequality have multiple causes, but one contributory factor is the growth of zero hours contracts, an arrangement in which the employee makes himself or herself available to work each day but where the employer does not guarantee any fixed number of hours and pays only for hours worked. These contracts are becoming widespread in sectors such as social care and hospitality where hourly pay rates are typically close to the minimum wage and their effect is thus to compound low wages with earnings fluctuations and insecurity (TUC, 2014).

Wage stagnation, the growing contrast between rich and poor and the brazen exploitation inherent in zero hours contracts can all be interpreted as signs of fundamental flaws in national systems of wage determination. Historically, one of the main constraints on the size of the low pay sector and, to a lesser degree, on top earnings was the process of collective bargaining conducted by strong trade unions operating in tight labour markets. Unions may therefore be able to position themselves both as the champions of wage growth and income equality and as part of the solution to the problems arising from egregious forms of exploitation. In some cases, unions may be able to act directly, as in the recent campaign by New Zealand unions against zero hours contracts in fast foods (Unite the Union, 2015); in other cases, unions have successfully sought to persuade governments and parties, such as the UK Labour Party, to ban most forms of zero hours contracts (Labour Party, 2015). Unions in the UK, the USA and New Zealand have also acted in concert with other organisations, running campaigns to persuade large private and public sector employers to pay a ‘Living Wage’ (£9.15 per hour in London 2015) to their minimum wage staff (£6.50 per hour in 2015). This type of coalition building with community and faith groups does not necessarily produce many additional union members but it does, albeit indirectly, extend the reach of unions into areas of the labour force where they have few members and little or no collective bargaining presence (Newman, 2013).

Living wage campaigns have often involved unions moving up the supply chain, away from the contractors who employ low wage workers, in cleaning or food processing, for example, and towards the powerful, and often highly profitable, corporations that place the contracts and largely set their terms (Chen, 2014). Campaigns to improve terms and conditions within low pay firms have often collided with the employers’ claims that their scope for improvement is limited by the stringent terms in their contracts. Targeting large finance companies or food retailers has sometimes proved a more effective method of boosting pay and conditions further down the supply chain. It does, however, depend in part on the susceptibility of the contracting firm to the reputation damage associated with campaigns connecting the firm to low pay (Wright and Brown, 2013). The supply chain literature in conjunction with the organizing literature, suggests that recruiting groups of workers in one workplace at a time has proved to be extremely onerous, delivering very modest membership gains for significant cost and without preventing the ongoing erosion of aggregate union membership. Increasingly, unions in the United Kingdom and the USA, but elsewhere too, have begun to consider ways in which union influence over terms and conditions of employment can be increased in the absence of a significant resurgence of membership (Ewing and Hendy, 2013). While they are considering ways of moving towards the continental European model of high bargaining coverage with relatively low union density, paradoxically, some West European union movements are beginning to commit resources to organising, as they worry that the growing gap between high bargaining coverage and falling union density will eventually prove unsustainable (Gumbrell-McCormick and Hyman, 2013).

Conclusion

The trade union movement around the world remains in the throes of a prolonged and deep decline, whether measured by membership and density, bargaining power in relation to employers or political influence over the ubiquitous neoliberal narrative that underpins the policies of many governments. Decline has not been arrested or reversed by the many strategic initiatives undertaken in recent years, such as organising campaigns or coalition building, although it is possible that the state of the unions would be even more parlous if these initiatives had not been pursued. Yet, positive signs include the continuing high density of some occupational unions and successful general strikes launched by union confederations in many parts of Europe against unpopular government reforms to pensions and welfare benefits.

It can be argued that the raw material for union revitalisation, a pervasive and shared sense of injustice, is developing within many advanced capitalist economies in the form of wage stagnation, growing income inequality and labour market insecurity, to name only three issues. These are clearly problems for working people but also threats to employers and governments in the form of inadequate economic demand, limited income tax revenues and social discontent. Campaigns around the Living Wage and tax rates on high incomes as well as continuing protests about government welfare, pension and labour market reforms could all help to strengthen the general belief that unions are both relevant and effective. Unions can also seek ways of jointly regulating wages and conditions across particular sectors of the economy, particularly in the private service sector where membership levels are often very low.

Such initiatives will need to be underpinned by a ‘narrative’ that challenges the master themes of the neoliberal agenda and its endorsement of policies with dysfunctional economic and social consequences. Portrayal of trade unions as a source of labour market rigidity, as a constraint on employer management of the labour process and as a drain on corporate profits can only be challenged by an equally wide-ranging counter-narrative. This would assign a positive role to trade unions as an agent of social justice, equality and meaningful employee voice. While the neoliberal narrative is probably most potent in LMEs such as the UK, the USA and Australia, it has also gained considerable traction within recent European Union policymaking and in government circles within the classic CMEs such as Germany (Blyth, 2013). The pervasiveness of the neoliberal narrative creates major political, ideological and organisational challenges for trade unions. Yet, it also poses a major intellectual challenge to the core insight of the VoC literature, namely, that there are different and equally effective ways of organising national employment relations systems.