Earlier chapters have shown us that: the decline in union density can be partly explained by structural change in the labour market, but that this change cannot explain the deterioration in union membership in the 1990s; that declining union density cannot be blamed on falling sympathy for unions – indeed, it appears that sympathy for unions has increased since the early 1980s; and that what little evidence there is does not suggest that there has been a decline in employees' perceptions of union performance or union propensity. In this chapter we turn from structural to institutional influences on decline in union membership: changes in the approaches of employers and governments to unions, and in the fundamental determinants of union membership.

An Overview of the Institutional Break

Price and Bain (1989) proposed that, while relationships governing union membership would mostly be stable and cyclical (explicable by business-cycle variations), at particular times there could be fundamental changes or ‘paradigm shifts’ to those relationships. These institutional breaks emerge from particularly forceful conjunctions of social or economic events and powerful alliances of some of the participants in industrial relations, and alter the institutional arrangements surrounding the employment relationship. A paradigm shift creates ‘new patterns in the context of industrial relations’, principally changes in ‘labour laws and the powers and roles of regulatory agencies, employer policies towards unionisation and collective bargaining, and union structures, political activities and ideologies’ (Chaison & Rose 1991).

Declining union density has potentially wide implications for society and for the study of industrial relations. Continuing reductions in union density might be expected to lead, unless other forces intervene, to widening inequality and reduced ‘voice’ for employees in their working lives. As Barbash (1985) has pointed out, it is unclear whether the benign aspects of human resource management would continue in the absence of organised labour's countervailing power. Moreover, the traditional academic discipline of industrial relations is founded, ultimately, upon the study of the interactions between employers and collectively organised employees: the decline in union density in Britain has been claimed to be leading to the ‘end of institutional industrial relations’ (Purcell 1993), fundamentally affecting the way labour markets are studied. This book has therefore presented a framework for analysing the reasons for union membership and non-membership, and has examined the reasons for union decline.

In the preceding chapters, union membership in Australia has been examined within a ‘change-response’ framework. This model was used because it focused on the way in which union membership is determined by changes in the environment or in the strategies of major participants in the industrial relations system, and by the way in which other participants respond to those changes.

This chapter analyses union membership at the level of the individual employee. The data come from four surveys of employee attitudes: the 1990–91 Survey of Employees in Metropolitan Sydney Establishments (SEMSE), undertaken by the author; the 1995 Australian Workplace Industrial Relations Survey (AWIRS95); the 1996 Labor Council Survey (LCS-96); and the 1996 Australian Election Survey (AES) (Jones et al. 1996). These surveys are discussed in more detail in the appendix on research methodology. The first two surveys were restricted to employees in workplaces with 20 or more employees, with SEMSE also being restricted to certain industries and localities; the other two were household surveys with no restrictions on type of workplace or locality. In 1997 a second Labor Council Survey (LCS-97) was, like the first, undertaken by Newspoll; it had far fewer questions but some results are briefly discussed in this and other chapters.

This chapter looks at the reasons people have for belonging or not belonging to a union. The focus is initially on answers to some openended questions posed by the surveys, followed by a discussion of union propensity (whether employees prefer to be in a union); the instrumentality of union membership; satisfaction with unions; responsiveness of unions; and aspects of the management-employee relationship.

Union Membership in the Surveys

Before discussing attitudinal data from the employee surveys, we need to know something about union membership in each of the surveys.

While previous chapters have looked at various types of data concerning individual employees, much of the change in union membership is change that takes place in the workplace. It is the workplace, then, that is the focus of this chapter. What are the characteristics of workplaces that influence union decline and growth? What do they tell us about the influence of union and managerial behaviour and strategy on union membership? What do they tell us about the efficacy of the Australian union movement's amalgamationist strategy?

This chapter examines the issue of within-workplace change from four angles: first, changes in union density within workplaces; second, the extreme situations of deunionisation or union collapse; third, the recruitment of new union members and establishment of unions; and fourth, the effect of union amalgamations on union membership. The Australian Workplace Industrial Relations Surveys are the main sources of information, including the AWIRS95 ‘main survey’ of 2000 workplaces and the separate ‘panel survey’ of 700 workplaces that were surveyed in 1989–90 and reinterviewed in 1995–96, enabling comparisons over a period of almost six years to be made.

All ‘workplace’ data in this chapter refer to those with 20 or more employees. Although not representative of all workplaces, they account for the great majority of employees in unionised workplaces, and enable us to understand the forces for change within the workplace.

There are differences in union density between different industries and occupations; density varies according to the size of the workplace and between the public and private sectors, between full-time and part-time workers, between casual and permanent employees, and between the sexes. Structural change in the labour market might therefore be expected to lead to changes in union density. This chapter examines the influence of these factors on union density, and assesses the extent to which they might be said to be ‘structural’.

Industrial and Sectoral Change

In all countries, there are differences in rates of unionisation among different industries. Workers ‘are more likely to be union members in manufacturing, transport or public administration than in agriculture, (retail and wholesale) trade or financial services’ (Visser 1991:107).

The pattern of industry variation in union density in Australia is comparable to that overseas, but Australia appears to have larger gaps in density between the private and public sectors than many other countries (the average gap in OECD countries being 20 percentage points in 1988). Public-sector unionisation exceeds private-sector unionisation by 55 per cent to 23 per cent in the members survey. Interindustry variation in union density is greater in the private sector than in the public sector. That is, union density tends to be high amongst public-sector workers irrespective of their industry or employment; in the private sector, union membership is more dependent upon employees' industry.

Through the 1980s and into the 1990s union membership, as a share of the Australian workforce, has been falling. In the space of two decades from 1976 to 1996, union density (the proportion of employees belonging to a union) dropped by two-fifths. The union movement, in a country which had once enjoyed the highest density in the world, is facing a crisis of membership.

While union density is not the same as union strength, it is nonetheless one of the most important factors that affects a union movement's power. From overseas experience, particularly that of the United States, it appears that declining union density may raise major problems of legitimacy for the union movement as a whole. It can be used as a weapon by employer groups and others to argue that the ‘privileges’ afforded unions, and denied the majority of the workforce, should be withdrawn. Declining union membership may lead to cutbacks in staff and resources in unions and thereby to reductions in the union organising effort and in services provided to members. Success (or failure) in recruiting union members may influence the attitudes and actions of employees and employers and encourage success (or failure) elsewhere (Rose & Chaison 1992; O'Neill 1971; Western 1993b).

Whether the syndrome of self-perpetuating, deep decline evident in the United States will be repeated in Australia is difficult to predict, but it highlights the importance of understanding the reasons for union decline and the factors that work to reverse this trend.

A number of studies have distinguished between ‘ideological’ and ‘instrumental’ reasons for belonging to a union. Researchers have used a range of names with varying degrees of elegance – such as ‘enterprise unionateness’ and ‘social unionateness’ or ‘instrumental unionism’ and ‘social unionism’ (e.g. Prandy et al. 1974, 1982) – to distinguish between these or similar concepts. In this book, the term ‘union sympathy’ is used to describe the general, ideological views about unions held by employees, and ‘union instrumentality’ to describe the extent to which employees consider they will benefit or have benefited from union membership.

Union sympathy and instrumentality may vary between situations and countries (Gallagher & Strauss 1991). Perceptions that unions improve the welfare of all workers have been found to provide a substantial independent, altruistic reason for union membership in the US (Fiorito 1992). Some researchers suggest that changes in ideological views of unions can affect union density. Lipset (1986) argues that the success (or failure) of US unions in organising and recruiting is correlated with measures of the public approval of unions. Measures of the general image of unions have revealed image to have a strong influence upon the likelihood of union membership or propensity in the US (Getman et al. 1976; Schriesheim 1978; Youngblood et al. 1984; Deshpande & Fiorito 1989) and Belgium (Gevers 1992). A few Australian studies have shown the importance of ideological views of unions in influencing membership of Australian unions (Christie & Miller 1989:263–8; De Cieri 1991; Christie 1992; Deery & Grimes 1994).

The Accord played the central role in union policy in the 1980s and 1990s. Consequently it has played a central role in popular explanations of union decline in that period. This chapter critically examines the impact of the Accord on union membership and the further changes that might be expected in the light of the Liberal–National Party Government's enactment of the Workplace Relations Act 1996. It identifies the propositions underpinning the argument that the Accord was a significant cause in the decline in union membership, considers ways of testing these propositions, and discusses the relationship between the Workplace Relations Act and union membership.

The Two Halves of the Accord and Their Impact on Union Membership

There are several ways in which the possible impact of the Accord could be tested. In order to do this, we first need to identify the plausible mechanisms by which the Accord could affect union membership. There are in effect three main mechanisms by which the Accord could influence the decline in union membership.

The first is a potential ‘real wage’ effect. The decline in real wages associated with the Accord could have led to employee dissatisfaction with unions and declining union propensity. Members may then have left unions since they no longer saw unions doing what they were meant to be doing – raising real wages.

ABS surveys are described and discussed in the relevant ABS publications. This appendix focuses on other data sources.

The Australian Workplace Industrial Relations Surveys (AWIRS)

There are several major recent survey databases that have been produced under the sponsorship of the Commonwealth Department of Industrial Relations (DIR) (now the Department of Workplace Relations and Small Business). By far the most important of these are the 1990 and 1995 Australian Workplace Industrial Relations Surveys (AWIRS90 and AWIRS95 respectively). AWIRS90 was conducted from October 1989 to May 1990, and is reported upon in Industrial Relations at Work (Callus et al. 1991). It consisted of two surveys. The first was a personal interview survey of 2004 workplaces, involving the administration of between one and four questionnaires in each workplace, plus (in approximately 85 per cent of them) a self-completed questionnaire containing factual data on employment and other matters. The second was a telephone survey, using a shorter questionnaire, of 349 small workplaces with 5 to 19 employees. The sample frames for both surveys were designed by the ABS from DIR specifications and drawn from its register of establishments. The population from which the sample for the main, personal interview survey was drawn comprised all workplaces with 20 or more employees in all industries except two: agriculture, forestry, fishing and hunting, and defence. The response rates were: for the main survey, 86 per cent, and for the small workplace survey, 89 per cent.