Chapter 1
Introduction
Paul Willman and John Kelly
There is a very broad picture to be drawn about union organization in Britain across the last twenty-five years. More or less independent of the measure of union organization one takes â and we will discuss several below â there is less of it than there was in 1979. The task for the academic analyst may thus be seen as documenting the decline and attempting explanations for it and, indeed, much of this type of activity has taken place. However, the attempt to explain the decline in union organization has itself indicated the complexity of the phenomenon and the information loss which results if one fails to disaggregate the process of change in trade union organization.
The year 1979 is often seen as a watershed in this matter, with a change of government from one broadly collaborative with trade union organizations to one distinctly hostile coinciding with the beginning of a long period of decline in membership and density. In subsequent years, increasingly hostile legislation on union organizing and activity went along with membership loss and density reduction. However, the pattern of change is complicated and there are a number of dimensions of union organization and activity to be considered. We begin by describing the broad pattern of change.
Union membership in Britain 1975â2002
Union membership in Britain climbed rapidly during the 1970s, fell from 1980 and began to rise again, albeit briefly, at the end of the 1990s. Part of the explanation for the turnaround in membership after 1979 was the sharp rise in unemployment which began that year (see Table 1.1). A large component of the unemployment growth was the shakeout of labour in the heavily unionized manufacturing sector: in 1979 there were just over seven million employees in the manufacturing industry but by 1992 the number had fallen below four million (Blyton and Turnbull 1998: 49). Employment expansion occurred primarily in private services where union density has always been far lower than in manufacturing or the public sector. In addition, government policy towards unions became extremely hostile from 1979, measured for example by its legislative programme to curb union power, and remained so until the election of the Labour Party in 1997. The legal duty on employers to recognize and bargain with a union representing the majority of its workforce was abolished in 1980.
Table 1.1 Trade union membership and unemployment in Great Britain 1975â2002a
As well as sectoral shifts in employment, changes to employer policy within sectors became important. In general, employer policy towards union recognition became much tougher and the owners of new plants, in particular, became increasingly reluctant to deal with unions. For example, 59 per cent of workplaces opened in the 1970s had union recognition for collective bargaining within ten years; the comparable figure for workplaces opened in the 1980s was just 34 per cent and for the 1990s it was 27 per cent (Machin 2000: 635; Bryson et al., this volume). Some employers actively sought to remove unions altogether: in the mid-1980s the incidence of derecognition was running at approximately 25 cases per year but by 1992 this number had leapt to 72 per year before then tailing off to just seven cases in 1998 (Kelly 1996: 89; Gall 2003a: 11). Some of the causes of decline were endogenous to unions themselves. Unions devoted few resources to organizing and recruitment: at the start of the 1990s almost half the TUC unions with 3,000 or more members did not even have a national recruitment officer (Kelly and Heery 1994: 48).
However, by the early years of the following decade, the situation had changed significantly and in ways that should have favoured a significant resumption of membership growth. Unemployment had begun to fall again from 9.8 per cent in 1992 to reach just 5.2 per cent in 2002 (see Table 1.1). Job losses in manufacturing continued throughout the 1990s albeit at a slower rate than in previous years but public sector employment actually rose by 149,000 between 1998 and 2000, primarily in health and education (Bach and Winchester 2003: 294). The union recognition provisions of the Employment Relations Act (ERA 1999) came into effect in June 2000 and their passage through parliament was associated with a sharp rise in the number of new recognition agreements, from 108 in 1997 to 668 in 2001 (Gall 2003a: 11). Very few cases have gone to the Central Arbitration Committee (CAC) â just 233 between June 2000 and May 2003 â suggesting that in the first few years of its operation the new law has not triggered substantial employer resistance (Wood et al. 2003; Moore, this volume). Finally unions themselves now devote more resources to organizing than ten years ago: 62 per cent of unions now have a national recruitment officer and a substantial minority of unions have sent trainees through the TUC Organizing Academy, created in 1997 (Heery et al. 2003: 59).
It is impossible to make an exact prediction about the effect of this set of circumstances on union membership. There is no agreement either about the impact of any one of these variables on union membership or about the relative weights that can be attached to each of them (cf. Dunn and Metcalf 1996 and Freeman and Pelletier 1990 on the impact of legal changes). But what we can reasonably say is that the environment and union policies changed to such a degree in the late 1990s that between 1997 and 2002 unions should have returned to membership growth. Membership did increase in 1998 and 1999 but these gains were wiped out in the following two years.
It will be a central concern of this volume to discuss why union recovery has been so limited. Some of the factors may be environmental, but some may be internal to unions. In the findings from the British Worker Representation and Participation Survey (BWRPS) there is evidence to suggest that there is a demand for representation among non-members that unions do not meet (see Gospel and Willman 2003). Data from the Workplace Employee Relations Survey (WERS) on the reduction in union density under collective bargaining indicates that union services are not always sought when they are readily available. Other evidence on the union wage differential indicates that the instrumentality of unions may have declined over recent years (Blanchflower and Bryson 2003). In short, there is at least a case for arguing that some of the reasons for reduction in many measures of union organization may be endogenous.
However, because of this complexity in the factors which influence measures of union organization and the range of measures themselves, we need some form of framework for the analysis of union organization and activity. This framework must deal with the influence of the three major institutional actors in industrial relations â government, employers and unions. It must also consider a wide range of measures of union organization and activity.
Agency, action and outcomes
Our approach here is as follows. First, we identify three broad agents of impact on union organization and activity. These are employers, government and unions themselves, the main actors in industrial relations. They often act in concert or at least the consequences of their actions are interactive. However, their actions impact union organization and activities in ways which are conceptually and temporally distinct.
Second, we turn to the dimensions of union organization and activity which may be characterized as the set of dependent variables we are seeking to explain. Three are identified, as follows:
- Resources; what are the key resources available for trade unions to organize members, by which we mean the broad processes of recruitment, mobilization and retention?
- Processes; how do unions, in this broad sense, organize?
- Outcomes; what are the consequences of union organization?
We know that there is no simple relationship between the dimensions. For example, take resources and outcomes. Unions may have large resources in terms of existing membership and financial resources but be unable to secure a union wage differential. Or take the example of processes and outcomes. Unions may become involved in large-scale recruitment activity but secure few members. A final example, of imbalance between resources and processes, comes from France, where large-scale union action can be mobilized on the basis of very low levels of membership and participation.
The agents may affect the dimensions in very different ways. Governments may provide unions with resources for organization through legislating to allow check-off or time off for representatives. They may afford processes, such as the current recognition provisions. They may even guarantee outcomes, such as the provisions of the old Fair Wages resolutions or the National Minimum Wage. Similarly, employers may provide the resources which sow the seeds of permanent organization by provision of facilities or information, may give recognition and bargaining opportunities as processes and may give up wage premia as a result of bargaining. Here, as with the government as agent, it may be that concession is a better word for provision but it remains the case that factors which affect these agentsâ propensity to concede or provide will be vital for union organization. Unions themselves act as agents specifically in making decisions about resources and implementing mobilization processes which may generate organizational outcomes. As with governments and employers, these acts may have positive or negative impacts.
We have attempted a schematic summary of theses issues in Figure 1.1. The cell entries are examples not comprehensive lists. Our purpose is to show that the organizational data which, for example, are trawled by WERS, represent different phenomena with different implications for union organization. Specifically, an abundance of resources will not generate organizational processes in all cases and these processes are not always successful in generating outcomes. A full understanding of the past and an accurate prediction of the future trend of union organization and activity in the UK depends on an understanding of the complex interactions of agency and dimension.
Figure 1.1 Agency, process and outcomes
The book
The chapters in the book focus on different aspects of agency and dimension. In Chapter 2, Moore offers the first major examination of the operation of the statutory recognition procedure. Here the government acts as agent to secure union recognition as an outcome by offering a statutory route. On the positive side for trade unions, not only does this route work much of the time, but the legislation appears to cast a shadow raising the overall level of voluntary recognition. However, the process also allows contest between the mobilization resources of the union and those â often greater â of the employer directed at avoiding unionization; in some cases, the greater resources prevail.
Organizing resources are a central concern of Kelly and Badigannavar and Perrons in the next two chapters. In Chapter 3 Kelly and Badigannavar use case studies of successful and unsuccessful organizing campaigns to illustrate the difficulties in mobilizing employees. Employer hostility is an
important determinant of outcomes, but by no means the only one. Perronsâ Chapter 4 on the ânew economyâ highlights some of the difficulties unions have in organizing outside their traditional strongholds. She shows, first, the diverse and complicated nature of work in the new economy and, second, some of the difficulties unions have in offering services and representation to such employees.
In Chapter 5, Willman focuses directly on union agency. He argues that the form of conglomerate union structure which has come to dominate the set of large unions is poorly designed as a platform for revitalization. In fact, it represents a defensive and risk-averse response to decline leading to a situation where some of the barriers to renewed union growth are internal to unionsâ own structure.
The next two chapters focus on developments within a traditional union stronghold, the public sector. Bach and Givan, in Chapter 6, examine the ways in which public sector reform initiatives influence the processes of union restructuring. Here government acts as direct employer and its agency has clear effects on organizing outcomes. Difficult questions face public sector unionism and, as Badigannavar and Kelly show, partnership initiatives at local level may not offer much help. Their matched-pair analysis of workplaces with and without such agreements illustrates the limited transformative power of partnership and the range of contextual variables which can influence organizing outcomes.
Bryson, Gomez and Willman focus on the employer as agent in Chapter 8. Most employers want some form of voice mechanism in the workplace but increasingly they choose the non-union option. This chapter models the employer choice of voice mechanism in terms of its cost and the risks involved. Although on some measures of cost and risk, the union voice option should be more attractive, there are very strong cohort effects showing a preference for non-union voice. On the positive side, WERS data show that there is, among non-union employers, indifference rather than widespread hostility to union voice.
In Chapter 9, OâGrady and Nowak offer an insider assessment of the strengths and weaknesses of the TUCâs New Unionism initiative and map out a number of ways in which unions can progress the New Unionism project. The book finishes off with a concluding chapter by the editors.
