Jump to ContentJump to Main Navigation
Managing the UnionsThe Impact of Legislation on Trade Unions' Behaviour$

Roger Undy, Patricia Fosh, Huw Morris, Paul Smith, and Roderick Martin

Print publication date: 1996

Print ISBN-13: 9780198289197

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780198289197.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 24 October 2019

Changing Union Behaviour

Changing Union Behaviour

(p.31) 3 Changing Union Behaviour
Managing the Unions

Roger Undy

Patricia Fosh

Huw Morris

Paul Smith

Roderick Martin

Oxford University Press

Abstract and Keywords

Even before the changes, unions were already experiencing financial problems, since the income from membership subscriptions was not enough to cover the total expenses. Because of the immense drop in union membership, unions had to deal not only with aggravated financial issues but organizational issues as well. The non-ballot-related changes affected the decisions made by union leaders and also interacted with the balloting legislation. This chapter provides a discussion about membership, merger, and structure in order to better understand the impact of membership loss to union behaviour and government issues. Because unions initially adopted individual strategies to minimize the damages triggered by the changes, a movement was made to establish a more centralized government. The move to reform organizational structure made way for the development of new policies concerning employers and partnership.

Keywords:   membership loss, financial problems, centralized government, reform, merger


The industrial relations climate of the 1980s and early 1990s moved, as shown in the previous chapter, strongly against trade unions’ interests, particularly towards the end of the period. Partly as a result of this development, many unions experienced significant and unexpected falls in membership. Following the rapid growth of union membership and union influence in the 1970s, the unions were not well prepared, either organizationally or financially, to deal with such marked changes in their fortunes. There had been little reason in the 1970s to develop contingency plans for the hostile environment of the next decade and, despite the increases in membership between 1969 and 1979, the unions’ basic financial position in 1980 was not healthy. For, as Willman (1989) has shown, even in the halcyon days of the 1970s, income from membership subscriptions ran consistently below total expenditure. Unions were not therefore in a position to spend their way out of their difficulties even if they had known what new initiatives to fund. Moreover, the Conservative Government’s attempts to marginalize the unions politically and industrially compounded their difficulties and helped to serve notice that they could not expect much, if any, help from external sources. Unions were therefore thrown back onto their own rather limited resources. Membership loss was in these circumstances both a major change and the trigger for further changes in union behaviour.

As illustrated in Figure 1.1 (p. 5), the non-ballot-related changes both affected and reflected leaderships’ strategic choices and interacted with the balloting legislation to influence its effectiveness. Hence, before examining in detail in later chapters the balloting legislation and its effect on union elections and collective bargaining behaviour, we will consider some of the important non-ballot-related changes in unions which occurred over the same period. Given the severity of the impact of membership loss on other aspects of union behaviour, we first consider the closely related questions of membership, merger, and structure, before turning, second, to government. Issues of government include matters that affected, and in turn were affected by, both democratic and administrative rationalities. As we will show, changes in government generated tension between the two. (p.32) Third, and central to a union’s political rationality, we consider changes in objectives and means. We will examine each of these areas of change and comment on their causes. In conclusion, the interrelationship between these different aspects of union behaviour, and the relative influence of the various factors promoting change, will be considered.

Union Membership, Merger, and Structure

In the period of Conservative Government the number of unions and union members declined from the 1979 peak as shown in Table 3.1. In total, union membership fell by 28 per cent and the number of unions by 39 per cent. The TUC experienced a similar decline in its affiliation and membership. In 1979 the TUC had 112 affiliated unions and a membership of 12,128,078. By 1991 the number of affiliates was down to 74, a fall of 34 per cent, and its membership was 8,192,664, down by 32 per cent (see Table 3.2).

The 24 unions in our study (including the six in-depth case studies) experienced changes in membership over the period 1979–91 as shown in Table 3.3.

As this table shows, 20 of the 24 unions in our study suffered a loss of members between 1979 and 1991. The average loss per declining union was 30 per cent. The four unions gaining membership averaged an increase of (p.33) 17 per cent. Overall the 24 unions lost, in aggregate, 29 per cent of their 1979 membership. Thus, the general experience of the unions in our study of union practice was one of considerable membership loss. It was the exceptional union operating in the public—or recently privatized—sector that increased its membership: two of the unions with gains in 1989 recruited in the newly privatized BT and the other two recruited in the public sector. One of the unions enjoying a growth in membership, the NCU (23 per cent growth), was included in the case studies. However, it should be noted that the NCU secured a transfer of approximately 40,000 members from the CSPA in 1985 and that without this transfer it too would have experienced a marginal decline in membership. Also, as will be shown later, a number of other unions also mitigated their membership loss by merging in this period.


Table 3.1. Trade unions: numbers and membership, 1979–91


Number of unions at end of year

Total membership at end of previous year (thousands)

Percentage change in membership since previous year





















































Source: EG 5.93: 191.

Table 3.2. TUC affiliated unions and membership, 1979–89


Number of unions


Percentage change in membership since previous year





















































Source: TUC Congress Reports, 1979–91

Aggregate union density also declined over the period 1979–91. Changes in density are not easily calculated.1 However, there is no doubt that union density fell significantly throughout this period. Table 3.4 indicates the extent of the change in density in the UK between 1979 and 1987 (in this table the unemployed are included as members of the labour force used for calculating the percentage in trade unions). Using data from the 1989, 1990, and 1991 Labour Force Surveys, it has also been calculated that union density fell again to 38.8 per cent in spring 1989, 38.0 per cent by the following spring and to 33 per cent in 1991 (EG 1991: 342).

Table 3.3. Trade unions in ballots study: membership change, 1979–91




Percentage change















                   Changing Union Behaviour





















































                   Changing Union Behaviour



























(*) Case-study unions

(1) The figure for the NUR in 1991 is the combined membership of the NUR and NUS resulting from their merger in 1990 to form the RMT.

(2) The figure for the ASTMS in 1991 is the combined membership of the ASTMS and TASS resulting from their merger in 1988 to form the MSF.

The Workplace Industrial Relations Survey (WIRS) of 1990 also provides further information on changes in union density over the period 1984–90 (Millward et al. 1992). The survey found that union density declined from 58 per cent in 1984 to 48 per cent in 1990, with a more marked decline among manual than non-manual employees. One of the most striking features of the 1990 survey was that there was an increase in the proportion of workplaces with no members at all and a significant movement away from union membership across all workplaces in the (p.35) trading sector (the traditionally highly unionized sector) across the period 1984–90. This decline in density was experienced in workplaces with both high and low densities of membership in 1984.

Table 3.4. Trade union density


Per cent



















Source: Disney (1990).

Various explanations or hypotheses have been advanced for the above decline in union membership and density. These explanations vary according to the following interrelated dimensions of the studies conducted. The research method, or dominant research method, predetermines the nature of the findings to a great extent. If the researchers seek to explain aggregated data and work top-down, rather than explore disaggregated data and work bottom-up, they will tend to one set of explanations rather than another. Top-down analysis lends itself to quantitative and deductive methods and bottom-up to qualitative and inductive methods of research. If those seeking to explain the cause of aggregated union growth work on data from the top down, they invariably use econometric modelling, select their data, test their correlations for significance, and then search, often imaginatively, for an explanation of the process which supports their findings. The result is that they almost always conclude that it is the external (i.e. external to the union as an institution) variables, which they managed to quantify in order to include them in their model, that caused the changes to occur (see Disney 1990). In contrast, the more qualitative researchers, such as ourselves, tend to focus on the interaction between external and internal variables within a particular union setting. As Lincoln and Gubba (1985) argued, this can lead to well-grounded and hence more credible and dependable theories concerning the process of change, but it limits the confidence with which these findings can be transferred and used to explain changes in the wider population or aggregated data.

The external and internal explanations for union growth have recently been re-examined by a number of academics seeking to advance (p.36) understanding in this field of work by drawing on the hypotheses developed to explain the growth in the 1970s to also explain the decline of the 1980s. There have also been some attempts to explain union decline by developing a new set of variables or by significantly amending previous models. Also, the OECD (1991: 97–131) tested a number of hypotheses by reference to changes in union density across its member countries. The econometric modellers have continued to argue, often implicitly, that unions primarily, if not totally, ‘receive’ (Undy et al. 1981: 160–3) or lose members because of factors outside the union’s control. In their models, unions are generally found to have played little part in determining growth or decline. Members—or potential members—are assumed to be indifferent as between unions. Unions therefore grow or decline in response to an undifferentiated demand for their services or lose members from a lack of demand. Employees’ propensity to join a union may be found in these models to vary, inter alia, according to the level of employment, the composition of employment (gender, full-time/part-time workers, size of establishment, sector, geographic location, manual/non-manual workers, age of worker (Green 1992), the level of economic activity—primarily level of employment—inflation, and the real rate of wage increases (Disney 1990: 165–177). Political factors, and the employer’s/manager’s behaviour with regard to union membership, may also be said to influence union growth and decline. However, employers and managers are again assumed to be indifferent as between which unions they do or do not encourage or discourage in the recruitment of their employees.

We would see little grounds for disagreeing with Kelly (1989) that changes in employment composition have played a limited but significant role in reducing union membership in relatively closed industrial-type unions recruiting in the high-density territories of coal, steel, and rail in the early to mid-1980s. These unions have broadly sustained their high densities (coal 90 per cent; rail 94 per cent, and metal extraction and manufacture 64 per cent unionized) but still suffered a major loss of members due to the decline in employment in their territories. For reasons not satisfactorily explained by external factors alone, the compositional changes which increased employment in relatively low-density private service industries have also contributed to the reduction in aggregate union density. The Labour Force Survey (EG 6.91: 341–3) shows that in the largest single sector listed in 1989, i.e. the retailing of food, clothing, and household goods, only 17 per cent of the 2,206,220 employees were unionized. Similarly in hotels and catering, out of 939,209 employed only 11 per cent were unionized. A movement of workers, and therefore potential union members, to these traditionally low-density sectors has helped to reduce union membership. The OECD study (1991) further confirms that structural employment shifts have affected trends in unionization across its member countries, but that they do not alone (p.37) account for the changes in union density experienced in these countries. Also, a study by Green (1992) calculates that the composition effect contributed approximately 30 per cent towards the fall in aggregate density between 1983 and 1989. The WIRS study of the period 1984–90 also helped place the effect of compositional change in context by concluding that ‘changes in density within continuing workplaces appear to have been more important than compositional changes in the population of trading sector establishments in the overall decline in union membership’ (Millward et al. 1992: 68).

A number of quantitative studies point towards the business cycle’s having the major part to play in reducing union density. In the period of growing and high-level unemployment, between 1979 and 1986, the unemployment effect is claimed, when combined with real increases in wages, to account for a major part of the fall in union membership in this period. Kelly refers to a correlation of 0.72 between employment changes and union density for the period 1980–5, and Disney claims that ‘a period of rising unemployment, high real-wage growth and a Conservative Government is sufficient to explain the decline in density’ (Kelly 1989; Disney 1990: 168).

The above strong correlation between a number of economic and employment variables in the 1980s and the decline in union density suggests that external factors, and particularly the rise in unemployment, have in some manner not explainable by quantitative analysis reduced the propensity of workers to unionize. This is probably the result of some combination of economic and employment factors influencing the workers themselves directly, and indirectly affecting the employers’ and managers’ behaviour towards unionization and unions’ ability to recruit. However, between 1986 and 1989 there was an increase of 1.4 million in employees in employment (EG 6.9: 340), yet, as can be seen from Table 2.2, the TUC’s membership fell by 5.5 per cent in 1989 alone—the highest figure for four consecutive years of falling membership, i.e. –2.5 per cent in 1986, –3.5 per cent in 1987, and –1 per cent in 1988. This suggests either that there were new forces at work or the emphasis given to the importance of unemployment effects in the early part of the 1980s was misplaced.

Alternatively, it is suggested by some commentators (EG 6.91: 337–43) that the post-1986 decline in density is largely due to the compositional effect, as the increase in employment was predominantly in areas of low union density. This does not of itself explain why these areas have been and remain low union density. Moreover, when various authorities have attempted to quantify the impact of compositional effects they have tended to suggest that it accounts for considerably less than half of the loss of density in the period 1980–9.

A somewhat different explanation for the fall in union density, but with its origins in the same econometric system of analysis, is offered by (p.38) Freeman and Pelletier (1992). They conclude: ‘the vast bulk of the observed decline in union density in the UK is due to the changed legal environment for industrial relations’ (ibid. 165). Moreover, these changes are not the result of the legislation in general, but of those laws that #x2018;directly impact on organizing (ibid. 148)’. Although recognizing that the authors give due recognition to the role of unions and the resources they may, or may not, direct to organizing, it is still difficult to accept their conclusions as a satisfactory explanation for the decline in union density. As we will show, the key balloting parts of the legislation operated in a complex manner and require careful analysis before pronouncing on their effects on both the processes and outcomes they sought to influence. Freeman and Pelletier’s assumption that the far more limited legal attempts to reduce the union’s ability to organize had the intended effect on union and management behaviour is justified neither by studies of the impact of the Labour Government’s previous legislation on recognition and unionization (Bain and Price 1983: 19–20) nor by our findings on the working of the much more substantial and significant balloting legislation. Furthermore, it is necessary to distinguish between achieved and received union growth, and to recognize that the unions’ ability to achieve growth is far more dependent in the UK on factors other than the law (Undy et al. 1981: 127–66).

The question of union growth and decline therefore requires an examination of why it is that the trade unions themselves have not acted more effectively to stem the outflow of members, or recruit more members. Trade unions have continued to supply the services required by potential new members and have achieved an increase in new members by their own actions in this period: without this recruitment activity they would have lost considerably more members. For example, USDAW recruited 118,000 new members in order to add 10,000 to its total. Further, USDAW actually increased the percentage of members entering the union each year in the late 1980s and 1990 as compared to the early 1980s, as Table 3.5 shows.

The TGWU recruited 220,438 members in 1985, 238,093 in 1986, 241,450 in 1987, and 244,910 in 1988. Despite this recruitment of 944,891 new members over four years, its membership fell over the same period by 177,702 members. Unions are therefore recruiting new members, whether by their own achievements, or by merely receiving them into membership due to external factors. What they are not doing is recruiting sufficient numbers to replace those leaving. This is despite a number of recruiting campaigns launched by individual unions, such as the GMB and ‘Decision 84’, which committed this previously restrictive-passive2 union to a positive recruitment strategy and campaign. Also, the TGWU’s ‘Forward TGWU’ (1988 and 1989 versions) gave a high priority to recruitment. This followed and superseded the TGWU’s ‘Link-Up’ Campaign in 1987, which sought to (p.39) consolidate the union’s position in areas of existing strength while also extending into new areas of recruitment. Also, the TUC has co-ordinated and targeted some recruiting activities.

Table 3.5. USDAW: new members as proportion of total


Per cent





















Source: Upchurch and Donnelly (1992).

In a study of full-time officials’ recruitment activities Kelly and Heery (1989) concluded that despite a developing national commitment to active recruitment by the national leadership of some general worker unions there were considerable constraints on recruiting in ‘new job territories, where there is little tradition of union organization’ (ibid. 210). Full-time officials were found to devote little time to this kind of recruitment activity, in comparison with their other responsibilities. These findings are particularly important when taken in conjunction with the adverse compositional effects of the 1980s and the manner in which unions achieved growth in the 1970s. In the 1970s, unions positively oriented towards recruitment achieved growth by projecting a particular union image, effective recruitment campaigns, provision of more local recruiting agents, and increased security of membership and other methods (Undy et al. 1981: 161–6). ASTMS, in particular, worked hard to project a particular national image for itself and its General Secretary Clive Jenkins, while in practice it was locally highly flexible and ready to adjust to meet the needs of its new members. A militant image and successful strikes also brought rapid immediate increases in membership for a number of other unions, as they demonstrated their ability to ‘deliver the goods’.

Recruitment campaigns which targeted non-unionized groups in existing job territories were also markedly effective in the 1970s (ibid. 169). In some parts of the country, sole recognition agreements negotiated in new firms were important for achieving growth. Increased shop steward activity (p.40) in the TGWU, previously dominated by its full-time officials, also assisted recruitment. Finally, membership was made more secure and some of the very high losses which occurred through not following up on initial recruitment were staunched, if not stopped, by the more extensive use of check-off and the increased use of the closed shop.

In the industrial relations climate of the 1980s and early 1990s many of these means of achieving growth were not available to the unions. The advances made by increasing security of membership were largely ‘one-off administrative changes which could not be repeated. Shop steward power and influence were relatively muted and shop steward activity generally reduced. Union image—underpinned by the traditional values associated with combination and collectivism—may have been blunted by the Conservative Government’s promotion of individualism. Instead of strikes demonstrating union effectiveness, the most prominent strikes had the opposite effect. ‘The goods’, in terms of real wage increases, seemed available without recourse to industrial action. Moreover, having ‘mopped up’ many of the non-unionized workers in their own job territories, the full-time officials were incapable of pursuing, or reluctant to pursue, recruitment campaigns amongst the non-unionized groups in those industries previously shown as expanding, as work-force composition changed in the 1980s (Kelly and Heery 1989). Many unions also lacked the financial resources necessary to support the staff increases needed if more resources were to be put into direct recruiting in new territories. Finally, as we will show below, individual unions which could have been expected to pursue more vigorous recruitment campaigns became as much, if not more, concerned to increase their membership through mergers. Thus, in the circumstances of the 1980s and early 1990s, the means of achieving growth previously exploited by the expansionist unions of the 1970s were not generally appropriate for recruiting in the newly expanding job territories which were traditionally difficult to unionize.

In future studies of union growth and decline there may therefore be advantage in focusing attention on both received and achieved growth in the same research framework. The assumption that all potential members (and managers and employees) are indifferent as between unions should be abandoned. It should be accepted as a departure point for such research that unions may, to some degree, influence their own rate of growth and decline and in so doing affect aggregate growth and decline. Further, the often implicit assumption that the same exogenous factors affect all phases of growth and decline should also be questioned.3 The research method used could probably be improved by examining flows of members in and out of unions and by assuming that different factors may affect the ability of unions to recruit and retain members within workplaces with different densities, i.e. at different phases of growth and decline. The relative importance of the different phases for union growth and decline should be (p.41) considered. This may be represented diagrammatically as a progression, in either direction, through four quadrants or phases, as shown in Figure 3.1.

Figure 3.1 is intended to provide a framework for examining union growth at both the workplace and subsequently the aggregate level of analysis, by combining consideration of both achieved and received growth (and decline) at each of four interrelated phases of union organization or density. Each of the quadrants represents a phase of union development within the workplace. The levels of density allocated to each phase were (p.42) chosen rather arbitrarily. However, the underlying assumption is that there are some critical points in union growth and decline which, if achieved, may produce further significant changes in the level of density at the workplace. In the diagram these are: (1) between phases 1 and 2, representational rights, involving the union gaining or losing its right to represent its individual members in, for example, grievances and disciplinary procedures; (2) between phases 2 and 3, recognition rights, involving the union gaining or losing the right to bargain collectively on pay and related issues; (3) between 3 and 4, the power which the union may accrue (or lose) which enables it to ‘mop up’ free riders by exerting pressure for a de facto closed shop, or similar arrangements guaranteeing a very high level of union membership.

                   Changing Union Behaviour

Fig. 3.1. Framework for research into union growth and decline

Whether or not these phases are interrelated and ‘breakthrough’ points do occur at the levels of density ascribed to them in this model requires further research. The notion that the phases may be so closely associated with each other at the breakpoints that relatively minor movements of employees (represented by the broken curves—clockwise for growth and anti-clockwise for decline) at the margin will subsequently produce much larger movements in density, and possibly move the workplace into a new phase of growth or decline (represented by the unbroken curves—clockwise for growth and anti-clockwise for decline), should be explored. If this is so, those factors working at the margin to move workplaces into new phases of growth and decline will have the most impact. Thus, not all combinations of received and achieved growth will have the same effect. Research into the factors which move workplaces between different phases may therefore offer considerable advantages over research that treats all movements in membership as being around a common density, i.e. the national average. The relative importance of achieved and received growth is likely to differ according to the phases studied.

In phase 1, achieved growth is probably the main source of the initial unionization, through full-time officers’ or activists’ direct recruitment of the non-unionized, or, in combination with received growth, through the union signing single-union-type agreements with the employer prior to workers being hired. This phase, if the result of direct recruitment, is also likely to be the most fluid, with fairly rapid movements back into non-unionism should the initial recruitment fail to produce sufficient members to claim at least representational rights.

Phase 2 and movement into the critical phase 3, or vice versa, is likely to be determined by a combination of achieved and received growth factors. In the circumstances of the 1980s and early 1990s it is difficult to see how unions in phase 2 could force recognition for the purposes of collective bargaining on a recalcitrant employer, although it may be easier to organize resistance to its withdrawal. The power balance of the 1980s and early 1990s has moved against unions, and in the absence of a neutral or (p.43) sympathetic employer or management it is unlikely that a union with a density of less than 50 per cent could achieve recognition and break through into phase 3. The danger for the union is that failure to achieve such a breakthrough may leave it with representational rights, but these may not be sufficient to retain the support of existing members. The union could therefore find itself moving rapidly back into phase 1 and through it into non-unionism. On the other hand, a union in phase 3 may well be able to mobilize opposition to the withdrawal of bargaining rights. It could well be the case that the low incidence of the withdrawal of bargaining rights in the case of manual workers is at least partially accounted for by the latent power which managers perceive unions as retaining in reserve. Thus, the maintenance of high levels of density found in some workplaces by the Workplace Industrial Relations Survey 1990 (Millward et al. 1992: 60–70) may well be the result of strong union organization and prove a major factor in maintaining overall density at its present level.

Movement between phases 3 and 4 and the enforcement of 100 per cent unionization on the ‘free riders’ has probably been made more difficult by external factors and particularly the Conservative Government’s actions in outlawing the closed shop. At least some of the achieved growth of the 1970s was the result of union pressure for and management acquiescence in such agreements. However, some unions have managed to maintain 100 per cent membership in some workplaces despite these changes. Why this is possible, and why other unions have failed to stop a movement of members out of other high density workplaces into non-unionism, is also an area for further research.

The framework for analysis suggested above therefore starts with recruitment in the workplace and perceives union growth and decline as the product of a series of interrelated phases. It is a dynamic rather than a static scheme of analysis. It further assumes that each phase of development may be determined by different combinations of achieved and received growth and that, at critical points, movements of members between phases may have disproportionate effects on the next phase of growth or decline.

In conclusion, as regards union membership: many unions, including most of the leading unions in the TUC, suffered from a substantial reduction in union membership in the 1980s and 1990s for which the ‘good years’ of the 1970s had left them ill-prepared. At the same time, aggregate union density also fell significantly. There has been limited success in explaining the causes of this overall decline in density. A number of studies have shown that the environment, including, inter alia, unemployment, compositional changes in employment, and the Conservative Governments’ legislation, adversely affected trade union growth. However, attempts to quantify the extent to which each of these developments reduced union density have produced conflicting results.

(p.44) It is probable that a combination of the above external factors encouraged employers or managers hostile to unionization to resist attempts at recruitment of their non-unionized workplaces, and to seek lower levels of membership amongst those already unionized. Unions, weakened by the developments of the 1980s, appeared largely incapable of making inroads into the traditionally hard to unionize areas of employment, such as private services. Moreover, unions also seemed to find it difficult after 1984 to prevent significant outflows of members (and reductions in density) in workplaces in those sectors previously seen as strongholds of unionization. Nevertheless, there is evidence that even unions losing overall membership continued to recruit new members at reasonably high levels during this period. How this inflow and outflow of members was actually affected by the adverse external factors of the 1980s and 1990s is unclear and requires further research on the process and dynamics of union growth and decline. However, the resulting loss of members, and the uncertainty as to how, if at all, it could be reversed, had a most marked effect on union behaviour. Besides adopting new schemes of direct recruitment they sought to mitigate the loss of members and associated problems, including growing financial difficulties, by other means. For some unions one of these means was to seek mergers.

Union Mergers4

Trade union mergers were given a significant boost by the Trade Unions (Amalgamation etc.) Act 1964. This eased the previously restrictive balloting process required for amalgamations and introduced merger by transfer of engagements, whereby only the minor union in a merger was required to ballot its members and secure a simple majority of those voting for the merger to be agreed. The major or acquiring union in a transfer is not required to ballot its members. In an amalgamation all unions involved have to ballot their members. As shown in Table 3.6, between 1967 and 1989 there were 331 mergers (this includes all unions involved in amalgamations but only transferring unions in transfers of engagements). Also, as shown in Figure 3.2 there was, in respect of TUC-affiliated unions, a sharp upward movement in the proportion of unions merging from 1979 onwards. As we will show later in this chapter, union mergers, but particularly amalgamations, had a significant effect on the internal organization (or government) of unions.

The post-1979 merger movement as identified in Figure 3.2 at present under way, shows no sign of abating. In a survey of TUC affiliates conducted in 1990 and 1991, 60 per cent (42) of the unions surveyed did not reject merging within the next five years and 23 unions indicated that they were intending to search actively for merger partners in this period. Hence, (p.45) the slight upturn in merger activity shown in Figure 3.2 for 1990 is likely to continue its upward movement into the mid-1990s and beyond.

Table 3.6. Union transfers of engagements and amalgamations by years, 1967–89


No. of transferring unions

No. of transferring members

No. of amalgamating unions (no. of unions after amalgamation)

No. of amalgamating members

Total members




9 (4)






4 (2)






2 (1)






9 (3)






8 (3)






5 (3)






7 (2)










5 (2)












2 (1)






2 (1)










3 (1)










10 (4)






2 (1)






6 (2)






3 (1)










2 (1)






4 (2)






2 (1)









Total unions involved (244 + 87) = 331

Per cent of total transferring or amalgamating 1967–89:







Note: The figure for the total number of unions and union members used in the calculation of percentages were taken from EG (5.89 and 5.90) and from Marsh (1988).

Quantitative research into mergers stresses the importance of exogenous factors in shaping the merger movement and seeks to correlate aggregate mergers with economic factors such as the business cycle, wages, retail prices, and also, in some cases, political factors (Waddington 1988). There (p.46) are a number of problems inherent in using this approach. One key difficulty is that a merger is, in many cases, the end-product of extended negotiations, in some cases covering several years. Although the date on which mergers are legally finalized is a matter of recorded fact, the starting point of the negotiations is, in major mergers, frequently a question of dispute requiring research in its own right. Hence, to seek to correlate mergers (end dates) with the economic and political developments that may have triggered the merger search some years before is unlikely to shed much light on what actually determined individual mergers or, in aggregate, the merger movement. Further, external factors which may trigger the merger search are mediated through the union leadership and may change in importance during the merger process. They are, therefore, indirect and complex in their effect and vary from merger to merger in their importance according to the weight given them by the negotiators.

                   Changing Union Behaviour

Fig. 3.2. TUC merger activity: 5-year moving averages 1969–91

External factors tend to exert most influence on mergers at the start of the merger process. In particular they can trigger a merger search when they combine to produce a decrease in a union’s membership (or growth expectations) and/or adversely change the composition of a union’s membership. In these circumstances the major merging union may seek a merger as a means of restoring, partially or fully, its membership loss and/or as a means of extending its interests into a more promising job territory. The minor union’s search for a merger may also be initiated following a loss of members, particularly if the loss threatens the minor union’s survival. Nevertheless, membership loss alone is not necessarily the determining factor in merger search. Unions vary widely in terms of their merger intentions and actions when faced by significant membership loss. In the survey of TUC unions referred to above, 28 (40 per cent) of respondents expressed no future (five-year) interest in mergers. These (p.47) were mainly small unions. Their average size was 12,927 (median 3,770), and 21 of them had lost members since 1978 and 15 had lost over 30 per cent of their membership in this period. In contrast, 42 (60 per cent) of the unions did not reject mergers over the next five years and 23 of these indicated an intention to search actively for mergers in this period. The 23 actively searching unions included almost all the large and medium-sized unions in the TUC and accounted for 83 per cent of the TUC’s total membership. These unions had, on average, suffered only an 11 per cent loss of members over the period.

The interrelationship between membership loss, financial performance, and merger intentions is therefore quite complex and varies between unions (Little 1991). As Table 3.7 shows, there is no overall positive correlation between intention to merge and membership loss, reduction in investment income, declining subscription income, or change in total income.

Thus, the external factors most likely indirectly to trigger a merger search were not sufficient in themselves to cause a merger. Indeed, unions affected by widely different changes in membership and financial well-being may react in a similar fashion to questions concerning their merger intentions. For example, three unions with over 20 per cent membership growth rates in the TUC between 1979 and 1989—the British Actors Equity Association (BAEA; 60 per cent), the National Association of Probation Officers (NAPO; 30 per cent), and the Fire Brigades Unions (FBU; 21 per cent)—perhaps not surprisingly expressed no future interest in mergers, but so did 9 small unions each experiencing a reduction of over (p.48) 40 per cent in members. Similarly, unions intending to search actively for mergers included two unions with a loss of over 40 per cent in members, i.e. the AEU and the Iron and Steel Trades Confederation (ISTC), and two unions with an increase of over 20 per cent, the NCU and the Banking Insurance and Finance Union (BIFU). There were also marked variations in the financial status of unions amongst those both intending and not intending to merge. Hence, while some merging unions may be driven to search for a merger by external factors which adversely and indirectly affect their size and financial position, not all unions respond to the same factors in the same way. It is therefore how union leaders interpret these developments and their reasons for pursuing mergers which need to be explored.

Table 3.7. Merger intentions, membership and finance: percentage average changes 1979–89 in membership, investment and subscription income


No future interest in mergers

Actively interested in mergers

Not rejecting mergers

All trade unions






Investment incomea





Subscription income





Total income





(a) All prices used were adjusted to 1989 values.

(b) This calculation omits the TSSA, which showed an increase of 4200 per cent and in consequence distorted the group’s mean. If included it raised the group’s increase to 327 per cent.

In examining why unions merge, it is important to distinguish between the major and minor merging unions and between transfers of engagements and amalgamations. The leader of the minor union negotiating a transfer of engagements is normally involved in surrendering autonomy in exchange for some gain such as security. In contrast, the major union’s leaders in a transfer of engagements are primarily seeking to enhance in some way the standing of the larger union and they will probably not change significantly their present system of organization and power structure in order to accommodate the incoming minor union. However, in an amalgamation, the two parties are likely to see themselves as having similar status or as equals—otherwise why not merge by transfer of engagements? In merger by amalgamation both parties may be required to make significant changes in their rule books and organizations in order to agree the formation of a new union. In the following discussion, the major merging unions’ reasons for merging will be examined before turning to examine the motives of the minor merging unions (amalgamations will be taken as involving only major unions).

Between 1978–89 a total of 98 unions were merged by transfer of engagements with TUC-affiliated unions. As regards amalgamations, nine TUC affiliates joined with eight other unions. The transferring unions had on average 8,328 members and there were 64,327 members, on average, in the smaller of the amalgamating unions. Most of the major merging unions sought mergers primarily to increase membership and/or change their job territory. A majority of the major unions considered their mergers to be primarily consolidatory5 inasmuch as the merger built on an existing presence of some strength in an industry or amongst a particular occupational group. Of the 27 major merging unions examined, 14 fell into this category, i.e. were consolidatory. Three of the remaining major unions had been involved in defensive mergers, attempting to safeguard a position previously weakened by loss of members or increased recruitment competition from some other union. Two other major merger unions tended to adopt a predominantly aggressive or expansionist approach to (p.49) mergers. For these unions mergers were a vehicle for increasing the membership and expanding the territorial base. At certain points some of the primarily consolidatory unions also entered into expansionist mergers. The remaining eight unions did not seem to fit any of the above categories. They appeared not to have a merger policy. They were generally opportunistic or unplanned in their merger activities. Hence, the majority of the major merger unions did not use mergers to expand outside their existing industrial territories but to improve their situation within existing industrial boundaries by recruiting the same or new occupational groups. Where they involved new occupational groups these tended to be the recruitment of craftsmen into predominantly white-collar/technician unions and the addition of white-collar workers to predominantly blue-collar unions.

The above comment on the consolidatory nature of most major unions’ merger activities needs to be balanced, however, by reference to the actions of those unions which pursued a more expansionist merger policy. In particular, the Association of Scientific Technical and Managerial Staffs (ASTMS) and the Technical, Administrative and Supervisory Section (TASS) (prior to their amalgamation and the formation of the MSF) and the GMB (formerly the GMWU) were more expansionist than other unions in their approach to mergers. Also, the TGWU, although not showing the same vigorous pursuit of mergers as it displayed in the 1970s, was still keen to exploit any opportunities which arose for expansion by merger. Between them the ASTMS, TASS, the GMB, and the TGWU accounted for 46 mergers in this period. The ASTMS, GMB, and TGWU covered a relatively wide area of job territory at the start of this period, although the ASTMS was mainly confined to white-collar recruitment. Thus their merger searches, even if they remained within existing industrial territories, still had a wider industrial scope than that enjoyed by more ‘closed’ unions. Further, the GMB and TASS deliberately used a number of mergers to take them into new occupations and industries. TASS (a draughtsmen’s and technicians’ union), for example, merged with a series of four small blue-collar engineering unions and in 1986 with the Tobacco Workers Union. The GMWU’s break with its past, rather passive, approach to mergers occurred in 1982 when it finally amalgamated, at its second attempt, with the craft-based Boilermakers, and in 1988 it again broke new ground when the GLC Staff Association joined it by transfer of engagements. The GMWU’s amalgamation with the Association of Professional, Executive, Clerical and Computer Staff (APEX) in 1989 to form the GMB was also intended to expand radically the union’s rather minor white-collar membership. Thus, the most active merger unions expanded the scope of their industrial and occupational territory significantly whilst other major unions followed more conservative, consolidatory, and defensive mergers.

(p.50) Apart from territorial gains the major merging unions also sought other advantages from mergers. Seventy-seven transfers of engagements and 14 amalgamations were examined from the perspective of the major merging union in an attempt to identify non-territorial advantages. Three key areas of possible advantage were identified, as shown in Table 3.8. An open-ended question was also posed seeking information about ‘other advantages’, but respondents referred to mainly territorial-type gains in answering this question.

As can be seen from Table 3.8, only the question concerning an increase in bargaining power received over 50 per cent positive responses. However, in subsequent interviews this positive response was frequently qualified, and only in a minority of cases was a merger mentioned specifically as having a direct and positive effect on bargaining power. For example, the TGWU noted that its merger with the Northern Textile Allied Workers Union in 1984 gave it a presence in national negotiations previously denied it. Also, the NGA argued forcibly that its merger with SLADE considerably enhanced its bargaining power. There were also a number of general comments about improved co-ordination of bargaining following a merger. Nevertheless, mention was frequently made in interviews of the difficulties in achieving an increase in bargaining power when the minor or acquired union was experiencing a decline in the demand for its members’ services, and in many cases sought a merger because of a serious loss of members. Hence, claims for an increase in bargaining power in the difficult economic circumstances of the 1980s need to be treated cautiously.

The increase in political power took two main forms. Some unions saw the merger creating a stronger political alliance of like minds, for example, the TASS–Tobacco Workers Union merger. Secondly, amalgamations between relatively large unions were thought to increase the new union’s political influence in the wider labour movement. The GMWU’s merger with APEX to form the GMB and the AEU’s merger with the EETPU to form the AEEU came into this category. In respect of financial gains, the major unions overall referred more to the financial cost of mergers (p.51) than to the savings. It would therefore appear to be the exception rather than the general rule that mergers lead to immediate financial gains.

Table 3.8. Non-territorial advantages of mergers




Increase in bargaining power

48 (77%)

13 (21%)

Increase in political influence

21 (35%)

39 (65%)

Financial savingsa

23 (37%)

40 (63%)

(a) Some respondents qualified a yes answer with reference to achieving savings in the longer term.

One of the key influences on large amalgamations was the drive to reduce inter-union competition for members and inter-union disputes vis-à-vis a common employer. The UNISON, GPMU and MSF amalgamations respectively involving NALGO, NUPE, and COHSE (1993), SOGAT and the NGA (1991), and the ASTMS and TASS (1988) were influenced by such considerations. Technological change in the print industry created regular and fierce demarcation disputes between SOGAT and the NGA and resulted in intense competition for members. The merger was seen as one way of helping solve this problem. The ASTMS–TASS merger was also influenced by attempts to end the two unions’ damaging competition for members. NALGO, NUPE, and COHSE started merger talks after a number of disputes between NALGO and NUPE in particular, and complaints by NALGO that NUPE was poaching its members. For example, at the 1988 NALGO conference a motion supporting the merger referred to such ‘inter-union disputes…threatening effective trade unionism as a whole’ (NALGO 1988: motion 116).

Lastly, some major unions also sought or used mergers as a means of reorganizing their own internal system of government. Mergers, and in particular large amalgamations, provide an opportunity for, and could be used to legitimize, a radical review of a union’s system of internal government. A number of union leaders with ‘problems’ of internal organization—for example, too high a proportion of funds under the control of local organizations, or powerful regional- and/or district-level officials—may see mergers as a way of resolving these difficulties via the rule changes required to accommodate the merging unions’ different internal structures and processes of decision-making. Moreover, in order to make the major unions more attractive to minor merging unions faced with a number of potential partners, the major union may adjust its own organization to provide the incomer with more autonomy and in this process also ‘correct’ some perceived governmental shortcomings. The GMB in particular, when pursuing mergers, took the opportunity to adjust its system of government for other purposes.

To conclude: in the 1980s and continuing into the 1990s there was an increase in merger activity. This ran parallel with and was influenced by the loss of members discussed previously. However, not all unions losing members and suffering the consequent financial insecurity merged. Some small unions preferred to remain independent, even though adversely affected by external developments. These unions would appear to prefer a slow death to merging with a larger organization. Major merging unions tended to seek mergers to consolidate their position in existing job territories. However, a small number of more expansionist unions had a disproportionate effect on the union map. They used mergers to break new (p.52) ground and in some cases to demonstrate their effectiveness as future potential merger partners. Further, the mega-mergers between the AEU and the EETPU in 1992, and between NALGO, NUPE, and COHSE in 1993, added further impetus to the merger movement in the 1990s. They will intensify pressure on other large unions to merge in order to protect their relative influence within the wider labour movement. Thus with little sign of a recovery in union density, and with the potential minor merging unions having little if any chance of penetrating new job territories, the merger movement will continue as the major merging unions search for new partners in the late 1990s.

Union Structure

Union structure, following convention, is taken as being the ‘morphology or external structure…the coverage by each union of industries and occupations’ (Clegg 1979: 165). There have been various attempts to classify union structure. The intention of the classifications has been to help ‘separate the elements in a situation that would otherwise appear confused and chaotic’ (Turner 1962: 24). The terms ‘craft’ (occupation), ‘industrial’, and ‘general’ were initially used to distinguish between unions which recruited, respectively, a particular group of craftsmen or several related crafts, those which recruited (or attempted to recruit) all workers within a particular industry, and those which would recruit any and all workers. These ideal types were subsequently amended, or dropped and replaced by such concepts as open and closed unions, sectoral unions, public and private sector unions, white collar unions, conglomerate unions, and enterprise unions.6 Undy et al. (1981) attempted to produce a more dynamic model of union structure in terms of the degree to which unions became more or less occupationally and industrially diverse over the period 1960–75. In sum, it was found that virtually all the major unions in the TUC became more industrially and/or occupationally diverse over this period. In terms of other classifications they became more ‘open’ and moved towards general unionism.

The degree of union diversity was increased by a small number of highly significant unions which expanded beyond their 1970s industrial boundaries by merger. Also, the Government’s privatization policies took public sector unions into the private sector. Hence, the 1970s trend towards more occupationally and industrially diverse unions gathered pace in the 1980s and 1990s. However, the kind of mergers agreed did not generally offer the major unions much scope for penetrating low or non-unionized territories. An exception to this was BIFU, which was successful in attracting into mergers a number of Staff Associations which offered considerable scope for direct recruitment of non-unionists in the growing employment field, in the mid-1980s, of banking, finance, and insurance.

(p.53) In the 1980s and early 1990s union structure therefore became more and not less occupationally and industrially diverse. The previous consensus around industrial unionism was dead. Each major merging union, in what was frequently a highly competitive search for minor merger partners, therefore sought a merger for its own purpose. Some of the major union leaders saw scope for intervention by a higher authority—perhaps the TUC—in someone else’s merger negotiations, but they did not wish to surrender their own sovereignty. Also, the presence of expansionist merger unions meant that a large proportion of the minor merging unions had more than one merger option and they too would not have welcomed being pressed, by some third party, into mergers on less favourable terms than could be negotiated freely. Thus, although external factors both reduced union membership and made the recruitment of replacement members more difficult, it was the manner in which union leaders responded to these and other external and internal pressures that determined union morphology. Some small unions faced with loss of members and financial difficulties chose to do very little; others sought refuge in larger organizations. Some major unions sought to expand occupationally within existing industrial boundaries; others sought to break into new industrial territory. In a small but influential number of larger mergers, political alliances seemed more important than territorial gain, and in some mergers the opportunity they offered for increased solidarity, vis à vis the main employer, was critical in determining the choice of merger partner. Union leaders, although constrained by financial pressures and in some cases by their internal decision-making machinery, therefore exercised considerable discretion in choosing their merger partner. In the absence of a widely accepted notion of what was an ideal structure, the result was that many unions chose partners for internal reasons and in consequence the external map became more and not less diverse. Nevertheless, a new kind of industrial or trade group structure did emerge from this process. This was because the major unions involved encouraged and facilitated those changes in job territory with greater internal differentiation between different groups of members.

Union Government

In using the term ‘union government’, we mean unions’ internal decision-making. This may be described in various ways and is an important aspect of both democratic and administrative rationalities. But for the purpose of this discussion, and prior to examining the effects of the balloting legislation on union government, it will be considered according to the analytical taxonomy developed by Undy et al. (1981: 37–60) for the analysis of changes in unions’ internal decision-making. First, we will (p.54) consider the vertical bifurcation of union government into bargaining and non-bargaining decision-making channels and processes. Second, the degree to which decision-making is centralized or decentralized within these channels will be discussed. Third, the degree to which decisions at the different levels are concentrated or diffused will be examined. Reference will also be made in this latter discussion to the informal processes and bodies of decision-making.

First, the merger movement of the 1980s and early 1990s caused a number of previously predominantly single-channel unions to adopt a bifurcated system of government, with responsibility for some bargaining decisions and industrial or trade-specific issues being allocated to a new separate channel, or channels, of decision-making. This development was most notable in the case of the GMB, and to a lesser extent in TASS prior to its merger with the ASTMS. The AEU also made a more tentative move in this direction before its merger with the EETPU.

The GMB in particular further extended and formalized the movement it tentatively first made towards vertical bifurcation in 1969 when it formed advisory national industrial committees. Following its merger with the Boilermakers in 1982 and more importantly its merger with APEX and the Greater London Staff Association in 1989, the union initiated a major reorganization of its internal structure. As the union commented in its special merger report, Shaping up for the Next Century, it decided ‘not to try to merely bolt together the two existing Unions [i.e. the GMWU and APEX] and their various ruling bodies and institutions but to set out to create a new integrated Union’ (GMB 1988: 4). The solution was to form a number of sections or trade groups combining the members of all the merged unions according to their occupation or trade. It also served to make the GMB itself more ‘merger-friendly’ by providing greater potential autonomy for incoming unions than would have been provided under its previous form of government. Other large unions involved in mergers also tended to adopt a similar sectional or trade group system as a means of accommodating the incoming union. Hence, although mergers led to greater diversity of membership within the individual union, the new larger unions became more highly differentiated internally in the vertical plane in order to meet and service more effectively the needs of their enlarged membership’s different occupational and industrial interests. There was therefore a movement towards what may be termed a ‘Transport & Generalization’ of larger unions as they copied to varying degrees the TGWU’s long-standing system of vertical bifurcation.

Second, the degree of centralization/decentralization in the single or dual channels of decision-making varied between unions in the 1970s and continued to do so in the 1980s and 1990s. As regards the bargaining channel: in the 1970s a number of large unions, for example the TGWU and NUPE, encouraged decentralization, whereas some, such as UCATT, (p.55) preferred to retain centralized bargaining. In the 1980s and 1990s, employers tended in the main to devolve pay bargaining away from national or industry-wide negotiations, the major exceptions being general printing and some parts of the public sector.7 Some unions, such as those in British Rail, opposed such changes and had some success in delaying or changing management’s proposals. However, in the private sector in general and in some parts of the public sector, unions were faced with little choice but to accept management decisions to decentralize pay bargaining. At the same time, unions faced with major financial problems were not in a position to provide more full-time officer support to service the growing number of local negotiations. Indeed, a number of unions were cutting back on the number of local officials and merging district offices in order to save money. The AEU, for example, introduced a package of cuts in 1990, including a reduction in the number of full-time officials and a reorganization of its branch structure in an effort to reduce its deficit. Some unions, for example the NCU and NUR, attempted to ‘shadow’ bargaining devolution by reshaping their own internal bargaining machinery (Steele 1990: 62).

Overall, since the early 1970s there has therefore been a trend within unions towards further decentralization of the level of pay bargaining. This was primarily a response to employer initiatives to lower the level at which pay was determined and union inability to sustain higher-level negotiations against the employers’ wishes. However, as will be shown in the later chapter on balloting in industrial disputes, the degree of control exercised by the leadership over the process of decision-making on industrial action has itself been centralized. Thus, the level of pay bargaining has been devolved, while the national leadership’s control over the use of sanctions in support of bargaining has been centralized.

Decision-making in non-bargaining channels has tended to become more centralized. In the 1980s centralization of non-bargaining issues occurred in the AEU, GMB, and the Civil and Public Services Association (CPSA), and within the TGWU in the early 1990s. The AEU’s executive promoted rule changes in the 1980s which gave it greater powers over the reorganization of the union’s branch and district boundaries and the movement of full-time officials and the determination of their duties. The executive also used the referendum to bypass intermediary bodies and pose questions directly to the membership on the controversial question of accepting government funds for ballots. Further, the executive sponsored a rule change which limited the rights of incumbent officials to contest other posts should their own post also be up for re-election. In sponsoring some of these changes the dominant ‘Group’8 within the executive also appeared to seek political advantage in elections affected by their boundary changes and by the restrictions placed on certain incumbent officials.

The GMB and CPSA both started the 1980s more centralized than the (p.56) traditionally and relatively highly devolved AEU. The GMB’s reliance on a strong regional system of government has its origins in the creation of the GMWU in 1923 and it was still identifiably the same union in 1980. In the early 1990s the full-time regional secretaries continued to play a pre-eminent role in the union’s decision-making processes, but there were signs of movement towards more central direction of the union’s policy and strategy. The merger process produced a trade group type structure with the result that independent sources of influence developed at the national level outside the regional secretaries’ control. Further, the executive, and therefore in practice the general secretary, began to play key roles in the appointments of the regional secretaries in the late 1980s. While the regional secretaries continue to exercise considerable influence, they are now in a weaker position—at least numerically—on the executive. Finally, the election of an influential General Secretary (John Edmonds) with strong notions of where he wished to take the union itself served to centralize decision-making.

For the CPSA the major centralizing tendency in its non-bargaining activities has been associated with declining reliance on local branch organization and a movement towards regionalization. In 1989 a survey of its area co-ordinating committees found that one-third had not held a meeting in the previous eighteen months. Also, reduced time off for lay representatives under the National Facilities Agreement and an increased turnover amongst its branch officers suggested that the union’s local system of organization was in difficulties. The preferred solution adopted in the late 1980s was to appoint regional officials. However, this policy was frozen in 1991 because of financial difficulties.

The TGWU in 1992 ‘undertook a programme of substantial internal re-organization’.9 This followed a major review by external consultants and the executive led by Bill Morris, the General Secretary. One of the key areas of concern was the relationship between the regions and the head office. Tensions between regional (or district) organization and the centre are common to many unions. In the TGWU this was intensified by the union’s financial problems. The union had a significant deficit from 1989–92, including an operating deficit of £12 million in 1991. The General Secretary election in 1991, between Morris and Wright, was also seen by many as a head office versus regions contest. Some supporters of regional organization believed that the head office spent the ‘regions’ profits’ extravagantly while others, more supportive of the centre (or Transport House), contended that the regions behaved like eleven separate unions and did not adhere to national policy. The debate continued in 1992 and 1993 with a series of consultative meetings at all levels of the union around the strategy document ‘One Union T&’. An important part of the proposal is greater head office control of the union’s total finance, its regions, and their expenditure.

(p.57) The reasons for centralizing decision-making in the non-bargaining channel varied between unions. However, it is very clear that administrative concerns were very influential. The need to cut costs in the face of declining subscription income caused a number of unions to review their systems of government and question their efficacy. Drawing power back to the centre was seen as a means of exercising greater control over expenditure. It was often associated with attempts to reduce the money held at local and branch level and transfer all, or part, of this money to the head office. Further, economies were also made in expenditure incurred by the lower levels of the organization by reducing the frequency of meetings of intermediary bodies. Also, the pursuit of mergers in a competitive environment resulted in a higher degree of centralized control in the non-bargaining channel; in order to accommodate the interests of the incoming union and thus counter the competing union’s offer, the national leadership of the major unions needed greater discretion to adjust their own organization. Not surprisingly, in some unions this resulted in adjustments in the newly merged union which enhanced the power of head office over other levels of organization and raised questions about the union’s democratic rationality.

Third, in terms of the concentration or diffusion of decision-making there is little evidence that power became more concentrated in the hands of full-time officials, or the bureaucracy, in the bargaining channel, with the important exception of the influence of balloting legislation. Indeed, the decentralization of bargaining levels initiated by employers, and unions’ general inability to fund increased numbers of full-time officials to service the increase in bargaining units, point to a greater diffusion of power in the bargaining channel. This is also borne out by other studies of shop stewards’ independence of full-time officials in the 1980s (Fosh and Cohen 1990).

In the non-bargaining channel the question of whether or not there was a general concentration or diffusion of power at the national level, as decision-making became more centralized, depends on the relationship between formal and informal processes of decision-making and in particular the role of factions and parties. There appeared to be no generalized movement in this dimension of union government in the 1980s and early 1990s. However, changes associated with mergers, and, as we will show in Chapter 5, the imposition of postal ballots for elections, did serve to reduce the influence of the formal checks and balances provided by intermediate bodies. As a consequence the balance of power at the national level between conference, executive, and leading full-time officials broadly moved in favour of the executive and General Secretary.

The informal processes underwent significant changes in some unions, for example in the TGWU, and to a lesser extent in the MSF, as factionalism developed greater influence within both unions. The AEU, (p.58) NCU, and CPSA continued, much as before, to be influenced in their policy and rule-making at the national level by party or factional considerations. The AEU’s diffused but stable political system continued to be dominated by the moderate ‘Group’. The NCU and CPSA remained relatively unstable political organizations at the national level in the 1980s—if anything they became more diffused and volatile in this period. Both unions experienced political shifts in control of the executive and changes in the political composition of senior national full-time officials. The GMB and EETPU, by comparison, were politically stable and power remained relatively concentrated at the national level. EETPU resembled a one party—moderate—system of government. The GMB had no nationally organized party or faction, and the established Labour loyalist leadership still continued to dominate the decision-making processes despite the radical change in government structure associated with its aggressive merger policy. (The relationship between factional organization and balloting will be more fully discussed in Chapter 5.)

Union government in the 1980s and early 1990s therefore saw a movement towards more common forms of decision-making across unions with greater vertical bifurcation, primarily due to mergers; greater centralization of power in the non-bargaining channel, primarily due to financial problems; and more decentralized bargaining activity, primarily due to the actions of employers, leaving aside the impact of the balloting legislation on industrial action. At the national level there was also, in the major unions studied, either a continuation of, or a growth in, factional organization, except for the GMB and EETPU. However, the formal checks and balances on the powers of national leaders were weakened in some unions as the conference became less influential and power gravitated towards the executive and the General Secretary. It may be suggested that centralization and greater concentration of power in view of the turbulent environment and the associated difficulties facing the unions studied was an appropriate response given the need to protect the institutional framework of the union. It was, however, also associated with continuing or growing party or factional organization in some unions. In terms of democratic and administrative rationalities there were therefore some conflicting movements. Generally the drive for greater efficiency in government elevated the importance of administrative concerns, but the growth of factionalism and the continued use of existing checks and balances to diffuse power helped strengthen unions’ democratic tendencies. In Chapter 7 we will examine how these particular governmental developments interacted with the balloting legislation to change union government and union democracy.

(p.59) Objectives and Means

Unions’ political rationality includes objectives and means. These have been defined and classified in various ways.10 For the purpose of this discussion we distinguish between ‘business’ and ‘reformist’ unionism as a means of broadly categorizing overall union objectives (Dabscheck and Niland 1981: 98–9). Business unionism refers to unions whose objective is improving their own members’ limited sectional interest within the existing social system. In practice the union is therefore a bargaining agent seeking economic improvement for its members within the workplace. This leaves the union largely, if not solely, relying on collective bargaining as a means of achieving its goals. In contrast, reformist unionism, while also conscious of its members’ immediate economic needs and committed to collective bargaining on their behalf, also seeks to act in the wider interests of its members and others within society. In the UK this may involve an association with the Labour Party and the TUC.

The unions studied combined elements of both kinds of unionism. They did not consistently choose to occupy one or other of the extreme positions, fully business or fully reformist. Nevertheless, in some respects the ideological stance of the unions’ national leadership varied significantly, although they tended to adopt more pragmatic and opportunist positions on employment issues when it was to their advantage. Hence, in order to distinguish between unions we refine the use of ‘business’ and ‘reformist’ by introducing the notion of union character (Blackburn et al. 1968). ‘Character’ will be used to show movement, first, by reference to a union’s preference for becoming more or less employment- or society-focused and, secondly, by reference to a union’s preference for the use of moderate or militant strategy and tactics in pursuit of its employment objectives. As we will show, a union which adopts a moderate political position does not necessarily adopt the same stance consistently in its employment activities. Also, it should be noted that these political differences were almost always contained within a generally socialist or social democratic mould. Thus, the main characteristic was some form of association with the Labour Party; all the unions which were affiliated to the Labour Party at the start of the period reaffirmed their affiliation in the imposed ballots.

The political and industrial character of the EETPU was the most ostensibly moderate of all the unions studied. The EETPU is frequently referred to as the leading exponent of union ‘new realism’, which is in turn equated with business rather than reformist unionism. It involves movement towards employment-centred activities and the adoption of a moderate job regulation strategy and tactics. The emphasis is on co-operating with the employer and not assuming an adversarial posture. It reached its apotheosis in the EETPU’s single-union strike-free deals. In (p.60) these deals the EETPU offered employers, mainly on greenfield sites, a degree of functional flexibility and a disputes procedure incorporating pendulum arbitration. In return the EETPU wanted sole recognition, harmonization of terms and conditions, and some form of consultative machinery. By 1989 the EETPU claimed to have forty such agreements. Further, the EETPU was dominated nationally by moderates, or right-wing Labour supporters, and therefore represented the most politically conservative of the trade unions. This stance was also closely associated with a strong commitment to protecting its own sectional interest rather than those of the wider Labour movement. The union’s expulsion from the TUC in 1988 for failing to accept a decision of the TUC’s disputes committee reflected this commitment and further isolated the EETPU from the mainstream union organization. Nevertheless, it remained politically reformist through its continued affiliation to the Labour Party and, following its merger with the AEU in 1992, it rejoined the TUC.

The EETPU was not exceptional in signing single-union deals that excluded other unions from recruiting on greenfield sites. All the other large general unions in the study—the AEU, GMB, and TGWU—sought single-union deals and competed with the EETPU in union ‘beauty contests’. The AEU, in particular, had considerable success signing major deals with the Japanese vehicle builders Nissan in 1985 and Toyota in 1991 (FT 23.4.85 and 1.11.91). The AEU also sought a single-union strike-free deal with Ford in Dundee in 1987, bringing itself into conflict with other major unions, leading to a complaint to the TUC’s disputes committee (TUC 1988: 48–50). The craft-conscious AEU also negotiated Task Flexibility Agreements’ removing long-standing demarcation lines between different craftsmen in its willingness to co-operate with management in job regulation in the 1980s. Also, in common with the EETPU, it was led by politically moderate leaders throughout this period, but it did not assume a similarly combative stance in relation to the TUC.

The GMB, a similarly politically aligned although more centrist union, also supported single-union deals on greenfield sites as the best option for both employer and union. The GMB developed its own model agreement with provision for training for employees, a consultative committee, flexibility, and the right to go to arbitration in disputes. In 1989 the GMB claimed to have thirty-six such agreements. In ‘Shaping up for the Next Century’ the GMB argued for a new approach stressing the need to cooperate with employers—a theme fitting its new slogan, ‘Working together’. Moreover, by the mid-1990s the GMB was a leading exponent of ‘industrial partnership’. Its General Secretary, John Edmonds, joined the President of the AEU, Bill Jordan, in signing the Involvement and Participation Association’s brochure ‘Towards Industrial Partnership: A New Approach to Relationships at Work’ (IPA 1992). This stressed the (p.61) advantages of social partnership between management and employee representatives and joint commitment to the success of the enterprise.

The TGWU stood politically more towards the reformist end of unionism. It did not reject single-union deals but consistently very strongly opposed no-strike deals. At the national level, new realism was therefore consciously rejected. As neither of the factions in the TGWU were as closely aligned with the right of the Labour movement as those in the EETPU or AEU, there was little internal political conflict over this policy. However, in such a large and geographically dispersed union, some regional secretaries and local officials did pursue job regulation practices more akin to those adopted by the AEU and GMB.

Neither the CPSA nor the NCU were consistent in their political objectives over this period. In both unions the political differences between the factions were quite marked. This was particularly so in the case of the CPSA, where the moderates and the left both held more extreme positions than their counterparts in other unions. Thus, as the controlling group changed in each union, so did their industrial and political policies and strategies. For example, within the CPSA there was a tendency to adopt a more adversarial stance in the field of job regulation under the Broad Left and to seek negotiated compromises under the moderates. The Broad Left in the CPSA also pressed for the establishment of a political fund and affiliation to the Labour Party—a policy eventually overwhelmingly accepted by the membership in a ballot in 1987 (73 per cent for and 27 per cent against) even though the Treasury’s refusal to deduct such political fund contributions prevented the actual creation of the fund.

Other policy shifts initiated by unions which may be thought to reflect a movement in objectives and methods towards business unionism included the introduction or expansion of services directed towards the individual member rather than the collective membership. Virtually all the unions studied, regardless of their political leanings, moved to strengthen their services to the individual. The package provided by the AEU included assistance with mortgage arrangements, home contents insurance, motor insurance, and life assurance. This was generally a low-cost image-building exercise which made use of the newly computerized membership list to negotiate discounts with a number of financial institutions. In introducing these services, blue-collar unions were following a lead given earlier by white-collar unions, and in particular, unions with a professional or staff association origin. However, there was universal agreement within the unions’ national leadership that these were supplementary to, and not replacements for, collective bargaining.

The strategy and tactics used to implement job regulation policies throws further light on the character of the unions involved and in doing so questions the degree to which the above developments indicate a ‘new (p.62) realism’ and the adoption and practice of business unionism. First, while the combination of single-union and strike-free deals is a relatively new innovation, the practice of unions negotiating favourable terms with employers on greenfield sites is not. For example, the GMWU’s northern regional secretary in the 1960s and 1970s actively pursued closed-shop agreements with firms considering moving into his region, prior to their actually employing people in his region (Undy et al. 1981: 136). In making such approaches, one of the advantages sometimes offered by the GMWU would be the exclusion of the TGWU. It is therefore the strike-free part of the agreement which is new, not the use of such agreements to exclude competitor unions. Second, the number of single-union strike-free deals agreed in the period studied was miniscule compared to the number of conventional agreements in existence. Even the EETPU’s forty such agreements in 1989 only covered a very minor part of that union’s membership. Also, similar deals signed by the AEU and GMB covered only a small fraction of their membership. The agreements were therefore more symbolic than representative of the unions’ general form of collective agreements. However, in signing such agreements the unions exposed themselves to counter-claims from other employers, particularly those in the same industrial sector, for no less favourable agreements. For instance, unions in Rover came under considerable pressure and eventually agreed to sign a ‘new deal’ agreement which was clearly influenced by agreements made with Japanese vehicle builders (IRS 1992: 12–15). It did not, however, include a strike-free clause as such.

Lastly, although unions signing strike-free deals have rarely resorted to militant industrial action11 under such agreements, they have been militant in other circumstances. Unions governed by parties or factions on the right of the labour movement have tended to be more sympathetic towards some aspects of business unionism but have not necessarily chosen to be more moderate than other unions in the field of job regulation when conducting bargaining under more conventional agreements. Indeed, the militant behaviour of the AEU in leading12 the Confederation of Shipbuilding and Engineering Unions’ (CSEU) campaign and selective strikes for the shorter working week in 1989–90 serves as a clear demonstration of a politically moderate union’s willingness to undertake major industrial conflict. The determination of the AEU’s national and moderate ‘Group’ leadership and not least the campaigning of Bill Jordan, who was approaching re-election for the Presidency, was critical for the relatively successful outcome. By September 1990, 1,034 agreements covering approximately 400,000 workers had been signed (providing for a 37-hour week) ‘without accepting the workforce and temporal flexibility clauses which the EEF insisted upon during the 1983–87 negotiations’ (McKinlay and McNulty 1992: 211).

Also, the arch-proponent of new realism in greenfield sites, the EETPU, (p.63) did not reject the use of more militant methods in more traditional plants. In a series of pay claims in electricity supply in the 1980s and 1990s the EETPU, as the leading negotiator, pursued a policy of tough negotiations, joining other unions in calling a series of strike ballots (FT 19.4.90). Moreover, in the Ford pay claim of 1989–90 the EETPU instigated an official and indefinite strike against the introduction of flexible working practices and conditional allowances for skilled workers.13 The strike was called and sustained for three weeks despite all the other unions supporting the proposed agreement, and resulted in 4,000 workers being laid off in the UK and Belgium. During this strike TGWU members crossed official picket lines for the first time in Halewood’s long history of industrial disputes.

It can be concluded, therefore, that even in the case of the craft-conservative and traditionally politically moderate EETPU and AEU there was no movement to abandon reformism and embrace business unionism. The EETPU’s expulsion from the TUC was temporary. It also owed much to the General Secretary’s—Eric Hammond’s—provocative approach to the pursuit of sectional industrial interests at the expense of more left-inclined unions (Lloyd 1990: 598–633). However, after amalgamating with the AEU in 1992 it rejoined the TUC, and it remained in the Labour Party throughout the period. Moreover, it also pursued an aggressive and militant job regulation strategy when it suited the leadership and had the support of members, as at Ford. Similarly, the AEU used single-union strike-free-type agreements to win recognition on greenfield sites. But it also used considerable skill in leading one of the few successful militant job regulation campaigns of the late 1980s and early 1990s. The more left-inclined TGWU, although protesting at strike-free agreements, itself signed single-union deals with Japanese companies, particularly in Wales.

Thus, while unions sought to project a new image thought to be attractive to non-unionized employees and employers alike, this did not automatically lead to a new approach towards conventional job regulation if the circumstances were thought to warrant a more traditional response. Unions adjusted their approaches to suit the particular contexts in which they operated. They generally recognized that greenfield sites required a different approach if they were to be unionized in the climate of the 1980s. New realism in the form of no-strike deals met the needs of such employers and helped the unions gain recognition in a particular segment of their membership territory. Subsequently, by the mid-1990s unions adopted a more co-operative strategy and some moved beyond this to promote industrial partnership. This was prompted, at least in part, by the leadership’s recognition of their more limited power base.

(p.64) Summary and Conclusions

The causes of the non-ballot-related changes were complex and interrelated, but the dramatic fall in membership and density was clearly a key factor driving a number of other changes, particularly when it was associated with a financial deficit. Changes in union government frequently originated in such developments. However, the causes of the fall in membership and density were themselves complex and are not satisfactorily explained by existing research. While quantitative and deductive studies have provided some evidence of the general impact of external factors on changes in union density, they have failed to explain satisfactorily the interaction between these factors and the dynamic process of membership gain and loss, and why it is that unions have not recruited more effectively in both existing territories and the new expanding areas of employment. As a consequence the present schemes of analysis offer little guide to how union behaviour has itself influenced this crucial change.

In circumstances of falling membership and union failure to penetrate new areas of employment, mergers were seen by some leaders as offering a more certain and possibly cheaper way of increasing the major union’s membership and saving the minor unions from extinction. They may, however, have also diverted attention and energy away from a more determined attempt at the retention of existing members and the recruitment of the non-unionized. The mergers that followed were, in the absence of any widely accepted form of external union structure, largely determined by the internal concerns of the merging unions. Moreover, the major unions involved in amalgamations had a particularly important impact on union structure. They not only took large groups of members into new organizations but also caused other large unions to be sensitive to their own relatively diminished standing. In this way the large or mega-mergers of the 1990s look set to become the catalyst for future major mergers in the late 1990s.

Changes in union government, as they affected non-bargaining decisions, probably owed more to mergers and financial difficulties than any other factors. The major merging unions accommodated the incoming minor unions by becoming more vertically bifurcated around trades or industrial sectors. This in turn tended to diminish the influence of intermediate bodies, such as regional or other geographically based systems of decision-making. Also, in making cuts in expenditure and seeking to maximize income there was a tendency to draw decisions back to, and concentrate them at, the centre. Contrary to these developments, employers were primarily responsible for bargaining becoming more decentralized.

The role of the union leader (the inner ring in Figure 1.1) in exercising choice in determining the new forms of government was also important for (p.65) reshaping union government. However, in some unions, such as the centralizing TGWU, the exercise of this discretion became more constrained as factional activity limited the discretion conventionally exercised by the national full-time officials and the General Secretary. Similarly, in the factionalized and volatile CPSA, and to a lesser extent in the NCU, no faction secured a dominant position for more than a few years: in these unions the centre’s decision-making powers did not strengthen. In contrast, the non-factionalized GMB, the one-party EETPU, and the two-party but Group-dominated AEU, experienced relatively consistent leadership. In these unions the process of adjusting internal decision-making in response to the new climate also tended to produce changes which enhanced the national leadership’s powers relative to other bodies within their unions.

The unions’ commitment to some combination of reformist and business unionism remained largely intact throughout the period. Support for the Labour Party was solid. There were, however, some important variations between unions in their pursuit of business goals. In particular, unions on the centre-right of the political spectrum showed themselves more willing to accommodate the interests of employers on greenfield sites. Nevertheless, even the most craft-conservative and politically moderate and stable unions did not radically change their character; if a conventional bargaining approach and militant tactics were required by the circumstances they were used. Unions thus tended to adopt a contingent approach to strategy and tactics, provided it came within the limits of what was ideologically sound, or at least acceptable when the questions arose. However, towards the end of the period, as unions came to realize that the powers enjoyed in the 1970s were unlikely to be restored, unions began to develop a more strategic approach to their relations with employers. Notions of social and industrial partnership were explored and, in some areas, integrated with other changes to produce a new proactive strategy for regenerating the unions.

To conclude: in the unexpectedly hostile environment of the period studied, unions were initially inclined to adopt damage-limitation policies and strategies. Each union tended to defend its own interests and to give priority to institutional survival. Unions did not generally undertake a radical review or create a new movement-wide consensus around the issue of what unions were—or should be—for. Neither did they agree on a commonly desirable union structure. The general movement towards a more centralized system of government and greater vertical bifurcation was the result of a number of unconnected individual actions, although some leaders did use the opportunity these presented to draw decisions and financial control back to the centre. For much of the period, unions and their leaders therefore concentrated on dealing with internal problems as they arose, on the basis of purposeful ad hocery. Given the relatively (p.66) rapid change in their fortunes and the external environment, this may have been the best they could hope for in the short term. Towards the end of the period, however, more considered and overtly strategic approaches were being developed. A number of unions initiated radical reviews of their organizational structure and in some unions this coincided with the development of new policies towards employers, including the promotion of industrial partnership.


(1) . See British Journal of Industrial Relations, vol. 28, no. 2 (1990) for three articles which analyse changes in union density: Freeman and Pelletier; Disney; and Bailey and Kelly. Also see Waddington (1992). In this article Waddington calculates union density by reference to the labour force, excluding employers, the self-employed, members of the armed forces, and the registered unemployed. The result of this calculation is to raise the union density figures as compared to those quoted on p. 35, e.g. on p. 35 we quote figures of 41 per cent for union density in 1987 whereas Waddington produces a figure of 46.3 per cent for the same year.

(2) . The GMWU (forerunner of the GMB) was categorized as ‘restrictive-passive’ in its orientation to membership recruitment by Undy et al. (1981: 64–6) because in comparison with other large general unions it did not pursue membership growth as a priority in the 1960s and 1970s.

(3) . The initial idea concerning the process of recruitment was triggered by Disney’s reference (1990: 1.71) to different factors affecting different stages of union membership, quoting a mimeo by F. Green of the University of Leicester. This mimeo has not been seen by the authors.

(4) . The material used in this section is taken largely from research conducted by Roger Undy and Lord McCarthy for the TUC. The research was primarily conducted between 1989 and 1994 using postal questions and interviews of TUC affiliates, and covered 90 per cent of the TUC’s affiliated unions and 98 per cent of the TUC’s total membership in 1989.

(5) . See Undy et al. (1981: 167–219) for a discussion of these terms.

(6) . See Salamon (1987: 166–7) for a discussion of these terms.

(7) . See Purcell (1991) for a discussion of changes in employers’ bargaining levels in the 1980s.

(8) . In the AEU the ‘Group’ was the politically moderate party or faction which had dominated the government of the AEU since the introduction in 1972 of postal ballots for the election of all full-time officers.

(9) . TGWU (1993), a TGWU strategy paper used to initiate debate on further reforms at the TGWU’s 35th Biennial Delegate Conference in 1993.

(10) . Dabscheck and Niland (1981). Chapter 4 of their book discusses the various classifications of unions according to their behaviour and how they seek to achieve their goals. We did not include reference to socialist revolutionary (p.67) unionism (Darlington (1994)) as it had little relevance to the behaviour of the unions studied.

(11) . LRD, vol. 78, no. 11 (10.89) identified five companies with ‘no-strike’ deals which had also experienced industrial action (p. 9).

(12) . For a brief discussion of the AEU’s role in leading this dispute, see McKinlay and McNulty (1992).

(13) . See the Financial Times on the following dates for reports of this dispute: 30.1.90; 6.2.90; 7.2.90; 14.2.90; 28.2.90; 6.3.90; 7.3.90; and 8.3.90.