11.1 Introduction

The landscape of British industrial relations has changed markedly during the last 25 years from its collectivist highpoint in 1979. Since then, union density and the coverage of collective bargaining have almost halved (Cully et al. 1999; Edwards 2003), and even where pay is still determined by collective bargaining, its scope is typically much narrower (Kelly 1997; Towers 1997). The problem is particularly marked in ‘new’ establishments, where it has proved very difficult for unions to gain recognition and the experience of being a trade union member is increasingly rare (Machin 2000; Waddington 2003). Although joint consultative committees and union–management (or employee–management) partnerships are now a more important part of the agenda (Bacon and Storey 2000; Guest and Peccei 2001; Marchington 2001; Terry 2003), they are less able than collective bargaining to provide workers with an independent voice. Indeed, JCCS are only found in just over one-quarter of establishments (Cully et al. 1999). In short, because the UK system developed through voluntary arrangements at establishment or enterprise level, rather than through the creation of legal supports at national level as in much of mainland Europe, workers' rights are particularly vulnerable to a loss in trade union power.

Most readers will find nothing new in this short résumé of changes in UK industrial relations. However, as we saw in Chapter 3, this analysis is based on an important and implicit assumption about how the subject is defined, one which is constrained by a focus on the relationship between a single employer and its employees. Because most commentators have been focusing on the extent to which employees working for a particular employer have ‘lost their voice’, they have overlooked the fact that many organizations now choose to use subcontracted labour rather than employ people direct. This has been particularly apparent in the public sector where large swathes of work have (p. 240 ) been transferred to the private sector via Compulsory Competitive Tendering (CCT), Best Value, and Private Finance Initiatives (PFIs). Private sector firms have also cut back on work such as catering, cleaning, haulage, and security to focus on what they see as core activities. Some have gone even further by contracting out lower value-added manufacturing operations so as to concentrate on customer service work. In other words, the ‘typical’ employer of today is very different from 25 years ago, and many of what were characterized as standard, internalized jobs are now undertaken by subcontractors, agency workers, or the self-employed. Moreover, there are numerous situations in which workers employed by different organizations, or by agencies, work alongside each other at the same workplace, often employed on quite different terms and conditions. Some are likely to be union members while others are not, some will find that previous alliances with other union members at the same workplace have now fragmented or disintegrated, and yet others will be denied the opportunity to join unions by hostile employers. Here, we are particularly interested in analysing whether and how workers are able to make their voice heard when organizational boundaries are redrawn.

There are few studies that conceive of industrial relations beyond organizational boundaries (Rubery et al. 2003). Heery and Abbott (2000) argue that trade unions must move away from their traditional, relatively narrow agenda of specifically defending the rights of their own members and adopt a wider concern with workers as a whole. They suggest (2000: 167) that, ‘putting it crudely, the trade union movement has to decide whether it will speak primarily for the economic “insiders”, many of whom feel embattled, or whether it will accept a broader representative remit which encompasses the substantial proportion of employees with non-standard working arrangements’. Following on from this, Heery et al. (2002: 13) examine what trade unions have been doing in order to represent four sets of ‘contingent workers’—part-timers, workers on fixed term contracts, agency staff, and freelancers. Based on a postal survey of fifty-six trade unions, they find (2002: 13) there has been rather more progress in relation to part-time workers than any other group, but overall most unions do not invest in dedicated recruitment activities for contingent workers. There have been some attempts to set up specialist branches for agency workers or provide them with a dedicated full-time officer, but this is limited to unions facing sizeable problems in this regard—for example, NUMAST and the CWU. In conclusion, Heery et al. (2001: 29) argue for a ‘form of trade unionism that breaks with the dominant workplace model and seeks to represent workers “beyond the enterprise” ’.

In this chapter we address the concept of worker voice and representation, as well as the forms they take, in the context of inter-organizational relationships. In these situations, we have to rely on alternative and more complex conceptions of employment and organizational relationships that allow for influences from ‘non-employers’ over the nature of work, as well as focus on the (p. 241 ) problems workers face in trying to express their voice beyond organizational boundaries. This requires us to examine the ways in which inter-organizational contracting changes existing mechanisms for voice or makes it difficult for some workers to express their opinions. For example, agency workers might feel excluded from the channels available to other staff or find it difficult to develop close relations with union members working for different employers there. Workers who are employed by the same firm on contracts at one workplace could experience major problems forging links with their colleagues at other workplaces, even if they are in the same union. Yet other workers might have little chance to join unions and find voice only through a managerially dominated consultation system. In short, the blurring of organizational boundaries has major implications for worker voice.

Four cases are examined here, including two different parts of the Customer Service network, and the principal points about each case are summarized in Table 11.1. This is developed from Willman et al. (2003), and it differentiates between (a) employees' propensity to join a union, (b) the union's propensity to organize a workplace, and (c) the employer's propensity to deal with a union. This is useful because it demonstrates clearly how patterns of voice are dependent on the attitudes and behaviours of a range of different actors and the interplay between them. It is often assumed that employers represent the major stumbling block for unions trying to recruit new members but it is also clear that some groups of workers are not especially interested in joining unions either. Rather less common, but nevertheless important, is the situation where unions do not consider it cost-effective aiming to recruit new members due to workplace size, perceived employer hostility, or worker indifference. Useful though this framework is however, it ignores the fact that more than one employer may have an influence on employment relations at a particular workplace, either as a client or as a major employer at the same place of work. It also ignores the role of the union prior to contracting, a factor that is especially important if transfers have taken place between organizations or if permanent and agency staff—perhaps employed on different terms and conditions—work alongside each other and interact at work. Indeed, by treating each actor as if it were a single, unified agent, major simplifications have to be made when analysing voice across organizational boundaries. Accordingly, in Table 11.1 we attempt to incorporate the attitudes of employers and clients as well as between historical union traditions and those of workers employed by different organizations at the same workplace. Moreover, we believe it is crucial to examine patterns of worker representation both within organizations and across organizational boundaries.

Although precise developments in each case differed, depending on the interplay of different forces, the common theme was a reduction in worker voice following changes in contracting arrangements. One of the major consequences had been to displace risk from one organization to another, and in (p. 242 )

(p. 243 ) (p. 244 ) particular to groups of workers who were less able to articulate their views and resist the changes imposed on them. Their lack of power resulted in worse terms and conditions, either in terms of lower pay, less employment security, and limited career development—as we have already seen in earlier chapters. So, for example, the security staff who worked for an agency but employed at the Scotchem site were disenfranchised from any form of meaningful voice because not only did the agency fail to involve them, they were also ignored by the unions at the site where they worked. While the supply teachers generally retained their union membership, they had little opportunity to become involved in union activities at the schools and felt their voice was fractured from traditional channels both at the school and at the agency. Industrial relations at the airport had become fragmented following the requirement to end monopoly provision of services and instead seek these from a series of different contractors. Even though workers remained in the same unions and membership levels were sustained, conflicts arose repeatedly between those working for different firms and on different conditions. Finally, voice in the customer service case was disconnected, partly due to geographical distance and multiple sites but also due to sharply contrasting union traditions and the disparate nature of inter-organizational relations on different contracts. In each case, therefore, interrelationships between different employers, unions, and workers had major consequences for worker voice. In effect, opportunities were limited in contexts where work (and sometimes workers) was transferred to new employers, where organizational boundaries were blurred and work was fragmented. In the following analysis, we consider each of these cases in turn.

11.2 Disenfranchised voice

There were strong traditions of trade unionism at Scotchem (Marchington 2001; Marchington et al. 1992) and ample evidence that management had adopted a de facto partnership approach since the 1990s. The company recognized several unions (GMB, TGWU, and AEEU), membership levels were high, and senior shop stewards had access to management through a range of consultative and bargaining mechanisms at establishment level. For the most part the unions worked well together, although there have been some tensions between craft and process workers. During the research period, a new communications group had been set up comprising the senior managers and senior stewards, and additional training sessions had taken place for this group under the auspices of a local trade union college. Both sides viewed management–union relations at plant level in a generally positive light: ‘I talk with them (the senior stewards) about the manufacturing plan for the next month and what the issues are. It gives them a chance to express one or two of their concerns about the future’ (Scotchem, Director 1, male).

(p. 245 ) In the late 1990s, Scotchem decided to outsource various aspects of its work, including security, to external contractors. Prior to this, security work had been undertaken by manual workers who were finding it increasingly difficult to cope with the physical demands of working on the plant, and they were redeployed to this work in the run-up to retirement. They were retained on the main negotiated rates for chemical workers, and the move to security work was regarded as recognition for long-term service. As part of a worldwide drive to control costs in Multichem, senior management at Scotchem decided it was no longer cost-effective to continue with existing arrangements and people working on the gatehouse were offered relatively generous early retirement packages. The contract was won by Securiforce, which brought in its own staff to run the system. The trade unions went along with this largely because they feared cuts elsewhere, and instead devoted their energies to negotiating severance terms and developing strategies to help make the plant more viable in the future.

The task of providing security work was therefore undertaken by a quite different set of people from those previously employed by Scotchem, and they had no links with other staff on the site. Securiforce did not recognize trade unions, and since most people working for the agency were ex-service personnel they lacked any union traditions. Pay levels were much lower than for other Scotchem workers and for those who previously worked in the gatehouse, largely because most Securiforce workers already received pensions from their previous work. Moreover, they were employed on a short-term basis, rarely if ever met up with other agency staff, and had limited contact with management. As one of the security staff noted:

It's not a job, career-wise, that anybody would consider, but if you wanted to fill a gap for six or seven months you could certainly do worse. (Securiforce, guard 1, male)

In such a situation, it was unlikely trade unionism could flourish, and indeed it did not. The security guards felt isolated from the rest of the workforce, they were not invited to social or communications events with Scotchem workers and they felt there was little interest in them or their needs. Indeed, given the fact that their employment effectively closed off a potential future job for Scotchem workers, they felt resented and marginalized. The site unions did not feel any responsibility to speak for this group either, and the de facto partnership agreement did not include Securiforce workers because they were not covered by the terms of collective agreements. In short, the lack of representative machinery both with the agency and within the Scotchem regime rendered staff in the gatehouse almost totally voiceless. At best, they had limited communications with Scotcheiris Administration Manager. While Scotchem management were positive towards unions for their own employees, this did not extend to those working for other companies on the site and, given the lack of interest at Securiforce and among the agency workers they employed, channels for voice were absent. The contracting arrangement effectively disenfranchised the people employed on security work.

(p. 246 ) 11.3 Fractured voice

The teacher supply case contrasted sharply with this, largely because union membership was relatively high among supply teachers and at similar levels to staff employed on permanent contracts. According to the main teacher unions at least 30,000 supply teachers were on their books (about 75 per cent of those employed in this capacity) and numbers grew sizeably during the last few years. Subscriptions for supply teachers were on reduced rates, recognizing they are not likely to be employed as many days as permanent staff, but benefits were similar and the NUT, the NASUWT, and the ATL all argued this indicated unions were equally committed to both groups. Even though they may spend only a few days at each school, supply teachers were generally union members because many had been employed on permanent contracts in the past and/or saw it as valuable for professional or instrumental reasons. Indeed, among the sample of supply teachers interviewed only one said that they were not in a union, with the remainder split between the three mentioned above. The unions were in an ambivalent position about agencies because while they wished to recruit members involved in this work, they were also concerned that agency work could undermine the terms and conditions of permanent staff. However, none of the unions had negotiated a deal with the agencies, so union membership was both individualized and distanced from their legal employer, the agency. Nevertheless, no distinctions were made between members:

Once they have subscribed and been accepted for membership they are members and we don't differentiate between a supply teacher and a full-time (sic) teacher. We don't say you can only hold branch office, for example, because of this. (ATL, union rep. 2, female)

While there may be high levels of union membership among supply teachers, opportunities to exercise voice were fractured for several reasons. First, under the ‘old’ system Local Education Authorities (LEAs) kept a register of interested teachers who, like permanent staff, were on nationally negotiated pay scales and had similar levels of holiday pay. Now that supply teachers came through agencies, they could not claim parity with their colleagues employed by the schools where seniority was rewarded by additional pay increments. They were paid a flat rate as a Newly Qualified Teacher (NQT) although this rose for teachers with more than five years' experience, and it may also be supplemented on an individual basis at the discretion of the agency (Grimshaw et al. 2003). There was no collective route to the agencies to determine pay, thus making it difficult to improve terms and conditions for the whole group.

Second, they were marginal both to the schools and the agencies. In the case of the former, teachers were typically employed on short-term contracts that made it hard to establish close and trusting relations with other union members at the schools, and there was less incentive to stay behind to attend union meetings because of the temporary nature of their work. Union activism was problematic given the lack of a base in a particular school and the fact that there were a number of competing unions in the sector. The dominant union (p. 247 ) tended to vary depending upon region and whether the school was primary or secondary, and even if an agency supply teacher was a member of a large union there was no guarantee it would have sizeable representation in the schools where they worked. There were also problems in getting their voice heard at the agency because, although there were regular contacts to offer work, and in some cases check how things were progressing at a school, there was no formal mechanism for supply teachers to meet up and discuss issues related to work. This meant their voice was unlikely to be heard unless things went wrong. Raising a grievance was not easy either. Four of the supply teachers that were interviewed (out of twenty-four) claimed to have been assaulted by pupils—because they were regarded as ‘easy game’—but only one felt confident to challenge the school and they gained support from the agency rather than the union. Most felt powerless if there were grievances because the school did not owe them anything and ‘making waves’ could jeopardize chances of a permanent post at the school (Grimshaw et al. 2003: 276). Another respondent (TeacherTemp, supply teacher 11, male) said he would be reticent to take up a grievance either with the Head or the agency for fear of being blacklisted by the school or by the agency. Yet, another saw no point in being in a union:

All the union would have done would be to cause problems between myself and the school and indirectly that would have caused problems between the agency and myself. It wouldn't have done my employment opportunities much good. (TeacherTemp, supply teacher 13, male)

Third, there were often tensions between permanent staff and agency teachers that made it difficult to establish a common stance. Quite a few permanent staff that were interviewed expressed resentment about supply teachers, arguing they were not committed to the school, the pupils or the teaching profession. For example, two said:

Most supply teachers come in with their newspaper, their coats, umbrella and expect every lesson to be sorted out for them…if they arrive at a classroom and there is no work set they won't do anything. (School B, senior teacher 3, male)

No, they don't come to parents' evenings, they don't write reports, they don't do break duties…The head of department would prepare their work and say to the supply teacher ‘this is what you are teaching in this lesson, go in there and deliver it.’ (School A, ATL union rep. 1, male)

In these circumstances, it was highly unlikely that union representatives in the schools, who were usually permanent teachers, would have much incentive to progress supply teachers' grievances or would be able to generate support from the rest of the membership. Even if they did, by the time the issue came out in the formal arena, in most cases the supply teacher would have moved on. In short, therefore, despite this group of agency workers having high levels of union membership, there were a number of major obstacles—multi-unionism, lack of bargaining opportunities with the agency, and tensions in the schools—that caused their voice to be fractured.

(p. 248 ) 11.4 Fragmented voice

At first sight, the airport case might appear quite simple because all activity took place on one site and the majority of the workers, although employed by different firms, were in the same union (the TGWU). However, the historical context for this case is critically important to understand how the processes of contracting and recontracting led to divisions between workers and the fragmentation of voice. Airportco had high levels of union membership, a long history of management–union relations based upon local authority agreements and the routine involvement of councillors. Industrial relations had traditionally been characterized by conflict, especially in those areas where workers were prepared to use their power, as in the fire station or among baggage handlers at times of high throughput, and militant union traditions had been bolstered by the adoption of a no redundancy policy by the airport, in line with that of the local authorities. This tradition developed when airport operations were relatively integrated, but in the early 1990s application of competition legislation required various airport activities to be put out to tender. This enforced fragmentation was also used by management at the airport to implement major changes to the employment relationship, by limiting coverage of the ‘no redundancy’ policy and by using the threat of competitive tendering to force the union to accept lower pay rates for new hires.

The restructuring of industrial relations involved several interrelated developments. First, to meet the competition requirements, Airportco set up a wholly owned baggage handling subsidiary (BH) with an independent management system and lower rates of pay so it could tender successfully for work against the existing handling company (FH). Existing staff were redeployed under the no compulsory redundancy policy to other jobs within Airportco but employment security for new staff was explicitly tied to securing contracts from the airlines. This brought some operations at Airportco into direct competition with the private sector. The same union had traditionally represented both airport and FH staff¹ but the creation of BH led to direct competition between the operations; now the same union represented workers from Airportco as well as BH and FH staff. Second, management at Airportco used its new-found confidence associated with the establishment of BH on lower terms and conditions to apply the same logic to its internal operations and pushed for changes to existing employment arrangements. The threat of future competition was used to introduce the so-called market rates for all areas, thereby creating a two-tier system of wages depending upon date of recruitment and not skill level. Furthermore, failure to secure change at the fire station led to the establishment of an independent subsidiary that hired new staff for a second shift on worse terms and conditions. In this case, therefore, the fragmentation of the industrial relations and employment system was brought about both by the involvement of new private sector organizations and by the creation of subsidiaries owned by the airport itself. Moreover, (p. 249 ) because Airportco retained its no compulsory redundancy policy under the influence of the local authority, a third tier to the workforce was introduced comprising temporary and seasonal workers, often provided by agencies, who were on part-time contracts to meet peak demands.

Given this much more fragmented and divided employment system, it was perhaps surprising that union membership remained high among staff at Airportco, in BH and in FH, all of which recognized the same union. In principle high union density might be expected to enhance the likelihood of effective site-wide representation so as to avoid damaging internecine conflicts. However, for a number of reasons, this did not happen and worker voice and representation was fragmented across organizations at this multi-employer site. While some of this can be attributed to a failure of union imagination and perspective, to a large extent the problems reflected the legacy of previous industrial relations and the rather cumbersome and divisive way in which this system was changed. Instead of producing a strong solidaristic system of trade union organization—as it did in the past—the current arrangements tended to please no one. Long-serving employees either resented the loss of union power or blamed the union for its continuing and unhelpful militant stance. New recruits blamed the union for failing to prevent them being employed on inferior market rates, while the third tier of temporary, seasonal workers remained excluded from representation. Blame appeared to be allocated evenly between management who were felt to use unreasonable threats to force through market rates and the union that was perceived as lacking commitment to new staff:

‘They (the management) keep threatening us, “you will do this or we will lose the contract.” ’ (Airportco, union rep. 1, female). ‘The union only represents the ex-airport personnel and older people, and not the new people’. (Airportco, supervisor 1, male)

This was particularly evident among security staff where the union agreed to the hire of new staff at market rates that involved a pay rate 40 per cent lower than for existing staff. However, the union continued to oppose the deployment of market-rate staff on gate duty, insisting that their pay should be increased to the old rates. New staff felt this action was blocking their access to progression within the company. A security officer hired on market rates named one of her main grievances against the union:

Not giving any information to ourselves about what exactly market related pay involves and why we can't progress. (Airportco, security officer 1, female)

Many permanent staff, both at BH and airport security, were not convinced the union was taking a sufficiently proactive stance to protect the interests of temporary workers, focusing instead on enforcing job demarcations or getting time off for union duties rather than the bigger questions of market rates and the share of permanent jobs. This led to disillusionment among staff:

I can't be really bothered about the union. They seem to fight over petty little things like rosters…they're not fighting for permanent jobs. (BH, baggage handler 1, male)

(p. 250 ) It was clear the union had not made a very successful transition from representing a largely integrated workforce to adopt new approaches that protected members from, rather than exposed them to, being used as part of a process of inter-organizational competition that had the effect—in part by design—of weakening worker voice. Not only did the union fail to find an effective way of regulating and moving towards harmonization across the three-tier workforce, but it also adopted a narrow, sectionalist stance to inter-organizational competition. It was clear from interviews at BH and FH that union representatives in both organizations encouraged management to take a more aggressive stance vis-à-vis the other handling company even though the baggage handlers in both organizations were represented by the same union and were competing for the same volume of work. According to the BH General Manager, the union was leading a general staff reluctance to cooperate with FH staff when dealing with an airline contract that BH had just lost. FH management also appeared willing to take advantage of the hostility between representatives in the two firms as the union had

distanced itself from the grievances of the part-time workforce because (it) supported FH management's plans to expand its business into Terminal X—which had been the traditional sphere of BH. (FH, Manager 3, male)

However, while the union can be criticized for its failure to adapt to new conditions, the opportunity to develop a more progressive response was undoubtedly hampered by shifts in management policies and practices—both at Airportco and elsewhere. The approach at the airport had changed a number of times over the past decade; several managers and workers referred to a policy in the early 1990s called variously ‘excellence through partnership’ or ‘quality through partnership’. There was general agreement that this experiment had been abandoned, but for different reasons; for some managers it had just led to a loss of control and a further deterioration of industrial relations. For union representatives, the abandonment of partnership signalled the change to a hard-line stance by managers at Airportco, associated with the drive for market rates and the move away from regular fortnightly joint meetings with the union to quarterly meetings. Not all managers were convinced of the wisdom of this approach, feeling the new market rate system was divisive and unfair, and may even have pushed rates too low and had an adverse effect:

You have people virtually doing the same job…on a substantial amount of money less than somebody else…You then get somebody who's on market-rated pay doing a bloody good job and they're standing next to somebody who's not doing quite as well as they are, and they're being paid a lot more money. (Airportco, Manager 3, male)

At BH there was evidence of a closer partnership between management and the union, a policy that had worked well according to top management and involved regular formal meetings between management and the shop stewards, up to twice monthly, along with frequent informal contact. The unions (p. 251 ) were represented by a full-time convenor and eleven other shop stewards. The Managing Director spoke warmly about the quality of the relationship:

We do have a close working relationship with the trade union and we are proud of that. Without that relationship we would not have been able to achieve some of the successes we have achieved in the five or six years since I've been with the organization. We've not had a major industrial dispute, we've not had any major operational disruptions and we've been able to negotiate some quite significant restructuring proposals…We have introduced market pay. We have, because of the commercial realities, seen staff wages fall quite substantially. (BH, Managing Director, male)

However, middle management appeared less impressed and felt the union was restricting initiatives. One suggested that

although there are lots of meetings going on, the union is not necessarily co-operative…initiatives are not implemented because the union does not like them. (BH, Manager 3, male)

These conflicting accounts of industrial relations at BH depend on the point of comparison; relations were more cooperative and harmonious than at Airportco but the union at BH had considerably more operational control than at other organizations which had never been part of the main airport—for example, FH and CleanCo. Engagement with the union at these firms was more arm's length and there were fewer positive and fewer negative comments on the industrial relations climate, as well as more diverse views about what the union should do to represent its members more effectively.

Thus, despite having high membership density within the same union, workers lost immediate and ongoing contact with colleagues in other organizations. As we have seen, tensions arose between workers at BH and FH due to battles for a set amount of work from the airlines. One of the terminal supervisors at BH noted:

The unions are very disintegrated. Each department and group looks after its own interests no matter what happens to other departments or groups. If the Indians fight amongst themselves, the Chiefs can sit back and laugh. (BH, supervisor 1, male)

These inter-organizational divisions reinforced those between the three tiers of the workforce, leading to a fragmented and divided system of worker representation. A telling example of this was the failure of union representatives to keep each other informed about developments. The two female shop stewards representing the new market-rate staff in security felt excluded from the inner circle of the union:

There is a lot going on, on both sides (TU and management), but some TU reps who have been to the meetings would not tell other reps what was going on. We (TU reps) all go separate ways. The other lady shop steward and I have not been kept informed of what has been going on. (Airportco, union rep. 1, female)

The airport case illustrates clearly the problems faced by unions in a system where groups of workers, albeit in the same union, were placed in direct (p. 252 ) competition with each other after organizational restructuring and the introduction of new management initiatives. Not only were there potential tensions between workers employed by different firms as they fought for work, there were also divisions and contradictions between different types of worker (permanent versus temporary; men versus women; established and market rates) employed by the same company. In such circumstances, there was little chance that a coherent and unified voice could develop across the airport, even for members of the same union. Instead, because the system relied on cost-cutting and financial gains through contracting and recontracting, voice became fragmented.

11.5 Disconnected voice

11.6 Discussion and conclusions

There are two sets of points to make in the conclusions to this chapter. First, though the Willman et al. categorization is useful in directing analysis to how the propensities of different sets of actors (employers, unions, and employees) shape industrial relations outcomes, it inevitably simplifies the situation. It is rare for managers, unions, or employees to be unified in their attitudes and behaviours at any single workplace, so once we go beyond the boundaries of the organization, processes are much more complex and multifaceted. For example, the influence of more than one employer was apparent at every case we examined here, ranging from the relatively straightforward situation in the chemicals case, where just two employers played a part in marginalizing voice for security workers, through to the airport case where a range of historical traditions and changes in organizational forms had splintered employment across a wide range of firms. Similarly, it was hard to conceive of a cohesive union position in any of the cases, most obviously where several different unions were recognized and worker voice was fragmented rather than focused on a common cause. In multi-employer workplaces however, it was even harder for unions to build coalitions between groups of workers who were engaged on different parts of the contract; for example, in the customer service case, (p. 259 ) despite being members of the same union and in the same branch, hostilities between housing benefit workers employed by the Council and TCS made it very difficult to establish and sustain a shared identity Moreover, there was no evidence that union representatives had combined together across TCS, and in some cases (e.g. the call centre) UNISON appeared to accept it was unlikely to gain recognition in the short term—though other unions have elsewhere (Bain and Taylor 2001). To some extent this was due to the fact that the workers themselves showed little interest in joining unions, and for agency workers in particular there was little knowledge of what unions did. However, this was not universal and the agency workers in teaching and housing benefits administration had a much greater propensity to join unions, either for professional or instrumental reasons. Accordingly, just like employers and unions, it is unlikely that workers attitudes can be simply classified as positive or negative; much depends on their occupational allegiances and their employment status (temporary or permanent, first or second tier in the workforce).

Second, while worker voice was marginalized in each case, the precise form that this took varied depending on circumstances. The security workers were disenfranchised because neither their own employer nor the unions showed much interest in helping them articulate their views, even though Scotchem workers at the same site were able to take advantage of a wide range of EI practices. The outcome for the supply teachers was different in that they tended to have reasonably high levels of union membership, but their voice was fractured across organizational boundaries. Previously, supply teachers tended to come from a local authority pool, where they were both subject to similar sets of conditions to other teachers employed by the same LEA and covered by the same bargaining arrangements. Now that supply teachers were provided by agencies and they could work in schools across a number of LEAs, they found it much harder to become involved in union activities at the schools. Moreover, their employer (Teacher Supply) did not recognize unions and there was a feeling that future work prospects would be harmed if they raised grievances either with the agency or with Heads in the schools. The whole airport system had been fragmented following anti-monopoly legislation such that workers, even though they might still be in the same union, were divided in terms of what they saw as their best interests, and there were examples of shop stewards at one company urging their managers not to cooperate with other companies even though their fellow members' jobs might be at risk because of this. Furthermore, there were tensions across different groups of workers, in particular between those on historically determined (higher) rates of pay and those on lower, market-rated pay. In short, the reorganization of business systems had helped to fragment worker solidarity. Given the wide range of sites and types of business contracts involved in the customer service case, worker voice was disconnected both within TCS as well as across organizational boundaries. This was apparent in the divisions between union members working for Council X and for TCS in the housing benefits case, and between TCS and agency workers at the call centre.

(p. 260 ) Each of these cases illustrates well the problems workers face when trying to establish or maintain voice in the context of inter-organizational contracting. The transfer of risk from one employer to another, and then on to the individuals who provide the service, is starkly apparent when temporary contracts are not extended or the demand for labour declines—as is now the case with supply teachers following reduction in school budgets. Other cases are less dramatic but nevertheless demonstrate the consequences of inter-organizational contracting for workers' wages and conditions, as well as their career development, when they no longer have a strong countervailing source of power at work through trade unions. In short, while workers employed by some organizations in the network might have clearly defined channels for voice in order to articulate their grievances and protect their conditions, those employed on precarious contracts across organizational boundaries lack the collective strength to make their voice heard.