Corps de l’article

Just as the best settlement of a dispute is a voluntary settlement, so the best way for employees and employers to settle differences of their working together is by an agreement voluntarily arrived at.

Industrial Canada (1943: 124)

On October 17, 2007, Buzz Hargrove, President of the Canadian Auto Workers (CAW), and Frank Stronach, Chairman of Magna International, stunned labour and business groups when they announced a comprehensive labour relations agreement called the “Framework of Fairness Agreement” (FFA). Long-time adversaries, Hargrove and Stronach claimed that there were “a surprising number of issues on which we now see eye to eye” (Stronach and Hargrove, 2007: A21). Observers labeled the deal historic given the potential implications it could have for union-management relations in Canada (Armstrong, 2007). The primary feature of the FFA is a Magna-sponsored process that allows 18,000 non-unionized Magna workers in Canada to decide whether they want to be represented by the CAW. In cases where a majority of a plant’s workers support unionization by a secret ballot vote, Magna will voluntarily recognize the CAW as exclusive bargaining agent at that plant.[1]

This paper will show how elements of the FFA are a manifestation of the historically significant, and often overlooked, principle of industrial voluntarism—the principle that the contours of union-employer relations ought to be a product of voluntary agreement between the principal parties (i.e., union and employer) without state intervention. We trace the historical development of industrial voluntarism in Canada between approximately 1900 and today, using the FFA as a contemporary example.

Prior to the adoption of legislation to establish labour relations boards which regulated union organizing, collective bargaining, and dispute resolution, industrial voluntarism flourished in Canada (Fudge and Tucker, 2001). Voluntarism’s hallmark was voluntary contracting between unions and employers. Consistent with this practice, the state adopted a non-interventionist stance; furthermore, in this environment, employers often successfully resisted union demands for formal recognition. However, the balance of power shifted dramatically with the passage of federal and provincial labour laws in the mid-1940’s; henceforth, governments assumed an active role in industrial relations through labour relations boards which could compel employers to recognize and collectively bargain with unions supported by the majority of workers. Under this new regime, it seemed that voluntarism no longer served a purpose. Indeed, at times some scholars even claimed that it was defunct (e.g., Fink, 1973). However, evidence indicates that voluntarism waned but never vanished, and is still relevant. At present, the principle of voluntarism informs neutrality agreements, voluntary union recognition, and other non-statutory approaches.

The paper is organized into four parts: Part one defines industrial voluntarism and examines its practical significance in the context of union organizing between 1900 and 1944. Part two shows how voluntarism was accommodated in post-war labour laws. Part three examines various employer and union actions to contract outside of Canadian labour law and how the logic of voluntarism informs such agreements. Finally, part four uses the recent Magna-CAW agreement as a case study to examine similarities and differences between historical and contemporary manifestations of industrial voluntarism. Deals like the FFA indicate the continued relevance of voluntarism as an organizing logic.

Industrial Voluntarism Defined

The term “industrial voluntarism” has been used to describe the norm that dominated union organizing and, more broadly, union-management relations in Canada during most of the first half of the 20th century (Fudge and Tucker, 2001). In practical terms, the principle of industrial voluntarism defines situations in which unions and employers initiate, develop, and enforce agreements without state assistance or compulsion (Flanders, 1970; Hawkins, 1971).

Industrial voluntarism was first used to describe industrial relations in early twentieth century Britain (Flanders, 1974; White, 1978). In the early American experience and tradition of American Federation of Labor leader Samuel Gompers, the term has been described this way:

“Voluntarism” is […] the philosophy said to characterize American unionism prior to the New Deal era under which the labor movement committed itself to work within the laissez-faire capitalist economy, relying on its economic power to protect and promote the interests of workers and rejecting government aid and intervention (Fink, 1973: 805).

Unfortunately contemporary usage of industrial voluntarism in scholarship is not always consistent (Thompson, 1986; England, 1986). For instance, England (1986) noted that the following definition of voluntarism has currency in some circles even though it seems at direct odds with traditional meanings:

[Laws] which protect union organization, enforce collective agreements, control timeliness of economic sanctions, and establish the procedures under which bargaining takes place have traditionally been regarded as consistent with voluntarism because, it is believed, their thrust is procedural rather than substantive (264).

However, England (1986) rejected that this defines voluntarism because, as he argued, it is impossible to distinguish between “permissible procedural intervention,” which is assumed in the quote above and assumed to enable voluntarism, and “impermissible substantive intervention,” which undercuts voluntarism (270). More recently, Slinn (2008) noted that the “relatively interventionist reality of our labour relations system and the widely-held understanding that it is, at least apart from procedure, a voluntary system” (725) makes the concept difficult to define. Others have treated state-sponsored industrial relations systems and the norm of voluntarism as competing paradigms which gain advantage under particular circumstances (Brudney, 2005).

In this paper, we do not define voluntarism as a binary concept—that is, that it is either strictly present or strictly absent. Rather, based on the evidence marshaled here, we propose that industrial voluntarism has varied in potency over time. Further, we assume that aspects of the contemporary industrial relations systems in Canada reflect voluntarism more so than others. For example, collective bargaining negotiations seldom require substantive state intervention (e.g., first contact arbitration) and are more or less conducted on a voluntary basis. In other domains however voluntarism seems especially weak, particularly in relation to union organizing (Slinn, 2008) where labour board imposed remedies for employers’ unfair labour practices are relatively more common than interventions in collective bargaining. Our analysis is primarily concerned with voluntarism in the context of union recognition from 1900 to present.

Industrial Voluntarism (1900-1944)

Prospects for Canadian workers wishing to unionize during the early 20th century were generally bleak due to strong employer resistance to organizing and collective bargaining. For their part, many employers refused to recognize a union formally unless workers demonstrated their collective strength by striking for recognition; even in these cases, recognition was not a guaranteed result. Furthermore, in the early 20th century, governments were unwilling to compel employers to recognize unions with majority support. Politicians would let power relations between unions and employers determine the outcome of organizing campaigns, and here, the preponderance of management’s power militated against union organization. Self-regulated industrial relations were supported by judicial interpretations of common law which severely constrained union organizing activity (Fudge and Tucker, 2001). Other factors that negatively affected union activities included anti-union employer tactics: blacklisting suspected union supporters, dismissing employees, using detectives to spy on union members, and employing strike-breakers (Manley, 1986). All of these activities were ironic features of early twentieth-century industrial voluntarism, which supposed that employers and workers alike were free to choose.

Voluntarism was an ideal consistent with liberal democracies (Glasbeek, 1986). However, while historically voluntarism meant that unions were “free” to contract wages and employment matters on behalf of their members, union recognition, a necessary precursor to collective bargaining, was left to the discretion of employers. In 1903, the President of the influential Canadian Manufacturing Association (CMA) succinctly and forcefully articulated how its members interpreted the central role of voluntarism amid growing union organizing activity: “The employer of Canada must [never forfeit the freedom] to purchase without interference such labor as he requires” (quoted in Craven, 1980: 125).

Notwithstanding turn of the twentieth-century statutes, such as the Conciliation Act of 1900 and Industrial Disputes Investigation Act (IDIA) of 1907, which set out procedures for state conciliation of strikes and lockouts, as well as periodic statements of industrial relations’ principles and goals, most politicians took a laissez-faire attitude towards matters of industrial relations.[2] These circumstances meant industrial voluntarism flourished. Boards of investigation appointed under the IDIA and legislation that mandated state-sponsored conciliation before a strike or lockout were limited to issuing non-binding reports to disputants. At most, unions were afforded temporary recognition during a work stoppage (under the IDIA). In response to industrial unrest that threatened the public interest, Mackenzie King, author of the IDIA, took an active role in personally mediating strikes. Consistent with the spirit of voluntarism, he believed that state-appointed conciliators should act as ‘impartial umpires’ during disputes, favourable to neither unions nor employers. In practice, however, King himself showed bias towards employer interests in some situations, and compromise was his preferred approach to ending strikes (Fudge and Tucker, 2001; Craven, 1980). In this legal vacuum, many employers were able to resist demands for union recognition even when faced with strong worker support for collective bargaining.

Craven (1980) noted that King may have believed “that unions which deserved recognition would get it because they would convince management of their reasonableness and indispensability [to] a good working relationship” (235). Hence, voluntarism was most likely to succeed where unions were ‘reasonable’ in their demands and respected the law, and in situations which involved enlightened employers who were willing to negotiate.

There is evidence that some employers accepted union recognition, and that stable collective bargaining relations developed. This was especially true for railway workers, longshoremen, and other craft workers who occupied strategic or “respectable” positions in industry, such as, typographical workers and locomotive engineers (Fudge and Tucker, 2001). Railway companies, for example, sought stability in their operations and thus tried to establish solid collective bargaining relationships at the outset with independent (non-company dominated) unions. More generally, employers were more inclined to grant union recognition when the union and its leadership were conservative, responsible, free of “foreign agitators,” and averse to strike activity (Craven, 1980).[3]

Sam Herbst, organizer for the International Ladies Garment Workers Union (ILGWU), was one such union leader who embraced the norms of industrial voluntarism to achieve gains for members. During the late 1930s, Herbst negotiated union recognition contracts with all of Winnipeg’s garment manufacturers by promising (and delivering) an end to intense price cutting and industrial instability. Herbst’s top-down autocratic brand of organizing, which riled some of his fellow union leaders, involved first signing labour agreements with manufacturers and, later, having workers join the ILGWU (Smith, 1985; Gray, 1966).

For the most part, voluntarism continued to be the dominant form of labour relations in Canada until the mid-1940s. During the Second World War, the federal government assumed jurisdiction for labour relations in most industrial sectors. Strike activity soared, with employers’ refusals to recognize unions a leading cause of non-wage related strikes.[4] Voluntarism was so deeply entrenched in Canada that even when Prime Minister King faced intense political pressure to change course in the early 1940s, his government responded only by issuing non-binding labour relations principles—and even to these irresolute guidelines his ministers would not adhere (Fudge and Tucker, 2001).

Industrial Voluntarism (Post-War Era)

With strike activity at historic record levels in 1944, King gave into pressure for US Wagner-like legislation and introduced wartime order-in-council PC 1003. The preamble of this historically significant piece of Canadian legislation declared unequivocally that voluntarism alone was no longer sufficient to govern labour relations. It stated that “both employers and employees should be free to organize for the conduct of negotiations between them and that a procedure should be established for such negotiations” (emphasis added; Labour Gazette, 1944: 136-137). The “procedure” mentioned in the legislation not only guaranteed workers the right to organize collectively, but also compelled employers to recognize bona fide unions, offered workers state protection against employers’ unfair labour practices, and made collective bargaining compulsory (McInnis, 2002). Although the federal government was not the first Canadian government to adopt labour law reforms (British Columbia, Ontario, and Saskatchewan passed similar laws in 1943), its action had the most immediate impact on labour organizing activity because federal control of industry during the war was extensive. Pursuant to the passage of the legislation, recognition strikes dropped to two percent of all strikes between 1944 and 1946 (Labour Gazette, 1945 and 1947).

Interestingly, voluntarism was accommodated in wartime labour laws despite explicit language to the contrary in these laws’ preambles. In effect, the new federal and provincial laws offered dual procedures for union recognition. An important provision of PC 1003 stated that upon an application from a union, “the [Labour] Board shall by an examination of records, by a vote or otherwise, satisfy itself that an election or appointment of bargaining representatives was regularly and properly made” (emphasis added; Labour Gazette, 1944: 8). This clause effectively accommodated voluntary union recognition schemes without a formal representation vote monitored by a labour board. This situation was similar to that in the US, where, since 1935, elections supervised by the National Labour Relation Board have not been a prerequisite for union recognition (Brudney, 2005). Fudge and Tucker (2001) conclude that while passage of new labour legislation “marked a rupture from the individualism of the common law and the absolutism of property rights,” […] “instead of replacing the regimes of liberal and industrial voluntarism, industrial pluralism was grafted onto them” (302 and 306).

Labour laws in Ontario and some other provinces were circumspect on the contracting-out of union recognition to industrial relations parties. However, legislation in other provinces was not so chary of voluntarism; for instance, voluntarism was explicitly condoned in Quebec’s Labour Relations Act (1944). The Quebec law allowed “unrecognized associations” to enter into collective agreements; however, “an agreement so entered into shall become void the day another association is recognized by the Board” (Labour Gazette, 1944: 38). Thus, in Quebec, while voluntarism was accommodated, contracting outside of the law was accorded a lower status than state-sanctioned processes.

Two factors explain why voluntarism was accommodated in wartime labour laws. First, the allowance of voluntarism likely appeased employer concerns about statutory coercion at little cost. Not surprisingly, during the war, employer groups, such as the influential Canadian Manufacturing Association, declared that “compulsion [in industrial relations is] neither desirable nor necessary” (National War Labour Board (NWLB), 1943: 160). In terms of collective bargaining, one employer reasoned that “the word “compulsory” and the word “bargain” cannot be joined together since the word “bargain” means the right of each to bargain with the other freely and without compulsion. The moment compulsion steps in, bargaining ceases to exist” (NWLB, 1943: 1185). Thus, offering dual methods for union recognition—one explicit and one implicit—may have been a measure of compromise for employers.

Second, it is possible that Canadian law-makers sought to encourage the development of a more orderly and balanced form of industrial voluntarism that established clear procedural boundaries which in the long term would ultimately necessitate less government intervention (Doorey, 2007; Fudge and Tucker, 2001; Glasbeek, 1986). Whatever the reason why voluntarism was accommodated in post-war Canadian labour law, the vast majority of unions, having been treated unfairly by employers and with indifference by governments, abandoned the principle for institutional measures of protection. Thus, union commitment to state-sponsored recognition systems—both the card check and mandatory votes—remained strong in Canada in subsequent decades.[5]

The arrival of meaningful state participation in industrial relations led many in the field to assume that the norm of voluntarism became extinct. In support of this view, Fudge and Tucker noted the following: “By 1950, the ideology and practice of industrial pluralism [i.e., participation by unions, employers, and the state] had become hegemonic; everything that had come before it was treated as pre-history, hardly worthy of attention” (2001: 302). Writing in the mid-1980s, a distinguished group of Canadian industrial relations scholars wrote a series of articles related to the question of the “death of voluntarism” in industrial relations (England, 1986). Indeed, in terms of union recognition, post-war labour laws provided capital and labour with the institutional mechanisms adequate to support the formation of unions. However, as Doorey has noted, the “scheme of voluntary [union] recognition […] remained in place, waiting patiently in the wings to be exploited by the industrial relations actors” (2005: 22).

The Resilience of Voluntarism: Voluntary Recognition, Neutrality Agreements, and other Non-Statutory Procedures

Voluntarism has manifested in several ways in the post-war period. Here we describe neutrality agreements, voluntary recognition schemes, and other non-statutory developments. We observed differences not only in the nature and scope of these contracts, but also in the number of actors involved. Generally, industrial voluntarism is embodied in relations between one union and one employer. However, it has also been demonstrated in labour-management relations at the sector level, and in rare cases, at a provincial level as it did in British Columbia in the late-1980s. We turn to this historically significant example first.

In 1987, the Premier of BC, Bill Vander Zalm, adopted sweeping reforms to the province’s labour law. Known as Bill 19, the Industrial Relations Reform Act renamed the Labour Relations Board the Industrial Relations Council (IRC) and granted it considerable power to intervene in labour relations. For instance, the IRC could designate private sector workers as essential in the event of a strike, allow employers to openly engage in electioneering during organizing campaigns, restrict successorship rights, and limit picketing, to name a few features of the controversial legislation (see Leslie, 1991; Goldberg-Hiller, 1996). The BC Federation of Labour (BCFed) and other unions organized a boycott of the IRC. The central challenge of the boycott was to develop a long term strategy that demonstrated that “legality and order [in industrial relations] could remain consonant in the context of illegality and resistance” (Goldberg-Hiller, 1996: 330).

To meet this challenge, BCFed affiliated and some non-affiliated unions devised a parallel system that guaranteed continuation of stable industrial relations. Although the boycott was never total while it was in force between 1987 and 1992 (e.g., the IRC was used by some unions for certification applications and for certain exemptions), voluntarism played a central role in the functioning of the counter administration. Instead of using the IRC, affiliated unions relied on independent mediators, arbitrators, and lawyers to monitor certification and strike votes, and to assist in resolving disputes with employers. Further, private negotiations between employers and unions occurred more often under this regime.

An “exemption committee” comprised of union officers affiliated with the BCFed was established to hear special requests to use the IRC’s administrative services. As Goldberg-Hiller (1996) noted: “For the boycott to remain strong, the boycott code, produced within the exemption committee, had to effectively regulate industrial relations, usurping the role of the competing labour code and labour relations board” (342-343). He identified several cases for which an exemption committee decision to allow access to the IRC was successfully used as leverage to resolve disputes with employers before actually resorting to the IRC. For example, this order from a committee’s decision regarding a bargaining unit dispute illustrates the role of voluntarism in the effective functioning of the temporary industrial relations system.

Ask your counsel to attempt to get the employer to agree to an independent arbitrator to hear the case and render a binding decision based on previous Labour Relations Board jurisprudence. If this measure fails, the panel has granted an exemption to appear before the IRC pursuant to this matter” (Boycott Exemption Files, BCFed Offices, Burnaby, BC, cited in Goldbert-Hiller, 1996: 346, fn. 89).

Goldberg-Hiller (1996) notes that strike activity remained low during the boycott, which coupled with the overall approach to industrial relations “effectively legitimated the boycott in the eyes of business” (347). While voluntarism proved to be an effective stopgap measure during the boycott, some local unions bore significant financial costs associated with using the rival system. Delays for arbitrators were commonplace and fees for arbitrators, mediators and lawyers were a burden, especially for smaller unions. The voluntary system of industrial relations was sustained until 1992 when a more balanced labour code was introduced by Premier Mike Hartcourt.

In the vast majority of cases in which non-statutory voluntarism has emerged in the post-war period, it has been far less elaborate and far-reaching than the experience in BC. The most common manifestation is voluntary recognition agreements between an employer and union. In their simplest form these agreements describe the bargaining unit and the fact that an employer agrees to recognize a union without any preconditions. Historically, this has been used in the construction industry when non-union contractors wish to work on a project (Carrothers, 1968). In the retail food industry, Loblaws, owner of Superstore and other grocery stores, has voluntarily recognized workers prior to opening new stores (e.g., ALRB, 1992). In these scenarios, evidence of majority support for a union is required, and often the card check method or private vote is used.

More recently voluntarism in Canada has manifested in neutrality agreements, which Doorey defines as “privately bargained contracts between unions and employers that define a set of rules by which union organizing campaigns and voluntary recognition will take place” (2007: 42-43). To form these agreements, a union or employer approaches the other party to negotiate terms; or, alternatively, an established union may instigate neutrality agreement negotiations for a non-unionized subsidiary as part of wider collective bargaining with the unionized parent company. For instance, during collective bargaining with DaimlerChrysler in 2000, the Canadian Auto Workers Union (CAW) negotiated a neutrality agreement for organizing Freightliner, a non-union Daimler subsidiary. Such agreements also require employers to remain neutral during organizing campaigns (Brudney, 2005; Eaton and Krieksy, 2001). Most salient is that the state usually has no direct role in regulating the union certification process.

An example involves the United Steel Workers which recently negotiated a wide-ranging agreement with ArcelorMittal, the parent company of long-time non-union steelmaker Dofasco. The deal negotiated in 2007 outlined a process whereby Steelworker representatives would be given access to the Dofasco shop floor to gauge worker support for unionization. If employee support was present, the next step involved electing workers to a bargaining committee and conducting contract negotiations with Dofasco. If a tentative collective agreement was reached, workers would then have an opportunity to vote on the agreement and, implicitly, union representation. Ultimately these provisions were not enacted because the USW found insufficient support for unionization among Dofasco workers (USW, 2008).

Neutrality agreements surfaced in the US in the late 1970s (Brudney, 2005). Their proliferation was due in large part to unions’ dissatisfaction with the National Labour Relations Board’s (NLRB) union recognition procedure, which employers’ legal action can seriously protract. Delays in the formal certification process, increasing use of anti-union tactics by employers, “right to work” legislation, and other such factors resulted in a sharp decline in union density (Eaton and Krieksy, 2001). This led to experimentation with other approaches to revitalizing the union movement (Eaton and Krieksy, 2006; Rose and Chaison, 2001). Accordingly, neutrality agreements premised on voluntary contracting are now accepted by some unions as a strategic option for union renewal. However, some US employers have challenged the legality of such agreements (Cohen, Santucci and Fritts, 2006).

Neutrality agreements have been adopted in Canada since the 1980s for similar reasons. First, in some Canadian jurisdictions, union organizing has become more difficult since the adoption of secret ballot elections in place of the card certification method (Martinello, 2000) and election delays (Campolieti, Riddell and Slinn, 2007). These changes have increased the potential for employer interference in organizing campaigns, which has been shown to lower union certification rates (Bentham, 2002; Johnson, 2004), especially in the private sector, where density has sharply declined in the past three decades (Godard, 2003).

Doorey (2007) argues that contemporary labour law has played a subtle yet critically important role in legally validating the above mentioned non-statutory approaches.[6] To test this claim, we searched provincial and federal labour codes for provisions explicitly related to voluntary union recognition agreements. In some jurisdictions, legislation is silent on voluntary union recognition (Federal, Newfoundland and Labrador, Manitoba, Saskatchewan, and British Columbia). In these jurisdictions, however, labour board decisions accept voluntary recognition agreements. For example, in a recent decision, the BC Labour Relations Board (BCLRB) stated that voluntary union recognition schemes “are widely used and accepted in labour relations in BC and are consistent with Code principles” (BCLRB, 2008: 4). An earlier landmark BCLRB decision makes clear the Board’s jurisdiction: “Voluntary recognition originates outside of the Board’s jurisdiction. It is born out of the parties’ agreement as opposed to statutory compulsion” (Delta Hospital, BCLRB No. 76/77, [1978]: 367).

As artifacts of the pre-WWII industrial voluntarism, statutory provisions related to voluntary recognition appear in several jurisdictions (Prince Edward Island, Nova Scotia, New Brunswick, Ontario, and Alberta). Across these provinces, though, there is variation both in the degrees to which the law recognizes voluntary agreements and to which voluntarism is advertised as an alternative to statutory processes. For example, in Ontario in 1993, the New Democratic government engendered conditions that were more favourable for the negotiation of neutrality agreements when it amended the Ontario Labour Relations Act so that voluntarily recognized unions could legally strike and access state-sponsored conciliation (Doorey, 2007). The Alberta Labour Relations Board (ALRB), which publishes detailed information about the potential benefits and costs of voluntary recognition, notes that one of the main advantages of voluntary recognition schemes is that such pacts can create a “spirit of cooperation” between employers and unions (ALRB, 2003). Moreover, the landmark BCLRB decision compared voluntary union recognition agreements to statutory certification this way:

The principal advantage of [statutory compulsion] is that it is an orderly, statutory process and one which is overseen or monitored by an independent tribunal having the responsibility to protect the legitimate interest of employers, trade unions, and employees. The most commonly noted advantages to recognition by agreement are that the parties have come together initially on amicable terms rather than as adversaries; that the parameters of the bargaining relationship are determined by the parties themselves rather than by some external agency which may or may not fully understand the intricacies of their work situation; and that expense and delay are avoided. (Delta Hospital, BCLRB No. 76/77, [1978] 367 and 371)

The language used in the BCLRB decision and in the ALRB guide is strikingly reminiscent of the spirit of pre-PC 1003 era language, when employers and the state equated ‘reasonable’ unionism with cooperation and shared interests, and ultimately, voluntary contracting. While these examples are not concrete evidence that all Canadian governments deliberately use “legal norms, procedures, and sanctions to ‘frame’ or ‘steer’ the process of self-regulation” (Barnard, Deakin and Hobbs, 2005: 4) in industrial relations, the evidence does suggest, at a minimum, that voluntarism has been accommodated and given legal status. In Ontario and Alberta, voluntary recognition is overtly presented as an alternative to statutory mechanisms. In the next section, we focus on the recent neutrality agreement between Magna International and the Canadian Auto Workers (CAW), known as the Framework of Fairness Agreement (FFA). Further, we use this case to compare contemporary and historical manifestations of voluntarism.

A Case Study of Voluntarism: The Magna-CAW FFA

Within this legal context, Frank Stronach, founder of Magna International, and Buzz Hargrove, then President of the CAW, negotiated the FFA. Announced in October 2007, the wide-ranging deal brought an end to years of sometimes bitter conflict between Magna and the CAW. Throughout the 1970s and 1980s, the US-based United Auto Workers (UAW) (the predecessor to the CAW), attempted to unionize workers at Magna’s Canadian plants, but with little success. In 1999, the CAW made limited inroads when it managed to organize workers at the Magna-controlled Integram Seating plant, and, subsequently, two additional Magna plants. It also negotiated a “neutrality letter” with The Big Three, a key part of which stated that Magna workers should “decide whether or not to join a union in an atmosphere free of intimidation, interference or risk of reprisal” (Keenan, 1999: B9). But, despite significant investments by the CAW/UAW in organizing Magna workers, little further was achieved.

Magna has successfully staved off traditional union organizing attempts because of its corporate culture, human resource management system, and subtle resistance to unions. The Magna “Employee Charter,” a social contract established by Stronach, guarantees workers “a safe and healthful workplace,” “fair treatment,” “competitive wages,” and “employee profit participation” (Magna, 2008)—many of the provisions workers might look to a union to provide through collective bargaining (Lewchuk and Wells, 2007a). In addition, an “Open Door Policy” and a corporate telephone Hotline allow employees to voice workplace concerns. In the minds of some workers, these protections, and compensation on a par with that of workers at unionized auto plants, may obviate the need for a union. Lewchuk and Wells (2007a), though, found evidence that this view is not universally shared, especially among Magna workers who fear repercussions for speaking up about work-related problems. Other evidence suggests that, in the past, some Magna managers deliberately tried to undermine CAW organizing campaigns (Lewchuk and Wells, 2007b).

In September 2005, Stronach initiated informal negotiations with Hargrove to develop a new labour relations system, one that would be based on co-operation rather than conflict. Following two years of private talks, Stronach and Hargrove announced the FFA.[7] Publicly, the two men stated that they entered the agreement for the sake of the competitive position of the Canadian auto manufacturing industry, which faces increasingly cut-throat global competition (Stronach and Hargrove, 2007). In the spirit of co-determination, Stronach and Hargrove jointly authored an editorial published in the Globe and Mail:

Magna accepts the CAW as a genuine partner, with a crucial role to safeguard the interests of Magna’s workers as the company grows and changes. And the CAW accepts Magna’s culture of ‘fair enterprise,’ and the unique structures we’ve put in place over the years to make decisions and resolve concerns with maximum worker participation (Stronach and Hargrove, 2007: A21).

The motivation for the deal is undoubtedly strategic, and both Hargrove and Stronach believe that the FFA can help them achieve disparate objectives. For Hargrove, the most pressing challenge is to maintain (and potentially increase) private sector union density in Canada, which has declined in the last decade.[8] There is grave concern that union density in Canada may follow the US experience, where public and private unionization has sharply declined since 1980 (Kumar and Schenk, 2006). In the Canadian auto sector, the decline in unionization has been pronounced over the past two decades as General Motors, Chrysler, and Ford have lost market share and shifted from vertically-integrated auto manufacturing to automobile assembly with components produced by non-unionized subcontractors. At the same time, auto manufacturers Toyota and Honda have avoided unionization by using human resource management practices which share some similarities with Magna’s employee relations system.

For Stronach, too, concluding the FFA was a matter of strategy. In many ways, Stronach’s motivations are more intriguing, especially since he initiated the negotiations leading to the FFA. The deal represented an end to industrial instability caused by untimely union organizing, legal battles, and potential labour unrest. By organizing with the CAW, Magna’s founder was also able to select the union he wanted and keep out potentially more militant rivals that might emerge in the future.[9] More importantly, by organizing voluntarily, Stronach may have been able to negotiate for more favourable terms than could be reasonably expected in publicly-followed collective bargaining with the CAW (Van Praet, 2007). Finally, and in a related vein, the FFA preserves Magna’s human resource management system now and will continue to do so when Stronach relinquishes control of the company. Since Magna uses an innovative employee relations system and, at the time of the FFA negotiations, provided similar wages and benefits offered by unionized competitors (Lewchuk and Wells, 2007a), unionizing under terms set out by the FFA it may have been assumed that Magna would not incur significant financial costs. These reasons align closely with Eaton and Kriesky’s (2006) findings about employers’ motivations for negotiating neutrality agreements.

Such ostensibly radical concessions on the parts of both Hargrove and Stronach were sure to draw sharp reviews, both negative and positive. Opposition to the FFA centres on several key provisions, but has not questioned the principle of contracting outside statutory processes. A controversial provision of the FFA states that impasses in collective bargaining will go to binding arbitration instead of a strike or lockout. Ed Broadbent decried this no-strike clause, stating that “workplace democracy becomes a sham when unions lose their only instruments of power, the right to withdraw their labour and the right to select, without management approval, their workplace representatives” (2007: A25). The FFA also stipulates that there will be no union stewards in Magna plants; instead, “employee advocates” will be chosen by union members and management. On these points, some have called the deal a sell-out to independent unionism and have warned that it would likely set a precedent which could negatively affect the interests of unionized and non-unionized workers alike (Gindin, 2007).[10] Despite these charges the FFA was approved at the CAW annual meeting in December 2007.

In contrast, proponents of the FFA see the pact as an opportunity to improve Magna’s competitive position, and by extension, the conditions for workers (Armstrong, 2007). Furthermore, greater co-operation between the CAW and Magna may improve their ability to lobby governments to re-negotiate global trade agreements (Armstrong, 2007). On the criticism regarding the union’s concession of the fundamental right to strike, supporters of the FFA argue that Magna workers can vote no to the agreement, and so retain this right (CAW, 2007). More recently, Hargrove publicly stated that the FFA “could be a model for other non-union operations like Honda, Toyota and Dofasco if they would be willing to invite [the CAW] in and have a neutrality agreement that says they welcome the union and workers would make the decision” (Van Alphen, 2007: B5).[11]

Comparing Historical and Contemporary Manifestations of Voluntarism

The most obvious similarity between the voluntarism today and that of yesterday is that contemporary voluntarism represents a form of private contracting for union recognition, which was the norm in the early 20th century. Mackenzie King, who for decades refused to support legislation guaranteeing workers compulsory union recognition, would likely approve of today’s modest experiments with self-regulated industrial relations.

However, the circumstances that gave rise to neutrality agreements and other non-statutory mechanisms reveal important differences between contemporary and historical voluntarism. At present, Canadian unions have a choice between, on the one hand, initiating negotiations with employers, or, on the other hand, attempting to gain recognition through state-sponsored certification procedures. The boycott of the IRC in BC is interesting in that the leadership of the BCFed openly declared that areas of the boycotted legislation “that may prove to be of benefit to unions will be used” (Goldberg-Hiller, 1996: 337). In the case of the CAW, it certainly benefited from having an alternative in the Ontario labour board’s certification process. It is unlikely that Magna would have agreed to the FFA in the absence of a real threat of continued organizing. One of the appeals of voluntarism may be the potential reductions in legal expenses associated with employer attempts to resist unionization. Hargrove (2009) noted that if unions and employers can avoid forming an adversarial relationship “then naturally [they] start out with a more mature relationship and that will even improve over the years.”

Related to this, the ability of the state to enforce voluntary recognition agreements is greater now than in the past; the legal standing of voluntary union recognition agreements does, however, vary across jurisdictions. According to Doorey (2007, 2009), in Ontario, successful voluntary recognition has the same “practical effect” (2007: 92) as the formal Board certification system; however, Section 66 of the OLRA stipulates that any time during the year after a voluntary recognition agreement was signed, an employee or other union member representing an employee in the bargaining unit may apply to the Board to have it declare that the union is not entitled to represent the bargaining unit. Similarly, in Manitoba and Alberta, voluntary recognition agreements do not enjoy the “same rights and privileges in law” (Manitoba Labour Board, 2006: 19). The Alberta Labour Relations Code provides unions with less legal protection because either a union or employer can sever recognition with six months notice before the expiry of a collective agreement.

Overall, contemporary labour codes provide significantly more protection than was afforded to unions in the early 20th century model of self-regulated industrial relations. It is still too early to tell whether the dual process for union recognition acts to rebalance power to enable voluntarism to be something more than a fig leaf for managerial power. Further, in the case of the FFA, it is unclear what direct role Ontario labour law had in shaping the deal.[12]

Recent manifestations of voluntarism have shown a remarkable ability to work within the contours of the law. Historically, the legal vacuum that existed with respect to labour law meant that employers and unions could fashion agreements which today would be illegal (e.g., formation of company unions). Looking ahead, union-and-employer-driven voluntarism may be opposed if non-unionized workers do not have meaningful opportunities to shape and, if desired, reject such deals. As the ALRB stated: “Voluntary recognition is not a way of circumventing the employees’ freedom to choose union representation, but of facilitating that choice (24).”

Lastly, and importantly, deals like the FFA cover a broad range of labour-management issues beyond union recognition. Publicly, the CAW leadership framed the negotiations and concessions leading to the FFA this way:

[Like] any bargaining, negotiations over voluntary recognition involve give and take. The union can’t dictate the deal. And the deal will never be perfect. But if it allows the union to get its foot in the door, without the employer crushing us before we start, then it’s a victory (CAW, 2007: 12).

Indeed, the BCLRB’s statement regarding the potential positive relational and other benefits of voluntary union recognition rings especially true in the context of the FFA: “Parties [that] come together initially on amicable terms rather than as adversaries [may realize benefits because] the parameters of the bargaining relationship are determined by the parties themselves rather than by some external agency which may or may not fully understand the intricacies of their work situation.” First, the FFA suggests that employers can extract significant concessions as a quid pro quo for voluntary union recognition. Primary among these is negotiating terms of a collective agreement before employees have an opportunity to consent to union recognition. Second, and consistent with unions and employers coming together on amicable terms, Hargrove (2009) believes that parties “can have a mature relationship by starting out on a mature basis.”


This paper attempts to understand post-war developments in non-statutory union recognition in historical context. For much of the first half of the 20th century in Canada, politicians and employers tried and (mostly) failed to have unions accept the ideology of industrial voluntarism. Having experienced the institutional model of union recognition in the post-war period, some unions have shown a willingness to appropriate strategically the practice of voluntarism. For the union movement, there are inherent risks in maintaining the status quo, just as there are to contracting outside labour laws. In contrast, it is clear why non-union employers, when faced with a real threat of union organizing, might prefer to negotiate, for example, a neutrality agreement. Not only can such deals allow them to select a preferred union, but also, as in the case of Magna, they can legitimate human resource management systems, guarantee industrial stability, and constrain the scope of collective bargaining. Overall, the Canadian experience indicates that the strength of voluntarism in the post-war period is linked to the perceived efficacy of institutional certification processes and the compulsion employers feel to engage in such negotiations. Given the current economic uncertainty, most employers would probably be reluctant to enter such negotiations.