Corps de l’article

Canadian policy is not as favourable to the promotion of collective bargaining... [T]here has been a more positive attitude toward collective bargaining in the United States than in Canada.

H.D. Woods, Professor of Industrial Relations, McGill University, 1962.

Over the past four decades, union membership in the United States has slumped from one third of the working population to a post-depression low of 13.5 percent, while Canadian membership rates have remained surprisingly stable at about 31 percent of the non-agricultural workforce. Today, in both the private and public sectors, Canadian union membership is about double that of the U.S. (Greenhouse 2001; Akyeampong 1999). Numerous labour scholars have identified public policy and employer opposition as critical factors contributing to the divergent fortunes of the two union movements (Rose and Chaison 1995; Taras 1997). Since the 1970s, the American Federation of Labor-Congress of Industrial Organizations (AFL-CIO) has argued that Canadian labour policy functions in a manner that is closer than its U.S. counterpart to the original intent of the National Labor Relations Act (NLRA) because it regulates aggressively employer opposition to unionization. Thus, it has looked to Canadian policy for a model for the reform of U.S. labour law (AFL-CIO 1985).

Previous explanations of the differences between U.S. and Canadian labour policy have stressed Canada’s social democratic political culture and institutions, decentralized federalism in labour relations, provincial policy experimentation, and parliamentary system of government (Lipset 1989; Bruce 1989). While these accounts document how, in recent decades, organized labour has contributed to the enactment of “pro-union” national and provincial laws, they fail to recognize the extent to which these laws are consistent with the principles and practices of state intervention in labour-management established in the 1940s and 1950s. This article analyzes the wartime and post-war development in Canada of two critical differences between Canadian and U.S. labour policy: trade union certification and state regulation of collective bargaining (including the regulation of striker replacements). Since the 1970s, Canada’s “pro-union” policy on certification and striker replacements has insulated organized labour against employer opposition, a major cause of union decline in the U.S.

But the importance of state regulation of collective bargaining and trade union certification extends beyond their contribution to the recent fortunes of organized labour. The development of public policy on these issues illuminates the fundamental principles of the post-war system of state intervention in labour-management relations. In the name of preserving industrial peace, Canadian lawmakers have circumscribed the economic weapons of unions, especially their right to strike, and established stringent requirements for trade union certification; but they have also restricted employers’ recruitment of striker replacements and severely limited management involvement in the certification process. In the name of protecting the free choice of employees, U.S. policymakers, in contrast, have imposed less onerous requirements for trade union certification and placed fewer restrictions on strikes. However, they have also deregulated employer behaviour, allowing management to permanently replace economic strikers and electioneer aggressively during National Labor Relations Board (NLRB) elections. In the immediate post-war decades, when outright employer opposition to collective bargaining was relatively rare, unionists and their political allies lambasted the onerous requirements of federal and provincial certification policies and intrusive regulation of the process of collective bargaining. Instead, they preferred the lower certification requirements and less intrusive system of state intervention in the U.S. But during the past three decades, an era of heightened international competition, deregulation, capital mobility, and political neo-liberalism, Canada’s extensive regulation of labour relations has protected workers against market-driven anti-unionism and helped preserve the institutions of collective bargaining. Thus, instead of looking admiringly to the U.S., Canadian unions have struggled against the creeping Americanization of federal and provincial labour policies in recent years.

The article is based on original research in government, labour, and management archival collections (especially those of the Labour Department, the Canadian Labour Congress, Canadian Manufacturers Association and Chamber of Commerce), other contemporary materials (such as newspaper and conference proceedings), and secondary sources. It is divided into two main parts. The first part examines the major developments in the post-war history of trade union certification, focusing on three issues: “cards versus ballots,” the thresholds of support that unions are required to demonstrate for certification, and state regulation of employer opposition during representation campaigns. It argues that the fundamental principles of Canada’s post-war certification system have continued to influence the policy debate on trade union recognition during the past three decades. The second part analyzes state regulation of the process of collective bargaining, including federal and provincial policy on the use of striker replacements. Beyond examining the impact of striker replacement policy on the outcome of economic strikes, we must consider the issue within the larger framework of the wartime and post-war debates over restrictions on the economic weapons of unions and management. The article also documents labour and employer efforts to influence public policy on certification and replacements, with particular reference to events in Ontario and the federal jurisdiction. While it focuses primarily on the post-war decades, the article extends its discussion of certification and striker replacements up to the 1990s in order to demonstrate the underlying continuity between developments in the 1940s and 1950s and more recent policy innovations.

Trade Union Certification in Post-war Canada

During the past half century, Canada’s system of trade union recognition has developed in a very different direction from that of the U.S. In contrast with the mandatory elections, protracted campaigns, and aggressive employer opposition in America, post-war Canadian certification procedures have not presented the same obstacles to union recognition. Several provinces issue certifications on the basis of union membership cards. Those provinces that currently require secret ballots conduct elections quickly after a union’s application for certification—thereby limiting the duration and effectiveness of management campaigns—and enforce tough sanctions against employer interference with workers’ free choice. Between 1950 and 1995, Canadian unions consistently won close to 70 percent of their applications for certification, and several commentators have attributed their greater organizing successes, when compared with U.S. unions, to these national differences in certification policy (Martinello 1996; Thomason and Pozzebon 1998).

Cards versus Ballots and Voting Quotas

The fundamental features of Canada’s current system of trade union certification were established during the 1940s and 1950s. In response to escalating wartime strikes, labour’s growing political influence, and the recommendations of the National War Labour Board (NWLB), the Liberal government enacted PC 1003 in February 1944. This provided the basic legal framework for federal and provincial labour law for the next half century. Despite the striking differences between Canadian and U.S. certification procedures in recent years, PC 1003’s system of statutory recognition was modelled closely on that of its southern neighbour. Like the NLRA, PC 1003 incorporated the principles of exclusive representation and majority rule, and allowed the NWLB to certify unions on the basis of documentary evidence demonstrating that over 50 percent of eligible workers desired union representation or by conducting secret ballot elections. Prior to certification, however, the labour board required that unions win the support of over 50 percent of workers eligible to vote in elections (MacDowell 1978).

Canada’s two main labour federations, the Canadian Congress of Labour (CCL) and Trades and Labour Congress (TLC), attacked this rule because non-voters were effectively counted as votes against the union. Instead, labour leaders proposed reforming wartime election policy to conform with U.S. certification policy, which required that unions win only a majority of votes cast in NLRB elections.[1] But the wartime Liberal government dismissed outright labour’s election proposals. The Labour Department claimed that lower voting quotas for certification would “intensify” problems encountered under PC 1003 when unions collected authorization cards “under dubious circumstances,” then petitioned for elections “in which they had nothing to lose and everything to gain.” Every province except Saskatchewan required that, prior to certification, unions win the votes of a majority of eligible workers, and the government concluded that this rule had “worked very satisfactorily” and therefore “should be continued” in future certification procedures. Thus, the Industrial Relations and Disputes Investigation Act (IRDIA), passed by parliament on June 17, 1948, included the provision that unions must win the support of a majority of eligible voters in representation elections.[2]

Union leaders continued to attack the “unduly high percentage” of employee support required for both card and election certifications throughout the 1950s and 1960s. The Canadian Labour Congress (CLC) warned that employer coercion had frequently frustrated the true wishes of “vulnerable” employees. When an employer was “openly hostile to the formation of a union,” the CLC argued, it was “often impossible to get a substantial number of employees to stick their necks out by signing any union document, even though a majority of them may actually want the union to represent them.”[3] The CLC also justified its demand for lower voting quotas by contrasting the standards that operated in representation elections with those of political elections. The CMA replied that its analogy was “unsound” because, unlike political elections, certification was a “change in the form of government for employees” and thus it should be supported by a clear majority of all employees.[4]

Until the late 1970s, national and provincial governments generally accepted employers’ arguments against lowering certification requirements. Along with the federal jurisdiction and several other provinces, Ontario finally lowered the levels of employee support required for both card and election certifications in the 1970s. Certain jurisdictions, including Ontario, lowered the level of support that unions were required to demonstrate in order to obtain certification elections from 45 percent to 35 percent and the support required for card certifications from 65 percent to 55 percent. Ontario, Manitoba, Nova Scotia, and the federal jurisdiction also allowed the certification of unions that won a majority of votes cast in recognition elections (Arthurs 1967; Woods 1973). Unions and their political allies justified the new certification rules as the “best way” to prevent employer coercion and “in the tradition of normal democratic practices,” while management representatives attacked their “overt bias,” which demonstrated that the government had abandoned its “customary position of neutrality.”[5]

Employers have opposed card certifications throughout the post-war decades. Between 1944 and 1948, anti-union employers resisted government attempts to certify unions and impose bargaining (Fudge and Tucker 2001). After the enactment of the IRDIA in June 1948, many large firms reluctantly accepted the inevitability of collective bargaining with outside unions but demanded that Canada’s largest employer organization, the Canadian Manufacturers’ Association (CMA), lobby parliament for mandatory certification elections. They insisted that union organizers frequently pressured “contented” workers into signing authorization cards, and thus argued that labour boards should not certify unions on the basis of “fraudulent” evidence that “did not reflect the true wishes of the employees.” Voicing the opinion of many irate employers, one manager complained that organizers had “their means” of persuading employees to sign cards: “Employers should be entitled to know that the majority of the employees do, in fact, desire union representation and we know from experience that this can be only determined through a properly conducted secret ballot.”[6] Employers also complained that card certifications denied them adequate opportunity to respond to “untruthful” union propaganda; secret ballots, they insisted, would both protect vulnerable workers from union coercion and give employers a chance to explain to employees their opposition to collective bargaining.

The CMA agreed that cards were “unreliable” evidence of employees’ desire for union representation, but it discouraged employers from demanding elections in all representation disputes, pointing out that the IRDIA had introduced “more stringent requirements” for card certifications. Shortly before the enactment of the IRDIA, Labour Minister Humphrey Mitchell had warned parliament that certifications based on authorization cards were PC 1003’s greatest failing. Thus, over labour opposition, he secured the inclusion in the IRDIA of a provision requiring that unions must demonstrate that a majority of employees were dues-paying members, whereas PC 1003 had required only that workers sign authorization cards.[7] Hailing this amendment as “a great improvement in the law,” the CMA opposed mandatory elections because it believed that many workers, who would fail to meet the new law’s criteria for “members in good standing,” might nevertheless vote for unions in secret ballots. The CMA explained to one disgruntled employer that it had rejected the idea of mandatory elections “because you would have a vote where a minority were union members, but the non-members would vote for the union.” And it cautioned another manager that mandatory elections would “hurt the employer’s position, because if there are not sufficient members in good standing, the Board is likely to throw out the certification without a vote.” Thus, insisting upon elections might, in certain instances, “harm employers and promote unions,” even where support for collective bargaining was weak.[8]

Although the CMA subsequently supported mandatory elections, its steadfast opposition to them in the late 1940s undermined employers’ protests during this critical period of policy formulation. By the time the CMA had changed its position to one of demanding elections in all representation disputes, the Ontario Labour Relations Board (OLRB) resolutely supported the well-established practice of card certifications, and it certified mostly on the basis of membership cards for the next half century. The pro-management amendments to the Ontario Labour Relations Act (OLRA) enacted in 1970 were designed to encourage more elections and discourage card certifications, but they had limited impact and were reversed by further amendments to the law in 1975 (Carter 1992). Employers fared no better with the federal government. Management representatives argued for mandatory elections under the federal labour code in their response to the recommendations of the Woods Task Force (1968), but the government rejected their proposal (Woods 1968).

During the past two decades, Canada’s trademark system of card certification has started to unravel, as several provinces, including BC, Alberta, and Ontario, have enacted laws requiring mandatory elections. Prior to the election of the Progressive Conservative government of Mike Harris in 1995, the OLRB had relied overwhelmingly on card certifications for almost half a century. The new government did not share this enthusiasm for card certifications, however, and a few months later, it introduced mandatory elections for the first time in Ontario’s history; union victory rates and applications for certification declined in number almost immediately (Martinello 2000). But even under the Harris government, the labour board has held expeditious elections and enforced tough sanctions against unlawful interference with employees’ choice of bargaining representatives. And despite sustained opposition from management organizations, several other provinces have continued to certify unions on the basis of documentary evidence of union membership. Initially conceived as part of an overall system of trade union recognition that required that unions demonstrate high levels of membership support, card certifications have become a critical and distinctive characteristic of Canada’s system of labour-management relations.

State Regulation of Employer Opposition During Representation Campaigns

In February 1947, the Liberal government terminated PC 1003, originally an emergency wartime measure, and provincial governments again assumed responsibility for labour relations. Shortly before the termination of PC 1003, employer associations and provincial premiers argued for provincial control of labour relations, while most unionists and their political allies favoured a strong national labour code (Coates 1973). The premiers and employers prevailed and by the early 1950s all provinces except Quebec and Saskatchewan had enacted laws modelled on the IRDIA, covering the workplace rights of almost 80 percent of Canadian workers. Over the next several years, provincial labour boards developed complex certification provisions that differed in their details from one another. But decentralization of labour policy has not produced eleven markedly different systems of certification and in the post-war years, a distinctly “Canadian” system of trade union certification developed at both the federal and provincial levels. Even under conservative governments, most provincial boards have preferred card certifications, but have required that unions demonstrate between 55 and 65 percent union membership prior to certification. If unions provided evidence of between 25 and 55 percent membership, labour boards ordered representation elections (but only Saskatchewan and Ontario required that unions provide less than 50 percent proof of membership). And most provincial laws included provisions allowing for the automatic certification of unions that had demonstrated between 50 and 55 percent support if unlawful interference had rendered impossible a free election (Logan 1956).

Prior to the enactment of the IRDIA, the CMA and the Chamber of Commerce lobbied for a provision in the new law protecting employers’ freedom to communicate with employees on the issue of unionization. When the U.S. Congress enacted the Taft-Hartley Act in June 1947, which included an employer free speech provision [section 8(c)], they campaigned for an identical provision in Canadian law. The Chamber of Commerce argued that without a provision protecting employer speech, the IRDIA’s restrictions on management opposition to unionization would “constitute an unwarranted interference with the necessary right of an employer to manage his own business.” But the Liberal government categorically rejected its demands for a Canadian Taft-Hartley, insisting that a free speech provision was unnecessary because, unlike U.S. labour policy, Canadian law had “not gone to extremes.... The pendulum has not swung so far in either direction.”[9] Thus, the IRDIA and six provincial laws initially contained no direct reference to election conduct beyond a general prohibition of employer interference with workers’ right to organize, while a minority of provincial laws explicitly prohibited aggressive electioneering by either management or unions (Woods 1973). Instead, national and provincial labour boards developed a strict policy against aggressive employer opposition during representation campaigns in the years after 1948.

State Regulation of Employer Opposition in Ontario

Developments in Ontario, the most populous and most industrialized province that often established precedents in labour policy, have been particularly important. The OLRA initially made no direct reference to election conduct, and in the late 1940s the OLRB adopted a relatively permissive position on campaign electioneering, overturning election results only in cases involving clear evidence of employer coercion (Millar 1981). In response, the CMA encouraged its members to fight vigorously against unionization, but acknowledged that “this matter of speaking to employees” during organizing campaigns was “a bit tricky.”[10] By the early 1950s, however, the OLRB was consistently ruling against aggressive employer electioneering, much to the chagrin of the CMA. In Underwood Manufacturing Co. Ltd. (1952), the OLRB ordered the immediate certification of the union after the company laid off several workers, including several union activists, at the start of an organizing campaign.[11] The board overturned an election victory for the employer after a company union had circulated leaflets slandering its CCL opponent in Joseph Gould & Son (1952).[12] And the OLRB refused to nullify an election that the union lost after the employer had delivered a “captive audience” speech in Savage Shoes (1960), but also stressed that its decision “must not be construed as a ‘carte blanche’ to employers” to hold captive speeches under any circumstances. On the contrary, in future cases, the board warned that it would subject to close scrutiny both the context and the content of management communications.[13]

Incensed by the board’s tough line on employer electioneering, the CMA campaigned energetically for a free speech amendment to the OLRA. It complained that, under the existing rules, employers “get the impression from the rulings of the labour board that they cannot lawfully discuss union matters in any way with their employees or make any statements whatsoever on this subject.”[14] Its lobbying eventually paid off. In 1958, the Ontario Legislature’s Select Committee on Labour Relations recommended the addition to the law of a provision protecting employer speech, and two years later, the provincial legislature amended the OLRA to include such a provision.[15] Unlike Section 8(c) of Taft-Hartley, however, Ontario’s free speech provisions did not significantly benefit employers who were intent on defeating organizing campaigns. Indeed, in its first free speech disputes under the amended law, the OLRB again ruled against aggressive management electioneering. In Sun Tube (1962), the board certified the union without an election after the company held several captive speeches predicting dire economic consequences if workers voted to unionize.[16] The OLRB ruled against an employer’s interrogation of employees about their union sympathies in Piggot Motors (1962).[17] And the board ordered the immediate certification of a union because of both the context and the content of the employer’s captive speeches in Wolverine Tube (1963).[18] As a result of these landmark decisions, the OLRB avoided replicating the U.S. certification system which, by the 1950s, had allowed management greater freedom to oppose unions than ever envisaged by the authors of the NLRA.

The OLRB has enforced an even tougher policy against unfair management practices during representation campaigns during the past three decades. Amendments to the OLRA in 1975 significantly enhanced the power of the OLRB to impose automatic certifications if unfair management practices had rendered it unlikely that the true wishes of the employees would be disclosed through an election. Designed to prevent recalcitrant companies from benefiting from their own unlawful practices and to deter future misconduct, the new amendments produced an immediate change in board policy. In several landmark decisions, the OLRB imposed automatic certification on employers who had predicted plant closures, relocations or job losses during certification campaigns. Following the distribution of anti-union letters that were “deliberately calculated to play on employee fears for their job security,” the OLRB certified the textile workers’ union, even though it had lost the representation vote at Dylex, Ltd. by an overwhelming margin.[19] In Viceroy Cont. Ltd. (1978), the OLRB found that an employer’s statements that jobs may be lost if the plant were to unionize, “however factual ... could reasonably be perceived by the employees as a clear threat to their jobs.”[20] And, for the first time, the OLRB ordered the certification of a union without opening ballot boxes after Lorain Products Ltd. had provided employees with a “Hobson’s Choice” by warning in letters and speeches that if the union won, they would have to strike in order to maintain existing levels of jobs and benefits.[21]

Ontario employers condemned these “punitive” decisions, but their protests went unheeded. A few years later, in its landmark Radio Shack decision (1979), the OLRB incorporated “the most comprehensive set of remedies... ever fashioned by a labour board in Canada.”[22] During an organizing campaign and subsequent contract negotiations, the company had dismissed two activists, spied on the union, threatened to “move out west,” refused to reinstate one employee, disparaged the board’s procedures, and failed to bargain in good faith. Finding the company guilty of “flagrant” transgressions of the law, the board ordered Radio Shack to introduce a compulsory dues check off, bargain in good faith, pay compensation to both union and employees, provide names and addresses of employees, and refrain from its campaign to undermine the union.[23] While the OLRB rejected the union’s request to impose a first contract, its sweeping remedies demonstrated the expanded powers of the board after the 1975 amendments to the OLRA. Management representatives castigated the Radio Shack decision as “an extremely harsh one” that had “far surpassed” previous decisions in disputes involving unfair management practices (Filion 1980). But former OLRB chairman Don Carter supported the decision and argued that the expansion of the board’s remedial powers was “unquestionably the most important” legal development of the 1970s, one that had “breathed new life” into long-existing prohibitions on employer interference with the right to organize (Carter 1976, 1992).

During the past two decades, the OLRB has continued to impose tough penalties on employers who commit unfair practices, while management associations have continued to lobby on the question of employer communication during certification campaigns. After the election of the Progressive Conservative government in 1995, the CMA once again complained that employers were prohibited from communicating with their employees “in anything other than the most limited manner.... Employers must be allowed to make any statement or express any opinion that is reasonably held.”[24] In response, the government introduced Bill 7 in November 1995, a central purpose of which was to “encourage communication between employers and employees in the workplace.” The CMA also recommended, without success, that the government allow three weeks (rather than the proposed five days) between the posting of a certification application and the holding of an election because this would ensure a “reasonable period of time to disseminate necessary information” for employees to make an informed choice.[25] After the CMA lost its battle for longer certification campaigns, one commentator complained that the five-day limit deprived employers of “the practical opportunity to wage an effective counter-campaign” (Levitt 1999). But the OLRB has defended the five-day limit as the “most critical characteristic” of the new certification scheme and it has largely adhered to this strict time frame, thus restricting both the duration and intensity of management campaigns (Blaikie 1998).

Like the 1960 amendment to the OLRA protecting employer speech, however, the 1995 amendment did not immediately alter the OLRB’s policy on employer communications, as clearly demonstrated by its Wal-Mart decision. In February 1997, the Ontario board imposed a union on a Wal-Mart store in Windsor, Ontario, even though the employees had voted by more than three to one to reject union representation. The OLRB found that Wal-Mart had refused to say whether the store would be closed if employees voted to unionize, thus contravening its own much-vaunted “open door” policy, which the company had called its “greatest barrier to union influences trying to change our corporate culture and union-free status” (Wal-Mart 1997).[26] Welcomed by union advocates as entirely consistent with twenty years of case law in Ontario, the controversial Wal-Mart decision caused widespread consternation among employers’ representatives. Once again, however, the Conservative government was quick to respond. On June 4, 1998, the government introduced Bill 31 (the Economic Development and Workplace Democracy Act), the so-called “Wal-Mart bill,” which removed the board’s authority to issue mandatory certifications, making Ontario one of only two provinces (with Alberta) that lacked this powerful sanction against unlawful interference.[27]

But the OLRB still severely regulates management opposition and imposes tough penalties against illegal interference, as demonstrated by its recent Barons Metal decision. In May 2001, the OLRB ordered a re-run election at Baron Metal Industries after the company had hired two members of a Sri Lankan criminal gang to intimidate its mostly Tamil employees shortly before an election. In addition to mandating a second election, the OLRB ordered the firm to pay compensation to employees and union, provide the union with an office inside the plant, and to pay employees for regular meetings with the union on company premises.[28] As a result of such decisions, one employer representative has recently cautioned that “simply because automatic certification is no longer available, the OLRB is definitely not powerless” when it comes to regulating management opposition during representation campaigns (Boniferro 1998).

State Regulation of Employer Opposition in the Federal Jurisdiction

Employers have fared no better on the issue of election communications under the federal jurisdiction. While federal law covers fewer than one tenth of the nation’s workers, Canada Labour Relations Board (CLRB) decisions have often established important policy precedents that were later adopted by provincial labour boards, particularly in the decades prior to the 1970s. Throughout the 1950s and 1960s, the CMA and Chamber of Commerce lobbied for a free speech amendment to the national labour code. In 1968, one study of management associations reported that a leading policy objective was “freedom of expression for employers” during representation campaigns, but the Woods Task Force firmly rejected their demands for greater liberty to communicate with employees (Patterson 1968; Woods 1968). Indeed, the Task Force stressed that the one situation in which restrictions on employer speech were unequivocally justifiable was the representation campaign.

The CLRB ruled upon surprisingly few free speech cases in the post-war decades, but in those cases it has tackled, the board has strictly regulated management communications (Carter 1992). In one landmark case, Bank of Montreal (1985), the CLRB defended its requirement that employers remain neutral during representation campaigns after the employer had held captive meetings and personal interviews intended to dissuade employees from voting for unionization. The board ruled that, because of the economic power of the employer over its employees, simply by holding a captive meeting, the bank had run foul of the law: “The economic power of the employer puts it in a position to compel an audience. The purpose of making a captive audience meeting an unfair labour practice is to remove that leverage.... Even though there may be no actual threat, the employer’s taking the initiative to convey its anti-union views amounts to undue influence because it restricts the employee’s real freedom of choice.”[29] Thus, the Canada Labour Code has generally adopted a stricter standard for regulating employer communications during certification campaigns than has been the case in most provinces, where labour boards have enjoyed greater liberty to balance employers’ freedom of expression against employees’ freedom of association.

In recent years, the national board has continued to impose tough penalties on recalcitrant employers. Most recently, in December 1999, the Canada Industrial Relations Board (CIRB, formerly CLRB) imposed union recognition on Transx Ltd. after the company had refused to reinstate employees, spied on union activists, and “aimed at establishing grounds for the dismissal of union supporters.” While it stated a preference that certification be based upon a secret ballot, when the possibility of free choice had been “seriously compromised” by illegal practices, the board asserted its “duty” to certify the union despite a lack of evidence of majority support.[30] Not surprisingly, the national board’s strict policy on employer communications has frequently attracted the ire of employer representatives. One employer publication recently lambasted the “particularly restrictive” policies of the CIRB under which employers were “not permitted even to express a preference to remain union-free during an organizing campaign” (Levitt 1999). But, despite employers’ complaints, the national board, like its Ontario counterpart, has exhibited few signs of abandoning its tough stance on management opposition during representation campaigns.

The Legacy of Post-War Certification Policy

Post-war Canadian certification policy satisfied neither unions nor employers. Incensed by the high thresholds of support required for both election and card certifications, labour leaders attacked the “anti-union” bias of Canada’s certification procedures and advocated their reform along the lines of U.S. certification policy (Kidd 1960). Employers, in contrast, opposed the “chronic unreliability” of card certifications and complained that labour boards interpreted too broadly legal prohibitions against management interference with workers’ right to organize, thereby effectively suppressing their arguments against collective bargaining (Patterson 1968). But during the post-war years, when relatively few employers fought aggressively against unionization, most academic observers believed that Canada’s tough certification policy favoured management because Canadian unions were required to demonstrate “a much more solid foundation of worker support” to win certification than were U.S. unions (Woods 1962). However, in the context of open employer opposition to collective bargaining in the U.S. since the 1970s, few unionists doubted that Canada’s card certifications, expedited elections, anti-electioneering policies, and tough penalties against unlawful interference provided greater protection for workers’ free choice than did American labour policy.

Although the policies of national and provincial boards did not always favour labour, they established a consistent policy against aggressive electioneering by either management or unions and kept certification campaigns brief. Probably no more friendly to unions than their U.S. counterparts, Canadian employers have found their actions consistently constrained by a labour policy much less hospitable than U.S. law to aggressive management opposition to unionization. As a result of this strict electioneering policy, Canadian employers have seldom enlisted the services of anti-union consultants, so conspicuous in representation campaigns south of the border during the past three decades. Indeed, in 1980, an organization that had provided numerous referrals to anti-union consultants throughout the U.S. stated that it “knew of no knowledgeable, expert consultant in Canada. The idea of union avoidance is so new to Canadians, that no consultant (to our knowledge) has been operating in Canada.”[31] Canada’s post-war certification system—based on card certifications, high thresholds of employee support, strict regulation of employer resistance, and tough penalties against unlawful interference—has created an inhospitable environment for the aggressive campaigns advocated by anti-union consultants.

But Canada’s strict certification policy has not deterred entirely employer opposition to unionization. Canadian union density peaked at 37 per cent of the non-agricultural workforce in the early 1980s, and the participants at a government conference concluded that the country’s certification process was “not working as well as it might primarily because of employer attitudes and behaviour” (Bain 1981). Management opposition to unionization during certification campaigns has probably intensified during the past decade (Yates 2000), yet many provincial boards have proved reluctant to exercise their authority to certify unions if employer misconduct has rendered impossible employee free choice. And despite some notable (but fleeting) union victories at such venerable “union free” corporations as MacDonald’s, Starbucks, and Wal-Mart, management opposition is likely to prevent any significant expansion of collective bargaining into the private service economy in Canada for the foreseeable future (Adams 1993; Godard 2001).

Striker Replacements and State Regulation of Collective Bargaining

In Canada, the legal status of striker replacements has also developed in a very different direction than in the U.S. Since the 1970s, national and provincial labour boards have provided economic strikers with much greater legal protection, and replacement workers with considerably less security, than their American counterparts. As a result, employers’ recruitment of permanent replacements has rarely been a prominent feature of economic strikes north of the border during the past three decades. In high-profile strikes since the 1970s in which employers have recruited replacements—such as the Gainers strike in Alberta, the Fleck strike in Ontario, and the Canada Post strike in the federal jurisdiction—they have generally rehired strikers at the end of the disputes (Starkman 1993). The significance of the striker replacement issue, however, extends beyond its often-critical impact on the outcome of economic strikes. Rather, it is best considered as part of the larger question of the extent to which the state regulates the process of collective bargaining. While provincial laws explicitly limiting management’s right to recruit replacements are a feature of the past three decades, Canadian law-makers established the principle and practice of state regulation of the economic weapons of both employers and unions during the process of collective bargaining in the 1940s and 1950s.

The Policy Debate on State Regulation of Collective Bargaining

Although the political origins and policy objectives of PC 1003 were quite different from those of the NLRA, the extent to which the state should regulate the economic weapons of management and unions was central to the debates over both labour laws. In the 1940s, Canadian lawmakers rejected the NLRA’s “hands-off” approach to state regulation of the economic weapons of unions and management. Prior to the enactment of PC 1003, the Liberal government criticized the American system of free collective bargaining—the preferred choice of most union leaders, particularly those of the CCL—as an “insufficient safeguard” against the threat of industrial unrest, especially during the wartime emergency.[32] Instead, the government advocated extensive legal restrictions on unions’ strike weapon and a continuation of Canada’s long-established practices of compulsory investigation and mediation of industrial disputes (Fudge and Tucker 2001). While it incorporated several of the NLRA’s central provisions, PC 1003 omitted the American law’s protection of the right to strike and provided for a greater degree of state intervention in the actual process of collective bargaining. It restricted both labour’s right to strike and employers’ right to lock workers out, stated that collective agreements must include “no strike” and grievance arbitration provisions, and at the end of agreements, required that management and unions proceed through extensive conciliation procedures before resorting to economic force. Thus, under the post-war legal regime, the state not only established the institutions and procedures of collective bargaining (as was the case in the U.S.), but it also regulated “extensively” the process of bargaining and required that unions “earn” the right to strike.[33] Dissatisfied with PC 1003’s compromise position on state regulation of economic weapons, labour leaders attacked its “burdensome” constraints on unions’ freedoms, while management associations advocated even more extensive restrictions on economic strikes. If employers’ freedom to oppose unions during organizing campaigns were to be severely restricted, insisted the Chamber of Commerce, unions’ right to strike after certification “must be correspondingly limited.”[34]

While it reduced drastically the number of strikes called over union recognition (which were now illegal), PC 1003 failed to deliver industrial peace. In 1946 over 40 percent of Canadian workers participated in work stoppages, mostly over issues of wages, union security, or national bargaining (Logan 1956). Enacted shortly after the massive reconversion strike wave, the 1948 IRDIA and provincial labour laws modelled on it continued state regulation of the economic weapons of unions and management.[35] In the 1950s and 1960s, conservative provincial governments—such as those in BC, Alberta, and Manitoba—introduced additional legal constraints on unions’ freedoms, requiring that they conduct ballots or give notice prior to calling strikes (Carrothers 1965). The federal government also passed emergency legislation in an attempt to ensure the uninterrupted operation of the railways and several provincial governments restricted the ability of workers in “essential services” to strike (Starkman 1993). In response, labour leaders and their allies attacked the “excessive intrusiveness” of provincial and federal labour laws. When asked in 1960, “What do unions expect from governments in the realm of industrial relations?” one labour leader replied: “First of all, a minimum of interference” (Kidd 1960). And most academics agreed that Canada’s intrusive labour policies “tend to affect unions more adversely than they do employers” (Jamieson 1971).

Unions and employers continued to clash over the issue of state regulation of the economic weapons of unions and employers throughout the post-war decades. If labour laws were to limit the right to strike, unionists argued, they should also proscribe the use of replacement workers. Indeed, labour leaders “expected” that “a government which believes in collective bargaining” would enact legislation “forbidding an employer from attempting to open his plant by the hiring of strike-breakers” (Kidd 1960). While unions insisted that blanket anti-replacement legislation was a “legitimate objective,” certain academic experts believed that it would distort the “test of economic realities,” which was at the core of Canada’s system of collective bargaining. Rejecting the notion that the right to recruit replacements had provided employers with “unwarranted advantage” during contract negotiations, one legal scholar argued that “labour’s successes” under the existing system of collective bargaining had imposed “a heavy onus on union spokesmen who advocate legislation outlawing the hiring of replacements” (Arthurs 1967). The CMA and the Chamber of Commerce have also defended the notion of the strike as a test of the economic strength, but simultaneously advocated severe limitations on unions’ right to strike (Patterson 1968). But while national and provincial governments have continued and frequently expanded legal regulation of labour’s strike weapon through back-to-work legislation and other restrictive measures, particularly in the public sector, the policy debate on replacement workers had turned decisively against employers in most provinces by the early 1970s.

State Regulation of Striker Replacement Prior to the 1970s

The IRDIA and provincial laws initially made no direct reference to the legal status of striker replacements, and Canadian labour boards ruled on few cases involving replacement workers in the immediate post-war years. The use of striker replacements diminished significantly in the late 1940s and 1950s as formal collective bargaining relationships became more widespread and most firms abandoned their “more overt and provocative anti-union tactics” (Jamieson 1971). Even in high-profile stoppages where employers recruited replacements, such as the Asbestos strike in Quebec and the Brandon Meat Packers strike in Manitoba, they usually reinstated the strikers at the end of disputes. Nevertheless, the CMA reassured its members that nothing in Canadian law “prevents an employer from hiring replacements,” while in response to inquiries about how to unload existing unions, it advised that as a result of the recruitment of replacements, “many a union has been forced out following an unsuccessful strike.”[36] But in several important disputes in the 1940s and 1950s, provincial labour boards restricted employers’ right to replace economic strikers, thereby rejecting the CMA’s position on replacements. During a province-wide moulders’ strike in Nova Scotia in 1947, for example, the board ruled that under the provincial law, striking workers retained their employee status while participating in a lawful stoppage and thus could not be permanently replaced.[37]

Considerable uncertainty surrounded the legal status of striker replacements at both the provincial and federal levels for the next few decades. Although always more reluctant than their U.S. counterparts to interfere with labour boards, Canadian courts within individual provinces have sometimes overturned board decisions on the legal status of striker replacements. In the landmark Brandon Packers case (1960), for example, the Manitoba Supreme Court overruled the provincial board, deciding that economic strikers could vote in a decertification election, but their replacements could not.[38] Even the federal government seemed somewhat unsure about the status of economic strikers and their replacements. In response to an inquiry about the reinstatement rights of economic strikers in 1961, the Labour Department wrote: “The question of whether an employer may terminate the employment of an employee who is on strike, by a positive action on his part, has not been firmly settled.”[39]

The legal status of economic strikers and replacements was no clearer in Ontario. The Supreme Court of Canada appeared to rule that the OLRA allowed the permanent replacement of economic strikers in the celebrated 1962 Royal York Hotel case, Canadian Pacific Railway Co. v. Zambri. When the hotel union struck for higher wages after conciliation had failed to produce a new contract, management sacked the strikers and hired permanent replacements. The initial magistrate in Royal York ruled in favour of the employer, but the High Court overturned this decision and stated that dismissing workers for participating in a lawful strike constituted an unfair labour practice. Both the Court of Appeal and the Supreme Court of Canada subsequently upheld the High Court’s decision. Almost two decades after the introduction of compulsory collective bargaining, Royal York finally established that economic strikers participating in a legal stoppage retained their employee status and thus could not be fired.[40] But could they be permanently replaced?

The central issue in Royal York was whether or not an employer violated the law by sacking economic strikers for participating in a lawful stoppage; only one Supreme Court justice discussed employers’ right to hire permanent replacements in the absence of unfair practices. At the end of economic strikes, argued Justice Locke, employers were “not obliged to continue to employ their former employees.... [Employers] are at complete liberty ... to engage others to fill the places of the strikers.” Referring directly to the U.S. Supreme Court’s 1938 Mackay decision, which had established the right of U.S. employers to recruit permanent replacements, Locke declared: “That is the law in Canada also, in my opinion.”[41] Thus, in disputes not involving unfair practices, according to Locke, employers were required to reinstate striking workers only if positions were available—they were not obliged to sack replacements in order to create jobs for strikers. And in the years immediately following Royal York, most commentators concurred that “the employer has the right to hire replacements for his striking employees” (Arthurs 1967). The Royal York ruling did not finally settle the legal status of striker replacements in Canada, however, and during the past three decades, national and provincial governments and labour boards have restricted employers’ use of striker replacement through both legislation and jurisprudence.

State Regulation of Striker Replacements Since the 1970s

During the past three decades, the position of the national labour board on striker replacements has undergone a complete reversal. In the 1970s, the CLRB ruled, in several economic strikes, that employers could recruit permanent replacements providing they had committed no unfair labour practices, thus apparently accepting Justice Locke’s position on replacements.[42] However, in the 1980s and 1990s, the CLRB overturned its previous position and ruled that an employer’s refusal to rehire strikers at the end of a dispute constituted bad faith bargaining in several disputes involving employer demands for concessions to existing agreements. The CLRB ruled that an employer’s insistence on retaining replacements and its refusal to reinstate economic strikers at the end of a dispute, in which the union had acceded to its demands, constituted an unfair labour practice in General Aviation Service (1982).[43] In Eastern Provincial Airways (1983), the board rejected the notion that employers enjoyed an “unfettered right” to permanently replace economic strikers and again ruled that an employer’s failure to discharge replacements to create positions for strikers amounted to unfair bargaining.[44] In Royal Oak Mines (1993), the board ruled that, because the interests of replacement workers were “squarely opposed” to those of the permanent workforce, replacements were ineligible to vote in decertification elections, thereby capping a series of decisions on voting rights for replacements dating back to Brandon Packers (1960).[45] Thus, during the 1980s and 1990s, the CLRB adopted the position that the Canada labour code guaranteed limited reinstatement and voting rights to workers who had participated in failed economic strikes.

Striker replacements policy has also undergone fundamental change in several provinces during the past three decades and Ontario has seen greater upheaval in its law and jurisprudence on replacements than most other provinces. The provincial legislature amended the OLRA in 1970 to state that workers participating in a lawful strike could apply in writing to their employer to return to work within six months of the start of the dispute.[46] The OLRB immediately applied the new policy (and, indeed, extended it) in several strikes in which employers with recently certified unions had recruited striker replacements in an effort to avoid negotiating first contracts or resist union security demands. The OLRB ordered the reinstatement of economic strikers in Becker Milk (1978), a violent dispute in which the company hired a strike-breaking security firm, and stated that strikers enjoyed an “unequivocal right” to bump their replacements during this six month period.[47] In 1981, the OLRB ordered Fotomat to reinstate economic strikers after the company had hired replacements during a strike over first contract negotiations and assured them that they would retain their positions at the end of the strike. In the Fotomat dispute, the OLRB extended for a further seven months the period during which strikers could reapply for their positions because, it ascertained, the company’s unfair practices had prolonged the strike.[48] And after the longest strike in Ontario’s history, the OLRB ordered Shaw-Almex Industries to rehire economic strikers after finding that the company had discriminated against the strikers, interfered with the union, and bargained in bad faith. Facing certain defeat in the three-year long strike, the union had acquiesced to all of the company’s demands but insisted that it rehire the strikers. In response, Shaw-Almex withdrew its settlement offer and announced that the replacement workers would retain their positions. However, in a decision that, according to the Steelworkers’ union, “breathed new life into the right of strikers to return to work after a prolonged strike,” the OLRB ordered the company to rehire the strikers and pay retroactive wages.[49] Thus, like the national labour board, the OLRB has repeatedly defended the reinstatement rights of economic strikers and restricted employers’ use of replacements.

Ontario was not alone in enacting legislation restricting striker replacements. Since the 1970s, a majority of provinces have passed legislation severely limiting the replacement of economic strikers, just as U.S. employers were starting to recruit permanent replacements en masse. In the 1970s, Alberta, Manitoba, and Prince Edward Island also passed laws providing limited reinstatement rights for economic strikers, while the following decade Manitoba and PEI banned permanent replacements. And following several violent strikes involving both replacement workers and strike-breaking security firms in the 1970s and 1980s, Alberta, BC, Manitoba, New Brunswick and Ontario enacted laws proscribing the use of “professional strike-breakers” (Budd 1996).

Most provincial striker replacement laws have distinguished between temporary and permanent replacements, allowing employers’ use of the former but severely limiting their recruitment of the latter. But in Quebec (enacted in 1977, strengthened in 1983), Ontario (enacted in 1993, reversed in 1995), and British Columbia (1993), social democratic governments outlawed the recruitment of both temporary and permanent replacements. Moreover, several other provinces have considered enacting blanket anti-replacement legislation during the past decade. In 1994, Liberal governments in New Brunswick and Nova Scotia indicated that they were contemplating blanket anti-replacement legislation, but following vigorous employer campaigns against the proposals, both governments backed down.[50] In 1997, the Saskatchewan Labour Board issued an interim ruling stating that Pepsi had violated provincial law by recruiting replacements during a lockout, but subsequently reversed its decision after the NDP government made clear its reluctance to sponsor anti-replacement legislation.[51] Finally, in the late 1990s, the federal Liberal government created a new labour board, the Canada Industrial Relations Board (CIRB), and gave it the power to ban replacements if it deemed that the employer had recruited them with the purpose of undermining a union’s “representative capacity.” In 1998, the government formally amended the Canada Labour Code to codify this policy change in law, but the CIRB has yet to rule on the question of employer intent in a particular strike context.[52]

The Legacy of Post-war Regulation of Collective Bargaining

The question of state regulation of the economic weapons of unions and management has been central to recent debates over striker replacements. Certain academic commentators criticized bans on temporary replacements in Quebec, BC, and Ontario for interfering with “legitimate” management weapons and for tilting the balance of power in collective bargaining “too far in labour’s direction” (Weiler 1984; Gould 1993). Management representatives claimed that “intrusive regulation” had left these provincial laws out of step with those of their major competitors for capital investment and thus would result in a “mass exodus” of businesses from Ontario and BC to neighbouring provinces and to the U.S. The BC Business Council, for example, claimed that the province’s anti-replacement law “amounts to excessive state intervention” in the process of collective bargaining and warned that “capital and labour are mobile.... Canadian businesses cannot be restrained by regulatory disadvantages that are not faced by competitors.”[53]

Supporters of the “anti-scab” legislation, in contrast, have argued that the laws are entirely consistent with a central principle that was enshrined in the legislation of the 1940s, i.e., promoting peaceful labour relations through state regulation of the economic weapons of management and unions. Existing labour policy, they argued, already limited strikes to the period following the expiration of a collective agreement and after the failure of investigation and conciliation procedures, and thus the new laws simply “restored balance” to the brief period during which strikes were permitted (Hopkinson 1996). Thus, the tradition of state regulation of the process of collective bargaining, established in the 1940s and 1950s, has continued to influence federal and provincial policy on the use of striker replacements during the past three decades. We should, therefore, consider the striker replacement issue within the larger context of the debate over restrictions on the economic weapons of unions and management.

Just as Canada’s strict certification policy has failed to prohibit aggressive employer opposition during organizing campaigns, however, its pro-union striker replacement policy has not prevented employers from engaging in activities designed to undermine the effectiveness of economic strikes. In particular, employers have frequently recruited the services of strikebreaking security firms in their efforts to continue operating during economic strikes. Strikebreaking security firms have operated in Ontario since the 1960s and were well integrated into the fabric of labour-management relations in the province by the early 1980s. The prohibition of the use of professional strike-breakers in 1983 and the enactment of anti-replacement legislation a decade later created a much less hospitable environment for their activities, but security firms have enjoyed a boom in business since the repeal of the province’s anti-replacement law in November 1995. Firms and organizations that have hired strike-breaking firms in recent years include the Toronto Star, the Salvation Army, and several municipal governments. As a result, one national newspaper lamented recently: “Is the state of labour relations in Canada really so retrograde that even governments and left-leaning newspapers feel that they have to call in a modern-day Pinkerton’s? And whatever happened to the mediating role of government?” (Carlson 1999). Thus, while lawmakers have prohibited the permanent replacement of economic strikers and prevented employers from exploiting economic strikes as a means to unload existing unions, they have not stopped firms from using replacements to win concessions from existing agreements and to impair union security.


Crucial differences between Canadian and U.S. labour laws in recent years have contributed significantly to the divergent fortunes of the two labour movements, but Canadian unionists have not always viewed favourably federal and provincial labour policies. Indeed, during the past half century, organized labour’s perception of whether Canadian or U.S. law is more favourable to unionization has undergone a dramatic reversal. In the post-war decades, unions attacked the “excessive” requirements of Canadian certification policy and legal restrictions on the right to strike, while today the AFL-CIO looks to Canada for an example of how to reform U.S. labour law. By lowering voting quotas for trade union certification and restricting severely the use of replacement workers in the 1970s and 1980s, Canadian lawmakers reinforced the legitimacy of unionization at precisely time when many U.S. employers were launching a full-scale assault on the institutions of collective bargaining. Over the past three decades, Canadian labour boards have frequently certified unions on the basis of membership cards, held quick elections, restricted severely employer opposition during representation campaigns, imposed tough penalties on firms that interfere with employees’ free choice, and prohibited the permanent replacement of economic strikers.

But pro-union national and provincial policy innovations during the last three decades should not be viewed as a radical departure from post-war practices. Rather, as this analysis of the development of labour policy on certification and striker replacement demonstrates, the underlying principles and basic features of Canada’s system of state intervention in labour-management relations have remained remarkably stable since the 1950s, while the larger economic and political contexts in which the laws operate have changed dramatically. During the past three decades, a period of increased international competition, economic restructuring, and neo-liberalism, Canada’s “anti-union” labour policy—which has imposed stringent requirements for certification and restricted the right to strike, but also limited employer opposition during representation campaigns and prohibited permanent replacements—has protected workers against employer hostility better than has its U.S. counterpart. In the past decade, however, national and provincial governments have tolerated, rather than encouraged, collective bargaining, so whether or not Canada’s interventionist labour policy will continue to protect unions against market-driven anti-unionism remains to be seen.