Corps de l’article

What is union revival?

There is no widely accepted definition of union revival or the criteria that should be used to evaluate revival efforts. There are some sweeping prescriptions for change. For example, Lipsig-Mumme (1999: 4) defines union revival in the broadest possible terms as “the ability of a union or union movement to adapt and survive; to continually adjust its goals, its practices, its priorities, its alliances in relation to the working class and the economic and socio-political world in which it operates, so that it continues to defend workers’ interests.” Some writers see the essence of revival as nothing less than the transformation of unions into organizations devoted to promoting human rights and social justice and redressing the imbalance of power between workers and employers. Others see union revival in more immediate and less ambitious terms. Revival means more effective use of the law and of members, allies and financial resources, for example by exploring ways to organize outside of labour board procedures and among workers without a tradition of unionization; the use of information technologies to better communicate with members, unorganized workers and the public; political and organizing alliances with community organizations; sophisticated research on employers to gain bargaining and organizing advantages; and the education of members in the values of unionism and encouraging their participation in organizing and union governance (e.g., Bronfenbrenner et al. 1998; Mantsios 1998; Century Foundation Task Force 1999; Nissen 1999; Juravich 2000). Union revival may also mean changing employment relationships in which members assume a greater role, either by force of bargaining or the law, in workplace and top management decisions. Or revival can mean new union influence nationally over government decisions that affect workers’ income and job security, for example by imposing restrictions on international trade agreements (Bureau of National Affairs 1997; Newland 1999).

Claims to union revival may have external components (adjusting to relationships of power with employers, the state, the local community, coalition partners, the international community, and so forth) and internal ones (e.g., new union priorities and better use of resources). But whichever components are emphasized, there remains the inherent limitation that revival is equated with processes rather than outcomes. Thus, observers can point to a change in activity (e.g., greater membership participation in political canvassing, the formation of a coalition) or a union victory (a successful organizing campaign, a major strike that was settled mostly on a union’s terms), and claim revival, even if it does not produce significant and lasting changes in the fundamental dimensions of unionism, i.e., increased membership and union density (the portion of the workforce in unions). Furthermore, we have to be wary of dramatic assertions of union revival; they may be self-serving, intended by unions to reassure members and potential members that the union is energized and heading in the right direction.

We believe that revival is more than an awakening or a proclamation of new union programs and priorities, or a change in members’ attitudes and participation. Revival should have clear, measurable outcomes. It should show itself as significant and lasting gains in membership and density that can translate into greater bargaining strength and political influence. In this article we examine such outcomes and probable outcomes in Canada and the United States in terms of three related dimensions—organizing, collective bargaining, and political action. We begin by briefly examining the state of the unions in the two countries.

Comparative overview of Canadian and U.S. unionism

In the mid-1950s, union density stood at about one-third of the non-agricultural workforce in both Canada and the United States. Since then, the direction of unionism has diverged. U.S. union membership expanded up to 1975, but failed to keep pace with employment growth. As a result, union density dropped below 25 percent by the late 1970s and, by 1999, had plummeted to just below 14 percent. U.S. unions lost nearly 6-million members between 1975 and 1999 (a decline of over 25 percent). In contrast, union density in Canada experienced modest increases (ranging between 35 percent and 37 percent throughout most of the 1980s), before stabilizing at between 32 percent and 35 percent throughout the 1990s. Between 1980 and 1999, Canadian unions expanded by more than 600,000 members (an 18 percent increase), but these gains lagged employment growth. The magnitude of the changes in union membership and union density in Canada and the United States appear in Table 1.

Whereas union decline in the 1980s was prevalent among eighteen OECD countries, the United States experienced one of the steepest declines in union density and now has one of the lowest national density rates. Union density in Canada, on the other hand, has remained relatively stable, preserving its ranking as a middle-density country (Western 1997).

The divergence in union density between the United States and Canada has been attributed to three broad factors related to the ability of unions to recruit new members. First, both the levels of organizing activity and organizing success rates have been substantially higher in Canada than in the United States. To a considerable extent, this has been associated with differences in the legal environment. The American system of labour board certification of unions relies on elections among workers and typically results in protracted campaigns in which employers use union-avoidance tactics ranging from procedural delays to discrimination against union supporters. In contrast, certification in Canada normally is based on signed membership cards (and occasionally expedited elections), thereby minimizing the opportunity for employers to resist unions. As well, the incidence of employer unfair labour practices is substantially higher in the United States and legal remedies to combat employer interference are not as strong or as effective as those in Canada. We have argued elsewhere (Rose and Chaison 1996) that declining union density perpetuates itself through its adverse impact on organizing priorities and resources. In the aftermath of steep membership losses, many American unions had financial difficulties and were forced to focus on servicing existing members rather than embarking on organizing activity, with its usual high costs and low yields. Conversely, with a more favourable legal environment and a higher density rate, organizing has remained a high priority investment for Canadian unions.

Table 1

Union Membership and Density in the United States and Canada Selected Years (index 1956=1.00)

Union Membership and Density in the United States and Canada Selected Years (index 1956=1.00)
Sources: Canada: Workplace Information Directorate. Various years. Directory of Labour Organizations in Canada. Ottawa: Canada Government Publishing; Human Resources Development Canada. 2000. Workplace Gazette, Vol. 3, No. 3. United States: Labor Research Association. 2000b. “Union Trends and Data: Union Statistics.” www.laborresearch.org.

-> Voir la liste des tableaux

Second, higher unionization has also enabled Canadian unions to outperform their American counterparts in collective bargaining, particularly during the turbulent 1980s. Concession bargaining was far more prevalent in the United States where union membership losses and aggressive employer bargaining strategies put unions on the defensive. The inability of American unions to repel employer demands for concessions and secure collective bargaining gains did little to increase the attractiveness of unionism among nonunion workers (Rose and Chaison 1996).

Third, the affiliation between organized labour and the social democratic party, the New Democratic Party (NDP), has enhanced the political influence of Canadian unions (Rose and Chaison 1996). The institutionalization of the NDP within a highly federalized parliamentary system has resulted in the liberalization of collective bargaining laws (Bruce 1989). In contrast, union-political party linkages in the United States are less formal and unions have not been able to persuade Congress to reform the national collective bargaining law.

Efforts to explain differences in union density rates have also considered other factors, such as structural shifts in employment and public support for unions. Neither of these factors offers a satisfactory explanation for the divergence in U.S.-Canada density rates. Although Troy (2000) attributed the divergence in union density rates to Canada’s larger public sector, this fails to recognize that the Canadian public sector density rate is nearly double that of the United States and that private sector unionism has declined at a much slower rate in Canada (Rose and Chaison 1996). Further, structural differences in the economies and labour forces have been shown to account for only 15 percent of the U.S.-Canada unionization gap (Riddell 1993). This finding is broadly consistent with studies showing that structural changes are not responsible for much of the union density decline among OECD countries (Western 1997). Similarly, public opinion data offer little insight into this divergence in union density. The trends in the two countries are often ambiguous, sometimes contradictory and usually changing over time (Chaison and Rose 1991). Public approval of unions as institutions and of union leaders may fall, while public support for what unions do rises. Unions may be seen at one time as powerful but needed, at other times no longer powerful but still needed. They may be perceived as good for workers, but not necessarily as good for the economy or the public generally. Moreover, workforce surveys as distinct from public opinion polls, show a significant reservoir of support for unions, but such support often dissipates during organizing campaigns where there is stiff employer opposition (e.g., Freeman and Rogers 1999).

When comparing the U.S. and Canadian labour movements, we see a symbiotic relationship between union density and union effectiveness. Union organizing activity, collective bargaining performance and political influence can be regarded as determinants as well as consequences of union density levels (Rose and Chaison 1996: 100–101):

declining density in the United States has lowered union performance in organizing, bargaining and political activity and this brought about further declines in density. The Canadian union movement has avoided a cycle of decline; its continued growth and stability have enabled it to confront challenges in resourceful and energetic ways and density level changed little during the turbulent 1980s.

Analysing union density rates enables us to not only consider what unions do and how well they do it, but we can also reflect on the prospects for union revival. First, revival movements in the two countries start on a different basis; Canadian unions must attempt to expand beyond their state of relative stability, but U.S. unions must somehow turn around a severe and continuing decline in unionization. Second, the causes of low density rates, particularly in the United States, are the very issues that must be addressed in proposals for union revival—increasing organizing, regaining bargaining power, and changing labour laws to enable the first two to happen. Third, union density rates provide a measure to gauge the effectiveness of union revival efforts; unions must do more than win some high profile organizing drives; they must organize enough workers to offset attrition and raise net membership levels and union density.

Canadian and U.S. unions: prospects for revival

The Organizing Challenge Remains

Organizing is at the very heart of union revival and it is widely recognized as the highest priority of the Canadian and U.S. labour movements. Organizing success, measured as either overall membership gains or a few major victories, is frequently cited by the unions as a proof that they have turned a corner. Indeed, the simplest definition of a revived labour movement is one that grows and looks like a movement, rather than a group of organizations.

Canada. The stability of union density in Canada over the past three decades reflects the considerable organizing success of unions, most notably in the public sector. However, if Canadian unions are to grow and prosper, major initiatives will be required. The organizing challenges faced by Canadian unions are daunting. They include the need to make significant inroads into what are commonly referred to as the “harder to organize” segments of the economy, including (1) the service sector, notably private services, (2) small and medium-sized enterprises and (3) components of the labour force that are expanding at above average rates, e.g., knowledge workers in the private sector and contingent workers (Finlayson and McEwan 1998).

While some successful organizing drives have been heralded as breakthroughs (e.g., McDonalds and Starbucks Coffee), they have not expanded the frontiers of collective bargaining. Union density rates among purportedly harder to organize workers have remained stagnant. For example, there has been a very modest improvement in the union density rate for part-time workers, rising from 20 percent in 1987 to 22 percent in 2000 (Mainville and Olineck 1999; Akyeampong 2000). Despite the substantial rise in part-time employment, the density rate for part-time workers remains at about two-thirds the overall density rate. In addition, there has been no noticeable improvement in union density rates in private services. Density rates for service producing sectors (e.g., finance, insurance, real estate and leasing; accommodation and food; and professional, scientific and technical) have remained relatively constant at less than 10 percent for the past decade (Akyeampong 1998, 2000). These results not only reflect the organizing difficulty, but may also be indicative of tradeoffs among union priorities. According to a recent survey of Canadian unions, protecting the current level of members’ wages and benefits was more important than organizing. In addition, a higher priority was attached to organizing in the union’s traditional jurisdiction than in new areas of growth (Kumar, Murray and Schetagne 1998).

A recent survey of union organizing in Ontario (Yates 2000a, 2000b) indicates the difficulty of increasing density rates among harder-to-organize groups. It showed that while union efforts to organize nonstandard workers was not high, some success had been achieved in the private service sector. In private services, organizing effort increased, the organizing success rate was just below the provincial average and the total number of employees organized ranked second only to manufacturing. These gains, however, lagged total employment growth in private services, a trend consistent with the economy as a whole.

Although union membership growth occurs in various ways, e.g., through voluntary recognition by employers and employment expansion pursuant to union security clauses in collective agreements, new organizing through certification is recognized as the most important dimension of union expansion. It is also measurable. Recent evidence reveals the number of certifications granted in Canada declined between the mid-1970s and early 1990s (Martinello 1996). The trend undoubtedly has been influenced by early union success in organizing most of the public sector (where employment has recently declined) and the failure to make substantial inroads among smaller business establishments. There are both notable gaps and large annual variations in the data respecting the number of employees covered by new certifications, but there was no discernible trend in the number of new union recruits between 1976 and 1988. Thus, although the commitment to new organizing remains strong and the drop off in new organizing has been modest compared to the U.S. situation, organizing new recruits has lagged growth in the workforce. According to a recent report, the unions’ share of total employment growth (employees added to payrolls and self-employment) between 1998 and 1999 was about 4 percent (Little 1999).

In addition to the challenge of organizing the unorganized, unions have had to commit resources to retaining or recapturing members affected by restructuring and downsizing. In the public sector, privatization and contracting out have become more prevalent. When privatization involves the sale of an enterprise’s assets, union representation often is retained through the doctrine of successor rights (e.g., air traffic control). Other forms of privatization (i.e., those not protected by successor rights) and contracting out of public sector work to the private sector have required public sector unions to mount organizing campaigns in an attempt to recapture lost members (Thompson 1995). Similarly, the outsourcing of private sector union jobs has posed a major challenge. It was a contentious issue in the 1999 contract negotiations in the automobile industry (discussed below).

The organizing challenge has not been eased by labour law reforms facilitating the spread of collective bargaining. In recent years, there have been no new legislative initiatives to facilitate union organizing, even in jurisdictions where the NDP, the social-democratic party, is in power. Indeed, political shifts—the rise of neo-conservatism in Canada and decline of the NDP—has not only reduced the impetus to initiate labour law reforms, but it has led to the repeal, in some cases, of union-friendly legislation. This occurred in Ontario when the Conservatives were elected in 1995 and repealed the legislative initiatives of the NDP and introduced mandatory certification elections. The effect of legislative changes introduced by the NDP and the Conservatives on union organizing success was significant. The level and success of union organizing increased substantially and decertification declined in the wake of new NDP legislation whereas there was a significant decline in organizing attempts and success, and a rise in decertification following passage of the Conservatives’ legislation.

According to Martinello (2000), the above results demonstrate the importance of the political environment on organizing activity. First, the political regime alone—before any new legislation is passed—has significant effects on the levels of certification and decertification activity. Second, the labour legislation produced by the political regime also has significant effects on organizing success. These two conclusions can be combined to draw a third, more general conclusion, namely that government ideologies and initiatives have an impact on union growth and decline.

For there to be a major and sustained upturn in union organizing success, the following will have to occur. First, there will have to be renewed popular support for the NDP and a commitment to progressive labour laws by the NDP. According to Western (1997: 8), one of the factors critical to achieving union growth has been the manner in which “working-class parties have enlisted the power of the state to promote union organizing”. The introduction of labour law reforms would not only enhance the opportunity for new organizing initiatives, but limit employer opposition to unions. Although both the level and impact of employer opposition on union organizing performance has been negligible compared to the U.S. experience, there has been a rise in unfair labour practices—from 0.05 per certification in 1967 to 0.70 per certification in 1986 (Robinson 1994). Second, there would have to be major effort and commitment of resources to organizing by individual unions and union federations. Third, the labour movement will have to sell the attractiveness of unions to unorganized workers. In a recent survey, non-union workers were asked if they would vote for a union. Nearly one-third of the Canadian respondents said they would opt for a union (Lipset and Meltz 1998). Unions will have to tap into the latent demand for their services among non-union workers.

United States. In the United States, organizing also occurs in the absence of legislative initiatives that might simplify or expedite the union certification process, or impose greater penalties on employers who interfere with workers’ rights to unionize. But with a density rate that is already so low and nearly two decades of precipitous decline in both density and membership levels, the organizing scene is far gloomier than in Canada. Arguably, the situation is now beyond the influence of political regime and legislation. For years, the unions had followed a simple strategy of “reform labor law first, then organize” (Moberg 1998: 11). But the Dunlop Commission, charged by President Clinton with appraising the state of labour law, did not recommend any substantive reforms in its 1994 report (Commission on the Future of Worker-Management Relations 1994) and the election of a Republican Congress in that year ended any chances of incremental improvements in the law of organizing and bargaining.

A call to organizing is found in virtually all union and federation plans for union revival in the United States. The AFL-CIO leadership has exhorted affiliated unions to “rebuild the culture of organizing” (John Sweeney quoted in Burkins 1997: A20), to aim for a target of three percent annual growth, and to allocate thirty percent of their budgets for organizing (Cohn 1997). Union strategy committees invariably stress the need for a transformation into an “organizing union” by increasing the resources for organizing, targeting new groups of workers (e.g., part-timers) for recruitment, and exploring new approaches to organizing (e.g., joining with other unions in community—and industry-based campaigns).

Looking beyond all the excitement surrounding new campaigns and the unions’ recognition that something must be done soon to jump-start organizing, the organizing record is less than impressive. Few unions have answered the call to devote greater resources to organizing and in 1999 less than half of union locals had organizing programs (Greenhouse 1999a; Karr 1999). Moreover, the emphasis on renewed organizing deals with process rather than outcome. Over the past five years there have been a number of policies and programs aimed at rejuvenating organizing including “Union Summer” programs that recruit young and older persons to assist organizers; new organizing alliances between unionists, intellectuals and the clergy; and experimentation with organizing outside the law by convincing employers to grant union recognition based on evidence of majority union membership rather than labour board certification elections. These have attracted a great deal of attention and may suggest to some that union revival is underway. But they have not resolved the fundamental problem of organizing enough workers to significantly change the membership and density trends.

Masters (1997) estimates that most unions devote 2–4 percent of their budgets to organizing; to reach the target of one-third of these budgets they would need to allocate roughly $6.6 billion. His analysis also shows an annual cost of growth that must be intimidating to union officers and members, ranging from “more than $1.6 billion to maintain steady-state membership at $500-per-recruit price tag, to more than $22 billion to achieve 10-percent annual growth at $1800-plus per head.... To the extent any of these cost estimates are on target, funding such an organizing program within existing budgeting conditions is unrealistic” (Masters 1997: 182). Because of high annual attrition in membership (mostly due to shrinking employment in union firms), unions would have required over 12 million new members for the period 1996–2000 to raise union density from 11 percent to 18 percent (Masters 1997).[1]

Languishing organizing efforts have resulted in low-density rates and membership concentration in the public sector and the traditional and declining strongholds of unionism in the private sector, rather than recently expanding sectors. Between 1987 and 1999, union density rates for part-time workers rose only slightly (from 6.5 to 6.7 percent) and declined for workers in service industries (from 6.3 to 5.5 percent) (U.S. Department of Labor 1989; 2000). A recent AFL-CIO report found that from 1984 to 1997, the 30 fastest growing sectors of the economy (e.g., finance, retail trade, hotels) added 26 million jobs, but only about 5 percent of the workers in those industries joined unions. Furthermore, in the eight industries with the greatest job losses (e.g. steel, autos) about 80 percent of the 2.1 million jobs lost belonged to union members. These losses were due mostly to competition from imports or from employers who relocated overseas to take advantage of lower labour costs (Greenhouse 1999d). In other words, underlying low density rates in the private sector and making the organizing task all the more difficult is the fact that “job growth is fastest in industries where unions are weakest while job losses are greatest in industries where unions are strongest” (Greenhouse 1999d: A19).

The demand for unionization in the United States appears quite stable. Survey evidence reveals about one-third of non-union workers expressed a desire for union membership over the past three decades (Freeman 1989; Freeman and Rogers 1999). Unlike Canada, the predominant impediment to capturing the demand for unions is management opposition. According to Freeman and Rogers (1999: 89): “Looking upon unionization as a good produced by the market, the United States is producing too little.... As far as the evidence goes, the main reason that these workers [workers who want unions] are not unionized is that the managements of their firms do not want them to be represented by unions.”

Summary. Although there are significant differences in the level and success of union organizing in Canada and the United States, unionization has not kept pace with employment growth. In both countries, union revival will require legal changes to facilitate organizing (albeit the need is more acute in the United States). As well, unions will have to make organizing their highest priority and commit more resources to it.

Regulating Labour Markets: The Limits of Collective Bargaining

Unions rely extensively on collective bargaining to regulate labour markets. Although differences have been found in the bargaining performance of Canadian and U.S. unions, the collective bargaining model limits the ability of unions to retain and expand membership.

Canada. The most obvious difficulty is structural. The structure of organizing and collective bargaining is enterprise-based and thus highly decentralized. This means there are many unions and employers engaged in autonomous negotiations. As a result, there is an absence of centralized and coordinated bargaining by peak employer and union organizations aimed at establishing wages and working conditions across labour markets. Accordingly, unlike many European countries, there is no centralized wage-setting system that enhances union influence. Western (1997) observed a close association between centralized labour market institutions and high levels of union organization. Centralized labour markets are advantageous to union growth because they lower employer resistance to unions, permit union confederations to coordinate union organizing activity and give union confederations a voice in developing economic policy.[2] The absence of centralized bargaining structures has been the principle barrier to expanding union influence, membership and density in Canada.

In many countries, globalization has contributed to the erosion of centralized bargaining systems and hence union influence. Although decentralization has also occurred in a number of industries in Canada, its cumulative effect has been modest given the preponderance of enterprise-level bargaining. It would appear that globalization and free trade have made it increasingly difficult for unions to take wages out of competition. The union-nonunion wage differential is estimated to have shrunk from approximately 25 percent in the late 1970s to about 8 percent in 1997 (Renaud 1997; Gunderson, Hyatt and Riddell 1999). Changes in the economic and political context have increased employer bargaining power and resulted in management aggressively pursing cost reduction strategies. As a result, union bargaining goals have become more defensive. A recent survey found that protecting current wages and benefits was the most important union bargaining priority by a wide margin (91 percent of the respondents). Reflecting increased concerns about job security, the next highest priority was layoff protection (64 percent of the respondents) (Kumar, Murray and Schetagne 1998).

For the most part, Canadian unionism is characterized by strong workplace organization and the absence of macro-corporatist institutions. A possible labour market strategy for unions would be pursue legislation requiring sectoral collective bargaining as practiced in the construction industry. Under this approach, newly organized firms would automatically be incorporated into multi-employer, sectoral bargaining units. This would facilitate broader-based bargaining and master collective agreements, and create economies of scale in the negotiation and administration of collective agreements. While such an approach would facilitate union expansion, it would face some formidable hurdles, including persuading government to adopt such a measure and overcoming union fragmentation.

From a comparative perspective, another limitation has been that Canadian unions do not exercise control over the market position of the jobless through union-run unemployment funds. The high union density found in countries with union-run unemployment funds (e.g., Sweden and Belgium) reflects the ability of unions to attract and organize workers with high unemployment risks (Western 1997). For Canadian unions, collective bargaining provides some protection against job loss (layoff and recall decisions are influenced by seniority) and offers benefits to members who become unemployed. Even so, unions have not significantly limited management discretion to reduce the size of the workforce or otherwise reorganize work that undercuts their organizational strength. For example, they have not been successful in limiting the use of contingent (part-time and temporary) workers, even though contingent employment continues to expand at the expense of full-time standard employment. Similarly, unions have not succeeded in imposing major limitations on, let alone halting, the ability of enterprises to outsource work, which results in the erosion of union jobs and the expansion of employment in small—and medium-size enterprises that are typically nonunion (Finlayson and McEwan 1998). Despite some gains, just over half of the major collective agreements in 1999 impose some restrictions on contracting out (e.g., if it results in layoffs). Only a small minority provides an outright ban on contracting out (Giles and Starkman 2000).

The outsourcing of parts and modular production (where suppliers assemble entire components of a vehicle) have been contentious issues in the automobile industry and have led to the erosion of union jobs and membership. Although the Canadian Autoworkers Union (CAW) has negotiated limits on outsourcing (e.g., in 1996 General Motors agreed there would be no out-sourcing of major operations during life of the collective agreement), the union has had considerable difficulty organizing auto parts manufacturers. Estimates indicate the proportion of organized workers among the major auto producers is 90 percent compared to only 45 percent in the independent auto parts industry (Kumar 1999). Accordingly, in the 1999 bargaining round, the CAW demanded that the Big Three manufacturers issue a “neutrality letter” to Magna International Inc., a major auto parts manufacturer with 55 plants in Canada, tens of thousands of employees and not a single collective agreement with the CAW. The union succeeded in obtaining a provision designed to get the major automobile manufacturers to use their influence to persuade Magna to adopt a responsible labour relations posture and to respect the right of workers “to decide whether or not to join a union in an atmosphere free of intimidation, interference or risk of reprisal” (Keenan 1999: B9). While the neutrality letter may prove to be largely symbolic, it underscores the priority attached to recapturing lost union jobs.

Unions have achieved some gains that cushion the economic impact of job displacement, including severance pay plans, relocation and retraining opportunities, and supplementary unemployment insurance or SUB (an allowance topping up publicly-funded unemployment payments). Despite the increased emphasis on job security issues in collective bargaining, there have been no widespread gains in these areas of security (Giles and Starkman 2000). In addition, few unions aside from those in the construction industry have established referral systems or placement agencies (“Union Finds New Jobs” 1999). Such an approach might assist displaced members find employment and maintain union commitment. Moreover, even though there is nothing comparable to the union-run unemployment schemes found in Europe, unions could expand their role in assisting the unemployed. At its 1999 convention, the Canadian Labour Congress recognized the potential for labour councils to be identified as advocacy centres on behalf all workers requiring assistance with work-related problems such as unemployment insurance and workers’ compensation.

Outside the realm of collective bargaining, unions have used their financial resources to maintain or expand the employment of union members. For example, in the construction industry, unions have invested pension funds in building projects to create jobs for members (Menter 1999). Additionally, various labour-sponsored funds (investment funds offering tax breaks to investors) have been approved by governments to maintain and create permanent jobs. Another institutional arrangement, which is exclusive to Quebec, is the decree system. This provides for the extension of collective agreement provisions to nonunion employees in a region or sector and has the effect of eliminating “competition over wages and working conditions among firms operating within the same industry” (Déom and Boivin 2000: 489). Again, while these institutional arrangements are important, their overall economic impact is relatively modest.

United States. Unions in the United States engage in a bargaining process similar in form and structure to their Canadian counterparts and consequently the prospects for reviving collective bargaining appear limited. Growth is limited by decentralized bargaining, the absence of union controls over unemployment benefits, and the employers’ reliance on out-sourcing. In 1999, labour pointed to its success in negotiations in the automobile industry and claimed a revival in collective bargaining. Most obvious was the contrast with the past and the magnitude of gains. Negotiations between the United Autoworkers and the automobile companies were far different from the concession bargaining of the 1980s, which broke with a long-standing wage pattern and resulted in wage and benefits freezes or cuts, and the loosening of work rules. The latest negotiations reinstituted the earlier wage formula (inflation plus three percent), added a signing bonus, imposed restrictions on layoffs and increased pensions. Rather than denoting revival, the auto settlements demonstrate what a union can achieve in negotiations with employers experiencing rising sales and profits and who can ill afford a strike. Moreover, the auto companies could absorb the costs of the agreement because of expanding production and the adverse impact of foreign currency values on their competitors’ prices. Underlying the agreement, however, is the spectre of global competition. It allows for manufacturers to reduce the size of their workforces within agreed-upon limits and improve pensions to entice retirements. Thus, companies can be leaner, more productive and more profitable in the world market (Greenhouse 1999c; Franklin 1999).

The auto settlement will not bring about general union revival in negotiations because union density is so low in other industries. One observer commented:

Corporations, facing intense competitive pressures from globalization, have become tougher about agreeing to wage increases pushed by unions. Corporate America has changed its way since the 1950’s and 1960’s, when many companies rolled over in the face of union wage demands, knowing that they could always pass on added costs to the consumer. Back then, companies hardly had to worry about competition from imports.

Greenhouse 1999c: C8

Because of low density, unions can expect stiffer resistance from employers outside the automobile industry. Moreover, if strikes occur, employers will be more willing to use striker replacements (Greenhouse 1999c).

Surveys reveal the vulnerability of unions in negotiations. Cutcher-Gershenfeld, Kochan and Wells (1998: 30) questioned union and management negotiators and found that “management threats regarding replacement workers and plant closings are now a key part of the collective bargaining landscape”. In particular, first-contract negotiations often involve employer threats to use replacement workers or close the plant, and almost a quarter of first-contract negotiations did not produce an agreement. A survey of employer bargaining objectives carried out by the Bureau of National Affairs (BNA) found that among those employers to negotiate in 1999, large portions said they were likely (20 percent) or somewhat likely (21 percent) to find replacement workers if there was a strike (Bureau of National Affairs 1998). This may be easier to threaten than to do given that the present low unemployment rate makes it difficult to recruit skilled workers as striker replacements and strikes threaten rising company profits (Verespej 1999).

The collective bargaining record is mixed and does not provide compelling evidence of union revival. The BNA survey for the first half of the 1999 shows small wage increases in collective bargaining settlements (Bureau of National Affairs 1999b). Moreover, the union-nonunion wage differential has narrowed considerably from a peak of 31 percent in 1978 to 20 percent in 1999 (Bureau of National Affairs 2000). Union critics maintain the narrowing gap indicates that unions have been unable to protect members’ wages from intense global competition—conditions under which unionized employers have difficulty passing on wage increases to consumers (Deavers 1998).

Job security has remained a key issue in negotiations because of recent downsizing and mergers. To address job security issues, unions are adopting a strategy of requiring companies to invest capital to enhance the long-term stability of their operations. This approach has been most prevalent in automobile and steel industry negotiations and often entails the union agreeing to longer-term contracts in return for investment commitments from companies. Employers also seek changes in work rules and increased productivity in return for investment promises (Bureau of National Affairs 1999a).

In the broadest sense—beyond any particular strike or settlement—bargaining in the United States will not lead to union revival for much the same reason as it will not in Canada. It remains highly decentralized (there are over 100,000 agreements in force) (Cutcher-Gershenfeld, Kochan and Wells 1998) and it occurs primarily at the plant and company levels. Its autonomous character reduces the ability of unions to coordinate national activity and leaves unions vulnerable to employer relocation within and outside the United States (Towers 1997). Individual negotiations, such as those in autos, might seem to be trend-setting victories but they signify the ability of unions to use prosperity to their bargaining advantage rather than deal directly with foreign and domestic nonunion sectors. Despite some notable reversals of past trends, bargaining remains primarily defensive, seeking to restore past patterns and protect jobs, but structurally incapable of responding effectively to global competition and the huge domestic nonunion sector (86 percent of the workforce).

Summary. Collective bargaining in both countries is fragmented, i.e., decentralized and uncoordinated. As a result, organizing and collective bargaining activity occurs on a piecemeal basis, thereby limiting opportunities for union acceptance and expansion. Although unions in Canada may be perceived as more legitimate actors in the economy than their counterparts in the U.S., unlike their contemporaries in Europe, they have not been fully integrated into the institutional framework. The prospects for increased participation and integration, which would likely facilitate union expansion, do not appear favourable at this time.

Unions and Political Influence

In both Canada and the United States, labour unions are important political players, capable of shaping to varied degrees the policies of major parties and affecting election outcomes. But their power is transitory and diffused; it leads to protection of workers in general, but falls short of bringing about the changes needed to revive the unions as institutions.

Canada. Unions and politics in Canada are best understood in terms of the role of social democratic parties. Internationally, high levels of unio-nization have been associated with the success of social democratic parties. In particular, social democratic governments have typically enacted legis-lation to foster union organizing, allowed unions to participate in shaping economic policy and developed generous welfare states (Western 1997). Although the Canadian NDP has never achieved the political status of social democratic parties in many European countries (e.g., it has never held power federally), its influence both as an opposition party and as the governing party in several provinces contributed to the stability of union density during the 1980s when union decline was evident world-wide (Bruce 1989).

In other industrialized countries, economic recession, global compe-tition and economic uncertainty have contributed to the electoral failure of social democratic parties. The NDP is no exception. It lost party status in the wake of the 1993 federal election. Although it increased its total seats in Parliament in the 1997 election, public support for the NDP has plummeted. Public opinion polls show support for the NDP peaked around 40 percent in the mid-1980s, but fell to 11 percent in the 1997 federal election (Lawton 1999). In the 2000 federal election, the NDP lost 6 of its 19 seats, barely retained party status and attracted only 8.5 percent of the popular vote. At the provincial level, the NDP was swept aside in the 1995 Ontario election and failed to win enough seats in the 1999 election to maintain party status. In other 1999 provincial races, the NDP failed to hold their leads in public opinion polls in Saskatchewan and Nova Scotia. In Saskatchewan, the NDP Government was reduced to minority government status and in Nova Scotia, the NDP was defeated by the Conservatives. Meanwhile, the unpopular and imploding NDP Government in British Columbia is unlikely to survive the next election (Urquhart 1999). The only bright spot on the 1999 election scene was the NDP victory in Manitoba. The NDP premiers in both Manitoba and Saskatchewan have moved toward the political centre by adopting elements of Tony Blair’s “Third Way” as their theme (i.e., stressing fiscal prudence). In Saskatchewan, this shift has sparked labour disenchantment with what it sees as the NDP’s abandonment of social democratic principles, its fiscal conservatism and its use of back-to-work legislation to curtail legal strikes (O’Hanlon 1999).

Conversely, the relationship between social democratic parties and organized labour remains strong in Quebec. The Parti québécois (PQ) has won four out of last six elections and has governed continuously since 1994. The political influence of Quebec unions is not based on political affiliation, but derives from a neo-corporatist ideology, centralized collective dealing and support for separatism. Although tensions exist between the PQ and Quebec unions, the PQ maintains close ties with labour and has involved the union movement in various tripartite arrangements even in the face of stiff employer opposition. These ties, along with union support for the Bloc québécois (BQ) at the federal level, have been directed at maintaining sup-port for separatism as well as social democracy (Déom and Boivin 2000).

More broadly, the economic crisis of the 1990s has sparked a fissure in relations between organized labour and the NDP. Initially, it involved disputes with public sector unions over wage restraint policies and legislative restrictions on collective bargaining. It has spread to philosophical questions involving social democratic principles and electoral strategy. At the NDP’s 1999 national convention, the party endorsed a policy of fiscal responsibility based on balanced budgets and tax cuts. Labour leaders strongly criticized this initiative and decried the party’s drift toward the political centre (P. Adams 1999) The poor showing of the NDP in the 2000 federal election has prompted the Canadian Labour Congress to undertake a review of its commitment to the NDP.

Even though the NDP forms the government in three of the ten provinces, questions remain about the political fortunes of the party. Its status, as well as labour’s political influence, appear to have ebbed. While labour continues to play an important role in the NDP in terms of financial and election support, rank and file support for the party is weak as only 17 percent of union households voted NDP in the 1997 federal election (Laghi 1999). Previous voter surveys indicate that approximately 20 percent of union members vote NDP in federal elections, about twice the level of support received from nonunion members. Whereas union members belonging to affiliated local unions provide greater support for the NDP (32 percent) than non-NDP affiliated union members (17.5 percent), local union affiliation with the NDP has been consistently low and declined from its peak of 14.6 percent in 1963 to 7.3 percent in 1984 (Archer 1990). Perhaps more significantly, union members provide greater support for the Liberal party than for the NDP and this preference holds for those in NDP-affiliated local unions. Thus, while it is evident that affiliation shapes union member voting preferences, “the NDP’s lack of electoral success can be attributed in part to the inability of union locals to either deliver more votes to the party or to stimulate higher rates of affiliation with the party” (Archer 1990: 63).

It appears that recent efforts to enhance the electoral prospects of the NDP are rooted in broadening its base. Thus, even if there is a revival of the electoral fortunes of the NDP, the influence of labour over policy issues may be less prominent or diluted. Given the symbiotic relationship between social democratic parties and unionism, i.e., social democratic parties assist union growth and strong unions enhance the electoral success of social democratic parties, the prospects for union revival in Canada will be hampered by low public support for the NDP and the party’s drift toward the political centre. While some critics decry efforts to create a “new-style” social democrat, others point to the shift in the blue-collar vote from the NDP to the Conservative Party. In Ontario, for example, issues such as lower taxes and welfare reform are popular with blue-collar voters, suggesting that the NDP has not gone far enough to the political right. The dilemma for the NDP is whether to attempt to recapture the blue-collar vote or to write it off in favour of focussing on social activists, visible minorities, academics and other professional groups (Urquhart 1999).

The near universal support and commitment of union leaders to the NDP has also wavered recently. The rise of neo-conservatism and the adoption of market-driven solutions to economic ills have further alienated organized labour. In this context, unions have been largely ignored, forgotten or portrayed as a special interest group. With the declining electoral success of the NDP, organized labour has pursued alternative strategies, albeit largely defensive ones. A recent example was “strategic voting” in the 1999 Ontario election. A coalition of unions decided the way to defeat the antagonistic, right-wing Conservative government was to support opposition candidates with the best chance of winning (typically Liberals rather than NDP candidates) in key ridings or election districts. This represented a major shift away from the usual effort to elect a labour-friendly government to a pragmatic attempt to defeat the party most hostile to unions. Whereas some success was achieved in the targeted ridings, the strategy failed to dislodge the Conservatives. Equally significant, the downside to strategic voting is it undermined the already modest “labour vote.” Indeed, the popular vote for the Ontario NDP in the 1999 election was a 62-year low (Urquhart 1999).

Whereas labour is capable of providing vast resources to support the NDP, it has not been able to deliver the union vote. Securing unified support for the NDP is constrained by the fragmented organizational structure of the labour movement, social and political differences among unions (e.g., regional, national-international, and private-public) and political cleavages within Canada. In contrast, Quebec independence is a unifying theme for the PQ and Quebec unions. Thus a distinguishing feature of the NDP-labour relationship is the absence of a “unified and integrated approach to politics similar to that which exists between the unions and social democratic parties in other democracies where unions are highly integrated with the social democratic party” (Archer 1995: 36).

One positive development has been the labour’s commitment to social unionism and the establishment of coalitions with non-labour groups (e.g., community, religious and environmental groups). These efforts, which have suffered some notable setbacks (e.g., the passage of Canada-US Free Trade Agreement (1989) and NAFTA (1992) despite coalition opposition), will have to be intensified and sustained if labour is to help shape social and economic priorities. One step in that direction has been the involvement of the Canadian Labour Congress (and the AFL-CIO) in promoting high labour standards under a labour side agreement to NAFTA known as the North American Agreement on Labour Cooperation (R. Adams 1999). More broadly, labour’s involvement with social partners to improve the social well being of Canadian society should help dispel the myth that labour is a narrow special interest group. In addition, it can enlist community support for collective bargaining initiatives and organizing campaigns. This has been evidenced in campaigns opposing the restructuring of the postal system and in efforts to improve employment conditions for homeworkers in the garment industry. These campaigns highlight the potential to turn labour issues into social justice issues. Although these innovative strategies are still in their infancy, their continued development would offer a ray of hope for union revival in politics (Tufts 1998).

United States. In the United States, unions assign a major role to political activity in their program for union revival, yet remain uncertain about how to best use their political influence in the absence of a distinctively social-democratic party. They have been and will remain important players in national elections and some state and city elections, but their power is often exaggerated by their opponents and themselves. Furthermore, the limited and largely defensive objectives of labour’s political strategy, even if skilfully defined and energetically pursued, will not bring about union revival.

John Sweeney, the AFL-CIO President, recognized the interrelationships between organizing, bargaining and political activity in a labour movement with low-density rates:

There are two keys to our success or failure in rebuilding this labor movement—one is organizing and one is politics. We cannot succeed at rebuilding our membership base without winning in politics, and we cannot win in politics without substantially increasing our numbers.

quoted in Verespej 1999: 59

But unions’ political action remains largely defensive. It aims to restore what was lost during previous administrations (e.g., increasing the budgets of the labour board), deter opposition attempts to weaken collective bargaining (e.g., blocking legislation that would have permitted employer-dominated unions), and reduce threats to members’ job security (e.g., restricting foreign competition). Its groundbreaking campaigns, often undertaken through coalitions, deal with general workers’ concerns (e.g., health care reform and minimum wages).

It has also been successful in pressuring the Clinton Administration to negotiate a trade agreement with Jordan that mandates compliance with international labour norms. The direct inclusion of labour issues in a trade pact rather than as a side agreement has been a major objective of the unions since 1992 when they were defeated in their campaign against the North American Free Trade Agreement. Unions believe that tying new trade opportunities to respect for labour standards would discourage employers in the United States from relocating to take advantage of poor labour conditions and the inability of workers to unionize. The AFL-CIO was involved in drafting the labour sections of the agreement (Kahn 2000). This recent success, however, is in sharp contrast to labour’s major defeat when Congress passed legislation allowing permanent normal trade relations with China, despite intense opposition (Burkins 2000).

Trade pacts can lead to membership losses, which have to be offset by accelerated union organizing. The changes in the laws that would facilitate greater organizing are highly unlikely, at least during the next national administration. This is certainly not because unions are politically impotent. Just the opposite, unions showed impressive strength in the 2000 national elections and their members remain the Democrat’s most powerful constituency and the source of many of the party’s campaign workers (estimated to be 1,300 full-time organizers and 100,000 rank-and-file volunteers in the field during the last election) (Dine 2000). They registered 2.3 million new union household voters and made 8-million phone calls to their households (Labor Research Association 2000a). Fifty-nine percent of the vote from union households went to Al Gore, the same as Bill Clinton four years earlier, despite the former being far less popular among voters (Connelly 2000). Gore was the overwhelming choice over Bush in heavily unionized and populous states (e.g., California, New York and Pennsylvania) and union support led Democrats to victory in several congressional seats, including a closely contested Senate seat in Michigan (Labor Research Association 2000b).

Regardless of the unions’ power, there is little chance of passage of significant labour law reform. First, specific changes in the law related to organizing were never emphasized by major national candidates because of they were concerned about antagonizing business supporters. Only Bill Bradley, who was defeated in the Democratic presidential primary, appeared receptive to union certification based on card checks (Harwood 1999). Second, even if some in Congress were to support major changes in organizing, it is highly unlikely that bill could be passed because of the narrow margins of party control of the presidency and congress.

As substantial as union political influence has been at times, it has not produced the reforms in certification needed for a surge in organizing. Nor can it insulate the unions from the forces of the global product and labour markets. The unions’ approach to global markets has been to make labour issues an integral part of the global trade agenda under the premise that “unless globalization works for working people it’s not working” (Suzman 1999: 14). The unions were the main force behind the broad coalition that demonstrated against the World Trade Organization at its Seattle meetings in December 1999. As a clear sign of their political clout, the unions succeeded in having President Clinton propose their major objective—the linking of trade agreements with sanctions against nations that violate labour standards (e.g., restrictions on child labour, protection of workers’ rights to organize). Nevertheless, labour still had to fight the Clinton administration on other trade fronts, including the continuing attempt to extend the North America Free Trade Agreement to other countries, and overcome the major defeat it suffered when the House passed a bill allowing a permanent normal trade relationship with China despite intense union opposition (Greenhouse 1999b; Greenhouse and Kahn 1999; Burkins 2000). However, even if all labour’s efforts had succeeded, they still would not have stemmed the membership loses that must be replenished each year by new organizing.

Summary. Changes in the political context have reduced the political leverage of organized labour. In Canada, given the growing antipathy of most incumbent governments toward unions (and occasional hostility toward public sector unions), the divisions within the NDP and the waning political influence of unions generally, and the tenuous support for the NDP among union members, the political climate is not likely to be supportive of union revival. In the United States, the absence of a social democratic party and the fact that the realization of major labour law reforms are not on either political party’s agenda means that organizing will remain difficult and probably prohibitively expensive for labour.

Conclusions

The evidence leads us to pessimistic conclusions about the capacity of Canadian and U.S. unions to successfully pursue programs of sustained revival. In essence, union revival in both countries depends on high levels of organizing activity and success. The organizing task that lies ahead is far greater for U.S. unions than Canadian ones because of the significant differences in union density. But both labour movements lack the institutional frameworks and public policies that have been shown elsewhere (e.g., Scandinavia) to extend membership and help cushion unions from global product and labour markets (Western 1997). With low-density rates and severe competition from foreign and domestic nonunion sectors, organizing becomes an all-consuming task. Major expenditures of funds and staff time and effort are needed on both sides of the border merely to forestall membership decline. Yet, the primary mode of organizing—the union certification process—does not lend itself to quick and major membership gains.

The chances of more union-friendly organizing procedures in both countries, though particularly in the United States where the organizing dilemma is more pressing, is highly unlikely. Political action without a strong social democratic party seems incapable of addressing the inherent difficulty of union growth and its threat to union institutional stability. Political successes, often the result of coalition activities and rank-and-file mobilization, deal mostly with the workplace rights of all workers (e.g., safety standards, minimum wages). Finally, collective bargaining is so highly decentralized in both countries that it precludes the national coordinated action needed to protect union members’ jobs in the long-term and to expand beyond the most heavily unionized industries into growth sectors of the economy.

Does the pendulum not swing both ways? Have unions in Canada and the United States not experienced hard times before, for example during the bleak 1920s, only to rebound in dramatic fashion? This is not an uncommon view. For example, at a 1982 conference of industrial relations scholars and practitioners in the United States, some saw union revival generated by the conditions of low density: “once unions declined to the point of having little impact, discontent [among workers] will rise. Exit will replace voice, and the social and personnel problems now handled by unions will no longer be resolved. Union membership and activity would ultimately revive” (“The Future of Industrial Relations: A Conference Report” 1983: 128). Ten years later, an American labour historian wrote of a probable return to the mass union organizing of the 1930s: “What made trade unionism compelling to American workers in the past—and is likely to do so in the future—was its job conscious capacity to link itself with their aspiration for industrial justice” (Brody 1992: 40). When an economic crisis is reached, the argument goes, the stage is set for a union rebound because workers will return to unions to achieve fair treatment at the workplace and in the firm.

We do not believe in the inevitability of union rebound in either Canada or the United States. In both countries, employers are now much more sophisticated in fighting union organizing and finding ways to satisfy workers’ needs (psychological as well as financial) that might otherwise be fulfilled by union representation. Globalization leaves unions less effective than ever before in protecting members from low-wage, non-union workers elsewhere. Large and growing segments of the labour force either have jobs in industries or occupations without traditions of unionism (e.g., technical and service workers in the computer-related and biomedical industries) or are employed under arrangements considered ill-suited for collective bargaining—what Newland (1999: 56) described as “part-time, short-time, sporadic, contractual, or home-based”. Finally, workers increasingly receive or expect to receive industrial justice (e.g., protections against discrimination, layoffs or wrongful discharge) through the law as individuals rather than as group members during collective bargaining. In short, there is no assurance in either Canada or the United States that past pendulum swings favouring unions will be replicated or pursued with the same intensity.

Historically, union revival has followed an economic crisis and/or favourable legislative initiatives. The rise of industrial unionism in Canada and the United States followed the Great Depression and the enactment of legislation promoting the Wagner Act model, i.e., the positive right of employees to join and form unions. In the 1960s and 1970s, an explosion in public sector unionism in both countries followed the passage of enabling collective bargaining legislation. But now the prospects of an organizing revolution appear remote because of the absence of economic upheaval (e.g., a depression, declining confidence in market-based economies, and/or a resurgence of the welfare state) and the slim chances of major legislation favouring union organizing.

Union revival in terms of sustained increases in union membership and density—the ultimate measures of overall union success or failure—requires nothing less than a paradigm change in the industrial relations systems in both countries. This might encompass outright government sponsorship of workers’ rights to union membership and collective bargaining (equating these rights with fundamental civil rights), the protection of domestic markets and institutions from the adverse effects of globalization, and providing a larger voice for labour in macro-economic management. There must also be a broadening of the scope and breadth of union membership; intense efforts to organize temporary and part-time workers, new ways of representing the unemployed and low wage workers when bargaining status has not be achieved (e.g., through lobbying and enforcement of legal employment rights). Unions would resemble social movements as they seek to protect and speak for broad segments of the labour force; at the same time, they could not forsake their traditional roles as bargaining agents.

Union revival also means revival in political action. This will not only require the allocation of greater resources for political action and a continuation of the use of members as campaign volunteers, but a search for new ways of influencing the political process, for example, less emphasis on traditional party alliances and creating coalitions of like-minded organizations for grass-roots and community-based mobilization of members and other workers. Coalitions involving unions must be lasting and based on the parties’ acceptance of the missions and social value of each other, rather than transitional and purely pragmatic alliances directed at achieving a particular goal (e.g., passing a law, winning an organizing drive, stopping a threatened plant relocation).

We do not mean to diminish the significance of changes in unions in recent years. In the United States, where the union movement seemed moribund for decades, there has been a reawakening of sorts since the mid-1990s within the AFL-CIO and some major unions, e.g., the Teamsters and the AutoWorkers. Changes in leadership brought a higher priority to organizing and a new interest in politics. American unions are now building new alliances with environmentalists, consumer groups and students, and restoring old ones with clergy. Canadian unions have had a longer history of success in coalition activities. In both countries, coalitions have proven effective in support of strikers and organizing campaigns, particularly among the low-paid workers, and they are crucial for political action. New priorities and coalition activities have energized the federations and many unions on both sides of the border and have given them a new sense of optimism. This is a significant development by itself. Such an awakening is needed for union revival, but by itself does not constitute revival.

Much of the discussion of union revival has been either too diffused (defining revival in broad terms incapable of measurement) or too specific (focusing on individual events, such as strikes or organizing contests as evidence of revival). We consider revival as the task of expanding membership and union density through organizing and related bargaining and political activity, while also reversing the negative impact of low membership and density rates on those activities. Absent major changes in what unions do and for whom they do it, we remain pessimistic about the prospects for union revival in Canada and the United States.