What Workers Say: Employee Voice in the Anglo-American Workplace, edited by Richard B. Freeman, Peter Boxall and Peter Haynes, Ithaca: Cornell University Press, 2007, 244 pp., ISBN 978-0-8014-7281-7.

When this book arrived I eagerly read a few of the key chapters. Many of the contributors are colleagues and friends that I know and respect so I welcomed an opportunity to praise their work. But this book, while reporting useful survey results on workers attitudes and conditions with regard to voice at work, has some serious problems of conception and execution.

The inspiration for What Workers Say is the Workers Representation and Participation Survey (WRPS) originally carried out by Richard Freeman and Joel Rogers and reported in What Workers Want (1999, updated in 2006). That survey was remarkable for its finding that nearly all American workers wanted some form of collective representation in the establishment of their conditions of work. Some 32% of nonunion workers wanted a union. Most of the rest wanted “an organization with more limited independence from management.” Only 7% wanted “no organization.”

Unfortunately, in my opinion, What Workers Say does not effectively move forward the agenda of revealing the general failure of society to provide working people with collective representation of the sort that they want and to which international human rights standards say they are entitled. Instead, although there is some discussion of (so-called) “nonunion forms of representation,” What Workers Say focuses primarily on traditional union density and the unfulfilled demand for traditional union representation. Alternative voice mechanisms are discussed to an extent but, for the most part, are conceived of as employer-instituted competition for independent unions rather than as alternatives that employees might choose instead of conventional adversarial unionism. There is no clarity of awareness of the important distinction between independent local unionism and employer-dominated employee representation schemes. Into the nonunion conceptual basket goes everything from management briefing schemes to joint consultative committees to whatever the researcher considers to be unconventional worker representation.

The contributors to What Workers Say collected similar (but not exactly the same) data from Anglo-American countries other than the United States. In the United Kingdom, Ireland, Australia, and New Zealand fresh surveys modeled generally on the Freeman and Rogers study were carried out. In Canada the researchers (Michele Campolieti, Rafael Gomez and Morley Gunderson) mined further the database assembled originally by Noah Meltz and Seymour Martin Lipset and reported in The Paradox of American Unionism (Ithaca: ILR Press, 2004). And for the US, Freeman reprises the WRPS survey and reviews additional relevant surveys that have appeared since the WRPS findings were first published. In an early chapter Freeman, Peter Boxall and Peter Haynes argue that these “Anglo-American” economies are enough alike to constitute a “distinct capitalist model” and present relevant data in support of that claim.

With regard to the union situation in the covered countries, David Peetz and Ann Frost summarize the results as follows:

“First, union density has declined, but appears to have mostly stabilized. Second, in four of the countries (Australia, New Zealand, the United Kingdom and Ireland) union reach [defined by Boxall, et al. as “proportion of respondents who report the presence of a union of some kind at their workplace, and those reporting a union at their workplaces that they can join.”] is considerably higher than the union density figures indicate – which highlights a significant free-rider problem. Third, those who are members of a union overwhelmingly report high levels of satisfaction with their union. Fourth, many nonmembers report their willingness to join a union given the opportunity – importantly, this includes nearly a third of all workers in nonunionized workplaces. Fifth, employees in these six countries report wanting a degree of influence over decision making at work that they do not yet have. Finally, the employees surveyed consistently report (with perhaps the exception of Canadian employees) that they desire a cooperative relationship between the union and management at the workplace – but this does not only apply to the union because they also expect management to reciprocate by cooperating with the union to solve workplace problems.” Other than the first two, all of these findings are consistent with what Freeman and Rogers first reported for the US and suggest broad attitudinal consistency across the Anglo-American world.

John Purcell and Konstantinos Georgiadis, in their chapter on why employers should be interested in employee voice, add that “there has been a marked growth in all countries in employers use of direct forms of employee involvement,” such as “briefing systems or communication chains, quality circles or problem-solving groups, meetings with all employees (town hall meetings), types of team work giving employees some discretion, employee surveys, and other types of upward communication and regular information sharing.” In all of these countries too (except, the contributors tell us incorrectly, the US) there exists what Purcell and Georgiadis call “indirect forms of employee voice.” Among the varieties they mention: “worker representatives meet with management, such as employee forums, works councils, or joint consultative committees…” Consistent with the ambiguity in the volume as a whole, they make no distinction between management instituted and controlled forms of indirect voice and forms independent of management and under the control of workers.

Some of the observations in Richard Freeman’s article on the US are puzzling. He tells us unequivocally that “the United States has developed no new institutions to close the gap in representation and participation created by declining unionism,” which seems to directly contradict comments he makes in another part of the chapter about the appearance of organizations such as Alliance@IBM, the AFL-CIO’s Working America and SEIU’s Purple Ocean, all alternatives to traditional exclusive agent, monopoly representation unionism. Also according to Freeman “Section 8 (a) (2) of the National Labor Relations Act outlaws company unions, which rules out the company-sponsored labor committees that cover 14 percent of the workforce (and 20 % of the nonunion workforce) in Canada.” This passage gives the impression that Canada is different from the US in this regard but in The Paradox of American Unionism Lipset and Meltz reported essentially the same amount of nonunion representation in both the US and Canada (p. 165).

Few Canadian researchers and practitioners will recognize the representation system reported by Campolieti and company. They tell us that 55.7% (or, after some statistical fiddling, 55.9%) of Canadian workers have access to some form of representative voice in the workplace, a figure that appears in none of the major Canadian textbooks on industrial relations. It is generally understood that about 30% of Canadian workers are union members and that “collective bargaining coverage tends to be a couple of percentage points higher than the membership rate.” (G. Murray, “Unions: Membership, Structures, Actions, and Challenges,” Union-Management Relations in Canada, M. Gunderson, A. Ponak and D. G. Taras, eds., Toronto: Pearson Addison Wesley, 2005, p. 82). The 20+% difference is the result of some creative manipulation of the Lipset and Meltz data set. Campolieti, Gomez and Gunderson, in one part of their article, report that that either 42.1% (or possibly 46.6%) of Canadian workers enjoy “union voice.” Apparently 42.1% of the survey respondents reported being covered by collective agreements while another 4.5% enjoyed “dual channels.” Although the term “dual channels” is not clearly defined it would seem to suggest people enjoying both union and nonunion voice. Since they do not tell us otherwise, the 4.5% might be part of the 42.1% or it might be a completely independent category – thus the confusion over total coverage.

To further obfuscate the issue, in another part of the same article the authors, drawing on other surveys, state that the “actual unionization rate or probability of being a union member” for the economy as a whole is “roughly 34%.” In yet another paragraph they revert to the more broadly embraced statistic when they state that “unions appear ‘alive and well’ at about 30% of the potentially ‘organizable workforce’…” So, in one paragraph 42.1% (or 46.6%) is the percent of Canadian workers enjoying union voice, in another the figure is apparently 34% and in another it is estimated at 30%.

The authors also calculate a “potential unionization rate” of 42% by adding the 19% of nonunion workers who say they would like to have union voice and subtracting the 12% of unionized workers who say they would prefer to represent themselves to the “roughly 34%.” This produces the apparent absurdity that percent of workers already enjoying union voice (42.1 or 46.6%) is greater than the “potential unionization rate.” This may all be explainable as a function of the statistical manipulations that were applied but, in my judgment, it is very unenlightening and extraordinarily confusing.

Contrary to traditional practice in Canada, the authors define union voice as “the presence of a union in the workplace, rather than simply to membership and/or coverage.” While that might make some sense in countries where the scope of representation is a matter of negotiations between labour and management, it is a suspect notion in Canada where bargaining units are customarily legally defined. Does it mean that white-collar workers at Stelco, who have never enjoyed union representation, are counted as having union voice because the Steelworkers Union has organized and bargains for Stelco’s blue collar factory workers? If so, union voice has been awarded to a lot of Canadian workers who don’t have a clue about their representation status.

Also using the Meltz and Lipset survey, Campolieti, et al. tell us that “57.8 percent of Canadians prefer direct over collective forms of voice, at least in dealing with their own workplace problems.” But that statistic differs radically from what Freeman and Rogers found. In the original WRPS survey Freeman and Rogers sampled Canadian workers as well as US workers but only reported the US findings. However, Freeman provided me with the Canadian results and, contrary to what Campolieti and company report, they indicated that nearly all Canadian workers wanted some form of collective representation (Labour Left Out, Ottawa: Canadian Centre for Policy Alternatives, 2006, p. 24). Those data surely were available but the authors of this study offer no explanation for ignoring them.

As a final point of confusion, in the summary section of the article the authors state: “We find that roughly 30% of all workers (in the employed non-self-employed population) had no formal voice…” (emphasis added). As far as I can make out, this conclusion was not previously discussed and does not follow from the analysis.

I am sorry to say that I found still more of which to be critical in this volume, but a full delineation of all of the problems would require an article far too long for a book review. So, I will conclude with just two more observations.

In their final chapter Boxall, Haynes and Freeman, are very perplexed that unions in the UK, Ireland, Australia and New Zealand have a major free rider problem. They ask “Why do unions have such a large infill problem in these countries?” and then cast about for ad hoc explanations. But this is a phenomenon that has been well understood by international and comparative IR experts for decades. An extensive treatment of it is included in my Industrial Relations Under Liberal Democracy (Columbia: University of South Carolina Press, 1995).

My final complaint is that the contributors to this book have, for the most part, wholeheartedly bought into a vision of unionization and collective representation that sees it as purely an issue of economic self-interest. In the final chapter, for example, we are told that in order to “renew membership, unions need more ‘products’ than one-size-fits-all membership linked to collective bargaining.” The title of Freeman’s chapter is “Can the United States Clear the Market for Representation and Participation?”

There is little awareness in this book of the struggle of the international labour movement in the 19^th and 20^th centuries to secure worker dignity by insisting on the right of all workers to co-decide their conditions of work. There is no awareness of the strong contemporary, international movement building on that tradition by embracing unionism and collective bargaining as fundamental human rights. Surely one of the key reasons for the low representation rates in all of these countries is conceptual – the downgrading of unionism and collective bargaining from fundamental rights of all human beings to economic crutches for disgruntled individual labour market traders. So long as the representation and participation problem is considered to be one of conventional unions coming up with new products for potential new customers it is hard to see how unionism will once again be embraced by society as critical to the larger democratic project. And until genuine employee-controlled collective representation is seen, once more, as essential to democratic society, it is difficult to see how serious “union renewal” will occur.