This file is also available in Adobe Acrobat PDF format
Labor-based political parties have been an important electoral force in every advanced capitalist country. Every one, that is, except the United States. Elsewhere, these parties were established in the late nineteenth or early twentieth centuries, and, ever since then, there has been a great debate about why the American experience was different.
The late nineteenth century was also a critical period in the United States. Indeed, in the early 1890s, amid a wave of social and political unrest, the American union movement came close to establishing a labor party. At its annual convention in 1894, the American Federation of Labor debated a “Political Programme” that sought to commit the unions to independent political action. The Programme had been referred back to the Federation’s affiliates by the previous year’s convention, and many delegates were mandated to vote in favor of it. In fact, some unionists had already begun to build party organizations in a number of key cities and states. But AFL President Samuel Gompers and his allies were strongly opposed to the Programme, and with the help of some procedural machinations, they prevailed on the Federation to reject any foray into party politics. However, it was more than just procedural machinations that produced this result. For just one year later, delegates voted overwhelmingly for a resolution that declared that “party politics whether democratic, republican, socialistic, prohibition, or any other, should have no place in the convention of the A.F. of L.”1 Moreover, subsequent conventions repeatedly confirmed the AFL’s opposition to any form of partisan political action (whether through the establishment of a labor party or through involvement in one of the existing parties). In spite of the efforts of a substantial minority of unionists, the rejection of labor party politics became firmly entrenched.
The failure to establish a labor party had fundamental and wide-ranging consequences, not only for the political development of the United States, but also for its subsequent social and economic development. If a labor party had been established, it is highly likely that business interests would have had less influence over public policy, that income and wealth would have been more equally distributed, that trade unions would have been stronger, and that a more comprehensive welfare state would have developed. This last point can be made with particular confidence. After more than two decades of comparative research, it is now widely accepted that there is an important causal link between the influence wielded by labor-based parties, and the extent, type, and timing of welfare state development. Indeed, this “working-class power resources” or “social democratic” model of welfare state development has become a kind of orthodoxy. Like all orthodoxies, it has its challengers, and its supporters accept the need for various revisions and modifications. But even after all due weight has been given to a range of additional factors, there is good evidence that the political influence of organized labor is a key part of the explanation for some of the most important variations in social and economic policy.2
The failure to establish a labor party also lies at the heart of a wider debate about “American Exceptionalism,” and it provides an important vantage point from which to assess what, if anything, is distinctive about American politics and society. Some scholars now rail at the very mention of this debate, but there can be little doubt about its longstanding centrality in American intellectual life, or about its enduring influence over broader popular perceptions of the United States. Understanding what, if anything, is distinctive about American politics and society, and the nature of its institutional and ideological traditions, is, of course, a matter of great interest to Americans. But it need hardly be said that it is also of far more than local significance. Given the power the United States has to influence the rest of the world, the task of understanding the forces that shape its development is a matter of global importance.
In this book, I want to address a series of nested questions. At the center of the book is the title question about why there is no labor party in the United States. But the book also addresses both a more specific and a more general question. The more specific question concerns the decision that the American Federation of Labor took at its crucial convention in 1894, as well as the failure of various state-level initiatives at that time. The more general question concerns the longstanding effort to identify distinctive characteristics of American politics and society, and to offer an account of their origins and effects. Some of these questions will be familiar. But the approach to them will be novel, and it will, I hope, produce some unexpected answers.
The standard explanations for why there is no labor party rely on comparison with Europe. They point to various characteristics of the United States, like its high standard of living, its well-entrenched democracy, and its culture of liberal individualism. Explanations based on factors like these have become a kind of received wisdom, and they frequently appear in public commentaries, college textbooks, and scholarly debates.
But are they correct? In this book, I want to take a fresh look at these explanations. I propose to reassess them, and develop a new explanation, not by comparing the New World with the Old, but rather by pursuing a “most similar” comparison of one New World country with another. In particular, I propose to compare the United States with Australia. Many of the conventional explanations look much weaker when the United States is compared with Australia. For Australia had most of the same New World characteristics as the United States, and yet it produced one of the earliest and most electorally powerful labor parties in the world.
Comparison with Australia is especially appealing because Australian unions established their party in the early 1890s, just when American unions came closest to establishing a party of their own. In each case, discussions about the establishment of a labor party took place against a similar backdrop of events. In the early 1890s, both countries suffered from the worst depression of the nineteenth century, and a series of major industrial confrontations took place in which governments sided with employers and left the unions completely defeated. But in each case, the response of the union movement was different. After some initial vacillation, the American Federation of Labor rejected any form of party political involvement, and opted to remain committed to “pureand-simple” unionism. The Australian unions, on the other hand, put aside their longstanding apolitical traditions and decided to launch a party.
In order to clear the way for the systematic comparison that I envisage, I will begin by addressing a number of preliminary issues. First, I will clarify what I mean by a “labor-based party.” Second, I will outline the comparative explanatory strategy I propose to pursue. Third, I will consider a number of possible objections to my approach. Fourth, I will provide a brief account of the history of the labor movement in the United States and Australia in the late nineteenth century. Fifth, I will consider whether unionists in these countries were aware of each other’s activities. And finally, I will offer a short guide to the topics discussed in the chapters that follow.
The question about why there is no labor-based party in the United States needs to be distinguished from a number of related questions with which it is often confused. These include, most famously, the question about why there is no socialist party in the United States. But they also include questions about why there was not a more class-conscious labor movement, and why there was not a revolutionary labor-based party. In the extensive literature on these questions, the distinctions between them have often become blurred. I will frequently draw on the insights and causal hypotheses that emerge from this literature, but my focus will remain on the question about why there is no labor-based party. After all, it is only in this respect that the experience of the United States is distinctive. There were other advanced capitalist countries that did not establish a socialist party, or that did not have a very class-conscious labor movement, and there were many countries without significant revolutionary parties.
The labor-based parties that emerged in the late nineteenth and early twentieth centuries took different forms in different countries. Some were social democratic parties (like in Germany), some were socialist parties (like in France), and some were labor parties (like in Britain). But all of these parties saw themselves, and were seen by others, as members of a common political family. Indeed, they established an international organization—the Second International—that formalized and institutionalized this family relationship. And ever since they emerged, they have been regularly treated this way in numerous academic studies, newspaper commentaries, and public debates. What all these parties had in common was the uniquely privileged position they attributed to workers. Parties can be defined and categorized in terms of their ideology and identity, their organizational structures, and the social groups they represent. Labor-based parties attributed a uniquely privileged position to workers in all three of these respects, and it was this combination of characteristics that set them apart. In some cases, this privileged position was an informal and de facto one. More often, it was codified and formally entrenched.
The ideological pronouncements of labor-based parties made the pursuit of workers’ interests the central focus of their objectives, and the symbols they adopted made their self-image as the party of workers the centerpiece of their identity. In most cases, their ideology was a form of socialism. But this was not invariably so. In Britain and Australia, labor-based parties were initially established without a socialist objective, and in France, Italy, and Spain, anarchist currents were influential (Bartolini, 2000, 66–87). Labor-based parties were distinctive because of their labor-based ideology, not because of the radicalism, socialism, or leftism of that ideology. The organizational structure of labor-based parties gave a uniquely privileged place to trade unions. This manifested itself through the interpenetration of party and union organizations and the cross-linkages between them. These cross-linkages usually took the form of interlocking organizational ties, although they were occasionally more informal and contingent (Bartolini, 2000, 241–262). In some cases, the party was predominant; in some cases, the unions; and in some cases, neither. But in all cases, labor-based parties were distinctive because of the central importance attributed to their organizational ties with unions, and the priority these were given over relationships with other organized social groups.3 Workers were also the most important source of support for labor-based parties. Lack of polling data before the second half of the twentieth century makes it difficult to be precise, but there is a great deal of evidence that the predominance of working-class support was a common feature of these parties (Geary, 1981, 94–97). Labor-based parties might also appeal to other social groups (like small farmers or middle-class intellectuals), but workers remained the single most important group they represented.
The kind of labor-based party that was most likely to emerge in the United States was a labor party. This was the kind of labor-based party that emerged in all the other English-speaking countries. And although some socialists persisted with other models, this was the kind of party that the main proponents of labor-based party politics sought to establish in the United States in the late nineteenth century. It is thus also the kind of party with which I will be principally concerned.
Labor parties were established as the political wings of union movements. Unlike both social democratic and socialist parties, where Marxism and other socialist or revolutionary doctrines had an important influence, they were initially motivated more by pragmatic objectives than doctrinal ones. But labor parties themselves could take a number of more specific forms. In both the United States and Australia, the kind of labor party that union leaders sought to establish in the early 1890s was a labor-populist party, in which unions aimed to build an alliance with small farmers. In the United States, the main proponents of labor party politics sought to achieve their goal by building on and remolding the People’s parties that had recently been formed by farmers’ organizations. The attempt in 1894 to turn the People’s party in Chicago into a vehicle for labor party politics was meant to provide a model of how this could be done.
This attempt highlights the fact that there were two ways in which a labor party might have emerged. It might have emerged as the result of a national decision by the American Federation of Labor. But it might also have emerged as a result of the successful establishment of a model party in one or more key states: a model that was then emulated and spread to others. It will be important to keep both these possibilities in mind. In order to do this, I will have to pay careful attention, not only to the national decisions of the union movement, but also to the decisions of union leaders in states like Illinois.
The claim that there was no labor-based party in the United States requires some qualification. After all, a number of socialist and labor parties did appear in the late nineteenth and early twentieth centuries. But these organizations never attained the significance that labor-based parties attained in the rest of the advanced capitalist world. All statements about party systems require some criteria against which to determine the significance of individual parties. Just as we can meaningfully characterize the United States as having a two-party system, despite the existence of numerous minor parties, so, too, the claim that there has not been a labor-based party captures an important truth. This claim is really shorthand for the claim that there has not been an enduring electorally viable labor-based party. The most important labor parties—the United Labor parties in 1886, the labor-populist parties in 1894, and the farmer-labor parties after the First World War—were briefly able to garner significant electoral support, but they were not able to endure. The most important socialist parties—the Socialist Labor party in the late nineteenth century, and the Socialist party in the early twentieth—were enduring organizations, but despite a handful of local successes, they were not electorally significant. In other advanced capitalist countries, labor-based parties almost always became one of the main contenders for government office: backed by the support of between a third and a half of the electorate (Sassoon, 1996, 42–43). Even where they did not command this level of support, they always developed into important political actors. At the very least, they acquired “coalition” or “blackmail” potential (Sartori, 1976). Socialist parties in the United States met none of these criteria.
But what about the Democratic party? A third-party insurgency was not the only way in which a labor-based party might have emerged in the United States. In principle, such a party might also have resulted from an attempt to change the nature of one of the two main established parties. An early national move to align the unions with the Democratic party began in 1906 (Greene, 1998, and Sarasohn, 1989). But the most important move in this direction occurred during the New Deal, when, following the rise of the Congress of Industrial Organizations (the CIO), much of the union movement set aside its rejection of partisan politics. It is sometimes said that the resulting realignment turned the Democrats into a quasi-social democratic party. But while workers and their unions did become an important part of the New Deal coalition, they did not acquire a uniquely privileged position within the Democratic party.4 In terms of ideology and identity, New Deal liberalism was built around an eclectic mix of messages and measures, in which working-class interests had to jostle for position with the interests of other social and economic groups. In terms of organizational structure, unions and their Political Action Committees had to compete with the continuing importance of urban machines associated with ethnic and religious organizations, as well as with deeply entrenched southern conservative power brokers. In terms of party support, workers became one of at least four important electoral bases: taking their place alongside various ethno-religious groups, southern whites, and African-Americans. The Democratic party may have developed a quasi-social democratic tincture, but it did not become a labor-based party. The fact remains that the United States has never had an enduring electorally significant party of that sort.
Explaining by Comparing
A century or more of debate has thrown up a great many factors that may potentially help explain why there is no labor party in the United States. How are we to decide which, if any, of these factors really did have a significant effect? Natural scientists deal with this kind of problem by conducting experiments to control for the effects of different potential agents. Experiments enable them to isolate the effect of one agent by holding the others constant. But the application of this method is rarely possible for those seeking to explain social outcomes. We cannot just go back to 1894 and run that year again with, say, a new electoral system, or a different set of cultural values. Social scientists and historians have to rely instead on the comparison of cases that have actually occurred in the “laboratory” of history.
One way to try to control for the effects of different factors under these circumstances is to undertake the comparison of closely matched or “most similar” cases. This method of comparison attempts to approximate some of the advantages of experimental control by making the best possible use of the real-world cases that history has provided. It does this by comparing cases that have been carefully selected to ensure that, while they differ with respect to the outcome to be explained, they are as similar as possible in other respects. The aim is to control for as many potential explanatory factors as possible in order to identify the critical factor (or small cluster of factors) that differentiates the cases, and that might thus have helped cause the outcome in question.5
A most-similar comparison of the United States and Australia lies at the heart of this book. The basic justification for undertaking this comparison is methodological. It rests on the claim that systematic comparison with Australia makes it possible to use the most-similar method to maximum effect.
But comparison between the United States and Australia is not the only comparison that plays an important role in the book. One reason for this is that comparison with Europe, though it is perhaps less clearly visible, appears in the background throughout. Studies based on this conventional comparison provide most of the principle initial explanatory hypotheses that I will be seeking to test and reassess. Additional hypotheses suggested by comparison with Australia will also be examined, but it is the conventional comparison with Europe that underpins the study by providing its starting point. Thus, the two-country comparison that lies at the heart of this work does not stand on its own. Rather it self-consciously builds on over a century of scholarship and comparative effort.
There is also a more explicit sense in which the book rests on the comparison of multiple cases. For I will regularly supplement the principle intercountry comparison with the comparison of intracountry cases. Some of these are intertemporal comparisons, such as the comparison of labor party experiments in the United States in 1894 and 1886. Others are comparisons of territorial units within the United States and Australia. In this respect, I pay particular attention to the case of Illinois, but I also consider other states, especially in the industrialized Northeast and Midwest. Within Australia, I pay particular attention to the case of New South Wales, but I also make use of the cases of Victoria, Queensland, South Australia, and other states.
The choice of comparative methods invariably involves trade-offs.6 The best we can hope to ensure is that these trade-offs are reasonable, given the question at hand. In addition to the possibility of exercising “semi-experimental” control, the close comparison of a small number of similar cases offers some other significant advantages. This kind of comparison makes it possible to pay careful attention to detail, complexity, and the timing of developments. It helps avoid the inaccurate or trivial conclusions that can result from conceptual “stretching” or the retreat into high levels of abstraction. And it facilitates the examination of interaction effects at critical conjunctures. Furthermore—and this is especially significant—it makes it possible to complement the search for correlation with the search for causal mechanisms.
However, this kind of comparison also has some well-known limits. It may, for example, provide insufficient cases to test all the potential explanatory factors. Here, the mix of comparisons on which my argument rests, and, in particular, the ability to make use of supplementary temporal and territorial cases, helps compensate for this. Intracountry cases provide good material for additional comparisons, because they share most of the characteristics of the main country cases of which they are a part. They also make it possible to increase the number of cases without incurring the prohibitive costs of time and effort that would be required to fully characterize a number of completely new countries. Another well-known limitation of the comparison of a small number of cases concerns the difficulty of generalizing findings to other cases. Here, I will be primarily concerned with explaining developments in the United States, rather than with formulating a general theory of the emergence of labor-based party politics. Nevertheless, the fact that the book builds on the outcome of earlier comparisons with European countries goes some way towards addressing this problem.
Overall, the comparison I propose to pursue seems to offer a reasonable way of negotiating the trade-offs involved in choosing between different methods. It not only allows me to take full advantage of the most-similar method. It also enables me to take advantage of the close analysis that the comparison of a small number of cases makes possible, while retaining at least some of the advantages that come from the comparison of a larger number of cases.
In each chapter, I will consider one or more of the main potential explanatory factors. First, I will consider the extent to which each factor was present in the United States and Australia. Then, I will consider its effects. I aim to offer a nuanced and authoritative causal account (underpinned by carefully documented archival and secondary research), while at the same time retaining a clear analytical and explanatory focus.
In order to assess the extent to which each factor was present in the two countries, I need to bring together a great deal of information about the economic, social, political, and cultural characteristics of the United States and Australia in the late nineteenth century. Though simple in concept, in practice this is a difficult and labor-intensive task. So much so that one practitioner has warned that “no sane person would attempt it” (Fredrickson, 1997, 11). Undertaking this task, however, makes it possible to do two things.
First, it enables me to check whether the explanation in question rests on an accurate characterization of the United States. It may seem unnecessarily pedantic to insist on this, and I will not dwell on this point in every chapter. But, as we will see, a number of conventional explanations rely on unduly simplistic, partial, or otherwise misleading characterizations. Indeed, some rely on characterizations that are incompatible with each other.
Second, and most importantly, it enables me to cast doubt on the significance of a great many potential explanatory factors by demonstrating that these factors were present in both the United States and Australia. Indeed, as we will see, many of the best known and most widely accepted explanations appeal to factors of this sort. It is possible that one or more of these factors may still retain some explanatory significance because of their interaction with other factors, and I will need to bear this in mind when considering their effects.7 But provisionally at least, though only provisionally, I can set them to one side.
In order to assess the effects of each of the main potential explanatory factors, I need to identify whether or not there is a plausible causal mechanism that links the factor in question to the failure to establish a labor party.
A causal mechanism consists of a chain of causes and effects that connects a given explanatory factor to a given outcome, and accounts for how the causal impact of the factor is exerted. Identifying a plausible causal mechanism involves (a) breaking down the relationship between the factor and the outcome into a series of component relationships with one or more intermediate factors, and (b) showing that each of the factors in this series—both the initial explanatory factor and the various intermediate factors—have well-known or widely accepted effects, which, when taken together, help generate the outcome. As its name suggests, the concept of a causal mechanism is based on an analogy. It appeals to the idea that we should be able to specify the metaphorical cogs and levers that connect causes to outcomes.8
I will typically attempt to identify these “cogs and levers” by considering the way in which each potential explanatory factor altered the opportunities and constraints facing labor leaders or workers. I will try to offer a more fine-grained account of the causes and effects at work, by showing how various factors both shaped the interests and identities of key individuals and groups, and generated incentives that altered their choices and actions.
Comparison with Australia helps identify plausible causal mechanisms in the United States. Sometimes it contributes to establishing the plausibility of an already identified causal mechanism. Sometimes it points to a more plausible alternative. And sometimes it draws attention to the importance of particular intervening factors. But comparison with Australia is not, of course, the only way to identify plausible causal mechanisms, and where a potential explanatory factor is absent, it is less likely to help.
Where a factor is common to the United States and Australia, examining its effects helps determine whether the earlier provisional finding that it is not causally significant should be confirmed. A simple most-similar comparison of the extent to which a factor is present can cast doubt on its causal significance. But only after the additional examination of its effects is it warranted to reach the stronger conclusion that it can be ruled out. Comparison with Australia not only helps confirm that many factors can indeed be ruled out. It also helps show that some of these factors had effects in the United States that were quite different from those that are usually ascribed to them. In this way, it challenges some of the most entrenched conventional wisdom about American political development, and provides a vantage point that has the potential to alter the way we view the United States as a whole.
Where a factor differentiates the United States from Australia, examining its effects helps determine whether it really is causally significant, for correlation alone is insufficient to establish causation. Even if a potential explanatory factor is a distinctive characteristic of the United States, it can only be causally significant if there is a plausible causal mechanism linking it to the outcome we are seeking to explain. This stipulation enables us to distinguish between spurious or coincidental associations and real causal effects. Although comparison with Australia (supplemented by the comparison of various time-periods and states) does a better job of controlling for the effects of potential explanatory factors than comparison with Europe, it still leaves us with a number of contending explanations. There is no reason we should expect to find a mono-causal explanation. But we still need to determine which of the remaining factors really did have a significant effect. This is a problem that faces almost every attempt to use the most-similar method. Some countries, time-periods, or states are more similar than others. However, when dealing with the explanation of complex social phenomena, no cases are so similar that they enable us to rule out all bar one of the potentially relevant factors. The effort to identify plausible causal mechanisms helps surmount this problem. It enables us not only to confirm whether common factors can be excluded, but also to test which of the remaining factors really are causally significant, and to rule out those that are not.
Some Possible Objections
There are, of course, a number of possible objections to the project I propose to pursue. In this section, I want to try to preempt some of those that are most likely to be raised. Two concern comparison with Australia, one concerns the temporal focus, and another two concern the nature of the project itself.
The first objection concerns the political status of Australia. According to this objection, comparison with Australia is inappropriate, because, in the late nineteenth century, each Australian state was still a British colony. As a result, it might be thought that Australia is unable to provide a separate case with which to test explanations that derive from comparison between Europe and the United States. There is no doubt that the Australian colonies were not fully sovereign, legally independent states in the late nineteenth century. But, for my purposes, this is much less important than it might at first seem. For although the Australian states were still colonies, they were self-governing colonies. Not wanting a repeat of the Boston Tea Party, the British authorities in the mid-nineteenth century had sought to retain Australian loyalty (and minimize their own costs) by accepting demands for a form of self-government within the British Empire. Each colony had its own independently elected parliament and government, which not only controlled taxation, tariffs, and almost all other aspects of domestic economic and social policy, but also controlled immigration and its own police and military forces.9 So the Australian colonies were sufficiently independent political units to make comparison with them meaningful. And this was particularly true with respect to the matters of most concern to the labor movement. In any case, when unionists established labor parties in Australia, they were certainly not emulating or importing a British institution. For the British labor party did not yet exist.10
The second objection concerns the choice of Australia as a most-similar case. According to this objection, while Australia may indeed provide a similar case for comparison with the United States, there are other even more similar cases on which we should focus on instead. In particular, it might be thought better to compare the United States with Canada. The establishment of a stable labor-based party in Canada in 1932 was a somewhat belated development, and it did not become one of the two main national parties.11 Nevertheless, like other advanced capitalist countries, Canada did establish a labor-based party, and so it is eligible to be considered as a possible most-similar case. Canada undoubtedly shares some important underlying similarities with the United States, and comparison between the two has been fruitfully pursued in a number of studies, most notably those of Seymour Martin Lipset.12
For my purposes, however, the Australian case has two important advantages. The first stems from the fact that it not only shared many underlying similarities with the United States, but also shared many important proximate similarities because of the common circumstances that unions experienced in the two countries in the early 1890s. In part, this was simply because Australian unionists and their American counterparts were deciding whether or not to establish a labor party around the same time. The second advantage stems from the fact that Canadian unions were not independent organizations. Most were actually branches of unions in the United States, and the Canadian Trades and Labour Congress consisted almost entirely of Canadian locals of the unions that constituted the American Federation of Labor. As a result, the decisions Canadian unionists reached were heavily influenced by the decisions of their American counterparts: the very decisions we are trying to explain. Like the decisions of unionists in Illinois or other American states, the decisions of Canadian unionists can provide additional explanatory leverage (King et al., 1994, 222). But the organizational independence of the Australian union movement makes Australia a better principal case around which to build a most-similar comparison.
The third objection concerns the decision to focus on the late nineteenth century, and especially on the early 1890s. The economic and social turmoil of the early 1890s, and the loosening of political ties that accompanied it, provided the proponents of labor party politics with particularly propitious circumstances. However, others have pointed to the importance of different periods. Some have focused on the period around 1886 (Fink, 1983, and Voss, 1993), and some on the period after 1906 (Greene, 1998, and Sarasohn, 1989), while others have focused on the New Deal realignment of the 1930s, or the decade leading up to it or just after it (Brody, 1983, and Lichtenstein, 1989). Much can be learned from these periods, and I, too, will discuss them at points. But the 1890s are especially worthy of attention for at least three reasons. This was when the American union movement as a whole gave the question of whether to establish a labor party its most sustained and serious attention. This was when it came closest to actually establishing such a party. And this was when the rejection of this option became firmly entrenched as settled official policy.
The fourth objection I want to consider concerns the “Why no labor party?” question itself. Proponents of this objection argue that questions that seek to explain an absence or a non-occurrence are not proper questions for historical inquiry. They complain that, instead of calling for the explanation of what actually happened, this kind of question focuses on the failure of people to act in accordance with certain theoretical predictions or presumed norms. This objection has been made in various ways. Some complain of “negative questions” and an “epistemology of absence.” Others complain of “presumed norms of historical development,” “deviations from an expected trajectory,” and “a priori” or “essentialist” assumptions.13 There are a number of ways to respond to this objection, but the most powerful response is a very simple one. In asking why there is no labor party in the United States I am not trying to explain an “absence” or a “non-occurrence.” Rather, I am trying to explain something that did happen. I am trying to explain a decision that American labor leaders made in 1894, and reaffirmed repeatedly thereafter: a decision not to establish a labor party.
The objection to comparing the experience of the United States against presumed international norms is often associated with a fifth objection that concerns the “American Exceptionalism” debate as a whole. This objection has now itself become a kind of norm, with various authors noting that the debate about American exceptionalism is in “ill repute,” declaring themselves to be “against exceptionalism” and arguing that “the first thing” that comparative studies must avoid are “the ideas of norms and its corollary, exceptions.”14 It is easy to see why many writers are wary of the American exceptionalism paradigm. It can seem uncomfortably close to a misleading and self-congratulatory rhetorical posture, which often comes to the fore in American public life. The ritual incantations of politicians and journalists whenever a new president is inaugurated provide a typical example. These commentators invariably speak in awed tones about the peaceful transfer of power that is taking place, and the genius of the American system that allows it, seemingly oblivious to the fact that there are other countries that have managed this trick for generations. Some have even managed it without assassinations and civil wars.
I am sympathetic to the spirit of this objection. I, too, will reject the claim that the United States is exceptional in a great many of the ways that it is usually thought to be. Indeed, a recognition of just how similar the United States was to at least one other country lies at the heart of my entire approach. But there is no escaping the fact that some aspects of American political development have been exceptional. And the question that I am addressing—a question that lies at the heart of the traditional debate about American exceptionalism—concerns one such aspect. As the critics point out, it is wrong to assume the existence of international norms as a matter of pre-empirical commitment. But it is just as wrong to make a pre-empirical assumption that there cannot be any such norms. As a matter of fact, labor-based parties were established in all of the other advanced capitalist countries in the late nineteenth and early twentieth centuries. So, in this respect, there is an international norm, and the experience of the United States is exceptional.15 The exceptionalism with which I will be concerned is simply a matter of observed empirical fact. It is a fact that calls for explanation.
Since not all readers will be familiar with both of the two main countries that I will be comparing, I want to provide a brief overview of the development of the labor movement in each.
The United States
The labor movement first emerged in the United States in the 1820s and 1830s after some sporadic earlier activity. However, it was not until the 1850s and 1860s that stable unions were established. The printers were the first to form a durable national union organization. They were followed by other groups of skilled craft workers, especially after a revival of union activity that took place during and after the Civil War.
In the 1830s, and again in the 1860s, these unions formed Trades Assemblies or Trades and Labor Councils to coordinate their activities in major towns and cities. But it was only after the depression of 1873 to 1878 that city-based Trades Assemblies began to have a continuous existence. A number of short-lived national union organizations were also founded. But in 1881, craft unions laid the basis for a more enduring organization when they met to establish the Federation of Organized Trades and Labor Unions of the United States and Canada. This body was the direct predecessor of the American Federation of Labor, and it was relaunched with that name in 1886.
In the late nineteenth century, unskilled and semi-skilled workers also began to organize unions, especially in the mining and railroad industries. The repeated bursts of industrial action on the railroads were particularly significant because this industry lay at the heart of the American economy. The growth of organization among unskilled and semi-skilled workers was connected to the sudden rise and fall of the Knights of Labor, which organized local assemblies open to all workers. However, these were not, strictly speaking, union organizations since, with a few exceptions, non-employees could also join. The Knights, which had begun as a small secret society in 1869, experienced rapid growth in the mid-1880s. At its high point in 1886, the Knights claimed 703,000 members (up from 104,000 the previous year). But its loss of membership was just as precipitate, and within a few years it had lost most of its influence.
The rise and fall of the Knights of Labor was closely associated with the “Great Upheaval,” which came to a head in 1886. In that year, there was a wave of industrial unrest (following an earlier victory for the Knights against a railroad magnate), the newly relaunched American Federation of Labor began a national campaign for an eight-hour workday, and there was an upsurge in political repression (especially following the Haymarket bombing in Chicago). In the wake of these events, United Labor parties were formed to contest elections in many cities and states, but despite some impressive results, these parties did not survive long.
In the early 1890s, a number of developments combined to produce a new and unprecedented period of social conflict. Industrywide unions of coal miners and railroad workers were established. A series of major strikes took place, in which governments deployed military forces to intervene on behalf of the employers, and in which the unions were completely defeated. Populist farmers established People’s parties, which made gains in a number of states and appealed directly for the support of organized labor. And the economy sank into the worst depression of the nineteenth century.
It was in this context that the AFL came close to establishing a labor party. The 1893 AFL Convention voted to consider a “Political Programme.” The Programme was modeled on the recently adopted program of the British Independent Labour party, and it called on the unions to establish the capacity for independent political action. The Programme was referred back to the AFL’s constituents, who were asked to instruct their delegates on whether or not to vote in favor of its final adoption at the following year’s convention. Many unions had long been opposed to the establishment of a labor party, but a marked change in attitude began to appear.
The favorable reception the Programme was receiving gave a boost to those unionists who supported the idea of forming an alliance with small farmers, and saw the People’s parties as providing a vehicle that offered the best hope for establishing a new labor-based party. The supporters of this labor-populist strategy were especially strong in Chicago. First there, and then in a number of other cities and states, they sponsored conferences that saw the People’s party adopt most of the policy planks set out in the AFL’s putative Programme, and then began to put their plan into practice.
As the AFL’s 1894 convention approached, it seemed that a majority of delegates may have been mandated to support the Political Programme. However, AFL President Gompers and his allies remained firmly opposed to it, and using a series of arguments and procedural maneuvers, they managed to prevail. Supporters of the Programme retaliated by replacing Gompers as president. But at its 1895 convention, the AFL reconfirmed its position of opposition to the establishment of a labor party, and re-elected Gompers as president. Thereafter, both the position and the president became firmly entrenched.
With the AFL’s position decided, the People’s party fell under the control of its more conservative wing. The party then dropped most of labor’s demands, and placed its main emphasis on the demand for the free coinage of silver, which was offered as an all-purpose panacea. These free-silver populists then merged with the Democratic party, and in the elections of 1896, a major electoral realignment occurred, which enabled the Republican party to dominate national politics for most of the period until the New Deal.
The first unions were formed in Australia in the 1830s and 1840s, but it was only after the turmoil of the gold rushes in the early 1850s that stable union organizations began to emerge. These were craft unions formed by skilled workers, and a number of them, especially in the building trades, had important early successes with the achievement of an eight-hour workday.
Largely as a result of cooperation in the eight-hours movement, unions began to establish Trades and Labour Councils (TLCs) to coordinate their activities in each of the colonial capitals, beginning with the Melbourne Trades Hall Council in 1856, and the Sydney TLC in 1871. From 1879, the unions also began to coordinate their affairs on an Australia-wide basis at meetings of the Intercolonial Trades Union Congress. However, it was the TLCs that remained the main focus for cooperation.
In the 1870s, and even more so in the 1880s, the union movement expanded beyond its original craft base to include large numbers of unskilled and semi-skilled workers. This “new unionism” grew rapidly on the waterfront, on the railways, in mining, and in the pastoral industry. The formation of the Amalgamated Shearers’ Union (ASU) in 1886 was a particularly important development because of the pivotal role that the export of wool played in the Australian economy.
In the early 1890s, just before the onset of the same great depression that afflicted the United States, these unions became involved in a series of full-scale showdowns with employers. The first of these, the so-called “Maritime strike,” had two immediate sources: the attempt by ship owners to force the Marine Officers’ Association to disaffiliate from the Melbourne Trades Hall Council, and the attempt by shearers to establish a closed shop. The dispute spread throughout the economy as miners refused to supply coal to the ship owners, and waterside workers refused to handle non-union wool. The strike lasted for two months and was followed by further struggles in the mining and pastoral industries. In each case, governments sided with employers and deployed armed forces—and in each case the unions were defeated.
It was against this background that labor parties were formed in a number of colonies. Although they had occasionally experimented with the election of individual candidates, most unions had long been opposed to the idea of entering the electoral arena. This now began to change. The first labor party was established in NSW. In early 1890, before the Maritime strike had started, the NSW TLC had drawn up a plan to run candidates in the general election that was due the following year. However, most affiliates remained unenthusiastic about the plan, and it was only in 1891, after the defeat of the Maritime strike, that they agreed to establish a labor party. Similar proposals were being formulated in Queensland and South Australia, and the 1891 Intercolonial Trades Union Congress endorsed all these initiatives and called on other colonies to emulate them.
In its first electoral test in 1891, the new NSW Labor party—the Labor Electoral League—won 22 percent of the votes and 25 percent of the seats, leaving it holding the balance of power between the established parliamentary groupings of Free Traders and Protectionists. However, the party was soon dogged by defections and divisions. This experience led it to adopt a stricter system of party discipline, which a number of its parliamentarians refused to accept. In the 1894 elections, the party’s vote was reduced to 16 percent, and it won only 11 percent of the available seats, but the more disciplined party structure ultimately proved a more solid base on which to build support. In 1895, the party regained the balance of power, and in 1904 it became the official opposition.
Labor parties also began to build support in other colonies, and, after its establishment in 1901, in the new federal parliament. By 1899, the Queensland Labor party was able to form a minority government—the world’s first labor government—although it only lasted for six days. In 1904, and again in 1908–1909, Labor was able to form longer-lasting minority governments in the federal parliament, and the success of the party forced its two main opponents to merge. By 1910, both federally and in NSW, the Labor party was able to win power with a majority in its own right.
Developments in the United States and Australia did not take place in complete isolation from each other. The labor press covered news of developments in other countries, and labor leaders referred to these developments to an extent that readers may now find surprising. In both the United States and Australia, news of developments in Europe, and especially Britain, received particular attention. But labor leaders in these two countries were also aware of each other’s major developments.
Australian labor leaders learned about developments in the United States through reports of major American industrial and political conflicts in the Australian labor and daily press, by reading American labor journals, and in some cases through individual travel and contacts. When, in far outback New South Wales, a leader of the Broken Hill Miners’ strike obliquely urged fellow unionists to take direct action against strikebreakers by acting “how strikers acted in America,” he clearly felt able to assume that his audience was aware of recent confrontations there.16
Given the size and growing importance of the United States, not just economically but also as a political and ideological model, it is not surprising that Australian unionists followed developments there with interest. Many Australians looked at the United States as another English-speaking settler society that offered an image of their own future. The influence of developments in the American labor movement was part of a wider cultural influence of the United States as a whole. Indeed, in certain respects, the influence of these developments rivaled that of developments in Britain.
The very name of the Australian Labor Party provides some evidence of this influence. In Australia, the word labor is spelled “labour” with a “u” as in Britain. But, to this day, in the name of the Australian Labor Party, it is spelled without one, as in the United States. The exact reason for this is unclear, and the spelling varied in the early years of the party’s existence. But it seems to have reflected the influence of American experiments in the minds of some of the party’s main backers.17
Labor leaders in the United States were also able to follow developments in Australia through reports in the labor press, and, again, in some cases through individual contacts. But the influence of Australian developments on the labor movement in the United States was not as broad or deep as the influence of American developments in Australia. Outside the Pacific Coast, it was largely limited to leaders, activists, and intellectuals. In the minds of these leaders, Australia was notable because it was seen as a similar society in which the labor movement was unusually strong, well-developed, and advanced. AFL President Gompers, for example, claimed that he sometimes read the Melbourne papers in order to follow the success of the movement for an eight-hour workday. He told the 1892 AFL convention that “comparatively speaking” unions in Australia are “the most extensive, general, and perfected.” And during the debate over the Political Programme in 1894, he published a letter from the President of the Melbourne Trades and Labour Council that gave details of the number of labor members in the various Australian parliaments.18
The most detailed source of information about the Australian labor movement in the early 1890s was the Coast Seamen’s Journal. This highly professional and well-edited publication was the official organ of the Sailors’ Union of the Pacific. Once a month it published “Our Australian Letter”, which often appeared on its front cover, while on other weeks it frequently carried a lengthy column of “Australian News.” The letter, which was written by the Secretary of the Seamen’s Union in Sydney, provided detailed information about the latest industrial and political developments affecting the Australian labor movement. The amount of information about Australia in this journal was quite exceptional. However, the journal certainly circulated beyond San Francisco and the Pacific coast. It must have been read in the Chicago headquarters of the National Seamen’s Union, and from there, it presumably reached others in Chicago—a city that was at the center of labor reform movements and labor party experiments.
The 1893 World’s Fair in Chicago provided another conduit through which information about the Australian labor movement reached labor leaders in the United States. The commissioners who organized the NSW exhibit published a number of accompanying pamphlets, including one on “Social, Industrial, Political and Co-operative Associations.” A wide range of American labor leaders attended the “World Labor Congress” that took place under the auspices of the Fair, and at which they discussed a number of “profoundly stimulating” reports on labor movements in Europe and the Antipodes. Henry Demarest Lloyd, the Chicago-based intellectual who helped organize these meetings, and who was at the center of the effort to establish a labor party, was certainly aware of the emergence of the Australian Labor party, which he specifically alluded to in some influential speeches. In March 1893, he began a sustained investigation of developments in Australia and New Zealand, and by 1894 he had come to see the Antipodes as providing a model for the United States.19
By the early twentieth century, Australia came to be seen, in both the United States and Europe, as something of a social laboratory. The consolidation of the Labor party, the introduction of arbitration courts, and the beginnings of a welfare state seemed to place Australia in the advanced guard of an international movement for social reform. And a number of social scientists and labor-oriented intellectuals studied developments there and published their findings. This interest waned after the First World War, although occasional articles continued to be published. In the mid-twentieth century, the best-known scholars of American exceptionalism were certainly aware of the importance of testing their arguments against the Australian case, and the significance of the case is sometimes briefly acknowledged in more recent comparative treatments of the American labor movement. However, despite its methodological advantages, there has never been an attempt to offer a systematic comparison of the development of labor politics in the two countries.20
Each of the chapters that follow considers one or more of the main factors that may potentially help explain why there is no labor party in the United States. As Table 0.1 shows, these factors can be grouped into three loosely defined categories. However, these categories should not have too much importance ascribed to them, for many of the factors could easily be placed in more than one category. The arguments in each chapter build on those that precede it. But the chapters have also been written so that each can be read as a relatively self-contained whole by those principally interested in one particular aspect of social and political development.
Chapter 1 considers the economic interests of workers. First, it examines the argument that the level of prosperity in the United States ensured that economic grievances were insufficient to support the establishment of a labor party. Second, it examines the impact of different types of union organization, paying special attention to the weakness of “new” unions that organized large numbers of unskilled and semi-skilled workers. And third, it examines whether it was possible for unions to facilitate a labor-populist alliance between workers and small farmers. In the process, the chapter also aims to provide important background information about workers and their organizations.
Chapter 2 considers questions of race and immigration. It examines the argument that racial consciousness and racial conflict reduced the viability of class-based movements. In particular, it examines the claim that anti-black and anti-Chinese sentiments generated divisions between workers that hindered the establishment of a labor party, both directly (by weakening its electoral viability) and indirectly (by making it more difficult to organize mass industrywide unions). The chapter also considers the conflicts and changing attitudes that accompanied the growth of immigration from southern and eastern Europe in the late nineteenth century. It examines the claim that the racialization of hostility towards these immigrants strengthened intraclass divisions between skilled and unskilled workers and further undermined the prospects for the establishment of a labor party.
Chapter 3 considers some basic institutional features of the American political system. First, it examines the argument that the early removal of property qualifications and the introduction of manhood suffrage for whites removed the kind of class-based political grievance that a labor party needed in order to mobilize support. Second, it examines the argument that the electoral system reinforced the position of the two main existing parties and made it extremely difficult for any third party to gain legislative seats. Third, it examines the argument that federalism hampered the emergence of a labor party by dispersing political authority among different political units and by multiplying the number of elections the party would have to contest. And fourth, it examines the argument that presidentialism, and the two-party reinforcing effects of presidential elections, generated incentives that undermined the independent political strategy many labor leaders hoped to pursue.
Chapter 4 considers the role of the courts. In the late nineteenth century, unions in the United States experienced a wave of intense judicial hostility. The chapter examines the argument that this led unions to conclude that it was either futile or foolish to engage in electoral politics. According to this argument, unions reached this conclusion because judicial review gave the courts the final say on the political decisions that mattered most to them, and because the courts were largely immune to external political influence. The chapter also summarizes the evolution of labor law and traces the development of union attitudes to politics. It highlights the fact that these attitudes varied along more than one dimension.
Chapter 5 considers police and military intervention, as well as the overall impact of repression. There are two conventional theses that see repression as having had an important impact. According to the soft repression thesis, there was so little repression in the United States that unions did not have sufficient incentive to engage in political action, while according to the hard repression thesis, there was so much repression that unions were cowed into adopting an apolitical stance. This chapter examines which, if either, of these arguments is right. The chapter also discusses the development of police and military forces, and provides an account of the main strikes that unions experienced in the early 1890s.
Chapter 6 considers liberal ideas and the weakness of feudal traditions. First, it examines the argument that the prevalence of the idea of social egalitarianism—an egalitarianism, not of economic resources, but of social status—minimized or eliminated the status-based grievances and class consciousness that might otherwise have made it possible to build support for a labor party. Second, it examines the argument that the prevalence of the idea of individual freedom delegitimized the interventionist political goals of the labor movement, and underpinned a sense of American identity that was inimical to these goals. The chapter considers both how labor leaders interpreted these ideas, and whether their interpretations were likely to seem plausible to ordinary workers. In assessing the plausibility of these interpretations, it considers the social behavior of capitalists and other elites, the involvement of the state in economic development, the “mateship” ethos, the influence of the “New Liberalism,” and the growth of industrial concentrations and monopolies.
Chapter 7 considers religion. It examines the extent of religious involvement, the nature of religious beliefs and practices (including the level of support for different denominations and the strength of evangelicalism and revivalism), the attitudes of the Protestant and Catholic clergy, and the response of labor leaders to these attitudes. The chapter pays special attention to the relationship between religion and the party system. It examines the political salience of religious sectarianism in the late nineteenth century, and considers conflicts over temperance, education, and organized anti-Catholicism. In particular, it examines the argument that religious sectarianism fostered intense Democratic and Republican loyalties among ordinary workers, and that this led labor leaders to fear that union organization would be severely disrupted if they entered the electoral arena.
Chapter 8 considers left-wing factions and their reform ideologies. It examines the influence of socialists, anarchists, populists, proponents of Henry George’s single tax, “nationalist” followers of Edward Bellamy, cooperative colonists, Knights of Labor, and “pure and simple” unionists. The chapter looks at the level of support for each of these schools of thought among labor leaders and activists, and considers the impact that tensions between them had on labor party experiments. The chapter also pays special attention to the influence of Marxian socialism, and the extent of factional conflict between those who were influenced by it. It examines the argument that this conflict led many labor leaders to fear that the establishment of a labor party would produce an outbreak of socialist sectarianism that could destroy the unions themselves.
The conclusion draws together the main findings of the book. It emphasizes that these findings rule out many of the best-known conventional explanations for the fate of labor party politics, and it suggests that some of the factors to which these explanations appeal had effects that were very different from those that are usually attributed to them. It then sets out which factors really do help explain why there is no labor party in the United States, and examines the interaction between them. Finally, it considers the consequences of these findings for the wider debate about the character of American politics and society.
Return to Book Description
File created: 11/28/2007
Questions and comments to: email@example.com
Princeton University Press