CHAPTER 11

THE AMERICAN DREAM

More than any nation, America has celebrated itself as the land of opportunity. Immigrants came to the New World to escape European class barriers. In America, wealth and position were to be organized differently—open to every person of talent and hard work. The frontier beckoned to those seeking a new chance. The fabulous wealth of the continent promised abundance for all who were willing to work. It was the natural setting for an ideology of individualism.

The roots of the American Dream can be found in colonial America; it gave Poor Richard’s Almanac its distinctive American character. But it was especially during the rise of industrialization in the late nineteenth century that the imagery of upward mobility came to dominate America’s discussions of itself (Thernstrom, 1964:57–59). It is certainly no accident that just as a massive industrial proletariat was being created on these shores, America worked hardest at convincing itself of its openness and classlessness.

The American Dream seemed the perfect immunization against the dangers of a militant class consciousness. It promised a common vision to all Americans—workers and bosses, the poor as well as the rich. In an open America, class struggle would be unnecessary. Discontent with one’s position would inspire workers to change their positions within the system, rather than trying to change the system itself.¹

Class struggle, if not unnecessary, would at least be crippled by the individualistic pursuit of gain. To the extent that workers succeeded in getting ahead, their mobility deprived the working class of its ablest leaders. To the extent that workers failed, they were left blaming themselves for their own deprivations (Lane, 1962; Sennett and Cobb, 1972). Either way, the struggle to get ahead fragmented workers and prevented them from developing a collective class consciousness. In this way, the American Dream reinforced the class system and protected it from the challenges of a radical working class.

The image of America as the Land of Promise is such an indelible part of the national heritage that it has been a favorite explanation for the failure of American class consciousness. Americans both believed in individual opportunity and lacked a radical working-class consciousness; the two characteristics must surely be linked as cause and effect.² The pieces of the theory fit so well together that research often neglected to look for rigorous evidence, convinced that closer scrutiny would only demonstrate the obvious.

There were some objections. American Dream theories especially outraged European Marxists (Moore, 1970).³ According to Marxian theory, the spectacular development of American capitalism implied that American workers should be especially impoverished. Therefore, Marxists quickly challenged the reality of the American Dream. Sometimes they based their objections on little more than anecdotes describing individual cases of poverty or on travelogues of visiting Europeans.⁴ Sometimes they undertook more rigorous compilations of wage rates, cost-of-living estimates, and standard-of-living comparisons (Kautsky, 1905–6, cited in Moore, 1970:118).

But the Marxist dissent seemed, even then, more the desperate defense of a prophet that failed than an analysis of actual American conditions. The American Dream had more resilience than complex wage calculations could dislodge. The country’s founders had passed down a faith in the opportunities of the New World, and the elite consensus simply dismissed the carping of radical critics as biased political ideology.

Latter-day academics with better tools, though hardly the same radical purpose, have taken up a similar line of questioning. Are American workers really more prosperous than their European contemporaries? Do they really have more chances to advance to the middle class? Or is the American Dream only a mirage that obscures the existing class divisions?

The prevalence of these empirical inquiries could easily distract us from the more fundamental question of the relevance of opportunity or affluence for working-class consciousness. Even if American workers were more mobile or prosperous, would that necessarily make them less class conscious? What is the connection between “objective” realities such as mobility rates and “subjective” reactions such as class consciousness? The theory that mobility or prosperity deters class consciousness is plausible enough, but nobody has adequately tested it.⁵ Stephan Thernstrom, after two major histories of American mobility (1964, 1973), leaves to “future research” the question whether that mobility actually “impeded the formation of class-based protest movements” (1973:259). But this, of course, is the fundamental question.

Again, the exercise of constructing the opposite hypothesis is instructive (see Katznelson, 1981:12–13). Perhaps mobility promotes dissension rather than stability. After all, mobility disrupts the existing pattern of personal ties that usually keep subordinates in their place. Mobility raises expectations that were never imagined in traditional societies. The new expectations aggravate the sense of deprivation of those workers left behind. And the relative success of the few who succeed provides them also with greater resources with which to challenge those still above them.

Such destabilizing consequences of mobility are as plausible as the more conservative theory. A more Durkheimian theory that change creates unrest would explain why the early stages of industrialization are the most dangerous for capitalism. The great modern revolutions, the French, Russian, and Chinese, erupted not in advanced industrial economies but during capitalism’s first stirrings. It is in the first stages that change is most severe and workers are freed from traditional restraints.

Clever reasoning could perhaps reconcile the two theories: perhaps both mobility and stability weaken working-class consciousness. Our purpose in raising the countertheory is to question a facile assumption that opportunity and wealth always weaken working-class consciousness. We would like proof and are disturbed by how little evidence has been mustered on behalf of such a theory.

Our contribution to this debate is quite modest. We merely wish to show that individual opportunity plays little part in obscuring the perception of class in the United States. American workers know who is working class, and the American Dream does not delude them into imagining that they are anything more or less than they are.

The American Dream concept provides at least three distinct explanations of American exceptionalism: the frontier, social mobility, and wealth. There is even a rough progression through history in the popularity of these three theories. The first explanations focused on the frontier as the outlet for working-class discontent. But as the frontier closed and the working class remained impotent, other explanations had to be sought.⁶ The new theories equated the social mobility of industrial society with the geographic mobility of the old frontier (Thernstrom, 1964:61): the opportunity to move up replaced the opportunity to move West. And throughout our history but especially after World War II, various theorists have cited the American workers’ affluence as the guarantor of working-class conservatism: it was not only the promise of mobility to the few but the reality of generous incomes for the many that forestalled any sense of mass injustice. We shall discuss each of these three theories in turn, distinguishing them for separate analysis, but recognizing that they form an interrelated set of explanations about how the American Dream has anesthetized working-class consciousness.

Geography

The Frontier

For Frederick Jackson Turner (1920) the American frontier explained weak American class consciousness. According to Turner’s theory, the frontier was a great outlet that drained away the most discontented and reinforced an individualistic ideology of achievement. Class consciousness—not to mention class warfare—required a stable working class stuck in its position with little hope of escape.⁷ The frontier opened up the escape valves and prevented the urban industrial pressure cooker from building up too much steam. The solidarities forged in the crowded cities of the East were dissipated in the great open spaces of the West.

In fact, the frontier thesis ignores some of the most violent battles in American labor history.⁸ Most of the disorder in the 1894 Pullman strike occurred in the West (Taft and Ross, 1969:286). At Coeur d’Alene, Idaho, dynamite twice became the medium of exchange between striking miners and mine owners (Jensen, 1950). The Western Federation of Miners emerged from these conflicts as one of the most radical unions in American history, the progenitor of the Industrial Workers of the World (see Dubofsky, 1969). The Wobblies had their greatest strength in the West and fought some of their most violent conflicts there. Besides the IWW, the Pacific coast longshoremen’s union has long distinguished itself as the major Communist-led union in the country.

Colorado probably exceeds all other states for the most violent labor disputes. During the Colorado Labor War of 1904, violence killed 17 and wounded 23. Ten years later, coal strikes around Ludlow ended with 74 dead. In 1920, seven more died in a Denver street railway strike, and in 1927, eight were killed in another coal strike (Taft and Ross, 1969).

The West also pioneered the use of the general strike, often considered to be labor’s most class-conscious weapon (see, for example, Luxemburg [1906], 1970; Brecher, 1972:233–63). By this standard, Seattle’s general strike of 1919 was perhaps the most thorough demonstration of working-class solidarity in American history. At the time, the mayor called it a “Bolshevik insurrection” (Friedheim, 1964). General strikes were also called in Waco, Texas, and in Kansas City, Missouri, during World War I (Bing, 1921:30), and in San Francisco in 1934. In Billings, Montana, employers turned the tactic around and called a general lockout in 1918 (Bing, 1921:30).

All these labor battles were fought in the West, where the American frontier supposedly dissipated working-class consciousness. Despite the popularity of the thesis about the pacifying influence of the frontier, the historical data do not support it.

Nor do our survey data suggest any important regional differences in class perceptions (see Table 11.1). Western workers are about 3 percentage points more likely to see themselves as working class than workers elsewhere in the country. This difference is in the opposite direction from an expectation of weaker working-class consciousness. Moreover, managers in the West are also about 3 points less likely to see themselves as middle class, so the gap between managers and workers is almost exactly the same size in each region of the country: class divisions are no more or less clear in the West than in the more industrialized areas of the country.⁹

The South

If any region is known for its weak working-class movement today, it is the South, not the West. Low rates of unionization have combined with antiunion politics to reduce southern workers to the most poorly paid and weakest sector of American labor. In 1977 only 13 percent of southern workers belonged to unions, compared to 24 percent in the rest of the country (Freeman and Medoff, 1984:27).

Labor’s weakness in the South cannot be blamed on weak working-class consciousness. As Table 11.1 shows, southern workers perceive the working-class division in exactly the same way as do workers in the northeast and north-central regions. Moreover, southern workers may want unions as much as workers elsewhere. Freeman and Medoff (1984:29) report that southern blue-collar workers not yet in unions are more likely (46 percent) to say they would vote for a union than would similar workers in other regions (28 to 38 percent). (The ABC/Washington Post poll, however, found no regional differences among nonunion blue-collar workers in their willingness to join a union.) The attitudinal evidence, therefore, suggests that low unionization in the South may more reasonably be attributed to the hostility of management in low-profit southern industry and such barriers as the right-to-work laws prevalent in the South.

Internal migration

Nineteenth-century workers moved not only to the frontier but within the eastern metropolises as well. Boston in the 1880s underwent a “dizzying” turnover rate of 64 percent (Thernstrom, 1973:221).¹⁰ Footloose workers rarely put down roots long enough to build a stable working-class movement. A continually fluctuating population obstructs labor organization, no matter how class conscious the workers are.¹¹

Geographic movement has also been thought to encourage a psychology of individual achievement rather than a class-conscious solidarity, to reduce the sense of common class experience that creates strong working-class loyalties. Neighborhoods become less stable and work groups more temporary. As mobility severs the ties between one worker and another, it offers the hope of individual advancement.¹²

Neither the historical nor the attitudinal evidence supports this otherwise plausible theory. Migratory workers, such as loggers and maritime workers, have created some of the most radical and militant unions in the United States.

Our survey data also provide no support for any linkage between physical mobility and weak working-class consciousness. Of course the interviews are all contemporary, whereas the geographic mobility theories developed from the early industrialization of the nineteenth century. If workers’ reactions to geographic mobility were different then than they are now, we have almost no way of finding out what they were.

Should this movement have any effect on people’s class perceptions? It is hard to predict its impact. The fluidity theory might suggest that class perceptions would be less stable for the movers. Compared to those settled in the same place all their lives, the movers might be less certain of their position in the class structure and therefore less class conscious. This greater uncertainty would be detected in our data as a less clear division between managers and workers.

In addition, geographic mobility itself may imply some middle-class status. In the contemporary United States, unlike in the nineteenth-century cities Thernstrom studied, geographic mobility is now associated with the middle class (Blau and Duncan, 1967).¹³ People move out in order to move up. In our own sample, the managers are more often migrants: 66 percent of the managers are now living in places different from where they grew up, compared to 52 percent of workers. Migrants may therefore have some reason to assume a middle-class status even if their current jobs don’t yet reflect middle-class positions.

There is little support for either of these possible implications of internal migration. “Movers” are slightly more middle class than “stayers” but by only a couple of percentage points (see Table 11.2). Both managers and workers who move show the same slight tendency to middle-class self-placement, so the gap between them is unaffected by migration. Thus there is no evidence at all that lifelong residents are any clearer about their class position than the transients. In sum, internal migration appears to be no more relevant than region in determining class perceptions.

Social Mobility

The frontier has long since ceased to be a significant factor in American class relations. Even geographic mobility is not the distinctive characteristic of the working class that it once was. As these outlets for class tensions became less available, more attention focused on a different kind of mobility—not physical movement across the country but social movement up the status ladder. The extent of this social mobility reinforced the image of America as the land of opportunity (Bottomore, 1966:50).

The rags-to-riches story has become the central element in the American Dream ideology. Even if most social mobility was only short-distance (rags-to-respectable-working-class, and working-class-to-small-proprietor), it still rewarded individual initiative and thus was believed to drain the American working class of the kind of collective resentment that created Old World militancy (de Tocqueville [1835, 1840], 1954 [vol. 2]:269; Commons, 1908:760–61; Coser, 1956:36; Lipset, 1960:267; Lane, 1962: 218; Thernstrom, 1964:58, 1973:258; Sweezy, 1967: 160–61; Burawoy, 1979:106–7; Karabel, 1979:212). In Europe, workers were stuck in their lot in life and could turn only to collective protest for any hope of improving their position.

Arthur Shostak, for example, in his Blue-Collar Life, asserted—without offering any documented evidence—that in the United States,

blue-collarites are further discouraged from class-conscious politics by the underrecognized presence in their ranks of two types of mobile individuals, those “displaced” workers skidding down from above, and those Horatio Alger types aspiring up and out. [The skidders], who either were raised in white-collar families or were once possessors of white-collar work . . . often remain optimistic about their chances to recoup status losses; as such they function to reduce working-class solidarity from below. . . . Blue-collarites who are busy making their way up and out of the ranks alternate ties among blue-collar peers, and in other ways undermine class cohesiveness. (Shostak, 1969:226)

Such assertions have never been systematically proven. Do skidders maintain middle-class values? Are potential mobiles less class conscious? American sociologists have not often asked about the effects of opportunity: does rapid mobility in fact reduce working-class consciousness?¹⁴ Instead, they have questioned the extent of opportunity: do U.S. workers really have such great chances for advancement, and are they better chances than European workers have? Our main interest is in the effects of mobility although we can only begin to sketch an answer. But first we should look at what social science has learned about the rates of social mobility in the United States.

Rates of mobility

The verdict on American mobility rates is not yet in, despite vast amounts of research in the last two decades. The question proved more complicated than was first imagined. A number of false starts provided lessons about distinctions that must be made before we can compare mobility rates in the United States with those of other industrial countries.¹⁵ Much of American mobility into the new middle class is a consequence of the extraordinary growth of professional and managerial positions in the American occupational structure. As the percentage of middle-class positions increased from 15 to 30 percent in a generation, somebody had to fill those new positions. The openings thus created pulled up the children of working-class families. This kind of structural mobility was not a reflection of the openness of the system itself—that is, it did not reflect any of the equality of opportunity that Americans pride themselves upon. A society that guarantees middle-class positions to all the children of the middle class and accepts working-class children only in order to fill vacancies can hardly be judged to be very open and fluid. Therefore, the openness of the system—the degree of circulation mobility—came to be defined as all the residual movement up and down and social hierarchy once the mobility mandated by structural changes had been subtracted out.

The studies that separated structural from circulation mobility concluded that most American mobility was structurally required by the growth of middle-class positions (Hauser et al., 1975). With that growth-determined movement subtracted, U.S. rates of circulation mobility were above average but not exceptionally so, and gave every indication of being quite constant throughout history. One cross-national comparison of blue-collar-white-collar circulation mobility (Tyree, Semyonov, and Hodge, 1979) placed U.S. rates somewhat lower than Canada’s and higher than Great Britain’s.

Consequences of mobility

Our interest, however, is in the consequences of those mobility rates, and the study of these consequences has additional complications. The most important is the separation of the effects of mobility itself—the movement from one position to another—from the simple effects of destination and origin. The issues are necessarily intertwined, much like the age, period, and cohort effects discussed in Chapter 6.

We begin with a computation of working-class and middle-class placements for 20 separate categories according to four levels of the person’s own occupation and five levels of the father’s occupation. The divisions—professional/managerial, other white-collar, skilled blue-collar, other blue-collar, and farm (for fathers only)—incorporate the actual class divisions (manager-worker-farm) and the most often noted status distinctions within the working class. (More detailed breakdowns would be possible but would result in such small samples in some of the cells of the matrix that we would not get reliable estimates of their subjective class placements.)¹⁶ The results of these calculations are presented in Table 11.3. The highlighted cells forming a diagonal in these tables represent the occupationally stable Americans whose work falls in the same broad category as their fathers’. Those above that diagonal are upwardly mobile: their occupational position has higher status than their fathers’.¹⁷ They constitute 46–52 percent of the total sample, a large segment that reveals the structural shift upward between generations. The downwardly mobile, some 20–22 percent of the samples, are below the diagonal.

If we sort the 20 categories into these three broad groups, we find that the downwardly mobile are more working class than the stable, as we would expect. The upwardly mobile, however, are not more middle class than the stable; among women the upwardly mobile are notably less middle class than the stable. But these comparisons say little about the effects of mobility, because we cannot tell whether the class placements are determined by the direction of mobility or merely by the occupational destination. By definition, none of the downwardly mobile are managers, but substantial proportions of both the upwardly mobile and the stable currently hold managerial positions. Thus the upwardly mobile and the stable are more middle class because they are more often managerial, not because they have been mobile.

What we have to do is look at the effects of mobility while holding constant the person’s present position. For instance, we should compare the stable skilled workers with those who were upwardly mobile into skilled work and with those who were downwardly mobile into skilled work. It is obvious from Table 11.3 what the results of those comparisons would be. The upwardly mobile are less middle class than the stable; the stable are less middle class than the downwardly mobile. According to such an analysis, upward mobility makes people more working class and downward mobility makes people more middle class! But this result is also misleading because it is not the act of mobility per se that determines those class placements; it is the direct effect of parental origins. All persons with unskilled working-class fathers have more working-class identifications than people with skilled-worker fathers, regardless of their current positions; that is, there is a consistent (main) effect of father’s occupation on class placement, regardless of the person’s current position. Therefore, parental origin effects must also be subtracted from the mobility comparisons in order to determine the effects of movement itself. In other words, we must control for both parental origin and current occupation. To do this we have to estimate the effects of each of the “father’s” and “own” categories, subtract these estimates from the observed percentages in Table 11.3, and then look at the differences due to mobility. These estimates are reported in Table 11.4. What is clear from this table is that mobility makes very little difference; virtually all the differences in Table 11.3 can be explained by the overall effects of “father’s” and “own” occupation, beyond this, whether any mobility was entailed by the particular conjunction of the two is irrelevant.¹⁸

The conclusion about the effects of mobility is very similar to the conclusion about the rates of mobility: what matters is the number of middle-class and working-class positions, not the process by which people move into these positions. Working-class identifications will increase or decrease depending on the number of working-class positions and working-class (or farm) fathers. Mobility itself has little effect on subjective class placements.

The importance of one’s own occupation is little more than a restatement of what we observed in Chapter 4: the person’s position in the class structure is a primary determinant of class perceptions. We have now added a new component that requires some discussion: apparently one’s father’s position also has an effect on class perceptions. Those who had managerial fathers are especially likely to see themselves as middle class. Why is this? Why should family origin—removed by years or decades from the experience of most of our respondents—continue to have an effect on their subjective class placements? Granted, the parental effect is much weaker than that of the person’s own job, but why should it remain at all? What relevance to positions in a class order could parental position have?

We suspect that the main reason the father’s position affects class placement is that it also affects the chances for future advancement. We know from the status attainment research that the father’s occupation continues to have an effect on the son’s career mobility throughout the son’s working life. Sons with middle-class fathers are likely to advance further than sons with working-class fathers. Perhaps the middle-class self-placements of the sons (and daughters) of middle-class parents anticipate this greater future advancement. The son may not have a truly middle-class position yet, but he can more realistically anticipate holding such a position if his father was middle-class.

Affluence

Sombart’s best-known explanation of the lack of U.S. socialism was the affluence of the American worker:

This much is certain: the American worker lives in comfortable circumstances. He is well fed. . . . He dresses like a gentleman and she like a lady and so he does not even outwardly become aware of the gap that separates him from the ruling class. It is no wonder if, in such a situation, any dissatisfaction with the “existing social order” finds difficulty in establishing itself in the mind of the worker. . . . All Socialist utopias came to nothing on roast beef and apple pie. (1906:105–6)

Countless defenders of the American Way have echoed this answer (Gulick and Bers, 1953:528; Potter, 1954; Lipset, 1963:203 and 1979:25; Wilensky, 1966; Laslett, 1970:135, 302; Wattenberg, 1974), and even some who are not such staunch defenders (Bottomore, 1966:53–55).¹⁹

Two problems beset the affluence explanation. First, the facts about American workers’ prosperity may be incorrect. Once historians incorporated cost-of-living adjustments into their calculations, the position of American workers looked less privileged. Phelps-Brown and Hopkins (1950) found that in the crucial period for labor organization between 1860 and 1913, the real wages of the U.S. worker rose less than the real wages of Swedish, German, British, or French workers.²⁰

The second problem with the affluence explanation is more disturbing because it goes to the heart of our understanding of the causes of social protest. It turns out that social theorists have only assumed that affluence produces more conservative workers; rarely has anyone attempted to support the assumption with any evidence.²¹ Nevertheless, the idea flourishes as if it were self-evident. For example, liberals implicitly accept it whenever they prescribe a dose of economic development as the best inoculation against Third World revolutions. The assumption is that poverty causes unrest and that eliminating poverty will eliminate the unrest.²²

The fallacy in this theory, as Barrington Moore (1966) pointed out, is that poverty is as old as human society, but revolutions are rather rare events; indeed, they may be more common in the modern era, even though there is no evidence that poverty has grown. Moore reminds us of the obvious fact that revolutions are first and foremost struggles over power, it is the distribution of power, not the amount of wealth, that incites rebellion.

We have a similar objection to the affluence theory of American exceptionalism. Embourgeoisement arguments assume that affluence makes a difference—that workers who live comfortably enough will begin to see themselves as middle class even though they do not share any of the control over society that is characteristic of a true middle-class position. But this argument confuses status and power: affluence is a dimension of social status, and while it may be important for many things, it does not substitute for power. Workers will not, for the most part, be bought off by greater incomes; what makes them working class is their subordination, not their modest lifestyles. Again, one of Studs Terkel’s workers expresses this as well as any sociologist:

The almightly dollar is not the only thing in my estimation. There’s more to it—how I’m treated. What I have to say about what I do, how I do it. It’s more important than the almighty dollar. . . . I can concentrate on the social aspect, my rights. And I feel good all around when I’m able to stand up and speak for another guy’s rights. That’s how I got involved in this whole stinkin’ mess. Fighting every day of my life. And I enjoy it. (1974:189–90)

We can test the affluence argument in two ways: first, by taking a closer look at the role that income plays in people’s perceptions of their class position; second, by looking at the role of suburbanization. This is the only measure available to investigate directly the question of lifestyle—not just the amount of income earned but how it is spent to approximate a middle-class ideal. Fortunately, suburbanization has been a central item in the mythology of middle-class affluence. As Ben Wattenberg puts it in his celebration of the “massive majority middle class”:

For at least a quarter of a century, the idea of “middle class” in America has been associated with the idea of “suburbia.” . . . It should come as no surprise to see . . . a massive increase in the rate and number of suburban dwellers. . . . There is no other nation in the world where suburbia has become the plurality lifestyle nor where it is moving, apparently inexorably, toward majority status. The case can be made, in fact, that the suburban lifestyle is the first new and major residential life pattern to emerge since the rapid growth of the cities during the early years of the Industrial Revolution. (1974:105–7)

This typically overblown statement suggests that suburbanization has weakened the class divisions that characterized industrial urban America.²³ But there are no data on suburbanization’s effects; Wattenberg’s data document the growth of suburbs, not their consequences. What we need to discover is whether suburbanization in particular and affluence in general really make any difference for workers who are otherwise thoroughly proletarian.²⁴

In Chapter 4 we found that income is an important determinant of middle-class placements. In that sense the affluence argument is sustained: the more money one makes, the more likely one is to be middle class. But how important is income relative to class differences? Can workers earn their way into the middle class? Here the support for the affluence argument is more modest.²⁵ Figure 11.1 plots the proportions of middle-class placements separately for managers and workers (of equivalent education, occupational prestige, age, and so on). At all levels of income there is a substantial gap between managers and workers. In fact, workers making $16,000 would have to increase their income eight times before they had the same probability of being middle class as managers with $16,000 incomes. Thus, while higher incomes do tend to make workers more middle class, they do not make up for the basic class division that separates workers from managers.

Suburbanization does seem to have some effect on people’s perception of their class position, but the effect is so modest that it is hardly worth the fuss that has been made over it. Relative to equivalent city and rural residents, suburbanites are 4 percent more likely to see themselves as middle class. The difference is larger than we could attribute to chance alone, but it cannot be of much substantive significance. Suburban workers are about 41 percent middle class; urban workers are 37 percent middle class. Equivalent managers are 60 percent and 55 percent middle class. Suburban workers may appreciate the more open spaces and amenities of suburban life, but they are not fooled into thinking they are middle class.

Our analyses therefore remind us again of the importance of the distinction between class and status to an understanding of how Americans see their society. Status differences such as income levels and lifestyles may be important to the way Americans rank themselves along a scale from low to high status, but they are less significant in the way Americans determine positions across the class division separating the working class from the middle class. That division, like the division between both those classes and capital, is primarily a division based on power. The American Dream, insofar as it includes visions of comfort and affluence, is largely irrelevant to the course of class conflicts.

Notes

1. Among many possible expressions of this view, Tom Bottomore’s is typical: “America was the ‘land of opportunity,’ a vast, unexplored and unexploited country in which it was always possible, or seemed possible, to escape from economic want or subjection by moving to a new place, acquiring land or some other property, and adding to it by personal effort and talent” (1966:49). See also Lipset, 1960:253, 1963:193–233.

2. Usually, the causal linkage is implicit; only a few have attempted to formalize it into a general theory. The most interesting effort comes from the inventive economist Albert Hirschman (1970), who links mobility and revolt to the choice of Exit, Voice, or Loyalty. The discontented (those who reject Loyalty) are left with two alternatives: Voice (protest) and Exit (mobility). Hirschman, however, points out that Voice and Exit are sometimes complementary processes: workers seeking individual mobility may be precisely the ones who also lead collective protest.

3. However, a few European Marxists were impressed by American conditions. Wilhelm Liebknecht, a founder of the German Social Democratic Party, claimed that “generally the badly paid worker here [in the United States] is better off than our well-paid worker” (1887, cited in Moore, 1970:28–30).

4. Marx’s daughter Eleanor and her husband, Edward Aveling, reported on their 1886 trip to the United States: “At the one end of the scale is the millionaire, openly, remorselessly crushing out all rivals, swallowing up all the feebler folk. At the other end is the helpless, starving proletarian” (cited in Moore, 1970:31).

5. Melvyn Dubofsky has made the same complaint: “One major assumption is that the possession of property in the form of homes or savings satisfies individuals with their place in the existing social order. Another premise assumes that limited occupational mobility for the parent and somewhat greater opportunity for the children tends to the same effect. Both assumptions seem logical but remain untested” (1975:12–13).

6. “In American thought an ingenuous faith in the open road westward had long supported belief in an open road upward. The [1880s] cast a shadow over both ideas at the same time. A new sense of ‘closed space’ compounded the emerging fears of a closed society” (Higham, 1955:38).

7. Even earlier, Engels (Marx and Engels [1894], 1953:294) had endorsed a similar view: “Land is the basis for speculation, and the American speculative mania and speculative opportunity are the chief levers that hold the native-born worker in bondage to the bourgeoisie. Only when there is a generation of native-born workers that cannot expect anything from speculation any more, will we have a solid foothold in America.” In a letter to Sorge on January 16, 1895, he describes American class consciousness as zigzagging between “the mind of the industrial worker [and] that of the pioneering farmer” (Marx and Engels [1894], 1953: 270).

8. It also ignores the fact, as has long been noted, that not many eastern industrial workers ever migrated to the frontier (see Shannon, 1945). See the collection in Hofstadter and Lipset (1968) for other critiques of Turner. Some attempts to resurrect the frontier thesis have been based on the consequences for American economic growth (e.g., Simler, 1958; Murphy and Zellner, 1959; Karabel, 1979:214).

9. Further examination of the distinctiveness of the West reveals that only people who were raised there are more working class. Both managers and workers raised in the West are about 4 percentage points more likely to take a working-class label—a somewhat bigger difference than among current residents but still a trivial result.

10. Our GSS data confirm substantial geographic mobility—they show 56 percent of the sample living in places other than those they were raised in, 30 percent in different states—but not Thernstrom’s 64 percent turnover rate per decade.

11. “The extreme transiency of the urban masses must have severely limited the possibilities of mobilizing them politically and socially, and have facilitated control by more stable and prosperous elements of the population. Effective organization demands some continuity of membership, and this was glaringly absent among the poorest city dwellers of nineteenth- and early-twentieth-century America” (Thernstrom, 1973:232. See also Karabel, 1977:213.)

12. “The cause of this lack of psychological cohesiveness in American labor is the absence, by and large, of a completely ‘settled’ wage earning class. The opportunity of the ‘West’ has never ceased. In this vast country, several historical industrial stages are found existing side by side, though in demarcated areas. There is, therefore, the opportunity to migrate from older to newer and less developed sections, in which a person without much or any inherited property may still find the race for economic independence a free and open race” (Perlman, 1922:166). See also Lee (1961), who emphasizes all geographic migration as the true safety valve.

13. Thernstrom reports that for nineteenth-century Newburyport (1964) and Boston (1973), those workers who could not find good jobs tended to move out; residents well established in middle-class positions tended to stay where their jobs were. Dawley (1976:155) confirms this negative relationship of migration and status for nineteenth-century Lynn. Thus, for this analysis, it would be especially hazardous to project the results back to earlier times.

14. Lopreato and Hazelrigg (1972:115, 442–55) claim that “mobility is the crucial variable interfering with the formation of classes” and provide some evidence that upwardly mobile Americans are more conservative and Republican. Design problems confound their conclusions, however.

15. The simplest distinction to be made is between inter- and intragenerational mobility. Most studies have concentrated on intergenerational mobility—the changes between parental families and current positions. Intragenerational mobility—the career changes between first job and current position—has been less well studied, primarily for lack of adequate data. Our own research is forced into the same neglect, despite plausible arguments that the workers’ own chances for advancement—not their children’s—would have a more depressing effect on class conflict.

16. We have also computed all analyses of mobility with controls for the extraneous factors of age, education, income, and spouse’s education and occupation, so that we can look directly at the effects of occupational position and mobility between those positions. Those adjusted results are reported in the chapter appendix, Tables 11.A-11.C. Because of the additional controls, the mobility effects are more muted in these tables; nevertheless, the same general conclusions may be drawn.

17. The determination of upward direction requires an assumption about the status level of farmers, whose position is otherwise ambiguous in an industrial order. We have placed them at the bottom, below the unskilled-semiskilled workers—a position that is justified by the resulting class placements of their children.

18. An even more thorough test can be made by comparing how well the estimates for “father’s” and “own” occupation reproduce all the cells in Table 11.3. In fact, summing the simple effects of both occupations will almost exactly reproduce each of the cells of the table. No cell is any different from what might be expected as a result of the two components. Adding additional coefficients to represent the individual cells increases the chi-square a negligible amount.

19. Laslett repeatedly claims that affluence undermined labor militancy, but he rarely ties the historical facts together as cause and effect. For instance, he claims that the garment workers’ gains after the 1910 Protocol of Peace had a conservative effect “over the long run,” but then finds it difficult to explain why in the next union elections those workers chose an even more radical union leadership.

20. This was Kautsky’s (1905–6) conclusion at the turn of the century. Husbands (1976: xxiv) has also attacked Sombart for his description of U.S. working-class affluence. Moreover, Sweden had the most rapidly improving standard of living in the Phelps-Brown and Hopkins (1950) study, and yet Sweden had one of the most radical labor movements (Rosenblum, 1973).

21. David Brody (1968) argues that the labor peace of the 1920s supports the assumption. The welfare capitalism that flourished during the decade (e.g., company insurance programs) failed only after the Depression, when the corporations were no longer able to keep the benefits coming. However, it seems more reasonable to explain the 1920s by the dramatic defeats of major strikes at the beginning of the decade (the coal strikes, the steel strike of 1919), a situation that turned around only during the New Deal, when the government ended its hostility toward unions.

22. Again, theory would be improved if we experimented with the opposite statement. Columnist George Will (1985) provides a convenient formulation—what he calls “Will’s Law of Discontent” or the “Paradox of Prosperity”—that to us seems closer to the truth: “Discontent increases with the opportunities for acting on it. There is a lot of discontent going around among middle-aged people in the middle classes of affluent societies. These are people who have the ability to imagine other ways of living and the disposable income to act on their imaginings. A 13th-century peasant toiling from sunup to sundown behind an ox, in the shadow of a castle, tilling fields owned by the owner of the castle, never said to his spouse, ‘Hey, let’s chuck this and open a beer garden.’” Michael Harrington (1976:x), who probably agrees with George Will on little else, also questions the simple poverty-equals-unrest assumption. He points out that the great growth of the German Social Democrats before World War I coincided with working-class prosperity, and the student New Left in the 1960s was based on prosperous youth. And Howard Wachtel (1974:109, 119) argues that younger workers are now more militant because their incomes are more secure than were those of their Depression-burdened parents.

23. Eli Chinoy (1955:126) also emphasizes that possessions, particularly a home, convince his autoworkers that they are moving forward and thus have no need of collective protest. See also de Tocqueville [1835], 1954 [vol. 2]: 270; Lipset, 1960:269; and Lane, 1962: 80, 250.

24. Radical analyses (e.g., Parenti, 1978:101) have also blamed consumerism in the working class for falsely allying workers with property-owning capitalists.

25. We also do not know much of the income effect may be a proxy for differences in class position not captured by our three class indicators. In Chapter 4 we noted the weakness of the class indicators, especially the measure of authority. How much of the difference between a $30,000 truck dispatcher and a $15,000 truck dispatcher might be unmeasured differences in the class position of the two jobs (e.g., greater authority or more planning duties)?