Time for Things Labor, Leisure, and the Rise of Mass Consumption

Olney argues that the growing expenditure on consumer durables cannot be explained by increasing incomes in conjunction with decreases in the prices of those goods. Comparing prices in the 1920s to those of the pre– World War I era, she finds that consumer durables were actually relatively more expensive in the later period than other goods and services. So the evidence suggests that there was a real shift in demand for consumer durables in the post–World War I period. What sense are we then to make of this change in consumer behavior? Olney suggests, like Avner Ofer, that we should think of the accumulation of consumer durables as a form of saving. On this interpretation, acquiring a consumer durable is an act of investment in an asset, akin to investments in fixed capital undertaken by firms. Consumers bank the utility-generating potential aforded by their artifacts for future use. Olney thus follows Becker (1981) in conceiving of households as utility-producing factories, with consumer goods as their working capital. Yet this analogy only highlights the oddness of the underlying pattern. It would be perverse for a firm to buy a piece of capital equipment and then leave it unused (or underutilized) while its value is inexorably eroded by depreciation. Also, in general it makes sense for firms and households alike to keep a good amount of capital in its fungible money form, as cash reserves, since future needs and wants cannot be easily predicted.45 Moreover, holding capital in the form of assets incurs a penalty if, as a result of an unforeseen calamity, a firm or a household is forced to sell those assets at distress sale prices. Under specific market conditions—for example, when people expect a shortage of some good, or under conditions of very high inflation—it might be a good idea to stockpile capital goods and materials. However, although there have been intermittent periods of high inflation in the United States— for example, in the period immediately following World War II and in the 1970s—on the whole the rate at which prices have increased over the century has been fairly gradual. Nor is it the case that hoarding consumer durables can be explained by fears of future scarcity. There were shortages during the wars, which led to price controls, and distributional problems during the Depression, but apart from those periods, America has been for a majority of its citizens, and for much of the twentieth century, a “land of plenty,” with low inflation and an abundance of goods. Given these historical conditions, the explanation given by Olney for the accumulation of consumer durables seems questionable.

72

Time for Things

The evidence thus indicates that consumers began in the early twentieth century to accumulate durable goods, and they did this in lieu of investing money or working fewer hours. But what kinds of goods did they acquire? Three categories immediately stand out: automobiles, houses, and appliances. The importance of car ownership in the first half of the twentieth is clear in Robert and Helen Lynd’s (1929, 1937) ethnographic research. The Lynds found ample evidence of the new obsession with cars in their studies of “Middletown” (Muncie, Indiana). Time and again, the Lynds’ informants speak of their obsession with owning and driving automobiles. During periods of hardship, the Lynds were informed, the very last expenditure to be cut back on is gas and other car-related expenses: “I’ll never cut back on gas! . . . I’d go without a meal before I’ll cut down on using a car,” one informant tells them (1929, p. 63).46 The Lynds’ impression is borne out by aggregate data. Between 1916 and 1926, car sales tripled from 1.6 million to 4.3 million per year. The proportion of the population owning cars between 1910 and 1930 increased from 1 percent in 1910 to 26 percent in 1920, and by 1930, it had reached 60 percent. According to the 1929 census of distribution, Americans spent 18.08 cents of every dollar on automobiles and their upkeep—as much as on food. There are of course good practical reasons for owning a car given the progressive dispersion of urban form in modern America, with the growth of suburbs, in conjunction with a generally inadequate public transportation infrastructure. Yet this poses a chicken-and-egg question: Which came first, the taste for automobiles or the reorganization of social space such that they became necessary? I will consider the significance of automobiles as consumer durables at greater length in Chapter 5 of this book.

A second consumer durable that showed dramatic growth over the twentieth century was housing.47 Houses are not always immediately thought of as consumer durables, because they are often regarded as investment assets. However, except for a fairly small class of speculative real estate investors, houses are, for most people, primarily consumer durables. At the beginning of the century, only 36.5 percent of Americans owned their own homes. By 1960, homeownership had spread to 62 percent of the population. The stock of rental accommodation actually shrank in the post-1945 era as Americans collectively pursued the dream of homeownership. At the same time, in the postwar period there was a huge building boom. A quarter of the housing stock in the United States after 1960 was built in the 1950s (Cohen 2003, p. 122). In part, the switch to homeownership was in response to significant federal subsidies, in the form of income tax deductions for mortgage interest and property taxes, the origins of which date back to the tax reforms of 1913 (Prasad 2012).48 Yet even given homeowner-friendly tax policies, it is dificult to make sense of this striking trend in terms of the narrow economic benefits of homeownership. Home prices track markets more generally, so it is unclear why investing in housing makes more sense than other kinds of investment. Indeed, some studies indicate that renting and investing, as opposed to buying and paying a mortgage, is a superior strategy for wealth creation (Beracha, Skiba, and Johnson 2017). Furthermore, the United States has tended to have high levels of geographical mobility. Although we lack good time series data for geographical mobility prior to 1935, more contemporary data indicate that more than one-third of all home buyers move within five years, which is generally too soon for them to have been able to build enough equity to cover transaction costs.49 Under those conditions renting housing would seem to make more sense than buying it. More generally, switching from renting to owning, for many, corresponds to a decline in the quality of life, as households become highly leveraged—house rich and cash poor. The strength of the preference in the United States for home buying is, therefore, quite puzzling.

Empirical Pattern in the United States

73

At the same time as homeownership became increasingly prevalent over the course of the twentieth century, the size of the homes occupied by Americans got progressively larger. Between 1950 and 2006, the average home size in the United States doubled from around 1,200 square feet to close to 2,400 square feet. This was largely facilitated by suburbanization, which opened up cheap land for development. Yet it is tempting to interpret the growth in the size of American abodes as connected to the increasing accumulation of material durables, as larger homes were required to store and organize increasing quantities of things. Homes, seen in this light, are commodities used for the storage of other commodities.

Another category of consumer durable the demand for which took of in the early part of the century was domestic appliances. One possible explanation, touched on earlier in the chapter, for the failure of the working week to fall as wages increased, is that the commodities and services acquired with the wage had the efect of reducing necessary labor time outside work. So according to this line of reasoning, as income went up, the amount of time spent on unpaid domestic labor went down, with the net result that the amount of truly free time (left over after all necessary labor, paid and unpaid) per household increased. This explanation would seem to solve our puzzle. If increasing income at constant hours of work has the efect of reducing total household labor hours (domestic plus paid work) because the income is used to acquire domestic appliances that save time, then it might make sense to choose money over less paid work time so that more and better labor-saving devices can be acquired. However, it is not at all clear that the time-saving efect of commodities is suficient to account for the degree to which increasing productivity has not been used to reduce paid work time. As discussed previously, it is questionable whether the use of new domestic appliances significantly reduced time spent on housework, as opposed to increasing standards of cleanliness and order within the household (Cowan 1983). Moreover, research on the rates of difusion of diferent kinds of consumer commodities casts some doubt on the notion that the acquisition of time-saving consumer durables explains the preference for earnings over free time. Economic historians Susan Bowden and Avner Ofer (1994) compared the spread of time saving and time using consumer durables in the United States and the United Kingdom during the twentieth century and found that in both societies, time-using commodities difused significantly more rapidly than time-saving commodities. Consumers evidently directed their discretionary spending first at products whose use demanded, rather than saved, time. The prioritizing of time-using durables makes it implausible that any preference workers had for more income over fewer hours can be explained by the use of that income to acquire appliances and other products that free up more time.

74

Time for Things

The US case thus illustrates the kind of pattern that is quite typical of high industrial capitalism. After an initial period during which working hours decreased, further increases in productivity were relentlessly channeled into an expansion of output, while the rate of decrease in the length of the average working week diminished almost to zero. The pattern of trends in productivity, consumer spending, and hours of work described in this chapter is certainly hard to make sense of within the broad set of assumptions that neoclassical economists bring to their analyses. At the same time, explanations based on behavioral economics, class power, identity-making, or status-seeking, while making sense of certain aspects of the pattern that emerged in twentieth century capitalism, fail, for reasons set out in the last chapter, as full explanations for that pattern. An alternative approach, which I will take up in the next chapter, is to make sense of these trends in terms of changes in the normative framework underpinning wage-labor exchange. These afected the meaning, and especially the justification, of work under capitalism in a way that had significant consequences for attitudes toward wage goods and consumption.

Empirical Pattern in the United States

75

Before pressing forward with the development of a theory along those lines, as an addendum to this chapter some attention should be paid to the question of whether the US pattern of productivity and work time is, in fact, atypical among capitalist societies and best explained in terms of national traits. Some analysts have suggested that the slow rate of decline in work time in the United States during the afluent postwar years reflects the cultural and historical idiosyncrasies of America, those that make it “exceptional” when compared to the rest of the capitalist world. Tibor Scitovsky (1992) and, to a lesser extent, Juliet Schor (1991) both make this kind of argument. Indeed, Scitovsky (who was by background a European aristocrat) suggests that the solution to what he regarded as a peculiarly American problem of overwork and overproduction is for Americans to emulate European practices of self-cultivation and savoir vivre. However, it is not clear how much of the diference in work hours between Europe and America is a result of a diference between the two in preferences for free time, one ultimately attributable to “culture,” and how much is a result of other factors. For example, during economic downturns, US firms typically shed workers. Those who remain work the same number of hours or sometimes even more, as they are forced to do some of the work of those who were made unemployed. By contrast, in Europe the response to economic downturns has, in the postwar period, been more likely to have involved a reduction of hours per worker rather than in numbers of employees. Reduced hours under those conditions clearly do not indicate a preference for leisure but rather a stronger commitment in Europe to a corporatist welfare capitalism that extends special protections to those in work (Esping-Andersen 1990). Similarly, in France the thirty-five-hour workweek was introduced in 2000 primarily as a way to address the problem of structural unemployment, by redistributing work.50 At the same time, marginal income tax rates in Europe have tended to be higher than in the United States, which decreases the incentive for working longer hours. These diferences in labor market regulation, industrial policy, and taxes have the efect of depressing work hours in Europe as compared to the United States.

76

Time for Things

Taking these, and other, factors into account, Robert Gordon (2010) estimates that only one-third of the diference in work hours between the United States and Western Europe can be attributed to diferent preferences for leisure. The remainder Gordon considers to be “involuntary leisure,” forced on European workers by conditions beyond their control.51 This would suggest that those who have used diferences in work hours between Europe and the United States as evidence of cultural diferences in attitudes to work and free time have been overstating their case.

Moreover, while there are certainly diferences between the work time regimes in Western Europe and the United States, the trend in both places is similar: increasing productivity has predominantly been used to increase output rather than to reduce work time. Although Western Europeans in the late twentieth century on average worked fewer hours than Americans, the diference between the two, relative to the huge increase in productivity in both regions, is not great. European productivity improved at roughly the same rate as American productivity in the postwar period, yet even taking length of vacations into account, the working week in Western Europe was longer than in the United States until at least the mid-1970s. In 1970, average hours in the fifteen countries that made up the European Union before its eastward expansion in 2004 were actually marginally higher than they were in the United States (Blanchard 2004).52 In the mid-1970s, work hours were between 10 and 15 percent longer in Britain, France, and Germany than they were in America (Prescott 2004). It is true that by 1987, hours were 15 percent shorter in Europe than in the United States, and by 2000, the gap had increased to 23 percent (Maddison 1991; Blanchard 2004).53 Yet since hours of work are calculated by dividing total hours worked by number of workers, the greater prevalence of part-time work in Europe distorts the picture.54 Some of the diference in work hours toward the end of the century can be accounted for by underemployment in the EU—a consequence of the dual labor market structure created by European corporatism, which secures stable employment for some, while consigning others to permanent casual, part-time labor. Recent data from Eurostat (2017) indicate that full-time hours per week have been very similar in Europe and in the United States since 2006—around forty-one hours in Europe and forty-two hours in the United States. Furthermore, according to one study, once hours of domestic labor are taken into account, work hours in a sample of three European countries (Germany, Italy, and the Netherlands) between 1985 and 2003 were at most only 12 percent lower (at the end of that period) than in the United States (Burda, Hammermesh, and Weil 2006). Americans spent more time working in the market than Europeans, and Europeans put more time into domestic labor, while the total time Americans and Europeans gave to labor (market plus domestic) was not that diferent.55 Some of the diference in market work hours between Europe and America thus reflects the tendency for domestic labor to be done in-house in Europe, while it tends to be outsourced through the market in the United States.56

Empirical Pattern in the United States

77

In sum, the contrast between leisured Europeans and workaholic Americans, to the extent that it has any empirical basis, is essentially a feature of the last quarter of the twentieth century and the years since the turn of the twentyfirst century. Moreover, the glaring disparity during the twentieth century between growth in productivity on the one hand and the rate of reduction of work time on the other holds in Europe as well as in the United States, albeit to a slightly lesser extent in the years toward the end of the twentieth century.57 More striking than any diference in work time between the United States and Western Europe is the failure of work time in both places to fall to a greater extent than it did over the course of the twentieth century. Keynes’s 1930 prediction, that the industrial world would witness a radical contraction of labor time, is not much closer to being realized in Europe than it is in the United States.58 It would seem that there is a quite general bias in capitalism toward turning improved productivity into more output rather than more free time.

78

Time for Things

FOUR

A Theory of Mass Consumption as Wage-Labor Commensuration

One way to make sense of the pattern of work and consumption described in the previous chapters is in terms of the changing experience and understanding of wage labor—in particular, by paying attention to the conditions under which wage labor came to be seen as a fundamentally legitimate form of exchange. Wage labor is, after all, the dominant feature in the organization of the economy under capitalism, engaging most of the population for much of its waking life, while consumers are almost always also wage earners. In order for wage labor to become broadly acceptable as a potentially fair economic practice, work had to be plausibly commensurable with its wage.1 This is because commensurability is a prerequisite for a fair exchange of labor for a wage. Commensurability between labor and wage became possible in the context of an approach to commodities that unhinged their use value from the concrete contexts in which those commodities get used, so that use value becomes to a significant extent hypothetical rather than something closely tied to the practical realities of particular lives. This hypothetical understanding of object utility could help explain the seemingly perverse patterns of consumer commodity accumulation and work time that is characteristic of advanced industrial capitalism.

The early history of capitalism was characterized by a good deal of political struggle and ideological contention over wage labor as a form of exchange. The idea that free activity could legitimately be exchanged for a wage was anathema to large sections of the labor movement, and in the earlier stages of industrial capitalism, both in the United States and in Europe, wage labor was subjected to widespread criticism and condemnation.2 As industrial capitalism advanced, however, the exchange of labor for a wage increasingly came to be accepted as part of the natural order of things. The focus of critical attention began to shift from the legitimacy of capitalism itself, and especially of the institution of wage labor, first to conditions of work, such as the length of the working week, and then to levels of compensation. Over time, wage labor became construed as a potentially fair exchange, with much efort put into thinking about the criteria by which its fairness can and should be gauged.

There have been various ways of thinking about what constitutes a fair wage, and various schemes for implementing diferent conceptions of fair compensation. For some economists, fair wages are just those delivered by the untrammeled operation of market forces. The marginalists, for example, held that, under free market conditions, wages reflect the value of labor’s contribution to the product. Because, according to this view, wages simply measure the value labor adds to output, they are fair by definition. For some labor activists and radical political economists, by contrast, fair wages are those that capture the full value of the product, on the grounds that only labor creates value. The way in which labor has been used to calculate wage rates in practice has varied across time and place. Compensation has been figured, for example, as a rate for discrete physical actions per unit of work time or, alternatively, in terms of labor’s material output, in the form of piece rates (Biernacki 1995).

What the diferent ways of thinking about and determining fair compensation for work under capitalism have in common, however, is a conception of fairness as objectively measurable. That is, fairness is regarded not as a matter of convention or custom but, rather, as an objective fact about an exchange, rooted in the properties of the substances being exchanged. An exchange is objectively fair, in this view, when a rational observer, without knowledge of the specifics of the set of norms governing the exchange, would judge it to be so. For example, the exchange between feudal lord and vassal, free peasant, or serf might be entirely in accordance with custom and therefore conventionally proper. A modern observer, however, would be unlikely to judge feudal exchange as fair, because of the apparent disparity in the services exchanged (not to mention the degree of coercion involved) and the apparently arbitrary nature of the underlying convention. Similarly, modern conceptions of fair economic exchange are quite diferent from the customary rates of pay described by historians as typical of the early industrial period (Hobsbawm 1964). This is not to say that within a customary framework, the subaltern party would necessarily consent to what would appear to modern eyes to be an unequal exchange—feudal peasants, for example, certainly expressed discontent, and not infrequently rebelled. But resistance tended to be justified in terms of the violation of other conventions and customs. For example, in late medieval and early modern England there was a counterhegemonic discourse, deployed against both feudalism and nascent capitalism, that asserted the customary rights of “free-born Englishmen.”3 Similarly, in England, enclosure was resisted on the basis of customs established “since time immemorial” and “beyond memory of any man,” which gave commoners right of access to resources such as grazing lands (E. P. Thompson 1993; Wood 1997).4

80

Time for Things

By the later nineteenth century, the very general principle of objective fairness had become central to the understanding, both on the left and on the right, of what it is that makes an economic exchange morally acceptable.5 Clearly however, convergence on the ideal that work should receive objectively fair compensation did not mean that there was consensus on a particular definition of fair wages. On the contrary, the inherent vagueness of the notion of a fair wage resulted in a proliferation of diferent interpretations of the idea. The dificulty and vagueness of the idea of fair exchange of labor for a wage was a consequence of the problem of commensuration.

If fairness is an objective fact about an exchange, determined by the ratio obtaining between the exchanged substances, then those substances must be commensurable. That is, there must be some nonarbitrary procedure for translating the quantity of goods or services flowing in one direction into the quantity of goods and services flowing in the other. If the substances being exchanged are not commensurable, then clearly it would not be possible to “objectively” judge whether that exchange is a fair one. In order for fair exchange to be possible there must, therefore, be some metric by virtue of which the exchanged substances can be seen as standing in some definite quantitative relationship to one another. Moreover, if an exchange is to be objectively fair, rather than just fair according to convention, the metric must measure some essential quality the exchanged substances have in common.6 The notion of an objectively fair exchange, as opposed to an exchange that is fair according to convention, thus involves the presupposition that the substances being exchanged share qualities such that they are commensurable.7

A Theory of Mass Consumption as Wage-Labor Commensuration

81