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

Part of the appeal of owning a consumer durable is, then, the prospect of possessing a notional quantity of use value surplus to likely actual use, therefore indexing a surplus of free time. Indeed, the activity time indexed by a set of goods can potentially exceed the absolute amount of time remaining in a person’s life, thereby perhaps suggesting that finitude has been transcended. In owning objects, possibilities seem to stretch beyond those realizable under average conditions of real use, and this makes the work-time-consumption regime typical of high industrial capitalism at least acceptable and perhaps even quite appealing.

1.5. Explains the Correlation among Saving, Spending, and Labor Decommodification across Diferent Capitalist Societies

Moving to a very diferent, much more macrohistorical and comparative, level of analysis, interpreting mass consumption as wage-labor commensuration suggests an explanation for the correlation between the degree to which labor has been decommodified (measured by the extent of development of the welfare state) and savings rates across the modern industrial capitalist world. It seems intuitive that capitalist nations with less developed welfare states should, in lieu of dependable social insurance, have higher savings rates than those with more developed welfare states. The larger the gaps are in a social safety net, the more reason individuals have to save money, in case they should fall on hard times. In fact, however, the countries with the weakest welfare states have the lowest saving rates. This is a surprising finding—yet one that is consistent with the commensuration theory.

In his important book, Three Worlds of Welfare Capitalism, Gosta Esping-Andersen (1990) categorizes capitalist societies according to the extent of decommodification of labor produced by state policies.26 Esping-Andersen defines decommodification as the mitigation of pressure on people to sell their labor in the market. In providing alternative means of subsistence, welfare states lessen their population’s dependence on the labor market, in efect, making capitalism less coercive for labor. The more dependent people are on the labor market in a given society—that is, the fewer alternatives they have for getting a living apart from selling their labor—the less decommodified labor is in that society.

102

Time for Things

Esping-Andersen uses a variety of measures to create a composite score indicating the extent to which diferent nations have efectively decommodified labor. He finds that countries can be grouped into three categories, characterized by particular histories of policy orientation. The three categories, in ascending order of level of decommodification, are liberal welfare states, corporatist welfare states, and social democratic welfare states. The liberal welfare states are the Anglo-Saxon nations, the corporatist states are those of continental Europe (excluding Scandinavia and including Japan), and the social democratic set comprises Scandinavia (excluding Finland), Austria, Belgium, and the Netherlands.27

From a standard economic viewpoint, stronger welfare states—those that to a greater extent decommodify labor—should have lower saving rates and therefore higher rates of spending. This is because the presence of a strong safety net lessens the need to save defensively. Although the correlation between saving / spending and the extent of welfare provision is not an issue that Esping-Andersen investigates, data from other sources indicate that saving rates have in fact been lowest in the liberal states—those that are the least decommodified according to Esping-Andersen. In the Euro area since 1991, for example, household saving rates have been between two and a half and three times higher than in the United States, a member of Esping-Andersen’s least decommodified group, while saving in Japan (categorized as part of the corporatist group) has been between four and six times higher. Net household saving rates between 1985 and 2009 in the countries that make up Esping-Andersen’s liberal group, which is the least decommodified, are consistently among the lowest in the OECD (Garon 2012). Comparing the subset of nations that are members of the OECD from each of Esping-Andersen’s groups, average saving rates between 1985 and 2009 are 11.6 percent for the corporatist group, 9.6 percent for the social democratic group, and 5.6 percent for the liberal group (see Appendix 2 for table of OECD figures).28 While the corporatist and social democratic welfare states have quite similar levels of saving, the saving rates of the liberal welfare states—those with the least in the way of a safety net—are significantly lower.29

Once more, this pattern is surprising, as there would seem to be more reason to save in contexts in which people are more dependent on the labor market for their economic survival. Yet it becomes less surprising when interpreted in terms of the theory of consumption as wage-labor commensuration. If the commensuration theory is correct, then the pattern would follow as a matter of course, and the grounds for surprise would be removed. The theory entails that consumer materialism—defined as the propensity to turn increasing hourly wages into increased consumption, rather than reduced work time or increased savings—should be positively correlated with the extent to which labor is commodified (or, in Esping-Andersen’s terms, not decommodified). The reasoning for this goes as follows:

A Theory of Mass Consumption as Wage-Labor Commensuration

103

I. The extent to which labor is commodified in modern industrial capitalist societies is, in part, a reflection of the degree to which wage labor is seen as a legitimate form of exchange. The less legitimacy wage labor has, the more pressure there will be on the state to provide economic support independent of labor market participation. This makes subjection of individuals to the labor market more conditional (on being able-bodied and of sound mind, on general macroeconomic conditions, etc.).

II. According to the commensuration theory, seeing commodities as commensurable with labor contributes to the legitimacy of wage labor as a form of exchange. In order for the wage to become commensurable with the work that is given for it, it needs to be spent on commodities. Kept in money form, the wage is too abstract to facilitate commensuration.

III. Therefore, the degree to which labor is commodified (or, in Esping-Andersen’s terms, not decommodified) should, ceteris paribus, be positively correlated with rates of consumer spending and negatively correlated with saving rates. Higher rates of consumer spending should go along with more commodified labor. Capitalist nations with developed welfare states, which to some degree decommodify labor, should, therefore, have lower rates of spending and higher rates of saving.

So, where the institution of wage labor has greater legitimacy, people will tend to be more commodified (that is, subjection to the labor market will be less conditional). At the same time, the means by which that legitimacy is secured—the accumulative consumption of time-indexing goods—results in higher rates of spending and therefore lower savings rates. Conversely, it follows from the commensuration theory that decommodification should be inversely correlated with spending rates (and, by default, positively correlated with household saving rates). And, as we have seen, cross-referencing Esping-Andersen’s empirical analysis with comparative data on rates of saving and spending reveals a pattern that provides some support for this prediction.

104

Time for Things

The connection between the rise of wage labor and the increasing propensity to spend can also be found going further back in history. If the commensuration theory is correct, then, as labor becomes more commodified (assuming this is because wage-labor exchange becomes more legitimate), so spending rates should increase. And indeed, at least according to economic historian Jan De Vries (2008), the historical process of the commodification of labor was accompanied by a greater propensity to consume.30 However, where the labor process is to a greater extent under the control of workers, while at the same time what is being sold is the product of labor rather than labor power—as was the case in the protoindustrial phase of capitalist development—the economic exchange is somewhat diferent. Since, under those conditions, workers in theory retain ownership of their time, there is not the same imperative to turn increases in hourly wages into more wage goods (construed as an accumulation of congealed free time), as opposed to increases in actual free time.31 So in earlier periods of capitalism, before the full commodification of labor power, a more balanced attitude to the tradeof between free time and money was likely to prevail.

1.6. Accounting for Certain Findings of Behavioral Economics

Changing levels of analysis once again, another way in which the theory of consumption as wage-labor commensuration sheds light on modern economic behavior is its capacity to explain some of the more perplexing experimental discoveries of behavioral economics. As mentioned in the second chapter, over the past three decades research in behavioral economics has led to experimental findings that challenge the homo economicus model of the economic agent. The rationality presupposed in that model is not easily compatible with findings like hyperbolic discounting, the endowment efect, and nontransitive preferences, which behavioral economists have suggested are widespread features of economic behavior. Behavioral economists and psychologists have tended to interpret their findings in an ahistorical fashion, as indicating general patterns in human behavior. Daniel Kahneman, one of the founders of behavioral economics, takes the various biases and heuristics discovered in the field to be psychological primitives, natural facts about the human species that, as such, are universal in scope, explicable in terms of the selective advantages they conferred in our evolutionary past. In a synthetic theoretical work, for example, Kahneman (2011) explains the disparate findings of behavioral economics through a theory of the architecture of the human brain.32 Yet surely it is premature to leap so quickly to such a general level of explanation, without first considering the mediating influence of sociohistorical context. The experimental subjects of behavioral economics are, after all, almost always drawn from modern capitalist societies. It is not unreasonable, therefore, to interpret the discoveries of the field as reflecting the more proximate sociocultural and historical setting of the subjects on whose behavior they are based. Indeed, two prominent examples of supposedly irrational behavior discovered by behavioral economists are quite compatible with the theory of economic behavior under capitalism developed in this book.

A Theory of Mass Consumption as Wage-Labor Commensuration

105

The first of these is the so-called endowment efect. One of Kahneman’s most famous experimental discoveries is the tendency for subjects to value more highly an object that is owned than the same object when it is not owned. The mere fact of possession seems to add value to an object for its owner. Kahneman and his collaborators call this the “endowment efect” (Kahneman, Knetsch, and Thaler 1990, 1991; Knetsch 1989; Thaler 1980). Within behavioral economics the commonly accepted explanation of the endowment efect is just that human beings are hardwired to feel aversion to loss. The endowment efect is interpreted as indicating a marked asymmetry between the response of subjects to prospective losses and prospective gains. In essence, people dislike losing a given valued thing much more than they like gaining the same thing. However, it is not at all clear that the theory of loss aversion does much more than ofer a diferent way of describing the behavior it purports to explain.

The theory of consumption as labor commensuration, by contrast, suggests a substantive sociological and historical explanation for the endowment efect. The endowment efect can be explained by the tendency, in a society dominated by wage-labor exchange, for owned objects to become saturated with the sunk costs represented by the labor exchanged (and, therefore, free time given up) for the money used to acquire them. If acquiring commodities is, in part, an efort to make the wage equivalent to the free activity given up in exchange for it, and if that free activity is inalienable and qualitatively particular, then the characteristics of free life activity (sacrificed in order to work) might well get transferred to the wage goods received as compensation for its loss. Coming to own commodity objects by exchanging a wage for them then gives those objects a particular meaning by virtue of the particular bits of life exchanged for them. Wage goods, once possessed, thus become inscribed with the sacrifice represented by the work that went into acquiring them. To put it another way, if labor is indeed inalienable, then that inalienability might to some degree get transferred to the objects received in exchange for labor. Those objects then become relatively inalienable, such that their owners will be reluctant to part with them. That then gets expressed in the higher value owners place on them, compared with identical goods that are not owned—Kahneman’s endowment efect.

106

Time for Things

A second important finding from behavioral economics that the theory of consumption as commensuration helps explain is the tendency, discussed in Chapter 2, for subjects to discount the utility of future rewards hyperbolically as opposed to at an exponential rate. To recap, in studies of intertemporal choice, behavioral scientists have found that subjects tend to place very high value on a reward expected in the short-term, after which its value is steeply discounted. However, the same reward, with an equivalent delay, when expected in the longer term, is discounted to a much smaller degree. In terms of consumer behavior, hyperbolic discounting implies a tendency to spend in the short term as opposed to saving strategically (to “smooth” consumption over the long term), since the utility attached to satisfying an immediate desire dwarfs the envisioned utility of satisfying more temporally distant wants. When graphed, the discount function that describes this valuation of rewards over time produces a hyperbolic curve.

Hyperbolic discounting has been a source of great puzzlement to behavioral scientists and philosophers, as it would seem to violate certain canons of rationality.33 Rational consumers are supposed to think about the future mainly in terms of the utility cost of delayed consumption, modified by degrees of uncertainty. Economists assume that utility is discounted over time at an exponential rate, producing a discount curve that is much less bowed than a hyperbolic one. Similarly, although more distant points in time are more uncertain than closer ones, assessments of uncertainty over time should, other things being equal, be much more consistent than is suggested by the hyperbolic function that describes much economic behavior. If hyperbolic discounting simply reflects increasing degrees of uncertainty about the future, then the uncertainty attributed to the future would increase very rapidly over short intervals of time, then, oddly, flatten out over the longer term. It thus seems implausible that hyperbolic discounting has to do with the epistemic status of the future.34

A Theory of Mass Consumption as Wage-Labor Commensuration

107

At the same time, agents, especially institutions, do sometimes discount utility in a supposedly rational, exponential fashion. For example, when people save, the usual expectation is that interest on savings (that is, the returns for forgoing consumption) will be paid at an exponential rate. So, it would seem that humans are endowed with two discrete discount curves— one hyperbolic and the other exponential. Behavior in some contexts is best described by the exponential curve, while in others the hyperbolic curve gives a better fit.

As mentioned in chapter 2 of this book, one strategy to explain this behavioral complex, advocated by George Ainslie (1992, 2001), has been to posit multiple competing selves within the agent, corresponding to distinct neurophysiological structures in the brain, each endowed with a particular orientation toward the future, reflected in its own discount curve. The behavior of a person is then a kind of equilibrium outcome, resulting from competition between these sub-self-level actors, at what Ainslie (1992) describes as the “picoeconomic” scale. Ainslie’s theory, like Kahneman’s, seeks maximal scope, and is consequently very insensitive to historical context. The objective is to come up with a robust theory of human behavior, understood on the most general of levels. Yet, as noted in the discussion of Kahneman, the experimental subjects whose behavior such theories attempt to explain are invariably subjects living in, and formed by, modern capitalist societies. We might do well, therefore, to look for an explanation of their behavioral tendencies in the proximate historical context of modern capitalism.

It seems quite possible that the high value people place on spending in the short term, compared with spending over the medium or longer term, reflects a sense of urgency, in a society organized around wage-labor exchange, about converting money into use values. Money wages form the most important component of income for the vast majority of people under capitalism. Consequently, for most people, money mostly means wage earnings. The disposition to privilege short-term consumption, rather than smoothing consumption over time, might well, therefore, have at least something to do with the identity of money wages as compensation for work. If commensurate compensation for labor requires that concrete potential activity be returned to the worker in the form of commodity objects, then there will be a strong impulse to spend earnings (and possibly also make use of credit, if wages are deemed insuficient to do the job of commensuration) to acquire such objects. Accumulating money, as opposed to things, by comparison works quite poorly as commensurable compensation for labor, even where the idea is to save in order to fund later consumption. Of course, people can fantasize about what they could do with their wage rather than spending it. Contemplation of the potential use values that a money wage can aford might for some time stave of the urge to convert that wage into commodity objects, especially if wages are being saved with some particular product in mind, such as a new home or car. But the use value of money wages per se has a second-order quality about it. The usefulness of unspent money subsists in the set of use values it can in theory command, which is, in efect, a potential to command stores of potential use value. For that reason, fantasies about how money could be spent are ultimately not concrete enough to be efective in making the exchange of labor for a wage seem like a fair exchange of commensurable substances.35 In order for a sense of commensuration to be achieved, the wage must be converted into actual stuf, endowed with use power. Wage earners will, therefore, tend to turn money wages into things, and do so sooner rather than later.

108

Time for Things

As an alternative to the quite radical theoretical move of positing multiple sub-self-level agents—biologically ingrained homunculi competing for control of the person—we can thus account for the behavioral complex described by Ainslie and others in terms of overarching institutional and normative logics within capitalism. The tendency to discount hyperbolically makes sense when seen in the context of the logic according to which wage labor is justifiable. The tendency to discount exponentially, on the other hand, expresses the rationalized, calculative dimension of modern capitalism, most famously described by Max Weber (1992, 2013). If this interpretation is correct, then rather than holding that the two kinds of discounting reflect the operation of discrete cognitive systems in the brain, we can connect each to behavioral norms associated with a particular social position. To the degree that the logics underlying each rate of discount might have some normative claim on the behavior of economic agents, we can expect to observe the behavioral inconsistencies that Ainslie interprets as evidence of the existence of multiple selves. In some contexts, for example when investing profits, the exponential style of reasoning is dominant, while in other contexts, for example when deciding what to do with money wages, the hyperbolic one prevails.

A Theory of Mass Consumption as Wage-Labor Commensuration

109

1.7. Explains Pattern of Consumption over the Life Course

The labor commensuration theory of mass consumption also works as an explanation for the distribution of consumption over the life course. As noted earlier in the chapter, and also in Chapter 3, the prediction of the life cycle hypothesis—that people will attempt to maximize utility across the life course by becoming indebted when young and poor, saving when at the height of their earning power, and then spending after retirement—is not borne out by the evidence. The actual empirical pattern is as follows: When young, people underconsume relative to the predictions of the theory, because they are resistant to borrowing. When they are old, they also underconsume, because they spend a smaller part of their income than they should do, according to the theory. When they are middle aged, at the peak of their earning power, they overconsume (that is, undersave) (Deaton 1994). The peculiar thing, therefore, is the arc of spending over the life course—which would seem to indicate irrational risk aversion in youth, irrational risk taking in middle age, and irrational risk aversion in old age.

This pattern, which economists find puzzling, fits well with the theory of consumption as wage-labor commensuration. The young are reluctant to borrow in order to spend because they are much more likely to be in fulltime education or training.36 Even if they are working, their work is more likely to have strong element of on-the-job training (as formal or informal apprenticeship) and, for that reason, is something other than a straightforward wage-labor exchange. That people spend, rather than save, when middle aged, at the height of their earning power—when they have fully settled into the workforce—can straightforwardly be accounted for by the felt impulse to make their wage equivalent to the work done for it, by converting it into commodities. Moreover, the impetus to borrow, in order to finance consumption, takes on more force when people are in full-time work and, in particular, where wage levels seem insuficient to compensate for work. The old spend less than is predicted by the life cycle hypothesis because, as suggested in the discussion earlier in this chapter about the diferent uses of diferent kinds of income, postretirement income is one step removed from wage-labor exchange. This removes the wage-labor commensuration imperative to spend for retirees, which opens up space for other attitudes to income—for example, saving it to pass it on to heirs or else being risk averse and, therefore, reluctant to spend. There are of course clearly other possible explanations for what is going on here. But the fact that the life cycle pattern of consumption fits well with the predictions of the commensuration theory provides at least some support for the theory.

110

Time for Things

1.8. Explains Waste

The theory of consumption as commensuration provides a way to make sense of the huge amounts of waste produced under industrial capitalism. Commodities require time to produce utility—yet as output and real wages increase, while the working week stays constant, the time available to make use of each thing produced tends to decline. For that reason, the actual utility realized per thing produced (as opposed to the potential use power each item contains) likewise tends to decline. The diference between the aggregate potential use value of things and the use value actualized through using them is, in efect, waste. This becomes literally so as underutilized objects drift in the direction of landfills. If commodity acquisition was informed by sensible assessments of how much time and energy is available for making use of commodities rather than by the hypothetical use value contained in commodities (as a function of their physical properties), then the production of waste itself would be self-limiting. As per capita output (and hourly income) increased beyond a certain point, there would be increasing demand for time, as opposed to things, thus reducing demand and hence working as a negative feedback mechanism. If, however, as the commensuration theory suggests, the point of consumption is not just to realize actual use value but also to accumulate notional durations of activity indexed by commodity objects, to retroactively justify the labor given in exchange for the wage, then the connection between material output and actual need is attenuated. In such circumstances, natural constraints on the production of waste get removed, with dire environmental consequences.

1. Some Possible Objections to the Commensuration Theory

The commensuration argument thus works quite well as an explanation for certain puzzling aspects of consumer behavior under capitalism. It is also consistent with macrohistorical patterns in twentieth-century industrial capitalism. There are, nonetheless, objections that can be raised against it, as well as some questions about how certain significant developments in capitalist economies fit with the theory. It is to these that I now turn. Objections and questions are posed in the heading of each subsection, followed by my responses.

A Theory of Mass Consumption as Wage-Labor Commensuration

111