Consumption is not a goal in itself. As Schumacher put it, since consumption is merely a means to human well-being, the aim should be to obtain the maximum of well-being with the minimum of consumption.
While there is no doubt that some consumption is essential to welfare, there is much evidence to suggest that any link between the two is the exception rather than the rule in rich countries: evidence from historical and cross-cultural comparisons, from studies of self-reported happiness, and from commonly accepted social indicators. If that is the case, then it is perhaps possible to reduce consumption while maintaining or even improving welfare, even leaving aside the ecological costs of consumption.3
Orthodox economics encourages us to see the wish to consume as much as possible as almost part of human nature. Historical accounts of how consumer society came about and comparative studies of consumer and non-consumer societies suggest a very different picture: consumerism is historically specific. Sahlins' (1974) analysis of the 'original affluent society' documents the situation of Stone Age hunters and gatherers who, even in the marginal habitats to which they are nowadays restricted, require only a few hours a day to meet their needs, and thereafter stop work. This type of economy, while only possible at low population densities, suggests that, whatever its roots, consumerism does not stem from human nature. Moreover, there is much evidence of resistance to luxury in many traditional cultures (Stearns, 2001, ch.3).
The Industrial Revolution saw a transition from 'enough is enough' to 'the more the better', rooted in production for the market rather than for oneself.
Without mechanisms for expanding needs, people are inclined to work less rather than earn more, as demonstrated by the need for early industrialists to cut pay in order to force workers to work a 'full' working day (Gorz, 1989, ch.9). Or, as Sismondi expressed it in the early years of the Industrial Revolution, 'Luxury is not possible except when it is paid for by the labour of others' (quoted in Smith, 1993: 188); self-employed artisans chose leisure over luxury.
In fact, consumerism4 only really became established during the post-Second World War period. Until then the finiteness of consumer wants — and therefore the likelihood that productivity growth would result increasingly in leisure rather than consumption increases — was largely taken for granted, and the Victorian and interwar periods saw the popularity, in many countries, of ideas about leisured utopia. At the same time there were fears on the part of the elite about what people would do with their free time. Increasingly there were other problems, such as selling an ever-growing volume of goods, largely unrelated to needs, and redistributing available work in the face of rising productivity and unemployment; meanwhile the 1930s slump undermined the desire for free time. The 'solution' adopted was based on engineering increasing demand for consumer goods. Working hours were to be somewhat shorter, though still far longer than they need be, thus providing both the time and income required for bouts of concentrated, commercialised leisure (Cross, 1993).
As a result a set of consumer values has emerged, which Lewis Mumford characterised as follows:
There is only one efficient speed: faster; only one attractive destination: further away; only one desirable size: bigger; only one rational quantitative goal: more.
(quoted in Sachs, 1992: 120)
Of course adopting these values is a recipe for frustration, since they require attaining not any particular state of affairs, but a continuously receding goal.
Subjective well-being and income: the evidence
Perhaps the simplest approach to investigating whether, within consumer society, well-being improves along with increased consumption levels, is to ask people. A large body of research on income and 'subjective well-being' (SWB) has accumulated over the past four decades, based on surveys of self-reported happiness. In these surveys, respondents are asked to grade themselves on a scale of happiness or satisfaction. For example, in the Eurobarometer Survey carried out in the EU since 1973, people are asked whether they are 'very satisfied, fairly satisfied, not very satisfied, or not at all satisfied' with their lives (Oswald, 1997: 1819).
A variety of possible factors explaining well-being have been investigated, in particular income. Although there is some variation in methods as well as in the detailed findings, there is broad consensus regarding income and well-being (one recent summary is Frey and Stutzer, 2002):
1 In a given country at a given time, there is some correlation between income and well-being: the rich are happier than the poor. A typical partial correlation coefficient however is only about 0.13 (Frank, 1999: 112), indicating that other factors are more influential. Factors such as health, social relationships and leisure activities seem to be particularly important (Argyle, 1999).
2 Comparisons between countries yield a more mixed conclusion. For low-income countries, there does seem to be some correlation — though again, rather low — between income and well-being; but there is no correlation among high-income countries. One typical finding is that:
Above $13,000 in 1995 purchasing power parity there is no significant linkage between wealth and subjective well-being.
(Inglehart and Klingemann, 2000: 171)
3 The most striking result however is that where income and SWB are tracked over time, there is no correlation whatever between the two. The proportion of US respondents who declare themselves 'very happy' has fluctuated between 30 to 40 per cent over the period 1972 to 1991. Japanese self-reported happiness was flat for nearly three decades (1961— 1987) in spite of more than quadrupled income (Frank, 1999: 72-73). In contrast, EU data (1973-1998) show large fluctuations (and big differences among member countries) but no clear trend and no relationship to income (Inglehart and Klingemann, 2000: 167).
The pattern shown by these results was first pointed out by Easterlin (1972), and dubbed the 'Easterlin paradox': happiness appears to be related to income on the basis of cross-sectional, but not time series, data. The explanation, suggested by Easterlin himself, and many others since, seems to be that it is principally income relative to others which determines happiness. Diener, for example, concludes that 'the influence of income is largely relative; it is not the absolute level of goods and services that a person can afford' (1984: 553).
This explanation in turn raises other questions, discussed below, in particular why relative income should be so important, but also what the relevant comparison group is.
It has also been suggested (for example, by Myers, 1992, ch.3) that income, relative not only to others but also to one's own previous experience, may have some importance for happiness. Equally there is strong evidence that pay rises (and indeed both 'good' and 'bad' events in general) have only a temporary effect. Moreover, fast-growing economies seem to show the same invariant happiness level as slow-growing ones, as the example of Japan, referred to above, illustrates. It is worth adding that many other, non-consumption, goods escape the logic of relative comparison (Frank, 1997): when individuals enjoy such things as more free time, shorter commuting, reduced stress, more extensive social networks and better environmental conditions, they generally experience greater well-being irrespective of others' experience. But these are things which often become scarcer as consumption increases.
The SWB approach: criticisms and limitations
Such studies of course raise methodological problems. SWB research is based on the idea that reported well-being is an accurate and unbiased measure of actual well-being. But this requires that there is, at all times, a particular level of well-being that individuals experience; that they are aware what that is; and that, faced with the survey situation, they simply decide to tell the truth. Yet we know little about the mental processes that may lead respondents to give particular answers. Some attempts have been made to establish a more reliable link between reported and actual well-being. There have been some findings that reported happiness is related to brain activity (Inglehart and Klingemann, 2000: 165). Reported well-being has also been correlated with various measures of physical health, as well as other reported feelings such as stress or self-confidence. Some care has also been taken to ensure that linguistic differences do not affect international comparisons (Layard, 2003).
However, research also suggests there are major weaknesses in the SWB approach. First, the concept of SWB itself is far from satisfactory. If measured at one point in time (as it usually is), it is extremely volatile. Survey respondents only form a judgement about their well-being once they are asked about it, and their response is highly dependent on how the question is framed, what other questions may have preceded it, and what past events they call to mind when answering (Kahneman, 1999; Schwartz and Strack, 1991). Second, respondents commonly deceive themselves and others by claiming they are happier than they really are (the 'social desirability' effect) (Hagedoorn, 1996; Phillips and Clancy, 1972); although, to complicate matters, such deception may in fact enhance actual well-being (Naess, 1994). Third, there may be a 'satisfaction treadmill' rather than a 'hedonic treadmill', where people evaluate their well-being, rather than their income, by comparison with others (and/or with previous experience) (Kahneman, 1999); in this case actual well-being may increase with income, without this being reflected in reported well-being.
It is thus likely that reported well-being partly reflects factors other than actual well-being. It is difficult, for example, to exclude this kind of possibility when considering the large differences among EU countries referred to above. Such problems, while serious, are unlikely to invalidate completely SWB comparisons over time in the same country, so that the key finding, that well-being does not increase with income, in the aggregate, over time, still seems persuasive. However, in the absence of a reliable relation between actual and reported well-being, it would be unwise to rely exclusively on the SWB findings.
In view of doubts about the validity of the most direct approach — attempting to measure SWB itself — it is worth looking at other evidence regarding income and welfare. Here the obstacle is that there are no generally agreed objective measures of overall well-being. However, commonly used indicators of basic welfare (such as life expectancy, literacy, adequate diet) tend to stabilise at much lower levels of consumption than those of present-day Western Europe or North America. In some areas (family and community life, urban environment) welfare may decline as consumption rises.
A common pattern, illustrated by Figure 4.1 in the case of life expectancy, seems to be that for low levels of per capita gross domestic product (GDP) (and consumption) there is quite a strong relationship between consumption and welfare; past a certain point however there is little or no relation. But there are also many examples of low-income countries that nevertheless achieve high welfare levels. On the basis of these data a per capita GDP of about purchasing power parity (PPP) $5,000 (the world average is over PPP$7,000) would be more than enough to ensure a similar level of welfare as even the most overdeveloped countries manage. And perhaps a far lower per capita GDP level would be sufficient where other circumstances are conducive to well-being
Per capita GDP (PPP$)
Figure 4.1 Life expectancy vs. GDP Source: Based on UNDP 2003. Data are for 2001.
(including in particular relatively equal distribution of the GDP). A similar pattern exists for other basic indicators, such as literacy, per capita daily calorie intake, and access to safe drinking water.
Of course such indicators of basic needs satisfaction cannot be taken to indicate well-being as a whole; they merely reflect a consensus about its most basic aspects. The problem is that as one moves from the most basic aspects of well-being, conceptual and measurement problems increase, and there is little agreement about what well-being measures should include. That said, indicators such as suicide rates, crime rates, marital breakdown rates and rates of mental illness do not suggest that rising income and consumption are associated with improved well-being.
What the evidence about consumption and welfare suggests
Evidence about the relation between consumption and welfare is thus not absolutely conclusive, and perhaps can never be, given the general and intuitive nature of concepts like welfare and well-being. Nevertheless, all the evidence points in the same direction, and taken together it is very persuasive. The goal of ever-increasing consumption, far from being universal, is mostly a feature of post-1945 industrialised societies. Even where consumerism rules, factors other than consumption are more important for welfare and — most important for our argument — insofar as welfare is related to consumption at all, once basic material needs are satisfied, it is an individual's relative, not absolute, consumption level that counts for his or her welfare. It follows that in rich countries increases in consumption do not and cannot, in the aggregate, lead to improvements in overall welfare.
If increased consumption does not lead to greater welfare, the search for welfare on the other hand seems to lead to consumption: if, for whatever reason, individuals are concerned about their consumption relative to others, then a (quite rational) individual wish for greater welfare will drive a process of ever-increasing consumption, even though it cannot lead to any increase in welfare in the aggregate. Each individual increases his or her own consumption so as to improve his or her relative position; but the aggregate result of each individual's efforts is to maintain his or her relative position, and thus welfare, on average unchanged.
But there is one kind of evidence that appears to contradict the rest, and which finds expression in the economic doctrine of 'consumer sovereignty': most consumers apparently freely choose, at considerable sacrifice in work and effort, to consume as much as opportunities allow and certainly more than is required to meet their basic needs. This suggests that perhaps, in spite of the evidence presented, consumers may experience welfare as a result of rising consumption. If the view that consumption does not contribute to welfare is to be defended, and also if we are to find acceptable mechanisms to reduce consumption, we must answer the question: If consumption does not bring welfare, why do we consume?
Was this article helpful?