Review of Political Economy

ISSN: 0953-8259 (Print) 1465-3982 (Online) Journal homepage: https://www.tandfonline.com/loi/crpe20

The Dependence Effect, Consumption and

Happiness: Galbraith Revisited

Amitava Krishna Dutt

To cite this article: Amitava Krishna Dutt (2008) The Dependence Effect, Consumption
and Happiness: Galbraith Revisited, Review of Political Economy, 20:4, 527-550, DOI:
10.1080/09538250802308919

To link to this article: https://doi.org/10.1080/09538250802308919

Published online: 27 Sep 2008.

Submit your article to this journal

Article views: 1159

Citing articles: 17 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=crpe20
Review of Political Economy,
Volume 20, Number 4, 527 –550, October 2008

The Dependence Effect, Consumption

and Happiness: Galbraith Revisited
AMITAVA KRISHNA DUTT
Department of Economics and Policy Studies, University of Notre Dame, USA

ABSTRACT In his analysis of the affluent society, Galbraith argued that advertising and
the sales promotion activities of firms create wants for people, which makes them consume
more without making them better off, because their wants were artificially created. Thus,
in the affluent society, ever-increasing levels of production (and consumption) do not
increase welfare. This paper considers three criticisms of Galbraith’s analysis: first,
firms cannot ‘create’ wants for consumers without their consent, because consumers
are not mere pawns in their hands; second, even if people’s wants are created, they
may be better off by consuming more; and third, that expansion of consumption can
make people better off by expanding aggregate demand. It draws on the recent
literature on consumption, income and happiness, and develops a simple model of
growth and distribution, to argue that Galbraith’s analysis holds up against these
criticisms.

1. Introduction
Half a century ago, John Kenneth Galbraith argued in The Affluent Society that
even when an economy becomes rich according to conventional measures, such
as real production and income, it is beset with many problems. ‘Conventional
wisdom’ in economics (to use Galbraith’s phrase) gives a paramount position to
production and takes it to reflect the wants of the sovereign consumer. In this
view, since consumer wants are never satiated, increasing output of goods and
services is good for society. While this view may be relevant for situations in
which many people are poor and economically insecure, Galbraith argued that
it is inappropriate for affluent societies. In his view, in such societies, since
increases in production create consumer wants, the fulfillment of such artificially
created wants makes consumers no better off. On the contrary, the creation
of wants for private goods reduces people’s willingness to spend on public
goods and services such as education, transportation facilities, and health and
sanitation, which results in a social imbalance – private affluence and public
squalor – leading to social disorder. He also pointed out that the preoccupation

Correspondence Address: Amitava Krishna Dutt, Department of Economics and Policy Studies,
University of Notre Dame, Notre Dame, IN 46556, USA. Email: [email protected]

ISSN 0953-8259 print/ISSN 1465-3982 online/08/040527– 24 # 2008 Taylor & Francis

DOI: 10.1080/09538250802308919
528 A.K. Dutt

with increasing production results in a greater acceptance of inequality both

because inequality is supposed to increase growth by providing incentives for
saving and effort, and because increasing production can reduce poverty
without reducing inequality. The implication is that this leads to greater inequality,
which makes society worse off.
A central element of Galbraith’s indictment of the affluent society is his
analysis of how production increases wants and how this means that increasing
production makes society no better off.1 He wrote:

As a society becomes increasingly affluent, wants are increasingly created by the

process by which they are satisfied. This may operate passively. Increases in
consumption, the counterpart of increases in production, act by suggestion or
emulation to create wants. Expectation rises with attainment. Or producers
may proceed actively to create wants through advertising and salesmanship.
Wants thus come to depend on output. . . . [I]t can no longer be assumed that
welfare is greater at an all-round higher level of production than at a lower
one. It may be the same. (Galbraith, 1958, p. 158)

Galbraith termed the way wants depend on the process by which they are satisfied
‘the Dependence Effect’, and proposed the squirrel wheel as an apt model for the
affluent society, which runs only to stay at the same place.
Galbraith’s claim has been disputed in various ways.2 This paper focuses on
three major criticisms, which are as follows. First, the increases in consumption in
affluent societies that Galbraith observed, and which continue to occur in affluent
and less-affluent societies today, are not due to the advertising and sales-
promotion activities of firms. Firms may try to persuade consumers, but cannot
force them to buy their products. Consumers are the ones who make buying
decisions, and they should not be thought of as passive pawns in the hands of
firms. Moreover, there is no clear evidence that advertising has an effect on
total levels of consumption, and at best it makes specific producers of particular
goods increase their sales at the expense of other producers. Second, even if
firms increase consumption by creating needs, it does not necessarily follow
that welfare is not increased by increases in consumption. The fact that advertising
makes us buy a particular product does not imply that its purchase does not make
us better off, and therefore increase welfare. Third, even if an increase in
consumption does not make us better off directly, it is possible that it has other

1
Both supporters and detractors of Galbraith’s book focus on Chapter XI, entitled ‘The
Dependence Effect’. Hayek (1961, p. 346), a critic, writes: ‘I believe the author would
agree that his argument turns upon the “Dependence Effect” explained in Chapter XI of
the book’, and titles his critique ‘The non sequitur of the “Dependence Effect”’. Stanfield
(1983, pp. 590– 591), in summarizing and pointing out the relevance of Galbraith 25 years
after the appearance of the book comments that nothing else in the book has invited more
controversy.
2
For a discussion of the book and a review of early criticisms and appraisals of it, see
Hession (1972).
 The Dependence Effect, Consumption and Happiness 529

consequences that result in increases in welfare. For instance, it can increase the
demand for goods, make firms produce more, and thereby increase employment
and reduce unemployment, improving wellbeing.
The purpose of this paper is to evaluate these three criticisms by examining
some recent theoretical and empirical literature on consumption and happiness
and by examining some simple macroeconomic models of consumption, pro-
duction, employment and growth, and relating them to Galbraith’s analysis in
The Affluent Society. By so doing, it attempts to provide a critical evaluation of
Galbraith’s complaints about the affluent society.
The rest of the paper proceeds as follows. Section 2 examines the criticism
that firms do not actually increase consumption by creating needs, and discusses
it in terms of recent analytical developments in the analysis of consumption
and happiness. It also briefly comments on the empirical argument that advertis-
ing does not increase consumption, merely reallocating sales between firms.
Section 3 turns to the criticism about whether consumption actually increases
welfare by examining recent empirical work on the determinants of happiness
as reported by people themselves, especially on whether increases in consump-
tion and income increase happiness, and by discussing explanations of the empiri-
cal findings. Section 4 examines the argument that advertising has a favorable
impact on the economy because it increases aggregate demand, thereby reducing
unemployment and promoting economic growth. Section 5 concludes.

2. Advertising, Marketing and Consumption

The effect of advertising on consumption has been intensely debated. Some have
followed Galbraith in arguing that advertising and other sales promotion activities
of firms increase consumption, creating wants for goods and services that are not
needed, wants that do not naturally exist. However, the view attributed to
Galbraith that firms can simply manipulate consumers into buying more has
come in for some severe criticism.
In his early criticism of Galbraith, Hayek (1961, p. 347) argues that firms do
not create wants. He states that:
The tastes of man . . . are shaped in a great measure by his cultural environment.
But though in some contexts it would perhaps be legitimate to express this by a
phrase like ‘production creates the wants’, . . . [it] . . . would clearly not justify
the contention that particular producers can deliberately determine the wants
of particular consumers. The efforts of all producers will certainly be directed
towards that end; but how far any individual producer will succeed will
depend not only on what he does but also on what the others do and on a
great many other influences operating upon the consumer. The joint but uncoor-
dinated efforts of producers merely create one element of the environment by
which the wants of the consumers are shaped.
Although Hayek seems to concede that firms can play some role in influen-
cing consumer preferences, later critics appear to deny them any important role.
It is argued that consumers are the ones who decide what they will buy and
what they will not: firms do not have the power to just produce anything and
530 A.K. Dutt

force them on consumers against their free will.3 To underscore this point, the
American Association of Advertising Agencies put out an advertisement with a
picture of a woman using shaving cream and a razor on her face, stating that
‘[d]espite what some people think, advertising can’t make you buy something
you don’t need’ (see Belch & Belch, 2007, p. 733). It is pointed out that many
expensive advertising and sales promotion campaigns by firms have failed,
which demonstrates that firms cannot manipulate consumers. For instance,
the attempt by Coca-Cola to reformulate Coke in 1985 failed, as did Ford’s
attempt to market a radically new car, called the Edsel, despite expensive adver-
tising campaigns.4
It is argued that although the earlier approach of marketers to advertising
involved trying to make consumers buy whatever firms produce, the more recent
approach is for marketers and firms to find out what consumers really want
(perhaps in a latent manner) and to supply it to them. Firms want to sell by creating
brand loyalty; this cannot happen if consumers are not getting what they really
want and being satisfied with their purchases. A leading principles of marketing
text, by Kotler & Armstrong (2004, p. 5), states that ‘[t]oday, marketing must be
understood not in the old sense of making a sale – ‘telling and selling’ – but in
the new sense of satisfying consumer needs’ (italics in original). The text continues
to point out (p. 6) that ‘[o]utstanding marketing companies go to great lengths to
learn about and understand their customer’s needs, wants and demands. They
conduct consumer research and analyze mountains of consumer sales, warranty
and service data. Their people at all levels – including top management – stay
close to customers.’
Finally, although advertising seems to affect the demand for particular pro-
ducts in some cases and contexts (see Albion & Farris, 1981), it is argued that
there is little evidence to suggest that advertising expenditure causes an increase
in aggregate consumption. Although Taylor & Weiserbs (1972) report that adver-
tising increases consumption, most studies find negligible effects. Simon (1970)
summarizes data collected during newspaper strikes and on the distribution of
advertising expenditures and consumption expenditures that do not find strong
effects, although spending appears to fall during newspaper strikes when people
are less exposed to advertising. Schmalensee (1972) finds that increases in sales
increase advertising, but advertising does not affect the total consumer spending
on goods and is inconsequential in driving aggregate demand. Quarles & Jeffres
(1983) find that increases in disposable income lead to increases in consumption,
which in turn leads to increases in advertising expenditures. The review of the

3
Katona (1960) suggests the image of the powerful consumer in affluent societies, who
makes consumption decisions using discretionary income based on his or her level of con-
fidence. He draws attention to the instability of the consumption function despite high
levels of advertising by large corporations.
4
I mention the case of Edsel here because Galbraith (1998) refers to it in his introduction to
the 40th anniversary edition of the book, saying that earlier, textbooks used to argue this
point with the example of Ford’s Edsel, but later jettisoned it. I return to this case in Note 5,
below.
 The Dependence Effect, Consumption and Happiness 531

literature in Luik & Waterson (1996) suggests that advertising positively affects
consumption demand in the early stages of a product’s life cycle, but with the
passage of time primarily affects a brand’s share rather than total consumption.
There are many problems with these arguments. First, the simple ‘proof’,
which points to the failure of some advertising and sales promotion campaigns,
does not hold water. This can be shown with the following analogy: just because
some lies are not believed by people, it does not follow that people never believe
lies or even that the majority of lies are not believed. It would be rather surprising
if such simple proofs could, in fact, resolve the issue.5 Indeed, the simple proof of
the claim that advertising increases sales, with the argument that if advertising
does not work firms would not advertise, is also erroneous. Firms may be adver-
tising not to lose their market share when their rivals advertise; this does not
imply that if there is no advertising by anyone sales would be lower.
Second, the fact that consumers are the ones that ultimately make the pur-
chasing decision does not imply that firms cannot make them buy more than
they would otherwise.6
It can be argued that if consumers are consciously aware that advertise-
ments are trying to persuade or influence them they can employ their cognitive
defenses to protect themselves form going against their own interests. However,
growing evidence suggests that much of human behavior, including consumer
behavior, is not under conscious control but occur automatically without cogni-
tive intervention. This is particularly evident in the writings of behavioral econ-
omists, who have drawn on the findings of cognitive psychologists and decision
theorists.7 The implications of these contributions, including the different types
of cognitive biases of consumers, has been explored by Hanson & Kysar
(1999a), who argue that sellers can make use of the cognitive failures of
buyers to manipulate markets influencing, if not determining, market outcomes
by controlling the format of information, the framing of choices, and the setting
in which purchases are made. Hanson & Kysar (1999b) amass a large body of
empirical evidence, which shows the extent to which such manipulation

5
There are also problems with specific examples. I am grateful to an anonymous referee for
pointing out that the Edsel disaster is actually a complex tale of many shortcomings, and
should not be seen as a failure of advertising in isolation. See, for instance, Brooks (1963)
and Bonsall (2002).
6
Katona’s (1960) argument, noted earlier – that consumer spending is unstable and depen-
dent on consumer confidence despite high advertising expenditures – does not imply that
advertising does not positively affect consumption spending. The fact that, as Katona
argues, consumers use their discretion to replace durable goods long before they must
be replaced because they are completely worn out and beyond repair, implies the possi-
bility that the timing of replacement, and hence the level of consumption, can be affected
by marketing. Moreover, the argument that consumption expenditure is affected by adver-
tising and other forms of marketing does not imply that it is not also affected by consumer
confidence.
7
Some of the relevant literature in economics (and psychology) is surveyed in Earl (1990)
and Rabin (1998).
532 A.K. Dutt

occurs, some of it purposefully by sellers of products with knowledge of the

relevant literature on buyer behavior.
Some marketing scholars accept the idea of the persuasive potential of
marketing rather than subscribing to the ‘market concept’, or the belief that the suc-
cessful business has to be oriented to satisfying the needs of the buyer, based on its
acceptance of the ideas of consumer rationality and consumer sovereignty (see, for
instance, Dickinson et al., 1986). Even advertising principles texts concede that
subtle techniques and subliminal messages can activate consumers,8 even though
subliminal cuts are ineffective.9 However, such priming or activation is argued
to be possible only if the topic is compatible with the consumer’s current need
states: a consumer will be activated by subtle messages suggesting that a particular
product will enhance attractiveness only if a need to be attractive is salient to the
consumer. But since consumers have a variety of salient needs, such as the need to
have friends, meaning in life, to relieve boredom, and be esteemed by others – as
included, for instance, in Maslow’s ([1954] 1970) hierarchy of needs – in addition
to the need to be attractive, advertisers have many options open to them. Moreover,
it is argued that such priming is not effective in affecting judgments over a long
period of time. Even if this is true, with impulse buying and frequent activation
through a variety of media, product placements in stores, and the internet, consu-
mers can be influenced to buy products in the short run, after which habit and others
factors may take over. The role of the internet, which allows online shopping with
the touch of a single button, is particularly important here.10
Moreover, the recent literature on why people consume suggests that an
important reason why consumers buy more is that other consumers buy more
(see, for instance, Hirsch, 1976; Schor, 1998; Frank, 1999; Lichtenberg, 1998;
Layard, 2005). This happens for a number of reasons. With imperfect information
about products, the fact that others are buying things can signal to others that they
are good to have. When others buy some things with network externalities – such
as cell-phones and internet connections – one is left out of the loop when one does
not have these things. When many people have something, social norms require
others to have it as well and not having it can result in a loss of dignity; Adam
Smith’s example of leather shoes makes this point well. A very important
reason is status: in their attempt to have high status people try to consume more
than others, and this leads to a race to consume more (see Dutt, 2001).

8
See, for instance, Shimp (2007, p. 63).
9
James Vicary created a myth by claiming that consumer choice can be affected by adver-
tisements (using words) superimposed repeatedly in single frames within movies (even
though the eye could not consciously register them). Although this scam has been
subsequently discredited (see Rogers, 1992 – 93), subliminal advertisements involving
the use of symbolic messages that prey on the viewers’ subconscious to affect their con-
sumption behavior is an effective marketing tool.
10
Amazon.com’s practice of providing personalized suggestions based on what other
customers who have bought the item ordered have also purchased, can be thought of as
helping out customers in a world of imperfect information and bounded rationality, but
arguably also promotes, and is designed to promote, impulse buying.
 The Dependence Effect, Consumption and Happiness 533

Galbraith himself stresses the importance of relative consumption in

explaining increases in consumption. He approvingly cites the writings of
Keynes and Duesenberry, who stressed the role of emulation in making consu-
mers keep up with, or get ahead of, the consumption of others, and writes that
‘[o]ne man’s consumption becomes his neighbor’s wish. This . . . means that
the process by which wants are satisfied is also the process by which wants are
created. The more wants are satisfied the more new ones are born’ (Galbraith,
1958, p. 154). Thus, for Galbraith, emulation is an important aspect of the depen-
dence effect. However, his subsequent analysis focuses on the corporation,
stating that ‘[t]he even more direct link between production and wants is provided
by the institutions of modern advertising and salesmanship’ (Galbraith, 1958,
p. 155). In so doing, he does not adequately recognize the connections between
the two elements of the dependence effect, that is, that advertising and salesman-
ship take into account the relative consumption effect and try to exploit it. One of
the ways in which advertising persuades people to buy more is by creating social
norms, and especially by stressing the importance of status and by showing that
the consumption of some goods brings more status. Some advertisers can even
use some peoples’ dislike for status competition by associating their products
as not providing status!
The recent literature, by stressing the role of relative consumption and by
underplaying the role of the corporation, appears to have gone to the other
extreme, compared with Galbraith’s greater emphasis on the corporation. This lit-
erature seems to suggest that it is a part of human nature to emulate others, and
does not take into account the fact that even if human beings are driven by the
motives of emulation and status-seeking it does not follow that they would seek
to do so by specific acts of consumption. Nor does it explain why these motives
are stronger and consumption higher in some times and places than in others. His-
tories of consumption can turn to new patterns and levels of marketing efforts of
corporations to explain these trends (see, for instance, Stearns, 2001; de Graaf
et al., 2002).
In addition to appealing to status and related relative consumption motives,
advertisers can appeal to other ‘needs’ of consumers. As mentioned earlier, they
can try to convince people that by buying a product they obtain friends,
novelty, excitement, and the fulfillment of their dreams. They may even resort
to fear tactics, scaring people into buying things to overcome their fears (as in
the case of burglar alarms, insurance and large cars). These ‘needs’, of course,
exist in individuals, but what does not exist, and which advertising tries to
create, is the notion that these needs can be satisfied by the specific products
being advertised. Marketing texts refer to the distinction between needs and
wants in saying how needs are a basic part of the human makeup, which can be
translated into specific wants of products, including brands, by marketing,
among other influences (Kotler & Armstrong, 2004, p. 6).
It should also be noted that even if advertising only spreads information about
products, it may still increase sales beyond what consumers would buy in the
absence of information.
Third, the empirical literature, which tries to show that advertising does not
increase consumption, is problematic for a number of reasons.
534 A.K. Dutt

First, it interprets advertising and sales promotion in a narrow way, focusing

on the direct advertising expenses of firms. Firms rely increasingly on public
relations activities rather than on advertisements to boost their image and increase
their sales (Ries & Ries, 2004). Moreover, sales promotion expenses go well
beyond standard marketing expenses: expenses on product innovation are import-
ant for increasing consumption. Firms, of course, spend enormous amounts of
product development which are not included in sales promotion and advertising
expenditures.
Second, the pressure to consume more comes from many other things, such
as films and television shows which, in addition to increasing the demand for
specific products due to ‘product placement’, increase the urge to consume in
general by ‘softening’ people up to make them easy prey for advertisers (see
Adorno & Horkheimer, [1944] 2000). Some of these effects come about not
because advertising of specific products leads to increasing sales, but because
of the broader social effects of advertisement and the other aspects of consumer
culture, often unintended by advertisers, which promote values – such as materi-
alism, selfishness, social competitiveness, and sexual preoccupation – which help
to increase consumption (see, for instance, Pollay, 1986).
Third, even if advertising and other sales promotion expenses have a
positive effect on consumption and sales at the level of the individual firm, it
is unlikely that the effect is a simple one that makes current advertising (even
with lags) affect sales. Marketing textbooks point out that advertising is not
just a current expense but an investment. Shimp (2007, p. 246), for instance,
cites Proctor and Gamble’s ex-chief executive officer, Jennifer Lawrence, and
a major marketing consultant, likening advertising and other sales promotion
expenditures to exercising and writing: ‘Advertising momentum is like exercise.
Stop exercising, and you will lose conditioning and probably gain weight. Stop
advertising, and your brand likely will lose some of its equity and market share
as well’.
Fourth, as noted earlier, there is evidence to show that advertising is crucial at
the early stages of the development of products. Later on, the markets of advertis-
ing firms may become saturated, or sales may depend more on emulation effects
than on advertising. These comments imply that the empirical studies that inter-
pret sales promotion narrowly, and which examine simple cause and effect
relationships without taking into account the complex interdependencies
between the activities of corporations and the media on the one hand, and other
social factors – such as emulation effects and social and cultural norms – on
the other, are seriously flawed.
The discussion here has focused mainly on the role of advertising in affecting
consumption. However, as mentioned in the previous paragraph, advertising is
only one way in which marketing increases consumption. We end this section
with just one example of how sellers affect consumption in ways other than adver-
tising (as narrowly defined): the design of shopping malls. Malls attract not only
people who have specific buying plans, but also browsers who have no clear
purchasing plans, but who may have ‘latent’ demands. Earl & Potts (2000)
argue that the way in which products are displayed and stores are located in shop-
ping malls is designed to increase spending by the browsing shopper by grabbing
 The Dependence Effect, Consumption and Happiness 535

their attention, rather than to reduce search time and other transactions costs for
the optimizing shopper with well-defined preferences.

3. Consumption and Happiness

Galbraith (1958, p. 160) argued that:
our concern for goods . . . does not arise in spontaneous consumer need. Rather,
the dependence effect means that it grows out of the process of production itself.
If production is to increase, the wants must be effectively contrived. In the
absence of the contrivance the increase would not occur. This is not true of
all goods, that that it is true of a substantial part is sufficient. It means that
since the demand for this part would not exist, were it not contrived, its utility
or urgency, ex contrivance, is zero. If we regard this production as marginal,
we may say that the marginal utility of present aggregate output, ex advertising
and salesmanship, is zero.
However, it can be argued that just because firms create wants for people, and
people can be persuaded by firms to consume more, it does not follow that
increases in consumption will not satisfy some strongly-felt wants or increase
their happiness. Hayek (1961, p. 346) argues that Galbraith’s suggestion that if
wants are created by the process by which they are satisfied they will not increase
happiness, is a non sequitur:
Very few needs indeed are ‘absolute’ in the sense that they are independent of
social environment or of the example of others, and that their satisfaction is an
indispensable condition for the preservation of the individual or of the species.
Most needs which make us act are needs for things which only civilization
teaches us to exist at all, and these things are wanted by us because they
produce feelings or emotions which we would not know if it were not for our
cultural inheritance. Are not in this all our esthetic feelings ‘acquired tastes’?
Littlechild (1982), also from an Austrian perspective, draws on the work of G.L.S.
Shackle, who emphasizes the role of imagination in choice and the limited role of
reason. Shackle argues that consumers choose between subjectively-imagined
future pleasures and pains associated with different products rather than
between objective properties of different products. Littlechild (1982, p. 34) inter-
prets Shackle as rejecting the ‘rational’ consumer approach with imperfect infor-
mation (in which advertising serves to provide consumers with more information
and therefore allows them to make more informed choices), yet suggesting that
advertising helps consumers to exercise their imagination and has a role in ‘stimu-
lating the creation of opportunities for satisfying tastes’. The implication is that the
consumer’s preferences prior to exposure to advertisements are not the ‘real’ ones,
but rather immature ones that have not developed to their potential.
Hayek and Littlechild are right to imply that Galbraith oversells his notion
that wants are in some way unimportant because they are created by the process
with which they are satisfied, and therefore their satisfaction necessarily does
not increase utility or social welfare. However, data on subjective well-being
that is now available – but not available when Galbraith wrote his book –
seem to suggest that Galbraith was not far off the mark. Economists and other
536 A.K. Dutt

social scientists have produced empirical studies that question the fact that
increases in consumption and income – at least significantly – affect happiness
as evaluated by the consumers themselves. The pioneering contributions of
Easterlin (1973, 1995, 2001), and subsequent work by Oswald (1997), and Frey
& Stutzer (2002) among others, suggest a number of empirical regularities.11
Time series data for individual countries (such as the US, and Japan, with high
levels of per capita income) do not reflect significant (and in some cases any)
increases in the average level of self-reported happiness over time, despite signifi-
cant increases in income and consumption. Panel data on specific groups of indi-
viduals over their lives suggest that despite large increases in income, these
individuals do not show significant increases in self-reported happiness. Cross-
sectional studies across countries suggest that countries with higher levels of
per capita income and consumption do not have higher average levels of self-
reported happiness beyond a certain level of income that is far below the
income of the rich countries of the world. Even individuals who win lotteries
have been found to report no greater happiness after a few years. To be sure,
there is some support for the income/consumption –happiness connection.
Cross-sectional studies within countries seems consistent with it: people
in higher income groups with higher levels of consumption report higher levels
of self-reported happiness than people in lower income groups; it seems that it
is better to be rich than poor in a particular society at a particular point in time.
Cross-country studies do suggest a positive income–happiness link at low
levels of income. Some studies suggest that people are happier – even if
temporarily – if their consumption and income increases. However, the bulk
of the evidence seems to contradict the consumption– happiness relationship.
What all this suggests is that increases in production and consumption, at least
in affluent societies, may not increase subjective well-being significantly, and in
that sense have a low or zero marginal utility. This holds even in the presence
of advertising and salesmanship, contrary to, but in fact strengthening, Galbraith’s
suggestion!
There are good reasons to believe that increases in consumption caused
by marketing efforts may not result in lasting increases in happiness. If adver-
tising is to be successful it should not only induce people to buy things by
making them believe that they lack something that they need to have, but
by continuing to make them believe that they lack something to make them
buy even more. Thus, to be effective, advertising needs to keep people in a
state of discontent. Moreover, to the extent that advertisements and other
kinds of marketing methods try to increase sales by making people insecure
and fearful, and if the resultant purchases do not actually overcome some nega-
tive feelings (one does not necessarily obtain friends by buying beer, contrary
to what is suggested by commercials!), discontent can actually increase with
increases in consumption.

11
See also Easterlin (2002) for an edited collection of key contributions and Layard (2005)
for a recent overview.
 The Dependence Effect, Consumption and Happiness 537

The recent literature on consumption and happiness provides a number of

additional reasons why increases in consumption may not significantly increase
happiness in affluent societies. If, as discussed earlier, consumption increases
because people consume more when others consumer more in order to improve
their status or conform to social consumption norms, it is possible that increases
in consumption will not increase happiness. For instance, if people consume
more because they seek higher status (because they want to maintain their
status relative to others or to catch up to them) through consumption, as all
people consume more, their relative consumption, and hence status, does not
change, so that they are not able to obtain higher status or utility. They may
even end up being worse off if they work more to increase their consumption,
and work cuts into their enjoyable leisure time. If advertising works through the
status motive, this argument can explain why advertisement-driven consumption
increases bring about little or no increases in subjective well-being.
There are other reasons why consumption increases may not increase happi-
ness or utility, some of which were alluded to by Galbraith himself. Although
Galbraith did not explicitly discuss their effect on subjective well-being, such
effects have more recently been found to exist. For instance, Galbraith emphasized
the role of consumption in increasing consumer borrowing in his prescient discus-
sion of the problems of consumer indebtedness. He presented data on trends in
consumer debt in the 1950s to argue that rising debt makes consumption
demand more volatile, because an interruption in consumer borrowing can
cause swings in consumption spending, thereby destabilizing the economy
much in the way that investment demand does. Recent discussions of consumer
debt confirm some of Galbraith’s fears (see Carroll & Dunn, 1997) but, in
addition, show that increases in consumer debt have a negative effect on self-
reported happiness (Brown et al., 2005). Other channels include spending more
time on work to increase income to buy more, spending less time with family
and friends, and choosing higher pay over other considerations such as social
relationships. To the extent that family, friendships, and other social relationships
are important determinants of happiness, increases in consumption and income
can adversely affect happiness (see Lane, 1991, 2000). Finally, rising income
can result in many kinds of stresses for many people from new technologies
and from having to deal with new market segments and new social codes as
they get promoted, which can reduce self-reported happiness (Lane, 1991).
The negative effects of increases in consumption may be measured not just in
terms of self-reported happiness, but in other ways as well, as suggested by the
effects of public consumption. Galbraith noted that increases in private consump-
tion, driven by emulation and advertising and salesmanship, crowds out public
consumption spending on goods, which are not subject to the status motive or
to sales promotion, and this results in the loss of social balance between private
and public consumption. Private consumption and public goods are complemen-
tary, since – for instance – more cars require more roads and more packing
requires more effort in disposing garbage. The consequence of the loss of social
balance is that objective indicators of well-being, such as education and health
indicators, in addition to subjective well-being, are adversely affected (see
Frank, 1999, for a discussion of some of these effects).
538 A.K. Dutt

4. Consumption and the Macro-economy

If advertising increases aggregate consumption, it may have a positive effect on
the economy by increasing aggregate demand and reducing unemployment,
even if the increase in consumption itself does not have a positive effect on the
happiness of consumers. Increases in aggregate demand, in turn, increase output
and reduce unemployment, and may even increase investment and the growth
rate of the economy, increasing employment creation and reducing poverty if
growth is equitable.
The fact that high consumption demand made possible by greater sales pro-
motion can reduce unemployment has important implications for happiness.
Empirical studies on happiness show that unemployment has a significant negative
effect on it even if income and consumption do not. Clark & Oswald (1994) find
that in Britain ‘[j]oblessness depressed wellbeing more than any other single
characteristic including important negative ones such as divorce and separation.’
Di Tella et al. (2001), using data from 12 European countries over 1975–1991,
found unemployed people are 0.33 units less happy on the scale from 1 to 4
when they are unemployed. Frey & Stutzer (2002), using data from Switzerland,
find that the marginal effect of being unemployed, rather than employed, other
things constant, is a 20.6% loss in happiness.12 These findings confirm ideas
that joblessness leads to significant effects due to psychic costs (such as feelings
of anxiety and loss of self-esteem) and social cost (due to the stigma of jobless-
ness, especially in societies in which work defines a person’s position in life),
even holding income constant. There is also evidence that a general increase in
the unemployment rate reduces the general level of happiness in society (Di
Tella et al., 2001), for instance, by increasing the fear of unemployment among
the employed.
The notion that sales promotion activities can positively affect aggregate
demand is recognized by a number of economists. Baran & Sweezy (1964,
pp. 23– 24) write that:
There can be little doubt . . . that the chronic underutilization of resources which
has plagued the United States for more than a generation now would have been a
great deal more severe if it had not been for the spectacular growth of advertising
during this period. If this is correct, it follows that attempts to establish or curtail
advertising could have seriously adverse effects unless accompanied by compre-
hensive and effective planning for full and socially desirable employment. This
is a point which critics of advertising consistently neglect . . .
This is consistent with their view that because of increases in monopoly power in
the economy, the share of profits increases at the expense of the wage share, which
reduces consumption demand and reduces output (see Baran & Sweezy, 1966).
Galbraith himself was aware of the contribution advertising makes. He wrote,
within a decade of the publication of The Affluent Society:

12
This compares with a marginal loss of 3.1% for people with low and middle income,
13.3% for unhealthy people.
 The Dependence Effect, Consumption and Happiness 539

Advertising and its related arts . . . help develop the kind of man the goals of the
industrial system require – one that reliably spends his income and works
reliably because he is always in need of more . . . In the absence of the
massive and artful persuasion that accompanies the management of demand,
increasing abundance may well have reduced the interest of people in acquiring
more goods . . . Being not pressed by the needs for these goods, they would have
spent less reliably to get more. The consequence – a lower and less reliable
propensity to consume – would have been awkward for the industrial system
(Galbraith, 1967, p. 219).

However, in The Affluent Society, Galbraith did not discuss this aggregate
demand-creating role of advertising. He seemed to worry, instead, about the
problem of inflation caused by a high level of aggregate demand (due to govern-
ment policies aimed at providing economic security) and the wage-price spiral
(due to the coexistence of oligopolistic firms and strong labor unions). In the
1998 edition, Galbraith states in his introduction that his focus on inflation and
the wage-price spiral was, with hindsight, misplaced because of weakened
unions and technological change. So it is important to take aggregate demand
and unemployment into account in analyzing the consequences of sales pro-
motion. Is it not possible, then, that increases in consumption demand will
increase aggregate demand and growth and reduce unemployment and poverty,
thereby increasing happiness and welfare?
These issues can be examined using the simple textbook aggregate demand-
aggregate supply (AD-AS) model.13 The economy’s position is shown in the usual
way by the intersection of a downward-sloping AD curve and an upward-rising AS
curve on price and real output space. An increase in advertising expenditures (and
other activities of firms and those media that increase sales), which increases the
level of consumption at each price level, shifts the AD curve to the right, increas-
ing real output (as well as the price level). It will therefore have the effect of
increasing employment and reducing unemployment, and the effect on prices
will be small if there are no significant wage pressures when unemployment falls.
The textbooks, however, also point out that these effects of increases in con-
sumption are only valid in the short run, when involuntary unemployment can
exist. If there is such unemployment, wages will fall, which will reduce the
price level, increase real money supply, and increase aggregate demand as
people spend their excess real money balances, or lend it out, which reduces the
interest rate and encourages spending by others. In the medium run, therefore,
the economy will be at the so-called natural rate of unemployment or the
NAIRU. Starting from medium run equilibrium, an expansion in consumption
demand due to the activities of firms will only increase output and reduce unem-
ployment in the short run; wage and price increases will move the economy back
to its medium-run equilibrium. Consumption spending will merely crowd out
investment spending, thereby jeopardizing the expansion of the economy in the
future.

13
See, for instance, Blanchard (2006).
540 A.K. Dutt

The argument that the effect of increases in consumption are only temporary,
and negated in the medium run, while subscribed to by many – in fact most –
mainstream macroeconomists, is neither theoretically persuasive nor empirically
accurate. There is no automatic necessary tendency for the economy to return
to some natural rate of unemployment through wage and price adjustments. For
instance, price reductions can result in debt deflation and an adverse effect on
the financial position of debtor firms, which can reduce investment spending,
and price changes may lead to endogenous changes in money supply with no
(appreciable) effects on interests rates (unless there are policy changes made by
the monetary authorities). Moreover, economic expansion can result in endogen-
ous changes in technology, which increase the so-called natural level of output
(see Dutt, 2006a). There are many empirical studies which suggest that the real
output in many countries does not follow a stationary process, implying that
temporary shocks can have long-term, even permanent, effects (see, for instance,
Dutt & Ros, 2007, for a review).
If aggregate demand changes can have effects beyond the short run, are we to
conclude that the sales promotion activities of firms necessarily reduce unemploy-
ment? It may not do so if such activities increase people’s desire to obtain more
income and work more, increasing the supply of labor.14 Abstracting from this
possibility, we confine attention to the aggregate demand side. That increases in
sales promotion expenditure need not necessarily increase aggregate demand
and reduce unemployment can be seen from a simple explicit version of the
AD-AS model.
Assume that firms set the price level according to the markup-pricing rule,
so that
P ¼ (1 þ z)bW (1)
where P is the price level, W is the money wage, b is the fixed (for simplicity) unit
labor requirement, all other variable inputs are ignored, and z is the level of the
markup, determined by factors such as the level of industrial concentration,
held fixed for now. Firms spend an amount A in real terms on advertising. The
rate of profit of firms is given, using equation (1), by:
z Y A
r¼
(2)
1 þ zK K
where K is the stock of capital. The level of consumption demand is assumed to
depend positively on real income and the level of advertising expenditures. We
write the consumption function in a linear form as:
C ¼ c0 A þ cY (3)
where c is the marginal propensity to consume out of income and c0 the marginal

14
See Schor (1998), for instance, for the argument that an increase in consumption may
increase labor supply, and that unemployment may be reduced if workers decide to
reduce their supply of labor and increase their leisure time, with more people sharing
employment.
 The Dependence Effect, Consumption and Happiness 541

effect of advertising expenditure on consumption. Investment demand is assumed

to depend on the rate of profit and, assuming that the investment as a ratio of
capital stock depends on the rate of profit linearly, we have:
I
¼ a0 þ a1 r (4)
K
where I is real investment and ai are positive investment parameters. Equilibrium
in the goods market requires:
Y ¼CþI (5)
where we abstract from government fiscal policy for now. Substituting equations
(2) through (4) into equation (5), and denoting the rate of capacity utilization, Y/K,
by u, we find the equilibrium rate of capacity utilization to be:
c0 a þ a0
a1 a
u¼ (6)
1
c
a1 ðz=1 þ zÞ
where a ¼ A/K. An increase in advertising expenditures, reflected in an increase
in a, is seen to have an ambiguous effect on capacity utilization, the sign of the
effect depending on c0-a1, if other things are equal. It is possible that if the profit-
reducing effects on investment are stronger than the consumption-increasing
effects of advertising, capacity utilization, and hence output, will decline. If the con-
sumption effect is weak, as some critics of Galbraith argue, then the negative effects
will dominate. In other words, the AD curve may shift to the left.
Other things, however, may not be equal. First, when advertising expendi-
tures rise the markup is likely to rise. Firms may try to recover some of their
lost profits by increasing their markups. Kalecki (1971, p. 50) argues that increases
in sales promotion activities replaces price competition by competition in adver-
tising, which will increase the degree of monopoly and hence the markup. It has
also been argued that advertising increases industry concentration because of high
capital costs of creating brand loyalty and for scale economies in advertising,
which also create barriers to entry, and this makes it possible for firms to increase
their markup. Moreover, if advertising competition is able to tip consumers
towards one or a few products, and away from competitor products, industrial
concentration will increase. All of this, by increasing z, and hence the rate of
profit, will make it more likely that capacity utilization will increase with adver-
tising. This makes it more likely that the AD curve will shift to the right. However,
the increase in the markup also implies an upward shift in the AS curve which may
reduce output. So, even if the AD curve shifts to the right, the upward shift in AS
can reduce capacity utilization. Second, autonomous investment, a0, may depend
on the price level through the interest rate effect: a rise in the price level brought
about by the increase in the markup, by reducing the real supply of money may
increase the interest rate and reduce investment demand. Third, if we bring in gov-
ernment spending into the analysis, a rise in advertising expenditure, as discussed
by Galbraith (1958) and mentioned earlier, will increase consumption expenditure
but reduce government expenditure on public goods. This will tend to shift the AD
curve to the left. The overall effect of an increase in advertising, therefore, is
ambiguous.
542 A.K. Dutt

Fourth, income distributional issues, which were considered important by

Galbraith,15 can be incorporated into the model by assuming that there are two
classes, workers who receive wage income and capitalists and advertising firms
who receive the rest of the income, which can be called profits. If it is assumed,
as in most heterodox macroeconomic models, that workers consume their entire
income and saving occurs only out of profits, we can write consumption out of
profits as:

Cc ¼ co A þ cc sY (30 )

where cc is the marginal propensity to consume out of profits. In this case, equation
(6) will have to be replaced by:

a0 þ (c0
a1 )a
u¼ (60 )
½(1
cc )
a1 s

where s ¼ z/(1þz) is the share of income going to profits. The stability of short-
run equilibrium requires that cc þ a1 , 1, which implies that the rise in the
markup, by increasing the profit share, will have a tendency to reduce u by redis-
tributing income from profit recipients who save to workers who do not. This
raises the likelihood that an increase in sales promotion activities will reduce
capacity utilization.
One problem with the formulation just discussed is that sales promotion has no
effect on consumption by workers, which follows from the assumption that workers
consume all their income. We can allow sales promotion to have an effect on con-
sumption by workers if we incorporate another feature of the economy highlighted
by Galbraith, that is, consumer debt. To introduce this we allow workers to save and
accumulate wealth, as in models pioneered by Pasinetti (1962), but extend these
models to allow for consumer borrowing and debt.
The income of workers, net of interest payments or receipts, is given by:

Yw ¼ wbY þ iKw (7)

where w is the real wage, W/P, i the interest rate, and Kw is the stock of wealth
held by workers, which can be positive (representing assets) or negative (repre-
senting debt). The consolidated income of capitalists and advertising companies
(we refer to the two jointly as capitalists from now on) is given by:

Yc ¼ zwbY
iKw (8)

15
Regarding income distribution, the first edition noted that although inequality had faded
as an economic issue because of the emphasis on aggregate output and because it had not
worsened, though still high (Galbraith, 1958, pp. 84– 85). In the 40th anniversary edition,
however, it was stated that the problem of inequality had not only continued, but worsened
(Galbraith, 1998, p. 71).
 The Dependence Effect, Consumption and Happiness 543

Consumption by workers depends on consumption by capitalists because workers

try to emulate their consumption, and on their own income,
Cw ¼ gCc þ cw Yw (9)
The consumption by each worker also depends on the consumption level of other
workers, and this effect is captured in the sizes of the two parameters g and cw. It is
assumed that 0 , cw , 1, but that cw is large. If g ¼ 0, so that consumption by
workers does not depend on the consumption level of capitalists, saving by
workers is small but positive. However, if g . 0, with high levels of capitalist
consumption, it is possible for worker saving to become negative, and for
workers eventually to incur debt. Consumption by capitalists depends only on
their own income, since they do not try to emulate the consumption of workers,
only that of other capitalists, an effect that is incorporated into the parameter cc.
Thus, assuming a constant consumption propensity (which also depends on
sales promotion by firms), we get:16
Cc ¼ cc Yc (10)
For simplicity, we assume that the interest rate is fixed by the monetary auth-
orities and that banks operate as financial intermediates with zero costs (other than
interest costs) and profits.
For the short run we assume that investment, I, is given, as is the stock of
capital and the stocks of capital owned by workers (and capitalists).
Short run equilibrium, for given I, K and Kw, obtains when the goods market
clears, or equation (5) holds. Noting that C ¼ Cc ¼ Cw, and using equations (1)
and (7) through (10), we obtain the short-run equilibrium level of capacity utiliz-
ation,
½cw
(1 þ g)cc ikw þ g
u¼ (11)
1
(1 þ g)cc s
cw (1
s)
where, as before, s ¼ z/(1þz) is the profit share in production income, g ¼ I/K
and kw ¼ Kw/K. We assume that the denominator is positive, which is required for
the stability of short-run equilibrium. We also assume that the numerator is posi-
tive which, given the sign of the denominator, is necessary for positive levels of
capacity utilization. This implies that there is a lower bound for kw (when it is
negative) for a given value of g.
Changes in advertising expenditures affect a number of parameters, that is,
cw, cc, g, and s, positively. Ignoring the numerator for now, an increase in the con-
sumption propensities, cw and cc, of the two groups, increases capacity utilization.
An increase in g by making workers want to emulate capitalist consumption more
strongly, has the same effect. The effect of an increase in s, brought about by the
increase in the markup as firms increase markups to cover higher advertising

16
We do not distinguish between the consumption propensities of capitalists and those who
obtain income from advertising for simplicity. Allowing for differences would require us
to keep track of the stock of capital owned by each group separately, adding to the
complexity of the dynamic model developed below.
544 A.K. Dutt

expenses, depends on the sign of cw – (1þ g)cc. We assume this to be positive,

with workers having a higher marginal propensity to consume out of income
than capitalists, so that cw . cc, and with the induced effect of capitalist consump-
tion on worker consumption being positive but not too strong. With this assump-
tion, the rise in s reduces capacity utilization by redistributing income from
workers to capitalists and therefore reduces consumption. Additional effects
arise from the numerator of equation (11). There are no effects from investment,
which is assumed to be fixed in the short run. The effects of changes in the con-
sumption parameters, however, depend on the sign of kw. If it is negative, so that
workers are net debtors, an increase in cw reduces capacity utilization because
interest payments on debt have a negative effect on their consumption, which is
made more negative with a higher marginal propensity to consume. The effects
of the other consumption propensities in the numerator can be interpreted in a
similar way. If workers are net creditors, the effects will be reversed. An increase
in g increases u in the usual way, by raising aggregate demand. An increase in kw
increases capacity utilization by redistributing capital income from capitalists to
workers.
In the long run, investment, the stock of capital, and the stock of assets owned
by the two classes, and hence, g and kw can change. We assume that in the long run
the rate of investment (or the rate of growth of capital, in the absence of deprecia-
tion), changes according to the dynamic equation:
dg=dt ¼ l½gd
g (12)
d
where l . 0 is a speed of adjustment parameter, and g , the desired rate of invest-
ment, is given by
gd ¼ a0 þ a1 r þ a2 u (13)
where desired investment is assumed to depend positively on the rate of profit, r,
and on the rate of capacity utilization (see Steindl, 1952). The rate of growth of kw
is given by:
k^ w ¼ K^ w
g (14)
where the overhat denotes rate of growth. The rate of change in Kw is determined
by worker savings, so that:
dKw =dt ¼ Yw
Cw (15)
Substituting the expression for the rate of profit, which is given by equation
(2), and equation (13) into equation (12) we get:
dg=dt ¼ l½a0 þ a1 (su
a) þ a2 u
g (16)
Substitution from equations (7) through (10) into equation (15) and then into
equation (14) we get:
dkw =dt ¼ ½(1
cw )(1
s)
gcc su
½g
((1
cw ) þ gcc )ikw (17)
Equations (16) and (17) are the dynamic equations of the model for the long run, in
which u is given by equation (11).
 The Dependence Effect, Consumption and Happiness 545

The dynamics of this model can be analyzed by using phase diagrams in g, kw

space. The dg/dt ¼ 0 locus is positively-sloped. Starting from a position on the
curve, an increase in kw increases u as can be seen from equation (11), which increases
desired investment, which increases dg/dt. To bring dg/dt back to 0, we need to
increase g, which reduces dg/dt provided that [(a1s)/L] þ a2 , 1, where
L ¼ [1 þ (1 þ g) ccs 2cw (12 s)] is the denominator of the investment multiplier.
This condition is a stability condition, which ensures that the rise in actual investment
does not increase desired investment through its effect on capacity utilization more to
result in a cumulative increase in actual investment. Above the dg/dt ¼ 0 line, g falls,
and below the line g rises, explaining the vertical arrows in Figure 1.
The shape of the dkw/dt ¼ 0 line is more complicated. Substituting equation
(11) into equation (17) we obtain:
dkw =dt ¼ (1=L)f½(1
cw )(1
cc )i kw
Lg kw þ (L
cc (1
s)gg (170 )
Setting dkw/dt ¼ 0 we get:
(1
cc )(1
cw )ikw
g¼ (18)
Lkw
½L
cc (1
s)
The shape of the curve for this equation, which is the dkw/dt ¼ 0 line, depends on
the sign of [L2cc (12 s)], which is given by the expression:
G ¼ (1
s)(1
cw )
gscc (19)
which can take any sign. The expression measures the effect on saving by workers
of an increase in capacity utilization. The first term captures the positive effect of
an increase in the wage share and the marginal propensity to save out of wage

Figure 1. Long-run dynamics: the ‘happy’ case.

546 A.K. Dutt

income, while the second term captures the negative effect of an increase in the
profit share, the increase in consumption of capitalists, and the induced increase
in consumption by the workers.
If the expression in equation (19) is positive, the relation between g and kw
satisfying equation (18), and hence implying dkw/dt ¼ 0, is shown as in Figure 1.
The values of g and kw which are the asymptotes for the curve are given by:
(1
cc )(1
cw )i
gþ ¼
L
and
(1
s)(1
cw )
g s cc
kwþ ¼
L
The horizontal arrows, from equations (17) and (11), show how kw changes at differ-
ent points in the figure. The figure shows that the economy, unless it starts from very
low values of kw, tends to its long-run equilibrium at point E. In equilibrium there
will be a positive level of kw, so that workers will own a constant share of total
capital. For the model to be internally consistent the parameters should be such
that kw  1, and that the equilibrium values of g and kw do not imply the full utiliz-
ation of capital or the full employment of labor.
If the expression in (19) is negative, the dkw/dt ¼ 0, will be as shown as in
Figure 2. The values of the asymptotes will be as before, and the horizontal
arrows, showing movements in kw, are as shown. In the configuration shown in
the figure, there is no long-run equilibrium, and it is possible for the economy
to be on a path in which kw and g keep falling, implying greater indebtedness
of workers and lower rates of growth. If the investment parameters are sufficiently
large, it is possible for the dg/dt ¼ 0 line to intersect the dkw/dt ¼ 0 line, and for
there to be a stable long-run equilibrium (at the intersection with the higher growth
rate). Of course, kw cannot take any value. We have to ensure that g and kw do not
fall to such an extent that u becomes negative. Far before that point is reached,
however, some lower bound for kw will be reached at which workers will reach
their credit limits. In that case the equations of this model will cease to apply.17
It can be seen that it is possible the economy may experience a regime shift
due to increases in advertising. Increases in advertising expenditures, by increas-
ing the marginal propensities to consume or the level of g, can move the economy
from the case in which G . 0, so that Figure 1 applies, to the case in which G , 0,
so that Figure 2 applies. Starting from a long-run equilibrium shown in Figure 1,
it may seem that in the short run capacity utilization increases with more adver-
tisement expenditure (although, as discussed earlier, this may not happen).
However, if the economy experiences a regime shift, it will experience a fall in

17
If workers reach their credit constraint – as a fraction of their income net of interest
payments – the evolution of debt will no longer be as analyzed in this model, which
allows workers to borrow any amount they wish, given their income and consumption
plans. With credit constraints, consumption will be determined by how much they can
borrow and the evolution of debt will be as analyzed in Dutt (2006b).
 The Dependence Effect, Consumption and Happiness 547

Figure 2. Long-run dynamics: the ‘unhappy’ case.

the share of capital owned by workers, and eventually rising worker indebtedness,
which will imply a lower level of economic growth and a less equal distribution of
income and assets. Given an exogenously fixed rate of growth of labor supply,
unemployment will grow (faster) as well. The economy can then be said to be
less ‘happy’.
The intuition behind this effect of increasing advertising (and other market-
ing expenditures) on growth and distribution is straightforward. When workers
save, they obtain income from both wages and interest, and are able to own a
share of the capital in the long run. If advertising expenditures increase workers
consumption, they may experience a rise in income in the short run because of
the expansionary effect of higher spending by both workers and capitalists. But
with increases in capitalist income and the desire by workers to emulate them,
workers become borrowers and eventually net debtors. As their debt increases,
workers experience a decline in their net income, and with a redistribution of
income (from interest payments), there is a contractionary effect on aggregate
demand, because capitalist interest earners have a higher saving propensity than
workers. This contractionary effect reduces capacity utilization and growth,
implying a worsening in both growth and income distribution.
Economic growth may not make people, especially the rich, happier. But it
may do so if it makes the poor have higher levels of income, reduces unemployment,
and reduces inequality. However, our analysis implies that the sales promotion
expenditure of firms is unlikely to lead to this type of growth, and may even lead
to stagnation. There are much better ways of improving the macroeconomic per-
formance of the economy in ways that make people happier, for instance, by
increasing the supply of useful public goods, and improving the distribution of
income.
548 A.K. Dutt

5. Conclusion
This paper has attempted to evaluate Galbraith’s analysis of the affluent society in
which firms create wants and increase consumption without increasing well-being.
It has done so by examining three criticisms of Galbraith’s analysis: firms cannot
manipulate consumers, it is not true that higher levels of consumption induced by
firms do not increase welfare, and higher levels of consumption can result in
improved macroeconomic behavior, which can make people better off.
First, it has argued that even though consumers ultimately decide how much –
and what – to consume, firms affect consumption through their marketing and
related efforts. This argument has drawn on recent research on the determinants
of consumption, especially that which shows that consumption by people depends
on the consumption of other people.
Second, by reviewing empirical evidence and theoretical contributions, it has
argued that increases in consumption and income may not be associated with
increases in well-being even as evaluated by consumers themselves. Although
Galbraith may have oversold the argument that increases in output and consump-
tion will not increase utility since it is brought about by increasing sales
promotion, his argument is supported by empirical data and theoretical analysis.
Third, it has argued, by reviewing standard macroeconomic models and by
developing a new one, that increases in advertising and other sales promoting
expenditures may well have adverse long-run macroeconomic effects by increas-
ing consumer indebtedness and inequality, and reducing growth.
Galbraith’s analysis seems, 50 years later, to be right on the mark even
though, at the time, the argument could not be fully developed and firm empirical
evidence was not yet available.

Acknowledgments
An earlier version was presented at the Eastern Economic Association meetings at
New York, 2007 and at the Analytical Political Economy Workshop at Queen
Mary, London University, 2007. Comments from participants at these presenta-
tions, including Bill Gibson, Gilberto Lima, Simon Mohun, Malcolm Sawyer,
Gil Skillman, Martha Starr, and Roberto Veneziani are gratefully acknowledged.
I am also grateful to Michael Etzel for discussions and for pointing me towards
relevant marketing and advertising textbooks. Finally, I would like to thank the
anonymous referees for this journal for their extremely useful comments and sug-
gestions, and references to the literature.

References
Adorno, T.W. & Horkheimer, M. ([1944] 2000) The culture industry: enlightenment as mass decep-
tion, in: J.B. Schor & D.B. Holt (Eds) The Consumer Society Reader, pp. 3–19 (New York: New
Press).
Albion, M.S. & Farris, P.W. (1981) The Advertising Controversy: Evidence on the Economic Effects
of Advertising (Boston: Auburn House).
Baran, P. & Sweezy, P. (1964) Theses on advertising, Science and Society, 28, pp. 20–30.
Baran, P. & Sweezy, P. (1966) Monopoly Capital (New York: Monthly Review Press).
 The Dependence Effect, Consumption and Happiness 549

Belch, G.E. & Belch, M.A. (2007) Advertising and Promotion: An Integrated Marketing Communi-
cations Perspective, 7th edn (New York: McGraw Hill).
Blanchard, O. (2006) Macroeconomics, 4th edn (Upper Saddle River, NJ: Prentice Hall).
Bonsall, T. (2002) Disaster in Dearborn: The Story of the Edsel (Stanford, CA: Stanford University
Press).
Brooks, J. (1963) The Fate of Edsel and Other Business Adventures (New York: Harper & Row).
Brown, S., Taylor, K. & Price, S.W. (2005) Debt and distress: evaluating the psychological cost of
credit, Journal of Economic Psychology, 26, pp. 642–661.
Carroll, C.D. & Dunn, W.E. (1997) Unemployment expectations, jumping (S,s) triggers, and house-
hold balance sheets, NBER Macroeconomics Annual, Vol. 12, pp. 165–217 (Cambridge, MA:
MIT Press).
Clark, A.E. & Oswald, A.J. (1994) Unhappiness and unemployment, Economic Journal, 104,
pp. 648–699.
de Graaf, J., Wann, D. & Naylor, T.H. (2002) Affluenza: The All-Consuming Society (San Francisco:
Berrett-Koehler Publishers).
Di Tella, R., MacCulloch, R.J. & Oswald, A.J. (2001) Preferences over inflation and unemployment,
American Economic Review, 91(1), pp. 335–341.
Dickinson, R., Herbst, A. & O’Shaugnessy, J. (1986) The marketing concept and customer orien-
tation, European Journal of Marketing, 20(10), pp. 18–23.
Dutt, A.K. (2001) Consumption, happiness and religion, in: A.K. Dutt & K. Jameson (Eds) Crossing
the Mainstream: Ethical and Methodological Issues in Economics, pp. 133–169 (Notre Dame,
IN: University of Notre Dame Press).
Dutt, A.K. (2006a) Aggregate demand, aggregate supply and economic growth, International
Review of Applied Economics, 20(3), pp. 319–336.
Dutt, A.K. (2006b) Maturity, stagnation and consumer debt: a Steindlian approach, Metroecono-
mica, 57(3), pp. 339–364.
Dutt, A.K. & Ros, J. (2007) Aggregate demand shocks and economic growth, Structural Change and
Economic Dynamics, 18(1), pp. 75–99.
Earl, P.E. (1990) Economics and psychology, Economic Journal, 100, pp. 718–755.
Earl, P.E. & Potts, J. (2000) Latent demand and the browsing shopper, Managerial and Decision
Economics, 21, pp. 111–122.
Easterlin, R. (1973) Does economic growth improve the human lot? Some empirical evidence, in:
P. David & M. Reder (Eds) Nations and Households in Economic Growth: Essays in Honor of
Moses Abramovitz, pp. 89–125 (Palo Alto, CA: Stanford University Press).
Easterlin, R. (1995) Will raising the incomes of all increase the happiness of all?, Journal of Econ-
omic Behavior and Organization, 27, pp. 35–47.
Easterlin, R. (2001) Income and happiness: towards a unified theory, Economic Journal, 111,
pp. 465–484.
Easterlin, R. (Ed.) (2002) Happiness in Economics (Cheltenham, UK: Edward Elgar).
Frank, R. (1999) Luxury Fever: Why Money Fails to Satisfy in an Era of Excess (New York: Free
Press).
Frey, B.S. & Stutzer, A. (2002) Happiness and Economics (Princeton, NJ: Princeton University
Press).
Galbraith, J.K. (1958) The Affluent Society (London: Hamilton).
Galbraith, J.K. (1967) The New Industrial State (London: Houghton Mifflin).
Galbraith, J.K. (1998) The Affluent Society, 40th anniversary edn (Boston: Houghton-Mifflin).
Hanson, J.D. & Kysar, D.A. (1990a) Taking behavioralism seriously: the problem of market manipu-
lation, New York University Law Review, 74, pp. 630–749.
Hanson, J.D. & Kysar, D.A. (1990b) Taking behavioralism serious: some evidence of market
manipulation, Harvard Law Review, 112(7), pp. 1420–1572.
Hayek, F. (1961) The non-sequitur of the ‘dependence effect’, Southern Economic Journal, 27(4),
pp. 346–348.
Hession, C.H. (1972) John Kenneth Galbraith and his Critics (New York: New American Library).
Hirsch, F. (1976) Social Limits to Growth (Cambridge, MA: Harvard University Press).
Kalecki, M. (1971) Selected Essays on the Dynamics of the Capitalist Economy (Cambridge, UK:
Cambridge University Press).
550 A.K. Dutt

Katona, G.A. (1960) The Powerful Consumer (New York: McGraw-Hill).

Kotler, P. & Armstrong, G. (2004) Principles of Marketing, 10th edn (Upper Saddle River, NJ:
Pearson-Prentice Hall).
Lane, R.E. (1991) The Market Experience (Cambridge, UK: Cambridge University Press).
Lane, R.E. (2000) The Loss of Happiness in Market Democracies (New Haven, CT: Yale University
Press).
Layard, R. (2005) Happiness: Lessons from a New Science (London: Penguin Press).
Lichtenberg, J. (1998) Consuming because others consume, in: D.C. Crocker & T. Linden (Eds)
Ethics of Consumption: The Good Life, Justice and Global Stewardship, pp. 155–175
(Lanham, MD: Rowan & Littlefield).
Littlechild, S.C. (1982) Controls on advertising: an examination of some economic arguments, Inter-
national Journal of Advertising, 1(1), pp. 25–37.
Luik, J.C. & Waterson, M.J. (1996) J.C. Luik, J. Waterson & M (Eds.), Advertising and Markets
(Oxford, UK: NTC Publications).
Oswald, A.J. (1997) Happiness and economic performance, Economic Journal, 107, pp. 1815–1831.
Maslow, A. ([1954] 1970) Motivation and Personality, 2nd edn (New York: Harper & Row).
Pasinetti, L. (1962) The rate of profit and income distribution in relation to the rate of economic
growth, Review of Economic Studies, 29(4), pp. 267–279.
Pollay, R.W. (1986) The distorted mirror: reflections on the unintended consequences of advertising,
Journal of Marketing, 50(2), pp. 18–36.
Quarles, R.C. & Jeffres, L.W. (1983) Advertising and national consumption: a path-analytic
re-examination of the Galbraithian argument, Journal of Advertising, 12, pp. 4–13, 33.
Rabin, M. (1998) Psychology and economics, Journal of Economic Literature, 36(1), pp. 11–46.
Ries, A. & Reis, L. (2002) The Fall of Advertising and the Rise of PR (New York: Harper Business).
Rogers, S. (1992–93) How a publicity blitz created the myth of subliminal advertising, Public
Relations Quarterly, 37(4), pp. 12–17.
Schmalensee, R. (1972) The Economic Effects of Advertising (Amsterdam: North-Holland).
Schor, J. (1998) The Overspent American: Upscaling, Downshifting and The New Consumer
(New York: Basic Books).
Shimp, T.A. (2007) Advertising, Promotion, and Other Aspects of Integrated Marketing Communi-
cations, 7th edn (Mason, OH: Thomas South-Western).
Simon, J.L. (1970) Issues in the Economics of Advertising (Urbana, IL: University of Illinois Press).
Stanfield, J.R. (1983) The Affluent Society after twenty-five years, Journal of Economic Issues, 27(3),
pp. 589–607.
Stearns, P. (2001) Consumerism in World History: The Global Transformation of Desire (London:
Routledge).
Steindl, J. (1952) Maturity and Stagnation in American Capitalism (Oxford, UK: Basil Blackwell).
Taylor, L.D. & Weiserbs, D. (1972) Advertising and the aggregate consumption function, American
Economic Review, 62(4), pp. 642–655.