Literature Review of Consumption, Inequality, and Debt Dynamics

This paper reviews the theoretical concept of wage-led and profit-led demand regimes by incorporating relative consumption concerns, integrating the Veblenian concept of conspicuous consumption into the Bhaduri–Marglin model. It posits that if the profit share increases and the corresponding decrease in workers’ income is distributed unevenly, efforts to ‘keep up with the Joneses’ may increase consumption, leading to a consumption-driven profit-led regime. The review highlights the potential complementarity between the Institutionalist/Evolutionary concept of conspicuous consumption and the (Post-)Keynesian concept of effective demand, suggesting that increases in the profit share leading to unequally distributed losses within the working class can make a temporary profit-led regime theoretically possible even without positive effects on investment or net exports. The analysis emphasizes the role of social emulation, interpersonal comparisons, and behavioral economics in understanding consumer behavior and the formation of social identity.

CONSPICUOUS CONSUMPTION, INEQUALITY AND
DEBT: THE NATURE OF CONSUMPTION-DRIVEN
PROFIT-LED REGIMES
Jakob Kapeller and Bernhard Schütz*
University of Linz
(May 2013; revised April 2014)
ABSTRACT
This paper extends the theoretical concept of wage-led and profit-led demand regimes by incorpor
relative consumption concerns. Specifically, it integrates the Veblenian concept of conspicuous co
sumption into the Bhaduri–Marglin modelby assuming that relative consumption concerns matter
primarily within the working class. If in such a framework the profit share increases and the corre-
sponding decrease in workers’ income is distributed unevenly, efforts to ‘keep up with the Joneses
increase consumption and, hence, lead to a consumption-driven profit-led regime.
1. INTRODUCTION
Ever since its publication, the seminal article by Bhaduri and Marglin (1990
on demand regimes has sparked a lively debate on whether the current gro
path of a given country or region is determined by a wage-led or a profit-le
demand regime.1 While the empiricalresults are often ambiguous (see,for
example, Bowles and Boyer, 1995; Gordon, 1995; Stockhammer and Onara
2004; Naastepad, 2006; Naastepad and Storm, 2006–7; Hein and Vogel, 20
Stockhammer et al., 2009; Hein and Tarassow, 2010), the theoretical conce
itself is clear-cut: a country finds itself in a profit-led demand regime when
fall in wages and a corresponding rise in the profit share induces sufficient
* We would like to thank Eddy Beckers, Octavio Fernandez-Amador, Eckhard Hein, Michael
Landesmann,Markus Marterbauer,Susanne Pech,Miriam Rehm, Martin Riese, Christa
Schlager, Sepp Zuckerstätter, as well as two anonymous referees for a series of very helpful
inspiring comments. Any remaining errors are ours.
1 Initially Bhaduri and Marglin (1990) called them stagnationist (wage-led) and exhilarationist
(profit-led).
bs_bs_banner
Metroeconomica 66:1 (2015) 51–70
doi: 10.1111/meca.12061
© 2014 John Wiley & Sons Ltd

additionalinvestment (Bhaduriand Marglin,1990) and/or export demand
(see, for example, Blecker, 1999) to compensate for the decrease in consum
tion demand. If these channels are not strong enough, the negative effect o
consumption dominates and the demand regime is wage-led.2 The recent
decade has cast doubt on the empirical relevancy of Bhaduri’s and Marglin’s
theoreticalconcept,since in many western economies—the USA being the
most prominent example—rising profit shares did not lead to high rates of
investment growth (Stockhammer, 2005–6). Taking economic development
Germany as an example, it seems that the export-driven scenario is a much
more eligible case for profit-led demand. Moreover, rising profit shares were
accompanied by high consumption growth in many countries, pointing to an
unexplained variation in the marginal propensity to consume. This variation
is sometimes explained by referring to wealth effects due to increasing hous
prices (e.g. Zezza, 2008). A complementary argument, that has become wid
accepted, attributes the rise in the marginal propensity to consume in the U
to the rise in income inequality taking place over the last decades and the ro
of relative consumption concerns (see, for example, Cynamon and Fazzari,
2008, 2013; Stiglitz, 2009; Barba and Pivetti, 2009; Evans, 2009; UN Commi
sion of Experts,2009;ILO and IMF, 2010;Kumhof and Rancière,2010;
Rajan, 2010; Hein, 2011; Kumhof et al., 2012; Stockhammer, 2012, 2013; va
Treeck, 2012; Kapeller and Schütz, 2013). Here it is argued that those house
holds that were falling behind in relative income tried to keep up in terms of
relative living standard (as depicted by the level of consumption of goods an
services) by financing consumption through debt. Thereby deregulation and
innovation (e.g. Credit Default Swaps (CDSs), collateralized debt obligations
(CDOs)) on financial markets as well as a housing price bubble that inflated
collateral values made sure that this demand for credit was accompanied by
sufficient supply.3
In what follows, we focus on the demand side of these developments and
take a sufficiently large supply ofconsumer creditas given (see for this
Kapeller and Schütz, 2014). Thereby, our contribution to the theoretical Post
Keynesian literature on demand regimes is inspired by the before-mentioned
pre-crisis developments. In our view, these developments make it necessary
amend the existing theoretical arguments on profit-led regimes by providing
simple and clear-cutmodelincorporating consumer debtand increasing
inequality—two main features of the recent turmoil—to illustrate a potential
2 See also Kurz (1990) who arrives at a similar result by introducing the normal rate of profit
into the investment function.
3 Already before the crisis Minsky (1986) provided a quite well analysis of the pattern of event
(see on this also Kapeller and Schütz, 2013).
52 Jakob Kapeller and Bernhard Schütz
© 2014 John Wiley & Sons Ltd

explanation for the recent economic developments in the USA. In doing so,
suggest the possibility of a (naturally short-lived) consumption-driven profi
led regime (in addition to the already existing investment- and export-drive
possibilities). We illustrate in this paper that once we allow for (1) relative
consumption concerns within the working class and (2) increases in the pro
share leading to unequally distributed losses within the working class,a
temporary profit-led regime becomes theoretically possible even in the abs
of any positive effects on investment or net exports. Particularly in the case
the US economy assumption (2) seems to be in line with statistical facts, as
look at the development of the wage share (which mirrors the developmen
the profit share) and real family income growth suggests: as the wage shar
been declining over the last 30 years (except for a short-lived upward mov
ment during the Clinton era, see figure 1),4 there has also been an increase in
income inequality within the working class, as the diverging growth rates o
real family income across quintiles 1 to 4 suggest (figure 2).
From a paradigmatic perspective we exploit the potential complementari
between the Institutionalist/Evolutionary concept of conspicuous consump-
tion (Veblen,1970 [1899])and the (Post-)Keynesian conceptof effective
demand.In demonstrating ourcaseof a consumption-driven profit-led
regime we thereby also contribute to a Pluralist Paradigm in the spirit of
Dobusch and Kapeller (2012) that seeks to create new insights through the
4 For a discussion of potential reasons see Stockhammer (2009).
61
62
63
64
65
66
67
58
59
60
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Figure 1.Adjusted wage share.
Source:Annual macro-economic database of the European Commission (AMECO). Here the
adjusted wage share is defined as the ratio of compensation of employees to the number of
employees (full-time equivalents) divided by the ratio of nominalGDP to total employment
(full-time equivalents).
Conspicuous Consumption, Inequality and Debt 53
© 2014 John Wiley & Sons Ltd

exploitation ofcomplementary conceptsas they are found in different
schools of thought (see on this also Kapeller and Schütz, 2013).5
2. THEORETICAL BACKGROUND
While the Keynesian principle of effective demand is well known and does
not have to be further explained, the concept of conspicuous consumption
(or more general: relative consumption concerns) as well as its specific unde
standing employed in thispaperneedssome more emphasis:following
Veblen (1970 [1899]), consumer preferences are socially mediated, implying
that the hesitation to reduce consumption is not necessarily due to a direct
loss of comfortor pleasure,but is related to questions ofsocialstatus.
5 There are a variety of reasons why exploring potential complementarities between different
heterodox approaches seems beneficialfor the further development of economic theory (see
Dobusch and Kapeller,2009,2012;Garnettet al.,2010;Kapeller,2010).With respectto
Institutionalist/Evolutionary and Post Keynesian economics there exist many acknowledged
conceptual similarities, which are often applied in combination when answering practical ques-
tions. In fact many Post Keynesians utilize concepts from Institutional economics when discuss
ing microeconomic issues and Institutional economists often refer to Post Keynesian argument
on macroeconomic issues (e.g.Lavoie,1992,2009;Arestis,1996;Dosi et al.,2010).In sum,
however,these contributions often do not point at theoreticalcomplementarities that lead to
genuine interactions between theoretical arguments, but rather rely primarily on similarities or
on the need to fill obvious blind-spots in one’s own tradition.
117.4%
97.9% 103.5% 104.7%
88.7%
49.0%60.0%
80.0%
100.0%
120.0%
140.0%
1947-1973 1979-2009
-7.4% 3.7%
11.2%
22.7%
-20.0%
0.0%
20.0%
40.0%
Lowest fifth Second fifth Third fifth Fourth fifth Highest fifth
Figure 2.Real family income growth by quintile.
Source:State of Working America; Economic Policy Institute (EPI) analysis of Census
Bureau data.
54 Jakob Kapeller and Bernhard Schütz
© 2014 John Wiley & Sons Ltd

Specifically, Veblen (1970 [1899]) asserts that the social emulation of prefe
ences is coined by ‘a desire to live up to the conventional standard of dece
in the amount and grade of goods consumed’ (p. 49), where the convention
standard has a slight upwards bias, since ‘[. . .] each class envies and emul
the class next above it in the social scale, while it rarely compares itself wit
those below or with those who are considerably in advance’. This viewpoint
corresponds wellwith empiricalresults on the formation of socialidentity
(Hogg and Terry, 2000), where it is argued that those people with whom we
share parts ofour daily life (like colleagues or neighbors)have a strong
impact on our perception of an adequate, ‘conventional’ living standard. In
turn, such people have a higher probability to serve as reference points
(Rabin, 1998) for determining an individual’s consumption aspirations. This
assumption is reinforced by theoreticalresults on the evolution of socially
mediated preferences within populations, where great gaps in consumption
expenditures might lead to mutual ignorance between a minority of very ri
and a mass of ‘common’ people (Kapeller and Steinerberger, 2014, Chapte
5). We find similar results also in behavioral economics, where it is argued
that ‘the reference state usually corresponds to the decision maker’s curre
position’(Tversky and Kahneman,1991,p. 1046),while our ‘perceptual
apparatus is attuned to the evaluation of changes or differences rather tha
to the evaluation of absolute magnitudes’(Kahneman and Tversky,1979,
p. 277).In terms of socialemulation,interpersonalcomparisons should,
therefore,occur relative to subjects close to ‘the decision maker’s current
position’.In the same vein behavioraleconomics supports the idea of an
upward bias in socialemulation,since visible relative losses of income or
standard of living are subject to a ‘status quo bias’ (Kahneman et al., 1991
Following these arguments we employ the assumption of conspicuous con-
sumption in our modelby assuming that people compare themselves with
those people who are similar to them, but still slightly in advance (for empi
cal evidence see also Stutzer, 2004).
Hence,when conspicuous consumption6 plays an important role in con-
sumer spending, aspired consumption levels depend not only on income, b
6 It is sometimes assumed that Veblen (1970 [1899]) used this term only to denote consumpt
activities of a specific group (the nouveau riche of his times) devoted to signaling one’s wealt
others. However, a close reading of Veblen (1970 [1899], esp. Chapters 4–5) shows that Vebl
explicitly asserts that reputational concerns play a decisive role for all income groups and tha
the means of satisfying these concerns through conspicuous activities (consumption, leisure
vary over time. Thus, the nouveau riche only represent an illustrative case of a far more gene
principle. In line with this reading of Veblen we employ the concept in a general way as depic
the fact that consumption preferences are not intrinsically given but rather developed throug
interaction with and comparison with others (socially mediated preferences).
Conspicuous Consumption, Inequality and Debt 55
© 2014 John Wiley & Sons Ltd

also on the consumption level of other, associated groups. Thus, a fall in the
wage share occurring at the expense of one group of workers will not neces-
sarily lead to a decrease in aggregate consumption. As long as those worker
who lose income can somehow afford to hold on to their aspired consump-
tion level (which is in turn influenced by other workers whose income did no
decrease),the immediate consequence willrather be a lower saving rate of
those workers falling behind in their income. If income is no longer enough
to afford consumption aspirations, consumption-driven credit-arrangements
arise and saving rates might even turn negative. However, since saving rate
can hardly be negative forever,such a constellation is only a temporary
possibility, pointing to the inherent instability of such a scenario.
Duesenberry (1962 [1949]) also makes this connection, although he emph
sizes the pivotal role of a continual improvement of consumption goods (i.e.
the creation of ‘superior goods’), because ‘[. . .] for any particular family the
frequency of contact with superior goods will increase primarily as the con-
sumption expenditure of others increase. [. . .] The result will be an increase
in expenditure at the expense of saving’(Duesenberry,1962 [1949],p. 27).
Empirical evidence for the importance of relative consumption concerns for
determining credit demand is provided by Christen and Morgan (2005), who
show that increases in income inequality (measured by the Gini-Index) con-
tributed significantly to the rise in consumer credit in the USA.Similarly,
Krueger and Perri (2006) find that an increase in income inequality does not
lead to a corresponding increase in consumption inequality. Further empiri-
cal evidence for relative consumption concerns can be found in Boushey and
Weller (2006),Bowles and Park (2005),Neumark and Postlewaite (1998),
Pollin (1988,1990)and Schor (1998).See also van Treeck (2012)for an
overview related to contributions emphasizing this relationship in the contex
of the current crisis.
Arguments from Institutionalist/Evolutionary consumption theory have
lately been addressed in a series of other Post Keynesian models. Dutt (2005
2006,2008,2012)has shown in severalcontributions how debt-financed
consumption, caused by relative consumption concerns, can increase outpu
in the short run but may have negative consequences for output in the long
In Dutt (2005, 2006) it is assumed that conspicuous consumption motives ar
directly related to the willingness of consumers to incur debt. Here conspicu
ous consumption concerns increase workers’ desired level of debt and there
stimulate the economy in the short run. Long-run effects can be positive or
negative,since in the long-run interestpaymentsredistributeincome
from workers to capitalists (who have a lower propensity to consume).
Dutt (2012) introduces a managerial class in addition to the already familiar
classes of workers and capitalists. Here the way to account for conspicuous
56 Jakob Kapeller and Bernhard Schütz
© 2014 John Wiley & Sons Ltd

consumption by workers is structurally similar to Dutt (2006, 2006) and the
effect of debt-financed consumption is again positive in the short run but c
turn negative in the long run. Dutt (2008) introduces conspicuous consump
tion by assuming that workers try to emulate the consumption behavior of
capitalists. Similar to that, Barba and Pivetti (2009) also offer a model wher
workers try to imitate capitalists and emphasize the initialexpansionary
impact of debt-financed consumption. Hein (2012) assumes that the amou
debt-financed consumption depends on capitalists’ savings. In his framewo
each period a fraction of capitalists’ savings goes to workers in the form of
credit,where conspicuous consumption concerns are interpreted to have a
positive impact on the share of capitalists’ savings that goes to workers. He
receives what he calls a debt-led demand regime.7 Long-run implications in
Hein (2012) are consistent with Dutt (2005, 2006). Zezza (2008) presents a
stock-flow consistent model where wealth effects through rising house pric
as wellas conspicuous consumption effects cause an increase in aggregate
consumption. Like in Barba and Pivetti (2009), Dutt (2008) and Hein (2012)
is workers thatemulate the consumption behavior ofcapitalists.Finally,
Palley (2010)—building on thecontributionsby Keynes (1997 [1936]),
Duesenberry (1971 [1948]) and Friedman (1957)—proposes an alternative
way to model individual consumption behavior, calling it the ‘relative perm
nent income theory of consumption’. According to the RPI theory, an indi-
vidual’s marginal propensity to consume is a negative function of the avera
permanentincome within the society.Davancatiand Pacella (2010)and
Kapeller and Schütz (2014) assume that relative consumption concerns ma
primarily within classes, where the latter show how increasing income ineq
ity within the working class can in combination with a Minskyian banking
sector lead to boom and bust cycles.
In contrast to these contributions, the present paper attempts to relate th
issue of conspicuous consumption to the concept of wage-led and profit-led
regimes, which means introducing it into the Bhaduri–Marglin model. Fur-
thermore, in line with Davancati and Pacella (2010) and Kapeller and Schüt
(2014) and in contrast to the other models mentioned before, we will explic
itly model income inequality within the working class, thereby assuming th
relativeconsumption concernsmatterprimarily within a certain socio-
economic group (i.e.among workers).In accordance with Bhaduriand
Marglin (1990) we restrict our focus to the short run. For an analysis how th
exploitation of such a demand regime can lead to crisis in the long run see
Kapeller and Schütz (2014).
7 See also Hein and Truger (2010) and Stockhammer (2011) who argue among others that
pre-crisis expansions in many countries have been debt-led.
Conspicuous Consumption, Inequality and Debt 57
© 2014 John Wiley & Sons Ltd

The rest of this paper is structured as follows.Section 3 introduces the
standard framework ofwage-led and profit-led demand regimesfirst
proposed by Bhaduri and Marglin (1990). Section 4 then modifies this frame
work along the lines just discussed, creating the possibility of a consumption
driven profit-led demand regime. Section 5 discusses the results and puts th
into perspective.
3. THE BASELINE SCENARIO: A SIMPLE MODEL WITH
WAGE-LED DEMAND
In this section we introduce the basic concept of wage-led and profit-led
demand regimes which was first introduced by Bhaduri and Marglin (1990),
leaving conspicuous consumption concerns still unconsidered. Several auxil-
iary assumptions are made to keep the model as illustrative as possible.
As we show later in the paper,the theoreticalpossibility of a profit-led
regime does not rely on positive effects coming from net exports. Therefore
we assume a closed economy to keep the illustration as simple as possible.
equilibrium, consumption and investment demand must be equal to aggre-
gate output, wheretotal consumption consistsof consumption ofthe
working class and the capitalist class:
Y C C I= + +w c (1)
Prices p are set as a mark-up on unit labor cost:
p m w N
Y
w N
Y
= +( ) +
⎛
⎝⎜ ⎞
⎠⎟1 1 1 2 2
n n
(2)
where m denotes the mark-up, wn the nominal wage rate and N the number
of workers. Our working class consists of two groups: workers of type 1 and
type 2,where we assume thatboth groups initially earn the same wage
(w1 = w2). The levelof output determines employment and productivity is
assumed constant over time as well as across groups. From (2) it follows tha
the share of profit income is given by
h w N
Y
w N
Y h= − − = − − − −( )1 1 11 1 2 2
1 1Ω Ω (3)
where w stands for the real wage rate, Ω1 (=w1N1/Y) for the wage share of type
1 workers and the income share of type 2 workers is expressed as residual o
the other two income shares. In what follows we assume (just like Bhaduri
58 Jakob Kapeller and Bernhard Schütz
© 2014 John Wiley & Sons Ltd

and Marglin,1990),an increase in the profitshare.However—trying to
capture pre-crisisdevelopmentsin a stylized way—we assume thatthe
increase in the profit share does not affect all workers in the same way. In
particular we assume that the rise in the profit share happens exclusively a
the expense of type 2 workers, while type 1 workers are able to attain their
share in national income. Thereby a rise in the profit share increases incom
inequality within the working class. Independently of how this relative redu
tion in income of type 2 workers comes about, increasing the profit share b
the amount ψ in such a way must fulfill the following condition:
h h+ = − − − − −( )ψ ψ1 11 1Ω Ω (4)
where we interpretthe leftside of the equation as the new profitshare
(h′ = h + ψ)and the term in brackets as the new income share oftype 2
workers (again expressed as a residual).
Capitalists consume a constant fraction cc of their income:
C c h Yc c= +( )ψ (5)
Workersconsume a constantfraction cw (where 1 > cw > cc > 0)of their
income:
C c Yw w1 1= Ω (6)
C c h Yw w2 11= − − −( )Ω ψ (7)
Investment depends on the rate of return, where the latter can be decomp
into the determinants profit share and capacity utilization (z = Y/Y*, where
Y* denotes output at full capacity utilization) to yield the following invest-
ment function:8
I I h z= +( )ψ, (8)
Here we apply the usual definition that production capacity (i.e. the capital
stock) is fixed in the short period and investment only adds to capacity onc
the short period (= period that firms need to install the additional machine
has elapsed (for a discussion see King, 2008). Furthermore it is assumed th
there always exist a sufficient number of available workers and that firms
finance investment entirely by means of credit. Furthermore we presume a
8 See Bhaduri and Marglin (1990). The rate of profit can be written as
r = Π/K = (Π/Y)(Y/Y*)(Y*/K) = (h + ψ)za,where Π denotes totalprofits and the ratio of full
capacity output to the capital stock (a) is assumed to be constant.
Conspicuous Consumption, Inequality and Debt 59
© 2014 John Wiley & Sons Ltd

endogenous credit money economy in which granting a loan does not requir
the transfer of some kind of loanable funds (see, for example, Lavoie, 1992;
Godley and Lavoie, 2007).
Case 1: A wage-led regime
We assume that full capacity output Y* is given in the short period and equa
to 1 (which means normalizing all relevant variables as proportions of full
capacity output; see Bhaduri and Marglin, 1990). Inserting (5)–(8) into (1)
and dividing by Y* (with z = Y/Y*) yields
z c h z c z c h z I h
C C C
= +( ) + + − − −( ) + +c w w
c w w
ψ ψ ψΩ Ω1 1
1 2
1 ,, z( ) (9)
Differentiation gives the following classical result:
d
d
w c
z
I c c z
S Iz zψ
ψ
=
∂
∂ − −( )
− (10)
Here Sz = (h + ψ)(1 − cc) + (1 − h − ψ)(1 − cw) and Iz =∂I / ∂z are related to
the Keynesian stability condition, according to which the existence of a stab
equilibrium requires thatthe reaction ofsaving to a change in capacity
utilization (Sz) exceeds the reaction ofinvestment(Iz), which gives us a
positive denominator in the above equation. Since income distribution is not
allowed to change endogenously,any setof parameters for which Sz > Iz
assures the existence of such an equilibrium (see Bhaduri, 2008, 150).9 For
any parameter constellation that violates the Keynesian stability condition,
the goods market multiplier would be infinite and output would go towards
either zero or infinity. This basic structural feature applies to all variants of
our model—hence,the model’s results are only viable when the relation
Sz > Iz actually holds.
The above equation describes the effect of a rise in the profit share on
capacity utilization that happens exclusively at the expense of type 2 worke
Since the Keynesian stability condition requires a positive denominator in th
above equation,the sign of (10) solely depends on the numerator:If the
positive effect on investment (∂I / ∂ψ) exceeds the negative effect on con-
sumption (−[cw − cc]z), the fraction in equation (10) will be positive and the
demand regime profit-led.
9 See also Schütz (2012) who discusses this issue for a structurally similar model that allows fo
endogenous variations in the profit share.
60 Jakob Kapeller and Bernhard Schütz
© 2014 John Wiley & Sons Ltd

However, if we assume that the empirical relevance of the positive invest
ment effect is rather small(as stagnating realinvestment growth in many
countrieswith increasing incomeinequality—among them theUSA—
suggests, see Stockhammer, 2005–6)—to state our point clear let us assum
I = I(z)—equation (10) reduces to (11) and the possibility for a profit-led
regime disappears.
d
d
w cz c c z
S Iz zψ =− −( )
− (11)
For reasons of exposition we keep the assumption of I = I(z) also for the ne
section.10
4. AN EXTENDED MODEL WITH PROFIT-LED DEMAND:
INTEGRATING CONSPICUOUS CONSUMPTION MOTIVES
In accordance with Veblen’s concept of conspicuous consumption we assum
that type 2 workers and type 1 workers share a common social identity (Ho
and Terry, 2000) and, hence, type 1 workers serve as a reference group of
2 workers implying that the consumption aspirations of the latter will depen
on the consumption of the former. In particular, as type 2 workers fall behin
in terms of income, they will be concerned about their consumption relativ
to type 1 workers. As already emphasized we assume that relative consum
tion concerns are mostly relevant among workers as compared with betwe
workers and capitalists. While this is obviously a simplification, we think tha
the first channel is more relevant as the argument of a common social iden
is stronger in this case. Furthermore we assume that there is only upward-
comparison,meaning thatconsumption decisions oflow-income workers
(type 2) do not influence consumption decisions of the higher income work
(type 1).
Therefore, we assume that the consumption function in (7) describes the
consumption behavior of type 2 workers as long as their income is not less
than those of type 1 workers. As soon as the income of type 2 workers falls
below that of type 1 workers equation (7) willonly describe a part of the
consumption decision, since it does not account for any potential desire to
10 This actually takes us from the Bhaduri–Marglin model (whose contribution it was to allow
for profit-led demand regimes) back to the Amadeo (1987) version of a Rowthorn–Dutt mode
(Rowthorn, 1981; Dutt, 1987), which only permits wage-led regimes. Those readers who pref
to stick to the assumption of a positive∂I/∂h can do so by just adding∂I/∂h to the numerator of
the results presented in the next section, since this is the only way how it would enter.
Conspicuous Consumption, Inequality and Debt 61
© 2014 John Wiley & Sons Ltd

keep up with their peers (type 1 workers). Therefore we introduce β = N2/N1
to account for the proportion of workers whose income is depressed (our typ
2 workers) relative to those workers whose income stays constant (type 1)
and replace (7) by the following equation once type 2 workers fall behind in
income (Kapeller and Schütz, 2014):
C c h Y Cw w w2 1 11 1= −( ) − − −( ) +α ψ α βΩ (12)
Consumption behavior as described in (7) is also reminiscent in (12), but its
influence is weakened. If α = 1, workers would want to exactly hold on to th
consumption level of type 1 workers, while with α = 0 equation (12) reduces
to (7) and we would exclude this kind of relative consumption concerns. The
higher the desire to keep up with the other group is, the larger is α.
Case 2: A consumption-driven profit-led regime
As long as consumption aspirations of type 2 workers do not exceed their
disposable income, these workers do not have to incur debt. In this case the
equilibrium condition as written in (9) takes the following form (where we
already use the assumption of∂I/∂ψ = 0):
z c h z c z c h z C
C C C
= +( ) + + −( ) − − −( ) +c w w w
c w w
ψ α ψ α βΩ Ω1 1 1
1 2
1 1 + ( )I z (13)
Differentiating gives the following result:
d
d
c wz c c z
S Iz zψ
α
= − −( )[ ]
−
1 (14)
where Sz = (h + ψ)(1 − cc) + Ω1(1 − cw) + (1 − h − Ω1 − ψ)[1 − (1 − α)cw] −
αcwΩ1β and Iz = dI/dz.
As we can see, the result in (14) can be positive or negative depending on
the size of α.If the relative consumption effect is rather small,we get the
standard wage-led result.Conversely,if relative consumption concerns are
strong enough—i.e.if (1 − α)cw < cc—the demand regime willbe profit-led,
giving us the consumption-driven profit-led demand regime mentioned at th
beginning.11
11 See on this also Wisman and Baker (2010) who claim that the period preceding the 2008 cris
was characterized by a profit-led regime, however without presenting any theoretical framewo
62 Jakob Kapeller and Bernhard Schütz
© 2014 John Wiley & Sons Ltd