JOURNAL OF POST KEYNESIAN ECONOMICS

2018, VOL. 41, NO. 2, 260–283

https://doi.org/10.1080/01603477.2017.1348902

none defined

On the causes of the changes in income shares: Some

reflections in the light the United States experience
Antonella Stirati

ABSTRACT KEYWORDS
The last 30 years have witnessed a dramatic change in the Changes in income shares in
distribution of income, with the wage share falling in all major the United States; income
industrialized countries. Main-stream analyses, including New shares; post Keynesian
analyses of income
Keynesian ones, which retain the notion of factor substitution
distribution
leading to a “factor intensity” inversely related to its rate of
return, have encountered some difficulties in the interpretation JEL CLASSIFICATIONS
of this change. Nonmainstream approaches present an B50; D33; E25; J30
advantage in the explanation of the phenomenon, consisting
in the fact that they entail no a priori connections between the
changes in distribution and the changes factor proportions.
Hence if a change in institutions or in the bargaining strength of
the parties affects distribution, income shares may vary
significantly (i.e., changes in wages need not be accompanied
by changes in labor to output ratio in the opposite direction as
in mainstream analyses). Yet empirical observation may
question also some of the analyses that have been advanced
outside the mainstream. The article will explore the ways
in which nonmainstream approaches have interpreted the
described changes in distribution, and assess them from an
analytical viewpoint and with reference to U.S. data. The
purpose is that of pointing at some open questions and
problems.

The main purpose of this article is to critically discuss the explanations of

the observed changes in income shares that have been advanced by non-
mainstream scholars, and in particular to identify the channels through which
the factors identified in this literature would have acted. That is, whether they
are envisaged to have acted directly on workers bargaining power, or mainly
through an exogenous increase in the pure remuneration of capital, or an
increase in business profits and monopoly rents.
The focus will be on the changes in the wage share and its complement, the
overall profit share, without entering in the question, discussed in several
articles,1 of the repartition of the latter among dividends, interests, retained
profits, and rents or between returns to financial and non-financial
capital—even though the issue of the effects of financialization on the wage
share will be discussed in some detail. The question of what has determined

Antonella Stirati is with the Dipartimento di Economia, Università Roma Tre.

Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/mpke.
1
See, for example, Epstein and Power (2003), Palley (2007), and Kim (2013), among others.
© 2018 Taylor & Francis Group, LLC
 JOURNAL OF POST KEYNESIAN ECONOMICS 261

the rise of the profit share as a whole, independently of its repartition between
financial and nonfinancial sector and among different types of capital
incomes, is relevant in itself, and is also connected with the issue of increasing
inequality (Glyn, 2009; Jacobson & Occhino, 2012; Hein, 2015). In addition,
such increase in the overall profit share must originate in the repartition of
the value added produced in the economic system—that is, an increase in
unproductive (noncapital) wealth cannot, per se, increase the profit share as
a whole (for a discussion of this point, see Stirati, 2017, pp. 53–55; see also
Palley, 2007).
The arguments will be developed as follows. In the next section, some
difficulties of the mainstream in explaining the decline in the wage share will
be recalled, and the advantages of an alternative approach whereby income
distribution is explained by the bargaining power of the parties, and employ-
ment levels depend on effective demand, will be highlighted. The subsequent
three sections are aimed at clarifying the alternative theoretical framework
and the different ways in which the causal relationships between distributive
variables have been conceived by nonmainstream economists, particularly
within the framework of the surplus approach. The subsequent section
discusses the arguments advanced in heterodox and post-Keynesian literature
concerning the effects of globalization and financialization on income distri-
bution and attempts to disentangle the ‘channels’ through which they would
affect the latter. Finally, Some Evidence from the US Economy over the last
50 Years section presents descriptive data concerning the U.S. economy and
discusses them in the light of the arguments surveyed in the previous sections.
The data are consistent with an increase in net returns on capital between the
late 1970s and late 1990s driven by high real interest rates, whereas during the
2000s we observe a somewhat puzzling divergence between the two variables.
Overall however, direct influences on workers bargaining position also seem to
have been relevant.

Changes in income distribution and alternative approaches to

economic theory
The last 30 years have witnessed a dramatic change in income distribution,
with the wage share falling in all major industrialized countries (Figure 1).2
These changes do not appear to be even partly attributable to a
trend increase in the ratio of the value of the capital stock to value added,
which is not evident in the data for the major economies (Figure 2). An
2
The data reported here refer to the wage share adjusted by imputing to the self-employed a labor income equal to
that of employees; by this definition, the share is not affected by changes in the proportion between employees
and self-employed workers. GDP at factor costs is net of taxes on production. The adjusted share therefore is the
same as the ratio of average labor income per worker (or standard labor unit) to average product per worker (or
standard labour unit): Adjusted wage share = [(Compensation of employees/employees) × total employed]/VA
(value added) at factor cost.
262 A. STIRATI

Figure 1. Adjusted wage share as a percentage of gross domestic product at factor costs,
1960–2015.

Figure 2. Net capital stock on gross domestic product at constant market prices, 1960–2015.

increase in that ratio would obviously entail, for any given profit rate, an
increase in the profit share (and a fall in the wage share); however, this
would not reflect an increase in the return on capital and hence a change
in distribution as it is properly understood, but rather a structural change
in the economy.3

3
In discussing the data, I am taking the classical rather than neoclassical view. Therefore, I am not supposing any
predetermined relation between the desired capital output ratio (which will depend on economic structure and
relative prices) and the rate of profit. What I am considering here is merely the accounting fact that for a given rate
of profit, an increase in the value of the capital stock would entail increased total profits. There is sometimes
ambiguity in post Keynesian literature, because changes in income shares may, in a sense, always be treated
as changes in income distribution from a macroeconomic point of view—but the underlying causes may be
different in nature. For example, in the formal analysis presented in Panico et al. (2012) the change in income
shares is entirely caused by a change in the proportion between the value of capital and output and not by a
change in the rate of profit. The analysis therefore does not address the interpretation of the changes that have
actually been taking place in most countries in recent decades. A similar problem arises with Piketty’s (2014)
interpretation of changes in income shares (on this, see Stirati, 2017).
 JOURNAL OF POST KEYNESIAN ECONOMICS 263

Note that the wage-share data tend to underestimate the change in

distribution, because the wage share comprises the salaries of top managers
that are included in firms’ payrolls,4 which rose sharply, particularly in the
Anglo-Saxon countries, and which may be regarded as capturing part of
business profits rather than as labor incomes (Piketty & Saez, 2003; Palley,
2007; Glyn, 2009). In fact, personal income distribution data (cf. Piketty &
Saez, 2006) and the trends in wages of nonmanagerial workers suggest a
sharper change in income distribution in the USA than is shown by income
shares data. The level of real hourly earnings of production workers in the
private sector has actually fallen in the USA since the mid-1970s (see Figure 3),
despite a steady increase in productivity in the same period (gross domestic
product [GDP] at constant prices per person employed rose by 70% between
1973 and 2012; see also Pollin, 2003, p. 43).
The interpretation of such changes in income distribution represents
a challenge to economic analysis. Mainstream analyses, including New
Keynesian ones, which retain, at least in the long run, the notion of factor
substitution leading to a factor intensity5 inversely related to its rate of return,
have encountered some difficulties in the interpretation of the changes
in income distribution we have described. In these models a fall in the
equilibrium real wage rate would have to be associated with a rise in the
labor-to-capital and labor-to-output ratios determined by optimization in
consumption and production.6 This is particularly important in the context
of the explanation of changes in income shares, because if factor substitution
was supposed to take place in the manner described by marginal theory, a fall
in wages (or in the proportion between wage and product per worker), owing
to the increase in the above-mentioned ratios between labor and output and
labor and capital, would probably give rise to small or nil (depending on the
values of elasticity of substitution) changes in income shares (Rowthorn
1999). This, by itself, creates an underlying difficulty in all mainstream
attempts to explain the observed changes in income shares, which are usually
overcome by attributing such changes to labor-saving technical innovation.
These technical changes are, however, hard to identify empirically, and the
methods adopted in applied works have been criticized, among other things,
as often amounting to assuming that changes that cannot be otherwise
explained according to the theory must be due to technical innovations of
4
The share includes compensations in the form of salaries, but not payments in the form of stock options, or as
compensation of professional services.
5
Sraffa (1960) and Garegnani (1970), among others, have shown that the notion of factor intensity itself is devoid of
meaning, because—for any given underlying technique—factor intensity is not independent of relative prices and
distribution.
6
This is the case when, in the familiar model of a price equation and a bargained wage equation, the former is
decreasing in the real wage-employment space because it reflects a decreasing marginal product of labour. If labor
marginal product is assumed constant, as in some textbooks, a change in the bargaining strength of the workers
would shift the bargained wage equation, determining lower equilibrium unemployment, but no changes (given
productivity) in the equilibrium wage. In this case, therefore the model could not be used to explain changes in
distribution.
264 A. STIRATI

Figure 3. Hourly real wage level of production and nonsupervisory workers in U.S. private
sector, 2010 dollars.

the appropriate type (Stockhammer, 2009, pp. 19–22; see also Kohler et al.,
2015, for econometric results that undermine the role of technical innovation
in negatively affecting the wage share in Organization for Economic
Co-operation and Development [OECD] countries).
Besides the role of labor-saving technical innovations, another, more
appealing, explanation of changes in income distribution lies in the role of
globalization in increasing unskilled labor supplies through various channels
(immigration, off-shoring of intermediate production, imports of products
from emerging economies). Again, however, when this is treated within
mainstream economic models that retain well-behaved factor substitution
and the consequent tendency to full employment of factors, the expected
results are at variance with facts in some important respects. In particular,
although according to mainstream trade theory, globalization should lead to
a fall in relative unskilled labor incomes in advanced economies, it should also
improve the relative wages of unskilled versus skilled workers in emerging
ones, a phenomenon which is not generally observed (International Monetary
Fund, 2007, p. 176). In addition, it must be noted, as pointed out by Krugman
(2007), that what has been witnessed in industrial countries and particularly
the United States, in recent decades, is not a change in the relative positions
of skilled and unskilled workers, but rather the inability of both groups
of workers to benefit from increased productivity, to the advantage of a tiny
minority of the population.
Although the heterodox approaches are multifaceted, many post Keynesian
and nonmainstream scholars can be represented as holding a conflict and
institutional explanation of income distribution between wages and profits,
while output and employment are determined by effective demand. The
advantage of such an approach to the phenomenon of distribution and
 JOURNAL OF POST KEYNESIAN ECONOMICS 265

employment is that it entails no a priori connections between the changes in

distribution and the changes in the proportion between labor and output, or
between the value of the capital stock and the value of output; accordingly,
decreases in nominal and real wages are not seen as the route that can lead
to full employment, because employment levels depend, even in the long
run, on the principle of effective demand.7 Therefore if, for example, a change
in the bargaining strength of the parties affects income distribution, the effects
on income shares may be significant.
Yet empirical observation may pose some questions also on the analyses of
distribution that have been advanced within nonmainstream approaches. In
the following sections I will be exploring the ways in which the perspective
just described might be used to interpret the observed changes in distribution.
In doing so, I will adopt the analytical standpoint that can be labelled for short
Classical Keynesian, because it combines the classical surplus approach to
income distribution with the principle of effective demand as the theory of
output. My objective, which is primarily to frame the questions rather than
finding the answers, is based both on theoretical considerations and empirical
evidence. The latter however is not intended to provide a test for theory, but
rather to contribute to an initial reflection on the open questions.

Mutual relations and causality between distributive variables

We can start considering possible interpretations of changes in income
distribution in the light of the revival of the classical approach to income
distribution by looking at the system of relative price equations:
P ¼ PAð1 þ iÞ þ PAq þ lw
Where P is the relative price vector expressed in terms of a numeraire, A is
the matrix of the production coefficients, i is the interest rate, representing the
pure remuneration of capital, l is the labor inputs vector, w is the real wage,
and finally ρ is the diagonal matrix of the rate of profits of enterprise or
business profits that remunerate risk and illiquidity associated with
productive activity. Note, however, that although risk and liquidity premia
can be supposed proportional to the capital stock, this is not necessarily the
case for top-managerial remuneration, though the latter may be regarded as
a component of business profits adding to risk and illiquidity premia.
The business profit term is there to represent the fact that the return on
capital invested productively must be higher than the interest rate obtained

7
The main analytical premise of these views is to be found in the criticism of the principle of factor substitution
(Garegnani 1970). In contemporary macromodels the tendency of the economy to potential output is generally
attributed to the so-called real balance effect and Keynes effect. However, the former cannot by itself be regarded
as capable of ensuring that tendency, and the inverse relationship between aggregate investment and the interest
rate (the Keynes effect) must in the end rely precisely on factor substitution (see Petri, 2004, Chapter 7). Concern-
ing the role of effective demand in determining growth see Cesaratto and Mongiovi (2015).
266 A. STIRATI

on safe financial investments (such as long-term public bonds) to compensate

for risk and illiquidity associated with productive investments, which are
likely to differ across industries. Thus although competition must equalize
the pure remuneration of capital across all types of investment, total profit
rates, including both interest rate and business profits, do not necessarily
equalise. In addition, the diagonal matrix ρ may also contain monopoly extra-
profits earned in some industries because of barriers to entry.8
In the classical tradition, the relation between these distributive variables
has been conceived in different ways. Most of the old classical economists
thought, like Ricardo, that the real wage was the given variable, determined
by socially established subsistence requirements and by the relative bargaining
position of the workers (Stirati 1992, 1994, 1999). Thus, profits would be
determined residually. Given the normal rate of business profits, the resulting
pure remuneration of capital used in production would ultimately govern the
rate of interest in credit markets (since a rate of return on capital higher than
the rate of interest would stimulate a demand for loans too high with respect
to savings and hence a tendency for the rate of interest to increase—and vice
versa in the opposite case—cf. Ricardo, 1821/1951, p. 364). It was recognized,
however, that the interest rate could differ from the remuneration of capital
determined in production for some length of time: According to Ricardo,
for example, “for the last twenty years, the Bank … have lent money below
the market rate of interest’ (1821, p. 364, italics added for emphasis; the term
“market rate” means the remuneration of capital obtained in production).
Thus the causal links went from real wages to the profit rate and from the
latter (with given business profits) to the interest rate. Marx’s conception
was different, though not always spelt out entirely clearly and consistently:
He retained the notion of a given real wage rate determined by necessary
consumption and class relations, and hence a rate of profit in its entirety
determined residually but saw the rate of interest also as a magnitude that
could be independently determined, leaving the business profits to be treated
as the ultimate residual variable. This conception indicates the possibility of
a tension between the interests of financial and productive capital (for a
discussion of these approaches, see Pivetti, 1991, pp. 61–69; Panico, 1988,
Chapters 1–3).
In the current revival of the classical approach to distribution two main
lines have been pursued, which will be briefly described below: that of
determining the rate of profit as a whole by means of the so-called
Cambridge equation and that of accepting the broad view of the classical
8
Extra profits may be earned in some firms/sectors not only owing to barriers to entry, but also thanks to better than
dominant technique, that is to some form of more or less transitory technological monopoly allowing reduction of
costs below those associated with the dominant technique. In this case such extra-profits would have the nature
of rents and are not price-determining, that is, they are not included in the term ρ in the price equations which
must refer to the dominant (i.e., price-determining) technique. On the other hand, such rents are necessarily part
of the empirically observed size of the operating surplus (see note below for a definition).
 JOURNAL OF POST KEYNESIAN ECONOMICS 267

economists who saw income distribution as the outcome of class conflict,

but attributing the role of the independent variable mainly to the interest
rate.

The rate of growth as a determinant of the rate of profit

One very controversial way of determining distribution within the Classical
Keynesian approach is the so-called Cambridge equation (Kaldor, 1956;
Pasinetti, 1962), which establishes a link between the rate of growth of the
economy and the rate of interest. Under the simplifying assumption of zero
savings out of wages the equation takes the form:
G ¼ sp r
where G is the rate of capital accumulation, sp is the propensity to save out of
profit income, and r is the profit rate.
The analytical criticism of this way of determining the rate of profits that
has been voiced is that, if G is the actual rate of accumulation, r appearing
on the righthand side is not the normal, but the ex-post actual profit rate
reflecting the actual degree of utilization of capacity. The ex-post profit rate
changes with capacity utilization, totally irrespective of changes in the real
wage, given the techniques in use. This contrasts with the nature of the rate
of profit that appears in the price equations, which is the normal profit rate,
associated with (a) the dominant technique, described by the technical
coefficients in the equation, and (b) a normal or desired degree of capacity
utilization. Only under the latter conditions the normal rate of profits has a
definite, inverse relation to the real wage rate, and can therefore, if given,
univocally determine it, as in Sraffa (Ciccone, 1990a; Garegnani, 1992;
Garegnani & Palumbo, 1998; Aspromourgos, 2013). The equation is therefore
irrelevant to the theory of distribution.
Of interest here is that although the Cambridge equation is still used in
Post-Keynesian formal macromodels, to my knowledge no one has attempted
to interpret the changes in income distribution in recent decades with refer-
ence to changes in the rates of accumulation or in the share of investment on
value added, as would be in the logic of the model under discussion. Indeed,
such a suggestion would be at variance with observation. In major industrial
countries the decline in the wage share has in fact gone together with a decline
in the rates of accumulation and GDP growth. It may be noted in this regard
that even the ex-post actual profit share and profit rate that are measured by
the national accounting statistics do not in the long run exhibit the relation
expressed in the equation. The explanation lies in the short-run nature of
the connection between the actual rate of growth and the ex-post rate of
profits, reflecting changes in capacity utilization. The relation is therefore
visible in the data for changes in the rates of GDP growth over the cycle
268 A. STIRATI

(e.g., the profit share tends to fall during recessions),9 but not necessarily so
when looking at longer-term tendencies, since situations of under-utilisation
or over-utilisation of capacity tend to be corrected – albeit not necessarily ever
fully eliminated – by changes in capacity; though some decline in the profit
share associated with lower than desired capacity utilisation might actually
turn out to be rather persistent when the economy is stagnating or declining
over relatively long periods of time.
Given the notoriety of his work, it may also be worth mentioning here an
argument recently advanced by Piketty (2014) concerning the effects of
the growth rate on the profit share. The analysis behind his argument is
embedded in Solow’s growth model, and runs as follows: stagnating
population causes slow growth, and this in turn must determine a (moderate)
fall in the profit rate and, with an assumed elasticity of substitution greater
than one, a more than proportional increase in the capital to output ratio
and hence a rising profit share. These neoclassical arguments however are
outside the scope of the present discussion and have been shown to suffer
from severe shortcomings on both empirical and analytical grounds (Stirati,
2017, pp. 49–51)

Bargaining power and social relations

Most economists adopting the Classical Keynesian perspective tend to regard
changes in income distribution as determined by changes in the bargaining
position of the parties, in turn associated with changes in the economic, social,
and institutional set-up. Even from such a broad perspective, however, the
main channels through which the changes tend to occur may be regarded
as acting primarily directly on real wages or on the interest rates.
As described above, the old classical economists and Marx tended to regard
the real wage rate as the given variable, determined by social and economic
conditions (among which the unemployment and underemployment rates
are of primary importance), with the profit rate determined residually. Some
contemporary economists, however, have argued that this point of view is ill-
suited to represent the process of determination of real wages, because in fiat
money economies workers can only bargain over nominal wages, while firms
can and will generally pass increased money costs of production on to prices
9
Operating surplus or profits in national accounts are a residual concept—what remains of net (or gross) domestic
income after subtracting the (adjusted) wage share. If the wage share is adjusted for self-employed imputed labor
incomes, it measures by definition the ratio between the real wage (in terms of the GDP deflator) and the value
added at constant prices per person employed. The latter is empirically known to move procyclically (Okun’s law)
and therefore for a given real wage, the wage share would tend to rise and the profit share to fall during a
recession. However, there is much empirical evidence that real wages also tend to vary procyclically, so that in
actual fact the wage share may rise or fall in recessions according to which effect prevails. Generally, however,
particularly in the initial phase of a recession, one can observe a fall in the profit share. Note that these empirical
considerations have little to do with mainstream debates over changing mark-ups over the cycle, which confront
problems and use concepts and empirical methods (such as increasing marginal costs, the neoclassical aggregate
production function, total factor productivity) that are entirely embedded in mainstream theory (Stirati, 2016).
 JOURNAL OF POST KEYNESIAN ECONOMICS 269

(Pivetti, 1991, pp. 36–37). On this basis, and following the suggestion by Sraffa
(1960, p. 33) that the rate of profit may be regarded as determined by the rate
of interest, it has been argued that, in general, distribution may be seen as
determined primarily via interest rate policy, acting on the minimum com-
petitive return on capital appearing in the price equations. According to this
view, which is close to the Keynesian tradition of regarding the interest rate as
a conventional and institutional variable, the interest rate is independently
determined by the monetary authorities subject to a number of objectives
and constraints. Because business profits, regarded as risk and illiquidity
premia, must also be conceived as given magnitudes in a competitive
economy, the residually determined variable is the real wage. This remains
true even when business profits earned above the interest rate include, besides
risk and illiquidity premia, extra profits due to the existence of obstacles to
free competition and barriers to entry, because these extra profits cannot
be regarded as arbitrary, but as a definite addition to the competitively
determined benchmark. Thus, for a given nominal wage rate and output
per worker, and given the rate of business profits, an increase in interest rates
causes an increase in the opportunity cost of capital and hence the monetary
costs of production10 and raises the price level vis-à-vis the money wage
(Pivetti, 1991). As a consequence, monetary authorities are regarded as
generally capable of governing income distribution. The role of the interest
rate in determining the price level relative to the nominal wage has gained
acceptance also in wider post Keynesian literature, which sees the interest rate
as a component of the mark-up charged by firms over the wage component of
the costs of production (see, e.g., Hein & Schoder, 2009). In this latter litera-
ture, however, sometimes only actual interest payments made by firms are
regarded as costs entering price formation (Hein 2015, p. 928). However,
the notion that real interest rates must represent a minimum return required
on investments, whatever the source of financing, appears to be analytically
sounder, as it results from consideration of the forces of competition and
profit seeking which are constantly at work in the economic system.
In regard to the connection between the rate of profits and the rate of
interest, a somewhat more cautious and detailed argument has been made
by Garegnani. He agrees that competition in product markets tends to ensure,
over sufficiently long periods of time, that the real rate of interest and the rate
of profit move in step, and that, if the rate of interest is a conventional vari-
able, largely determined by monetary authorities, then it can determine the
normal rate of profit and the corresponding real wage rate. However, accord-
ing to Garegnani, “The policy of the monetary authorities is not conducted in
a vacuum, and the movement of prices and of the money wages determined in
the wage bargain will be amongst the most important considerations in the
10
An increase in the interest rate would increase the opportunity cost of productively invested capital, even for firms
using internal rather than borrowed funds to finance investment.
270 A. STIRATI

formulation of that policy” (Garegnani, 1983, p. 63). This more cautious

position is also supported by Ciccone (1990b) and Stirati (2001), who argue
that the reaction of money wages to a price increase may affect the real
interest rate, for a given money interest rate established by the central bank
and financial markets. Although a once-for-all increase in nominal wages
would have—with a lag—a permanent effect on the price level, with no
consequences for the real interest rate, a continuous increase in nominal
wages would cause price inflation and affect the real interest rate (see Stirati,
2001, for a formal analysis). The causation between the interest rate and real
wages therefore cannot be conceived too mechanically, and the direction of
causation may actually vary according to circumstances.
It is perhaps also worth recalling at this stage that Sraffa explicitly refers
to the surplus component of wages, because the subsistence requirements,
according to Sraffa, are necessary costs and as such should most appropriately
be treated in the same manner as other production costs, and included in the
coefficient matrix (Sraffa, 1960, pp. 9–10). If so, there must be limits within
which wages may be affected by monetary policy, depending on the size of
the surplus component of wages above the subsistence requirements. The
terms subsistence and necessaries are also used by the old classical economists,
whose standpoint Sraffa in his preface claims to be adopting. Necessaries
therefore has the same meaning as it did for them, including, to quote Smith
(1776/1976, V.ii.k.3) “those things which the established rules of decency have
rendered necessary.” Thus, even in advanced economies, the notion of a
subsistence floor continues to be relevant, and may not be so far distant from
current wages. On the other hand, the tendency in contemporary economies
toward a continuous increase in productivity creates scope for increases in the
rate of profits that do not require a fall in real wage levels.

What is the role of globalization and financialization?

Recent discussions and applied analyses of the changes in income distribution
have tended to emphasize the role of globalization and financialization in
affecting income distribution over recent decades. Some of the arguments
that have been advanced, particularly in non-mainstream literature, will
be summarized and commented on here with a view to establishing how
(i.e., through what channels) they can have affected class relations and
economic institutions and hence the bargaining position of the parties.
The phenomenon of globalization, involving increased competition from
emerging economies in product markets, off-shoring and immigration, can
be expected to directly affect the bargaining position of workers in advanced
economies by increasing job losses and unemployment, or by representing a
powerful threat that jobs will be lost (because of delocalization or imports
from other countries) as a consequence of higher wages or improved work
 JOURNAL OF POST KEYNESIAN ECONOMICS 271

contracts and conditions (Bronfenbrenner 2001).11 In addition, one could argue

that the impact of globalization works through other channels as well; of
particular importance are the constraints on macroeconomic policies that are
imposed by free capital mobility (e.g., on public budget and public debt manage-
ment), which can then affect the formation of aggregate demand. All in all, it
thus would appear that many of the factors that are synthesized as globalization
are likely to directly affect the bargaining position of workers through their
impact on employment growth and unemployment and the threat of
delocalization of production and job losses. Several observers, for example, have
referred to job insecurity as a main factor in subdued nominal wage dynamics in
the United States even in the 1990s, a period of low and falling unemployment:
Greenspan (then governor of the Federal Reserve) referred to “traumatized
workers” and their job insecurity as a factor in explaining low inflation even
in a period of sustained growth and low unemployment rates (Greenspan,
1997; Choi, 2001; see also Pollin, 2003, pp. 50–56) and Gordon (1997, p. 30)
in 1997 wrote in this context, “The 1990s have been a time of labour peace,
relatively weak unions, a relatively low minimum wage” and pointed to inten-
sified world competition in product and labor markets, and the increased
inflows of immigrant labor in the United States as likely explanations.
Many discussions of changes in functional income distribution carried out
by nonmain-stream, post-Keynesian economists point to the role of financia-
lization (see Hein 2011, 2015; Palley 2007; Kohler et al. 2015). Yet my
impression is that in this regard the analysis of the causal mechanisms
envisaged, apart from the effects of higher interest rates discussed in the
previous section, are not yet entirely clear, or satisfactory, though some of
the econometric exercises carried out confirm a negative correlation
between some indicators of financialization of the economy and the wage
share (Stockhammer, 2009; Kohler et al., 2015).
With regards to the role of financialization in affecting income distribution,
the following considerations arise. First, it is perhaps worth clarifying that,
from a macroeconomic point of view, the increases in the wealth of top man-
agers or investors in financial markets in the form of capital gains cannot be
regarded as affecting the distribution of income. The moment any individual
agent wishes to use that wealth to finance current consumption the assets
must be sold, so that the aggregate effect on the private sector as a whole is
nil. On the other hand, if increased wealth obtained in the form of capital
gains can be used as collateral to obtain additional credit, as was largely the
case in the recent past, the outcomes of this for the distribution and uses of
current output differ according to whether it is assumed that the latter is
(on the average over economic cycles) given and equal to potential output
11
At the same time, in contrast with predictions deriving from mainstream theory, globalization does not necessarily
improve the relative position of unskilled vs. skilled workers in emerging economies, owing to large labor reserves
and/or to general political and institutional conditions in those countries.
272 A. STIRATI

or is instead considered elastic, even in the long run, to changes in aggregate

demand (Garegnani 1992, 2015/1962). Under the latter assumption, which is
consistent with the approach taken here, the transformation of capital gains
into credit is the same thing as any credit-financed increase in autonomous
expenditure, and would not therefore encroach upon someone else’s income
and consumption, but cause an increase in output. A similar conclusion can
be drawn with regard to the effects of increased asset value on the propensity
to consume.
However, there may be a sense in which the high earnings of individual man-
agers and firms in the financial sector, even in the form of capital gains, might
have affected income distribution – i.e., by affecting the social norms concern-
ing income levels, and the opportunity costs and expected earnings of indivi-
duals and firms operating in other sectors of the economy. In this regard,
however, the question within the present framework of analysis is whether
and to what extent such change in the social norms concerning the legitimate
remuneration of managers may actually be regarded as an independent factor in
affecting distribution, that is, whether it can be regarded as a cause rather than
as a consequence of changes in the bargaining position of the workers.
Another connection between financialization and a higher normal profit
rate might lie in the possibility that the high returns (mostly in the form of
capital gains) realized by means of financial or real estate transactions have
produced a change in the rate of return on capital regarded as the minimum
competitive benchmark, from the interest rate on riskless long-term bonds to
the (higher) returns on some different portfolio of assets.
A further argument connecting financialization with higher normal profit-
ability much emphasized in Post-Keynesian literature concerns the changes
that it is believed to have brought about in the governance of firms towards
stock-holder value orientation, focusing on short-term realization of high
profits. Institutional changes such as the possibility of hostile takeovers and
the performance-related remuneration of managers would have changed the
latter’s priorities from the growth of the firm (financed by the re-investment
of retained profits) to distributing high dividends to stock-holders and in
general would have led to a greater power of stock-holders in defining the
objectives of the firm. This is more an argument about the utilization of prof-
its than about an increase in profitability; however, there are considerations in
this literature about the effect of this changes in the governance of firms on
employment and workers’ bargaining positions (Palley 2007; Stockhammer
2004; Hein 2011).
On the other hand, the enormous increase in financial transactions and
leverage that have been made possible by the deregulation and expansion of
financial markets seems likely to have increased the profits of financial
corporations for any given capital stock, out of commissions and interest
payments, thus capturing an increasing share of national income. If, however,
 JOURNAL OF POST KEYNESIAN ECONOMICS 273

the financial sector is largely a monopolistic sector, and if, as has been argued,
many of the financial products and services can be regarded as luxury,
nonbasic goods12 (Barba and de Vivo 2012), then such high profits would
not have affected the normal rate of profit in the other sectors of the economy,
though they would show in national accounting data for aggregate income
shares in proportion to the weight of the financial sector. According to OECD
national accounting data, however, in the United States between 1987 and the
beginning of the 2000s the proportion of gross operating surplus to value
added in the financial sector increased from 35% to 45% (roughly in line with
what happened in some of other sectors such as manufacturing and transports
and communications, see Figure 6)—so that from the point of view of the
trends shown in aggregate national accounts, and in consideration of the fact
that this sector weight in the aggregate GDP at its maximum was just below
8%, the financial sector has not provided a determining contribution to the
upward trend in the overall ‘profit’ share.13
Summing up, the arguments about the effects of financialization on
changes in manager compensation norms in the economy as a whole and
the change in the benchmark return on capital and shareholder value would
imply, within the approach outlined above, that for any given real long-term
interest rate there will tend to be an independent increase in normal profita-
bility. On the other hand, increased monopoly profits in the financial sector
would not affect normal profitability in the economy at large, but would in
principle result in a fall of the aggregate wage share (and a mirror increase
in operating surplus) proportional to the weight of the financial sector in
the economy. U.S. data, however, do not indicate that the latter was a major
cause of the changes in observed income shares.

Some evidence from the U.S. economy over the last 50 years
The purpose of this preliminary examination of empirical evidence and
historical reconstruction is to provide some insights into two questions.
The first is whether the evidence is consistent with the idea that real interest
rates and profit rates actually tend to move in step. The second is whether
interest rate policy can be regarded as an independent factor in determining
changes in income distribution. Recalling the analysis summarized in the
Bargaining Power and Social Relations section, an increase in nominal interest
rate, according to the modern revival of the classical approach, can affect
distribution by causing a rise in the price level vis-à-vis the nominal wage.
12
That is, they do not figure in the costs of production of wage goods or other basic goods, so that the interest and
fees paid to obtain them may be conceived as expenditures in luxury consumer goods or services, such as
gambling games.
13
The financial sector has increased its weight in GDP between 1987 and 2015 from 6.3% to 7.2%, and because its
operating surplus as a proportion of value added is higher than the economy average, it has contributed to the
increase in the profit share through a composition effect, which is however of small impact.
274 A. STIRATI

On the other hand, nominal wage inflation might be able either to resist
the distributive effect of higher nominal interest rates or to determine an
autonomous push toward higher real wages. However, it appears rather
difficult to disentangle these different directions of causation by looking
directly at nominal variables. I will therefore simply try to assess (a) whether
real wage increases are indeed associated with periods of higher nominal wage
inflation (which would suggest that bargaining over nominal wages matters
for income distribution); and (b) whether there are elements in the overall
institutional, economic and labour market conditions that suggest the
relevance and priority of conditions directly affecting wage bargaining with
respect to autonomous changes in the policies of the monetary authority.
The reconstruction, however, is descriptive and is not intended to represent
a test of any particular theory, but rather aims at stimulating questions and
further reflections on these complex issues.
Naturally, statistical data cannot provide information on the normal profit
rate as defined above, but only on the actual rate affected by the degree of
capacity utilization and earned on the existing capital stock, which is not
necessarily of the kind associated with the dominant technique and includes
quasi-rents on economically or technically obsolete plants as well as rents or
extra-profit connected to elements of technical monopoly. However, data on
long-term trends in income shares (i.e., on the proportion of real wages to
output per worker) and ex-post profit rate may provide an acceptable
indication of the direction and order of magnitude of the changes.
The choice of the United States reflects that country’s leading role and the
fact that it is accordingly the least constrained in fixing interest rates.
The long-term real interest rate on government bonds began to increase
sharply in 1980 and reached a peak in 1983, as a result of the change in
nominal interest policy beginning in 1979 (Figure 4). It subsequently fell,

Figure 4. Nominal and real long term interest rates in the United States, 1961–2015.
 JOURNAL OF POST KEYNESIAN ECONOMICS 275

Figure 5. Adjusted wage share in the United States, 1960–2015.

stabilizing at around 4.5% in the period 1989–1997, about two points above
the values observed in the early 1960s and about 3.5 above the—historically
low—average values in the period 1969–1979.14 After 1997 real interest
rates began to fall and have remained below 2 per cent since 2003, despite
a moderate recovery between 2005 and 2007.
If we now look at the long-term tendencies in income distribution over the
same period (Figure 5), we see that the wage share for the entire economy
tended to decrease since the late 1970s; the declining long term trend in
manufacturing was sharper, though it apparently began later; this however
may be due to the stronger countercyclical behavior of the wage share in
manufacturing, which tends to increase at the outset of a recession: in fact
after the mid-70 s we observe three short run increases: in the severe recession
of 1980–1982, in 2000–2001, and 2008.
Data on gross operating surplus as a share of value added in various
subsectors of the economy (available as a continuous series only since
1987) also confirm that operating surplus and mixed income15 share of output
increased—albeit to a different extent—in all sectors of the economy (Figure 6;
note that although the mining industry exhibits a very sharp increase in the
gross operating surplus, its weight in the economy is limited (1% in 2009)
so that it is not a major factor in the overall trend).
Consistently with the picture just outlined, the national accounting data
on the net returns on the net capital stock (Figure 7) indicate, after the
1980–1983 fall associated with the recession, an increasing trend which
14
The average real interest rate of the period 1969–1979, calculated excluding the 2 years in which interest rates
became negative, was 0.9%.
15
Longer historical series for compensation of employees, operating surplus and capital stock by industry are
available from the U.S. Bureau of Economic Analysis. However, U.S. Bureau of Economic Analysis statistics are
based on national definitions that differ from the internationally accepted system of national accounts and are
not comparable with the data of other countries. I have therefore chosen to use only the statistics for the United
States provided by international sources such as the OECD, which follow the internationally comparable
definitions, at the cost of unavailability of the industry capital stock series and the availability of other series only
since 1987. The operating surplus in Figure 6 is not adjusted (see Footnote 1 above) and hence comprises the
entire incomes of self-employed workers (mixed incomes).
276 A. STIRATI

Figure 6. Gross operating surplus and mixed incomes as a proportion of gross domestic
product.

reverses the decline experienced between the mid-1960s and the end of
the 1970s.16
Thus, overall the change in net returns on net capital stock between
the late 1970s and the period 1995–1997 appears broadly consistent with a
gradual adjustment of income distribution to the rise in the real interest
rate—data not shown in the figure show that the return on capital increased
from about 5.4 on average in the 1970s to an average 7.5 in 1995–1997
(Ameco Database, 2011), somewhat less than the increase in real interest
rate. These data, however, must be considered in view of the fact that the
large increases in top managers’ salaries are included here in the aggregate
of the compensation of employees and tend therefore to underestimate
the redistribution of income away from non-managerial labour (see the

16
Though the return on fixed capital increased, it could be wondered whether this has to do with an increase in the
cost of intermediate inputs (circulating capital), particularly primary commodities, per unit of output. Some
prudence is necessary when evaluating data on intermediate inputs, because on the one hand, they depend also
on the degree of vertical integration in production, and on the other, and more importantly, the speed at which
intermediate inputs circulate is certainly much higher than the accounting period (1 year). This means that the
actual proportion of circulating capital over value added on average during the year amounts to a much lower
figure than that in the data (that is, if the velocity of turnover for intermediate inputs were one month, the
proportion based on annual data should be divided by 12 to obtain the average proportion of advanced
circulating capital on value added during the year). It should also be considered that turnover rate may vary
significantly over time and across sectors owing to several circumstances, including demand conditions. This said,
national accounting data show that the proportion of intermediate inputs to value added has fluctuated over the
period without reflecting on income shares. After 2005 the proportion rises somewhat above previous historical
peaks up to a maximum of 2 percentage points—this does not seem to be a relevant figure in the light of the
above considerations (OECD, structural statistics).
 JOURNAL OF POST KEYNESIAN ECONOMICS 277

Figure 7. Net returns on net capital stock in the United States, 1960–2015; 2010 = 100.

Changes in Income Distribution and Alternative Approaches to Economic

Theory section).17
Since 2000, however, real interest rates begin to decline significantly and
after 2003 tend to stabilize at historically low values (below 2%), whereas
on the other hand the net return on net capital stock stabilises and even
increases after the 2008–2009 crisis.
The evidence for the last decade might thus be consistent with an
exogenous increase in the mark-up such as could have been brought about
by financialisation (e.g., by changes in the norms on managers’ compensation,
and benchmark return on capital; see the What is the Role of Globalization
and Financialization? section), but also, alternatively, with the view that both
real wages and interest rate can vary independently for some length of time,
thus leaving business profits to be determined residually. In the period under
discussion, nominal wage dynamics remained the same as in previous years,
fluctuating between rates of increase of 2% and 4% according to the cycle
(Figure 8), and was likely affected by the same economic and institutional
framework: The continuing adverse effects of globalization, an increase in
unemployment from 4% to 6% between 2000 and 2003, and a Republican
presidency may all have contributed to the insecurity of workers noted by
several observers during the 1990s. At the same time, for a number of reasons,
since the beginning of the 2000s the Federal Reserve successfully pursued low
nominal and real interest rates. An interest rate lower than the returns
on other forms of investments may have contributed to the acceleration of
financialization, since it provided an incentive to borrow and reinvest in
financial markets or real estate (much less so in production, because aggregate
productive investments are influenced, and constrained, by aggregate
demand). It also provided a rational foundation for the expectation of increas-
ing stock market and real estate values, favoring the then self-reinforcing
17
The influence of rising real interest rates on profit rates is supported, albeit indirectly, by Epstein and Power (2003)
who find that there is no evidence of an inverse relation between non-financial and financial profits.
278 A. STIRATI

Figure 8. Annual rates of change: Hourly wages of production and nonsupervisory workers in
the United States, 1965–2015.

bubbles. It is generally recognized that a major objective of the Federal Reserve’s

low interest rate policy was precisely that of sustaining financial markets. This,
however, leaves open the question of why a tendency toward a lower proportion
of the price level vis-à-vis nominal wage did not manifest itself as a consequence
of decreased interest rates and hence monetary costs of production. With sub-
dued wage dynamics and a central bank aiming to establish low real interest
rates, such a tendency would be expected to manifest itself via price deflation,
whereas in fact price dynamics remained very much the same as in the 1990s.
As far as the second question is concerned, that is, whether it is possible to
say something about the respective roles of wage bargaining and monetary
policy as independent causes of changes in income distribution, the first
observation is that periods of higher nominal wage dynamic often are also
periods in which real wages rise, as can be seen from Figure 8. This is true
of the early phase up to 1980, except when a significant exogenous inflationary
impulse came from the oil shocks of 1973 and 1979, but also after 1997. On
the other hand, between the mid-1980s and the mid-1990s real wages initially
fell (at a roughly constant rate of about 1% a year) and then stagnated, inde-
pendently of the fluctuations in nominal wage growth rate (Figure 8). It is of
some interest that in the second half of the 1990s the moderate reversal of the
trend in real wage growth took place after a number of years (the long Clinton
boom) that were characterized by sustained GDP growth and a continuous fall
in the unemployment rate (Figure 9). Hence as a result, it would seem, of
conditions directly affecting the labor market and wage bargaining.
With regard to the phase beginning in the early 1980s, it is quite clear that
after 1979 there was a determined change in monetary policy aimed at imple-
menting high nominal and real interest rates, and the data are broadly compat-
ible with a causation going from higher long-term interest rates to changes in
 JOURNAL OF POST KEYNESIAN ECONOMICS 279

Figure 9. Unemployment rate in the United States, 1960–2015.

income distribution. Yet things might be more complex. The increase in

interest rates was preceded by firm-level reactions to the enhanced bargaining
position of workers. In the early 1970s firms were already undertaking restruc-
turing processes featuring delocalization of production and union avoidance
(Bluestone & Harrison, 1988). In fact, in the United States the first signs of
a change in labor relations can be dated from the reversal of the trend of
production workers’ real wages in the early to mid-1970s (Pollin 2003, pp.
42 and ff.; see Figure 3) and a fall in the wage share. In parallel with this,
the unemployment rate, which had been falling steadily between 1960 and
1969, initiated a reversal of the trend. Thus, it could be suggested that Volker’s
policy of high nominal and real interest rates was actually made possible by
indications that a weakening of labor was already under way. In addition,
the change in interest rates and monetary policy after 1979 was accompanied
by a huge surge in the unemployment rate, which was in turn conducive to
further institutional changes affecting unions and the labor market. There
are therefore elements that indicate that labor market conditions and the
institutional set-up directly affecting the bargaining position of workers may
be important in determining not only nominal but also real wages.18

Conclusions
The Classical Keynesian approach appears better equipped than mainstream
theory (in all its variants) to provide a consistent explanation of the dramatic
changes that have taken place in recent decades. This is due to the combi-
nation of criticism of the analytical foundations of decreasing factor demand
curves, the revival of the classical approach to distribution as determined by
norms, institution and power relations, and the consequent classical separ-
ation between the determination of output and employment and distribution.

18
This is also consistent with econometric results concerning OECD countries—see Hein (2015) and Kohler et al.
(2015); for similar conclusions concerning Italy and some other European countries, see Levrero and Stirati
(2006) and Stirati (2010, pp. 132–139).
280 A. STIRATI

On the other hand, these large changes in income distribution add to the
collection of empirical observations that are difficult to account for within
mainstream theory—unsurprisingly, in view of the flawed theoretical
foundations of the approach.
Even some mainstream new Keynesian economists now appear to acknowl-
edge that not only personal, but also functional income distribution essentially
depends on norms and institutions, and that only changes in the latter can
explain, first the absolute and relative improvement in labor earnings until
the mid-1970s, and subsequently, the reversal of that trend (Krugman
2007). Note that this is a novelty since, despite the emphasis on labor market
institutions, in standard New-Keynesian macromodels the equilibrium real
wage is determined residually, given labor (marginal) productivity and the
mark-up. This in turn is determined by the elasticity of product demand
curves in imperfectly competitive markets. Thus, changes in labor market
institutions in these models affect not the equilibrium wage (as a proportion
of output per worker), but the equilibrium unemployment rate.
Within the classical approach, however, different emphases can be found as
to whether the changes in institutions and power relations affect the
distribution of the surplus (over and above the necessary requirements of
the workers) by acting primarily on interest rate determination or wage
bargaining. In this regard, the evidence from the United States seems to sug-
gest that Garegnani’s cautious position may be the most appropriate: In
principle, either variable may be subject to be determined residually according
to circumstances, and monetary policy concerning interest rates may be
influenced and constrained, among other things, by power relations and
institutions in the labor market, so that the latter can be very important in
determining distribution, even in contemporary, fiat money economies. In
the U.S. experience, the circumstances directly affecting wage bargaining
appear to have been important and are often emphasized in narrative accounts
(see, among others, Harrison & Bluestone, 1988; Pollin, 2003; Krugman, 2007).
This seems confirmed by recent work showing a negative relationship between
an indicator of labor market conditions (based on unemployment size and
duration) and the wage share in the United States (Shaikh 2016, chapt 14).
Finally, U.S. data suggest that for more than a decade now interest rates and
the rates of profit on capital have not been moving in step—as expected (in the
long run) if the working of competition tends to equalize the pure remuner-
ation of capital in different uses, with given normal rates of business profits.
In principle this could be explained by exogenous changes in normal business
profits, exogenous in this context meaning independent of economic and
institutional factors (e.g., the weakening of unions) that directly affect wage
bargaining. The origins of such changes in turn would need to be carefully
identified. Alternatively, it should be acknowledged that interest rates and
profitability can actually diverge, at least for some time, owing to independent
 JOURNAL OF POST KEYNESIAN ECONOMICS 281

forces acting at the same time and in the same direction on the interest rate and
real wages, leaving business profits to be determined residually. In this case, the
circumstances that may have prevented or delayed a tendency toward price
deflation as a result of competition, and the consequences of the gap between
profitability and interest rates, should be further investigated.

ORCID
Antonella Stirati http://orcid.org/0000-0002-4195-9416

References
Aspromourgos, A. 2013. “Sraffa’s System in Relation to Some Main Currents in Unorthodox
Economics.” In Sraffa and the Reconstruction of Economic Theory: Volume Three: Sraffas
Legacy: Interpretations and Historical Perspectives, edited by Enrico S. Levrero, Antonella
Palumbo, and Antonella Stirati, 15–33. Basingstoke, UK: Palgrave Macmillan.
Barba, A., and de Vivo, G. 2012. “An ‘unproductive labour’ view of finance.” Cambridge
Journal of Economics 36 (6):1479–96. doi:10.1093/cje/bes048
Bluestone, B., and Harrison, B. 1988. The great U Turn – Corporate restructuring and the
polarizing of America. New York, NY: Basic Books.
Bronfenbrenner, K. 2001. The effects of plant closing or the threat of plant closing on the right
of workers to organize, Report to the US Department of Labour.
Choi, M. 2001. Threat effect of foreign direct investment on labor union wage premium.
Amherst, MA: PERI. Working paper no 27, http://www.peri.umass.edu/fileadmin/pdf/
working_papers/working_papers_1–50/WP27.pdf, accessed 25 January 2013.
Ciccone, R. 1990a. “Accumulation and capacity utilization: Some critical considerations on
Joan Robinson’s theory of distribution.” In Essays on Piero Sraffa – Critical perspectives
on the revival of the classical theory, edited by K. Bharadwaj and B. Schefold, 417–29.
London, UK: Unwin Hyman.
Ciccone, R. 1990b. “A Comment to Pivetti.” In Essays on Piero Sraffa – Critical perspectives on
the revival of the classical theory, edited by K. Bharadwaj and B. Schefold, 454–56. London,
UK: Unwin Hyman.
Epstein, G., and Power, D. 2003. “Rentier incomes and financial crises: An empirical
examination of trends and cycles in some OECD countries.” Canadian Journal of
Development Studies 24 (2):229–48. doi:10.1080/02255189.2003.9668914
Eurostat, Ameco database spring 2011 in Ameco archive http://ec.europa.eu/economy_
finance/db_indicators/ameco/archive_en.htm
Garegnani, P. 1970. “Heterogeneous capital, the production function and the theory of
distribution.” Review of Economic Studies 37 (3):407. doi:10.2307/2296729
Garegnani, P. 1983. “Notes on consumption, investment and effective demand,” Cambridge
Journal of Economics, 1978, 2:335–53, reprinted in J. Eatwell and M. Milgate (eds) Keynes’s
Economics and the Theory of Value and Distribution (London: Duckworth).
Garegnani, P. 1992. “Some notes for an analysis of accumulation.” In Beyond the steady state,
edited by J. Halevi, D. Laibman, and E. Nell. London, UK: Macmillan.
Garegnani, P. 2015/1962. The problem of effective demand in Italian economic development:
On the factors that determine the volume of investment. Review of Political Economy
27 (2):111–33.
282 A. STIRATI

Garegnani, P., and Palumbo, A. 1998. “Accumulation of capital.” In The Elgar companion to
classical economics, edited by H. D. Kurz and N. Salvadori. Cheltenham, UK: Edward Elgar.
Glyn, A. 2009. “Functional distribution and inequality.” In The oxford handbook of economic
inequality, edited by B. Nolan, W. Salverda, and T. M. Smeeding. Oxford, UK: OUP,
http://www.oxfordhandbooks.com/10.1093/oxfordhb/9780199606061.001.0001/oxfordhb-
9780199606061-e-5.
Gordon, R. J. 1997. “The time-varying NAIRU and its implications for economic policy.”
Journal of Economic Perspectives 11 (1):11–32. doi:10.1257/jep.11.1.11
Greenspan, A. 1997. “Testimony of chairman.” The Federal Reserve’s semiannual monetary
policy report, Before the Committee on Banking, Housing, and Urban Affairs, U.S. Senate,
July 22, 1997, http://www.federalreserve.gov/boarddocs/hh/1997/july/testimony.htm
Hein, E. 2011. “Distribution, ‘Financialisation’ and the financial and economic crisis -
Implications for post-crisis economic policies.” Munich Personal RePEc Archive Paper
No. 31180, http://mpra.ub.uni-muenchen.de/31180/.
Hein, E. 2015. Finance-dominated capitalism and re-distribution of income: A Kaleckian
perspective. Cambridge Journal of Economics 39:907–934.
Hein, E., and Schoder, C. 2009. “Interest rates, distribution and capital accumulation – A Post
Kaleckian perspective on the US and Germany.” Working Paper No. 04/2009, Institute for
International Political Economy.
International Monetary Fund. 2007. World economic outlook. Spillovers and cycles in the
global economy, Washington, DC, International Monetary Fund. http://www.imf.org/
external/pubs/ft/weo/2007/01/pdf/text.pdf
Jacobson, M., and Occhino, F. 2012. “Labour’s declining share of income and rising
inequality.” Economic Commentary, Federal Reserve Bank of Cleveland, September.
Kaldor, N. 1956. “Alternative theories of distribution.” Review of Economic Studies 23:83–100.
Kim, Y. 2013. “Household debt, financialization and macroeconomic performance in the
United States, 1951–2009.” Journal of Post-Keynesian Economics 35 (4):675–94.
doi:10.2753/pke0160-3477350408
Kohler, K., Guschanski, A., and Stockhammer, E. 2015. “How does financialization affect
functional income distribution? A theoretical clarification and empirical assessment.”
Economic Discussion Papers 2015-5, University of Kingston, London, UK.
Krugman, P. 2007. The conscience of a liberal. New York, NY: Norton & Company.
Levrero, E. S., and Stirati, A. 2006. “The influence of unemployment, productivity and
institutions on real wage trends: the case of Italy 1970–2000.” In Wages, employment,
distribution and growth – International Perspectives, edited by E. Hein, A. Heise, and
A. Truger. Basingstoke, UK: Palgrave Macmillan.
Palley, T. I. 2007. “Financialization, What it is and Why it matters.” PERI Working paper
series no 153, University of Massachusetts, Amherst.
Panico, C. 1988. Interest and profit in the theory of value and distribution. Basingstoke, UK:
Macmillan Press.
Panico, C., Pinto, A., and Anyul, M. P. 2012. “Income distribution and the size of the financial
sector – A Sraffian analysis.” Cambridge Journal of Economics 36 (6):1455–77. doi:10.1093/
cje/ber022
Pasinetti, L. L. 1962. “Rate of profit and income distribution in relation to the rate of economic
growth.” Review of Economic Studies 29 (4):267–79. doi:10.2307/2296303
Petri, F. 2004. General equilibrium, capital and macroeconomics. Cheltenham, UK: Edward
Elgar.
Piketty, T. 2014. Capital in the 21st century. Cambridge, MA: The Bellknap Press of Harvard,
Harvard University Press.
 JOURNAL OF POST KEYNESIAN ECONOMICS 283

Piketty, T., and Saez, E. 2003. “Income inequality in the United States, 1913-1998.” The
Quarterly Journal of Economics 118 (1):1–39.
Piketty, T., and Saez, E. 2006. “The evolution of top incomes: A historical and international
perspective.” American Economic Review 96 (2):200–05. doi:10.1257/000282806777212116
Pivetti, M. 1991. An essay on money and distribution. London, UK: Macmillan.
Pollin, R. 2003. Contours of descent – U.S. Economic fractures and the landscape of global
austerity. London, UK: Verso.
Ricardo, D. 1821/1951. The principles of political economy. In Works and correspondence,
edited by P. Sraffa, with the collaboration of M. Dobb, vol I. Cambridge, UK: Cambridge
University Press for the Royal Economic Society.
Rowthorn, R. 1999. Unemployment, capital-labour substitution and Economic Growth.
Washington, DC, International Monetary Fund Working Papers, no 99–43.
Shaikh, A. (2016). Capitalism—competition, conflict crises. Oxford, UK: Oxford University
Press.
Smith, A. 1776/1976. An inquiry into the nature and causes of the wealth of nations. In The
Glasgow edition of the works and correspondence of Adam Smith, edited by R. N. Campbell,
A.S. Skinner, and W. B. Todd, 869–70. Oxford, UK: Oxford University Press.
Sraffa, P. 1960. Production of commodities by means of commodities. Prelude to a critique of
economic theory. Cambridge, UK: Cambridge University Press.
Stirati, A. 1992. “Unemployment, institutions and the living standard in the classical theory of
wages.” Contributions to Political Economy 11 (1):41–66. doi:10.1093/oxfordjournals.cpe.
a035608
Stirati, A. 1994. The theory of wages in classical economics. A study of Adam Smith, David
Ricardo, and their contemporaries. Aldershot, UK: Edward Elgar.
Stirati, A. 1999. “Ricardo and the wages fund.” In Value, distribution and capital: Essays
in honour of Pierangelo Garegnani, edited by G. Mongiovi and F. Petri. London, UK:
Routledge.
Stirati, A. 2001. “Inflation, unemployment and hysteresis: An alternative view.” Review of
Political Economy 13 (4):427–51. doi:10.1080/09538250120099944
Stirati, A. 2010. “Changes in functional income distribution in Italy and Europe – Service
sector prices, labour market conditions and institutional change.” In The global economic
crisis: New perspectives on the critique of economic theory and policy, edited by E. Brancaccio
and G. Fontana. London, UK: Routledge.
Stirati, A. 2016. “Real wages in the business cycle and the theory of income distribution: An
unresolved conflict between theory and facts in mainstream macroeconomics.” Cambridge
Journal of Economics 40 (2):639–661.
Stirati, A. 2017. “Wealth, capital and the theory of distribution: Some implications for Piketty’s
analysis.” Review of Political Economy 29 (1):47–63.
Stockhammer, E. 2004. “Financialisation and the slowdown of accumulation.” Cambridge
Journal of Economics 28 (5):719–41. doi:10.1093/cje/beh032
Stockhammer, E. 2009. Determinants of functional income distribution in OECD countries,
IMK at the Hans Boeckler Foundation studies, Macroeconomic Policy Institute, Berlin,
September.
Copyright of Journal of Post Keynesian Economics is the property of Taylor & Francis Ltd
and its content may not be copied or emailed to multiple sites or posted to a listserv without
the copyright holder's express written permission. However, users may print, download, or
email articles for individual use.