Schumpeter's Business Cycle Theory Its Importance For Our Time

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Schumpeter's Business Cycle Theory

Its Importance for our Time


Gunther Tichy, Graz

1 Introduction

Schumpeter is well known as a business-cycle economist. His work, however, has


had limited influence on the theory of the business cycle, as well as on business-cy-
cle policy. One may say that only a few superficial aspects of his work are familiar
to most economists. This paper tries to review Schumpeter's business-cycle theory,
to work out its fundamental ideas as well as its weak points, and to investigate its
usefulness for modern business-cycle theory.
Given Schumpeter's gigantically large work on business cyclesl, it may seem
strange to state that Schumpeter apparently had little interest in explaining what we
today consider business cycles, that is four to five year fluctuations in overall capac-
ity utilization. What he apparently was interested in were the laws of economic de-
velopment, especially the laws of capitalism's development, or to put it in the words
the book Business Cycle starts with: "Analyzing business cycles means neither more
nor less than analyzing the economic process of the capitalist era ... Cycles ~re not
like tonsils, separable things that might be treated by themselves, but are, like the
beat of the heart, of the essence of the organism that displays them. 2" In the course
of economic development innovational impulses and their consequences may in-
duce waves of different frequencies. Why some frequencies show up more often
than others, if and why the causation is different for waves of different frequency,
the role of innovations for short waves, other causes of business cycles, all these
questions so important for today's business-cycle theory, are of secondary impor-
tance to Schumpeter. In fact, business cycle theory was a side aspect to him. His pri-
mary interest was to join static and dynamic theory, to join theoretical and historical
analysis, to join Walras' equilibrium theory and Schmoller's historical description.
The object Schumpeter used for experimentation to demonstrate his theoretical
concerns, were the ten to sixty year waves, called Juglar and Kondratieff by
Schumpeter. The three-to-four-year ones, called Kitchin by him, are the ones prob-
ably most similar to today's business cycles, but they were of secondary interest only
to Schumpeter. He treats them in footnotes without any deeper interest. But even
Schumpeter's explanation of the longer waves is strangely abstract, primarily inter-
ested in logical consistency and it is in principle unsuited to any statistical test. The
second volume of Business Cycles nevertheless attempts the major logical and
methodological jump: It tries to test historical business cycles as to their compatibil-
ity with the first volume's (untestable) theory and historical facts. It is evident that

1 Approximately 2500 pages, of which the book Business Cycles comprises about two fifth's.
2 Schumpeter, Business Cycles, V.

C. Seidl (ed.), Lectures on Schumpeterian Economics


© Springer-Verlag Berlin Heidelberg 1984
78 Gunther Tichy

Schumpeter's intentions were completely different from those of Keynes, his great
competitor.

2 A Reader's Digest Version of Schumpeter's Business-Cycle Theory

Schumpeter's business-cycle theory centers, as it is widely known, around the con-


cept of innovation, " ... a change in some production function which is of first and
not of the second or a still higher order of magnitude3", " ••• the introduction of new
commodities, ... technological change in the production of commodities already in
use, the opening up of new markets or of new sources of supply, Taylorization of
work, improved handling of material, the setting up of new business organizations
such as department stores - in short, any 'doing things differently' in the realm of
economic life ...4". Preconditions for innovations are that the economy has been in
a state of near-equilibrium long enough, so that the pressure on entrepreneurs' prof-
its enforces additional efforts, that relative and absolute prices have been stable
long enough to make it possible to calculate completely new projects5, and that
some pressure has developed, due to the previous lack of innovations. The upswing
starts, when innovations bunch6, for three reasons: Firstly because the same precon-
ditions (as mentioned above) confront all entrepreneurs, secondly because the proj-
ects are connected technically and commercially, and thirdly, because the first step
of the pioneering entrepreneur acts as an example for the followers and makes their
calculation easier7• As Schumpeter starts from equilibrium or quasi-equilibrium,
which implies "normal" utilization of factors of production, the additional factors
of production have to be provided materially by withdrawal from other uses, and fi-
nancially by credit creation. The production of investment goods can rise only if the
production of consumer goods shrinks during the transition period. Shifts of rela-
tive prices and an increase of absolute prices are necessary as well as unavoidable.
The primary wave caused by innovation is followed by a secondary one caused
by induced purchases of inputs and consumer goods 8• It may be strengthened by
reactions of the old entrepreneurs, by speculation, expectations, and reaction of
stocks. Quantitatively the results of the secondary wave may be more important
than those of the primary wave9• When the goods produced with the new equip-
ment come to the market, supply increases due to the increased productivity. Prices
fall and entrepreneurs - according to Schumpeter - repay the previously created
credits at least partially, out of their profits.

3 Schumpeter, Business Cycles, 94.


4 Ibid. 84.
5 Ibid. 131.
6 Schumpeter, Theorie der wirtschaftlichen Entwicklung, 334-342.
7 Schumpeter, Business Cycles, 100f.
8 Ibid. 145.
9 Ibid. 146.
Schumpeter's Business Cycle Theory 79

The upswing ends for four reasons: First, as the innovations are used up gradu-
ally and the respective markets are satiated and second, because entrepreneurs' de-
mand for newly created credit shrinks - as mentioned above - the process of auto-
deflation starts. Third, prosperity ends because during the secondary wave several
enterprises are founded, which are not viable in the long run. Most relevant, how-
ever, appears that fourth, uncertainty increases, because of higher likelihood of
shocks. The innovation induces changes in prices and quantities not only of the
goods directly concerned by it, but sets into motion an adjustment process of all
prices and quantities over the entire economy, directed towards restoring of a new
Walrasian equilibrium, which is, however, not attained in most cases. During the
adjustment process current relative prices cannot act as a basis of calculation of new
products, of the gains or losses of innovations and investment. This increases the
risk and so the entrepreneurs reduce their investment and especially shy away from
innovations.
The downswing - at least in Schumpeter's two-phases cycles - tends to attain a
new Walrasian equilibrium. It may, however, easily overshoot this equilibrium if the
secondary wave was bolstered up as a cause of expectations or speculation, or if the
liquidation of firms, due to the fall of prices to, turns the recession into depression.
Schumpeter distinguishes between recession and depression not so much by their
amplitude - as we tend today - but basically: Recession leads back to an equilibri-
um, which is as a rule an equilibrium different from the one the prosperity had start-
ed from, depression leads away from equilibrium. Recession is the phase of well-
being, as the innovation has increased the supply of goods, and - contrary to pros-
perity - the share of consumer goods is higher. Recession, like prosperity, is a nor-
mal and necessary part of the cyclical process of economic development. Depres-
sion, on the contrary, is not a period of well-being and not a necessary part of
cyclical developmentl1 • "Whether it occurs or not is a question offact and depends
on accidential circumstances, such as the mentality and temper of the business com-
munity and the public, the prevalence of get-rich-quick morals, the way - conscien-
tious or otherwise - in which credit is handled in prosperity, the ability of the public
to form an opinion about the merits of propositions, the degree to which it is given
to believe in phrases about prosperity plateaus and the wonders of monetary man-
agement and so on. Moreover, no theoretical expectation can be formed about the
occurrence and severity of depressions. 12"
The end of depression, the lower turning point, and the beginning of the revival
occurs, when the movement towards equilibrium again begins to dominate. "For
saying that firms will not act in a way which will lead to recovery and eventually to a
neighbourhood of equilibrium, would be synonymous with saying that they will de-
liberately forgo gains and incur losses which it is in their individual power to make
or to avoid, and scrap plant and equipment which could be profitably used. It is
sometimes objected that scramped lower-level equilibria may arise from which peo-
ple will not of themselves move. This may be so in individual cases, particularly in
imperfectly competitive situations. But the probability that this state of things

10 Schumpeter, Business Cycles, 148.


11 Ibid. 150.
12 Schumpeter, Business Cycles, 150 (emphasis in the original).
80 Gunther Tichy

should prevail all over the system, in all industries and with all concerns - for that
would be necessary to justify the inference - is indistinguishable from zero,13"
The tendency towards equilibrium has several causes. The increasing market
share of surviving enterprises when others are forced to liquidate, the fact that shifts
of relative prices imply chances not only for losses but for gains as well, policy ac-
tions "or some favorable chance event14" are among the causes. But Schumpeter
himself had to emphasize "that, theoretically, the system may never conquer the
breathing space in which it could recover by itself'S". It has to be added that
Schumpeter's arguments as to the likelihood of an underemployment equilibrium -
implicitly directed against Keynes - are hardly convincing.
Schumpeter emphasizes several times, that the phase of revival is the last and
not the first one of the cycle, as it leads back to an equilibrium. The danger of over-
shooting is small in this phase16• Back in equilibrium the economy may stay without
waves for an indefinite period of time; sooner or later, however, the likelihood of
new innovations increases for the reasons given above. When innovations occur is,
however, indeterminate, as they have deterministic as well as stochastic character.
For Schumpeter innovations are not only impulses of cyclical fluctuations and
their cause, but their character and their period of implementation determine the
length of the cycles as well. The longer the period of implementation of innovations,
the longer the wave. Adaptation in the structure of the entire economy due to large
scale innovations such as railways cannot be made all at once. Before they are com-
pletely implemented, the structure of relative prices becomes heavily distorted and
entrepreneurs find it increasingly difficult to calculate further investment projects.
Hence they refrain from innovations, reduce investment, and thereby bring about
recession. In addition to this, the period of implementation of innovations and with
it the length of the cycles is determined by the structure of the economy which reacts
on innovations and the financial conditions, as well as the habits prevailing in that
economyl? Neither theory nor logic can give reasons for a uniform length of cy-
cles18• "As a convenient descriptive device19", however, Schumpeter designs his
well-known three-cycle scheme, forty-month Kitchins, ten-year Juglars, and sixty-
year Kondratieffs, which implies the nice numbers-game: Three Kitchins per Jug-
lar, six Juglars and therefore eighteen Kitchins per Kondratieff. As already men-
tioned, Schumpeter selected this scheme as a convenient devise for description. He
emphasized that a five-cycle scheme would be equally plausible2°.
The integral number of shorter cycles within the longer, however, arises not out
of chance but is theoretically necessary in Schumpeter's framework. At the end of
each cycle, that is the end of recession in the two-phase cycles, and the end of reviv-
al in the four-phase ones, stands an equilibrium or quasi-equilibrium. One can
hardly conceive a Kondratieff equilibrium and a Juglar and/or Kitchin disequilib-

13 Schumpeter, Business Cycles, 152 f.


14 Ibid. 154.
15 Ibid. 154.
16 Ibid. 156.
17 Ibid. 143.
18 Ibid. 167.
19 Ibid. 170.
20 Ibid. 169.
Schumpeter's Business Cycle Theory 81

rium at the same time. What should the system of market-clearing prices look like
under such conditions?
It is interesting to note, that the ftrst 160 odd pages of the book - before the in-
troduction of this "empirical" three-cycle scheme - deal with "the" cycle only, ex-
plain "the" cycle; this necessarily implies that the forty-month, ten-year and sixty-
year cycles have the same causation. This monocausality contains several
unplausibilities. For the Kondratieff it is difficult to understand the constant fre-
quency. Why do large-scale innovations bunch just every sixty years? For the Kitch-
in it is not easy to understand the role of innovations for its causation. It would
seem easier to explain it as a secondary-wave phenomenon, as the consequence of
disturbing the system of relative prices and the resulting uncertainty of investment.
This, however, would be a Keynesian rather than a Schumpeterian explanation.
Schumpeter, however, does not exclude that the short waves may be of the adaptive
type21 or caused by reactions of consumer expenditure22•
Incidentally it may be mentioned that besides the innovation cycles of varying
length, Schumpeter considers other types of cycles, to which he attaches less impor-
tance, however: Waves produced by external factors, wars, gold production or
crops23, waves of adaption, waves produced by random effects on self-reinforce-
ment24, hesitations and vibrations, fluctuations in the aggregates 25 , and replacement
waves26•
The great theoretical achievement of Schumpeter is to have realized that busi-
ness fluctuations are not a phenomenon of their own, but constitute an integrating
part of the development of the (capitalist) economy. Growth does not simply mean
linear enlargement. It implies development, structural change, new dimensions. It
has to materialize in jumps, as the innovations pursuing the pioneer one and the
hardly avoidable over-reactions of the secondary wave enforce adjustment reac-
tions of all prices and all quantities. It includes the variability of the price system, a
necessary consequence of the upswing following innovations. It necessarily in-
creases the uncertainty about prices to be achieved and quantities to be sold and in-
creases the uncertainty in calculating new innovations and new investment. Innova-
tion and investment are therefore reduced. The system then ftnds its way back to an
equilibrium with no growth and no proftts. When the innovators have realized that
the price system is stable again, they can calculate new innovation projects and they
will sooner or later realize them.
Cyclical fluctuations therefore have a function in Schumpeter's model of devel-
opment, and they can be prevented only at the cost of a renunciation of growth27 •
Only the overheating of the secondary wave could be reduced without negative con-
sequences for development, especially the increase in the price level. The likelihood

21 Schumpeter, Business Cycles. 171, 174, 183.


22 Ibid. 531.
23 Ibid. 175.
24 Ibid. 179.
25 Ibid. 188.
26 Ibid. 189.
27 Schumpeter, "The Explanation of the Business Cycle", 308. A successful "Keynesian policy" is
possible for Schumpeter if, and only if the task is to prolong the state of a "moving equilibrium";
this, however, is not the normal task. A management of global demand is restricted to the case in
which supply can easily adjust to demand. Neither when innovations lack will this be the case nor
when they are under way. Modern supply-side economics followed Schumpeter in this point.
82 Gunther Tichy

of a success of such a policy is very small to Schumpeter: "Dealing with the former
(i. e. speculative banking causing overheating of expansions) would, in order to be
effective, require a policing power, which to this day has always been quite beyond
central banks. This inability of capitalism to police itself is a striking as its inability
to protect itself - it always requires both a policeman and a protector of nonbour-
geois complexion, who regulate, shield, and exploit it. This is as true of the times of
Queen Elisabeth as it is today. But it is largely this inability that produces crises as
distinguished from mere depressions.28"
The process of development, according to Schumpeter, does not need the purg-
ing power of crises29, but it does need the purging power of depressions 3o• Once the
economy has come into a crisis, there is little that can be done to subdue it. The re-
vival is sound only if it grows out of itself31 ana this to Schumpeter is in no way
problematical, as it normally happened in the past within four to five years 32• We
have noticed, however, in the relevant chapters of Business Cycles, that Schumpe-
ter's arguments for the necessity of a revival are not very convincing. But one should
not overemphasize this last point, as Schumpeter never discussed questions of busi-
ness-cycle policy in detail: The memories of all his students agree that he refused to
give policy advice in his later years33 •

3 The Problematical Aspects of Schumpeter's Business-Cycle Theory

The merits of Schumpeter's business-cycle theory are identical with its deficiencies.
They both lie at the base of his research program. Schumpeter's ultimate goal was to
amalgamate statics and dynamics, to give an all-inclusive explanation of the eco-
nomic development of capitalism, from short-term fluctuations over medium- and
long-term thrusts of growth to its decline. Schumpeter tried not only to combine
equilibrium theory, development theory and vision, but to use the sum of all three
directly to explain economic development year after year. Three main areas of criti-
cal comments arise: The character and the central position of the concept of equi-
librium, the dominating rOle of innovations, the monocausality, and the concept
using the same explanation for cycles of widely varying lengths. Not one of these
central points of Schumpeter's business-cycle theory is cogently founded in logical
or theoretical reasoning, and none is proved empirically.
Schumpeter's amalgamation ofbusiness-cycle theory and equilibrium theory is no
more than an interesting idea if taken literally. That every cycle swings out into a
state of equilibrium is neither theoretically nor empirically cogent. In addition the
meaning of the concept of equilibrium is not at all clear since Schumpeter gives no
explicit theory of employment. The level of employment consistent with equilibri-

28 Schumpeter, Business Cycles, 660.


29 Schumpeter, "The Historical Approach to the Analysis of Business Cycles", 309; Schumpeter,
Theorie der wirtschaftlichen Entwicklung, 366 f.
30 Schumpeter, Theorie der wirtschaftlichen Entwicklung, 354-369.
31 Schumpeter, "Depressions", 117.
32 Schumpeter, "Depressions", especially 117.
33 This was, however, not the case in his younger years; cf. Seidl, this volume, pp. 201-205.
Schumpeter's Business Cycle Theory 83

um is unclear. Apparently Schumpeter assumed some sort of natural rate of unem-


ployment. As any of the three types of cycles ends in an equilibrium, and - accord-
ing to Schumpeter - the curves of the longer waves represent the equilibria for the
shorter ones34 a Kitchin equilibrium during a Juglar upswing must imply another
level of employment and other price relations than the same equilibrium in a Juglar
downswing. But as, in the same way, the Juglar equilibria differ from one another
whether they occur in a Kondratieff up- or downswing, 18 different equilibria nec-
essarily exist, according to the phase of the two longer cycles during the Kitchin
equilibrium.
The position of rest in the upswing between revival and prosperity seems not
very plausible from the empirical point of view of the stylized facts of business cy-
cles. As innovations are at least partly stochastic in character, this period of equilib-
rium necessarily is of indeterminate length. This, however, is not consistent with the
fact that, once the depression is overcome, the upswing normally proceeds, or - if
weak - turns around in a new downswing after inventories are filled up, but rarely
does it put in a rest before continuing.
The second topic open to critique is Schumpeter's emphasis of the central role of
innovations. In today's terminology, one could say that Schumpeter neither had an
explicit theory of invention, nor an explicit theory of innovation, nor an explicit the-
ory of diffusion of technical progress and of innovation. Even rough ideas about the
time pattern of these processes are lacking in his writings. Even in our days we still
do not have these theories, but this does not gloss over the problem that Schumpe-
ter's remarks on innovation were too poor a basis for the gigantic structure of theory
built on it. Some of Schumpeter's intuitive ideas about innovations are plausible
and consistent with our mainstream knowledge. We may, e. g., mention the hypothe-
sis that there always exists a stock of inventions as a basis for innovations35, or the
hypothesis that firms concentrate on investment for enlarging capacities in phases
of high demand, while they prefer rationalization investment and innovation in
phases of slack demand (equilibrium ?). Some of Schumpeter's statements are im-
plausible or do not seem to conform to our evidence, such as the hypothesis that in-
novations concentrate in large firms 36, or the assumptions about the speed of diffu-
sion: Empirical investigations seem to imply that the diffusion of innovation is
much to slow for an explanation of Kitchin waves and much to variable for con-
stant-length Kondratieffs37• It could probably explain Juglars, but we have no evi-
dence for ten-year waves in our time, apart from the weak observation that every
second of our four- to five-year fluctuations turns out a bit stronger38• Schumpeter
does not supply any evidence. He enumerates several years in which new technolo-
gies or innovations appeared, but he could not adduce any counter-evidence that
there were no innovations in the other years. For an explanation of the Kondratieff
waves a very explicit theory of the diffusion process would be required: Why it lasts
- roughly or exactly - 60 years until the economy returns to a new Walrasian equi-
librium after large innovations.

34 Schumpeter, Business Cycles, 173.


35 cr., e.g., Schmookler(1966), or Jewkes, Sawers and Stillerman (1969).
36 For a counterevidence cr., e. g., Mansfield (1977), or the survey ofliterature in Aiginger and Tichy
(1982) 86-92.
37 cr., e.g., Ray (1969) or Hay and Morris (1979) 465-471.
38 Aiginger(1974).
84 Gunther Tichy

So it is a major point of criticism that Schumpeter tried to explain business cy-


cles of such enormously varying length as from 40 months to 60 years with exactly
the same theory. In the theoretical part of his work he deals with "the cycle" only. In
a few instances, however, he does indicate that the Kitchin may be explained by
lagged adjustment or by phenomena of the secondary wave39 • But there is no at-
tempt whatsoever to differentiate between the theoretical explanations of Juglar
and Kondratieff.
Special problems arise with Schumpeter's explanation of the turning points.
That the evidence for the necessity of a revival - and so for the explanation of the
trough - is weak, has been mentioned already, and this weakness was more or less
accepted by Schumpeter. But what about the upper turning point? That the innova-
tion may be exhausted after a while is plausible. But why don't the innovative entre-
preneurs look for other ones? Why do they repay bank loans and so start the auto-
deflation? Schumpeter's explanation for this is the increasing variability of all
prices during the process of adaption towards a Walrasian, or - according to our
critique - a temporary equilibrium. This is not completely convincing for two rea-
sons. First, the entrepreneur needs only a few prices and the observation of a few
markets for his calculation. He needs to know how these few move relative to each
other, which, though being more complicated in a boom, is not intractable. Other
uncertain factors, such as the satiation of markets, decreasing profits, limits of ca-
pacities, or the concentration on capacity-widening investment, may be more detri-
mental to innovations than the variability of prices. Second, one could imagine a
completely different explanation of the innovation: For Schumpeter, innovations
grow out of states of equilibrium, the only phase of the cycle that permits them. The
theory of the product cycle combined with portfolio theory could imply another hy-
pothesis, that is that the entrepreneurs innovate when they consider the risk of inno-
vation not higher than the risks in continuing to produce products in the last phase
of their product cycle. For the firm as a whole even a risky innovation need not in-
crease the total risk, as the dispersion of risks may decrease or may become less
skew. According to this hypothesis innovation should bunch in periods of market
deterioration and of structural problems. The evidence for this hypothesis is no
worse than for Schumpeter's one of the bunching of innovations in periods of equi-
librium.
It should be emphasized that this critique does accept Schumpeter's model of
business cycles in chapters I to IV of volume I of his Business Cycles. This model
correctly abstracts from all factors, less important for the explanation it provides.
What has to be criticized is the mono causal theoretical explanation and the attempt
to verify it empirically confronting the turning points derived arithmetically from
the three-wave scheme with the turning points derived in a rather crude way from
historical data and facts 4o•

39 Schumpeter, Business Cycles. 171 IT., 183,531; Schumpeter , "The Historical Approach to the
Analysis of Business Cycles", 315.
40 Cf. Kuznets (1940) 266: "One cannot well escape the impression, that Professor Schumpeter's
theoretical model in its present state cannot be linked directly and clearly with statistically ob-
served realities; that the extreme paucity of statistical analysis in the treatise is an inevitable result
of the type of theoretical model adopted; and the great reliance on historical outlines and qualita-
tive discussion is a consequence of the difficulty of devising statistical procedures that would
correspond to the theoretical model."
Schumpeter's Business Cycle Theory 85

To repeat what had already been mentioned before: The deficiency of


Schumpeter's work on business cycles is deeply rooted in Schumpeter's intentions.
It has no justification of its own, but was designed and therefore exists only as a
part, as a mechanism of transmission of Schumpeter's theory of economic develop-
ment. Beauty and logical consistency enforce monocausality. Had Schumpeter per-
mitted a certain life of its own to his business-cycle theory, had he integrated ele-
ments of contemporary explanations of the business cycle, the result might have
been less original and more eclectic, but it certainly would have been of greater in-
fluence.
Grandseigneur Schumpeter, however, esteemed intuition, subjectivism and orig-
inality, as his very last book proves41 • He appreciated logical consistency, elegance,
and daring interpretations. He pressed thoughts as well as facts into a beautifullogi-
cal framework, intuitively found before: "The line of reasoning of this gigantic
building of ideas appears sometimes too artificial to me42", SpiethofT remarked.
Schumpeter's attempts to verify this gigantic building of ideas in historical reality
proved unavailing, as he stuck too much to the methods of the Historical School.
Collection and processing of statistical, as opposed to historical, data facts was ap-
parently too tedious to him and he also ignored contemporary research. The main
cause of the limited influence of Schumpeter's work on business cycles apparently
was that he disappointed all his friends and partisans at the same time: The statisti-
cians, to whom the subtitle of Business Cycles promised a statistical analysis, were
disappointed by the lack of any statistical methodology and the determination of
most turning points by "graphic portrayal43 " of the series44• The econometricians
were disappointed that a founding member and then (1939) president in office of
the Econometric Society ignored the methods developed by Ragnar Frisch and Jan
Tinbergen in good time to be used by Schumpeter. The mathematicians found no
vestige in the purely verbal description of the man, who said, that every day without
mathematics (and Greek) was a lost day for him45 • The theoreticians looked in vain
for a discussion of contemporary business cycle theory. They were astonished - to
say the least - that Schumpeter explicitly left the "systematic comparison of my an-
alytic scheme with others (in particular those of Mr. Keynes, Professor Haberler,
and Mr. Harrod) to the reader46", especially as Schumpeter did not refrain from at-
tacks in the footnotes. The historians lastly found no serious historical analysis, but
instead the attempt, to press historical facts into the rigid pattern of the formal
three-cycles scheme. Contrary to what Schumpeter said several times, he did attach
highest importance to counting the number of years between turning points, the sec-
ond volume concentrates almost exclusively on this topic.

41 Schumpeter, History ofEconomic Analysis; cf. also Mann (1958) 103 f.


42 Spiethoff (1949150) 292: "Die Gedankenfiihrung dieses riesigen Ideenbaues dunkt mich oft zu
gekiinstelt."
43 Kuznets (1940) 265.
44 As far as Schumpeter used formal statistical methods at all, they differ from case to case with no
reason discernible: For marking turning points once the method of "normal points" is used
(Schumpeter, Business Cycles, 469), once the method of trend deviations (Schumpeter, Business
Cycles, 554), twice the method of moving averages (Schumpeter, Business Cycles, 468, 474), in the
other cases "graphical portrayal".
45 Cf. among others Schneider (1970) 20f.
46 Schumpeter, Business Cycles, VII.
86 Gunther TIchy

These critical points were clearly recognized by contemporary economists. Kuz-


nets summarized: "The critical evaluation above of what appear to be important
elements in Professor Schumpeter's conclusion, viewed as a systematic and tested
exposition of business cycles, yields disturbingly destructive results. The association
between the distribution of entrepreneurial ability and the cyclical character of
economic activity needs further proof. The theoretical model of the four-phase cy-
cle about the equilibrium level does not yield a serviceable statistical approach. The
three-cycle scheme and the rather rigid relationship claimed to have been estab-
lished among the three groups of cycles cannot be considered, on the basis of the
evidence submitted, even tolerably valid; nor could such validity be established
without a serviceable statistical procedure. The core of the difficulty seems to lie in
the failure to forge the necessary links between the primary factors and concepts
(entrepreneur, innovation, equilibrium line) and the observable cyclical fluctuations
in economic activity.47"

4 Schumpeter's Merits for the Business Cycle Theory of the Future

Schumpeter's attempts to build an integrated theory failed. Hence it is not his re-
sults that we can learn from, but rather his attempts and ideas. And we need to learn
since we still have not yet an integrated theory of cyclical development, nor even a
satisfying theory of the business cycle. The standard paradigm of cumulative de-
mand fluctuations lacks consistency as well as microfoundation and is not able to
take account of expectations, growth, and structural problems. The alternative con-
cepts of New Classical and New Keynesian Macroeconomics are no more than first
sketches of ideas and cannot, up to now, give convincing answers to policy ques-
tions. The attempts to construct a consistent and relevant theory of the business cy-
cle can learn from several aspects of Schumpeter's theory. They have done this al-
ready in the past. Some of the most important of these aspects are:
(1) the ase of a plurality of methods;
(2) elements of a theory of expectations, at least a critique of oversimplifying as-
sumptions of destabilizing expectations48 ;
(3) the explicit consideration of the non-homogeneity of aggregates, the resulting
caution in aggregation and the consideration of structural aspects;
(4) the attempts to endogenize spurts and ruptures in development;
(5) the consideration of spill-overs from one disturbed market to others;
(6) the emphasis put on the variability of relative prices, their relevance for the
working of the spot markets and the generation of expectations, their influence
on uncertainty and investment decision if futures markets are absent;
(7) some elements of a theory of innovation, where innovation correctly comprises
not only technical but also organizational and marketing aspects;
(8) the accentuation of the trade-off between static and dynamic efficiency;

47 Kuznets (1940) 270.


48 Schumpeter, Business Cycles. 44.
Schumpeter's Business Cycle Theory 87

(9) the differentiation between the central position of the entrepreneur as an inno-
vator and the function of the capitalist;
(10) the emphasis put on the destabilizing factors of supply, the basic elements of a
theory of the product cycle as well as of a theory of bounded rationalitf9;
(11) the implicit solution of the persistence problem of modern equilibrium busi-
ness-cycle theory, as the shocks of innovations are necessarily auto-correlated
as pilot innovations are followed by imitators;
(12) the distinct differentiation of superficial cyclical phenomena and underlying
causes5o, which could drastically improve business-cycle policy, which in Bas-
tard-Keynesian tradition today concentrates on fighting against the conse-
quences of recession instead of against its causes, a danger which Schumpeter
recognized and which he attacked51.
It would be no problem to extend this list, but it makes more sense to have a
closer look at the very point that saw Schumpeter most ahead of his time and which
has not found adequate attention up to now: Schumpeter's concept of a relative
equilibrium and the resulting elements of a theory of innovations, a theory, however,
which cannot explain the three-cycle-scheme. Schumpeter evidently believed that
he had integrated Walrasian equilibrium theory. Of course, he didn't achieve this
ambitious goal, but instead created something completely new. Formally Schumpe-
ter started from Walras's equilibrium of clearing markets and transactions at mar-
ket-clearing prices only. This equilibrium can be found, however, in Schumpeter's
world once every sixty years only, at the end of the Kondratieff. In the meantime
there are "neighborhoods of equilibria52", every ten years at the end of the Juglar,
and more distant neighborhoods every forty months at the end of each Kitchin. In a
book which explicitly refers to Walras again and again a concept such as a "neigh-
borhood of equilibrium" necessarily remains unclear, as the price system can either
clear the market or not.
Schumpeter's concept of equilibrium is unclear in yet another reaspect: Nor-
mally a state of equilibrium is characterized by the fact that it does not breed any
tendency for change out of itself. Only exogenous shocks can disturb the equilibri-
um. But the innovations which disturb the equilibrium in Schumpeter theory are not
exogenous - jf they were they could not explain the periodicity of the cycles - but
result necessarily from the characteristics of the equilibrium, i. e. constant relative
prices and zero profits. So Schumpeter's equilibrium is an equilibrium only as long
as supply and demand functions are unchanged. Schumpeter's theory implies, how-
ever, that in such a situation the supply function has to change sooner or later, when
the pilot innovation occurs. The equilibrium therefore harbours in itself the factors
for its own overthrow. Combined with the fact that most "equilibria" are only
"neighborhoods", Schumpeter's concept of equilibrium does not make much sence
if understood in Walras' way; Schumpeter himself spoke of the "chronic disequilib-
rium of reality53".

49 Schumpeter, Theorie der wirtschaftlichen Entwick/ung, 356.


50 Schumpeter, "The Historical Approach to the Analysis of Business Cycles", 312.
51 Schumpeter, Business Cycles, 1044.
52 Schumpeter, Business Cycles, 71, 173.
53 Schumpeter, Das Wesen des Geldes, 269: "chronisches Ungleichgewicht der Wirklichkeit".
88 Gunther TIchy

As Schumpeter surely knew, everything he was interested in - development, cy-


cles, Credit, interest rate, profits, capital, innovations - does not exist in a Walrasian
world. The concept of Walrasian equilibrium could have had only the character of a
reference model to him. On the basis of this reference model Schumpeter developed
something completely new, a theory of the automatic disturbance ofany equilibrium:
In their search for profit innovating entrepreneurs disturb the economic equilibri-
um, which is statically Pareto-optimal but dynamically suboptimal for them. By
their innovation entrepreneurs create a disequilibrium which leads to uncertainty
concerning quantities, prices, and profits. It therefore impedes further innovations
and thereby creates further disturbances. Eventually, the system swings back to a
new equilibrium but Schumpeter did not work out explicitly the preconditions and
the details of this process. In fact, most transactions must happen in disequilibrium.
Equilibria are rather short periods of stable relative prices that create innovations
and so quickly disturb the basis for the calculation of further innovations and so the
conditions which made them possible. Schumpeter's term "creative destruction"
gets a new meaning if attached to his attitude to equilibrium.
Interpreted in this way, Schumpeter's concept of equilibrium is an expression of
the degree to which the adjustment of relative prices has come to a rest. Instead of
the dichotomy: equilibrium - disequilibrium, Schumpeter developed a hierarchy of
speeds of adjustment, depending on the degree of the disturbance. Equilibrium the-
ory becomes an instrument for explanation, not the explanation itself54. The change
of the structure is an integral part of the adjustment process. Relative prices may
change but slowly in such situations, as agents cannot easily distinguish systematic
and stochastic components. So trading at false prices may prevail for a while. All
this has become an integral part of modern microeconomics in the meanwhile.
The concept of the unavoidable endogenous disturbance of equilibrium and the
permanent tendency to return to this state is the great, and lasting achievment of
Schumpeterian theory. Schumpeter's business cycle theory therefore does not need
the exogenous impulses, which are the most important ingredients of its contempo-
rary and later competitors; these exogenous impulses can, however, be easily inte-
grated in Schumpeter's theory if necessary.
If one takes Schumpeter's ideas seriously, the necessity of stable relative prices
and of the calculability of investment as preconditions for innovation and growth,
then Austrian economic policy of the last thirty years has indeed acted in the spirit
of Schumpeter, even if it did this by referring to Keynes. Austrian economic policy
tried to stimulate investment by reducing entrepreneur's uncertainty as far as possi-
ble, by making all the data, relevant for investment decisions, forseeable and easier
to calculate, last but not least by means of a steady policy55. In addition, Austrian
economic policy never forgot that policy has to fight against the causes of distur-
bances and not against the disturbances themselves. I am therefore happy to be able
to draw the conclusion, that Austrian economic policy could find a synthesis be-
tween the two great opponents of the thirties, a synthesis between Schumpeter and
Keynes, a political synthesis, while the theoretical synthesis is still far ahead.

54 Schumpeter, Business Cycles, 30ff.; concerning modern business-cycle theory cr. Lucas (1981).
55 TIchy (1982), (1983 a), (1983 b).

You might also like