<<

. 15
( 25 .)



>>

any endogenous variables, then the (modified) long-run equilibrium has not cism in economic models, see Boland [1982a, Chapter 7; 1989, Chapters 1, 7
been reached. and 8].
7 For more on the methodological question of explaining dynamics, see Boland
[1982a, Chapter 6] and for a discussion of the technical requirements of
explanation as distinguished from description, see Boland [1989, Chapter 6].
8 I say ˜methodological aspects™ to distinguish them from empirical aspects, such
as the truth of the assumptions about the relative variability of the givens used
to distinguish the short run from the long run.
9 The equilibrium price system is one instance of such a social institution; other
institutions include the laws governing trade and advertising practices and tax
laws. The extent to which the social knowledge provided (such as norms,
guidelines and legal limits) is necessary is directly related to the power of the
institution.
10 Such a situation was recognized by Plato in his dialogue ˜Laches™. It is
observed at the beginning that ˜some laugh at the very notion of consulting
others, and when they are asked will not say what they think. They guess at the
wishes of the person who asks them, and answer according to his, and not
according to their own, opinion.™
11 Clearly, it does not attempt to be relevant for the explanation of the observed
behaviour of a hermit or anyone else who opts out of a society (although it
would apply to a group that opts out). In other words, it does not attempt to
apply to an asocial situation.
12 In correspondence, Ludwig Lachmann noted to me that he offered a similar
theory of social institutions in his 1970 book. His illustration of the differences
between consensus and concrete institutions is the difference between ˜the
market™ and the stock exchange.
13 Of course, not all solutions are invented or designed “ some may be
˜discovered™.
14 Let me define these two different views of the explanatory relationship between
institutions and individuals. Psychologistic individualism is the methodological
requirement that says all explanations of institutions must recognize that only
individuals can make decisions and that the only exogenous variables allowed
are nature-given, including the psychological states of the decision-makers.
Institutional holism would allow other exogenous variables such as the ˜destiny
of the nation™, class interest, etc. In the extreme, institutional holism would
deny a role for the individual in determining the social outcomes [e.g. Sraffa
1960].
It is commonly thought that if an explanation is not psychologistic-individu-
alist then it is ˜holist™ (or ˜collectivist™). This is a mistake. The distinctions to be
drawn are between individualism and holism and between psychologism and
institutionalism. This means that there are four distinct views. Economists since
© LAWRENCE A. BOLAND

Part III

Some missing elements
© LAWRENCE A. BOLAND

9 The foundations of Keynes™
methodology




By ˜uncertain™ knowledge ... I do not mean merely to distinguish what
is known for certain from what is only probable... Even the weather is
only moderately uncertain. The sense in which I am using the term is
that in which the prospect of a European war is uncertain, or the price
of copper ... twenty years hence... About these matters there is no
scientific basis on which to form any calculable probability whatever.
We simply do not know...
I accuse the classical economic theory of being itself one of these
pretty, polite techniques which tries to deal with the present by
abstracting from the fact that we know very little about the future.
John Maynard Keynes [1937, pp. 214“15]

Liquidity is freedom. When a firm takes action that diminishes its
liquidity, it diminishes its freedom; for it exposes itself to the risk that
it will have diminished, or retarded, its ability to respond to future
opportunities. This applies both within the financial sphere and
outside. I have myself become convinced that it is outside the financial
sphere (very inadequately considered, in relation to liquidity, by
Keynes) that liquidity is potentially of the greater importance...
Liquidity preference, for the financial firm, is a matter of marginal
adjustments, as Keynes very rightly saw. But the liquidity problem of
the non-financial firm is not, as a rule, a matter of marginal
adjustments.
John Hicks [1979, pp. 94“5]

Generality pursued too avidly leads to emptiness. As scientists we
must be willing to live dangerously. What we must seek is no
inadmissible specialisations and no unnecessary generality.
Paul Samuelson [1950, p. 374]


Keynes said that the readers of his book would have to endure a ˜struggle
of escape™ if his critical assault upon them was to be successful. This
chapter is about his ˜assault™ strategy, its comportment relative to common
views of what Keynes was trying to do, and its logical possibilities of
© LAWRENCE A. BOLAND
132 Principles of economics The foundations of Keynes™ methodology 133
success. Since Keynes was arguing against the then predominant variables to change. When we are discussing Keynes™ assault it is important
Marshallian neoclassical method of economic analysis, we will have to also for us to keep the Marshallian logical continuum in mind since it is directly
give some more time to considering the essentials of Marshall™s methods in relevant to the significance of the ˜general vs special case™ debate and it is
order to determine where Keynes might have thought he was placing the indirectly but more fundamentally relevant to the intellectual background
most telling blows.1 against which Keynes was directing his assault.
Since a longer time period is being considered whenever one adopts a
methodological perspective further to the right on Marshall™s continuum,
GENERAL VS SPECIAL CASES
more and more variables can be made endogenous instead of exogenous “
The claimed thrust of Keynes™ assault was to show that ˜classical™ that is, more variables can be considered to have been chosen by
economic theory was merely one special case on a more general continuum maximizing individuals whenever there has been enough time allowed to
of possible cases. Unfortunately, this way of presenting his assault can be make any needed adjustments or ˜substitutions™ (to use Marshall™s term).
very misleading. Whenever we are dealing with formal models we are
Y
always dealing with arbitrary frameworks defined in terms of specified sets
of variables. What may be a special case in one framework of given
exogenous and endogenous variables can often be seen as the general case S
in another merely by rearranging the allocation of those variables between
Whigh
being considered endogenous or exogenous. I think the arguments of
˜Keynesian™ economists such as Patinkin [1956] demonstrate this. As long
as the only variables allowed are natural givens and the aims of individuals
We
(i.e. no social variables are allowed if they are not reducible to the logical
Wlow
consequences of individual choices), their interpretation of Keynes™
D
˜general vs special™ case argument will always see Keynes™ assault as a
failure.
L high L low L e
For Keynes, generality refers to a methodological-cum-historical X
continuum. On this continuum any current state of equilibrium is a special
Figure 9.1 Observable levels of employment
case, as it is merely one point on a historical-time continuum. Similarly,
any realistic state of disequilibrium is also just a specific point on that
If we leave aside the long-run temporal aspects of Marshall™s
continuum. A state of disequilibrium is more general in the sense that there
continuum, and instead maintain a market-run perspective, then we can
are many more possible states of disequilibrium than there are possible
appreciate a different continuum. Specifically, the typical labour market
states of equilibrium.
can be seen to form a continuum of prices (see Figure 9.1). At any point in
In the other camp, which includes followers of Marshall and the so-
time a wage-rate and a level of employment will be observed. Observable
called Keynesian Counter-revolutionaries [see Clower 1965], generality is
points (i.e. points representing levels of actual employment at the going
seen differently because they are referring to a different continuum. It is
wage rate) will be located on the demand curve whenever the wage rate is
different because Marshall™s method of explanation uses a logical
above the equilibrium rate and they will be on the supply curve when it is
continuum of time periods which runs from a zero point at the left end
below that rate. Along the continuum of observable levels of employment,
representing an infinitely small instant to a point at the right end
the maximum observable level of employment (without exploitation) will
representing an infinitely long period of time. In between the extremes are
be that one point where demand equals supply. Thus, there is then a
his various temporal perspectives “ ˜market periods™, ˜short periods™, ˜long
continuum running from high wage rates to low rates with just one rate
periods™ and the inter-generational ˜secular™ periods. For Marshallian
being the equilibrium rate.
advocates of neoclassical economics, whenever one is considering points
further to the right one is automatically considering periods of time which
allow more and more variability “ that is, which allow for more time for all
© LAWRENCE A. BOLAND
134 Principles of economics The foundations of Keynes™ methodology 135
would have us do.2
GENERALITY FROM KEYNES™ VIEWPOINT
Neoclassical economics can accommodate psychologistic individualism
Keynes™ argument was more than a petty dispute over historical vs logical
only in long-run explanations. In the neoclassical short run, according to
time-continuum viewpoints. He argued that there are important non-
psychologistic individualism, all non-natural variables may be considered
individualist, non-natural givens facing the real-time individual decision-
˜exogenous™ only temporarily as an arbitrary matter of methodological
maker. A main thrust of Keynes™ argument is that these short-run ˜macro™
perspective. For example, in a short-run model one will see many variables
variables are necessary for adequate explanations even in the usual
that cannot be changed in the short run (e.g. available capital, technical
neoclassical micro model. In particular, there are ˜aggregate™ variables such
knowledge, the income distribution, the interest rate, the market structure,
as GNP, the general price level and expectations which do not depend on
etc.) and that are thus exogenous constraints for the individual decision-
any specific individual™s psychological state but on the behaviour and
maker. Such a short-run perspective can never be an adequate neoclassical
expectations of all other individuals. At any point of time these are
explanation since neoclassical methodology requires that all such
contemporaneously determined variables which the individual cannot
temporary, non-individualist variables be transformed into endogenous
choose, yet they are variables whose states affect the decisions made.
variables by simply broadening one™s logical-time horizons. As a
Keynes™ concept of generality seems to rest, then, on the methodological
consequence, the only acceptable neoclassical explanation will be a long-
position that considers a model with more exogenous givens to be more
run model in which it is logically possible to reduce all endogenous
general. Any methodological strategy that restricts the list of permitted ex-
variables to matters of individual choice guided by psychologically given
ogenous variables would be considered a ˜special case™ in Keynes™ classical
aims [e.g. Lucas 1980].
framework. This is contrary to the usual neoclassical perspective which
In any Marshallian long-run model everything will be in equilibrium
measures generality by the number of endogenous variables explained.
because there will not be any non-natural constraints artificially preventing
Whenever enough time is allowed in any neoclassical model, all
the individual from adjusting his or her situation to its optimum. Often any
variables, including ˜aggregate™ variables, can be shown to be the ultimate
short-run constraints that are neither non-natural nor non-individualist will
result of individual choice. But it is also important to realize that in
be explained away as being the results of past (optimizing) choices. In
Keynes™ argument no amount of realistic time would ever be sufficient to
neoclassical methodology, disequilibria caused by intervening constraints
explain ˜aggregate™ variables away as the neoclassical methodologists
are either temporary states of affairs or they are illusions [see further,
would have us do. So it is important to keep Keynes™ arguments restricted
Archibald and Lipsey 1958]. In any neoclassical model, a disequilibrium is
to the Marshallian ˜short run™ since the definition of that time period
temporary merely because enough time has not been allowed to pass for the
requires the needed exogeneity of variables.
relaxation of the intervening non-natural constraints. 3 As I discussed in
Chapter 5, a disequilibrium will be an illusion in Coase™s sense whenever
NEOCLASSICAL METHODOLOGY AND PSYCHOLOGISTIC one can show that it is really an equilibrium and that its reality would be
INDIVIDUALISM apparent if we were to properly perceive that the intervening constraints are
the logical consequences of the natural givens (viz. of externalities).
It is a central methodological feature of any neoclassical theory that the
It is unfortunate that most neoclassical economists confuse psychologis-
only exogenous variables allowed are those natural constraints such as
tic individualism with methodological individualism and the situation is not
resource availability and naturally given psychological states of individuals
helped by Keynes™ reliance on such things as subjective probabilities.
such as their tastes or preferences. This limitation on acceptable exogenous
Referring to his theory of the consumption function, he says, ˜This psycho-
variables is much stronger than mere ˜methodological individualism™ which
logical law was of the utmost importance in the development of my own
requires only that neoclassical explanations be individualist “ that is, be
thought™ [1937, p. 220]. But perhaps Keynes™ insistence on taking a
based on the notion that only individuals make decisions. As I noted in
psychologistic view of decision-making is only because he wants his criti-
Chapter 8, the stronger version, which is called ˜psychologistic
cism accepted. In particular, he wants to avoid its being automatically
individualism™, should not be confused with individualism per se.
rejected by proponents of neoclassical economics. He surely realized that it
Individualism per se does not require any commitment to reduce all
is all too easy for them to think his view might entail the abandonment of
economic explanations to matters of psychology as John Stuart Mill [1843]
neoclassical theory.
© LAWRENCE A. BOLAND
136 Principles of economics The foundations of Keynes™ methodology 137
KEYNES™ MACRO-VARIABLES VS NEOCLASSICAL to be on the boundary of their capabilities? This question, we shall see,
INDIVIDUALISM reveals the importance of Keynes™ idea of ˜liquidity™. What if an individual
chose some degree of ˜liquidity™? By choosing to have liquidity individuals
Keynes implicit insistence on a necessary role for macro-variables in the
deliberately choose not to operate on the boundary of production
explanation of individual decision-makers could create methodological
possibilities. But, most important, there is no way to rationalize the choice
problems for any ˜counter-revolutionary™ Keynesian model. Macroeco-
of liquidity in a neoclassical framework since the existence of liquidity
nomic variables (those whose values depend on the behaviour of all
itself is inconsistent with maximization (as maximization requires being on
individuals in the economy) do not present a problem if we restrict our
the boundary).
analysis to long-run equilibria. But this requirement supposedly leads to
highly unrealistic models (˜in the long run we are all dead™) and thus the Y
need to look at short-run models. The important question here is whether P
y
restricting economics to short-run models necessarily violates the require-
P
ments of methodological individualism. x
S
R
To say that Keynes insists on a short-run perspective for economic
explanations is not to criticize Keynes for not being individualistic. In a
very important way he was more individualistic than typical neoclassical
W
economists. As Spiro Latsis [1972] has argued, the neoclassical
T
maximization model suffers from not truly allowing free choice by the
individual decision-makers in question. If an individual in the long-run PPC
equilibrium is given a utility function by nature and the constraints are also
given by nature, the choice option which maximizes utility is
X

<<

. 15
( 25 .)



>>