<<

. 16
( 25 .)



>>

mathematically predetermined and only needs to be found by the
Figure 9.2 Production possibilities curve
individual. There is no free choice in long-run equilibrium. The only
question is whether the individual is smart enough to know when his or her
Before I examine the idea of liquidity I need to reconsider Marshall™s
utility is maximum. Of course, the concept of ˜constrained maximization™
world without the phenomenon of liquidity “ namely, the textbook world of
has always had its methodological problems. 4
Marshallian-neoclassical maximization where all predictions and
explanations are based on one or more boundary functions. I will do so by
THE MARSHALLIAN BACKGROUND OF CONSTRAINED- briefly looking at the object of Keynes™ assault: Marshall™s methodological
OPTIMIZATION METHODOLOGY approach to economic explanations. As I discussed in Chapters 2 and 3,
Marshall™s methodology is quite straightforward and involves the
Latsis™ view of neoclassical methodology may be too severe. Nevertheless,
application of the Principle of Substitution subject to the requirements of
there is a difficulty with any neoclassical framework which makes
the Principle of Continuity in his economic explanations. Recall again that
˜constrained maximization™ the keystone, and this difficulty is a concern of
the Principle of Substitution merely says that every individual makes a
Keynes™ assault. The difficulty is that with a neoclassical model one cannot
choice between options by selecting the one option which maximizes a
explain the existence of ˜liquidity™. In neoclassical maximization models
given objective function. The Principle of Continuity is co-requisite with
all optima are necessarily points on a boundary formed by the natural
the other principle because deliberate maximization presumes that the
constraints, much as the textbook Production Possibilities Curve (PPC)
options lie on a continuum. Any finite endpoint usually represents one of
forms the upper bound on the possible mixes of output combinations
the constraints facing the individual decision-maker. The chosen option
limited only by the available resources and technologies (see Figure 9.2).
must not be at one of the endpoints of that continuum “ that is, the chosen
We are to explain the state of an economy by showing that the economy is
(maximizing) option must be somewhere between the endpoints. If the
at a point on such a boundary (point R) and that the shape of that boundary
optimum were at an endpoint it would not be clear whether the chosen
(viz. its slope) at the chosen point explains prices. Why would anyone want
option was the most desirable or simply accepted.
© LAWRENCE A. BOLAND
138 Principles of economics The foundations of Keynes™ methodology 139
While Marshall™s methodology of explanation can easily be based on his point on the PPC, the corresponding point on the locus of tangencies in the
two principles, the task of using it runs into some procedural difficulties. Edgeworth“Bowley box must have been chosen. To be an optimum point
One cannot explain everything in the universe all at once. Every on the PPC, the slope of the PPC must equal the ratio of the prices for the
maximization situation involves constraints of which some are irrelevant two goods illustrated and at the point on the tangency locus in the
endpoints and others merely define the situation. For example, in the Edgeworth“Bowley box the slopes of the respective iso-quants must both
consumer maximization model, the budget line is a constraint but is not equal the given ratio of factor prices. These are all necessary conditions for
an equilibrium allocation.6
always an exogenous variable. Given enough time, the individual consumer
chooses it, too [cf. Clower 1965]. So, as I have noted before, Marshall™s Now, if a point interior to the PPC were chosen, the relationship
strategy is to lay out a continuum consisting of ever longer time periods in between prices and marginal productivities would break down since the
which more variables become endogenous. Again, it needs to be pointed shape of the boundary will be irrelevant. If an interior point is chosen, all of
out that when discussing long-run decisions “ those which require a lot of the neoclassical marginal productivity theories of income distribution
time “ the firm will always be in a position where it has been able to would be in serious jeopardy if not completely lost if the individuals did
optimize with respect to the shorter-run variables. One might say that not operate on their respective boundaries. I shall argue below that this
Marshall™s explanatory methodology is all a matter of peeling the temporal breakdown is the importance of Keynes™ introduction of ˜liquidity™. The
onion. usual neoclassical assumptions and results cannot be maintained if
˜liquidity™ is to be accommodated.
W
Available capital




THE KEYNES“HICKS METHODOLOGY OF OPTIMUM
T ˜LIQUIDITY™
X Let us now turn to the matter of Keynes™ concept of liquidity. As a student
R
I was once taught that ˜liquidity™ was the key contribution of Keynes. Later
Y
I was taught that liquidity was only important in terms of the effectiveness
of monetary policy. In these terms, Keynes would seem to have little to say
S
except in a severe depression where interest rates were so low that further
monetary stimulation of investment would not be possible. These views of
Available labour Keynes™ liquidity are quite unsatisfactory. Nevertheless, the concept of
Figure 9.3 Edgeworth“Bowley box liquidity is the source of all the alternative views which say that Keynes
introduced one particular variable or another. For example, there is the
The Marshallian Principle of Substitution methodology always claim that all that matters is Keynes™ assumption that the labour market is
considers the decision-maker to be facing something like a short-run not in equilibrium (and hence the employment is less than maximum “ see
production possibilities curve. The curve forms a continuum and its Figure 9.1). It was sometimes claimed that all that matters is the ˜liquidity
position is limited by given constraints. Note that the PPC represents the trap™. And, of course, many still claim it is just the recognition of
Pareto-optimal allocations of fixed resources which can be represented in a ˜expectations™. All of these can be seen to be merely instances of what
two-factor world by the height and width of an Edgeworth“Bowley box Hicks now recognizes as a general form of liquidity, 7 as I will try to show.
(see Figure 9.3). Specifically, it is a one-to-one mapping between the points
on a locus of tangency points between two opposing production iso-quant
Hicks™ theory of Keynes™ liquidity concept
maps and points on the PPC representing the (maximum) output levels
indicated by the two iso-quants that are tangent. The correspondence A more general view of the concept of liquidity is the key to the
between Figures 9.2 and 9.3 shows that the position of the PPC is limited methodological strategy of Keynes. In his 1979 book, Causality in
by the available amounts of the two factors. If the size of the box is Economics, Professor Hicks has carefully explained his view of the concept
increased, then the PPC will be located further from the origin. 5 To be at a of ˜liquidity™. While Hicks is more concerned with the quasi-Austrian
© LAWRENCE A. BOLAND
140 Principles of economics The foundations of Keynes™ methodology 141
aspects of real-time decision-making, he reveals the importance of why clear the market, there is excess supply. Such excess supply may very well
there may be good reasons for an individual to be choosing an amount of represent a desirable state from the standpoint of the employer. For some it
liquidity. Here the importance of an individual™s choosing an amount of is always desirable to be able to expand production immediately whenever
liquidity would be that the individual is choosing to be inside his or her necessary. Similarly, whenever the wage is below the market-clearing
possibilities boundary. wage, a thirty-five-hour work week may be optimum for an individual even
The point raised by Hicks is that in a world that is either static or moves though he or she could work a sixty-hour week. Having some free time to
in a sequential fashion (step-by-step, as in Marshall™s world of comparative pick up some emergency side money when it is needed may be more
statics where there is always enough time allowed to make any desirable than working to one™s limits according to an inflexible contract.
adjustments), there really is no need for liquidity. However, in a world Good business may also require the ability to choose one™s speed of ad-
where many things are happening simultaneously, the presumption of justment to changing conditions. Sometimes a quick response is better than
optimization is usually misleading. Every decision involves an actual a slow response and at other times it is the reverse. Flexibility is the key
decision situation (a set of relevant givens “ income, prices, technology, here. But it is not a variable that can be chosen in the same way one would
availability, etc.) and a time lag. Since every decision takes time to choose a quantity of food or a quantity of capital to achieve a given current
implement, during that time the original givens (which depend on the objective. The reason is that one™s choice of liquidity, be it financial as
actions of other people) might have changed and thus the implemented Keynes discussed or non-financial as Hicks noted, always depends on
choice decision might not actually be the optimum for the new givens. 8 variables which cannot be easily determined. However, knowledge of them
For example, if one thinks the future will favour large fuel-inefficient would be essential for the usual neoclassical explanation.
personal automobiles and that there will be an unlimited amount of fuel,
then specializing in the production and marketing of such autos might be
THE CONSEQUENCES OF ˜LIQUIDITY IN GENERAL™
the optimum choice regarding one™s production technology. If the market
should suddenly shift in favour of small efficient autos or if the availability While Keynes focuses his idea of liquidity on the narrower concept of
of cheap fuel disappears, then one™s profit potential would be drastically financial liquidity, it is easy to see that the idea of liquidity can be extended
altered. The same would be true in the less dramatic case where a certain to all situations where the decision-maker is placed inside the boundary of
size of market is anticipated but there is a sudden increase in demand due his or her capabilities. The classic example is that of ˜excess capacity™
to a strike at a competing firm. If the previous level of output was the usual which is a position where the firm has enough capital to increase
neoclassical long-run optimum (price equals average cost) then the firm production without raising unit costs (i.e. it is within the infinitely rising
would not be able to respond competitively by producing more unless there cost limit at the absolute boundary of production capabilities). Whenever
was more production capacity. To increase capacity would take time and the firm operates with ˜excess capacity™ the economy must be inside the
might not even be the optimum after the strike is over. It would seem that PPC and, being inside, small adjustments in the chosen point may not
zero excess capacity for the firm in the Marshallian short run “ that is, no affect the costs or productivities.
liquidity in the non-financial sense “ would not be an optimum situation. To understand the significance of stressing the desirability of liquidity
However, the appropriate optimum (with regard to excess capacity or we need also to see why it is not part of the usual neoclassical model.
liquidity) may not be knowable by the firm since knowledge of it depends Consider again the textbook PPC of Figure 9.2. For the sake of discussion,
on unknown contemporaneous actions of other people as well as on the let us think of a firm producing two different goods, X and Y, with two
unknown future. factors, L and K, such that the firm™s production decisions include deciding
on an allocation of the available factors between the two production
processes and thereby a point within the production possibilities set. The
Keynes™ use of liquidity
boundary of this set is the PPC.9 So long as more is always better, any
Allowing for liquidity as a deliberate choice variable is central to Keynes™ individual facing the limitations represented by such a curve will want to
assault. From Keynes™ viewpoint, such liquidity is simply good business. be producing on the boundary of possibilities as represented by the curve.
For example, usually, whenever the labour market is in a state of To produce on that boundary, all available resources will need to be fully
˜disequilibrium™ where the current real wage is above the one which would employed by definition of the PPC. If one does not use all resources fully
© LAWRENCE A. BOLAND
142 Principles of economics The foundations of Keynes™ methodology 143
then necessarily the chosen point will be inside the boundary. the above equalities between relative prices and relative marginal
Whenever the firm is producing on its PPC optimally (i.e. maximizing productivities will be satisfied. Point W in Figures 9.2 and 9.3 represents
its ˜profit™ or net revenue) we know that the relative marginal productivities the misallocation of fully employed factors. If we wish to consider a case
of those resources in the production of X will just equal the relative where not all of the available factors are being employed then we need to
marginal productivities of those resources in the production of Y since both determine a different PPC for the under-employed case. So, I have
ratios must be equal to given relative prices of those inputs. Similarly, for reproduced the PPC of Figure 9.2 in Figure 9.4 such that the under-
any resource, the ratio of its marginal productivity in the production of X to employment PPC (PPCue ) will be inside the full-employment PPC of
that in the production of Y must just equal the same ratio for any other Figure 9.2. I illustrate the relationship in Figure 9.4 where W is an interior
input since these ratios will all equal the given relative price for the two point for both PPCs and may correspond to a misallocation of the
products. What is significant about all this is not that these well-known employed factors in each case. Point V represents an output mix that is
equalities are achieved but that the individual™s decisions must be optimum for the given prices but still implies an under-employment of
responsive to changes in the given prices. Note that this is why the issue of factors. At point W profit (or net revenue) is not being maximized with
˜stickiness™ of wages is so important since whenever any price is artificially respect to all inputs (see Figure 9.3). As a result the income distribution
restricted from changing in response to different market conditions, that will not likely reflect the indirect demand for productive services. Since
price no longer provides useful information for any decision-maker. there is more than one way to be at an interior point (e.g. excess capital,
Generally speaking, prices are easier to change than quantities. A fixed excess labour or any combination of these), and since by being there the
price only slows down any adjustment process. Although it may take much firm may not be maximizing profit with respect to at least one of the inputs,
longer, in the usual neoclassical model it is at least logically possible to predicting where the firm will be if it has chosen to respond to any change
find values for the quantities such that all of the equations can be restored in the prices would be difficult. Similarly, if the firm has chosen a point
as equalities. inside the boundary, restricting any input may not have immediate effects
What is most important here is that whenever the given prices change on the individual firm™s output level. For these reasons not only is there no
there is an explainable shift from one point on the boundary to another on guarantee that individual firms (or individual consumers) will be doing
that boundary since we can calculate the point on the boundary at which all what society wants, but any attempt by government to alter their behaviour
the equalities are satisfied. And almost always there will be a shift by changing tax rates or by manipulating interest rates may prove to be
quite ineffective in the short run. 10
whenever one of the prices changes. The whole importance of the
competitive market is that everyone should take prices as the appropriate
Y
signal concerning what to produce or buy. That the price of fuel-inefficient
autos should be falling relative to efficient autos is important social
information. In responding to such a price change by reducing the output of
inefficient autos, the firm is doing what society wants “ just as indicated by R
the change in relative prices.
Let us now consider a firm that is not on the PPC defined by its amounts
available of the two factors. Note that there are two ways to be at an W V
interior point. One way is by not maximizing with respect to all the givens
PPC
“ such would be the case if the allocation point W in Figure 9.3 were
chosen since the slope of at least one of the two iso-quants cannot be equal PPC ue
to the given ratio of factor prices. The other way is by not using all of the
available factors, perhaps for the purpose of providing flexibility (i.e. room X
to maneuver). Figure 9.4 Under-employment PPC
Now what happens when the firm is not operating on its possibilities
boundary “ that is, when, for example, it is deliberately providing liquidity Keynes™ discussion of expectations (when expressed in terms of
in the form of excess capacity? For one thing, except by accident, not all of methodological and epistemological questions) raises similar issues. In his
© LAWRENCE A. BOLAND
144 Principles of economics The foundations of Keynes™ methodology 145
1937 QJE article about the General Theory he explicitly identifies decision maximization. In the face of uncertainty, liquidity is a means of avoiding
processes which are not optimizing. Collecting all the available information the difficult determination of maximizing choices. Thus, when it comes to
to make an investment decision may be uneconomical even if it is logically liquidity (which, in the face of uncertainty, is offered as a necessary short-
possible. Simple rules-of-thumb (˜conventional judgement™) may be run endogenous variable in the General Theory), there may not be any
adequate but may not be optimizing even for the state of limited good reason to doubt the presumption that liquidity has been chosen
knowledge. Follow-the-leader behaviour may be easier to justify than optimally “ except one. If liquidity could be chosen like any other variable
maximization. Since all investment decisions involve estimations about there would be no need for liquidity! So, I am arguing that Keynes™
future states of affairs, relying on the going interest rate as an indicator primary assault lies in the empirical claim that in any individualist model
about the appropriate relative price for future-vs-present consumption of an economy liquidity (or excess capacity) is a necessary object of choice
decisions (following Irving Fisher) presumes that it has been determined in and thus all long-run models must be empirically false. The reason why it
a free market of buyers and sellers with perfect foresight. If buyers and is necessary is that so many of any individual™s decisions depend on the
sellers are, instead, using information from sub-optimizing decisions, what status of what we might now call ˜macro-variables™ “ variables which
does the market interest rate indicate to an individual decision-maker? High depend on the contemporaneous actions of many other individuals.
interest rates may only reflect the current state of optimism rather than Stressing the aggregate or macro aspect of the variables only emphasizes
known investment possibilities. this dependence.
The point of Keynes™ assault is that he wishes to challenge the advocates
of neoclassical economics on their own terms “ namely, in a world where
ON EFFECTIVE CRITICISM
only individuals make decisions. If he were to try to criticize them on
It is unfortunate that the so-called Post-Keynesians as well as the counter- radically different terms, his views could too easily be dismissed as being
revolutionaries consider the General Theory to be a ˜blueprint™ for an alter- irrelevant for questions addressed by neoclassical economics. In this case it
native to neoclassical economics. Such a viewpoint leads readers to miss is not clear that Keynes was successful; the only apparent change in main-
the sophisticated criticism and challenge that Keynes offers neoclassical stream economics since the publication of the General Theory has been the
believers. Despite what many critics of neoclassical economics might like introduction into the curriculum of a course called macroeconomics and
to believe, the introduction of liquidity or excess capacity into an otherwise with it the implicit claim that Keynes was dealing with questions that are
neoclassical model does not always conflict with the usual assumption of different from those addressed by microeconomics. Keynes is entirely to
maximization. For all we know the individual firm may have inadvertently blame for this means of avoiding his criticism. He is the one who stresses
chosen the optimum amount and thus have all its marginal productivities the necessary role of macro-variables in the theory of the individual
equal to their respective factor prices. That is to say, whenever there is decision-maker. Perhaps he only introduced ˜macro-variables™ because he
excess capacity, maximization is not logically precluded. What Keynes accepted the psychologistic version of individualism that underlies all of
argued was simply that there is no good reason to think that firms have neoclassical methodology, yet the introduction of such variables was
consciously chosen the optimum amount in accordance with neoclassical against the neoclassical methodological individualist rules. Had he avoided
models. Furthermore, to say firms may not be optimizing does not deny psychologistic individualism he would not have had to stress the
any conscious attempt on their part to choose the optimum amount of ˜aggregate™ variables “ that is, had to emphasize the active role of variables
liquidity “ although, in the face of uncertainty it is unlikely that they could which cannot be explained as being reflections of only the aims of individ-
uals in real time.11 But of course, this conjecture is silly. Had he not
ever succeed. In other words, all the usual elements of neoclassical choice
theory and methodology are here since only individuals are making choices followed psychologistic individualism, as most neoclassical theorists do, he
and those choices are intended to be optimizing. would have been dismissed on these grounds alone “ without ever dealing
For many objects of immediate choice (consumable goods, direct with his criticism. Until mainstream neoclassical economics drops its
services, etc.) there is no good reason to doubt neoclassical maximization. dependence on narrow psychologistic individualism, Keynes™ assault will
However, for objects of choice involving judgements about the future state not be much of a struggle for neoclassical economic theorists.
of the economy (such as investments, capacity, etc.), it is difficult or
impossible to see the decision process as that of straightforward
© LAWRENCE A. BOLAND
146 Principles of economics
NOTES
10 Individualism without psychology
1 The arguments presented here were those I gave in a conference at Cambridge
University in 1983. Most of the proceedings of that conference were sub-
sequently published in Lawson and Pesaran [1985].
2 As a form of individualism, institutional individualism still maintains the view
that only individuals make decisions yet allows Keynesian-type macro-vari-
ables to play a role in the individual™s decision process.
3 All other variables are just ˜independent™ endogenous variables with respect to
the individual decision-maker but ˜dependent™ endogenous for the system as a
whole [see Chapters 2 and 3 above]. Note also that in a broader sense (e.g.
general equilibrium theory) only the variables which are exogenous in the long-
run models are truly exogenous [see Hicks 1979].
4 But not all of the problems are usually discussed [see Chapter 1 above].
5 For an explanation of the relationship between PPCs and the Edgeworth“
[Mathematical Psychics involved] considerations so abstract it would
Bowley box, see Samuelson [1950].
of course be ridiculous to fling upon the floodtide of practical politics.
6 In the special case of the price-taking individual consumer with no market
But they are not perhaps out of place when we remount to the little
power, the possibilities ˜curve™ will always be a straight budget line since that
rills of sentiment and secret springs of motive where every course of
individual does not affect the given prices. The location of the curve is
action must be originated.
determined or constrained by the limited available resources or income. The
Francis Edgeworth [1881/1961, p. 128]
constraints may not be naturally given but only difficult to change in the time
period under consideration. But what is most important here is that the chosen
All human conduct is psychological and, from that standpoint, not
option must be a point on the boundary formed by the ˜curve™. In a set-theoretic
only the study of economics but the study of every other branch of
sense, a possibilities curve is the positive boundary of a convex set of available
human activity is a psychological study and the facts of all such
options.
branches are psychological facts.
7 Specifically, he refers to ˜financial™ and ˜non-financial™ liquidity [Hicks 1979,
Vilfredo Pareto [1916/35, sec. 2078]
94ff].
8 This may not have been what Hayek [1933/39] intended but one can certainly

<<

. 16
( 25 .)



>>