<<

. 11
( 25 .)



>>

6 That is, if [5.1], [5.2a] and [5.3] hold, [5.2b] must also hold.
erature, however, one can find many critiques that are focused on what are 7 That is, MC ≡ W / MPPL.
claimed to be essential but neglected elements of neoclassical explanations. 8 The calculation follows from the definitions of these terms:
µ ≡ (‚Q/Q)/(‚P/P) ≡ (P/Q)·(‚Q/‚P)
There are two particular exogenous elements that have received extensive
and
critical examination. One is the question of what a decision-maker needs to
MR ≡ ‚(P·Q)/‚Q ≡ Q·(‚P/‚Q) + P·(‚Q/‚Q)
know to be a subject of the maximization assumption. The other involves
© LAWRENCE A. BOLAND
86 Principles of economics Axiomatic analysis of disequilibrium states 87
” P·[1 + (Q/P)·(¶P/¶Q)]. ANPK·K ” X “ (W/Px )·L
Thus, we can transform [ii] into the following:
MR ” P·[1 + (1/e)] 2
(¶ANPK /¶K ) = [(¶X/¶K)/K ] “ [(ANP ·K )·(¶K/¶K )]/K . [iii]
K
and since P = AR, the relationship between AR and MR follows.
Since (¶X/¶K ) ” MPP and (¶K/¶K ) = 1, we can further obtain:
K
9 See above, p. 66.
(¶ANPK /¶K ) = (MPP “ ANP )/K. [iv]
10 The implausibility of the firm being a perfect competitor with regard to output K K
prices does not necessarily imply an implausibility of the firm being a perfect With [iv] we can see that if the slope is positive (i.e. ANPK rising) then
competitor with respect to input prices. That is, a few big firms in one industry (MPPK “ ANPK) > 0, which implies MPPK > ANPK. And, if the slope is zero
still may compete with many other industries for labour (or capital). This, of (i.e. the slope is horizontal when ANPK is maximum) then
course, assumes at least a minimum degree of homogeneity or mobility of (MPPK “ ANPK) = 0, which implies MPPK = ANPK.
labour “ that labour could easily move from one industry to another. If for any QED
reason this is not the case, then we will have to include the elasticity of labour 15 Note that the marginal and average cost curves are short-run curves in Figure
supply, x, in the calculation of Marginal Cost. If we do this, we will get (for the 5.2(b). I will not try to define an intermediate version since it will not add much
short-run equilibrium): to the analysis.
MC = (W/MPPL ) [1 + (1/x)] 16 As Samuelson [1972] noted, for there to be a net profit for an entire economy
begs the question of whether there is a Santa Claus [see further Boland 1986a,
But since I wish to keep things as uncomplicated as possible here I will not
Chapter 2].
develop this type of imperfect competition further.
17 Lange uses m to represent the output of machines but here I will use K, to
11 The difficulties with combining the notion of imperfect competition with a
maintain the notation of this chapter.
long-run or general equilibrium model are not new. Recent discussion [e.g. Hart
18 It should be noted here that Lange does not state equation [L5¢] since he derives
1985; Bonanno 1990] have complained that most attempts to do so [e.g.
both [L4] and [L5] using Lagrange multipliers and thus implicitly assumes [L4]
Negishi 1961] usually have involved compromising assumptions that leave the
and [L5] are both true. By recognizing [L5¢], I am making it possible to treat
end results far from being an ordinary general equilibrium model augmented
[L4] and [L5¢] separately while still recognizing that Lange™s equation [L5] is
with the assumption of imperfect competition. John Roberts and Hugo
also a necessary condition of a maximum X.
Sonnenshein [1977] seem to be going further by arguing that such an
19 According to Lange, the real rate of interest is zero when X is maximum [p.
augmentation is impossible. In my simple-minded arguments which follow it
169]. It should be noted here that my representation of Lange™s model is
seems that the problem is not just a question of coming up with a clever
slightly different from what he explicitly states. Lange takes equation [L4] as
modelling technique but rather a fundamental logical obstacle.
obviously true such that any disequilibrium can only be the result of my
12 That is, by analogy we can see that using equation [5.5] yields:
equation [L5¢] not being true. All of Lange™s propositions still follow from my
“ b = (¶Q/Q)/(¶AC/AC) ” (AC/Q)·(¶Q/¶AC)
representation of his model.
and since MC ” ¶(AC·Q)/¶Q ” Q·(¶AC/¶Q) + AC·(¶Q/¶Q)
20 See note 8.
” AC·[1 + (Q/AC)·(¶AC/¶Q)]
21 If instead we assume the profit-maximizing firm has two inputs, L and K, then
MC ” AC·[1 “ (1/b)].
we get
the measure (1/b) is increased by the factor [1 + (Pk ·K )/(W·L)].
13 Consideration of the intermediate-run equilibrium makes it possible to entertain
an alternative assumption for the goal of the firm in the intermediate run even
when the firm may wish to maximize profit in the short run. While it will be
easy to show that maximizing the rate of return makes sense only when
comparing equal amounts of investment (i.e. it is possible to make more profit
at a lower rate of return when the amount is not fixed), it is not uncommon to
find people bragging about high rates of return achieved as if this were optimal.
14 Consider the relationship between MPPK and ANPK. In particular, let us show
that
ANPK = MPPK whenever ANPK is maximum (with respect to K), and
ANPK < MPPK whenever ANPK is rising as K increases.
By definition:
ANPK ” [X “ (W/Px )·L] / K. [i]
Now let us determine the slope of the ANPK curve (¶ANPK /¶K ) by
differentiating equation [i]:
2
(¶ANPK /¶K ) = [(¶X/¶K ) “ 0]/K + [X “ (W/P )·L]·(“ 1)·(¶K/¶K )/K . [ii]
x
Since by [i]:
© LAWRENCE A. BOLAND

Part II

Some neglected elements
© LAWRENCE A. BOLAND

6 Knowledge in neoclassical
economic theory




the economist™s advice to policy-makers must often appear crude and
be misleading ... he gives the impression that investment policy is a
matter only of millions spent per year, no matter on what. Efficiency is
for him a matter of best mixes, not of best shapes. He seems to treat
knowledge as a stuff, obtainable in measurable quantities for a known
expenditure, and guaranteed to produce effects knowable in advance;
he believes that we can know in advance precisely what it is, in all
essentials, that we are going to find out. Better a contradiction in terms
than acknowledge a chink, let alone a gaping rent, in the armour of
rationality.
George Shackle [1972, pp. 114“15]

Whatever assumptions about knowledge we may attribute to it, general
equilibrium does not seem to stand up well to a critical inquiry. In
modern Austrian economics, by contrast, we find the problem of
knowledge to be a matter of fundamental concern.
Ludwig Lachmann [1976, p. 55]


Neoclassical economic theory is often criticized for neglecting an essential
element of knowledge in models of economic decision-making. The most
common critiques would have us reject all neoclassical models because
they are claimed to be based upon ˜perfect knowledge™ and the like. Often
it is argued that neoclassical explanations are incomplete without a formal
treatment of uncertainty and information search. The distinguishing feature
of such critiques is the presumption that assumptions regarding
imperfections in knowledge can be recognized in the neoclassical world
without, at the same time, completely undermining other desirable
methodological properties of this framework, such as internally stable
equilibria, consistency with ˜rational™ decision-making and in general an
˜explicitness™ regarding explanation. Other more radical critics find such a
proposal for piecemeal reform untenable. Many neo-Keynesian thinkers,
among others [e.g. Clower 1965; Leijonhufvud 1968; Kornai 1971], argue
that any systematic programme to incorporate imperfect and incomplete
© LAWRENCE A. BOLAND
92 Principles of economics Knowledge in neoclassical economic theory 93
knowledge into economic theory must sacrifice the traditional neoclassical guished from methodological) features of a decision-making environment
concepts of (general or long-run) equilibrium. George Shackle™s [1972] to play a major role in determining the decision-maker™s response to
lengthy critique is even more uncompromising. He argues that such a knowledge ˜shocks™. At the end of the chapter I will consider a possible
programme must imply the sacrifice of all the methodologically desirable solution to the problem of knowledge dynamics which sees theories of
properties listed above. knowledge (˜epistemologies™) as an autonomous foundation for any expla-
Shackle™s critique is perhaps the most interesting (and dangerous) of the nation of an individual™s methods or decisions which are based on histori-
above critiques and it is therefore of interest that the Austrian economist, cal observations or expectations.
Ludwig Lachmann [1976], has argued that Shackle™s critique constitutes a
successful and decisive challenge to neoclassical economics. Lachmann
MAXIMIZATION AS ˜RATIONALITY™
sees Shackle™s results as even grounds for rehabilitating Austrian views
which give a more fundamental role to knowledge and changes in It is common to find economists using the term ˜maximizing™
knowledge, notably the Austrian theories of Hayek and Ludwig Mises. interchangeably with ˜rational™. As Samuelson noted many years ago
In this chapter I will focus primarily on the arguments of Lachmann and [1947/65, p. 98], what most philosophers might call ˜rationality™ is a much
Shackle and in particular on Lachmann™s overriding view that, since the stronger concept than what is required for decision-making. For
theories of Shackle and the Austrians give a more strategic role to Samuelson, ˜consistency™ was sufficient “ the Axiom of Revealed
knowledge and its limitations, they possess a clear-cut advantage over Preference is merely an expression of consistency. While in many cases
neoclassical economics in explaining economic phenomena. While one can one could substitute ˜consistent™ for ˜rational™, it would be misleading
agree with both authors that no economic theory can be methodologically when the stronger notion is intended. The stronger notion of rational is
complete without a careful specification of the knowledge considerations often a confusion between the mechanics of giving an argument in favour
lying behind all decision-making, their arguments are insufficient for either of some proposition and the psychology of the person stating the argument.
the rejection of neoclassical economics or the resurrection of Austrian The psychology version is not what economists usually mean by ˜rational™
economics. Rather, what Lachmann™s essay reveals is that neither the even though they sometimes refer to a failure of an argument as evidence
neoclassical nor the Lachmann“Shackle viewpoint under discussion is of the ˜irrationality™ of the decision-maker. The accusation of ˜irrationality™
explanatorily complete with respect to knowledge. Specifically, neither is but a left-over artifact of the eighteenth century rationalism which
provides a satisfactory solution to what might be called ˜the problem of Voltaire parodies in Candide. The eighteenth century rationalists would
knowledge dynamics™ “ the problem of defining an explicit and non-trivial have us believe that if one were rational one would never make a mistake
role for changes in knowledge to play in the explanation of the transition and thus whenever we make a mistake (e.g. state a false argument) then we
between short-run (temporary) equilibria and long-run (general) equilibria. must be irrational [see further, Agassi 1963].
Before I begin discussing Lachmann™s and Shackle™s viewpoints I need One does not have to take such a strong position to understand what
to explain why these critics seem to have an excessive concern for the economists mean by a rational argument. All that is intended is that
requirements of ˜rationality™ rather than the more mundane notion of whenever one states an argument “ that is, specifies a set of explicit
maximization that I discussed in Chapter 1. Once this distinction is clari- assumptions “ the argument will be rational if and only if it is logically
fied, I will examine the failures of the neoclassical and Lachmann“Shackle valid. Logical validity does not require that the argument be true but only
viewpoints to provide a satisfactory solution to the problem of knowledge that the assumptions are logically sufficient, that is, that the conclusions
dynamics. In doing so, I will have to discuss two other important distinc- reached are necessarily true whenever the assumptions are all true. But
tions. One is the methodological distinction between exogenous and why the concern for ˜rational™ arguments? One reason for the concern is
endogenous knowledge in decision-making. The other is the distinction the universality and uniqueness provided by rational arguments. The
between epistemology and methodology which will play a major role in promise of ˜rationality™ is that once the assumptions are explicitly stated,
this chapter. Using these distinctions it will be argued here that a central anyone can see that the conclusions reached are true whenever the
shortcoming of both the neoclassical and the Lachmann“Shackle view- assumptions are true. That is, if the argument is rational, everyone will
points resides in their failure to permit the epistemological (as distin reach the same conclusions if they start with the same assumptions. It is
this universality of rational arguments that forms the basis of our
© LAWRENCE A. BOLAND
94 Principles of economics Knowledge in neoclassical economic theory 95
understanding behaviour or phenomena. If the behaviour or phenomena can in short-run models is a fixed and exogenous parameter. The assumption of
be ˜rationalized™ in the form of a rational argument for which the behaviour exogeneity guarantees that knowledge considerations can determine short-
or phenomena are logical conclusions, then anyone can understand the run choices, but not vice versa. The assumption of fixity guarantees that
behaviour or phenomena if one accepts the truth of the assumptions. any variation in economic choices can be fully explained by variations in
In the nineteenth century this notion of universality was captured in the objective factors (i.e. factors other than knowledge).
notion of maximization since both notions involve similar mechanics. If we The assumption of fixity has often been defended on the basis of either
can specify an appropriate objective function for a decision-maker who is a of two propositions:
maximizer then we can understand the choice made. This is because, if the
(a) That there exists unlimited ˜perfect knowledge™ “ that is, if knowl-
objective function (e.g. a utility function) is properly shaped so that there is
edge (of past, present and future) is perfect, then it follows trivially
a unique optimum, then everyone using this function while facing the same
that it cannot change, or
constraints will make the same choices. Thus, again, it is the universality
(b) That the time period being considered by the theory is too short to
and uniqueness that form the basis of our understanding. Every
permit any knowledge change whatsoever, the limiting case being
neoclassical theory is offered as an intentionally rational argument. The
explanation at a point in time (˜statics™).
explicit assumptions include those which specify the shape of the objective
function, the nature of the constraints and, of course, the assumption of Knowledge, in the sense of (b), is thus analogous to capital in the short run.
maximization. But on either grounds, the assumption of fixity leads directly to
The criticisms discussed in this chapter focus on how the decision- Lachmann™s (and Shackle™s) major criticism of the theory. In a world of
maker knows his or her objective function or the constraints. The question actual uncertainty, knowledge cannot be stable but must inevitably be
asked by this type of criticism, which presumes that rationality is always volatile; thus short-run equilibria are extremely temporary. Of course, no
the stronger notion, is whether there is also a rational way to acquire this explanation of a short-run equilibrium per se is sufficient for the
essential knowledge. Many people apparently still think that one can determination of the eventual long-run equilibrium. Since a long-run
inductively acquire knowledge by means of an inductive logic “ a logic equilibrium is merely a special short-run equilibrium, the attainment of a
which uses singular observations as assumptions and reaches general, long-run equilibrium presumes the existence of the one state of knowledge
universally true conclusions. Trying to show how one acquires true knowl- appropriate for that special short-run equilibrium.
edge in this way always involves what is called the ˜problem of induction™. The methodological problem which neoclassical economics presumes to
Unfortunately, this is not a solvable problem since there is no inductive be solved is: How does knowledge change to that which is necessary for
logic that will meet the requirements of universality and uniqueness in the long-run equilibrium state, that is, to the one state of knowledge which
every case as implied by the notion of the ˜rationality™ of an argument. is appropriate for the special short-run equilibrium which holds in the long
Whether one thinks the ˜problem of induction™ is solvable or not, the run? A complete explanation of long-run equilibrium must provide an
questions raised by Lachmann and Shackle do not require induction or explanation of knowledge dynamics [see Arrow 1959b; Gordon and Hynes
rationality in the stronger sense. Maximizing decision-making does require 1970]. But, if the acquisition of the knowledge appropriate for long-run
knowledge of the objective function and of the constraints (e.g. prices) and equilibrium is explained, knowledge ceases to be exogenous. 2 In the long
if we are to explain the choices made we must somehow deal with the run, knowledge is an endogenous variable (like prices or capital) hence
decision-maker™s knowledge.1 knowledge does not play a decisive role “ at least not in the sense of the
role played by individuals™ tastes and the current state of technology. This
means that, for the purposes of determining or calculating the long-run
THE METHODOLOGICAL PROBLEM OF KNOWLEDGE
equilibrium, (endogenous) knowledge is irrelevant.
As Lachmann notes, identifying precisely what assumptions concerning Lachmann™s and Shackle™s criticisms of the above view may be seen to
knowledge distinguish neoclassical theory ˜is anything but easy™ [1976, p. be more than just a plea for ˜realism of assumptions™. First, if Lachmann™s
55]. Nevertheless, it is still possible to identify the basic methodological criticism of neoclassical theory is simply that it does not take knowledge
flavour of the neoclassical view of knowledge and this may be conveyed by into account in any explicit form even though we clearly know that states
the proposition that, no matter how knowledge is characterized, knowledge
© LAWRENCE A. BOLAND
96 Principles of economics Knowledge in neoclassical economic theory 97
of knowledge do determine the properties of short-run economic choices, economics can be seen to be based on the unwarranted fusion of the
then it follows trivially that the neoclassical explanation must be incom- methodological problems connected with the fixity of knowledge and the
plete and its predictions arbitrary (in the sense that neoclassical results epistemological problems connected with the perfection of knowledge.
must vary indeterminately with an unspecified state of knowledge). This These two types of problems are easily made indistinguishable whenever
would be an unfair representation of neoclassical economics and therefore one insists that perfect knowledge must have an indispensable place in the
an unfair criticism. As suggested above, a better way to characterize the neoclassical framework “ for example, when it is presumed that, without it,
neoclassical tradition is as one where, in the long run, knowledge is (complete) rationality is impossible. A commitment to a perfect knowledge
explanatorily irrelevant and in the short run it is specified explicitly as assumption on these confused grounds thus constrains neoclassical theory
fixed and exogenous. We may then judge this specification against other to use an assumption involving the fixity of knowledge. Yet this constraint
alternatives, specifically against those of Hayek, Mises and Shackle. only exists because of the presumption that perfect knowledge is a
Shackle™s critique [1972] may be seen as an attempt to show that the precondition for rationality “ a view which has been criticized [e.g. Tisdell
assumption of fixed and exogenous knowledge is unsatisfactory primarily 1975]. In short, if we separate rationality from perfect knowledge, the way
because it means relinquishing the explanation of economic processes over is made clear for the introduction of ˜imperfect™ and thus potentially-
time (˜economic dynamics™). Shackle reaches this conclusion in two ways. variable knowledge into neoclassical theory. A possible key to this
His first argument, and the one he stresses, introduces an additional separation is the rejection of the Marshallian long-run vs short-run
proposition: dichotomy.
Shackle does not make these arguments. His critique is essentially in an
(c) That perfect knowledge is possible only at a single point in time
earlier tradition of showing that any explanation which requires the
[1972, p. 165].
assumption of perfect knowledge must be inconsistent with any theory
It then follows directly that, in so far as neoclassical theory depends upon which incorporates dynamically-variable knowledge. It should also be
the assumption of ˜perfect knowledge™ to explain the ˜fixity™ of knowledge, noted that while Shackle is especially wary of the epistemological
neoclassical theory can only be rationalized for a point in time and not over problems which are entailed by the perfect knowledge assumption, he is
time. It is thus only if unlimited perfect knowledge could exist over time notably lax on developing the epistemological rationale for his own
that a neoclassical theory based on fixed knowledge could produce viewpoint. This will be the major critical theme of the following sections.
meaningful dynamic explanations. Would the incompatibility between I discuss Shackle™s version of the Austrian arguments in some detail
neoclassical theory and dynamic explanation be removed if we settled for a here because, according to Lachmann [1976, p. 57], Shackle™s arguments
view of knowledge as limited, incomplete or otherwise imperfect (i.e. represent a convenient modern expression of much of the Austrian
˜expectational™) and in turn introduced the view that this knowledge was viewpoint and stand as a major source of criticism of neoclassical theory.
fixed or rigid over an acceptable, yet small, duration? Shackle™s second And the essence of this viewpoint is that knowledge is better specified as
argument is that the answer to this question is ˜No™. Since Shackle [1972, an exogenous, yet highly volatile, item in our economic explanations. Note
pp. 77, 180, 436] sees expectations as subject to moment-to-moment that from Shackle™s perspective the conflict between the neoclassical and
instability and thus as perfectly volatile, even a neoclassical theory which Austrian viewpoints involves only the variability of knowledge over time;
sacrifices the assumption of perfect knowledge is still limited to a point in it does not involve its exogeneity. While Lachmann refers to the possible
time. If Shackle is suggesting that neoclassical economics faces only two ˜endogeneity™ of knowledge in Shackle™s theory [p. 56], it is important to
alternatives, then it must choose to analyze either static situations or recognize that this concept is interesting only when it is defined relative to
situations of perpetual change and instability, but not both. a dynamic process such as learning dynamics. The appropriate relation
One can easily agree with the general spirit of Shackle™s criticism, since between changes in knowledge and adjustments in economic choices is that

<<

. 11
( 25 .)



>>