The Competitiveness of Nations in a Global Knowledge-Based Economy
Brian J.
Loasby
Management
Economics and the Theory of the Firm
[1]
Journal of Industrial Economics,
15 (3)
July 1967, 165-176.
Content
The Assumption of Profit Maximization
THE origin of this paper is a belief that, whatever
its virtues as a major component of the theory of value, the theory of the firm
as it exists at present is an inadequate basis for analysing
the process of managerial decision-making.
Professional economists working in industry are ready enough to declare
(to take two examples) [2] that “the theory of the firm based on the marginal
analysis is most unlikely to have any relevance to the real world and is of
negligible use to economists earning their livings in it”; or that “in such
analyses as I have attempted of the actual processes of decision-making in
business I have frankly found the theory almost completely irrelevant”. Perhaps, indeed, few academic economists would
nowadays dispute this point. Of course
many of the concepts used in the theory of the firm are valuable; but the
integrated theory is not.
Customary forms of theorizing about the firm require
that the firms of the real world be treated as ‘black boxes’, which transform
information inputs into decision outputs by managerial processes which need not
be properly understood provided that their effects can be calculated. Now this is, in principle, a perfectly
legitimate intellectual procedure; it is not necessary to be able to explain in
order to be able to predict. [3] The fact that no one yet understands exactly what happens
within a blast-furnace does not prevent blast-furnaces from being used; and the
establishment of statistical relationships does not depend upon the ability to analyse the processes which create those relationships. However, it is precisely the managerial
processes within these ‘black boxes’ which form the subject matter of
management economics. If management
economics is to make a more adequate contribution to explaining business behaviour, to increasing the efficiency of management, and
to the development of management education, economists must begin to look more
carefully at managerial processes. For
this purpose, the assumptions of the traditional theory are hopelessly inadequate.
1. My thanks to all
those who have commented on various drafts of this paper are not intended to
disclaim my responsibility for its present form; in particular, the implied
interpretations of P. W. S. Andrews’s work do not necessarily carry his approval.
Our arguments are quite different.
2. Both privately communicated.
3. For a discussion of
this argument, see K. J. Cohen and R. M. Cyert, Theory
of the Firm (Englewood Cliffs, N.J., 1965), Chap. 2.
165
The Assumption of Profit Maximization
The assumption that the objectives of the firm can be
adequately summarized as the maximization of its profits has had a long life. Perhaps its persistence can be best explained
by the aura of rationality which clings to it, and to the kind of behaviour that apparently follows from it. Since the behaviour
postulated by the marginal theory of the firm is rational behaviour,
then that is the way firms should behave. Profit maximization therefore becomes a test
of managerial competence; it is indeed virtually a moral issue. Since the logic of the theory is impeccable,
evidence that any actual firm does not behave in this way cannot disprove it;
it merely shows that the firm is badly managed. Why, then, should one bother with empirical
testing?
This paper is not concerned with the question, to
which P. W. S. Andrews has directed his attention, whether the commended behaviour does indeed lead to maximum profits, even on its
own assumptions. [4] But it does need to be remembered, by those who extol the
virtue of profit-maximizing behaviour, that such behaviour is rational only if the object is to make as much
money as possible. To argue the
rationality of profit maximization as an objective is strictly meaningless;
rationality is an efficiency measure, concerned with the relationship of means
to ends. It can say nothing about the
ends themselves. If businessmen choose
to prefer other aims to the maximization of profit, the preference may be
inconvenient for the theorist, but it cannot be called irrational.
The assumption of profit maximization also appears to
have practical advantages. The argument
that no better assumptions are available has, it is true, now been undermined
by progress in other branches of the social sciences - which economists, as a
class, too readily ignore; but since not many attempts have yet been made to
work out the economic consequences of alternative assumptions, most economists,
being interested in teaching, using or refining the existing corpus of theory, rather
than labouring on major new constructions, continue
to behave as if the argument was still valid.
Around this unchanged position one is nowadays likely
to find another, and much stronger, defence. Since all assumptions must be unrealistic in
varying degrees, the validity of the assumptions must be judged by the validity
of the results to which they lead; and it is claimed that profit maximization,
while avoiding the complications introduced by alternative assumptions,
produces results which are valid enough for the purposes for which the theory
4. P.
W. S. Andrews, On Competition in Economic Theory (London, 1964).
is to be
used. This article is not intended to
challenge that defence; but it must be pointed out
that the purposes for which the theory of the firm is to be used explicitly
exclude the explanation of decision processes within the firm.
This was not always so; the restriction represents a
forced withdrawal from ground which was once thought to be unmistakably within
the province of the economist. This
abandoned ground has now been invaded by organization theory and operational
research, among others, and the occasional forays of economists into the area
are unsupported by any substantial body of general theory. As long as this situation persists, any claims
by economists that their subject should be at the centre of the study of
management looks distinctly shaky - it has the appearance of the discredited
argument that specialists need to be co-ordinated by
someone who does not have the disadvantage of specific relevant expertise.
Profit maximization has, finally, a strong
intellectual appeal. It provides a
unique solution for each problem; more complex criteria do not. This point is well illustrated by two recent
attempts to construct a theory which is based on other assumptions. In developing his theory of managerial
capitalism, in which managers attempt to maximize the growth of their
companies, subject to certain constraints, Marris
finds himself relying heavily on the takeover bid and the threat of a bid to
ensure that all managers behave in the same way - a job much more elegantly
performed in the accepted theory by the assumption of free entry. [5] O. E. Williamson’s proposition that managers
attempt to maximize a utility function which has as its three principal
components retained profits, staff and ‘managerial slack’ (approximately
equivalent to managerial rent) appears to allow for different relative
valuations of the three components by different managers, and therefore for
different decisions in identical situations. The only hope for unique solutions in Mr.
Williamson’s world appears to be an economy managed by fully conditioned
organization men. [6]
But the intellectual appeal of profit maximization is
spurious. If, instead of assumptions,
one starts with the observed behaviour of firms, then
the virtue of simplicity becomes weakness, and the defects of complexity become
virtues. All firms do not behave in the
same way in similar circumstances, and a theory which helps to explain why they
do not is perhaps to be preferred to one which asserts that they should.
5. R. Marris, The Economic
Theory of ‘Managerial’ Capitalism (London, 1964).
6. O. E. Williamson, ‘A Model of
Rational Managerial Behaviour’, in R. M. Cyert and J. G. March, A Behavioral Theory of the Firm (Englewood
Cliffs, N.J., 1963), pp.
237-52.
167
The weakness of traditional theory is well displayed
in situations of oligopoly, where it has tended to adopt the attitude of the
legendary Scottish preacher: “Brethren, here is a great difficulty. Let us look it firmly in the face and pass on.”
The difficulty is that, within the
traditional framework, the solution depends on the behaviour
of individuals, and the theory has no way of handling this behaviour.
The tendency to argue that only
collusion is rational, and that other kinds of behaviour are irrational and therefore beyond the reach of
analysis is a counsel of despair, resulting from the naive equation of
rationality with profit maximization.
The great attraction of perfect competition theory,
which was inherited by monopolistic and imperfect competition theories too (but
which oligopoly theory lacks), was its determinism; it had no need of the
individual. It is time that he was
allowed back. It would be foolish to
claim that the decision-makers in industry have more than a limited discretion.
In some instances - for example, the
special case of perfect competition - they have none at all; but, in general,
it is safest to assume that there is a constrained area within which they have
some freedom of choice. Occasionally,
that choice can be as crucial as a casting vote. In order to explain and predict these choices
we need to examine more carefully the objectives of the decision-makers.
‘Profit maximization’ is in fact not one assumption
but two: first, that the objective is profit, and
second, that the desired level of performance is the optimum. These assumptions, although closely related,
are best considered separately.
It has generally been accepted that profits, like
other forms of income, are desired, not for themselves, but for the
satisfactions over which they can give command. But although it has long been recognized that
the activities by which wages are earned may include important direct sources
of satisfaction or dissatisfaction in their own right, no similar refinement
has been applied to the concept of profits. Allowance is made for variations in risk and
occasionally for leisure-preference, but that is all. It seems just as likely, however, that those
who make the major decisions within a firm should be interested in the total
net advantages - including possibly substantial satisfactions from the job
itself - deriving from their activities as that wage-earners should be
interested in the total net advantages of their jobs: indeed, given that the
formers’ incomes are usually much bigger and therefore provide a much bigger
margin over the
cost of
necessities, they might be thought to be more interested in the non-pecuniary
factors, as their cost in terms of the value of income foregone would probably
appear less. [7] The possible net advantages to be sought from the control
of a business - and therefore the possible range of objectives of the business
- may be conveniently grouped into four sets, which may be labelled
economic, social, psychological, and organizational.
The set of economic objectives will obviously include
profit; the other objectives will usually be partly complementary to profit and
partly competitive with it. Cyert and March work with a set of five: in addition to
profit they list objectives for production (including both level and stability
of output), stock levels, sales and market share. [8] In The
Practice of Management, Drucker, after declaring that
a business needs objectives in every key area, lists these ‘key areas’ as
follows: market standing, innovation, productivity, availability of resources,
profitability, managerial performance and development, worker performance and
attitude, and public responsibility. [9] It should be noticed that, whereas Cyert and
March are describing, Drucker is an advocate; but his
advocacy is directed squarely at business practice.
Forgetting for a moment the last two on Drucker’s list, we may notice a striking contrast between these
two sets of economic objectives. Profit
is the only one that is included in both. Market standing is by no means the same as
market share: it is much more a matter of the firm’s reputation with its
customers and distributors, and the relation between its range of existing and
potential products on the one hand, and its existing
and potential markets on the other. [10] The distinction between the two lists is simple, and one
that is very familiar in economics: the former is a short-period set, the latter
a long-period set. This is as it should
be, for Cyert and March are concerned essentially
with the day-to-day behaviour of the firm while Drucker is concerned rather to lift the businessman’s eyes
from the pressures of the day and to consider major issues of policy. The fact that short-period objectives can be
described whereas long-period objectives apparently need to be advocated has a
significance of its own in explaining business behaviour.
For each of these sets of objectives, it would be
possible to argue that profit is the fundamental objective, and that all the
others are
7. Douglas McGregor, The
Human Side of Enterprise (New York, 1960),
discusses the manager’s job as a direct source of satisfaction, and
considers the organizational - but not the economic - consequences.
8. Cyert and March, pp.
40-3.
9. P. F. Drucker, The Practice of
Management (London, 5955), pp.
52—3.
10. Drucker, p. 55.
169
objectives
which will be set in such a way as to contribute to its achievement. It would be possible; but it would not be very
plausible. Apart from the argument to be
adduced shortly, that particular members of the firm will associate themselves
- and be required by their organizational responsibilities to associate
themselves - with particular objectives, the performance of a firm in terms of
for example, its share of the market or its record of technical innovation may
well produce satisfactions to the decision-makers which they may value more
highly than the profit foregone by attracting dearly bought custom or by
pursuing expensive ingenuity. If a more
single-minded firm makes higher profits, why should they worry? The satisfaction they desire cannot be
achieved in that way. For them, the
attempt to maximize profits would be irrational.
This first group of objectives may legitimately be considered as the proper concern of the economist. The remaining three sets lie outside his particular territory; but he should be prepared to accept the findings of those who have investigated them. Two of the sets are suggested by the last two items on Drucker’s list. Almost all large companies, at least, would accept in principle that they have certain public responsibilities, which it is not realistic to dismiss as ways of obtaining long-run profits. [11] This is particularly true of companies operating overseas in emergent countries. It is also obviously true of nationalized industries, which are not expected to be profit-maximizers, and are therefore implicitly excluded from the theory of the firm - a striking omission, one would have thought. [12]
Fogarty has recently argued that some at least of these responsibilities should be written into a new Corporation Act. [13] Without necessarily accepting that argument, one may certainly accept its implication that this is a set of objectives which is different from the economic set just discussed. To be a good employer, to assist the local community, to be, in general, a well-behaved and respected member of society - these are aims which are widely accepted by firms, and nowadays widely expected by the public. In fact, the company that appears to be seeking nothing but profits is likely to have a poor reputation with the public - and there are few firms which do not put some value on their reputation in its own right, and not simply as an aid to profit. What this value is the economist can attempt to measure in terms of profit foregone; why it should be what it is does not necessarily concern him. What does concern him is that his assumptions about it should be reasonable.
11. R. Eells, The Meaning of
Modern Business (New York, 5960), p. 70. Much of this
book is concerned with these issues.
12. I owe this point to Mr. D. K.
Clarke of the Department of Politics, University of Bristol.
13. M. P. Fogarty, Companies
Beyond Jenkins (London, 1965).
‘Worker performance and attitude’ may be considered in
part as falling within the definition of a firm’s responsibilities to society. In part, however, it falls within the category
of more specifically psychological satisfactions desired by the firm’s
decision-makers. These satisfactions may
be socially desirable, as, in general, good human relations are, or
undesirable, as the wish to exert power may usually be; what is certainly true
is that these satisfactions are not simply means to an end, but important ends
in their own right. (It is to be hoped,
for example, that we have now passed the stage at which good human relations
were thought to be rewarded - indeed often measured - by increased
profitability.) The firm’s
decision-makers bring their own objectives to work with them. Some of these objectives are sought through
the pursuit of the economic objectives of the business, but others can be
achieved only by modifying these objectives. [14] For example, “many owners of small firms are reluctant to
let any share of their firm leave their own hands... the price of such
independence may well be a failure to grow”. [15] Once again, the economist can attempt to measure the cost
of these psychological objectives, but for his assumptions about their nature
and force he must rely on the social and industrial psychologists.
There remains one more set of objectives to be
considered. The content of this set has
been hinted at in the use of the phrase ‘the firm’s decision-makers’. Existing theory is still built around the idea
of the entrepreneur. Of course it is
realized that the business run by the owner-manager is no longer typical of
industry, and there has been some uneasy recognition that the manager who is
not the owner may be interested in other things than profit. That this may also be true of the
owner-manager has escaped general notice; but it will be observed that our
arguments so far do not depend on whether or not the business is managed by its
owners. There has also been some search
for the identity of the ‘entrepreneur’ in modern business. Some are still to be identified; but it is
significant that they are to be found predominantly among the creators of new
business or among the successful takeover-bidders. In the great majority of large firms the
entrepreneur does not exist. In his
place there is an organization; and an organization is not a collective
entrepreneur. It has distinctive
features of its own, which are the concern of organization theorists.
Hitherto, the interest of economists in organization
has been confined to the relationship between organizational size and effi-
14. Cyert and March, Chap. 3; McGregor, Chaps.
3-4.
15. J. A. Bates, The Financing of Small Business (London,
1964), pp. 20-1.
171
ciency; and it is remarkable, looking back, to see how well
the debate over the existence of managerial diseconomies of scale was kept
going without anyone inquiring how organizations actually work. But for some time now other social scientists
have been observing that members of an organization may have varying objectives
of their own, and that conflicts of interest are not confined simply to those
between management and the shop-floor. In addition, the manager’s terms of reference,
and the demands of the job itself, provide a set of objectives for his sphere
of responsibility; and as a practical guide to departmental decision-making an
objective of maximizing company profits is useless. Moreover, it is much too facile to believe
that the sum even of all the departmental objectives - let alone the sum of all
managerial objectives - is equal to the objective of the organization as a
whole. To get this equation even
approximately right is one of the major problems in running any large
organization. Economists, when they have
recognized this problem at all, have assumed that it can be solved by an
appropriate organization structure and management control system. Such an assumption is not often justified; and
the relationship between departmental and company objectives deserves the
specific attention of economists.
It has been argued so far that the simple assumption
of profit as the objective should be replaced by a composite assumption
including economic, social, psychological, and organizational sets of
objectives. Some of these objectives
will become effective, not so much by influencing the choice between
alternative courses of action, but rather by suppressing one or more of these
alternatives altogether; they will not even be considered, because they
conflict with company policy. But how is
this company policy formulated? There is
no reason to believe that all firms have identical sets of objectives, or
indeed that any one firm uses the same set for all decisions. Is there anything one can say about the
formation of a firm’s preference system?
A possible answer seems to be that it can be discussed
in terms similar to those used in analysing the
preference system of the individual consumer. The objectives of the organization, like the
goods and services demanded by the individual consumer (which are his
objectives) are subject to substitution at the margin. Absolute priorities are perhaps as rare in the
one case as the other (though continuous marginal adjustment, as will be argued
later, is unlikely either). When
considering objectives other than profit, one can distinguish income and
substitution effects (research and development objectives gaining weight
relatively to sales as profits increase,
for example); and one can analyse the forces
which tend to produce particular patterns of objectives as one can analyse the forces which tend to produce particular
patterns of consumption. These patterns
may vary not only between firms but also from time to time within the
individual firm, as circumstances change. This is not an area in which the economist has
any special skill; but it is work which comes naturally to psychologists,
sociologists and organization theorists. Unfortunately, psychologists, sociologists and
organization theorists agree rather less often than economists, so that the
economist who wishes to use their ideas is faced with the problem of choosing
between them.
At this point it is necessary to consider a
fundamental objection to this type of argument. Is not the advocacy of more complex
assumptions about objectives open to the retort by E. A. G. Robinson to
Andrews’s theory which he misread as an account of simple full-cost pricing behaviour? If some
firms base their actions on objectives of this sort, while other firms go
simply for profits, then surely “the wicked nonconformists will triumph, and
the respectable ritualists will suffer”. [16]
It is possible, but it is not particularly likely. It is not particularly likely for four
reasons, the first two of which refer to friction within the system, the other
two to the working of the system itself. First, there are so many ‘respectable’ firms about
that the successes of the pure profit-seekers are unlikely to make much impact.
Second, the advance of the
profit-seekers is impeded by elements of monopoly and by the limited mobility
of resources in relation to the rate of change in the equilibrium position. Third, success in achieving these other
objectives, as has been stressed, may be explicitly regarded as a partial
substitute for profit. (Robinson’s
original argument against Andrews assumed that all firms were profit-seekers.) The fourth reason is that it is not at all
easy to be an efficient profit-seeker; as this reason also affects the
optimization assumption it must be considered separately.
The theory of the firm was originally developed on the
assumption of perfect knowledge. That
assumption has now been substantially modified, and its modification has
allowed the development of sophisticated theories of decision-taking under
conditions of uncertainty. But this
uncertainty relates simply to the future outcome of alternative courses of
action; and it is uncertainty of a probabilistic kind. But uncertainty extends much wider than this. Companies just do not have the information
which the traditional theory
16. E. A. G. Robinson, ‘The Pricing
of Manufactured Products’, in Economic Journal, December 1950, p. 774.
173
assumes
that they have. A shrewd economist
working in the intelligence department of a major company reports, for example,
that despite the efforts of a team of economists, “we know very little indeed
about the particular demand curve facing our particular firm. Similarly we know very little about our cost
curves”. [17] That firms should be unable to construct the marginal
functions required by customary forms of theory is not surprising; but if these
functions cannot be constructed, of what use are the simple decision rules for
profit maximization? Moreover, if a firm
does not know (except in an accounting sense, with its semi-arbitrary
allocations) what its costs are, how much less likely is it to know what they
ought to be? Yet it is this ‘ideal’ cost
curve which is assumed to be known in traditional theory - which thus cannot
recognize cost control as an economic problem. Optimization - or, more probably,
sub-optimization - may be possible in some situations; but it is difficult to
see how a firm can be a thoroughgoing conscious optimizer without a great deal
more information than it commonly possesses.
Profit maximization is not, therefore, usually an
operational objective. In practice, it
must be replaced by objectives that are operational, and these generally
possess one or both of two features: they are partial objectives, of the kind
listed above (often departmental objectives), or they are expressed in terms of
satisfactory rather than ideal performance. The first feature has already been discussed;
it is now time to discuss the second.
The argument that firms characteristically formulate
their objectives in terms of satisfactory levels of performance is associated
particularly with the work of Simon. [18] To many
economists this approach appears less intellectually satisfying than the
optimization assumption, for while there is only one optimum level there may be
many levels which might prove satisfactory to different people; and while the
optimum level is objectively determined, what determines the appropriate level
of satisfaction? [19]
This objection can be met in two different ways. Let us begin with the more conservative of
the two. If one retains for a moment the
usual assumptions about consistency, rationality, and knowledge, then the
situation facing a firm at a particular time will determine an ideal level of
achievement for each of its objectives, just as the situation facing a consumer
will determine an ideal level of expendi-
17. Private communication.
18. H. A. Simon, Administrative
Behavior, 2nd edn.
(New York, 1961), p. xxv.
19. Marris,
pp. 267-8.
ture on each desired commodity. But what happens if we withdraw these
assumptions? In order to remain
continuously at the highest possible level of satisfaction in the face of
continually changing circumstances, a consumer needs far more knowledge and far
more mental ability (for this seems to be a major component of the assumption
of ‘rationality’) than he can be expected to possess; and the same is true of
the multi-objective firm. Therefore, even if we postulate an original consistency of
objectives (which is most unlikely to be found in practice), this will soon be
eroded. It is because
optimization is unworkable that satisficing takes its
place. The practical alternative is
usually sub-optimization. There is no
reason to believe that this will in general produce better results, and it is
certainly likely to precipitate more conflicts within the organization.
A second ‘conservative’ argument in favour of using levels of satisfaction instead of ideal
standards is that decisions are not in practice costless, and that they should
not therefore be taken unless the cost of taking them is exceeded by the
increased satisfaction that results. Continuous
marginal adjustment, it has just been argued, is usually impossible; if it is
possible, it is probably too expensive. What
an organization (like the individual consumer) needs is a set of indicators
which will tell it when to review its present pattern of behaviour.
This set of indicators is provided by
the standards of performance established in all the important decision areas.
On the basis of these ‘conservative’ arguments it
appears that satisficing behaviour
(as it has been styled by Simon) can be justified as the best practical
approximation to optimal behaviour, in a world where
the assumptions underlying optimal behaviour are
invalid. Satisficing
can be rational. (This is not to argue
that every example of satisficing behaviour
found in practice can be thus justified.) But does this conclusion not mean that
management economics can continue to use the assumption of optimization, while
admitting that the frictions of imperfect knowledge, imperfect rationality, and
the cost of decisions will prevent the behaviour of
firms from more than approximating to the results so obtained? It does not; because these factors are not
mere frictions. They are important
elements in the managerial decision process.
The best answer to the objection of the optimizers is
far more radical. They complain that the
level of satisfaction cannot be determined by the accepted methodology of
economic theory, which requires that general analytic conclusions be derived by
logical argument from initial assumptions. But, as always, the wide extension of a
general conclusion has to be paid for by the shallowness of its intension; it
is possible to say something about all firms
175
only by
saying little about any particular firm. It is thus perhaps inevitable that any theory
which claims to be the theory of the firm should regard any particular firm as
a ‘black box’. But in looking at a
particular firm, to ask how the level of satisfaction is logically determined
is to ask the wrong question. Instead of
deriving equilibrium conditions by logical argument from initial assumptions,
the satisficing approach explains a specific,
possibly transient, state in terms of development by a dynamic process from
specified initial conditions. Differences in the levels that are judged
satisfactory by different firms - and in different countries - do indeed exist.
Optimization theory has difficulty in
explaining them logically; satisficing theory can
explain them as the product of an analysable process.
The weakness of traditional theory in dealing with
specific situations is shown by its inability to discuss efficiency except as a
function of scale or of location. This
weakness may bias recommendations for public policy. An economic evaluation of location policy,
while considering the relative advantages of different sites, and the possible
effects of dividing a firm’s organization, is liable to ignore the (possibly
more important) impact of a move in stimulating a review of the firm’s policy
and operating methods. [20] Proposals for improving industrial efficiency concentrate on
the size and degree of specialization of firms. The efficiency of management, instead of being
recognized as a major determinant of the firm’s performance, is treated as a
function of the scale of its operations. If all firms are assumed to be maximizing
their profits, what else can one do - except perhaps to make some unhelpful
remarks about managerial competence? This
is a concept which the traditional theory is unable to analyse,
but which satisficing theory can approach in terms of
the levels of performance which are accepted as satisfactory.
Thus, whatever its virtues as a major component of the
theory of value, the theory of the firm as it exists at present is not only an
inadequate basis for analysing the process of
managerial decision-making; it is also an inadequate basis for the analysis of
some major topics which economists still (rightly) regard as lying within their
province. It is outside the scope of
this article to do more than suggest how the alternative assumptions of
multiple objectives and standards of satisfaction can be used to develop an
appropriate analytical apparatus for tackling these problems. But since the problems are undeniably
important, perhaps even the suggestion of a solution is enough to stimulate
effort.
University of Bristol
20. This point is developed in my
article, ‘Making Location Policy Work’, Lloyds Bank Review, January
1967, pp. 41-2.
176