The Competitiveness of Nations in a Global Knowledge-Based Economy

Brian J. Loasby

Decision Premises and Economic Organisation

DRUID 1998 Summer Conference

Bornholm, June 9-11



The Premises of Theory

The Bounds of Rationality


Authority, Trust and Strategy




In The Mechanisms of Governance (1996), Oliver Williamson compares the decision premise, as proposed by Herbert Simon (1957, p. xii), with the transaction as rival units of analysis for developing the economics of organisation, and celebrates the victory of transaction cost economics as a theory of organisational forms.  No usable analytical structure, by contrast, has been built on decision premises.  In this paper I intend to examine the significance of decision premises, and to indicate how they may help us to understand economic behaviour.  I shall also suggest why economists have ignored Simon’s proposal.


The Premises of Theory

It is appropriate to begin by recognising that Williamson’s own argument for basing analysis on unitary transactions relies on the implicit proposition that the choice of decision premises for economic theories may be decisive for the development of knowledge, though as is usual among economists that proposition is never formally stated.  It is certainly true that the premise that economists should focus on transactions, and the costs of transactions, in order to explain the allocation of economic activities between firms and markets has had important consequences for the development of economic theory and for economists’ perception of its applicability.  This is just one example of the ways in which the development of economics has been shaped by the decision premises from which economists have worked.  If theories are logical structures, this should not be surprising; logical reasoning can only make explicit what is already contained within the premises on which it is based.  What one can get out depends on what one has put in.  Since no economist has any difficulty in pointing to mistakes in the basis of reasoning by some other economists, it ought not to be difficult for all economists to agree that the choice of premises for economic reasoning is an important topic for discussion.  That it is nevertheless difficult to get agreement on this conclusion indicates that there is something wrong with its premises.  What is wrong, I believe, is that most economists have very little idea how such a discussion could be organised, and, more fundamentally, that they assume, and are encouraged by their training to assume, that the basic questions have all been settled.  One consequence, as Leijonhufvud (1998) has shown, is that economists who are thoroughly trained to use modern decision premises have no possibility of understanding Keynes’ theory of unemployment.

It is consistent with the assumption that all the basic questions have been settled to construct theories which implicitly assume that the basic questions about the decision premises on which economic agents base their rational choices have also been settled.  Everyone knows how to decide what to do; and everyone decides in the same way.  In macroeconomic theory, that allows analysis to proceed by focussing on a single representative agent; in game theory and transaction cost economics it is necessary to have at least two agents, but they are two representative agents, and can be transposed without affecting the analysis in the slightest degree.  As Douglas (1995, p. 102) observes of Williamson’s work, “firms vary, but not individuals”.  That, fundamentally, is the reason for “the absence of surprise, victims, and the like” (Williamson, 1996, p.46); it is also the reason why economic theories, including transaction cost theories, so rarely mention entrepreneurship, and why (with some notable exceptions known to us all) the modelling of innovation is so inadequate.  The impression of the development of economics given in the textbooks (when any impression at all is given) is misleading for precisely the same reason.


One advantage of this commonality of issues, which I have tried to use on previous occasions, is that discussions of decision premises in the economy and in the economics profession can illuminate each other; and though the primary focus of this paper is the former I would like next to consider an earlier example of a choice of decision premises which has had widespread effects, including a major impact on economists’ understanding of the rationale of the firm.  The first half of Nero Sraffa’s celebrated article of 1926 on “The laws of return under competitive conditions” was based on a paper published in Italian in the preceding year, which is now available in an English translation (Sraffa 1998 [1925].  At the head of this 1925 paper stands a quotation from Marshall’s Principles (1920, p. 461): “The Statistical theory of equilibrium is only an introduction to economic studies; and it is barely even an introduction to the study of the progress and development of industries which show a tendency to increasing return”.  There is abundant textual evidence that Marshall believed the choice of premises for economic analysis to be of prime importance, and that he was anxious that the wrong choice should not be made; but Adam Smith might have warned him that his own proposal to begin with static equilibrium and then switch to evolutionary models for advanced work created a ‘gap’ that was likely to disconcert the imagination of economists too severely to be tolerated (Smith 1980).

Sraffa demonstrates in considerable detail the difficulty of reconciling increasing, and also decreasing, return with perfectly competitive equilibrium, and then without further argument ends the paper with this conclusion: “we must then concede that, in general, commodities are produced under conditions of constant costs” (Sraffa 1998 [1925], p. 363).  Perfectly competitive equilibrium is a fundamental and undiscussable decision premise, and therefore the concept of increasing return can have no place in economic theory.  Sraffa’s conclusion in 1926 was somewhat different: then it was perfect competition that had to be rejected, but static equilibrium remained untouchable.  Premises have to be chosen to match the competence of the analyst, as Hicks (1939, pp. 83-5) was characteristically open in admitting.  Theoretical possibilities are constrained by the perceived bounds of rationality.

Let us compare this choice of decision premise with that of Allyn Young in 1928.  As Currie (1997) and Ravix (1997) have recently reminded us, Young extended Smith’s theory of economic progress through the division of labour and Marshall’s theory of internal and external economies through internal and external organisation into a vision of economy-wide interactive development.  “New products are appearing, firms are assuming new tasks, and new industries are coming into being.  In short, change… is qualitative as well as quantitative.  No analysis of the forces making for economic equjilibrium… will serve to illumine this field, for movements away from equilibrium, departures from previous trends, are characteristic of it” (Young 1928, p. 528).  For Young, it was static equilibrium that was an unacceptable decision premise; he too did not argue the case, but assumed, in opposition to Sraffa, that premises should be factually correct rather than theoretically tractable.  (There is, I believe, no simple operational rule by which premises should be chosen; the choice of premises is a scientific art.)  Young declared that “[e]very important advance in the organisation of production… alters the conditions of industrial activity and initiates responses elsewhere in the industrial structure which in turn have a further unsettling effect” (Young 1928, p. 533); and he observed that the costs of this developmental process are “not the ‘costs which figure in an equilibrium of costs and advantages” (Young 1928, p. 535).  Some of them are clearly what Langlois (1992) has called ‘dynamic transaction costs’: the costs of changing both production methods and the organisational arrangements, within and between firms, in order to institute and then improve upon those changes.


Thus the challenge left by Marshall to his successors evoked orthogonal responses, with potentially orthogonal consequences.  From Sraffa’s 1925 paper we can derive the Chicago insistence both on using perfect competition and on ignoring the plausibility of assumptions, and from his 1926 paper we can derive imperfect competition theory, structure-conduct-performance models, the identification of widespread market failure and the hostility to increasing returns so memorably expressed by Samuelson (1972 [1967]).  From Young’s premises we can derive his own conclusion that the interactive process of increasing return is the source of economic progress, which cannot be adequately considered at the level of the firm alone - as Marshall’s treatment clearly indicated.  Thus Sraffa’s juxtaposition of increasing returns and the equilibrium of the perfectly competitive firm is the logical consequence of inappropriate premises, and the theory of the firm which arose from it the logical response to an incorrectly-formulated problem.

Penrose’s (1959, 1995) non-equilibrium theory of the firm as an agent of development can, however, be encompassed within Young’s premises, if it is set in the context of Richardson’s (1972) vision of the evolving organisation of industry.  We can also interpret Chamberlin’s (1933) theory of monopolistic competition as a dynamic analysis of specialisation, innovation, and market development, presented in a static guise (Robinson 1970).  I will just pose the unanswerable question of how Chamberlin might have developed his ideas had Young, who had been his thesis supervisor, not died and been succeeded by Lionel Robbins, whose own decision premises were quite different.  Ravix’s (1997) argument that Young’s emphasis on the growth of productive knowledge matched his conception of the most effective means of progress in economic science might bring this paper within Young’s research programme.  That it is within Adam Smith’s cognitive, sociological, and economic research programme (Loasby 1996) I am certainly prepared to claim.


The Bounds of Rationality

If the premises from which economists work have significant effects on the development of economics, is it not likely that the premises from which economic agents work will have significant effects on the choices that they make?  There is plenty of evidence of such effects, some of which I have cited myself, and much more has been presented by, for example, Martin Fransman, Neil Kay and Dick Langlois.  But here we should notice how “thinking has been shaped by terminology” (Richardson 1960, p. 69), or perhaps fundamentally by the premises to which that terminology has to conform.  A notable feature of Sraffa’s papers of 1925 and 1926, to which Denis O’Brien (1984) has drawn attention, is the assimilation of Marshall’s ‘free competition’ to perfect competition, in which competitive activity is constrained out of existence.  As Shackle (1961, p. 272) observed, the agents in economic models do not make choices but act according to necessity.  Why, then, should anyone be concerned about the decision premises on which they act?

There is one reason which should concern economists whose own methodological decision premises include theoretical coherence.  Within such a tightly constrained model it is impossible to find any explanation for firms, or markets, or money.  All that we observe is a complete set of interlocking individual routines.  Just as Hicks (1982, p. 7) discovered sixty-five years ago that we must introduce uncertainty as a premise before we can consistently introduce money, so Knight in 1921 and Coase in 1937 recognised that uncertainty was a logical prerequisite for a theory of the firm (though Coase disagreed with Knight’s specific reasoning).  As Oliver Williamson has repeatedly emphasised, transaction cost theory cannot get started without some restrictions on


rationality.  The difference between Williamson and Coase, to which we will return, is that Williamson believes that what is decisive is the threat to incentive-compatibility which is posed by particular conjunctures of bounded rationality, whereas Coase believes that these threats do not provide any systematic basis for discriminating between market transactions and internal governance.

Simon’s conception of bounded rationality may perhaps be understood as truncated rationality, in which the set of decision premises has to be abbreviated and simplified in order to be amenable to the limited human capacity for logical processing.  If the processing is strictly logical but limited, it is an obvious inference that decisions are to be explained by the way in which the premises are truncated; hence Simon’s proposal to take the decision premise as the unit of analysis.  In the background, as indicated by Simon’s frequent use of chess as the archetypical example, there is usually the shadow of a fully specified problem that would be accessible to anyone with sufficient cognitive capacity.  A similar shadow lies in the background of the analysis of ‘information’ in economics, including transaction cost economics; a full information set always exists but is costly to acquire and may be inaccessible to some agents.  That incomplete information, unlike bounded rationality, does not preclude formal optimisation is the consequence of some incoherence in the analytical premises, as has often been pointed out.  The truncated rationality which substitutes rational expectations for the unattainable stability theorems of intertemporal general equilibrium, makes co-ordination failure undiscussable (Leijonhufvud 1998, pp 18 1-2).  That is not an issue for this paper; it may be dismissed with Knight’s (1933, p. xiv) observation that he was “puzzled at the insistence of many writers on treating the uncertainty of result in choice as if it were a gamble on a known mathematical chance”.

My own view is that a conception of truncated rationality as an approximation, good or bad, to a fully-specified problem greatly understates the difficulties that we face through the intersection of complexity, interdependence, and cognitive limitations; rational choice is impossible, not because of difficulties with formal reasoning but because there is no way of knowing that the premises for such reasoning are correctly specified.  Our logical processing must therefore be based on a problem formulation which cannot replicate the phenomena for which we substitute it.  The normal situation is not one in which we have to truncate a known full set of premises (such as the rules of chess) but one in which we have to impose a set of premises that we believe, or hope, will be adequate.  “It is clear that to live intelligently in our world… we must use the principle that things similar in some respects will behave similarly in certain other respects even when they are very different in still other respects” (Knight 1921, p. 206).  It is obviously important that the similarities should be important and the differences irrelevant; but whether this is so for any decision must depend, in ways which can never be known for certain, on the particular circumstances of that decision.  But since logical reasoning cannot do more than reveal what is already contained within the premises -though that can sometimes be a great deal - it must follow that we determine, consciously or unconsciously (and usually the latter), the conclusions that we reach when we select our premises.

Our conclusions may therefore be in error because the selection of premises is faulty; and this selection must be made by non-logical means.  “[T]he identification of ideal with empirical statements is not deductive.  Having neglected the uncertainty in our premises, we can never be sure of the logical necessity of our conclusions.  Every theoretical calculation becomes metaphorical” (Ziman 1978, p. 27).  This is true of economic theory, as of all sciences, and of decisions in business and in our daily lives.  The choice of premises for scientific analysis sets limits to our knowledge;


but without such constraints no knowledge is possible.  Similarly, the choice of decision premises by economic agents sets limits to their actions, but without such limits no reasoned action is possible.  Moreover, knowledge may be false, and actions may be mistaken.  The recognition of these limitations has logical implications for the conduct of scientific research and of economic activities.

At this point I should make clear three drastic restrictions on the premises for this paper.  First, I shall be concerned only with decisions that are taken for what are considered to be good reasons, though certainly not only with decisions that turn out well.  This excludes many decisions, which need to be understood for any comprehensive explanation of the working of an economy - or the working of the economics profession; but ‘rationality’ is such a core concept in economics that it is appropriate to focus here on reasoned decisions.  Like Williamson, my presumption in what follows is that agents are seeking efficiency; but it is obvious that they do not always achieve it.  Like Hayek, I believe that we need to understand success before we can explain failure; but I also believe that what we need to model are processes which are inherently fallible, and not simply because of an error term.  Human decision processes are fallible, and it is the need to impose premises that makes them so.  Second, we should note that routine behaviour is also governed by the equivalent of decision premises, in the form of a classification system to which phenomena are assigned and a set of actions to which categories are matched (as is illustrated by classifier models).  An investigation of these premises would therefore be entirely appropriate for an examination of the importance of decision processes, but in order to focus on rationality it will not be attempted here.

The third restriction is that, in conformity with the concept of truncated rationality, I shall assume the faultless application of logic (which is rarely observed in practice) and concentrate on the variety of ways in which problems may be truncated, or framed.  Because no framework will be best for all purposes there is always a potential for failure through applications which turn out to be inappropriate; the ‘logic of appropriateness’, correctly applied, ensures internal but not external coherence.  But in addition to examining specific situations, we should recognise that an efficient system response to bounded rationality, and specifically to the insufficiency of any single set of premises, is offered by the division of labour; the differentiation of knowledge sets may be the source of substantial gains from specialisation.  (Smith’s, Marshall’s and Young’s arguments were much wider than this, but we will keep to the narrower line, which is represented by Knight.)  The division of labour is, in part, a means of economising on rationality through the specialisation of decision premises, and thereby increasing its productivity.  But one consequence is that knowledge becomes increasingly dispersed and asymmetric.  Not only will different specialists know different things; they will formulate their knowledge in different ways, which may make communication difficult, and even sometimes dangerous because the receiver’s interpretation may be inconsistent with the sender’s intention.  Information has no significance without context.  Co-ordination may therefore be problematic even in a world of universal benevolence.

We can now identify four themes for analysis and empirical study.  The first is the distribution of different clusters of premises within the economy, or the pattern of specialisation in decision making.  (This we might call Knight’s theme.)  The second is the relationship among those within a particular specialism: how are they grouped and what contacts do they have?  (The obvious reference here is to Marshall’s Darwinian emphasis on variation within a trade, and the process of critical examination and amendment within industrial districts.)  The third is the pattern of premises currently favoured in that specialism, and the implications for what those who work within it can and cannot do, and for what they can and cannot perceive: compare Sraffa’s premises with those of


Young.  (This theme is represented in this Conference by Martin Fransman and Dick Langlois, among others.)  The fourth is the relationship between different categories of specialists.  (We think naturally of George Richardson.)  Here are the basic issues of economic organisation.  It is my view that a fundamental premise for analysing them is the context of the growth of knowledge; reasoning in terms of fixed information sets is as likely to lead to misleading conclusions as reasoning in terms of given products and technologies in the 1930s.  Opportunism, on the other hand, although significant does not seem to me fundamental to an explanation of industrial organisation.  It needs to be incorporated at an early stage, but the problems and opportunities of developing and co-ordinating knowledge deserve priority.



An exploration of these themes is far too large a task for the remainder of this paper.  I will strictly limit my present agenda to indicating some of the implications of bounded rationality for the relationships between individuals and the groups of which they are members or with which they interact.  It is convenient to begin with the problems faced by each individual in trying to behave reasonably.  How is a situation to be framed in order that we may select an action; how are we to close our model in order to put logic to work?  Our cognitive powers are so limited in relation to the complexity around us that we very often feel in need of some assistance in filling the gaps.  If we always had to give conscious thought to our decision premises we would make very few decisions.  We therefore often look around for help, hoping to follow the example of someone who may be more skilful in coping with a particular kind of situation.  It is in this search for good practice that Choi (1993) finds the origins of institutions, which we use as ready-made decision premises.  Institutions help us to close our decision models and thus permit logical deduction within a plausible framework.  Furthermore, shared conventions help us to understand the behaviour of those who share them, and create the possibility of tacit co-ordination.  What frameworks are plausible may vary between groups and over time; and institutions may accordingly be local or general, and may change.

Any society requires some institutions of very broad scope; more specialised groupings, such as the community of economists, develop more specialised premises which both assist the individual and facilitate co-ordination within particular areas of interest.  All effective human interaction depends on intersubjectivity - for example, the shared attribution of ‘moneyness’ to certain pieces of paper, the shared interpretation of marks on other pieces of paper as the tracks of subatomic particles, the shared belief that replicator dynamics provides insight into biological processes or the fortunes of a group of firms, the shared response that a report from a firm’s research department defines a productive opportunity.  The division of cognitive labour focusses the attention of each category of specialist on particular phenomena, on particular ways of classifying them, and on particular ways of formulating problems in order to make them accessible to particular logical techniques and the experimental techniques which are associated with them it allows those within each category to learn from each other, even if they are in competition.

A firm is such a specialised grouping, which must draw on the institutions of a wider society in order to get started but then develops particular institutions on which individuals come to rely for simplifying their own decision making and as a basis of co-ordination within the organisation.  If we begin, like Coase, with a theory of market co-ordination, but with a more detailed appreciation than Coase had acquired in the early 1930s of the reliance of rigorous theory on prices that are both


costlessly known and known to be equilibrium prices, then we can regard the firm as a substitute for a set of contingent claims markets (Loasby 1976, p. 65) and as an arena for equilibrating processes; but if we also take account of Young’s vision of economic progress, we can also regard each firm as part of the mechanism which propels economic systems away from previously-existing equilibria.  They are local centres of specialision in coping with particular kinds of uncertainty.  This is Knight’s (1921, p. 255) explanation of the firm as an entrepreneurial creation, which Casson (1982, p. 373) has acknowledged as the basis of much of his theory of entrepreneurship.  “Men differ in their capacity to form correct judgements as to the future course of events in the environment. This capacity, furthermore, is far from homogeneous, some persons excelling in foresight in one kind of problem situations, others in other kinds, in almost endless variety” (Knight 1921, p. 241).  Casson (1982, p. 23) defines an entrepreneur as “someone who specializes in taking judgmental decisions about the coordination of scarce resources”, and begins his analysis with the absorption of uncertainty; but, recognising the contributions of Schumpeter and Penrose, he then develops the role of the entrepreneur as innovator.

“With uncertainty present,… the primary problem or function is deciding what to do and how to do it” (Knight 1921, p. 268), and if powers of deduction are assumed to be infallible but limited, the quality of decisions depends on the quality of the premises.  Knight notes that this quality improves with experience of particular kinds of situations, and that this improvement takes time (Knight 1921, p. 243), thus providing a basic reason why the boundaries of a firm should be fairly stable and why performance within those boundaries may get better; the possibility of such improvement was crucial to Smith’s and Marshall’s theories of economic development.  Casson’s emphasis on the importance of each entrepreneur’s distinctive way of organising information as a basis for distinctive rational decisions, which has been a continuing feature of his work (Casson 1997), is also an emphasis on the importance of the scope of each business and the accumulation of experience, and therefore implicitly on the premises for decision making.

Search is organised on the basis of locally-corroborated conjectures, and the costs of decision making are reduced by postponing decisions and then simplifying the postponed decisions by a commitment to particular relationships, and thus to particular decision premises.  Though Coase (1988 [1937]) explains that the creation of a firm leads to the avoidance of subsequent market transactions, he does not go beyond the foundational contract by which a firm is constituted to discuss the costs of creating an organisation, and does not refer to Marshall’s (1920, p. 377) emphasis on the capital which any successful business must invest in building up its internal and external organisation.  He therefore never quite states that the costs of a series of transactions, or a series of acts of governance, may be reduced by appropriate investment, as Casson (1982) does in explaining why firms make markets.  Both Coase and Casson, however, are in effect writing about the organisation of knowledge, which provides credible premises for later decisions.

In his second article on the firm, Coase (1988 [1972], p. 63) recognised that “the costs of organizing an activity within any given firm depend on what other activities the firm is engaged in”; that is presumably why “General Motors [is] not a dominant factor in the coal industry or why A & P [does] not manufacture airplanes” (Coase 1988 [1972], p. 65).  It should be noted that Coase’s interest in the scope of a firm is not focussed on vertical integration.  However he does not say why the costs of organising activities should not be additive.  The explanation lies in the possibility of reusing knowledge - in the present context in the possibility of applying the same set of decision premises over a wider range or a longer series of decisions (Loasby 1976, pp. 72-3).  A firm provides


an agreed framework of decision premises which both facilitates decision making and generates a pool of experience which may be used to simplify these premises and make them more precise (Williamson 1967, p. 136), thus making additional managerial services available, as Penrose (1959, 1995) has explained.

Penrose has shown how the emergence of such additional managerial services may lead to the extension of the scope of a firm’s activities, which in the present context may be expressed as the extension of well-established decision premises to a new line of business.  Contrary to a common assumption in economic theory, it is never possible to reuse knowledge without cost; but the costs of reuse within a well-functioning organisation can often be much lower than the costs of reuse outside.  Thus, as Langlois (1992) has pointed out, extensions of scope may be motivated less by the desire to protect rents than by the high costs of transferring knowledge to other organisations which are unfamiliar with the appropriate decision premises.  A firm may also have advantages as an instrument of local co-ordination because it can internalise important externalities through the specification of the agenda and the criteria for decisions (Loasby 1976, pp. 76-8).

The organisation of resources within an administrative framework (Penrose 1959, 1995, p. 149) is in part the organisation of compatible decision premises for the effective use and further development of the knowledge which is distributed within the firm.  Not all firms succeed in establishing such compatability; and if they do they face two insidious problems: first, they may fail to make adjustments as their environment changes, and second, the price of compatability may be the suppression of enterprise within the firm.  The danger of suppressing variety in the process of maintaining internal coherence may however be mitigated by the variation between the firms within an industry in the minor premises which they use to interpret their experience.  This variation not only promotes the development of the economy; firms may be able to benefit from the activities of their rivals.  The importance of interfirm variety used to be well recognised within the chemical industry, in which access to knowledge created by rivals through cross-licensing, and the role of research departments as importers of technology were major premises of strategy; but it is often overlooked in the pursuit of global scale.

The suppression of necessary variety may be particularly serious if a firm seeks to internalise a cluster of activities which though complementary are sharply dissimilar (Richardson 1972).  The difficulties of matching disparate ways of thinking while maintaining the match between each way and its own particular field of application may be very great; and the concentration of the power to decide in order to prevent the opportunism that specificity of knowledge appears to make possible is liable to impede both the present use of that specific knowledge and its further development.  Interfirm collaboration, of varying degrees of formality, is often a sensible response to the dilemma of organisational versus activity-focussed compatability, because this minimises the interface between very different ways of structuring problems and facilitates the development of very specialised skills to cope with these difficulties, with the aim not of reducing the costs of this class of transactions but of maximising their net value (Kay 1997).

All organisational design is conjectural (Egidi 1992).  In order to make problems manageable it relies on a combination of opposing principles of decomposition and aggregation: issues which are believed to be sufficiently similar are bundled together (remember Knight’s necessary but dangerous principle for living intelligently), and separated from other issues by the definition of each manager’s responsibility, which is at once a grant of independence and a prohibition of


tresspass.  Problems and perceptions are framed in a way that, it is hoped, will keep down the costs of decision without significantly damaging their quality.  The organisational definition of decision spaces creates the local environments within which the trial and error learning takes place that drives economic progress according to Smith’s basic principle.  But this learning is itself subject to error: the inescapable difficulties of identifying causality in any complex system (the Duhem-Quine problem) are aggravated by the falsehoods that are necessarily incorporated in any organisational design (Levinthal 1996) and which are reinforced by the natural human tendency to take credit for success and assign the blame for failure elsewhere.  This propensity to error strengthens the argument for variety across firms, and also supplies a cognitive underpinning to Young’s (1928) association of increasing returns with a changing organisation of industry.


Authority, Trust and Strategy

As Marshall emphasised, the importance of firms in economic development lies in their organisation of knowledge.  (That this organisation might become seriously defective was a problem that can readily be discerned in Marshall’s (1919) Industry and Trade).  In the process they are sometimes - though certainly not always - more effective than market relationships in “transforming a conflict system into a co-operative system” (Levitt and March 1995, p. 12); but this is not to be explained solely, or even mainly, by their power to curb opportunism.  Indeed, what is striking to anyone who observes human behaviour is how many occasions for opportunism are not taken; and this no doubt requires explanation.  Adam Smith, we should not forget, had a theory of moral sentiments which is still applicable to small and relatively stable communities, which may include organisational groupings; but self-interest may often be sufficient to explain why we so often accept other people’s premises for our actions . It is so much simpler than constructing our own; indeed, as has already been pointed out, we do not have the capacity to supply more than a small proportion of our own premises for making sense of the situations that we encounter or for deciding what to do about them.

The more complex our economy and our society become the more dependent we are on premises which are supplied by other people.  It is often a relief rather than a cause for complaint to accept the decision premises of an organisation rather than provide our own; many people prefer employment to independence.  When we recognise the cognitive problems of constructing a coherent preference ordering it is not even difficult to understand how we may come to internalise organisational preferences, notably in the form of corporate culture, which for good or ill inhibits challenge to the current institutions, and may even make those institutions undiscussable (Argyris 1994).  Many economists, it may be observed, seem to be very comfortable with the decision premises in which they have been trained, and some find it hard to perceive what there might be to discuss about them.

Chester Barnard (1938) emphasised that it was the recipient of any communication who decided whether it was authoritative; and all of us are very happy to find sources of communication to which we can grant authority, even when we could apparently do better for ourselves by ignoring them (Reynaud 1996).  Firms provide an environment in which members may develop good reasons for deciding whose authority they can rely on as premises for their own decisions - or as we might say, who can be trusted - for no organisation can work effectively unless each of its members is prepared to accept the word of many others, most of whom are not their formal superiors.  As Kay


(1997, p. 215) has pointed out, it is very often the case that people agree to work together not because they trust each other but in order to discover whether they can reasonably trust each other, which of course cannot be done without exposing oneself to opportunism.  But if knowledge is specialised, trust is indispensable.  What is missing in the analysis of organisational equilibria which is based on transaction costs is the effect of process on attitudes and arrangements; and it is missing because the premises of the analysis exclude it.

A firm is an interpretative system, within which people develop common ways of understanding, and common ways of responding to what they believe they have understood.  Especially in a new firm, there may be a strong entrepreneurial element in this interpretation; as Martin Fransman (1995) has emphasised, a firm may be the location for the construction of beliefs by which interpretative ambiguity is resolved.  Ambiguity, of course, is also the arena for opportunism, and it is not unknown for people to construct beliefs in which they have no faith but which they use to deceive others; but the desire to impose order is so insistent that almost any pattern which is not incompatible with our presently serviceable set of connecting principles may seem preferable to fragmented knowledge, and crowd out opportunism.  Often it crowds out the entrepreneur’s own opportunism; Schumpeter was probably right in his implicit assumption that entrepreneurs believed their own promises.

In models of efficient and inefficient allocation, strategy can hardly be other than an efficient set of paths through a decision tree or a set of actions by which to gain some monopolistic advantage.  But business strategy, like military strategy, belongs in a world of uncertainty; it is a means of resolving ambiguity into a coherent set of major premises which promotes the compatability of decisions over time and across decision makers.  Soon after he had joined the research departmnent of a Division of ICI, my former colleague Frank Bradbury was writing down the formulae for some chemical compounds that he thought might have some commercial application when his manager walked past; after scanning the list he simply asked “Where’s the chlorine, Frank?”  At that time the Division produced chlorine in large quantities as a by-product, and its researchers had considerable expertise in chlorine chemistry; it was therefore natural to impose the use of this by-product and these skills as decision premises for research.  In a similar fashion, at a time when BP had an unmatched reputation for discovering productive oilfields but not for marketing the products, the Research Director of BP Chemicals told his research staff “The objective of this Department is to shift crude”.  Another aspect of this overall strategic emphasis was the readiness to divert an oil tanker already approaching the Grangemouth refinery near Edinburgh to Rotterdam in response to a slight rise in the spot market price.

Strategic decision premises provide structure to a complex set of decisions which might otherwise be incoherent.  But no choice is without opportunity costs, and the pathology of the focus which is provided by strategy is bias.  Both technological foresight and technological oversight may often be explained by the decision premises on which crucial decisions were based.  For example, neither the top management of major corporations nor a technologically-based firm of consultants could see any reason why anyone should buy a photocopier to do what was already accomplished by a typist using carbon paper (Brown 1997, pp. 97-8); the contract by which IBM bought DOS from Microsoft, and which now appears to have been a disastrous blunder by the former and a brilliant coup by the latter, was carefully tailored to the localised decision premises of both parties, neither of whom was blessed with rational expectations (Porac 1997); and the precisely-calculated performance limits of optical lithography rested on well-defined physical, technical and production


constraints which proved not to be binding (Henderson 1995).  On the other hand, a novel set of decision premises may lead an entrepreneur to deduce the possibility of a major innovation.  It is not necessary for any single premise to be new; as Schumpeter (1934) claimed, what matters is the combination, though what he did not emphasise is that this combination may follow a good deal of development in what are to become the components of the new vision.



Decision premises are required for decision processes.  Theories of equilibrium states, even when based on the concept of rational choice, have no place for decision processes, for in equilibrium there are no decisions to be made.  The best theorists have recognised that behaviour out of equilibrium cannot be analysed with the standard and type of rigour that is now expected, and so they choose theoretical premises which exclude it.  One of the attractions of game theory is that there are no out-of-equilibrium moves; and one of its limitations is that there is no credible method of dealing with multiple solutions or with inconsistency within the logic of backward induction.  Transaction cost economics, as a theory of efficient allocations, is essentially an equilibrium theory, and also avoids behaviour out of equilibrium; thus the transaction is indeed the appropriate unit of analysis and the decision premise is not.

If economics is to be defined as the study of fully-defined allocation problems, in which it increasingly appears that the key requirement is incentive compatability based on rational expectations, then that is the end of the matter - except perhaps to tidy up a few internal inconsistencies by deleting such concepts as money and the firm as an organisation.  But modern economics began with a theory of economic development, propounded by a man who was deeply concerned with the possibilities of human understanding, human communication, sensible action and the possibilities of self-deception.  Within that tradition it seems important to pay attention to the formation and development of the premises on which people reason and the knowledge and skills which they may acquire as a consequence of economic organisation.

However, although I hope to have shown that thinking about decision premises can be enlightening, I do not think that this should be the primary unit of analysis for the study of economic organisation.  That place I believe should be filled by the concept of the activity, linked with the concept of capability as it has been by Penrose and Richardson.  Activities and capabilities are highly problematic, very diverse, and may be combined and developed in many ways; therefore they merit a great deal of attention, in which the causal links with organisation should be bi-directional.  Activities are linked by interfaces, which are justifiably the special concern of transaction cost economics.  However, I agree with Coase, and with Demsetz (1997), that problems of knowledge which do not qualify as agency problems provide important theoretical premises for explaining the choice of transaction mode.  I would go one stage further, and argue that the management of a class of transactions, or of a particular governance relationship, is itself an activity, and the effectiveness of that management depends on the evolved capability of the person, or persons, who are doing the managing.  Since management requires decisions, its quality depends upon the managers’ decision premises.  Therefore the analysis of transaction costs and of decision premises should be incorporated within the analysis of activities and capabilities, which is fundamentally an analysis of processes and institutions.



Argyris, C. (1994) ‘Good Communication that Blocks Learning’, Harvard Business Review, July-Aug, 77-85.

Barnard, C. 1(1938) The Functions of the Executive, Cambridge, MA: Harvard University Press.

Brown, J. 5. (1997) ‘Changing the Game of Corporate Research: Learning to Thrive in the Fog of Reality’, in Garud et al. (1997), 95-110.

Casson, M. (1982) The Entrepreneur: An Economic Theory, Oxford: Martin Robertson.

Casson, M. (1997) Information and Organization, Oxford: Clarendon Press.

Chamberlin, E. H. (1933) The Theory of Monopolistic Competition, Cambridge, MA: MA: Harvard University Press.

Choi, Y. B. (1993) Paradigms and Conventions: Uncertainty, Decision Making, and Entrepreneurship, Ann Arbor: University of Michigan Press.

Coase, R. H. (1937) ‘The Nature of the Firm’, Economica, N. 5. 4, 386-405; reprinted in Coase (1988), pp. 33-55. References are to this reprint.

Coase, R. H. (1972) ‘Industrial Organization: A Proposal for Research’, in V. R. Fuchs (ed.), Policy Issues and Research Opportunities in Industrial Organization, New York: National Bureau of Economic Research, 59-73; reprinted in Coase (1988), 57-74. References are to this reprint.

Currie, L. (1997)’Implications of an endogenous theory of growth in Allyn Young’s Macroeconomic Concept of Increasing Returns’, ed. R. J. Sandilands, History of Political Economy, 29, 4 13-43.

Demsetz, H. (1997) ‘The Firm in Economic Theory: a Quiet Revolution’, American Economic Review, 87,426-9.

Douglas, M. (1995) ‘Converging on Autonomy: Anthropology and Institutional Economics’, in Williamson (1995), 98-115.

Egidi, M. (1992) ‘Organizational leaming and the Division of Labour’, In H. Simon, M Egidi, R. Marris, and R.Viale, Economics, Bounded Rationality and the Cognitive Revolution, Aldershot: Edward Elgar, 148-73.

Fransman, M. (1995) Japan’s Computer and Communications Industry, Cambridge: Cambridge University Press.

Garud, R., Nayyar, P. R., and Shapira, Z. B. (eds) Technological Innovation: Oversights and Foresights, Cambridge: Cambridge University Press.

Henderson, R. (1997) ‘On the Dynamics of Forecasting in Technologically Complex Environments: the Unexpectedly Long Life of Optical Lithography’, in Garud et al. (1997), 147-66.

Hicks, J. R. (1939) Value and Capital, Oxford: Oxford University Press.

Hicks, J. R. (1982) Collected Essays on Economic Theory, Volume II: Money, Interest and Wages, Oxford: Oxford University Press.

Kay, N. M. (1997) Pattern in Corporate Evolution, Oxford: Oxford University Press.

Knight, F. H. (1921) Risk, Uncertainty and Profit, Boston: Houghton Mifflin. Reprinted Chicago: University of Chicago Press, 1971.

Knight, F. H. (1933) ‘Preface to the Re-issue’, Risk, Uncertainty and Profit, Reprints of Scarce Tracts in Economics and Political Science No. 16, London: London School of Economics, xi-xxxvi.


Langlois, R. N. (1992) ‘Transaction-Cost Economics in Real Time’, Industrial and Corporate Change, 1, 99-127.

Leijonhufvud, A. (1998) ‘Mr Keynes and the Moderns’, European Journal of the History of Economic Thought, 5, 169-88.

Levinthal, D. A. (1996) ‘Organizations and Capabilities: the Role of Decompositions and Units of Selection’, presented to Workshop at the University of Trento, 17-19 June.

Levitt, B. and March, J. G. (1995) ‘Chester I. Barnard and the Intelligence of Learning, in Williamson (1995), 11-37.

Loasby, B. J. (1976) Choice, Complexity and Ignorance, Cambridge: Cambridge University Press.

Loasby, B. J. (1996) ‘The Division of Labour’, History of Economic Ideas, 4, 299-323.

Marshall, A. (1919) Industry and Trade, London: Macmillan.

Marshall, A. (1920) Principles of Economics, 8th edn, London: Macmillan.

O’Brien, D. P. (1984) ‘The Evolution of the Theory of the Firm’, in F. H. Stephen (ed.), Firms, Organization and Labour, London: Macmillan.

Penrose, E. T. (1959, 1995) The Theory of the Growth of the Firm, Oxford: Basil Blackwell (1959), Oxford University Press (1995).

Porac, J. F. (1997) ‘Local Rationality, Global Blunders, and the Boundaries of Technological Choice’, in Garud et al. (1997, 129-46.

Ravix, J. L. (1997) ‘Fondements Critiques pour une Analyse de la Dynamique Industrielle: la Méthode de Allyn Young’, Revue Economique, 48, 965-88.

Reynaud, B. (1996) ‘Types of Rules, Interpretation and Collective Dynamics: Reflections on the Introduction of a Salary Rule in a Maintenance Workshop’, Industrial and Corporate Change, 5,699-721.

Richardson, G. B. (1960) Information and Investment, Oxford: Clarendon Press. 2nd edition 1990. Richardson, G. B. (1972) ‘The Organisation of Industry’, Economic Journal, 82, 883-96; reprinted in Richardson (1990), 224-42.

Robinson, R. (1970) Edward H. Chamberlin, New York: Columbia University Press.

Samuelson, P. A. (1972 [1967]) ‘The Monopolistic Competition Revolution’, in R. E. Kuenne (ed.), Monopolistic Competition Theory: Studies in Impact, New York: Wiley, 105-38; reprinted in R. K. Merton (ed.), The Collected Scientific Papers of Paul A. Samuelson, Cambridge, MA and London: MIT Press, 18-51.

Schumpeter, J. A. (1934) The Theory of Economic Development, Cambridge, MA: Harvard University Press.

Shackle, G. L. 5. (1961) Decision, Order and Time, Cambridge: Cambridge University Press.

Simon, H. A. (1957) Administrative Behavior, 2nd edn., New York: Free Press.

Smith, A. (1980) ‘The Principles Which Lead and Direct Philosophical Enquiries: Illustrated by the History of Astronomy’, in Essays on Philosophical Subjects, ed. W. P. D. Wightman, Oxford: Oxford University Press, 33-105.

Sraffa, P. (1926) ‘The Laws of Returns under Competitive Conditions’, Economic Journal, 36, 535-


Sraffa, P. (1998 [1925]) ‘On the Relations Between Cost and Quantity Produced’, trans. J. Eatwell and A. Roncaglia, Italian Economic Papers, 3, 323-63. First published 1925 in Annali di Economia, 2, 277-328.

Williamson, O. E. (1967) ‘Hierarchical Control and Optimum Firm Size’, Journal of Political Economy, 75, 123-38.

Williamson, O. E. (ed.) (1995) Organization Theory: From Chester Barnard to the Present and Beyond, expanded edition, New York and Oxford: Oxford University Press.


Williamson, O. E. (1996) The Mechanisms of Governance, New York and Oxford: Oxford University Press.

Young, A. (1928) ‘Increasing Returns and Economic Progress’, Economic Journal, 38, 527-42.

Ziman, J. M. (1978) Reliable Knowledge, Cambridge: Cambridge University Press.