Entrepreneurship and the Market Process

ness will here be regarded, therefore, not only as a property possessed to a greater or lesser degree by all (conscious or unconscious) mental processes, but as the basis of man’s capacity to move successfully in a world very imperfectly known to him – an adaptation to his ignorance of most of the particular facts of his surroundings. (Hayek 1973:30; emphasis added) Entrepreneurs must formulate theories in order to create and to discover potential market opportunities for improved coordination. Prior theoretical frameworks are required for the selection and interpretation of market data. Abstract relations – which are equivalent to theories – ‘enable us to discriminate between different things and to respond to them differentially’ (Hayek 1978:37). Entrepreneurs are only able to recognise something as new and different (e.g. changes in consumer tastes, technology or resources) if they are equipped with expectations of what is ordinary. In addition, their perceptions of similarity in two or more market events presuppose the adoption of a point of view, that is, the adoption of a theory (Popper 1959:420–421). If entrepreneurs did not have any system of theories, market events would seem to them to be of such an indistinguishable homogeneity that they would not be able to make sense of what is happening in the market. (Of course, this is itself a theoretical proposition.) Entrepreneurial action would be purposeless. If entrepreneurs are to have any possibility of understanding the world and of effecting changes within it, they will be forced to abstract and to simplify. Entrepreneurs must select those aspects of a situation which are related to their specific purposes and to the particular problems in which they are interested. Their theories and models assist them in focusing their attention on specific segments of the dense complexity which characterises their concrete existence. ‘We never act, and could never act, in full consideration of all the facts of a particular situation, but always by singling out as relevant only some aspects of it’ (Hayek 1973:30). For example, an entrepreneur marketing computer software does not typically consider the colour © 1996 David A. Harper of customers’ hair in the development of marketing strategy, although hair colour is indeed one of countless variables to which the entrepreneur could conceivably attend in this empirical situation. One potential reason for such selectivity is that the entrepreneur assumes that the colour of customers’ hair is not relevant to his or her problem: it is hypothesised that this factor has no effect or no detectable effect upon the phenomenon of interest. More specifically, the entrepreneur implicitly makes the negligibility assumption that the colour of customers’ hair does not affect the success of policies to market computer software (cf. Musgrave 1981). However, for a completely different problem, such as the marketing of cosmetics, spectacles, clothing, or jewellery etc., hair colour may be considered especially pertinent and worthy of the entrepreneur’s attention. 5.2 GENERAL CHARACTERISTICS OF ENTREPRENEURS’ THEORETICAL SYSTEMS The theoretical systems that entrepreneurs construct (and the entrepreneurial action programmes they derive from them) possess several important properties: • a hierarchical structure and near-decomposability; • openness to change, revision and replacement (i.e. theories are not static); • limitations of scope; • a conjectural (non-verifiable) and potentially objective character; • individuality (i.e. theories are specific to individual entrepreneurs, of whose imaginations they are a product). The following discussion will concentrate upon the first of these – the hierarchical structure of the entrepreneur’s system of knowledge – because it contributes most to explaining aspects of entrepreneurial behaviour, such as why an entrepreneur may act as if his or her conjectures in one domain are not interlinked with those in another. The entrepreneur’s theoretical system, like that of the scientist, is a hierarchically organised structure. Among other things, an entrepreneur’s theoretical system comprises individual theories at different epistemological levels of abstraction: theories about interactions at one level (e.g. within a particular market or industry) and theories at higher levels (e.g. a set of markets) or lower levels (e.g. specific market segments). Those of the entrepreneur’s theories © 1996 David A. Harper on higher levels of abstraction describe deeper layers of reality, even though the states of affairs these theories describe are more hypothetical or conjectural (cf. Popper 1963:115–116). Thus, the more general the entrepreneur’s empirical theories about the world, the greater the depth in the entrepreneur’s claims about reality (see section 6.3.1 on the falsifiability principle). Each entrepreneurial theory is in turn a hierarchical system of deductively linked hypotheses which incorporate different degrees of generality (though by no means does the entrepreneur make all deductive linkages explicit).2 At the apex of the scale of generality are the highest-level postulates and hypotheses which the entrepreneur builds into a particular theory. They are the propositions from which the entrepreneur derives all others in elaborating that theory. In a sense these assumptions determine and entail all the lower-level claims made by the entrepreneur about reality. However, like reallife scientists, entrepreneurs often have some difficulty in assessing whether one hypothesis is really of a higher or lower level than another. Each entrepreneurial theory comprises a number of individual elements conjoined together into a more or less coherent whole. Entrepreneurs try to minimise the mutual contradictions and irrelevancies between these elements. However, because the environment is in a state of flux and entrepreneurs are continually changing their conjectures, it is very rare for an entrepreneur’s theoretical system to exhibit perfect unity and total consonance at any point of time. At the risk of oversimplification, it may be fair to say that the criticism, plastic control and correction of the entrepreneur’s hypotheses proceed upwards from lower to higher levels of the hierarchy of knowledge, whereas prediction goes in the opposite direction: ‘Reasoning downwards in generality’ means deducing consequences – making predictions – from hypotheses; while reasoning upwards [to higher-level statements] corresponds to the hypothetico-deductive ‘testing’ process, where a hypothesis that is not directly testable is checked by reference to a directly testable consequence (or where a more general hypothesis is assessed by reference to one that is less general). (Stewart 1979:72) More specifically, the entrepreneurs theoretical system is a multilevel hierarchy which consists of the following major structural components: metaphysical principles, empirical theories, and specific © 1996 David A. Harper predictions.3 Each of these is discussed separately in the remainder of this chapter. I also have some comments to make about the entrepreneur’s action programmes. 5.3 THE ENTREPRENEUR’S METAPHYSICAL PRINCIPLES The entrepreneur’s metaphysical ideas are purely speculative and can often be quite vague and fuzzy. They are articles of faith which the entrepreneur holds firmly. The configuration of these ideas is thus likely to be highly stable over quite a long period of time. Included in this category are the following sorts of assumptions: • the cluster of hard-core assumptions that the entrepreneur makes about the world, its regularity, and his or her location in space, time, society and nature; • the entrepreneurs basic beliefs about people, their openness to change, their needs (including their needs for self-esteem and social regard), the driving forces that motivate them, their honesty and trustworthiness (including the extent to which they act in good faith when unanticipated events occur), their altruism, asceticism and obedience, and the entrepreneur’s beliefs about the attributes of human nature which are important to contracting processes. Two divergent sets of assumptions about human nature are neatly exemplified by McGregor’s (1960) Theory X and Theory Y. The particular pattern of these assumptions will in turn shape the entrepreneur’s theory of leadership; • the entrepreneur’s beliefs about wealth, material gain and economic security; • the entrepreneur’s beliefs about the forces or constraints, if any, which delimit market events, and his or her beliefs about the control of fate in business (and especially about his or her influence over the outcomes of ventures) (Gasse 1986:52); • the entrepreneur’s assumptions about the sovereignty of consumers and the autonomy of consumer preferences (including assumptions about the extent to which marketing strategies can alter, reveal or create consumer tastes and thereby change consumers’ behaviour to suit the entrepreneur’s own ends); • the proverbs about ‘the economy’ that the entrepreneur accepts, the entrepreneur’s preconceptions about the economic process as a whole and about what is causally or teleologically significant within it, and his or her common-sense ideas about economic © 1996 David A. Harper theory. Also included here are the entrepreneur’s convictions about the extent to which market prices reveal the private information of other agents (Bacharach 1986:175–176, 181). These fundamental presumptions correspond to what Kelly in his theory of personality (1955, 1963) refers to as an individual’s ‘core constructs’ and to what Converse (1964) calls ‘central idea-elements’ in a person’s political ‘belief-system’. Taken together, they also resemble Boulding’s (1956) notion of ‘the Image’ and Gasse’s (1986) concept of an ideology of business. It should be noted that it is not so much the effect of one particular metaphysical idea taken in isolation that is important in explaining entrepreneurial behaviour, but rather the effect of the metaphysical programme as a whole (cf. Rokeach 1960). Another point is that entrepreneurs are often not aware of their own metaphysical speculations and rationalisations. Metaphysical theories are not testable, either because of their logical form or because of dogmatic decisions made by the entrepreneur to allow them to go unquestioned for the time being. Entrepreneurs may develop them into testable empirical theories at later stages of the venture process or of their careers. Although metaphysical ideas may at times slow down the entrepreneur’s learning, it must be remembered that they are a prerequisite to entrepreneurial discovery: they are an important source of entrepreneurial inspiration and perform a useful heuristic role in guiding the entrepreneur’s search for information. New entrepreneurial ventures would be impossible without them. 5.4 THE ENTREPRENEUR’S THEORY OF LATENT DEMAND An entrepreneurs theoretical system (T) is a superordinate construct which encompasses at least three subordinate empirical theories: • a theory of latent demand (td); • a theory of production (tp); • a theory of governance or, of the organisation of, transactions (tg). The entrepreneur’s theory of latent demand is discussed in detail in this section. The other two components are examined in the following sections. The entrepreneur’s theory of latent demand involves conjectures about the most urgent of the as yet unsolved problems of consumers and hypotheses about the new bundles of product characteristics © 1996 David A. Harper that will satisfy these latent demands. It also includes conjectures about the price-quantity-quality configurations by which target consumers will be willing to buy the entrepreneur’s new product over a period of time. More specifically, the entrepreneur forms a complex set of conjectures about: • the factors which determine the demand for the new product; the rate of growth and variability over time of demand, including the significance of seasonality in demand; • total market demand for alternative forms of the entrepreneur’s new product, that is, the total quantity that will be bought by the target customer group in a selected geographical area during a specified time period under a given marketing programme (especially product and pricing strategies) and assuming given environmental conditions; the maximum price that the entrepreneur thinks possible without inducing production by potential or actual rivals; the sales profile of the entrepreneur’s venture as consumer learning from experience with the product takes place; minimum and maximum possible first-time sales, replacement sales and repeat sales of the new product in each period (these conjectures in turn will depend upon the entrepreneur’s assumptions about the factors which affect the actual timing of replacement and repeat sales, such as the survivalage distribution of the product and possible competitive offerings available at the replacement date etc.); • the characteristics of potential end-use customers (i.e. the structure and composition of market demand): who will buy, where they will buy, when they will buy, how often they will buy, why they will buy; the degree of heterogeneity of potential customers in terms of their goals, resource endowments, knowledge and problem situations; the possible dimensions for segmenting markets (e.g. geographic, demographic, psychographic/life-style, benefit, volume, situation-specific and person–situation segmentation); the characteristics which a resulting market segment must exhibit for it to be profitable to develop a separate marketing programme tailored to it (e.g. measurability, accessibility, responsiveness and substantiality); taxonomic principles for classifying the problem situations of groups of consumers; the knowledge that potential consumers have about the products in existence, the terms on which they are available, their qualitative characteristics and their power to cater to consumers’ requirements; how to take consumers’ © 1996 David A. Harper transaction costs of exchange into account when designing alternative marketing strategies and pricing schemes; the purchasing methods employed by buyers (e.g. haggling, acceptance of list prices, solicitation of quotes); • the elasticities of demand with respect to changes in relative prices or product characteristics; how elasticities of demand can depend upon complementarities – linkages between consumers’ activities – and the associated costs of disrupting consumers’ existing patterns of choice; how demand elasticities vary across groups of potential customers; how the price-elasticity of demand facing the entrepreneur will change over the product life cycle, especially with increased product standardisation and consumer experience with the product; • the extent, if any, to which the entrepreneur can engineer people’s individual demand functions by marketing strategies and by particular tools in the marketing communications mix (especially advertising, sales promotion, personal selling and publicity); whether the demands of individual consumers are independent of one another or whether there are interaction effects; • the events and situations which cause problems for consumers (e.g. the failure of existing products); how entrepreneurs can stimulate potential consumers into recognising the urgency of hitherto overlooked problems; • how consumers create a choice agenda of possible activities: that is, how they identify a relevant set of options that could conceivably be solutions to their problems; • the methods consumers use to evaluate and choose between rival possible schemes of action; the sets of decision rules and evaluative criteria used by consumers to choose between competing products (conceived as bundles of characteristics); the relative importance different segments of consumers assign to these criteria in particular use-contexts; whether consumers evaluate product characteristics simultaneously or sequentially and the number of characteristics that they consider; in particular, whether target customers for the entrepreneur’s new product employ neoclassical, compensatory or non-compensatory decision procedures, and if the latter, then whether they use disjunctive, lexicographic or conjunctive choice models (or some hybrid form). Following Cantillon (1931), the growth-of-knowledge (GK) theory rejects the notion that demand functions are in any way given to the entrepreneur. Rather it is assumed that the constellation of demand © 1996 David A. Harper must be discovered (however tentatively) by the entrepreneur through a process of trial and error-elimination. This process of learning is not confined to the discovery of the demand function for a prespecified new product which possesses a given set of physical attributes. The reason for this is that the entrepreneur must also try to discover the combinations of product characteristics that are valued by particular groups of potential consumers in particular situations. (More on this shortly.) In order to test hypotheses about which combinations consumers value most, the entrepreneur must typically experiment in a piecemeal fashion with the elements of the marketing mix. Not surprisingly, the entrepreneur is considered unlikely to subscribe to formal neoclassical theories of demand, for reasons that Richardson has elaborated: Formal demand theory assumes that consumers are able to determine an order of preference between different combinations from a fixed list of distinct commodities; it represents no more than a simple application of the logic of choice. If we wish to consider, as does an entrepreneur, which kinds and qualities of goods to produce, then this model of consumers’ behaviour needs to be replaced. (Richardson 1960:102; emphasis added) Even Lancaster’s (1966) modified neoclassical theory of demand, which to its credit portrays consumers as wanting goods for the characteristics they expect them to contain, is not satisfactory because it fails to emphasise the situations of decision-makers sufficiently.4 Consequently, his theory cannot account for the immense variability in consumer behaviour that the entrepreneur must somehow attempt to explain and harness. Furthermore, it retains much of the neoclassical apparatus, such as utility maximisation and convex preference orderings, and thus leads to deterministic models of consumer behaviour. According to the GK theory, the entrepreneur applies the method of situational analysis in order to explain and to predict consumer behaviour and to segment market demand. Consequently, the entrepreneur analyses the situations of groups of consumers because demand is conjectured to be a function of the interaction of individuals with their situations.5 The entrepreneur subdivides market demand into distinct groups of consumers who are conjectured to have different demand schedules and to require © 1996 David A. Harper substantially different product characteristics in different usage situations. The entrepreneur seeks explanatory theories of consumer behaviour which appeal not to psychological elements (such as consumer preferences, tastes, wants and motives, etc.) but rather to objective elements of the situations in which consumers find themselves (such as consumers’ goals and knowledge, and the physical, social and temporal dimensions of purchase situations). For any particular case, the entrepreneur reconstructs two distinct problem situations: the usage situation as it is, and the situation as it appears to consumers (i.e. consumers’ subjective perceptions of the situation). Both reconstructions are of course conjectural but may be corroborated by independent evidence. This method of explanation has even found favour with some marketing specialists: ‘buyers can be just as naturally described and categorised in terms of the situations they will primarily use the product in, as describing them in terms of demographics, personality-traits or their attitudes’ (Dickson undated: 5). Indeed, in the marketing literature there is a growing recognition that individual consumer characteristics are by themselves severely limited in their ability to explain variation in buyers’ behaviour. This has prompted marketers to investigate explicitly the role of objective situations (and consumers’ perceptions of usage situations) in determining consumer choices (Belk 1974, 1975a, 1975b; Fennell 1978; Lutz and Kakkar 1975). Consequently, both the ‘productuse situation’ and the ‘person within situation’ can be regarded as potentially useful bases for explaining and segmenting market demand and for targeting marketing strategy. From the perspective of person–situation demand analysis, the entrepreneur views each consumer as a problem-solver and thus interprets consumers’ actions in any specific situation as attempts to solve problems: ‘[The ultimate objectives of consumers’ activity] can scarcely be described as the satisfaction of individual “tastes”’ (Richardson 1960:103). In particular, most problems arise for consumers in their (disappointed) attempts to predict, explain and control events in their lives. It is the entrepreneur’s contention that goods and services are demanded not so much for themselves but rather as possible means for obtaining the ends of prediction, explanation and control (Alderson 1971:143; Earl 1986b: 94; 1990a: 732; Gutman 1982). The entrepreneur therefore treats a new product as a novel bundle of situation-specific problem-solving services. (Conversely, any novel bundle of such characteristics may be thought © 1996 David A. Harper of as a joint product in that it may be able to solve several problems in various situations.) As part of the method of situational analysis, the entrepreneur may form conjectures about the net implications to individual users of utilising his or her new product in particular problem situations. From the subjective standpoint of consumer i in product-use situation j, the key issue is whether or not the positive implications appear to exceed the negative implications of making the switch to the entrepreneur’s new product (see Figure 5.1).6 In the case of markets for durable capital goods, the positive implications for potential industrial buyers (as construed by the entrepreneur) may comprise expected reductions in production costs and/or in transaction costs as a result of using the entrepreneur’s new capital item, and/or possible increases in sales revenues. By way of example, a new machine may be expected to lead to lower direct operating costs for the user if it automates some stage of the manufacturing process. If the new capital good is designed to reduce the uncertainty of physical output quality, it may also be anticipated to reduce the user’s transaction costs (especially, control, monitoring and measurement costs). The entrepreneur’s new good may also be expected to increase the user’s sales revenues if demand-enhancement effects are generated by using it in the production process – for example, if the new machine makes possible improvements in product performance and design which are valued by consumers. On the other hand, the conjectured negative implications of starting to use a new good may consist of the amount the consumer must pay for the entrepreneur’s new product as well as the consumer’s expected transaction costs of organising this exchange. (Indeed, the common distinction made in the marketing literature between convenience goods, shopping goods and specialty goods is based upon the price of the consumer product and the level of transaction costs the consumer bears in evaluating, selecting and obtaining the appropriate item.) Because consumers do not have proven true knowledge of the merits of every product or characteristics-bundle compared to another, they will often make experimental purchases as part of their continual quest to discover new and better ways of solving their problems and achieving their specific goals. In this way, their patterns of choice can be expected to change over time. Entrepreneurs’ theories of consumer behaviour will have a major impact upon their theories of production, and especially upon their product differentiation strategies. For example, if an entrepreneur © 1996 David A. Harper holds the theory that most target customers will use a disjunctive model of choice in the relevant class of purchase situations, the entrepreneur will seek to design the new product so that it performs better than rivals with regard to the single evaluative attribute which is conjectured to dominate consumer choice. (See Earl (1983b: 108– 112) for further implications of non-compensatory choice models for product design.) The entrepreneur’s theory of consumer demand, and especially the market segmentation strategy the entrepreneur derives from it, may be a source of competitive advantage. If the entrepreneur’s rivals interpret market demand in terms of traditional categories of consumers (e.g. geographic and demographic segments), they may not be alert to the existence of hitherto uncatered for market segments. Suppose then that on the basis of his or her novel segmentation theory, the entrepreneur develops a new product (or characteristics-bundle) specifically tailored to the situation-dependent problems of users in that segment and that the entrepreneur subsequently dominates the target market at a profit. If competitors do not share the same theory of market segmentation, they may be unable to discover the reasons for the entrepreneur’s success. And since they cannot explain the source of the entrepreneur’s success, the types of people purchasing the new product and the specific problems they are seeking to solve by consuming it, then rivals will experience some difficulty in encroaching upon the entrepreneur’s dominant market position, unless they can construct new and better theories of market segmentation. Consequently, the entrepreneur’s Figure 5.1 The conjectured implications of using a new industrial product © 1996 David A. Harper profits may continue for some time (Haley 1968:34; O’Driscoll and Rizzo 1985:105). 5.5 THE ENTREPRENEUR’S THEORY OF PRODUCTION The entrepreneur’s theory of production contains conjectures about the technological possibilities of combining given inputs into novel products or of obtaining given product concepts from novel combinations of inputs. More specifically, the entrepreneur forms hypotheses about: • how to characterise the productive transformations of which the entrepreneur’s venture may be capable; the nature of the relation between inputs and outputs (e.g. many-to-many relation between sets of inputs and outputs because labour services may be of a variable specification and yield a variable performance); • alternative possibilities of plant size and production capacity; for a given plant, the level of capacity utilisation that can be achieved during specified periods under normal working conditions and the plant’s maximum technical capacity; the appropriate scale of initial entry and the effect of scale of entry on the entrepreneur’s relative cost position; the appropriate length of production runs and how it may change over the product life cycle; the length of production runs necessary if the entrepreneur is to be able to experiment with new product designs so as to test the reactions and needs of consumers; the conditions under which it is appropriate to employ batch mode production rather than continuous process manufacturing, or vice versa; possible adjustments to plant capacity as production and product-mix change during the life of the venture; • the sequencing of, and the activities involved in, different stages of the production process; the specific input requirements for various production stages; the time schedule of production (e.g. start-up, initial plant tests, full capacity production); the time pattern of production (i.e. whether the new product is to be produced to order or delivered from inventory); possible procedures for quality control, production control and inventory control within a multistage production process; • possible techniques of production; the elasticity of input substitution; the extent to which the production process will tend to be labour or capital-intensive; the extent to which relatively © 1996 David A. Harper skilled workers are required who can participate in the problemsolving process and in the modification of product designs; • the specificity of inputs used in the production process (e.g. the mix of general purpose machinery and specialised machinery); how production costs change as asset specificity increases; the divisibility of inputs (especially of the manufacturing plant and of major machines); the degree of integration of the production process; the flexibility of alternative technologies and the rate of technological change in the entrepreneur’s industry; • the categorisation of production costs, that is, what to include under the rubric of production costs and how to classify them. Entrepreneurs form conjectures about the amounts they will have to ‘pay out’ to all factors to produce the planned level of output, including payments to shareholders and any imputed costs (e.g. depreciation and a sufficient rate of return on the money capital invested); • the relevant dimensions for distinguishing between different kinds of production costs: the components into which total production costs can be divided (e.g. total fixed costs and total variable costs, both of which can in turn be further divided into sunk and salvageable parts); • the structure of production costs, and the complex set of supplyside mechanisms which continually alters it; the structure of relative factor prices and its determinants; the quantity and quality of factors of production to which the entrepreneur has access at particular factor prices; the effects of organisational slack and Xinefficiency on production costs; • how production costs vary in the short-run as the level of output changes, together with which costs are constant over a specified range of output; the potential for reducing average production cost over time through the exploitation of possible economies of scale and of scope and also through learning by doing (the ‘experience-curve effect’); the lowest level of output which can be expected to yield the envisaged minimum average cost (i.e. the conjectured minimum efficient scale); the determinants of economies of scale and of scope in production; the sacrifices in economies of scale necessitated by alternative market segmentation strategies; • the production costs of rivals, existing or potential, who could produce near-perfect substitutes for the entrepreneur’s product; the likely volume of both competitive and complementary output; © 1996 David A. Harper • the ranking of, and the interaction between, the various factors which are critical to the selection of an industrial location (e.g. the location and ownership of essential inputs, the proximity of principal product markets, the existence of infrastructural investment, the socioeconomic environment, public policies, etc.); possible locations and specific sites suitable for the particular venture (especially for the erection and operation of any manufacturing plant); • the array of possibilities for qualitative variation in output (i.e. for product differentiation). The GK theory excludes the possibility that entrepreneurs characterise their firms as well-defined, fully specified production sets. It denies that entrepreneurs choose points on a unique production frontier which exists independently of economic agents and the firms which they create. By referring to the entrepreneur’s theory of production, the GK approach emphasises the experimental elements involved in the production process – the circumstance that the outcome of any particular input decision is not known with certainty. In this connection, Leibenstein (1968:73) deserves special mention as having identified two subtle assumptions of the neoclassical theory of the production function which he considers to be a major impediment to a satisfactory theory of entrepreneurship: namely, that ‘the complete set of inputs are specified and known to all actual or potential firms in the industry, and that there is a fixed relation between inputs and outputs’. He objects to economists treating the production function as a datum. Implicitly pursuing a subjectivist Hayekian argument (Hayek 1937:39), Leibenstein takes issue with observing economists assuming that actors know facts they do not and cannot actually know: ‘Where and to whom in the firm this knowledge is supposed to be available is never stated. In fact, there are great gaps of knowledge about the production function’ (1968:73).7 It is clear that the entrepreneur’s theory of production is intimately connected to the entrepreneur’s other theories. For example, the entrepreneur chooses a capital equipment rig and a target level of capacity utilisation in the light of conjectures about strategic aspects of pricing, such as those related to possible competitive reactions and latent market demand. Economies of scale can be realised only if the output is more or less continually sold, though inventory buffers can be used to some extent to absorb variations in demand. In addition, the envisaged pros and cons of producing to order and of © 1996 David A. Harper producing to stock will depend, among other things, upon the entrepreneur’s conjectures about the importance of being able to satisfy customers quickly (which favours producing to stock) and his or her conjectures about the significance of being able to custom design products which precisely match the buyer’s specifications (which favours producing to order). 5.6 THE ENTREPRENEUR’S THEORY OF GOVERNANCE Different entrepreneurs will hold different theories about the critical dimensions with respect to which economic transactions differ, and they will form different hypotheses about the most efficient and flexible ways of organising certain types of transactions. Different entrepreneurs will have different ideas about the circumstances under which market transaction costs will tend to be relatively high, and they will therefore have different conjectures about the circumstances under which internal organisation is appropriate. As outside observers, we see no a priori reason why the governance theories to which entrepreneurs subscribe must necessarily coincide with or conform to Williamson’s transaction cost approach, especially since the entrepreneur is entirely absent in Williamson’s framework.8 (In other words, there is no need to invoke some Muth-like rational expectations hypothesis, according to which the forecasts of economic agents are assumed to correspond to the predictions of the ‘relevant’ economic theory.) As a result, two entrepreneurs, who are otherwise identical except for their theories of governance and who are exposed to the same objective situation, are likely to choose different ways of arranging production and marketing, and their ventures will probably exhibit different patterns of growth. The entrepreneur’s theory of governance comprises conjectures on at least some of the following factors: • the sets of environmental and human factors, and combinations thereof, that generate obstacles to trade and hence transaction costs; • the definition of transaction costs (e.g. ‘resource losses due to lack of information’ (Dahlman 1979:148)) and the boundaries of the concept of transaction costs; • the classification of transaction costs into functional categories (e.g. search and information costs, bargaining and decision costs, and policing and enforcement costs); © 1996 David A. Harper • the specificity of the inputs which are required in the production and marketing of the entrepreneur’s new product (that is, the degree to which assets are fixed in terms of location, function or user). In addition to conjectures on how governance costs vary with asset specificity, level of output and governance structure, the entrepreneur also makes hypotheses about the demandenhancement effects (e.g. design or performance benefits) and the production-cost consequences (e.g. savings in transportation and inventory costs) that may be yielded by greater asset specificity. Thus, entrepreneurs’ theories of governance are linked inextricably to their theories of latent demand and of production; • the frequency with which transactions recur in bringing the new product to market (i.e. one-time, episodic, highly recurrent); • the uncertainty in economic transactions, and the sources of such uncertainty (e.g. whether it is exogenously or endogenously created); • the extent and variety of appropriability, measurement and divisibility problems that can be expected during the development of the venture; in particular, the extent of externality problems which can be expected to arise when independent distributors do not appropriate all the benefits of their efforts to enhance the quality of the entrepreneur’s new product or when they do not bear all the costs of quality debasement; • the cognitive competence of decision-makers with whom the entrepreneur will be engaged in contracting (i.e. the limits to their rationality), and the sophistication of their judgmental decisionmaking; • the different types of self-interest seeking to which other parties are prone, including the extent to which the propensity to act opportunistically varies among members of the contracting population, and the difficulties of distinguishing opportunistic from non-opportunistic actors. These conjectures of course are derived from the entrepreneur’s metaphysical beliefs about human nature, which were discussed earlier; • the importance placed on human dignity by parties to different kinds of transactions, and the types of transactions for which dignity is most crucial (especially transactions in internal and external labour markets); • the potential conflicts which can emerge in contracting processes, the sources of such conflict, and the major transactional milestones which are most critical to the success of the entrepreneur’s venture. © 1996 David A. Harper Included here are the entrepreneurs conjectures on how competitive (large-numbers) exchange relations can be transformed into small-numbers bargaining conditions, with the associated possibility of ‘hold-up’ problems at recontracting stages; • the existence of first-mover advantages which arise when the winner of the original contract acquires a cost advantage (e.g. by virtue of the acquisition of a unique location) (Williamson 1971:116); • the comparative merits of alternative governance structures (e.g. firms and markets) as institutions for acquiring knowledge and for coordinating diverse transactions across space and time; • the kinds of governance structures (and alternative forms of hierarchy) which will: promote a culture of creativity and of critical inquiry into highly complex market problems; safeguard the venture’s transaction-specific assets; permit the entrepreneur to appropriate the economic gains from innovation, speculation and arbitrage; help the entrepreneur to cope with the structural uncertainty inherent in competitive market processes; and permit adjustment (at the appropriate speed) to changes in technology, consumer requirements and resource availabilities; • the possible trade-off between economic efficiency in a static sense (i.e. the minimisation of the sum of production and transaction costs at any particular conjuncture) and flexibility (i.e. the potential for effective adaptation to unanticipated environmental changes); how the nature of what is transaction-cost efficient depends upon whether the entrepreneur perceives the environment to be in equilibrium or disequilibrium; the relative efficiency and the relative flexibility of alternative governance structures under specified circumstances (e.g. under disequilibrium conditions when markets and technologies are constantly changing); and the pathdependency of transaction costs; • the potential transaction-cost savings that can be realised by switching from one governance structure to another, and the conditions under which such economic gains can be expected to arise; • the determinants of different forms of divestment and disintegration (i.e. the splitting up of integrated structures into different parts); • the potential trade-offs between production-cost economies of scale and scope on the one hand and governance-cost economies on the other; • the interdependence between organisational forms and techniques of production. © 1996 David A. Harper At the hub of the entrepreneur’s theory of governance are his or her ideas about the spectrum of governance structures that can be devised to effect different kinds of entrepreneurial transactions (including arbitrage, speculation and innovation). For any particular economic activity, the entrepreneurs choice of governance structure is not limited to the polar cases of pure markets or hierarchical organisation. There are complex hybrid modes of governance located between both ends of the continuum, and entrepreneurial activity in this intermediate range is widespread (see Best (1990); Joskow (1985); MacMillan and Farmer (1979); Richardson (1972); Williamson (1983a, 1984c, 1985a). Belonging to this middle group of hybrid transactions are venture capital contracts, venture nurturing (including corporate new venture divisions and R & D partnerships), venture spin-offs, joint ventures, franchising, reciprocal trading (including product exchange agreements), share contracts, quasi-vertical integration and other forms of non-standard contractual arrangements. These sorts of transactions can sometimes involve the use of ‘hostages’ to support the exchange relationship. From this standpoint, the entrepreneur regards the firm as only a nexus of interrelated contracts, so that the line between the firm and the market becomes blurred. In particular, both the outer boundaries of the new firm, which the entrepreneur sets up as a vehicle for product innovation, and the internal structure of the firm itself are not given but are open to entrepreneurial initiative and imagination. Upon deciding to internalise some sets of related transactions, the entrepreneur must then apply his or her theory of governance in order to design the internal organisational structure of the venture, and especially its degree and form of hierarchy. In addition, the entrepreneur’s theory of governance is used to design the external organisation of the venture firm, that is, the linkages and relationships of cooperation between the entrepreneur’s venture and other firms which are completely separate in ownership and control. (As omniscient economists, however, we may observe ex post that the actual structure of the venture is not the sole result of conscious design on the part of the entrepreneur: although of pragmatic origin, much of the structure – such as connections of goodwill and other institutions of inter-firm cooperation – will evolve at least in part spontaneously over time. Because of structural uncertainty, the development of an innovating firm will depend upon many actual circumstances and economic changes which the entrepreneur could not have predicted.) © 1996 David A. Harper The transaction cost literature identifies the condition of asset specificity as the most important attribute of transactions and thus as the major determinant of the most efficient governance structure for organising transactions.9 (Indeed, because the transaction cost approach is mainly presented as a descriptive rather than prescriptive analysis, it implicitly assumes that entrepreneurs and corporate strategists also consider asset specificity to be the critical transactional dimension (cf. Earl 1984:21).) Asset specificity is defined as the condition which arises when inputs cannot be ‘redeployed to alternative uses or by alternative users without the sacrifice of productive value’ (Williamson 1986d: 15). For the growth-of-knowledge theorist, this definition should make it clear that the degree of specificity of durable inputs is not, as is typically assumed in transaction cost economics, an exogenous given known with certainty by the economic agent.10 Rather it is matter of conjecture on the part of the decision-maker. The conjectured degree of asset specificity depends on the entrepreneur’s expectations at time t1 (the moment of the decision to acquire the asset or to set up the governance structure) of the secondbest use of the asset at some future date, t1 + n’ when the entrepreneur may foreseeably wish to dispose of that very asset.11 (At time t1 + n’ of course, the entrepreneur’s expectations of the asset’s second-best use may differ from those held at time t1.) Because of structural uncertainty regarding the nature of transactions, both present and future, the entrepreneur may hold false hypotheses about the opportunities for redeploying malinvested assets to other uses or users, and hence the entrepreneur may also hold false conjectures about the specificity of the assets employed in any particular production technique. Such conjectures are most severely tested whenever the entrepreneur actually goes about turning capital goods to their perceived second-best purposes or whenever the entrepreneur tries to sell or lease out the assets to other potential users. The entrepreneur’s theory of governance is predicted to be one of the most important sets of conjectures crucial to entrepreneurial success. Many entrepreneurs fail to consider transaction costs explicitly, or they consistently underestimate the relevance of transaction costs (Picot 1986:4; Picot and Schneider 1988:27–31). Transaction costs are conjectured to be the hardest category of costs for entrepreneurs to measure and to predict. In particular, this class of costs is expected to be much less predictable than production costs (cf. Walker and Weber 1984:378). Consequently, it is conceivable that some entrepreneurs may not even possess a theory of governance, © 1996 David A. Harper with the result that they will omit transaction costs entirely from their evaluations of the profitability of competing innovative activities. Such entrepreneurs are likely to view their firms as mere production functions (however crude or ill-defined), rather than as governance structures. The assessment of potential transaction costs, and the importance assigned to them, can be expected to depend in part upon the previous functional specialisation of the entrepreneur. More specifically, entrepreneurs with a technical background in engineering are predicted to emphasise production costs and to downplay or ignore governance costs to a greater extent than entrepreneurs with extensive experience in purchasing. The failure of entrepreneurs to include transaction-cost factors in their formation of business strategies may result in entrepreneurs making incorrect and ill-fated decisions about where to draw the boundaries of their firms.12 Indeed, false theories of governance are predicted to be a frequent cause of entrepreneurial error and losses.13 ‘Manufacturers sometimes operate on the mistaken premise that more integration is always preferred to less’ (Williamson 1983c: 116). Williamson (1983c: 112–116) reports on the losses suffered by manufacturing firms which had made mistaken decisions to integrate forwards into distribution or backwards into the supply of raw materials. Among other things, substantial penalties from full vertical integration are conjectured to have arisen when the integration strategy involved a large sacrifice in economies of scope (and to a lesser extent, When it led to the failure to exhaust scale economies fully). The emphasis placed upon the entrepreneur’s theory of governance is also consistent with the results of an empirical study by Picot, Laub and Schneider (1988, 1989) into innovative business start-ups in Germany. They found the organisation of transactions to be a decisive determinant of the performance of newly founded ventures in their sample. Compared with less successful innovators, the very successful entrepreneurs appeared to adapt their governance structures much more sensitively (and in the direction predicted by Williamson’s transaction cost theory) to changing economic conditions. Whereas very successful innovative start-ups tended to shift to external procurement in factor markets as soon as conditions were favourable, less successful new firms tended to do the reverse and brought the production of components in-house. Furthermore, very successful firms utilised more sophisticated and specialised modes of contracting for externally procured resources than did the less successful enterprises. (Less successful firms were also more rigid in their procurement policies © 1996 David A. Harper and tended to manufacture standardised inputs themselves to a greater degree than very successful firms.) Very successful innovative entrepreneurs made more extensive use of long-term supply contracts, cooperation agreements, written contracts, and complex contracts involving intensive negotiations. In this way, the very successful ventures were able to focus their limited pools of resources upon their hardcore innovative activities which could not be delegated to the external market, and at the same time they were able to govern their contractual relations with external suppliers more efficiently and flexibly. On the basis of their results, Picot et al. concluded that the ‘design and development of transaction-cost efficient coordination modes seem to be a very important property of innovative entrepreneurs in a dynamic environment’ (1988:28).14 5.7 THE ENTREPRENEUR’S SPECIFIC PREDICTIONS So far I have described the three main types of empirical theories which each entrepreneur puts together in developing an entrepreneurial idea for a new business venture: namely, the entrepreneur’s theories of latent demand (td), of production (tp) and of governance (tg). Entrepreneurs combine their insights into latent demand with their theories of production and of governance in order to make their discoveries and to derive their specific predictions about the existence of profit opportunities. I will now discuss this point in more detail. From the conjunction (T) of these three empirical theories (together with any auxiliary hypotheses he or she makes), the entrepreneur derives the prediction (P) that a new product embodying a particular bundle of characteristics can be marketed to a target set of potential consumers in a definite spatial region over a specified future time period to yield revenue which will more than cover the sum of production costs and governance costs.15 More formally and succinctly, the entrepreneur predicts: T·A → P or T·A → π, π = R (PC + GC) such that π > 0, where: T = some conjunction of the entrepreneur’s theories of © 1996 David A. Harper latent demand, of production, and of governance, that is, T = td·tp·tg; · = the conjunction (simultaneous assertion) of hypotheses; A = the entrepreneur’s set of auxiliary hypotheses, especially his or her account of the initial conditions of a market situation (i.e. the entrepreneur’s ‘knowledge of the particular circumstances of time and place’ (Hayek 1945:521)); → = the relation of deducibility or analytical implication; P = the entrepreneur’s set of specific predictions; π = anticipated profit; R = total expected revenue from selling the new bundle of product characteristics; PC = total expected production costs; GC = total expected governance costs. The entrepreneur’s predictions that opportunities for profit exist are low-level elements in his or her theoretical system which are in closer empirical contact with reality than higher-level elements. The entrepreneur specifies these predictions in such a way that they apply only to finite ‘regions’ of space and time. The entrepreneur’s prediction (P) is thus that ‘There is an opportunity for profit in the space–time region k’. The entrepreneur’s prediction (P) has the logical form of a singular existential statement or a singular ‘there-is’ statement: it belongs to that class of propositions that asserts that ‘There is an x in the space–time region k’ or that ‘An (market) event of type x is occurring in the region k’. It should be made clear that in the present context the term ‘predictions’ does not just comprise the entrepreneur’s forecasts of unknown future events. As used here, the term also includes assertions about past market history (i.e. retrodictions) as well as descriptions of known economic events which the entrepreneur wants to explain (i.e. explicanda) (cf. Popper 1959:60; 1960:133). 5.8 THE ENTREPRENEUR’S ACTION PROGRAMMES The entrepreneur also possesses an incompletely specified hierarchy of action programmes which are applied in particular types of situations. These programmes organise the entrepreneur’s actions © 1996 David A. Harper into patterns of various levels of complexity, generality and abstraction: ‘configuration is just as important a property of behavior as it is of perception [and I might add, of conjecture]’ (Miller, Galanter and Pribram 1960:12–13).16 Entrepreneurs’ action programmes are punctuated by long-range goals and short-run objectives. In addition, entrepreneurs develop policy recipes to guide their actions towards the attainment of prescribed goals. An entrepreneur may consider major policies, procedures and rules to be the hard core of one action programme which is itself part of the protective belt of another, higher-level action programme, thereby giving rise to a kind of core/demicore map of entrepreneurial action programmes (see Remenyi 1979). Langlois provides a novel matrix approach which is useful for depicting the hierarchical structure of economic agents’ plans: We can . . . think of the agent’s actions as organized in hierarchical fashion, with the set of actions A divided into subsets Ai composed of subactions aij, which are, in turn, composed of sub-subactions aijk, etc. . . . An action or plan at the highest level of the hierarchy is abstract in the sense that it is oriented toward a typical situation and consists in a general pattern of response. As we examine the plan at lower levels of the hierarchy, we see that more details have been filled in. The plan is increasingly more concrete in its orientation and more specific in the response it embodies. (Langlois 1986a: 183–185) The common distinction between business strategy (e.g. A i ) and tactics (a ij ) reflects the hierarchical nature of the entrepreneur’s planning process. The end of this chapter concludes the introduction to the basic concepts of the GK theory of entrepreneurship. The next three chapters apply these concepts to the question of how entrepreneurs operate within the market system. In particular, they elaborate the notion of the ‘sophisticated falsificationist’ entrepreneur.