THREE VIEWS OF ENTREPRENEURIAL OPPORTUNITY

1
THREE VIEWS OF ENTREPRENEURIAL OPPORTUNITY
Saras D. Sarasvathy
University of Washington
S. Venkataraman
Nick Dew
Rama Velamuri
University of Virginia
Invited book chapter in the Entrepreneurship Handbook edited by Acs et.al.
(Revised January 4, 2002)
2
"Although we are not usually explicit about it, we really postulate that when a market could be
created, it would be."
-- Kenneth Arrow (1974)
For almost fifty years now, following the trail of issues raised by economists such as
Hayek, Schumpeter, Kirzner and Arrow, researchers have studied the economics of technological
change and the problem of allocation of resources for invention (invention being the production
of information). The bulk of this literature simply assumes that new technical information will
either be traded as a commodity or become embodied in products and services (hereafter called
‘economic goods'), without addressing any specific mechanisms or processes for the
transformation of new information into new economic goods or new economic entities (such as
new firms and new markets). It is inside this gap that we begin our quest for the concept of an
“entrepreneurial opportunity”.
In a recent interview with CNN, Whitfield Diffy, the inventor of public key encryption
(currently an employee of Sun Microsystems), explained that although his entire subsequent
career had benefited from his invention and he had done very well financially in the process, it
did not occur to him to start a company to commercialize his invention. In fact he expressed
astonishment at the "hundreds and hundreds of people trying to turn a buck on it". The designers
of the MIR space station would no doubt express similar astonishment at the venture capitalists
who recently bid (in vain) several million dollars to turn it into an advertising/tourist resort --
just as the scientists working with DARPA did not foresee the age of e-commerce. The history
of technological invention is full of unanticipated economic consequences. And, yet, the study
3
of the economics of technological change is full of "just-so" stories1 that seemingly demonstrate
the inevitability of commercialization of all new technologies through familiar recurring patterns
such as the technology adoption curve. Unfortunately, of course, we do not have any data on all
the new products and markets that were not created to commercialize new technologies in the
past.
This paper challenges the assumption underlying current theories of technological
change, laid out so pithily by Arrow in the initial quote, viz, "when a market could be created, it
would be". Instead, it focuses on Arrow's exhortation to researchers to tackle one of the central
problems in economics today: "... the uncertainties about economics are rooted in our need for a
better understanding of the economics of uncertainty; our lack of economic knowledge is, in
good part, our difficulty in modeling the ignorance of the economic agent."
We begin our exposition with a definition of entrepreneurial opportunity. Then we
delineate its elements and examine it within three views of the market process: i.e., the market
as an allocative process; as a discovery process; and as a creative process (Buchanan &
Vanberg). Within each stream, we examine the assumptions about the knowledge (ignorance) of
the decision maker with regard to the future, and the implications of those assumptions for
strategies to recognize, discover, and create entrepreneurial opportunities. We end the essay with
a set of conjectures that challenge the inevitability of technology commercialization and argue
for a more contingent approach to the study of the central phenomena of entrepreneurship.
1 1 Just so stories (based on Rudyard Kiplings collection of short stories of the same title) are stories that explain why things are the way they are.
Such stories also tend to celebrate things the way they are -- subscribing to the fallacy that because certain things came to be, there is some
element of “optimality” or “correctness” attached to their origin and structure. This approach leads us to discount the significance of pre-histories
because if existence by itself is the starting point of theory building, almost any story could ex-post serve as sufficient explanation for the prehistory.
One delightful example is the story of an arbitrage struggle between an elephant and a crocodile that explains how the elephant came to
have a long trunk! Relatedly, almost all the social sciences seem perfectly capable of explaining every creation after the fact, but can predict
nothing before the creation.
4
ENTREPRENEURIAL OPPORTUNITY
The Oxford English Dictionary defines opportunity as “A time, juncture, or condition of
things favorable to an end or purpose, or admitting of something being done or effected.” If we
believe that that ends are never specified prior to the pursuit of an entrepreneurial opportunity,
but may emerge endogenously over time, we can draw our definition of entrepreneurial
opportunity from the second part of the above sentence. An entrepreneurial opportunity,
therefore, consists of a set of ideas, beliefs and actions that enable the creation of future goods
and services in the absence of current markets for them (Venkataraman, 1997). For example, the
entrepreneurial opportunity that led to the creation of Netscape involved (a) the idea of a userfriendly
Web browser (Mosaic); (b) the belief that the internet could be commercialized; and, (c)
the set of decision-actions that brought together Marc Andreesen (the creator of Mosaic) and Jim
Clark (the ex-founder of Silicon Graphics) to set up base in the small town of Mountain View.
In sum, an entrepreneurial opportunity consists of:
1. New idea/s or invention/s that may or may not lead to the achievement of one or more
economic ends that become possible through those ideas or inventions;
2. Beliefs about things favorable to the achievement of those ends; and,
3. Actions that implement those ends through specific (imagined) new economic artifacts (the
artifacts may be goods such as products and services, and/or entities such as firms and
markets, and/or institutions such as standards and norms).
THREE VIEWS OF ENTREPRENEURIAL OPPORTUNITY
Drawing upon three streams of economic literature pertinent to entrepreneurial
opportunity – i.e., market as an allocative process, market as a discovery process, and market as
5
a creative process -- we could model an entrepreneurial opportunity as a function, or a process or
a set of decisions, respectively. The antecedents for the three views presented here specifically
draw upon three works, i.e., Hayek (1945), Knight (1921), and Buchanan and Vanberg (1991) --
all of which grapple with the central problem demarcated by Arrow (quoted earlier) in terms of
understanding uncertainties in the economy and modeling the ignorance of the economic agent.
In an important essay in 1945, Hayek postulated the concept of dispersed knowledge
where no two individuals share the same knowledge or information about the economy. Hayek
distinguished between two types of knowledge: first, the body of scientific knowledge, which is
stable and can be best known by suitably chosen experts in their respective fields; second, the
dispersed information of particular time and place, whose importance only the individual
possessing it can judge. Hayek pinpointed the harnessing of this latter type of knowledge as a
key and underestimated element in the economic development of society. This dispersion has
two extremely important implications as far as entrepreneurial opportunities are concerned.
First, dispersion of knowledge is a root explanation for the presence of uncertainty, which gives
rise to opportunities in the first place. Second, dispersion of knowledge is another root
explanation of the nexus of the enterprising individual and the opportunity to discover, create
and exploit new markets (Venkataraman 1997, Shane 2000). Without this nexus of the
individual and the opportunity, most inventions will lie fallow. Frank Knight (1921) clearly
realized the implications of uncertainty for economic organization.
In his seminal dissertation, Risk, Uncertainty, and Profit, Knight distinguished between
three types of uncertainties about the future that an economic agent may face:
?? The first consists of a future whose distribution exists and is known, and therefore decisions
would only involve calculating the odds of a particular draw and placing one's bets based on
6
the analysis. In this case, risks can be reduced through diversification. This assumes that all
the possible outcome scenarios are all equally likely, ex ante.
?? The second consists of a future whose distribution exists but is not known in advance. The
agent, in this case, has to estimate the distribution through repeated trials and can then treat it
the same as the first case. Furthermore, as the environment changes dynamically, successful
strategies evolve through adaptive processes including careful experimentation and learning
over time. Although we do not know the probabilities attached to each of the outcome
scenarios, the probabilities do exist, and their distribution can be uncovered over time.
?? The third type of uncertainty, which Knight called true uncertainty, consists of a future that is
not only unknown, but also unknowable -- with unclassifiable instances and a non-existent
distribution. The economic agent, or entrepreneur, who takes on this true uncertainty, gets
compensated for it through "profit" -- a form of residual return after the normal factors of
production are paid for and all market contracts fulfilled.
Knight did not explicate how the entrepreneur deals with this true uncertainty. But, instead he
argued that: The ultimate logic, or psychology, of these deliberations is obscure, a part of the
scientifically unfathomable mystery of life and mind. We must simply fall back upon a
“capacity” in the intelligent animal to form more or less correct judgments about things, an
intuitive sense of values. We are so built that what seems to us reasonable is likely to be
confirmed by experience, or we could not live in the world at all. In this third case of Knightian
uncertainty, there is no meaning to the attachment of probabilities to the opportunity vectors.
Instead, we need to understand the process through which the different levels of actors interact.
The benefits get created endogenously, in the very unfolding of those interactions.
7
Later researchers, especially Austrian economists such as Von Mises and Kirzner, and
subjectivists such as Lachman and Wiseman, have tried to tackle this problem of Knightian
uncertainty. Fixing a rather penetrating philosophical gaze on the works of these economic
theorists since Hayek and Knight, Buchanan and Vanberg (1991) contrast the three views of
economic theory presented here as follows: "The market as an allocative process, responding to
the structure of incentives that confront choice-makers; the market as a discovery process,
utilizing localized information; or the market as a creative process that exploits man's
imaginative potential..." They argue that “the perceptual vision of the market as a creative
process offers more insight and understanding than the alternative visions that elicit
interpretations of the market as a discovery process, or, more familiarly, as an allocative process.
In either of the latter alternatives, there is a telos imposed by the scientist's own perception, a
telos that is nonexistent in the first instance. And removal of the teleological inference from the
way of looking at economic interaction carries with it significant implications for any diagnosis
of the failure or success, diagnosis that is necessarily preliminary to any normative usage of
scientific analysis."
Recent empirical evidence that examines in detail how expert entrepreneurs make
decisions when faced with building a firm for a new product without any given existent market
(i.e., when faced with Knightian uncertainty), provides considerable support for the view of the
market as a creative process that neither ignores teleology nor assumes it a priori in the research
endeavor. Sarasvathy (1998) found that the entrepreneurs not only did not assume the existence
of the market, but explicitly expressed their belief that the existence of the market cannot be
demonstrated or known in advance. Rather than using the causation-based logic 'To the extent
you can predict the future, you can control it.', the subjects in the study overwhelmingly (74% of
8
the subjects over 63% of the time) followed the logic of effectuation that 'To the extent that you
can control the future, you do not need to predict it.' (Sarasvathy, 2001). This logic overcomes
the problem of true Knightian uncertainty in a curiously paradoxical way: On the one hand, it
eschews prediction altogether – i.e., eliminates the need for prediction, and on the other, it
transforms the future into near certainty – i.e., makes it highly predictable -- by “creating” the
distribution.
But, both in Buchanan and Vanberg's theoretical exposition of the market as a creative
process and Sarasvathy's empirically grounded conceptualization of effectuation, the key issue is
not which of the three views is "right", but rather which view is more useful under what
conditions of uncertainty. Such a pragmatic approach allows us to utilize the three views
explicated so far to construct a rather simple typology of entrepreneurial opportunities based on
the pre-conditions for their existence, as follows:
1. Opportunity Recognition
If both sources of supply and demand exist rather obviously, the opportunity for bringing them
together has to be "recognized" and then the match-up between supply and demand has to be
implemented either through as existing firm or a new firm. Examples include arbitrage and
franchises.
2. Opportunity Discovery
If only one side exists -- i.e., demand exists, but supply does not, and vice versa -- then, the nonexistent
side has to be "discovered" before the match-up can be implemented. Examples include:
Cures for diseases (Demand exists; supply has to be discovered); and applications for new
technologies such as the PC (Supply exists, demand has to be discovered).
9
3. Opportunity Creation
If neither supply nor demand exist in an obvious manner, one or both have to be "created", and
several economic inventions in marketing, financing etc. have to be made, for the opportunity to
come into existence. Examples include Wedgewood Pottery, Edison's General Electric, U-Haul,
AES Corporation, Netscape, Beanie Babies, and the MIR space resort.
Table 1 presents a summary comparison of the three views along several different
dimensions. In the next three sub-sections, we look at the nature of entrepreneurial opportunities
and effectuation costs from each of these perspectives and develop questions for future research.
THE ALLOCATIVE PROCESS VIEW
Neoclassical economic theory discusses several efficiency properties of markets –
allocative, productive, coordinative, and informational. We will focus in this section on the
allocative efficiency of markets and its implications for opportunity recognition. Allocative
efficiency is achieved when: a) the income of consumers is optimally allocated to consumption,
i.e., they are able to buy the goods and services that they value most; and b) resources (factors)
are optimally allocated to production, i.e., they are used to produce the goods and services that
consumers desire.
Allocative efficiency is achieved in a perfectly competitive market, whose characteristics
are as follows: there is a very large number of buyers and sellers, all of whom are so small that
none of them individually can affect prices; prices of homogeneous goods and factors are
uniform throughout the economy; all factors are perfectly mobile; returns to scale are constant;
and all economic agents have perfect knowledge about available alternatives. There is an
assumption of complete markets, i.e., there are markets for all possible products and services.
Furthermore, agents are free to enter and exit the market. Disequilibria are short-term
10
phenomena, and are quickly cleared to bring the situation back to equilibrium through the
tatonnement process - prices go up when demand exceeds supply and down when supply exceeds
demand – which functions through the mythical figure of the Walrasian auctioneer. There are
further requirements for the achievement of an optimal allocation of resources, such as the
absence of any divergence between private and social costs and the existence of perfect
competition in all sectors of the economy. When a market has achieved allocative efficiency, it
complies with two conditions: first, price is equal to marginal cost, which is also equal to
minimum average cost (P=MC=minAC); and second, Pareto optimality is achieved, which
means that resources cannot be redistributed to make anyone better off without making someone
else worse off.
The allocative view concerns itself with the optimal utilization of scarce resources. In this
view, an opportunity is any possibility of putting resources to better use. At equilibrium, there
are no opportunities, because resources have been optimally allocated. However, profits can arise
in two ways. First, to the extent that a perfectly competitive market is not in equilibrium,
opportunities for short term profits are available, but they quickly disappear when new firms
enter the market attracted by the profits. Second, if we assume that all information is available in
the system but is randomly distributed, and therefore acquiring information involves a costly
search process, then the opportunity for profit is simply the difference between the benefit of the
information and its cost. However, the random distribution of information means that no agent
has the possibility of systematically benefiting from superior information. The core idea is that
all products and ideas that can potentially exist are all known to be feasible but costly to produce.
When the cost problem is solved (for example, due to scientific breakthroughs in laboratories),
opportunities arise. However, opportunity is not specific to any one person because there is no
11
informational advantage within this view. Thus there is no heterogeneity between economic
agents that enables one agent to be systematically better than another in acquiring information,
and consequently in the recognition and pursuit of opportunities. Which agent recognizes the
opportunity is therefore a purely random variable. Moreover, since there is no divergence
between private cost and social cost (that is, the opportunity cost for an individual agent of a
resource in a particular use is the same as the social opportunity cost of the resource in that use),
any possibility of a Pareto improvement at the system level is equivalent to an opportunity at the
individual agent level.
Arrow (1962) discussed three reasons why a perfectly competitive market could lead to a
suboptimal allocation of resources to invention: inappropriability, indivisibility, and uncertainty.
In what follows, we analyze how allocative efficiency is compromised as a result of these three
reasons.
Inappropriability
An issue that has been debated for many decades is whether there is any incentive to
innovate in a perfectly competitive market, because it does not, by definition, permit the
appropriation of rents in a sustained fashion. Kamien and Schwartz (1975) study the relationship
between market structure and innovation, and conclude that “few, if any, economists maintain
that perfect competition efficiently allocates resources for technical advance” (p. 2). Arrow
(1962) argued that the incentive to innovate could exist even in perfectly competitive markets:
“It may be useful to remark that an incentive to invent can exist even under perfect competition
in the product markets though not, of course, in the “market” for the information contained in the
invention. This is especially clear in the case of a cost reducing invention. Provided only that
suitable royalty payments can be demanded, an inventor can profit without disturbing the
12
competitive nature of the industry. The situation for a new product invention is not very
different; by charging a suitable royalty to a competitive industry, the inventor can receive a
return equal to the monopoly profits” (p. 619).
For Arrow’s point to be valid, the assumption of all sectors of the economy being in a
perfectly competitive equilibrium must be relaxed. Schumpeter (1942) was of the opinion that
the propensity of a firm to innovate was directly proportional to its size and market share. He
based his view on the considerable resources required to innovate and the incentive of adequate
return. Nutter (1956) disagreed – “Desire and necessity drive competitive and monopolistic
producers alike to innovate: desire for better-than-average profits motivates the venturesome and
industrious to introduce new products and techniques; loss of profits forces the cautious and
passive to imitate or perish” (p. 523).
Villard (1958) offered a view that ran counter to that of Nutter, concluding that
innovation was unlikely at both extremes. “Industries where “competitive oligopoly” prevails
are likely to progress most rapidly and that therefore “competitive oligopoly” may well be the
best way of organizing industry. The basic point is that progress is likely to be rapid (1) when
firms are large enough or few enough to afford and benefit from research and (2) when they are
under competitive pressure to innovate – utilize the results of research” (p. 491). Scherer (1967,
“Market Structure and the Employment of Scientists and Engineers”, AER, vol 57, pp. 524-531)
agreed with Villard, arguing that moderate levels of concentration lead to the highest levels of
innovation.
Indivisibility
Blaug (1985) defines indivisibility as follows: “If two productive agents are perfect
substitutes of each other when used in combination to produce a given output, they are
13
necessarily infinitely divisible: the isoquants in this case are straight lines, meaning that the
marginal rate of substitution of the two factors is a constant” (454).
Arrow (1962) argues that “a given piece of information is by definition an indivisible
commodity, and the classical problems of allocation in the presence of indivisibilities appear
here” (p. 615). He goes on to explain the problems: “In the absence of special legal protection,
the owner cannot, however, simply sell information on the open market. Any one purchaser can
destroy the monopoly, since he can reproduce the information at little or no cost. Thus the only
effective monopoly would be the use of the information by the original possessor. This however,
will not only be socially inefficient, but also may not be of much use to the owner of the
information either, since he may not be able to exploit it as effectively as others” (615).
Economic theory assumes that in the absence of property rights, the original creator or
discoverer of particular information would lose control of it once it was reproduced and
accessible to other parties. Thus a large part of the discussion on appropriate institutional
structures revolves around establishing the right incentives – copyright laws, patent laws etc - for
agents to innovate. However, there may be some classes of information that can be used only in
combinations with other assets, such as human and physical capital. For this reason the rents
from the use of such information may not accrue to parties who do not possess these assets, and
this difficulty may provide adequate protection for the innovator, even in the absence of specific
legal protection. There are many industries in which firms do not patent inventions in spite of
the existence of patent laws. The distinction between information and knowledge becomes
relevant here. Brown and Duguid (2000) argue that knowledge differs from information in three
ways: first, knowledge is tied to a knower; second, it is harder to detach than information; and
third, it is hard to give and receive because it requires more by way of assimilation. They also
14
distinguish between the explicit and tacit dimensions of knowledge. “[S]trategy books don’t
make you into a good negotiator, any more than dictionaries make you into a speaker or expert
systems make you into an expert. To become a negotiator requires not only knowledge of
strategy, but skill, experience, judgment, and discretion. These allow you to understand not just
how a particular strategy is executed, but when to execute it. The two together make a negotiator,
but the second comes only with practice” (Brown and Duguid, 2000: 133-134).
Thus, although information is indivisible and the costs of reproducing it are close to zero,
we may relate it to a resource, as defined in the resource based view of the firm. Knowledge, on
the other hand would be a capability in that it represents a combination of information, physical
capital and human capital. Focusing exclusively on raw information makes us view
opportunities as arbitrage possibilities, which are not agent specific. On the other hand, focusing
on knowledge opens up rich vistas of agent specific opportunities, whose recognition depends
upon already owned knowledge and other assets (Shane 1999).
Uncertainty
Akerlof (1970) argued in his famous “lemons” paper that an extreme case of information
asymmetry could lead to a complete market failure. Information asymmetry leads to uncertainty
that causes a downward bias in demand and supply. This is because, at very high levels of
uncertainty, agents will need concessions so large from the other party to the transaction that
neither will recognize any opportunity in the exchange. Institutional support is then often needed
to overcome the uncertainty and to restore trade in the market. For example, organizations such
as the SEC ensure certain minimum levels of transparency and fair play, which benefit all
participants in the form of an increase in the volume of trade. Markets themselves can correct
for this asymmetry – firms specializing in information gathering, analysis, and dissemination
15
pervade all markets. These firms lower an individual agent’s search costs while increasing the
quality of information. Institutions such as guarantees, brand names, and licensing practices are
some of the other ways of overcoming the uncertainty caused by information asymmetry.
The other major reason for uncertainty according to Arrow (1974a-Limited Knowledge
and Economic Analysis) is the nonexistence, except in a very limited number of commodities, of
futures goods markets. “Hence, the optimizer must replace the market commitment to buy or
sell at given terms by expectations: expectations of prices and expectations of quantities to be
bought or sold. But he cannot know the future. Hence, unless he deludes himself, he must know
that both sets of expectations may be wrong. In short, the absence of the market implies that the
optimizer faces a world of uncertainty” (p. 6).
According to Arrow, this uncertainty leads to the economic agent taking steps to reduce
risks, such as the holding of inventories, preference for flexible capital equipment etc. It also
leads to the creation of new markets for the shifting of risks, such as the equity market.
However, while conceding that probabilities are subjective, because different agents have access
to different information, he implies that each agent can know his own distribution of
probabilities from his own past. He states that uncertainty means “that we do not have a
complete description of the world which we fully believe to be true. Instead, we consider the
world to be in one or another of a range of states. Each state of the world is a description which
is complete for all relevant purposes. Our uncertainty consists in not knowing which state is the
true one” (1974b-The Limits of Organization: 33). The views of Frank Knight (and perhaps
more importantly, the different interpretations of what he actually meant) on the distinction
between risk and uncertainty become very relevant here.
16
In summary, there are several implications of viewing the market as an allocative
process. First, the focus is on the system and not on individuals or firms, which are all
homogeneous in their access to technology and in their cost structures. Second, ex ante, all
economic agents are equally likely to detect a given opportunity. Opportunity recognition is thus
a purely random process. Third, the term competition is as appropriately applied to factor
markets as it is to the market for goods and services. In both cases, the markets are assumed to
be in competitive equilibrium.
THE DISCOVERY PROCESS VIEW
Two factors influencing the distribution and use of new information have therefore
attracted attention from researchers. The first is that access to information sources is extremely
important, leading some researchers to suggest that the prime determinant of entrepreneurship is
whether the entrepreneur has an advantageous network position from which informational
advantages accrue (Burt, 1992). For instance, information is often “sticky” (von Hippel, 1994)
in that it is tacitly accumulated by users, which means that access to the relevant information for
discovery to occur is only available to a few individuals who have direct and intimate contact
with users. Second, new information or knowledge often requires complimentary resources in
order to be useful, such as a prior knowledge (Venkataraman, 1997, Shane, 2000) that is also
often tacit in nature. Such prior knowledge creates the “absorptive capacity” necessary for an
individual to make use of new information (Cohen & Levinthal, 1990).
The second reason why people possess different beliefs about the prices at which markets
should clear is because, as Kirzner (1973) has observed, the process of discovery in a market
setting requires the participants to guess each other’s expectations about a wide variety of things.
However, the regular supply of new information from endogenous sources creates uncertainty
17
(Knight 1921) owing to the fact that the discovery of genuinely novel information by other
agents can affect the value of resources. Such discoveries cannot be known ahead of time and
may add previously unimagined categories of usage for particular resources, thus changing the
structure of the decision problem the entrepreneur faces (Langlois 1984). Since it is impossible
to have accurate expectations about inventions that have yet to be made, people form
expectations based on hunches, intuition, heuristics, and accurate and inaccurate information,
leading their expectations to be incorrect some of the time.
The problem of forming accurate expectations given the genuine uncertainty caused by
the endogenous supply of novel information is compounded by some characteristics of human
decision-making. All individuals utilize knowledge that is subjectively held, incomplete and
tacit. Entrepreneurs therefore form beliefs and expectations about future events that are
indeterminate for at least three reasons. First, because much knowledge is tacit (Polanyi 1967)
other individuals -- upon whose actions the correctness of the entrepreneur’s expectations depend
-- often base their decision-making on invisible elements of experience that are hard to verbalize,
but are observed instead only as hunches, intuition and judgement. Second, situations calling for
prediction are not given self-evidently because the essence of any situation is how it is enacted
by individuals (Weick 1979). People often produce part of the situation they face (they “enact”
it). The dependency of enactment on tacit cues imposed on a situation by individuals means that
there is an indeterminacy in how individuals produce situations, just as there is an indeterminacy
to how they react to them. This is especially so when multiple actors interact, making the
production of a situation dependent on an “inter-enactment” process. The third reason why
outcomes are indeterminate is because interaction among individuals gives rise to emergent
outcomes. One example of an emergent outcome of the interaction of many individuals in a
18
market is a structure of prices, but many other emergent outcomes are not so predictable, hence
their discovery as an aspect of market processes. One of the traits of complex adaptive systems
such as market processes is level differences: observed patterns of behavior differ dramatically
between the micro and macro levels (Kauffman 1995). In other words, macro level phenomena
are often indeterminate from micro-level observations. Hence the opportunity to discover is an
outcome of the very inability to predict, or form accurate expectations, about such complex
dynamic phenomena.
Since entrepreneurial opportunities depend on asymmetries of information and beliefs,
entrepreneurs’ buying and selling decisions are not always correct and this process leads to
“errors” that create shortages, surpluses, and misallocated resources. An individual alert to the
presence of an “error” may buy resources where prices are “too low”, recombine them and sell
the outputs where prices are “too high”. The notion that individuals can make these genuine
discoveries about misallocated resources has led some researchers to stress the role of “surprise”
(Kirzner 1997) in this process. The nature of overlooked profit opportunities is that they are
completely overlooked, and therefore individuals are genuinely surprised when they identify a
hitherto unexpected profit opportunity. Such surprises are not searched for at the cost of a
deliberate search process. Instead, individuals are totally ignorant of these misallocated resources
and their total ignorance precludes a deliberate search process. Given that uncertainty and
indeterminacy make expectation formation difficult, it is reasonable suggest that regular
surprises will be a feature of the discovery process.
One factor that leads to stability in expectations is the role of institutions, which are
routinized patterns of action. The presence of routines makes expectation formation a
possibility, since certain patterns of human behavior can be reasonably predicted based on the
19
observation of routines (Heiner 1983). Given the limitations pertaining on human cognition
(Simon 1996), routines are an essential aspect of human action for two reasons: first, because
they allow each particular individual to preserve scarce decision-making resources for
application to non-routine decisions; and second, because they allow all other individuals to
economize on scarce decision-making resources because they can make reasonable predictions
about the actions of others based on observation of their routines.
Routines are therefore pervasive at the individual level, where we usually describe them
as habits, as well as at the organizational level. Every individual has a particular regime of
unreflective habits that are accumulated over a lifetime of experience and experimentation
(Tsoukas 1996). The particular habits of an individual amount to a specialized collection of
routines. Organizations such as firms also accumulate specialized collections of routines
(Nelson and Winter 1982). In fact, one example of a predictable routine is the entrepreneurial
process described here: people can reasonably forecast that some other people are conjecturing
resources are undervalued in their current use and can be purchased and recombined and put to
more valuable use. On the other hand, people can also reasonably forecast that many other
individuals are simply carrying on with their daily lives: being a fireman, or minding their
children, or relaxing in their old age. In fact, were it not for the presence of imperfect
information and a wide variety of routine modes of behavior (i.e. non-alert, non-entrepreneurs)
the entrepreneurial discovery process would not work (Loasby 1999).
Institutions are important because they impose structure on the world, and as we have
already seen, an absence of structure creates the kind of uncertainty that makes forming accurate
expectations an impossibility. But to the extent that institutions do exist, expectation formation
is a reasonable possibility. Institutional routines therefore are an important part of the discovery
20
process in two ways: first, because routines create a stable interpretative scheme, they enable the
entrepreneur to impose order on and make sense out of the “bloomin’ buzzin’ confusion” of
experience (James 1908); and second, because individuals know what a stable structure is, they
are able to notice exceptions. In essence, the notion of surprise only makes sense because an
individual knows when he/she is not surprised. Since cognitive limits mean individuals cannot
be attentive to everything at once, entrepreneurial alertness (Kirzner 1973) is a function of what
is not given attention; that is, it is a function of other routinized modes of behavior. In other
words, entrepreneurial alertness comes with the opportunity cost of that which has been taken for
granted. Given that opportunity cost is the essential feature of choice, this economic calculation
ought to come as no surprise to us.
Of course, as the structure of a particular market becomes well established and
routinized, eventually entrepreneurial opportunities become cost inefficient to pursue. This
occurs for two reasons. First, the opportunity to earn entrepreneurial profit will provide an
incentive to many economic actors. As opportunities are exploited, an externality is created:
information diffuses to other members of society at no cost or low cost, and these individuals can
imitate the innovator and appropriate some of the innovator’s entrepreneurial profit. This
diffusion through imitation is one of the most important yet under-researched aspects of the
entrepreneurial process (Nelson and Winter 1982, Baumol 1992). Although the entry of
imitating entrepreneurs may initially validate the opportunity and increase overall demand,
eventually competition begins to dominate (Hannan & Freeman, 1984). When the entry of
additional entrepreneurs reaches a rate at which the costs from new entrants exceeds the benefits,
the incentive for people to pursue the opportunity is reduced because the entrepreneurial profit
becomes divided among more and more actors (Schumpeter, 1934).
21
The second reason entrepreneurial opportunities eventually become cost inefficient to
pursue is that the exploitation of opportunity provides information to resource providers about
the value of the resources that they possess, leading them to raise resource prices over time to
capture some of the entrepreneur’s profit for themselves (Kirzner, 1997). In short, the diffusion
of information and learning about the accuracy of decisions over time, combined with the lure of
profit, will reduce the incentive for people to pursue any given opportunity.
The duration of any given opportunity depends on a variety of factors. The duration is
increased by the, “inability of others (due to various isolating mechanisms) to imitate, substitute,
trade for or acquire the rare resources required to drive down the surplus” (Venkataraman, 1997:
133). For instance, the provision of monopoly rights, as occurs with patent protection or an
exclusive contract, increases the duration. Similarly, the slowness of information diffusion, or
lags in the timeliness with which others recognize information, also increase the duration,
particularly if time provides reinforcing advantages, such as occur with the adoption of technical
standards (network externalities) or learning curves.
What makes the discovery process metaphor powerful is that the dual premises of a
continuous supply of new information and a continuous process of realizing information about
the “errors” of prior expectations suggest the market process will be a continuous one. This view
of the market as a process distinguishes the discovery view from the allocative view, where the
metaphor of equilibrium leads to the perception of markets in static terms. In contrast, the
discovery process illustrates how the market is necessarily “alive” and a hive of human activity.
THE CREATIVE PROCESS VIEW
The origins of the creative process view are more recent than the older views based on
the market as a discovery process and the even older and established view of the market as an
22
allocative process. Consequently, this view is not yet as well developed as the other two. The
key idea in this view, as Buchanan and Vanberg (1991) point out, is that telos is neither ignored
nor imposed on the phenomena concerned. Instead, ends emerge endogenously within a process
of interactive human action (based on heterogeneous preferences and expectations) striving to
imagine and create a better world.
The origins of the allocative process view lie in the philosophy of Adam Smith and the
equilibrium-based calculus of Marshall, Walras, Arrow & Debreu and others; the development
of the discovery process view owes its origins to the philosophical roots of evolution going back
to Darwin, and is steeped in the calculus of asymmetric information explicated by Hayek, Nelson
& Winter and others; similarly, the creative process view originates in the philosophy of
pragmatism professed by James and Dewey, and takes its cue for shedding a large portion of
historical and even evolutionary determinism, instead moving toward a calculus of contingency
based on the notion of human “free will.”
In 1992, founding his arguments on the work of pragmatic philosophers, and drawing
from reputed scholars in a variety of social sciences, Hans Joas sought to establish the creative
nature of all human action. Key to his theorizing is a triad of arguments that demonstrate that
action (as an empirical fact) is: (a) always situated (i.e., cannot presuppose purposes or be
divorced from the sources of the actor’s intentions); (b) intrinsically corporeal (i.e., cannot be
freed from the constraints and possibilities of the body of the actor); and, (c) essentially social
(i.e., cannot originate or occur meaningfully in the absence of others). The three sets of
arguments challenge the existing conceptions of human action based on formal or normative
models of based on “rationality” (For example, models of suhjective expected utility). In Joas’
own words, “… I have argued that some approaches towards a conceptualization of human
23
creativity have actually drawn an artificial rift between creative action and the totality of human
action. My intention is therefore to provide not a mere extension to, but instead a fundamental
restructuring of the principles underlying mainstream action theory.” (1996: 145)
Joas shows that to the extent that an actor is incapable of purposive action, lacks control
over his own body, and is not autonomous vis-à-vis his fellow human beings and environment,
his actions are creative. In other words, they end up creating novelties in our world. Hence, in
Joas’ conception, instead of being anomalies to be explained, surprise and novelty become
natural desiderata of a theory of human action that is not confined to so-called “rational” action.
The creative process view urged by Buchanan & Vanberg (1991) asks us to speculate on
this alternative model of human action, and to develop non-teleological theories of economics.
In other words, if human beings are not assumed to be “rational” actors, but instead if human
behavior is deemed inherently creative, what kind of an economics (or any other social science,
for that matter) would we get?
Joas (1996) and Buchanan & Vanberg (1991) are not isolated in their exhortation to
scholars to pursue this line of inquiry. March’s garbage can model of decision making contains
one such set of attempts (March, 1994). In his own words, “In a garbage can process, it is
assumed that there are exogenous, time-dependent arrivals of choice opportunities, problems,
solutions, and decision makers. Problems and solutions are attached to choices, and thus to each
other, not because of any means-ends linkage but because of their temporal proximity” (1994:
200). Examples of garbage cans include committee and board meetings where a variety of
problems, solutions, and decision makers come into temporal proximity with or without
particular means-ends chains being involved in the coming into being of particular choices.
24
Building further upon such attempts, March urges us to build a “technology of foolishness” or
theories of decision making in the absence of pre-existent goals (March, 1982).
Other attempts in this direction include the empirical work based on Weick’s theories of
enactment and sensemaking (Weick, 1979; 1995). Just as March’s oevre on decision making
highlights the endogeneity of goals, Weick in his theory of enactment focuses on the endogeneity
of the environment. He points out how theorizing about “organization” and “environment” as
two separate entities prevents organizational scholars from asking important questions. In his
own words, “But the firm partitioning of the world into the environment and the organization
excludes the possibility that people invent rather than discover part of what they think they see.”
(1979: 166)
As early as 1969, Simon (1996) had talked about designing or planning without final
goals and the artificial nature of the world we live in. His exposition brought out the role of
current action in the design of future environments. In his own words, “The real result of our
actions is to establish initial conditions for the next succeeding stage of action. What we call
"final” goals are in fact criteria for choosing the initial conditions that we will leave to our
successors.” Therefore, how we want to leave the world for the next generation becomes an
important question in theories based on the creative view.
In sum, the crux of the creative process view is the need to build non-teleological theories
of human action, wherein values and meaning emerge endogenously. Recent empirical work in
expert entrepreneurial decision making (Sarasvathy, 2001b) has led to the development of such a
non-teleological theory in entrepreneurship. This theory posits an alternative to predictive
(causal) rationality, called effectuation, that underlies decisions made by entrepreneurs in
bringing new firms and markets into existence (Sarasvathy, 2001a). Starting without any given
25
goals, effectuation inverts the key principles and logic of predictive rationality to carve out an
alternative paradigm to rational choice. In this view opportunities do not pre-exist – either to be
recognized or to be discovered. Instead they get created as the residual of a process that involves
intense dynamic interaction and negotiation between stakeholders seeking to operationalize their
(often vague and unformed) aspirations and values into concrete products, services and
institutions that constitute the economy.
SUMMARY AND CONCLUSION
In the foregoing exposition we have outlined and briefly discussed three views of
entrepreneurial opportunity under the broader umbrella of the three views of the market process
as allocative, discovery, and creative. We now turn to the question of how to integrate the three
views into our practice and pedagogy and future scholarship, particularly in the area of
entrepreneurship.
One way to look at the three views would be to simply consider them three equally valid
and non-overlapping modes of thinking about entrepreneurial opportunities. Such an approach
focuses only on the distinctions between the views and overlooks both the possibilities of
relationships and interactions between them, and also the fact of empirical confounding in the
way they are embodied in economic phenomena. Table 1 sets out all three views along certain
key dimensions and allows us to discuss from a bird’s eye view, as it were, both distinctions and
overlaps.
For example, looking at the operationalization of the three views as the recognition,
discovery, and creation of opportunities suggests that the creative view might be more general
than and prior to the other two views. This is because creative processes contain recognition and
discovery as necessary inputs, while recognition and discovery can do without most key aspects
26
of creativity. A simple example of this point is that before we can “recognize” or “discover”
great art, that art has to have been created. Similarly, entrepreneurial opportunities may be
posited to have been “created” through the decisions and actions (conscious or unintended) of
economic actors before someone can “recognize” or “discover” them. For instance, once
specific goals, values and preferences have been formed through the creative process, discovery
processes can discover various means to achieve the goals. And when both ends and means
become manifest, allocative processes figure out which particular means can best achieve which
particular ends.
Another way to integrate the three views would be to recognize that they are extremely
context-dependent. In other words, each view is useful under different circumstances, problem
spaces and decision parameters. For example, when resources are clearly specified and goals are
given, the allocative view will be the most appropriate. In contrast, when the problem spaces are
characterized by enormous uncertainties, and value criteria for making choices are highly
ambiguous, a creative approach might be called for.
The essence of our exposition is not to establish the superiority of any one of the three
views or even to completely characterize them in all their possible relationships. Rather, our
explicit intention here is to demonstrate that the study of entrepreneurial opportunity is a far
richer and substantially more textured and interesting area of inquiry than it has hitherto been
supposed to be. Furthermore, it derives its interest and promise as much from the practitioner’s
desire to earn higher profits as from the philosopher’s and artist’s dreams of creating a better
world. But perhaps most importantly, an inquiry into entrepreneurial opportunity has the
potential to unlock one of the greatest intellectual puzzles of our time, namely the creation of
new value in society.
27
In conclusion, every invention2 engenders opportunities for the creation of several
possible economic (as well as other types of socially significant) effects. In the foregoing
sections we have examined three sets of views with regard to how these effects come to be.
Approaches based on the view of the market as an allocative process focus entirely on the final
effects of opportunity creation, treating the processes leading to these final effects as mere detail;
approaches based on the view of the market as a discovery process emphasize only the origins of
the opportunity for creation, treating the final effects as inevitable products of competitive
markets; and finally, approaches based on the view of the market as a creative process emphasize
the decisions and actions of the agents, making both origins and final effects contingent upon
those decisions and actions.
In our view, if we are to deepen our understanding of entrepreneurial opportunity, we
need to integrate these three approaches, emphasize contingencies rather than inevitabilities in
each. As a first step in that direction, we offer the following fundamental argument for the study
of the central phenomena of entrepreneurship – viz, entrepreneurial opportunities.
Conjecture 1:
The set of all possible economic goods based on any invention is larger than the set of
economic goods actually created within a finite period of time after the invention.
Conjecture 2:
Not all actual economic goods created from an invention will be created by existing
economic entities. In other words, the creation of new economic goods often entails the
creation of new economic entities such as new firms and new markets.
Conjecture 3:
2 The term "invention" need not be limited to technological (i.e., science-based) inventions. Inventions can occur in
all spheres of human activity -- in the arts (surrrealism), in sports (snowboarding) and in philosophy (pragmatism),
28
From the point of view of economic welfare, not all actual economic goods and economic
entities arising out of any invention are equally “desirable”.
Ergo, the lags (temporal and otherwise) between any invention and the creation of new economic
welfare enabled by it, require not only the ability and alertness to recognize, and the perception
and perseverance to discover opportunities for the achievement of pre-determined goals such as
increasing profits and larger market shares, but also necessitate decisions and actions based often
only on human imagination and human aspirations, that may or may not in time lead to new
products, firms and markets.
to name only a few.
29
Table 1: Comparing the three views of entrepreneurial opportunity
View Allocative View Discovery View Creative View
What is an
opportunity
Possibility of putting
resources to good use to
achieve given ends
Possibility of correcting
errors in the system and
creating new ways of
achieving given ends
Possibility of creating new
means as well as new ends
Focus
Focus on System
Focus on Process
Focus on Decisions
Method
Opportunities “recognized”
through deductive processes
Opportunities “discovered”
through inductive processes
Opportunities “created” through
abductive processes
Domain of
application
When both supply and
demand are known
Only one or the other
(supply or demand) known
When both supply and demand
are unknown
Distribution of
opportunity
vectors
Opportunity vectors are
equally likely
Existent, but unknown
probability of opportunity
vectors
Probabilities for opportunity
vectors are completely nonexistent
Assumptions
about
information
Complete information
available at both aggregate
and individual levels
Complete information at the
aggregate level, but
distributed imperfectly
among individual agents
Only partial information even at
the aggregate level, and
ignorance is key to opportunity
creation
Assumptions
about
expectations
Homogeneous expectations
both at the micro and macro
levels
Homogeneous expectations
at the macro level;
heterogeneous expectations
at the micro level
Heterogeneous expectations at
both micro and macro levels
Management
of uncertainty
Uncertainty managed
through: Diversification
Uncertainty managed
through: Experimentation
Uncertainty managed through:
Effectuation
Definition of
success
Success is a statistical artifact
Success is outliving failures
Success is a mutually negotiated
consensus among stakeholders
Unit of
competition
Resources compete
Strategies compete
Values compete
Outcomes
Strategies for:
Risk management
Strategies for:
Failure management
Strategies for:
Conflict management