Explaining entrepreneurship

To many in the field, at its core, the phenomenon of entrepreneurship captures a basic process that is almost as ubiquitous as evolution. The basic process involves the realization of new value (e.g., by constructing new lower cost processes) resulting from action. That action can be both mental (e.g., involving the construction of beliefs, making decisions, and learning) and physical (e.g., involving searching and prototyping). That action is sparked by some change in the environment (e.g., an exogenous shock that changes the information available to the market). Of course, such a process is not exclusive to entrepreneurship, or to technological progress, or even to humans. It is a process also considered in fields like technology innovation management, in creative outlets beyond the technological, and in non-human biological systems (although that outside research – e.g., in biology – is seldom cited in the entrepreneurship literature).²

One existing, common and core explanatory model that depicts the basic process in entrepreneurship phenomena consists of three related pieces: 1) a way (or the belief of a way) to increase profits; 2) a market imperfection (e.g., any condition impeding the increase of efficiency or of value in supplying a good) that is addressed by that way; and 3) a specific entrepreneur who embodies some heterogeneity in her differentiated ability to address that market imperfection. We detail this basic process on pp. 23–28 by providing a relevant context, by defining the terms involved, by clarifying the drivers, and by stripping away all of the excess in order to re-establish what academics who study entrepreneurship focus on.

The basic process labeled as entrepreneurial activity (or entrepreneurship) is not new or exclusively commercial or human-specific. Market imperfections (referred to as MIs) have existed throughout time and space, and continue to do so, and are developed and exploited by differentiated entities. That basic process, as the most important part of the engine of capitalism, explains how most recent and successful empires were built. That basic process occurred in the first ancient exploitation of the giant species in Australia by homo sapiens – where giant mammals (mainly marsupials) did not know the dangers of the newly arrived species of man.³ That basic process has even been applied by less-intentioned parties (e.g., the wheat plant, and the canine animal) to exploit the preferences of the genes of mankind, resulting at times in the centuries-long detriment to the living conditions of an average person for the gain of an increase in the population of certain genes (Harari, 2015). In other words, the basic process of an entity trying to improve its existence by exploiting an MI using its differentiation (i.e., its heterogeneous advantages) is a natural one, and does not require a human or a monetize-able opportunity to exist.

Definitions and explanations

Drawing on the description of the core explanatory model underlying existing entrepreneurship activity, we now define the important terms of that process, and further describe its elements that are referred to in recently proposed partial theory (and, that given these are widely understood terms or straightforward logical statements, we state them without fanfare, understanding that some scholars will not be in agreement even if citations were provided). Figure 3.1 depicts the summary of terms explaining the basic process (as commonly understood in the field).

The definition of entrepreneurship: Under the basic process that defines the core of the field, entrepreneurship can be defined and explained (unoriginally) as heterogeneously addressed market imperfections. The MI provides the possibility to increase profits.⁵ The heterogeneity provides the basis for a competitive advantage to the focal unique entity. Together, these form the basis for a belief (or a natural programming for non-humans) that underlies the intention of an entity to act to appropriate some part of that profit.⁶

For-profit MI types: One of the most important identity relationships in business is that profit equals margin times quantity sold less fixed costs. It is our field’s touchstone, much like, in physics, force equals mass times acceleration. Entrepreneurial activity, as an engine of capitalism, seeks to increase profits in two main ways: i) for an existing good – by lowering costs (variable or fixed), by increasing price, or by growing quantity sold [or some net positive combination of these variables]; or ii) for a new good, by having sufficient net income to cover fixed cost investments. Thus, every for-profit MI involves possibilities and actions relevant to either i) or ii). The entrepreneur seeks to increase profits in order to capture some profits privately (or, for social entrepreneurs, to increase in value for its intended targets). The entrepreneur faces some level of uncertainty that necessarily comes from newness and from the changes that are supposed to lead to the increase in profitability (where the uncertainty can entail unknowns regarding technical success, market demand, rival reaction, and competency in the new venture to gain legitimacy and deliver). With success, the entrepreneur and her investors reinvest some of the increased profits in further attempts to increase future profits, thus keeping the virtuous cycle of capitalism driving.

Figure 3.1 Basic process underlying existing entrepreneurship theory

The origins spurring entrepreneurial activity: There exist only two MI origins – those directly based on unforeseeable shocks, and those that are not. In the former case, the entity that profits is differentiated in its speed or effectiveness in responding to the external shock’s effects on profitable possibilities (e.g., in exploiting a new technology to lower manufacturing costs of an existing good). In the latter case, the focal entity’s actions expose a new profit possibility (e.g., in the firm’s experiments with existing technologies), where such actions are based ultimately upon endowed beliefs or abilities that were then leveraged in experiences that the entity had with its environment (some of which may have involved stochastic processes). Note in either case, the MI arises directly or indirectly from shocks or endowments that were exogenous.

The heterogeneity of the entity: The heterogeneity that differentiates the entity trying to address the MI can take many forms. Consider the heterogeneity captured in the form most commonly associated with human entrepreneurial activity – the differential advantage of a new venture relative to an incumbent. For the focal new venture to have an advantage, incumbency must be detrimental. Incumbency must involve relative disadvantages due to sunk costs, commitments that lock them out of new technologies, outdated mental maps, slow bureaucracies, cannibalization concerns, and so on. Additionally, it must be the case that the focal new venture can pursue its course while being able to sidestep those existing traps (i.e., the ones the incumbents cannot avoid) to make different choices and take different actions. Furthermore, the focal new venture must be able to do so without the requirement for much legitimacy, complementary assets, experience, brand, supply chain, and so on. And, still further, that focal new venture must be able to do so in a context where the parties in the relevant industry value chain (especially the customers) understand the new, alternative offering provided by that venture quickly and painlessly (e.g., without prohibitive switching costs, training, disposal costs, and so on). In other words, the heterogeneity involved in this process is multi-dimensional, and involves several affected parties; it is not simply that the entity differs from others, but it is how and to whom those differences matter (as well as possibly why, when, and where those differences matter).⁷

Endogeneity and exogeneity: It is important not only to define the MI and the heterogeneity in the core process that defines entrepreneurial activity, but also to be explicit about other terms related to the process, as these have been used as wedges to promote proposed new concepts. One such clarification that is important to make relates to the endogeneity and exogeneity involved in the process. No entrepreneurial activity is an island; all such activity involves both exogenous and endogenous influences. Endogeneity is implied by the heterogeneity part of the core process, where a differentiated entity (internally) takes actions to address the MI. Exogeneity is implied in both the external origins of the entity’s heterogeneity (i.e., in some endowment) and in the spark (i.e., external shock) that identifies the MI is now ready for the entity to act upon. Thus, entrepreneurial activity defined as heterogeneously addressed market imperfections is a combination of both an internal and an external driver at work in the process. (It is a definition that recognizes that entrepreneurship is a complex process, open to a wide variety of factor influences – both hard and soft – combining in interdependent ways and sequences.)

Factor types: Another clarification to make relates to the nature of the factors involved in providing the increase in value: specifically, relating to the way such factors exist – e.g., objectively, subjectively, or otherwise. Recent proposed theorizing sets this question up as a choice between objective reality and social construction. While objective reality is readily understood, social construction is a more complicated idea. Social construction is best understood in an evolutionary, historic perspective; it is an integrating, cooperative method for humans to collectively imagine a thing (i.e., to believe in it) so that they can operate efficiently and effectively in large numbers (i.e., at over the 150-person tribe size – see Harari, 2015). The most useful and obvious example thing is money – where specific objects (e.g., metal disks or shells or paper) represented transferrable general value that many people who wanted to trade other goods accepted in exchange (e.g., money for labor units, for crops, for spices, etc.). Money is a thing that provided the basis for geographically wide trade, for the separation of people from business in the form of a limited liability company, and so on (Harari, 2015). Socially constructed things are real given that others believe in them and even buttress those beliefs by force (e.g., by force of the crown who backed the value of the coins). Two consequences of social construction should be noted then: First, it does not apply to small groups – like an entrepreneur and her initial partners and customers – where a common set of beliefs can be supported through personal trust rather than through constructs that rely on external enforcement. In other words, the beliefs in the focal things that are not immediately tangible are sufficiently real within that group to be considered tangible, given that the minds that can touch them are housed in touchable members of that small group. Second, even when the relevant group gets big, and the idea of socially constructed things actually applies, those things (e.g., money, loans, stock, wave-particle duality, etc.) are sufficiently real to the entrepreneurs to have equivalence to tangible things. For example, even if the entrepreneur suddenly does not believe in money because it is socially constructed, she will still be thrown in a very tangible jail if she steals some. Her ability to combine a set of existing factors – whether those factors are tangible things or not – in order to address an MI is the only important issue in the process. A philosophical debate about the immediately tangible nature of the factors is a secondary concern, at best, to real entrepreneurs.

Existing factors are combined: To be clear, the process that the heterogeneous entity takes to address an MI involves decisions and actions. Those actions involve combining existing factors in new ways to increase profits. Those factors can be existing tangible things (e.g., materials), or existing intangible things (e.g., knowledge). No matter what, however, they must already exist somewhere or they would not be available to be combined anew. Every new invention uses existing factors. A confusion may arise when there are multiple steps involved, where intermediate newly created factors are used to get to the final combination of factors embodying the final product that addresses the MI. In that case, the focus can then be shifted to the existing things that were combined to produce the intermediate factors. In other words, no new thing can be created from nothing (i.e., a vacuum); a new thing must be created by combining existing things. The point is that those existing things (that are later combined in new ways to address the MI) are all real in the same sense to the entrepreneur – all exist sufficiently in the entrepreneur’s mind to apply, regardless of their immediate tangibility. And, given that all combinations involve knowledge – an intangible thing – all entrepreneurial activity involves some form of subjective reality.

Informational gaps: The process that the heterogeneous entity takes to address an MI involves decisions and actions that take place where vital information is notimmediately available (or no MI would exist, as it would be known as would be its solution). The assumed heterogeneity across entities implies that not everyone knows everything that everyone else knows. In the real world, any shock will create ripples of informational updates from those who know what is new to those who want to know what is new, and so on. The process that allows some differentiated entities to combine existing factors into something new to address a new MI is messy, dynamic and complex. One of the details not well understood is the sub-process involving identifying which factors to combine and how (and when and where and why). This is a sub-process that involves teasing information out of others in order for the focal entity to translate contextual ambiguity into risk, and addresses a set of informational gaps and asymmetry problems in order to move from a position where not all states and their probabilities are known (i.e., ambiguity) to a position where the states and their probabilities are known (i.e., risk). Only under the latter state can optimal decisions be made.

From ambiguity to risk: This translation from ambiguity to risk that the entrepreneur must undertake to make the decisions to combine the factors to address the MI involves getting nature (e.g., its physical properties in terms of technological possibilities, like super-conductance, or its deposits of materials, like oil) and people (e.g., customers, investors, suppliers, and so on) to respond with answers to the actions taken that represent the entrepreneur’s hypotheses and questions. Obviously, better answers are more likely when something is at stake for the respondent (e.g., an attractive product behind the offered prototype) and when the offering being presented is more complete (so the respondent does not have to use as much imagination and can focus on specifics). When nature and people provide feedback, then the possible states and their probabilities become increasingly known to the entrepreneur. Essentially, the informational gaps that need to be filled involve information that nature and people hold (e.g., involving technological feasibility and market demand levels) that the entrepreneur tries to pry from them using certain actions (e.g., testing prototypes for performance and asking for pre-order commitments). Given that it is standard to assume that, over time, firms end up making decisions under risk, it must be that ambiguities over such commercial opportunities are always resolvable once nature and people provide that initially missing information. In other words, the standard assumption is that ambiguity can always be translated into risk by drawing out the existing needed information from nature and people. Successful entrepreneurs seem to do this earlier and then to exploit the first-mover advantages involved.

Note that we are arguing that all entrepreneurial activity, given its necessary condition of doing something new entails gathering new information from nature and from people to formulate a risk-based decision. Regardless of whether the underlying innovation is incremental (i.e., with few unknowns to resolve) or radical (i.e., with many unknowns to resolve), the process is the same. Any separation of entrepreneurial activity based on how many unknowns need to be resolved alone does not appear warranted at first blush.

The divide between ambiguity and risk: It is a standard assumption that there is a clear division between ambiguity and risk, even though (in our literature) ambiguity always reduces to risk over time (except in the special cases, like those of wicked problems). That assumption implies that the only challenge is whether ambiguity can be reduced to risk before it is ready to be so reduced. (The challenge is met through the mechanism of nature and people providing feedback to actions to provide the missing states and their probabilities.) But is that a false or artificial challenge? It is simple to prove that any ambiguous context can be reduced to a risky one even without any significant action by the entrepreneur. Given the power to choose the states, an entrepreneur can always choose the currently known knowable state of the status quo (given she has experience with this), with the alternative state of not status quo covering all else, and be able to assign a probability to each. Thus, that divide is not a natural one; it is an artificial one. (On the other extreme, one could argue that all possible states of a real world can never be known down to the sub-atomic level, and so risk is never possible either.) In fact, the important divide has to do with the application to the real world, and the idea that the real world effectively chooses the level of detail of the states (and not the entrepreneur, although she may try to). It is the level of the decision that defines the necessary level of detail of the states (e.g., one needs to know rain or not rain if one cares about the decision to take an umbrella, whereas one needs to know more detail like the inches of precipitation over a set of hours in a specific geographic location if one needs to plan for flooding, and so on). The real world alone should define the decision levels that are useful and the states that need to be known. (Given that the decision level may depend on the possible states, it is likely that this uncovering of information is a co-evolving process, but that is not the point here. The point is that the divide between ambiguity and risk is artificial in a specific way – i.e., contingent on the decision – and, as such, any proposed theory that uses such a divide as a wedge is likely to be standing on shaky ground.)

It may be helpful to envision a theoretical world where the process unfolds, one much like Kirzner implies (1973, 1979). It is a world full of possible combinations of existing factors along different sequences, some of which can lead to new profits. To identify those few in the latter category, the entrepreneur needs to ask the right questions of information-holders to get their attention and motivation and reaction, at a specific time. There is no lasting ambiguity but instead it is a lack of specific knowledge at any one time about un-addressed MIs (where quality effort in quantity is needed to shed light on a particular corner of that unknown space to determine if it can be profitable and how so). But, the entrepreneur cannot just try any combinations at hand without knowing somewhat deeply what is being tried or she is unlikely to learn the right lessons to progress. The learning approach is captured in humankind’s recent cognitive revolution where ignorance is finally seen as a kind of truth (i.e., knowing the unknowns), and as opportunity (i.e., to make those identified unknowns then known).

The choice of how to explain entrepreneurial activity

Prior to testing the newness of recently proposed partial theories, it is useful to describe the main approaches for explaining phenomena used in the social sciences. Burrell and Morgan (1979) provide some guidance here regarding the choices along ontological, epistemological, methodological, and human nature dimensions, where one consistent approach is considered subjectivist (defined by nominalism, anti-positivism, ideography, and voluntarism), while its contrasting consistent approach is considered objectivist (defined by realism, positivism, nomothety, and determinism). If the core model has been traditionally described in terms of one approach (objectivist), while the proposed partial theories all take the contrasting approach (subjectivist), then that cross-approach translation issue might explain why tests for newness have been lacking.

Theory is valuable because it provides the understanding behind science’s purpose as a means to understand, predict, and control phenomena. Theory explains the how and the why focal outcomes occur (e.g., outcomes between variables, or along a process sequence), as well as how, when, where, and who. Once the theory is provided, humankind can then act to influence the phenomena to its benefit (e.g., by directing the outcomes, or by preparing for the outcomes given an accurate prediction). Influencing the phenomena, however, means adding some specificity to the description, and that is done through explanation. The explanatory model specifies, in applicable and explicit terms, what is controlled for, what variables or events are focal, what the boundaries of applicability are, and so on. The explanatory model provides an academic, and often a practical, context where the theory should work. For convenience, let’s consider the explanatory model as the means of identifying the assumptions, drivers, outcomes, relationships, controllable levers, dangers, and so on of the partial theory.

The approaches for explanation are many, but are anchored by two choices, objectivist and subjectivist. The objectivist approach maximizes simplicity while capturing the focal relationship that is theorized. It is abstract yet robust and tends to provide maximally explicit (mathematical) definitions of all terms, assumptions, and mechanics in a simplified version of what is supposed to be the real world. The subjectivist approach maximizes reality, attempting to capture the full complexity, messiness, and unknowns involved in the focal phenomenon. Its description is based on small sample experience, and is informed by heuristics (e.g., gained through trial-and-error). The objectivist approach provides a maximum understanding (e.g., a set of mathematical equations) of a minimal set of factors (e.g., variables and events). This is done in an assumed ideal laboratory setting where items that are uncontrollable in the real world are assumed controlled in this theoretical world. It is mostly deductive. The latter anchor provides at least a minimally useful understanding of a maximum set of factors combining. This is done in a real-world setting where an attempt is made to build theory based on trying to recognize patterns in that world. In is mostly inductive (although sometimes abductive).

Based on the descriptions of the approaches, it follows that each approach has its limitations. Each involves a trade-off between how much reality to capture and how simply the focal relationship can be described. The more frictions and second-order effects and variables that are captured in the explanatory model, the more complex and difficult to optimize (or difficult to predict) the focal relationships become. A problem arises when one approach is held up as a straw man to argue for the other approach’s superiority as an explanation (e.g., without noting its own limitations). Unfortunately, that problem seems to manifest itself in several recently proposed partial theories.

Given that most of the theory in entrepreneurship uses one of the two approaches, we focus on the contrast between them. It is a choice between the mathematical, abstracted explanatory model of the objectivist approach, and the experiential, real-world explanatory model of the subjectivist approach. The objectivist approach often provides mathematical precision – e.g., explicit definitions, simplifying assumptions, and a focus on mostly isolated single relationships. The value of this approach is that it is very testable, accessible, and understandable, and provides a clearly bounded benchmark for further works (Such further work would add complexities in an orderly fashion, say, by considering specific contingencies to the focal relationship.) The limitations of this approach mainly involve its forced simplicity, based on explicitly excluding real-world complexities like additional variables and their interactions with the focal relationships. Thus, this model cannot provide a holistic understanding of a phenomenon. That said, it is an approach that is universally used in science, is precise, is objective, and isolates the hypothesized primary drivers of a phenomenon. By contrast, the subjectivist approach often involves experiential, descriptive richness of a real phenomenon as observed in the field. The value of this approach is that it provides detail, qualitative information, patterns at multiple system levels, and a big picture of a noisy phenomenon. The limitations of this approach mainly involve small sample size, subjectivity, possible interference with the system through the act of observation, susceptibility to noise, potential lack of immediate generalizability, imprecise definitions, unexplained black boxes, and imprecise boundaries.⁸ That said, it is an approach that is common in social science, provides a more holistic version of the phenomenon, and is more relatable to the practitioner.⁹

Of the two approaches, the subjectivist, experiential approach appears to hold the most danger of being exploited (although any data supporting either approach could be falsified). Description of a phenomenon is useful, but it is not scientific theory; it is only data (Whetten, 1989). Unless that description is translated into a story of relationships among factors (or sequences of events), no theoretical contribution exists. Besides the danger of never explaining the why of the focal relationships over describing the how, there is the danger of the subjective specification of the story based on what is normally proprietary data. This experiential approach is more susceptible to exploitation because it is often inductive and, thus, involves an open system, with imprecision and no immediate validation checks on the initial observations (e.g., Denzin & Lincoln, 2000). It involves a relatively younger methodology, making it more susceptible to unexposed issues.¹⁰ Further, the approach may be accompanied by a type of salesmanship that cannot be applied to an objectivist, more mathematical and deductive approach. The experiential modeling is often accompanied by rich details of the initial observations that appeal to senses rather than to cold scientific logic. This approach appears to leave open a temptation to describe a magic show rather than to explain the mechanics of the illusion behind it (e.g., the art of core skills in DCV micro-foundations as described by Teece, 2007). And that temptation is strong because the data from which proposed inductive theory is drawn are difficult to replicate and to test independently. As such, it is relatively more likely for less scrupulous researchers to let their self-interests guide the field observations they report, the strong patterns they find, and the novel explanations they propose. Such temptation appears more likely when the lead researchers accept a cultish position of the person who is an expert on the initial observations that led to the new proposed theory. Such cultishness is less likely in an objectivist, mathematical approach because there is a cleaner break of model from modeler, given that the premises used are almost always standard, accepted assumptions and, if not, are argued by drawing upon a wide set of empirically tested results where the modeler would not usually have absolute personal influence.¹¹

Testing proposed theories

We test three different types of recently proposed partial theories for newness based upon what is known about entrepreneurial activity and explanatory approach choices. We do so focusing on each type’s recent example in the literature, while commenting on why each has been considered as new theory and why. To be clear, the test for newness we conduct is not a test for contribution. Research that fleshes out an existing theory can be a contribution. (It could introduce an interesting contingency, or fill in some missing steps in a process, though, but it would not serve to differentiate the field).

We test bricolage, effectuation, and creation opportunities as the examples of the three different types of proposed theories. The first type details one MI variant occurring under a specific context. For bricolage, it is the underpricing of internally available factors (and/or their combinations). Bricolage occurs when the entrepreneurs uncover those arbitrage possibilities – the ones embodied in new, though often temporary, uses for those available factors – under conditions of necessity, as being the mother of invention. The second type of proposed theory details an individual’s journey in finding her MI to address, focusing on ambiguous contexts, and suggesting helpful heuristics to use in those contexts (e.g., heuristics for risk-mitigation, for information access, for adaptability, and so on). Effectuation is presented as a more realistic journey than that reflected in the simplified model of the existing basic process. The third type of proposed theory leverages the link between the MI and the heterogeneous entity, focusing on contexts characterized by higher levels of informational uncertainties. In contrast to the second type, there is less focus on any individual’s story and more focus on generalizable actions for the entrepreneur to take in her social world (i.e., the world made up of real people and their beliefs as consumers, suppliers, investors, partners, and so on). The partial theory of creation opportunities provides a contrast to the existing model in a different way than the other two types. It challenges the discovery assumptions of the basic process of entrepreneurial activity (e.g., that factors underlying MIs are objectively real and searches for them are optimizable).

While all three types of proposed partial theories differ in significant respects, they also share several common characteristics that helped make each of them highly marketable: First, they all confound their processes with the wicked research area of creativity. This is problematic for numerous reasons, including that creativity is an area of study that is much wider than commercial entrepreneurship, and so does little to help define the field. As well, creativity entails its own golden rule precluding optimization – i.e., that creativity can never be fully understood or else it would no longer be creativity (because everyone would do it and nothing would then stand out as creative). Thus, any proposed theory centered on the black box of creativity is likely to be widely interesting but also very unlikely to provide clear differentiation for our field. Second, they all choose the more inductive, observation-based, realism-rich subjectivist approach to explain the phenomena – i.e., where small samples are drawn upon to find discernable patterns of actions and outcomes. This leads to interesting narrative stories of active heroes who confront surprising obstacles on their paths to successes. Unfortunately, that may lead to a capturing of the process pieces and patterns based on surprise rather than on strategic importance. Third, they all sell their proposals by contrasting them to a straw-man super-rational model of the basic process, often keying in on the known limitations of such an abstract approach. This can lead to a downplaying of the limitations of the subjectivist approach, and even to some selective – and sometimes inaccurate – characterizations of existing explanations (e.g., to create theory as gap-filling). It may also indicate that the proposed theory cannot stand on its own merits under its own theorized boundaries. Fourth, all provide seductive new labels for their constructs or process steps that give an appearance of having more understanding of the unknown issues – like creativity – than is warranted. The admission of ignorance of such issues (i.e., of the creativity process) would be more helpful in guiding future research, but may hinder the marketing of the new proposals. Fifth, in the end, all redescribe the existing basic process outlined on pp. 23–28 – where an MI is addressed by a heterogeneous entity. None of the proposals provides propositions that lie outside this known story.

Prior to testing those proposed types of partial theories, we describe a set of premises that formalize relevant parts of the existing explanations that we can then draw upon to explain specific testing analyses (again, based on existing, common understandings in the field).

• Premise 1: one cannot make something from nothing – addressing any MI entails combining existing factors, regardless of whether those factors are easy or difficult to find, or are tangible or intangible.
• Premise 2: all for-profit MIs are of known types – addressing MIs for profit, by definition, entails known types of goals – i.e., to reduce costs, to increase price or quantity, or to combine of these activities to produce new monetized value.
• Premise 3: entrepreneurial success is explained through the relationship between the addressed MI and the focal heterogeneous entity – a specific success must rely on some recognizable differentiated advantage of the pursuer of the MI that allows that MI to be addressed in a manner that provides new profits in a competitive environment.
• Premise 4: all MIs are addressable by several entities ex post, using existing factors – in order to be economically commercializable at a reasonable scale by the entrepreneur’s firm or licensee, the ability to address the MI must be replicable by others, or able to be routinized.¹²

Testing the first type of proposed theory, using the example of bricolage: this is the most straightforward test, given that the concept of bricolage is admittedly an established idea (Baker & Nelson, 2005; Levi-Strauss, 1966). Providing details about the process of how a specific type of MI is addressed in a specific context (i.e., a permissive one) by differentiated entities (i.e., through enacting) is a contribution, but it does not constitute a new theory. Further, the underlying creativity involved in combining immediately available factors to make something new, spurred by necessity, remains a black box. Although the entrepreneur is more active in discovering and exploiting the arbitrage opportunity than in a traditional, simplified model, the pieces and relationships don’t change regardless of whether the story has become more exciting from references to successful visible cases (e.g., the Danish wind turbine industry – Garud & Karnøe, 2003).

Testing the second type of proposed theory, using the example of effectuation: given this proposed theory has been recently comprehensively critiqued regarding its contributions to the entrepreneurship field (Arend, Sarooghi, & Burkemper, 2015), we will keep the testing here relatively short. Effectuation applies an interesting wedge involving Premise 3 to leverage the necessary link between the MI and the individual (Sarasvathy, 2001). It does so to support the notion that every addressed MI emerges only because an individual took action. Under that notion, effectuation can then logically focus on the individual herself and her unique path to actively finding her own MI, which makes for some very interesting and marketable stories. Unfortunately, it does not make for new theory given that the basic process remains the same as captured in the existing model. There may be some contribution in the focus on ambiguous contexts, where traditional, simplified decision-making rules do not initially apply, but where heuristics and alternative approaches like experimentation are more applicable to reduce the uncertainty into risk. However, there may also be some confusion in the approach. The implicit assumption that every individual has her own guaranteed profitable MI is actually not implied by the premise that every profitable MI has its individual. Further, the creativity involved in how the individuals successfully exploit contingencies to make a new artifact remains a black box, regardless of the risk-mitigation, cooperation-focused heuristics prescribed. There do not appear to be any propositions emerging from the proposed theory that conflict with the existing basic process of an MI being addressed by a heterogeneous entity. Effectuation describes different prescriptions to use in more uncertain environments than in more certain ones, but that is not new (e.g., experimentation is a well-known approach to reduce uncertainty), nor does that alter the pieces of the basic process or the relationships involved regardless of whether the individuals seem more involved in uncovering the relevant information to address the MI.

Testing the third type of proposed theory, using the example of creation opportunities: this example is not old like bricolage, nor has it been comprehensively critiqued like effectuation, and so we focus more attention on it here. This example also applies the Premise 3 wedge to link the individual to the MI, but does so for a different reason. It does so to focus on the importance of the social interactions of the entrepreneurs to other relevant people in identifying and addressing MIs (Alvarez & Barney, 2007). This also means less of a focus on personal utility as an outcome and more of a focus on profitability (i.e., on the goals outlined in Premise 2) compared to effectuation, making this type of proposed theory closer to the existing model. Despite that proximity, the proposed theory argues its differentiation based on a questionable assumption that violates Premise 1. The assumption is that different factors and relationships describe the basic process when an MI is more difficult to address – i.e., when the informational requirements are harder to uncover through social interactions, when there are more factors that are needed to combine and some of these are intangible, and so on. Essentially, the assumption is that a different theory is needed to explain the basic entrepreneurial process when it involves creating an online community-based brokerage like eBay than when it involves wild-cat drilling to discover oil (i.e., a different theory is needed to create something from not quite nothing – from intangibles or from harder-to-do tasks). The main explanation is a philosophical one – one that may resonate with academics but would be meaningless to practitioners (which poses somewhat of an irony for a proposed theory that is supposed to be more realistic). The explanation is that, while contrasting MIs labeled discovery are based on objectively real factors (e.g., existing oil deposits), those labeled creativity are based on socially constructed factors (e.g., on belief-based transaction-supporting systems and factors like currency and stocks). Of course, all of these factors are equally real to the real entrepreneur; she knows that her digitized money will buy other goods just like her analog oil will.

While philosophical differences may exist over types of factors, those do not translate into the necessity for either a different process or a different theory (Ramoglou, 2013). The process is the same because all MIs are addressed by individuals (Premise 3) and thus all involve social interaction and reliance on socially constructed factors. In other words, a continuum exists that describes the characteristics of the factors and steps involved in the process, but it remains as only one process of combining existing factors to address an MI in a differentiated fashion. While some MIs may involve a more complex combination of factors, and more intermediate steps to reduce the informational gaps (e.g., through experimenting in order to reduce uncertainties into risks), the pieces and relationships remain the same. There is no new theory in the example of creation opportunities.

As with the other types of proposed theory, a subjectivist approach has been used effectively to provide rich, marketable stories of entrepreneurial journeys and to identify patterns in the complex actions taken (although with more deductive and philosophical discussion than with effectuation). As with the other examples, the black box of creativity remains unopened, although this example highlights the type of ingenuity used by entrepreneurs in their social activities to alter the beliefs of others to market their products and to solicit the true beliefs of others to understand the possible states and probabilities of new commercial territories. Although unopened, the patina of a greater understanding of creativity is provided in the labeling of the shocks that initiate this process as endogenous versus exogenous (Barreto, 2012). As explained previously, such a division is unjustified, as both internal and external changes drive all attempts to address an MI when it is actually addressed. The creativity is hidden in the suggestion that the entrepreneur manufactures the shock through her own actions, endogenously, to uncover the MI. That relabeling of creativity is unhelpful, and confuses types of factor characteristics (e.g., factor tangibility) with effort characteristics (e.g., their physical, social or mental dimensions; their informational load; and, so on). And, as with the other examples, there is a contrasting straw man version of existing theory to compare against – here, of so-called discovery opportunities (i.e., those based solely on objectively real factors).

If there is some contribution to existing explanations provided by this example, it is in the detailing of the deeper issues involved in the MIs (and the factors) that take more effort, are less obvious, have greater informational problems, and so on. However, any such contribution appears to be based on some shaky ground. First, as noted earlier, the reference to the term social construction is unusual because the size of the society needed to share beliefs initially to address the MI is likely small. Regardless, building shared mental models within a network is not a new phenomenon and it only draws on existing theory (e.g., sociology). Second, the proposed contrast to discovery appears to be inaccurate at best. The Austrian economists (e.g., Hayek, Kirzner, Schumpeter) who describe versions of the concept all appear to describe a process that is not the rote search procedure alluded to in the creation opportunities literature, but instead a process that is more creative (Kor, Mahoney, & Michael, 2007). For example, as Kirzner (1997, p. 72) explains: “the notion of discovery, midway between that of the deliberately produced information in standard search theory, and that of sheer windfall gain generated by pure chance, is central.” In other words, this type of proposed theory may be more effective in providing a contribution to existing theory by specifying what its boundaries are (e.g., regarding the reality of factor types involved) and by accurately using existing labels (e.g., about what discovery is, what social construction is, and what an opportunity versus a business idea is).

Discussion

We now discuss the preceding testing of proposed theories in entrepreneurship. We comment upon the potential for harm to the field’s theory-building when newness is misidentified (e.g., lost legitimacy), and propose possible ways to mitigate the occurrence of such misidentification (e.g., acid tests for newness that editors could apply). We suggest alternative approaches for creating new theory, and we provide an example to prove that this feat can be accomplished. Table 3.1 summarizes the testing of the three proposed theories we explained on p. xxx, and the differences of the new theory example we describe on p. xxx.

Table 3.1 Proposed theory characteristics and testing

Are there harms from popularizing the proposed partial theories as new when they are not? While all three types of proposed partial theories, as illustrated by their recent examples, provide contributions to the existing model, none provide a new, alternative theory. So, while each provides value to the field, it should be noted that each may also entail harm when any one of them is mistaken as an alternative theory. First, there may be harm caused by adding confusion about the basic process. That may lead to misallocations of research resources, to cannibalistic relabeling of concepts, and to misleading prescriptions to practice. Second, there may be harm to the legitimacy of the field when such false starts occur repeatedly. Outsiders start to question the motivations and abilities of the gatekeepers, critics and academic bodies to do their job to properly vet new proposals. This may lead to questions about who owns the field if no one seems responsible for, at a minimum, ensuring proper critical debate over new proposals. To be clear, the continued proposal of new partial theories is healthy for the field, as it provides potential contributions to existing theory and also possibly leads to alternative views. However, it needs to be tempered with timely, multi-perspective and critical reviewing in order to help test each proposed theory’s differences and to identify the true contributions. And, this seems especially necessary for inductive theory proposals where the main assumptions and relationships are relatively less clearly defined and entail less historic methodological context.

Possible solutions: To avoid the unjustified promotion of proposed new theory, two things need to exist: i) a fair acid test to vet new theory; and ii) a set of responsible parties in the field to apply the test early. To simplify the testing, any proposed new theory needs to be presented in a form that clearly states its differences (in propositions), its foci (of core relationships), its boundaries, its assumptions, its detailed definitions, its modeling choice, how it should be tested, and how it stands alone as an alternative explanation of relevant phenomena. Prior to allowing its authors to repeat its message (versus to test it) in subsequent outlets, at least one critical assessment (beyond what the reviewers and editorial team did) should be published by independent scholars. The acid tests should involve ensuring that the proposed theory meets the criteria for standard theory-building (e.g., Dubin, 1969), and that it provides clearly differentiated relationships, some of which are in the form of propositions that conflict with existing theory. As for responsible parties, perhaps there is a need for an explicit code to be written and then followed in the field by those with more power (e.g., journal editors), enforced through peer pressure. In other words, instead of moving the field, through the use of power, towards a particular view or model choice, instead we should at least agree on a framework for the process for debating (and evaluating) the viable options proposed for that movement.