Most directors today recognize the importance of robust oversight, but it is unclear whether boards, as they are currently constituted and operate, are up to the task. The increasing size and complexity of companies, the expanding array of risk areas, and the difficulty boards have in getting the information needed to exercise effective oversight all bode poorly for a positive answer to this question.

— Paine and Srinivasan (2019, p. 16)

Introduction

The theories underlying modern corporate governance practices have grown stale. Whereas modern corporations have found ways to successfully expand far beyond the scope of traditional theory, our theories of governance have failed to keep pace. In fact, despite explicit recognition of and ever-expanding research on governance problems—issues as theft, fraud, and other misconduct (Hersel et al. 2019)—these wrongdoings continue to plague businesses at essentially a steady rate. It seems that despite firms’, regulators’, and scientists’ efforts, increasingly complex governance mechanisms are met with increasingly complex ways of hiding from or skirting these efforts. Thus, as Paine and Srinivasan note in the introductory quotation, it seems doubtful that corporate governance theory, as presently conceived, is up to the task of such oversight.

The theoretical impetus for the present-state corporate governance structure has been, primarily, Jensen and Meckling’s (1976) foundational work on Agency Theory, wherein agency problems are unpacked, the role of principal monitoring is developed, and other solutions, such as ‘bonding’ (i.e., the bondholders’ use of external auditing services) or compensating agents with ‘inside equity’ to align incentives and mitigate the desire for misconduct, are proposed. While we can certainly agree that the theoretical insights purveyed by Agency Theory are important and useful, it is also quite clear by now that the implications drawn therefrom have been largely unsuccessful in resolving those agency problems. It may be worthwhile at this point to start afresh, to reassess the role of governance and the problems it deals with using a distinct (meta-) theoretical lens.

Our goal for this article is to build a new theoretical framework for corporate governance. Motivating this effort is, in large part, recent developments in the philosophy of science, specifically regarding social scientific meta-theory. As social science continues to grapple with the reproducibility crisis (Bergh et al. 2017; Camerer et al. 2018; Open Science Collaboration 2015), growing work in social ontology is pushing management theory onto new foundations that suppose social science to be of a distinct nature and character than the natural sciences (McBride 2018; Packard 2017; Packard and Bylund 2021; Tsoukas and Chia 2002). Fortunately, essentially new foundations are not required of us, as various social scientists, especially from sociology (Dilthey 1989; Ricœur 1981; Weber 1978, 2011) and economics (Menger 2009; Mises 1962, 1981), have long been sounding this alarm and have laid subjectivist foundations for the social sciences that are strong and apposite.

Thus, we build our new foundations for corporate governance atop the subjectivism of the Austrian School of economics. To do so, we adopt the ‘praxeological method’ of theory building (Hoppe 1995), a rationalist methodology developed within the Austrian school whereby theoretical conclusions are deduced a priori from essentially true first principles. From these foundations, along with various insights already devised by the Austrian school (Hayek 1937, 1945; Mises 1998), we elaborate a new theory of agency, organization, and governance that concludes with the important insight that growing organizational complexity requires greater dispersion of governance mechanisms rather than more centralized oversight. In the simplest of terms, we turn to a bottom-up approach to governance theory building to either replace or complement the top-down approaches that comprise almost the entire body of modern governance theorizing. We elaborate this insight into specific implications and propositions for corporate governance.

‘Crusoe’ methodology

Let us begin our theory-building effort with a brief explanation of our methodology, which has become rather atypical. Our approach is deductive, building upward from basic axioms and assumptions, rather than inferentially downward from empirical observation. It is the so-called Crusoe methodology, a classical approach to theory building in which a theorist begins with the simplest form of economy—the economy of one—to establish the basic economic mechanics and then add complexity thereto (Söllner 2016). While virtually all classical schools of economics have established their basic premises with this method, the Austrian school has formalized and expanded upon it.

The praxeological method

The foundations for our methodology, which has been termed the ‘praxeological’ (Mises 1962, 1998) or ‘Austrian’ (Hoppe 1995) method, are in self-evident axioms. Kant (1998) observed that there are several necessary syntheticFootnote 1 truths that cannot be refuted without contradiction—basic principles, such as principles of causality and spatio-temporality, are necessarily known or deduced a priori or else no other understanding could be derived therefrom. One such necessary synthetic truth, Mises (1962, p. 6) asserts, is the ‘cognition of action, that is, the cognition of the fact that there is such a thing as consciously aiming at ends.’ This ‘human action axiom’ asserts ‘the immutability and universality of the categories of thought and action’ (Mises 1998, p. 35):

This axiom, the proposition that humans act, fulfills the requirements precisely for a true synthetic a priori proposition. It cannot be denied that this proposition is true, since the denial would have to be categorized as an action-and so the truth of the statement literally cannot be undone. And the axiom is also not derived from observation—there are only bodily movements to be observed but no such things as actions—but stems instead from reflective understanding. (Hoppe 1995, p. 22)

From this human action axiom and the axiom of causality (Kant 1998), we can then deductively construct economic propositions, both necessary (if they fully derive from a priori true propositions) and contingent (if they rely on one or more contingent states of affairs). Contingent propositions are still logically ascertainable, but only partially so and must be assessed under the explicit assertion that the contingency is given. For example, there are categories of human action, such as consumption. Not all actions are consumption, and so to study consumption via this method, we must understand the context in which consumption is the given action and construct a consumption function logic atop the givenness of consumption (i.e., given that an action is consumption, what are its nature and consequences?).

In following this method, we begin with basic axioms and build them logically upward into contingent propositions, building as much as possible from those necessary truths to inform the nature and consequence of any presumed state of affairs. Insofar as we maintain strict reliance on necessary states of affairs, and provided we maintain logical integrity, the conclusions we deduce are contingently true or, at the very least, logically plausible (depending on the nature of the contingencies included).

Basic premises and definitions

To begin our theoretical development, we must first put forth some basic foundations and essential definitions. Foundationally, we build from the human action axiom, from which we must extrapolate basic categories of action. Mises (1998: 252–257) elaborated various subcategories of the human action parent construct, which would include the action functions of consumer, owner, investor (capitalist), entrepreneur, manager, and worker. All intentional (economic) action, Mises argues, is encapsulated by one (or more) of these economic subfunctions.

More recently, Mitchell et al. (2021) refine these categories into two parent action functions: ownership, and governance. Under the ownership category are the entrepreneurship and consumption functions, which have their own subfunctions. Specifically, the entrepreneurship subcategory has, under it, the management and labor subfunctions. Under the governance parent function is the oversight subfunction. Each of these action categories requires definition for our theory-building.

Defining the categories of action

Following Mises (1998), Mitchell et al. (2021) define the distinct categories of human action as ideal types characterized by their economic function. The ownership function entails ultimate right of control over a resource. Ownership is obtained by taking a resource out of the ‘state of nature’ (i.e., homesteading; Locke 1689) or by just acquisition through voluntary exchange. By right, other claims to the control of an owned resource are superseded by the owner’s. Control or authority over an owned resource can be delegated to another by its owner, but the owner retains the ultimate right to revoke such authority at their discretion.

The primary and ultimate end of owning and using resources is the consumption subfunction (Hutt 1990) of ownership, which we define as the expenditure of resources toward an intended gain in subjective well-being (Packard and Bylund 2021). The second and secondary ownership function is the entrepreneurship (or producer) subfunction, which is tasked with judgment over the preferred productive use of resources and the construction of specific plans for their use (Foss and Klein 2012). Under this entrepreneurship subfunction is the management sub-subfunction, which is tasked with the enactment of the entrepreneur’s judgment and plans. Under the management sub-subfunction is the labor sub-sub-subfunction, which is tasked with the execution of specific production activities.

The governance function is not an ownership subfunction (although it is often performed by owners), but ‘entails both the identification and acceptance (or not) of involuntary boundaries [of action], as well as the determination of what voluntary boundaries to impose on themselves’ (Mitchell et al. 2021: 13) or, in other words, the conscribing of the ownership function to certain boundaries, whether moral, strategic, or political. As part of this function, the oversight subfunction polices and enforces these boundaries.

These definitions provide the foundations and clarity necessary for the following theoretical elaboration.

Rebuilding corporate governance theory

Having explicated our rationalist methodology and set forth our foundational definitions, let us now proceed with the theory-building process. Our starting point is, again, Mises’s human action axiom—‘the proposition that humans act, that they display intentional behavior’ (Hoppe 1995, p. 61), with the categorical functions just defined. Admitting the methodological individualist position that socioeconomic activities are comprised of intentional individual actions and their intentional interactions, we can begin our theory construction with a standard Robinson Crusoe scenario.

Autarkic self-governance

Let us suppose that an actor (e.g., Crusoe) is found in the ‘state of nature,’ i.e., alone and isolated. Although it is tempting to leap to a conclusion that such a person is outside of economic analysis, such a conclusion would be rash. Indeed, the ‘autarkic economy’ or economy-of-one is just as real as the ‘catallactic’ or social economy within which it is embedded (Packard 2020). ‘Economy’ is defined as ‘the productive pursuit of needs satisfactions through purposeful action’ (Menger 2007; Packard 2020, p. 398). Thus, a single, isolated actor acts economically, producing supply to satisfy demand, despite there being no market exchange.

Assuming the Robinson Crusoe scenario, then, what can we say about corporate governance? Does a single actor have need of such? Clearly, the qualifier ‘corporate’ becomes moot and can here be discarded. But is there autarkic governance? Certainly, we are prone to talk about self-governance, e.g., in discussions of human psychology and self-discipline, but let us be precise. Is ‘self-governance’ literal or metaphorical?

Recall that ‘governance’ is here defined as that function which conscribes the ownership function. Socioeconomically, the principle of ownership in the autarkic context is trivial, if not useless. In autarky, there can be no competing claims for control over resources, and so all resources can be thought to be controlled and, thus, owned by the autark. However, in a stricter Lockean sense,Footnote 2 such resources are not ‘owned’ by the autark until taken out of the state of nature to be used. Robinson Crusoe does not own the whole island, even though he has it all to himself, but owns only those things he has taken from it and used. Although this may appear a trivial distinction, it will prove a necessary foundation as we begin to build theoretically upward.

Within the ownership function is the entrepreneurial function, which concerns judgment over the optimal use of owned resources. Intuition may rebuff the idea that Robinson Crusoe could be an entrepreneur. However, Packard (2020) argues that autarkic entrepreneurship is in fact a critical and vastly overlooked aspect of the modern economy and, of course, of the Robinson Crusoe economy. If entrepreneurship is ‘the intentional pursuit of new economic value’ (Packard 2017, p. 544), then that function can be performed by Robinson Crusoe or any other actor regardless of whether or not they make additional economic gains through a division of labor and market exchange. To elaborate somewhat on the idea of autarkic entrepreneurship, we recognize that intentions are rarely well-established, priority-ranked, or clear-cut, but are instead often fuzzy, temporally dependent, nested, and often contradictory. The task of the autarkic entrepreneur, then, is to allocate resources, including time and effort, to the highest-valued ends or, more specifically, to the tasks that will bring about the greatest value over time.

Also, because they are performed by the self-same person, the subfunctions of management and labor also appear trivial, subsumed by the entrepreneurial function. In other words, because Crusoe is the only one to act, he must not just make judgments and plans but also enact them. Yet, in a strict sense, we can still distinguish the various roles. Autarkic management is thus, simply, the temporal allocation of one’s own knowledge, skills, and efforts toward enacting predetermined entrepreneurial plans. Autarkic labor is, of course, performing those self-planned and self-managed activities.

Returning to the question of autarkic governance, then, self-governance can now be defined as self-imposed constrains on their own ownership decisions. Said differently, Crusoe might judge it imprudent to spend all of his scarce time in endless experimentation in search of better solutions to his various needs. Were Crusoe to have such a proclivity, he would need self-discipline to allocate his time and resources effectively toward, foremost, ensuring survival and some minimum level of well-being and, then, whatever additional time available toward capital accumulation and innovation as investments toward greater future productivity and well-being. It is this conscribe and constrain function that is the essence of governance.

Two-person orders

Let us now expand our analysis to the two-person context. The introduction of a second person allows social interactions that can increase learning and productivity, but also introduces meaning to the concept of ownership. Because resources are scarce, and because consumption of resources may entail their destruction, the ownership of resources becomes a critical and potentially contentious issue here. Thus, whereas governance in the Robinson Crusoe example was an issue of mere self-governance, the addition of another requires governance to include both self-governance and other-governance, or governance of interactions between the actors. Such expanded governance can take many forms, each a variant of five basic forms: autarky, collective, cooperative, market, and hierarchy (see Table 1).

Table 1 Types of two person orders
Full size table

Separated autarkies

The first possible social arrangement of two economic actors is economic isolation—that is, they can each have their own properties and pursue their own affairs independently of each other. In this case, the economic structure is generally the same as above, but the concept of ownership becomes a much more central issue. Where two autarkies are close in proximity, questions of ownership can become contentious. Ownership implies the right to exclude or prohibit another from its use, whether in production or consumption. Under such circumstances, trespassing, theft, robbery, and other property disputes and violations become potential issues.

The task of ownership in this case is not constrained merely to production and consumption, as in the sole autarky scenario, but must now also include protecting that which is owned due to the possibility of property rights violations. Thus, the role of both ownership and of governance are expanded to include potential interactions with the other. In this specific case of uncooperative autarkies, other governance focuses specifically on property protections. Such protections may include legal innovations (e.g., agreements), defense innovations, policing activities, trust-building efforts, and so forth. Because of these costs, separated autarkies are comparatively inefficient.

Collective

A second possible arrangement is a collective, where individual ownership is partially surrendered, some or all alienable (i.e., non-personhood) resources instead being conjointly ‘owned.’ Because ownership is herein understood as the control of a resource, we do not mean by conjoint ownership that one or the other actor cannot control a resource without the permission of the other, although such an arrangement might be made. Instead, what conjoint ownership means is that the rights to exclusion implied by ownership are removed. Thus, like the single autarky scenario, the notion of ownership here is trivial.

The abolishment of private ownership entails significant ownership and governance difficulties in dealing with scarcity and resource consumption. The collaborative actors must somehow allocate those scarce resources aptly—determining both what they should and should not be used for—according to their conjoint objectives and charge, such as to minimize conflict between them. Such collective ownership and governance are difficult, as decisions regarding resource allocation must be done conjointly or, else, one must defer to the other. This may cause problems of incentives and equity.

The governance of collective resources is further impeded by a key incentive problem, colloquially referred to as the ‘tragedy of the commons’ (Hardin 1968). This references the incentive problem that arises when productive efforts are untethered from the consumptive benefits that such efforts engender. When one does not reap the benefits of their own productive efforts, for better or for worse, they are disincentivized to put forth additional effort. In a collective, this is naturally the case where productive capacity is not equivalent between partners or where their individual consumption needs vary. While perfect equidistribution may seem the likely candidate for resource purveyance, such a distribution may not be equitable in terms of its outcome, especially when individual needs and capabilities vary significantly. A collective of both a physically large actor and a small, petite actor may imply an imbalanced distribution of foodstuffs to provide full and proper aliment to each person. But what such a distribution ought to be can be difficult to ascertain. For example, there are significant hurdles in understanding (Packard 2019; Witt 2001) and communicating (Packard and Burnham 2021; Polanyi 1962) idiosyncratic needs. What one can understand of another’s needs is comparatively little and difficult to ascertain. Thus, where the collective is not universally relieved of strongly autarkic motives, the governance function necessarily breaks down and production withers.

Cooperative

The third possible arrangement is a cooperative, where economic production is cooperatively conjoined. In a cooperative, ownership of property remains autarkic, but the entrepreneurship function and the governance of production does not. Such forms of organization facilitate more complex production activities, as many tasks are too difficult or complex for a single actor to accomplish alone, and even more simple tasks can often be performed more efficiently through a division of labor (Durkheim 2014; Smith 2007). Management of more complex projects may (but need not) require that one actor take on the managerial role, while the management of simpler projects will tend to be done conjointly.

The principal problem overhanging this form of organizing is the sharing of the entrepreneurship and governance functions. That is, who in the partnership determines what is (and is not) to be produced, as well as when, and how? Thus, this organization runs into similar challenges as the collective, as it is not always clear who merits control over production efforts under what circumstances. In an ideal case, the decision should be unanimous but, in reality, the idiosyncratic needs of the partners are in tension. Thus, should entrepreneurship and governance be allocated according to knowledge and ability or to need, and how would such factors be fairly ascertained? Even if an acceptable balance of the entrepreneurship and governance functions could be struck, the order of productive operations may also induce tension within the cooperative, as the partner who benefits first may benefit more and longer than the partner who gets control of production later. Here, Hayekian knowledge problems (Hayek 1945) begin to arise—each partner’s knowledge is distinct, much of it tacit (Polanyi 1958), which makes efficient governance and management of the cooperative by one or the other necessarily inefficient, the active governor unable to efficiently employ the other’s knowledge.

Finally, because property is individual, there may be need for personal property protections, as in separated autarkies. However, because violations of properties would upset partners’ willingness to participate in the cooperative and could thus lead to the less-efficient order of separated autarkies, there is some natural dampening of the incentive to violate the partner’s property claims.

In short, a cooperative is viable but inefficient, and must rely to a great extent on selflessness in order to be sustainable.

Market

A fourth possible organization is a market order, which entails autarkic entrepreneurship by both partners, the individual outputs of which can then be cooperatively exchanged in mutually beneficial trade. Thus, both production and governance remain autarkic, while engaging in limited cooperation post-production to gain the benefits of a division of labor.

The benefits of a market organization are highly similar to a cooperative, while resolving the problems of conjoint entrepreneurship and governance by leaving the entrepreneurship function autarkic, while the governance function is intersubjective. By ‘intersubjective,’ we mean that the parties, while free to self-govern, may voluntarily agree to some institutional rules to govern the exchange relationship, which are overseen and enforced individually. Because the entrepreneurial function is autarkic, complex productive tasks that require the efforts of both actors may become difficult, requiring negotiated agreements that can be difficult to reach when the value output is not easily divisible. Intersubjective governance also introduces transaction costs to the order (Coase 1937), which can reduce total economic efficiency under certain circumstances.

The market order also necessitates the same need to protect one’s property that is observed in separated autarkies and in cooperatives. But again, there is a natural dampening of the incentive to violate properties in order to maintain the more efficient market order—violations would result in the comparatively inefficient separated autarkies.

In general, economists have found market orders to be comparatively efficient in most economic circumstances, and generally more efficient than the three previously outlined organizational forms. However, when transaction costs are high, they can cause inefficiencies that may be overcome with a hierarchical form of organization (Williamson 1967, 1975).

Hierarchy

The final possible arrangement of two actors, then, is a hierarchy, such as employment. In the hierarchical relationship, ownership is autarkic,Footnote 3 but one partner is subjugated beneath the other such that only one performs the entrepreneurship and governance functions, while the other performs subservient duties. The subjugated actor (agent) acts at the behest of the entrepreneur-governor (principal), who grants to the agent some compensation for their productive efforts.

The hierarchical form occurs when one actor voluntarily subjugates themself to the governance of the other. One might subjugate themself in a hierarchical form, rather than some other form of organizing, for at least three reasons: (1) if the agent is less talented or capable than the principal and, thus, would do better under the principal’s leadership and tutelage (i.e., apprenticeship), (2) if the two actors’ risk preferences vary significantly, or (3) if the principal owns superior properties such that the agent would benefit more from such a relationship than via any other form. First, the apprenticeship organizational format is effective where one possesses highly valuable knowledge and skills that the other does not. This organizational form allows the ‘apprentice’ to learn these skills over time, while giving the ‘master’ significant gains from the division of labor in the meantime. Second, one may voluntarily subjugate themself in an employment relationship in exchange for a guarantee of income, any additional gains (losses) being kept (surrendered) by the principal (entrepreneur) (Mises 1951). Finally, employment may be chosen out of necessity, where a cooperative is infeasible due to significant differences in owned resource. For example, if actor A owns fertile land and actor B’s properties are infertile, B may voluntarily enter an employment arrangement to reap benefits from A’s properties. Subjugation is likely to be preferred in this case only if B cannot find some other productive activities that could engender a sustainable market arrangement.

The hierarchical form introduces the well-known principal-agent problem and the mechanics of agency theory (Jensen and Meckling 1976). More specifically, hierarchy is faced with critical challenges due to the idiosyncratic motives (Menger 2007; Rothbard 1956) and knowledge (Hayek 1937, 1945) of the principal and agent. This agency problem—that ‘agency conflicts arising from a divergence between agents’ and principals’ utility functions, creating potential for mischief’ (Lan and Heracleous 2010, p. 294)—has been widely discussed in both the agency theory (e.g., Bosse and Phillips 2016; Eisenhardt 1989; Jensen and Meckling 1976; Lan and Heracleous 2010) and transaction cost (e.g., Argyres and Zenger 2012; Bylund 2021; Klein 2010; Williamson 1994) literatures. It is also the primary impetus for arguments for decentralization, i.e., the relegation of decision authority to lower-level employees (Foss et al. 2015; Hempel et al. 2012).

Also and less widely recognized, Hayekian knowledge problems are not constrained merely to interorganizational interactions, but applies equally to intra-organizational processes—i.e., there are ‘internal Hayekian knowledge problems’ (Ng 2020, p. 464). These problems arise within hierarchy from the fact that the principal cannot fully know and thus exploit the idiosyncratic knowledge of the agent, and that the agent cannot effectively employ their own knowledge and skills most effectively toward the principal’s ends due to a thick tacit knowledge barrier. All that can be communicated is an explicit task and its purpose, which are detethered to some necessary extent from the tacit needs and aims underlying the requested task. This knowledge barrier must inhibit the agent from employing the entirety of their own knowledge toward innovative solutions to the principal’s needs. The principal is also vastly limited in their innovative capacity to only that knowledge they themselves possess as well, perhaps, as a limited amount of knowledge about the agent’s knowledge and skillset, curtailing possible tasks to a limited and inefficient set. As a result, hierarchical governance generally leaves a vast trove of knowledge unused. This problem also inhibits the principal’s knowledge of the agent’s needs, which may inhibit satisfactory compensation for work done on the principal’s behalf, which can lead to the breaking of the relationship.

Conclusions

Of the five possible economic arrangements of two collocated actors, each has key drawbacks that generally derive from the benefits and complications of sharing versus dividing the entrepreneurship and governance functions. The prevailing view today generally holds that markets and hierarchies are the most economically advantaged and least disadvantaged, and so these tend to attract the bulk of scholarly attention (e.g., Anderson and Brown 2010; Bylund 2014; Halevy et al. 2011; Halevy et al. 2012; Williamson 1973, 1975). Most modern economic production in fact occurs via some hybrid of these organizational forms (Foss 2003; Makadok and Coff 2009), to be elaborated later.

Adding a third actor

Let us continue our deductive analysis by adding one more actor. A third actor brings up several new problems and possibilities. Moving forward, we will not further elaborate the less efficient autarky, collective, or cooperative forms of governance, as we admit that an efficient order would not take any of those forms. We will merely observe that adding additional actors to these forms only further exacerbates their weaknesses (although a third would also augment a collective and cooperative’s benefits of task complexity). In contrast, adding additional actors can augment the efficiency of market and hierarchy orders, which will thus tend to be preferred.

Market

Adding a third actor to a market form, in contrast to the other forms, significantly magnifies its strengths, while only moderately exacerbating its weaknesses. Specifically, the complexity of productive tasks that the market could achieve is increased, and the division of labor increases. Hayek (1988) argues that it is no coincidence that the most densely populated societies also enjoy the highest average standard of living, as the efficiencies of a division of labor intensify exponentially.

The problems of markets also grow with increased market actors. Primarily, the threat of property infringements increases with more possible malefactors. This may divert greater productive effort toward oversight efforts, such as property protections, which can reduce market efficiency. Certain transaction costs (e.g., search costs) may also increase somewhat with a growth in market participants, but such increases are not foregone.

Hierarchy

Adding a third person into the hierarchical form exacerbates both the agency problems and the Hayekian knowledge problems, as it introduces a third idiosyncratic will and more idiosyncratic and tacit knowledge. It also introduces new ‘structural’ opportunities. First, the hierarchical structure may remain ‘flat,’ i.e., the principal may govern both agents as their manager and supervisor. The management task entails judgment over the resources and tasks assigned to the agents. The task of supervision is the monitoring of activities to limit shirking (Jones 1984), partially alleviating the agency problem. A second option is for the function and subfunctions of entrepreneurship and/or governance to be delegated, creating a ‘tall’ or ‘vertical’ hierarchy. This offers different structural configuration possibilities. For example, one agent may be tasked with the management and/or oversight of the other, the principal maintaining the roles of entrepreneur and of governor over the former. Alternatively, the tasks of entrepreneurship might be delegated, the agent effecting ‘derived judgment’ (Foss et al. 2007) over the principal’s resources, including the other agent. By delegating the entrepreneurial task to a ‘middle manager’ and granting them authority to make entrepreneurial judgments, the internal Hayekian knowledge problem is partially alleviated, but only to an extent. The middle agent still cannot know the lower agent’s full knowledge, nor can they fully know the principal’s mind and intent.

The problem of production

Continuing our analysis, we will build our governance theory atop the theory of the firm. Bylund (2016a) advances the basic insights of ownership and entrepreneurship, the division of labor, and the separation of knowledge into a theory of the firm based in an organizational problem he terms the ‘specialization deadlock.’ In short, the specialization afforded by the division of labor is limited by the extent of the market. Contract laborers are incentivized to be generalists in order to attract many bidders. Specialization in complex and specific skills can have extremely high asset specificity. Thus, specialization itself suffers from the ‘holdup’ problem (Williamson 1985). While this is not a problem in simple markets, advanced societies run into this holdup, the specialization deadlock, quite often.

As entrepreneurs innovate increasingly complex solutions to consumer problems, the production of such complex solutions becomes extremely costly. Their costs can be severely curtailed by an advanced division of labor, but this runs into the specialization deadlock problem. Contracted laborers are unwilling to develop such asset-specific skills without some guarantee of long-term employment. Thus, to overcome this specialization deadlock, an entrepreneur may form a firm as an ‘island of specialization’ (Bylund 2016a, p. 6), which is in essence a ‘nexus of contracts’ of employment (Jensen and Meckling 1976; Kim and Mahoney 2010). As such, the entrepreneur promises guaranteed, long-term wages in exchange for the laborer’s development and application of ‘firm-specific human capital’ (Hashimoto 1981), assuming on themselves the risks of entrepreneurial loss (Mises 1951).

In short, in large societies, the typically optimal economic organization is a market economy, with increasingly large and complex firms (hierarchies) as islands of specialization. But here, the Hayekian knowledge problems of hierarchy previously described become a central issue.

Knowledge as an organizing problem

Knowledge is always and necessarily individual (Polanyi 1958), and it is always at least partially subjective (Scheler 1980). There is no ‘social knowledge’ and what is widely known is always known individually through a unique, subjective interpretive lens (Gadamer 2006; Ricœur 1981). For example, while we all know that ‘the sky is blue,’ it can never be ascertained whether the experience that another has of ‘seeing the blue sky’ perfectly reflects one’s own self-same experience (e.g., do we all see the same ‘blue’?). Because of this, subjective knowledge is necessarily tacit, i.e., it cannot be communicated, at least not in its entirety (Polanyi 1962). Beyond the problem of literal incommunicability of tacit knowledge, there are also practical limitations to knowing others’ explicit knowledge. While one can communicate their explicit knowledge, certainly not all of it can be communicated, and so one must be very selective in deciding what knowledge and information is worth sharing.

The idiosyncratic nature of knowledge led Hayek (1937, 1945) to discuss the problems of economizing such knowledge effectively. He concludes:

‘[I]t would seem to follow that the ultimate decisions must be left to the people who are familiar with these circumstances, who know directly of the relevant changes and of the resources immediately available to meet them. We cannot expect that this problem will be solved by first communicating all this knowledge to a central board which, after integrating all knowledge, issues its orders. We must solve it by some form of decentralization’ (Hayek 1945, p. 524).

In a firm, as an island of specialization, entrepreneurs invest in labor with firm-specific human capital, which is then managed by the entrepreneur or a hired manager to enact the entrepreneur’s plans and vision for production. The separation of the entrepreneurial vision from the skills and knowledge to achieve it can cause complex management problems. These are resolved either through strict supervisory tactics to ensure a particular vision is enacted or else by relaxing the vision to be more adaptive to the inputs and ideas of the laborers who enact the vision (Burgelman 1983b; Ouchi 1977). These distinct approaches (as well as hybrids) offer distinct advantages with respect to the inherent Hayekian knowledge problems but are weak in others. For example, strong supervision ensures that the entrepreneur’s knowledge is successfully applied by the hierarchy but leaves most of the rest of the hierarchy’s knowledge and skills on the table as untapped potential. On the other hand, decentralized approaches allow much more of that various knowledge to be used productively but can easily result in diversions of resources toward too many and too unrelated projects that stretch resources thin and delay production. Thus, appropriate governance of knowledge and resource use becomes vital.

Modern governance theory

Having laid the logical groundwork of organization and the role of governance therein, let us now add complexity and realism to our basic toward developing a more useful and robust theory of corporate governance. From our distinctive aprioristic approach, we have derived two generalized socioeconomic value mechanisms underpinning different forms of organizing. First, productivity is enhanced through cooperation and a division of labor to coordinate individual (often tacit) knowledge. However, a division of labor also implies that there be exchanges, which have transaction costs. And second, because ownership means control over resources, private ownership is exclusionary, while shared ownership and control is problematic and difficult to govern. Private ownership, in conjunction with colocation that accompanies a division of labor, requires complex governance practices to protect properties and interests.

To remind the reader, governance has herein been defined somewhat more strictly and narrowly than is common in modern governance theory. Traditionally, governance theorists have adopted the Cadbury Report’s (1992) definition of corporate governance as the ‘system by which companies are directed and controlled,’ or ‘the formal structures, informal structures, and processes that exist in oversight roles and responsibilities in the corporate context’ (Hambrick et al. 2008, p. 381). Here, however, we have adopted Mitchell et al.’s (2021) definition: ‘governance establishes norms and values, defines and prioritizes purposes and aims, and sets the operative rules within which the ownership functions must operate… [and] also oversees and polices these regulatory boundaries.’

This definition helps us avoid conflating the decision-making responsibilities of the entrepreneur regarding production activities with the governance activities that constrain those activities within certain regulatory boundaries, externally dictated or self-imposed. Governance includes political, market (social), industry, and corporate sources of regulation of firms’ activities. Although external constraints are ‘involuntary,’ internal governance must choose to comply or else skirt such regulations. Thus, corporate governance entails the determination of which rules and constrains, externally or internally sourced, the firm and its operatives must operate within and the policing of those constraints.

Building from these foundations, and from the theoretical scaffolding laid in the prior sections, we can easily see that modern corporate governance theory is premised upon a presumption of hierarchical organization. Thus, as we have observed, firms suffer from knowledge and agency problems, which lead to governance breakdowns.

Misconduct in hierarchy

Let us adapt Hayek’s (1945) problem of ‘the use of knowledge in society’ to the use of specific, governance-related knowledge within a firm. Specifically, we shall assume, realistically, that governance-relevant knowledge is heterogeneous, at least partially tacit, and that explicit knowledge is (at least somewhat) costly to communicate. By governance-related knowledge we mean idiosyncratic knowledge of external rules, of personal values, and of the various resources that the firm controls. This knowledge is gained idiosyncratically and, so, is necessarily unique to each individual—particularly their highly tacit knowledge of personal values. Because this knowledge is idiosyncratic and heterogeneously dispersed, challenges arise in effectively establishing and enforcing effective governance policies.

In fact, it is precisely this knowledge heterogeneity, in conjunction with hierarchical production structure that underlies the motivation and opportunities for misconduct. Misconduct ranges from simple and comparatively inconsequential ‘shirking’ to egregious and consequential misconduct such as theft and fraud. Agency theory (Jensen and Meckling 1976) asserts that these issues are, essentially, Hayekian knowledge problems, or information asymmetries (Akerlof 1970), between principal and agent. The corporate context is almost universally hierarchical to some extent, the principal delegating the firm’s complex tasks to specialist agents. The agent, with delegated access to the firm’s resources, are often in a position to exploit their privileged access to and knowledge of resources.

Cressey’s (1953) well-known fraud triangle pinpoints motive or pressure, opportunity, and rationalization as necessary to corporate misconduct. Others have added capability as a fourth factor (Schuchter and Levi 2016; Wolfe and Hermanson 2004). However, we might leverage entrepreneurship theory to distinguish the ‘third-person opportunity’ for someone from the ‘first-person opportunity’ for me (Haynie et al. 2009; McMullen and Shepherd 2006), which implies capability. First-person opportunities for misconduct arises from an individual’s unique, idiosyncratic knowledge of a firm’s governance of its resources—and particularly the flaws in that governance—that enable the exploitation of others’ ‘unknowledge’ (Shackle 1983) of those same weaknesses.

Hierarchical oversight

This agency problem has been the centerpiece of decades of research into effective governance practices to mitigate misconduct. Standard agency theory highlights the role of oversight to ensure that agents are not misbehaving (Jensen and Meckling 1976). Oversight can take many forms, with managerial oversight and audits being the most common.

The goal of such oversight is to reduce opportunity for misconduct (Gomulya and Boeker 2016; Pierce et al. 2015). When a manager is regularly looking over an agent’s shoulder, it is difficult for that agent to find opportunity to misbehave. Regular audits have a similar effect—fear that an auditor will discover the misconduct ensures that such misconduct is avoided (Neville et al. 2019; Velte 2021).

Firms also attempt to reduce the motivation to misbehave. Firms will often compensate their top agents (executives), who have the greatest access to the firm’s resources, with equity, stock options, restricted stock, or long-term incentive payouts (Burns and Kedia 2006). These forms of compensation reduce managers’ motivation to misbehave, as misconduct would come at the expense of the firm’s performance and would, thus, hurt their own financial well-being. Beyond this, firms will often attempt to mitigate misconduct at the lower hierarchical levels by establishing a positive culture and employee buy-in to the firm’s purpose (Liu 2016). Where there is cultural buy-in, agents are less likely to undermine the firm’s efforts through misbehavior.

These top-down approaches tend to be costly and only partially effective. Even strict and severe oversight cannot inhibit all possible misconduct. There are cracks in every system that can escape even the most careful manager or auditor. Indeed, evidence suggests that, over time, improving oversight methods and technologies are being met by equally improving deception and obfuscation techniques (Paine and Srinivasan 2019). Efforts to mitigate misbehavior motivation are also only partially successful—it is impossible to ensure sufficient buy-in from all to avoid misconduct.

Thus, top-down governance solutions have been remarkably unsatisfactory in mitigating corporate misbehavior. Yet, despite these flaws, they remain the standard for corporate governance.

Decentralized governance theory

We now arrive at the crux of our theoretical argument. The challenges of modern governance just reviewed are, again, inherent to hierarchical forms of organizing. However, as our theoretical analysis has shown, hierarchy is not the only viable form of social organizing—market orders are also viable. This insight is the centerpiece of decentralized governance theory, which we shall now outline.

Let us offer some clarification up front, as our intention here is to theoretically explore the possibilities of market orders within firms’ boundaries. This, of course, seems to fly in the face of common understanding, which holds internal firm organization as an explicit alternative to market orders (e.g., Williamson 1975). But the essence of the market order is autarky, as we have outlined above. The artificial boundary of a firm is not inherently obviating of autarky—many firms are decentralized, ‘flattening’ their hierarchy. Some firms (e.g., Valve, The Morning Star Company) have no hierarchy whatsoever, operating essentially as an internal market.

A mountain of research extolls the virtues (and costs) of decentralizing production processes, allowing for more autonomy (autarky) in determining productive activities.Footnote 4 Some of these virtues include greater job satisfaction (Carpenter 1971; Wheatley 2017; Worthy 1950) and less stress (Ivancevich and Donnelly 1975) and turnover (Liu et al. 2011), more proactivity (Den Hartog and Belschak 2012) and innovativeness (Damanpour 1991; Klein et al. 2019; Pierce and Delbecq 1977), and better teamwork (Griffin et al. 2001; Yang and Choi 2009).

Virtually no research yet exists, however, with regard to the decentralization of governance practices. In large part, this is because no firms, to our knowledge, practice such governance decentralization. But this is something of a chicken or egg problem—it is not clear whether no such practices exist because it is inefficient or due to mimetic isomorphism to standard practices of centralized governance. We propose that it may, in fact, be the latter, which is bolstered by the fact that conformity to standard centralized governance practices is, in many countries, required by law for public corporations.

But let us pull on the thread a little more to theoretically unravel the processes of governance were these systems to be decentralized.

Decentralized governance: an introduction

Let us begin our conclusionary analysis with an introduction to what decentralized governance means and what such governance might look like in practice. At a glance, it may seem strange to conceive of governance as potentially decentralized, which is essentially the practice of allowing employees to ‘govern themselves.’ Yet, self-governance, far from an absurdity, is in fact thought to be economically efficient (Barzel 1987).

In decentralized governance systems, the rules of governance are emergent rather than imposed. Members of the organization voluntarily agree to the rules and are able to propose revisions. Such rules are ‘softer’ than top-down governance, as the body of members attend to the ‘spirit of the law’ rather than to the minutiae of codified rules. This allows for some variance in rule abiding. Questions and concerns about rules and rule breaking are brought before the entire organization or else a designated committee to determine an appropriate ruling.

Policing of the rules is done by the group members themselves. In some cases, the self-reporting of violations may be encouraged through leniency mechanisms. For example, some regulatory agencies offer reduced penalties for voluntary disclosure of regulatory violations, which lower policing costs (Short and Toffel 2007; Toffel and Short 2011) and tend to result in more effective remediation (Innes 1999). Rather than risk discovery and the more severe consequence it entails, some may instead voluntarily disclose their misbehavior. However, much of the policing and enforcement of rules would be done by constituent members of the organization. Impetus for such policing efforts is not just preservation of the organization’s resources but also the maintenance of the organization’s integrity and reputation. Penalties for rule breaking would vary according to the severity of the infringement, ranging from internal discipline, to expulsion and professional ostracism, to legal prosecution.

The London stock exchange

As a compelling exemplar of self-governance, Stringham (2002) offers the history of the London Stock Exchange (LSE). After an act was passed in 1696 ‘To Restrain the Number and the Practice of Brokers and Stockjobbers’ (ibid: 5), a number of London’s quickly growing stockbrokerage profession began to skirt the new law by leaving the Royal Exchange and participating in informal exchanges, often in coffeehouses. Outside of the formal regulations of the State, informal market governance practices began to emerge to deal with the presence of fraudsters and defaulters. The first self-governance mechanisms entailed shunning and banning of defaulters, but its enforcement over time was difficult. In an attempt to better exclude defaulters and fraudsters from the profession, the more reputable brokers formed an exclusive club in 1762. After an ousted member brought suit against the club, the group was forced to open admission to any who paid a daily fee. But it was soon realized that the daily fee was not enough to keep out disreputable brokers, so the group began to instead require a hefty annual subscription fee. Subscription also meant agreement to abide by common rules of order, with fines levied on rule-breakers. Because some were uncooperative with the new rules, the group was disintegrated and a new exclusive exchange, the LSE, was formed. Because its rules were self-enacted and self-enforced, various rules were tried and scrapped until it codified its rulebook in 1812. In 1877, the government recognized that the LSE’s decentralized self-governance was ‘capable of affording relief and exercising restraint far more prompt and often satisfactory than any within the read of the courts of law’ (quoted in Stringham 2002: 13).

The Maghribi traders

Another interesting example is the 11th Century Maghribi Traders’ Coalition (Greif 1989, 1993). During the Middle Ages, trade across the Mediterranean regions was fraught with peril. Land shipping was slow, and sea shipping was risky. Storms were a severe threat, as was piracy. Furthermore, it was easy for a hired shipping agent to simply steal the goods—they could easily claim the goods were lost or stolen. There was at best a crude legal system that was incapable of policing and prosecuting such malefactors. Thus, traders would personally accompany their goods on such journeys, costing them dearly in productive time. It was this inefficiency that the Maghribi coalition addressed:

The Maghribi traders overcame the contractual problems associated with agency relationships by organizing such relationships through a nonanonymous organizational framework, the coalition. Within the coalition an internal information-transmission system served to balance asymmetric information and a reputation mechanism was used to ensure proper conduct. This reputation mechanism explains the observed ‘trust’ relations among the traders. The ‘trust’ did not reflect a social control system or the internalization of norms of behavior (although these factors play a role in any economic system). Rather, the Maghribi traders established a relationship between past conduct and future economic reward. As a result, agents resisted the short-term gains attainable though deception, since the reduction in future utility resulting from dishonest behavior outweighed the associated increase in the present utility. Since this fact was known beforehand to all traders, agents could acquire a reputation as honest agents. (Greif 1989: 881)

The trade coalition did not operate with a top-down hierarchical governance structure, but self-policed with reputation and trust penalties or even, when necessary, expulsion and ostracism. As a result, it enjoyed a high degree of reliability and strong reputation, putting them at significant advantage in the shipping industry, for an enduring period of time.

Decentralized governance and agency and knowledge problems

To theoretically validate the promise of decentralized governance, let us reconsider the principal-agent problem and other Hayekian knowledge problems through market organizational mechanics. In any multi-person collaborative system (collective, cooperative, market, hierarchy), there are knowledge asymmetry problems. In collectives and cooperative, as previously reviewed, these asymmetries tend to result in a tragedy of the commons. In hierarchies, these can manifest as principal-agent problems as well as other systemic flaws that result in opportunities for corporate misdeeds.

For markets, however, knowledge asymmetry problems are largely mitigated due to the maintaining of autarkic rights to property, production, and self-governance, with intersubjective governance by mutual agreement or government. As a result, each individual’s distinct knowledge is placed at that individual’s own behest, their utility maximization efforts further bolstered by mutually beneficial exchange. Not all such asymmetries are thus mitigated, however—intellectual property protections and lemons problems (Akerlof 1970) can leave persistent inefficiencies. Yet, in Akerlof’s used car example, even these asymmetry problems have been largely mitigated over time with innovative market solutions.

Internally to organizations, decentralization of governance produces many of the same benefits, but only to the extent that organizational agents are motivated to optimize the organization’s governance. As with the LSE and the Maghribi traders, market competition typically provides this impetus. Stakeholders are incentivized to mitigate misconduct within the organization so that it can effectively compete in the market and preserve its long-term financial prospects, and thus their own financial stake. Misconduct undermines competitiveness and can sully the organization’s reputation. While such behaviors may be appealing to individual agents with little personal stake in the organization’s long-term success, those in the organization that have a greater stake are incentivized to protect their organization from such malefactors. The success of decentralized governance, then, hinges on the benefit and interest of agents in the long-term viability of the organization.

The primary advantage of governance decentralization is that greater and more specific knowledge of the resources and access points can be employed in governance practices. In typical centralized governance practices, managers and auditors watch for standard signals of wrongdoing, these signals are often untethered from the specific resource access systems and mechanisms that agents actually use, leaving governance gaps that are opportunities for misbehavior. If governance is decentralized, agents far more familiar with and knowledgeable of those systems can place that individual knowledge toward better oversight, reducing or eliminating those gaps that centralized governance audits so often overlook.

Decentralization and governance innovations

One of the more interesting features of decentralized governance is that the ‘softness’ of established rules allows for governance innovations. Hayek (1978, pp. 62–63) explains:

There is an advantage in obedience to such rules not being coerced, not only because coercion as such is bad, but because it is, in fact, often desirable that rules should be observed only in most instances and that the individual should be able to transgress them when it seems to him worthwhile to incur the odium which this will cause. It is also important that the strength of the social pressure and of the force of habit which insures their observance is variable. It is this flexibility of voluntary rules which in the field of morals makes gradual evolution and spontaneous growth possible, which allows further experience to lead to modifications and improvements. Such an evolution is only possible with rules which are neither coercive or deliberately imposed—rules which, though observing them is regarded as merit and though they will be observed by the majority, can be broken by individuals who feel that they have strong enough reasons to brave the censure of their fellows. Unlike any deliberately imposed coercive rules, which can be changed only discontinuously and for all at the same time, rules of this kind allow for gradual and experimental change. The existence of individuals and groups simultaneously observing partially different rules provide the opportunity for selection of the more effective ones.

As a result of such flexibility and the innovations it affords, decentralized governance is expected to improve its governance practices over time, as the history of the LSE exhibits.

Certainly, it is possible for centralized governance to innovate also. But, like centralized production decisions (Burgelman 1983a, 1991; Ouchi 1977, 1980), innovativeness in governance is severely hampered by the bureaucracy of centralization, where new ideas are liable to be shot down by upper-level managers wary of and resistant to change. Moreover, in centralized governance systems, new rules are imposed from the top, leaving the rule-making to regulators that often know very little of the systems that they regulate. Thus, changes to rules—such as the Public Company Accounting Reform and Investor Protection (Sarbanes–Oxley) Act of 2002 (Romano 2005; Zhang 2007) and the Wall Street Reform and Consumer Protection (Dodd-Frank) Act (Bainbridge 2010; Dimitrov et al. 2015)—tend to be inefficient and ineffective. While such inefficiencies in centralized governance systems tend to persist and, if anything, are prone to ‘correction’ through increased complication, which tends toward exacerbation, decentralized governance is more capable of correcting inefficiencies by abandoning bad rules and innovating better solutions.

Discussion

Our aim herein has been to address the arms race between improving corporate governance and corporate misconduct, both of which are growing in complexity and effectiveness. It appears that corporate governance is losing this race.

‘The recent spate of behavioral complaints against senior corporate leaders has raised questions about board oversight of executive conduct and caught numerous boards off guard. On a different front, various companies have suffered serious breaches of cybersecurity that have exposed a lack of preparedness and resulted in significant reputational damage; others have been tripped up by data privacy concerns and are facing political and user backlash. Environmental disasters, labor abuses in the supply chain, mistreatment of customers—these are other examples of the new breed of risk management issues that are consuming the attention of boards. The broadening menu of risks has created a challenge for traditional practices of internal controls and is testing the ability of boards to provide adequate oversight.’ (Paine and Srinivasan 2019, p. 15).

To turn the tables and get corporate governance practices in front of its malcontents, we will need a more radical rethinking of standard governance practices.

Toward this end, we have developed, through rationalist analysis, a theoretical framework that encompasses the general array of socioeconomic ordering alternatives by which agents can structure and govern their productive interactions. Perhaps unsurprisingly, the two dominant socioeconomic ordering forms for economies of many agents are markets and hierarchies (Williamson 1973, 1975). What is interesting to us, however, is that the virtues and challenges of these distinct socioeconomic ordering forms are almost always delimited at the boundary of the firm itself. Williamson’s (1973, 1975) classical treatment in particular discusses the impetus for integrating activities within an organization’s hierarchy versus leaving or disintegrating such activities to the market—the classic make-or-buy decision. Certainly, hybrid forms have been put forth that integrate components of both market and hierarchical features within the firm (e.g., Foss 2003; Makadok and Coff 2009). Also, there is growing interest in and practice of the decentralization of production activities in ‘flat’ hierarchies (Foss et al. 2015; Hempel et al. 2012; Klein et al. 2019).

Yet, despite this growing interest and advocacy of employing a market order form within the firm for more optimal productive output, there has been no discussion of the possibility of such market orderings for the performance of the governance function. We have herein defined the function of ‘governance’ to be ‘to constrain, conscribe, and regulate’ production to within ethical and strategic boundaries, and to police and enforce those boundaries (Mitchell et al. 2021, p. 14). As our analysis reveals, there is nothing innately distinctive between the production and governance functions that would make one more effective via market order and the other via hierarchical order. Instead, the advantages and disadvantages of these distinct organizational forms are highly similar for both economic functions. Thus, it is possible and even advisable in some (perhaps many) circumstances for governance to be decentralized.

The decentralization of governance may seem somewhat counterintuitive given the long-standing tradition of centralized governance. Are we really to expect employees to govern themselves? Of course, at an organization level, a market economy is essentially defined in terms of allowing firms to govern themselves—central (political) governance of economic production is the hallmark of a socialist political structure. It is also widely accepted that sole-proprietor entrepreneur self-governance is possible. But we rarely see such decentralized governance as viable or even possible at an individual level within firms. This is likely due to principal-agent problems that are inherent to employment and which can undermine the incentive to self-govern in accordance with the principal’s governance requirements.

Certainly, the challenges of the principal-agent problem are real and consequential, which may justify the use of hierarchy to monitor agent behavior. However, hierarchy as a solution to the principal-agent problem brings with it other challenges, such as information asymmetry problems (Keil et al. 2004) and disempowerment problems (Hempel et al. 2012). It is not altogether clear, then, that hierarchy is always the best governance structure for even large organizations.

As with incomplete contract theory (Grossman and Hart 1986; Hart and Moore 1990), which argues that contracts can never account for all contingencies, top-down governance is also always incomplete. There are endless possible ways to skirt established rules and their enforcement. As a result, governance mechanisms have to adapt to ever-changing rule-breaking and are necessarily always lagging.

As corporate governance practices continue to lag behind innovative malefactors, what becomes revealed is that misbehavior is always an issue, first and foremost, of self-governance. Assuming that agents know and understand the principal’s governance goals and standards, and those of society (e.g., government regulations), misbehavior is always a case of some agent willfully contravening those standards for personal gain, broadly defined. Modern corporate governance, then, is intended to catch or prevent such malefactors after they have had a self-governance breakdown.

Recognizing this, decentralized governance theory seeks to address (1) how to best improve employee self-governance and (2) how to most effectively catch self-governance breakdowns before they result in substantial resource misallocations or losses. Regarding the former, we posit that governance decentralization proffers alternative ways to think of and motivate self-governance. In fact, we suggest that centralized governance may do much to undermine self-governance, much in the same way that overbearing political governments can undermine socioeconomic self-determination (Bylund 2016b; Easterly 2013; Holcombe 2018).

Regarding the latter, decentralized governance theory suggests that centralized governance suffers from knowledge asymmetries that may, if organized effectively, be partially mitigated by decentralizing governance oversight, to the lower levels of an organization’s hierarchy. As a result, those closest to and most familiar with resource allocation processes can monitor the gaps in such processes and ensure the protection of those resources. Said differently, those closest to the resource allocation processes can best monitor those resources allocations to ensure no loss or leakage. Centralized governance is, simply, too removed from those specific processes to allow sufficient knowledge of whatever gaps there may be in the resource allocation process, which present opportunities for misbehavior.

Space and scope constraints preclude us from pursuing an elaboration of what decentralized governance might look like in practice—what are its primary governance mechanisms, what are the incentives that compel self-governance and peer monitoring, and how it might be most effectively implemented. We leave such elaborations to future research. Fortunately, decentralized governance is not the eccentric meanderings of theoretical unrealism—there are various real-world examples of such decentralized governance in practice that may offer opportunities for inductive theorizing on and empirical testing of such practices and theories. The rise of completely flat organizations, such as Valve software, and of ‘decentralized autonomous organizations,’ such as Bitcoin, provide empirical evidence that decentralized ‘community governance’ can work in practice. These examples of decentralized governance warrant deeper scientific examination.

Also, we expect that this research may cause a stir among corporate governance scholars as we proffer new theoretical foundations for the field. In some ways, our theory rebuilding efforts through the praxeological method have confirmed (and better explained) long-standing theory on the challenges of governance within the context of hierarchical organization. Agency theory (Jensen and Meckling 1976) in particular stands the test of deductive rigor, but only for the hierarchical context. Our analysis reveals that hierarchy is only one organizational option, and that internal market forms are viable, not only for organizing productive activities, but for performing the governance function also. Thus, our efforts better theoretically contextualize prevailing governance research and open new avenues for additional research. For example, we note with others (e.g., Foss 2003; Makadok and Coff 2009) that hierarchy and markets are ends of a spectrum, with gradients between them. Hybrid governance forms, then, should be the subject of future research.

Conclusion

Corporate governance theory is in a deep rut. Despite decades of significant and rigorous academic work, severe political and regulatory interventions in the wake of some of the largest corporate scandals in world history, and widespread corporate interest in misbehavior prevention, corporate misbehavior has remained surprisingly constant. Efforts to mitigate such misdeeds have either been altogether ineffective or else malefactors have been as innovative in misbehaving as the overseers have been at detecting such misconduct. It might be time to fundamentally rethink standard corporate governance practices.

We have herein developed, through rationalist-deductive methodology, a refined and elaborated framework for understanding governance practices. This framework implies that market-type decentralization of governance is not only possible but may in fact be advantageous to traditional centralized governance practices, if effectively implemented. This is not to say that all organizational contexts call for decentralized governance—the most effective form and type of governance will expectedly depend on the type of organization. But it is time we move on from a ‘gold standard’ view of centralized governance and to a recognition of a broader array of governance approaches that reflect the distinct advantages and disadvantages of centralization versus decentralization.