1. Introduction
Uncertainty, according to Knight, reflects the imperfect knowledge of the future that guides all human conduct and decision-making in the present, regardless of institutional conditions. Although all individuals ‘live by knowing something about the future’, most of the ‘problems of life or of conduct’, Knight observes, ‘arise from the fact that we know so little’ (Knight, Reference Knight2006: 199, emphasis in the original). We act now, in the present, but our choices are guided by our anticipations of the future: by a perception of how the future will unfold without our interference, and by an idea of how our actions can alter these conditions as they emerge. In formulating these expectations, individuals lack the ability to ‘perceive the present as it is and in its totality’ (Knight, Reference Knight2006: 202); and given this partial knowledge of the present, they often fail to infer how the future will unfold ‘with any degree of dependability’ (Knight, Reference Knight2006: 202). For these reasons, decision-makers are not omniscient regarding the future and cannot make predictions with certainty, regardless of whether they act within markets, democratic government, or civil society. In this sense, all human action involves Knightian uncertainty and thus raises the question of how actors within different institutional settings engage in productive and allocative decision-making in its presence.
Given the ubiquity of uncertainty, this paper explores how different institutional settings allow for the bearing of that uncertainty, and, more importantly, it explores the impact of uncertainty on productive and allocative decisions within these different institutional settings. In the process of exploring these questions, the paper also undertakes a comparative institutional analysis, comparing calculative or market-based institutions to non-calculative institutions, primarily those of democratic politics and civil society.
Within markets, entrepreneurs exercise judgment and bear Knightian uncertainty while making production decisions that aim to satisfy consumer preferences. The role of profit and loss as, it is revealed through monetary calculation, provides us with a mechanism to correct errors within markets. Within democratic government, a multiplicity of competing interests and groups compete with one another for access to resources as well as decision-making power over those resources. Meanwhile, within the realm of civil society, the decisions of nonprofit entrepreneurs, as well as those of donors/benefactors and beneficiaries, play an important role in influencing how resources are allocated in the presence of uncertainty.
We outline the main implications of Knightian uncertainty and calculation in markets and then ask three key questions of democratic politics and nonprofit civil society: (1) Who undertakes the role of judgment under uncertainty? (2) Which goals matter and whose preferences are satisfied? And, (3) How is error detected and corrected without monetary profit and loss? We find that coordination of allocative plans with underlying ‘consumer’ preferences is likely to only appear in special cases in democracy whereas it is more likely in nonprofit civil society. However, in neither realm can one conclude that there is an inherent alignment of interests between judgment-bearing entrepreneurs and consumers due to the presence of Knightian uncertainty and the lack of calculative decision-making.
2. Knightian uncertainty, judgment, and the market process
Knightian uncertainty and Knightian risk
Different scenarios throw up different types of knowledge imperfections. This, Knight argues, leads the uncertainty of the future to take on different forms, and gives rise to the distinction between Knightian risk and Knightian uncertainty. Although there are different interpretations of Knight's views on the difference between risk and uncertainty, this paper builds upon the work of Boudreaux and Holcombe (Reference Boudreaux and Holcombe1989), Langlois and Cosgel (Reference Langlois and Cosgel1993), and Foss and Klein (Reference Foss and Klein2012). The distinctive aspect of the interpretation of Knight's work put forward by these authors is the emphasis they place on his theory of imperfect knowledge. Langlois and Cosgel (Reference Langlois and Cosgel1993: 458–460) and Foss and Klein (Reference Foss and Klein2012: 84–86) argue that Knightian risk refers to those situations where individuals can calculate a priori or statistical probabilities in the face of uncertainty, whereas the concept of Knightian uncertainty refers to those situations where the decision-maker can only estimate likely outcomes and must rely on his (or her) subjective judgments to navigate the prevailing uncertainty.Footnote 1
Knightian risk emerges in scenarios, such as rolling a fair die or tossing a fair coin, where individuals possess complete knowledge of the possible outcomes from an event and can form an ‘absolutely homogenous’ (Knight, Reference Knight2006: 224) classification of events. This allows the individual to calculate the likelihood of each possible outcome a priori and to assign an equal probability to each outcome.
In a second type of situation, individuals may have complete knowledge of the set of possible outcomes from an event but may not be able classify these events into groups that are absolutely homogenous. Consider, for example, an individual trying to assess the likelihood that a building he owns will catch fire. He knows the set of possible outcomes for each event but cannot rule out the fact that certain causes operate on one or on a few of the buildings in question but not in the case of the others. Such a scenario is also characterized by the presence of Knightian risk. But, the calculation of probability, instead of being done a priori, must rest on ‘tabulating the results of experience’ (Knight, Reference Knight2006: 215). The individual must conduct a statistical analysis of a group of events that, while not absolutely homogenous, are roughly similar in many respects. Based on the statistical evidence gathered, the individual can calculate the statistical or empirical probabilities associated with the different outcomes that are involved.
A third type of imperfection of knowledge emerges in situations where an individual lacks complete knowledge of the possible outcomes that may result from an event. In contrast to the knowledge imperfections that give rise to Knightian risk, where the possible outcomes are known with certainty, in such scenarios of Knightian uncertainty the individual is in a situation where he does not know the outcomes that may result from a particular decision. He is, thus, forced to try and guess at what the possible outcomes may be and finds that he must form ‘the best estimate’ that he can ‘of the outcome of his actions’ (Knight, Reference Knight2006: 226). After having arrived at such an estimate, he then turns to consider the likelihood associated with these estimated outcomes and tries to ‘estimate the probability that his estimate is correct’ (Knight, Reference Knight2006: 226).
The nature of Knightian uncertainty lends a subjective character to these estimates of outcomes and their associated likelihoods: an individual might arrive at different estimates when placed in a similar situation at two different points in time, and two individuals may arrive at different estimates when confronted with similar situations. For this reason, the instance in question ‘is so entirely unique’, argues Knight, that ‘there is no valid basis of any kind of classifying instances’ (Knight, Reference Knight2006: 225–226). Instead, in handling such uncertainty, the individual decision-maker relies on ‘two separate exercises of judgment, the formation of an estimate and the estimation of its value’ (Knight, Reference Knight2006: 227). In such situations of estimated probability, he must use his subjective judgment in arriving at an estimate of the possible outcomes and at an estimate of the probabilities involved.Footnote 2
A recent strand of the literature on uncertainty has developed and built upon these concepts of Knightian risk and uncertainty. Dequech (Reference Dequech1997, Reference Dequech2000, Reference Dequech2011) identifies not two but three distinct conceptions of uncertainty. Situations involving Knightian risk, Dequech argues, are characterized by the presence of weak uncertainty, where the decision-maker can form ‘a unique, additive and fully reliable probability distribution’ over possible future outcomes (Dequech, Reference Dequech2011: 622). In scenarios involving Knightian uncertainty, on the contrary, decision-makers are faced with fundamental uncertainty: they find themselves unable to form probability distributions over possible outcomes, not because they lack the necessary information and knowledge to do so, but because, given the unpredictable nature of the changes taking place in their environments ‘some relevant information cannot be known, even in principle’ (Dequech, Reference Dequech2000: 48). Between these two conceptions of uncertainty lies the concept of ambiguity, where decision-makers are unable to form a probability distribution over possible outcomes merely due to a lack of relevant information.
Packard et al. (Reference Packard, Clark and Klein2017) refine this tripartite classification of uncertainty by drawing on the work of Shackle (Reference Shackle1949, Reference Shackle1961, Reference Shackle1972). In situations of Knightian risk and ambiguity, they argue, decision-makers are faced with a closed set of possible options (means) and outcomes (ends). The difference between the two concepts lies in the availability of the relevant knowledge needed to form probability distributions over this closed set of options and outcomes: in the case of risk, the knowledge is available to the decision-makers and the relevant probabilities can be identified, whereas in the case of ambiguity the lack of knowledge means that individuals are unable to form probability distributions over the known set of possible outcomes.
In addition to developing the distinction between risk and ambiguity, Packard et al. also further extend the concept of Knightian uncertainty (or fundamental uncertainty). Although all situations involving Knightian uncertainty are characterized by an open set of options and/or outcomes, they identify three different types of such uncertainty: situations of environmental uncertainty, where ‘the set of options is closed but the set of outcomes is open’ (Packard et al., Reference Packard, Clark and Klein2017: 5); scenarios characterized by creative uncertainty, where the opposite holds true, i.e. ‘the set of outcomes is closed and the set of options is open’ (Packard et al., Reference Packard, Clark and Klein2017: 6); and finally, situations of absolute uncertainty, where ‘both option and outcome sets are open’ (Packard et al., Reference Packard, Clark and Klein2017: 6).
This broadening of the concept of Knightian uncertainty does not undermine, but instead reinforces Knight's claim regarding the use of judgment to deal with this type of uncertainty. As Knight argues (Knight, Reference Knight2006: 224–225), and as Packard et al. (Reference Packard, Clark and Klein2017: 7) note, in situations involving Knightian risk, decision-makers can be guided by calculations of numerical probabilities, and they can do likewise when a situation involving ambiguity presents itself repeatedly over time. In all situations involving Knightian uncertainty, however, whether it involves environmental uncertainty, creative uncertainty, or absolute uncertainty, decision-makers cannot base their decisions on the calculations of numerical probabilities. Instead, they must resort to making judgments regarding the ‘possibility’ (Packard et al., Reference Packard, Clark and Klein2017: 6) or likelihood of the different combinations of options and outcomes that they consider feasible, and then must use these judgments to make their decisions (Knight, Reference Knight2006: 227; Packard et al., Reference Packard, Clark and Klein2017: 7–8).
Judgment and entrepreneurship in calculative settings: profit, loss, and the market process
Entrepreneurial judgment in the marketplace and the theory of profit and loss
Knight (Reference Knight2006: 272–282) developed a theory of profit and loss based on his view that the function of entrepreneurship consists of exercising judgment in the face of uncertainty. As Knight (Reference Knight2006: 271) notes, in the marketplace the entrepreneur bears this uncertainty and uses his subjective judgment in the allocation of scarce resources. The fact that entrepreneurs differ in their ability to exercise judgment helps explain how their decisions to allocate factors of production leads to the emergence of monetary profit and loss. This Knightian conception of entrepreneurship as the exercise of judgment within the marketplace has recently been rediscovered and developed, as is evidenced by a growing body of literature (Brown et al., Reference Brown, Packard and Bylund2018; Foss et al., Reference Foss, Klein and Bjornskov2019; Foss and Klein, Reference Foss and Klein2012, Reference Foss and Klein2015; Godley and Casson, Reference Godley and Casson2015; Hallberg, Reference Hallberg2015; Lounsbury et al., Reference Lounsbury, Gehman and Glynn2019; Mcmullen, Reference McMullen2015; Packard and Bylund, Reference Packard and Bylund2019).
The burden of Knightian uncertainty is not borne equally by all individuals in a market economy: it is a select group of individuals who exercise the function of entrepreneurship and the ability to form and exercise judgment is thus unevenly distributed. As Knight notes, there are two significant ways in which individuals differ in their ability to make decisions. First, they ‘differ in their capacity by perception and inference to form correct judgments as to the future course of events in the environment’ (Knight, Reference Knight2006: 240, emphasis in original); and second, they differ in their capacity ‘to judge means and discern and plan the steps and adjustments necessary to meet the anticipated future situation’ (Knight, Reference Knight2006: 241, emphasis in original). It follows that some individuals are better at dealing with Knightian uncertainty and can make less erroneous judgments when faced with it.
This uneven distribution in the ability to make correct judgments in the face of Knightian uncertainty implies that not everyone has to own a firm. Instead, these ‘differences in individuals in relation to uncertainty’ give rise to ‘a tendency to specialize the function of meeting it in the hands of certain individuals and classes’ (Knight, Reference Knight2006: 243–244). More specifically, they give rise to the ‘enterprise and the wage system of industry’ (Knight, Reference Knight2006: 271), where there is a specialization of functions within firms: ‘a special social class’, i.e. the entrepreneurs, ‘direct economic activity’ (Knight, Reference Knight2006: 271), own and control resources, and have the wherewithal to exercise their judgments, whereas ‘the great mass of the population’ (Knight, Reference Knight2006: 271) decide to simply work and earn wages, thereby opting out of the need to deal with Knightian uncertainty as far as the production of goods is concerned.
Situations involving Knightian uncertainty are unique and cannot be insured against. Instead, the relative accuracy of judgments made by entrepreneurs in these situations result in them either earning a profit or suffering a loss. Consider the case of entrepreneurs bidding to hire the services of a factor of production (Knight, Reference Knight2006: 271–281). Given differing expectations among multiple entrepreneurs, the factor price that emerges is a result of the forces of the competition. The monetary profit (or loss) than each entrepreneur earns reflects ‘the difference between the market price of the productive agencies he employs […] and the amount which he finally realizes from the disposition of the product which under his direction they turn out’ (Knight, Reference Knight2006: 277). Those entrepreneurs whose judgment turns out to be more accurate earn a greater monetary profit, or suffer a smaller monetary loss, whereas those who err to a greater degree are punished more severely.
The market process, error correction, and economic calculation
Knight's theory of profit and loss bears a striking resemblance to that advanced by Mises (Reference Mises1998: 286–291; Reference Mises2008), which is not surprising given that the Misesian conception of uncertainty and the function of entrepreneurship are similar to Knight's.Footnote 3 Mises (Reference Mises1998: 291) stresses that the difference between cost and revenue ultimately stems from the imperfect knowledge that an entrepreneur has of future market conditions. Thus, incomes earned by entrepreneurs reflect the quality of their anticipations and judgments: the ultimate source of an entrepreneur's profit ‘is his ability to anticipate better than other people the future demand of the consumers’ (Mises, Reference Mises1998: 288, emphasis added).
Mises (Reference Mises1998: 330–335) and Kirzner (Reference Kirzner1973, Reference Kirzner1997, Reference Kirzner2011) also take Knight's analysis further. They provide an analysis of a market process, or a process of error correction, that emerges from the erroneous judgments of entrepreneurs.Footnote 4 In the market, entrepreneurs are motivated to bear uncertainty and make judgments to improve their well-being: the lure of possible monetary profit impels them to look into the future and adjust their actions in line with their expectations (Kirzner, Reference Kirzner and Kirzner1982). Thus, they have an incentive to form the best possible judgments of the uncertain future (including underlying consumer preferences). The monetary profit and loss system rewards those entrepreneurs that are more successful in doing so, while penalizing those who are less capable of bearing uncertainty. It is this same lure of possible profit, Kirzner (Reference Kirzner and Kirzner1982: 148–151) argues, that also incentivizes entrepreneurs to recognize the errors that they have made, and to correct these errors by reformulating their judgments of the future.Footnote 5
Thus, it is monetary calculation enabled through the price system that allows for emerging factor prices to reflect the less than perfect judgments of competing entrepreneurs. Profits and losses earned in market exchange then allow entrepreneurs to use their judgment to make changes to their production decisions (e.g. either expanding production or reducing production to cut losses). Entrepreneurs can thus make use of money prices to engage in economic calculation (Hayek, Reference Hayek1945; Mises, Reference Mises1920): they can use money prices ex ante, while planning their actions, as well as ex post, in reviewing the outcome of their decisions. In this sense, markets inherently enable the coordination of production decisions and entrepreneurial judgments with consumer preferences. There is only one distinct class of entrepreneurs that undertakes judgment under Knightian uncertainty and only one class of wage earners/consumers whose preferences get satisfied. Furthermore, the interests of these two groups are naturally aligned with one another within the institutional setting of markets.
3. The implications of Knightian uncertainty in non-market settings
Non-calculative action can be defined as action that proceeds without the guiding force of monetary prices or profit and loss. Without the ability to calculate monetary profit, judgment must proceed with the help of other guiding principles, under ubiquitous Knightian uncertainty. In this section, we consider in detail how judgment and error-correction proceed under the two prominent non-calculative institutional settings of democratic politics and civil society.Footnote 6
Our analysis shows that in these settings, one cannot come to similar conclusions about the inherent tendency for coordination of productive and allocative judgments undertaken by entrepreneurs with underlying ‘consumer’ preferences, for two key reasons: (1) There are either multiple groups or classes of actors (often with conflicting goals) that assume the role of judgment and entrepreneurship while aiming to serve the preferences of multiple classes of ‘consumers’ within these two realms. (2) The lack of monetary profit and loss not only severely constrains the ability to aggregate ‘consumer’ preferences in a meaningful way but also hampers with the incentives and the ability of entrepreneurs to systematically detect and correct errors over time.
Thus, we find that within democratic politics, coordination, while not impossible, is not likely to be the norm but more of a ‘special case’. Furthermore, it follows that within nonprofit entrepreneurship, coordination of entrepreneurial decisions with consumer preferences is more likely than democratic politics and less of a ‘special case’, but not an inherent characteristic as it is in markets, for reasons analyzed below.
Uncertainty and judgment in democratic politics
Three key questions arise when we analyze how entrepreneurship functions in the realm of politics: (1) Who exercises judgment in the political realm, i.e. who are the political entrepreneurs? (2) What goals are these political entrepreneurs trying to accomplish and whose preferences get satisfied? And, (3) How can political entrepreneurs detect and correct errors?
Who exercises judgment?
First, who exactly is it that exercises judgment in the political realm? In democratic politics, judgment is divided and distributed among multiple groups of actors including, but not limited to, politicians, legislators, unelected bureaucrats, administrators, and private-sector lobbyists. Klein et al. (Reference Klein, Mahoney, McGahan and Pitelis2010: 2) provide a definition of public entrepreneurship as ‘any action in pursuit of the public interest’. However, they acknowledge that this definition comes with its drawbacks. ‘A caveat in this endeavor’, they write, ‘is that public interests may not be well defined’. For starters, the ‘alignment of individual objectives into a public interest is a complex problem’. Citing works by Downs (Reference Downs1957) and Buchanan and Tullock (Reference Buchanan and Tullock1962), they acknowledge that divining the will of the public is a task that is far easier said than done. ‘Mechanisms such as majority voting, arbitration, and consensus-building are highly imperfect aggregators of individual interests’, they note. ‘In some situations’, as was most famously outlined by Arrow (Reference Arrow1951), ‘such aggregation may not even be possible’.
However, a broader notion of political entrepreneurship has been examined by numerous authors (Bylund and McCaffrey, Reference Bylund and McCaffrey2017; Holcombe, Reference Holcombe2018; McCaffrey and Salerno, Reference McCaffrey and Salerno2011; Wagner, Reference Wagner2016a, Reference Wagnerb). A common finding in this literature is that market and political entrepreneurship cannot be treated as separate and distinct phenomena. Private and political entrepreneurship are often mutually dependent and co-evolving (Ostrom, Reference Ostrom1990, Reference Ostrom2005). To pursue their objectives, private entrepreneurs must constantly work with politicians and regulators; politicians and bureaucrats must likewise work closely with private actors to achieve their ends (Holcombe, Reference Holcombe2018; Wagner, Reference Wagner2016a, Reference Wagnerb). This symbiotic relationship becomes increasingly evident over time as governments grow to play a more active role in the economy (Bylund and McCaffrey, Reference Bylund and McCaffrey2017). We proceed with this broader definition for the remainder of the paper.
What goals are political entrepreneurs trying to accomplish? Whose preferences get satisfied?
A second question that arises is: what goals do political entrepreneurs try to accomplish? Voters are often treated as the core group of ‘consumers’ whose preferences political action is aimed at satisfying. If we accept the broader definition of political entrepreneurship as encompassing a broader network of actors in both the public and private sphere, however, it follows that the goals of multiple groups or classes of judgers must be accounted for, and not just those of voters.
Three branches of the existing literature help to explain what motivates political entrepreneurs. First, the public choice literature emphasizes how political entrepreneurs are motivated by self-interest (rather than public interest) and a desire to capture rents, as when private firms lobby the government for special privileges (Krueger, Reference Krueger1974; Tullock, Reference Tullock1967). This rent-seeking behavior can also emanate from various other actors/judgers within government. Wagner (Reference Wagner1966) argues that the primary goal of regulators and bureaucrats is to satisfy the wishes of interest groups and thus expand or maximize their own budgets and authority. In this view, bureaucrats and public officials should not be treated as mere passive intermediaries of the public's interests; like private actors, they often act to advance their own self-interest.
The second branch that examines the goals of political entrepreneurs is the entrepreneurship literature, where the emphasis is on what political and market entrepreneurs share in common. Like market entrepreneurs, political entrepreneurs are primarily motivated by the desire for profit. The core difference is that rather than competing in market settings to earn monetary profits, they must compete in the political realm, often in the absence of explicit price signals. Klein et al. (Reference Klein, Mahoney, McGahan and Pitelis2010: 3–4) summarize the parallels between political and market entrepreneurs: like Knightian entrepreneurs, political entrepreneurs also exercise judgment under uncertainty to ‘invest resources, tangible and intangible (time, effort and reputation), in anticipation of uncertain future rewards’ (ibid.: 4). Like entrepreneurs in the marketplace, political entrepreneurs who exhibit superior judgment are rewarded, whereas those whose judgment is relatively inferior are penalized. In some cases, these rewards may come in the form of monetary profits (e.g. a solar company lobbying politicians to pass a tax rebate on solar panels to increase its sales). In other cases, like the entrepreneurship conducted by a politician or bureaucrat, the reward may come in non-monetary form such as higher re-election chances or increased budgetary autonomy. In either case, these rewards constitute what we might call a ‘political profit’ (Holcombe, Reference Holcombe2002).
The third branch comes from the literature on new institutional economics (NIE). Traditionally, NIE scholars have analyzed how market entrepreneurs respond to changes in a society's rules of the game. More recent work, however, has focused on the active role that political entrepreneurs play in devising and altering political institutions to advance their interests.Footnote 7 A key idea in this literature is that entrepreneurs often play an active role in altering institutions: as such, there is a bidirectional relationship between institutions and entrepreneurship (Elert and Henrekson, Reference Elert and Henrekson2016). Entrepreneurs don't merely abide by rules; they can also alter or evade these rules in actively pursuing their goals (Bylund and McCaffrey, Reference Bylund and McCaffrey2017; Coyne and Leeson, Reference Coyne and Leeson2004; Henrekson and Sanandaji, Reference Henrekson and Sanandaji2011). This is especially true when the rules of the game themselves are unclear or subject to change – that is, when entrepreneurs must cope with ‘institutional (or regime) uncertainty’ that is similar to the uncertainty discussed by Knight (Higgs, Reference Higgs1997). In these situations, entrepreneurs have an incentive to act not just within the rules but at the institutional level to alter rules in ways that might mitigate this uncertainty.Footnote 8
Whose preferences do political entrepreneurs aim to satisfy? Voters as ‘consumers’ represent the core group whose preferences politicians' actions are aimed at. But even if politicians do win elections, they cannot assume that voters are giving them a mandate to pursue all their campaign platforms.Footnote 9 Furthermore, political entrepreneurs do not have an inherent incentive to meet voter or ‘consumer’ preferences if political activity is tax-funded or relies on rent-seeking interest groups and campaign contributions. Hence, interest groups and other private-sector lobbyists are also a class of ‘consumers’ in democratic politics as far as politicians and legislators directly or indirectly seek to satisfy their preferences. In fact, interest groups oftentimes have an incentive to decrease access to accurate information available to voters, who are rationally ignorant (Boudreaux, Reference Boudreaux2014; Caplan, Reference Caplan2011; Downs, Reference Downs1957).Footnote 10
How do political entrepreneurs detect and correct errors?
Third, how are ‘errors’ identified and corrected in the realm of politics? The feedback signals that political entrepreneurs rely on are far weaker, and the selection mechanism for allocating resources toward more successful political entrepreneurs is ‘complex and poorly understood’ (Klein et al., Reference Klein, Mahoney, McGahan and Pitelis2010: 4). Political entrepreneurs are not subjected to the same discipline that is provided by market competition and are rarely able to refer to money prices or profits and losses to assess their performance.
McCaffrey and Salerno (Reference McCaffrey and Salerno2011: 554), however, argue that by shifting the economy's production patterns from what it would have been in an unhampered market, political entrepreneurs can capture ‘quasi-profits’. This mirrors Holcombe's (Reference Holcombe2002: 147) notion of ‘political profits’ that politicians can reap in the form of higher approval ratings and reelection odds or that bureaucrats can reap in the form of a promotion or securing more funding and power for their agency. Similarly, Wagner (Reference Wagner2016a, Reference Wagnerb: 535) notes that ‘to be successful, political enterprises must raise sufficient revenue to return profits to investors. Those profits are disguised through indirect transactions and ideological formulations but are profits all the same’. These political profits, although they are often psychic, and are not the monetary profits that market entrepreneurs use to engage in economic calculation, still play a vital role in guiding the actions of political entrepreneurs.
Moreover, there are two important structural mechanisms to constrain poor decision-making and check bad outcomes in the political realm: election cycles (typically every 4 years) and terms limits for executive and legislative administrators (Erler, Reference Erler2007; Reed et al., Reference Reed, Schansberg, Wilbanks and Zhu1998; Will, Reference Will1992). In terms of whether there exists a systematic tendency to correct errors over time, Boettke et al. (Reference Boettke, Tarko, Aligica, Boettke and Storr2016) build on Hayek (Reference Hayek1945) and argue that there are indeed parallels between the market process that informs the actions of private actors and the democratic process that informs public actors. In much the same way that the market process generates the very knowledge that is required to engage in rational economic calculation, the democratic process helps generate the knowledge necessary for solving collective action problems. This, they argue, helps explain why Western democracies have done better than non-democratic nations in getting rid of the most harmful economic policies (e.g. nationalization of industries, price controls, etc.).
However, although market and political entrepreneurship share some of these similarities, there are also many important differences between the two. Market entrepreneurs receive more constructive and immediate feedback than their political counterparts (Mises, Reference Mises1944).Footnote 11 Price signals allow them to quickly adapt to constantly changing circumstances and to assess their performance in real time. They also help entrepreneurs identify errors and convey valuable information to guide them on the best ways to correct these errors moving forward. The feedback signals that political entrepreneurs rely on, in contrast, are far weaker than market price signals and often come with longer lags. Politicians, for instance, have to wait years for elections to take place to receive concrete feedback from voters. Furthermore, the inability to meaningfully aggregate consumer/voter preferences also reduces the coordinative capacity of democratic elections.
The selection mechanism for allocating resources toward more successful political entrepreneurs is also more difficult than it is in markets. Market entrepreneurs typically own and exercise full control over the resources they exercise control over and thus directly bear the burden of uncertainty in the form of profits and losses. Political entrepreneurs, in contrast, don't typically own or fully control their resources, nor do they directly bear the financial costs of their decisions because the resources they control are jointly- or publicly-owned, and because the very definition of what is ‘public’ often shifts over time. In short, political calculation is beset by more obstacles than market calculation, particularly in a world of Knightian uncertainty.
Again, none of this implies that political entrepreneurs receive no meaningful feedback signals. Political entrepreneurs do have access to a number of non-market feedback signals. For example, politicians can reference election results, and bureaucrats can refer to the size of their budgets. Our core argument is that non-market signals do not provide feedback that is as accurate and immediate as that provided by market price signals. In markets, price signals help one class of market entrepreneurs coordinate their actions with the interest of one class of consumers whose interests are naturally aligned with theirs. Such coordination in politics is more difficult because it involves several disparate classes of entrepreneurs (politicians, bureaucrats, interest groups, and legislators) and consumers (voters, lobbyists, etc.) whose interests are not naturally aligned.
Uncertainty and judgment in the nonprofit sector
Who undertakes judgment?
The role of the non-profit or charitable entrepreneur is distinguishable from that of the social entrepreneur (Frank, Reference Frank, Shockely, Frank and Stough2008). Social entrepreneurship is defined as market entrepreneurship with an additional social end in mind, such as a coffee shop that donates a percentage of its revenue to local foodbanks and commits to employing homeless people or those temporarily unemployed. It thus earns monetary revenue as well as aims at achieving a well-specified ‘social’ end while doing so. Thus, social enterprises and social entrepreneurs can and do calculate profit as well as strive for social goals simultaneously and hence do not belong in the non-calculative action category.
Boettke and Prychitko (Reference Boettke and Prychitko2004) further distinguish between nonprofit entrepreneurship and not-for-profit entrepreneurship. Hospitals and universities fall within the latter category in that they can and do calculate monetary revenue and accounting cost, yet any profits generated are not distributed among stakeholders but are channeled into the operating costs of the enterprise. In this sense, the goal of not-for-profits is not to maximize monetary profit (Ben-Ner, Reference Ben-Ner and Rose-Ackerman1986; Ben-Ner and Gui, Reference Ben-Ner, Gui, Anheier and Ben-Ner2003; Frank, Reference Frank, Shockely, Frank and Stough2008; Hansmann, Reference Hansmann1980, Reference Hansmann and Powell1987) as they are legally bound by the non-distribution constraint.
True nonprofit or charitable entrepreneurship, however, does belong in the realm of non-calculative action. Here, the explicit goal is the achievement of a certain social goal such as charity or direct transfers to the poor, disaster recovery with the aid of community members, voluntary housing associations, religious groups that seek to bring spiritual awareness, philanthropic endowments that seek to promote education or awareness about environmental issues, etc. Such nonprofit enterprises make up an important part of civil society and play an important role in developing as well as in developed countries (Acs and Desai, Reference Acs, Desai, Acs and Stough2008; Acs and Phillips, Reference Acs and Phillips2002; Harvey et al., Reference Harvey, Maclean, Gordon and Shaw2011). In this case, the nonprofit entrepreneur plays the crucial role of coordinating between two disparate groups: benefactors/donors and beneficiaries. In order to do this, the nonprofit entrepreneur may act alone or through the creation of a nonprofit organization. More recently, nonprofit entrepreneurship has been enabled through technology and peer-to-peer giving through online platforms in the absence of a nonprofit organization (Nair and Miller, Reference Nair and Miller2020).
The crucial difference between socially useful action carried out by the nonprofit entrepreneurs and entrepreneurs who exercise judgment in government is the role of voluntarily raised or private funding. Not only do nonprofit entrepreneurs bear uncertainty and carry out judgment with regard to the preferences and ends of beneficiaries, but they must also solicit and raise funding privately from donors/benefactors interested in investing in the social cause at hand.
Tullock (Reference Tullock1966) provides an economic analysis of the similarities between charities and bureaucracies. He models the typical donor as analogous to a self-interested voter in a democracy, who tends to be rationally ignorant, as discussed above. Donors are thus unlikely to have strong incentives to acquire accurate information from the charitable entrepreneur about the true state of the project or the true state of the intended beneficiaries. Given the behavioral assumption of self-interested donors, Tullock (Reference Tullock1966) finds that charities are likely to maximize expenditures on promotional materials, marketing and soliciting donations rather than maximizing expenditures on improving beneficiary outcomes.
Tullock (Reference Tullock1966) posits that either ‘bureaucratic charities’ become similar to government agencies whereby the goal is to maximize donations/budgets, or they can tend toward becoming ‘ideal charities’ whereby the goal is to maximize expenditure on beneficiaries. Tullock (Reference Tullock1966) concludes that significant donor involvement, feedback and monitoring of the nonprofit entrepreneur is essential to avoid the fate of the nonprofit enterprise becoming overly ‘bureaucratic’. Haeffele and Storr (Reference Haeffele and Storr2019) provide a similar assessment: the effectiveness of the nonprofit organization and the nonprofit entrepreneur, they argue, is explained by whether the organization resembles a public or a private enterprise.
Which goals matter? Which preferences get satisfied?
As discussed above, the charitable or nonprofit entrepreneur bears Knightian uncertainty in the non-calculative realm of socially beneficial action. Either acting on his own, as in the case of peer-to-peer crowdfunding through an online platform, or embedded within a nonprofit organization, the nonprofit entrepreneur must judge the need/demand for a specific social good as well as secure/raise funding in order to bring it about.
In markets, consumers are easily identified, and market entrepreneurs seek to discover and satisfy their preferences (Kirzner, Reference Kirzner1997). Analogously, beneficiaries or the intended recipients of charitable endeavors are, in a sense, the true consumers of the nonprofit enterprise, as the main actions of the enterprise are directly aimed at them. In this sense, charitable or nonprofit entrepreneurs must engage in discovery and learning of what the needs of their beneficiaries are as well as seek meaningful feedback in order to learn about their own impact (Frank, Reference Frank, Shockely, Frank and Stough2008).
However, donors are rightful consumers as well (Andreoni, Reference Andreoni1990; Andreoni and Payne, Reference Andreoni, Payne, Auerbach, Chetty, Feldstein and Saez2013; Hochman and Rodgers, Reference Hochman and Rodgers1969; Sugden, Reference Sugden1982, Reference Sugden1984; Warr, Reference Warr1982). Altruism on the part of a donor implies the existence of inter-dependent utility, in that the donor is made better off when the beneficiary is made better off. In this sense, donors are consumers of the utility gained from knowing that the beneficiary was made better off and are willing or more likely to donate to those charities that they believe will bring about such results (Hochman and Rodgers, Reference Hochman and Rodgers1969).Footnote 12 Thus, the problem faced by the uncertainty-bearing nonprofit entrepreneur is made more complicated in that he must seek to satisfy the preferences of two distinct groups of individuals simultaneously: donors and beneficiaries.
However, the coordination of charitable activities with ‘consumer’ preferences is more likely than under democratic government, because there is only one class of entrepreneurs who must satisfy the ‘consumers’ (donors and beneficiaries), using privately raised funding. Furthermore, it follows that if donor and beneficiary interests are aligned with each other, in that donors are altruistic and genuinely care about beneficiary welfare, coordination of resources is more effectively achieved than when the two groups' interests are not aligned.Footnote 13
Error correction within nonprofits
Non-price learning and feedback mechanisms in the absence of monetary calculation become prominent within the context of charitable or nonprofit entrepreneurship and the existence of Knightian uncertainty (Boettke and Coyne, Reference Boettke, Coyne, Shockely, Frank and Stough2008; Boettke and Prychitko, Reference Boettke and Prychitko2004; Chamlee-Wright and Myers, Reference Chamlee-Wright and Myers2008). Perhaps the most significant error-correcting mechanism that exists in the nonprofit entrepreneurial realm is the voluntarily funded nature of project (Boettke and Coyne, Reference Boettke, Coyne, Shockely, Frank and Stough2008). Nonprofit entrepreneurs must be able to solicit and raise funding from voluntary sources and do not have the power to tax or direct access to tax revenue. In this sense, nonprofit entrepreneurs are more constrained than entrepreneurs in the government and closer to market entrepreneurs.
Short of losing funding, however, error correction or feedback must take place along non-monetary dimensions. Several such mechanisms are suggested in the literature on nonprofits. Boettke and Coyne (Reference Boettke, Coyne, Shockely, Frank and Stough2008) borrow from the literature on microfinance and suggest the use of reputational collateral as a useful device to evaluate nonprofit entrepreneurs. For the purpose of credibility, the entrepreneur should stake his own reputation on the charitable project that he chooses to invest in. Chamlee-Wright and Myers (Reference Chamlee-Wright and Myers2008) discuss the role of embeddedness in social networks as a useful enforcement mechanism for socially motivated endeavors. Evidence of this mechanism is found in real world social learning situations but also in online fundraising endeavors where research finds that online donors are much more likely to donate to a project where they already know and are familiar with the fundraiser in real life (‘relational altruism’), or are inspired by the fundraiser and identify with him rather than the project itself (‘the champion effect’) (Nair and Miller, Reference Nair and Miller2020).
4. Conclusion
Knightian uncertainty has been taken into consideration explicitly when it comes to markets and market process theorizing. This has not necessarily been the case when it comes to theorizing about democratic politics and civil society, despite the well-established understanding that Knightian uncertainty is ubiquitous and applies to all human action. Inside market-based decision-making, entrepreneurs bear Knightian uncertainty and undertake judgment while being guided by monetary prices and by the error-correcting mechanism of profit and loss. Within the institutional setting of markets, there is a strong inherent tendency for the coordination of productive and entrepreneurial decisions with underlying consumer preferences. This paper extended this analysis to ask three key questions of the institutional mechanisms present in democratic politics and nonprofits, which represent non-calculative decision-making: (1) Who undertakes the role of judgment under uncertainty? (2) Which goals matter and whose preferences are satisfied? And, (3) How is error detected and corrected without monetary profit and loss?
The paper found that both within democratic politics and nonprofits, there are strong conceptions of the political entrepreneur and the nonprofit entrepreneur as decision-maker and bearer of judgment. However, because there are multiple entrepreneurial classes such as politicians, interest groups, legislators, and bureaucrats, often with conflicting goals, as well as multiple ‘consumer’ groups such as voters (who tend to be rationally ignorant) and lobbyists, coordination of allocative decisions with consumer preferences in democratic politics is likely only in ‘special cases’, with such cases being the exceptions that prove the rule. For instance, this result is borne out by a well-developed literature in political science that stresses the advantages of local government over unitary or central government in terms of effectiveness, satisfying voter preferences and constraining government expenditure (Ostrom, Reference Ostrom2005). In nonprofits, our analysis suggests that coordination is much more likely, given that there is only one class of entrepreneurs that exercise judgment and the existence of private decision-making, albeit there are two distinct groups of ‘consumers’ (benefactors and beneficiaries). However, when benefactor and beneficiary interests are aligned (e.g. through benefactor altruism), effective coordination of resources is most likely to exist.
However, neither in the case of democratic politics nor in that of nonprofit civil society can one conclude that there is an inherent alignment of interests between entrepreneurs and consumers, in the way that these interests are aligned within the calculative institutional setting of markets with prices and monetary profit and loss. Because markets allow for the meaningful aggregation and comparison of dispersed consumer preferences and local knowledge through the common denominator of monetary prices, there is a systematic built-in tendency as well as a strong incentive for entrepreneurs to correct their own errors over time. Given the ubiquity and persistence of Knightian uncertainty, however, erroneous judgment is inevitable and understanding the role of error correction processes in non-calculative settings is crucial.
Perhaps somewhat paradoxically, democratic government with its structural mechanisms and constraints of term limits and election cycles contains more built-in error correcting mechanisms than does nonprofit civil society. Charitable nonprofits are neither constrained by having to maximize profit and avoid monetary loss, nor are they subject to the ‘public will’ through the constraints of elections and voter preferences. Instead, they are only constrained by private donor preferences. This suggests that the persistence of entrepreneurial error in nonprofits is also possible over time (though not necessary), given the lack of an in-built mechanism to correct error, even though nonprofits resemble market-based institutions on other margins (private cost and private decision-making).
Further research might look into applying the recent literature on the different types of Knightian uncertainty with its emphasis on options (means) and outcomes (ends) explicitly into research on civil society or democratic politics (Dequech, Reference Dequech2000; Packard et al., Reference Packard, Clark and Klein2017). Applying judgment-based analysis and Knightian uncertainty to study the internal organization of the government and charitable firm is also likely to be a fruitful endeavor (for instance, see Miller et al., Reference Miller, Klein and Nairworking paper). However, our analysis suggests that there are limits to enhancing efficiency and outcomes in government and charity through merely improving internal organization given institutional structures and error correcting mechanisms. Finally, epistemic humility points to the need to continue taking Knightian uncertainty explicitly into account in studying the ‘hard’ institutional cases of democratic politics and civil society (as well as other non-calculative settings), not just in understanding markets.