An Institutional Theory of the Nonprofit

Distinguishing Sectors From Institutions

This article views sectors and institutions as two very different types of organizing principles. According to Abzug (1999), “nation-states have a need to establish a sectoral system to refine regulation, taxation, and a wide range of public policies” (p. 131). The nonprofit sector is, therefore, usefully and predominantly defined according to characteristics identical or similar to the legal requirements set forth in Section 501 of the U.S. tax code. Under this sector-based approach, social enterprises that are incorporated as private, for-profit businesses are not considered part of the nonprofit sector no matter how much their activities may resemble those of nonprofit sector organizations. Likewise, such a definition does not include the so-called nonprofit “dark matter” that includes both informal voluntary activity and formal activity that occurs without registration with a government entity (Grønbjerg et al., 2010; Smith, 1997; Van Til, 2000).

Elinor Ostrom defines institutions as “the prescriptions that humans use to organize all forms of repetitive and structured interactions including those within families, neighborhoods, markets, firms, sports leagues, churches, private associations, and governments at all scales” (E. Ostrom, 2005, p. 3). These prescriptions are comprised of rules, norms, and strategies employed by individuals and organizations to achieve a wide variety of individual and group objectives. To observe nonprofit institutions, therefore, we are interested in what types of interactions are repetitively employed in what we would collectively identify as “nonprofit-type” activities, and when such interactions might be prescribed.

This distinction between institutions and sectors is not trivial. Although the institutional lens has been previously used to examine individual nonprofit organizations or service areas (Bushouse, 1999; Cooney, 2006; Feeney, 1997; Newhouse, 1970), only a very few nonprofit theorists suggest origins of nonprofit organization as a whole based on principles of institutional theory outside of the traditional economic market failure/government failure couplet (notable exceptions include Kramer, 2000, Ben-Ner, 2002, and Amendola et al., 2011). This means that although current sector theory may provide powerful narratives for the protections, restrictions, and incentives legally associated with nonprofit status, it does little to explain how and why people voluntarily organize themselves to carry out nonprofit-type activities even when no such legal protections, restrictions, and incentives exist (Brody, 1996a). Without institutions, nonprofits are a mere figment of our legal imagination.

Empirically, sector alone appears largely insufficient to make meaningful distinctions between organizations. In many cases, research has demonstrated that even on social outcome dimensions, sector does not predict important differences in performance (Heinrich, 2003; Leviten-Reid, 2010; Luksetich, Edwards, & Carroll, 2000; Rosenau & Linder, 2003; Witesman & Fernandez, 2013), management (Moore, 2000), purpose (Child, Witesman, & Braudt, 2014), or even public perception (Handy et al., 2010). Organizational form appears to be more an artifact of tax regulations than truly descriptive of unique organizational characteristics, motivations, functions, or even history (Brody, 1996a, 1996b). Given their failure to identify unique characteristics of nonprofits in the present, legalistic, sector-based definitions of economic activity seem inadequate to provide clear hypotheses about how nonprofits may evolve in the future.

New institutional economic theory—closely related to political economy and the public choice movement—focuses largely on the rules, norms, and strategies that human beings use to deal with the presence of certain biophysical characteristics of the world in which we live (E. Ostrom, 2005). In particular, these rules, norms, and strategies generally provide some mechanism for distributing goods within and/or across groups. It is from this perspective that this article seeks to enhance our understanding of nonprofit-type activity through propositions about the nature of nonprofit institutions—independent of sector.

Institutional Characteristics of Nonprofit-Type Activity

This article proposes a pair of criteria for defining nonprofit-type institutions. These criteria are not entirely novel; in fact, both have been made in various forms previously by nonprofit theorists (Ben-Zion & Spiegel, 1983; Hansmann, 1980; Salamon, 1999; M. Sefton & Steinberg, 1996 Sen, 1993; Steinberg, 1987). The contribution of this article is to identify these criteria as the two central characteristics of nonprofit-type activity, to justify their utility in defining the nature of nonprofit institutions, and to expand on their implications for the nature and role of nonprofit-type activity for distributing goods in society irrespective of sector.

Consumption Versus Redistribution

Among the core activities carried out by institutional actors are payment for the production/provision of the good and consumption of the good (E. Ostrom, 2005). In typical private market exchanges, the individuals who pay for the good are also generally the consumers of the good. We will call such arrangements “consumptive.” There are two distinct types of consumptive arrangements. The first is when all citizens are both payers and consumers (e.g., all citizens pay taxes for a certain benefit and all receive that benefit), and the second is when citizens are payers if and only if they are also the consumers (i.e., private market exchanges). In either case, consumers “get what they pay for” in such institutional arrangements.

In contrast, however, some institutions are arranged such that the payer relinquishes consumption of the good to another party, separating the roles of payer and consumer. Such an arrangement is termed redistributive. Separation of the roles of payer and consumer is central to both redistributive public policy (Lowi, 1979; T. Sefton, 2008) and the practice of donative redistribution (Hansmann, 1980; Salamon, 1999). In governments, taxpayers must frequently pay for public services under threat of sanction even if they are not beneficiaries of those services (T. Sefton, 2008). In nonprofit-type activities, donors frequently pay for services voluntarily to promote the availability of services to others and to express their own social preferences and values (Ben-Zion & Spiegel, 1983; M. Sefton & Steinberg, 1996; Steinberg, 1987).

Redistribution can take any of three forms: (a) The person who pays for the good never consumes it but allows one or more others to consume the good, (b) the person who pays for the good consumes the good and then relinquishes his or her ownership of the good to one or more others so that they may derive value from the good, or (c) the person who pays for the good consumes the good in conjunction with others who simultaneously draw value from it. In each of these cases, the individual who pays for the good is not the sole beneficiary of the good (or may not be a beneficiary of the good at all), and in all three cases, someone other than the payer derives value from the good. This redistribution of goods distinguishes nonprofit-type activities both from consumptive government regimes and also from pure private interactions, in which the purchaser of the good is the sole beneficiary.

Thus, in addition to the voluntary criterion, we define nonprofit institutions as those containing some form of redistribution from payer to consumer:

The redistributive criterion: Nonprofit-type institutions involve redistribution, meaning that the individual who pays for the good is not the (sole) beneficiary of the good.

These two dimensions—coercion versus voluntary action and consumption versus redistribution—can be used to define key differences between the primary macroinstitutions of society using a two-by-two matrix as represented by the top and left elements in Figure 1.

OPEN IN VIEWER

Figure 1.

Nature of goods and nature of institutions.

Exchanges that are both voluntary and consumptive in nature are identified as “market-type” arrangements, and are consistent with a free market economy in which individuals maximize personal utility. The use of a central authority to coerce either payment for goods or consumption (or nonconsumption) of goods is viewed as a “government-type” interaction. Government-type interactions are further divided into two types: those that do not involve redistribution (government-type I) and those that do (government-type II). Finally, voluntary and redistributive exchanges are identified as nonprofit-type activities.

Invoking these two criteria for nonprofit institutions will create different boundaries around nonprofit-type activity than do legalistic sector criteria. A legally recognized nonprofit organization or government agency could sell services using a fee-for-service model, in which consumers choose to purchase them and no redistribution would occur (e.g., museum souvenir shops, nonprofit sale of fairly traded goods, sale of postal products and services). Such a model would involve a voluntary consumptive institutional arrangement, or market-type exchange. Likewise, a government agency or for-profit organization could solicit donations to support a service it wanted to provide (e.g., donations to support a public park of for-profit hospital). Because donors voluntarily provide funding for benefits that accrue to others, this would describe a nonprofit-type or voluntary redistributive institutional arrangement, even though the organizations involved would be considered outside the formal “nonprofit sector.”

The Nature of Goods

Institutional rules, norms, and strategies do not develop in a vacuum. Instead, they generally evolve in response to specific challenges posed by the biophysical world (Bromley & Meyer, 2014; E. Ostrom, 2003, 2005). Several scholars (Ben-Ner, 2002; Billis & Glennerster, 1998; Bushouse, 1999; Ferris & Graddy, 1999; Kingma, 2003; Weisbrod, 1977) have noted the need for distributing specific types of goods as a defining reason for the presence of the nonprofit sector.

One useful approach to understanding the nature of goods and their inherent relationship to the development of distribution mechanisms is a matrix describing their subtractability and excludability. Ostrom and Ostrom proposed this framework in 1977 based on the work of Samuelson (1954) and Olson (1965). Excludability is characterized by a situation in which it is “costly to exclude individuals from enjoying benefits” (E. Ostrom, 2005, p. 24). This article will employ the following operational test for “costly” versus “not costly” exclusion: If the default position of a good is exclusion—that is, a proprietor must take some form of positive action to make the good available to others—then it is said to be “excludable” because the positive action required to share the good with others would impose a cost.

For excludable goods, this cost associated with inclusion may create a barrier to the distribution of some goods, which might otherwise be desirable. This excludability problem is a hallmark of some persistent issues in society: There are goods that we want to make available to more people, but there are high associated costs. Such goods include intangibles such as health, education, job skills, and well-being; they also include tangible goods such as housing and vaccines.

According to the nature of goods framework, subtractability is “when the use of a flow of [goods or] services by one individual subtracts from what is available to others” (E. Ostrom, 2005, p. 25). A good is subtractable if consumption of that good by one person substantively diminishes the value of the good either for other simultaneous consumers or for subsequent consumers. This concept is not perfectly binary, but rather exists on a scale from perfectly subtractable to perfectly non-subtractable. A good is perfectly subtractable if consumption of the good completely precludes any value from being derived by any other potential consumer. However, a good is perfectly non-subtractable when an unlimited number of consumers can derive the whole value of the good either simultaneously or serially. Between these extremes lie cases in which use of a good by one consumer diminishes—but does not preclude—the value that can be derived by others.

The lower and right elements of Figure 1 create a two-by-two matrix describing the excludability and subtractability of goods. Excludable and subtractable goods are “private” goods, non-excludable and non-subtractable goods are “public” goods, excludable and non-subtractable goods are “toll goods,” and non-excludable and subtractable goods are “common-pool resources.” Each type of good is associated with its own distributional challenges. These include challenges in encouraging payment for public goods (the “free rider” problem), preventing over-consumption of common-pool resources (the “tragedy of the commons”), and in the case of publicly desirable toll goods, the problem of exclusion.

Nature of Goods and Nature of Institutions

There is a logic to the association of institutional arrangements with specific types of goods. Because each type of good may be associated with its own unique distributional challenges, the evolution of institutions surrounding the distribution of each type of good is likely to aim at ameliorating those challenges (E. Ostrom, 2005). In general, non-excludability of goods provides a potential justification for coercive institutional regimes, excludable goods are suited to distribution through voluntary regimes, subtractable goods can be logically associated with consumptive regimes, and non-subtractable goods are naturally suited to redistributive regimes.

Where goods are non-excludable, a free rider problem may exist that markets and other voluntary arrangements are ill-equipped to manage. Because it is difficult for proprietors of non-excludable goods to prevent others from consuming them (creating a free rider problem for public goods and over-consumption for common-pool resources), use of a central authority with coercive power may be valuable for assigning the roles of payer and consumer (Brennan & Buchanan, 1980; Stoker, 1998). This “failure” of market-type institutions to usefully distribute non-excludable goods is one common justification for the presence of government-type institutions (Bator, 1958; Stiglitz, 1989; Weimer & Vining, 2005; Weisbrod, 1977). Where goods are non-excludable, we can expect the evolution of coercive institutions to identify and enforce the roles of payer and consumer to protect against free riding and over-consumption.

When goods are excludable, a central authority is not necessary for identifying the roles of payer and consumer because such roles are self-enforceable. In other words, if a good is excludable, it is possible for the proprietor of the good to prevent others from consuming it—or to require payment for it in the case of exchange—so a central authority is not as necessary (Bushouse, 1999). For excludable goods, voluntary institutional arrangements allow individuals to identify for themselves how much they value a good and to make exchanges accordingly.

Subtractable goods are naturally suited to consumptive institutional regimes. Because the value of subtractable goods decreases or disappears upon consumption, regimes in which the people who pay for the good also consume the good are logical because basic economic theories of the rational actor assume that a person will only pay for a good if they derive some value from it. Although exceptions to this “self-interested, rational actor” hypothesis will be discussed later, traditional market theory presumes that consumptive regimes evolve in response to the distribution of subtractable goods on the basis that it is both logical and equitable that people who value a good should be able to consume it upon payment in an appropriate formal exchange (Weimer & Vining, 2005).

When a good is non-subtractable, a proprietor may consume the good without diminishing its value. Once the good is paid for, therefore, additional consumers may consume the good without detracting from the proprietor’s ability to derive value from the good. The proprietor may derive as much value as she prefers, and because the good is non-subtractable, a Pareto-improving solution would be to allow others to benefit from the good as well. Thus, non-subtractable goods are naturally suited to redistributive regimes, because adding more beneficiaries without detracting value from the proprietor is Pareto-improving. In the words of Powell and Steinberg (2006), allowing additional consumers access to such goods would “help them, hurt no one, and consume no additional scarce resources” (p. 121).

We can extend the logic of the preceding paragraphs to determine the institutional characteristics we might expect to evolve in response to each of the four types of goods identified by the nature of goods framework. Namely, we would expect market-type regimes to evolve in response to the distributional characteristics of private goods, government-type regimes to evolve in response to the distributional characteristics of both public goods and common-pool resources, and voluntary redistributive (nonprofit-type) regimes to evolve in response to the distributional characteristics associated with toll goods. To demonstrate these “institutional matches,” the nature of goods framework and institutional characteristics are overlaid into the same matrix in Figure 1.

This set of proposed matches between the nature of goods and institutional regimes suggests the possibility that the primary macroinstitutions of society—governments, markets, and nonprofit-type institutions—evolved in parallel rather than in sequence, each in response to the distributional characteristics and constraints of a different set of goods. Although this observation in no way negates the contributions of market failure, government failure, and other major theories contributing to our conception of nonprofit theory, it does suggest a paradigmatic shift in how these theories can and should be understood. Centrally, markets are not appropriately viewed as the dominant, original, or primary distributional mechanism for all goods. Neither are government or nonprofit institutions secondary institutions in any sense. Each type of institution has its own independent rationale for existence.

These observations suggest explanations for the significant role of churches and other nonprofit-type organizations throughout the history of society, and their largely parallel development with both market institutions and the state (Morris, 2000). They also suggest predictions about the circumstances in which nonprofit-type activity (voluntary redistributive exchange) would be both predicted and prescribed: Such would be the case whenever the good to be distributed could be appropriately described as both excludable and non-subtractable—a toll good.

Toll Goods and Club Theory

Toll goods have received various treatments in public choice and related fields, most frequently being loosely equated with public goods. Toll goods have been alternately termed club goods (Buchanan, 1965; Sandler & Tschirhart, 1997), excludable public goods (Swope & Janeba, 2005), local public goods (Batina & Ihori, 2005; Cornes & Sandler, 1996), semipublic goods (ten Raa & Gilles, 2005), and impure public goods (Sandler & Tschirhart, 1997), depending on the specific good characterization framework applied by the authors.

Perhaps the most common formulation and cohesive expression of toll good theory is “club theory,” including research that has examined aspects of the nonprofit sector from this perspective (Bushouse, 1999, 2011; Kushner & King, 1994). Club theory and the more general notion of toll goods differ in that for many toll goods, exclusion is the problem associated with the distribution of goods rather than a purported solution (Bushouse, 1999). This assertion has been supported in at least one economic model of club goods and other papers that tackle the distributional problems associated with toll goods (Bushouse, 2011; Kennedy, 1990; Schweik, Mergel, Sandfort, & Zhao, 2011).

Recent evolution in club good theory has shed new light on the class of toll goods. Traditionally, toll goods and club goods have been viewed as a very limited set—either as “special cases” of public goods (such as public roads or bridges that have been excluded from general public use through the application of toll booths) or as benefits that derive from participation in formal membership organizations. More recently, scholars have begun to return to the definition of toll goods identified by V. Ostrom and Ostrom (1977), recognizing that a wider variety of goods may meet this definition (Bushouse, 1999; Potoski & Prakash, 2009). Potoski and Prakash (2009) suggest that even in the absence of a formal membership organization, adherence to a shared value set creates a “brand” that behaves as an excludable, non-exhaustable club good. “Members” are those who share the characteristic (in their case, participation in voluntary regulation schemes), and “nonmembers” are those who do not share the characteristic.

This back-to-basics approach to toll goods can be expanded to include other “clubs” that might denote shared privileges or resources that are not necessarily distributed by a central structure or organization, but which are both non-exhaustible and excludable. Under this paradigm, any intangible good that can be simultaneously enjoyed by multiple users would meet the non-exhaustibility criterion, and any such good that requires positive effort to extend its benefit to new consumers would also meet the excludability criterion. As examples, a wide variety of knowledge, skills, and values would all qualify as toll goods. In each of these cases, more than one person can “own” or “consume” the good simultaneously without diminishing the value of the good for other users. Also, in all cases we would expect some sort of positive action to extend ownership of the good from one individual to another (Bushouse, 1999). Knowledge must be shared, skills must be taught, and values, attitudes, and beliefs require significant effort to be communicated to others.

While traditional club theory suggests that the primary problem associated with clubs is congestion or crowding (Reiter & Weichenrieder, 1999; Sandler & Tschirhart, 1997), new club theory turns this idea on its head. Potoski and Prakash suggest that rather than making other members of clubs worse off, “. . . one firm’s membership can often enhance the value that other members receive” (Potoski & Prakash, 2009, pp. 2-3, emphasis added). This suggests that for beneficial toll goods, the natural distributional problem is that because of their excludability, they are by default inaccessible to members of the relevant public who might benefit from them. Powell and Steinberg (2006) term this distributional inefficiency an “overexclusion market failure.”

If we recognize that the main problem with distributing toll goods is the need to overcome their de facto state of exclusion, we can work to devise institutional mechanisms to overcome the barriers to inclusion (Bushouse, 1999). Among these mechanisms may be the employment of voluntary redistributive institutions. Rather than central institutions with coercive power being the primary distributors of such goods, efficiencies may be found in decentralizing the responsibility for the distribution of goods, allowing individuals and firms to voluntarily disseminate or redistribute toll goods.

The Lack of Biconditionality

In this article, it has been proposed that toll goods are naturally suited to distribution through voluntary redistributive (nonprofit-type) institutions. But clearly, it is not the case either that all toll goods are or should be distributed through nonprofit-type activity, or that toll goods are the only type of good that can or should be distributed in this way. In other words, the biconditional logic that only toll goods are distributed by nonprofits and that nonprofits only distribute toll goods is false. There are two primary reasons this is the case.

The first is that the approach of identifying “natural” matches between the nature of goods and specific characteristics of institutions relies heavily on an assumption of economic self-interest. In other words, we assume that actors will only pay for and value a good if they themselves can consume it. We know that this is not always the case—sometimes people derive personal value from providing goods to others, or seek to contribute to the overall well-being of community groups of which they are part (Rose-Ackerman, 1996; Schwartz & Howard, 1984; Simon, 1993). The role of altruism in the selection of voluntary redistributive regimes should be obvious: If a person can derive personal or social value from voluntarily purchasing a good for someone else, then a regime of voluntary redistribution can occur for any good that is the subject of such altruism. Thus, in the presence of altruism, we would expect to see goods other than toll goods distributed via voluntary redistribution.

Second, institutions are dynamic, subject to sociological forces that allow for complex and unpredictable learning processes. People and organizations tasked with the distribution of specific goods may seek models and ideas from other organizations or sectors, and may innovate to identify new solutions to old distributional problems. This process, which includes the dynamics of isomorphism, institutional learning, social innovation, entrepreneurship, and organizational sociology may override both tradition and prescription (Bromley & Meyer, 2014; DiMaggio & Anheier, 1990; DiMaggio & Powell, 1991). Thus, we would expect toll goods to be distributed in a variety of institutional arrangements. While this does not negate the potential value of theoretical prescriptions about institutional matches between the nature of goods and specific types of institutions, it should be clear that such observations are among the many influences actually shaping the practice of distributing goods in society.

The Dark Side of Voluntary Redistribution

Although a main objective of this article is to suggest a natural match between toll goods and voluntary redistributive institutions, this is not to say that voluntary redistribution is without its problems. The two primary problems stem directly from the nature of these institutional regimes. Specifically, voluntary regimes are subject to the idiosyncratic preferences of donors and consumers, and redistributive regimes are subject to significant information asymmetries.

Voluntary regimes—including both market-type and nonprofit-type institutions—rely on individual preferences to aggregate into a satisfactory social whole. Even when voluntary distribution mechanisms are efficient in an economic sense, they may be subject to “public value failure,” in which the result does not meet the value-based preferences of the community or society as a whole (Bozeman, 2002). Without a coercive mechanism to monitor and ensure social outcomes at a holistic level, it is possible that the microeconomic decisions of individual actors will not yield optimal levels of desirable goods in society. When production and distribution of goods relies on voluntary mechanisms, the outcomes are subject to individual—rather than community—needs and values. This may produce unsatisfactory results.

Redistributive regimes can also create problems for achieving optimal results. The separation of payer and consumer roles results in information asymmetry (or “contract failure”): Because the payers do not consume the goods, they have less information than those who do consume the goods about its value (Billis & Glennerster, 1998; Bushouse, 1999, 2011; Hansmann, 1980; Weimer & Vining, 2005). When the goods in question are intangible, they may already be subject to significant information asymmetries that make it difficult to assign value (Weimer & Vining, 2005). Thus, the redistributive separation of consumer and payer may exacerbate informational problems that already exist.

Hansmann (1980) uses these issues as justifications for the legal characteristics of the nonprofit sector in the United States, though it is unclear how effective such mechanisms as the non-distribution clause and board governance really are at resolving them (Brody, 1996a; Bushouse, 1999, 2011; Valentinov, 2008) or whether other solutions may exist.

Implications for Theory and Practice

The theory presented here suggests that nonprofit-type activity may be viewed as an institutional response to the excludability problem in the distribution of publicly desirable toll goods. This proposition suggests important revisions to our view of the relationships between government, markets, and the voluntary sector. Market failure theory, government failure theory, contract failure theory, and voluntary failure theory each contribute elements to the understanding presented here, but the present theory suggests revisions to—and connections between—each of these preceding theories.

One important theoretical contribution suggested by the institutional theory of nonprofits is that markets, government, and voluntary institutions are parallel rather than hierarchical, in contrast to the market failure-leads-to-government-failure-leads-to-nonprofits narrative that has been so prevalent in the literature. Markets should only be considered the default institution when private goods are the subject of discussion. By extension, “market failure” suggests specific problems associated with the assumptions central to the distribution of private goods through market institutions, and “government failure” refers to problems associated with coercive regimes in distributing non-excludable goods. If these assumptions are broken—including assumptions about the nature of the good—inefficiencies may result.

Voluntary failure, then, should be extended to suggest the consequences inherent in relaxing the assumptions necessary to result in efficient distribution of toll goods through voluntary redistribution. This is a marked difference from its current construction, but potentially compatible with the intent of its author to correct both the “gross neglect of the nonprofit sector in most assessments of American social welfare policy” and to “acknowledge the significant and inherent shortcomings of voluntary organizations” (Salamon, 2003, p. 183).

A second major contribution of the institutional theory of nonprofits is that it suggests a rationale for the role of the nonprofit and voluntary sector even in informal settings or when markets and government institutions might be underdeveloped or in a significant state of flux, where other traditional theories of nonprofit and voluntary action have fallen short (Salamon & Anheier, 1998). Whereas market, government, and contract failure theories suggest that other key social institutions precede nonprofit and voluntary action, this institutional theory suggests circumstances in which we might expect to see the emergence of institutions that are characterized by voluntary redistribution. Where we see toll goods with marked overexclusion failures, we would expect to see voluntary redistributive regimes evolve in response.

Finally, perhaps the most important result from the theory presented here is the idea that the core problem associated with the distribution of desirable toll goods is overexclusion. This insight may provide greater traction in understanding what types of distributional solutions—including voluntary redistributive regimes—may work to overcome the exclusion barrier to raise the quality of life for all members of society.