Rereading Salamon: Why Voluntary Failure Theory is Not (Really) About Voluntary Failures

1 Introduction

As the contributions to Nonprofit Policy Forum’s Lester Salamon Memorial Issue (vol. 14: 3&4) suggest, Voluntary Failure Theory, also known as Interdependence Theory, remains among the most salient of Salamon’s conceptual contributions to nonprofit studies globally. Contributors attest to the theory’s usefulness in illuminating contexts as diverse as the founding period of Israel (Kabalo and Almog-Bar 2023), local public service provision in Iowa (Shafiq, Albrecht, and LeRoux 2023), and others in between. Yet, the theory also has its limitations—both conceptually and empirically. Empirically, there have been questions about how the theory can be tested, or whether it is even testable in the first place, as raised by Julian Wolpert’s (2003) review of it. Conceptually, the theory is focused on service provision, at the expense of other important functions and activities of nonprofits (e.g. advocacy and philanthropy), and draws its core strength from relationships in the health and human services fields that may differ quite significantly from other fields of nonprofit activity (Grønbjerg and Smith 2021). This short note seeks to problematize some of these issues in hopes of stimulating further engagement with, and deeper discission of, his theoretical legacy.

2 Voluntary Failure Theory: Viewing It Correctly

When Salamon departed the Carter Administration for the Urban Institute in 1980, he set out to develop his Third-Party Government Theory (Salamon 1981; Salamon 1995, Chapter 1) with no particular intention to make nonprofits a major research focus until approached by the leadership of Independent Sector to conduct some work in this area (see Editorial of NPF 14:3). As he later recounted during staff meetings of the Johns Hopkins Comparative Nonprofit Sector Project, he agreed to do so because nonprofits were part of the third-party government theory anyway and he might as well have started there—without anticipating that the study of the nonprofit sector would become a focal point of his life’s work. To fully understand the origins of Voluntary Failure Theory, it is important to appreciate that “two strands of theorizing form the core of this voluntary failure line of thought” (Salamon and Toepler 2015: 2162; see also Salamon 2003): the first is his earlier Third-Party Government Theory as the basic foundation; and the second is an extension in reaction to Weisbrod’s (1975) market failure/government failure argument.

2.2 The Voluntary Failure Extension

Salamon’s core contention that it is crucially important analytically “to differentiate between government’s role as a provider of funds and direction and government’s role as a deliverer of services” (Salamon 1995: 41) is a direct implication of the third party government conceptualization, which provides the conceptual underpinning of Voluntary Failure Theory. Indeed, Voluntary Failure Theory should be understood as an extension of Third-Party Government and an elaboration on the role of nonprofit providers in it (Salamon and Toepler 2015).

The motivation for extending Third-Party Government Theory was Salamon’s (1995) contention that Weisbrod’s argument that nonprofits exist because of the “twin failures” of both markets and government led to erroneous conclusions about the nature of sector relations. Specifically, Weisbrod’s theory suggested a “paradigm of conflict,” marked by a zero-sum game between the sectors, that was not consistent with empirical reality and the conceptualization of Third-Party Government Theory. In contrast, Salamon posited a “partnership paradigm” in which a mutually beneficial relationship exists, based on a division of labor between the sectors, in which government often assumes the financing of public services and nonprofits take on the delivery function.

For nonprofits, the benefits lie in access to public resources and the ability to scale operations, as well as legitimacy gains deriving from partnering with government. The benefits for government to utilizing pre-existing nonprofits on the ground and the expertise these organizations already have lie in lower transaction costs vis-à-vis establishing new government agencies to provide similar services and in political expediency. Beyond that, nonprofits have certain shortcomings—the four voluntary failures—that additionally can serve to “justify government support” (Salamon 1995: 44): philanthropic insufficiency, paternalism, particularism, and amateurism.

While these voluntary failures suggest why nonprofits cannot carry public service provision at scale without help from the government, government has the ability to counter these failures through the power to tax and overcome free ridership (philanthropic insufficiency); enforcing equity considerations and creating entitlements (philanthropic particularism); prioritization through democratic decision-making (philanthropic paternalism); and imposing standards and certification requirements (philanthropic amateurism). As such, the “interdependence theory moves away from the zero-sum thinking that characterized some of the economic theories …, shows nonprofit–government relations less in a competitive light, and emphasizes collaboration instead” (Anheier and Toepler 2023: 173).

As part of the argument, Salamon (1987, 1995 argues that it is nonprofits rather than government that emerge first to attend to new collective action needs that are not addressed by the market. The reason for this is that voluntary action is more easily mobilized than government responses, which require that “substantial segments of the public must be aroused, public officials must be informed, laws must be written, majorities must be assembled, and programs must be put into operation” (Salamon 1995: 44). Clearly these obstacles remain in force as voluntary failures limit the nonprofit response from scaling up. As suggested in Figure 1, the voluntary failures primarily explain why nonprofit responses typically remain limited without government involvement. The principal reason is the free ridership market failure (i.e. Salamon’s philanthropic insufficiency) which can be overcome in very small groups through social pressure and selective incentives (hence nonprofits respond initially at a small scale), but requires coercive government action to overcome in larger groups.

Figure 1:

Configuring Salamon’s voluntary failure and third-party government arguments.

However, without a majority-approved public policy interest in addition to the expectation of transaction cost savings and service quality gains vis-à-vis direct government provision, there is no compelling reason for government to extend support in order to remedy voluntary failures for their own sakes. Weisbrod’s (1975) median voter argument still applies and majorities may not be persuaded to tackle free ridership on all occasions. In such cases, government has no mandate to intervene and remedy voluntary failures (Figure 1). If a majority can “be assembled” to address the collective need, however, then government can either directly provide the goods or services, which presumably will happen if there is no appropriate or sufficient nonprofit presence on the ground; or the Third-Party Government route will be chosen, if the use of existing nonprofits reduces transaction costs and furthers political expediency. In either case, the benefits to government—rather than the voluntary failures per se—motivate public support and set the collaborative relationship in motion.

3 Is Voluntary Failure Theory About Voluntary Failures?

Seen this way, the voluntary failures are a rather ancillary aspect of Voluntary Failure/Interdependence Theory, which is at its core about justifying positive, collaborative government-nonprofit relations in line with Third-Party Government Theory. Featuring voluntary failure prominently in the title of his 1987 article (Salamon 1987) was a neat way to connect his argument to the prevailing “failures” discourse in the theorizing about the emergence of the nonprofit sector, i.e. Weisbrod’s (1975) twin failures and Hansmann’s (1980) contract failure. But in contrast to Weisbrod and Hansmann, Salamon’s focus was not on proposing a rationale for the existence of nonprofits but rather aimed at working out a rationale for collaborative relationships between the sectors.

The voluntary failures themselves are not crucial for the development of this rationale. A reflection of this is that Salamon himself tended to avoid the Voluntary Failure Theory moniker in favor of Interdependence Theory (see Salamon and Anheier 1998; Salamon 2003), which more accurately reflects the core of the theory. Moreover, Salamon and Anheier’s (1998) effort to test prevailing theories using Johns Hopkins Comparative Nonprofit Sector data used government social welfare spending as a measure to test Interdependence Theory rather than voluntary failure-based variables: the test was to see whether higher government social spending leads to large nonprofit sectors cross-nationally. Numerous subsequent studies (e.g. Bae and Sohn 2018; Jeong and Cui 2020; Kim and Kim 2018; Lecy and Van Slyke 2013; Liu 2017; Lu and Xu 2018) likewise successfully tested Voluntary Failure/Interdependence Theory without referencing the voluntary failures; or while primarily focusing on Weisbrod’s government failure, like Paarlberg and Zuhlke (2019: 117), find “evidence of complementary, interdependent relationships between government and the nonprofit sector,” the Interdependence Theory‘s core contention.

Wolpert’s (2003) key criticism of Voluntary Failure Theory was that it was difficult to operationalize and empirically test the voluntary failures and thus to validate the theory. While he was correct on the difficulty of operationalizing the voluntary failures, research since has disproven his contention that Voluntary Failure/Interdependence Theory cannot be validated.^[1] For testing purposes, the four voluntary failures are a bit like a red herring: the theory can be well operationalized without recourse to them.

A key reason for this red herring lies in the differences between Salamon’s intention, on the one hand, and Weisbrod’s and Hansmann’s, on the other. The idea of voluntary failure makes it tempting to equate them conceptually with market failures (public goods, information asymmetry) and government failure (median voter, majority constraint) and extend the twin failure into a triple failure of all three sectors (Steinberg, Brown, and Taylor 2023). Unfortunately, giving in to that temptation is a common, but false equivalency that even Wolpert (2003) fell prey to when he wrote that “the nonprofit theories of the type proposed by Salamon, Weisbrod, and Hansmann are courageous attempts to unify the field. Their fundamental contributions have provided important insights about the rationale of nonprofits among private and public sector activities and have uncovered some commonalities that merit a unifying framework” (p. 171); then further asking among his “annoying questions” about Voluntary Failure Theory: ”Are there both necessary and sufficient conditions for a nonprofit sector to emerge?” (p. 174). However and as indicated above, Salamon did not set out to create a failure rationale for the existence of the sector. That nonprofits exist, often before government action, is a historical given for Salamon (1987, 1995, explainable by the greater ease of mobilizing voluntary rather than governmental responses and supportable through historical case studies (e.g. Kabalo and Almog-Bar 2023).

Accordingly, the assumption that these theories serve as alternative, but equivalent rationales incorrectly elevates the voluntary failures to a presumptive explanation for the emergence of nonprofits (and hence to the core of Voluntary Failure/Interdependence Theory). If that had been their purpose, then a case for operationalizing and testing them as a condition for nonprofit growth and development could have been made. But it is not voluntary failures that explain nonprofit growth, but government partnerships that channel public support to the sector through various government tools, as Third-Party Government Theory suggests.

This is not to say that the concept of voluntary failure is without analytical merit. Many studies (Jones and Deitrick 2020; LeRoux 2009; Robbins and Lapsley 2008) and contributions to this Memorial Issue (Jang, Valero, and Ford 2023; Shafiq, Albrecht, and LeRoux 2023) employ the four failures in meaningful ways and provide new attempts at operationalizing them. Conceptually though, the voluntary failures remain problematic. Salamon himself never specified whether the four failures were meant to represent organizational failings, or shortcomings of nonprofits at the community, field or sectoral level. The distinction is clearly important, as, for example, particularism can be detrimental at the community level when it leads to duplication of services and/or the exclusion of minority groups, but at the organizational level it represents a desirable expression of the organization’s core values that gives it its distinctiveness (Anheier and Toepler 2023). Since Salamon characterized particularism as a failure, this might suggest that the voluntary failures should be seen as systemic rather than organizational problems.

Looking at voluntary failures as organizational level failures creates the additional problem of distinguishing between voluntary and other kinds of failures. While there has been consistent interest in organizational failure of nonprofits (Anheier 1999; Hager et al. 1996; Dolšak and Prakash 2022), and Seibel (1989; 2022 even suggested that nonprofits often fail successfully and permanently, Salamon’s voluntary failure concept specifically focused on failures that justify government to step in and help remedy them. Not every nonprofit failure is therefore a voluntary failure and sorting out which are which provides ample ground for a research agenda. As the voluntary failures are not quintessential for the Voluntary Failure/Interdependence Theory, such an agenda would also need to clarify where failure analyses could further inform research progress. Arguably, Salamon’s partnership paradigm has laid the basis for collaboration research, which might provide a suitable tie-in for organizational-level failure research.

4 Conclusions

Salamon’s partnership paradigm has fundamentally shaped nonprofit research globally for more than three decades, although it has increasingly come under pressure in recent years (Toepler et al. 2023). Even in strongholds like Germany, the continual assault of neoliberal policies has weakened the position of nonprofits and allowed the government to see itself as less interdependent with the sector than before (Zimmer and Priller 2023). Nevertheless, Salamon’s Voluntary Failure/Interdependence Theory continues to provide fruitful openings for future research and will remain a core concept shaping our understanding of nonprofits and their relationship with government.