How Collective Action Problems Suppress Participation in Nonprofit Trade Associations

Abstract

In seeking to understand how nonprofits participate in policymaking at the state level, scholars of the third sector tend to overlook or assume a barrier to this activity: collective action problems. I show that such problems suppress charter school participation in their trade associations. An analysis of original survey data and informant interviews combined with Internal Revenue Service data for the complete set of charter membership associations in the United States demonstrates that schools’ participation in these collectives follows a classic pattern of collective action problems: In states where the number of potential members is high, participation rates are lower. Across states, the size of the population of schools eligible for membership better explains variation in participation than other factors typically deemed important by scholars, such as organizational resources and policy environments. The finding supports the argument that large numbers inhibit participation in associations that pursue collective goods for their industry.

Keywords

trade associations collective action participation free riding nonprofit advocacy

Local governments provide critical venues for nonprofit involvement in the policy process (Berry & Arons, 2003), but capacity for policy advocacy on behalf of nonprofit organizations (NPOs) is uneven at the state level (Abramson & McCarthy, 2012). This study uses collective action theory (Olson, 1965) to consider variation in engagement in state associations of charter schools. From Olson’s standpoint, actors weigh participation costs against expected policy benefits as well as against the likelihood of being pivotal to policy change. As the chances decrease that one’s efforts will make the difference for policy advocacy collaboration, the collective action problem (CAP) strengthens and rational actors should prefer to free ride—that is, to forgo participation while reaping collective goods gained by those involved. Such goods benefit an entire population (see Tschirhart, 2006, p. 528), including advocates as well as those sitting on the sidelines. For example, a charter school serving special needs children benefits from favorable regulations that were negotiated by the state charter association, but the school chooses to not join the association. Studies have applied this framework to explain the character of nonprofit trade and infrastructure associations (e.g., Abramson & McCarthy, 2012; Esparza et al., 2014; Young, 2010); however, empirical support for the theory has been lacking in nonprofit advocacy literature. This want of evidence is unfortunate because scholars of the third sector question the validity of Olson’s argument (e.g., Chen, 2018, p. 98S) and remain reticent about the existence of CAPs even when offering insights consistent with CAP theory (see MacIndoe & Beaton’s, 2019, allusion to free riding, p. 80; and Salamon et al.’s, 2008, finding of lower participation in broad vs. issue-specific advocacy collaboration). Statistical findings from this investigation address this indeterminacy by offering a demonstration of CAPs suppressing participation.

Collective Nonprofit Advocacy

Nonprofit researchers call for careful consideration of collective forms of advocacy (Boris et al., 2014; Mosley, 2012). Scholars attentive to collaborative advocacy have typically taken the participating nonprofit as the unit of analysis (e.g., Donaldson, 2007; Li et al., 2017), so collectives seem to be extensions of the activities and needs of nonprofits. Collaboration is articulated as one strategy among many for pursuing policy (Almog-Bar & Schmid, 2014; Guo & Saxton, 2010; Mosley, 2010), and a technique for facilitating advocacy or enhancing advocacy capacity. Nonprofit managers report coalitions as the best way to overcome barriers to engagement (Mosley, 2014). Collectives provide opportunities to learn skills required for political activity, such as how to influence agency rulemaking (Leroux & Goerdel, 2009), and thereby increase capability for NPO advocacy (Schmid et al., 2008). Nonprofit policy collaborations are the way NPOs pool scarce resources and so participate in the policy process without cutting deeply into service delivery priorities (e.g., Leroux & Goerdel, 2009; Mosley, 2013) and are efficient vehicles for advocating (e.g., Boris & Maronick, 2012). These advances in the literature convey how collaboration amplifies the voice of NPOs while shaping their advocacy behavior, but are less concerned with exploring the nature of nonprofit advocacy collectives.

Given the scholarship on nonprofit associations (Tschirhart, 2006), we can expect NPOs to use a variety of organizational forms for collaboratively advancing their policy interests. Descriptive information, when pieced together, suggests a diverse ecology. For policy advocacy, NPOs join collectives that are formal coalitions, informal networks, infrastructure and trade associations, oriented broadly to the third sector, specific to their field or issue, based on memberships varying in size from a dozen to hundreds of organizations, and purposed for the national, state, and local levels (see Abramson & McCarthy, 2012; Berry & Portney, 2014; Fyall & McGuire, 2015; Salamon et al., 2008; Sandfort, 2014). The survey research articulates participation usually under general headings only such as alliances, partnerships, and associations (for summaries, see Almog-Bar & Schmid, 2014, p. 10; Boris & Maronick, 2012, p. 401) and with this inclusive approach has converged on the understanding that nonprofit advocacy commonly involves collectives. Large-N surveys report NPOs participating in some type of policy-oriented collaboration at high rates, for instance: 84% of human service nonprofits in Los Angeles (Mosley, 2011) and 87% of a national sample of community development, arts, and human service nonprofits (Salamon et al., 2008).

From a collective action viewpoint, we ask what explains this centrality of collaborations to nonprofit advocacy? We might reasonably conclude that the high percentage results from a set of major factors together: the strategic, technical, and efficiency advantages of advocacy through collectives; the multiplicity of networks and coalitions NPOs can join; the legitimacy that accrues to nonprofits when advocating collaboratively (e.g., Reitan, 1998); as well as field-level norms to participate (Mosley, 2014) and to be good nonprofit citizens (Abramson & McCarthy, 2012). If, additionally, CAP theory applies, then another valuable insight comes into focus: the arrangements that NPOs participate in represent, in varying degree, solutions to problems of collective action. For instance, smaller collectives are able to overcome CAPs using reputation mechanisms (Olson, 1965). Coalitions of particular constituencies may pursue special policies for just membership, that is, club goods (Esparza et al., 2014; McNutt, 1999), in contrast to broader coalitions that are more likely to pursue public goods (Godwin et al., 2013, pp. 204–205) and therefore face greater free riding. And long-standing collectives that enjoy institutional support help solve CAPs by mobilizing preexisting resources (McCarthy & Zald, 1977). However, claims along these lines rest on less-than-solid ground so long as CAPs remain to be demonstrated among populations of NPOs.

As a first step to exposing CAPs, one needs to account for joiners in policy collaboration as the share of total eligible participants—what Tschirhart (2010) refers to as membership density. Mapping the nonprofit association landscape, however, is fraught with methodological difficulties (Tschirhart, 2006) such as how to construct valid sampling frames of comparable advocacy associations (Andrews & Edwards, 2004) and how to quantify their potential versus actual members. In response to these challenges, this study turns to a particular type of collective, the nonprofit trade association, and asks what collective-level factors are associated with higher rates of NPO participation in such associations?

Nonprofit Trade Associations

Across the third sector, formal associations promote nonprofit industries—such as museums, day cares, and charter schools. For example, state associations of charter schools advocate for higher per-pupil funding, delivering a resource for the charter subsector that individual schools could not leverage acting alone. Scholars refer to these kinds of cooperatives as trade associations, subsector associations, and trade groups for 501c3s (Abramson & McCarthy, 2012; Berry & Arons, 2003; Walker, 1991), and offer complementary explanations for why they form on the state level: to manage the encroachment of competing industries, to cooperate with government priorities linked to public contracting, and to influence relevant regulatory environments (Esparza et al., 2014; Smith & Lipsky, 1993; Walker, 1991). These collectives are designed specifically for organizational membership (Young, 2010) and perform two basic functions: advocacy on field-specific issues and services for members—such as joint contracting with suppliers, professional opportunities, and capacity building (e.g., Abramson & McCarthy, 2012; Balassiano & Chandler, 2010). Tschirhart (2010) suggests accountability programs that signal members’ quality could become an additional function for such associations.

As champions of collective goods, often for substantial populations of NPOs, trade associations encounter CAPs that are rarely solved in full (Abramson & McCarthy, 2012). And as they mobilize for entire fields of nonprofits (Esparza et al., 2014), their advocacy occurs on the industry level, which broadens the scope of collective goods sought. Such associations should therefore routinely face CAPs harder to mitigate than those faced by joint actions of nonprofits advocating for their constituencies narrowly. To add to the challenge, social rewards that may motivate individuals into collective action are unreliable incentives to mobilize organizations (Tschirhart, 2006, p. 531). Associations based on organizational membership offset free riding by offering tangible benefits in the form of member services (selective incentives), and also count on participant satisfaction derived from sharing common policy goals (purposive incentives) (see Tschirhart, 2006, p. 528).

The inherent difficulty that trade associations face solving CAPs recommends them as the focus of this research. Trade associations that cover sizable populations of NPOs eligible to join should typically face stronger CAPs than ones that encompass smaller populations within the same industry. We would expect higher levels of involvement among fewer NPOs, where reputation effects compel cooperation and a single actor’s contribution could be calculated to have impact. In large populations, we would likely observe either lower levels of engagement or associations incentivizing participation with exclusive benefits for members (Olson, 1965). For instance, Table 1 reports the smallest population of charter schools (charter actor population, n = 8) participates fully in its association (association membership, n = 8), while the average participation rate across populations is 75%. We can anticipate most associations partially solve CAPs, yet also face challenges incentivizing full membership. Therefore, I expect that larger populations of NPOs will have a negative effect on participation.

Table 1.

Descriptive Statistics. Statewide Charter Associations.

Variable	M	SD	Minimum	Maximum	Expected effect on participation
Participation	0.754	0.226	0.248	1
Association membership	123	163	8	788
Charter population
Charter actor population	183	237	8	927	–
NA charter population	225	302	7	1,253	–
Charter students	103,350	140,413	1,300	604,700	–
Member services	0.412	0.166	0.100	0.900	+
Association resources
Association staff	8	18	1	115	+
Association expenses	1,920,498	3,966,170	60,945	22,800,000	+
Association age	13.000	5.715	1	23	+
Charter law age	20.475	4.641	1	26	+
Independent schools	0.728	0.173	0.400	1	+
Policy context
Policy score 2015	139.425	24.094	41	174	+
Policy score 2017	144.825	21.802	51	176	+
Change to favorable governor	0.150	0.362	0	1	+
Teacher union density	0.690	0.245	0.199	0.974	+
Legislative professionalism	0.842	2.176	−1.829	6.238	+

Note. NA = National Alliance of Public Charter Schools.

Hypothesis 1: As populations of NPOs eligible to join trade associations increase, rates of participation in these associations should decrease, all else equal.

CAP theory maintains that selective incentives have a positive effect on participation and are essential for organizing large numbers when leverage to compel participation is weak. By offering member services and other inducements, associations motivate nonprofits to join (Abramson & McCarthy, 2012; Balassiano & Chandler, 2010). Without incentives, the tendency increases to free ride on the advocacy of motivated subgroups (Olson, 1965). Therefore, I expect that associations devoting more resources to member services will enjoy higher rates of participation.

Hypothesis 2: As member services provided by trade associations increase, rates of participation in the associations should increase, all else equal.

Testing these hypotheses is feasible, because we can observe the complete set of trade associations in nonprofit industries and can quantify the NPOs eligible to affiliate.

The literature suggests other contextual factors, in terms of policy and industry resources, might affect rates of NPO participation. Favorable policy environments could promote nonprofit associations (Walker, 1991) because, as policy areas come to be perceived as legitimate and supported by public funding, state agencies will encourage the group system (Smith & Lipsky, 1993). This reasoning contrasts, however, with the argument that restrictive policy environments could motivate organizations to allocate resources to advocacy (e.g., Nicholson-Crotty, 2011). There is little doubt that changes in political context can stimulate participation—for instance, elections of new governments open policy windows that encourage advocates to push their proposals (see Kingdon, 1995).

Several ways of thinking about nonprofit capabilities point to how resources, in aggregate, could shape industry-level advocacy. Resources available to a sector’s advocacy organizations should, according to resource mobilization theory, help advocates overcome CAPs (McCarthy & Zald, 1977). Therefore, measures of trade association capacity—such as discretionary spending (Mosley, 2010), and the association’s staff size and age (Tschirhart, 2010) —could make a difference for rates of participation. Another approach is to estimate the share of constituent NPOs that face resource scarcity. Nonprofit advocacy studies identify low-resource NPOs as particularly in need of collectives to achieve their advocacy goals (Fyall & McGuire, 2015; Mosley, 2014), and so industries based on financially struggling nonprofits might see robust membership in their trade associations.

Empirical Setting

The charter school sector provides an accessible national set of NPO populations and trade associations to test the above hypotheses. Charter schools are nonsectarian, publicly funded schools that operate free from many regulations that apply to traditional public schools. By this study’s observation year, 2017, 44 states plus D.C. had enacted charter legislation following the passage of the first charter school state law in 1991, and charter populations varied from 3 (Iowa) to 1,253 schools (California). This variation arises mainly from differences in laws, density of metropolitan areas, and state population size. Nearly every charter school is an NPO, because all but two states mandate that charters incorporate as nonprofits on the school level. In 2016, a total of 40 state associations of charter schools existed across 34 states plus D.C., and all were politically active at the statehouse (or, in the case of D.C., the city government). Separate agencies known as charter authorizers oversee accountability and quality of the schools, leaving the state associations to their primary functions of policy advocacy and member services. The associations solicit charters to affiliate during annual membership drives, and the charters that take part in the association pay a membership fee.

This analysis is limited to associations that operated statewide (or in the case of D.C., citywide), that sought to recruit the entire state charter population, and that were functioning in 2016. I surveyed associations for rates of membership and for member services. These measures contributed to an original data set that describes the 40 state-level charter associations. Of these, all but two remained operational through 2017 when data were collected. Associations that ceased to function prior to this period (such as Oregon’s charter association, which became dormant in 2014) were excluded from the sample.

The large majority of charter associations operate as the sole membership-based charter interest group within state. However, four states plus D.C. each supports two charter associations that compete for school participation. All charter associations are registered nonprofits, and most maintain offices near state capitals. In aggregate, charter associations spend approximately US$75 million annually, employ 351 staff, and represent thousands of schools.¹

During the period of data collection, charter laws existed in 10 additional states where associations were absent. In several such states, charters had not yet been established; in two, the charter sector was relatively well developed, but membership associations failed to endure; and in another, the association had transformed into a staff-driven policy firm. Data from these states are included in tests for selection bias using Heckman selection models.²

Figure 1 shows the landscape of charter school associations. Charter laws are absent in the states represented in white, while charter laws are present but associations are absent in the states in black.

Figure 1.

Varying participation in charter school associations.

Across the charter environment, differences in school management are noteworthy. According to the National Alliance of Public Charter Schools, henceforth referred to as the NA, in the 2016–2017 school year, and in the 34 states plus D.C. with functioning associations, almost three out of four charters were independent schools—that is, fully governed by local boards of directors, freestanding from larger administrative structures. The remaining were administered, in varying degrees, by management firms. Most firms were nonprofit, claiming 17% of the sector, while for-profit firms administered 10% (David, 2018, p. 6). These differences in management might have implications for state associations, because independent schools generally operate on fewer resources than charters managed by firms and therefore could rely relatively more on their associations for advocacy services.

Data and Method

Using phone interviews, in-person meetings, and a web-based survey, I collected observations over 12 months beginning in spring 2017. Data collection involved four stages. First, executive directors of charter associations received an email requesting a phone conversation. Those who agreed were introduced to the project and then asked to fill out a survey or spend additional minutes by phone answering questions. Second, nonresponses received follow-up emails and phone calls, increasing the response rate. Third, those who remained unresponsive received requests for in-person meetings, and in many cases, requests were sent to association staff such as communication or membership directors. A number of association staff were then willing to share information by phone or web survey, increasing the response rate to 60% of the set of associations. Fourth, in-person meetings were required for the remaining associations. Many of these took place in the course of six charter school conferences, several occurred at the main headquarters of the association, and several occurred at locations near state government offices. These efforts resulted in observations for the complete set of charter membership associations. In most cases, data came directly from executive directors (60%) but also from senior-level association managers (33.5%) or from knowledgeable informants within the state’s charter sector (7.5%).³

Survey questions asked these informants to tell the number of charter school actors eligible to join their association, and the percent (or actual number) engaged by paying dues. The survey also asked respondents to estimate the percent breakdown of their association services into two categories: member services, defined as all services aiding the staff, programming, and operations of member schools; and policy advocacy. For instance, one association executive estimated her association as 20% purposed for member services and 80% for advocacy, while another estimated 50/50 as the ratio for his association.

The dependent variable examined in this study, participation, is the rate of charter school membership in associations. From this measure, we learn the average rate of participation is 0.754 (standard deviation [SD] = 0.223). For linear regressions, I perform a logistic transformation on this variable because participation is a proportion from 0 to 1, and its values are nonlinear when approaching the extreme of 1.

Two independent variables test the hypotheses. Measures for charter population are entered as the first independent variable and test Hypothesis 1. Because the number of charters in some states is highly skewed above average, for all population variables I employ the natural logarithm. To operationalize charter population in the main models, I use charter actor population—the number of charter school actors eligible to join the association, as reported by the association staff. This measure captures the size of school populations that can participate. But in case reports are unreliable, I rerun the full model operationalizing the population measure in additional ways: as the count of charters—NA charter population—published in the 2017 report of the NA,⁴ and as the number of charter school students by state—charter students—to proxy for the number of charter schools. These three variables are highly correlated. As shown in Appendix Table A1, the primacy of charter population over other factors remains, regardless of how the variable is operationalized.

As the second independent variable, I include selective incentives measured as member services. Missing data were a minor issue in the construction of this variable, and for two observations, the value was estimated. From the association staff, we learn that charter associations on average dedicate 41% of their resources to supporting membership, yet their center of gravity—on average 59% of their resources—tilts toward advocating for the charter movement.

The central concern is what collective-level factors are associated with higher rates of participation in trade associations. There could be additional contextual reasons why participation rates might rise or fall, and accordingly I control for other possibilities.

Resources available to associations could predict participation, and I take approaches similar to Mosley (2010) and Tschirhart (2010) to operationalize this factor: association expenses is a measure of total expenses reported on the association’s IRS Form 990 when available for 2016 (otherwise for 2015); association staff captures the number of employees listed on the association’s website; and association age represents the years the association has organized in the charter sector, calculated as the observation year minus the founding year. As well, I include charter law age, measured as the years since the passage of the state’s charter law, because interviews revealed that some associations grew out of previous attempts to organize charters and therefore could have learned from earlier efforts. Median values of these four variables depict the typical association: 14.5 years old, in a state that passed its first charter law in about 1996, employing 3.5 staff, and spending US$716,226 annually. Because association staff and expenses are both highly skewed measures, I employ the natural logarithm to transform these variables for linear models. I anticipate that each of these four variables will have a positive effect on participation.

The possibility that resource scarcity drives charters into trade associations is entered as the percent of independent schools comprising state charter populations (independent schools). At one extreme, all charters are independent schools, and 40% are independent at the other extreme. I anticipate that charter populations of mostly independent schools have higher rates of participation in their trade associations.

Because of the relevance of the policy context, I include policy scores from the NA’s 2015 and 2017 reports (policy score 2015 and policy score 2017). Every year, the NA scores state charter laws according to what national advocates consider to be the ideal charter legislation. A variable—change to favorable governor—is constructed to account for states that experienced a recent (2014 or 2016) switch from Democrat to Republican governor—a direction generally advantageous to charters during this period. A shift to a Republican governor could signal to advocates that a policy window has opened to press for preferred charter legislation; 15% of the observations saw such a change.

From interviews, I identified other contextual variables. Informants from several states described frequent direct contact between charter schools and officeholders, rendering assistance from the intermediary association less necessary. For instance, an association executive referred to the low cost of advocacy at the statehouse given high accessibility to legislators (Statement 1 in Appendix Table A3). Therefore, I explore models that include legislative professionalism as a control, using the scale developed by Bowen and Greene (2014), and anticipate that low scores drive down participation in the association.

Several informants mentioned conflict with teacher unions (e.g., Statement 2 in Appendix Table A3). Interest group scholarship expects alliances to form among organizations encountering disciplined opposition (e.g., Hojnacki, 1997). Therefore, I include the National Center for Education Statistics’ recent measure for teacher union density by state, from the 2017 to 2018 National Teacher and Principal Survey.

These summary statistics are reported in Table 1. In the following section, Figures 2 to 4 explore how participation links to charter population and selective incentives. The hypotheses are then tested using ordinary least squares (OLS) regressions (Table 2). In the models, I cluster standard errors on state because of state differences regarding whether charters may join one or two associations. In robustness checks, I control for states with two associations and replicate findings using subsets of observations (Appendix Table A2).

Table 2.

Effects of Collective Action Factors, Association Resources, Management, and Policy Context on Participation.

DV: Participation
IVs	Staff	Expenses	Management	Policy 2015	Policy 2017	Favorable governor	Full model
Charter actor population	−1.412***	−1.333***	−1.087***	−1.117***	−1.118***	−1.066***	−1.420***
Charter actor population	(0.227)	(0.230)	(0.294)	(0.260)	(0.254)	(0.272)	(0.233)
Member services	0.104	−0.855
Member services	(1.188)	(1.500)
Association staff	1.055***						0.954***
Association staff	(0.157)						(0.150)
Association expenses		0.707***
Association expenses		(0.136)
Independent schools			−1.413
Independent schools			(1.322)
Policy score 2015				0.017**
Policy score 2015				(0.008)
Policy score 2017					0.021**	0.031***	0.024***
Policy score 2017					(0.009)	(0.010)	(0.009)
Change to favorable governor						1.448**	1.091**
Change to favorable governor						(0.557)	(0.535)
Constant	6.682***	6.884***	7.771***	4.428***	3.767***	1.906	3.341**
Constant	(1.427)	(1.602)	(2.188)	(1.031)	(1.176)	(1.386)	(1.515)
Observations	40	34	40	40	40	40	40
Adjusted R²	.562	.530	.324	.371	.382	.426	.615

Note. Standard errors in parentheses. DV = dependent variable; IVs = independent variables.

**p < .05. ***p < .01.

Figure 2.

Charter school state associations: Participation and population.

Figure 3.

Charter school state associations: Participation–population relationship.

Figure 4.

Charter school state associations: Participation and selective incentives.

Results

The bivariate plots represent charter actor populations at different rates of participation in state associations, first Figure 2 using raw data and then Figure 3 using variables transformed for OLS.

From the below plot, several things are notable. The line, created with a LOWESS function, describes a negative trend in participation as the number of charters in each state increases, and associations that enjoy over 90% participation cluster within relatively small charter populations. All associations with 92% participation or higher inhabit populations of 85 charter actors or lower—settings that appear exempt from poor participation. The 24 associations rooted in charter school environments of fewer than 125 schools all enjoy participation rates over 60% and in most of these cases (19 associations) in the 80% to 100% range. However, above a threshold of 125 schools, membership rates fall across a wider spectrum of charter populations, with participation beneath 50% in a number of cases. As well, the picture reveals several outlying associations that engage charters at high participation rates even in contexts of many hundreds of charter schools.

With participation rescaled by logistic transformation, and with population logged, these same data are presented again in Figure 3.

The relationship is decreasing and supports Hypothesis 1. That is, the more charter school actors within a state, the fewer are involved in the association.

In contrast to the correspondence between participation and population, Figure 4 represents the relationship between participation and member services.

Here, participation fails to hinge on services offered to membership. Schools are not joining associations because of selective incentives, at least not in the way this study operationalizes the factor. OLS models below confirm the null relationship.

Table 2 reports main regression models. For the larger series from which Table 2 is drawn, see the Online Appendix Tables A4 to A6.

With regard to the effect of population, Hypothesis 1 is fully supported. The count of charter actors eligible to join an association has a negative and statistically significant effect on participation. The coefficient of −1.420 (full model) indicates that a typical association working in the average population of 183 charter actors could expect to have a participation rate of 82%, and if the population were to jump 1 SD above the mean to 420 schools, then participation would fall 24 percentage points to 58%, ceteris paribus.

With regard to the effect of selective incentives, member services do not influence participation, confirming the flat line in Figure 4. The staff and expenses models report this variable’s insignificance when conditioned on association resources. One explanation could be the absence of controls for other sources from which charters receive support services. For instance, one association executive pointed to a national charter resource center, located within state, to which local charters could turn for services (Statement 3 in Appendix Table A3; also see Statement 4).

Association resource variables (staff and expenses models) affect participation. The association staff coefficient of 0.954 in the full model indicates that an association with a relatively large staff of 26 employees (1 SD above the mean) benefits from a participation rate of 94%—approximately 12 percentage points higher than an association with the average number employees, ceteris paribus. This leads one to consider how developed workforces help associations that face severe CAPs. For instance, an association in the largest charter school population (927 charter actors) employs 115 staff and enjoys a solid level (85%) of participation. However, association age and charter law age (Online Appendix Table A4) show no sign of contributing to participation. And the percent of independent charters does not influence participation.

Variables linked to policy context attest to greater participation. Policy scores from several years before the study predict participation (policy 2015 model), as do scores of the study’s observation year (policy 2017 model). The policy score 2017 coefficient of 0.024 in the full model indicates that an association in a state with laws relatively favorable to charter schools (scores 1 SD above the mean) benefits from a participation rate approximately 7 percentage points higher than an association in a state with an average NA score, ceteris paribus. The finding comports with studies such as Walker (1991) and Smith and Lipsky (1993) that expect well-functioning associations in legitimized policy areas. The variable indicating change to a favorable governor also predicts participation—both in a bivariate regression (β = 1.437, p < .01, not displayed in the table) and as a control in the favorable governor model. The variable’s coefficient of 1.091 in the full model indicates that an association in a state whose governor’s office has recently switched from Democratic to Republican benefits from a participation rate of 92%—approximately 13 percentage points higher than an association in a state that did not experience such a shift.

In a series of robustness checks (Appendix Table A2), the pattern of results remains consistent. The population coefficient changes little and remains highly significant even when the model relies on just the 30 states with single associations, or on the 35 major associations excluding the smaller, competing associations. And controlling for whether an association operates solo in the state or in competition with another association has no effect on the models. The statistically significant variables (charter actor population, association staff, policy score, and change to favorable governor) are tested again, conditional on additional controls. However, inclusion of teacher union density and legislative professionalism are insignificant and barely influential on the meaningful variables (see Online Appendix Table A6).

Discussion

This study uses novel data on charter school membership associations to engage central aspects of the collective action framework and finds that CAPs, as articulated in pristine terms by Olson (1965) to explain individual behavior, shape charter participation in their trade associations, net of other relevant factors. Charters tend to participate in their associations at higher rates in states where only dozens, rather than hundreds, of charters operate. This effect holds regardless of policy environment, charter law year of enactment, association age, association staffing level, teacher union density, or professionalism of the state legislature. The insignificance and even low construct validity of the selective incentives measure does not contradict this story: Sources other than charter associations offer services to schools, and so associations lack leverage to incentivize with membership services—although they endeavor to do so. Some might think that charitable nonprofits generally would be above these kinds of difficulties. After all, nonprofit missions typically envision enlarging the common good (Bass et al., 2007, p. 39). But when faced with CAPs, charter schools are not exempt from the free riding observed among for-profits seeking public policy (Lowery et al., 2004) nor among individuals generally (Olson, 1965).

This analysis also suggests that policy context and association resources matter for school participation. It appears that state governments whose policies receive the NA’s imprimatur correspond to relatively well-mobilized charter schools, and that greater association resources mobilize schools. The latter point is particularly important to understanding CAPs because collectives with greater access to resources should be better equipped to sustain collective action on an industry level (McCarthy & Zald, 1977).

Interview data, used to identify alternative explanations for this study’s statistical models, speak in finer detail to why schools opt out of participation. Association managers report that charters avoid participation mainly because of the following: policy disagreements with associations, CAPs, and scarcity of resources in some settings but resource sufficiency in others. A separate article under review elaborates these findings. The regression analyses here complement the picture: Charters respond to policy environments, CAPs undermine participation, and patterns of school management fail to tell a straightforward story about charter population resources shaping participation.

The limitations of this investigation should be acknowledged. First and foremost, the data come entirely from the charter sector, so we must ask whether these findings apply to other nonprofit industries. There could be fields of nonprofits where motivations to collaborate are so strong that CAPs lose their relevance. Adding to the caution, Hager (2014) found differences between engineering and health association professionals on motivation to join.

The study relies on one period of observation, which limits the certainty of claims about mechanisms underlying factors of participation. It makes sense to ask whether elements correlated with charter populations also shape collaborative behavior, determining both jointly. For instance, perhaps the policy struggle has subsided in sizable charter populations and advocates have declared victory, so there is less need to band together into interest groups. And if this were true, participation could flag in states that had better charter policies. However, the measure of difference between NA policy scores from 2015 to 2017 does not predict participation. One might also wonder whether a low density of charter students motivates charters to work together to petition government for enlargement of their sector. But the percent of public school students that attend charters does not significantly predict participation and, when entered as a control, has little effect on coefficients for charter population. A similar concern could lead us to ask whether high engagement somehow relates to low numbers of eligible actors. For instance, if foundations help associations when charter populations are small, perhaps grants lower the reliance on dues thereby increasing participation. The majority of associations reported membership fees as a line of revenue on their IRS Form 990, enabling a sub-analysis,⁵ but there is no statistical relationship between dues and participation.

While this study demonstrates how lower numbers of charters correspond with higher levels of engagement, the observations include unusual cases, for instance, a population of 173 charters—nearly the average—with a participation rate of only 31%, almost 2 SDs below average. Also, we observe associations that maintain membership quite well even when severe free riding is to be expected. These examples call for identification of other plausible forces affecting participation. For example, an informant pointed out that association effectiveness could matter (Statement 5 in Appendix Table A3). While association staff, expenses, and age variables may proxy for effectiveness, much more could be learned. Leadership factors exert influence on member participation (Mason, 2016; Wilson, 1995), and one informant reported efforts to recruit outstanding executives to fill positions in prominent associations (Statement 6 in Appendix Table A3).

Conclusion

We may envision the executive of a charter association who prefers a context of hundreds, rather than dozens, of schools eligible for membership. The findings of this study join with Abramson and McCarthy (2012) in alerting them to expect worse CAPs in larger fields of schools. Free riding alone does not indicate a particularly cynical or uncooperative group of NPOs. Neither does it necessarily mean the association has failed to convey its policy work to stakeholders. In fact, a component of low participation is the understanding that the association provides collective goods.

While one could reassure leaders of nonprofit trade associations that a limited amount of free riding might not harm their ability to advocate effectively, the consequences can become damaging. This study finds charter associations petition state government for policy with backing generally from three quarters of their field. But at a certain point, extensive nonparticipation undermines efforts to represent the industry, exemplified vividly in several cases of association collapse during this investigation’s observation year.

Organizationally speaking, there are ways to counteract CAPs. While managers of nonprofit trade associations seldom have recourse to penalties, they will often discover chances to recognize volunteerism, to update knowledge of valued services, and to articulate advocacy in terms that resonate with membership. While rewarding and inspiring affiliation may seem sensible, it is worth stressing that staff who cultivate such skills are essential to a trade association’s survival and growth. This notion is reinforced by this study’s finding of strong correspondence between association staffing levels and participation.

By offering an empirical demonstration of the collective action challenge based on the charter sector, this study has addressed an indeterminacy in the literature. Nonprofit scholars have seemed either to assume CAP theory, perhaps given its application in adjacent disciplines, or to overlook CAPs for want of evidence. Observing CAPs among charter schools represents a step toward clarification. To sharpen this study’s contribution, future studies could test the approach in settings with similar scope conditions—specifically, varying population sizes of NPOs in pursuit of collective goods. As well, rich possibilities lie ahead in studying nonprofit trade associations using nonprofit-level factors in multilevel analysis to learn more about barriers and incentives to participation.

Supplemental Material

sj-pdf-1-nvs-10.1177_08997640211017662 – Supplemental material for How Collective Action Problems Suppress Participation in Nonprofit Trade Associations

Supplemental material, sj-pdf-1-nvs-10.1177_08997640211017662 for How Collective Action Problems Suppress Participation in Nonprofit Trade Associations by Clifford W. Frasier in Nonprofit and Voluntary Sector Quarterly

Footnotes

Appendix

Table A3.

Statements From Open-Ended Interviews.

1.	Some schools have their own advocacy, because advocacy is cheap here in [state S]. (Interview with association executive in state S)
2.	Yesterday, at our advocacy day, the unions showed up as a counterforce to us. Our teacher union has recently beefed up its outreach to our charter teachers. (Interview with association executive in state G)
3.	There is just one membership association for schools in [state O], but the National Resource Center is here in [state O], and they too offer operations. (Interview with association executive in state O)
4.	Other associations handle pieces of a charter’s business: the School Boards Association; the Principals and Supervisors Association. The School Boards Association does governance training for board members, so the [charter school] association does not do a lot of that. (Interview with school director in state R)
5.	Your study should measure an association’s effectiveness, because its effectiveness could attract membership. (Interview with national informant 04)
6.	In some of the big population states, the association leaders are expected to be high quality, and are high quality. There is national recruitment for them to apply. (Interview with state informant in state V)

Note. The statements are based on handwritten notes transcribed immediately following open-ended interviews.

Acknowledgements

The author thanks Anthony Bertelli, Patrick Egan, Steven Rathgeb Smith, and the anonymous reviewers for their valuable comments on drafts of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the American Society of Association Executives Foundation in cooperation with New York University.

ORCID iD

Clifford W. Frasier

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biography

Clifford W. Frasier recently received his PhD in Public Administration from New York University, where he teaches courses in public policy and American politics. His research considers third sector management and political activity, and focuses on how nonprofits collaborate in policy advocacy at the state and local levels.

References

Abramson

A. J.

McCarthy

(2012). Infrastructure organizations. In Salamon

L. M.

(Ed.), The state of nonprofit America (pp. 423–458). Brookings Institution Press.

Almog-Bar

Schmid

(2014). Advocacy activities of nonprofit human service organizations: A critical review. Nonprofit and Voluntary Sector Quarterly, 43(1), 11–35.

Andrews

K. T.

Edwards

(2004). Advocacy organizations in the US political process. Annual Review of Sociology, 30, 479–506.

Balassiano

Chandler

S. M.

(2010). The emerging role of nonprofit associations in advocacy and public policy: Trends, issues, and prospects. Nonprofit and Voluntary Sector Quarterly, 39(5), 946–955.

Bass

G. D.

David

Arons

K. G.

Matthew

Carter

S. R.

(2007). Seen but not heard: Strengthening nonprofit advocacy. Aspen Institute.

Berry

J. M.

Arons

D. F.

(2003). A voice for nonprofits. Brookings Institution Press.

Berry

J. M.

Portney

(2014). The group basis of city politics. In Pekkanen

Smith

S. R.

Tsujinaka

(Eds.), Nonprofits and advocacy: Engaging community and government in an era of retrenchment (pp. 21–46). Johns Hopkins University Press.

Boris

E. T.

Maronick

(2012). Civic participation and advocacy. In Salamon

L. M.

(Ed.), The state of nonprofit America (pp. 394–422). Brookings Institution Press.

Boris

E. T.

Maronick

Nikolova

(2014). Shaping the government-nonprofit partnership: Direct and indirect advocacy. In Pekkanen

R. J.

Smith

S. R.

Tsujinaka

(Eds.), Nonprofits and advocacy: Engaging community and government in an era of retrenchment (pp. 66–84). Johns Hopkins University Press.

10.

Bowen

D. C.

Greene

(2014). Should we measure professionalism with an index? A note on theory and practice in state legislative professionalism research. State Politics & Policy Quarterly, 14(3), 277–296.

11.

Chen

K. K.

(2018). Interorganizational advocacy among nonprofit organizations in strategic action fields: Exogenous shocks and local responses. Nonprofit and Voluntary Sector Quarterly, 47(4_suppl), 97S–118S.

12.

David

(2018). National charter school management overview: 2016-2017 school year. National Alliance for Public Charter Schools. https://www.publiccharters.org/our-work/publications/national-charter-school-management-overview-2016-17

13.

Donaldson

L. P.

(2007). Advocacy by nonprofit human service agencies. Journal of Community Practice, 15(3), 139–158.

14.

Esparza

Walker

E. T.

Rossman

(2014). Trade associations and the legitimation of entrepreneurial movements: Collective action in the emerging gourmet food truck industry. Nonprofit and Voluntary Sector Quarterly, 43(2_suppl), 143S–162S.

15.

Fyall

McGuire

(2015). Advocating for policy change in nonprofit coalitions. Nonprofit and Voluntary Sector Quarterly, 44(6), 1274–1291.

16.

Godwin

R. K.

Ainsworth

S. H.

Godwin

(2013). Lobbying and policymaking. CQ Press.

17.

Guo

Saxton

G. D.

(2010). Voice-in, voice-out: Constituent participation and nonprofit advocacy. Nonprofit Policy Forum, 1(1), Article 5.

18.

Hager

M. A.

(2014). Engagement motivations in professional associations. Nonprofit and Voluntary Sector Quarterly, 43(2_suppl), 39S–60S.

19.

Hojnacki

(1997). Interest groups’ decisions to join alliances or work alone. American Journal of Political Science, 41(1), 61–87.

20.

Kingdon

J. W.

(1995). Agendas, alternatives, and public policies (2nd ed.). Longman.

21.

Leroux

Goerdel

H. T.

(2009). Political advocacy by nonprofit organizations: A strategic management explanation. Public Performance & Management Review, 32(4), 514–536.

22.

C. W.

Tang

(2017). Nonprofit policy advocacy under authoritarianism. Public Administration Review, 77(1), 103–117.

23.

Lowery

Gray

Anderson

Newmark

A. J.

(2004). Collective action and the mobilization of institutions. The Journal of Politics, 66(3), 684–705.

24.

MacIndoe

Beaton

(2019). Friends or foes? How managerial perceptions of the political opportunity structure shape nonprofit advocacy. Public Performance & Management Review, 42(1), 59–89.

25.

Mason

D. P.

(2016). Common agency in nonprofit advocacy organizations. Nonprofit Management and Leadership, 27(1), 11–26.

26.

McCarthy

J. D.

Zald

M. N.

(1977). Resource mobilization and social movements: A partial theory. American Journal of Sociology, 82(6), 1212–1241.

27.

McNutt

(1999). Public goods and club goods. Encyclopedia of Law and Economics, 1, 927–951.

28.

Mosley

J. E.

(2010). Organizational resources and environmental incentives: Understanding the policy advocacy involvement of human service nonprofits. Social Service Review, 84(1), 57–76.

29.

Mosley

J. E.

(2011). Institutionalization, privatization, and political opportunity: What tactical choices reveal about the policy advocacy of human service nonprofits. Nonprofit and Voluntary Sector Quarterly, 40(3), 435–457.

30.

Mosley

J. E.

(2012). Keeping the lights on: How government funding concerns drive the advocacy agendas of nonprofit homeless service providers. Journal of Public Administration Research and Theory, 22(4), 841–866.

31.

Mosley

J. E.

(2013). Recognizing new opportunities: Reconceptualizing policy advocacy in everyday organizational practice. Social Work, 58(3), 231–239.

32.

Mosley

J. E.

(2014). From skid row to the statehouse: How nonprofit homeless service providers overcome barriers to policy advocacy involvement. In Pekkanen

R. J.

Smith

S. R.

Tsujinaka

(Eds.), Nonprofits and advocacy: Engaging community and government in an era of retrenchment (pp. 107–134). Johns Hopkins University Press.

33.

Nicholson-Crotty

(2011). Nonprofit organizations, bureaucratic agencies, and policy: Exploring the determinants of administrative advocacy. The American Review of Public Administration, 41(1), 61–74.

34.

Olson

(1965). The logic of collective action: Public goods and the theory of groups. Harvard University Press.

35.

Reitan

T. C.

(1998). Theories of interorganizational relations in the human services. Social Service Review, 72(3), 285–309.

36.

Salamon

L. M.

Geller

S. L.

Lorentz

S. C.

(2008). Nonprofit America: A force for democracy? Johns Hopkins University, Center for Civil Society Studies.

37.

Sandfort

(2014). Analyzing the practice of nonprofit advocacy: Comparing two human services networks. In Pekkanen

R. J.

Smith

S. R.

Tsujinaka

(Eds.), Nonprofits and advocacy: Engaging community and government in an era of retrenchment (pp. 222–253). Johns Hopkins University Press.

38.

Schmid

Bar

Nirel

(2008). Advocacy activities in nonprofit human service organizations: Implications for policy. Nonprofit and Voluntary Sector Quarterly, 37(4), 581–602.

39.

Smith

S. R.

Lipsky

(1993). Nonprofits for hire. Harvard University Press.

40.

Tschirhart

(2006). Nonprofit membership associations. In Steinberg

W. P.

Powel

(Eds.), The nonprofit sector: A research handbook (pp. 523–541). Yale University Press.

41.

Tschirhart

(2010). Self-regulation at the state level: Nonprofit membership associations and club emergence. In Gugerty

M. K.

Prakash

(Eds.), Voluntary regulation of NGOs and nonprofits: An accountability club framework (pp. 85–98). Cambridge University Press.

42.

Walker

J. L.

(1991). Mobilizing interest groups in America: Patrons, professions, and social movements. University of Michigan Press.

43.

Wilson

J. Q.

(1995). Political organizations. Princeton University Press.

44.

Young

D. R.

(2010). Nonprofit infrastructure associations as reluctant clubs. In Gugerty

M. K.

Prakash

(Eds.), Voluntary regulation of NGOs and nonprofits: An accountability club framework (pp. 85–98). Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.12 MB