Government Regulation and the Political Activities of Nonprofits

Abstract

We propose a conceptual model of the political activities of nonprofits that qualify for exemption under subsections of the Internal Revenue Code other than 501(c)(3), including social welfare organizations, civic leagues, social clubs, and so on, which considers three categories of explanatory factors: organizational capacity, financial strategy, and operating environment. Using a Heckman selection model with longitudinal IRS 990 data, we find government regulation to be an obstacle for nonprofits to engage in the policy process. Political activities of non-501(c)(3) organizations are also negatively associated with government support, suggesting these organizations perceive government intervention differently from 501(c)(3) organizations when engaging in political activities.

Keywords

government regulation nonprofit political activities non-501(c)(3) organizations Heckman selection model

Introduction

Nonprofits have a complicated, multidimensional relationship with the government. Government regulatory policies directly affect the operating landscape of nonprofits, from granting individuals tax deductions for donations to encourage philanthropy, from awarding contracts and grants to nonprofits to promote their goods and services, and from restricting the amount certain nonprofits are allowed to spend on political and lobbying activities to avoid risking loss of their beneficial tax exemptions.

Influencing government and public policy is an important role of nonprofits (Andrews & Edwards, 2004; Guo & Musso, 2007; Mosley, 2011, 2012). Government regulation plays an important role in influencing their scope of political activities, as many nonprofits become involved in the political process to represent their clients, donors, and members. However, our understanding of the government-nonprofit relationship is still under development even as the nonprofit sector continues to expand. There are competing views regarding the role of nonprofits: (a) they should remain neutral and focus on service provision and (b) they should actively represent their constituents and members.

We seek to better understand how government regulation influences the political activities of nonprofits. We use IRS 990 data to estimate spending on political activities by nonprofits that qualify for exemption under subsections of the Internal Revenue Code other than 501(c)(3)—which we term non-501(c)(3) nonprofits and includes social welfare organizations, civic leagues, social clubs, labor organizations, business leagues, and so on—as a function of government regulatory policy that determines whether their charitable contributions are tax deductible. Nonprofit political activities influence their abilities to represent public voice, utilize social capital, and enhance civic participation. Thus, better understanding the relationship between nonprofits’ political activities and government regulation is important for understanding democratic civil society (Boris & Mosher-Williams, 1998).

Extant literature on the government-nonprofit relationship has focused on resource dependence (Fyall, 2016; Lee et al., 2018; Lu, 2018a) or their financial connection (Cheng, 2019; Neumayr et al., 2015). While political activities of nonprofits have gained scholars’ attention in past years, most studies have concentrated on specific types of social service agencies (i.e., faith-based and secular; Fyall & Allard, 2017), “organizations,” (e.g., Andrews & Edwards, 2004; J. M. Berry & Arons, 2003; Boris & Mosher-Williams, 1998 ; Child & Grønbjerg, 2007 ; Guo & Saxton, 2010; Guo & Zhang, 2014), or 501(c)(3) charitable organizations (e.g., Clemens & Guthrie, 2010; Garrow & Hasenfeld, 2014; Saidel, 2002; Salamon, 2012; Schmid et al., 2008; Smith & Grønbjerg, 2006 ; Suárez & Hwang, 2008). However, there are important distinctions between political/lobbying activities of advocacy organizations and of nonprofits without a core mission of advocacy. These non-501(c)(3) organizations primarily engage in political and lobbying activities to promote the interests of their members and/or constituents but do not identify specifically as advocacy organizations.

The following sections discuss government regulation of nonprofit political activities, particularly how restrictions imposed upon 501(c)(3) charitable organizations differ from those for non-501(c)(3) nonprofits. We then present our conceptual model of the determinants of non-501(c)(3) nonprofits’ political activities. Next, we discuss our data and two-stage model specification before presenting our empirical results. Finally, we discuss the implications of our findings and contributions of this research.

Government Regulation of Nonprofit Political Activities

In the United States, government regulations are imposed on nonprofits in exchange for the benefits of tax exemptions. Nonprofits must register with the Internal Revenue Service (IRS) and request tax-exempt status as described in Title 26 of the Internal Revenue Code (IRC) under Section 501(c). Organizations choose the 501(c) organizational designation that best matches their mission, which establishes the tax exemptions they receive and which government regulations they encounter. Of these designations, 501(c)(3) and 501(c)(4) are most common, but there are many other subsections the IRS uses to regulate the remaining organizations.

Government regulations limit the amount nonprofits may spend on lobbying, discouraging 501(c)(3) charities from such activities as they are the most restricted because their charitable contributions are tax deductible for donors. The IRS defines lobbying as, “expenditures (including allocable overhead and administrative costs) paid or incurred for the purpose of attempting to influence legislation.”¹ Direct lobbying involves communications “with any member or employee of a legislative or similar body, or with any government official or employee who may participate in the formulation of the legislation,” and grassroots lobbying involves communications intended to “affect the opinions of the general public.”

Despite restrictions on lobbying, nonprofits are often engaged in other forms of political activities not regulated by the government. For example, providing public education about a social issue that does not target specific legislation is an advocacy activity not restricted by government. Nonprofits can also organize group actions on policy issues, such as boycotts, public protests, or letter writing campaigns. Finally, they can provide their members and the public with information about policy problems and recommended solutions.

Since Gais and Walker (1991) proposed inside and outside strategies, scholars have categorized various types of nonprofit political activities (e.g., Avner, 2002; Gormley & Cymrot, 2006; Guo & Saxton, 2010; Mosley, 2011). Using different terminology, these studies distinguished outside types (e.g., media overtures, protest, and public education) from inside types (e.g., legislative lobbying and legislative testimony). Following this literature, nonprofits have several ways of engaging in political activities even when they are restricted on direct lobbying.

While scholars have paid more attention to the hybridity of 501(c)(3) nonprofits, we know little about nonprofits that are not 501(c)(3)s. Considering stark differences in the characteristics of 501(c)(3) nonprofits compared to other 501(c) organizations, it is important to understand whether and how these other nonprofits participate in political activities. Compared to the strict regulations imposed upon 501(c)(3) organizations, there is more variation in government regulation of non-501(c)(3) political activities and deductibility of the donations they receive.

Distinguishing Non-501(c)(3) Nonprofits

Non-501(c)(3) nonprofits are less restricted by government from engaging in political activities, as they are not subject to lobbying limits or prohibited from campaigning, but their donations may not be tax deductible. Since strict prohibition on campaigning is not imposed on non-501(c)(3) organizations, we expect them to be more politically active than 501(c)(3) organizations. Table 1 highlights the key categorical differences between 501(c)(3) charities and non-501(c)(3) nonprofits.

Table 1.

Differences in Nonprofit Types.

Organizational characteristics	501(c)(3)	Non-501(c)(3)
Economic role	Provider of public goods and services; agent of government in provision of public goods	Contributor to political discourse; provider of private goods to members
Benefit structure	Benefits accrue to users, donors, and society	Benefits accrue to members and society
Government regulation of lobbying	Spending limited by IRS rules; campaigning prohibited	No limitations; campaigning permitted; not required to disclose donors
Government regulation of tax deductibility of donations	Tax deductibility of donations	Donations or fees/dues possibly tax deductible
Primary revenue sources	Higher reliance on donations and grants	Higher reliance on fees
Previous research on advocacy	Focused generally on advocacy as a primary function	Largely descriptive for finances; fee setting; member attraction and retention

Note. IRS = Internal Revenue Service.

As shown in Table 1, non-501(c)(3) nonprofits provide direct benefits to members in exchange for payment of fees/dues. Non-501(c)(3) nonprofits also produce goods and services with excludable benefits, so indirect benefits for the general public are minimal. In contrast, 501(c)(3) charities (often financed through tax-deductible contributions) provide services that indirectly benefit donors but directly benefit nondonors and/or the general public regardless of membership or payment. Commensurately, non-501(c)(3) nonprofits often have a primary revenue source of fees/dues from members and are generally less reliant on donor contributions or government support.

Non-501(c)(3) nonprofits have no statutory limits imposed upon lobbying; therefore, they can spend more on these activities without fear of losing their tax-exempt status. Conversely, 501(c)(3) charities are required to keep lobbying expenditures as an “insubstantial” portion of total spending or adopt the h-election (Grasse et al., 2021). Some non-501(c)(3) nonprofits, such as 501(c)(4) organizations, accept the benefit of unlimited spending on political activities but in return give up their ability to receive tax-deductible contributions. This greater variation in spending restrictions related to political activities, as well as the extent to which contributions are tax deductible for donors, highlights the appropriateness and importance of studying non-501(c)(3) organizations.

Conceptual Model of Non-501(c)(3) Political Activities

We offer a conceptual model of key factors influencing the political activities of nonprofits and then use this framework to specify our empirical model of the determinants of political activities by non-501(c)(3) organizations. In doing so, we draw upon extant research examining organizational, financial, and environmental factors that compel nonprofits to engage in political activities (Chaves et al., 2004; Guo & Zhang, 2014; Lu, 2018b; Walker, 1991).

Organizational Capacity

According to resource mobilization theory (McCarthy & Zald, 1977), political efforts take organizational resources; thus, better-resourced organizations have more chances to engage in political activities. Empirical studies consistently show a positive relationship between the size of human resources and nonprofit political activities (Child & Grønbjerg, 2007; Suárez & Hwang, 2008), and between financial resources and advocacy (Bass et al., 2007; Eising, 2007; Junk, 2016; Klüver, 2010; Mosley, 2010; Nicholson-Crotty, 2007). Non-501(c)(3) nonprofits with sufficient human resources may have larger staff sizes (Donaldson, 2008) or professional lobbyists (Gibelman & Kraft, 1996) dedicated to political activities without compromising mission-related service delivery. Likewise, organizations with more financial resources can secure and allocate more funds for political activities (Gormley & Cymrot, 2006). As such, we hypothesize that a greater level of organizational capacity will lead to more spending on political activities.

Hypothesis 1a: An organization’s political activities are positively associated with organizational human resources.

Hypothesis 1b: An organization’s political activities are positively associated with organizational financial resources.

Financial Strategies

The ultimate goal of nonprofits is to fulfill their missions (Lewis, 2005), and they employ different strategies to secure financial sustainability to do so. Many organizations earn revenues from government agencies (Salamon et al., 2004). However, resource dependency theory suggests government funding may suppress nonprofit engagement in political activities (Neumayr et al., 2015). While empirical studies provide inconsistent results on the relationship between government funding and nonprofit political activities (e.g., Lu, 2018a; Neumayr et al., 2015), it is apparent that nonprofits’ autonomy from government enables them to accomplish their missions and advocate for their members or constituents. As non-501(c)(3) nonprofits are less restricted from political activities than 501(c)(3) charities, they may actively engage in advocacy activities when they can independently secure funding to build advocacy capacity. We consider two different financial strategies: market-based and fundraising.

Some organizations employ a market-based strategy in which the majority of income comes from service or program provision. For instance, California Medical Association (CMA), which is a 501(c)(6), advocates for the medical profession and the patients they serve. Over 60% of CMA’s annual revenue in 2019 came from service programs; it did not receive any government or private funds. These types of organizations may attempt to generate revenues through fees-for-services instead of relying on government grants or donations (Nah & Saxton, 2013). As such, we expect that organizations with a market-based strategy are more likely to engage in political activities:

Hypothesis 2a: An organization’s political activities are positively associated with a market-based strategy.

Another way nonprofits secure financial capacity independently is through donations. Nonprofits’ efforts for charitable donations, indicated by fundraising expenses, can be defined as a donor-focused strategy (Graddy & Morgan, 2006; Nah & Saxton, 2013). Some organizations like the Alliance for Youth Action (a 501[c](4)) do not receive government support but rather rely 100% on private contributions. Unlike government funding, private funding usually does not require collaboration with government or other organizations (Jang & Feiock, 2007). As such, a nonprofit with a donor-focused strategy may be more likely to preserve its autonomy. Thus, we hypothesize that nonprofits with more revenues from donors have greater autonomy from government and are therefore more likely to engage in political activities:

Hypothesis 2b: An organization’s political activities are positively associated with a donor-focused strategy.

Operating Environment

Government policies that affect nonprofits include changes in deductibility or tax rates for charitable contributions, policies related to spending on contracts and grants, and laws or regulations regarding nonprofit financial management. Nonprofits’ political activities are regulated by government in exchange for preferential tax treatment, which operates as a regulation of non-501(c)(3)s, because of the government’s power to revoke its tax-exempt status if the organization is too involved in political activities. Even if an organization’s main role is to advocate for their membership in a particular policy area, they must still be concerned with regulatory spending limits to ensure their tax-exempt status is secure.

The regulatory and institutional environments within which non-501(c)(3) nonprofits operate—the political atmosphere—also influences the scope and intensity of their political activities (Chaves et al., 2004; Child & Grønbjerg, 2007). Existing literature shows inconsistent results about the effect of government funding on the political activities of nonprofits. Thus, financial support from the government alone may not reveal the reasons why some organizations engage in political activities while others do not. Political science literature proposes an “opportunities and threats” framework (e.g., MacIndoe & Beaton, 2019) to better understand the effect of the political environment on organizational engagement in political activities. A nonprofit is more likely to engage in political activities in an opportunity situation where “gain is likely, and over which one has a fair amount of control” or in a threat situation where “loss is likely, and over which one has relatively little control” (Dutton & Jackson, 1987, p. 80). Some studies find nonprofits lobby and engage in advocacy more agressively in a more favorable political environment (Guo & Zhang, 2014; Tarrow, 1998), whereas others suggest nonprofits engage in political activities when there is a political threat (Gais & Walker, 1991; Nicholson-Crotty, 2007). We hypothesize both government regulation (i.e., political threats) and support (i.e., political opportunities) will influence non-501(c)(3)s’ political activities:

Hypothesis 3: An organization’s political activities are negatively associated with government regulation.

Hypothesis 4: An organization’s political activities are positively associated with government support.

Another important aspect of the operating environment is external resources. As resource dependency theory assumes that “organizations become dependent of their environments for critical resources” (Guo & Zhang, 2014, p. 1161), it is plausible that nonprofit behaviors are conditioned by the extent to which resources are controlled by internal or external stakeholders, such as organizational members and donors. For those who seek more support and resources to advocate for their members, it is important for the organization to obtain tangible results that manifest in some degree of success (Coates & David, 2002), which may motivate them to engage in various political activities.

When it comes to organizations more dependent on donors, as it is difficult to observe the outcomes of political activities (Hoefer, 2005; Jackson, 2014), it is hard for organizations to convince donors that their advocacy efforts influenced policy outcomes. Donors tend to donate more for activities that provide immediate outcomes, such as human service delivery (Casey, 2011), and may not want to deal with controversial issues caused by policy advocacy (Teles & Schmitt, 2011). Consistent with resource dependency theory, we predict a positive relationship between dependence on membership fees and organizational involvement in political activities and a negative relationship between dependence on donors and organizational political activities:

Hypothesis 5a: An organization’s political activities are positively associated with its dependence on membership fees.

Hypothesis 5b: An organization’s political activities are negatively associated with its dependence on private donations.

Data and Model Estimation

We analyze data on 501(c) organizations for all subsections except those classified as 501(c)(3) incorporated public charities. Table 2 provides descriptions of the subsections included in 501(c) of the Internal Revenue Code. Many of these organizations are membership associations (i.e., subsections 5, 6, and 7) that are not open to the public but rather promote the interests of their members who are engaged in a common occupation, social or recreational activities, or other common interests. Others (i.e., subsections 4, 8, and 10) are actively engaged in the political arena by promoting civil rights and liberties or other advocacy efforts intended to promote social welfare and/or human rights. For only those organizations with reported spending (N = 17,430) on lobbying or legislative activites, Table 2 also provides the frequencies of organizations within each 501(c) subsection. The overwhelming majority of organizations with political, legislative, or lobbying spending are classified by the IRS as 501(c)(82) political organizations, which benefit from tax-deductible contributions. For those organizations that face less government regulation because their contributions are not tax deductible, we see the greatest numbers of organizations with political, legislative, or lobbying spending in the 501(c)(5) labor, agricultural, and horticultural subsector, followed closely by the 501(c)(6) subsector of business leagues, chambers of commerce, and real estate boards.

Table 2.

IRS Subsection Code Descriptions and Frequencies.

501(c) subsection code	Subsection code description	Frequency	Percent
0	Unidentified	182	1.04
2	Title holding corporation for a tax-exempt organization	50	0.29
3	Religious, educational, charitable, scientific, and literary organizations	108	0.62
4	Civic leagues, social welfare organizations, and local associations of employees	1,173	6.73
5	Labor, agricultural, and horticultural organizations	5,489	31.49
6	Business leagues, chambers of commerce, and real estate boards	4,313	24.74
7	Social and recreational clubs which provide pleasure, recreation, and social activities	319	1.83
8	Fraternal beneficiary societies and associations with lodges	256	1.47
9	Voluntary employees’ beneficiary associations including federal employees’ voluntary beneficiaries	177	1.02
10	Domestic fraternal societies and associations including lodges	43	0.25
12	Benevolent life insurance associations including mutual ditch or irrigation companies	193	1.11
13	Cemetery companies providing burial and incidental activities for members	39	0.22
14	State-chartered credit unions and mutual reserve funds offering loans to members	431	2.47
15	Mutual insurance associations providing insurance to members	16	0.09
17	Supplemental unemployment benefit trusts providing payments	3	0.02
19	Post or organization of war veterans	125	0.72
25	Title holding company for pensions	19	0.11
27	State-sponsored workers’ compensation reinsurance	1	0.01
50	Cooperative hospital service organization	9	0.05
82	Political organizations	4,484	25.73
	Total	17,430	100

Note. Data compiled by authors. Source: Return of Organization Exempt from Income Tax (IRS Form 990), National Center for Charitable Statistics (NCCS). N=17,430.

For analysis, we collected data for all non-501(c)(3) nonprofits who filed a Return of Organization Exempt from Income Tax (IRS Form 990) any year from 1989 to 2010, which we obtained from the National Center for Charitable Statistics (NCCS).² While many of these non-501(c)(3) nonprofits are observed each year, nearly as many organizations are not observed each year. In addition, the data provided by NCCS are not longitudinal, but rather a compilation of cross-sectional data reported each year. As such, we analyze the full population of non-501(c)(3) nonprofits and also split the data into a balanced panel versus cross-section (i.e., non-panel) and use mixed methods to compare the results. The full data set yields 3,727,477 observations with numbers of organizations ranging from 95,275 in 1997 to 226,667 in 1989. The most recent year of our data set (2010) includes 166,478 organizations. The largest numbers of nonprofits that spent at least $1 on lobbying or legislative activities occurred in years 2001 (n = 2,573), 2002 (n = 2,834), and 2003 (n = 2,657).

We use these data to estimate spending on political activities by non-501(c)(3) nonprofits (E) as a function of government regulatory policy determining whether charitable contributions are tax deductible (GR), organizational capacity (OC), financial strategies (FS), operating environment (OE), and industry controls (IC) as specified in Equation 1.³

l n (E_{i}) = α + {G R}_{i} β_{1} + {O C}_{i} β_{2} + {F S}_{i} β_{3} + {O E}_{i} β_{4} + {I C}_{i} β_{5} + ε_{i}

(1)

E_i is the sum total of a non-501(c)(3) nonprofit’s political expenditures (direct or indirect, line 81a), expenses paid for legislative activities (schedule A, part III, line 1), lobbying to influence public opinion (schedule A, Part VI-A), and lobbying to influence a legislative body (schedule A, Part VI-A); natural log values are used for analysis. GR_i is our main independent variable of interest and represents a dichotomous variable with a value of 1 if an organization’s contributions are tax deductible for the donor (i.e., IRS deductibility code = 2), and 0 otherwise. We hypothesize that non-501(c)(3) nonprofits in which charitable contributions received are tax deductible for donors are more regulated by the federal government and will therefore spend less on political activities in a given year, so we expect the sign of the coefficient for this variable to be negative.

Given that government officials might place stricter (or weaker) regulations on non-501(c)(3) nonprofits if they are more likely to engage in political activities, a nonprofit’s decision to engage in lobbying, advocacy, or other political activities might be endogenously determined. To overcome this potential endogeneity and correct for likely selection bias, we follow the Heckman selection model two-step approach. Our first stage predicts the unobserved preference of a non-501(c)(3) nonprofit to spend resources on political activities using a logistic regression model. This preference is estimated as a function of a vector of exogenous characteristics (Z), which must include at least one variable that influences the nonprofit’s decision to engage in political activities but does not influence the amount a nonprofit expends on such activities. We use three variables to serve this purpose (see Table 3): (a) Prior Political Activity Dummy (a dichotomous variable coded 1 if an organization attempted to influence legislation, had nonzero political expenditures, or had nonzero lobbying expenditures at least once during the prior decade [1989–1999], and 0 otherwise), (b) Prior Political Activity Count (the cumulative number of years an organization attempted to influence legislation, had nonzero political expenditures, or had nonzero lobbying expenditures during the prior decade [1989–1999]), and (c) Post-1990s Organizations (a dichotomous variable coded 1 if an organization’s Federal Employment Identification Number [EIN] is NOT observed at any point during the 1989–1999 time period, and 0 otherwise). We then use predicted probabilities from the first-stage logistic regression to estimate Equation 1 for the time period 2000–2010.⁴

Table 3.

Variables and Definitions.

Variable names	Variable descriptions
Dependent variables
Political, legislative, lobbying activity dummy	Dichotomous variable indicating if an organization had any political expenditures, expenses paid for legislative activities, lobbying to influence public opinion, or lobbying to influence a legislative body.
Political, legislative, lobbying activity amount	Summation of an organization’s political expenditures (direct or indirect, line 81a), expenses paid for legislative activities (schedule A, part III, line 1), lobbying to influence public opinion (schedule A, Part VI-A), and lobbying to influence a legislative body (schedule A, Part VI-A).
Exogenous instruments
Prior political activity dummy	Dichotomous variable indicating if an organization attempted to influence legislation (schedule A, part III, line 1), had political expenditures (part VI: other information), or lobbying expenditures (Sch.A.VI-A.37b or column B) at least once during the prior decade (1989–1999).
Prior political activity count	Cumulative number of times an organization attempted to influence legislation (schedule A, part III, line 1), had political expenditures (part VI: other information), or lobbying expenditures (Sch.A.VI-A.37b or column B) during the prior decade (1989–1999).
Post-1990s organizations	Dichotomous variable indicating if an organization’s Federal Employment Identification Number (EIN) is not observed at any point during the 1989–1999 time period.
Organizational capacity
Other salaries and wages	Dichotomous variable indicating if an organization reported nonzero dollar values of other salaries and wages (total, column A, line 26).
Officer/director compensation	Dichotomous variable indicating if an organization reported nonzero dollar values of compensation for officers, directors, etc. (total, column A, line 25).
Year-end assets	Dollar value of total assets at end-of-year (column B, line 59).
Organizational financial strategies
Program service revenue	Ordinal variable coded -1 if an organization reported negative dollar values of program service revenue (line 2), 0 for zero dollar values, and 1 for positive dollar values.
Professional fundraising fees	Ordinal variable coded -1 if an organization reported negative dollar values of professional fundraising fees (total, column A, line 3), 0 for zero dollar values, and 1 for positive dollar values.
Fundraising expenses	Ordinal variable coded -1 if an organization reported negative dollar values of fundraising expenses (line 15), 0 for zero dollar values, and 1 for positive dollar values.
Organizational operating environment
Government regulation of nonprofit contributions	Dichotomous variable indicating if an organization’s contributions are tax deductible for the donor (IRS deductibility code = 2).
Substantial government support	Dichotomous variable indicating if an organization is not classified as a private foundation because it receives substantial support from government or the general public (reason code = 7).
Substantial contributions & membership fees	Dichotomous variable indicating if an organization receives >33 1/3% of support from contributions and/or membership fees (reason code = 9).
Contributions revenue	Ordinal variable coded -1 if an organization reported negative dollar values of total public support (line 1d), 0 for zero dollar values, and 1 for positive dollar values.
Organizational controls
Organizational age	Number of years from 2020 since IRS ruling or determination letter recognizing an organization’s exempt status.
Political organization	Dichotomous variable indicating if an organization is classified by the IRS as a political organization (subseccd = 82).
Arts, culture, and humanities	Dichotomous variable indicating if an organization’s major operating subsector is arts, culture, and humanities (ntmaj5 = AR).
Education	Dichotomous variable indicating if an organization’s major operating subsector is education (ntmaj5 = ED).
Health	Dichotomous variable indicating if an organization’s major operating subsector is health (ntmaj5 = HE).
Human services	Dichotomous variable indicating if an organization’s major operating subsector is human services (ntmaj5 = HU).
Civil rights, social action, advocacy	Dichotomous variable indicting if an organization’s NTEE major group is civil rights, social action, advocacy (ntee1 = R).

Note. EIN = Employment Identification Number; IRS = Internal Revenue Service; NTEE = National Taxonomy of Exempt Entities.

In stage 1, the unobserved preference of a non-501(c)(3) nonprofit to engage in political activities is expressed as $E D_{d i}^{*} = Ζ_{i} Φ_{d} + ω_{d i}$ , where $ω_{d i}$ is a random error term and d represents a non-501(c)(3) nonprofit’s observed political activity status (E_i), which takes on a value of 1 each year the organization had any political expenditures, expenses paid for legislative activities, lobbying to influence public opinion, or lobbying to influence a legislative body, and 0 otherwise. The probability that a non-501(c)(3) nonprofit engages in political activities d is expressed as

P r (E D_{i} = d) = P r (ω_{d i} < Ζ_{i} Φ_{d}) = F (Ζ_{i} Φ_{d})

(2)

where F(•) is the cumulative distribution function of ω_d. Equation 2 is estimated using the observed status of a non-501(c)(3)’s engagement in political activities (Ed_i) and a vector of organizational characteristics (Zi) to determine the coefficients (Φ_d). From this first stage, we calculate the predicted probability of a non-501(c)(3) having any political expenditures, expenses paid for legislative activities, lobbying to influence public opinion, or lobbying to influence a legislative body (P_1i).

By allowing the relationship between nonprofit political activity and the vector of independent variables in Equation 1 to vary by the status of an organization’s engagement in political activities, the expected value of the natural log of political activity expenditures can be expressed as

E [l n E_{i}] = P_{0 i} [E [(l n E) | d = 0]] + P_{1 i} [E [(l n E) | d = 1]]

(3)

where

E [(l n E) | d] = α_{d} + {O C}_{i} β_{2 d} + {F S}_{i} β_{3 d} + {O E}_{i} β_{4 d} + {I C}_{i} β_{5 d} + ε_{i} η_{d}

(4)

By constraining the slope coefficients in Equation 4 to be equal across each political activity outcome d, Equation 3 can be simplified and expressed as

E [l n E_{i}] = α + P_{1 i} β_{1} + {O C}_{i} β_{2} + {F S}_{i} β_{3} + {O E}_{i} β_{4} + {I C}_{i} β_{5} + ε_{i} η

(5)

We use Equation 5 for the second-stage estimation. This approach corrects for selection bias in non-501(c)(3) nonprofit engagement in political activities and allows for unbiased estimates to be obtained in the second-stage estimation.

Empirical Results

First-Stage Estimation

Table 4 provides descriptive statistics for our first-stage estimation. The minimum and maximum values are provided for each variable. Then, the mean values of each variable are provided for the full population of non-501(c)(3) nonprofits filing IRS Form 990 anytime between 2000 and 2010 (i.e., population mean values), while compared to the mean values for only those organizations observed every year between 2000 and 2010 (i.e., panel sample mean values) and those organizations that are not observed each year (i.e., nonpanel sample mean values). All mean values of dichotomous variables are reported in percentage terms for ease of interpretation. The nonpanel sample contains 45.99% (N = 781,741) of the total 1,699,955 observations, while the panel subsample contains 54.01% (N = 918,214) of total observations.

Table 4.

First-Stage Descriptive Statistics.

Variables	Minimum	Maximum	Full sample mean values	Nonpanel sample mean values	Panel sample mean values
Dependent variable
Political, legislative, lobbying activity dummy	0	1	1.03%	1.12%	0.94%
Exogenous instruments
Prior political activity dummy	0	1	3.98%	1.98%	5.67%
Prior political activity count	0	27	0.0852	0.0342	0.1286
Post-1990s organizations	0	1	19.47%	39.62%	2.30%
Organizational capacity
Other salaries and wages	0	1	32.31%	20.28%	42.55%
Officer/director compensation	0	1	25.28%	16.53%	32.73%
Year-end assets (Ln)	$0.22	$27.34	$12.45	$11.81	$12.98
Organizational financial strategies
Program service revenue	0	1	46.28%	39.32%	52.%21
Professional fundraising fees	0	1	1.62%	1.78%	1.49%
Fundraising expenses	0	1	3.56%	3.83%	3.33%
Organizational operating environment
Government regulation	0	1	25.68%	33.25%	19.23%
Substantial government support	0	1	0.99%	1.60%	0.47%
Substantial contributions & membership fees	0	1	1.96%	3.05%	1.03%
Contributions revenue	0	1	38.46%	43.61%	34.07%
Organizational controls
Organizational age	9	120	48.5978	44.2626	51.7928
Political organization	0	1	1.98%	4.22%	0.08%
Arts, culture, and humanities	0	1	2.14%	2.37%	1.95%
Education	0	1	4.99%	5.96%	4.15%
Health	0	1	3.94%	3.70%	4.14%
Human services	0	1	32.96%	28.34%	36.90%
Civil rights, social action, advocacy	0	1	0.42%	0.52%	0.34%
Observations	1,699,955	1,699,955	1,699,955	781,741	918,214

Note. Mean values for all dichotomous variables are reported in percentage terms for ease of interpretation and discussion.

Table 4 shows that only 1.03% of non-501(c)(3) organizations reported any kind of engagement in political, legislative, or lobbying activities in 2000–2010. For organizations observed each year, only 0.94% reported any sort of political activity compared to 1.12% of organizations that were not observed every year. These mean values fall below those of the prior decade (1989–1999) when 3.98% of all non-501(c)(3) nonprofits reported some political activity. This higher overall participation rate in the prior decade is also unevenly divided between the two samples, with only 1.98% of nonpanel sample organizations reporting political activities compared to 5.67% of those in the panel sample. Relatedly, only 1.98% of total observations reflect organizations classified by the IRS as political organizations. Perhaps more interesting is that the organizations in the panel sample represent political organizations only 0.08% of the time, while the nonpanel sample reflects political organizations 4.22% of the time, suggesting shorter longevity. These descriptive findings are important, as they suggest the extent to which non-501(c)(3) nonprofits engage in political activities might be overstated anecdotally and/or within extant research.

In terms of government regulation, non-501(c)(3) nonprofits are somewhat restricted with 25.68% of organizations having charitable contributions that are tax deductible for donors. However, there is variation in this regulatory policy, as less than 20% of organizations in the panel sample and more than 33% of nonprofits in the nonpanel sample are allowed such a tax deduction. While a nontrivial proportion of non-501(c)(3) nonprofits are faced with government regulatory policy regarding their charitable contributions, very few of them are heavily reliant on government support. Table 4 shows that 0.99% of total observations are not classified as private foundations because they receive a substantial amount of financial support from the government or general public (i.e., substantial government support).

Table 4 also shows that non-501(c)(3) organizations are 48½ years old on average; the nonpanel sample organizations are a little younger at 44 years on average compared to the panel sample with an average age of nearly 52 years. Almost half the organizations (46.28%) generate program service revenue with the panel sample nonprofits even more likely (at 52.21%) compared to only 39.32% of nonpanel sample organizations. When it comes to contributions revenue, we see the opposite trend. Overall, 38.46% of non-501(c)(3) organizations solicit contributions revenue with the nonpanel sample more reliant (at 43.61%) than the panel sample reporting contributions revenue 34.07% of the time. Finally, non-501(c)(3) organizations provide human services the most (32.96%) out of the other industry categories. In fact, the organizations classified in the civil rights, social action, and advocacy industry sector represent only 0.42% of total observations.

Table 5 provides three sets of pooled logistic regression results. Model 1 analyzes the full population, and Models 2 and 3 provide the results for the panel and nonpanel samples for comparison.⁵ Each model is estimated with observed information matrix (OIM) standard errors and includes a series of n−1 dichotomous variables indicating each year and state (not reported) to account for variation in unobserved factors. All models display chi-square values that are statistically significant at the 99% confidence level and pseudo R² values ranging from 0.1410 to 0.2326. The coefficient values reported in Table 5 reflect odds ratios.⁶

Table 5.

Odds Ratios of First-Stage Logistic Regression.

Variables	Model 1	Model 2	Model 3
Variables	Full sample pooled logit	Nonpanel sample pooled logit	Panel sample pooled logit
Exogenous instruments
Prior political activity dummy	13.62***	12.89***	13.95***
Prior political activity count	1.05***	1.03**	1.06***
Post-1990s organizations	1.79***	1.63***	1.73***
Organizational capacity
Other salaries and wages	1.03	1.16***	0.98
Officer/director compensation	2.00***	1.94***	2.04***
Year-end assets	1.07***	1.03***	1.09***
Organizational financial strategies
Program service revenue	0.86***	0.78***	0.90***
Professional fundraising fees	1.10	1.24**	1.04
Fundraising expenses	1.30***	1.18**	1.39***
Organizational operating environment
Government regulation	-2.00***	-1.71***	-2.30***
Substantial government support	1.32***	1.43***	0.90
Substantial contributions & membership fees	2.24***	2.14***	2.24***
Contributions revenue	1.06***	1.07*	1.06*
Organizational controls
Organizational age	1.00	1.00	1.00
Political organization	13.09***	15.59***	9.87***
Arts, culture, and humanities	1.07	0.79	1.25**
Education	1.34***	1.34***	1.33***
Health	1.84***	2.34***	1.62***
Human services	1.39***	1.47***	1.36***
Civil rights, social action, advocacy	1.76***	1.69***	1.84***
Constant	0.0007***	0.0010***	0.0005***
Observations	1,372,725	570,699	802,026
LR chi-square	27,903.56	6,449.12	21,313.53
Prob > chi-square	0.0000	0.0000	0.0000
Pseudo R²	0.2024	0.1410	0.2326

Note. Reported values reflect odds ratios with observed information matrix (OIM) standard errors and statistical significance indicated as: ***p < .01; **p < .05; *p < .10. Although not reported, pooled estimation results include n−1 dichotomous variables indicating year and state “fixed effects.” Dollars were adjusted to 2020 constant values using the Consumer Price Index.

Table 5 shows prior engagement in political activities has a statistically significant (at the 99% confidence level) and nontrivial (in terms of magnitude) impact on the likelihood that a non-501(c)(3) organization will spend resources on these activities again. Overall, if a non-501(c)(3) nonprofit attempted to influence legislation or reported political or lobbying expenditures at least once during the prior decade (1989–1999), the organization is 13.62% more likely to engage in such activities again. Prior political activities make the nonpanel sample nonprofits 12.89% more likely to engage in political activities again. Among the panel sample, prior political activities make them 13.95% more likely to undertake such actions again. These are among the highest odds ratios reported in Table 5.

Table 5 also shows that the cumulative number of times these non-501(c)(3) organizations engaged in political activities during the prior decade increases the likelihood of repeating such actions by a little over 1% (1.03%–1.06%) each time. However, prior engagement in political activities does not singlehandedly motivate non-501(c)(3) nonprofits to do so. Table 5 shows that organizations not observed in the prior decade (either because they did not exist or did not file a 990 with the IRS) are also more likely (between 1.63% and 1.79%) to engage in political activities.

The positive impact of prior political activities is only matched in magnitude by the impact of non-501(c)(3) nonprofits being classified by the IRS as political organizations. Such organizations are between 9.86% (panel sample) and 15.59% (nonpanel sample) more likely to report political expenditures. Conversely, non-501(c)(3) nonprofits reporting program service revenue are 77.97% to 89.54% less likely to engage in political activities. Non-501(c)(3) nonprofits facing government regulation (i.e., contributions tax deductible for donors) are 1.71% to 2.30% less likely to engage in political activities. Among the remaining variables in Table 5, officer/director compensation, the value of assets at fiscal year-end, fundraising expenses, substantial (i.e., greater than 33 1/3%) income from contributions and membership fees, and an organization’s major operating subsector all increase the likelihood of a non-501(c)(3) nonprofit engaging in political activities.

Second-Stage Estimation

Table 6 provides descriptive statistics for our second-stage estimation, which includes only those non-501(c)(3) organizations who reported nonzero political expenditures, expenses paid for legislative activities, lobbying to influence public opinion, and lobbying to influence a legislative body. Again, both the minimum and maximum values are provided for each variable. Then, the mean values are provided for the population of non-501(c)(3) nonprofits filing an IRS Form 990 anytime 2000 to 2010, while compared to the mean values for the panel and nonpanel samples. The nonpanel sample contains 50.42% (n = 8,788) of the total 17,430 observations, while the panel sample contains 49.58% (n = 8,642) of total observations.

Table 6.

Second-Stage Descriptive Statistics.

Variables	Minimum	Maximum	Full sample mean values	Non-panel sample mean values	Panel sample mean values
Dependent variable
Political, legislative, lobbying activity amount (Ln)	$0.26	$17.79	$8.67	$9.19	$8.14
Organizational capacity
Other salaries and wages	0	1	42.31%	22.82%	62.14%
Officer/director compensation	0	1	43.33%	22.21%	64.80%
Year-end assets (Ln)	$0.26	$23.02	$12.60	$11.58	$13.61
Organizational financial strategies
Program service revenue	0	1	34.33%	20.23%	48.67%
Professional fundraising fees	0	1	3.68%	4.70%	2.65%
Fundraising expenses	0	1	6.73%	7.90%	5.54%
Organizational operating environment
Government regulation	0	1	34.15%	59.89%	7.97%
Substantial government support	0	1	0.75%	1.21%	0.29%
Substantial contributions & membership fees	0	1	2.52%	3.13%	1.90%
Contributions revenue	0	1	45.85%	64.70%	26.68%
Organizational controls
Organizational age	10	108	52.1686	46.3960	54.9372
Political organization	0	1	25.73%	49.82%	1.23%
Arts, culture, and humanities	0	1	1.22%	0.81%	1.63%
Education	0	1	3.43%	2.94%	3.92%
Health	0	1	4.90%	3.68%	6.14%
Human services	0	1	35.52%	20.41%	50.88%
Civil rights, social action, advocacy	0	1	0.65%	0.55%	0.76%
Observations	17,430	17,430	17,430	8,788	8,642

Note. Mean values for all dichotomous variables are reported in percentage terms for ease of interpretation and discussion.

According to Table 6, the non-501(c)(3) organizations reported spending for political, legislative, or lobbying activities in 2000 to 2010 ranging from $0.26 to $17.79 natural log dollars, which equals $1.30 to $53,159,639 actual dollars. On average, these non-501(c)(3) nonprofits spent $5,827 actual dollars. The panel sample organizations spent considerably less, on average, at $3,442 actual dollars, compared to the nonpanel sample nonprofits that spent $9,781 on average. These non-501(c)(3) nonprofits are classified by the IRS as political organizations 25% of the time with 49.82% of the non-panel sample organizations classified as political organizations compared to only 1.23% of nonprofits in the panel sample.

Table 6 illustrates another interesting trend regarding government regulation. These non-501(c)(3) organizations are a little more regulated overall with 34.15% having charitable contributions that are tax deductible for donors; less than 8% of organizations in the panel sample compared to nearly 60% of nonpanel sample organizations had charitable contributions that were tax deductible. Although a large proportion of organizations face this government regulatory policy regarding their charitable contributions, only 0.75% of total observations are classified as private foundations because they receive a substantial amount of financial support from the government or general public (i.e., substantial government support). Only 2.5% of total observations reported any income from contributions or membership fees with the larger portion (3.13%) among the nonpanel sample compared to 1.9% of panel sample organizations.

Finally, Table 6 shows non-501(c)(3) nonprofits are 52 years old on average; the nonpanel sample organizations are 46 years old on average compared to the panel sample at nearly 55 years. In addition, fewer organizations (34.33%) generate program service revenue; the more established organizations are even more likely (at 48.67%) to generate income from this source compared to only 20.23% of nonpanel sample organizations. Again, 45.85% of non-501(c)(3) nonprofits solicit contributions revenue; the nonpanel sample is more heavily reliant (at 64.70%) than the panel sample reporting contributions revenue 26.68% of the time. Finally, these non-501(c)(3) nonprofits still provide human services the most (35.52%) out of the other industry categories. Once again, organizations classified in the civil rights, social action, and advocacy subsector represent less than 1% of observations.

Table 7 provides three sets of pooled ordinary least squares (OLS) regression results.⁷ Model 1 analyzes the full population, and Models 2 and 3 provide the results of analysis on the panel and nonpanel samples for comparison.⁸ Each model was estimated with robust standard errors and includes a series of n−1 dichotomous variables indicating each year and state (although not reported) to account for variation in unobserved factors. All three models display F-values that are statistically significant at the 99% confidence level and adjusted R² values ranging from 0.1999 to 0.2446.

Table 7.

Second-Stage Regression Results.

Variables	Model 1	Model 2	Model 3
Variables	Full sample pooled OLS	Nonpanel sample pooled OLS	Panel sample pooled OLS
Exogenous instrument
Predicted probability of political, legislative, lobbying activity from first-stage results	7.82***	10.85***	6.85***
Organizational capacity
Other salaries and wages	0.20***	0.24	0.16*
Officer/director compensation	0.62***	0.82***	0.58***
Year-end assets	0.10***	0.09***	0.08***
Organizational financial strategies
Program service revenue	0.10	−0.06	0.12*
Professional fundraising fees	0.26	0.70**	−0.05
Fundraising expenses	0.54***	0.71***	0.43***
Organizational operating environment
Government regulation	−1.92***	−1.63***	−2.19***
Substantial government support	−0.82**	−0.77	−0.70
Substantial contributions & membership fees	0.20	0.19	0.42*
Contributions revenue	0.26***	0.30**	0.28***
Organizational controls
Organizational age	−0.01***	−0.01***	−0.01***
Political organization	3.35***	4.22***	1.42***
Arts, culture, and humanities	0.25	−0.30	0.41*
Education	0.44***	0.30	0.42**
Health	1.39***	1.54***	1.24***
Human services	0.38***	0.41***	0.35***
Civil rights, social action, advocacy	1.08***	1.60***	0.56*
Constant	5.05***	5.99***	4.82***
Observations	12,024	3,808	8,216
F-value	42.61	20.45	72.79
Prob > F	0.0000	0.0000	0.0000
Adjusted R²	0.2133	0.2446	0.1999

Note. OLS = ordinary least squares.

Reported values reflect coefficients with robust standard errors and statistical significance indicated as: ***p < .01; **p < .05; *p < .10. Although not reported, pooled estimation results include n−1 dichotomous variables indicating year and state “fixed effects.” Dollars were adjusted to 2020 constant values using the Consumer Price Index.

Table 7 shows the predicted probability of political, legislative, and lobbying activities (from the first-stage results) is statistically significant at the 99% confidence level across all three models with coefficient values that are nontrivial in terms of magnitude. In Model 1, a 1% increase in the predicted probability of a non-501(c)(3) organization having nonzero political expenditures, expenses paid for legislative activities, lobbying to influence public opinion, or lobbying to influence a legislative body increases spending for these activities of $7.82 (rounded from $7.823171) natural log dollars, which amounts to $2,489.91 in actual dollars (calculated as the inverse of the natual log). Among the nonpanel sample in Model 2, a 1% increase in the predicted probability of political expenditures increases spending on these activities by $10.85 (rounded) natural log dollars, which equals $51,353.58 actual dollars. Among the panel sample in Model 3, increasing the predicted probability of political expenditures by 1% leads to a $6.85 (rounded) increase in natural log dollars or $943.88 real dollars for these activities. This is the largest impact in terms of magnitude in Table 7.

Table 7 also shows that when these non-501(c)(3) nonprofits have charitable contributions tax deductible for their donors, which symbolizes more government regulation since such organizations are required to limit their spending on political activities, these nonprofits actually do decrease their spending. Non-501(c)(3) nonprofits with tax-deductible contributions spend $1.92 (rounded) natural log dollars ($6.82 actual dollars) less than organizations without tax-deductible contributions. For the nonpanel sample, spending on political activities decreases by $1.63 (rounded) natural log dollars (or $5.10 actual dollars) compared to the panel sample of $2.19 (rounded) natural log dollars (or $8.94 actual dollars) when their contributions are tax deductible. Although the magnitude of impact from this government regulatory policy is rather small, it is the third largest among all variables in Table 7.

The second largest impact in Table 7 comes from a non-501(c)(3) organization being classified by the IRS as a political organization, which leads to an increase in political expenditures of $1.42 (rounded) natural log dollars (or $4.14 actual dollars) for the panel sample and $4.22 (rounded) natural log dollars (or $68.03 actual dollars) for the nonpanel sample. Finally, Table 7 shows officer/director compensation, the value of assets at fiscal year-end, organizational age, fundraising expenses, having a fundraising-based strategy of contributions revenue, and an organization’s major operating subsector all increase spending by non-501(c)(3) organizations to support political activities.

Figure 1 presents the results of our hypothesis testing. Hypotheses 1a and 1b proposed that human and financial capacity would be positively associated with an organization’s political activity; the regression results support this prediction. This suggests that an organization tends to engage in more political activities as the organization has more resources to allocate to political activities (i.e., lobbying) beyond its mission-related programs and services. Hypotheses 2a and 2b stipulated that market-oriented and donor-focused strategies would be positively associated with an organization’s political activities; hypothesis 2a is partially supported with mixed results across the models, while hypothesis 2b is fully supported. Specifically, if an organization increases its fundraising expenses, the nonprofit is more likely to invest its resources in political efforts. Our results also fully support Hypothesis 3 that an organization’s political activities are negatively associated with government regulation, suggesting that a non-501(c)(3) organization’s tax deductibility of contributions may determine the scope of its political activities. However, Hypothesis 4 is rejected as government support is negative and significant in Model 1 but not significant in the other two models. Finally, we predicted that an organization’s political activities would be positively associated with donor (H5a) and member dependency (H5b). Hypothesis 5a receives full support, suggesting non-501(c)(3) organizations with more contributions from donors are likely to increase their political activities; however, hypothesis 5b is only partially supported as membership fee dependency is only significant in Model 3.

Figure 1.

Results of hypothesis testing.

Discussion

We examined factors that shape the political activities of nonprofits that are not 501(c)(3) organizations. This study makes several significant contributions to the literature and practice. First, this is one of the first empirical investigations into the political activities of non-501(c)(3) nonprofits. As such, we fill a current gap in research on government regulation and nonprofit political activity, which has largely focused on 501(c)(3) charitable or human service organizations. Our study indicates that only 17,430 or 1.03% of non-501(c)(3) organizations participated in political activities (i.e., lobbying and legislative activities) in the 2000s, which is lower than that of 501(c)(3) charitable organization in the United States (e.g., 67%, Guo & Saxton, 2010) and Singapore (e.g., 76.8%; Guo & Zhang, 2014). This difference may be due to the fact that we only captured a narrow aspect of political activities (i.e., lobbying and legislative activities), while other studies also included advocacy efforts such as research and public education, which are not regulated by the federal government.

Second, we proposed and empirically tested a conceptual model for understanding nonprofit political activities by non-501(c)(3) organizations that considered three categories of explanatory factors: organizational capacity, financial strategy, and operating environment. We proposed that government regulation (through the deductibility of charitable contributions) influences an organization’s political activities. Our empirical findings provide some initial support for our conceptual model and also show the need for improvement. Specifically, consistent with previous literature on nonprofit advocacy (e.g., Andrews & Edwards, 2004; J. M. Berry & Arons, 2003; Boris & Mosher-Williams, 1998 ; Child & Grønbjerg, 2007 ; Guo & Saxton, 2010; Guo & Zhang, 2014), the majority of our hypotheses received support from the regression results. However, dependency on government support was found to be either not significant or negatively associated with political activities. This result is somewhat surprising, as much prior research on 501(c)(3) charitable organizations has found a positive relationship. One possible explanation is that non-501(c)(3) nonprofits fundamentally differ from 501(c)(3) charities when making decisions about whether to engage in political activities as they tend to perceive “government funding as a weak catalyst rather than an obstacle” (Lu, 2018b, p. 203).

Our findings also suggest that organizations with past advocacy or lobbying experience are more likely to formally participate in the political process. This implies that to become an active participant in the development of public policy and to engage formally with government in the policy process, an organization must first develop the experience. In addition, we found that non-501(c)(3) organizations with tax-deductible contributions are less likely to engage in political activities than those without tax-deductible contributions. While the magnitude of this finding is rather small, we believe it is an important indicator of the relationship between government tax benefits and nonprofit political activities, which appears to be adversarial (Young, 2000). It seems nonprofits do not want to challenge existing government policy, because doing so may threaten the continuity of tax benefits received (Pfeffer & Salancik, 1978; Wolch, 1990; Lu, 2018b). In addition, nonprofits’ tax benefits could be considered more of a regulation, rather than a catalyst for non-501(c)(3) organizations (Lu, 2018b). It is also likely that complex government regulations on political activities make nonprofit leaders hesitant to engage in political activities (Berry & Arons, 2003; Reid, 1999).

While these findings might not be entirely surprising, combined with our findings that very few organizations spend resources on lobbying, this is notable. This normative concern is relevant because nonprofits should—and are often encouraged—to participate in the political process by following the examples and strategies of those organizations with advocacy missions (L. L. Berry, 1999). As such, we believe our findings make an important contribution to the dialogue about whether government financial benefits encourage or discourage nonprofits’ participation in political activities.

A limitation of this study is that our analysis is unable to capture informal engagements of non-501(c)(3) organizations. Nonprofits vary significantly in the types of political activities they engage in besides lobbying. Non-501(c)(3) nonprofits might also actively engage in other types of political activities such as public education, electoral advocacy (Reid, 1999), media advocacy, coalition building (Avner, 2002), developing consensus among experts (Hoefer, 2001), and expert testimony (Guo & Saxton, 2010). As such, the dependent variable of our study—the amount spent on political activities—is somewhat limited in capturing all potential political activities, particularly those that are not regulated by the federal government. As such, we invite future research to further explore these other advocacy activities of non-501(c)(3) organizations.

Finally, for the non-501(c)(3) organizations we examined, we expected more active participation in political and lobbying activities, particularly because they are less regulated than 501(c)(3) charities. Yet, we found only a small portion of these organizations participate. This raises many questions. One alternative is simply that the presence of nonprofits in the political arena may be overstated in extant literature. We think our results support a call for re-examination of the role of nonprofits in political discourse. Alternatively, there might be a much bigger, systemic reporting problem to be explored further. Prentice (2018) examines the incidence of misreporting of lobbying expenses on the IRS 990 form. While evidence of misreporting is found, his work identifies many characteristics of nonprofits that are less likely to have misreported.

Ultimately, our empirical findings have practical implications as we show that nonprofits’ human and financial capacity is positively associated with their political activities. We believe it is likely that organizations with lots of resource availability are likely to have more opportunities to engage in political activities. In addition, nonprofits more dependent on donors are also motivated to engage in lobbying and legislative activities to show their donors how their money is being used to represent their voice in the policy process.

Conclusion

We proposed a conceptual model of the determinants of political activity by nonprofits that qualify for exemption under subsections of the Internal Revenue Code other than 501(c)(3) (i.e., non-501[c](3) nonprofits). Our model considered three categories of explanatory factors—organizational capacity, financial strategy, and operating environment. Using a Heckman selection process with longitudinal data, we examined the role of government regulatory policy (measured by whether a nonprofit’s contributions are tax deductible for the donor or not) on its political activities as indicated by the amount of political, legislative, or lobbying expenditures reported on the organization’s IRS form 990. Overall, we found that non-501(c)(3) nonprofits that face government regulation are 1.71% to 2.30% less likely to engage in political activities. In addition, non-501(c)(3) nonprofits with tax-deductible charitable contributions spend $6.84 less on these activities than nonprofits not regulated by government. These findings suggest that government regulation could be an obstacle for nonprofits to engage in the policy process.

Footnotes

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Deborah A. Carroll

Suzette Myser

Notes

Author Biographies

Deborah A. Carroll is an associate professor in the Department of Public Administration and Director of the Government Finance Research Center in the College of Urban Planning and Public Affairs at University of Illinois-Chicago. Her research focuses on government and nonprofit financial management.

Suzette Myser is an assistant professor in the School of Public Administration at University of Central Florida with research focusing on nonprofit financial management.

Seongho An is an assistant professor in the School of Public Administration at University of Central Florida. His research focuses on E-advocacy of nonprofit organizations and digital technology in nonprofit management.

References

Andrews

K. T.

Edwards

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

Avner

(2002). The lobbying and advocacy handbook for nonprofits. Wilder Foundation.

Bass

G. D.

Arons

Guinane

Carter

M. F.

Rees

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

Berry

J. M.

Arons

D. F.

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

Berry

L. L.

(1999). Discovering the soul of service: The nine drivers of sustainable business success. Simon and Schuster.

Boris

Mosher-Williams

(1998). Nonprofit advocacy organizations: Assessing the definitions, classifications, and data. Nonprofit and Voluntary Sector Quarterly, 27(4), 488–506.

Casey

(2011). Understanding advocacy: A primer on the policy-making role of nonprofits. Center for Nonprofit Strategy, Baruch College, City University of New York. https://goo.gl/oegXGG

Chaves

Stephens

Galaskiewicz

(2004). Does government funding suppress nonprofits’ political activity? American Sociological Review, 69(2), 292–316.

Cheng

(2019). Nonprofit spending and government provision of public services: Testing theories of government–nonprofit relationships. Journal of Public Administration Research and Theory, 29(2), 238–254.

10.

Child

C. D.

Grønbjerg

K. A.

(2007). Nonprofit advocacy organizations: Their characteristics and activities. Social Science Quarterly, 88(1), 259–281.

11.

Clemens

E. S.

Guthrie

(2010). Politics and partnerships: The role of voluntary associations in America’s political past and present. University of Chicago Press.

12.

Coates

David

(2002). Learning for change: The art of assessing the impact of advocacy work. Development in Practice, 12(3–4), 530–541.

13.

Donaldson

L. P.

(2008). Developing a progressive advocacy program within a human services agency. Administration in Social Work, 32(2), 25–48.

14.

Dutton

J. E.

Jackson

S. E.

(1987). Categorizing strategic issues: Links to organizational action. Academy of Management Review, 12(1), 76–90.

15.

Eising

(2007). Institutional context, organizational resources and strategic choices: Explaining interest group access in the European Union. European Union Politics, 8(3), 329–362.

16.

Fyall

(2016). The power of nonprofits: Mechanisms for nonprofit policy influence. Public Administration Review, 76(6), 938–948.

17.

Fyall

Allard

S. W.

(2017). Nonprofits and political activity: A joint consideration of the political activities, programs, and organizational characteristics of social service nonprofits. Human Service Organizations: Management, Leadership & Governance, 41(3), 275–300.

18.

Gais

Walker

J. L.

Jr. (1991). Pathways to influence in American politics. In Jack

Walker

(Eds.), Mobilizing interest groups in America (p. 184). University of Michigan Press.

19.

Garrow

E. E.

Hasenfeld

(2014). Institutional logics, moral frames, and advocacy: Explaining the purpose of advocacy among nonprofit human-service organizations. Nonprofit and Voluntary Sector Quarterly, 43(1), 80–98.

20.

Gibelman

Kraft

(1996). Advocacy as a core agency program: Planning considerations for voluntary human service agencies. Administration in Social Work, 20(4), 43–59.

21.

Gormley

W. T.

Jr. Cymrot

(2006). The strategic choices of child advocacy groups. Nonprofit and Voluntary Sector Quarterly, 35(1), 102–122.

22.

Graddy

E. A.

Morgan

D. L.

(2006). Community foundations, organizational strategy, and public policy. Nonprofit and Voluntary Sector Quarterly, 35(4), 605–630.

23.

Grasse

Ward

K. D.

Miller-Stevens

(2021). To lobby or not to lobby? Examining the determinants of nonprofits taking the IRS 501 (h) election. Policy Studies Journal, 49(1), 242–267.

24.

Guo

Musso

J. A.

(2007). Representation in nonprofit and voluntary organizations: A conceptual framework. Nonprofit and Voluntary Sector Quarterly, 36(2), 308–326.

25.

Guo

Saxton

G. D.

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

26.

Guo

Zhang

(2014). Understanding nonprofit advocacy in non-western settings: A framework and empirical evidence from Singapore. VOLUNTAS: International Journal of Voluntary and Nonprofits, 25(5), 1151–1174.

27.

Hoefer

(2001). Highly effective human services interest groups: Seven key practices. Journal of Community Practice, 9(2), 1–13.

28.

Hoefer

(2005). Altering state policy: Interest group effectiveness among state-level advocacy groups. Social Work, 50(3), 219–227.

29.

Jackson

(2014). Evaluation of public policy advocacy: Challenges, principles and BEST-AC case study. International Journal of Public Sector Management 27, 272–280.

30.

Jang

H. S.

Feiock

R. C.

(2007). Public versus private funding of nonprofit organizations: Implications for collaboration. Public Performance & Management Review, 31(2), 174–190.

31.

Junk

W. M.

(2016). Two logics of NGO advocacy: Understanding inside and outside lobbying on EU environmental policies. Journal of European Public Policy, 23(2), 236–254.

32.

Klüver

(2010). Europeanization of lobbying activities: When national interest groups spill over to the European level. European Integration, 32(2), 175–191.

33.

Lee

Rethemeyer

R. K.

Park

H. H.

(2018). How does policy funding context matter to networks? Resource dependence, advocacy mobilization, and network structures. Journal of Public Administration Research and Theory, 28(3), 388–405.

34.

Lewis

(2005). The civil society sector: A review of critical issues and research agenda for organizational communication scholars. Management Communication Quarterly, 19(2), 238–267.

35.

(2018a). Fear the government? A meta-analysis of the impact of government funding on nonprofit advocacy engagement. The American Review of Public Administration, 48(3), 203–218.

36.

(2018b). Organizational antecedents of nonprofit engagement in policy advocacy: A meta-analytical review. Nonprofit and Voluntary Sector Quarterly, 47(4_suppl), 177S–203S.

37.

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.

38.

McCarthy

J. D.

Zald

M. N.

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

39.

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.

40.

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.

41.

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.

42.

Nah

Saxton

G. D.

(2013). Modeling the adoption and use of social media by nonprofits. New Media & Society, 15(2), 294–313.

43.

Neumayr

Schneider

Meyer

(2015). Public funding and its impact on nonprofit advocacy. Nonprofit and Voluntary Sector Quarterly, 44(2), 297–318.

44.

Nicholson-Crotty

(2007). Politics, policy, and the motivations for advocacy in nonprofit reproductive health and family planning providers. Nonprofit and Voluntary Sector Quarterly, 36(1), 5–21.

45.

Pfeffer

Salancik

(1978). The external control of organizations: A resource dependence perspective. Harper & Row.

46.

Prentice

C. R.

(2018). Misreporting nonprofit lobbying engagement and expenses: Charitable regulation and managerial discretion. Nonprofit Policy Forum, 9(2), 2018–0018.

47.

Reid

E. J.

(1999). Nonprofit advocacy and political participation. In Boris

E. T.

Steuerle

C. E.

(Eds.), Nonprofits and government: Collaboration and conflict (pp. 291–325). The Urban Institute Press.

48.

Saidel

(2002). Nonprofits, political engagement, and public policy. Nonprofit Advocacy and the Policy Process, 2, 1–18.

49.

Salamon

L. M.

(2012). The state of nonprofit America. The Brookings Institution.

50.

Salamon

L. M.

Sokolowski

S. W.

List

(2004). Global civil society (Vol. 2). Kumarian.

51.

Schmid

Bar

Nirel

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

52.

Smith

S. R.

Grønbjerg

K. A.

(2006). Scope and theory of government-nonprofit relations. In Powell

W. W.

Steinberg

(Eds.), The nonprofit sector: A research handbook (2nd ed., pp. 221–242). Yale University Press.

53.

Suárez

D. F.

Hwang

(2008). Civic engagement and nonprofit lobbying in California, 1998-2003. Nonprofit and Voluntary Sector Quarterly, 37(1), 93–112.

54.

Tarrow

(1998). Power in movement: Social movements and contentious politics. Cambridge. Transnational Politics: Contention and institutions in international politics. Annual Review of Political Science, 4, 1–20.

55.

Teles

Schmitt

(2011). The elusive craft of evaluating advocacy. Stanford Social Innovation Review, 9(3), 40–43.

56.

Walker

J. L.

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

57.

Wolch

J. R.

(1990). The shadow state: Government and voluntary sector in transition. The Foundation Center.

58.

Young

D. R.

(2000). Alternative models of government–nonprofit sector relations: Theoretical and international perspectives. Nonprofit and Voluntary Sector Quarterly, 29(1), 149–172.