Organizational Determinants of Political Involvement in Trade and Professional Membership Associations

Abstract Trade and business associations and professional membership societies like many other political interest groups in the United States advocate and lobby their interests in the policy arena. Yet, we lack the understanding of how organizational capabilities shape different forms of political engagement in trade and professional associations. Research on the political involvement of tax-exempt organizations has been focused on charitable nonprofits. Studying the elements that underlie political action strategies and tactics of trade and professional associations would expand our understanding of political engagement in the nonprofit sector. In our sample of occupational mutual benefit associations, both resource dependence and institutional factors such as board size, affiliated foundation, tax-exempt status, business classification, and membership type explained the likelihood of political strategies including advocacy, lobbying, grassroots activity, and having a political action committee (PAC). Organizational size characteristics such as budget, staff, and membership size mattered as well.


Introduction
Interest groups dominate the American political system (Berry and Wilcox 2018). Their abilities to mobilize resources and information influence the shape and direction of public policy, its actors, and its processes. Interest groups traditionally form around shared views or issues that they want to see reflected in public policy outcomes (Grossman and Helpman 2001). Some interest groups organize around commercial, economic, and occupational interests. These groups may include mutual-benefit nonprofits such as trade associations and professional societies (Spillman 2012;Saitgalina et al. 2016). Political engagement is a tool these organizations employ to promote change in public policy that advances their missions. Their members, who have a stake in the outcomes of public policies, typically support such involvement. Grossman and Helpman (2001) identified two common groups of activities that interest groups use for their interest representation: activities that entail some form of information dissemination and activities that provide resources to candidates and parties. The first group of activities includes such tactics as educating the public (advocacy), informing and persuading legislators (lobbying), and mobilizing the public to action (grassroots). The second group encompasses activities related to finance and campaign-giving to support or oppose candidates for political office through political action committees (PACs).
The study of political activities of interest groups through a normative lens, which emphasizes controversy and the seeming impunity of the actions, has been common (for review see Spillman 2012). Other studies focused on estimating the probability of success of various political strategies and tactics of interest groups (for a review see Berry and Wilcox 2018). Yet, less is known about what organizational and environmental factors shape the involvement strategies of professional and trade associations in the policy arena.
Further, there are many studies on public policy advocacy of charitable social and human service nonprofits (Almog-Bar 2018; Smith and Pekkanen 2012;Fyall and Allard 2017), whereas fewer looked at these activities in trade and professional nonprofit associations. Understanding how the organizational capabilities of these associations relate to their political engagement strategies would allow us to identify elements that underlie the composition of political action. This research evaluates the effects of interest group politics and considers financial, regulatory, and organizational conditions underpinning the political involvement of trade and professional associations.
This study considers two major theoretical approaches that have been used in examining human service nonprofit' advocacy efforts -institutional theory (Powell and DiMaggio 1991) and resource dependence theory (Pfeffer and Salancik 1978). These theories argue that organizations are bound by and can leverage internal and external factors and resources to achieve their goals. Thus, this study asks how resource dependence and institutional theory can explain the choice of political strategies of trade and professional associations. Kollman (1998) argued that professional and special interest groups representing businesses and industries traditionally use insider tactics that allow them to lobby legislators directly or campaign for preferred candidates, whereas public and social interest groups tend to use outside or indirect tactics of grassroots activism and broader advocacy. Given that policy advocacy encompasses a variety of strategies, this study provides insights into the interplay of resources and factors that leverage different public policy strategies of trade and professional associations.
This study considers whether institutional norms such as tax-exempt status, affiliated foundations, type of membership, and business classification as well as organizational resources like board size and revenue derived from membership dues are associated with the choice of public policy strategy -lobbying, advocacy, grassroots, or operating PACs. Organizations that serve professional, economic, or occupational interests and have a defined membership that pays annual dues, are the subject of this study. This research builds on previous research in several ways. First, we focus on mutual-benefit nonprofits rather than traditionally studied human service nonprofits. Second, we assess the relative impact of institutional factors and resources while controlling for the size of trade and professional associations. Third, we encompass a variety of political strategy choices rather than focusing on any single type.
This study uses a sample of associations from the dataset provided by the American Society of Association Executives (ASAE) and the Center for Association Leadership. The survey data is used to test institutional and resource dependence theory assumptions against the public policy strategies of trade and professional associations.

Background Associations
According to the U.S. Census Bureau, there were 309,150 religious, grantmaking, civic, professional, and similar organizations in 2016 according to the North American Industry Classification System (US Census 2016). Of these, about 59,400 are business, professional, labor, political, and similar organizations. Internal Revenue Service (IRS) recognized 63,866 501(c)(6) associations in 2016, which includes business leagues, chambers of commerce, real estate boards, boards of trade, and professional football leagues, among others. However, some professional associations can also be classified as 501(c)(3) educational, charitable, and scientific organizations, 501(c)(4) social welfare organizations and local associations of employees or have no tax-exempt status under IRS regulation. Thus, it is hard to estimate exactly how many occupational mutual-benefit nonprofits are out there.
Those that receive some tax exemption do so because they offer public or member benefits. These may include professional advancements such as in medicine or engineering, improved safety standards or occupational norms, and licensing and professional certifications that ensure more widespread quality in services that benefit all (Haynes and Gazley 2011). Another characteristic of these organizations is clearly defined membership, whether individual or corporate, which is distinguished through annually paid membership dues (Spillman 2012). The focus of this paper -trade and professional associations -are organizations that serve the industry, profession, or any other special interest, which includes a wide range of sectors from agriculture and construction to sports and culture (ASAE: The Center for Association Leadership 2012). memberships that express their interest in public policy. Coleman (1985) acknowledged that associations engage in policy advocacy activities out of competition for resources and benefits for their members. Berry and Wilcox (2018) and Spillman and Brophy (2018) recognized that trade and professional associations engage in political activity to varying degrees that relate to their mission in its many forms including direct and indirect lobbying, grassroots advocacy, voter education, and even campaign activity. The fact that trade and professional associations can be incorporated under 501(c)(3), 501(c)(4), and 501(c)(6) statuses, implies some restrictions on the amount and type of political activities they can perform.
Political activity at its core is participation in public discourse. It connects diverse groups to governmental institutions. It provides venues for civic engagement and contributes to the public representation of individual interests. Multiple political actors develop alternative channels for deliberation in the policy process to raise their concerns and promote their agendas (LeRoux 2007). Various forms of political activity range from indirect/outsider forms, such as advocacy and launching grassroots activity, to more direct and engaging insider activities such as lobbying and forming of PACs in support or opposition to candidates running for office (MacIndoe 2014).

Conceptual Framework
Scholars commonly used institutional and organizational characteristics and resources to explain the choice of public policy strategies and tactics of charitable nonprofit organizations (MacIndoe 2014;Mosley 2010;Donaldson 2007). Child and Gronbjerg (2007) found such organizational properties as size, funding sources, areas of interest, partnerships with other organizations, professional leadership, and administrative experience to be associated with measures of political advocacy in charitable nonprofits. Other research points to budget and staff size (Miller-Steven and Gable 2012; Mosley 2011), revenue size, IRS status and affiliated foundations (Chand 2013), revenue source and subsector designation (NTEE) (Moulton and Eckerd 2012), and active membership and mission of the organization (Smith and Pekkanen 2012), as determinants of nonprofit political activity.
To understand the use of various advocacy and lobbying strategies of trade and professional associations we adopt a similar approach. Two major theories are used to understand the choices of policy advocacy tactics -institutional theory (Powell and DiMaggio 1991) and resource dependence theory (Pfeffer, Gerald, and Salancik 1978). We test these two theories, which encapsulate various regulatory arrangements and available organizational resources, against four different policy advocacy and lobbying tactics -public policy advocacy, lobbying, grassroots activity, and having an affiliated PAC.

Resource Dependence Theory
Resource dependence theory posits that organizational behavior in its essence is driven by resource maximization and cost minimization (Pfeffer, Gerald, and Salancik 1978). From a resource dependence standpoint, organizations seek resources and strive to efficiently use these resources to secure and maintain the receipt of future resources (Mosley 2010). Organizations tend to go after resources that ensure organizational success and provide a competitive advantage. Thus, the availability of resources determines organizational decisions and the capacity to influence the environment, such as when organizations advocate or lobby in the policy arena.
Financial and human resources have traditionally been cited as capacity-building resources that enable action (Mosley 2010;Child and Gronbjerg 2007). However, financial and human resources can have a variety of impacts on organizational autonomy to make independent decisions (Walston and Khaliq 2012). For example, reliance on any one single source of funding or lack of professional expertise on a board or staff may limit the level and the choices of policy advocacy participation of a nonprofit organization (Mellinger and Kolomer 2013). More broadly, access to resources empowers an organization to mobilize for a broad range of actions in the policy arena, whereas the lack of resources impedes these actions.

Revenues
Multiple studies found that reliance on a particular source of revenue, such as the government funding, affects nonprofit policy advocacy (Buffardi, Pekkanen, and Smith 2015;Bass et al. 2007;Smith and Pekkanen 2012). According to Mellinger and Kolomer (2013), a higher percentage of restricted funding was associated with higher levels of state legislative advocacy in human service nonprofits. Further, Mosley (2011) found that dependence on government funding increased the use of insider tactics such as lobbying but had no effect on indirect tactics such as public policy advocacy. Trade and professional associations differ from traditional charitable nonprofits in that they derive some part of their annual revenue from membership dues and generally rely less on government funding (Steinberg 2007). Although membership dues revenue falls under the unrestricted category of funds, allowing the organization to make independent decisions about its expenses, heavy reliance on membership dues revenue may also constrain an association's engagement in the policy process. Thus, we can assume a curvilinear relationship between the amount of revenue that an association derives from membership dues and its involvement in policy-related activities.
H1: Professional associations with most of their revenues derived from membership dues are less likely to be engaged in all forms of political activity: advocacy, lobbying, the capacity to launch grassroots, and affiliated PACs. Bai (2013) and Aggarwal, Evans, and Nanda (2012) confirmed the connection between larger nonprofit boards and positive organizational performance and program diversity. Further, Jaskyte (2013) found that board size had an influence on administrative and technological innovations in a network of Communities In Schools organizations. Ni and Zhan (2017) established the relationship between board size, program efficiency, and revenue growth. Finally, Donaldson (2007) noted that board support was positively associated with a broad range of nonprofit advocacy activities. Although the literature is largely inconsistent on the effects of board size on political activism of nonprofit organizations (Gazley, Chang, and Bingham 2010;Saxton, Kuo, and Ho 2011), given the positive association between nonprofit board size and a variety of performance metrics, we hypothesize that board size will be positively associated with more active facilitation of members' interests through advocacy and lobbying.

Board Size
H2: Professional associations with larger boards are more likely to be engaged in all forms of political activity: advocacy, lobbying, the capacity to launch grassroots, and affiliated PACs.

Institutional Theory
Institutional theory suggests that organizations' behaviors are shaped by legitimacy concerns in the context of their environments (Powell and DiMaggio 1991). In particular, the presence of regulatory standards and institutional arrangements can determine the actions of organizations in response to environmental forces. The latter may include local, state, and federal regulations, industry standards, and expectations of other stakeholders. Organizational behavior in the policy arena may be shaped by normative expectations stemming from nonprofit organizations' affiliations or legal norms such as those that accompany charitable status. Legal restrictions may limit legislative and politically charged nonprofit activity. On the other hand, separation of these activities under different organizational entities may allow organizations to engage in a variety of political activities more independently and with fewer restrictions.
The field of activity in which an association operates may also determine not only the expectations for political behavior but also the policy environment itself. For example, manufacturing industries may respond to a heavier regulatory burden by engaging in more direct forms of political involvement. Associations' stakeholders may also exert pressure on their organizations to ensure that their interests are represented in public policy choices. Institutional theory explains how an organization's environment influences its roles and behaviors (Mellinger and Kolomer 2013).

Foundations
Associations may establish affiliated foundations to diversify their assets, obtain tax-deduction status for incoming donations, or expand visibility to the public and the media (Sidman 2010). Martin (2004) asserted that foundations provide a financial outlet to channel philanthropic giving into grassroots projects. Other common reasons to establish a foundation include funding research, expanding existing services to non-members, or establishing new programs that benefit the general public. Foundations promote broader benefits that associations can provide. A separate foundation may distinguish charitable expenses from lobbying to comply with IRS regulations (Moore and McAlexander 2005). Therefore, an association may feel more comfortable performing politically motivated activities, knowing it will not adversely affect its charitable activities. At the national level, Chand (2013) found that organizations with multiple tax-exempt statuses spent more on lobbying than those that only had one tax-exempt status, although this did not apply to organizations with 501(c)(3) and 501(c)(4) designations compared to 501(c)(3) organizations.
H3: Professional associations that have an affiliated foundation are more likely to be engaged in the lobbying form of political activity.

IRS Status: 501(c)(3) Vs. 501(c)(6)
Political activity, whereby an association seeks to represent the interests of its stakeholders in the policy arena, has always been considered quite sensitive to certain organizational attributes, such as IRS tax-exempt status . All nonprofits exist to fulfill their mission. These missions to a large extent determine their nonprofit tax-exempt status. Thus, there are differences in types of services provided, primary beneficiaries, and outcomes that benefit the public, across different tax-exempt nonprofits. Tax-exemption also determines the extent of public and community support and the range of associations' political engagement. Ultimately, tax-exempt status may affect how much influence these nonprofits exert in the political sphere. For example, a 501(c)(3) status may deter professional associations from getting involved with controversial political activities (Bass et al. 2007;. In particular, 501(c)(3) status imposes legal restrictions on lobbying and prohibits political campaign involvement.
H4: Professional associations that have 501(c)(3) status are less likely to be engaged in the following forms of political activity: lobbying and affiliated PACs.

Business Classification
The nonprofit literature suggests that mission drives nonprofit policy advocacy activities (MacIndoe 2014; Child and Gronbjerg 2007;Mosley 2010). Mission has been operationalized in other studies as a nonprofit's National Taxonomy of Exempt Entities (NTEE) code (Child and Gronbjerg 2007), nonprofit subsector designation (Mosley 2010), or specialist/generalist distinction (MacIndoe 2014). Spillman and Brophy (2018) and Hager (2014) contended that associations from different specialization fields have different professional cultures, and organizational and management standards. Technical professions such as engineering and manufacturing, as well as economics-related professions have a history of being the gatekeepers of their professions and industries more so than cultural and educational associations (Halliday 1987). Since gatekeeping may assume extensive filtering of incoming and outgoing information, one can expect that associations closely tied to areas of profitable business activity are more likely to control the information field and policy space that affects them. However, Buffardi, Pekkanen, and Smith (2015) did not find business classification to be related to a choice of venue strategies for advocacy activities.
H5: Professional associations with the engineering and manufacturing business classification are more likely to be engaged in the following forms of political activity: lobbying, capacity to launch grassroots, and affiliated PACs.

Type of Membership
Other stakeholder characteristics such as membership size and type were used to control for their effects on organizational political activity (Chand 2013). Membership type category has not been regularly used in the literature on nonprofit political involvement since occupational nonprofits were not commonly studied subjects. Yet, in the practice of professional associations the type of membership -individual or corporate -accounts for variance in membership values and benefits, primary associational activities, and organizational priorities (Rominiecki 2016). Corporate memberships may be more eager to see enacted policies or passed legislation that benefits their industries, compared to individual memberships. Thus, we can expect that corporate membership associations will be more engaged in direct political strategies that benefit their members.
H6: Professional associations that have organizations as their members are more likely to be engaged in the following forms of political activity: lobbying and affiliated PAC.

Methodology Data
This research uses the data provided by the American Society of Association Executives (ASAE), which was collected from the ASAE Foundation's Policies and Procedures survey in 2012. The ASAE represents 7,300 trade and business associations and individual professional membership societies from a wide range of economic sectors and professions. The ASAE Foundation, a charitable subsidiary, serves as a knowledge resource for their member associations in the areas of associational management. The Policies and Procedures survey provides a comprehensive look at professional associations and their activities in the U.S. The survey conducted in 2011-12 provides a range of information in different areas of management -finances, governance, human resources, membership structure and benefits, marketing, technology, education and development, meetings, government affairs, etc. The survey captures the ways that associations conduct their business and how they operate.
The population includes both individual membership associations, as well as trade groups that serve primarily companies and other business organizations. A sample of associations participating in the survey contains 30 different interest areas including scientific, banking, retail, healthcare, and culture organizations among others. The initial sample size of 750 observations had been reduced to 104 observations due to a significant amount of missing and incomplete responses to the political engagement questions. The questions on political engagement were located towards the end of the survey, which may have accounted for responders' survey fatigue and resulted in a significant percentage of missing answers (Olson 2014). Olson (2014) also suggested that survey design, mode of delivery, survey length, and the number of survey contacts can contribute to a low response rate. Finally, the nature of the questions, which may be considered sensitive, could have contributed to the low response rate (Porter and Ecklund 2012). We ran an independent-samples t-test to ensure that the smaller sample is representative of the original sample in its characteristics (Zimmerman 1997). The results of the t-test did not reveal statistically significant variances for the two groups. Thus, it is safe to use the smaller sample and generalize the findings to the larger sample of trade and professional associations.
All data came from the survey. Questions that were used included number of full-time employees, membership and budget size, proportion of revenues derived from membership dues, board size, presence of an affiliated foundation, professional classification, type of membership, and questions about political and public policy engagement strategies, namely advocacy, lobbying, grassroots efforts, and presence of affiliated PACs.

Dependent Variables
Public Policy Advocacy IRS defines nonprofit advocacy as a range of public policy engagement activities that fall outside of lobbying, which is more precisely defined by the federal regulations (IRS 2018). The Council on Foundations defines advocacy as encompassing all types of involvement with policy decision-makers that do not meet the definition of lobbying under the IRS. Other professional organizations and advocacy groups, for example Board Source or the Alliance for Justice, define advocacy as actions in relation to public causes, policy changes, and government decisions that express, communicate, educate, defend, recommend, and support specific public issues, causes, or opinions (Board Source 2016; Alliance for Justice 2016). Occupational associations may seek to advance their professions through advocacy. For example, the American Public Health Association influenced the creation of the children's health insurance program (Donaldson 2007). The public policy advocacy variable in this study was measured by the question: 'Does your organization engage in public policy or advocacy activities?'

Lobbying
According to the IRS (2018), "[d]irect lobbying refers to attempts to influence a legislative body through communication with a member or employee of a legislative body, or with a government official who participates in formulating legislation." Professional associations can provide input and feedback to Congress or state legislators on specific pieces of legislation affecting their practices. For example, the American Nurses Association (ANA) employs lobbyists to present and promote the views and concerns of its members at the congressional level to have favorable bills passed in the legislative process (Mason, Leavitt, and Chaffee 2013). The lobbying variable was measured by the following question: 'Does your organization engage in lobbying activities?'

Grassroots Political Activity
Grassroots political activity is known as indirect lobbying. It refers to engagement in issue advocacy or an attempt to influence legislation by mobilizing membership or the broader public concerning specific legislation, and encouraging the public to act on it (IRS 2018). Membership associations can engage in grassroots political activity, for example, through open letters with a call to action and petitions. They can also engage their members and encourage them to reach out to non-members to take action. An example of grassroots political activity is the way the ANA established the Nurses Strategic Action Team, a grassroots program, in response to emerging healthcare reform issues (Cohen et al. 1996). The question 'Does your organization have the capability and authority to launch grassroots political activity?' measured the capacity for grassroots activity variable.

Political Action Committee (PAC)
Some professional organizations utilize PACs to garner political influence. According to Grossman and Helpman (2001), professional societies representing the healthcare industry use members' taxable contributions to create affiliated PACs to support preferred political candidates. The largest proportion of contributions goes to candidates with the greatest ability to address the organization's interests regardless of their political parties (Ibid.). For example, the American Dietetic Association (ADA) created the American Dietetic Association Political Action Committee (ADAPAC) in 1981 to support federal office candidates that favor ADA policies (Maillet, Potter, and Heller 2002). The PAC variable is measured by the question 'Does your organization operate a political action committee (PAC)?' Each of these questions reflects a type of political engagement strategy. Each question has a binary Yes or No choice. Yes is coded as 1 to indicate that an association performs a distinct form of political activity or can get involved politically on behalf of its members. Table 1 shows the descriptive statistics and operationalization of all variables. Which of the following include the range of your organization's budget? (Choose only one) Less than $100,000 $100,000 but less than $500,000 $500,000 but less than $1,000,000 $1,000,000 but less than $5,000,000 $5,000,000 but less than $10,000,000 $10,000,000 but less than $25,000,000 $25,000,000 but less than $50,000,000 $50,000,000 but less than $100,000,000 Over $100,000,000

Independent Variables
The following variables measure organizational resources and institutional characteristics. The revenues derived from membership dues variable is measured by the question 'What percent of your organization's total revenue was derived from member dues in the last fiscal year?' The board size variable is captured using the 'Number of voting board members' question. The affiliated foundation variable is measured by the survey question 'Does your organization have a foundation?' with a binary answer option, where Yes was coded as 1 and No as 0. The IRS tax-exempt status variable is measured by the question 'What is the primary IRS tax status of your organization (the primary or parent organization)?' with answer options of 501(c)(6), 501(c) (3), 501(c) (4), other, and not tax-exempt. These responses were consolidated into three groups: 501(c)(6) coded as 0, 501(c)(3) coded as 1, and all the other categories coded as 2, with the first category being the baseline for comparison. The business classification variable was derived from the following question 'What is the primary interest/subject area of your organization?' The survey provides 31 options, which were consolidated into four categories: engineering/manufacturing (coded as 0), administrative/business (coded as 1), healthcare/medical (coded as 2), and other category (coded as 3), which included culture and arts, fraternities, sports, and education associations. The engineering/manufacturing category is used as a baseline for comparison. Finally, the membership type variable was measured using the question 'What kind of members belong to your organization?' Answer 'Individuals/Professionals' was coded as 0 and 'Companies/Institutions' answer was coded as 1.
We control for organizational size, using three dimensions: membership size, budget size, and staff size. Organizational size matters when considering the cost of political activism in charitable nonprofits (Child and Gronbjerg 2007;Mosley 2010;Fyall and Allard 2017). The size of the associations is measured by three questions. Membership size is gauged using the following question 'How many members are there in your organization (include all classes of membership)?' The staff variable is measured using the question 'How many full-time equivalents (FTEs) are employed in your organization (all locations combined but exclude subsidiary or foundation employees, if any)?'. The responses are reported in actual numbers. The budget variable is measured by a categorical question: 'Which of the following includes the range of your organization' budget?' with nine categories as answer options. Responses were converted to natural logarithms for three variables: membership size, staff size, and board size, to normalize the data. Finally, a square term of revenue derived from dues was added to all four regressions to account for a nonlinear relationship.

Analysis
Reduction in sample size has not significantly changed the composition across organizational characteristics. The sample of 104 associations, on average, has 13,000 members, 25 full-time employees, and a budget that ranges between one and five million dollars. The average size of the voting board is about 24 members. Associations in the sample derive about 39 % of their revenues from membership dues. The sample consists of 56 % associations with an affiliated foundation. About 64 % of associations have 501(c)(6) status, 26 % have 501(c)(3) status, with the rest falling in all other categories. Associations in the sample are distributed almost evenly among the four categories of business classifications, with nearly a quarter of organizations in each category. About 59 % of associations have organizational members and 41 % are individual member associations. Finally, about a third of the sampled organizations indicated that they engage in public policy advocacy activities or have the capability and authority to launch grassroots political activity, and less than half indicated that they engage in lobbying or operate a PAC. On average, surveyed associations engaged in one tactic out of four, 2 % engaged in two tactics, 6 % engaged in three tactics, and 8 % engaged in all four tactics, with 39 % indicating no political engagement across the four tactics. Table 2 presents Pearson's correlations for each pair of variables. Several paired correlations were significant with only one being moderate in strength. The only paired correlation that is both moderate and significant is between public policy advocacy and lobbying, which is not surprising since lobbying can be part of an advocacy strategy. Since none of the paired variables displayed high correlations no data manipulation was performed.

1.00
Grass-roots We use multiple logistic regression analysis since all four dependent variables are binary (Table 3). Logistic regression produces probability estimates for logarithmically transformed outcome variables, which account for nonlinear relationships between a categorical dependent variable and a set of independent variables (Baldauf and Santos Silva 2012). We found support of both resource dependence and institutional theories explaining the choices of political activities in trade and professional associations. Sample size = 104; p < 0.000, * sig. at 0.10, ** sig. at 0.05, ***sig. at 0.01 Our first regression explained about 33 % of the variation in public policy advocacy activity of trade and professional associations. Overall, organizations with larger staff and boards, those that have affiliated foundations, and those that operate in the healthcare field, are more likely to perform public policy advocacy activities. Specifically, larger boards increase the chances of an association being involved in public policy activity, which supports resource dependence hypothesis 2. Institutional theory hypothesis 3 is also supported since the presence of an affiliated foundation is positively associated with the probability of public policy activity. Further, associations that operate in a healthcare or medical field are more likely to perform public policy activities compared to engineering and manufacturing associations, a relationship which was not hypothesized. Finally, associations with larger staff are more likely to engage in public policy advocacy, which is consistent with the previous literature on the effects of organizational size on policy advocacy activity (Child and Gronbjerg 2007;Mosley 2010;Fyall and Allard 2017).
The second regression explained close to 20 % of lobbying activity. Notably, larger membership and budget size decrease the chance that associations lobby, whereas larger staff increases it. Larger boards increase the likelihood that an association lobbies, which supports resource dependence hypothesis 2. The presence of an affiliated foundation also increases the probability of lobbying activity, which supports institutional theory hypothesis 3. Finally, the squared term of the portion of revenue derived from membership dues is negative and significant, suggesting that there is an inverse curvilinear relationship between revenues that come from dues and associational lobbying, which supports the resource dependence assumption that overreliance on a single source of revenue may limit the overall engagement of associations in political activities. Hypothesis 1 is partially supported.
Three variables reached statistical significance in the third regression model and together explain about 29 % of the variation in capacity to launch grassroots activity. Overall, associations that have affiliated foundations, operate in the medical/healthcare field and have corporate members are more likely to have the capacity to launch grassroots activity, which supports hypothesis 3, but rejects hypotheses 5 and 6. Neither of the orga-nizational size measures nor the size of the board or IRS status had any effect on the capability and authority to launch grassroots activity.
The fourth regression explains about 33 % of the variance in having an affiliated PAC, with IRS status and business classification variables reaching statistical significance. Overall, associations that operate in the administrative/business sphere but not in culture, environment, or social and civic affairs, and those that are not 501(c)(3)s, are more likely to have affiliated PACs. Specifically, the odds of a professional association having an affiliated PAC are lower for associations with a 501(c)(3) IRS status compared to associations with 501(c)(6) status, supporting institutional theory hypothesis 4. In addition, the odds of having an affiliated PAC are higher for the administrative/business classification but lower, for the 'Other' category (which includes various culture, environment, social and civic affairs associations) compared to engineering/manufacturing associations. Thus, hypothesis 5 was both partially rejected and partially supported.

Discussion
Political engagement strategies draw on different organizational resources and can be constrained by the regulatory and industry environment. Our study reveals the relationship between resources and characteristics on one hand and the choice of political strategies on the other hand.
Our empirical examination bears evidence that both resource dependence and institutional factors are associated with the variance in policy advocacy tactics. The explanatory power of both theories varied across different forms of political engagement. In particular, we found evidence that supports resource dependence theory hypotheses with advocacy and lobbying tactics, but not with the capacity to launch grassroots activity or having an affiliated PAC. Larger boards are associated with increased advocacy and lobbying activity. On the other hand, overreliance on a single financial source, such as membership dues decreases the chances of lobbying activity.
Resource dependence theory highlighted the boundary-spanning capability of association boards. Board members' connections and expertise can help expand an organization's influence on its environment (Miller-Millesen 2003). Occupational interest groups with larger boards can be more visible and have more prominent members of their professions, and possess more cumulative expertise and knowledge, which can translate into more active advocacy and lobbying efforts.
Reliance on a single source of funding such as membership dues had a negative effect on lobbying activity. Previous studies found that government funding was associated with an increase in lobbying (Mosley 2011) and policy advocacy (MacIndoe 2014). Here we see the reverse relationship. As association relies more on membership dues for its budget it is less likely to engage in lobbying activity. Associations that rely heavily on membership dues may spend more on direct member benefits rather than external political engagement activities since their financial sustainability depends more on annual membership renewals.
Institutional theory factors such as affiliated foundation, IRS status, business classifications, and type of membership relate to various forms of political activity, both in expected and unexpected ways. Similar to Chand's Chand (2013) findings, the presence of an affiliated foundation facilitates all tactics except for having a PAC. Presence of an affiliated foundation can be a result of organizational growth and diversification. Affiliated foundations allow associations to expand programming and capture resources available only to charitable nonprofits. Organizational affiliates help to maintain autonomy and legal independence over different activities. Separate foundations can isolate charitable and political activities to allow more flexibility for associations to be more active in the public policy arena. Future research should examine how trade and professional associations create 501(c)(3) and 501(c)(6) affiliates and the purposes they serve.
As expected, IRS status is negatively associated with politically charged PAC tactics (Bass et al. 2007;. Creating PACs to endorse or oppose candidates for office is forbidden for 501(c)(3) organizations. Regardless of whether associations are created to advance professions or to serve social public missions, they operate in the same legal environment. Unlike 501(c)(6) associations, those with charitable tax-exempt status must abide by more stringent rules of political activity or potentially invite close IRS and public scrutiny.
Healthcare and medical associations are more likely to perform advocacy and have the capability and authority to organize and launch grassroots activity than engineering and manufacturing associations. The literature provides many examples of healthcare and medical professional associations being active in the political arena (Mahlin 2010;Maillet, Potter, and Heller 2002;Mason, Leavitt, and Chaffee 2013), suggesting a longstanding tradition of these organizations being an active voice in government decision making. This study provides further evidence that professional associations in the healthcare field engage more in policy advocacy.
Further, administrative and business associations such as those in commercial, banking, retail, and real estate industries are more likely to have affiliated PACs compared to engineering and manufacturing associations. Although both categories represent large industries, financial and banking industries are overseen more broadly by elected officials than narrow regulatory agencies and therefore better positioned to influence public policy by supporting candidates for office. Finally, associations in a broad spectrum of environmental, tourism, and sports industries, were less likely to have an affiliated PAC than engineering and manufacturing associations. Most of these associations may fall under the charitable category, hence be banned from engaging in partisan political activity.
Finally, although we expected corporate membership associations to be more active through direct forms of political action, we found the opposite. They were more likely to have the capacity to launch grassroots activity compared to individual membership associations. This can be attributed to their broader reach through organizational members. They have the capacity to appeal to employees of member organizations and by extension to wider public audiences. Therefore, these associations are better positioned to tap into the established resources of their corporate members to engage in public mobilization tactics such as grassroots.
Not surprisingly, political activity was a function of organizational size. Associations with bigger staff are more likely to engage in lobbying, suggesting that some of their staff were hired as professional lobbyists. Contrary to many other studies, however (Child and Gronbjerg 2007;Mosley 2010;Fyall and Allard 2017), budget and membership size have a negative relationship with lobbying tactics. It is possible that these associations primarily focus on engaging their members and spending more on membership activities. Considering wider access to resources, bigger associations may see less value in lobbying when compared to their other priorities. More research is needed to understand whether the amount of resources affects the choices of political strategies of trade and professional associations.
This research focused on characteristics described by two commonly cited theories that have been found to explain organizational behavior and strategies of charitable nonprofits (Moulton and Eckerd 2012;Mellinger and Kolomer 2013;Mosley 2010). Yet, other organizational and environmental characteristics could influence the political involvement of trade and professional nonprofit associations. Our findings also did not reveal the extent to which political tactics were used. Thus, it remains an open question of how much trade and professional associations engage in the policy arena. Associations can leverage their boards for advocacy and lobbying efforts and their corporate members to increase capacity to launch grassroots activities. However, just like their charitable counterparts, they are bound by regulatory limitations of political activity.

Conclusion
Trade and occupational associations that serves the interests of industries and professions, can be quite prominent players in the policy arena where they seek to represent the interests of their stakeholders. Learning about what defines and shapes their political involvement allows for a better understanding of the inner workings of this understudied subsector.
Overall, in line with the literature on charitable public policy advocacy, we found several connections between resource dependencies, institutional arrangements, and political tactics, suggesting that many factors affect the choice of political tactics among trade and professional associations. Yet, it is also clear that these factors do not explain everything. Future studies should investigate what other organizational and environmental characteristics may affect public policy advocacy strategy choices.
Our findings have important implications for nonprofit practitioners. Organizational leaders interested in increasing their advocacy tactics may learn from healthcare associations about how to step up their advocacy efforts, and from business and real estate associations on why they choose PACs to advance their interests. Further, associations can employ their corporate member connections to launch grassroots campaigns and use their boards to expand their advocacy and lobbying efforts. Finally, affiliated foundations may insulate parent associations to perform a variety of political activities. The importance of organizational resources and recognition of potential dependencies should not be overlooked when considering political engagement strategies.
These findings should be generalized with caution, however. Our limited sample of associations does not allow for broad generalizations of our conclusions to the rest of the trade and professional associations population, especially those that are not ASAE members. Further, the cross-sectional nature of the data only points at correlations among variables; future studies should attempt to collect longitudinal data to establish causations.
Nonetheless, our analysis provides an important step toward identifying what determines the political engagement strategy choices of trade and professional associations in the U.S. Our study contributes to the literature by providing evidence that resource dependence and institutional theories help explain the participation of professional associations in policy making activities. Future studies may explore other theories that explain political strategy choices of professional associations.
Professional associations are engaged in developing and shaping policies that directly or indirectly affect their stakeholders using various forms of activity including direct lobbying or issue advocacy, mobilizing members and other stakeholders or endorsing or opposing political candidates. Resources and environment both shape these organizations' decisions in the political arena and determine their ways of influencing public policies. This study has shed light on how organizational and environmental determinants of occupational interest groups shape their choices of political involvement strategies. Although this study does not encompass the full scope and range of all occupational nonprofits, it points the way to a fuller exploration of this important area of nonprofit policy research.