Abstract
This article aims to contribute to the long-standing discussion about nonprofit organizations’ (NPOs) dependence on public funding and its consequences on their advocacy role in modern societies. Drawing on resource dependence theory and data from a quantitative survey, the study investigates the impact of public funding and its extent on nonprofit engagement in advocacy. Traditionally, scholars have cautioned that NPOs reliant on public sources will hesitate to pursue political objectives and to engage in advocacy work. Yet, empirical findings are strikingly inconsistent. One of the reasons for these ambiguous findings may be the way advocacy is measured. To address this issue, we apply two different approaches to evaluate NPO engagement. Both sets of findings from our multivariate analyses of Austrian NPOs suggest that public funding does not have a negative impact on advocacy.
Nonprofits’ autonomy from the state is a core feature of these organizations. It is crucial for them to accomplish their mission independently and to follow their “civic virtues” (Chaves, Stephens, & Galaskiewicz, 2004, p. 292; Salamon & Sokolowski, 2004, p. 9). Especially in carrying out the essential function of engaging in advocacy and being a voice for citizens, nonprofits’ independence from the government has been considered pivotal (Frumkin, 2002; James & Rose-Ackerman, 1986; Kendall, 2003; Kramer, 1981; Salamon, Hems, & Chinnock, 2000). This is also true for financial independence (Salamon, 1995; Smith, 2004).
Nonetheless, today many nonprofits obtain large proportions of their revenues from public agencies. On average, some 31% of the total income of nonprofits in the United States and about 45% of those of European organizations stems from the state (Salamon & Sokolowski, 2004). This discrepancy between claim and reality has been the cause of lengthy discussion (Chikoto, 2010; Fetter, 1901; Fleischer, 1914; cited in Smith, 1998). Scholars and practitioners fear that nonprofit organizations (NPOs) will feel pressure on their political activities if they accept public funds—and, as a consequence, will reduce or even abandon advocacy.
Despite these long-standing debates, empirical support has been rare. There are many articles that focus on the (negative) influence of public funding on nonprofit engagement in advocacy, but only about 20 of these examine this relationship empirically—and almost all use data from the United States. The findings, moreover, are inconclusive, providing no unanimous evidence that public funding impedes advocacy. These inconsistent findings may be attributed to different measures of advocacy (Onyx, Dalton, Melville, Casey, & Banks, 2008).
This study investigates the relationship between public funding and nonprofit engagement in advocacy to stimulate further debate on this topic and to generate an empirical analysis from Central Europe. Drawing on previous literature and on resource dependence theory and organizational capacity building, we develop four hypotheses, which are tested using data from a 2007 survey of 250 Austrian NPOs. To establish a connection with earlier studies on the issue and to address the problem of contradictory results from previous research, we apply two different approaches to measuring nonprofit engagement in advocacy. The first metric is based on the resources devoted to advocacy by the NPO. The second depends on the advocacy practices in which it engages. We test each hypothesis using both measures. By doing so, we aim to disentangle contradictory results and to encourage discussion on this topic because these are the first empirical results for Central Europe.
Literature Review: Consequences of Public Funding on Nonprofit Engagement in Advocacy
Advocacy is commonly referred to as a nonprofit’s crucial civic function. According to Jenkins (2006), the overall aim of advocacy is to “correct imbalanced political representation by ensuring that a broader set of interests are voiced” (p. 308). Multiple activities intend to achieve this goal. They include talks with governments, the monitoring of government programs, legal advocacy in the courts, calls for boycotts and demonstrations, and the drawing up of petition letters, among others (Boris & Mosher-Williams, 1998; Jenkins, 2006; McCarthy & Castelli, 2001; Onyx et al., 2008). All are intended to do one of three things: To influence government decisions; to shape the policies of private institutions and corporations; or to encourage community and political participation (Jenkins, 2006).
With regard to the impact of public funding on nonprofit engagement in advocacy, two main lines of argument are discussed in the literature. The first, generally widespread view draws on resource dependence theory (e.g., Chaves et al., 2004; Moulton & Eckerd, 2012; Schmid, Bar, & Nirel, 2008). It implies that NPOs that receive public funding will eventually reduce their advocacy activities or even refrain from them altogether for fear of losing their funding. The second posits precisely the reverse effect, that is, the nonprofits who receive public funding will increase their engagement in advocacy (e.g., Chaves et al., 2004; Leroux & Goerdel, 2009; Mosley, 2011). On one hand, this proposition is based on the idea that such NPOs will have more resources with which to advocate (e.g., Bass, Arons, Guinane, Carter, & Rees, 2007; Mosley, 2011). On the other hand, it assumes that they will use their access to the government, which they have gained through obtaining funding, for advocacy purposes (e.g., Bass et al., 2007; Kelleher & Yackee, 2009).
The empirical studies that have investigated these two opposing assumptions concerning the consequences of public funding on nonprofit advocacy activities show a slightly different picture (see Table 1); their results are mixed. Although most of these analyses suggest that public funding will increase NPO engagement in advocacy, several studies come to the opposite conclusion. Furthermore, some do not find any association at all, whereas others reveal inconclusive findings. Possible explanations for these divergent results can be found in the data used (cross-section vs. panel data), the types of nonprofits analyzed, country-specific aspects, the models used to analyze the data (i.e., some studies are concerned with the probability of advocacy, and others with its extent or the range of advocacy activities); and, finally, the method used to measure advocacy.
Empirical Evidence on the Impact of Public Funding on Advocacy.
Public funding served as a control variable in this study.
Key Findings From Studies on Advocacy
The majority of studies we analyzed find government funding to be positively associated with advocacy. Most of these were based on regression analyses and cross-sectional data for nonprofits in the United States (see Table 1). To start with, O’Regan and Oster (2002) investigated whether or not NPOs engaged in advocacy by examining whether board members had undertaken any government-related advocacy measures. Similarly, Hwang and Suárez (2008) examined whether an organization “advocate[d] on behalf of its clients or cause through public information campaigns, legislative campaigns, or other activities” (p. 13)
Although using a similar method, Salamon (2002), Leroux and Goerdel (2009), Mosley (2011), and Moulton and Eckerd (2012) were not concerned with whether or not an NPO engaged in advocacy, but to what extent it did so. The approach used to measure advocacy activity in each case was, however, quite different. Salamon’s study, for example, analyzed advocacy activities as a percentage of expenditure, whereas Leroux and Goerdel measured the frequency of political-representation and political-mobilization activities. Taking yet another approach, Mosley asked whether nonprofits use indirect and insider advocacy tactics. Finally, Moulton and Eckerd used an index of three different advocacy activities.
The next three studies, though still based in the United States, drew on different data sources but also came to the conclusion that public funding relates positively to advocacy activity. Bass et al. (2007) examined the correlation between nonprofits’ share of public funding and the frequency of their advocacy activities. Second, Faulk (2008) applied regression analyses to panel data, measuring the extent of advocacy by NPOs’ online-media presence, that is, by the number of press releases during a given period of time. Last but not the least, Mosley (2012) demonstrated in her qualitative study that public funding motivates nonprofits to advocate the promotion of their organization rather than to effect substantive policy change.
Conversely, some studies did show a negative correlation between government funding and NPO engagement in advocacy. Sandfort’s (2005) case study of a U.S. shelter for battered women indicated that an increase in public funding went hand in hand with an increase in individual services and at the expense of political mobilization and empowerment. Likewise, Schmid et al. (2008) suggested in their cross-sectional analysis of Israeli nonprofit engagement across six different activities that public funding from local authorities was negatively associated with advocacy.
However, although all the studies mentioned above indicated either a positive or a negative correlation between revenues from public sources and advocacy, the next group of research found no relationship in either direction. All three studies drew on cross-sectional data from the United States, but where Leech (2006) used an additive index of a broad number of activities, Suárez (2009) examined the number of announcements published on NPOs’ web pages to determine whether the proportion of public funding had an impact on the nonprofit’s political activities as well as on its promotion of civic engagement. Taking a somewhat different approach, Garrow and Hasenfeld (2012) assessed whether nonprofits devoted any resources to advocacy.
The final group of studies presented mixed findings on the connection between public funding and advocacy. Chaves et al. (2004) analyzed cross-sectional as well as panel data from the United States. They measured whether or not nonprofits engaged in eight different kinds of advocacy activities and ran separate regressions on these activities. In most cases, no impact could be found; in some, a positive impact was apparent (Chaves et al., 2004). Likewise, Guo and Saxton’s (2010) regression analyses were not able to reveal any clear relationships. They found that public funding undermined nonprofits’ efforts when measured by the frequency of different activities (e.g., public education, media advocacy, grassroots lobbying). However, when measuring “the proportion of (an) organization’s resources (time, money, etc.) . . . spent” on advocacy, they found no relationship (Guo & Saxton, 2010, p. 13). Similarly, Onyx et al.’s (2008) case study on an NPO in Australia reported that government finance sometimes encouraged but sometimes suppressed political advocacy. The same ambiguity can be seen in the results of Nicholson-Crotty. In her first article (Nicholson-Crotty, 2007), she showed that public funding in the form of grants and contracts could produce either a positive impact or none at all. In her second study (Nicholson-Crotty, 2009), however, she found that public funding had a positive impact on advocacy but a negative one on grassroots lobbying, which she referred to as “aggressive advocacy.”
Implications of Literature Review
Our literature review has shown that the majority of previous empirical studies indicated that public funding has a positive impact on nonprofit advocacy engagement. Few articles observed a negative impact, three found no relationship at all, and five came to inconclusive conclusions.
This body of previous research raises several questions. First, because all existing empirical studies—with the exception of one from Australia and one from Israel—relate to nonprofits in the United States, it is not clear whether their findings are applicable to other countries, especially to countries belonging to other nonprofit regimes (Salamon & Anheier, 1998). Second, we wonder to what extent different ways of measuring advocacy contribute to differences in findings. Some studies examine whether NPOs engage in advocacy at all, whereas others consider the extent, scope, type, and intensity of advocacy. Furthermore, the metrics used may be based either directly on advocacy activities or on the resources devoted to these.
Authors who used advocacy activities as a measure asked nonprofits to identify the activities they had carried out from a given range of activities (e.g., mobilizing citizens to sign a petition, joining boycotts or demonstrations, or disseminating information) during the period concerned. Some researchers asked for responses to be given as a dichotomous variable. In some cases, these dichotomous variables were then added together and computed to create an additive index for the overall extent of advocacy. Others used Likert-type scales to measure the extent of individual activities (Bass et al., 2007; Leech, 2006; Moulton & Eckerd, 2012; Schmid et al., 2008). Another group measured advocacy in a nonreactive way by examining nonprofits’ media presence (Faulk, 2008; Hwang & Suárez, 2008).
The authors who concentrated on resources recorded either the financial or the personnel resources devoted to advocacy in each organization. Salamon (2002) looked at the proportion of expenditure used for advocacy activities, whereas Guo and Saxton (2010) studied the share of total NPO resources (time, money, etc.) devoted to them.
An examination of the various studies does not indicate an obvious relationship between measurement methods and findings about the impact of public funding on advocacy in most cases. However, Guo and Saxon’s (2010) findings contain evidence to suggest measurement-induced differences. These authors measured advocacy both in terms of resources and activities and found that public funding had no effect on the resources devoted to advocacy but did reduce the scope of nonprofits’ advocacy activities. Nicholson-Crotty’s (2009) findings suggest a similar conclusion. She found that although public funding had no impact on advocacy in general, it had a negative one on certain forms of advocacy, in this case, “aggressive” advocacy. From that, we might assume that government funding does not reduce the resources that nonprofits devote to advocacy—there is no evidence of that in the studies reviewed—but may negatively affect what these resources are used for.
We therefore directly address the following questions. First, we ask how public funding affects advocacy activities, how it affects the resources devoted to them, and whether its impact on these two measures can differ. Moreover, our article makes an important contribution by complementing existing research on the United States with an examination of NPOs in Central Europe, specifically in Austria, where the government and the nonprofit sector are closely intertwined.
Theory and Hypotheses
Most of the theoretical arguments that assume that nonprofits’ dependence on public funding reduces their engagement in advocacy are based on resource dependence theory. By contrast, arguments that support the opposite assumption derive from the concept of organizational capacity building.
According to resource dependence theory, organizations are not autonomous entities that achieve the desired ends at their own discretion, but bodies that rely on resources from their environment to secure their survival (Pfeffer & Salancik, 2003). To maintain access to these resources, they are prepared to adapt to the perceived needs and expectations of important resource providers, that is, the organizations they depend on financially (Oliver, 1991; Pfeffer & Salancik, 2003). As a result, nonprofits who receive public funding can expect or may have already experienced a cut in financial support in the event of any undesired, oppositional advocacy (Chaves et al., 2004; Evans, Richmond, & Shields, 2005). Furthermore, nonprofits may also refrain from advocacy for fear of other negative sanctions (e.g., being excluded from meetings, loss of access to the legislative process, and reduction in clients assigned; Chaves et al., 2004; Ryan, 1999). Moreover, some NPOs may believe that they are prohibited from engaging in advocacy-related activities while receiving public funds (Leech, 2006). In fact, some must sign contracts that explicitly restrict such activities. 1
The degree of perceived dependence is determined by the extent of public support. The more important the resources provided are, the more dependent the organization becomes. However, dependence decreases as the number of alternative resource providers increases (Pfeffer & Salancik, 2003).
Nonprofits rely on a variety of resources to achieve their goals. Financial resources, volunteers, members, information, recognition, public support, and access to policy-making processes are only a few examples (Johansson, 2003; Saidel, 1991). Furthermore, they depend almost entirely on their external environment for all these resources. It is arguable, however, that the resource exchange with the public sector seems to be of particular relevance. Public institutions depend on NPOs’ competence and their capacities to provide social services (Chaves et al., 2004; Saidel, 1991). In return, the public sector is the main provider of financial resources for nonprofits, at least in Central Europe. There, on average, public funding makes up about 57% of the sector’s income; in Austria, the figure is 50% (Salamon & Sokolowski, 2004).
Accordingly, resource dependence theory describes a highly relevant mechanism by which government funding might suppress nonprofit engagement in advocacy. This is true for work with the disabled in Austria. In that field, smaller organizations that are much less dependent on public funding are considerably more active in fighting for inclusion and against barriers than the large organizations, which benefit more substantially from public funds. We therefore suggest the following hypotheses:
The opposite effect, namely that public funding enhances nonprofit engagement in advocacy, is based on two very different arguments. First, public funding enables such organizations to build capacity, to professionalize and thus enhance their ability to engage in advocacy. As a consequence, the advocacy activities themselves as well as the resources dedicated to advocacy can be increased (Bass et al., 2007).
Second, because nonprofits that receive public funding are in regular contact with public organizations, they are more deeply involved in political issues than those that do not. In the event that public authorities need the expertise or opinion of nonprofits on a certain bill, they will tend to contact those who they know from the granting of public funds. Salamon (2002) labeled this explanation the paradigm of partnership (p. 8). This is particularly the case in Austria and in other countries in Central Europe, where NPOs are frequently invited to review draft bills and are thus embedded in political decision-making. In contrast to the research presented above, other empirical evidence shows that nonprofits engaging in advocacy receive a larger share of their revenue from the government than those who do not engage in advocacy (Bass et al., 2007). In Austria, some of the biggest multipurpose nonprofits who receive significant amounts of public money in specific fields, such as care for the elderly, are also the most active campaigners against public policies in other fields, for example, poverty reduction. Examples include the Caritas and the Red Cross, who have established themselves as major providers of social services and have utilized their growth to establish professional advocacy. Therefore, our alternative two hypotheses read as follows:
As we posited that the divergent ways of measuring advocacy were one reason for the inconsistent findings of previous research, we investigate the impact of public funding on advocacy by using activity-related and resource-related measures in our analyses. By comparing the two sets of results, we aim to establish whether or not the method of measurement matters.
Our resource-related measures are based on the number of working hours invested in advocacy. According to the capacity-building perspective, public funding will positively affect the amount of resources that could potentially be dedicated to advocacy. As a result, nonprofits could, for example, professionalize their advocacy management and establish specific posts or departments, thus enabling greater and more effective advocacy efforts. However, our activity-related measures mainly reflect the relative importance of advocacy activities within an NPO’s mission. According to resource dependence theory, these will become less important as public funding increases, because they are usually in opposition to government policy, and can thus be expected to decrease with a rise in public funding.
Consequently, the effects of public funding on advocacy may be paradoxical: A positive impact on the resources devoted to advocacy could go hand in hand with a decrease in advocacy’s relative importance within the mission and strategy of a nonprofit. At the same time, it is important to note that the power relation between nonprofits and government is not necessarily an asymmetric one, because mutual resource dependence is also possible, as mentioned above (Pfeffer & Salancik, 2003). Various factors, such as the presence of alternative service providers for the government or the availability of alternative revenues for the nonprofit, may also affect government-nonprofit interdependence and thereby mitigate the effects of government funding on nonprofits’ advocacy efforts. Therefore, resource dependence theory suggests the importance of additional variables and not just the level of government funding.
Data, Measures, and Method
For this study, we analyzed data from 250 Austrian nonprofits. This sample is representative of the composition of the Austrian nonprofit sector, which is characterized by a large percentage of organizations active in the fields of social services and health, education and recreation, as well as groups active in international affairs, interest groups, and environmental organizations (see Table 3). An additional feature of the Austrian nonprofit sector is that its members are closely entwined with the public sector. On average, 50% of their revenue originates from the public sector, and they serve as important partners in the provision of social services (Pennerstorfer, Schneider, & Badelt, 2013). In addition, they participate in political-decision making, because many established NPOs are integrated into the process of preparing laws. They are consulted on draft bills, and their comments are taken into account before a law is enacted (Neumayr, Schneider, Meyer, & Haider, 2007). Although there is no legal requirement for nonprofits to be involved in this process, it is nonetheless customary for legislators to call on their expertise (Heindl, 2003; Öhlinger, 2005).
In contrast to the legal context facing 501 (c) (3) nonprofits in the United States, there are no legal restrictions on nonprofit advocacy in Austria, where organizations are always allowed to engage in such activities regardless of their funding structure. Moreover, there are no negative connotations associated with the practice of advocacy or lobbying, as is the case in the United States (Bass et al., 2007). Indeed, advocacy by nonprofits is an uncontested feature of the corporatist culture in Austria and in many other European countries. Germany and France, for example, show a very similar relation between NPOs and the public sector (Salamon & Anheier, 1998). However, even in these corporatist nonprofit regimes, NPOs engaging in undesired advocacy activities are still in danger of losing public service contracts and public support, which might lead to a reduction in advocacy activities.
The data used were collected in 2007 and originated from two different sources. We randomly selected nonprofits from our first source, the Austrian business register, which includes all organizations with at least one paid employee. In order not to exclude nonprofits without paid employees, which, after all, form an important part of the sector, we also drew on a second source, the register of Austrian associations. From this, we randomly selected NPOs operating with volunteer staff only. We contacted 750 nonprofits in total and finally received data from 200 nonprofits drawn from the business register and from 50 organizations drawn from the register of associations. Thanks to this stratified approach, our sample was fully representative of this heterogeneous sector. Most importantly, it included some very small nonprofits to counteract the bias toward large, long-established organizations apparent in most empirical studies to date. The response rate for both samples was 33%. Because low response rates can produce biased samples, we checked whether our samples differed from the population of NPOs in the respective registers (Groves, 2006). With regard to size, funding structure, and field of activity, no significant deviation could be established.
Data were collected using a standardized questionnaire, which was divided into two parts, one administered via telephone, the other sent out by e-mail after completion of the phone survey. The first part of the questionnaire contained questions that respondents could answer spontaneously (e.g., questions on the organization’s aims, tasks, and goals). The second part of the questionnaire elicited financial aspects and other organizational characteristics. By using this split-survey technique, we also aimed to avoid common method bias (cf. Sharma, Yetton, & Crawford, 2009). Our respondents were the CEOs or other key representatives of the nonprofits in question.
Dependent Variables
Resources dedicated to advocacy
Our first dependent variable reports the relative importance of advocacy activities compared with the whole spectrum of activities in which nonprofits engage. To calculate this, we inquired about the proportion of total working hours—voluntary as well as paid—devoted to advocacy. We offered nonprofits a list of 10 activity categories 2 and asked them to report the percentage of total working hours devoted to each, assuming a total labor input of 100%. Together, 2 of these categories—“advocacy and representation of interests” and “public relations”—define the percentage of work dedicated to advocacy. 3 Although the first category captured the core of advocacy, the category “public relations” was included to gather information about public awareness campaigns, public education activities, and similar projects. As shown in Table 3, nonprofits in our sample devoted, on average, some 10.3% of their total working hours to advocacy.
Advocacy activities
To measure nonprofit engagement in advocacy using the activity approach, we handed out a list of five mission statements related to advocacy, for example, “to influence political and statutory decisions” or “to sensitize the general public on certain issues” (see Table 2). Respondents were asked to indicate the perceived importance of these statements in their organizations’ mission on a 5-point Likert-type scale ranging from 1 (not important) to 5 (very important). On the basis of these ratings, an overall score was calculated. As displayed in Table 2, the additive index shows a satisfactory level of reliability, with Cronbach’s α internal consistency value being .807.
Components of the Index to Measure Nonprofits Engagement in Advocacy.
Independent Variables
The independent variable for our H1 and H2 is a dichotomous variable that indicates whether or not nonprofits received funds from public institutions. As shown in Table 3, about 81% of all NPOs in the sample received public funding. The independent variable for H1a and H2a reports the share of public funding as a percentage of total revenue. On average, about 47% of total revenues stemmed from public sources (see Table 3).
Descriptive Information on Variables.
Control Variables
Because our sample was very heterogeneous, reflecting the broad variety of organizations in the sector, we controlled for nonprofits’ field of activity. The model, therefore, includes dummy variables for the fields of education, social services (including health care), and culture and recreation. The fourth field, henceforth referred to as the “other field,” comprises unions and chambers, as well as organizations focused on environmental issues, animal protection, or the representation of political or civil rights. As Table 3 displays, the organizations in our sample were distributed quite evenly across these four fields of activity. We also controlled for NPOs’ geographical range of operation (local, regional, national, or international), applying dummy variables for the four ranges. Furthermore, our analysis controlled for NPO size, expressed as the natural logarithm of total revenues. This control was useful for evaluating resource dependence, as larger organizations are more powerful, which might facilitate their flexibility and independence from a single source like public funding (Kendall & Knapp, 2000, p. 128).
In accordance with resource dependence theory, we used three more variables to control for the mutual dependence between nonprofits and government. First, we asked whether NPOs had access to alternative funding sources, as the availability of such resources lessens their dependence on public funding (Pfeffer & Salancik, 2003; Saidel, 1991; Schmid et al., 2008). We measured this availability using a dichotomous variable based on whether it was very easy (1) or very difficult (0) for the organization to gain alternative financial support in the event of financial strain. Second, we assessed nonprofits’ substitutability (Pfeffer & Salancik, 2003), either as a provider of services or as a provider of information. We estimated this by asking whether there were any other nonprofits, for-profits, or government organizations working on similar topics and activities, and operating in the same local area, which competed with the respondent’s own organization in some way. Finally, we controlled for the diversity of the NPOs’ revenue sources, because organizations with more diverse streams of revenue tend to engage more in advocacy (Bass et al., 2007; Froelich, 1999). To do this, we used a Hirschman Herfindahl Index. This index ranges from “0” to “1,” with “0” representing perfect diversity and “1” representing perfect concentration. For instance, if an organization’s revenue came equally from public, private philanthropic, private/commercial, and ecclesiastical sources, the index value was “0”; if it came from one source only, the value was “1.”
To test our hypotheses, we applied ordinary least square regression models, all including the same controls. Each hypothesis was tested by applying the two different approaches to measuring advocacy, namely the share of working hours devoted to it and the index reflecting the relative importance of advocacy activities. To facilitate an analysis of the contribution of the independent variable in explaining the variance (R2) of the dependent variable, we developed our models in three steps. The first included the government funding variables only. In the second, we added controls for size, industry, and geographic area. In the third model, we added the controls derived from resource dependence theory, namely, access to alternative sources of revenue, competition with other organizations, and diversity of revenue. By doing so, we aimed to discern the extent to which mutual dependence also matters.
Results
To start with the impact of receiving governmental funding, we can safely assert that this shows no relation to the advocacy activities of our sample. The first three columns in Table 4 display the findings when advocacy was measured by working hours. As Model 1 shows, the receipt of public funding does not explain the variance of advocacy at all. In Model 2, about 15% of the variance is explained by the sum of all controls. When three additional controls for mutual dependence between nonprofits and government are added (Model 3), the R2 increases from .154 to .164. It is thus clear that the mere receipt of public funding had no impact on advocacy in any of the three models.
Nonprofits’ Receipt of Public Funding and Advocacy.
Note. Dependent variable: Advocacy; shares of working hours devoted to advocacy: n = 242; advocacy activities: n = 249. OLS = ordinary least squares; ref. cat. = reference category.
p < .10. **p < .05. ***p < .01.
The last three columns in Table 4 present the findings when advocacy is measured with an index, which captures the importance of advocacy activities. Once again, none of the three models reveal a significant relationship between public funding and advocacy. Receiving public funding contributes very little toward explaining the variance in advocacy. In contrast, the field in which the nonprofit is active and the geographical range of its operations have the strongest impact (see Model 5). Adding controls for mutual dependence does not substantially alter any of the coefficients (Model 6).
In addition to the models presented in Table 4, we calculated interaction terms between public funding and individual fields, but the results remained unchanged. Clearly, neither H1 nor H2 can be confirmed by our analysis.
Similarly, regression analyses essentially reveal no significant relationship between the proportion of public funding in total NPO revenues and advocacy. As displayed in the first three columns of Table 5, the share of public funding does not contribute to explaining the variance of advocacy measured by working hours in any of the three models. Instead, about 12% of the variance is explained by the control variables.
Nonprofits’ Share of Public Revenues and Advocacy.
Note. Dependent variable: Advocacy; share of working hours devoted to advocacy: n = 242; advocacy activities: n = 249. OLS = ordinary least squares; ref. cat. = reference category.
p < .10. **p < .05. ***p < .01.
The last three columns in Table 5 present the findings for our activity-based measuring approach. In Models 4 and 5, the proportion of revenues from public sources is weakly but significantly and positively associated with advocacy (with a coefficient of .004 and .003, respectively). For example, after controlling for field and geographic range, an organization that is fully dependent on government funding is expected to rate advocacy activities as slightly more important than organizations that receive no government funding (.3 higher on a 4-point scale). Public funding explains about 1.45% of the variance in advocacy. However, in Model 6, which includes the controls derived from resource dependence theory, the relationship between the proportion of public-provided revenues and advocacy is no longer significant.
Only a few of the control variables affect the degree to which nonprofits engage in advocacy. First, the geographic range of operation had a strong and positive impact. Thus, advocacy was more important for nonprofits operating on the regional, national, or international level than for those only active locally. As shown in all models, nonprofits operating on the national level devoted about 7% more of their total working hours to advocacy than those active at a local level. The field of activity also affected the level of advocacy. Nonprofits pooled in the category “other field” are more engaged in advocacy than those active in the field of education. However, overall, our findings support neither H1 and H1a nor H2 and H2a.
Discussion, Limitations, and Recommendations for Further Research
Our empirical findings on NPOs in Austria are presented against the backdrop of the controversial debates about nonprofit dependence on public funding, which have arisen in recent decades. The findings suggest that there is no significant relationship between advocacy and the receipt of public funding or the proportion of an organization’s total revenue coming from public funding. Even when interaction terms for individual fields were included, we discerned no impact on advocacy, either negative or positive. Furthermore, the use of different methods to measure advocacy did not have any impact on our findings, either, as results did not differ between models using resource-related measures and activity-based measures. Hence, our findings do not support our initial assumption, namely that the method of measurement might cause conflicting results.
However, our data do suggest that the degree to which nonprofits engage in advocacy is influenced by the geographic range of operation and field of activity. NPOs that are active at national level devote a higher proportion of their total working hours to advocacy activities, and/or rate these as more important than organizations operating locally. These results reflect the arrangement of the nonprofit sector in Austria, because local units usually concentrate on the provision of services, leaving advocacy activities to national umbrella organizations.
As for the impact of public funding on advocacy, our results contradict the widespread assumption that public funding suppresses nonprofits’ engagement in advocacy. This finding is in line with other quantitative studies, particularly those that found no effect (Leech, 2006; Suárez, 2009). Chaves et al. (2004), for example, also found that public funding had no impact on advocacy, or merely a slightly positive one. However, we were unable to confirm their other supposition that the dependence of governments on nonprofits as service providers outweighs NPOs’ dependence on public funds. Indeed, we controlled for such mutual dependence by using three different variables. First, we examined whether organizations with access to alternative funding sources were less dependent on the government and therefore freer to engage in advocacy. Although the respective coefficients in all models were indeed positive, they lacked statistical significance.
Second, resource dependence theory would lead us to expect that existing competition with other organizations would reduce government-nonprofit interdependence and therefore lead to fewer incentives for nonprofits to engage in advocacy. Interestingly, we found negative coefficients regarding advocacy activities, but positive ones regarding the resources devoted to advocacy. Although the effect of competition was insignificant across all models, the differing impacts on advocacy activities and resources in the sample suggest that, under conditions of competition, nonprofits invest resources in advocacy without strengthening advocacy-related objectives. Because our measure included resources devoted to public relations, this finding indicates that competition brings about advocacy focusing on issues in the nonprofits’ own interests rather than those that support their clients. We call this tendency “forced advocacy” to reflect situations where NPOs must engage in advocacy activities purely to achieve public recognition, even though advocacy is not one of their key aims. We found this to be a result of increased competition but not, surprisingly, of public funding.
Finally, we expected that nonprofits with more diverse revenue streams would engage more in advocacy. Here again, our findings were mixed and insignificant. The fact that none of our controls were significant shows that, in contrast to the theory’s suggestions, resource dependence does not affect the advocacy activities of nonprofits in Austria. Furthermore, the increase in the R2 when these controls are added is extremely low, indicating that resource dependence was not very relevant for our sample.
However, some limitations have to be considered. First, we did not investigate how nonprofit advocacy is directed, in particular, whether it is in accordance or at odds with government policies. Qualitative studies argue that nonprofits do not necessarily reduce the amount of their advocacy, but rather change the content—probably away from advocacy not welcomed by government to less critical varieties, or to those it actively welcomes (Mosley, 2012; Onyx et al., 2008). Although NPOs in Austria do not seem to reduce the amount of work devoted to advocacy when they receive government funding, we cannot discern whether they changed the content of their advocacy. Resource dependence, however, could cause such behavior. In fact, the measurement of nonprofit engagement in advocacy in terms of the extent of such activities, rather than their type and quality, points toward a limitation of our study. Future empirical research would need to combine qualitative and quantitative approaches to tackle this issue, allowing quantitative studies to benefit from new measures of advocacy that capture its quality in terms of content and style.
Second, our study (as others before) is based only on cross-sectional data. Therefore, we cannot investigate causalities in general and the lagged effects of public funding, in particular. Drawing on resource dependence theory, we have assumed that nonprofits’ engagement in advocacy is a function of their revenue structure. However, causality could also work the other way round. For instance, organizations that concentrate on advocacy could search for alternative sources of revenue to remain independent from the government. Hence, longitudinal analyses are required to thoroughly assess whether public funding affects nonprofit advocacy or whether the functions preferred by nonprofits affect the funding portfolio (Suárez, 2009).
Finally, we are aware of limitations resulting from the specific characteristics of Austrian nonprofits. We argued that Austria is a good example of a so-called corporatist nonprofit regime, under which NPOs are closely entwined with the public sector. This interdependency is not only to be seen in financial matters but also in terms of practical cooperation, as nonprofits take on public responsibilities in many areas. A typical feature of this regime is that certain large Austrian NPOs have traditionally had close links to the major political parties (Christian Democrats, Social Democrats) and are still frequently coupled with those parties (e.g., politicians are board members, former politicians are executives, etc.). These NPOs thus have a strong tradition of advocacy, despite depending on public funding for their service-providing units. This arrangement might explain why our findings indicated that public funding does not alter nonprofit engagement in advocacy. Historical trajectories offer another explanation for the differences between our results and those based on samples from the United States. In Austria, and Central Europe in general, there has always been a relatively strict differentiation between the advocacy and the service function of an organization and a high level of tolerance toward NPOs that exert a sociocritical function. Consequently, Austrian NPOs do not seem to be reluctant to engage in advocacy even when large proportions of their revenue come from public agencies.
Footnotes
Acknowledgements
The authors would like to acknowledge three anonymous reviewers for their constructive comments.
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.
Other
This research was made possible thanks to grant funding from the Austrian Ministry of Science and Research under its program ‘New Orientations for Democracy in Europe - NODE’.