Does Government Funding to Public Charities Crowd Out or Crowd In Volunteers?

• In Event: The Public-Private Mix

Abstract

Both volunteer labor and monetary donations are important resources for nonprofit organizations in achieving their missions and providing services. The majority of public charities utilize volunteer labor (Hager and Brudney 2004). In terms of financing sources, government grants and contracts, taken together, accounted for about one-third of the revenue of public charities in 2013, while private contributions made up for about 13 percent (McKeever 2015). In spite of the growing “mutual dependence of government and nonprofit organizations” (Smith and Lipsky 2009: pp. 4-7), many nonprofit managers have been hesitant about public-nonprofit collaboration for concerns such as loss of independence, donations, and volunteers (Gazley and Brudney 2007).

The literature has most extensively examined the relationship between government funding and private monetary donations. Earlier studies on government funding suggest both “crowding-in” and “crowding-out” effects on monetary donations (e.g. Andreoni and Payne 2003; List and Lucking-Reiley 2002; Lu 2016). In a recent meta-analysis, Lu (2016) concludes that this tradeoff tension does not necessarily exist between grants and monetary donations. However, the answer to whether and how government funding affects volunteer labor remains much unclear.

The study seeks to understand the relationship between government funding and the use of volunteers in nonprofit organizations. Specifically, we focus on two hypotheses: nonprofit professionalization and the signaling effect. Existing literature has pointed out that government funding contributes to the professionalization of nonprofit organizations (Ebaugh, Chafetz, and Pipes 2005; Suárez, 2010). As a result, government funding may crowd out the use of volunteers as nonprofit organizations rely increasingly on professional paid staff. However, are potential volunteers affected by the signaling effect, particularly when individuals know little about the quality of the organization? In monetary donation, Heutel (2014) argues that government grants can function as a signal for charity quality when the individuals are less informed, and the author found a stronger crowding-in effect on younger organizations—which possibly are less known to the public.

This study uses the 2013-2016 Internal Revenue Service (IRS) Form 990 e-filer data from nonprofit charities under Section 501(c)(3). Because the supply and the demand for volunteers or government funding both vary within the nonprofit sector, in comparing the organizations, we ask—given a similar mission, are nonprofits with public funds more or less likely to get volunteers? To control for types of organization, we exploit the rich text data on nonprofits’ mission and main activities in the Form 990s and apply text-as-data methods to create measures of organization type.

Our findings support the crowding-in argument. Specifically, there may be a small signaling effect, meaning that organizations that are disadvantaged in showing their trustworthiness or reputation may be able to attract more volunteers with the support of public funds. Although in the monetary donation literature, the crowding-out effect tends to dominate, particularly among social services and health organizations (Brooks 2000; Guo 2007), our findings imply that possibly volunteers respond to organizations with government funding differently than donors. Furthermore, using the text data, we are able to identify areas that attract more volunteers. For example, consistent with previous studies (Smith and Lipsky 2009; Suárez 2010), nonprofits aiming at providing public and social services, such as homeless shelter, hospice care, and firefighters, engage more volunteers. Other services that require more professionalized knowledge and skills such as substance abuse, education (at various levels), and vocational training, on the contrary, are much less likely to use volunteers.

This study contributes to a deeper understanding of the funding-volunteer interaction. Methodologically, to our knowledge, this is the first study that uses text-as-data methods to analyze mission and activities information reported in Form 990s to better understand the causal relationship between government funding and private donations. In constructing comparable control groups, the existing studies generally rely on financial reporting information and the NTEE taxonomy (see Fyall, Moore, and Gugerty 2018). We show that the topic modeling technique adds value to causal estimation by extracting more information from text.

Our findings do not support the concern that government funding leads to increased professionalization and thus reduces the use of volunteers. Meanwhile, the data show significant differences in organizational dependence on volunteers, given organizational types manifested in their missions and main program service activities.

Authors

• Ruodan Zhang, Indiana University - Bloomington

• Haohan Chen, Duke University

• Jill Nicholson-Crotty, Indiana University