Abstract
Social entrepreneurship covers a broad domain, including social enterprise, defined as the use of for-profit strategies by nonprofit organizations. Driven by multiple factors, nonprofit organizations have increasingly turned to social enterprise in the hopes of funding their social missions. However, only limited research has fully delineated how the use of social enterprises affects overall funding and the conditions under which social enterprises are relatively more effective. This mixed-method, three-study project provides evidence that the introduction of a social enterprise negatively affects individual donations but that some of the negative effects can be mitigated when the social enterprise is perceived as mission consistent and competent. In addition, the results show that donor attitudes toward social enterprises moderate the effects. The results of these studies have important implications for marketing, social entrepreneurship, and public policy.
Nonprofit organizations undertake social enterprises for a variety of reasons (Dees 1998). First, social enterprises offer the potential to develop more sustainable sources of funding through the creation of a business rather than the reliance on charitable philanthropy, in which issues such as economic downturns and donor fatigue may negatively affect organizational funding (Weisbrod 1998, 2004). Second, social enterprises may not only increase funding but also help nonprofits carry out their social missions, especially when the social enterprise is directly related to that social mission. For example, Delancey Street Movers is a self-help organization in San Francisco that provides exconvicts, substance abusers, and homeless people—people who often have trouble finding employment—with employment opportunities through a financially sustainable moving company. In this way, the social mission of workforce development is carried out through the social enterprise of the moving company. Finally, social enterprises may allow nonprofits to diversify their funding stream to include not only grants and donations (Lazarevski, Irvine, and Dolnicar 2008) but also earned income. In summary, these issues contribute to the increasing trend of nonprofit organizations becoming more dependent on social enterprises (Dees 1998; Weisbrod 2004).
Although social enterprises offer potential benefits to nonprofit organizations, they may also lead to at least two challenges (Dees 1998; Forster and Bradach 2005; Weisbrod 2004). First, social enterprises may distract nonprofit organizations from their intended social mission (Alter 2006; Dees 1998). For example, some people view religious organizations (e.g., churches) engaging in gambling activities (e.g., bingo) as inconsistent with the mission of the organization. Such distractions may cause external stakeholders to question the identity of the organization and, therefore, to question their participation in the organization (Smith et al. 2010; Zietlow 2001). Second, nonprofit organizations may not have the skills, capabilities, and resources to develop successful for-profit social enterprises. As a result, social enterprises may generate little or no income while consuming financial resources of the organization (Foster and Bradach 2005).
Despite the increasing reliance on and the many positives and negatives associated with social enterprises, relatively little research has focused on how and why social enterprises affect other funding sources, such as grants and donations. Although the interrelationship of nonprofit funding sources has received empirical support, in which funding from one source (e.g., government grants) often affects funding from another source (e.g., donations; Brooks 2009; Weisbrod 1998), little attention has focused on the consequences of nonprofits engaging in social enterprises on other funding sources (see Kingma 1995).
In this article, we focus on one funding source—individual donors—to understand how different social enterprises may affect an individual donor's likelihood and willingness to make financial donations. Specifically, we draw on the concept of fit from the marketing literature to understand how the degree of fit between the social enterprise and the nonprofit organization may affect individual donor giving behavior. In this way, we both borrow from and contribute to the research agenda within corporate societal marketing that focuses on the role of fit and the effectiveness of a broad range of corporate social initiatives (Hoeffler, Bloom, and Keller 2010; Sen and Bhattacharya 2001). In addition, we examine how organizational competence and donor attitudes may affect individual donor giving behavior. These issues contribute to a broader understanding of how competence and attitudes affect similar social marketing issues, such as cause-related marketing (Pracejus and Olsen 2004) and corporate social responsibility (Sen and Bhattacharya 2001), and how nonprofit organizations engage in the promotion and marketing of their social enterprises. Given the relative size of individual donations to many nonprofit organizations and the increased usage of social enterprises, this mixed-method, three-study project has important theoretical and practical implications for marketing, social entrepreneurship, public policy, and the ongoing delivery of social services.
We designed the three studies to help answer questions about how social enterprises affect individual donor likelihood and intent. The first study is a qualitative study using semistructured interviews with executive directors of nonprofit organizations to sensitize us to key issues related to social enterprise and individual donors. The second study uses an experimental design to address the effects of different types of social enterprises on donation likelihood and donation intention. The third study introduces additional attitudinal constructs, which we hypothesize mediate and moderate the relationships among mission consistency, entrepreneurial competence, and donation behaviors.
Study 1
We conducted the first study to identify important themes associated with how and why donors react to the initiation of social enterprises by nonprofit organizations. This study served helped uncover some of the key issues of social enterprise and individual donor giving. The purpose of the study was to identify potential constructs and rationales about the influence of social enterprises on donations that we could test in quantitative experiments. Although interviews with individual donors might have been helpful, we focused instead on the executive directors of nonprofit organizations to gain a broader sampling of different reactions because each executive director interacted with the many different individual donors.
Methods
Sample
Study 1 consisted of semistructured interviews with 20 executive directors of nonprofit organizations actively involved in a social enterprise in the United States. The identification of appropriate nonprofits occurred through two channels as part of a larger study on social enterprises. First, we identified a publicly available database that listed nonprofit organizations engaged in social enterprises. The database included a search field of venture type, with options including clerical services, construction, curriculum sales, landscaping, light manufacturing, restaurants/catering, and thrift stores. Second, we used the snowball method of asking the participants involved in the study to identify other participants (Miles and Huberman 1994) who engaged in social enterprise. The age of the nonprofit organizations ranged from 6 years to more than 100 years, with an average of 20 years. The age of the commercial activities ranged from 2 to 20 years, with an average age of 7 years.
Data Collection
We collected data through two sources: archival data and use of semistructured interviews with the executive directors. The use of multiple methods of data collection allows for convergence and triangulation of the findings (Jick 1979). First, we collected archival data from the nonprofit organizations, including annual reports, press releases, and data available on the public database of social enterprises. Second, the semistructured interviews enabled us to ask the same questions of all participants and to ask follow-up questions as they arose (Lee 1999). Questions included topics such as the executive director's career background, history of both the nonprofit organization and the social enterprise, decisions to engage in social enterprises, and donor responses to social enterprises (for the interview protocol, see Appendix A). Two of the authors conducted all 20 interviews over the telephone, and each interview lasted approximately 60 minutes on average. A professional transcriber recorded and transcribed the telephone interviews verbatim. Interview transcripts averaged 17 pages and totaled 342 pages.
Data Analysis
In our data analysis, we were guided by the tradition of grounded theory (Glaser and Strauss 1967; Strauss and Corbin 1998). We used a two-step coding process, in which codes were determined inductively from the data and agreed on by the researchers. The process of coding is an established method of meaning condensation (Lee 1999) in which researchers abstract the most relevant themes (or codes) from the data and develop a dictionary during data analysis to identify their meanings (Kreiner, Hollensbe, and Sheep 2006).
On receipt of the interview transcripts, a two-step coding process ensued. In the first step, we independently read the interviews and developed codes for broad themes based on the content of the data. We assigned codes to a passage of text when we identified a phenomenon present in the transcript. On completion of the coding of the interviews, in the second step, we jointly coded the same interview. Disagreements were resolved through comparison and discussion. Although this emergent type of coding process makes traditional interrater reliability measures impractical, our coding process involved multiple perspectives from multiple researchers on each interview, helping reduce researcher bias in the data analysis (Kreiner, Hollensbe, and Sheep 2006).
Our ongoing data analysis was framed by ground theory techniques, which led to iteratively moving back and forth between the themes in the data and existing literature. This iterative process helped us become deeply involved in the phenomenon and match the themes to existing literature to identify gaps of knowledge. After coding 17 interviews, we found no new themes, suggesting that we had reached a point of theoretical saturation (Strauss and Corbin 1998).
Findings: Key Themes and Relationships
Responses from Study 1 helped us identify key themes and relationships about how and why social enterprises may affect nonprofit organizations and their individual donors. Overall, the results suggest that social enterprises are challenging for the key stakeholders of a nonprofit organization, including individual donors. Several different issues were raised, but the three most frequently cited issues, in which extensive variance occurred both positively and negatively, included (1) the perception of mission consistency of the social enterprise, (2) the perception of entrepreneurial competence of the organization, and (3) the attitude toward social enterprise. In the following subsections, we provide brief explanations and corresponding examples of how and why mission consistency, entrepreneurial competence, and attitudes affect individual donor views of social enterprises.
The Role of Mission Consistency
One of most common themes across the interviews with executive directors of nonprofit organizations engaged in social enterprise (identified by 90% of the respondents) was the perception of fit or consistency between the nonprofit organization and the social enterprise. While the issue of mission consistency was almost universally present, the view of mission consistency varied substantially. For some organizations, social enterprises were viewed as two mutually exclusive domains in which the nonprofit organization and the social enterprise were at odds with each other. For example, as one executive director suggested,
I think what nonprofits sometimes do is that they start to run these businesses, and they are serving two gods, if you will. They are serving the clients who come to the business and they are serving the employees who are getting the job skills or training. When you do that you create a dual and sometimes opposite set of needs and requirements.
As another executive director suggested, “It's a balancing act between managing the pressures of a small business and maintaining fidelity to the mission of the nonprofit.” However, several other executive directors did not find it challenging to managing the mission consistency of the nonprofit organization and the social enterprise. For example, one executive director commented, “We didn't start the social enterprise to serve our mission. Our mission is to run the small business so that people can work. So, there is nothing without the business.”
The issue of mission consistency was also identified as an important consideration for donors. For example, one executive director, who runs a manufacturing business that employs partially blind workers, highlighted how the nonprofit and social enterprise elements of an organization may cloud the understanding of the donor:
If we were just a social service organization that provided skill development, training and counseling, it's something the public can relate to easier. Funders and donors understand that. But, when you combine the fact that we're also a business and we also look to make profit in our business, that confuses our image to the public…. Our individual donor, who thinks that all we do is take care of blind people, now finds out about all of our manufacturing operations—it confuses them.
Extending the role of mission consistency, several executive directors identified that the mission consistency of the specific social enterprise mattered to individual donors. For example, one executive director talked about the shifting views of the donors based on whether the specific social enterprise was mission consistent or not:
Initially, I only got positive responses from donors. If it's mission focused, it's tangible. If you're doing it right, it's a tangible program component. I think there would be raised eyebrows if we had set up a Mailbox Plus, a national franchise … but Mailbox Plus is a neutral thing. It wasn't like it was a gambling casino.
While donors seemed to view a mission-consistent social enterprise positively, a mission-neutral (i.e., much less inconsistent) social enterprise raised concerns.
The Role of Entrepreneurial Competence
Another common theme emanating from the interviews with executive directors of nonprofit organizations (identified by 85% of the respondents) was the perceived entrepreneurial competence of the organization. As nonprofit organizations engaged in social enterprise, many executive directors highlighted the need for entrepreneurial skills and abilities. In many cases, the executive directors recognized that the skills of their employees and the skills required of the social enterprise differed. For example, as one executive director cautioned, “Don't hire a social service person to make your business a success. We'll give away everything we can and charge as little as possible.” Another remarked, “A lot of nonprofits do not have people that have entrepreneurial backgrounds…. There's a lot of good-hearted people, but business is just different than nonprofits.” An executive director, with previous entrepreneurial experience, put it this way:
There are a lot of really idealistic, utopian people about social enterprise and how to make this work. I'm an entrepreneur outside of here and I know the commitment, the blood, sweat and tears stuff that [go] into a start-up. In the nonprofit sector, there's a real naïveté about what's involved in starting a business, never mind a social enterprise, which is fraught with a whole level of other challenges conventional businesses don't even see—social service support, housing issues, legal issues, addictions.
The necessary entrepreneurial competencies often involve the abilities to identify profitable social enterprise opportunities, to raise necessary financial capital, and to execute the business opportunity. For example, one executive director suggested the difficulty in identifying social enterprise ventures: “We had been looking for an income-generating project, but we could never quite come up with anything that was going to generate enough money to make it worthwhile.” Finally, many executive directors expressed concern with executing the business opportunity even when the opportunity seemed straightforward. As one claimed, “It's just a franchise, but it's having the entrepreneurial expertise and money to set up a franchise and run it while you're trying to run the nonprofit.”
One of the primary concerns related to entrepreneurial competence was how it might affect an individual's image and potentially the donations to the nonprofit organization. Another concern was how donors might react if the lack of entrepreneurial competence led to an unsuccessful social enterprise. For example, as one executive director explained,
You're investing in inventory. What if it just sits on the shelf and no one buys it? Have we not been good stewards of the organization and misused the resources of the organization? Are people going to look at us and say, “How could you let this happen to our (nonprofit)?”
Another executive director was concerned about how the unsuccessful social enterprise history might affect donors’ future support: “We've done [social enterprises] in fits and starts…. It just doesn't generate enough revenue for us to be out there touting it to donors.”
The negative concerns were not shared by all executive directors. Instead, some believed that the entrepreneurial competence of the organization and its successful social enterprise would actually contribute to future donations. For example, as one executive director commented, “We've been able to successfully argue [to donors] that by them providing financial resources to us … to successfully grow our social enterprise, that those dollars are going much further than if they simply went to program delivery.” Ironically, even the executive directors with successful social enterprises were not convinced this would lead to increased individual donations. For example, one executive director said, “It's interesting in talking to other nonprofits. Some are very reluctant to even mention social enterprise or play up their earned income streams for fear that donors will reduce their donations.”
The Role of Attitude toward Social Enterprise
Beyond the issues of mission consistency and entrepreneurial competency, another common theme across the interviews with executive directors of nonprofit organizations engaged in social enterprise (identified by 80% of the respondents) was the attitude toward social enterprise. More important, the role of the attitude toward social enterprise was identified as a factor that may affect an individual donor's behavior.
Overall, the executive directors viewed the attitude toward social enterprise with intense feelings—both negative and positive. One executive director recalled the first time the concept was raised internally: “I remember one meeting where our program managers reacted so strongly to social enterprise. They asked, ‘What kind of organization are we becoming? Oh my God, poison just walked through the door. Put up the garlic.'” Although this response was from an internal member of the organization, it highlighted the potential for intense attitudes toward social enterprise. This type of response was illustrated by some donors who “restricted” their donations to the organization to be used only by the social service side of the organization and not the social enterprise.
In contrast with negative experiences, some executive directors experienced extremely positive feedback from donors based on their attitude toward social enterprise. For example, as one executive director explained,
[Social enterprise] has been very attractive [to donors], and there has been a core of individuals who have just really liked it, and really are passionate about it…. Many of them have made their money in business and they just think it's fantastic that business can be used to do good things, and you don't have to constantly be begging or working on the mind-set of where the next dollar is going to come from.
Another executive director described the attitude of the organization's initial donor as follows: “He's one of these guys who felt like asking for a grant was like sitting outside on the sidewalk with a cardboard sign saying, ‘Please give me money.’ He just didn't like the concept of asking for money. He was a pick- yourself-up-by-the-bootstraps kind of guy.” Echoing a similar issue, some executive directors pointed out how this issue was particularly important to some of their larger donors. For example, as one respondent remarked, “A lot of really significant donors are business people and they respect the fact that we're out there trying to do nontraditional things [such as social enterprise] in order to support our organization.”
In total, the qualitative observations from Study 1 helped inform our theoretical thinking and design for the quantitative experiments by highlighting the role of mission consistency, entrepreneurial competence, and attitude toward social enterprise in affecting individual donations when nonprofits engage in social enterprise. Drawing on these themes, we now turn to a literature review to develop our hypotheses and two experiments to further test and explore these issues.
Conceptualization and Hypotheses
The Crowding-Out Hypothesis
The relationship between social enterprises and donations is an important issue for many stakeholders in the field of social entrepreneurship because it directly involves the financial sustainability of these organizations and, thus, their ability to continue to provide critical social services. Faced with an increasingly competitive landscape for donations, “nonprofit organizations are becoming more dependent on commercial activities” (Weisbrod 1998, p. 16). Considering that nonprofit organizations are often funded through multiple sources of revenue (e.g., grants, donations, commercial income), the increasing dependence on commercial activities raises important questions about how social enterprises may affect other revenue sources of nonprofit organizations.
According to the crowding-out hypothesis, an increase in one source of revenue, such as a government grant, can lead to a decrease in revenue from other sources, such as private donations (e.g., Weisbrod 1998). The crowding-out hypothesis, from the economics literature, is based partly on the idea that donors may decide that their donations are no longer needed because of funding from other sources (Weisbrod 1998). Across seven major economic studies in the United States from 1982 to 1998, every $1.00 in funding from government organizations resulted in a displacement or crowding out of between $.15 and $.40 in private giving (Brooks 2000). Extending this logic to social enterprise, donors may reduce their individual contributions because the nonprofit organization has other sources of revenue and therefore is no longer as dependent on donor contributions (Segal and Weisbrod 1998). Kingma (1995) also finds evidence of crowding out using a cross-sectional design. If a nonprofit organization engages in a successful social enterprise, donors may perceive the nonprofit organization as less dependent on their donations and therefore may reduce their donations to the nonprofit organization. Thus, we anticipate a negative relationship between successful social enterprise and donor likelihood and intention to donate:
H1: When a nonprofit organization engages in social enterprise and revenues from social enterprise increase, donation likelihood and intentions should decrease.
Mission Consistency and Entrepreneurial Competence
Although H1 creates a baseline for social enterprises, many different types of social enterprises exist. Scholars have suggested that one way to envision different types of social enterprises is to examine their mission consistency and financial contribution to the organization (Wei-Skillern et al. 2007).
First, we examine different types of social enterprises that vary in their mission consistency between the social enterprise being pursued and the social mission of the nonprofit organization (Wei-Skillern et al. 2007). When a person considers donating, his or her perception of fit between the type of social enterprise being pursued and the social mission of the nonprofit organization is likely to affect donation likelihood and intentions.
In the management literature, the concept of fit refers to the transferability, synergy, or complementarity of products, skills, or markets (Porter 1987; Rumelt 1974). In the marketing literature, the perception of fit has been used extensively in areas such as brand extensions (Aaker and Keller 1990) and a wide range of corporate social initiatives (Hoeffler, Bloom, and Keller 2010), including cause-related marketing (Pracejus and Olsen 2004) and corporate social responsibility (Sen and Bhattacharya 2001). The basic logic regarding the role of fit is that people possess cognitive categories for certain objects. As additional objects are added, people compare the new objects with the previous objects to assess the number of shared associations (or similarities) between the objects (e.g., Bousch and Loken 1991). For example, effective brand extensions occur when consumers perceive a good fit between the original brand and the extensions (Aaker and Keller 1990).
Extending this logic to social enterprises of nonprofit organizations, the perceived fit between the social mission of the nonprofit organization and its commercial activity reflects the degree of mission consistency (Wei-Skillern et al. 2007). Whereas mission consistency is often viewed through the lens of the decision makers of the organization, we focus on a donor-based conception of mission consistency. We anticipate that the degree of mission consistency will affect donation likelihood and donor intentions, as we found in our qualitative data. As donors compare the new object of the social enterprise with the previous object of the social mission of the organization, they will perceive a degree of fit that will influence their donation likelihood and intentions. If donors perceive the commercial activity of the social enterprise as consistent with the nonprofit organization's social mission, they will likely be more supportive of the commercial activity through donations. Conversely, if they perceive the commercial activity of the social enterprise as inconsistent with the nonprofit organization's social mission, they will likely be less supportive of the social enterprise. Thus, we suggest the following:
H2: When a nonprofit organization's social enterprise is mission consistent (inconsistent), donation likelihood and intentions should be higher (lower).
Second, we examine different types of social enterprises based on their financial performance. In the entrepreneurship literature, theoretical (Man, Lau, and Chan 2002) and empirical (Erikson 2002) research highlights the relationship between entrepreneurial competence and financial performance. Entrepreneurial competence refers to the ability to identify opportunities, marshal resources, and execute business opportunities (Man, Lau, and Chan 2002). As we identified in the exploratory interviews in Study 1, donors may question the entrepreneurial competence of a nonprofit organization to develop a profitable social enterprise. When this occurs, donors will be less likely to support the nonprofit organization. Therefore, we suggest the following:
H3: When a nonprofit organization's social enterprise is entrepreneurially competent (incompetent), donation likelihood and intentions should be higher (lower).
Attitude toward Social Enterprise
Moving beyond the main effects of mission consistency and entrepreneurial competence, we were also interested in interaction effects related to attitude toward social enterprise. Building on the theory of planned behavior (Ajzen 1991), research across many different fields, including advertising and public relations, has found evidence of the relationship among attitudes, intentions, and behaviors (e.g., Ajzen and Fishbein 2005). As such, we were interested in understanding how a dispositional attitude toward social enterprise may affect donation behavior. According to the theory of planned behavior, the attitude toward a behavior would lead to the behavioral intention and, ultimately, to the behavior. In the current context, the attitude toward social enterprise should affect the behavioral intention and, ultimately, the intention to donate to a nonprofit organization engaged in social enterprise, as we found in our qualitative data. In line with the predicted main effects of mission consistency and entrepreneurial competence, we predict that the attitude toward social enterprise will moderate the relationship among mission consistency, entrepreneurial competence, and donation likelihood and intentions:
H4: When attitude toward social enterprise is low (high), the benefits of mission consistency and enterprise competence (i.e., higher donation likelihoods and intentions) should weaken (increase).
Study 2
We designed Study 2 to help understand the effects of social enterprise on nonprofit donations. Specifically, we investigated whether the introduction of a social enterprise to a nonprofit affects donation likelihood judgments and donation intentions. In addition, we examined the role of mission consistency (inconsistency) and entrepreneurial competence (incompetence) of the social enterprise. As discussed previously, we predict that donation likelihood judgments and donation intentions will decrease with the introduction of the social enterprise (H1) but that these judgments and intentions will be higher when the social enterprise is mission consistent and competent (H2 and H3).
Method
We employed a 2 (repeated measure: pre- vs. postexposure to a social enterprise) × 2 (mission consistent vs. mission inconsistent) × 2 (competent vs. incompetent) mixed experimental design. Mission consistency and competence were between-subject factors; we manipulated them using hypothetical scenarios developed through pretests. 1
We conducted pretests to identify the appropriate social enterprises to use in the scenarios. As a starting point, we referred to an archival database on social enterprises to better understand which social enterprises were common in practice. We then conducted pretests on a wide range of social enterprises (e.g., lawn service, dry cleaning, manufacturing, cleaning services) to identify the social enterprise scenarios to use.
Sample
Participants were 185 students recruited at a large midwestern university, who received extra course credit as an incentive to participate. They ranged in age from 18 to 30 years, with a mean age of 19 years, and 62% were women. To assess this population's donation practices, we asked participants if they had donated to a nonprofit organization within the past year and, if so, how they would rate themselves in terms of their donation behavior in comparison with others’ (five-point scale, from “I donate much less” to “I donate much more”). Of the participants, 68% indicated that they had donated within the past year, and 83% indicated that they donated the same or more than others.
Procedures and Measures
We gave the participants a questionnaire and told them that they would be asked to answer some questions about for-profit and nonprofit businesses. The cover sheet of the questionnaire also provided basic definitions of for-profit and nonprofit businesses to ensure that all participants understood the difference between these types of organizations.
The main part of the questionnaire was broken into two sections. First, participants received a description of a fictional nonprofit organization and were asked to imagine that they had been asked to donate to that organization. The description read as follows:
Imagine that you have been asked to donate to a well-known nonprofit organization, whose mission is to help poverty-stricken individuals learn to become self-sufficient through job training and placement in the food service industry. The organization also provides assistance to its clients in job search and placement. You like the work of this organization and have donated money to it in the past.
The participants were then asked how likely they would be to donate to this organization (on a seven-point scale, from 1 = “not all likely” to 7 = “extremely likely”). They were also asked how much they would donate to this particular organization if they had $100 to donate to any charitable organization. Participants indicated a dollar amount ranging from $0 to $100.
Second, participants read one of the pretested focal social enterprise scenarios. We manipulated mission consistency in the scenario using the type of social enterprise employed—namely, a full-service catering business (mission consistent) versus a check-cashing and loan service business (mission inconsistent). We manipulated social enterprise competence in the scenario using a description of the success or failure of the business (for all scenarios, see Appendix B). For example, the mission consistent/competent scenario reads as follows:
The nonprofit organization described above has now also opened a for-profit, income-producing full-service catering business for paying customers. After one year of business, the for-profit catering business is doing well and earning income that will be used to support the nonprofit programming side of the organization.
After reading the focal social enterprise scenario, participants again completed the main donation measures, donation likelihood judgment and donation intention in dollars. We also collected additional demographic measures, including gender, age, and ethnicity; we found no main effects or interactions across these variables.
Finally, the mission consistency manipulation involved a three-item, five-point scale (“It is consistent with their purpose/makes sense given their mission/is related to what they do”; 1 = “strongly disagree,” and 5 = “strongly agree”). The mission consistency manipulation check scale was reliable (α = .88) and was significantly different between mission consistent (M = 3.56) and inconsistent (M = 3.03; t = 3.74, p < .01) conditions. The entrepreneurial competence manipulation involved a two-item, five-point scale (“I perceive the business to be successful/perceive the business to be profitable”; 1 = “strongly disagree,” and 5 = “strongly agree”). The entrepreneurial competence manipulation check was reliable (α = .84) and was significantly different between entrepreneurial competent (M = 4.20) and incompetent (M = 2.30; t = 19.16, p < .001) conditions. Thus, both manipulations were successful.
Results and Discussion
Mixed analyses of variance on the focal dependent measures revealed a significant main effect of the repeated measure for donation likelihood (F(1, 182) = 148.08, p < .001) and donation intentions (F(1, 182) = 71.53, p < .001), with donation likelihood and donation intentions decreasing from time 1 to time 2 (M = 5.09 vs. 3.89; M = $52.84 vs. $38.60; for means across all conditions, see Table 1). Thus, H1 is supported; the introduction of a for-profit social enterprise can have a negative impact on donations.
Study 2: Donation Likelihood and Donor Intentions ($) as a Function of Time (Pre- vs. Postexposure to the Focal Scenario), Competency, and Mission Consistency
Mission Consistency
We found a significant repeated measure × mission consistency interaction for both donation likelihood (F(1, 182) = 12.67, p < .001) and donation intention (F(1, 182) = 3.95, p < .05). Follow-up tests showed that in mission-consistent (vs. mission-inconsistent) conditions, focal donation likelihood judgments at time 2 were higher, as we predicted (M = 4.21 vs. 3.57; t = 2.98, p < .01), and the difference between the time 1 and time 2 measures was lower, as we predicted for both donation likelihood judgments and donation intentions in mission consistent conditions (M = .853 vs. 1.557; t = 3.56, p < .001; M= $10.89 vs. $17.58; t = 1.99, p < .05). Thus, in general, H2 was supported. Still, it is worth noting that though having a mission-consistent social enterprise helped lessen the decline in donations, mission consistency was not sufficient to overcome the negative impact of the introduction of a social enterprise.
Entrepreneurial Competence
We also found a significant repeated measure × competency interaction for donation likelihood (F(1, 182) = 3.65, p = .058). However, contrary to our predictions, follow-up tests showed no significant differences across competent versus incompetent conditions at time 2 (M = 3.93 vs. 3.85; M = $39.52 vs. $37.70, n.s.), though the pattern of means was in the hypothesized direction; that is, donation likelihood and intentions were higher in the competent (vs. incompetent) conditions. Thus, H3 was not supported.
To better understand this finding, we conducted a post hoc examination of qualitative comments gathered after participants completed the dependent measures. We asked them to indicate why their donation behavior changed (if it did so) after the introduction of the for-profit business. In more than half the cases (53.5%), the participants indicated competency as one reason for changing their donation likelihood and intentions. For example, one participant said, “If the business is trying to make its own income and failing to be successful, I would begin to doubt how successful the job training and placement is going to be. Am I just wasting my money?” Another said, “I would feel that my money may be of better use somewhere else since they are now making their own money to support themselves.” Given this strong frequency, we developed a stronger manipulation of enterprise competence in Study 3 to explore this result further.
Our results are consistent with previous literature, which suggests that a crowding-out effect is likely to occur when a nonprofit engages in a social enterprise, resulting in lower donation likelihood judgments and intentions. However, if a social enterprise is mission consistent, supporting the nonprofit's core goal, this can mitigate, at least to some extent, the negative impact on donations. In considering this effect of mission consistency, we questioned whether inherent in the formation of judgments about mission consistency were attitudes toward whether the enterprise itself is good, ethical, and worthwhile. Thus, an additional goal of Study 3 was to explore the potential underlying cognitive processes of our effects.
Study 3
To better understand the relationship among social enterprise, mission consistency, and competence, in Study 3 we again evaluated these factors, using a nonstudent sample to further generalize our results. In addition, we examined general dispositional approval of social enterprises and attitude toward the specific social enterprise itself, to begin to uncover the underlying cognitive mechanisms, which may have influenced our results. As in Study 2, we predicted that donation likelihood judgments and donation intentions would decrease with the introduction of the social enterprise (H1) but that these judgments and intentions would be higher when the social enterprise was mission consistent (vs. inconsistent; H2) and competent (vs. incompetent; H3). In addition, we predicted that general dispositional approval of social enterprises would moderate the relationship between donation likelihood and intentions and mission consistency and competence, such that when dispositional approval was low (vs. high), donation likelihood judgments and donation intentions would decrease (H4).
Method
The design of Study 3 was similar to Study 2 with the following important exceptions: We employed a nonstudent sample, developed a stronger manipulation of competency to emphasize the commercial activity's performance, and included additional attitude and approval measures. We employed a 2 (repeated measure: pre- vs. postexposure to a social enterprise) × 2 (mission consistent vs. mission inconsistent) × 2 (competent vs. incompetent) × 2 (high vs. low dispositional attitude toward social enterprises) mixed experimental design. Mission consistency and competence were between-subject factors; we manipulated them using hypothetical scenarios (for a full description of all scenarios, see Appendix C).
Sample
We used a convenience sampling procedure with surveys distributed to adults in a midwestern metropolitan area. Participants were recruited from a local community organization and were randomly assigned to experimental conditions. The questionnaires were administered in a classroom-type setting. Participation was voluntary, and participants were not compensated. In total, 59 (45.4%) men and 71 (54.6%) women completed the questionnaire; the median age of participants was 41 years. We found no main effects or interactions across demographic variables. As in Study 2, to assess this population's donation practices, we asked the participants if they had donated to a nonprofit organization within the past year and, if so, how they would rate themselves in terms of their donation behavior in comparison with others. Of the participants, 98% indicated that they had donated within the past year, and 90% indicated that they donated the same or more than others.
Procedure and Measures
The procedure participants followed was similar to Study 2. Participants received a questionnaire and were told they would be asked to answer some questions about for-profit and nonprofit businesses. Definitions of for-profit and nonprofit organizations were included in the cover sheet. The participants then read a description of a fictional nonprofit organization and were asked to imagine that they had been asked to donate to that organization. We then collected the dependent measures of donation likelihood judgment (seven-point scale, from 1 = “not at all likely” to 7 = “extremely likely”) and donation intentions (dollar amount: $0–$100) used in Study 2. In the next section, the participants read one of the focal social enterprise scenarios. We again manipulated mission consistency using social enterprise type (catering business vs. check-cashing and loan service business). We manipulated social enterprise competence in the scenario using a description of the success or failure of the business, emphasizing the amount of money the enterprise was earning. After reading the focal social enterprise scenario, participants again answered the main donation measures, donation likelihood judgment, and donation intention in dollars.
Next, participants indicated their attitudes toward the social enterprise itself (i.e., the catering business or check-cashing and loan service). We assessed attitudes with a four-item, seven-point semantic differential scale (“bad/good,” “negative/positive,” “bad for society/good for society,” and “unethical/ethical”); the scale was reliable (α = .98). We summed these items to form an overall attitude toward the social enterprise strategy.
In addition to the focal dependent measures, participants completed a dispositional attitude toward social enterprises measure at the end of the study. This measure captures the extent to which people approve of nonprofits generally engaging in for-profit businesses. Those high in attitude toward social enterprise consider nonprofit engagement in social enterprise generally a good idea, useful, and congruent with the goals of the nonprofit. We measured the six Likert-type items (“Nonprofit organizations engaging in for-profit businesses is a good idea for nonprofit organizations/performs a useful function for nonprofit organizations/is consistent with the image of nonprofit organizations/fits the concept of nonprofit organizationis consistent with the mission of nonprofit organizations/distracts nonprofit organizations from their primary purpose”) on a five-point scale (from 1 = “strongly disagree” to 5 = “strongly agree”) with one item reverse scored. The scale showed both good reliability (α= .89) and excellent confirmatory factor analysis across this study and two pretests. In each case, the variance extracted exceeded .50, ranging from .55 to .70. Confirmatory factor analysis also yielded acceptable fit with the data (root mean square error of approximation < .05, and goodness-of-fit index = .99). We summed these items to form an overall attitude toward social enterprise score; we then assigned the participants to either high- or low-approval conditions using a median split.
Results and Discussion
Donation Likelihood and Donation Intentions
As in Study 2, the manipulation checks for mission consistency and competency were reliable and significant (αs > .88, ps < .001) Mixed analyses of variance revealed a significant main effect of the repeated measure for donation likelihood (F(1, 122) = 207.29, p < .001) and intentions (F(1, 122) = 86.28, p < .001), with donation likelihood and donation intentions decreasing from time 1 to time 2 (M = 5.25 vs. 3.25; M = $47.26 vs. $25.27; for means, see Table 2). Thus, H1 is again supported.
Study 3: Donation Likelihood and Donor Intentions ($) as a Function of Time (Pre- vs. Postexposure to the Focal Scenario), Competency, and Mission Consistency
Mission Consistency
We found a significant repeated measure × mission consistency interaction for both donation likelihood (F(1, 122) = 9.16, p < .01) and donation intentions (F(1, 122) = 4.71, p < .04). Follow-up tests showed that in mission-consistent (vs. mission-inconsistent) conditions, donation likelihood judgments and intentions at time 2 were higher, as we predicted (M = 3.70 vs. 2.66; t = 3.59, p < .001; M = $29.82 vs. $19.50; t = 2.23, p < .03), and the difference or gap between time 1 and time 2 measures was also lower, as we predicted for both donation likelihood judgments and donation intentions in mission-consistent conditions (M = 1.63 vs. 2.49; t = 2.90, p < .01; M = $17.22 vs. $23.25; t = 2.13, p < .04). Thus, H2 is supported. These results mirror those of Study 2; donation likelihood judgments and intentions are higher when mission consistency is high (vs. low), but overall donation intentions are still lower after initiation of the social enterprise.
Entrepreneurial Competence
There was also a significant repeated measure × competency interaction for donation likelihood (F(1, 122) = 18.45, p < .001) and donation intentions (F(1, 122) = 5.87, p < .02). Follow-up tests showed that in competent (vs. incompetent) conditions, donation likelihood judgments and donation intentions at time 2 were higher, as we predicted (M = 3.69 vs. 2.68; t = 3.49, p < .001; M = $29.35 vs. $19.99; t = 2.02, p < .05), and the difference or gap between time 1 and time 2 measures was lower for both donation likelihood judgments and donation intentions (M = 1.52 vs. 2.60; t = 3.61, p < .001; M = $17.54 vs. $28.81; t = 2.35 p < .03). Thus, unlike the findings in Study 2, follow-up tests showed significant differences across competency conditions. Thus, H3 was supported; when a nonprofit's social enterprise is competent, donation likelihood judgments and intentions are higher and able to lessen the drop in overall donations. 2
Donation likelihood and donation intentions are significantly correlated (r = .70, p < .001), and a similar pattern of results emerged across both measures. In the interest of brevity, we show subsequent results for donation likelihood only.
Although no support exists in the literature, an interaction of the repeated measure, mission consistency, and competence might be expected. Though not specifically hypothesized, the three-way interaction was nonsignificant. However, the pattern of results was consistent with our expectations. For example, for a mission-consistent and competent social enterprise, donation likelihood judgments at time 2 were highest (M = 4.19) and, conversely, were lowest in mission inconsistent and incompetent conditions (M = 2.14; t = 4.95, p < .01). Furthermore, for mission consistent enterprises, the negative impact of low competence was less (M = 3.21 vs. 2.14; t = 2.52, p < .05).
Dispositional Attitude toward Social Enterprises
For measured attitude toward social enterprises as a potential moderator, the results showed a significant repeated measure × attitude interaction for donation likelihood (F(1, 122) = 8.03, p < .02), with higher likelihood judgments for those high (vs. low) in dispositional attitude toward social enterprises (M = 3.59 vs. 2.89), as we predicted. More important, the results showed a significant interaction among the repeated measure, mission consistency, competence, and attitude toward social enterprise for donation likelihood judgments (F(1, 122) = 7.66, p < .01), in support of H4 (for means, see Table 3).
Study 3: Donation Likelihood as a Function of Time (Pre- vs. Postexposure to the Focal Scenario), Competency, Mission Consistency, and Dispositional Attitude Toward Social Enterprise
To further test the validity of this interaction, we also performed a dummy-coded regression on donation likelihood judgments at time 2, with mission consistency and competency as dummy-coded independent variables and attitude toward social enterprises as a continuous independent variable. Before the analysis, we mean-centered the attitude toward social enterprises measure to reduce multicollinearity between the main effect and interaction terms (Cohen et al. 2003). As we expected, there were significant main effects for mission consistency, competency, and attitude toward social enterprises (all t's > 2.20, ps < .05) and a significant interaction among mission consistency, competency, and attitude toward social enterprises (b = .218, t = 2.79, p < .01).
We predicted that when dispositional attitude toward social enterprises was low, donation likelihood and intentions would be lower but that those high in attitude toward social enterprise would be more forgiving of mission inconsistency or incompetence. This pattern of results emerged. Planned contrasts on donation likelihood judgments at time 2 showed that for those high in attitude toward social enterprises, donation likelihood did not decrease significantly when the commercial activity was either mission inconsistent or incompetent (all ps > .05); rather, this occurred only in the condition of both inconsistency and incompetence (M = 4.43 vs. 2.07; t = 4.22, p < .01). Conversely, for those low (vs. high) in attitude toward social enterprises, donation likelihood judgments were significantly lower when the social enterprise was incompetent, regardless of whether the mission was consistent (M = 3.94 vs. 2.53; t = 2.50, p < .02) or inconsistent (M = 3.94 vs. 2.21; t = 2.591 p < .01; see Figure 1). Thus, those high in attitude toward social enterprises had higher donation likelihood judgments overall, and they were more forgiving of a mission inconsistent or incompetent social enterprise.
Study 3: Donation Likelihood at Time 2 as a Function of Dispositional Attitude Toward Social Enterprise, Competency, and Mission Consistency
Attitude toward the Specific Social Enterprise
To further explore the influence of mission consistency, we conducted a post hoc mediation analysis of attitude toward the specific social enterprise (i.e., the catering business or the check-cashing and loan service business). Although we did not hypothesize an effect a priori, we believed that the underlying attitude toward the nonprofit's specific commercial activity would be inherent in judgments of mission consistency. Thus, an important additional research question involves the potential mediating role of this attitude. To investigate this question, a series of regressions that complement those we reported previously produced outcomes that met all of Baron and Kenny's (1986) mediation criteria for both dependent variables. Specifically, there were significant associations between attitude toward the specific social enterprise (mean centered) and donation likelihood judgments at time 2 (b = .595, t = 8.38, p < .001). A model in which we used both mission consistency and attitude toward the specific social enterprise to predict donation likelihood revealed a significant effect of the proposed mediator (b = .690, t = 7.57, p < .001), but the previously significant effect of mission consistency became nonsignificant (b = .150, t = 1.65, p = .102; see Figure 2). Finally, Sobel tests confirmed that attitude toward the specific social enterprise mediates the effects of mission consistency on donation likelihood (z = 6.22, p < .001). Overall, these results demonstrate that the underlying attitude toward the specific social enterprise mediates the relationship between perceived mission consistency and donation likelihood.
Study 3: Mediating Role of Attitude Toward the Specific Social Enterprise Between Mission Consistency and Donation Likelihood and Donation Intentions at Time 2
The findings from these two studies raise at least two important issues. First, although crowding out may be an important consideration regarding the interrelationship of some sources of nonprofit revenue streams, other issues may also be salient, particularly as they relate to an unsuccessful social enterprise. For example, it might be that donors attribute an unsuccessful commercial activity to the competency of the organization. As such, an unsuccessful commercial activity might reflect negatively on the organization as a whole. Further research is needed to understand how and why donors respond to negative outcomes of social enterprise. Second, the results provide additional insights into the relationship between social enterprises and donors. The results suggest that when the social enterprise is made salient, donors, on average, react negatively in terms of likelihood and intention to donate. This negative reaction is exacerbated when donors perceive the commercial activity of the social enterprise as inconsistent with the social mission of the organization and as unsuccessful in terms of profitability.
Discussion
Overview of the Findings
This research begins to address an important issue in the field of social entrepreneurship about the impact of the initiation of a social enterprise on donation likelihood and donor intentions. The results from the three studies contribute to an increased understanding of the relationship between social enterprises and funding by exploring how donors react to different types of social enterprises and the role of attitudes toward social enterprise in moderating these relationships. Qualitative interview data suggest that the perception of fit, the perception of entrepreneurial competence, and the attitude toward social enterprise all affect donor intentions and behavior. The results of experiments across two sampling contexts show that the introduction of a social enterprise has a negative impact on donor likelihood and intentions but that this impact can be somewhat mitigated when donor perceptions of mission fit, entrepreneurial competence, and attitude toward social enterprise are high. These findings contribute to the emerging literature on social entrepreneurship and the use of social enterprises by nonprofits in four important ways.
First, although prior research on crowding out has found a negative relationship between donations and nonprofit use of commercial activity, this research provides evidence of the directionality of the relationship. Specifically, the findings suggest a bidirectional relationship between the two variables. In other words, the initiation of a commercial activity may result from donor response (reduced donations) as Segal and Weisbrod (1998) report, but our research indicates that the initiation of a commercial activity may also affect future donor response. That is, the initiation of commercial activity may be either the response to or the impetus for donor reaction.
Second, across all scenarios, the average behavioral reaction by donors to social enterprises was negative; they reduced the level of their donations. Prior research on the relationship of commercial activity and donations with American Red Cross chapters finds a crowding out of $3.59 of donations for every $1.00 in commercial profit, suggesting that this result “is reasonable only if donors strongly dislike profit-making activities by the American Red Cross” (Kingma 1995, p. 29). On average, our results indicate that donors view social enterprises negatively, but there are many things organizations can do to mitigate these negative implications. However, additional research is needed to understand this issue more fully. In our exploratory qualitative interviews, we found some evidence in support of this view. The qualitative data also raise important situational variables that may prove useful in future inquiry particularly as they relate to how commercial activity may “crowd in” donations. For example, the executive directors specifically identified the role of donor fatigue and donor's prior business experience as variables that may affect donor support of nonprofit engagement in commercial activity.
Third, our research also indicates that donors may have predispositions or attitudes toward social enterprise that may explain why some favor and others reject such activity. Although such questions were beyond the scope of this research, additional research is needed to fully understand the situational conditions that affect donor reactions to social enterprise.
Fourth, this research suggests that donors are likely to continue to support nonprofits engaging in social enterprises that they perceive as competent. That is, donors may perceive nonprofit organizations with successful commercial strategies as practicing sound business and thus deserving continued support. As the interest in social enterprise continues to grow, it is necessary to understand the conditions under which donors may react differently to commercial activities. Further research is necessary to articulate this relationship clearly.
Implications for Marketing, Social Entrepreneurship, and Public Policy
Given the increasing scope and nature of social enterprise, the results of our research have several important implications for those engaged in social entrepreneurship. For nonprofit executives and board members, our results suggest that they should carefully consider social enterprises before initiating for-profit activities to prevent negative reactions from the individual donor community. The growth in social enterprise and social entrepreneurship, spurred in part by conferences, books, and successful examples, should be tempered to ensure the appropriate rationale for commencing a social enterprise (Foster and Bradach 2005).
Some observers have suggested that public policy should discourage social enterprises and encourage donations (Weisbrod 1998). Instead, we encourage careful consideration of social enterprises by focusing on the social enterprises that directly contribute to either the overall mission of the organization (e.g., through job training skills) or the commercial chances of success. In addition, our qualitative data suggest that organizations should communicate and market to donors, especially larger donors, about the potential profits available from social enterprises, which increase the social return on investment and potentially reduce the requests for future charitable contributions. From a marketing perspective, organizations that participate in social enterprise should segment their target market of donors more effectively and design specific communication strategies to both the individual and institutional (e.g., venture philanthropists) donors who are more likely to support social enterprise (Harvey 1990).
A similar issue is important for individuals and organizations that provide funding for social enterprise. At present, there is a major shortage of funding for social enterprises. However, as foundations and governmental organizations provide start-up and continued funding for social enterprises, the interdependence of alternative sources of revenue may be an important consideration for continued funding. Although we found evidence of crowding out of donations, we did not explore the degree to which this may occur. In many economic studies, the degree of crowding out is partial rather than full, suggesting that the increase in $1.00 of funding from one source results in the reduction of funding from another source (Brooks 2009). However, the reduction is often much less than $1.00, suggesting that the organization is still financially better off with the initiation of new activities even after some crowding out occurs. In addition, individuals and organizations that provide funding for social enterprises may need to provide appropriate business evaluation, training, and other forms of capital to increase the chances of success and scaling of the social enterprises (Bloom 2009; Bloom and Smith 2010).
Insights from the depth interviews also suggest that a nonprofit's involvement with commercial activities affects donation size. Contrary to crowding out, several executive directors identified how some large donors were particularly attracted to commercial activity, and donations actually increased as a result of the nonprofit's participation in commercial activities. As such, further research should examine how different sets of donors react to commercial activity. It might be that the average donor reacts negatively to commercial activities, but a few larger (or new) donors making large contributions may more than compensate for the loss of donations from smaller donors. This issue suggests that both the likelihood and the magnitude of donations may be affected when nonprofits participate in commercial activities.
Limitations and Further Research
As with all designs, our research has several limitations. One limitation is the extent to which we capped the donation size of potential donors. Total donations are a function of the number of donors and the size of their donations. In our quantitative research, we limited the ability of a single donor to affect the size of the donations by placing a ceiling on the size of the donation. Yet our qualitative data imply that such a ceiling may be an artificial constraint in the real world. For example, capital campaigns are often disproportionately affected by a small number of donors. As such, one important direction for research would be to understand the reaction to commercial activity of nonprofits by high net worth donors.
Another limitation of our research pertains to the generalizability of our findings. Given the use of controlled experiments, and specifically the use of only two focal scenarios that were diametrically opposed in terms of mission consistency, we cannot necessarily generalize our findings about the effects of social enterprise to what might actually happen in a field setting. Given the relatively limited knowledge of social enterprise, we employed controlled experiments to limit the number of alternative explanations. Further research is needed to test our results in real field settings. However, two issues did improve the overall research design. First, we moved beyond the student sample and used working professionals in Study 3. Second, we grounded our work and the identification of relevant constructs through in-depth interviews with executive directors of nonprofits that operated a social enterprise.
The use of a repeated measure design format may also be a limitation of our quantitative studies. Although we believed that collecting baseline measures of donation likelihood and donation intentions was appropriate to mitigate individual differences in donation proclivities and would provide a more direct measure of our manipulations, thus making a repeated measure study the appropriate choice (Greenwald 1976), the potential for demand or cueing bias remains. Although this research did not directly measure participants’ potential suspicion to questions, postexperimental debriefing did not indicate or suggest any such problem.
Given the relatively nascent work in the area of social enterprise and donations, another limitation of our research is the testing of mediators and moderators of the identified relationships. Our qualitative data suggest several potential avenues for further research in this area, including donor fatigue and organizational identification. One step in this direction was the introduction of attitude toward the social enterprise. Extensive application exists for this type of measure to assess a wide range of stakeholder attitudes, including board members, employees, and institutional funders. Because some donors can be particularly attracted to or turned off by social enterprises, understanding the conditions under which donors may be more (less) accepting of social enterprises is an important direction for research.
At a more theoretical level, the exact processes through which commercial activities affect donor likelihood and intention require additional research. In this research, we examined perceptions of fit and competence as two theoretical lenses to understand this relationship. These ideas highlight the need for additional consumer research on the cognitive processes through which donors make donation decisions.