The Role of Formal Third-Party Endorsements and Informal Self-Proclaiming Signals in Nonprofit Reputation Building

Abstract

In this study, we experimentally test the impact of a formal signal (a third-party certificate) and an informal signal (self-proclaimed management quality with respect to social entrepreneurship) on stakeholder supportive intentions and perceived organizational effectiveness. Our study confirms the social entrepreneurship advantage, but we find no proof of a convincing effect from the formal signal. However, complementary analyses and additional testing of control variables add new perspectives on the relative importance of the social entrepreneurship advantage and on potential moderators that could better explain in future studies the varying effects and specific contextual elements that influence formal and informal reputation-building signals.

Keywords

nonprofit reputation building formal signals informal signals social entrepreneurship advantage certification experiment

Introduction

Currently, one of the key challenges nonprofit organizations (NPOs) face is the lack of trust from external stakeholders. One major reason for this lack is information asymmetries between stakeholders and the organization (Haski-Leventhal & Foot, 2016; Sanzo, Álvarez, Rey, & García, 2015; Szper & Prakash, 2011). Stakeholders who provide resources can hardly control whether their contributions are allocated appropriately (Gugerty, 2009). On top of that the “unclear ownership structure . . ., the potentially weak disciplining power . . .” (Gugerty, 2009, p. 246) and few binding reporting regulations increase the (perceived) lack of accountability of NPOs even more (AbouAssi, 2015). Stakeholders cannot easily distinguish between high- and low-quality NPOs and can therefore develop a general distrust in all NPOs (Gugerty, 2009; Szper & Prakash, 2011; Willems, 2016). This emerging perspective on mistrust in NPOs is in contrast with the earlier claims of Hansmann (1987) that the NPO format in itself is a signal of trust for beneficiaries and other stakeholders. Hence, within the context of NPOs, various types of signals might influence the extent that trust or mistrust in organizations is fueled.

Both practitioners and academic researchers have been analyzing and discussing the question on how NPOs can actively find appropriate countermeasures to maintain and “restore public trust in the overall effectiveness, efficiency and honorableness of the sector” (Slatten, Guidry, & Austin, 2011, p. 114). Literature suggests the use of best practice approaches (Slatten et al., 2011, p. 125) and reputational signals that “provide credible information on nonprofit quality” (Gugerty, 2009, p. 246) to restore the accountability, transparency, and effectiveness of NPOs (Coule, 2015; Hyndman & McConville, 2016; Tremblay-Boire & Prakash, 2015; Willems, Jegers, & Faulk, 2016). Within this context, NPOs often rely on and/or mimic common practices and follow widespread trends that are considered favorable by the broad public, such as third-party certification and/or they (re)brand the organizational approach in new managerial paradigms (e.g., social entrepreneurship, shared values, etc.).

Within this ongoing research debate, this study aims at clarifying the effect of reputational NPO signals on stakeholder support intentions. In particular, we look at two types of signals: (a) informal, self-proclaiming signals and (b) formal, third-party endorsement signals. Second, the two types of reputational signals are tested in an experiment with respect to their impact on stakeholder support intentions and perceived organizational effectiveness. For this, we replicate the experiment by Andersson and Self (2015) and add important new elements with respect to (a) the distinction of self-proclaiming signals and third-party endorsements, (b) the interaction of both types of signals, (c) additional dependent variables focusing on stakeholder support intentions, (d) the relative importance of the signals compared to the control variables, and (e) the different context in which this experiment is conducted. This research note thus answers the call for more replication studies in nonprofit research (Helmig, Spraul, & Tremp, 2012).

Formal and Informal Reputation Signals

The formal signals focused on in this study are signals that are validated externally, such as third-party certifications that an organization can use in its communication to stakeholders (Tremblay-Boire, Prakash, & Gugerty, 2016). It is a form of accreditation, enabling NPOs to differentiate themselves from other organizations “on the basis of credibility and performance” (Slatten et al., 2011, p. 113), and serves as an indicator for managerial competence verified by an external third party. This way, organizations are able to improve their reputation and achieve legitimacy (Rao, 1994). The possible usefulness of such formal signals can be explained based on various theoretical perspectives. Agency theory, for example, argues that information asymmetries between a principal and an agent can lead to conflicts between the two parties, as the agent might not behave in the interest of the principal but instead makes decisions that serve his/her own benefits (Szper & Prakash, 2011; Van Puyvelde, Caers, Du Bois, & Jegers, 2012). This theory can be applied to the relationship between an NPO and its external stakeholders (Van Puyvelde et al., 2012), where potential donors fear the misallocation of their contribution as they have not the full information regarding the NPO’s distribution of resources. Therefore, a certificate that has been verified by an external and independent institute could serve as a way to decrease information asymmetries to external stakeholders (Slatten et al., 2011).

In addition, within institutional theory, the widespread accreditation through certificates can be explained. In a constantly changing environment, organizations seek to retain legitimacy to survive. Through isomorphism, organizations become more similar to each other as they strive to meet external standards and assessment criteria (Siddiki & Lupton, 2016). From this perspective, the organization’s external environment is able to impact the organization internally, and accreditation standards or certifications can provide widespread legitimacy to the organization (Slatten et al., 2011). Formal signals such as third-party certification are thus considered to be appropriate means and have become necessary conditions to signal accountability and quality of NPOs.

In contrast, informal signals are less clearly defined, and interpretations on their content might be very subjective. An example of informal signal is the (self-) proclamation of a specific quality such as being “social entrepreneurial,” which is also the signal focused on in this study. Although there is no clear definition of what social entrepreneurship actually means (Frank & Shockley, 2016), the concept is promoted as an effective practice for NPOs to be successful. Andersson and Self (2015, p. 2719) describe it as a way of “being proactive, willing to bear risk, challenge norms, and introduce novel ideas and solutions with the goal of enhancing its social mission and ultimate sustainability.” Others connect the term to more revenue-generating business activities while also achieving social, cultural, community, and/or environmental outcomes. Regardless of the absence of a globally agreed definition, “social entrepreneurship and the need to be more entrepreneurial have become [. . .] a highly regarded best practice approach for nonprofits to solve social issues, survive in a competitive marketplace, and generate social impact.” (Andersson & Self, 2015, p. 2720). This implies that organizations that proclaim to be social entrepreneurial are regarded to be different from traditional NPOs (Dart, 2004) yet morally legitimized by being “clearly (connected) to social values in the environment” (Dart, 2004, pp. 419-420). Being social entrepreneurial is thus considered to be positively impacting an NPO’s success (Andersson & Self, 2015) and therefore might be considered as a signal for quality by external stakeholders (Frank & Shockley, 2016; Grieco, Michelini, & Iasevoli, 2015; Raith & Starke, 2017).

In this study, we test whether the presentation of a third-party certificate (formal signal) and proclaiming to be social entrepreneurial (informal signal) have a positive impact on stakeholder support intentions toward an NPO. Furthermore, we test whether the combination of both signals leads to an even stronger positive impact on stakeholder support intentions. Confirming this interaction effect would indicate an amplifying mechanism when both formal and informal signals are combined in the overall reputation-building strategy of an NPO. Our main dependent variable of interest is stakeholder support intentions. To replicate the exact findings of Andersson and Self (2015), at least for the informal signal on proclaiming to be social entrepreneurial, we also tested the impact of both signals and their interaction on perceived organizational effectiveness. By testing both supporting intentions and perceived effectiveness, we can test the persistency throughout the presumed chain of effects, where signals influence perceived organizational effectiveness, which in turn and/or in combination influences supporting intentions (Mews & Boenigk, 2012; Sarstedt & Schloderer, 2010; Willems, Boenigk, & Jegers, 2014). We thus test the following hypotheses:

Hypothesis 1: A third-party certification has a positive effect on (a) stakeholder support intentions and (b) perceived organizational effectiveness.

Hypothesis 2: Proclaiming to be social entrepreneurial has a positive effect on (a) stakeholder support intentions and (b) perceived organizational effectiveness.

Hypothesis 3: The combination of proclaiming to be social entrepreneurial and a third-party certification has an even stronger positive effect on (a) stakeholder support intention(s) and (b) perceived organizational effectiveness.

Method

An online experiment has been conducted in which participants were asked to complete a survey after viewing a fictional NPO profile with two between-subjects manipulations (2-by-2 design). Data have been collected using an online platform (Qualtrics) and was mainly targeted at students and (young) professionals. A group of three students from a master’s course on Public and Nonprofit management took the lead in distributing the survey among fellow students who did not participate in the course themselves, friends, and former colleagues, as these students had earlier work experience. Such sampling is not representative for the full population or for particular predefined subgroups. However, given the fact that the focus of this study is on the distinct effects of experimental treatments, we preferred a cost- and time-efficient sampling over complete representativeness. Descriptive statistics are given in Tables 1 and 2 to make replication and comparison in further studies possible. In total, 191 completed surveys were obtained, and statistical analyses were done with SPSS.

Table 1.

Descriptive Statistics for Categorical Variables, for Overall Sample and Chi-Square Tests for (Un)relatedness With Experimental Treatments.

	Overall sample (N = 191)		Treatment: Certificate		Treatment: SE
	No	Yes	χ²	p	χ²	p
DV: Would you donate to ECOverse?	138	53	8.31*	.004	2.15	.143
DV: Would you volunteer for ECOverse?	127	64	0.59	.443	4.39	.036 *
DV: Would you do an internship at ECOverse?	125	66	0.03	.875	4.28	.039 *
MC: Did you see a quality certificate in the explanation of ECOverse?^a	48	53	45.66*	.000	3.14	.208
Have you ever donated to a nonprofit organization before?	55	136	3.76	.052	0.23	.630
Have you ever volunteered for a nonprofit organization before?	72	119	0.03	.866	0.04	.839
Have you ever done an internship at a nonprofit organization before?	151	40	0.22	.640	0.01	.924
What is your gender?			3.33	.068	1.59	.207
Female		90
Male		101
In which country do you reside?			2.67	.102	2.00	.157
Germany		175
Other^b		16
Please indicate your current main occupation			3.74	.291	0.47	.926
Student		133
Employee		50
Self-employed		5
Other		3
Please indicate your level of education			8.50	.075	1.35	.853
High school degree		22
Higher education (3 years)		41
Higher education (4 or 5 years)		65
Higher education (6 or more years)		56
Not applicable		7

Note. SE = social entrepreneurial; DV = dependent variable; MC = manipulation check; other variables are control variables.

p < 0.05.

The third answer category was I don’t know/I don’t remember, and 90 respondents choose this option.

All other countries are grouped in this category; these were (almost) all single observations in Austria, Canada, Egypt, Georgia, Ireland, Japan, Luxembourg, Malaysia, Moldova, Poland, Spain, Taiwan, United Kingdom, and United States.

Table 2.

Descriptive Statistics for Continuous and/or Scale Variables, for Overall Sample and t Tests for (Un)relatedness With Experimental Treatments.

	Overall sample				Treatment: Certificate							Treatment: Social entrepreneurship
						µ	SE	Lev. test		t test			µ	SE	Lev. test		t test
	µ	SD	Minimum	Maximum		µ	SE	F	p	t	p		µ	SE	F	p	t	p
DV: Interest to donate	−0.53	1.76	−3	3	No Cert.	−0.63	.17	0.44	.51	−0.80	.43	No SE	−.77	.19	0.50	.48	−1.86*	.07
DV: Interest to donate					Cert.	−0.43	.19					SE	−.30	.17
DV: Interest to volunteer	−0.38	1.86	−3	3	No Cert.	−0.29	.19	0.21	.65	0.66	.51	No SE	−.78	.19	1.57	.21	−2.96*	<.01
DV: Interest to volunteer					Cert.	−0.47	.19					SE	.00	.18
DV: Interest to do internship	−0.55	1.99	−3	3	No Cert.	−0.42	.20	2.15	.14	0.93	.35	No SE	−1.05	.20	0.38	.54	−3.43*	<.01
DV: Interest to do internship					Cert.	−0.69	.21					SE	−.09	.20
DV: Perceived effectiveness	0.26	1.35	−3	3	No Cert.	0.32	.14	0.01	.91	0.65	.51	No SE	.04	.15	0.32	.57	−2.12*	.04
DV: Perceived effectiveness					Cert.	0.19	.14					SE	.45	.13
MC: Perceived SE	0.68	1.28	−3	3	No Cert.	0.73	.12	3.23	.07	0.62	.54	No SE	.45	.14	2.98	.09	−2.43*	.02
MC: Perceived SE					Cert.	0.62	.14					SE	.89	.12
Age	26.90	7.05	18	69	No Cert.	26.38	.68	1.36	.24	−1.02	.31	No SE	27.70	.83	2.63	.11	1.52	.13
Age					Cert.	27.43	.76					SE	26.15	.61

Note. Age is a control variable. Lev. test = Levene’s test for equality of variances; t test = t test for equality of means; μ = mean; DV = dependent variable; No Cert. = no certification; SE = social entrepreneurial; MC = manipulation check. *p < 0.05.

The questionnaire started with a description of a fictional environmental NPO called ECOverse. The profile description of ECOverse consisted of two sections: (a) the general organizational background and specific activities and (b) additional information for the two treatments (see Figure 1). For the treatment with respect to formal signals, a fictional certification has been designed and was added to the organizational description. For the informal signaling, the word “social entrepreneurial” was added three times within the overall organizational description. As a result, this experiment had four different profile descriptions: (a) Group A without any manipulation, that is, with only the basic information on the organization; (b) Group B with a formal signal, that is with the certification symbol and text added; (c) Group C with the informal signaling; and (d) Group D with both the formal and informal signaling. Respondents were randomly assigned to one of the four groups. Randomization happened with the random question functionality of Qualtrics surveys (A: 40, B: 57, C: 52, D: 42). With 40 as lowest cell value in this experimental design, an effect size of about 0.64 standard deviations or larger can be detected (with 0.05 probability of Type I error, and 0.20 probability for a Type II error). Hence, with this design, we aimed at larger cell counts than the main study that we replicate (Andersson & Self, 2015), which is also higher than the minimal cell count of 20 observations, proposed by Simmons, Nelson, and Simonsohn (2011). In contrast, it has to be noted that this design can only detect relatively robust effects, which in particular serves a practical purpose. In other words, it has only limited practical value to recommend investing scarce organizational resources when potential effects of formal and informal signaling are only marginal.

After the organizational description, respondents could indicate their potential supporting intentions. Respondents could indicate for three distinct questions, how interested they were to donate, volunteer, and do an internship at ECOverse. We used a 7-point semantic difference scale with labels not at all interested and strongly interested, and the seven answer options were numbered from “−3” to “3.” Perceived effectiveness was measured on a 7-point Likert-type scale, and answer options were labeled and coded (Valeau, Willems, & Parak, 2016; Willems et al., 2016): very ineffective (−3), ineffective (−2), somewhat ineffective (−1), neither effective nor ineffective (0), somewhat effective (+1), effective (+2), and very effective (+3).

As measures of control, we also asked with three straightforward yes–no questions whether the respondent would donate, volunteer, and/or do an internship at ECOverse. This can give an additional insight into the supporting intentions as this is closer to the actual decision to be made by supporting stakeholders. As manipulation checks, respondents were asked on a next survey page whether they noticed a quality certificate in the ECOverse profile description (“Yes,” “No,” “I don’t know/I don’t remember”) and how social entrepreneurial they considered ECOverse, on a 7-point semantic difference scale (very low to very high), with the answer options numbered from −3 to 3. Closing questions focused on whether respondents had been donating, volunteering, or doing an internship at an NPO before and on respondents’ demographics like gender, age, country of residence, occupation, and education. These questions were asked to provide a detailed sample description and to make comparison with other samples possible (e.g., for further replication). Sample descriptions are given in Tables 1 and 2, along with information about the random distribution over the experiment treatments. t tests and chi-square tests show that no significant overrepresentation occurred in one of the treatment groups for (a) earlier donations, (b) earlier volunteering, (c) earlier internship, (d) gender, (e) country, (f) occupation, (g) level of education, or (h) age.

Results and Discussion

The chi-square tests in Table 1 suggest a higher likeliness to donate (yes/no decision), when a certificate is presented (χ² = 8.31, p = .004). A certificate has no effect on the decision to volunteer and/or to do an internship. In contrast, the social entrepreneurial signal results in a higher likeliness to volunteer and to do an internship. The t test in Table 2 also shows that the interest to volunteer and to do an internship is higher as a result of the informal SE treatment but not of the formal certificate treatment. However, and in contrast to the dichotomous measure of donation intentions, for the continuous measure of donation interest, no effect is found for the formal signal but for the informal signal. Hence, Hypothesis 2a, which expressed an effect of an informal SE-signal, is supported regardless the measure used (yes/no decision or 7-point Likert-type interest scale) for (a) volunteering, (b) doing an internship, and, only partially for, (c) donating. In contrast, Hypothesis 1a is not convincingly supported by the results of this study.

The t test of perceived effectiveness (Table 2; t = 2.98, p = .02) shows that the SE treatment results in a higher perceived effectiveness, supporting Hypothesis 2b. No support is found for Hypothesis 1b, as perceived effectiveness does not differ for the certificate treatment.

Table 3 shows the parameter estimates of four general linear model estimations, with both treatments as factors and the control variables as covariates. The respective dependent variables are (a) interest to donate, (b) interest to volunteer, (c) interest to do an internship, and (d) perceived effectiveness. This analysis can clarify whether a potential interaction effect exists (Hypothesis 3) even when no main effect for the formal certificate signal is found for this experiment, and it can clarify the relative explaining relevance of the SE treatment compared to control variables. However, the earlier found SE treatment effects are not large enough to be detected with the available statistical power in this analysis. Moreover, and despite the established unrelatedness of treatments and background variables (see Tables 1 and 2), the single effect detected (SE on internship interest) is not found when control variables are added to the model.

Table 3.

Parameter Estimates of General Linear Model, Explaining (a) Interest to Donate, (b) Interest to Volunteer, (c) Interest to do an Internship, and (d) Perceived Effectiveness.

	Interest to donate			Interest to volunteer			Interest to do an internship			Perceived effectiveness
	B	SE	p	B	SE	p	B	SE	p	B	SE	p
Basic model
Constant	−.80	.28*	.00	−.60	.29*	.04	−.93	.31*	.00	.18	.21	.41
Treatment (Cert.)	.05	.37	.89	−.32	.38	.40	−.23	.41	.58	−.23	.28	.41
Treatment (SE)	.29	.36	.42	.53	.38	.16	.85	.40*	.03	.25	.28	.38
Interaction (Cert. × SE)	.43	.51	.40	.49	.54	.36	.18	.57	.75	.31	.39	.43
Model summary
R²	.03			.05			.06			.03
Adjusted R²	.01			.03			.05			.01
Lev. test	.47			2.01			.47			.87
Significance (Lev. test)	.70			.11			.70			.46
Model with control variables
Constant	−.05	.51	.92	.45	.51	.38	.33	.55	.54	.71	.39*	.07
Treatment (Cert.)	−.13	.36	.72	−.64	.36	.08	−.41	.38	.29	−.28	.28	.31
Treatment (SE)	−.03	.36	.94	−.01	.35	.98	.40	.38	.29	.03	.27	.91
Interaction (Cert. × SE)	.66	.50	.19	1.07	.50*	.03	.67	.53	.21	.53	.38	.17
Age (centered)	−.01	.02	.67	−.02	.02	.22	−.01	.02	.61	−.01	.02	.45
Female (ref.)
Male	−.90	.25*	.00	−1.15	.25*	.00	−1.21	.27*	.00	−.98	.19*	.00
Student (ref.)
(Self-)employed	−.31	.31	.31	−.49	.31	.12	−.88	.33*	.01	.00	.24	.98
Education
High school (ref.)
Higher education: 3 years	−.33	.44	.45	−.20	.44	.66	.15	.47	.75	−.30	.34	.37
Higher education: 4 or 5 years	.06	.42	.88	−.06	.42	.89	.24	.44	.59	−.38	.32	.24
Higher education: 6+ years	−.50	.43	.24	−.63	.43	.14	−.43	.46	.35	−.29	.33	.38
Not applicable	−1.08	.75	.15	−1.81	.75*	.02	−1.50	.79	.06	−.01	.57	.99
German resident (ref.)
Not a German resident	.67	.45	.13	.78	.45	.08	.53	.48	.27	−.16	.34	.65
Donated before	.52	.28	.06	.04	.28	.89	−.28	.29	.34	.58	.21*	.01
Volunteered before	−.33	.26	.21	.31	.26	.24	.07	.28	.81	−.12	.20	.55
Did an internship before	.18	.31	.57	.12	.31	.70	.15	.33	.64	.13	.24	.58
Model summary
R²	.18			.28			.28			.20
Adjusted R²	.12			.22			.23			.13
Lev. test	1.07			1.17			.47			1.32
Significance (Lev. test)	.36			.32			.70			.27

Note. Age was centered on the mean, that is 26.9 years. Employment and self-employment were merged in a single category, versus students. Cert. = certification; SE = social entrepreneurial; Lev. test = Levene’s test for equality of variances.

p < 0.05.

The results of this experiment thus indicate a significant but rather (relatively) small social entrepreneurship effect, confirming the findings of Andersson and Self (2015). However, at least two additional remarks have to be made. First, the significant effect of the self-proclaimed, informal signal only explains a very small proportion of the variance in stakeholder support intentions (Table 3; adjusted R² values range between.01 and.05). Second, the introduction of additional variables shows that these results are dependent on potential additional elements. Further research on the social entrepreneurship advantage (Andersson & Self, 2015) can thus incorporate this concept in a broader set of antecedents, effects, and potential moderators. Such elaboration can provide additional insights in its relative importance but also in whether the social entrepreneurship advantage works differently for different stakeholder types and in different contexts.

In contrast to the SE treatment, this experiment does not support a formal certification effect. However, a critical note has to be made on this formal signal treatment based on the manipulation check. We asked respondents whether they remembered seeing a certificate, and we provided answer options: (a) “No,” (b) “Yes,” and (3) “Don’t know/Don’t remember.” A large proportion of people did answer that they did not know or could not remember (47.12%). This suggests that little attention was given to a visual representation of a certificate in the description of the organization, and, therefore, this treatment was not successful from an experiment design point of view. However, there was no overrepresentation of one of the two treatment groups in this subgroup of nonremembers (chi-square test, p = .247, N = 191), and for those who did give a clear choice (“Yes” or “No”), a majority did remember the information correctly (chi-square test, p < .001, n =101). Moreover, as there would be no academic nor practitioner value in giving certification information in such an extreme way beyond what any real organization would do, only to have a more convincing treatment effect in our experiment, we decided in our research design for a presentation of a certification close to a real-life situation. Figure 1 shows how the certificate took up a substantial part of the information provided to respondents. Hence, not finding an effect of a certification in fact is in line with earlier studies that have shown and/or suggested that the overall interest to donate is little or not related to the formal information available on the donation accountability of an NPO (Haski-Leventhal & Foot, 2016; Saxton, Neely, & Guo, 2011; Szper & Prakash, 2011).

As a complementary analysis, we tested for the subgroup of respondents that correctly remember (not) seeing a certificate whether a significant difference exist in (a) interest to donate, (b) interest to volunteer, (c) interest to do an internship, and (d) perceived effectiveness.

For interest to volunteer and interest to do an internship, a significant but negative effect is found, meaning that people who did observe the certificate were less interested to volunteer and/or to do an internship (interest to volunteer: difference is 0.851; t = 2.257, p = .027; interest to do an internship: difference is 1.173; t = 2.803, p = .006).

This negative effect is a preliminary finding for a specific subsample of this experiment and therefore needs further testing before it can be given substantial interpretation. Nevertheless, one possible explanation could be that we find this opposite effect because we used a nonexisting certificate. Our original aim was to have no unforeseen confounding factors with respect to reputation and/or recognition of existing certificates. However, respondents who actively took the (absence of a) certificate into account might have had earlier information and knowledge of real existing certificates and therefore might have been critical when observing an unknown or no certificate. An unknown certificate might even have been considered as a dubious signal, resulting in the negative effect for the subsample of people that well remembered seeing it. Further testing of this interpretation should be done by including additional moderating variables that focus on existing certificate reputation and on the extent that respondents are aware of certificates and the related information they signal. Only when this reputation and information signaling is actively associated with a visual certificate, the visual certificate might result in better stakeholder trust. Another potential interpretation is that respondents even consider it as window dressing, which is a practice where organizations focus on signaling quality with substitute information, like a certificate, that in itself has low transparency, rather than focusing on real quality in itself (Kolk & Perego, 2014).

The most remarkable control variable is the substantial gender difference in supporting intentions, where men indicate to be less interested to support the organization compared to women (Table 3; respective coefficients are −.90, −1.15, −1.21, and −.98 for interest to donate, interest to volunteer, interest to do an internship, and perceived effectiveness). This confirms the recent work of De Wit and Bekkers (2016) on gender differences in prosocial behavior, showing that women are more likely to engage in an additional form of prosocial behavior, for example, to diversify their contributions to a broader set of good causes. As De Wit and Bekkers find that men give less often but are more likely to give a larger amount, a further replication and elaboration of our experiment could not only take the willingness to donate but also the amount into account.

Furthermore, respondents who indicate being (self-) employed, compared to students, are also less likely to do an internship (Table 3; −.82, p < .01). This is likely the result of the fact that internships have less practical value for people who already work.

Conclusion

To contribute to the literature on reputation-building practices by NPOs, we aimed at testing the effects of a formal and an informal signal, and their combination, on stakeholder support intentions and perceived organizational effectiveness. With a 2-by-2 experimental design, inspired by the work of Andersson and Self (2015), we tested whether on one hand an accreditation certificate and on the other hand the wording “social entrepreneurial” had a relevant effect. Our analysis could only confirm the social entrepreneurship advantage as found by Andersson and Self but neither the potential effect of a third-party certification nor an interaction effect of both. However, the results of the complementary analyses suggest several concrete avenues for further research. Despite the fact that we could confirm the social entrepreneurship advantage, our results show that they might have limited explanatory power in combination with other variables. From this perspective, there is substantial value in further exploring the potentially more complex interrelatedness, mediation, and moderation of the social entrepreneurship advantage with other relevant variables that explain perceived organizational effectiveness and stakeholder support intentions. Potential research paths to continue can thus focus on research questions: Which stakeholder characteristics are related to, and/or moderated by, the social entrepreneurship advantage? Do particular subdimensions of social entrepreneurship have a differential effect on various types of stakeholder support intentions?

A critical evaluation of our manipulation check on the formal signal suggests either that our signal was not strong enough (despite being realistic) or that potential stakeholder support intentions are not significantly influenced by it. As a result, also for this type of signal, further elaboration and testing are needed. This elaboration could be in the area where widespread recognition of a certificate and previous knowledge by the potential stakeholder are included as potential moderating variables to explaining the difference between a positive or a negative effect of NPO certification.

In sum, this study has thus contributed by providing an extra verification of the social entrepreneurship advantage (Andersson & Self, 2015) and by adding interesting further avenues based on the complementing hypotheses and tests.

Footnotes

Acknowledgements

The authors thank the anonymous reviewers for their constructive way of giving feedback.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Deutsche Forschungsgemeinschaft (DFG)/German Research Foundation (project name and number: Reputations management für NPOs—DFG-WI4542. For additional information,

Author Biographies

Jurgen Willems is a post-doctoral research assistant at the University of Hamburg. His reseach focusses on managerial chalenges in nonprofit organizations, such as reputation management, goverance, and volunteer management.

Carolin J. Waldner is a research assistant at the University of Hamburg. Her PhD dissertation focusses on reputation building in and of nonprofit organizations.

Yasemin I. Dere is a master student in International Business and Sustainability (MIBAS) at the University of Hamburg.

Yuka Matsuo is a master student in International Business and Sustainability (MIBAS) at the University of Hamburg.

Kevin Högy is a master student in International Business and Sustainability (MIBAS) at the University of Hamburg.

References

AbouAssi

(2015). Testing resource dependency as a motivator for NGO self-regulation: Suggestive evidence from the global south. Nonprofit and Voluntary Sector Quarterly, 44, 1255-1273.

Andersson

F. O.

Self

(2015). The social-entrepreneurship advantage: An experimental study of social entrepreneurship and perceptions of nonprofit effectiveness. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 26, 2718-2732.

Coule

T. M.

(2015). Nonprofit governance and accountability: Broadening the theoretical perspective. Nonprofit and Voluntary Sector Quarterly, 44, 75-97.

Dart

(2004). The legitimacy of social enterprise. Nonprofit Management & Leadership, 14, 411-424.

De Wit

Bekkers

. (2016). Exploring gender differences in charitable giving: The Dutch case. Nonprofit and Voluntary Sector Quarterly, 45, 741-761.

Frank

P. M.

Shockley

G. E.

(2016). A critical assessment of social entrepreneurship: Ostromian polycentricity and Hayekian knowledge. Nonprofit and Voluntary Sector Quarterly, 45, 61S-77S.

Grieco

Michelini

Iasevoli

(2015). Measuring value creation in social enterprises: A cluster analysis of social impact assessment models. Nonprofit and Voluntary Sector Quarterly, 44, 1173-1193.

Gugerty

M. K.

(2009). Signaling virtue: Voluntary accountability programs among nonprofit organizations. Policy Sciences, 42, 243-273.

Hansmann

(1987). Economic theories of nonprofit organizations. In Powell

W. W.

(Ed.), The nonprofit sector: A research handbook (pp. 27-42). New Haven, CT: Yale University Press.

10.

Haski-Leventhal

Foot

(2016). The relationship between disclosure and household donations to nonprofit organizations in Australia. Nonprofit and Voluntary Sector Quarterly, 45, 992-1012.

11.

Helmig

Spraul

Tremp

(2012). Replication studies in nonprofit research: A generalization and extension of findings regarding the media publicity of nonprofit organizations. Nonprofit and Voluntary Sector Quarterly, 41, 360-385.

12.

Hyndman

McConville

(2016). Transparency in reporting on charities’ efficiency: A framework for analysis. Nonprofit and Voluntary Sector Quarterly, 45, 844-865.

13.

Kolk

Perego

(2014). Sustainable bonuses: Sign of corporate responsibility or window dressing? Journal of Business Ethics, 119, 1-15.

14.

Mews

Boenigk

(2012). Does organizational reputation influence the willingness to donate blood? International Review on Public and Nonprofit Marketing, 10, 49-64.

15.

Raith

M. G.

Starke

(2017). Negotiating the organizational effectiveness of social ventures among multiple stakeholders. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 28, 1473–1499.

16.

Rao

(1994). The social construction of reputation: Certification contests, legitimation, and the survival of organizations in the American automobile industry: 1895-1912. Strategic Management Journal, 15, 29-44.

17.

Sanzo

M. J.

Álvarez

L. I.

Rey

García

(2015). Business-nonprofit partnerships: Do their effects extend beyond the charitable donor-recipient model? Nonprofit and Voluntary Sector Quarterly, 44, 379-400.

18.

Sarstedt

Schloderer

M. P.

(2010). Developing a measurement approach for reputation of non-profit organizations. International Journal of Nonprofit and Voluntary Sector Marketing, 15, 276-299.

19.

Saxton

G. D.

Neely

D. G.

Guo

(2011). Web disclosure and the market for charitable contributions. Journal of Accounting and Public Policy, 33, 127-144.

20.

Siddiki

Lupton

(2016). Assessing nonprofit rule interpretation and compliance. Nonprofit and Voluntary Sector Quarterly, 45, 156S-174S.

21.

Simmons

J. P.

Nelson

L. D.

Simonsohn

(2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22, 1359-1366.

22.

Slatten

Guidry

B. N.

Austin

(2011). Accreditation and certification in the non-profit sector: Organizational and economic implications. Organization Management Journal, 8, 112-127.

23.

Szper

Prakash

(2011). Charity watchdogs and the limits of information-based regulation. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 22, 112-141.

24.

Tremblay-Boire

Prakash

(2015). Accountability.org: Online disclosures by U.S. nonprofits. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 26, 693-719.

25.

Tremblay-Boire

Prakash

Gugerty

M. K.

(2016). Regulation by reputation: Monitoring and sanctioning in nonprofit accountability clubs. Public Administration Review, 76, 712-722. doi:10.1111/puar.12539

26.

Valeau

Willems

Parak

(2016). The Effect of attitudinal and behavioral commitment on the internal assessment of organizational effectiveness: A multilevel analysis. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 27, 2913-2936.

27.

Van Puyvelde

Caers

Du Bois

Jegers

. (2012). The governance of nonprofit organizations: Integrating agency theory with stakeholder and stewardship theories. Nonprofit and Voluntary Sector Quarterly, 41, 431-451.

28.

Willems

(2016). Organizational crisis resistance: Examining leadership mental models of necessary practices to resist crises and the role of organizational context. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 27, 2807-2832.

29.

Willems

Boenigk

Jegers

(2014). Seven trade-offs in measuring nonprofit performance and effectiveness. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 25, 1648-1670.

30.

Willems

Jegers

Faulk

(2016). Organizational effectiveness reputation in the nonprofit sector. Public Performance & Management Review, 39, 476-497.