The Sustainability Performance of Social Enterprises in China: The Configurational Impacts of Ecosystems and Revenue Structures

Abstract

Despite the global development of social enterprises (SEs) over the past three decades, how to improve sustainability remains a challenging issue for most SEs. Although SE ecosystems have been recognized as crucial determinants of SE sustainability performance in the current literature, no empirical study has comprehensively examined the relationships among them. Additionally, prior studies have demonstrated that sustainability performance might vary among SEs of different revenue structures or across different contexts, suggesting that more attention should be devoted to the complexity of the causal mechanisms determining SE sustainability performance. To address these gaps in the current literature, this paper examines the complex, divergent, and asymmetric causal links among SE ecosystems, revenue structures, and the sustainability performance of SEs in China by conducting fuzzy set qualitative comparative analysis (fsQCA) of 274 typical cases of SEs. The results revealed alternative configurations for high and low levels of sustainability performance among SEs of different revenue structures. First, the fsQCA results indicated that SE sustainability performance was not determined by the impacts of single components of SE ecosystems but rather by the combined effects of multiple elements. Second, for SEs of divergent revenue structures, causal paths leading to high or low levels of sustainability performance showed notable discrepancies in terms of both number and composition. Specifically, commercial SEs receiving income mainly from market-based earned income were more likely to achieve higher levels of social and financial sustainability because of greater adaptability to SE ecosystems and less environmental dependence. Third, the impacts of different components of SE ecosystems on sustainable performance also varied with SE revenue structures. Three categories of components—policy environment, sociocultural setting, and industrial infrastructure—made more important contributions to SE sustainability performance in both the social and financial dimensions.

1. Introduction

Over the past three decades, social enterprises (SEs), defined as socially entrepreneurial organizations that pursue social aims and generate social value through exploiting market-based opportunities and leveraging business mechanisms [1], have emerged globally. The pursuit of organizational sustainability among SEs has become an area of great interest for academics, practitioners, and policy-makers. Scholars have used different terms to define SE sustainability performance, such as “sustainability” [2,3,4,5,6,7,8], “organizational sustainability or growth” [9,10,11], “sustainability or sustainable performance” [12,13], and “durability” [14]. Most scholars have focused on the social and financial dimensions of SE sustainability, particularly emphasizing the two fundamental ingredients of the concept, namely, consistently meeting social objectives and delivering social impact while maintaining commercial viability and achieving financial sustainability.

Previous studies have identified various factors that contribute to SE sustainability performance and devoted increasing attention to the impact of the SE ecosystem, which is defined as the supportive environment surrounding SEs that comprises a set of coordinated, interdependent actors and factors, enabling SEs to access supportive financial, social and cultural resources [3,15]. The literature has extensively examined the roles of various actors and elements that constitute SE ecosystems, such as government, law, and policy [10,16,17,18,19,20,21,22]; social impact investors [17,20,23]; accelerators and incubators [24,25]; socioeconomic environment [10,18,26,27]; institutional infrastructures [19]; stakeholders and social networks [3,28,29]; and public awareness and recognition [3,18,30]. More recently, growing attention has been given to the relationship between supportive environments and SE sustainable performance in the Chinese context, with a focus on the impacts of government and policy environments [31,32,33,34,35], market support [32], and social support [32,33].

Prior studies have also observed variations in the sustainability performance of SEs with different revenue structures and provided mixed evidence on the relationships between the sustainability performance and revenue streams of SEs, such as government support [10,21,36,37], earned income [38,39], and charitable donations [38,40]. Although there is a growing trend toward revenue diversification among Chinese SEs, the current literature has rarely investigated the impacts of revenue structures on their sustainability. As a notable exception, Yu and Bi proved that the earned income of SEs might amplify the positive contributions of certain types of scaling strategies to the scaling-up of SE social impact in China [41]. Nonetheless, no study has specifically examined the divergent contributions of different revenue sources to SE sustainability performance in the Chinese context.

Overall, the current literature has illuminated our understanding of SE sustainability performance by revealing the influence of both SE ecosystems and revenue structures. However, previous studies are subject to several limitations that are worth noting. Theoretically, despite the large body of literature on the linkages between SE ecosystems and SE sustainability performance, most studies have examined the impacts of different components of SE ecosystems in isolation rather than viewing them as an integrated system in which divergent actors and elements interact in a complex way. Although some scholars have empirically investigated this issue from a holistic perspective, by using composite indicators, such as “ecosystem” [3], “integrated support” [28], or “institutional fragility” [42], these studies have failed to examine the interactions between different components that make up these composite indicators. Moreover, no comparative study has assessed the relative importance of different components of SE ecosystems in contributing to SE sustainability performance. Additionally, current studies have demonstrated that SE sustainability performance might vary among SEs of different revenue portfolios or across different SE ecosystems, suggesting that more attention should be devoted to the complexity of the causal mechanisms determining SE sustainability performance. Methodologically, most empirical studies have utilized conventional quantitative statistical analyses, primarily assessing the net effects of independent variables while neglecting the multifaceted interdependencies among different predictors of SE sustainability performance and the asymmetric real-life relationships. Moreover, previous quantitative studies have typically adopted a deductive methodology yet have not looked beyond the hypotheses to discover new or unexpected causal relationships.

To address these theoretical and methodological gaps in the current literature, this study proposed a more holistic perspective of SE ecosystems. It performed a fuzzy-set qualitative comparative analysis (fsQCA) of 274 SEs to investigate the complexity of the causal links among SE ecosystems, revenue composition, and SE sustainability performance. The fsQCA identified divergent configurations for high and low levels of sustainability performance among commercial, donative, and government-supported SEs, which rely mainly on market-based earned income, charitable donations, and government support, respectively. As the first fsQCA study on the causal mechanisms of SE sustainability performance in the Chinese context, this study makes unique academic contributions to the SE literature in terms of its analytical perspective, research methodology, and variable measurement. Additionally, the findings provide important implications for practitioners, policy-makers, and other stakeholders who endeavor to promote the sustainable development of SEs.

The remainder of this paper is organized as follows. In Section 2, we describe the theoretical background and the configurational framework for fsQCA. In Section 3, we explain sample selection, measurements, and calibration. In Section 4, we present the main fsQCA results. In Section 5, we discuss the findings in light of the literature to outline their main theoretical contributions and practical implications. In Section 6, we draw conclusions, discuss the limitations of this study, and suggest directions for future research.

2. Background Literature and Configurational Framework

2.1. SE Ecosystem and Sustainability Performance

Previous studies have extensively considered the external environment surrounding SEs as an important determinant of SE sustainability performance [3,10,17,18,19,20,21,22,23,24,25,28,29,30], despite the use of different theoretical perspectives, such as resource dependency theory, the new institutionalism perspective, and the resource-based view. Most studies have used the term “SE ecosystem” to capture the impact of the external environment. For example, in their pioneering studies on the ecosystem of social entrepreneurship, researchers at the Center for the Advancement of Social Entrepreneurship (CASE) at Duke University emphasized that the success of social entrepreneurship requires a healthy institutional and social environment to support the practice. According to CASE, the ecosystem of social entrepreneurship includes two main components: capital infrastructure and context-setting factors. The former provides various resources, such as financial, human, social, political, and intellectual capital, which are essential for the success of social entrepreneurs and their organizations, whereas the latter forms environmental conditions, such as public policy and politics, media, and economic and social conditions, which define the legitimacy of SEs [43].

The current literature has devoted increasing attention to the impacts of various actors or elements in SE ecosystems on SE sustainability performance. The first stream of literature addressed the resource impacts of SE ecosystems, examining how financial, human, social, political, and intellectual capital become available to support the sustainable development of SEs, stemming from the contributions of governments [10,20,21,22], social impact investors [17,20,23], SE accelerators and incubators [24,25], and other stakeholders and social networks [3,28,29]. A second stream of literature addressed the legitimacy of SE ecosystems, analyzing how the regulatory and normative legitimacy of SEs are framed by different factors in macroinstitutional environments, such as law and policies [10,17,18,19], socioeconomic settings [10,18,26,27], and public awareness and recognition [3,18,30].

2.2. Revenue Structure and Sustainability Performance

An extensive body of literature has identified SE revenue structure as a crucial predictor of SE sustainability performance while providing mixed evidence on the contributions of different revenue streams of SEs, such as government support, earned income, and charitable donations. With respect to the impact of government support, studies have focused mainly on South Korea and drawn different conclusions. Kim and Moon demonstrated that government funding contributed positively to the sustainability of the economic and social performance of SEs [37], whereas Choi and Berry argued that government funding led to greater social performance but slower growth in economic performance [36]. Alternatively, Mekkaoui and Loukili concluded in their study on the social and solidarity economy in Morocco that government funding contributed positively to the annual income growth of cooperatives but negatively to their long-term survival [10].

With respect to the influence of earned income, Lu et al. revealed that commercial nonprofits in the U.S. deriving income primarily from charging prices for their services had less chance of dissolution than their noncommercial counterparts [39]. Alternatively, Erpf et al. made a different argument in the Swiss context that the proportion of market funding did not contribute significantly to the economic performance of SEs [38].

With respect to the effect of charitable donations, Erpf et al. noted that donation-based funding had a negative effect on the economic performance of Swiss SEs [38]. Nonetheless, Sahasranamam et al. discovered a positive and significant interaction effect between a founding team’s nonprofit entrepreneurial experience and philanthropic funding on SE economic performance in the global context [40].

2.3. Configurational Framework

Although previous studies have generated a wealth of insights into how SE ecosystems and revenue structures affect SE sustainability performance, several important limitations remain in the current literature. On the one hand, as most studies on SE ecosystems have concentrated on the contributions of single actors or elements, we still lack an understanding of how these divergent actors or elements interrelate and operate as a whole complex system to generate multiple pathways to SE sustainability performance. On the other hand, although the current literature has provided growing evidence on the heterogeneity of the causal links between revenue streams and SE sustainability performance across divergent SE ecosystems in different countries, prior studies remain silent on the combined effects of SE ecosystems and revenue structures and the possible interdependencies between them.

To address these knowledge gaps in the previous literature, this study integrated an investigation of the impacts of SE ecosystems and revenue structures in a single configurational framework. Specifically, by pursuing a fsQCA approach, this study examined the complexity of the causal mechanisms leading to SE sustainability performance, which might be potentially contingent upon the interaction among several factors in SE ecosystems and different revenue streams.

As a set-theoretical technique based on fuzzy-set theory and Boolean minimization [44], fsQCA investigates both within- and cross-case logics [45] to embrace complex causality [46] and, therefore, has several important methodological advantages. First, as an innovative technique bridging the gap between qualitative and quantitative methods, fsQCA attains highly accurate portraits of real-world processes while also enabling high generalizability across other contexts [47]. Second, fsQCA allows researchers to test equifinality, where multiple configurations of causal conditions can lead to the same outcome [48,49,50]. In other words, by using fsQCA, researchers can identify alternative causal paths producing targeted outcomes [45]. Third, unlike regression and other variable-oriented approaches that examine a limited number of two- and three-way interactions among variables, fsQCA investigates the potential interdependency of all antecedent conditions and can reveal additional fine-grained information about causal complexity [51]. Fourth, in contrast to conventional statistical techniques, which focus primarily on the symmetric linear relationships between variables, fsQCA assumes that real-life causal relationships tend to be asymmetrical and, therefore, emphasizes identifying substantially different configurations that predict both the presence and absence of an outcome [44,48,52]. Finally, while conventional quantitative studies are typically hypothesis-testing driven and deductive in design, fsQCA is an inductive, iterative method revealing patterns in the data at the case level that tend to be obscured by statistical analysis [51]. Thus, fsQCA is productive for discovering new and unexpected relationships between configurations and outcomes, improving subsequent theory-building and hypothesis-testing efforts [53].

Taking advantage of the methodological benefits of fsQCA, SE researchers have increasingly applied this method in studies on SE sustainability performance. Some studies focused on identifying the causal configurations determining the SE social sustainability performance, which is measured by different indicators, such as the durability of goal realization and community recognition [14], social effectiveness and institutional legitimacy [52], social sustainability [54], or social value creation [45]. Other studies concentrated on exploring the causal paths affecting the SE financial or economic sustainability performance, which is measured by various indicators such as sustained financial stability, independence, efficiency, profitability [14,52], or growth in revenue, profit, and return on investment [45]. Alternatively, another stream of literature took a holistic view to assess the impact on the SE organizational sustainability performance by using dual-dimensional indicators, such as organizational growth [2,55], organizational survival [56], and internationalization propensity and diversity [57]. Additionally, a small number of studies have also applied fsQCA to analyze SE sustainability performance in the Chinese context by examining the joint effects of business model innovation and legitimacy [58] or the combined contributions of different mission statement components [4].

Despite the growing attention devoted to using fsQCA to investigate SE sustainability performance, this study is the first in the SE literature to examine the complex and conjunctive causality that links SE ecosystems, revenue structures, and sustainability performance through alternative paths. Drawing on the relevant theoretical insights and empirical evidence in prior research, we propose a configurational framework for the fsQCA, as presented in Figure 1.

3. Research Design

3.1. Sample Selection

Theoretical sampling is highly recommended in the fsQCA literature, in which cases should be theoretically defined [44] and purposively selected on the basis of their potential to provide complementary insights into the analysis [59,60,61]. Accordingly, we obtained our sample via a theoretical sampling approach. Specifically, we conducted an online questionnaire survey in 2021 and obtained a database of 368 SEs, which included various types of SEs, such as entrepreneurial nonprofits, cooperatives, work integration social enterprises, public-private partnerships, and socially responsible corporations. Next, to ensure the satisfactory variability of cases across conditions and outcomes [44], we stratified all the cases into three subsamples according to their revenue structures. Then, cases were selected from each of the three subsamples on the basis of the variations in the social and financial sustainability performance. Finally, we obtained a sample of 274 cases that were diverse in terms of organizational size, age, sector, as well as registration form, and region, following simultaneously the principles of theoretical relevance and satisfactory variability. However, the representativeness of the sample must be assessed with caution, as the theoretical sample rather than probability sampling was applied in this study.

3.2. Variable Measurement

3.2.1. Outcome

We identified SE sustainability performance as the outcome variable in this study. In the current literature, SE sustainability performance is often measured in terms of both social and financial aspects. Therefore, we adopted a dual-dimensional conceptual framework and measured SE sustainability performance with two composite indicators: the sustainability of social performance and financial performance, comprising four components, respectively. Table 1 presents the indicators and measures of SE sustainability performance, along with the relevant literature that uses these measures.

Since the two composite indicators were composed of four variables with different units of measurement, we used the “min-max normalization” [62] method to transform all the raw data into those having an identical range [0, 1]. Therefore, we obtained the values of the two composite indicators, ranging from 0 to 1, as the mean values of the four normalized variables.

Table 1. Measurements of outcome variables.

Outcome	Indicators	Measures
Sustainability of social performance	growing beneficiaries [18,58,63,64]	1 = yes; 0 = no
	growing clients [63,64,65]	1 = yes; 0 = no
	accomplishment of social missions [14,45,52]	5-point Likert scale
	satisfaction of stakeholders [14,45,52]	5-point Likert scale
Sustainability of financial performance	financial breakeven [66]	1 = yes; 0 = no
	profit margin [8,66]	revenue minus expenditures and over revenue
	increasing income [5,10,45,56,67]	1 = yes; 0 = no
	increasing assets [41]	1 = yes; 0 = no

Table 1. Measurements of outcome variables.

Outcome	Indicators	Measures
Sustainability of social performance	growing beneficiaries [18,58,63,64]	1 = yes; 0 = no
	growing clients [63,64,65]	1 = yes; 0 = no
	accomplishment of social missions [14,45,52]	5-point Likert scale
	satisfaction of stakeholders [14,45,52]	5-point Likert scale
Sustainability of financial performance	financial breakeven [66]	1 = yes; 0 = no
	profit margin [8,66]	revenue minus expenditures and over revenue
	increasing income [5,10,45,56,67]	1 = yes; 0 = no
	increasing assets [41]	1 = yes; 0 = no

3.2.2. Conditions

We identified five components of SE ecosystems and three revenue streams of SE revenue structures as the condition variables in this study. The specific measures of all the condition variables and the corresponding literature are shown in

Table 2. Measurements of condition variables.

Conditions	Indicators	Measures
SE Ecosystem
Policy environment	a favorable policy environment where governments take various supportive measures to strengthen SE legitimacy and provide resources and services to SEs [3,10,18,21,34]	5-point Likert scale
Financial market	a favorable financial market that provides financial capital available and suitable for SEs [3,26,28]	5-point Likert scale
Labor market	a favorable labor market that provides human capital available and suitable for SEs [26]	5-point Likert scale
Sociocultural setting	a favorable sociocultural setting where the social norms and cultural values guide the general public and media to recognize the legitimacy of SEs and honor the contributions of SEs [3,18,32]	5-point Likert scale
Industrial infrastructure	a favorable industrial infrastructure where intermediary organizations provide incubation, resource linkages and capacity-building services, and promote alliances and cooperation among SEs [3,18,28]	5-point Likert scale
SE Revenue Structure
Commercial SEs	SEs relying mainly on earned income (sales of goods and services) [38]	1 = Yes; 0 = No
Government- supported SEs	SEs relying mainly on government support (grants, subsidies, and public purchases) [21,36,37]	1 = Yes; 0 = No
Donative SEs	SEs relying mainly on charitable donations (contributions from the public, foundations, enterprises or other donors) [37,38]	1 = Yes; 0 = No

.

3.3. Calibration

Drawing on the current fsQCA literature and on the characteristics of the sample distribution, we applied both percentile and manual calibration methods to define the three thresholds. With respect to the social and financial sustainability performance measured by continuous variables, we undertook the percentile method and set calibration values at the upper 95th percentile, median, and lower 5th percentile. With respect to the five conditions related to SE ecosystems, which were measured via Likert scales ranging from 1 to 5, we set the calibration values for full membership at 5, the crossover point at the mean, and full nonmembership at 1. With respect to the three conditions related to revenue structures, which were measured by dichotomous variables, a value of 1 corresponds to full membership, 0.5 to the crossover point, and 0 to the full nonmembership.

Table 3. Descriptive statistics and calibration values.

	Descriptive Statistics				Calibration Values
	Min.	Max.	Mean	S. D.	F. M.	Cr.	F. N.
sustainability of social performance	0.15	1.00	0.71	0.22	0.99	0.75	0.31
sustainability of economic performance	0.00	1.00	0.43	0.26	0.82	0.50	0.00
policy environment	1.00	5.00	2.57	0.26	5.00	2.57	1.00
financial market	1.00	5.00	2.70	0.31	5.00	2.70	1.00
labor market	1.00	5.00	3.03	0.25	5.00	3.03	1.00
sociocultural setting	1.00	5.00	2.79	0.23	5.00	2.79	1.00
industrial infrastructure	1.00	5.00	2.71	0.25	5.00	2.71	1.00
commercial SEs	0.00	1.00	0.57	0.50	1.00	0.50	0.00
government-supported SEs	0.00	1.00	0.21	0.41	1.00	0.50	0.00
donative SEs	0.00	1.00	0.15	0.36	1.00	0.50	0.00

Note: F. M. = full membership; Cr. = crossover; F. N. = full nonmembership.

lists the descriptive statistics of the raw data and predetermined calibration values for the points of the three threshold anchors.

4. Results

This study applied fsQCA through fs/QCA 3.0 software to investigate the complex effects of the two categories of causal conditions (SE ecosystems and revenue structures) on the targeted outcomes (high and low levels of sustainability of social and financial performance).

4.1. Necessity Analysis

We performed necessity analyses to assess whether the conditions involved in this study are necessary for producing the outcomes.

Table 4. Necessary conditions for SE sustainability performance.

	Social Performance				Financial Performance
	High		Low		High		Low
	Cons.	Cov.	Cons.	Cov.	Cons.	Cov.	Cons.	Cov.
Policy environment	0.61	0.70	0.59	0.61	0.65	0.65	0.54	0.64
~ Policy environment	0.66	0.64	0.71	0.62	0.64	0.54	0.71	0.71
Financial market	0.58	0.65	0.60	0.61	0.60	0.58	0.58	0.68
~ Financial market	0.65	0.65	0.66	0.58	0.67	0.58	0.64	0.66
Labor market	0.65	0.69	0.64	0.60	0.67	0.61	0.64	0.70
~ Labor market	0.63	0.66	0.67	0.63	0.67	0.61	0.65	0.70
Sociocultural setting	0.64	0.71	0.61	0.60	0.69	0.66	0.60	0.68
~ Sociocultural setting	0.64	0.65	0.70	0.64	0.66	0.58	0.70	0.73
Industrial infrastructure	0.62	0.70	0.61	0.61	0.66	0.65	0.58	0.67
~ Industrial infrastructure	0.66	0.65	0.70	0.62	0.66	0.57	0.70	0.71
Commercial SEs	0.61	0.56	0.53	0.44	0.60	0.48	0.55	0.52
~ Commercial SEs	0.39	0.48	0.47	0.52	0.40	0.43	0.45	0.57
Government-supported SEs	0.21	0.52	0.22	0.48	0.21	0.45	0.21	0.55
~ Government-supported SEs	0.79	0.53	0.78	0.47	0.79	0.46	0.79	0.54
Donative SEs	0.13	0.45	0.17	0.55	0.13	0.41	0.16	0.59
~ Donative SEs	0.87	0.54	0.83	0.46	0.87	0.47	0.84	0.53

Note: ~ indicates the absence of a condition; cons. = consistency; cov. = coverage.

presents the consistency scores, which range between 0.13 and 0.87. As none of the conditions exceeded the recommended threshold of 0.9 [44,49], the eight conditions (both their presence and their absence) were not necessary for causing high or low levels of SE sustainability performance. The results indicated that the outcomes of SE sustainability performance were not necessarily caused by any single condition, and further analysis of the synergistic effects of multiple conditions is needed.

4.2. Sufficiency Analysis

We performed sufficiency analyses to identify configurations sufficiently causing high and low levels of expected outcomes. In line with the recommendations from the fsQCA literature [44,48,59,60,61], we set the raw consistency threshold at 0.8, the PRI score at 0.7, and the frequency threshold at 2.

As shown in

Table 5. Configurations for the sustainability of financial performance.

	High							Low
	FH1	FH2	FH3	FH4	FH5	FH6	FH7	FL1	FL2a	FL2b	FL4
Policy environment
Financial market		•								•
Labor market									•
Sociocultural setting	•	•	•						•		•
Industrial infrastructure	•	•
Commercial SEs						•	•
Government-supported SEs									•
Donative SEs
Consistency	0.83	0.83	0.82	0.81	0.81	0.80	0.80	0.82	0.84	0.87	0.82
Raw coverage	0.08	0.05	0.21	0.22	0.21	0.24	0.23	0.08	0.15	0.06	0.21
Unique coverage	0.08	0.05	0.01	0.01	0.01	0.01	0.01	0.03	0.01	0.01	0.04
Overall consistency	0.76							0.71
Overall coverage	0.41							0.38

Note: indicates the presence of a core condition; • indicates the presence of a peripheral condition; indicates the absence of a core condition; indicates the absence of a peripheral condition; blank space indicates an irrelevant condition.

and

Table 6. Configurations for the sustainability of social performance.

	High							Low
	SH1a	SH1b	SH2a	SH2b	SH3a	SH3b	SH4	SL1	SL2	SL3	SL4
Policy environment										•
Financial market	•						•		•
Labor market									•
Sociocultural setting	•	•					•		•
Industrial infrastructure						•	•				•
Commercial SEs					•
Government-supported SEs										•
Donative SEs						•					•
Consistency	0.80	0.83	0.85	0.85	0.84	0.80	0.85	0.82	0.86	0.80	0.92
Raw coverage	0.28	0.22	0.21	0.19	0.22	0.05	0.04	0.09	0.09	0.05	0.04
Unique coverage	0.10	0.01	0.01	0.01	0.02	0.02	0.01	0.01	0.02	0.01	0.02
Overall consistency	0.78							0.81
Overall coverage	0.45							0.25

Note: indicates the presence of a core condition; • indicates the presence of a peripheral condition; indicates the absence of a core condition; indicates the absence of a peripheral condition; blank space indicates an irrelevant condition.

, the results of sufficiency analyses revealed divergent causal paths leading to high and low levels of sustainability of social and financial performance among SEs of different revenue structures. The neutral permutations within solutions FL2a and FL2b, SH1a and SH1b, SH2a and SH2b, and SH3a and SH3b indicated second-order or within-type equifinality [48].

In terms of the configurations for the high sustainability of financial performance, seven paths sufficiently led to the outcome: five associated with commercial SEs, one with government-supported SEs, and one with donative SEs. Moreover, the five configurations for commercial SEs revealed three different causal paths in which policy environment (FH3), industrial infrastructure (FH4 and FH5), and sociocultural setting (FH6 and FH7) presented as core conditions respectively, even when other components of SE ecosystems were absent or irrelevant. In contrast, the fsQCA yielded merely a single solution for government-supported SEs (FH1) and donative SEs (FH2), in which policy environment commonly appeared as a core condition, while having sociocultural setting and industrial infrastructure as peripheral conditions.

In terms of the configurations for the low sustainability of financial performance, the results allowed us to identify four paths that sufficiently generated the outcome. Specifically, in the two configurations for donative SEs (FL2a and FL2b), in which policy environment was absent as a core condition, donative SEs were linked consistently to low financial sustainability, although industrial infrastructure was present as a core condition and financial market or sociocultural setting was present as a peripheral condition. Alternatively, in the single configuration for government-supported SEs (FL1), the absence of industrial infrastructure as a core condition and the presence of the other four elements of SE ecosystems as core or peripheral conditions jointly led to low financial sustainability.

In terms of the configurations for the high sustainability of social performance, the results revealed seven configurations linked consistently to the outcome: five associated with commercial SEs and two with donative SEs. Specifically, the five configurations for commercial SEs revealed three different causal mechanisms in which policy environment (SH1a and SH1b), industrial infrastructure (SH2a and SH2b), and sociocultural setting (SH3a) presented as core conditions, even when other components of SE ecosystems were absent or irrelevant. In contrast, in the two configurations for donative SEs, the presence of a sociocultural setting and the absence of a financial market were core conditions (SH3a), or the presence of a policy environment and the absence of a labor market were core conditions (SH4).

In terms of the configurations for the low sustainability of social performance, the results provided four configurations that sufficiently produced the outcome. Specifically, there were two alternative configurations for government-supported SEs (SL1 and SL2). As indicated in SL1, low social sustainability was observed among government-supported SEs where industrial infrastructure was absent, despite the presence of the other four elements of SE ecosystems as core or peripheral conditions. The results also revealed a configuration for low social sustainability among donative SEs (SL3), in which the absence of a sociocultural setting and the presence of a financial market were core conditions.

4.3. Robustness Test

Following the suggestions of the fsQCA literature [44,48,59,60,61], we performed robustness tests to assess the sensitivity of the resulting configurations to the selected parameters and thresholds. Specifically, we adjusted the crossover points from the 50th to 55th percentiles, the consistency thresholds from 0.8 to 0.85, and the frequency thresholds from 2 to 3 and then replicated the analyses. The robustness tests did not result in considerable differences between the regenerated and prior solutions, albeit with minor changes in the number, consistency, and coverage of the solutions and subsolutions. The results indicated that the methodological decisions of the key parameters were unlikely to pose a significant threat to the validity of the findings.

5. Discussion

This study aims to contribute to the SE literature by examining the complexity of the causal relationships among SE ecosystems, revenue structures, and sustainability performance in the Chinese context. The fsQCA generated divergent causal configurations for high and low levels of sustainability performance in social and financial dimensions among commercial, donative, and government-supported SEs. The results provided justification for the use of the fsQCA method by sufficiently demonstrating the equifinal pathways, asymmetric causation, and conjunctural interactions in the relationships among SE ecosystems, revenue structures, and sustainability performance. These findings are both consistent with and contrary to those of previous studies and extend the prior literature by revealing the complexity of the causal mechanisms that determine SE sustainability performance.

5.1. Complexity in the Relationship Between SE Ecosystems and Sustainability Performance

This study focuses on investigating the divergent contributions of different elements in SE ecosystems to SE sustainability performance, paying particular attention to the interactions among them. The results revealed that various elements in SE ecosystems differed in terms of the comparative salience of their roles in the causal configurations.

Specifically, policy environment was a core condition not only in configurations for high financial performance among SEs of the three different revenue structures (FH1, FH2, and FH3) but also in configurations for high social performance among commercial (SH1a and SH1b) and donative SEs (SH4), revealing the importance of supportive policy environment for improving SE social and financial sustainability. Moreover, the findings highlighted the salience of supportive industrial infrastructure, which was presented as a core condition in solutions for high social and financial sustainability among commercial SEs (SH2a, SH2b, FH4, and FH5) and as a peripheral condition in solutions for high financial and social sustainability among government-supported and donative SEs (FH1, FH2, SH3b, and SH4). Furthermore, the results emphasized the vital role of a supportive sociocultural setting, which appeared as a core condition in causal paths leading to high financial sustainability for commercial SEs (FH6 and FH7) and in those generating high social sustainability for commercial and donative SEs (SH3a and SH3b).

These findings align with previous studies that attributed SE sustainability performance to various environmental factors in different contexts, such as community and market acceptance in Israel [30], public support, the availability of potential allies and supportive public policy in Italy [18], financial and training support from external stakeholders in Malaysia and Singapore [26], tax and insurance incentives, networking, cooperation, financial and training support in Greece [28], and social impact accelerators in the global context [24]. Notably, policy environment, industrial infrastructure, and sociocultural setting also appear to be central to configurations for low sustainability performance, while simultaneously playing dominant roles in configurations for high sustainability performance, as indicated in solutions for donative (FH2, FL2a, and FL2b), and government-supported SEs (FH1 and FL1). Thus, these findings extend the literature by showing the asymmetry and conjunctional features of the causal links between various components in SE ecosystems and SE sustainability performance, which have been overlooked in conventional statistical analyses that focused mainly on symmetric linear causality.

Moreover, the findings pinpointed relatively marginal roles played by the financial market and labor market, which appeared as core conditions in the configurations for low sustainability performance (FL1, FL4, SL3, and SL4) and as absent core conditions in those for high sustainability performance (FH1, FH2, FH4, FH5, FH7, SH3a, SH3b, and SH4). These findings conflict with prior studies that highlight the positive contributions of favorable financial and labor markets [17,20,23,41,68]. Nevertheless, we should not take it as evidence that downplays the contributions of financial and labor markets to SE sustainability performance. For example, as shown in solution FH2, when combined with a favorable policy environment, sociocultural setting, and industrial infrastructure, the financial market became a positive contributor to a high level of financial sustainability for donative SEs. This finding further demonstrated the complexity of the causal links among SE ecosystems, revenue structures, and sustainability performance, which has not been captured in the current literature.

5.2. Discrepancies in Configurations Across SEs with Different Revenue Structures

This study also focuses on examining how the causal relationships between SE ecosystems and sustainability performance are contingent on revenue structures. The fsQCA results revealed that the configurations for high and low levels of sustainability performance showed notable discrepancies across SEs with different revenue structures.

Specifically, there were five pathways for commercial SEs to achieve high sustainability performance in both the social (SH1a, SH1b, SH2a, SH2b, and SH3a) and financial dimensions (FH3, FH4, FH5, FH6, and FH7). However, donative and government-supported SEs obtained apparently fewer pathways to achieve high sustainability performance—one in the financial dimension (FH2) and two in the social dimension (SH3b and SH4) for donative SEs and only one in the financial dimension (FH1) for government-supported SEs. These findings indicated that commercial SEs were more likely to achieve high sustainability performance than government-supported and donative SEs, partially echoing Lu et al.’s findings in the U.S. context that commercial nonprofits had a greater chance of survival than their noncommercial counterparts [39].

In addition to the difference in the number of causal paths, the comparative advantage of commercial SEs also manifested as their ability to acquire more chances to achieve high sustainability performance by possessing greater flexibility and adaptability to SE ecosystems than government-supported and donative SEs. Specifically, commercial SEs were able to achieve high sustainability performance in a wide range of SE ecosystems, which were relatively simple in terms of composition and often embraced only one (SH2a, SH2b, SH3a, FH4, FH5, FH6, and FH7) or two supportive components (SH1b and FH3). In contrast, donative SEs achieved high sustainability performance only in more complicated SE ecosystems, which contained at least four supportive components (SH4 and FH2). Moreover, government-supported SEs had the weakest adaptability to SE ecosystems, obtaining only one pathway for high financial sustainability (FH1) but no pathway for high social sustainability.

These findings are consistent with those of studies that attribute SE sustainability performance to organizational capabilities from a resource-based view or to organizational resources from the perspective of resource dependency theory. Previous studies have shown that SEs with stronger marketing capabilities [69], greater abilities to generate earned income [16], or a higher percentage of revenue from earned income [41] have more chances to reduce resource dependence and enhance organizational autonomy, enabling them to make better use of resources and pursue sustainability. Moreover, previous studies have also demonstrated that government-supported and donative SEs often encounter significant difficulties in maintaining sustainability due to excessive resource dependency on governments [21,36,70,71] or donors [38]. Although our findings are in line with those of existing studies that emphasize the advantages of commercial SEs and the disadvantages of government-supported and donative SEs, the differences in sustainability performance captured in this study across commercial, government-supported, and donative SEs should not be interpreted as the singular influence of revenue structures. Instead, the finding of discrepancies in configurations across SEs with different revenue structures further illustrated the conjunctural interactions among SE ecosystems, revenue structures, and sustainability performance, which has not been examined in previous quantitative studies.

6. Conclusions

This study investigates the complexity of the causal links among SE ecosystems, revenue structures, and sustainability performance in the Chinese context using the fsQCA approach. The analyses generated multiple alternative configurations for high and low levels of sustainability performance among commercial, donative, and government-supported SEs, which rely mainly on market-based earned income, charitable donations, and government support, respectively. First, the findings demonstrated there was no single condition that led to high or low levels of sustainability performance of SEs, which instead were the outcomes of the combined effects of SE ecosystems and revenue structures, especially the conjunctural interactions among them. Second, the results revealed that although different elements in SE ecosystems worked jointly and interactively to predict sustainability performance, three categories of elements, namely, policy environment, sociocultural setting, and industrial infrastructure, played relatively more important roles than financial and labor markets. Third, the results showed considerable discrepancies in configurations across SEs with different revenue structures, indicating that commercial SEs had a comparative advantage in terms of gaining more pathways and having greater adaptability to SE ecosystems to achieve high sustainability performance than donative and government-supported SEs.

6.1. Theoretical Contributions

As the first fsQCA study that examines how SE sustainability performance is generated by the complex causal configurations of SE ecosystems and revenue structures in the Chinese context, this study makes unique academic contributions to the SE literature in terms of the analytical perspective, research methodology, and variable measurement.

First, theoretically, previous studies concerning SE sustainability performance have often examined the influence of various elements of SE ecosystems in isolation, without considering how different elements of SE ecosystems are interdependent and how SE ecosystems interact with other organizational conditions to produce sustainability performance. This study addresses the limitations of the analytical perspective by involving SE ecosystems, revenue structures, and sustainability performance within a single, integrated analytical framework. It enriches the SE literature by offering the first configurational understanding of the complexity of the causal mechanisms that contribute to SE sustainability performance.

Second, from a methodological perspective, this study supplements conventional quantitative statistical analyses by using fsQCA to address the issues of the multifaceted interdependencies among different predictors of SE sustainability performance and to shed light on the asymmetric relationships among them in real-life practices. Therefore, this study extends previous discussions on the causal links among SE ecosystems, revenue structures, and sustainability performance from the symmetric linear relationships between variables to the conjunctional causation, equifinality, and asymmetry of causal paths.

Third, in terms of variable measurement, this study contributes to measuring the SE ecosystem by synthesizing conceptual perspectives and empirical indicators from the relevant literature on entrepreneurial ecosystems, social entrepreneurship, and social enterprises. Additionally, this study proposes an operationalized definition of SE sustainability performance and offers specific indicators to measure both the social and financial dimensions of the concept.

6.2. Practical Implications

This study has important implications for SE practitioners, policy-makers, and other stakeholders who seek to improve the sustainability performance of SEs. The findings on the complex causal relationships among SE ecosystems, revenue structures, and sustainability performance will enable SE managers to make informed decisions on sustainable development strategies according to their specific revenue structures and SE ecosystems. For example, commercial SEs should pay particular attention to leveraging the supportive socio-cultural setting and industrial infrastructure in their operational and strategic posture, whereas government-supported and donative SEs should attach more importance to utilizing the positive contributions of policy environments. Additionally, the findings suggest that SEs can enhance their sustainability performance by adjusting their business models and resource strategies to improve the compatibility between their revenue structures and SE ecosystems.

The findings will also help policy-makers and other stakeholders (such as social investors, intermediary organizations, and certification institutions) improve the effectiveness and efficiency of policies and initiatives aimed at improving SE sustainability performance. For instance, this study offers novel insight that commercial SEs might outperform government-supported and donative SEs in terms of achieving sustainable social and financial performance by learning to be more flexible and adaptable to SE ecosystems. Accordingly, policy-makers and other stakeholders should devote more attention and resources to help SEs enhance their organizational capabilities to engage in market-oriented and commercialized operations (such as marketing and branding, business planning, investment, and financing, etc.) or incubate more commercial SEs rather than government-supported and donative SEs. Recent studies have shown that SE development in China still relies heavily on government support [31,32,33,34,35]; nonetheless, this study suggests that government-supported SEs have relatively poor sustainability performance. This indicates that instead of simply providing financial support, the government should experiment with alternative supporting measures, such as providing favorable taxation treatments to launching capacity-building programs, enabling legitimacy construction, and encouraging alliances and cooperation among SEs.

6.3. Limitations and Future Research

This research is not without limitations, which also opens avenues for future research. First, this study focused on examining the combined effects of SE ecosystems and revenue structures on sustainability performance, without considering other factors that have causal relationships with the conditions and outcomes investigated in this study. For example, as Sharir and Lerner suggested, the long-term sustainability of social ventures depends on their ability to gain resources and legitimacy, create cooperation between institutions, and develop internal managerial and organizational capabilities [72]. Therefore, future research could integrate other relevant conditions, such as organizational capabilities, legitimacy, stakeholder relationships, and governance models, into the configurational framework to obtain a deeper understanding of the complexity of the causal paths that predict SE sustainability performance.

Second, this study performed primarily static and cross-sectional analyses, without addressing the issue of how causal links among SE ecosystems, revenue structures, and sustainability performance change over time. Thus, investigating the ever-changing causal mechanisms among multiple conditions by applying fsQCA to longitudinal panel data becomes an important avenue for future research.

Third, as this study used China-only data, the findings potentially were context-specific, suggesting the uncertainty of their generalizability to other contexts. Therefore, future research could examine the relevance of our findings by conducting comparative studies across different contexts or settings, which might offer a richer understanding of the causal complexity of the issue.