Introduction

The field of social entrepreneurship (SE) is growing rapidly (Hu et al., 2020; Stephan et al., 2016; Wry & York, 2017), but “enthusiasm has outpaced conceptual development and refinement” (Miller et al., 2012, p. 630) and controversy on the meaning of SE persists (Ranville & Barros, 2021; Saebi et al., 2019). While previous conceptualization and operationalization studies undeniably provide valuable contributions to advance knowledge in this bolstering academic field, the adopted classical conceptual structure based on necessary and sufficient attributes fails to capture SE as an essentially contested concept (Choi & Majumdar, 2014). Often led “by advocacy worldviews of the researchers themselves” (Lehner & Kansikas, 2013, p. 198), scholars tend to either treat discrete features of SE as equivalent to the broader concept, or use a rigid, monothetic conceptual structure based on a single basic principle that potentially excludes some actors who may also identify themselves with SE. The plethora of SE definitions has not only been related to differences in how attributes are combined to define SE, but also associated with distinct ethical views and normative conceptions of SE (Bhatt, 2022; Bruder, 2021; Dey & Steyaert, 2016; Hota et al., 2020; Ranville & Barros, 2021; Zahra et al., 2009).

“Concept formation lies at the heart of all social science endeavors” (Gerring, 2012, p. 112), not only because it addresses the fundamental question of what we are talking about, but also because in doing so it allows for knowledge to accumulate and progress. Conceptual confusion (Dacin et al., 2010) and persistent disparity in terminology create serious obstacles to accumulate knowledge because what is presented under the banner of SE is different from one publication to another, setting an urgent call to systematize the scattered knowledge developed so far.

Following an “(…) organized set of activities that set priorities for what needs to be done in order to develop a strong conceptual definition” (Podsakoff, MacKenzie & Podsakoff, 2016, p. 169), our study aims to advance the discussion on conceptualizing SE by answering three research questions: (1) which attributes identify SE? (2) how are the attributes organized in existing academic conceptualizations? And (3) how do different sets of attributes relate to distinct ethical views? Based on an extensive sample of 209 academic definitions of SE and their supporting academic articles, we identify 12 attributes that comprehensively define the SE concept, reveal the existing multitude of attribute combinations, and shed light on the distinct ethical views that co-occur with specific choices of attributes. Previous studies have already investigated different definitions of SE to conclude that either it is an essentially contested concept that cannot be defined beyond a set of clusters of other overlapping concepts (Alegre et al., 2017; Choi & Majumdar, 2014) or a classical one holding a specific set of necessary and sufficient attributes (Bacq & Janssen, 2011; Peredo & McLean, 2006; Santos, 2012) and implicitly excluding divergent perspectives of SE. Toward a more inclusive model, this study refutes both the undefinable nature of SE and the classical structure of concept definition. Rather, we propose that SE follows a family resemblance conceptual structure (Rosch & Mervis, 1975; Wittgenstein, 1953), in which membership is not defined by a case possessing a universally accepted set of criterial features but by carrying shared attributes with other cases.

There are important implications for SE research in acknowledging the family resemblance structure. We untangle the fallacy of assuming that different phenomena are the same just because they carry the same SE label and caution against causal homogeneity assumptions among distinct SE subtypes that result from different combinations of attributes. Furthermore, our results contribute to set a sound path to operationalize SE because differences in definitions lead to differences in measurement. By clearly identifying the attributes of SE, this study supports the design of rigorous measurement instruments to empirically capture the different attributes and to investigate theoretically relevant causal relationships. Finally, this study also contributes to the ongoing reflection on the normative ethical foundations of SE and argues for the merits of looking at those foundations from a descriptive approach.

Theory

Conceptualizing Social Entrepreneurship

A considerable number of conceptual studies on SE emerged at the beginning of the century stressing different aspects of the phenomenon (Nicholls, 2010) and reaching a peak around 2006 (Alegre et al., 2017), when several highly cited papers were published. Following the abundance of publications in the academic field, scholars presented several excellent reviews (Saebi et al., 2019), some of which focus specifically on SE’s conceptual definition (Appendix). While some of these reviews provide a unique definition of SE (Bacq & Janssen, 2011; Peredo & McLean, 2006), others conclude that “a universally accepted definition of social entrepreneurship is hardly possible” (Choi & Majumdar, 2014, p. 364). Arguing that a cluster concept would enable systematic future research on SE, the same authors recognize that these overlapping clusters are also contested, complicating the applicability of the cluster concept.

As conceptualization articles tapered off, operationalization efforts began to emerge building on different definitions and attributes of SE. Lepoutre et al. (2013) measure ‘social entrepreneurial activity’ based on individuals’ self-reported social mission, innovation, and market-based revenues. Stevens and colleagues define SE as “entrepreneurship with an embedded social purpose (…), which is sustainable through trade (…) and not limited to a particular organizational form” (2015, p. 1053), developing a scale to measure the social and economic missions of a social enterprise. Kannampuzha and Hockerts (2019) suggest that ‘organizational social entrepreneurship’ is a formative construct based on three components—social change intentions, commercial activities, and inclusive governance. Within the nonprofit sector, Morris et al. (2011) suggested adapting the ‘entrepreneurial orientation’ construct to the nonprofit context. The ‘social entrepreneurship orientation’ (SEO) scale has four dimensions (social innovativeness, social risk-taking, social proactiveness, and socialness) according to Kraus et al. (2017), and five attributes in Dwivedi and Weerawardena’s study (2018) (innovativeness, proactiveness, risk management, effectual orientation, and social mission orientation).

The use of divergent sets of attributes reflects the enduring lack of consensus. Most of these efforts use a classical approach to conceptualize SE, claiming a universal set of singly necessary and jointly sufficient attributes, clashing with previous arguments for its contested nature (Choi & Majumdar, 2014). Recognizing the valuable contribution of these studies in refining the SE construct, the pursuit of operational definitions without acknowledging a broader frame of the concept risks legitimizing powerful actors in the field rather than capturing the existing diversity of perspectives. We share other scholars’ concerns that in the quest to find a consensual definition, perspectives defended by less powerful actors might become marginalized, and agree that “over time, this imbalance might be expected to undermine and perhaps even destroy the normative and cognitive legitimacy of social entrepreneurship to a wider audience” (Nicholls, 2010, p. 626).

Relating Ethical Theories and the Concept of SE

Scholars relate the existing diversity of SE definitions with distinct ethical views and normative conceptions of SE. While a few studies may acknowledge the diversity of perspectives in SE and different philosophical lenses underlying the concept (Ranville & Barros, 2021), most studies tend to assume a normative posture toward SE. For example, Bruder (2021) describes utilitarian tendencies and economistic confusions in a vast number of SE conceptions, and proposes an integrative economic ethics, based on Kantian deontological theory, to ground the SE concept as good entrepreneurship. Bhatt (2022) suggests that social enterprises are “market-based hybrid organizations with a dual mission of financial sustainability and social value creation” and seem to exclude other organizational formats from SE (2022, p. 743).

Normative ethical theories are generally presented in three broad branches: consequentialism, deontology, and virtue ethics (Arnold et al., 2019; Brink, 2006). In consequentialism, the moral worth of actions is determined by their consequences (Brink, 2006). The most familiar form of consequentialism is classical hedonistic act utilitarianism, which claims that an agent ought to perform that action that produces the most net pleasure for everyone concerned (Brink, 2006). An essential feature of utilitarianism is the principle of optimal productivity through efficiency (Arnold et al., 2019). In rule utilitarianism, moral behavior requires adopting rules based on the consequences they may have (Arnold et al., 2019). Finally, economism is based on the maximization of self-interest as the rational mechanism to promote economic welfare (Bruder, 2021; Ulrich & Thielemann, 1993).

Deontological ethics postulates that the morality of an action is based on whether that action itself is right or wrong, rather than based on its consequences (Arnold et al., 2019). Within deontology, Kantianism argues that actions must satisfy the categorical imperative of human dignity, claiming that a person should be treated as an end and never purely as a means to the end of others (Arnold et al., 2019; McNaughton & Rawlings, 2006). Informed by Kantian theory, Habermasian discourse ethics (Habermas, 1992) is a process oriented ethical perspective that emphasizes the participation of all affected parties in fair dialogues to establish moral norms (Beschorner, 2006). Moral rights theory is a current deontological ethical view, which grounds ethical theory in an account of personal claims and enforceable duties that render other persons liable (Steiner, 2006), which are not reducible to a theory of obligations or virtues (Arnold et al., 2019).

Distinct from the previous ethics of conduct, virtue ethics does not rely on a mechanical algorithm based on the consequences or on the acts themselves for making the right decision (Koehn, 1995). Based on an ethics of character, virtue ethics views morality’s primary function to cultivate virtuous characters with the disposition to do the right thing for the right reason in an appropriate way (Annas, 2006).

Extending previous ethical reflections on SE, our study focuses on ethics to understand persistent conceptual controversies that still prevail in defining SE. In this paper, we depart from previous normative viewpoints and take a descriptive stance to portray how distinct ethical views relate to different conceptions of SE.

Family Resemblance Versus Classical Conceptual Structures

The family resemblance structure is an alternative to the classical conceptualization approach and defines complex concepts for which no single set of characteristics is common to all variants (Komatsu, 1992; Rosch & Mervis, 1975; Wittgenstein, 1953):

“A family resemblance relationship consists of a set of items of the form AB, BC, CD, DE. That is, each item has at least one, and probably several, elements in common with one or more other items, but no, or few, elements are common to all items.” (Rosch & Mervis, 1975, p. 575)

The concept of ‘games’ and the common attributes in its subsets board games, card -games, ball games, and Olympic games represents an example of a complicated network of similarities overlapping and crisscrossing among the subsets (Podsakoff et al., 2016). Concepts holding a family resemblance structure are not bounded entities in which membership is defined by an item or case possessing a simple set of criterial features (Rosch & Mervis, 1975). Rather, cases that share many attributes with other cases bear greater family resemblance and are more ideally representative of the concept than members sharing only a few attributes (Podsakoff et al., 2016). Every attribute must be shared by more than one case from the same conceptual space, and those attributes shared by many cases are more central than others shared by only a few cases (Komatsu, 1992; Podsakoff et al., 2016). The explicit application of the family resemblance structure has been scarce in the management literature (Podsakoff et al., 2016), with a few notable exceptions, such as practice theory (Nicolini & Monteiro, 2016) and ‘privacy’ (Solove, 2019).

There are important implications from assuming the family resemblance as the underlying structure of a concept. First, unlike classical concepts, family resemblance ones accommodate members possessing different sets of attributes. The ‘or’ rather than the ‘and’ structural logic that underlies the family resemblance structure (Podsakoff et al., 2016) allows for membership arguments stressing differing features, aggregating distinct perspectives on the concept under the same conceptual space without ignoring existing variance (Yeung et al., 2012). Additionally, conceptual subtypes in a classical structure capture variation by adding an extra and non-essential attribute to the original concept. In contrast, to form subtypes using a family resemblance structure, Barrenechea and Castillo (2019) propose a method of grouping the existing attributes used in defining the concept and recommend “turning to the subtypes to gain the differentiation that is lost by the inherent nature of these structures” (2019, p. 123). Finally, one of the defining features of the family resemblance structure is attribute substitutability (Podsakoff et al., 2016), which hampers causal homogeneity that characterizes classical concepts (Barrenechea & Castillo, 2019), implying that antecedents and consequents may not be the same across conceptual subtypes.

Following literature’s guidelines to develop good conceptual definitions, our study allows for the conceptual structure to emerge from the data. In the next section, we describe the research methods employed. Then, we present and discuss our findings. Finally, we acknowledge the limitations of this study and summarize our main contributions.

Methods

Overall Research Design

To answer our research questions—which attributes identify SE, how are they organized, and how are they related to distinct ethical views?—we designed and guided our research based on the stages for strong conceptual development recommended by Podsakoff et al.:

“(a) identify potential attributes of the concept by collecting a representative set of definitions, (b) organize the potential attributes by theme and identify any necessary and sufficient -- or shared ones—[emphasis added], (c) develop a preliminary definition of the concept, and (d) refine the conceptual definition.” (2016, p. 169)

We started by collecting a representative sample of 209 definitions of SE because “in those cases where several different conceptual definitions already exist (…) then conducting a thorough review of the literature, identifying the key attributes based on these definitions, and organizing these attributes into meaningful conceptual themes before developing their definition may be the most important activities.” (Podsakoff et al., 2016, p. 180). Then, we analyzed the content of the definitions to identify the potential attributes of SE, generating tentative categories or themes from asking the specific question “which themes do scholars use in defining SE?” and refining them until we distilled 12 attributes. Next, we looked for patterns in how the themes were organized across different definitions. Finally, realizing that SE definitions organized attributes in a variety of distinct combinations, we reviewed the corresponding articles to understand how divergent ethical stances relate with the plethora of SE conceptualizations.

Data Collection

We collected our sample of definitions in three different periods to ensure geographical and temporal representativeness. First, through snowball sampling and starting with definitions identified by Santos’ (2012), we found Dacin et al.’s (2010) list of 37 definitions and to Zahra et al.'s (2009) list of 20 definitions, adding their own definitions. In these works, we found two additional lists: Masseti’s (2008) with 11 definitions and Weerawardena and Mort’s (2006) with 20. This initial sample of 60 unique definitions covered the period in which most conceptualization studies were published and contained highly cited definitions. However, this list included definitions prior to 2012 and was potentially biased toward American perspectives. To circumvent these limitations, we considered two additional sets of SE definitions. Alegre et al. (2017) systematically reviewed 307 articles until April 2015, from which we added 94 non-overlapping definitions to our sample. Subsequently, we conducted a systematic review of articles from 2015 to September 2018 searching titles, abstracts and keywords for the expressions social entrepreneurship, social enterprise, and social entrepreneur. We included articles from top journals important to SE research selected by Short et al. (2009) and those identified by Alegre et al. (2017), collecting 107 additional articles. Not surprisingly, most scholars in this last period used earlier conceptualizations as working definitions. After adding 55 new definitions, our final sample comprised 209 unique definitions of SE.

Data Analysis

Our data analysis evolved in three different stages, mostly based on going through qualitative data to generate categories that would answer our research questions. To explain the analysis conducted, we build on the analytical moves identified by Grodal et al. (2021) because “researchers can demonstrate rigor by detailing more precisely how they have purposefully drawn on a broad and diverse set of moves to engage with their data” (2021, p. 593).

In a first stage, we investigated both the 209 definitions and respective articles to identify and distill the attributes of the SE concept. We started analyzing the content of the 209 definitions by asking questions, the move in which researchers draw on their existing categories to select and approach data, with specific questions to which they would like to answer (Grodal et al., 2021). The authors explain that asking questions is a core part of early discovery because it enables the creation of initial categories for the problem at hand. Specifically, we started looking for categories in our sample by asking the question ‘which themes do scholars use in defining SE?’.

From our sample of 209 definitions, we extracted 651 expressions by breaking up each definition in different themes. For example, Bornstein’s (2004) definition was separated in five different expressions that relate to different themes: ‘a social entrepreneur,’ ‘is a pathbreaker,’ ‘with a powerful new idea, who combines visionary and real world problem solving creativity,’ ‘who has a strong ethical fiber,’ and ‘and who is totally possessed by his or her vision for change.’ Being aware that not all information is equally important in the categorization process (Grodal et al., 2021), we were nevertheless careful to analyze and code all the words as stated in each academic definition, because each definition already synthesizes each scholars’ perspective on the key attributes of the concept. As such, leaving out parts of the text could lead to missing out less salient attributes.

Then, to reduce data, we merged categories by uniting expressions to create superordinate categories (Grodal et al., 2021), based on similarities and differences among the expressions extracted. For example, we merged expressions such as ‘with a powerful new idea, who combines visionary and real world problem solving creativity’ with ‘new ways,’ ‘pioneers of innovation,’ ‘new programs, services, and solutions’ into a superordinate category we labeled ‘innovativeness.’ Through this merging process, we progressed from lower to more complex categories, significantly reducing the number of expressions to 15 tentative categories, of which 14 themes referred to potential attributes of SE and one category related to the unit of analysis.

Next, to distill SE attributes from the tentative categories, we read and analyzed the articles from which the definitions were extracted to understand what academics mean in each theme they draw on to define SE. Using NVivo to aid in the qualitative analysis of the articles, we started coding them not only using the tentative categories that emerged from the previous step as the initial codes, but also adding new categories to code data that could bring additional definitional clarity to the SE concept. For example, we also coded how academics define conceptual boundaries between SE and related constructs, such as CSR, business or commercial entrepreneurship, social activism, and traditional nonprofits. The codes referring to tentative categories of attributes were further expanded as different meanings were identified for an initial category. For example, employing the qualitative move splitting categories (Grodal et al., 2021), we separated the ‘sustainability’ category into two different codes related to ‘sustainability of the impact’ and ‘self-sustaining activities,’ which was then merged with the code ‘commercial activities.’ Refinement of tentative categories also involved relating and contrasting categories (Grodal et al., 2021). For example, we identified that ‘innovativeness,’ ‘proactiveness,’ and ‘risk-taking’ were related to the ‘entrepreneurial behavior’ category, as these categories are often mentioned together in academic papers. We also related the category ‘commercial activity’ with ‘entrepreneurial behavior’ after contrasting the distinct meanings of the latter category. Based on the articles, we shed light on the different meanings assigned to each attribute, highlighting conceptual controversies. We identified these controversies either from explicit references to different schools of thought (for example, Choi and Majumdar (2014) explicitly identify different schools of thought regarding SE attributes), or implicitly from the existence of different academic perspectives on a certain attribute (for example, the requirement of profit reinvestment in Weerawardena and Mort (2006) conceptualization paper versus profit distribution in Battilana et al’s (2012) paper). We concluded this first analytical stage by distilling, from the initial 14 themes, the 12 attributes that exhaustively cover the definitions analyzed.

In the second stage, Podsakoff et al. (2016) suggest investigating how attributes are organized among definitions. Using the initial 14 attribute-related themes, we saw them as distinct building blocks that scholars pick and choose to combine and provide their own definition of SE. We mapped all combinations of themes by creating a table in excel, with each definition in a row and each of the 14 themes in a column. For each row, themes were identified as present with “1” and missing with “0.” After concatenating the results of each row, we counted the frequency each combination appeared in our sample of 209 definitions. Despite the diversity of existing combinations, we analyzed the patterns of attribute association to identify the prominent SE conceptual subtypes. To that end, we counted the frequency of each pair of attributes appeared simultaneously in SE definitions. We also investigated how the 14 themes related to different units of analysis. To that end, we classified and grouped the different expressions referring to units of analysis in five levels: individual, organization, process, context, and other.

In the final stage, Podsakoff et al. (2016) suggest refining and providing a tentative definition for the concept. Building on previous research relating SE conceptualizations and ethical stances, we propose refining the SE concept through clarifying relationships with ethical views explicitly or implicitly stated in academic papers. To do so, we went back to the analysis of the academic papers. Using the three broad branches of normative ethical theory as templates, we reviewed the data previously coded in tentative categories associated with SE attributes and re-coded them according to ethical positions explicitly stated in the articles associated with the choice of attributes and their meanings. In this process, we split the three broad categories in more specific ethical views. For example, we split the category ‘consequentialism’ in ‘act utilitarian,’ ‘rule utilitarian,’ and ‘economism.’ Whenever the explanation or relevance of an attribute was not clear, we re-read the papers in search for additional arguments that would identify the scholar’s position. In the next section, we report the results of our analysis.

Results

Stage I – Identification of Attributes

The content analysis of the 209 definitions of SE is summarized in Table 1. We identified 14 initial themes used in characterizing SE in academic definitions and an additional category related to the unit of analysis. For clarity purposes, we grouped the 14 themes in three main elements of the SE concept: the social element, which includes social goals, transformational change, sustainability of the impact, and virtuosity; the entrepreneurial element, which relates to innovativeness, entrepreneurial behavior, proactiveness, commercial activity, resourcefulness, and willingness to take risk; and the managerial element, which includes profit reinvestment, business-like approach, scalability, and collective governance. Next, we provide an account of what scholars understand of each of these themes.

Table 1 Expressions and tentative categories from SE Definitions
Full size table

Themes Related to the Social Element

We identified four different themes that academics refer to in defining the social element of SE. First, most scholars agree that having a social mission (Dees, 1998; Seelos & Mair, 2005), addressing social problems (Bornstein, 2004; Mair & Martí, 2006; Waddock & Post, 1995), or creating social value (Austin et al., 2006; Peredo & McLean, 2006) is a key attribute of SE. Conceptions of SE often state that the mission needs to be social, but do not spell out exactly what a social mission is (Bruder, 2021). What constitutes ‘social goals’ is still controversial. Some scholars specify social goals by normatively enumerating them (Fowler, 2000). Seelos and Mair (2005) propose the widely accepted United Nations Millennium Development Goals as the ultimate goal of SE. In the same vein, Kroeger and Weber (2014) argue that the goal of social interventions is to create positive changes in the life satisfaction of individuals disadvantaged in a specific life domain. Other academics define social goals as opposed to economic ones, arguing that SE creates social value while commercial entrepreneurship creates value for personal or shareholders’ wealth (Austin et al., 2006; Mair & Martí, 2006). Drawing on economic theory, Santos (2012) counterargues that a clear dichotomy between social and economic value is complicated because economic goals are inherently social, as they improve social welfare through a better allocation of resources. In addition to the lack of consensus defining what social goals are, scholars also discuss the centrality of the social mission vis-à-vis economic goals (Peredo & McLean, 2006) or other legitimate aspects extraneous to profitability and mission success (Bruder, 2021).

The second theme related to a social element that emerged from our sample was ‘transformational change,’ cited in 38 definitions. Some SE definitions refer to the importance of having a social objective (Harding, 2004) or social purpose (Haugh, 2006) without explicitly referring the SE actor as a change agent. In this view, any kind of social goal is sufficient to identify SE phenomena and “[t]here is no exact way of fixing the border below which the importance of social goals fails to qualify something as social entrepreneurship” (Peredo & McLean, 2006, p. 64). Other authors argue that more than alleviating social needs, the social mission should seek to change the status quo, achieving a new equilibrium that will solve a social problem (Dees, 1998; Drayton, 2002; Martin & Osberg, 2007), implying structural shifts that transform behavior and promote systemic social change.

Thirdly, in defining SE, sustainability refers either to the organization’s ability to self-finance its activities (Lasprogata & Cotten, 2003), a topic we address ahead together with other entrepreneurial related themes, or to the sustainability of social impact and the provision of lasting social benefits (Santos, 2012). Regarding this latter meaning related to social goals, the aim to provide lasting benefits is sometimes seen as a distinctive feature of SE (Sharir & Lerner, 2006; Werawardena & Mort, 2006). More than ‘a quick hit’, social entrepreneurs are concerned with sustaining their impact by creating lasting improvements (Dees, 1998), which requires developing a solution that eventually eliminates the problem permanently (Santos, 2012). In this sense, ‘sustainability of the impact’ overlaps with the previous discussion on ‘transformational change’ and is not an additional attribute of SE.

Finally, still related to the social element of SE, a few scholars cite ‘virtuosity’ as an attribute of SE because it focuses on solving social problems rather than pursuing a profit or donating profits to a cause (Thompson, 2002). Others argue that “there is considerable need to research further the ethical context of social entrepreneurship and enterprise” (Chell et al., 2016, p. 621). Santos posits that “social entrepreneurs do not need to be defined as good or moral agents that want to help others” but rather as “economic agents who, due to their motivation to create value without concern for the amount they capture will enter areas of activity where the more severe market and government failures occur.” (2012, p. 344). Opposing to this claim, Bruder (2021) states that “(…) the prefix ‘social’ carries the normative validity claim of being good entrepreneurship” (2021, p. 500), proposing that a utilitarian view of overemphasizing either a social mission or economic goals may lead to unethical practices.

Themes Related to the Entrepreneurial Element

The analysis of the 209 SE definitions also evidenced six themes associated to an entrepreneurial element of the SE concept. The manifestation of entrepreneurial qualities (Drayton, 2002; Thompson, 2002), behaviors (Weerardena & Mort, 2006), or strategies (Lasprogata & Cotten, 2003) is a core theme defining SE. However, the meaning of entrepreneurship varies among different people (Zahra & Wright, 2016). In a minimalist sense, it might refer to starting a business based on commercial activity, or, in a popular sense, borrowing business methods (Peredo & McLean, 2006), or even from its French origin, accepting a challenging task, which involves innovativeness, resourcefulness, proactiveness, and risk-taking (Dees, 1998). To avoid tautology and confusion, we disaggregate the discussion on the ‘entrepreneurial behavior’ theme in its specific meanings: first, referring to entrepreneurship’s French origin, we examine the themes related to ‘innovativeness,’ ‘resourcefulness,’ ‘proactiveness,’ and ‘willingness to take risk.’ Then, in line with a minimalist sense, we debate the theme ‘commercial activity’ and the concern with organizational self-sustaining strategies. Finally, we discuss the popular interpretation of entrepreneurship—‘business-like approach’—in the next section, together with other managerial related themes.

Starting with ‘innovativeness,’ scholars repeatedly use terms such as new programs, new services, new solutions (Korosec & Berman, 2006), new activities (Austin et al., 2006), new ideas (Bornstein, 2004; Drayton, 2002), or new combinations of resources (Mair & Martí, 2006) to characterize SE. Innovativeness, or the quality of introducing new ideas, reflects an important means by which new opportunities are pursued (Lumpkin & Dess, 1996; Zahra et al., 2008). Innovativeness in SE does not necessarily mean creating a new product or service, and is often associated with developing new models, approaches, methods of production, strategies, and ways of organizing in order to increase effectiveness and efficiency in tackling a social problem (Dees, 1998).

Akin to the ‘innovativeness’ theme, ‘resourcefulness’ is a topic cited in 28 SE definitions. Unlike their business counterparts, social entrepreneurs leverage whatever resources available regardless of norms or conventions that usually limit more traditional sectors (Dees, 1998; Seelos & Mair, 2005). Profit-seeking firms typically mobilize resources that are valuable, rare, inimitable, and non-substitutable, which they seek to deny to competitors to obtain a competitive advantage (Barney, 1991). Social entrepreneurs, in contrast, often leverage abundant resources outside the organization’s boundaries (Austin et al., 2006), using them collaboratively with other players (Dacin et al., 2010).

Next, proactiveness means taking the initiative by anticipating and pursuing new opportunities (Lumpkin & Dess, 1996). Scholars often mention proactiveness as a core attribute of SE, referring to the direct action involved in pursuing an opportunity (Bornstein, 2004; Peredo & McLean, 2006). Earlier definitions viewed the social entrepreneur closer to a social activist, playing “critical roles in bringing about catalytic changes in the public sector agenda and in the perception of certain social issues” (Waddock & Post, 1995, p. 393). More recently, scholars argue that SE differs from social activism as it requires the creation of an organizational context (Mair & Martí, 2006) and direct action in solving a social problem (Martin & Osberg, 2007; Santos, 2012). An SE opportunity differs substantially from a commercial opportunity because the former is more prevalent and urgent, enables more accessibility and collaboration, and might require especially innovative approaches (Zahra et al., 2008).

Then, ‘willingness to take risk’ is a theme mentioned in only nine SE definitions. Tan and colleagues (2005) explain that willingness to take risk is a necessary but not sufficient condition of any entrepreneur. However, other academics note that, rather than ‘willingness,’ social entrepreneurs exhibit tolerance to risk and appear to adopt a highly cautious approach in dealing with risk and, rather than ‘risk,’ social entrepreneurs face ambiguous scenarios and probabilities in which nascent markets and unknown structures complicate the ability to predict revenue streams (Weerawardena & Mort, 2006).

Finally, cited in 41 percent of the SE definitions analyzed, ‘commercial activity’ is also an entrepreneurship related theme. Viewing entrepreneurship as starting a business with a commercial activity, some scholars argue that earned-income strategies to self-sustain the organization’s activities and the simultaneous pursuit of social and economic goals (Doherty et al., 2014; Hockerts, 2006) are fundamental attributes of SE. Despite the increasing inclination in SE toward structures that combine conflicting elements (Hota et al., 2020), for other scholars, social enterprises, defined as organizations “whose goal is to achieve a social mission through commercial activities” (Pache & Santos, 2013, p. 972), are not the only type of organizations pursuing SE goals. Dart (2004) suggests that requiring earned income models is a matter of ideology, because “from a rational perspective, social-sector innovations should all be equal whether they receive government funds or earn income” (2004, p. 420). Furthermore, SE is also commonly associated with innovative funding strategies as a means to extend the organization’s resource base, which may include sources of revenues other than commercial activities (Anderson & Dees, 2006), namely “‘voluntary’ or in-kind contributions and possibly donations and grant aid” (Chell, 2007, p. 13).

Themes Related to the Managerial Element

In addition to the previous social and entrepreneurial elements, typically associated with the concept of SE (Mair & Martí, 2006; Peredo & McLean, 2006), we also identified four themes related to a managerial element in the sample of definitions analyzed: profit reinvestment, business-like approach, scalability, and collective governance. We discuss each of these themes below.

First, while some authors argue that profit or surplus reinvestment is important to ensure the legitimacy of SE (Harding, 2004), a theme cited in 27 definitions, others claim that this is not always the case. Some social enterprises with commercial activities can legitimately decide to make profits (Zahra & Wright, 2016) and share them with their shareholders as they may target private investment to ensure funding requirements that require profit sharing (Battilana et al., 2012). Relaxing the requirement of profit reinvestment and allowing for profit distribution in SE is also aligned with the arguments that SE organizations can have diverse legal and organizational forms, ranging from nonprofit to for-profit organizations (Austin et al., 2006).

Second, although most scholars agree that social and business entrepreneurship differ from each other, some observe that a ‘business-like approach,’ or the employment of business principles (Dacin, et al., 2010) or methods (Peredo & McLean, 2006) such as strategy, structure, norms, and values (Dart, 2004) may be common to both types of entrepreneurship. The use of such professional practices is also critical to ensure accountability to the constituencies served and for the outcomes created because the discipline of the markets in which social entrepreneurs operate is often not closely aligned with their social mission (Dees, 1998; Ebrahim & Rangan, 2014; Molecke & Pinkse, 2020).

Third, ‘scalability’ is an appealing attribute of SE to some scholars because it enables reaching out to as many beneficiaries as possible (Bornstein, 2004) and legitimizing SE through success stories (Nicholls, 2010). However, other academics argue that in some cases, “growth may not be the best approach to achieve the organization’s goals or to have the greatest social impact” (Austin et al., 2006, p. 7) and that “many initiatives stay small and local, mirroring the tendency of many micro- and small businesses not to grow into medium-sized enterprises” (Thompson, 2002, p. 415).

Finally, we found the theme ‘collective governance’ in only six definitions. Reputed scholars, such as Defourny and Nyssens (2008), suggest that the “representation and participation of users or customers, influence of various stakeholders on decision-making, and a participative management are often important characteristics of social enterprises” (2008, p. 37). However, the same authors also observe that “empirical research has shown that the single-stakeholder character does not seem to jeopardize the multiple-goal nature of social enterprises” (2008, p. 7), suggesting that this attribute may not be a necessary one in conceptualizing SE. Other scholars even fear that a significant focus on stakeholders’ interests may contribute to lose focus on the essential social goal of SE, claiming that “it is often the case that the social entrepreneur becomes increasingly focused on organizational interests as a means to achieve social impact rather than on social impact itself” (Austin et al., 2006, p. 16).

Discussion of Stage I—Identification of Attributes

In sum, we observe that 12 attributes exhaustively cover the 209 definitions analyzed, excluding from the initial 14 themes the ‘entrepreneurial behavior’ and ‘sustainability of the impact’ themes because they seem to conceptually overlap with the remaining 12. Using frequency as a proxy of the centrality of each attribute and assuming that a central attribute is mentioned in more than ten percent of the total number of definitions, we found six central attributes referred to in more than 21 definitions: ‘social goals’ (184 definitions), ‘commercial activity’ (88), ‘innovativeness’ (76), ‘proactiveness’ (46), ‘transformational change’ (38), and ‘profit reinvestment’ (27). Previous studies mention as limitations the tautology often involved in conceptualizing SE (Santos, 2012), the different meanings contained in ‘social’ and ‘entrepreneurship’ terms (Dees, 1998; Peredo & McLean, 2006), and the controversy of the attributes proposed (Choi & Majumdar, 2014; Saebi et al., 2019). However, by systematically analyzing a comprehensive sample of SE definitions, we provide a more precise picture of the exhaustive list of SE attributes and how central each of them is in characterizing SE. The persistent controversy related to many of the attributes identified points to the contested nature of the SE concept (Choi & Majumdar, 2014). This contested nature complicates a straight-forward, unbiased selection of attributes to conceptualize SE and requires additional attention to understand how academics organize them.

Stage II—Organization of Attributes

In the second stage of good conceptual development, Podsakoff et al. (2016) recommend identifying necessary and sufficient attributes or shared characteristics across subsets of cases. Definitions in our sample combined on average three and up to six themes. We found 101 different combinations of themes in our sample. There is a wide dispersion of definitions through the different combination of attributes, without a specific set of attributes standing out from the wide variety of combinations. The most frequent combination, repeated in only 29 definitions, included ‘social goals’ and ‘commercial activity’ attributes.

Emergent SE conceptual subtypes

We then analyzed the frequency with which pairs of SE attributes appear together in SE definitions (Fig. 1).

Fig. 1
figure 1

Relationship between pairs of SE attributes and prominent SE subtypes

Full size image

The attributes ‘commercial activity’ and ‘innovativeness’ seldom appear together despite being frequently cited in SE definitions, suggesting there are two different schools of thought. Combinations between ‘commercial activity’ and ‘innovativeness’ with other attributes differ and present little overlap. Among the plethora of existing attribute combinations, it is possible to discern four emergent conceptual subtypes, which do not however cover all the spectrum of combinations (Fig. 1). Within the ‘commercial activity’ school of thought, the most representative subtype results from combining only that attribute with ‘social goals’ or linking them both with ‘proactiveness.’ Examples of SE definitions within this first subtype are “[t]he process of employing market-based methods to solve social problems” (Grimes et al., 2013, p. 460) and “[t]he process of identifying, evaluating, and exploiting the opportunities aiming at social value creation by means of commercial, market-based activities and of the use of a wide range of resources” (Bacq & Janssen, 2011, p. 376). A second conceptual subtype also based on ‘commercial activity’ results from requiring ‘profit reinvestment’ as a key characteristic of SE. Conceptualizations within this subtype relate to Yunus’ (2010) definition of social business, which is described as a self-sustaining organization whose “owners never intend to make profits for themselves” (Yunus, 2010, p. 310). In the same vein, Hartigan (2006) defines SE as businesses in which “revenues beyond costs are reinvested in the enterprise in order to fund expansion” (2006, p. 45).

Considering the ‘innovativeness’ school of thought, the variety of attribute combinations is wider and subtypes are less clear-cut. Notwithstanding, it is possible to identify a third prominent SE subtype combining ‘social goals,’ ‘innovativeness,’ and ‘proactiveness’ as key attributes of SE, which eventually includes further attributes such as ‘resourcefulness,’ ‘business-like approach,’ and ‘scalability.’ Within this subtype, Austin et al. (2006) define SE as an “innovative, social value creating activity that can occur within or across the nonprofit, business, or government sectors” (2006, p. 3). Adding ‘resourcefulness’ to the previous combination, Mair and Marti (2006) define SE as “a process involving the innovative use and combination of resources to pursue opportunities to catalyze social change and/or address social needs” (2006, p. 37). A fourth, distinct subtype emerges within the ‘innovativeness’ school of thought by requiring ‘transformational change’ as a key attribute of SE. Drayton’s (2002) definition of a social entrepreneur is an exemplar within this subtype: “What defines a leading social entrepreneur? First, there is no entrepreneur without a powerful, new, system change idea. (…) There are four other necessary ingredients: creativity, widespread impact, entrepreneurial quality, and strong ethical fiber” (2002, p. 123). Martin and Osberg (2007) and Santos (2012) also provide definitions within this fourth subtype.

Considering the units of analysis, we find that, in definitions explicitly including terms related to the ‘commercial activity’ theme, the most common unit of analysis refers to the organization-level (65% of the cases) versus less common individual-level (15%) or process-level (10%) units of analysis. However, in definitions referring to ‘innovativeness,’ academics tend to use individual-level (38%) or process-level (31%) units of analysis rather than those referring the level of the organization, mentioned in 14% of the cases in this SE subtype.

Discussion of Stage II—Organization of Attributes

The patterns observed are in line with evidence presented by Chliova et al. (2020), who argue that the SE category has historically originated from two different schools of thought based on either ‘commercial activity’ or ‘innovativeness.’ Our evidence shows that the combination of attributes differs and presents little overlap between the two main schools of thought. Based on the vastness of existing attribute combinations found, conceptualizing SE through a universally accepted set of sufficient and necessary attributes is hardly possible. Definitions based on a classical approach necessarily leave out part of SE’s conceptualization diversity. From the complicated network of similarities overlapping and crisscrossing among different definitions of SE, as in the example of ‘games’ provided by Wittgenstein (1953), we suggest that the SE concept has an underlying family resemblance structure. Much of the conceptual confusion that prevails in the SE field results from interchangeably using the same SE term to label each of the conceptual subtypes, promoting the fallacy of erroneously assuming that two different things are the same because they bear the same name. Scholars are thus invited to clarify the SE definition in use and synthesize conclusions within a specific SE subtype, rather than amalgamating them under the overall umbrella of the family resemblance concept.

Stage III—Development and Refinement of the Conceptual Definition

In the last stages of the concept development process, Podsakoff et al. (2016) suggest generating and refining a definition that describes the general nature of the concept. In family resemblance concepts, Barrenechea and Castillo (2019) “recommend turning to the subtypes to gain the differentiation that is lost by the inherent nature of these structures” (2019, p. 123). Building on other scholars’ research who noted normative implications underlying the definition of SE (for example, Bruder, 2021; Ranville & Barros, 2021), we investigated how distinct normative ethical positions relate to different SE attributes (Table 2).

Table 2 Ethical positions underlying the choice of SE attributes
Full size table

Noteworthy, we do not aim to oversimply the complexity that characterizes the SE concept with this exercise. Our aim in uncovering different ethical stances is rather to gain understanding on the co-occurrence of specific attributes in different SE conceptualizations.

Consequentialist Views

We found that a significant number of scholars normatively identify SE by stressing the social consequences or outcomes of entrepreneurial activities (e.g., Dacin et al., 2010; Dees, 1998; Leadbeater, 1997; Mair & Martí, 2006; Peredo & McLean, 2006; Santos, 2012). Although most ethical stances are implicitly stated, Peredo and McLean (2006), within a consequentialist frame, clearly argue that “(…) the pursuit of socially valuable outcomes is something worth identifying and fostering, whereas probing the mysteries of motivation is not only difficult but of little practical consequence” (2006, p. 63, emphasis added). In more radical conceptualizations, social goals are “an additional happy outcome” (Peredo & McLean, 2006, p. 64) of entrepreneurial activities. As such, several scholars ask if there is a difference between SE and other forms of entrepreneurship (Dacin et al., 2010) or if the difference is a result of “(…) neglect[ing] to mention the social outcomes and benefits of entrepreneurship – work, employment, belongingness, community, friendship, self-respect, social standing, and development of one’s capability” (Chell, 2007, p. 17). Within consequentialism, we also identified other stances that relate to more specific ethical views.

Assuming an act utilitarian view, academics often stress the relevance of maximizing value while minimizing harms for society (e.g., Austin et al., 2006; Nicholls, 2010; Santos, 2012) and of resource efficiency requiring the understanding of the competitive environment in which SE develops (Lasprogata & Cotten, 2003; Leadbeater, 1997; Weerawardena & Mort, 2006). While both arguments relate to a utilitarian stance, the centrality of social goals may differ. On one hand, conceptualizations focusing on the maximization of social value tend to view social goals as a central attribute in SE. On the other hand, from the point of view of resource efficiency, some academics don’t see the social element as necessarily central in SE, stating that “anyone who claims “the cause is all” is wrong” (Thompson, 2002, p. 427) and “(…) the social mission is not a sacred goal as traditionally has been believed [but] must be understood within the competitive environment within which the organizations operate.” (Weerawerdena & Mort, 2006, p. 30). To define SE, it is argued that ‘innovativeness’ or ‘resourcefulness’ are key attributes to rearrange resources in a productive way (Lasprogata & Cotten, 2003; Leadbeater, 1997; Weerawardena & Mort, 2006), proposing that “cost efficiency suggests that society could get far more innovative social welfare, delivered at a lower cost” (Leadbeater, 1997, p. 22). Through a utilitarian lens, holding a ‘business-like approach’ is critical because “[s]trong leadership and good management of socially entrepreneurial initiatives is important. There is always an opportunity cost for the resources being utilized. Achievement below that which could be achieved is a lost opportunity, an unmet need.” (Thompson, 2002, p. 427). Furthermore, related with a utilitarian view, ‘proactiveness’ often assumes that, rather than being motivated by moral obligation, the agent pursues those activities with potential to create positive social and economic outcomes because “(…) opportunity seeking behavior goes hand in hand with the financial viability of the opportunity” (Weerawardena & Mort, 2006, p. 31). Finally, ‘scalability’ might also be a relevant SE attribute to attain economies of scale (Tracey & Jarvis, 2007) and further pursue resource efficiency.

Regarding the type of good SE should pursue, we discerned two distinct stances. On one hand, a view more related to hedonic wellbeing referring social goals as life satisfaction domains (Kroeger & Weber, 2014), or quality of life (Zahra & Wright, 2016). Following a Benthamian tradition, some academics propose a formula to calculate and compare the amount of social value created among different social interventions because “markets do not do a good job of valuing social improvements, public goods and harms, and benefits for people who cannot afford to pay” (Dees, 1998, p. 3). On the other hand, some conceptualizations stress that SE should aim at the empowerment of beneficiaries, defined as the “process of increasing the assets and capabilities of individuals or groups to make purposive choices and to transform those choices into desired actions and outcomes” (Santos, 2012, p. 346). This understanding seems closer to the concept of eudemonic wellbeing, which corresponds to the degree to which a person is fully functioning and feels alive, thriving and authentic (Stephan, 2018). In this vein, academics stress the ‘transformational change’ attribute of the social element that targets the “(…) underserved, neglected, or highly disadvantaged population that lacks the financial means or political clout to achieve the transformative benefit on its own” (Martin & Osberg, 2007, p. 35). In this light, ‘innovativeness’ plays a key role in finding new solutions for current social problems (Dees, 1998; Drayton, 2002; Mair & Martí, 2006; Martin & Osberg, 2007).

Academics defending an economistic stance note that social goals may be viewed as “the means by which profitability is achieved” (Peredo & McLean, 2006, p. 64). Scholars acknowledge that “(…) previous contributions suggest both the importance and uniqueness of combining social and economic missions” (Stevens et al, 2015, p. 1052) and that social entrepreneurial opportunities “(…) may emerge from a reframing, which encourages seeing people in need as clients instead of beneficiaries” (Dorado, 2006, p. 331). In this frame of thought, ‘commercial activity’ is a required attribute of SE. Some academics go beyond the need to ensure commercial activities to fund social goals, echoing economic rationality arguments of self-interest and profit seeking as legitimate aims of SE (Gruber & MacMillan, 2017; Zahra & Wright, 2016). Although few authors explicitly mention the attribute ‘willingness to take risk,’ creating economic value implies taking risks under an economistic view (Peredo & McLean, 2006; Tan et al., 2005).

Finally, denoting a rule utilitarian stance, in which moral behavior requires adopting rules based on the consequences they may have, some scholars include ‘commercial activity’ and ‘profit reinvestment’ as attributes of SE. Commercial activities and markets are needed to ensure resource efficiency and legitimate surpluses. However, unlimited profit distribution is sometimes viewed as a cause of mission drift (Dorado, 2006) and the condition to reinvest surpluses is critical as it “reverses the profit maximization principle by benefit maximization principle” (Yunus, 2006, p. 3).

Deontological Views

When conceptualizing SE, a group of academics tends to emphasize the morality of the actions under a series of rules, rather than defining SE based on its consequences as in previously described views. We found different deontological ethical stances focusing on motivations that seem to underly the choice of specific attributes of SE.

First, several scholars refer to helping people in need as the motivation for SE (e.g., Hartigan, 2006; Santos, 2012), which seems to relate to Kant’s imperfect duty of beneficence (Hill, 1971). However, some of such definitions also seem to add a utilitarian twist, explaining that, to be pursued, actions must create a positive net value. Assuming a clearer deontological stance, other academics seem to implicitly hold a Kantian view in conceptualizations that refer to moral obligations to attend to social goals as the motivation for entrepreneurial action (Hockerts, 2017; Mair & Martí, 2009; Mair & Noboa, 2006). In this light, ‘proactiveness’ can be understood as the direct action motivated by moral obligation, rather than the pursuit of opportunities with value creating outcomes.

Second, SE conceptualizations that normatively require a ‘collective governance’ attribute may relate to a Habermasian discourse ethics. Several scholars argue that assuming a collective governance structure contributes to manage tensions among different stakeholders (Bacq & Janssen, 2011; Defourny & Nyssens, 2008) and ensure an ethical conduct through transparency (Bruder, 2021).

Finally, evidencing a rights-based ethical stance, some academics view the social element as claims of individuals to the State or other supra entity, explicitly and normatively enumerating legitimate social goals as a set of specific human rights. For example, Seelos and Mair (2005) propose that the scope of the social element is defined by the MDGs and argue that “[u]nless we set boundaries to the scope of SE, it may be impossible to define the unique characteristics that differentiate it from traditional or business entrepreneurship” (2005, p. 244).

Virtue Ethics

The number of scholars that explicitly refer to the virtuous character of the social entrepreneur as a requirement to identify SE is scarce and refers to earlier articles in the literature. Martin and Osberg (2007) speak of courage and fortitude as fundamental attributes of the social entrepreneur. Perhaps more explicitly, Drayton (2002) claims that social entrepreneurs must hold an ethical fiber and be a “(…) good person you instinctively know you can trust” because “(…) social change usually requires those affected to make several leaps of faith—which they won't do if they intuitively do not trust the champion of the proposed change” (Drayton, 2002, p. 124).

Discussion of Stage III—Development and Refinement of the Conceptual Definition

Most SE conceptualizations relate to a consequentialist paradigm and are seldom associated with deontological or virtue ethics. Indeed, most definitions within the two main schools of thought—based either on ‘innovativeness’ or ‘commercial activity’– relate to utilitarian views, identifying SE phenomena through the social and/ or economic outcomes produced, in line with Bruder’s (2021) argument. However, our data show noteworthy differences among ethical stances related to attributes (the columns in Table 2). Additionally, there is a revealing overlap between the different ethical stances (the rows in Table 2) and specific combinations of attributes described in the four previously identified SE subtypes.

The first subtype, combining ‘social goals’ and ‘commercial activity’ and representing the most common combination of attributes, is often associated with an economistic stance, assuming the use of competitive markets as a mechanism that ensures ethical behavior. In ‘market-based’ conceptualizations, economic profits are a positive result of value creation and thus ethically acceptable if not desirable. Second, the ‘social business’ subtype includes ‘profit reinvestment’ in addition to ‘social goals’ and ‘commercial activity’ as key components of SE. Relating with a rule utilitarian view, such definitions require the condition of reinvesting legitimate surpluses to differentiate SE from commercial entrepreneurship and ensuring businesses focus on maximizing social benefits rather than personal wealth. The third SE subtype is typically associated with a hedonic act utilitarian view and includes conceptualizations focused on maximizing social value creation, rather than on economic value creation or capture. This subtype, labeled ‘efficiency-driven’ SE, requires ‘innovativeness,’ ‘scalability,’ and a ‘business-like approach’ as important ingredients to achieve resource efficiency and optimal productivity. Fourth, SE definitions that combine ‘social goals,’ ‘innovativeness,’ and ‘transformational change’ also place a central role on the attribute ‘social goals.’ However, social goals are here understood as beneficiaries’ empowerment and avoidance of their dependence, which relates closer to eudemonic wellbeing. Finally, both virtue ethics and deontological views are less popular stances found in SE conceptualizations.

While uncovering these ethical stances provides a fruitful ground in understanding different combinations of attributes, it is hardly expected they are reconcilable. The prevalence of distinct and irreconcilable normative views defining SE is an argument against the applicability of a classical structure to SE and supports the family resemblance structure of the SE concept.

Contributions, Conclusion, and Limitations

At an academic level, our work contributes to both SE and ethics literatures. Regarding the SE literature, we propose SE follows a family resemblance conceptual structure (Rosch & Mervis, 1975; Wittgenstein, 1953), a less common and often overlooked conceptual structure in the management field (Podsakoff et al., 2016). The improbable reconciliation of prevailing SE conceptualizations, previously acknowledged as a contested concept (Choi & Majumdar, 2014) driven by different worldviews (Lehner & Kansikas, 2013), is a strong argument against the use of a classical conceptual structure that defines SE as a universal set of necessary and sufficient attributes (Bacq & Janssen, 2011; Peredo & McLean, 2006; Santos, 2012).

The family resemblance structure is inclusive of the various perspectives of SE by uncovering how different attributes are combined into prominent SE subtypes and thus allowing the systematization of its scattered understandings. Our study identifies twelve attributes that comprehensively characterize the SE family resemblance concept and relate them with four emergent SE subtypes by systematically analyzing the content of 209 academic definitions, setting a sound path for operationalizing the SE concept and developing rigorous measurement instruments. We build on previous studies that suggested clusters of concepts to advance research in SE (Alegre et al., 2017; Choi & Majumdar, 2014). However, the attributes identified are characteristics of SE with clearer boundaries and less problematic overlaps. Furthermore, our study also described more nuanced meanings of attributes among distinct conceptualizations.

These findings caution researchers regarding a potential causal heterogeneity among SE subtypes. Although collectively part of the same conceptual space, SE subtypes may not be substitutable for causal purposes. For example, organizational tensions are potentially significant in the process of creating an organization that simultaneously pursues social and economic goals, but probably less important in other entrepreneurial processes that do not resort to commercial activities to fund operations. Likewise, social innovation is an expected outcome of innovative SE activities that promote the creation of new approaches to tackle intractable social goals but is not an obvious consequence of organizations that fund their social goals through commercial activities.

The family resemblance structure untangles the fallacy that results from using the same SE term to label different combinations of attributes. Much confusion in the SE field is a consequence of erroneously assuming that different things are the same because they share the SE label. Additionally, assuming a family resemblance structure accommodates the variety of existing SE definitions without oversimplifying it in only a few more popular subtypes based on, for example, commercial activities. Rather, this structure accommodates less common and potentially marginalized definitions, such as those based on virtuosity or collective governance.

This study also contributes to the ethics literature, specifically to the ongoing debate on how different normative ethical positions relate to distinct conceptualizations of SE. Previous studies in this field, under a consequentialist lens, argue that a social mission can only be attested by looking at the practical outcomes of interventions and discuss ethical decision-making under the influence of societal power relations (Bhatt, 2022; Dey & Steyaert, 2016). Defending a deontological stance, Bruder (2021) concludes that SE should be normatively grounded on deliberative democracy practices that can legitimately qualify as social beyond a mission-centric maximization principle. In our study, however, by adopting a descriptive stance, we uncover a diverse set of ethical views related to the selection of different attributes to define SE. We conclude that when considered as a family resemblance concept, SE is not bounded to a unique ethical reasoning but to several. Our findings indicate that academics predominantly use consequentialist ethical views in defining SE, specifically relying on act utilitarian and economistic stances. Deontological and virtue ethics are less frequently reflected in SE definitions. Yet, departing from previous studies (e.g., Peredo & McLean, 2006; Bacq & Janssen, 2011; Bruder, 2021), we refrain from normatively concluding which SE conceptualization and respective ethical stance should prevail. Rather, we assume that different ethical views provide legitimacy to distinct SE definitions and avoid arguments for the superiority of a specific conceptualization.

This inclusive understanding of SE has important implications to the study of its ethics. By incorporating different ethical views, predominant and less so, researchers can bring to the table several important ethical aspects of SE, adding to the ethical discussion about the consequences of SE, its motivations, and the virtuosity of social entrepreneurs. For example, future research could study SE’s social mission, a prominent attribute across SE conceptualizations by considering consequentialist or deontological ethical claims. Related to the former, scholars may study how SE maximizes the good and minimizes harm to society, is a source of hedonic and/or eudemonic wellbeing or creates social value while maximizing resource efficiency. On the other hand, considering SE and its mission through the deontological ethical lenses, future studies may illuminate motivations emerging from attending to human dignity as a moral obligation, or respecting human rights when addressing personal claims upon society. Assuming a deontological perspective, scholars may also investigate how deliberative democracy contributes to solving stakeholders’ tensions and promoting ethical behavior. This study offers academics and practitioners a valuable map of the breadth of existing ethical views underlying the concept of SE.

We acknowledge two noteworthy limitations in our study. First, the attributes identified reflect the conceptualizations present in a sample of academic definitions. Despite the vastness of definitions collected and analyzed, other attributes may exist in SE definitions not included in our sample. Second, the nature of a family resemblance concept is inherently complex, and we do not aim to oversimplify it by directly associating the choice of each attribute or combination of attributes with a unique ethical stance. In addition, distinct ethical stances not referred to, as, for example, the ethics of care, may provide fruitful ground to understand the SE conceptualizations and are potentially worth exploring in future research.