Social entrepreneurship and social change: a practice-based study in non-governmental organizations

Purpose – This paper aims to examine how social entrepreneurship (SE) practices give rise to social change in thecontextof urbanBrazil. Design/methodology/approach – The study draws on a broader inductive, ethnographic and iterative practice-based studyconducted inthreeBraziliannon-governmentalorganizations. Findings – Social change is established through intertwined practices that involve active interplay of ambivalent positive and negative feelings associated with the social mission pursued by the social enterprise; ﬂ at organizational structures that encourage participation and taking of ownership among all stakeholders; andfocused organizationalobjectives (social purposes). Research limitations/implications – The paper presents an analytical framework composed of ﬁ ve propositions that may be used in future research aimed at maturing and re ﬁ ning the understanding of SE. The study also provides a methodological contribution for future studies of new phenomenon and young ﬁ elds of research that often must rely on inductive methodologies, by demonstrating how an iterative thematic analysiscanbeused inpractice-basedstudies. Practical implications – This paper has practical implications directly connected to its social implications, because understanding how social change is achieved may enhance the effectiveness of SE practitioners in bringing desired changes about. Furthermore, the discussion also provided insights for practitionersto re ﬂ ectupon theparadoxical nature of practices aimed at social change. Originality/value – The study suggests a set of propositions and an original de ﬁ nition of SE that mitigates conceptual inconsistencies found in literature drawing on empirical data and by incorporating the political lens found in practicetheory.


Introduction
Over the past 30 years, the idea of social entrepreneurship (SE) has both emerged and gained notoriety within a wide variety of fields (Rey-Martí, Ribeiro-Soriano, & Palacios-Marqués, 2016). This is especially because of: SE being perceived as having the potential to tackle a wide range of the worlds' social ills (Barinaga, 2013;Kenny, Haugh, & Fotaki, 2020); a shared belief that SE may potentially improve human life on the planet (Barber a-Tom as, Castelĺo, De Bakker, & Zietsma, 2019;Cho, 2006;VanSandt, Sud, & Marne, 2009); and to a growing awareness of social inequalities and environmental issues (Hoogendoorn, Pennings, & Thurik, 2010;Margolis & Walsh, 2003).
This suggests that although SE is still a contested concept (Choi & Majumdar, 2014), it is a phenomenon rendered important given its promise of delivering desired changes (Dey & Mason, 2018;Lumpkin, Bacq, & Pidduck, 2018;Nicholls, 2010). However, how SE leads to social change remains little known (Dey & Teasdale, 2013;Haugh & Talwar, 2016;Lumpkin et al., 2018;Seymour, 2012;Spicer, Kay & Ganz, 2019). This gap undermines the development of this (still) young field of study (Dwivedi & Weerawardena, 2018;Short, Moss, & Lumpkin, 2009). To address this issue, it has been suggested that research should focus on the actual work of social entrepreneurs (Granados, Hlupic, Coakes, & Mohamed, 2011). This should allow us to address theoretical inconsistencies, especially the monological and tautological definitions of SE that are centrally bounded to taken for granted notion of social change (Cho, 2006;Spicer et al., 2019).
Therefore, the purpose of this study is to examine how SE practices accomplish the social changes implied by their social purposes. To do so, we used practice theory and data generated in a broader ethnographic study carried out in three non-governmental organizations (NGOs) in the Brazilian state of São Paulo. This data was analyzed through an iterative (Eisenhardt, 1989) thematic analysis that allowed for the generation of an analytical framework composed of five propositions. These results were, then, analyzed using a practice theoretical lens that allowed us to arrive at an original definition of SE.
Our contribution is threefold. First, the proposed analytical framework may be tested in future research aimed at maturing and refining our understanding of SE. Second, this paper provides a methodological contribution for future studies of new phenomenon and young fields of research that often must rely on inductive methodologies, by demonstrating how an iterative thematic analysis can be used in practice-based studies. Finally, we address the field's challenge of reducing its conceptual inconsistencies by offering a theoretically informed and data-induced original definition of SE.
The article is organized as follows. In Section 2, we review literature on SE by critically engaging with the emergence of the field and its underlying issues. We highlight how these issues render it necessary to extrapolate existing works. Section 3 explains the tenets of a practice-based lens and the ontology that inspired the methodological design of our study. In Section 4, we present an analytical framework composed of five propositions and their intertwined relationships that are then analyzed using practice theory to suggest an original definition of SE. Section 5 concludes by discussing theoretical contributions and implications for practice.

Literature Review
Although the practice of SE is said to exist for centuries, the concept of SE is relatively new (Rey-Martí et al., 2016). Authors disagree as to exactly when this concept was coined; however, it is widely agreed upon that in the last 30 years it has both emerged and gained an impressive notoriety within and across diverse fields (Bacq & Janssen, 2011;Bloom, 2012;Certo & Miller, 2008;Dahles, Dey & Steyaert, 2010;Lehner & Kansikas, 2013;Rawhouser, Cummings & Newbert, 2019).
The growing and widespread notoriety of SE are evidenced in, for example, the emergence of foundations that recognize, promote and support the practice of SE, such as Ashoka (founded in 1980(founded in ), EMES (founded in 1996, the Schwab Foundation (founded in 1998) and the Skoll Foundation (founded in 1999), which have been operating continuously since their respective foundations (Defourny & Nyssens, 2010;Haugh, 2005;Hoogendoorn et al., 2010;Partzsch & Ziegler, 2011); the establishment of academic programs and research centers dedicated to SE in business schools in some of the world's most influential universities (e.g. Harvard, Columbia, Yale, NY, Duke and Oxford University) (Defourny, 2010;Nicholls, 2010); and policy-makers and governments have also addressed this phenomenon (Choi & Majumdar, 2014;Zeyen et al., 2013).
Nonetheless, the general agreement that SE is a "good" and important thing set a fertile ground for the emergence of theoretically dissonant and unproblematic accounts of this phenomenon that trigger relevant challenges for research. First and foremost, it gave birth to a pluralistic field that has to struggle with conflicting definitions of SE that are often and tautological/monological (Cho, 2006). Even though authors suggest that SE is an essentially contested concept that should remain broadly defined (Choi & Majumdar, 2014;Darabi, Soltani, Nasari, & Emami, 2012), we argue that these may leave room for pernicious inconsistencies that extrapolate conceptual grounds.
Our suggestion is based on the argument that we should exclude profit-centered ventures and even criminal or terrorist organizations from a working definition of SE (Abdukadirov, 2010). We suggest that this can be achieved by including ethical and political criteria in the definition of SE that may be found in practice theory, as we will show in our discussion. Others have already stressed the necessity of discussing SE's ethical and political aspects (Dey & Steyaert, 2014;Cho, 2006).
Moreover, we argue that a working definition of SE should challenge the commercial entrepreneurship literature, instead of simply appearing as an accessorial branch of it. This is not, however, an easy venture. First and foremost, this task requires a re-appropriation of the concept of entrepreneurship itself by challenging the dominance of the economic paradigm that has been persistent in this notion (Diochon, Durepos & Anderson, 2011;Shane & Ventakamaran, 2000).
One way to achieve this would be to further investigate what is social in entrepreneurship: if entrepreneurship is not solely about creating economical value, what is it about? As pointed out by Valeau (2010), once we define and explore what are the noneconomic assumptions located within the entrepreneurial process, it may be possible to change its nature and exponentially build on the potential SE has shown to influence practitioners and policy-makers. This potential of SE has already been recognized by Seelos, Mair, Battilana and Dacin (2010).
In addition, it has been suggested by Fayolle and Matlay (2010) that SE is linked to significant evolutions in the entrepreneurship literature, because it can be perceived as an evidence of this field's growing ability to address changes and to balance economy and social well-being. However, Sundin and Tillmar (2010) point out that the strong connection between entrepreneurship and the private/business sector has led SE to theoretical, political and practical shortcomings.
Nonetheless, we acknowledge that there is a general tendency in the field of management and organizations of broadening its economical spectrum and trying to build understandings of social and natural issues that are embedded in the functioning of today's organizations (Suddaby, 2012). To enclosure entrepreneurship solely in economic terms would, therefore, contradict this general tendency, and do a disfavor to the development of SE as plural academic field of inquiry it clearly has the potential to be. Still, we argue that stablishing a clear difference between social and commercial entrepreneurship is necessary, as the application of entrepreneurship in its dominant form (based on an economic paradigm) to the social sphere has been contested conceptually, practically and ideologically (Parkinson & Howorth, 2008). Santos (2012) counterpoints this affirmation, arguing that bringing SE into economic and strategy theory may be helpful to these theories by showing their more humane side. Although we do not disagree with Santos (2012), we argue that, at the present moment, it seems more urgent to further mature the field of SE before making it useful for other fields of studies. The reason for this is simple. The aforementioned fields have a much bolder and more developed theoretical body and, thus, could easily trap SE inside their own domains and logics.
We do not believe that business ventures are necessarily incompatible with social value creation. However, a contradiction may lie on how these profits are distributed: they can either be generated to be accumulated in the hands of one individual [what could possibly contradict the purpose of SE, Spicer et al. (2019)] or they could be generated to be used to further extent the organization's social impact. Furthermore, NGOs, for example, may engage in commercial activities to support their social pursuit. Also, private businesses may generate profits that are solely distributed to support social welfare.
In short, we do not agree with the exclusion for profit organizations or even public organizations from a working definition of SE. Quite the opposite, we suggest that we could develop a definition of SE that embraces multiple manifestations of this phenomenon in different organizational settings. Thus, we propose initially using the definition of Cho (2006) that SE is a "set of institutional practices combining the pursuit of financial objectives with the pursuit and promotion of substantive and terminal values" (p. 36). This preliminary definition, however, is still not fully satisfying, as it does not account for social change and will be subjected to a revision in the discussion section.

Methods
A practice-based study raises methodological issues given the difficulties involved in researching emerging practices (Nicolini, 2012;Gherardi, 2012). Nonetheless, this is a valuable approach for young fields still in need of considerable theoretical advancements, as it is the case of the field of SE. Although practice approaches are diverse, they share a few common orientations that guided the methodological design of this research.
First, practices are not mere activities. Rather, they are an instance where activities form a pattern that embraces both a materiality and sense/meaning (Nicolini, 2012). Second, a practice-based study must focus on contexts. Although practices are the center of analysis, they are not its only focus. Quite the opposite, practices can only be understood if contextualized, they are not conceived as a readymade ontological reference that exists "out there" independently (Schatzki, 1996). Third, practices are socially sustained doings and not individual ones (Gherardi, 2012).
In short, this study has extracted from practice theory the following assumptions: It is real, concrete and living individuals who constitute reality through their practices.
Practices are socially constituted. Practices have a context or a background ["loosely understood as an arena or set of phenomena that surrounds or immerses something and enjoys powers of determination with respect to it" (Schatzki, 2005, p. 468)]. An individual acquires practical understanding by participating in practices.
Practices are recurrent phenomena.
Because a research strategy is deeply connected to a theoretical framework that justifies it (Oliver, 2011), the assumptions above translated into the following methodological orientations that were adopted by this study: SE practices are everyday achievements, recurrent phenomenon and, therefore, must be investigated in a bottom-up, emergent and regular manner. SE practices are socially constituted, therefore, this study must have multiple informants who are involved in them and not only Social Entrepreneurs (SErs). SE practices occur against a background. Therefore, the contexts where SE practices occur must be observed in a closed and detailed manner.
To understand SE practices, one must participate in them. In other words, participation is an essential element for the investigation of practices.
These general orientations, therefore, gave birth to this study's research strategy and design. To achieve our objective, we analyzed data collected in a broader ethnographic research that was conducted in three NGOs: NGO A (five volunteers, aimed at providing work opportunities to exconvicts); NGO B (25 volunteers, aimed at building houses for low-income families); and NGO C (49 volunteers/employees, aimed at providing full-time education for vulnerable children). Data collection lasted for 12 months, from March 2014 to March 2015, and consisted of regular and multiple weekly visits to each of the researched organizations. The participants of this study included founders (or the SErs, except in NGO C, as the founder had passed away prior to the start of the study); director, volunteers; employees; family members (of the participant organization's members); donors; and beneficiaries.
Data collection during field work was done mainly through participant observations and through the engagement in reflexive conversations with research participants (Watson, 2011). During these observations, data was collected mainly through the use of field notes. Documents were collected whenever they were available or when they were of public access (such as documents, books, website content and advertising materials, videos and survey results). This secondary data was gathered to help the researcher understand the institutional and/or material context of the field.
In short, the following data was gathered: a total of 37 h of recorded conversations with key informants; three notebooks with field notes containing observation data made during 48 visits to the studied organizations; and additional institutional documents (legal documents; reports; books; website content; e-mails; institutional videos; interview videos; newspaper; and magazine articles).
Our analysis was conducted after data was collected in its entirety, when an iterative thematic analysis was undertaken with the assistance of the NVivo 10. This iterative process relied both on data and was informed by practice theory, in a process Locke, Feldman, and Golden-Biddle (2020) have summarized in Table 1.

Results: crafted propositions
In the initial stage of the thematic analysis, 16 themes that were coded from the data 10 times or more were identified. Codes that either had an overlap of coded references of more than half in the code matrix and/or that could be reduced to a narrower single analytical category were merged, producing seven second-order themes: structure; resources; external relationships; moral authority; stakeholders' mobilization and participation; feelings; and social purpose ( Table 2). The thematic analysis produced a holistic view of SE practices, allowing us to explore several of its facets. It also allowed us to make five propositions that aggregate the findings. As noted by Gioia, Corley and Hamilton (2012), the use of formal propositions is not a necessarily positivistic approach. Quite the opposite, building and presenting propositions may strengthen the contributions made by inductive studies such as this.
Based on an iterative analysis of the coded references and analyzing how the codes overlap, as described in our methods section, we were able to suggest the following proposition model (Figure 1 P2. Participants' moral authority extracted from displays of spirituality and strong beliefs shape their capacity to convince stakeholders to participate in SE practices.
P3. The interplay between negative and positive affect by effectively showing or portraying social ills enhances stakeholder's participation in SE practices.
P4. There is an internal tension between the objective social purpose of the organizations (their end "product") and the aspired social change, with a focused and effective "social purpose" working as vector that enhances the possibility of achieving social change.
P5. Social change is achieved in the unfolding of SE practices in an ongoing processual and cyclic manner through the participation of stakeholders in the organizations' activities.
5. Discussion: analyzing findings under a practice lens In light of practice theory, representing findings in the form of themes and propositions is not enough to build a critical account of SE practices. Rather, "It isn't simple, but the result is fabulous, I've never seen anything more beautiful in my entire life" "We face really sad stories, but there are also very beautiful stories" "This makes me feel awful . . . The talking is very beautiful, it's wonderful, but in practice. . ." "This is a tough situation. These children who are born in these poor neighborhoods, poor them, they have very little chance. What an awful thing, tremendously. We don't even know what to think" ". . . sometimes I feel very angry" "You must have sensitivity, to look at the situation and be moved" "Sometimes I cry; such beautiful feedbacks we receive" Material/economic Structure (69) Resources ( "Citizenship is what we seek to teach when we donate a house. We are not a solely welfare Project. We give assistance and donate a house [which is citizenship lesson] " "Our main purpose is to give these young people opportunities, because most of the times they lack having perspectives and opportunities in life" "Here we try to do things differently, and it's never easy to introduce in society a different way of doing things, a way in which everyone can contribute politically. Even simple things, like to try to implement something for everyone's wellbeing in your own apartment building, if in order to implement it you need to involve everyone" Source: Research data SE and social change emergent patterns found in the data. Therefore, they allow us to capture various aspects of practices. In this study, we explored seven themes, and each of these theme has made SE intelligible from a different perspective. In this sense, we suggest that using themes was a crucial analytical tool to achieve our main objective, because it allowed us to explore SE practices' "multifaceted nature" (Erden, Schneider, & Krogh, 2014) and also mitigate the difficulties that are recognizably present in the operationalization of a practice-based research. Nonetheless, the division of different dimensions of practice is not a precise one, because the three dimensions are intertwined and often overlap (Kemmis, Wilkinson, Edwards-Groves, Hardy, Grootenboer, & Bristol, 2014, p. 34), as Figure 2 illustrates.
While approaching this study's research question, it is important to remind readers that the key gap we found in SE literature is the recurrent tautological

What is social change?
Social change is a complex and polyphonic notion that may pertain to many aspects of society (which may be technological, economic, religious, ideological, demographic, stratificational, etc). and that has always been at the heart of sociological inquiry (Etzioni & Etzioni, 1964). In the wide spectrum of thinkers who have given social change a central place in their ideas, we have two who are located in its opposing ends: Marx and Weber. Although Marx stressed that social changes were to be sought in the modes of production and exchange (a materialistic approach to it), Weber demonstrated the importance of the spiritual/religious sphere in determining the economic structure itself (thus creating a cultural or symbolic approach to social change) (Etzioni & Etzioni, 1964). Although we will not build on this discussion, we argue that practice theory may provide us with a way of tackling and balancing both the economic/ material and discourse/cultural dimensions of social change.
Practice theory provides us the following assumptions that will reverberate in how we conceive social change: Practices constitute the locus where human discursive and material instances become possible and acquire meaning. All practices are social, and their resulting structures and systems in constant process of being enacted and maintained.
Therefore, social change must be regarded as an intrinsic element of practices. With these assumptions born in mind, we adopt the following preliminary definition of social: social change is any change that involves the violation of some aspect of the social status quo (LaPiere, 1965).

Social change and social entrepreneurship
As suggested by our findings, although the social purposes of the studied NGOs were diverse, they were intrinsically connected to a general aspiration for social equality among participants. Our findings suggest that this characteristic differentiates the studied organizations apart from purely charitable or from welfare organizations, which cannot be labeled social enterprises. The aspired social change was achieved not as end product or as an output, but in a processual and constant manner as the SE practices unfolded. It is important to highlight that being processual seems to be the very nature of SE, as their goals are never fully realized, mainly because this realization would mean that these organizations would be no longer necessary. Still, participants in this study were aware that in their "ideal" world, their work would not be necessary. In fact, in the case of NGO C, they expressed the intention to hand over their organization to be run by a state institution if it could keep its operations: Even though SE organizations have different social purposes, their idea of social change is commonly based on a general aspiration for social equality.
This draws us back to the discussion on what makes SE different from commercial entrepreneurship: commercial enterprises do not have the (almost utopic) aspiration to make themselves obsolete or part of the state apparatus. Quite the opposite, commercial entrepreneurship organizations are constantly seeking ways to perpetuate themselves in time and space. This point also makes so-called "businesses with social impact" or certain notions of for-profit SE troublesome, as they may be contributing more to the maintenance of a certain social status quo that gives them the reason to exist than to its effective change. In this sense, there is a clash between the "profit logic" and the logic of SE practices and, if assuring effective participation and involvement of stakeholders is key for the success of SE practices, it is the locus where social change takes place, the latter is inevitably connected with democratic practices.
We suggest that this is evidence of the centrality of the subject of social change in the realm of SE, as it may show an inner contradiction of such practices under certain circumstances. Giddens (1979) asserts that: This may explain why such contradictions exist in some organizations engaged in SE: although actors believe to be driving certain social change, they may in fact be reproducing precisely the social structure that they wish to change. In fact, the author explains that this is a paradoxical tendency that is present in all social practices, precisely because they are recursive, reproduced and routinized phenomenon.
This suggests that one must be very careful when accepting, for example, the theoretical claim made by Calas, Smircich and Bourne (2009) that the act of engaging in entrepreneurship itself (or in our case, SE) produces social change. Although this may be supported by our findings, it cannot be overly generalized or accepted without caution. In the case of this cited work, it is also important to observe that the authors do not analyze SE specifically. Furthermore, they use feminist theory and not practice theory to make this claim and restrict the idea of "entrepreneurship as social change" as "one social activity among others in a masculine-dominated society, and its capability to promote or hinder positive social change for women" (p. 558). Although we would not like to relativize the importance of feminist perspectives also for the study of SE, and although we acknowledge that Brazilian context is highly patriarchal and male dominated, we must also acknowledge that gender and male domination did not emerge as substantial issues surrounding SE practices during our field study. We must also note that one of the founders of the studied NGOs was a woman, and the main researcher did not perceive any significant gender-related differences in the functioning of these three organizations.
Practice theory may also help build upon the claims of Calas et al. (2009) with further implications to SE literature. Although Giddens (1979) warns us about the tendency of practices to reproduce social structures, there is still a revolutionary potential to them, which may in fact support the claim and explain how engaging in SE itself can be viewed as a social change process (Schatzki, 1996, p. 68).
Another clue that this participation effectively changes the previous status quo was the many times SEr's narrated they were recurrently portrayed as "crazy," "insane" or "out of their minds." Given this recurrent reaction, we suggest that the initiators of the studied organizations were indeed proposing something "out of the ordinary" to their peers and close social circle, which itself may suggest that this involved effectively changing their shared beliefs, ways of thinking and ways of doing. This is also evidence that once participants let go of this initial resistance to join the efforts of the studied organizations, this engagement of was more like a "leap of faith" than a rational decision.
Furthermore, this study showed that a constant reflection surrounding the phenomenon of SE (both in the theoretical and in the practitioners' realm) should be: Are so called SE practices effectively achieving social change? Or are they contributing to maintaining the very social problems they are addressing?
To incorporate this reflection, we propose rephrasing the definition of Cho (2006) that SE is a "set of institutional practices combining the pursuit of financial objectives with the pursuit and promotion of substantive and terminal values" (p. 36). Although Cho (2006) has inserted an ethical criterion in this definition when he mentions substantive and terminal values, we argue that there is still a gap in this definition concerning the political aspects of SE that are inevitably present in the concept of social change.
To find this political criteria, we combine the latter definition of social change with the description made by Schatzki (1996) of local politics as a set of activities that recognize "a variety of shifting sources and structures of oppression, misery, and discontent that aim to alleviate specific sufferings through opposition to the particular formations responsible for them" (p. 5). Therefore, we propose the following original definition for SE: SE is a set of organized institutional and political practices, which violate the social status quo by operating against a variety of sources and structures of oppression, misery and social inequalities through the bridging of the interests of multiple stakeholders.
We suggest that this definition may encourage a critical reflection upon the need to set apart organizations that are effectively pursuing to implement actual social change from other types of organizations that are more consonant with maintaining current social structures stable. However, we acknowledge that this is still a broad definition, because social change may occur in many forms. In the specific studied context, social change meant achieving social equality, as class inequality is a major part of the status quo of the studied context (namely, the context of urban Brazil). Besides, we were able to identify that this was an ongoing achievement triggered by the participation of multiple stakeholders, from antagonistic social groups, who were convinced to engage in SE practices. In short, we were able to elaborate the following definition of social change: Social change is accomplished in a processual and ongoing manner as stakeholders from antagonistic social groups participated in SE practices.
It can be said, therefore, that practice theory contributes to the literature of SE and also to the literature on new forms of organizing because it brings about a reflection with an undeniable political taste: it puts light on the paradoxical nature of practices and their role in changing and/or maintaining current social structures. This approach is provocative as it portrays the thin line that divides practices that sustain the social status quo and those that challenge it.

Conclusion
The theoretical contribution of this paper consisted of the proposal of a nontautological/monological definition of SE that incorporated the political lens found in practice theory and also a contextualized notion of social change. Furthermore, we presented an analytical framework composed of five propositions that may be used by future research on SE. These studies could either test the propositions in other contexts or deepen the discussion of these findings with theoretical papers on the subject.
This article also provided a methodological contribution by demonstrating how an iterative thematic analysis can be used in practice-based studies. We showed that the found themes served as circumscribed units of analysis that helped make sense of the complexity of practices.
The bottom-up and emergent elucidation of how participants in practices of SE bring social change about also provided an empirical contribution, as understanding how this process actually occurs may enhance the effectiveness of SE practitioners in bringing desired changes about. Furthermore, the discussion also provided insights for practitioners to reflect upon the paradoxical nature of social practices.