When given two choices, take both! Social impact assessment in social entrepreneurship

ABSTRACT

This paper examines how social entrepreneurs construct impact arguments as they begin to assess social impact. We examined the experiences of 68 social entrepreneurs in Chile and discovered that the construction of arguments for the purpose of thinking about and experiencing impact is different than the arguments constructed to establish dialogues around it. We explain this dual argument construction as arguments for worth and arguments for legitimacy. We expand scholarship on argumentation by clarifying social entrepreneurs’ efforts to pursue adherence facing competing demands and reinforcing their willingness and ability to engage with social impact assessment. We advance the understanding of social impact assessment in social entrepreneurship across three areas: tensions, accountability and performance and extend Nicholls’ general theory by explaining what precedes the discursive space where the assessment of social impact reconciles facticity and validity to establish materiality.

KEYWORDS:

• Social impact
• social impact assessment
• argumentation
• claims
• social entrepreneurship
• accountability

1. Introduction

Social entrepreneurs are considered central to society because they generate responses to social and economic inequality (Hjorth 2013). What comes along with these efforts is an onus of assessing the social impact thereof (Haugh 2005). This is the space of social impact assessment (Molecke and Pinkse 2017; Muñoz, Gamble, and Beer 2022; Ormiston 2023), which involves assessment practices aimed at ‘understanding, measuring, and reporting the intended or actual contribution of actions that focus on addressing societal problems’ (Ormiston 2023, 990). This contribution often materializes as positive changes-in-condition of the beneficiaries. These practices are particularly important for social entrepreneurs, as they can help them communicate impact (Grieco, Michelini, and Iasevoli 2015), spot un-realistic venture ideas (Barton and Muñoz 2023), offer evidence of change (Branzei et al. 2018; Molecke and Pinkse 2020, accelerate impact or value created (Kroeger and Weber 2014; Wry and Haugh 2018), and lay a discursive ground to construct future opportunities (Discua Cruz, Hamilton, and Jack 2021).

For social entrepreneurs and those tied to the social ventures (recipients, donors, funders, etc.) social impact assessment (SIA) poses particular challenges because the data used to capture changes-in-condition and the empowering processes has an uncertain nature attached to it (Nicholls 2009, 2018). SIA requires social entrepreneurs to capture and communicate evidence about social impact based on a set of socially embedded values. For example, the number of books given to children in need as a measure of improved educational outcomes is based on the belief that equitable social progress is important. Yet, equitable social progress can mean many different things, e.g. educational access, health outcomes, and threshold income levels. It can also be viewed differently across geographic regions, and thus measured in many ways over the life of a social entrepreneur. The collection and communication of relevant and meaningful information about impact is generally a challenge for social entrepreneurs, who continue to struggle to make sense of and showcase the important work they do.

Unfortunately, there are more questions than answers in academic literature. In trying to implement social impact assessment, Nicholls’s (2018) theory suggests that social entrepreneurs are required to engage in dialogue and consensus building, which can only be structured around the impact arguments constructed by those orchestrating efforts. This triggers tensions, which remains to be the primary focus of academic literature (e.g. André, Cho, and Laine 2018; Molecke and Pinkse 2017), and complicates the delineation of what SIA involves in social entrepreneurship across three areas: tensions, accountability, and performance. This opens a gap in the literature, because in the absence of guidance, SIA needs to be informally constituted through the interactions between social entrepreneurs and stakeholders, which tends to perpetuate tensions. So, in this study we ask how impact arguments are constructed by social entrepreneurs as they engage in social impact assessment. This gap needs to be resolved to better understand tensions in SIA and advance a significantly under-theorized space in social entrepreneurship (Rawhouser, Cummings, and Newbert 2019).

Our study was conducted in Chile, where we collected data through interviews with 68 social entrepreneurs and several engagement activities between 2016 and 2019. We explored social entrepreneurs’ engagement with SIA, particularly how arguments regarding changes-in-condition were being made and supported. We ground our work in argumentation theory (Rieke et al. 2008), where we paid particular attention to the making of claims, the evidence provided in their support, and the links between the two.

From our analyses, we derived two different ways in which arguments are constructed by social entrepreneurs: 1. Arguments for worth, which are constructed upon the enfoldment and projection of changes-in-condition, which they use in their attempts to be held accountable to themselves. 2. Arguments for legitimacy, which are constructed upon the objectification and compression of changes-in-condition, through which they attempt to be held accountable to others. These findings stem from a bifurcation of social entrepreneurs’ reference points for impact as they construct arguments. Leveraging argumentation theory, we argue that the construction of arguments as a way of thinking about and experiencing impact is different from the ones constructed to establish dialogues around it. Although this process creates dissonance and frustration for social entrepreneurs, dual argumentation allows them to pursue adherence when facing competing demands and reinforces their willingness to engage with SIA.

We make several contributions to theory. Firstly, we advance our understanding of impact assessment in social entrepreneurship. We offer two ways in which impact arguments are constructed as social entrepreneurs engage with SIA and show how they might contribute to thinking about, experiencing, and building dialogues around impact. In doing so, we offer a new way of seeing tensions in social entrepreneurship as contrasting arguments for understanding and communicating impact, which in turn lay the ground for the construction of materiality in social entrepreneurship. By revealing new tensions in SIA, we provide substance to the plasticity and the vulnerability of SIA, seen as central to the establishment of normative frameworks for the assessment of social impact (André, Cho, and Laine 2018; Molecke and Pinkse 2017, 2020). Secondly, we expand on Nicholls’ (2018) work by revealing novel aspects of social impact assessment in social entrepreneurship and explaining what underlies the construction of communicative frameworks used to assess and communicate relevant and meaningful information about impact. These insights provide analytical and theoretical blueprints to be used across entrepreneurship research as it pertains to the introduction of impact assessment practices and accountability for social outcomes (Power 2015; Rawhouser, Cummings, and Newbert 2019). Ultimately, SIA is not just about counting units, indicative of impact, but about constructing arguments to make sense of, understand and communicate impact. In this way, we encourage researchers and practitioners to embrace the expressive and emancipatory role that SIA might play in social entrepreneurship.

2. Theoretical grounding

Social impact assessment

Social impact assessment (SIA) refers to the processes and efforts to applicably describe, measure, report, and guide, the resources, inputs, processes, and/or policies, to arrive at some prosocial outcome (Grieco, Michelini, and Iasevoli 2015; Nicholls 2009, 2018; Rawhouser, Cummings, and Newbert 2019). The role of SIA, in the process of improving societal and environmental interests and outcomes (Gamble and Caton 2023), has garnered significant international interest, because it helps social entrepreneurs, funders, and policymakers determine whether interventions create intended positive changes-in-condition for beneficiaries (Beer and Micheli 2018; Hall, Millo, and Barman 2015; Hehenberger, Mair, and Metz 2019). The benefits of SIA include its capacity to inform strategic choice, develop legitimacy, and guide resource allocation decisions (Rawhouser, Cummings, and Newbert 2019). SIA originated as an accountability mechanism for monitoring the social benefits of public interventions and more recently is being applied by policymakers and impact investors to determine whether social enterprises are delivering on their promises to make a positive contribution to society (Beer and Micheli 2018; Hehenberger, Mair, and Metz 2019; Stephan et al. 2016). However, unlike financial accounting, SIA has no agreed-upon standards of measurement (Nicholls 2018).

Early scholarly work on SIA has pulled from practitioner-based works and offered normative contributions (Ebrahim and Rangan 2014). This has led to a plethora of SIA tools and frameworks for impact and investment measurement (Stone, Carttar, and Colby 2004), such as blended practitioner disclosures to engage with their stakeholders.¹ Even in, Maas and Liket 2011 uncovered more than 30 different approaches to SIA, varying in temporal dimensions, perspectives, and purpose. More recently, leading scholars have begun to criticize the normative focus of SIA scholarship and have made calls to fill the gaps in SIA theorization (Cooper, Graham, and Himick 2016; Hall and O’Dwyer 2017; Hall, Millo, and Barman 2015; Nicholls 2009, 2010). SIA continues to be contingent, contested, and under-theorized.

Nicholls’s (2018) general theory is a critical step towards SIA theorization, as he elaborates on the distinctiveness of the accounting practices involved in SIA and several of the inherent tensions. Muñoz, Cacciotti, and Cohen (2018, 2024) reinforce this point by showing how tensions are likely to be experienced as social entrepreneurs perceive and attempt to make sense of incipient prosocial efforts and be held accountable. These are efforts to describe, measure, report, and control the short-term and long-term aspects and changes in their impact efforts (Power 2015). Nicholls (2018) argues that these tensions emerge from a lack of clarity on boundaries, units of analysis, processes, shared understandings, and intended engagement targets. This is particularly so in settings where reference points for social performance and feedback mechanisms are ambiguous (André, Cho, and Laine 2018). In the scholarly literature, the focus has largely been placed on the tensions emerging in the early stages of SIA formalization (e.g. André, Cho, and Laine 2018; Molecke and Pinkse 2017) to explain how social entrepreneurs enact accountability systems facing uncertainty around data and measurement tools. These authors hint at an important gap, namely, that in the absence of guidance, social impact assessment tends to be informally constituted through the interactions between social entrepreneurs and stakeholders, which tends to perpetuate tensions.

Drawing on a well-established assessment tool – social return on investment - Nicholls (2018) explains how the tensions can be reconciled in the assessment of social impact. Nicholls proposes that this is done through communicative actions, which are used by stakeholders to find a consensus on the values of social life. This occurs through dialogue, free speech, and continuous interaction, which facilitates a common understanding instead of having the values imposed by an external order. He emphasizes how determining and communicating facts with SIA requires stakeholders to operate under socially derived values and forms of consensus building. In his view, Nicholls sees the novelty and promise of SIA in its ability to enact facts and values such that the emancipation of those suffering is possible. Through this process, what constitutes positive social change can be established.

Nicholls’ (2018) theory makes a significant contribution by explaining how tensions are resolved and how accountability can be established in the assessment of social impact. Yet, it does not get so far as to explain how entrepreneurs enact SIA in the process leading to normative frameworks. Where Nicholls’s general theory frames a big picture of a well-ordered ‘here and now’, the ‘before here and now’, when entrepreneurs begin to engage with impact assessment, remains hidden and thus under-theorized. We know that tensions surface in the definition of what information is relevant and meaningful for the assessment of social impact (Molecke and Pinkse 2017) and that these can be resolved through argumentation and consensus building (Nicholls 2018). However, we do not know how arguments for social impact are constructed by social entrepreneurs, despite their centrality to the development of communicative actions. This, in turn, limits our understanding of how SIA is constructed by social entrepreneurs and the empowering processes by which accountability is established.

Argumentation theory and the construction of impact arguments in social entrepreneurship

To explore the construction of impact arguments in SIA, we ground our work in argumentation theory (Rieke, Sillars, and Peterson 2008; Salmivaara and Kibler 2020; Toulmin 1994). Making an argument involves the construction of a coherent series of claims intended to establish a point (Rieke, Sillars, and Peterson 2008; van Werven, Bouwmeester, and Cornelissen 2015), which in our context of interest, is at the core of the communicative process of advancing, supporting, criticizing, and modifying impact arguments (Toulmin 2003). Arguments, in Toulmin’s (1994) original model, have three main components: claims, data, and warrants. A claim is a statement of something to be true, which, to be accepted as plausible, needs to be supported by data that act as the foundational reason behind the claim. A warrant links information to a claim, legitimizing the claim by showing the information to be relevant. Rieke et al. (2008) argue that the main objective of argumentation is to gain adherence, which can be thought of as the informed support of others. In this sense, arguments are context-specific since the acceptance or rejection of a claim is made within groups of people who have previously agreed upon (explicitly or implicitly) what constitutes an acceptable reason. For an argument to be accepted or rejected, it first needs to be made and supported. This is the focus of our work.

Construction of arguments in the context of SIA

Through this lens, we argue that as social entrepreneurs begin to assess social impact, they also engage in the construction of arguments to understand and communicate impact. This is more salient in emerging settings, where actors begin to informally think about social value and structure how such value can be captured, assessed, and communicated. We further argue that social entrepreneurs can create and activate alternative arguments, as part of their thinking and reasoning and in their attempts to engage with internal and external stakeholders. These are formed in context as they perceive and organize information about social problems, beneficiaries, and social interventions. Arguments construct meaning around how social value – materialized through changes-in-condition of beneficiaries – is delivered, assessed, and communicated. As entrepreneurs begin to conceptualize impact assessment mechanisms, such as SIA, they will likely produce a plurality of arguments given the many ways in which people understand the world and make meaning. In the case of SIA, it might depend on how people understand social problems affecting communities and how they think about and communicate their interventions and related impacts. Given the probable multiplicity of arguments, the emergence of a particular SIA is likely to be more complex than counting and reconciling social and economic outcomes. This line of reasoning underpins our examination.

3. Research methods: sample and data

Our study was conducted in Chile’s emerging social economy. Like most of Latin America (Bryer 2011), Chile offers a unique context in which to observe social entrepreneurs engaging with social impact assessment. In the late 1990s, Chile witnessed a surge in social demands leading to multiple civil society initiatives, which in turn led to the rapid expansion of social entrepreneurship, impact investment, and support programmes. The movement grew stronger from 2011 onwards (Muñoz, Kimmitt, and Dimov 2020), with a new ‘social enterprise economy’ marked the foundation of Sistema B (B Corps in Latin America), the first social investment fund, the first international festival for social innovation, the first Social Lab in Latin America, and the first nation-wide incubator for social entrepreneurs (Footprint Program). Social impact assessment is becoming an increasingly institutionalized practice in social entrepreneurship (Ormiston 2023). Chile appears to be the exemption, as the country still lacks relevant legal frameworks and formal rules for social industry accountability. The result is a social industry lagging behind many similar countries when it comes to formalizing approaches and tools for assessing social impact (Kimmitt and Muñoz 2018). This turns the Chilean social industry into an intriguing case, that allows us to observe how social entrepreneurs come up with their plans, tools, and actions to assess their social impact.

Our data collection took place over four years (2016–2019). We purposively selected 68 social entrepreneurs running social ventures that have been operating in markets for more than 12 months and have begun to assess the social impact of their actions. These social entrepreneurs operate across a range of industries including social finance, communication, and design, culture, sports, packaging, software development, health, business consulting, hospitality, apparel, recycling, amongst others; whilst tackling a diverse range of social and environmental problems, for example, poverty, drug addiction, deforestation, lack of education, financial exclusion, and mental health. An overview of the sample can be found in Table 1.

Table 1. Overview of participants.

Industry	Number of interviewees
Agriculture	2
Construction	2
Consulting	4
Craft	1
Education	14
Entrepreneurship support	5
Food	6
Health	12
Law	1
Marketing	1
Media	4
Music and arts	2
Property development	1
Recycling	4
Sanitation	1
Software	6
Urban development	2

Semi-structured interviews

We conducted 68 semi-structured interviews with the founders of the selected social enterprises. The interviews lasted between 50 and 70 minutes. These were recorded and transcribed in Spanish and subsequently translated by a native Spanish speaker, which allowed us to retain cultural nuances and the richness of the local language. The interview guide was divided into five sections with a total of 19 guiding questions, all related to decisions and circumstances during the early stages of their social enterprise. Additionally, we asked questions about the history of the social enterprise and its beneficiaries, social mission; resource acquisition, and institutional context; and we explored particularly their efforts to understand social impact and its measurement. This included the perceived social value being created by their social enterprise, their intended and perceived impact on beneficiaries, their future expectations, and how they make sense of, measure, and communicate such impact. Finally, we gathered substantial information on the problem the social entrepreneurs sought to resolve and the products or services they offered to customers and beneficiaries.

Participant observation and engagement

We organized three follow-up discussion groups and one collaborative workshop in Chile. The first two in March and October 2017 were led by one of the authors of the study, and the third one by the three authors simultaneously in May 2018, lasting, on average, two hours each. These follow-up discussion groups allowed the authors to dig deeper into the perceptions of the social entrepreneurs and their prior discussions about SIA. In March 2019, we organized one last gathering with a group of 20 founders and 10 managers of social enterprise incubators. Unlike the previous three events, during this workshop, we presented our early findings to the audience to observe their initial reactions and to garner their feedback. This allowed us to further corroborate the validity and practical relevance of our findings and go beyond the ‘façade’ of published communications.

4. Data analysis

For an argument to be accepted or rejected, it first needs to be constructed. Argumentation theory directed our attention to statements and supporting evidence that social entrepreneurs use in the construction of impact arguments, which is the focus of our data analysis. Our data analysis procedure draws on Toulmin’s work and was guided explicitly by Salmivaara and Kibler’s (2020) approach. We moved from an open examination of passages of text expressing impact arguments to the formal examination of the structure of impact arguments. We looked at claims and support information, exploring patterns in claims/data combinations, which allowed us to assess and differentiate the distinct features of impact arguments. We then proceeded to interpret the reasoning behind impact arguments, leading to different argumentation types. In the final stage, we further abstracted from our findings and sought to explain what might motivate the different ways of constructing impact arguments. In the following, we explain these four analytical stages in detail and show how our analysis was conducted. In what follows, we present our analyses in detail.

Preparation and structuring of text

We began our analysis by reading the transcripts, focusing on mission statements, explanations of problems and solutions, reflections on beneficiaries’ needs and required changes-in-condition, explicit impact statements and accountability tools used to capture, understand, and communicate impact. This allowed us to gain a comprehensive understanding of the overall content and most specifically of the claims relating to impact and assessment. We were also able to capture how the data to support claims was collected and processed, which allowed us to better understand the kind of reasoning underlying their construction. From here, we extracted all relevant segments of text from the 68 interview transcripts which contained an explicit reference to impact and accountability, e.g. statements explaining changes-in-condition – current and intended – and the evidence provided to support those claims. Since the arguments contained in the interview data are unstructured, we proceeded to structure impact statements following Toulmin’s (1958/2003) delineation of claims, data, and warrants. We looked at what impact is being claimed by the venture, what data is used to support that claim, and the rationale used to connect data to claims.

Given the nature of interview data, claims and data are not always connected to each other in a single quote or in an immediate manner. Impact claims can sometimes be made in the explanation of the mission statement and data provided in the explanation of how impact is made sense of or measured. We also found that impact arguments can be of complex or simple elaboration. Some entrepreneurs provide a broad idea of the kind of change they would like to see in the world. Prison education, for example, states: ‘we seek to make a change, to try and get people to have access to better education’. Others provide a much more detailed explanation. Reading programme explains ‘The social return of our initiative doubles the government’s impact because every peso we get, goes to the children. We complement public education and improve performance (SIACE test) in language and mathematics’.

Considering the above, we structured each argument using the following logic: “Venture X claims to make an impact Y [claim] because of Z [data] and the argument is valid since Z logically links to Y [warrants] in the context of X. In Table 2, we provide examples of data preparation, showing how we moved from raw data to structured argumentation. In the case of Virtual Reality, the venture claims to be making a positive impact on children who have not yet seen the ocean. They offer evidence relating to the use of virtual reality, which allows users to recreate the experience of being close to the ocean. This is a valid argument because offering such opportunities to disadvantaged children is seen as inherently positive.

Table 2. Data preparation: illustration.

Raw data	Claim	Data (because)	Warrant (since)
We have seen the impact that it creates when a child, who does not know the ocean, uses our VR visor to live the experience of being at the ocean, and that is an impact that excites you, it is very strong, so that type of impact is what we want to promote, a positive impact (Virtual reality).	Venture can make a positive impact on children who do not know the ocean.	Use of VR visors can recreate live the experience of being at the ocean.	Offering opportunities to disadvantaged children is positive.
The impact can be tremendous, if we manage to make these tools and technologies available to the most vulnerable schools, particularly to the first six years of primary education, it would be a transcendental impact on society (Fab Lab).	Venture can make a transcendental impact on society.	Tools and technologies are not currently available to the most vulnerable schools.	Accessible tools and technologies can significantly improve education in vulnerable areas and learning in the first six years of primary education
I think the impact is quite multidimensional. The impact is created around three main benefits, which is a person with a disability who had never had the possibility of accessing a job and begins to work, it changes their life. It means becoming great, independent, having the chance to earn your salary, and from then on, a world of possibilities opens up for you. For the family, they can also see the capabilities of a son who until that moment, was probably seen as a much younger kid (Disability).	Venture can make a multidimensional impact on people with disability and their families.	Opening job opportunities for people with disabilities who had never had the possibility of accessing a job can change their lives in many ways.	Accessing a job and beginning to work allows them to become independent, having the chance to earn a salary, and from then on, a world of possibilities opens up for them.
Our impact involves [pet] adoption rates and caring. We saw people downloading our app […] then [they] told us that they were now taking better care of their pets. That was very rewarding and validated our ideas. So, that is the only impact we had experienced and when we lived through that experience [seeing people taking better care of animals], it was rewarding, for us, for those who have supported and have collaborate with this project (Lost Pets).	Venture can increase pet adoption rates and improves how people treat their pets.	Users who downloaded the app are now taking better care of their pets.	App can change how people take care of their pets. The higher the number of app downloads, more users will start treating their pets better.

Examination of the features of argumentation

In the second stage, we assessed each structured argument in detail to identify categories of claims-data combinations as the basis for understanding how impact arguments are constructed. In our categorization, we focused on the distinct features of the argumentation, rather than its specific content as this allowed us to compare and find argumentation patterns across cases. We noticed a conceptual bifurcation, with claims-data combinations pointing in different directions and reflecting a duality in the construction of impact arguments.

On the one hand, entrepreneurs were putting forward value judgements that could not be checked against data. They were also relying on their ‘intuition’ and ‘experience’ to understand impact. We noticed entrepreneurs using subjective terminology, such as ‘transfers of karma’, ‘spirituality as social contribution’, or ‘new linguistic archetypes’. This reflects an inward-looking, close connection between the individual, the social problems, and the radical (future-oriented) changes they were trying to produce, make sense of, and communicate. Here, impact claims are based on the entrepreneurs’ purpose and value judgements regarding desired changes-in-condition. These changes involve radical societal transformations. Cultural Space for example, explains that ‘Because it is related to growing up, with spirituality, with my beliefs, we are all linked and we are all the cast, so we have to be partners and make a social contribution and at the same time I am immersed in the same environment and those same people are part of me’. Here, information about impact – used to support impact claims – can be perceived as embodied within social reality and obtained via close interaction with beneficiaries and their circumstances. The same case continues ‘It fills my heart and soul what I am doing, it allows me to meet people, I am happy with what I do’. Through it, social reality can be considered holistically. Radical transformations are situated in the future (outcomes), which can only be described and imagined prospectively, as an alternative to the present. Scholarship supports their claim by arguing ‘The main impact, I would say, is to support the development of the country, it is to support people to learn and improve and to deliver more of their capabilities to the country’.

On the other hand, entrepreneurs were making claims to affirm that certain elements existed relying, not on their ‘intuition’ or ‘experience’, but rather on ‘something they could count’ to understand impact. For example, claims about changes in the ‘work population’ were linked to ‘an indicator that things are going well’ or ‘numbers and quantities’. This reflects an outward-looking, distant connection with beneficiaries and a rationalization of gradual (present-oriented) changes. In these instances, objective terminology was used, such as ‘how many people contact us’, ‘sensible currencies’, ‘number of parks built’, or ‘number of people attending a policy meeting’. The conceptual bifurcation suggested duality in the construction of impact arguments.

Here, impact claims are based on descriptions of a selection of aspects of social reality, reflecting narrow changes in conditions. These changes communicate gradual progress expressed often in units of service delivered that make reference to immediate outputs. Craft in prison, for example, explains impact as ‘ … the numbers of women we have worked with and the income they receive. Those are our indicators, because the issue of training is a little more difficult to measure, we have never measured it’. Here, information about impact – used to support impact claims – can only be observed and detached from the self – distant interaction, which affects the extent to which information about problems, actors, actions, and changes are owned by the entrepreneur and how the self is involved in what is being perceived. Through it, social reality is atomized and condensed into representative parts. Units delivered, a reflection of gradual progress, are situated in the present (outputs) and are communicated through the quantification of social facts. Science Education, for example, justifies impact through ‘We can write it [impact] down in numbers. The new network has 54 teachers, 390 children, four ministries, two social enterprises, and three countries’. Despite the complexity of educating vulnerable children in science, impact data to support the claim refer merely to the size of the network. In Table 3 we offer illustrative evidence of this bifurcation, showing how claims-data combinations were pointing in different directions, within and across cases.

Table 3. Bifurcation in claim-data combinations: illustrative evidence.

Venture’s impact claims	Bifurcation in claim-data combinations (quotes)	Claim-data combination interpretation	Distinct features of impact arguments
Test better prepares low-income students for national tests.	1- There are parents who at the end of the year give us cookies because we changed their lives, 70-year-old gentlemen who thought he would have never been able to take the test again and it was incredible, like those there are thousands of stories. It all happens at a personal level.	Impact claim draws on experience and close interaction (talking with parents and exchanging stories and gifts)	Embodied social values
	2- We recruited a group of economists and concluded that for every hour that a student spends studying mathematics for the National Test, it rises about one point in the PSU, so basically if you dedicate 100 hours to it, you will increase an average of 100 points.	Impact claim draws on objective accounts and distant interactions (economists measuring study hours and test performance).	Stylized social facts and gradual outputs
Ecology Education improves quality of life by promoting ecological lifestyle.	1- In the end one realizes what the needs are in the world, it is a sensitivity that some people are born with, and I would not know how to explain it, it is a willingness to help or do something to improve the situation in the world.	Impact claim aligns purpose and radical changes needed in the world (I want to help change the world)	Embodied social values and transformational outcomes
	2- Impact measures could be … having our projects running, delivering workshops, having more people joining the organization, I also hope to be able to pay salaries, that would be a super good indicator that things are working well.	Impact claim aligns expectations with gradual progress (communicating small gains).	Stylized social facts and gradual outputs
Indigenous media restores peace and social justice by covering positive news in the Mapuche region.	1- [In the context of violence and social conflict amongst indigenous communities] I humbly believe that we have managed to contribute to generating a new visual and archetypal language of the indigenous world in Chile.	Impact claim reflects personal contribution to promotion of radical transformation (new visual and archetypal language for the indigenous world)	Embodied social values and transformational outcomes
	2- I believe that first, the number of projects executed are many, there is a lot of content created, very prolific, there are more than five hundred audiovisual productions, more than 50 events, 57 public tenders, which also point to this line, to this matter, sponsorships, donations, it has been a very prolific set of actions	Impact claim focuses on external engagement and immediate outputs (numbers of projects executed).	Stylized social facts and gradual outputs
Support Network promotes alternative model and improves solidarity and well-being in the workplace.	1- The impact would be to see people helping each other, strangers […] that people start to trust each other a little more. There is not going to be any great change, no cultural transformation if we do not begin to see the stranger as a close one, almost like our friends and our family. As long as we experience this division of us ‘friends’ against them ‘strangers’, it will be difficult for us to achieve societal change.	Impact claim draws on appreciation of the complexity of social reality through close connections (building trust in communities)	Embodied social values
	2- [… But] our only KPI we use is the ‘transfers of karmas’, that is our number, and we are careful with that. We also measure how many people contact us and any other [number] we can come up with.	Impact claim draws on a reduction of social reality into representations (karma points).	Stylized social facts
Responsible consumption increases awareness of responsible consumption.	1- The issue here is how to build a better economy. By promoting responsible consumption, we can create a much more participatory, more social, and fairer economy. And we believe that the state and companies must realize that.	Impact claim is presented as an imagined alternative future (building a new economic system)	Embodied social values and transformational outcomes
	2- We create indicators such as tests, collaboration agreements, and school or teacher satisfaction.	Impact claim is presented as quantifiable present outputs (satisfaction surveys).	Stylized social facts and gradual outputs
Science Education improves science education in Latin America.	1- We participate in roundtable discussions, to collectively understand what kind of education we need for climate change.	Impact claim draws on experience and close interaction (participatory roundtables)	Embodied social values
	2- Our videogames [used in environmental education] have been downloaded many times, I don’t know the numbers, but I know that we have them somewhere.	Impact claim draws on objective accounts and distant interactions (videogames downloads).	Stylized social facts

By cross-examining what underlies the bifurcations shown in Table 3 (column 3), we were able to identify four claim-data combinations configuring four distinct features of impact arguments (Table 3, column 4). The first feature - Embodied social values - centres the point of reference for impact on internal values, suggesting that change-in-condition can only be appreciated by experiencing complex reality. The second feature – Stylized social facts equally – centres a point of reference for impact but around external facts, which can only be appreciated by observing a simplified reality. The third feature - Transformational outcomes - explains the kind of change-in-condition intended (transformational) and how it might materialize in the future (outcome). The fourth feature - Gradual outputs - explains the kind of change-in-condition being delivered (gradual) and how it materializes in the present (outputs).

Clustering of distinct features

We then proceeded to cluster those features based on the patterns previously identified. From our observations, we noticed that outcome-oriented arguments containing transformative claims rely on value judgements and the data to support those claims was obtained through close engagement. We thus grouped Embodied social values and Transformational outcomes under Type 1 argumentation (Table 4). Conversely, we noticed that output-oriented arguments containing claims reflecting gradual progress rely on factual claims, and the data to support those claims was obtained through distant interaction and observations of specific aspects of social life. We thus grouped Stylized social facts and Gradual outputs under Type 2 argumentation (Table 4). Table 4 provides an overview of the distinct features.

Table 4. Structure of impact argumentation: distinct features and types.

Distinct features	Claims	Data	Illustrative examples
Type 1 argumentation
Embodied social values	Impact arguments contain claims that reflect the entrepreneurs’ values, purpose, and sense of worth, regarding desired changes-in-condition.	Impact claims are supported by granular experiential accounts relating to complex reality, captured through inter-personal connections with beneficiaries.	We are all thankful and we managed to deliver that peace that is super difficult to achieve, here it just happens. There are always people who are unhappy but the truth is that, what one does, is needed and generates an impact on certain people. Feeling that you are their support network, and that they feel that way, and that we have already managed to reach a balanced relationship where neither of us [beneficiaries and entrepreneurs] take advantage of each other (Nursery). Claim: Venture brings peace to communities Data: Feeling that the venture is the beneficiaries’ support network. Warrant: Emotional support creates balance and deliver peace.
Transformational outcomes	Impact arguments contain claims that reflect future outcomes leading to transformational change.	Impact claims are supported by a portrayal of a better future, expressed as holistic outcomes catalysed by intervention.	Our impact is that we seek to make a radical change [in Chile], to try and get people to have access to better education and through that to sustain their lives (Education). Claim: Radical change in education Data: Better quality of education Warrant: Better education can transform lives
Type 2 argumentation
Stylized social facts	Impact arguments contain claims that reflect stylized facts about itemized changes-in-condition in the life of beneficiaries.	Impact claims are supported by numerical evidence relating to aspects of social reality, captured through observation and distant interaction.	Impact to us is that everyone has free and equal access to quality information and independently. It [impact] is the kilobytes of content added to the internet by individualized users, basically their contribution, we also measure website hits and how much information is reposted (Wiki). Claim: Equal access to information. Data: Kilobytes of content added to the internet by users Warrant: More content added and reposted by individual users signals access.
Gradual outputs	Impact arguments contain claims that reflect present outputs proving gradual progress.	Impact claims are supported by numerical evidence relating to units of service, expressed as materialized atomistic outputs directly resulting from intervention.	We are making tools and technologies available to the most vulnerable schools. We are capturing our impact today basically by looking the amount of people served. We are going to work with 10 schools across the region. We have progressed from having three projects last year to having more than seven today. Then our impact would be assessed by looking at the courses delivered (Fab Lab) Claim: Tools and technologies improve education outputs Data: Number of schools served, and projects executed Warrant: More schools served, and projects executed can improve education outputs in vulnerable schools.

Interpretation of the reasoning of the argument types

From the identification and grouping of claim-data combinations, we proceeded to uncover and examine warrants to explore the different reasoning connecting claims and data (See example in Table 1, right-hand column). Toulmin explains that warrants represent the (implicit) reasoning underlying the argument structures and are crucial for understanding the logic embedded in and intentionality of the argumentation. In this sense, to explore warrants, we evaluated the reasoning that explained the nature of the impact claim: what underlies a valid impact statement and the role of the data: how the evidence offered supports the impact statement.

Following this procedure, we abstracted the reasoning underlying the two argumentation types (claims-data combinations) and elaborated them as logical statements. By assessing the abstracted reasoning, we were able to add meaning to the two impact argumentation types, suggesting two different ways in which arguments are constructed by social entrepreneurs as they engage with SIA: one that enfolds and projects and one that objectifies and compresses social problems, actors, and actions. Table 5 shows the two types of impact argumentation, the reasoning behind each of them, and their basic argumentation structure.

Table 5. Impact argument construction types.

Type of impact argumentation	Argumentation structure (claim)	Argumentation structure (data)	Reasoning (warrants)	Implications for SIA
Type 1 argumentation: Enfolding and projecting	Enfolding and projecting argumentation focuses on values and holistic transformational change, catalysed by value-based intervention.	Enfolding and projecting argumentation is supported by contextual and experiential accounts obtained through close interaction.	Because impact reflects social values underlying intended transformational changes-in-condition, impact arguments should be constructed by enfolding and projecting social problems, actors, actions, and change.	Impact can be captured, understood, and communicated by experiencing reality of beneficiaries and embracing values that underlie and catalyse transformative changes-in-condition. Arguments are constructed to think about and experiencing impact. E&P argumentation allows for pursuing internal adherence, aligning impact claims and purpose, and reinforcing entrepreneurs’ willingness to engage in SIA.
Type 2 argumentation: Objectifying and compressing	Objectifying and compressing argumentation focuses on stylized facts about aspects of social life and gradual change- in-condition, resulting from discrete outputs of the intervention.	Objectifying and compressing argumentation is supported by data reflecting objectification of social life, obtained through distant observation.	Because impact involves social facts about gradual changes-in-condition, impact arguments should be constructed by objectifying and compressing social problems, actors, actions, and change.	Impact can be captured, understood, and communicated by measuring and expressing output-based facts about present changes-in-condition. Arguments are constructed to establish dialogues around impact. O&C argumentation allows for pursuing external adherence, aligning impact claims and expectations, and reinforcing stakeholders’ willingness to engage in SIA.

Abductive theorizing

To help explain these divergent ways of constructing arguments, in the last analytical stage, we returned to the literature on social impact assessment and argumentation for inspiration. Leveraging argumentation theory, we revisited how arguments were being made and supported by social entrepreneurs and the reasoning underlying the two types. This led us to theorize on a dual construction of impact arguments in SIA. This duality is not trivial as the two types get tangled in the process of constructing arguments, where clashes appear between what seems meaningful to the social entrepreneur and what seems meaningful to others in terms of impact creation. It is, ultimately, a complex ordering of information that requires bringing the personal sense of purpose and experiences of impact into alignment with external expectations and the beneficiaries’ actual needs. This is clear in the quote below, showing tensions between who the entrepreneur is and what s/he thinks her/his role is when it comes to conceptualizing impact as they engage with SIA.

We [social entrepreneurs] don’t like to waste time on that [because] the motivation of the social entrepreneur is not to measure the impact, it is to make an impact. I know what my impact is [referring to changes in education in Latin America] and you are ok if tell you ‘I have 800 thousand users who spend more than half an hour on the site’. If it is going to cost me 10 million pesos to do an impact study to be able to show that indeed our users improve, it is not interesting to me, because with that same money, I can make an investment to develop an app so that they can study more. This is why social entrepreneurs are in this game of never measuring their impact. (Test)

Our theorization of dual construction of impact arguments involves contrasting ways of (1) Thinking and experiencing impact, and (2) Working toward establishing dialogues about impact, which are aligned with different attempts to be held accountable. Through the first type of argument construction – enfoldment and projection – social entrepreneurs attempt to be held accountable to themselves and their entrepreneurial sense of purpose (Cardon et al. 2009, 2017; Muñoz, Cacciotti, and Cohen 2018). We term this Arguments for worth. Through the second type of argument construction – objectification and compression – social entrepreneurs attempt to be held accountable to others and face external pressures for impact (Barman and MacIndoe 2012; MacIndoe and Barman 2013; Muñoz, Gamble, and Beer 2022). We term this Arguments for legitimacy. We summarize these insights in the discussion section.

5. Findings: the dual construction of impact arguments in social entrepreneurship

The analysis of claim-data combinations and reasoning allowed us to conceptualize two ways in which arguments are constructed as social entrepreneurs begin to engage with SIA. Next, we elaborate on each impact argument construction type presented in Table 5.

Constructing impact arguments by enfolding and projecting

The enfolding and projecting argumentation focuses on values and holistic transformational change, catalysed by value-based intervention. Here, argumentation is supported by contextual and experiential accounts obtained through close interaction. The main rationale for this type of impact argumentation follows the idea that because impact reflects social values underlying intended transformational changes-in-condition, impact arguments should enfold (experience and own) and project (see into the future) social problems, actors, actions, and change.

Enfolding: embodied social values

Through enfolding, the intricacy of social reality is embraced by social entrepreneurs and is perceived by experiencing its nuances. In constructing impact arguments, social entrepreneurs described social impact in their organization as ‘filling their heart and soul’ and driven by a ‘willingness to help’ or to ‘improve the situation in the world’. We observed the participants’ perception and understanding of impact emerging from their embracement of self-realization and building connections between social problems and individual purpose. Impact encompasses intrinsic motivations for a cause, and the felt experience of meaning by pursuing it. These motivations drove the work they did within their socially oriented ventures and were an integral component of the impact being created in each one. For example, a participant from the film industry was inspired to use culture as an outlet to address social issues such as addiction and healthy eating habits. The impact claim that the venture can resolve addiction through films stems from the founders’ spirituality and beliefs:

Because it is related to growing up, with my spirituality, with my beliefs, we are all linked and we are all part of the cast, so we have to be partners and make a social contribution and at the same time I am immersed in the same environment and those same people are part of me.

Impact claims thus leverage previous experience with a particular setting, issue, or target group to achieve greater purpose and meaning in their own life. When speaking only about impact, the social entrepreneurs also discussed a range of past and present experiences, showing how they use their experiential knowledge and passions for other people or spaces to address social problems. Many of the social entrepreneurs in our study told us how their past experiences, whether home or work-based, fed into and informed the impact of their social enterprise. For example, a participant originally from a small town noticed how disconnected inner-city individuals were from each other. Therefore, the participant strove to rebuild trust and connection with an online platform for exchanging kindnesses in his community. Impact information was obtained through personal connections. In his words:

When you go to the city and have many options, actually you can solve many things through money, [but] you lose that sense of community and everything is focused on having more and more money. So that would be an antecedent that I come from a small town where people help a lot, and I want to recreate that sense of belonging to a community. (Community consulting)

Experiences drive the perception of impact. While the number of products or services created, sold, or distributed matters to a certain degree in all these social enterprises, it was not the same thing as the experience of impact itself – the visible complex changes-in-condition of the beneficiaries. A venture supporting women’s empowerment in rural areas reflects on how data to support claims is obtained. S/he argues that objective measures (number of products sold) are restrictive and human development can only be understood by looking at the person behind the product:

That is our super strong focus, human development, we do not get so much out of the product itself, but rather out of the development of the person behind that product to be able to provide them with the means to reach economic sustainability that is very necessary for rural areas”. (Women’s Empowerment Enterprise)

The conditions that were changed across the ventures in our study included physical, mental, emotional, spiritual, and cultural. How these social entrepreneurs knew the impact had been created was something fundamentally felt:

It [impact] has to do with benefiting my client once they begin to experience the service they celebrate and they thank them in a very important way, it is the impact on lives, I can actually see, literally, happier people, I see the fulfillment of the mission (Integrated Elderly Care)

The essence of impact is thus characterized by inward-looking emotional processing. Unlike the cognitive activity elicited by measures to describe and communicate impact, impact arguments for social entrepreneurs were supported through the recognition and processing of experiences, feelings, and emotions felt during and as a result of social interventions.

In processing information about changes-in-condition, impact arguments are thus supported by looking at the nuances of social reality. This is done through establishing relationships and interactions with beneficiaries experiencing and recognizing internal and external transformations. This is how this argument organizes information. In the words of a Cultural Educator:

My objective is for the beneficiaries to live the experience. That’s why I opted for this movement because something is delivered, so for me the experience is important, feel the experience that the project leaves them, that they can expand, as ambassadors of this project.

Projecting: transformational outcomes

Arguments for impact can rely on and require different time markers to explain change-in-condition. Projecting situates change-in-condition as an incommensurable better future. The entrepreneurs expressed that social impact is something that would occur over the long term, not something that would happen overnight. Elderly care’s impact, for example, is situated in 50 years and the underlying intention is to radically change the health system: ‘[We want to] eradicate the paternalistic approach to aging care […] we work for the happiness and independence of the elderly and their families’.

Long-term impact requires multiple routes to progress, thus the embracement of self-realization interplays with a commitment to social healing, both aimed at constructing a radical societal change. Radical social or environmental restoration is the social aspect of impact that manifests as equalization or normalization of imbalances in society or the environment. Social entrepreneurs constantly highlighted that true impact would result in drastic changes to the economic or institutional structures of society. As explained by the founder of a social enterprise with an educational focus:

We are looking for a better quality of life for people who do not have access to education or culture and because they do not have access to information, it costs a lot more to get out of there. So, by using the least resources as recycling, great things can be achieved, we believe that it is possible to have another kind of world.

The latter presents an irreconcilable separation between an imagined alternative future and a realistic present, a ‘post-neoliberal dream versus a neoliberal today’, as many of our interviewees make claims about social impact. The effect of a social intervention is therefore visualized and measured differently depending on the relative position of the agent, which brings to light the inner ideological tensions when the different impact argumentations are brought together. An argument about social impact, in this sense, communicates radical change, which is done by projecting impact into the future, as a desired state, using prospective thinking. In the words of Easy tenancy ‘[…] we envision a place that they can call home, where they want to continue living, going to the same places, talking with the greengrocer, following the life they have always had’.

As seen in the quote below from a Media enterprise, which strives to include and communicate the culture of endangered indigenous communities, impact is discussed as fixing a problem in a neoliberal system that they argue is broken:

[about the Mapuche conflict], no one seems to understand the social and cultural aspects of it, it is always political, criminal, or economic. We don’t see the injustices [in the system]. Multiculturality has always been part of us, but it is not visible […] I humbly believe that we are contributing to creating a new visual and archetypal language for the indigenous world in Chile.

The entrepreneurs spoke of the various ways in which the impact of the enterprise targeted ‘vulnerable groups’ or the ‘poorest in the community’, or those ‘with a disadvantage’ to create some form of holistic transformation, which changed the lives of individuals. For example, in one venture where they rebuilt community parks, the founder spoke of how such activities can transform multiple aspects in highly vulnerable areas:

It is the lack of sociability, the deficiencies that exist in the most vulnerable neighborhoods that have an impact on the quality of social, family, individual life, lack of spaces of encounter, to generate trust among the neighbors … They are the communities in which we develop projects, they are urban communities, with a certain degree of social vulnerability that they want to transform where they live that can be a landfill, or an abandoned place where crime occurs.

For another participant who worked with women from impoverished Chilean neighbourhoods, impact was not just about ensuring that these women had economic livelihoods, however important that is, but also that they felt empowered, and ‘changed’ in their worldview of the value of their own life:

In Chile, women have always been at a disadvantage in comparison to men when it comes to finding a job… Women simply cannot get into the working world or it is very difficult, so we believe that it is important to generate flexible work and it is what has brought us good results … In addition, they learn to do things that generate well-being … super-important self-image … or know that they are women who have never ever worked, then feel able to make a necklace, a bracelet that changes the vision of themselves and that is the most relevant. (Craft in prison)

Likewise, for a community networking enterprise, social impact would mean that neighbours in the targeted areas started to trust each other, a holistic achievement that would not happen on the first encounter: ‘Our impact? We have not seen it yet. This is a long-term project. Impact would be when strangers help each other, that is our aim. Fostering trust among people’.

In sum, enfolding and projecting involves argument construction practices whereby entrepreneurs develop a sense of ownership over problems, actors, actions, and change, hold them closely whilst appreciating their complexity (enfold), and direct impact forward into an imagined, incommensurable future (project). It thus appears as claims that social entrepreneurs use to become accountable to themselves as they connect their values to others in the construction of a better future.

Constructing impact arguments by objectifying and compressing

In the objectifying and compressing argumentation social entrepreneurs are focused on stylized facts about aspects of social life and gradual change- in-condition, resulting from discrete outputs of the intervention. Here, argumentation is supported by data reflecting the objectification of social life, obtained through distant observation. The main rationale for this type of impact argumentation follows the idea that because impact involves social facts about gradual changes-in-condition, impact arguments should objectify (commensurable aspects of that reality) and compress (outputs that materialize in the present) social problems, actors, actions, and change.

Objectifying: stylized social facts

In making impact arguments, the social entrepreneurs perceive changes-in-condition by simplifying and distancing themselves from social realities. Social complexity is objectified, thus rendering it manageable by observing representative aspects thereof. This form of claim-making directs them towards a homogenous and detached view of reality. It constructs statements about changes-in-condition by focusing on outputs, which are observed through a singular lens, such as parks built, money raised, courses delivered, or apps downloaded. As expressed by the founder of a social crowdfunding platform:

We want a better and more democratized world for social entrepreneurs in Latin America, [but we measure that] we have raised 3 million dollars for 12 social entrepreneurs in the region in the last three years, from 668 investors.

In this sense, objectifying is characterized by outward-looking cognitive processing. It helps entrepreneurs understand impact in a cognitive sense by providing formulas and language for the physicality of its drivers and outcomes. In the case of sustainable farming, for example, the self is removed from the change-in-condition, and the measurement indicators set by the founders – indicative of impact – would only enable them to understand how many resources were being generated. Similarly, in speaking about a new impact strategy being developed with a group of sociologists, the founder of Automation mentions:

This is a new strategy for social impact […] the environmental part, the one that involves waste, there will also be indicators: how many kilograms we reuse a day and how much biofertilizer and biofuels we generate.

Whereas enfolding and projecting construct arguments about building a better future, the objectifying reduces lived experiences of beneficiaries into fractional representations of reality. It does so by reducing organizational activities and objects to numbers, features, and properties, as it directly relates to the desired change-in-condition. By committing to reductionism, the objectifying and compressing argumentation structures the organization of information that provides the foundational reasons behind the claims made.

When asked about SIA, the participants also spoke of focusing on particular aspects of the impact they were creating, but also, intriguingly, noting how this did reflect the impact of their enterprise. For one participant who worked with children to develop their understanding of environmental problems (and how to combat them), it was possible to count certain organizational objects involved such as the numbers of teachers and students, but not the impact of learning about these subjects. He explains:

For the impact of understanding environmental problems, we are already building this network of teachers who teach climate change in their schools and we are linking more or less 60 teachers participating in the process. We have also worked with 130 children from Chile, 130 children from Colombia, 130 children from Peru but it was not considered in the project an evaluation of impact on learning … but we can write impact in figures. (Environmental education)

Likewise, a female entrepreneur who worked directly with individuals suffering from HIV, women who had had an abortion, and pregnant teenagers, told us that certain things can represent the impact that the enterprise creates, but that any of them only signalled aspects of success towards ‘integrating these people into the conventional life system’:

You can study in this population how much their quality of life has improved after participating in the foundation’s programs or for example how many women who experienced abortion have improved their post-abortion syndrome with the treatment that we could deliver, that kind of things … that is a way to measure the impact that can have. (HIV care)

Compressing: gradual outputs

Social enterprises highlighted another aspect of impact – how it allowed for representational signals of progress in their enterprises, even if those representations were not considered to be the impact created. A social enterprise, that supported social entrepreneurs in the country with seed funding [Crowdfunding], recognized that the impact of their work stemmed far beyond any monetary exchange, but in effect what they were measured on were numbers of pesos exchanged:

It is more difficult to measure the impact and we have never done it, we have never asked him ‘hey it has been good for you to be in [our firm] ?’, on the other hand, what we have achieved in income, is super measurable, we have raised 1,580 millions of pesos for 12 Chilean enterprises in the last three years, from more than 680 investors.

Compressing guides the communication of gradual progress. It is utilized to contribute to incremental improvements, generally framed in operational terms. Unlike projecting, which makes claims as fixing a problem in a neoliberal system that they argue is broken, compressing makes claims by using the language of the same system and linearity to communicate how impact would be achieved in a way that anyone can understand. Ultimately, ‘[impact claims are …] public, open, easy to understand, and it is a language that is common to everyone’, emphasizes the founder of Eco-heating. The participants described how measures enabled the monitoring and diagnosing of the amount and state of resources, and the use of language deemed legitimate by external stakeholders, as described by the founder of an education service provider:

Impact? … we are considering it […] it would be super interesting to be able to measure our impact, mainly because it is a factor that is asked a lot when applying for funding, and yes, it would be very good to have that.

Impact arguments seemed to rely on and require different time periods to explain progress. Unlike enfolding and projecting, in this type of argumentation, the time between the action and its stated effect is discounted. For instance, in the context of closing educational or economic gaps across disparate social groups, the internal transformations required, and the many circumstances affecting an individual’s development path are unlikely to be expressed by a single measure of numbers in an area registered on the programme, nor the time they spent on it. As evidenced in the quote below, from the founder of an educational platform, this unnoticed long-term vs. short-term tension triggers a disconnection between what they say the impact is and what they actually measure:

[Our impact is] to close the educational, social and economic gap in Latin America … we [measure that we] have around 800 thousand registered users, they spend like 25 minutes in our platform, they improve academically, learn new things, and can perform better in the national tests. (Test)

This form of impact argument construction, in this sense, situates changes-in-condition as a quantifiable present. Whereas enfolding and projecting support claims by making reference to the construction of a better future, objectifying and compressing argumentation guides people to focus on present deliverables, because knowing whether that impact has been achieved requires situating measures in the present and the necessary information can only be captured retrospectively: ‘benefiting our clients once they begin to experience the service’.

While relevant to make sense of the world around them, the limitations of representational work were acknowledged. The founders interviewed all expressed that although SIA was occurring inside their enterprises, it was not the impact itself that was being captured and communicated by these measures. Some founders expressed how the construction of arguments for impact was impossible, because such things as poverty reduction, for instance, were impossible to reduce into any one variable to measure:

The bottom line is to reduce poverty in indigenous communities or anyone in a poor sector, and I do not put numbers because obviously, it is difficult to have a goal because different variables usually influence it.

Similarly, the founder of University Entrepreneurship believes that the impact of building entrepreneurial bridges between universities and markets can be transformative, but he argues that ‘it can’t be measured, at least quantitatively […] we can measure the number of beneficiaries, workshops, and attendance, following Corfo’s logic (government funding), but that’s not it’.

The arguments used by the enterprises, while deemed legitimate, do not portray for our participants the desired changes-in-condition. Change in social habits, mindsets, and behaviours is iterative and messy (Leiserowitz, Kates, and Parris 2006), not linear. Arguments for social impact simply aggregate units arithmetically instead of dealing with discontinuities or exponential change required for deeper transformations. This type of argument construction involves reducing impact to certain figures or representing isolated aspects of success. ‘If you look at the data, we have helped 22,000 children. I reduce everything to that’, stresses the founder of Reading Program. As evidenced in the quote below from a social enterprise that teaches IT skills, this leads to serious causal attribution errors in relation to what a single social action can achieve, particularly when causality in social complexity is overly simplified:

We see that global change, as a phenomenon, is a complex problem. But at school, you are taught to think in linear terms, so we lean toward mono causality, at most we associate two variables. But the systemic vision is not taught, and that was also a contribution of what we try to work on like children can deliver their systemic capacity of understanding.

Such endeavours necessitate a range of changes, that are not then considered when they are asked to discuss the measurement of their impacts. Impact is framed as a change-in-condition for beneficiaries which is intended to lead to new fundamental societal structures, but social enterprises struggle to connect these broader societal changes to features captured by this type of impact argumentation. The focus becomes therefore the direct changes occurring in activities, processes, and people around the organization. As explained by a founder of a water sanitation social enterprise: ‘[Discussing whether there is a] government-independent sustainable water system for rural areas (vs.) counting the number of people having access to clean water annually [is not the same thing]’. (Clean water).

In sum, objectifying and compressing involves argumentation practices whereby social entrepreneurs begin to observe problems, actors, actions, and change from a distance, structure them into manageable parts whilst focusing on commensurable aspects of that reality (objectify) and look backward for impact explanations as outputs materialize in the present (compress). It thus appears as claims that social entrepreneurs use to become accountable to others as they show how external expectations are being met using social facts relating to their present actions that people can understand.

6. Discussion

In this paper, we discovered two different ways in which arguments are constructed as social entrepreneurs engage with social impact assessment, which stem from a bifurcation in their reference points for impact as they construct arguments. To elevate our findings, in the following section we abductively elaborate an explanation of what this dual argument construction means for social entrepreneurs and what allows them to do.

Towards a theory of dual construction of impact arguments

Combining our findings with literature on social impact assessment and argumentation, we abductively theorize that the dual construction of impact arguments is linked to differing attempts to be held accountable – to themselves and others. The dual argumentation comes from confronting assumptions about what the social entrepreneur is and ought to achieve in the world – its sense of purpose (Muñoz, Cacciotti, and Cohen 2018) – and how this purpose can be used to generate legitimate impact from the perspective of outsiders, such as institutions, investors, and beneficiaries (Barman and MacIndoe 2012; Muñoz, Gamble, and Beer 2022). From here, we expand our exposition of types, explaining what these two types of construction impact argument are for, with the former being Arguments for worth and the latter being Arguments for legitimacy.

Arguments for worth refers to inward-looking practices that allow social entrepreneurs to make and support claims about impact, through which they are held accountable to themselves by referencing impact as a way in which they connect their values to others to positively influence society (Cardon, Post, and Forster 2017; Stephan et al. 2016). Through these inward-looking practices, social entrepreneurs understand social impact, and their role in it, by referring to an embodied understanding of lived realities and changing conditions (i.e. previous lived experience, physical closeness to beneficiaries) (de Rond, Holeman, and Howard-Grenville 2019). This allows them to appreciate the complexities of the different social realities they interact with through their work. They further determine whether impact has been created by articulating their commitment to creating an incommensurable better future, which is portrayed as a radical change. This form of accountability is deeply tied to the social entrepreneur’s purpose and sense of self-realization (Cardon et al. 2009, 2017), as it pulls from previous experiences with social problems, close relationships with and time spent listening to beneficiaries, and the entrepreneur’s emotional responses to present circumstances. Meaning for self-accountability is thus made based on the social entrepreneurs’ developing principles (Cardon, Post, and Forster 2017) and reactions to experiences with others in need, as they co-construct ideal futures for themselves and their beneficiaries (Beckert 2021) and create pathways to fulfil their ambitions.

Arguments for legitimacy refers to outward-looking practices that allow social entrepreneurs to make and support claims, through which they are held accountable to others using measures to interpret what are the external expectations for impact and to determine whether those expectations are being met. Through this outward-looking reference point, social entrepreneurs make claims about impact, and their role in it, by distancing themselves from it in pursuit of neutrality and objectivity (Sandberg and Tsoukas 2011). Claims are made based on what others’ needs are from the social enterprise, and the institutional pressures concerning priority impact areas (Arvidson and Lyon 2014, Barman and MacIndoe 2012; Muñoz, Gamble, and Beer 2022). Social entrepreneurs understand impact as their commitment to delivering outputs, constructed as numbers, features, and properties. To do so, they explore complex social situations by observing chunks of evidence, explain the future in present terms using retrospective objects (in place of prospective potentials), and communicate large societal transformations by pointing towards granular progress. Here, social entrepreneurs paint pictures of whether they are meeting external ideas of impact, and how operations can be rearranged to better achieve them (Molecke and Pinkse 2020).

Our findings and inferences tell us that the arguments social entrepreneurs construct for themselves to think about, and experience impact, are different from the ones they construct to work towards establishing dialogues around it. Since the main objective of argumentation is to gain the informed support of others Rieke et al. (2008), we argue that dual argumentation allows social entrepreneurs to pursue adherence at two levels. Internally, social entrepreneurs reinforce their commitment to their purpose, and, externally, they can engage stakeholders so that they can understand the impact of the social venture and grant legitimacy to the social venture’s efforts. In the former, impact claims and purpose can be aligned, and, in the latter, impact claims and expectations can be aligned.

Yet, tensions can be triggered when these arguments are articulated together. Here, we theorize that the dual argument construction can reinforce the willingness and ability of social entrepreneurs to engage with SIA facing uncertain data and the willingness and ability of stakeholders to engage with SIA facing uncertain values. If social entrepreneurs believe that factual data does not provide valid grounds to understand impact, the engagement with SIA is reinforced and adherence can be pursued based on value judgements. On the other hand, if stakeholders believe that purpose alone does not provide valid grounds to show impact, the engagement with SIA is reinforced and adherence can be pursued based on factual claims. In Table 6, we summarize these propositions in laid-back language to make them accessible for a wider audience.

Table 6. Dual construction of impact arguments in social impact assessment.

	Arguments for worth	Arguments for legitimacy
What are social entrepreneurs use it for?	To think about and experience impact	To establish dialogues about impact
Where is it coming from?	Attempts to become accountable to themselves	Attempts to become accountable to others
When are mostly needed?	When entrepreneurs believe that factual data does not provide valid grounds to understand impact	When stakeholders believe that purpose alone does not provide valid grounds to show impact
Why are they doing it?	Because these arguments allow them to increase sense of worth, aligning impact and purpose.	Because these arguments allow them to increase sense of legitimacy, aligning impact and expectations.
What allows them to do?	Gain internal adherence (entrepreneur) and strengthen alignment between values and impact.	Gain external adherence (stakeholders) and strengthen alignment between expectations and impact.
What is the result?	Reinforces the social entrepreneurs’ engagement with SIA, facing data uncertainty	Reinforces external stakeholders’ engagement with SIA, facing value uncertainty

Contributions

Our study makes two contributions to theory. First, we advance our understanding of social impact assessment in social entrepreneurship, across three areas: tensions, accountability, and performance.

Tensions in social impact assessment

Previous research has explored tensions in social impact assessment, focusing mostly on how tensions are navigated and reconciled (e.g. Molecke and Pinkse 2017, 2020; Nicholls 2018), yet they do not go far enough to explain what underlies those tensions. Existing studies highlight that social entrepreneurs face tensions whilst engaging with SIA, mostly because it requires collaboration across stakeholder groups (Ebrahim and Rangan 2014), and it is used to bridge financial with social goals (Battilana et al. 2015). As a result, social entrepreneurs ‘make do’ in the process of formalizing SIA practices by delegitimizing formal methodologies and reconstructing SIA using material and ideational bricolage (Molecke and Pinkse 2017). Our examination goes one level down to explore perceptions and experiences as entrepreneurs construct impact arguments, which antecede dialogue and negotiation. We uncover what actually is in conflict and provide a structured understanding of how arguments behind SIA can be shaped by differing reference points for impact and for what purpose. Through it, we can offer an alternative explanation for why tensions emerge and persist as social entrepreneurs engage with SIA, which in turn points towards the need to further develop ideas of the usefulness and validity of SIA plasticity rather than structure them out (Molecke and Pinkse 2017, 2020).

Social impact assessment and accountability

In taking a step back to reveal tensions in impact argumentation, we also reveal what being accountable as a social entrepreneur means and how that is supported by distinct argument construction practices. The dual construction of impact arguments leads to dual accountability, to themselves and others, whereby social entrepreneurs can align impact claims to their purpose and broader institutional mandates. Our findings suggest that SIA, at least in social entrepreneurship, should be conceptualized and studied both as a technical practice (legitimacy) and as a mechanism that shapes important work and life experiences (worth). This creates a challenge worth noting. The entrepreneurs spoke of the relevant referential experiences for understanding social impact as being based on their identified passions and previous exposure to similar circumstances and pains as those presented by beneficiaries. In some ways, the arguments for legitimacy thwarted this knowledge, by creating a sense of tension and confusion about how best to make claims about ongoing social efforts. It should not be assumed that such silencing is neutral, but instead actively investigated by scholars as a potential direct influencer of passion and enacted purpose (Cardon et al. 2009).

Social impact assessment and performance

The dual impact argumentation also reveals how social entrepreneurs can use impact claims to show different levels of performance. One form of argumentation enables social entrepreneurs to count and predict the trajectory of organizational phenomenon (i.e. people, resources, and processes) based on present ideas of success (Micheli and Mari 2014), whereas the other one is about imagining and co-constructing a new version of the world where peoples’ various conditions are improved. The entrepreneurs in our study referred to, drew from, and expanded upon their previous life experiences to construct arguments about social impact outcomes. For the most part, these experiential (rather than analytical) abilities to detect and describe social performance outcomes are absent from our theories in social entrepreneurship, even if they have already been alluded to in wider management scholarship (e.g. de Rond, Holeman, and Howard-Grenville 2019; Wally and Baum 1994). In this sense, our findings and theorization offer a better understanding of how performance measurement works in emergent social enterprises, the many conflicts that might emerge in its implementation (Cooper, Graham, and Himick 2016), and how performance measurement can play an expressive role in social entrepreneurship – beyond capturing and communicating financial and non-financial information.

Second, we extend Nicholls’ (2018) general theory by explaining what precedes the discursive space where the assessment of social impact reconciles facticity and validity to establish materiality. Much like the work of Power and Laughlin (1996), our findings shed light on several points of non-reconciliation, at which Nicholls’s facticity and validity appear to exist in sharp contrast. However, the dual construction of impact arguments allows us to go beyond the depiction of facticity and validity as the positivist and interpretive sides of social impact assessment. Impact arguments constructed through enfolding and projecting go beyond values and interpretative accounts, and the impact arguments constructed through objectifying and compressing go beyond social facts and positivist accounts. We find that truth, rightness, and sincerity (Nicholls 2018) of impact are not only established by numbers and indicators in this context, but through intricate and intimate reflections on the entrepreneurs’ life stories, present experiences, beneficiary experiences, and imagination. In their complexity, tensions configure distinct points of reference and argumentation practices. They explain the early stages of SIA and how social entrepreneurs can distinctively realize further constructions of meaning through SIA. In this sense, we elaborate further on the distinctive aspects of social impact assessment in comparison to mainstream accounting and performance assessment practice and highlight the importance of developing practices, which support multiple forms of argumentation in this context.

Finally, our study has implications for practice. Several studies have developed normative frameworks aimed at recognizing and measuring the impact of social enterprises to identify and measure the strategic drivers of social change (Stone, Carttar, and Colby 2004). Not surprisingly, many of these studies have catalysed industry-wide excitement to advance social impact assessment techniques and practices.² As a result, a host of tools and frameworks are now available to companies, governments, and social enterprises seeking to monitor and communicate evidence of their social impact. These are also promoted and used by social entrepreneurs, without any further reflection on the tensions they might experience as they engage with SIA, and we now know that requiring social entrepreneurs to use SIA may complicate their decision-making capacities. Ultimately, one argumentation device, in which claims are embodied, nuanced, and radical, must be let go of in favour of claims that are disembodied, narrow, and incremental to meet the legitimacy needs and resource demands of funders and policymakers (Muñoz, Kimmitt, and Dimov 2020). The tensions and argument construction practices tend to remain unknown to the founders and their social enterprises, triggering frustrations in the examination and formalization of SIA. Further, it is unclear what precisely is lost by allowing one dominate the other.

Future research

We envision three avenues for future research. First, although our explanation is grounded in tensions, we can only speculate whether these tensions prevail as the two argumentation forms become formalized, disappear, or are masked as social entrepreneurs elaborate new communicative frameworks. We do not know how tensions would evolve along with other factors, such as the age of the organization or the background of the entrepreneur. New research is needed to connect the tensions and argumentation practices we discovered with frameworks that, it is assumed, resolve tensions in SIA. In our follow-up observations, we noticed that often social entrepreneurs lean towards objectification while having to silence the enfoldment as an argumentation device, as funders and donors do not recognize and legitimize such inward-looking experiential accounts (Hehenberger, Mair, and Metz 2019).

Second, future research can also explore what is lost or gained when SIA is formalized, and what happens as one form of argumentation dominates the other. At the core of the tension, we also observed a state of discomfort. When the entrepreneurs were asked to bring the two arguments together, they felt embarrassed for not being able to do so. We realized that it is not necessarily that they do not have the technical skills to bring them together, but rather that the two arguments trigger cognitive dissonance and emotional tensions. This insight reveals an interesting effect of the tensions observed, which has not been studied in social entrepreneurship before. Future research can thus explore emotions and affectivity in SIA real time situations.

Finally, the dual impact argumentation opens a myriad of opportunities for reimagining what are ‘effective’ performance accountability mechanisms in social entrepreneurship – away from technocentric and positivistic ideals (Taylor 2004), towards more practice-based and embodied forms. A move in this direction requires a reconsideration of which skills and competencies are relevant for measurement when social impact is the focal object (Micheli and Mari 2014). If indeed the experiential aspects of knowledge (see e.g. Wally and Baum 1994) are a large contributor to the ability to detect and act upon social impact in a way that brings a comprehensive understanding to the process and outcomes, then there is a great amount of work that must be done to understand the implications for professional development, social impact assessment education, and ongoing training in social enterprises (Hwang and Powell 2009; Renko 2013).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the Corporación de Fomento de la Producción [15PES-44402].

Notes

1. For example: Social Return over Investment, Outcomes Star, and B Impact Assessment.

2. For example, the Robin Hood Foundation and the Robert Enterprise Development Fund (REDF) were early participants in the cost-benefit ratio methodologies for social programs, to understand and communicate impact (Emerson 2003). New Philanthropy Capital’s Inspiring Impact in the United Kingdom and Social Value International have pushed forward knowledge and praxis on SIA at a practitioner level.