Multiple paths for being recognized as a high impact firm in the banking sector

Abstract There is a call for firms in general and banks in particular to contribute to the generation of positive impacts (social, economic and environmental), paired with an increasing relevance for banks to communicate to their stakeholders the positive contributions they generate and their adherence to Corporate Social Responsibility (CSR) principles, for example, through recognized certifications of their positive impact generation. Subsequently, in order to contribute to knowledge of banks’ choices for generating a positive impact, the research question driving this research is: “What are the dimensions of impact (community, governance, workers, environment and customers) and/or their combinations and the degree of focus on each one which are necessary or sufficient for banks to achieve (or not) recognition of their high impact generation?” To answer this question, a sample of bank entities (n = 38) in the B Corp database is analyzed with fsQCA. Results show six combinations of focus on different impact dimensions that are sufficient for banks to achieve a high impact score, with the focus on the corporate governance dimension being the most prevalent among them. In addition, four combinations of scarce focus on different impact dimensions were sufficient for not achieving a high impact score, with the customers dimension being the most prevalent among them.


Introduction
The problems driving this research are twofold: a) the increasing importance given to the non-financial performance of organizations as a consequence of stakeholderdriven pressures for a transition to more sustainable and socially responsible business practices (Bengo et al., 2016); b) the necessity to contribute to the business case of non-financial impact created by the private sector (Silva et al., 2019) in order to foster such as "social performance measurement", "impact accounting" or "sustainability reporting". Another commonality is the use of the term social impact, despite in many cases the conceived impact also including the environmental dimension. Indeed, impact assessment has been considered as the process by which social and/or environmental tangible benefits generated by a company are shown (Emerson, 2003 cited by Grieco et al., 2015). This definition points to the relation between impact assessment and the concept of CSR, defined by Lu et al. (2020Lu et al. ( , p. 1639 as "a company's voluntary decision and determination regarding the changing needs of society and business interest groups to respect social, ethical (moral), legal, economic, environmental aspects in its activities, aiming to contribute to the concept of sustainable development." Considering this, in this study impact is an inclusive term understood as social, economic and environmental impacts. That is, impact is aligned with the increasing relevance of CSR for banks (Buallay, 2019) as an response to a general loss of confidence in the sector after the last financial crisis (Herzig & Moon, 2013) and the increasing importance for financial institutions of considering social and environmental issues in their decisions (Gambetta et al., 2017) in order to be transparent and responsible.
This in turn relates to stakeholder theory (Freeman, 1984) which states that organizations can obtain competitive advantages if they develop complex relationships with multiple stakeholders, with the sustainability of such relationships, especially in the banking sector, relying on the trust between parties, which in turn is dependent on firms' capability to satisfy different stakeholders' needs. That can be further understood by considering two relevant perspectives within stakeholder theory: instrumental and stakeholder-agency. For the first, the satisfaction of stakeholder needs is deemed as instrumental for the financial performance of firms, while for the second, the aim to avoid opportunistic behaviors in stakeholder-manager relationships is facilitated by CSR, which acts as a mechanism of monitorization and enforcement (Kim & Oh, 2019). Additionally, the different combinations of focus on CSR dimensions has been deemed as an indicator of how firms evaluate the power, legitimacy and urgency of stakeholders' expectations (Jamali, 2008 cited by Gangone & G anescu, 2014). Therefore, interest in the different ways in which companies pursue impact generation lies not only in the fact that the perceived ethics of banks are relevant for their reputation, but also in the difficulties in terms of coordination and demand of resources that arise in the process of aligning ethical practices with daily workflow and heterogeneous stakeholder demands (Azevedo & Ferreira, 2019).
Finally, the above is related to the recent phenomenon of the emergence of firms that obtain B Corp certification and are known as CBCs. These are companies that accept voluntarily third-party environmental and social audits performed by the nonprofit institution B Lab (Moroz et al., 2018) which appeared in 2006 to audit and certify the overall impact of companies using a wide range of indicators covering firms' impacts on communities, environment, employees, customers and governance. Although the project is still at a very early stage, it can be very beneficial for sustainable development (Cao et al., 2017) with more than 3.200 companies across 71 countries and 150 industries already certified 1 . Those firms may certify to provide socioeconomic agents with otherwise hidden information about their positive impacts in society and environment, as well as a mean to mitigate regulatory risk, signal their quality to consumers and/or to improve efficiency (Lytton, 2014). However, considering that there are more than 500 private sector national and transnational non-governmental organizations involved in globally certifying for-profit and non-profit ventures by conducting voluntary, third party social and environmental audits of their activities and impacts (Moroz et al., 2018), the use of some other approaches as assurance on sustainability reporting to better understand how companies build their corporate reputation (Simnett et al., 2009) and generate impacts should not be disregarded.

The importance of impact generation and the role of banks in it
The OECD. (2018) highlight the importance at national, regional and business levels of impact generation and its measurement in order to monitor activities, agreements, investments and communications which contribute to sustainable development.
From firms' perspective, a further reason for being concerned about impact is to avoid people perceiving a gap between their financial and non-financial interests, especially if they perceive that companies mainly or solely focus on value generation for their shareholders. That concern is paired with the increasing pressure faced by banks to report their commitment to governance, environmental and social aspects, a trend that is partially motivated by the increasing interconnectedness among organizations across the world and the rise of multinational companies from developing countries, with both factors leading to the higher relevance of banks sustainability reporting for multiple stakeholders (Buallay, 2019). In fact, stakeholders' information needs are deemed as a main driver of impact assessment (Grieco et al., 2015;Schiff et al., 2016).
Subsequently, in order to satisfy multiple and differing stakeholder needs, banks are evaluating sustainability from a multidimensional perspective (Raut et al., 2017), especially those aiming to contribute to sustainable development, for which social, environmental and business reasons are categories of corporate motives to advance social justice and cohesion (Zimmermann, 2019). In fact, since banks' activity is dependent on society's resources, they face a higher pressure to report their impacts on communities and other social areas (Wu & Shen, 2013) as a consequence of their influence in the economy, environment and society through their role as financial intermediaries, which entails the evaluation of companies and their assets, the provision of monetary resources and the management of financial risks and payment systems (Weber, 2012). Consequently, there has been an increasing engagement within the sector with sustainable activities linked to CSR strategies in order to produce value for multiple stakeholders (McDonald & Lai, 2011) and to improve reputation (Laguir et al., 2018), with banks generally holding good positions on international CSR investment ranking indices (Roy et al., 2015).

Constituents of impact: B Corp pillars and their relevance
B Corp establish five pillars to assess the impact of a company (governance, customers, community, environment and employees) defined as shown in Table 1. As regards the impact on communities, banks can promote integration and empowerment of people tackling the exclusion of disadvantaged groups (Teasdale, 2010). In that way, banks which mutually pursue financial and non-financial goals can obtain local credibility to encourage community support for their social mission (Shaw & Carter, 2007), for example through microcredits .
As regards the commitment with customers, it is positively correlated with organizational legitimacy (Chen et al., 2019) as a consequence of satisfying their needs and being aware of their interests, not just in terms of products and services but also by accomplishing activities directed to the generation of positive social and environmental impacts. In fact, when customers perceive the legitimacy of a company, they positively accept its services, products and business philosophy (O'Neil & Ucbasaran, 2016), especially when organizational legitimacy symbolizes a meaningful social identity that satisfies customers' social needs, potentially increasing the attractiveness of the firm for them (Bagger & Li, 2014). For the banking industry, there is a positive correlation between firms' CSR activities, financial performance and customer demands, especially when they are looking for long-term relationships with firms (Fatma et al., 2015). That is in line with recent works finding a positive relationship between the presence of CSR and consumers' trust, commitment, satisfaction and loyalty (Servera-Frances & Piqueras-Tom as, 2019).
As regards the employees pillar, banks can increase their motivation and retention, for example, thanks to an increase in their job satisfaction resulting from perceiving the importance given to them within the CSR strategy of the company (Jie & Hassan, 2018). Even more, contributing to society and/or the environment may result in an Measure the relation of the firm with the suppliers, implications in the community and other aspects related to diversity in the company. This also assesses whether the final aims of the product or services are to solve some social issue such as access to education, health, or basic services. Customer Assess the impact of a firm on its own customers, considering whether its products are designed to satisfy underserved populations and if they promote public wealth. This area also considers if the product or service aims to solve social or environmental problems (e.g. health improvement, environment protection, arts promotion). Environment Measure the environmental impact of a company due to the use of its own resources (facilities, inputs, energy). This considers all the elements related to their supply chain, including transportation and distribution channels. It also evaluates if the products or services are specifically laid out to deal with environmental problems (e.g. production of clean energy, waste reduction, conservation of wildlife, pollution reduction or better environmental education).

Workers
Evaluate the relationship between the company and its workers. This considers different aspects of the treatment of the firm's employees such as compensation, social benefits, training and opportunities to be owners of the company. This area of analysis gives great importance to the firm working environment, measuring elements such as communication between workers and managers, safety and health policies for employees or job flexibility. Governance Assess a company mission, transparency and accountability. This evaluates if there is a social or environmental content in the company mission and if the different stakeholders are engaged to accomplish it. It is relevant in this section to determine if the company shares financial information with its employees, if customers find it easy to give feedback, and if there is enough diversity in the governing bodies of the firm.
increase of perceived identification with the organization, which, coupled with an increasing focus on employee well-being can lead to an increase in their engagement with the firm (Zulfiqar et al., 2019). Finally, according to stakeholder theory, employees are especially relevant for realizing such contribution since at the end of the CSR implementation process, they are the ones who translate CSR initiatives into organizational outcomes (Boadi et al., 2020). With respect to governance, its relationship with organizational performance has attracted scholarly interest (e.g., Azevedo & Ferreira, 2019;Gangi et al., 2019) and has resulted in significant studies in multiple countries (Liang et al., 2013). In particular, governance has been deemed as a determinant of firm strategies and their relationships with stakeholders (Monks & Minow, 1995), especially in the banking sector (Turlea et al., 2010). That is in line with the conflict resolution approach which explains the positive effect of corporate governance on CSR, conceiving that effective governance facilitates the creation of a more systemic view of the dimensions of CSR and the interactions between them (Gangi et al., 2019). Subsequently, this reasoning provides support for the notion that corporate governance constitutes a relevant aspect of CSR (Jamali, 2008).
Finally, the assessment of impacts on the environment has been deemed by the United Nations in the Agenda 2030 as a relevant goal for sustainable business development. Regarding its relevance for banks, in a sample of 68 French banks studied over the period 2008-2011, Laguir et al. (2018) found a positive and significant bidirectional relationship between their corporate financial performance and their corporate environmental performance. Additionally, Buallay (2019) found that the environmental disclosure of European bank firms positively influenced their operational (ROA) and market (Tobin's Q) performance.

The relevance of impact assessment and the B Impact Assessment model
The traditional Weberian vision states that the main objective of assessment is informing and improving decision-making in order to answer the heterogeneous information needs of multiple stakeholders. For example, whereas private investors seek benefits and security, governments may want to assess organizations' contributions to society, whereas decision makers in companies may be interested in using assessment as a way of monitoring impact, and improving organizational performance and interactions with stakeholders (Bengo et al., 2016) through the identification of ways in which they can demonstrate their positive contributions, which in turn can improve organizational reputation. In that sense, the need of relevant and reliable information to assess past performance as well as to plan future actions make impact assessment a valuable tool for decision-making and for the adoption of systems which, addressing different forms of value (e.g., economic, social and environmental), help to cope with the need for legitimization to stakeholders (Luke et al., 2013). Following this perspective, impact reporting constitutes a resource at the organizational level, insofar as it improves credibility, therefore contributing to business through facilitating access to resources (Dowling & Pfeffer, 1975).
Considering the above, one might ask what is the method used for assessing the impacts of CBCs. The answer is the B Impact Assessment (BIA) model, which produces a final impact score resulting from the sum of the impacts on the five B Corp pillars using quantitative and qualitative data obtained through an audit process. During the certification process, the impact of firms is analyzed considering if the company was intentionally designed to generate positive outcomes for specific stakeholders. First, companies must fulfil and submit online the impact assessment form (BIA). Second, B Lab evaluates the aforementioned five impact areas, verifying if the company has obtained 80 points (out of a maximum of 200), which is the minimum score to be certified. Finally, to grant the certification, the company is required to make the necessary changes in its practices and legal structure in order to equilibrate financial goals with social and environmental ones. Additionally, firms that desire to keep the certification need to pass through the same process every three years, and every year B Lab audits 10% of the CBCs randomly.
Previous literature on firms' interest on the B Corp certification show a mainly external focus and a focus on how certification help companies to signal their commitment with different stakeholders (Cao et al., 2017;Dineen & Allen, 2016). Specifically, addressing multiple dimensions and considering both qualitative and quantitative indicators makes B Corp data a useful source for benchmarking and multisectoral and multicultural studies (Rawhouser et al., 2019), especially because the certification includes an evaluation at the social, financial and environmental level supported by the review of an advisory council whose nine members are independent (Chen & Kelly, 2015). Its independence is related to two facts: a) the B Assessment is a free, public service provided by B Lab, a non-profit organization; b) B Assessment standards are created and reviewed by the Standards Advisory Council which is a group of independent business and academic experts.

Qualitative comparative analysis
In order to analyze banks' strategies for impact generation and identify the different pathways that lead them to obtain a certification that recognizes their commitment, fsQCA has been chosen as the methodological tool for the empirical analysis. The fsQCA method tries to benefit from aspects of both qualitative and comparative approaches (Ragin, 1987) in three ways: a) instead of assessing correlations between variables, the goal is to identify causal conditions (conditions), as well as their possible combinations (configurations), leading to a specific result (outcome); b) instead of focusing on one or only a few case studies, fsQCA analyzes samples of low or middle size, between 15 or 20 cases (Fiss, 2011), but is equally well suited for analyzing bigger ones (Vis, 2012); c) fsQCA allows the existence of multiple conjunctional causality, which implies that it denies permanent causality and considers the possibility that different combinations of conditions can explain the same outcome (equifinality).
For this study, the above implies the assumption that banks can generate (or not) a high impact following different pathways. Considering the above, 38 bank entities (cases) are classified in terms of their generation (presence) or not (absence) of high impacts (outcome) in order to identify which are the impact dimensions, the degree of focus on each one and the combinations of them that have characterized banks achieving (or not) a high impact score.
Additionally, fsQCA allows identification of conditions and configurations that are necessary or sufficient for achieving an outcome (Ragin, 2008). A condition or configuration is deemed necessary when it needs to be present in order to have the possibility to achieve the outcome, whereas a condition or configuration is considered sufficient when its presence is enough for achieving the outcome. Therefore, the method allows identification of those strategies that are sufficient (or not) for banks in the sample to obtain a high impact score. Ragin (2008Ragin ( , 2017 and Schneider and Wagemann (2012) describe in a detailed way the process to follow in order to conduct an empirical analysis using fsQCA. First, the outcome and the causal conditions that could explain it must be identified in order to define what to consider as the relevant cases that will constitute the analyzed sample. For the purpose of this study, the two outcomes analyzed are represented by the high and low impact scores of CBCs banks with their certification still in vigor (n ¼ 38) according to the last evaluation of the company 2 .

Sources and calibration
For each condition, a high or low categorization according to the presence or absence of the different conditions (impact dimensions) and the achievement of the outcome (high or low impact score) is established. In order to be analyzed, conditions and outcome must be calibrated, which require definition of the rule for assigning the membership scores of the cases (bank entities) to each of the sets corresponding to the conditions and the outcome under study (Ragin, 2007(Ragin, , 2008. For this study, calibration has been conducted following the direct method proposed by Ragin (2008) in order to define three threshold points that correspond to: a) being "fully in" the set, b) being "fully out" of the set and c) being "neither fully out of nor in the set" (crossover point or point of maximum ambiguity). Ideally, calibration should be mainly based on external knowledge (Greckhamer et al., 2018;Ragin, 2017), nonetheless, in cases where data come from individual perceptions assessed by tools like questionnaires, as in the present case, choosing a criterion like the median as the rule for defining the crossover point can be a valid solution (Russo & Confente, 2019). Additionally, such a criterion has been successfully applied in previous works in high impact journals (Q1 and Q2) covering the areas of business and management (e.g., Misangy & Acharya, 2014;Gonz alez-Cruz et al., 2018). Subsequently, the median is chosen as the point of maximum ambiguity following the suggestions of Russo and Conferente (2019) and the aforementioned works. More specifically, as done by Misangy and Acharya (2014) and Fiss (2011), we have considered that a company pertains to a set when it is included in the upper quartile of the scores in the condition or outcome that represents such a set, whereas banks not pertaining to a set are those whose scores in the conditions and/or the outcome are below the lower quartile, with the median constitutes the point of maximum ambiguity. Table 2 shows the values for each condition and outcome.

Analysis of necessity
The last step of the fsQCA consists of the evaluation of the models and the analysis of their solutions. The evaluation of model validity depends on two indicators: consistency and coverage, both ranging between 0 and 1. Consistency refers to the degree that a relationship of necessity or sufficiency between a causal condition (or combination of them) and an outcome within a given data set is accomplished (Legewie, 2013). Instead, coverage indicates the degree to which the solution explains scores given to each case in the sample. The result of those analyses for models 1 (achievement of high impact generation) and 2 (failure to achieve high impact generation) is the identification of the necessary conditions (if any) for achieving (or not) a high impact. A condition will be considered necessary only if its consistency is equal to or higher than 0.9 (Ragin, 2008;Schneider & Wagemann, 2012) As shown in Table 3, none of the considered potential causal conditions for achieving (or not) a high impact generation are necessary. Such a result suggests that none of the considered dimensions is strictly required for obtaining (or not) a high impact score.
This result is in line with recent works on CSR which highlight how its contribution to the competitive advantage of firms stems from its role in balancing and managing multiple stakeholder needs, which in turn leads to an increase of firm value when a good value of CSR is achieved (Kim & Oh, 2019). That is coherent with stakeholder theory (Freeman, 1984) which posits that in order to obtain a higher value, firms need to satisfy the need of multiple stakeholders, which in turn is also aligned with the multidimensionality implicit in the impact and CSR concepts (Lu et al., 2020) with the consequence of that being that in order to obtain a competitive advantage from impact generation and CSR commitment, firms need to satisfy needs emerging in multiple areas. Additionally, Gangone and G anescu (2014) found that providing a higher degree of satisfaction to one stakeholder category contributes to the satisfaction of the rest, ultimately leading to an improvement of CSR initiatives. This result led the authors to state that "achieving a superior CSR performance is possible while meeting the expectations of all categories of stakeholders" (p.555).
Additionally, in their study of 33 Lithuanian companies, Lu et al. (2020) found that environmental and economic social responsibility positively influence the financial capacity of firms, whereas the social, shareholder and voluntariness dimensions positively influence productivity and work efficiency. Such results reveal the aforementioned perspective, that is: none of the specific CSR is determinant of a firm's value and performance, but all of them have a particular influence on its different aspects, which in turn points to the necessity of taking into account multiple CSR dimensions in order to obtain the competitive advantage resulting from satisfying multiple and differing stakeholders' needs. In turn, such vision leads to the following question: what is the degree of focus on different impact dimensions which is sufficient (or not) for achieving a high level of impact generation? The answer to this is given by the following analysis of sufficiency.

Analysis of sufficiency
As supposed, pathways leading to a high impact score (Model 1) are combinations of impact generation in at least two of the B Corp pillars. To complement such a hypothesis, the analysis of the configurations that are sufficient for failing in the achievement of a high impact score (Model 2) have also been identified. Both models are shown in Table 4. For both models, the consistency of the solution must be verified and only if its value is above 0.75 (Ragin, 2008) can the model be considered as valid. In that case, both models overcome the minimal value: Model 1 with a solution consistency of 0.86 and Model 2 with 0.81, both showing the intermediate solution which is one of the three possible solutions that fsQCA offers, the other two being the parsimonious and complex ones (a detailed discussion of that choice and the results for the parsimonious and complex solutions can be found in Appendix). For Model 1, it was assumed that the presence of conditions (high focus on B Corp pillars) should be the one that leads to a high impact score. Conversely, for Model 2, the assumption is that the absence of conditions (low focus on B Corp pillars) could be the potential cause explaining the failure of banks in achieving a high impact score.
Turning now to the solutions, in Model 1 the mainly present conditions in the configurations deemed as sufficient for achieving a high impact are governance (present in five of the six configurations) followed by customers and community (both present in four of the six configurations), with every sufficient configuration including a high degree of focus on at least two of the those pillars. Specifically, the identified pathways are as follows, with the number of banks exhibiting each configuration indicated in parentheses: Pathway 1a (n ¼ 7). The generation of impact in the pillars of clients, community and governance is sufficient for obtaining a high score. Pathway 1b (n ¼ 8). High impacts in the community, governance and environment areas are sufficient for achieving a high score. Pathway 1c (n ¼ 4). It is sufficient to focus on impact generation in clients, governance and environment to achieve the outcome. Pathway 1d (n ¼ 8). Similar to pathway 1b, the outcome is achieved focusing on communities and governance, nonetheless, instead of the environment, the customers are the third pillar which, in conjunction with the aforementioned two, constitutes the sufficient configuration. In conjunction with pathway 1b, this is the most common configuration. Pathway 1e (n ¼ 3). This differs from pathway 1c in that instead of focusing on workers, the generation of impact is most directed to the community. Pathway 1f (n ¼ 4). This is the only configuration that is constituted by four conditions with one of them being absent. In other words, this configuration suggests that banks who decide to generate a low impact in the workers dimension could still obtain a high score if they focus on its customers, the community and the environment. Additionally, another particularity of this pathway is that it is the only configuration that does not include the presence of high impact in the governance pillar.
In summary, the configurations mainly adopted by bank entities achieving a high impact score have in common the presence of two conditionsgovernance and communitywhile the rest of configurations, despite being a minority, include the presence of at least one of them. That is partially aligned with the results of Azevedo and Ferreira (2019) in which governance practices, ethical management of employees and employees' practices and behaviors were the factors with a higher degree of centrality and influence in ethical banking practices. Regarding governance, its relevance is confirmed by its presence in five of the six combinations. In contrast, despite the relevance of the employees' dimension, only two of the six combinations exhibited a high degree of focus on it. Additionally, Laguir et al. (2018) state that bank firms commit to the environmental dimension according to an instrumental logic, in the sense that when the short-term needs of their most relevant stakeholders (i.e. shareholders) are met, the firm may opt for increasing their efforts toward the environment. Conversely, when companies have a low degree of reputation among shareholders or a low financial performance, they will probably opt to focus less on the environment. This is exemplified in our results in that from the three configurations showing a high degree of focus on the environment, two of them also show a higher degree of focus on their customers, which in turn, are the main means of revenue for banks, therefore constituting an area of interest in order to cope with shareholder needs. That is aligned with the statement of Gangi et al. (2019) which considers that the main reasons for banks to pursue social and environmental issues are to: a) achieve their own strategic goals, b) respond to social and institutional pressures; and c) generate positive spillover effects. In that sense, future research could examine whether the differing degrees of focus on various pillars is correlated with financial indicators of banks. Finally and counterintuitively, despite the idea that employees are critical for CSR implementation (Boadi et al., 2020), the workers pillar is the one where a lower emphasis by bank entities has been shown, being present only in two of the six.
Before moving on, it should be noted that the main differences between Models 1 and 2 is that the configurations that are deemed sufficient to fail in the achievement of a high impact score are simpler than for achieving it. Specifically, the configurations shown in Model 1 require the presence of at least three conditions (except for pathway 1f), whereas in Model 2, despite existing less sufficient configurations, they only require the absence of at least two conditions (except for pathway 2d). As regards the pathways shown by banks not achieving a high impact score (Model 2), these are as follows: Pathway 2a (n ¼ 6). A low degree of impact on the clients and governance pillars is sufficient for not achieving a high score. Pathway 2b (n ¼ 10). Express that is sufficient to not generate a high impact in the pillars of community and workers to fail in the achievement of a high impact score. The higher presence of that combination is coherent with a recent work of Kok et al. (2019). In their case study of the struggles and interactions between departments of a Dutch bank trying to develop sustainability policies, they found that the mutual presence of two different logics (market and sustainability) lead to an increase of tensions and conflicts between departments which resulted in a mutual perception of incompatibility between logics that complicated the understanding between coworkers and therefore, the development and implementation of sustainability policies. In order to control for the above, a qualitative case by case analysis of the sample would be needed, which goes beyond the scope of this paper. Nonetheless, such results provide a possible explanation of the cases in which firms did not reach a high impact score. In particular, a middle degree of focus in some dimensions and a lower one in the rest could be a consequence of inner difficulties to coordinate and reach points of consensus regarding the way to pursue sustainability issues. Subsequently, without such a consensus, a higher degree of focus in any of the dimensions seems particularly difficult, which in turn is another aspect that points to the relevance of corporate governance as a framework of guidelines that can organize and simplify the way to approach other impact dimensions. Pathway 2c (n ¼ 9). Differs from the 2b only in that instead of showing an absence of high impact in the workers pillar, such an absence appears in the customers one. In that case, a low impact score could be the consequence of not obtaining enough revenue from customers, which in turn is consistent with the slack resource theory, which suggests that firms with a higher amount of resources beyond the minimum needed for their core activities would be facilitated in their aim to cope with multiple stakeholder needs (Kim & Oh, 2019). Pathway 2d (n ¼ 1). Differs from the rest in that instead of being constituted by two absent conditions, there are three: customers, environment and workers.
Finally, the main present condition across the pathways that leads to the failure in achieving a high impact score was the absence of focus in the generation of impact in the customers dimension. That is aligned with the findings of Romi et al. (2018), who analyzed the relationship between CBCs' commitment to social issues and their employee productivity and sales growth, finding that within the analyzed CBCs, those with a higher degree of focus on the customers dimension also exhibited a significantly higher level of employee productivity. Additionally, a lower degree of focus on customers could also lead to a lower degree of earning from them which in turn would complicate the allocation of resources across impact dimensions in order to exhibit a higher degree of focus on at least one of them. This idea is coherent with the slack resources theory (Nohria & Gulati, 1996), according to which, companies that benefit from resources beyond the minimum needed for their main productive process are advantaged in terms of developing their CSR (Kim & Oh, 2019). Nonetheless, the present study cannot confirm this reasoning which leads us to suggest that future research takes a look at possible relationships between CBCs' dimensions emphasized by companies and their financial performance.
In contrast, the less frequent condition in the configurations of Model 2 is the absence of focus on the governance pillar. This points again to the particular relevance of that dimension in the banking sector since, even in those companies with a lower impact score, it received at least a middle degree of attention. Nonetheless, it also suggests that a limited amount of effort directed to the governance pillar is not enough to fail in the achievement of high impact, whereas the focus on it is generally a constituent of pathways leading to achieving it. That is consistent with stakeholder theory in the sense that since there are multiple demands to satisfy, the failure in satisfying one of them should not necessarily lead to a low degree of satisfaction among the rest of the stakeholders, whereas in contrast, a higher degree of focus on corporate governance would facilitate satisfying the majority of stakeholders as a consequence of its role as a mechanism that facilitates the coordination of efforts among the CSR dimensions.

Conclusions
Obtaining an economic benefit derived from business activity is not enough for many banks, and academics as well as practitioners are increasingly concerned with their generation of positive impact on multiple areas related to the traditional pillars of CSR (Fatma et al., 2014), partly due to the social role that the industry has been developing in terms of financial inclusion (Gambetta et al., 2017) and community and sustainable development (Aras et al., 2018;. In this context, B Corp is considered as an auditor recognized by organizations interested in assessing the impact generated by their activity from a multidimensional perspective (Moroz et al., 2018), therefore constituting a relevant reference for exploring the different dimensions of impact on which banks have focused in order to being recognized as firms generating a high amount of positive impact. Such exploration complements previous works on CSR of bank firms (e.g., Gangone & G anescu, 2014;Lu et al., 2020) in the fact that it facilitates a better idea of the different ways in which banks afford CSR dimensions while distinguishing between those companies obtaining (or not) a high level of CSR commitment. In particular, the results of the present study led to the following conclusions.
First, the heterogeneity of impact generation manifests in the six different configurations leading to a high impact score, and its complexity resulting from being composed by a high degree of commitment with at least three dimensions of impact points to the necessity to approach the construct from a multidimensional perspective. This is coherent with the main tenet of stakeholder theory (Freeman, 1984), that companies will be able to realize competitive advantages when they successfully attend to the demands of multiple stakeholders. In that sense, banks could easily manage their stakeholders' demands through CSR commitment by directly asking them which are their concerns. In turn, it would facilitate their decision making regarding resource allocation, especially in developed economies where CSR disclosure is mainly motivated by the need to answer the needs of specific stakeholders like the state, investors or activists (Ali et al., 2017). This recommendation is in line with the thoughts of Stieb (2009), who suggested that committing with multiple stakeholders can lead to more equality and justice in terms of redistribution of benefits and decision-making power.
A second conclusion is that, neither the focus or not on a particular dimension of impact resulted necessary for achieving (or not) a high impact score. However, governance emerged as the first impact dimension across banks achieving it. This could be partially explained by at least two factors: first, the high demands of transparency, accountability and ethics to which banks are exposed (Herzig & Moon, 2013) and secondly, the relevance of corporate governance as a process-based management system which influences CSR (Shahin & Zairi, 2007) and determines relationships with stakeholders (Monks & Minow, 1995).
Counterintuitively, a low commitment with governance was only present in one of the four combinations that were sufficient for not obtaining a high impact score. A reason for that could be that due to the aforementioned relevance of governance inherent to the bank sector, the majority of companies need to show at least a middle degree of focus on it. According to this, a recommendation for banks concerned with CSR and sustainability would be to focus on their corporate governance in order to cope with institutional pressures for transparency whereas also benefitting from clearly defining a set of rules and business behavior guidelines that facilitate their decision-making regarding resource allocation among various CSR dimensions.
A third conclusion is concerned with the heterogeneity of focus on different pillars, with the presence of the community and clients in one in four of the six routes being added to the aforementioned importance of governance. The high presence of the impact on communities is in line with the role attributed to financial institutions as promoters of financial inclusion (Gambetta et al., 2017), which in turn benefit the economic growth of communities  while also improving the relationship with their customers, which suggests a potentially synergistic effect between both dimensions, and could explain their mutual presence in two of the six pathways that led to a high impact score.
Coherently, after customers, workers and community were the most prevalent dimensions that received scarce attention among the companies that did not obtained a high score. As regards employees, they are necessary for generating any kind of impact, but a scarce focus on them may result in lower job performance in comparison with organizations taking more care of them, an issue that can negatively influence organizational performance. Additionally, a scarce attention on communities could also reduce the potential benefits in terms of higher organizational identification of workers with the company , which in turn could also partially explain why it is paired with the workers dimension in one pathway. On the other side, the fact of not helping communities could result in less organizational attractiveness for customers when compared with other banks so that helping them may be perceived as more concerned with CSR and therefore they are better positioned for benefitting from a higher customer trust and identification with the firm (Homburg et al., 2013). This points to the importance for banks to take into consideration the social dimension of CSR, an aspect that has been recently highlighted by Lu et al. (2020) which found a significant and positive relationship between this dimension and the corporate image, the capability to satisfy customer needs, the productivity and the work efficiency of 33 Lithuanian companies.
Despite the above, this study is not without limitations. First, the use of cross-sectional data imply that causal relationships cannot be assumed, even if the fsQCA is an adequate method for affording complex causality (Schneider & Eggert, 2014). Secondly, the absence of consensus across literature and practice regarding impact assessment implies that the study of other models for impact assessment could yield different results, which suggests that future research could analyze banks with various models. Thirdly, the relatively small size of the sample (n ¼ 38) should be considered when generalizing results, nonetheless, it has been partially alleviated using the fsQCA method, which is well-suited for analyzing such samples (Ragin, 2008).
In summary, the main contribution of this research has been to identify the dimensions of impact and the degree of focus on them afforded by the banks that succeeded and failed in obtaining a high impact score that certifies their concerns with corporate governance, communities, employees, workers and/or the environment. Such results could be nurtured with future research looking at if the considered impact dimensions are related to financial performance indicators and other constructs at the organizational (e.g., performance, innovation capability, market value) and/or individual (e.g., job satisfaction and performance) level. An example of that kind of research is the work of Grimes et al. (2018) which analyzed the role of business owners' gender on the differences in the adoption of the B Corp certification. Additionally, the study could be replicated in other industries or from a crosscultural perspective in order to identify other strategies of firms committing to CSR and sustainable development.
Finally, it is worth noticing that companies exhibiting a lower impact score should not be considered as not contributing to sustainable development. In fact, all of them are trying to make some efforts in order to commit with different impact dimensions, with that effort being facilitated by certifying as B Corps. In that sense, a final suggestion for banks as well as for policy makers is to promote the adoption of CSR reporting standards determined by external agents (i.e. B Impact Assessment Model) which could help firms to more easily coordinate their efforts toward social and environmental issues whereas enforcing a trend that gets more relevant every day and to which the present work intends to contribute: the commitment of banks in particular and firms in general beyond financial goals, not just to cope with stakeholder pressures, but mainly to give back to the society and the environment a part of what they take.

Notes
Thirdly, the intermediate solution provides an answer to the aforementioned problematics since it assumes that only some of the configurations exhibited by the banks in the sample will lead to the considered outcomes. Therefore, this solution gives a way for practitioners, lecturers and policy makers to find ways used by banks to obtain (or not) a high amount of positive impacts, whereas opening the possibility for finding new configurations was not found in the present study, without using the aforementioned assumptions made in the complex and parsimonious solutions.
Nonetheless, despite the above, in order to be rigorous and facilitate the discussion of the results of the present study, the following tables offer respectively complex and parsimonious solutions for models 1 and 2.
Finally, a relevant aspect to consider is that the results of the parsimonious and complex solutions do not contradict those of the intermediate one since the major conclusions stemming from it can also be found in the other two solutions.
First of all, in the three cases, regarding Model 1, a high degree of focus on governance, followed by customers and community are the mostly present conditions among those banks obtaining a high impact score.
Secondly, in the case of Model 2, again the two solutions point to the critical influence of exhibiting a lower degree of focus on just two dimensions, which led to a low impact score. Additionally, a point worth mentioning is that in the complex and parsimonious solutions the relevance of corporate governance, which was already highlighted in the main text, is further Note: "" denotes presence of the condition whereas " " refers to its absence. emphasized, especially for the complex solution where in two pathways a low degree of focus on corporate governance, despite a high degree of focus on the other two pillars, constitutes two combinations leading to not achieving a high impact score. That is even more prominent given the fact that in the parsimonious solution, from the six configurations leading to not achieving a high score, three of them exhibit a low degree of focus on the governance pillar. Additionally, the higher relevance of the social dimension of customers is also shown in those two solutions, with a low degree of focus on that dimension being the most present condition among those configurations. Thirdly, and again coherently with stakeholder theory and the multidimensionality of the CSR and impact concepts, the pathways leading (or not) to a high impact score consist of different degrees of focus on multiple pillars, therefore being coherent with the realized analysis for the intermediate solution.