1 Introduction

The concept of social and environmental responsibility of companies has evolved greatly since Bowen (1953) presented institutional and normative reasons for business management to consider it. During the evolution and conceptualization of CSR, committed optimists, critics with this normative or moral vision of the economy, skeptics and pessimists have coexisted. Porter and Kramer (2006, 2011) were probably the most optimistic, affirming that companies can create economic value by creating social value, while critics have viewed CSR as a public relations policy or greenwashing for companies. Reich (2008) concluded that interest in CSR responds to companies’ interest in preventing the government from intervening in social and environmental matters. He argued that most companies cannot achieve social goals without generating costs for their customers or shareholders due to strong market competition. Fleming and Jones (2012) viewed CSR as a type of complex manipulation conducted by companies to undermine the political rights of society. The social impact of the crisis on society has once again led to normative and ethical arguments in the debate on CSR aims.

This chapter presents a critical vision of CSR based on normative and ethical approaches (see Garriga and Mele 2004). This vision involves a model of a company aligned with society in a slightly different way of understanding capitalism. Companies are considered socioeconomic entities and direct their mission and business model to generating added value for all their stakeholders, not only for (or even prioritizing) shareholders. It assumes a stakeholder theory approach defended by authors like Bowen (1953) and Freeman (1984).

CSR rhetoric has been criticized, because it has often promoted ‘ornamental elements’ which are not part of the core business and unbalanced reporting focused on positive results instead of negative impacts. CSR is defined as the ‘responsibility of enterprises for their impacts on society’ and involves managing and taking responsibility for every negative externality of the company (European Commission 2011). CSR is meaningful when it is applied to core business and its related risks, which is much more complex than developing it in a ‘perimeter scope’ (social sponsorship, public relations, reputation, etc.).

Few analyses of companies are based exclusively on the core business of CSR, which is related to the demands of stakeholders affected by the way products and services are produced and delivered, and the social impact they have (Pedersen 2010; Visser 2010; Öberseder et al. 2011). They offer a single ‘picture’ of CSR with a mixture of often incomparable elements like working conditions, environmental impact, advertising complaints or community sponsorship, some of which are not directly related to core business. For example, corporate governance is a fundamental part of CSR, but it cannot be considered external CSR, because it refers to an internal group of stakeholders (shareholders). Environmental impact can be observed from both an internal dimension (infrastructures, energy consumption, etc.) and an external dimension (e.g. in the banking sector, the impact of project finance related to low-carbon industries). External CSR is specific for every sector or even subsector—distinctive from the core business—while internal CSR can be defined by models applicable to different sectors (responsible human resources management, corporate governance, community involvement, etc.). Organizations which rate the social performance of enterprises have been criticized, as their metrics to quantify CSR are often invalid and misleading to stakeholders (Chatterji et al. 2016). Notwithstanding, we consider that rough data from sustainability rating agencies is the most suitable tool for constructing an external CSR index , as it allows us to select indicators only related to the commercial dimension. Rating agencies have also been widely used in academic research, particularly EIRISFootnote 1 data (Wu and Shen 2013; Cuesta-González et al. 2006; Scholtens and Dam 2007a).

In their review of the literature, Griffin and Mahon (1997) stated that the industry that the company belongs to can influence the CSR measurement method chosen by the company and the configuration of the stakeholder impact. The banking industry has a significant impact on socioeconomic development and has experienced an intense development of CSR management and transparency . Notwithstanding, the main research topic regarding this sector has been sustainability reporting .

Much of the research focused on sustainability reporting identified ethics, product responsibility , human resources (Weber et al. 2014) and environmental information (Gallego 2006) as weak points of banks’ CSR reporting, whereas its strengths were community involvement (Weber et al. 2014) and regulated corporate governance (Douglas et al. 2004). Branco and Rodrigues (2008) also found a positive correlation between the commercial visibility of banks (measured as size) and the spread of CSR information.

Cuesta-González et al. (2006) affirmed that the most important functions of a financial intermediary should be analyzed to establish the framework of its environmental and social responsibilities. They distinguished between the internal and external dimensions of CSR (the external dimension exclusively related to the core business) and argued that external CSR has been fundamentally developed by ethical banks. To boost external CSR, sustainability must be considered in the analysis of operative risks, but banks generally consider it a reputational risk . Schmid-Schönbein et al. (2002) identified three areas for development in terms of external CSR: sustainability criteria applied to project finance , sustainability applied to asset management, and retail banking activities and consumer relations.

In this study, we have focused on the external dimension of CSR, because we consider it to be the most relevant in terms of sustainability in the banking business. External CSR refers to the core business, the basic function of banking in the economy and the relational side of this sector, that is, their customers and the demands of society. The mixed assessment of external and internal CSR factorsFootnote 2 distorts the accurate evaluation of CSR advances in the banking core business. The construction of a self-built index exclusively oriented to this external dimension of CSR is proposed to focus on performance indicators and avoid those simply related to policies or ‘greenwashing’ purposes. We measure the external CSR performance of European banks and define to what extent size and the banking model or business specialization influence this performance.

2 CSR in the Commercial Banking Business

Commercial banks can be classified according to their mission and the diversification of their business. There are profit-oriented institutions (traditionally called ‘banks’ in Europe) and social-oriented institutions, which are often local or regional (‘cooperative’, ‘savings’ and ‘postal banks’ in Europe). We also differentiate between traditional commercial banks, focused on financial intermediation (savings and loans), and universal banks, which offer a wide spectrum of financial services beyond financial intermediation such as asset management, corporate banking, and so on. Universal banking is closely related to size rising, due to policies promoting the merging of entities and the centralization of decisions to minimize costs and maximize income.

Economic development relies on efficient institutions that reduce uncertainty related to markets and human transactions, which means reducing transactional and information costs. However, information is rarely perfect. Stiglitz and Weiss (1981) explained that financial markets are different due to serious principal agent problems, which include adverse selection—asymmetric information—and moral hazard, assumption of risks when another agent is supporting the consequences of those risks. They showed that information problems can lead to credit rationing and exclusion from financial markets even in equilibrium.

In the case of banks, generating complete information is easier when they manage local investments, and monitoring costs become lower than for interregional operations. Thus, local or regional banks should identify, control and finance local projects more efficiently. As moneylenders, banks specialized in retail segments are fundamental for borrowers whose projects remain unfinanced by financial markets (domestic economies, SMEs, microenterprises, etc.), because they can generate information regarding profitability and risk at a reasonable cost (Strahan and Weston 1998). Therefore, good performance in consumer care should be characteristic of local and regional entities, as they make greater effort in this area due to a greater branch network and higher ratio of employees per assets volume. Commercial banks rely on information, proximity and trust and meet the needs of local and less profitable markets, prioritizing results in the long-term over short-term liquidity and profit.

This basic function of the banking industry may be jeopardized under a universal banking strategy by consolidation and merging processes, where local and regional banks can be ‘swallowed’ in an aggressive competitive market. Martin and Minns (1995) affirmed that financial markets prefer short-term assets due to their rising liquidity to more productive long-term investments, such as credit banks operations. There can be excessive investment in trading speed, because speed allows trading venues to differentiate and charge higher prices (Pagnotta and Philippon 2011). Internationalization of financial management and securitization of markets promote higher liquidity due to the commissions generated by increasing portfolio rotation, which reduces the incentives of shareholders to promote the responsible management of invested companies (Levine and Schmukler 2006). Credit to the retail market is less attractive because of the higher costs and greater uncertainty associated with domestic economies and particularly SMEs, which is aggravated during downward trends of the economy. Short-term targets oriented to maximizing shareholder value have been cited as the cause of several irresponsible practices in commercial banking business like, for example, the mis-selling of products, interest rates-rigging and financial exclusion of less profitable segments of the market (Bowman et al. 2014).

If we look at commercial banking lending to SME, technology is different from other types of loans (Berger and Udell 1995). The business of SMEs requires greater control by a senior manager than the management of loans based on a simple ratio of a credit scoring model. Thus, there are diseconomies of scale that make doing business with SMEs more expensive. In countries where small and medium banks still have a relevant presence, the infrastructure to attend to the retail market is more expensive, but it is rewarded by customers and, consequently, financial margins and efficiency ratios are higher (Carbó and Rodríguez 2014).

After financial liberalization and globalization processes, banking institutions have increased their size through mergers and acquisitions, which have intensified during the financial crisis . On one hand, a large international bank should face higher risks in terms of sustainability and reputation (i.e. corruption, money laundering, impacts on human rights and environment in big projects in developing countries). However, size is also related to more sophisticated CSR policies and management (Wu and Shen 2013; Scholtens 2009). Some factors, such as company size or age, industry or risk exposure, could even explain the relationship between CSR and financial performance, becoming ‘mediator variables’ (Garcia-Castro et al. 2010; Hull and Rothenberg 2008), but they do not show a clear statistical dependence (Aupperle et al. 1985; McGuire et al. 1990; McWilliams and Siegel 2001).

Meyer (1998) anticipated that the consolidation of the banking industry would reduce the credit supply to retail business and that big banks would tend to allocate fewer assets to small businesses. They have moved toward a universal banking approach, combining retail and investment banking. This tendency was the result of economies of scale, the need for greater efficiency and increased competition. At the same time, it has produced a trade-off between less profitable, leveraged and risky business in a more competitive sphere (investment banking) and the profitable, simple business in a less competitive sphere (traditional retail banking). Thus, external CSR performance during the financing processes before the crisis is expected to be lower. Market pressure in a context of lower interest rates and financial margins incited the banking industry to obtain better financial performance ratios (ROE) by increasing their activity and placing financial products irresponsibly (Bowman et al. 2014). Higher ROE ratios are observed in countries with less institutional diversity and larger entities (UK, France, Sweden).

Thus, if recent internationalization processes have led to a more universal and financialized banking industry, we would expect commercial banks still focused on traditional financial intermediation to have better external CSR performance than large investment and universal banks.

3 Analysis and Empirical Results

In this section, we propose a methodology to measure the social and environmental responsibility of banks, based on the construction of a self-built model exclusively oriented to the external dimension of CSR and core business. We applied this model, which can be fed by rough data from sustainability agencies, to a quite homogeneous, meaningful target group: European commercial banks. We have focused the analysis on the outbreak of the recent financial crisis and the more stable period before the crisis to determine if the crisis was also an inflection point in terms of external CSR for both retail and universal banks.

3.1 Sample

The sample was comprised of Western European financial institutions categorized as banks during the studied period. This means that their main or defining activity is financial intermediation, although some institutions may carry out other kinds of financial activities. We only considered companies listed on equity markets, due to the limitations of sustainability rating agencies regarding non-quoted entities (Scholtens 2009; Wu and Shen 2013). Thus, we obtained a nonhomogeneous sample from 14 countries (12 EU countries).

3.2 Sources

The historical database EIRIS,Footnote 3 an international ESG rating agency, was selected as the main data provider. The total sample size of this portfolio is 49 commercial banks with longitudinal pool data from 2006 to 2009. EIRIS evaluates the quality of policies, management systems, reporting and performance in over 80 ESG areas. It was selected due to its broad coverage of external CSR issues, consistent research and compared assessment of peers.

We also carried out a complementary qualitative analysis with ASSET4 Database (A4),Footnote 4 as statistical research remains insufficient to explain some of the results or potential conclusions. A4, a Thomson Reuters database, provides additional ESG information for 48 financial institutions (years 2006–2014), especially relevant in terms of consumer and product information. We have selected information regarding controversies to contrast the results on the external CSR performance of companies. This database also allows a broader temporal analysis, including the post-crisis period. Finally, financial information was obtained from BankscopeFootnote 5 database, ASSET4, or alternatively from financial statements.

3.3 Procedure

We constructed an index based on the external CSR information provided by EIRIS, that is, information on the scope of products and services. To minimize the potential ‘misleading effect’ of erroneous CSR ratings due to ‘greenwashing’ campaigns and similar factors (Wu and Shen 2013), we tried to select performance indicators instead of those simply related to policies. Thus, indicators were selected based on the following criteria: (i) Defined scope: the indicator was only related to external CSR aspects; (ii) Specific: it was only related to a particular environmental or social risk; (iii) Independent: it had no correlation with other indicators.

The index constructed with EIRIS information comprised 27 indicators, classified into eight different areas (‘Environmental management’, ‘Climate change’, ‘Project finance and sustainability ’, ‘Corruption/bribery’, ‘Consumers’, ‘Human rights’, ‘Developing countries’ and ‘Armament’). Every area becomes an independent variable because a sub-score is assigned to the indicators that compose the area. Some examples of external CSR indicators are project finance and sustainability risk in the area of ‘Environmental management’, product-related litigation/recalls in the area of ‘Consumers’ or third world mining/commodities corrections, due to investment activities in ‘Developing countries’. Examples of unsuitable indicators are ‘General Environmental Policy’ or ‘Systems/practices for job creation and security’, as they belong to a mixed or internal CSR dimension.

3.4 Variables

The statistical analysis was based on the scores of banks and their ranking position. However, we also included distinctive variables. Considering some characteristics related to the diversity of institutions in the banking market (Bowman et al. 2014) and loyalty to financial intermediation and the retail market (Fernández-Olit and Cuesta-González 2014), we defined the distinctive variables as size and banking orientation (retail, investment, corporate, etc.), number of branches and employees (intensity of service to the retail market), and weight of loans and customer deposits in relation to total assets (Table 3.1).

Table 3.1 Variables and information sources included in the analyses
Full size table

3.5 Analyses

The data analysis included a cluster analysis, ANOVA and a discriminant analysis. We used cluster analysis to obtain two groups distinguished by different levels of financial characteristics related to banking business. ANOVA allowed us to contrast the profile of the dimensions measured by EIRIS and the obtained clusters. Finally, discriminant analyses were performed including the groups obtained in the cluster analysis as criterion variables and the dimensions of EIRIS as predictor variables.

3.6 Results

3.6.1 Cluster Analysis: Obtaining Groups of Banks

Table 3.2 shows the results of the cluster analysis. Considering the financial characteristics that outline the size and business model of banks (retail/corporate, commercial/investment/universal), we obtained the following clusters: Cluster 1 with high scores in size variables (mean total assets = 1,585,127; mean number of employees = 143,559; branches = 6078; etc.) and Cluster 2 with low scores in terms of absolute size. These results indicate a link between size and the business model of banking. Cluster 1 refers to entities whose mean total assets were 7.2 times larger than in the case of Cluster 2 entities. Thus, Cluster 1 contained entities belonging to universal banking, that is, banks that provide a wide variety of financial services, including both commercial and investment services. When we weighted the relative results of these variables against mean total assets, Cluster 2 had a higher ratio of relative employees and branches, suggesting a stronger orientation to the retail market. We also found a higher weight of net loans and customer deposits related to total assets, which is distinctive of banks more oriented to basic financial intermediation: savings and loans.Footnote 6

Table 3.2 ANOVA of variables in the clustering process
Full size table

3.6.2 External CSR Index by Year and Cluster (2006–2009)

We carried out a Spearman correlation between the CSR index position of the banks in each year from 2006 to 2009 (see Table 3.3). This correlation decreased when we considered more distant years, except for the correlation between 2008 and 2006, which was the lowest. This indicates that there have been changes in the CSR index, which are reflected in a consecutive reduction of the average score from 2006 to 2008. Thus, European banks have moved backwards in external CSR development. This could be due to the rising number of regulator penalties during this period. These results are also consistent with the data offered by A4 regarding consumer controversies (0 companies in 2006, 9 in 2009, and 23 in 2010). Other possible factors are the greater exposure of the sector to badly managed risks in terms of sustainable project finance and greater exposure to commodities in developing countries.

Table 3.3 Spearman correlation of CSR index position from 2006 to 2009
Full size table

In fact, 2008 was a critical point: it was the year of the crash of global financial markets and a wake-up call for the whole banking sector to reconsider their way of providing financial services. This was also a year of a serious world food price crisis and increase in famine in many countries, which resulted in a higher level of commodities controversies for companies. On the other hand, this index distortion could be partly explained by a structural change, as a new data dimension (corruption) became available for inclusion in the model in 2008. Although the global influence of this indicator on the index mean has been neutral, it has ‘shaken’ the position of companies. This factor would also explain the out-of-trend results of the correlation between 2006 and 2008.

Figure 3.1 shows the tendency of the average scores of the external CSR index for the clustered banks. External CSR of the banks in Group 1 declined steadily with a very low average score in 2009 compared to 2006. Greater exposure to external markets and diversification of activities seem to have been penalized in the case of the largest banks: higher exposure to corruption risks; greater involvement in bad banking practices like subprime mortgages and manipulation of interest rates; greater interest in developing countries, in unsustainable project finance and in speculative investment activities or financing armament; and so on. Smaller banks, less exposed, seem to have avoided or managed those risks and controversies better.

Fig.3.1
figure 1

Average CSR score of each bank cluster by year (2006–2009) (Source: Own elaboration based on information from EIRIS)

Full size image

3.6.3 ANOVA and Differences Between Clusters

Statistical differences were found between groups for all dimensions included in our index except corruption, consumer care and human rights management (Table 3.4).

Table 3.4 ANOVA for EIRIS dimensions
Full size table

The most meaningful indicator was the dependent variable ‘total index’, followed by the independent variable ‘project finance and sustainability ’. This endorses the consistency of our model and shows differences between clusters in the management of external CSR as a whole. Integrating sustainable principles, for example, Equator Principles ,Footnote 7 into project finance activities is considered one of the key points in the development of real CSR in the banking business (Scholtens and Dam 2007b; Cuesta-González et al. 2006). As international banks face higher risks in project finance, they need more sophisticated management systems, which are still being developed in many entities. In contrast, a local focus of banking activity could allow project finance to be carried out with a lower sustainability risk, as more information is available. We also obtained some less meaningful variables (‘developing countries’ and ‘armament’) that are quite related to asset management activity. The profile of the two clusters is summarized in Fig. 3.2.

Fig.3.2
figure 2

Bank profiles in external CSR index dimensions (Source: Own elaboration based on information from EIRIS)

Full size image

We found that entities with a higher score in the external CSR index had a higher probability of belonging to Cluster 2. On the one hand, this means that the constructed index provides a ‘balanced picture’ of CSR in terms of the business dimension. On the other hand, Cluster 2 entities (medium- and small-listed banks) showed better external CSR performance. This could be explained by the lower sustainability risk of this group regarding project finance (even though their related management systems are slightly worse). The entities in Cluster 2 have also received fewer penalties from regulators, and fewer controversies have been recorded regarding commodities in poor countries or armament. Although large and universal banks have more developed management systems, they do not compensate for their higher CSR risk. Thus, Cluster 1 generates more CSR controversies.

3.6.4 Analysis of the ‘Consumer and Product’ Dimension

After analyzing the main sustainability risks that differentiate retail banking from universal banking, we focused on the ‘Consumers’ area, where the CSR performance of the last group is comparatively better. This area also includes social and environmental responsibility related to products and services. The specificity of the analysis regarding one single area allows us to consider a broader time scope. Thus, we are studying a longer period—from pre-crisis to post-crisis years: 2006–2014Footnote 8—to determine if there has been an inflection point in terms of consumer responsibility. As previously mentioned, the A4 database is especially relevant in terms of consumer and product information and offers a broad range of indicators in this area.

The ‘Consumers responsibility index’ constructed with A4 information comprised 21 indicators, classified into three different areas (‘Client Loyalty’, ‘Product Innovation ’ and ‘Product Responsibility’). Each area is equally weighted in the total index and becomes an independent variable because each is assigned its own sub-score based on the indicators that compose the area. Some examples of indicators are customer satisfaction transparency in the area of ‘Client Loyalty’, product innovation/improvements in the area of ‘Product Innovation’ or social exclusion controversies in ‘Product Responsibility’. Table 3.5 summarizes the selected indicators.

Table 3.5 Indicators composing the consumer responsibility index
Full size table

Figure 3.3 shows the mean score of the index by cluster. Larger banks demonstrated an advantaged position in ‘consumer responsibility ’ down to 2009. From this year their mean score decreased, falling below the mean score of retail banks. As Wu and Shen (2013) and Scholtens (2009) stated, more sophisticated CSR policies and management could be related to company size. This may be particularly true in the case of consumers, one of the main stakeholders in the traditional management of companies. As seen in Fig. 3.4, ‘product innovation’ was strong in the case of Cluster 1—larger banks—during this pre-crisis period, for example, a wider development of sustainable project finance . Notwithstanding, retail banking shows leadership in client loyalty during this period, reflecting the reward of customers to a simpler model of business with a heavier infrastructure.

Fig.3.3
figure 3

External CSR index based on ASSET4*. Mean by cluster (*Index converted to 0–100 scale. Note: 2014 shows a lower index level due to database methodological changes. Source: Own elaboration based on information from ASSET4)

Full size image
Fig.3.4
figure 4

Scores by areas of ASSET4 CSR index . Mean by cluster (Source: Own elaboration based on information from ASSET4)

Full size image

There is an inflection point in the year 2009 in both clusters, indicating a one-year delay from the outbreak of the crisis. This translated into costs in terms of consumer responsibility . For more traditional retail banking, it implied a temporal slowdown, with the index reaching a steady performance similar to the pre-crisis level during the post-crisis years. This performance is the result of two complementary tendencies: an improvement in the area of ‘product innovation’—the crisis became a wake-up call for responsible innovation in medium and small entities—and a deterioration of ‘product responsibility’ due to the increase of customer and product compliance controversies during this period. For universal banking—Cluster 1—consequences were greater: the consumer responsibility index suffered a sharp fall and did not recover its previous level during the post-crisis period. The intensity of controversies related to customer and product compliance was much greater for this group, as were anti-competition controversies. The priority of allocating resources to these controversial issues may have also led to the decrease in product innovation.

We also find evidence of differences among countries. Most banks in Italy—a country with an important retail sector—were represented in Cluster 2. Bankscope data showed that they had a heavier cost structure than, for example, banks in the UK, which were mainly classified in Cluster 1. This is balanced by a higher ratio of return on interest and a larger ratio of profit before taxes in the case of the Italian entities. It could indicate that users are willing to pay higher prices for ‘proximity banking’, supporting our results regarding ‘client loyalty’, but further research should be carried out on this topic (Table 3.6).

Table 3.6 Scores by areas of ASSET4 CSR index
Full size table

4 Discussion

Far from the instrumental vision of CSR mostly followed by large companies, this study has adopted an advanced approach to CSR, which is closer to the EU’s most recent definition: CSR is related to the impacts that business has on society. We have demonstrated that available data are not sufficient to determine if the banking industry is playing their economic role responsibly. Nevertheless, considering the CSR information contained in two important databases and the European banking industry analyzed in them, we can affirm that institutions which are closer to local or regional demands had better external CSR performance than large institutions with a universal and international strategy.

The objective of the present study was to identify any correlation between the external CSR performance of banks and their size and banking model. We have avoided giving excessive weight to ‘policies’ in the construction of our model, as we consider it does not reflect actual performance in the business scope. ‘External CSR index ’ and ‘Project Finance and Sustainability ’ are the variables that best discriminate between the bank clusters defined in terms of banking model. Size, greater internationalization and business diversification seem to be risk factors for external CSR. We conclude that entities with the highest levels of external CSR are found in the group of medium-sized banks with diverse orientations and less diversified business. These results are consistent with those obtained by Cuesta-González et al. (2006) regarding the lower external CSR performance of larger banks and the significant lack of information regarding the external dimension of CSR compared to the internal dimension of CSR provided by financial institutions.

Our findings partly refute research regarding the positive influence of bank size on CSR engagement among banking entities (Wu and Shen 2013; Scholtens 2009). At least in terms of CSR applied to commercial business, we found evidence that larger entities had a lower performance, except in the consumer sphere. Medium- and small-sized banks should be closer to customers, as they have higher ratios of employees and branches, but this does not always result in better performance regarding consumers. This could be because consumer protection is highly regulated. As all banks have to comply with the normative, few differences can be expected when using indicators regarding consumer relations based on regulation , like customer satisfaction transparency.

The higher level of engagement of larger banks in sustainable project finance does not compensate for their higher impacts. This conclusion questions the effectiveness of incentives like signing Equator Principles and related policies and leads us to ask if they are reduced to a mere formalism to enter the market of large institutional projects.

Corruption is not a discriminant factor in our model. However, our index has coincided with the banks penalized by the European Commission in December 2013 for tampering with interest rates. All of these entities belong to the cluster defined by large banks, and their external CSR performance decreased during the studied period. The implications of these findings are related to size and transparency requirements in the banking sector.

5 Conclusions

CSR has often been considered a whole made up of very heterogeneous elements. Attempts have also been made to define CSR using the same standards for every industrial sector. However, when we focus on the external or commercial dimension of CSR, we find that common assumptions or tools cannot be applied to different industries. Structural elements like corporate governance can be addressed in the same way for banking and mining, for instance, but it is impossible to assess the sustainable performance of their products in the same way. Our findings demonstrate that the commercial banking sector has different models of entities with particularities that must be addressed from different perspectives, considering different regulation requirements (e.g. in reporting) and even different expectations from society.

This research shows that size and banking orientation are relevant variables for external CSR performance. The current trend of homogenization in the banking sector does not seem to search for greater social responsibility in business. We look to a horizon of large, quoted banks with highly diversified activities. However, the theory of operative inefficiencies in large mergers during the 1990s may have resulted in current inefficiencies in terms of CSR. Alternative governance models (cooperative, public, semipublic savings banks), small- and medium-sized banks, and entities focused on financial intermediation and the retail market are being undervalued and induced to disappear in Europe.

The quality and availability of data are limiting factors. Sustainability rating agencies do not usually manage information on non-quoted banks or analyze all specific risks of this sector in depth. In fact, several impacts, whose relevance emerged during the crisis, are not included in the available information. As company reporting is the main source of information for rating agencies, the improvement of databases and analysis quality requires a greater engagement of banks. Soft or hard regulation (depending on the bank’s risk level) would be desirable to guarantee the inclusion of core CSR information in management reporting. This information could be relevant for policy-makers and supervisor’s authorities. External banking CSR performance could be used for public aids, facilities for solvency requirements or new incentives for access to wholesale funding or branch network expansion. This new information could be demanded once the new directive on extra-financial information is in force. Investors may also appreciate this transparency to better analyze the risks associated with the business and the good governance of institutions.

Moving forward in more sustainable financial systems requires a return to a business model closer to customers. The crisis has shown that banks have failed to design and market products that meet the customers’ needs in transparent ways, given customers’ lack of financial literacy, asymmetrical information and inertia. Some post-crisis measures (like the pressure on banks to lend to SMEs) were opportunistically aligned with part of these wider concerns, like access to finance, but do not address the broader issue of how banks deal with their business customers. There is a need for CSR that has an impact on the behavior of the retail banking sector, as this sector supports the (productive) economy. Further research is needed to propose changes in the transactional banking model to obtain a more relational banking model.

Technology can be helpful on the road back to relational banking , as banks have access to large amounts of data which allows them to offer their clients customized products. This personalized, client-centered banking can help democratize financial services by offering advice to low-income clients using data techniques (data management, statistics and algorithms, big data). However, we must be careful because online banking also favors the standardization of products and services and risk management procedures based on predesigned credit scoring, which are not suitable for a low-income vulnerable population. The replacement of offices with technology requires the population to have adequate access, knowledge and confidence, as well as a medium or high level of financial knowledge. Therefore, the guidance of banking experts seems to be essential in the decision-making process of these consumers.

Although initiatives promoting social inclusion, such as the EU Directive on Payment Accounts guaranteeing access to banking services, are well received, they are insufficient. Other initiatives could also be promoted at the European level to encourage the breakdown of information on the banking business, areas and groups at risk of social exclusion. This reporting model has a long history since the approval of the Home Mortgage Disclosure Act of 1975 and the Community Reinvestment Act of 1977 in the United States. In this way, policy-makers would be able to evaluate if the banking industry is fulfilling its role of facilitating responsible and inclusive access to banking services.

In conclusion, a more relational and sustainable business model would help banking institutions to address the UN’s Sustainable Development Goals, specifically the 12th goal (responsible consumption and production) and the 10th goal (reduce inequality avoiding over indebtedness and financial exclusion ).