Don ’ t bank on it : Delineating the relationship between corporate social responsibility and retail banking affinity

This research examines the extent to which the moderating variable of awareness influences the relationship between Corporate Social Responsibility beliefs and consumers’ attitudes towards their banks, and whether this significantly affects their willingness to recommend the company. The research was limited to the four major South African retail banks. The study finds that consumers’ attitudes do not, in fact, mediate the relationship between CSR beliefs and willingness to recommend. However, a direct positive relationship appears to exist between attitude and willingness to recommend. The authors find that a superficial awareness of CSR initiatives has minimal impact on their behaviour. Whilst an increased intimate knowledge of their CSR activities may thus lead to business rewards, banks should focus on their core offerings as consumers see CSR as an added benefit.


Introduction
Banks are not the most loved companies at the best of times.This situation is further compounded by competing in an industry with a high degree of parity across product offerings.Banks are therefore constantly challenged to differentiate themselves from the competition.Some have sought refuge in positioning themselves as consumer champions, whilst others have positioned themselves as environmental champions.
Their performance in the league of domestic brands appears somewhat underwhelming.Table 1 reflects that South African banks have performed reasonably poorly in terms of popularity over the past few years.During both 2009 and 2010, no bank featured in the top ten favourite brands, while Nedbank is the only bank to feature in the top ten environmentally friendly brands (holding 3rd position in 2009 and 2010).Therefore, it would appear that banks' branding activities leave a lot to be desired.
Furthermore, Absa appears to be the only bank recognized for its efforts in uplifting the community.Table 2, below, tracks community upliftment between 2006 and 2010.It can be seen that Absa is the only bank to feature in the list fluctuating between fifth and seventh position.Again, it appears that the major banks fall short in terms of promoting themselves as companies who deeply care for the communities in which they operate.

Table 1: South Africa's top brands -Favourite & Environmentally Friendly (2010)
In order to excel in a competitive market place, banks are turning their attention towards Corporate Social Responsibility as an essential component of their marketing strategies (Jamali, 2007).The World Business Council for Sustainable Development defines CSR as "the continuing commitment by business to behave ethically and contribute to economic development while improving the quality of life of the workforce and their families as well as of the local community and society at large" (Moir, 2001) In the early twentieth century, wealthy companies were strongly criticized for their business practices.In the 1960s and 70s, American society began to expect businesses to voluntarily participate in solving society's problems, based on their belief that corporations had the legal and economic responsibilities to do so (Barnett, 2006).This trend surfaced in South Africa during the early 1990s.
Despite local banks increasing investment in CSR initiatives, it is largely unknown whether this has resulted in favourable customer sentiment and their willingness to recommend the institution.Furthermore, whilst numerous studies have considered CSR within a South African context (e.g.Irwin, 2003;Tustin & Pienaar, 2005;Skinner & Mersham, 2008), the authors were unable to locate substantial scholarly studies which focused specifically on the banking sector.It is hoped that this study will therefore shed new light on the effect on CSR within this particular domain.

Conceptual framework
Corporate social responsibility Donaldson and Preston (1995), Snider, Hill and Martin (2003), amongst others, suggest that CSR be employed as a marketing mechanism to stakeholders.An important aspect to consider, and the focus of this research, is the consumer perspective of CSR.In recent years, South African banks have invested heavily in sponsoring sporting events, such as the ABSA Currie Cup, Standard Bank Pro20 Series and Nedbank Golf Challenge.Standard Bank has also funded the National Arts Festival and Nedbank has been a longtime backer of wildlife initiatives.As noted by scholars such as Lantos (2002), such investment is seldom completely altruistic in nature.
Empirical research illustrates the various effects CSR has on employees, firm market value and consumers.Sen, Bhattacharya and Korschun (2006) postulate that CSR increases employees' resolve to commit their time and money to the firm's profit.Simpson and Kohers (2002), Orlitzky, Schmidt and Rynes (2003), as well as Griffin and Mahon (1997), suggest that a significant positive relationship exists between firms' CSR rating and their financial performance.Murray and Vogel (1997) have identified two categories of effects on consumers, namely, attitudinal and behavioural.
The process begins with the stakeholders becoming aware of a company's initiatives.In turn, this awareness leads to knowledge about the company and its corporate and social values.Then, over time, familiarity of the above evokes favourable attitudes and builds momentum in terms of support of, and commitment to, the firm.This process provides the basis on which this research will be based, focusing on awareness, attitudes and behaviour.

CSR's effects on consumers
Brown and Dacin's (1997) pioneering research indicated that a firm's CSR record can have a positive effect on a consumer's overall view of the firm and, through this, an effect on product preference.Porter and Kramer (2006), Jamali (2007) and Holme (2010) appear to concur with this notion.While many companies implement CSR strategies, the degree to which these actions contribute toward the brand varies greatly.This leads to the question of whether a company that focuses its brand more on CSR is likely to benefit from this endeavour.
Research suggests that a consumer's view of product quality and value can indeed be positively influenced more significantly if, in the eyes of the consumer, CSR is a core element of the brand.Taking a step further, it has been noted that brands which are positioned with CSR at their core are likely to elicit higher levels of loyalty and advocacy than those brands which appear to support social initiatives on isolated occasions.Thus, the authors hypothesise that: H 1 : CSR beliefs have a direct positive effect on consumers' attitudes towards the bank

Moderating effect of awareness
While the effects of CSR on consumers may be welldocumented, the issue of whether consumers are even aware of those activities has seldom been investigated.Sen, Bhattacharya and Korschun (2006) point out that a significant problem with research into CSR is that "awareness is either assumed or artificially induced." Building on the research of Sen and Bhattacharya's (2001) into consumer-company identification, Maignan and Ferrell (2001) propose that stakeholders' awareness of companies' impacts on specific issues is a requirement to organisational identification.However, they argue that this consumer identification relies on the degree to which firms communicate their CSR activities.
Du, Bhattacharya and Sen (2007) note that, naturally, awareness itself does not lead to favourable CSR beliefs, but rather the consumer perceptions of the company's motive for its CSR activities.These motives can be split into extrinsic or self-interested motives (with the goal of benefitting the company) and intrinsic or selfless motives (with the goal of benefitting the community, or part thereof).However, a trade-off between these two types of motives does not necessarily exist as CSR activities may be attributed to both extrinsic and intrinsic motives (Ellen, Webb & Mohr, 2006).The effect of CSR awareness on the consumer's CSR belief is moderated negatively by the perception of extrinsic motives and moderated positively by perception of intrinsic motives (Du et al., 2007).
As previously discussed, Murray and Vogel's (1997) five step process for generating goodwill from CSR demonstrates the need for consumer awareness before positive attitudes can be formed.Thus, focusing on corporate social responsibility, the following hypothesis is proposed: CSR awareness has a positive moderating effect on the relationship between CSR beliefs and attitude towards the bank.
Therefore, the authors expect that consumers with higher awareness will have more positive attitudes than those with lower awareness levels.

Level of consumer awareness
Despite the moderating impact that awareness seems to have on the effects of CSR, companies clearly have not communicated their activities effectively.In their field experiment, Sen et al. (2006) found that a fairly low proportion (17%) of respondents were aware of the CSR initiative being investigated.Those who were aware of the CSR project had heard via the normal channels such as newspapers or word-of-mouth and, as a result, had considerably more positive company associations and greater intention to purchase products.Boulstridge and Carrington (2000) also found that awareness of company activity in corporate social responsibility was very low despite increased media coverage and business activity in the area.The authors hypothesise: H 3 .South African consumers have a low awareness of their banks' CSR activities.

Consumer willingness to recommend
Typically when investigating the effects of consumer attitudes on their consequent behaviour, researchers focus on customer loyalty and customer satisfaction.However, high switching costs of the banking industry make these constructs difficult to measure.These high switching costs seem to be prevalent in both South Africa, as well as other countries, resulting in customers feeling locked-in, yet dissatisfied.While these switching costs may result in high retention rates, banks risk damaging their own brand.Being unable to switch service providers at will results in reduced number of recommendations and increased negative wordof-mouth (Lam, Shankar, Erramilli & Murthy, 2004).
It is clear that customer satisfaction and customer retention do not necessarily reflect attitudes of consumers towards the banking industry.As such, this study intends to use consumers' willingness to recommend their bank as a measure of behaviour that is likely to better reflect their state of mind.
Since customer attitudes affect overall satisfaction, and satisfaction directly affects future behavioural intentions (such as future visits to the branch and willingness to recommend), it follows that attitudes have the ability to affect future behavioural intentions.Baumann, Burton, Elliot and Kehr (2006) support this view in their research that revealed that empathy was a very strong predictor in customers' willingness to recommend and overall behavioural intentions.Customers who are positively biased to their banking establishments view the firms in a positive light resulting in greater customer satisfaction.Thus, they are willing to recommend these services to friends and family (Baumann et al., 2006).
In their investigation into how consumers handle and interpret persuasion communication, Friestad and Wright (1994) proposed that over time consumers develop personal knowledge about marketing communications.They do this through first hand experiences, conversations about how people can be influenced, and observing marketers and marketing campaigns.The personal knowledge they gain enables them to shape their attitudes towards the brand or company, ultimately, helping them respond to these persuasion attempts in order to achieve their own goals.This highlights the way in which attitudes mediate consumers' beliefs and behaviour.Therefore, the authors hypothesize: H 4 .Attitude towards the bank has a positive direct effect on willingness to recommend it.
H 5 .Attitude towards the bank mediates the relationship between CSR beliefs and willingness to recommend it.

Conceptual model
Literature has provided strong evidence that CSR has a positive effect on consumer attitudes, and that these attitudes evoke favourable behaviours, such as the willingness to recommend the company.However, the intensity of consumer awareness of those CSR activities is likely to affect this relationship.The conceptual model depicted in figure 1 is proposed.

Sample
A non-probability sampling technique was used in this study to obtain respondents for the field work.This cohort consisted of consumers from the four major retail banks in South Africa and was restricted to the residents of a single city.The sample was thereafter segmented according to bank of choice, age and race as these were the factors decidedly likely to influence results.

Questionnaire design
Based on our conceptual model, the following Likert scales were adapted from previous research as shown in Appendices 1a to 1e: CSR beliefs (Sen & Bhattacharya, 2001), attitudes towards company (Sen & Bhattacharya, 2001), and willingness to recommend (Zeithaml, Berry & Parasuraman, 1996).The original Likert-scale format was retained, allowing many multivariate statistical techniques to be employed for data analysis.In addition, a Likert-scale measuring attitude towards the company's CSR was developed.
Please note that hypothesis 3 is not causal in nature and, for this reason, was not included in the model The questionnaire was tested in various ways.Pre-testing by peers and consumers ensured that the questionnaire was error free and comprehendible.Academic consultants then confirmed its face validity, ensuring that the questions corresponded to the constructs being tested.Lastly, the institutional ethics committee confirmed that the questionnaire was not worded in such a way that it would offend any group of people.

Data collection
Although an online questionnaire may have been more convenient, some problems were anticipated including a high non-response rate, inadequate access to lower Living Standards Measure (LSM) groups and the lack of a large email database.As such, a hard-copy format was the primary source of data, with online surveys supplementing for the shortfalls.The online questionnaires provided an efficient and real-time means of retrieving data needed to make up the shortfalls of the white and middle-age categories.Although at times non-response was problematic, the reduced cost was beneficial.

Results
In general, the results did not match the predictions made.
As expected, generally there was a very low awareness of the banks' CSR activities.However, different levels of awareness did not have a bearing on attitudes towards banks, as was first thought, due to the fact that CSR beliefs did not appear to affect attitudes towards these institutions.It was, however, confirmed that attitudes had the power to influence consumers' willingness to recommend a bank.

Preliminary checks and controls
As was initially believed, the manipulation of respondents by providing some with intimate knowledge of their bank's CSR did generally have an effect on their responses to questions.The only construct that was not affected was willingness to recommend, in which the ANOVA was highly insignificant.Despite this, further statistical techniques across all constructs were split according to those respondents that were provided CSR knowledge (called the "informed" group) and those who were not (called the "uninformed" group).
The means (see Appendix 2) demonstrate the differences in the constructs between informed and uninformed sample groups, with CSR beliefs and attitude toward bank CSR increasing as knowledge increases.However, an inverse relationship exists in the attitude toward bank construct.Willingness to recommend shows little difference in means.Lastly, few differences between banks emerged across the constructs.

Measure validation
In order to assess the multi-item measures, both reliability and validity were assessed.
Confirmatory factor analysis found that three of the four constructs had a single factor, illustrating the commonality across each construct.Attitude toward bank CSR was found to have two factors, though, with two items loading onto a secondary factor.As most variance was explained through the first factor, this was retained whilst the second factor (containing two items) was removed from the study.
According to the Cronbach alphas and inter-item correlations (as seen in Table 3), all constructs were deemed highly reliable as even the lowest value (0,79) safely exceeded the cut-off of 0,7.Moderation: Table 5 displays a summary of the informed and uninformed moderation multiple regression effects.The low t-values and lack of significance of the interaction effect is indicative that no moderation effect exists.With no significant relationship between CSR beliefs and attitude toward bank, the presence of a moderating relationship with total awareness becomes irrelevant.This was supported by the general centring of the coefficients around zero.This resulted in the overall insignificance of the model demonstrated by the low F-stat values (0,460 and 0,134) and negative adjusted R 2 values (-0,015 and -0,020).

Verification of hypotheses
Mediation: Both informed and uninformed respondent subsets were tested for the mediation effect using four separate regression analyses.It was expected that mediation would exist for both cohorts.However, this was not found to be the case.
The insignificant relationships between CSR beliefs and attitude toward bank (p > 0,1) inhibit any significant mediation result.This is supported by the significance of the models when two independent variables are added (uninformed: p < 0,001, informed: p < 0,05).Table 6 displays the statistical parameters.
Both sample groups showed significant and positive relationships between willingness to recommend and attitude toward bank (uninformed: p < 0,001, informed: p < 0,05).The uninformed group's attitude had a relatively large effect (B = 0,597) on their willingness to recommend, while the informed group's effect was smaller (B=0,235).Interestingly, the uninformed group's CSR beliefs had a significant negative effect (p < 0,001, B = -0,268) on their willingness to recommend.Unsurprisingly, the adjusted R 2 values are low as other influential factors have not been included in the models.

Demographic variance
ANOVAs: While there were several additional significant differences within demographic factors, only key ones were extracted and explained by comparing the actual means.This is depicted in Table 7 below.The study confirmed our beliefs that race would influence consumers' attitudes and behaviours.
Firstly, white people in the uninformed sample group had worse attitudes toward their bank's CSR activities, whereas black people in the informed sample group had much better attitudes.Intimate knowledge appeared to have little effect on coloured people.It was surprising to find that black people had a much lower level of awareness of CSR activities than other race groups as this was the group of people most likely to benefit from many of the CSR initiatives.Lastly, white people in the informed sample group were less willing to recommend their bank than other race groups.
Although the largest difference occurred between the two wealthiest groups, the uninformed R 25 000+ income group was found to have much worse attitudes towards bank CSR activities than all other groups.However, the construct means indicated that they were impacted the most when told of the details of the CSR activities.Also, men were found to have better attitudes toward their bank's CSR activities when they have intimate knowledge.Thus knowledge appears to have a larger effect on men than on women.
Lastly, the uninformed sample group indicated that as people grow older, their attitudes towards their bank become more negative.However, this trend is not apparent in the informed sample.In addition, it was revealed that older age groups (42 and above) in the uninformed sample group have worse attitudes towards their bank's CSR activities than the younger age groups.This may be due to a heightened degree of cynicism.
Discriminant Analysis: Race was the only demographic variable that had prediction quality above 60%.
A common distinguishing factor of race groups across both sample groups (shown in Table 8) was their attitude toward their bank CSR.Another important distinguishing factor for the uninformed sample group was their CSR beliefs.However, when intimate knowledge of CSR is acquired this is no longer an important factor, but instead their attitude toward their bank is.While the prediction rates were moderate (uninformed = 63%, informed = 64%), more significant predictor variables are required to improve the model.

Discussion
Several conclusions may be drawn from the empirical analysis.
Effects of CSR beliefs: It was hypothesised that consumers' CSR beliefs would have a direct effect on their attitude towards their bank.However, this was found not to be true as consumers' attitudes towards their bank are seemingly formed by their evaluation of the core functions of their bank.It is debatable as to whether consumers' CSR beliefs have an effect on their willingness to recommend their bank, although intimate knowledge of CSR activities appears to have a significant impact considering the negative relationship between the factors in the uninformed sample group.Nonetheless, a crucial determinant of consumers' willingness to recommend their bank almost certainly remains its quality of service delivery.CSR cannot hope to be a replacement of such.
Mediating and moderating effects: Since no relationship was found to exist between consumers' CSR beliefs and their attitude toward their bank, it is clear that their attitude toward their bank cannot mediate this relationship.In addition, it was found that CSR awareness does not have a moderating effect on relationship between consumers' CSR beliefs and their attitude toward their bank.
Intimate CSR knowledge: Despite the fact that the study showed that a superficial awareness of CSR does not affect their attitudes, it was discovered that an intimate knowledge of the CSR activities does influence people in three ways: (1) it improves their beliefs regarding CSR, (2) it improves their attitudes toward the bank, and (3) it improves their attitude toward the bank's CSR activities.CSR awareness levels: As hypothesised, consumers had a considerably low awareness of their bank's CSR activities.This raises the question as to why banks engage in CSR initiatives (altruism or to market themselves) and whether banks are effectively showcasing their CSR investment.
Effect of attitudes: The fourth hypothesis that consumers' attitudes toward their bank has a significant and positive direct relationship with their willingness to recommend their bank was accepted, although attitudes toward banks appear generally poor due to the perceptions of substandard service.
The research also produced many findings relating to the differences across different demographic factors.
Race: It is clear that black people have a much lower awareness of CSR activities than other race groups as many peri-urban dwellers possibly see banks in a fundamental manner due to lack of knowledge and interest.It was also highlighted that white people, with or without intimate CSR knowledge, are less willing to recommend their bank.Lastly, white people who do not know about the CSR activities are indifferent about them and little loyalty is formed, whereas a great sense of loyalty is formed in black people who have knowledge of the CSR initiatives.
Income: The uninformed R25 000+ group was found to have much worse attitudes towards bank CSR activities than other groups.However they were impacted the most when told of the details of the CSR activities as the barrier of scepticism is overcome.
Gender: Although the general feeling of consumers is that gender makes little difference in how CSR is viewed, it has been suggested that women are more likely to be swayed by their emotions.However, it was actually found that the loyalty aspect of people's attitudes toward their banks' CSR was more likely to affect men in a favourable manner.
Conceivably, this may be due to a large proportion of females assuming responsibility for the budgeting of household needs.As this study was conducted during the 2008 recession, it is entirely possible that women were focused on financial prudence first and foremost.
Age: In the uninformed sample group, it is evident that as people grow older their attitudes toward their bank become more negative.This is likely to occur as people critically evaluate their bank as they become more involved with it.
Secondly, the study reveals that older age groups (42 and above) in the uninformed sample group have worse attitudes towards their bank's CSR activities than the younger age groups.This appears to be a result of the lack of disclosure as to where clients' money is being spent with regards to CSR.

Managerial implications
Strategic CSR: It is crucial for banks to clarify the objectives of their CSR initiatives.It should be clear as to whether their CSR activities are a means of building their brand or whether their motives are altruistic.If their intentions are profit-driven, banks should recognise that in order to gain rewards from CSR, it is necessary to provide customers with in-depth knowledge of their banks CSR activities.Simply creating superficial awareness of such activities does not convince consumers.In addition, banks need to align their CSR activities with both their target market and their products and services.If customers do not feel connected to CSR activities, they are less likely to retain knowledge of the CSR initiative and are thus less likely to exhibit loyalty for this reason.
Communicating CSR: The lack of consumer awareness demonstrates the need for banks to increase overall awareness of their CSR activities.Banks should create awareness of their CSR activities without it being perceived as a marketing ploy.This may be done through mailing informative brochures or magazines to consumers, or by using point-of-service advertising within banks linking the brand to causes without doing so in an ostentatious manner.
Service standards: In order to improve willingness to recommend their bank, these institutions must continue to improve their clients' attitudes by focusing on their core business functions, such as their products and customer service levels.This is important as consumers typically view CSR as a value-add but not a substitute for good service.

Limitations and further research opportunities
The study had a number of limitations which may serve to create opportunities for further research to be conducted.The diversity of consumers in South Africa posed a challenge as it was difficult to fully apply high-income country theory as different population segments reflect on companies' marketing in different ways.This research could not fully capture the cultural differences in the perspectives of consumers towards their banks.However, it clearly shows that these differences do exist.Further research into these differences would allow banks to develop superior marketing, and additionally CSR, strategies.
All research was restricted to a single city within South Africa, causing possible variance in opinions from South Africans from the other cities and provinces.As such, it is suggested that research should be expanded to the greater South African population reducing sampling error and discrepancies in partial generalisations.
The effects of high switching costs of South African banks on consumer attitudes and behaviour certainly merits further research in the context of CSR initiatives.Introducing all these variables into a single study may be highly beneficial.
Lastly, research may be conducted to ascertain whether this study's results apply solely to the banking industry or whether they can be applied to the entire financial services industry.

Figure 1 :
Conceptual modelTwo template questionnaires were developed for each bank.One type of survey, the control group, explored the consumer's view of their bank before introducing the topic of CSR.The other type of survey would introduce CSR from the beginning as well as elaborate on the CSR activities of their bank in an attempt to manipulate the attitudinal responses.Through this, comparisons can be explored between consumers that were manipulated and those that were not.The questionnaire also investigated both unaided and aided awareness of the CSR activities of the consumer's bank in an attempt to replicate the Du et al.(2007)  study which tested unaided and aided awareness a year after the particular campaign they investigated.The unaided openended awareness question was asked first and placed on a separate page to the aided CSR list to ensure that consumers did not simply copy the listed activities.