A reconstruction of Carroll’s pyramid of corporate social responsibility for the 21st century

Critique of prioritisation of economic responsibilities

Carroll’s justification for prioritising economic responsibility was that if a business does not make a profit then it will not survive, and then the other responsibilities become a moot point. However it can also be argued that if a business cannot be profitable without breaking legal or ethical norms, then it would be in the best interests of society if it were not in business at all (Kang & Wood 1995). Indeed, most commentators agree that it is the profit before ethics mentality that is at the root of most corporate misbehaviour. Businesses in a competitive environment are subject to many temptations to place profits before the law, and numerous studies demonstrate how this gives rise to illegal and/or unethical business practices (Bennett et al. 2013; Cai & Liu 2009; Shleifer 2004; Staw & Szwajkowski 1975). Goodpaster (2007) coined the term ‘teleopathy’ to capture the unbalanced focus on one goal above all others – typically profit – which he believed was the root cause of many business ethics failures such as WorldCom, Enron and Arthur Andersen. Goodpaster criticises the laws and social norms which are perpetuated by our business models and business education, which lead to a pathological profits-first mind-set. Instead, Goodpaster advocates the introduction of moral criteria in management decision-making as an antidote. This view is echoed by many commentators (Friedman et al. 2008; Greider 2003; Kolp & Rea 2006) who all argue that, in order to avoid repeated business scandals, CEOs need to refocus from prioritizing shareholder wealth or maximizing earnings per share, to realizing their ethical responsibilities i.e. creating an ethical corporation that attends to all stakeholder needs. Charles Handy (2002) claims that to put profit before all other goals is to mistake the means for the end arguing that: “The purpose of a business, in other words, is not to make a profit, full stop. It is to make a profit in order to enable it to do something more or better. What that ‘something’ is becomes the real justification for the existence of the business” (p.4).

Kang and Wood (1995) criticise the implicit hierarchy of responsibilities in Carroll’s pyramid of CSR, arguing that placing economic responsibilities before legal and ethical responsibilities makes it more likely that social welfare will be sacrificed for economic ends. Carroll (1991) was not unaware of this possibility, and posits that the four types of responsibility stand in in constant dynamic tension with one another, with the most critical tensions likely to be found between economic and legal, economic and ethical, and economic and philanthropic obligations. This contention is supported by one of the first studies that empirically tested the relative priorities given to the four types of responsibility by business (Aupperle 1984). Aupperle found negative correlations between the economic factors and the others (which all correlated positively with each other), in particular a strong inverse relationship between the economic domain and the ethical domains. This suggests that the more weight is given to economic concerns, the less to ethical concerns, a finding that was replicated by Clarkson (1988) in a follow up study. This raises the stakes for the well-being of society and the environment of perpetuating a model of CSR that prioritises economic concerns over ethical concerns.

Sachs and Ruehle (2009) agree, highlighting the example of Enron whose prioritisation of economic over ethical and legal duties resulted in a major business ethics scandal. As argued by Sachs and Ruehle, corporations cannot be seen as purely economic entities detached from the individual. So CSR is at heart primarily a moral concept, designed to highlight the responsibility of business to (as a minimum) avoid causing harm to society and the environment, or more proactively, contributing to the welfare of society and its stakeholders. The implications of this is that economic concerns should not be prioritised over social concerns as this entails treating people as a means to an end (i.e. profit) rather than ends in themselves which violates fundamental ethical principles. In particular, Kant’s maxim from his ‘Groundwork of the Metaphysics of Morals is that we should: “Act in such a way that you treat humanity, whether in your own person or in the person of another, always at the same time as an end and never simply as a means” (Kant 1785). This maxim is based on Kant’s claim that morality is based upon reason, and as humans we have a right to respect, and a rational person would never wish to be treated just as a means to someone else’s end. It follows, according to Kant’s principle of universality, that we should in turn not treat anyone else as a means to our end, unless we are also treating them as an end in themselves.

CSR scholarship and discourse

Carroll’s pyramid of CSR in context

As one of the first and most cited and most influential models of CSR (Schwartz & Carroll 2003) this precedence given to economic responsibilities has permeated the subsequent literature and the discourse surrounding CSR. Nevertheless, Carroll’s pyramid of CSR can clearly not be held responsible for the entire direction of the CSR discourse. Indeed, almost as prevalent in CSR textbooks as Carroll’s pyramid of CSR is Friedman’s assertions that “the social responsibility of business is to make a profit” ((Friedman 1970), p. 13). Brooks (2010) argues that the response of the CSR academy to Friedman’s pronouncement led to a wrong turn in the development of the CSR construct, as rather than justifying CSR on moral grounds, the academic community (of which Carroll was a part), responded by attempting to justify CSR in economic terms. Brooks argues that this has resulted in an almost obsessive quest to find a positive relationship between corporate social performance (CSP) and CFP that has dominated the CSR discourse ever since, leading to a marginalization of moral considerations. It has been argued that CSR has won the battle, but lost the war in the sense that although Carroll’s business-friendly model may have allowed CSR to have become accepted and even embraced by the business community, it does not allow ethical or even legal arguments to take precedence over economic arguments (Brooks 2010). Indeed if CSR was truly a moral responsibility then its relationship to CFP should be moot, yet while there are numerous articles relating to the relationship between CSR and CFP, there are very few exploring the social and environmental impacts of CSR (Baden & Harwood 2013), providing an indication of how far Carroll’s conception of CSR is from being a model of genuine social responsibility.

However Adam Smith in ‘The Theory or Moral Sentiments’ (Smith 1759) makes it clear that he is not advocating an economic system based solely on self-interest, and the free-market is only possible on the assumption that man is a social animal with moral sentiments designed by nature to live in harmony with society. Despite this, the notion that society is best served by an economic system based on self-interest is the idea that stuck, as illustrated by Heyne’s selective summary of Smith’s thought: “If Adam Smith was right, we are wasting our time attempting to define social responsibilities of businessmen. The public good will be promoted most effectively when each businessman assiduously pursues his own advantage” ((Heyne 1968) p. 7–8).

This widespread belief that allowing profit seeking businesses relative freedom to pursue their own interests is also to the benefit of society has been subject to numerous challenges, not only by political opponents of the free-market capitalism, but also increasingly by main stream commentators and indeed the general public, who have been witness and often victim to the destructive consequences of self-serving corporate activity (Mattick 2011; Palmer et al. 2012; Trudell 2012). An illustration of the prevalence of such views are the ‘Occupy Wall Street’ protests (Guardian 2011) which affected approximately 900 cities globally. Woller critiques the liberal business ethos of the US in particular and questions how managers can be expected to make trade-offs between environmental degradation and profitability when managers’ only stated obligation is to the shareholders. Woller contends that “the liberal business ethos thus acts as an intellectual prison that restricts the way in which managers, and society, for that matter, frame the issues, identify alternatives, and make decisions.”((Woller 1996) p. 313).

Empirical research on the Carroll’s four responsibilities of business

There is a distinction to be made between normative accounts of what CSR ought to mean and descriptive accounts of how businesses actually prioritise and perceive their business responsibilities. One of the first efforts to test Carroll’s pyramid of CSR empirically was conducted by Aupperle in 1984, based on CEOs’ weighting of items that reflect the four domains. The relative weights given the items conformed to the ordering of responsibilities presented by Carroll i.e. economic, legal, ethical and lastly, philanthropic. Pinkston and Carroll in their study of MNCs (Pinkston & Carroll 1996) confirmed this order of weightings. A study of black-owned businesses in the US suggested cultural differences, as a relatively higher weighting was given to ethical and philanthropic responsibilities than in previous studies. Economic responsibilities were still rated highest and philanthropic responsibilities lowest, but ethical responsibilities were rated higher than legal responsibilities (Edmondson & Carroll 1999). However all this research took place in the US and, more importantly, was based on a sample of top managers and CEOs who, while they may represent the interest of business, cannot be said to speak for society as a whole. In addition this research is now decades old and so not necessarily representative of current expectations of business, particularly bearing in mind the vastly greater domains of influence held by MNCs in the globalised economy.

A more recent study (Pedersen 2010) based on a survey of over 1000 business managers from international firms from a range of sectors and countries, on their “own views on the business unit’s responsibilities towards society” (p. 157), found that the most prevalent responsibility mentioned related to their firm’s environmental responsibilities. Other responsibilities listed in order of frequency pertained to product quality, employee well-being, community and society, with responsibilities related to shareholder value receiving the fewest mentions of all. This research contradicts the ranking presented in Carroll’s pyramid as ethical responsibilities appear to be primary.

There have been several studies comparing differences across nations in terms of their awareness of CSR, levels of CSR practices, and relative preference for economic, legal, ethical and/or philanthropic practices (Burton et al. 2000; Carroll 1999; Chapple & Moon 2005; Küskü & Zarkada‐Fraser 2004; Lindgreen et al. 2009; Maignan 2001). For example Ramasamy and Yeung (2009) argue that consumer awareness of CSR is positively associated with the level of economic development and that religious and cultural differences explain the increased role of philanthropy in Asian cultures. Cultural differences were also noted in a study by Maignan (2001) which found that French and German consumers tended to rate legal and ethical responsibilities of primary importance, whereas American consumers rated economic responsibilities as primary. These differences were attributed to the more communitarian or collectivist European approach versus the emphasis on the individual notable in US culture (Hofstede 1983; Lodge 1990).

Differences have also been attributed to institutional conditions, for example, philanthropy has been proposed to be more important in developing countries that lack a well-developed welfare state (Amaeshi et al. 2006). Visser (2006) also argues that in developing nations such as Africa, economic responsibilities will still be primary due to the high demand for foreign direct investment and jobs. This contention however is purely speculative and was not supported by empirical research. It could alternatively be argued that although there may be a lesser awareness and/or practice of ethical responsibilities in developing countries, this by no means indicates that ethical responsibilities are less important than economic responsibilities. Indeed an ongoing criticism of MNC business operations in developing countries is that the economic value gained by their operations has been appropriated by the MNCs and not shared with the host countries, with employees from the host companies typically being earning substantially less than a living wage (Hasmath & Hsu 2007).

Carroll’s global pyramid of CSR

Since Carroll proposed his original four levels of responsibility in 1979, and later in 1991, reworked them into a pyramid, business has become increasingly globalised. This has created new challenges for ethical business, not just due to differing cultural norms, but also as a result of the decreased level of regulatory control as MNCs can often choose to site their operations in countries where poor working conditions, low wages and lack of effective regulation can be exploited. Carroll (2004) accordingly incorporated global responsibilities into his pyramid of CSR, noting that “regardless of what is happening in individual countries, whether at home or abroad, the primary venue for ethical debates in the future will more and more be the world stage” (p. 114). Carroll amended the CSR pyramid simply by adjusting the wording of the responsibilities e.g. economic responsibilities entail doing what is required by global capitalism. It is regrettable that Carroll did not take this opportunity to update the prioritisation of business responsibility to reflect the changing institutional conditions in which business now operated. An oft-noted problem with globalised business is that the global reach of business is not matched by the global reach of the law, as legal standards vary from country to country, thus it would have made sense for Carroll to acknowledge this regulatory gap in amending his pyramid. It would be expected that the adequacy of the legal framework to protect public welfare would play a role in the relative ranking of responsibilities. For instance, contrary to developed countries “it is the rule rather than the exception that companies do not comply with existing legal frameworks related to corruption, payment of taxes, fair trade practices, respect for human rights, customer services, [and] environmental protection’ in developing countries” ((Prieto Carron et al. 2006), p. 978). Not only do many developing countries have an inadequate legal framework to protect society and the environment, the one that is in place is inadequately enforced. This means that avoiding harm to the environment and the community is effectively an ethical rather than legal obligation for the corporation. Thus I would argue that a more appropriate pyramid for globalised business would rank ethical responsibilities above legal responsibilities.

Summary of literature review

The literature review presents conceptual arguments for ranking ethical and legal responsibilities above economic responsibilities and highlights how the precedence Carroll gave to economic responsibilities potentially undermines the moral power of the CSR construct. However conceptual arguments alone have little weight in practice unless they are supported by relevant stakeholders, which would include both business and general society. In the following section, a study is presented which empirically tests the ordering of responsibilities implied by Carroll’s pyramid of CSR to see if they are reflective of current views of business responsibility. Samples of respondents drawn from both business and non-business backgrounds are compared to determine if there are any differences between the business and the public’s perception of what business is or ought to be responsible for. In addition, the potential moderating effect of perceptions relating to the adequacy of the legal system to protect public welfare is explored.

Measures

The aim was to determine how respondents ranked the four types of business responsibility: economic, legal, ethical, and philanthropic. However to avoid ‘pat’ unconsidered answers, the first section presented five examples of each type of business responsibility and asked respondents to rate them in terms of importance. These examples were based on Carroll’s original conception of the four responsibilities and typical business practices coming under each category (Appendix 1). The order in which the responsibilities were presented were randomised using a Latin Square and reverse Latin square design to avoid order effects (although all five questions relating to each type of responsibility remained together). A test of reliability using Cronbach’s alpha indicated that for each category the five questions could be combined to form a composite measure of each type of responsibility (all alphas were above .85). Once respondents had been exposed to five examples of each type of business responsibility, they were asked to rank them in terms of relative importance - again the order was counterbalanced.

In addition, to determine whether the perceived effectiveness of the legal system would affect the relative ranking of legal responsibilities, an additional question was posed asking ‘to what extent are the social and environmental impacts of business covered by laws?’

Cohort 1

This research aimed to access a representative sample of respondents from both business and non-business domains. Getting access to reasonably high level business managers from a wide variety of sectors is a challenge for researchers, and in this study it was decided to utilise the connections of students from the business school, many of whom had parents, relatives or friends of the family who worked in business. Thus the research was situated within the context of providing an example of empirical research in the field of CSR, and used as an opportunity to teach students how to formulate a research question, gather data and analyse results. Students were asked to survey 2–4 respondents either from business or non-business over their vacation period.

Although this enabled access to a more diverse and substantial sample that could otherwise be reached, it also raises issues about the integrity of the data. To minimise the likelihood of data fabrication and/or social desirability bias, the following measures were taken: This assignment was instead of students’ regular weekly assignment, so students were already expecting to have to do some kind of vacation assignment. Each student had to identify how they accessed their sample and provide a consent form with contact details so the integrity of the data could be checked. The assignment was preceded by a discussion of Carroll’s CSR pyramid and students were encouraged to take a genuine interest in the results, and told that it may form part of a research paper so data integrity was vital. No specific hypotheses were proposed, instead it was couched as exploratory research to be conducted with a genuine spirit of inquiry with no investment in any particular outcome. When inputting their results into an online survey, students were also required to answer a question on the perceived integrity of their result which allowed them to indicate if they thought the results were not valid e.g. due to respondent showing a strong response bias or lack of interest. It was made clear that no negative consequences would follow if students inputted results they were not sure were valid, as long as this was clearly indicated and could thus be omitted from the final analysis. An anonymous poll taken at the end of the course asking if any students had fabricated results also provided reassurance that students had shown integrity in data gathering. The instructions given to the students are presented in Appendix 2.

This cohort includes an additional 37 respondents from the IT and finance sector in Jersey (UK), recruited via a DBA student working in that sector. This sample comprised consultants, directors and high level managers, two-thirds male, predominantly in the 26–55 age range. The link to the survey was emailed to high level staff (N = 170) of an IT firm, who were told it was research into how certain CSR practices are perceived by business. All responses were anonymous and there were 37 completed surveys.

Cohort 2

A second study used paid participants recruited from the website Amazon Mechanical Turk (MTurk) which provides a large and diverse sample of potential respondents, typically from India and the US. MTurk is an online crowd-sourcing marketplace that matches workers with those requesting assistance on tasks that do not require specific training, such as survey completion. This method is a fairly recent development, but studies comparing results gained from respondents from MTurk to more traditional samples have found that they provide equally reliable and valid answers (as measured by test-retest, alpha etc.) yet provide a more demographically diverse set from a range of backgrounds, income brackets, ages, genders and occupations (Buhrmester et al. 2011; Mason & Suri 2012). The strengths and weakness of Amazon Turk samples have been researched, and the conclusions have been that MTurk has democratised social science and enabled a more hetergenous sample and produces generally reliable results, particularly when the tools to exclude low quality participants are used (Paolacci & Chandler 2014). For this study, only workers who met the criteria of 95 % + approval rate from previous tasks were eligible in order to screen out workers who did not have track record for being conscientious. To maximise the validity and reliability, those that completed in less than the minimum time the survey was expected to take were considered unreliable and their responses were omitted.

Mean rating of importance of responsibilities

Ranking of responsibilities

Both means of measuring the relative importance of responsibilities give rise to the same picture – based on the total sample including both business and non-business respondents, the first priority are legal responsibilities, with ethical and economic responsibilities tying for second place, and philanthropic responsibilities perceived to be the least important.

Perception of legal system

It was proposed that perceptions of the legal system may act as a moderator, influencing the relative order of rankings such that those who believe the legal system is inadequate to protect public interest would be more likely to rank ethical responsibilities before legal responsibilities. On a scale of 1 (inadequate) to 6 (adequate), most of the sample thought the legal system was adequate to protect public welfare: mean = 3.75, but 16 % of respondents selected 1 or 2 representing inadequate law. When the analysis was re-run using just those who thought the legal system was inadequate to protect public welfare, the relative rankings changed in the direction predicted. When looking at the rankings, there is a significant difference in the ranking order F(3187) = 13.88, p < .001, η² = .18, with ethical responsibilities now placed first (M = 1.85, SD = .96) followed by legal (M = 2.29) and economic (M = 2.51, SD = 1.08) responsibilities a joint second, and philanthropy significantly lower than the other three (M = 3.21, SD = 1.11).

Differences between business and non-business respondents

In terms of the relative ranking of the four types of responsibility, a MANOVA indicated that: non-business respondents rank philanthropic responsibilities significantly more highly than business respondents: F(1, 398) = 6.26, p = .01, η² = .02. Non-business respondents rank ethical responsibilities marginally more highly than business respondents: F(1, 398) = 3.23, p = .07, η² = .01. There is no significant difference in how the business and non-business respondents rank legal and economic responsibilities.

Age

There were significant differences in the age make-up of the business and non-business samples, in particular the non-business sample had a greater proportion of those in the 18–25 age range (n = 100) than the business sample (n = 36) presumably due to the high proportion of students in that sample. Also the business sample had a higher proportion of those in the 46–55 age range (n = 51) than the non-business sample (n = 29). These differences were significant: chi-square = 33.49, df = 5, p < .001. Although Age did correlate with rating legal responsibilities as important (r = 17, p = .001), there were no significant main or interaction effects of Age on the results and including Age as a covariate in the MANCOVA and in the repeated measures ANOVA did not affect the results in any substantive way.

Gender

The business sample had a significantly greater number of male respondents (n = 116) than female respondents (n = 54) while the public sample had more female respondents (n = 122) than male respondents (n = 103): chi-square = 19.775, df = 1, p < .001.

In terms of the relative ranking of the four types of responsibility, a MANOVA indicated that: female respondents rank philanthropic responsibilities significantly more highly than male respondents: F(1, 394) = 8.50, p < .001, η² = .02. Female respondents rank ethical responsibilities marginally more highly than male respondents: F(1, 394) = 3.60, p = .06, η² = .01. There is no significant Gender difference in the ranking of legal and economic responsibilities (Table 4.).

Controlling for Gender did not substantially affect the results. There were no significant Gender X Business/Non-business interactions when Gender was included as a fixed factor in the MANOVA comparing business and non-business ratings of importance of the responsibilities or ranking of responsibilities.

Nationality

A MANOVA indicated that there were some significant differences in how the different nationalities ranked the four types of responsibility F (12, 1176) = 4.02, p < .001 (Table 5). In particular, UK respondents rank legal responsibilities more highly than other respondents, US respondents rank ethical responsibilities more highly than respondents from other countries, whereas Indian respondents rank philanthropy more highly than UK respondents. Despite these differences it is clear from Table 5 that for all nationalities, responsibilities are rated in importance in a very similar order: legal, ethical, economic and philanthropic.

Nationality X Business/Non-Business

There were no significant Nationality X Business/Non-business interactions when Nationality was included as a fixed factor in the MANOVA comparing business and non-business ratings of importance of the responsibilities or ranking of responsibilities.