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

Abstract

As one of the most influential models of corporate social responsibility (CSR), Carroll’s pyramid of CSR has both reflected, and helped to perpetuate, a business-centric notion of CSR which implies that economic responsibilities take precedence over legal and ethical responsibilities. This paper argues that this conception of CSR needs updating to reflect the increased power of business in society. An empirical survey of 400 respondents from both business and non-business backgrounds indicated that current conceptions of the relative importance of business responsibilities differ greatly from Carroll’s rankings. Based on the empirical research and conceptual arguments, a revised CSR pyramid is proposed with responsibilities ranked as follows: ethical, legal, economic and philanthropic.

Due to the inability and/or reluctance of nation states to impose more stringent regulation upon businesses, many of which are multinational corporations (MNCs) who operate across national boundaries, the burden of ensuring more responsible business is borne increasingly by the construct of corporate social responsibility (CSR) (Scherer & Smid 2000). This recognition of the importance of CSR has led to a proliferation of academic and business outputs centred on CSR. However the power of the CSR construct to affect business practices and provide a moral counterweight to economic forces that may tempt businesses to irresponsible behaviour depends naturally upon how the construct is interpreted and applied.

In 1979 Carroll built upon the previous scholarship to define CSR as a construct that “encompasses the economic, legal, ethical, and discretionary expectations that society has of organizations at a given point in time” ((Carroll 1979) p. 500). In 1991 Carroll illustrated these four responsibilities in a pyramid, which ranks business responsibilities in order of relative importance, with economic responsibilities assumed as primary, followed by legal, ethical and finally philanthropic responsibilities. Despite numerous re-imaginings of the CSR concept by Carroll himself and by others (Carroll 2008), Carroll’s pyramid of CSR remains the best known model of CSR, present in almost all student textbooks relating to CSR and one of the most cited models in the CSR literature (Ma et al. 2012). Schwartz and Carroll (2003) list the numerous research and educational texts that have represented Carroll’s CSR pyramid and conclude that “such use suggests that Carroll’s CSR domains and pyramid framework remain a leading paradigm of CSR in the social issues in management field” (p.504). As such, this pyramid of CSR has both reflected, and also helped to determine, the way in which the construct of CSR is understood and applied.

In this paper it is argued that the changing role of business in society and the greater role and power of business relative to government in the 21^st century necessitates an updating and reviewing of the priorities suggested by Carroll’s pyramid of CSR. For example, foreign direct investment as a percentage of GDP has increased four-fold since Carroll originally proposed his pyramid of CSR (UNCTAD 2015). Many scholars are increasingly uncomfortable with the increasing power and influence of MNCs over governments (Baden 2015; Crouch 2011; Keys et al. 2013) and the capture of other institutions such as NGOs by business interests (Fortin & Jolly 2015; Utting 2000). Given the increased influence and power of business, for example over the use of limited resources, the environment, working conditions and government policy, the level to which the CSR construct can encourage and enable a socially responsible mind-set in business practitioners and business students is of vital importance to society. Therefore it is of concern that Carroll’s pyramid of CSR clearly implies that profit comes before legal and ethical responsibilities.

This paper considers the implications of the primacy placed upon economic responsibilities, particularly in the context of business education and the direction of CSR discourse and research. The paper goes on to present a study that empirically tests the ordering of responsibilities implied by Carroll’s pyramid of CSR to see if they are reflective of real-life current views of business responsibility. It is concluded, based on both moral arguments and empirical research, that Carroll’s pyramid of CSR now misrepresents the relative priorities of business responsibilities both as they ought to be, and as they are currently perceived by both business and non-business respondents.

Business education

There has been growing recognition that business education plays a key role in setting the norms, priorities and expectations of future business leaders. A survey on CEOs’ views relating to sustainability based on responses of 766 CEOs from nearly 100 countries across a range of sectors revealed that developing educational systems to develop mind-sets conducive to sustainable and responsible business was considered one of the top changes necessary to embed sustainability (Ashridge 2010).

However the observation that the culprits of scandals such as Enron were predominantly recruited from top US business schools focussed attention on the potential of business education to foster an unethical or amoral mind-set. In an open letter to the Deans and the faculties of American business schools, Mitroff admits: “we are guilty of having provided an environment where the Enrons and Andersens of the world could take root and flourish… we delude ourselves seriously if we think we played no part” ((Mitroff 2004) p. 185). Numerous commentators have criticised business schools for failing to provide sufficient attention to the ethical aspects of business and for their unbalanced focus on shareholder value (Kochan 2002; Leavitt 1989; Mangan 2006; Porter & McKibbin 1988; Putnam 2000; Starkey et al. 2004). Others have argued that the profits-first mind-set is explicit or implicit in all management models, and creates descriptive norms of businesses as typically amoral, which can become self-fulfilling (Ferraro et al. 2005; Ghoshal 2005).

The response has typically been to advocate the integration of issues of ethics and CSR across the business school curriculum (see www.UNPRME.org). However a recent review of the progress made in integrating CSR issues into the business school curriculum concludes that although progress has been made, changes are predominantly cosmetic (Baden & Higgs 2015). This is due in part to the fact that the gatekeepers of progress are the lecturers themselves (Ryan & Tilbury 2013) and there is a wide diversity in levels of interest and/or understanding of how to integrate CSR (Dean & Beggs 2006; McDonald 2004). For example, university teaching staff typically have a high degree of autonomy in terms of deciding course content and often assume that others will cover ethical issues (Cant & Kulik 2009). Full embeddedness of CSR across the curriculum is also hampered by the poor coverage of ethics and CSR in most business textbooks (Acevedo 2013; Baetz & Sharp 2004). As a result, the most common means to address CSR issues within business schools has been the inclusion of stand-alone CSR-related modules (Christensen et al. 2007; Matten & Moon 2004; Moon & Orlitzky 2011). This means that the CSR module often bears the sole responsibility for the task of instilling a socially responsible mind-set in business students.

It is therefore worth considering more closely one of the most prevalent models being used in CSR scholarship and education – Carroll’s pyramid of CSR (Carroll 1991); examining what assumptions it is based on and what assumptions it may give rise to about the role of business in society. Despite the visual depiction of legal responsibilities as secondary, Carroll explains that the two layers of economic and legal responsibilities should be seen as coexisting based on the precepts of the free enterprise system. Carroll also emphasizes that ethical and legal responsibilities are inter-related, as ethical issues are often a driving force behind the creation of laws and regulations. Lastly there are philanthropic responsibilities, which Carroll sees as discretionary – appreciated but not required. However it could be argued that Carroll’s additional proviso that economic and legal responsibilities are jointly obligatory is unlikely to prevail against the more commonly seen visual depiction of economic responsibilities as coming before legal responsibilities. Thus Carroll’s pyramid of CSR appears to perpetuate rather than redress the unbalanced focus on shareholder value that the business schools have been attempting to remedy through the inclusion of a CSR module.

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

The relative ordering of priorities of economic before legal and ethical responsibilities in Carroll’s pyramid of CSR also has implications for how CSR is perceived and justified. Many business scholars have been critical of the unbalanced focus on profit that pervades the CSR discourse, with CSR generally being justified in terms of leading to increased corporate financial performance (CFP) rather than as an obligation in itself. Blowfield (2005) argues that the domination of CSR discourse by economic values leads to discussions about social and environmental justice being based upon economic rather than moral values. He goes on to show how this can result in adverse consequences, particularly in the context of developing nations, where although the current conceptions of CSR may offer basic worker rights, they do not challenge the rights of companies to close down factories and disinvest from host countries without compensation. Justifying CSR on economic terms limits the ability of the concept to work for social and environmental justice, and can even be counter-productive, particularly in cases where voluntary CSR is being promoted as an alternative to effective regulation (Blowfield 2005; Frynas 2005; Reich 2008). Similarly many commentators agree that justifying CSR in economic terms allows corporations to cherry pick CSR projects that meet the business case, for example those that offer good marketing opportunities (e.g. cause-related marketing) or environmental projects that cut energy and material costs, while ignoring more pressing business impacts that cannot be justified by reference to the business case (Bondy et al. 2012; Burchell & Cook 2008; Parkes et al. 2010). Baden and Harwood (2013) draw attention to the consequences of the prioritization of economic responsibilities:

“If key negative impacts of business practices (e.g. sweatshop labour in the retail industry, destruction of communities as a result of the activities of extraction industries, unsustainable consumption encouraged by the marketing sector) are not addressed, as the costs of being genuinely responsible for their social and environmental impacts far outweigh any gains, then this has an on-going detrimental effect on entire communities, and our planet as a whole” (p. 617).

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.

It must also be appreciated that Carroll’s placing of economic responsibilities as primary did not happen in a vacuum and reflects the tenets of traditional economic theory based on the mechanisms of Adam Smith’s ‘invisible hand’, which posits that society benefits when business serves its self-interest:

“Every individual neither intends to promote the public interest, nor knows how much he is promoting it… By directing [his] industry in such a manner as its produce may be of greatest value, he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was not part of his intention. Nor is it always the worse for society more effectually than when he really intends to promote it” (Smith 1776).

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.

Methodology

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?’

Sample

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.

Total sample and demographics

Cohort1 includes 241 participants, four were removed due to suspect reliability leaving 237 participants 58 % male and 42 % female respondents. Seventy per cent of this cohort were British, with the rest predominantly from Europe. Cohort 2 (MTurk sample) includes a total of 171 participants, eight were removed due to suspect reliability (took an unfeasibly short time) leaving 163 participants with an even gender split (51 % male and 49 % female respondents). The MTurk cohort was split almost equally between US (N = 75) and Indian respondents (n = 79) and Other (n = 9). The Sample accessed by the students contained an even split of business (n = 124) and non-business (n = 113) respondents, whereas the MTurk sample had more non-business (n = 112) than business (n = 51) respondents.

The two studies were initially analysed separately, but as the results were very similar, for this paper they have been analysed as a whole and broken down by gender, nationality and age rather than by cohort. Thus the total sample is 400 respondents, 175 of these were from business and 225 classified themselves as non-business.

The business sample represented the full range of sectors, business types and levels. Firm sizes ranged from micro with < 10 employees (31 %), small < 50 employees (14 %), medium < 250 employees (24 %), and large > 250 employees (31 %). Nine per cent of business respondents classed themselves as high level leaders (CEO, director, chairman), 13 % as senior managers, 19 % as middle-managers, 12 % as low-level managers, 16 % as employees and 25 % as owner-managers and 6 % as self-employed.

Just over a third of the non-business sample comprised of students, and the rest were a mix of housewives, unemployed/retired, and those employed in the public sector such as health and education, and a variety of occupations.

There were 219 males and 176 females. Thirty four per cent of the sample were in the 18–25 age range, 24 % in the 26–35 age range, 11 % in the 36–45 age range, 20 % in the 46–55 age range and 10 % over 56. Forty one per cent of the sample were British, 22 % were Indian, 19 % were from the US and the remaining 18 % were classified as other, and included a mix of Cypriots, Austrians, Spanish, Italian, German, Swedish, Chinese, Bulgarian, Turkish and Russian respondents.

Details of demographic differences between the non-business and business sample are presented later in the demographic interactions section.

Results

Mean rating of importance of responsibilities

As indicated in Table 1, legal responsibilities are rated as being the most important, followed by ethical, then economic and philanthropic responsibilities bringing up the rear. Maunchly’s test indicated that the assumption of sphericity had been violated χ ² (5) = 174.27, p > .001, therefore the Greenhouse-Geisser corrected tests are reported (ε = .78). A repeated measures ANOVA indicates that ratings for each are significantly different F (2.35, 930.46) = 181.85, p = <.001, η² = .31. Post hoc contrasts showed that all levels differ significantly from each other although ethical and economic responsibilities differ marginally (p = .07).

Table 1 Averaged scores for philanthropic, ethical, legal and economic responsibilities (High = most important)

Ranking of responsibilities

The rankings confirm the results from the specific questions (Table 2). Maunchly’s test indicated that the assumption of sphericity had been violated χ ² (5) = 27.87, p > .001, therefore the Greenhouse-Geisser corrected tests are reported (ε = .96). A repeated measures ANOVA indicates that rankings for each are significantly different F (2.87, 1137.12) = 106.41, p = <.001, η² = .21. Post hoc contrasts showed that all levels differ from each other with the exception of ethical and economic responsibilities which are not significantly different.

Table 2 Relative ranking of philanthropic, ethical, legal and economic responsibilities (Low – highest rank)

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

As shown in Table 3, a MANOVA with the independent variable being respondent type with two levels (business/non-business) and the dependent variable being the mean ratings of importance for responsibility type, indicated that non-business respondents rate philanthropic responsibilities as significantly more important than business respondents: F(1, 397) = 6.59, p = .01, η² = .02. Non-business respondents also rate economic responsibilities as marginally less important than business respondents: F(1, 394) = 2.86, p = .09, η² = .01. There is no significant difference in how the business and non-business respondents rate legal and ethical responsibilities.

Table 3 Difference between Business and non-Business Sample in ratings and rankings of responsibilities

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.

Demographic interactions

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.

As shown in Table 4. a MANOVA with the independent variable being Gender and the dependent variable being the mean ratings of importance for responsibility type, indicated that that female respondents rated philanthropy as significantly more important than male respondents: F(1, 393) = 14.67, p < .001, η² = .04. Female respondents also rate ethical responsibilities as more important than male respondents: F(1, 393) = 3.96, p < .05, η² = .01. There is no significant difference in how the male and female respondents rate legal and economic responsibilities.

Table 4 Difference between Male and Female respondents in ratings and rankings of responsibilities

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 rated the importance of the four types of responsibility F (12, 1173) = 7.12, p < .001 (Table 5). In particular, UK respondents rate legal responsibilities as significantly more important than other respondents and Indian respondents rank philanthropy as significantly more important than other respondents.

Table 5 Difference between nationalities in ratings and rankings of responsibilities

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.

Discussion of empirical results

The overall picture from the empirical data is that there is clear agreement that legal responsibilities come before economic responsibilities. No sub-section of the sample thought that economic responsibilities came first (although for Indian respondents economic responsibilities tied with legal responsibilities). Legal responsibilities came first overall and joint first with ethical responsibilities for US respondents and female respondents. However those who believed the legal system was inadequate to protect public welfare (n = 64) ranked ethical responsibilities first and significantly higher than legal responsibilities, which is consistent with the proposal that ethical responsibilities will be deemed to be more important than legal responsibilities when the legal system is not trusted to protect public welfare.

Overall, ethical and economic responsibilities receive the joint second highest ratings of importance, after legal responsibilities, with female, non-business, UK and US respondents tending to place ethical responsibilities higher than economic responsibilities and male, business and Indian respondents tending to rank ethical and economic responsilities joint second. For all respondent types, philanthropy is rated as the least important. Female respondents, non-business respondents and Indian respondents rated philanthropy higher than male, business and other nationalities respectively. However these sub-samples still all clearly rated philanthropy as significantly less important than all other responsibilities and it was always ranked last.

So with both means of measuring the relative importance of responsibilities, the overall pyramid from a diverse sample that includes business and non-business respondents differs from Carroll’s (1991) pyramid – the first priority are legal responsibilities, with economic and ethical responsibilities tying for second place, the only area of clear agreement is that philanthropic responsibilities are least important.

These results present a very different priority of responsibilities from the one depicted in Carroll’s pyramid of CSR and from the ordering of results found in research conducted in the US in the 1980s and 90s, which suggested that economic responsibilities were considered to be primary (Aupperle 1984; Edmondson & Carroll 1999; Pinkston & Carroll 1996). However these results are consistent with later research conducted which suggests that ethical responsibilities are more primary (Pedersen 2010) indicating a change in attitudes over time.

Integration of conceptual arguments with empirical results

It was argued earlier that the models used in CSR scholarship influence how CSR is conceived by both the academic and business community. Attention has been drawn to the negative consequences of a profit before ethics mind-set, and to the growing criticisms of the way in which the potential moral power of the construct of CSR can be undermined by the dominant discourse of economic rationality. Thus any model of CSR needs to be a morally strong and consistent counterweight to the dominant paradigm of profits-first which is present in the typical business school curriculum - a curriculum which has been extensively criticised on the basis of its over-emphasis of shareholder value at the expense of societal and environmental welfare. The CSR construct also permeates the business discourse beyond education and similarly needs to provide a moral counterweight to the competitive environment that can present numerous temptations to put aside ethical and legal obligations when the economic arguments are strong enough. I would thus argue, in common with others (Kang & Wood 1995; Sachs & Ruehle 2009), that on conceptual grounds alone, both legal and ethical responsibilities should be placed before economic responsibilities. The empirical results from this study indicate that this argument would be consistent with the currently held views of both business and non-business respondents.

Furthermore, in the case of the relative prioritisation of ethical and legal responsibilities, it needs to be considered to what extent legal responsibilities provide adequate protection for human rights and environmental sustainability. Indeed, in countries where laws are not effectively enforced, adhering to the law could be seen more as a voluntary activity than legal compliance. There are also issues of laws lagging behind technology or in some cases reflecting the interests of the dominant class, rather than the ‘public good’ (Hosmer 2006; Kurbjeweit 2011). Bearing in mind the power that MNCs have over the regulatory environment, both in terms of business lobbying, and also in their ability in some cases to site their operations in countries with lax regulatory frameworks, it would appear prudent to place ethical responsibilities above legal responsibilities.

It has also been argued that CSR is and needs to be a moral concept, and indeed Kohlberg’s influential theory of moral development (Kohlberg 1981) places legal compliance as two stages lower on the moral continuum than the ultimate stage of moral reasoning which is based upon universal ethical principles such as justice. The results indicate that this contention would be supported by those who believe the legal system is inadequate to protect public welfare.

One area where there is more agreement in terms of the relative ordering of business responsibilities both in the survey results and in the literature is in the ranking of philanthropy as the least important.

The resulting pyramid shifts the construct of CSR from a more instrumental/descriptive construct to a more normative model of business responsibility. This is relevant to a debate touched upon earlier about whether the CSR construct should be seen as an instrumental, descriptive or normative account of business responsibility. A sense-making perspective would see such questions as naive due to the fact that social science models such as CSR are essentially socially constructed (Nijhof & Jeurissen 2006). The ‘double hermeneutic’ (Giddens 1987) describes how our beliefs about our world become self-fulfilling, so in this case the description of business as prioritising economic responsibilities does not simply describe a reality – it also norms and justifies it, making it more ‘true’. The implication of the double hermeneutic effect is that describing business as prioritising legal/ethical responsibilities over economic responsibilities, would in itself (assuming at least a minimal level of credibility) help make that description a reality. Thus, a more useful question is whether portraying CSR in a normative light is more or less likely to lead to more socially responsible behaviour than portraying CSR in an instrumental light.

Some have taken a pragmatic view that only an instrumental approach will enable business to engage in CSR (Amaeshi & Adi 2007) so it makes sense to construct CSR in the language of business rationality and strip it of its normative component. In the context of business education too, it has been argued that presenting the business case for issues such as CSR and sustainability is likely to make it more palatable to business students (Lourenço 2013). However, such as approach risks diluting the moral power of the CSR construct to the point it becomes an ineffective driver for genuinely ethical business, and could even be counter-productive, particularly in cases where voluntary CSR is being promoted as an alternative to effective regulation. It would be of interest to explore in future research the implications of a more normative model of CSR for business and business education. In particular, the question of whether normative model of CSR can better enable a more ethical mind-set among business students/practitioners than an instrumental model of CSR.

Based on the arguments above, it is concluded that a more powerful and effective CSR pyramid would be ordered as illustrated in Fig. 1: ethical responsibilities, legal responsibilities, economic responsibilities and finally philanthropic responsibilities. In other words, in order for business to have the license to operate and to be trusted with the production and allocation of scarce natural resources and inherently valuable human resources, it needs to first and foremost accept its ethical responsibility to do no harm and conform to society’s ethical norms and expectations. Secondly business needs to be compliant with legal responsibilities, and only once it has accepted these responsibilities is it free to make a profit and pursue its economic objectives.

Fig. 1

Proposed amended pyramid of CSR

Limitations and validity

Issues of data integrity with reference to the use of student contacts for cohort 1 may have occurred, although procedures were taken to minimise these by presenting no specific hypothesis, and allowing students to not contribute results with no penalty if difficulties were encountered. Further, the finding that a separate cohort recruited through different means (MTurk) and accessing a different sample produced almost identical results provides reassurance relating to the integrity of the data. In addition, the question of the relative importance of the four types of responsibility was assessed in two ways: firstly by asking respondents to rate the importance of a series of five examples of each type of business responsibility, which was then averaged to give an overall score of importance for that responsibility, and secondly by a simple ranking of the four types of responsibility. As both forms of approaching the question gave rise to similar results, this provides further reassurance with respect to data validity and integrity.

A further issue is that the non-business sample had a greater proportion of female respondents, which may account for some of the differences between business and non-business respondents. However it could be argued that this gender bias is reflective of a genuine gender bias in society which sees women more highly represented in non-business occupations such as health and education, than in more commercial occupations. Furthermore, including gender as an additional factor did not significantly affect the overall results. Similar issues arise with a lower age profile in the non-business sample, and again it can be argued that business respondents who are high enough up in the organisation to speak from a business perceptive are more likely to be older on average than respondents from the non-business sample. In addition, age did not show any interaction effect with respondent type when included as a covariate in the analysis.

The sample included respondents predominantly from UK, US and India. Further research could focus on cross cultural differences drawing upon a wider range of countries.

Conclusion

It is considered vital by most commentators that business managers adopt more socially responsible mind-sets and behaviours. CSR as a construct reflects this goal and thus needs to provide a moral counter-balance to the dominance of economic values pervading the rest of the management discourse and business school curriculum. Yet Carroll’s pyramid of CSR suggests economic values come before ethical and even before legal values. It has been argued that this focus has permeated the literature and the discourse surrounding CSR. It has been claimed that CSR has won the battle, but lost the war in the sense that although this business friendly model may have allowed CSR to have become accepted and even embraced by the business community, it does not allow ethical or legal arguments to take precedence over economic arguments. This has implications both in business education, where Carroll’s pyramid of CSR is one of the most cited models in CSR textbooks, and also in CSR scholarship, leading to an unhelpful focus on the business case for CSR, which it is argued, should be irrelevant on the basis that businesses should avoid illegal or unethical practices whether or not it financially benefits them to do so. It is concluded that the original pyramid of CSR needs updating to reflect the increased power of business relative to government in the 21^st century. The amended pyramid of CSR places ethical responsibilities as primary, followed by legal responsibilities economic responsibilities and lastly philanthropy.

Appendices

Appendix 1

Questionnaire (* = business respondents only)

Demographic Questions

• 1. Gender: Male/female

• 2. Age - circle one: 18-25; 26-35, 36-45, 46- 55, 56 – 65, 66+

• 3. * What sector is your business in? (free response)

• 4. * What size of company?

  • a. Micro < 10 employees,

  • b. Small < 50 employees

  • c. Medium < 250 employees

  • d. Large > 250 employees

• 5. What is your job role/title? (free response)

• 6. * What level employee are you?

  • a. Owner manager

  • b. Self-employed

  • c. Employee

  • d. Low-level manager/supervisor

  • e. Middle Manager

  • f. Senior level manager/Executive

  • g. High level leader – Chairman/CEO/Director

• 7. Nationality (free response)

Main Questionnaire

• 1. How important a business responsibility do you think the following philanthropic activities are? (ranging from “1” not at all important to “6” very important)

  • A. Allocating a percentage of profits towards charitable activities and/or benefitting the community

  • B. Managers and/or employees participation in voluntary and charitable activities

  • C. Sponsoring the arts/sports/community events

  • D. Provide free services/products to needy people

  • E. Assist voluntarily projects that enhance a community’s quality of life

• 2. How important a business responsibility do you think the following ethical practices are? (ranging from “1” not at all important to “6” very important)

  • A. Making sure that your business practices do not have a negative impact of society or the environment

  • B. Employ environmentally friendly technology and practices e.g. waste/recycling policies, reduce carbon footprint

  • C. Ensure a good working environment for your employees, e.g. profit-sharing, good working conditions, training

  • D. Fair trading and marketing policies, dealing honestly with customers and suppliers

  • E. Prevent ethical norms from being compromised in order to reach business objectives

• 3. How important a business responsibility do you think the following legal practices are? (ranging from “1” not at all important to “6” very important)

  • A. To comply with all environmental laws and regulations

  • B. Provide goods/services that meet minimal legal requirements

  • C. Ensure working conditions meet health and safety standards

  • D. Pay taxes according to the law

  • E. To be law-abiding

• 4. How important a business responsibility do you think the following economic practices are? (ranging from “1” not at all important to “6” very important)

  • A. Perform in a manner consistent with maximising earnings per share or profit for the owners

  • B. Maintain a strong competitive position

  • C. Maintain a high level of operating efficiency

  • D. To be committed to being consistently profitable

  • E. To maximise sales and revenue

• 5. To what extent are the social and environmental impacts of business covered by laws? Are there sufficient laws that are adequately enforced to protect public welfare?

  1. Insufficient laws/weak legal/regulatory framework – 6 Sufficient laws/strong legal/regulatory framework (circle)

  1 2 3 4 5 6

• 6. Please rank in order of importance which responsibilities you believe are most important for business (1 for most important, 4 for least important)

Philanthropic responsibilities …….

Ethical responsibilities …….

Legal responsibilities …….

Economic responsibilities …….

Appendix 2

Instructions to Students

Over the vacation we are all going to engage in some exploratory research to determine whether Carroll’s pyramid of CSR accurately represents current views of business responsibility. It will also be of interest to see if there any differences in the public’s view and the views of business regarding what are the most important responsibilities of business. There are no right or wrong answers, we may find that Carroll’s model is out of date, or his proposed ordering on the pyramid may still hold.

Instructions

There are two questionnaires, one to be completed by a business manager and one by a member of the public. Each student should get two respondents from business and two respondents from the public (feel free to get more if you can).

Ethics

Each respondent should complete a participant consent form that specifies what the research is for and that all results are anonymous and will be presented only as a whole, no responses will be attributable to specific individuals. Overall results may be written up and published.

Once you have the completed questionnaires – transfer the results to isurvey using the link below, and hand in the paper copy and participant consent form in the next lecture. Be aware that the order on the paper questionnaires is randomised and so questions will be in a different order online.

Further instructions regarding sample selection and interview technique

Overall I hope that as a group we will be able to access a fairly representative sample in terms of gender, age, sector type, size of company etc.

Business managers: choose someone who is high enough up in the business that they will be reflecting a business perspective – in other words, avoid low-level employees who may not consider the business interests to be the same as their own. It can be difficult to gain access to a suitable sample, so make the most of any contacts you have, including family, but also if you think their responses may be compromised by the fact they know you, they can always complete it directly using the online link. But in general make it clear there are no right or wrong answers, you just want their honest view.

Members of the public: again it’s fine to use people you know as long as you don’t feel their responses will be biased due to familiarity. Make it clear that you are asking for their view as a member of the public – not what they think business feels – so the question is what they think is most important for business, not what they think business thinks. Try to avoid any overlap with business people so choose people who are not business managers e.g. public sector workers, students (but not business/management), housewives, house husbands, low-level employees.

In both cases the survey starts with a question that asks if you think there was a strong social desirability bias present. Write yes if you have any reason to believe they were not giving their honest opinion, if you tick ‘yes’, then your results will not be included in the survey. But don’t make assumptions, e.g. those who are cynical about business may disbelieve any comments from business about ethics, even if they are in fact sincere. Conversely, you may bring your own assumptions about members of your family and disbelieve any comments that don’t fit your preconceptions.