What Do Business Executives Think About Distributive Justice?

Abstract

While there exist extensive literatures on both distributive justice and senior executive pay, and a number of authors (notably the French economist Thomas Piketty) have addressed the implications of high pay for distributive justice, the existing literature fails to address what senior executives themselves think about distributive justice and whether they consider high income inequalities to be morally acceptable. We address this gap by analysing a unique dataset comprising the views of over 1000 senior executives from across the world, which was constructed using a survey instrument designed by the authors based on a thought experiment resembling John Rawls’s original position. We report four main findings. First, executives conceptualise distributive justice in a pluralistic manner, endorsing different and sometimes apparently conflicting philosophical principles: to explain how this plurality can be accounted for we propose a novel field-theory framework for conceptualising beliefs about distributive justice. Second, executives support similar philosophical approaches at both society and company levels of analysis, thus countering the idea that companies should leave matters of distributive justice exclusively for governments to deal with via the tax system. Third, executives believe that they live in societies and work for companies that fall short of desirable distributive justice outcomes. Fourth, the distributive justice views of the executives in our sample fall into four distinct clusters that are correlated with certain socio-demographic markers. Finally, we note the distinction between distributive justice beliefs and behaviours, from which we derive a number of managerial and public policy implications.

Introduction

Over the past two decades, the compensation received by those employed at the top of companies (“supermanagers” according to Piketty 2014) has risen dramatically, continuing a trend that began in the 1980s. At the same time, the incomes of regular employees have more or less stagnated in real terms (Willman and Pepper 2020). In combination, these two developments have fuelled the now widely held belief that our societies, and the companies operating within them, have become increasingly unjust (Atkinson 2015).

Within the economics literature, there is an ongoing discussion as to whether the dramatic rise in wage inequality can be explained solely by economic factors: for recent reviews, see Lemieux et al. (2009) for the United States and Bell and Van Reenen (2013) for the UK. From a business ethics perspective, most contributions have consequently focused on whether the explanations proposed by economists might render large income inequalities morally justified or, more precisely, distributively just. Boatright (2009) and, more cautiously, Moriarty (2009) answer this question in the affirmative; Harris (2009) and Néron (2015) disagree. What is missing from the literature is an empirical investigation into what business executives think about distributive justice, and, by implication, whether they consider high income inequalities to be morally acceptable.¹ This article aims to fill that gap.

Establishing what senior executives think about distributive justice matters for at least two reasons. First, senior executives play a key role in shaping organisational pay policies (Sud and VanSandt 2011) as well as other policies that are of interest from the point of view of distributive justice. Second, their opinions can significantly affect how other employees think about distributional issues (see Schein 1985|2010). The views of senior executives thus affect both the pay policies within their companies and how these policies are evaluated by others within the firm.

In this article, we present the main findings from a research project designed to capture executives’ perspectives on distributive justice by matching their views with key philosophical principles. There is an extensive philosophical literature which aims to establish what principles a society would have to adopt in the distribution of its economic rewards to be recognisable as distributively just (for a review, see Konow 2000). We draw upon this literature in order to investigate what executives think about distributive justice in a philosophically informed manner. Specifically, we surveyed over 1000 respondents who work in high status leadership positions across the world to collect their views about what principles would lead to a fair distribution of income. The main objective of our research was to develop a thorough understanding of how senior executive conceptualise distributive justice. We sought, in particular, to explore the following sets of questions. (1) What philosophical principles figure most prominently in business executives’ conceptualisations of distributive justice? Are the views of most executives based on a single principle, or do executives try to accommodate several principles at once? (2) Is distributive justice seen differently at the level of the company as opposed to the level of society? Do executives conceive of distributive justice as primarily an organisational issue, primarily a societal issue, or do they see it as an issue that needs addressing at both levels? (3) Do executives believe that the principles that they endorse are enacted in the societies in which they live and in the companies for which they work? How large is any perceived justice deficit? (4) Do conceptualisations of distributive justice vary according to participants’ demographic characteristics, such as their age, gender, income level, or nationality?

In response to (1), we find that business executives conceptualise distributive justice in a pluralistic manner, endorsing different and sometimes apparently conflicting philosophical principles. To explain how this plurality can be accounted for, we propose a novel field-theory framework for conceptualising pluralistic distributive justice beliefs. With regard to (2), we find that business executives support similar philosophical approaches at both society and company levels of analysis, thus countering the idea that companies should leave matters of distributive justice exclusively for governments to deal with via the tax system. With regard to (3), we find that executives believe that they live in societies and work for companies that fall considerably short of desirable distributive justice outcomes. Taken together, these findings indicate that senior executives do not in general believe that the currently prevailing high pay inequalities are just and might therefore be expected to endorse smartly designed reforms aimed at lowering inequality. In response to (4), we find that, while executives’ beliefs about distributive justice are relatively homogenous across our sample, their views nevertheless fall into four distinct clusters that are correlated with a number of socio-demographic markers.

The remainder of this article is structured as follows. In the second section, we discuss the normative principles of distributive justice upon which our research is based. In the third section, we discuss our methodology. In the fourth section, we present our findings, which we discuss in the fifth section. The sixth section summarises and concludes.

Principles of Distributive Justice

The philosophical literature on distributive justice is vast. In designing our survey, our goal was to introduce business executives to a range of influential philosophical theories about distributive justice that also have intuitive appeal, as evidenced by empirical studies from economics, political science, and sociology. We included six principles in our questionnaire. The first and second principles emphasise individual merit as the proper basis for allocating economic rewards. The third and fourth concentrate on the significance of needs. The fifth and sixth principles stress the importance of voluntary transactions and economic efficiency. We focused on the justice of earnings from employment as opposed to the just distribution of wealth or other goods, as earnings from employment are a variable that senior executives can influence through the design of organisational pay policies. In what follows, we lay out the philosophical thinking behind the principles and the empirical support that they receive. For convenience, the six principles are summarised in Table 1.

Table 1 Six principles of distributive justice

First Principle: Desert

The first principle takes as its starting point the empirical insight that in the context of production individuals typically believe that effort and contribution should be rewarded. Under experimental conditions, when distributive stakes are made available not as “manna from heaven” (Cappelen et al. 2007, p. 818), but as the result of participants’ productive activities, the vast majority of people choose to reward effort and contribution by allocating rewards that reflect the differential productive inputs of different individuals (see, for example, Konow 2000). Consistent with these findings, Miller (1992) notes in an early survey of the sociological and social psychology literature that most people endorse inequalities in income if they reflect differential effort and contribution.

While the view that some people deserve to earn more than others is widely accepted by the general public, the position taken by philosophers tends to be more sceptical and nuanced. For example, Scheffler (2003) argues that desert is an individualistic concept—what a person deserves depends only on facts about him or her; however, distributive justice is really a holistic concept—the justice of a given person’s share of resources turns on facts about that person as well as other people. Rawls (1971|1999) and Dworkin (2000) both contend that ability and effort cannot be used to determine what different individuals deserve, as inherited talent and individual characteristics are morally arbitrary. Less sceptically, Sher (1987) and Lamont (1994) provide arguments to support the normative force of desert claims.

Second Principle: Equality of Opportunity

The second principle stipulates that a distributively just society is characterised by equality of opportunity. In such a society, equally talented and hardworking individuals have the same opportunities to be appointed to positions that come with economic advantages; no one is discriminated against because of their gender, ethnicity, socioeconomic background, or their membership in some other socially salient group. In the philosophical literature, equality of opportunity is widely recognised as important for distributive justice (see, for example, Mason 2006). Equality of opportunity is typically considered as a luck egalitarian principle of justice. Luck egalitarian theories are founded on the thought that differences in individual levels of welfare are distributively just only if they are traceable to personal choices for which people can be held morally responsible (Cohen 1989; Dworkin 2000).

Empirical studies show that many individuals try to control for luck in the distribution of economic advantages. Cappelen et al. (2013a) investigate how agents allocate economic gains when an individual’s contribution depends on two factors: first, the amount of money that the individual chooses to invest, which is under their control; second, the rate of return on his or her investment, which is not under their control. They find that only around 25% of participants reward luck by allocating a higher share of the overall economic gains to individuals whose investments enjoyed a higher rate of return. The other 75% of participants disregard the effects of differential rates of return when making their allocative decisions. In a related study, Cappelen et al. (2013a) find that many subjects choose to redistribute economic gains from better-off to less well-off participants when the differences in these participants’ economic welfare are due to luck, but not when the differences stem from differential choices.

Third Principle: Sufficiency

The principle of sufficiency demands that all members of a society should have an income that is high enough to lead a dignified life. What this means might vary from society to society. At a minimum, to lead a dignified life is to have one’s basic needs met and to be able to relate to other members of society on equal terms. In the philosophical literature, defenders of sufficientarian and related relational egalitarian theories of distributive justice include Anderson (1999), Frankfurt (1987), and Scheffler (2010).

Empirical studies confirm that many individuals take needs into account when they make distributive decisions or express their distributive preferences. In a series of empirical studies, Frohlich and Oppenheimer (1992) find that participants favour a principle for the society-wide distribution of income that maximises average income subject to a sufficientarian floor constraint. Participants overwhelmingly choose a sufficientarian distribution principle when they are asked which distributive scheme they would prefer to work under, and also when they are asked which scheme they deem fairest for society. In Frohlich and Oppenheimer’s experiments, participants were also asked to arrive at a consensus agreement in their group discussions, and collectively chose a sufficientarian principle 75% of the time. Similarly, Cappelen et al. (2013b) find that needs are taken into consideration in an experimental production task in which students from rich countries were paired with students from poor countries.

Fourth Principle: Maximin

Rawls (1971|1999) argues that a just society is characterised by institutions that promote equal opportunities and that redistribute income so as to make the worst-off members of society as well off as possible. Rawls arrives at this highly redistributive principle (which he calls the “difference principle” and we call “maximin”) by arguing that just principles are those that we would choose as in our self-interest when placed behind a “veil of ignorance”. Behind a veil of ignorance, we find ourselves in an “original position” where we do not know our identity—we are unaware of our socioeconomic status, our talents, our gender, and our ethnicity. Rawls argues that in such an original position, we would prefer institutions that make the worst-off members of our society as well off as possible. As people need to remain incentivised to work hard, maximin does not imply that everyone should end up with the same after-tax income.

Rawls relies on the assumption that we would be highly risk-averse when choosing institutions from behind a veil of ignorance. In the absence of such risk aversion, we might endorse different allocative principles (see, for example, Arrow 1973; Harsanyi 1975, who argue that a risk neutral individual would choose to maximise average welfare when placed behind a veil of ignorance).

While empirical studies confirm that individuals take needs into account when they make distributive decisions, the highly redistributive maximin principle finds only limited empirical support. Michelbach et al. (2003) identify behaviours consistent with a Rawlsian maximin principle in their experiments, but Frohlich and Oppenheimer (1992) observe that only 4% of their participants endorse a society-wide distributive principle according to which the worst-off members of a society should be made as well off as possible. Where participants had to reach a collective verdict, not one group settled for maximin. Similarly, in a more recent study, Bruner (2020) finds little support for the difference principle.

Fifth Principle: Efficiency²

The principle of efficiency proceeds from the assumption that in a market-based economy the prices paid for different types of labour should reflect the relative scarcity of that type of labour. This helps to allocate human capital efficiently, so that economic output is maximised. Efficiency is consistent with a utilitarian approach to distributive justice (Harsanyi 1975). From a utilitarian perspective, the right allocation of resources maximises general welfare. If resources are allocated efficiently, the monetary value of what is produced with those resources is maximised. Under certain idealised conditions, this plausibly enables the general welfare to be maximised also.

Among philosophers, Heath (2014) draws attention to the moral relevance of enabling the price mechanism to do its work. Like Hayek (1976) before him, Heath argues that market prices are neither just nor unjust, and that it is misguided to insist that companies should pay their employees a “just” income. Instead, wages should be made maximally free from distortions, thus reflecting the relative scarcity of labour. Heath’s focus on efficiency at the company level does not mean that he is opposed to redistribution at the level of society; indeed, Heath (2014, p. 10) accepts that redistributive policies are appropriate from a justice perspective if they help promote general welfare. From an empirical perspective, Miller (1992) observes that merit-based distributive principles are widely endorsed in an economic context not only because people believe that effort should be rewarded, but also because merit-based principles are thought to promote efficiency.

Sixth Principle: Entitlement

The sixth principle presumes that individuals are morally entitled to whatever income is voluntarily transferred to them. In the philosophical literature, this libertarian theory of entitlement is associated with Nozick (1974). He argues that, if there is a just starting point, then there can be nothing unjust about an unequal allocation resulting from a series of voluntary transfers. Moreover, if wealth that arises from voluntary transfers is appropriated by the state, then the state is unjustly depriving citizens of goods to which they are entitled.

Libertarian entitlement ideas receive some indirect empirical support. When the rules of experiments allow subjects to keep the total payoff to themselves, a significant proportion choose to hold on to the lion’s share of the payoff. In many experiments, up to a third of individuals act in a narrowly self-interested manner (see, for example Cappelen et al. 2007, p. 824; Frohlich and Oppenheimer 2004, p. 109; Krawczyk 2010, p. 136). On one interpretation of their behaviour, these individuals simply do not care about distributive justice, and maximise gains to themselves without feeling constrained by considerations of fairness. On another more libertarian interpretation, these individuals believe that they are morally entitled to keep everything to themselves because doing so is a permissible alternative that is offered to them under the rules of the experiment.

Methodology

Data Collection

The data for our study were collected with the help of a market research company and a professional services firm. Together they had access to a large panel of senior executives from around the world. The market research company was commissioned to collect data based on a list of pre-selection criteria (job title, earnings, industry sector, company size, nationality, and country of residence) with a view to ensuring that only “senior business executives” as defined for the purposes of the study were included in the sample.³ A panel-screener questionnaire was used to ensure that respondents who did not meet the pre-selection criteria were selected out. Quotas were agreed in advance to ensure that the sample was broadly representative of different ages, genders, industry sectors, and nationalities. The data were collected using an online survey during the period of October 2016 to January 2017, and then handed, anonymised, to the researchers in raw form for cleaning and analysis. Detailed demographic information about the sample is contained in Table 2. While data based on quota sampling cannot necessarily be guaranteed to be representative of the underlying population, there is some justification for using quota sampling given that high-level business executives are notoriously difficult to access (see Pettigrew 1992). Our sample is quite large (n = 1123), and well diversified, allowing for meaningful analysis.

Table 2 Demographics

Survey Design

To encourage executives to think impartially about questions of distributive justice, thus mitigating the effect of self-serving biases, we asked participants to engage in the following thought experiment before answering the questionnaire:

Suppose that from tomorrow morning onwards you find yourself in an imaginary society where the skills and talents that are relevant to economic prosperity are very different from the ones that you are currently familiar with. In this society you might, for example, have a successful career if you are able to carry heavy objects over large distances, or if you are prepared to perform monotonous tasks for hours at a time. Skills and knowledge may also be rewarded in this imaginary society – or they may not be – the point is that you don’t know when making your decisions.

This thought experiment was designed to resemble Rawls’s original position. Unlike Rawls, we did not ask respondents to imagine themselves as stripped of all knowledge about their particular identities. We asked them instead to imagine waking up into a strange world where their skills and talents might suddenly be useless. We believe that this second imaginary feat is easier to accomplish than the one that Rawls invites us to perform in his thought experiment. We preferred a Rawlsian “behind the veil” approach to a Smithian “impartial spectator” because we wanted to emphasise our respondents’ roles as subjects of the enquiry, rather than, for example, as policy advisers to a hypothetical state government.⁴

Our survey centred on the six principles of distributive justice described in the second section above. For each of the six principles, we provided a key claim and a rationale to introduce the principles to the survey participants. The principles, together with their key claims and rationales, are summarised in Table 1. In the questionnaire, the principles were assigned a random order and given neutral labels such as “Principle A” or “Principle B”. For each principle, we asked the respondents to what extent they agreed with each of the following three statements:

• (1) “I would want an imaginary society that I might have to find my place in to be governed in accordance with this principle”;

• (2) “A society governed by this principle would be a just society”;

• (3) “The society that I currently live in is governed by this principle”.

Participants were given the option to “strongly agree”, “agree”, “neither agree nor disagree”, “disagree” or “strongly disagree” with the different statements. In the second part of the questionnaire, we proceeded in a similar way, but asked participants to what extent they endorsed the different principles of justice as governing the distribution of income at company level. The order in which the different principles were presented was altered in the second part of the questionnaire to mitigate order effects. We also changed the terminology from “principles” to “statements” to make pairings between the first and second parts of the questionnaire less obvious.

In the third part of the questionnaire, we provided the following preamble to the questions:

Some commentators argue that ethical questions about the distribution of income are matters for governments to address, if at all, for example through the tax and welfare systems in society, and that corporations should only concern themselves with efficiency, profitability, and maximizing shareholder value. Other commentators argue that employees are also stakeholders in the corporation and that senior managers have to ensure that all organizational policies and processes, including those relating to pay and performance, are demonstrably fair. This section asks a number of questions about your views on these issues.

Participants were invited to respond to a number of propositions such as “corporations should take into account principles of fairness and distributive justice when determining their pay strategies”, and “questions of distributive justice should be settled exclusively at the level of society as a whole rather than also at the level of individual corporations”. Respondents were also given the option of providing narrative comments, which we translated, tabulated, and analysed by reference to the framework developed on the basis of our quantitative empirical work.

The final part of the questionnaire asked participants about their age, gender, nationality, main country of residence, indicative earnings from employment, the industry sector in which they work, the type of company they work for (i.e. public company, private company, or state-owned enterprise), their role within the company, and the size of company (in terms of turnover and number of employees).⁵

Use of a Survey Over a Lab-Based Experiment

Much empirical research on distributive justice is conducted “in the lab”—economists in particular tend to favour lab-based experiments over other methods of data collection. We chose a survey over a lab-based experiment for a number of reasons.⁶

The principal reason is a pragmatic one. Our objective was to reach a large international audience of high-level executives in order to study the distributive justice views of global business elites. It would have been extraordinarily difficult, if not impossible, to study such an audience in a behavioural laboratory. Thus, if lab-based studies are the only way of eliciting distributive justice views, our project would simply not have been feasible. In any case, while lab-based experiments have crucial advantages if the aim is to study behaviour, our objective was to study and conceptualise the distributive justice beliefs of senior executives, not to predict behaviour. Of course, for a perfectly virtuous individual, we might assume that there will be a close connection between behaviour and beliefs, so that their beliefs about distributive justice will guide their actions. However, while a less than perfectly virtuous individual might wholeheartedly believe that everyone should have an income sufficiently high to live a dignified life, they might at the same time struggle to give away part of their income to raise the standard of those finding it difficult to make ends meet. This potential gap between beliefs and behaviours gives rise to one of the practical implications of our study—distributive justice outcomes in society and organisations might be modified if the gap between beliefs and behaviours could be closed. We discuss this further below under the sub-heading “managerial and public policy implications”.

All this is not to say that experiments have no role to play in studying individuals’ justice views. Consider the fact that our survey uses a thought experiment resembling John Rawls’s original position to elicit distributive justice views. According to Rawls’s experiment, subjects are asked to imaginarily place themselves in an “original position” where they do not know their social position—they are unaware of their skills, talents, family background, ethnicity, religious views, etc. Stripped of this knowledge, they are then asked to choose those rules for the distribution of resources that they would consider in their self-interest. Rawls thinks that the rules that we choose as in our self-interest when placed in an original position are rules of justice. In our survey, we merely invite participants to engage in such a thought experiment. In the lab, by contrast, it might be easier to place subjects in an original position—it can be explained to them in more detail that their identity is veiled, that it will be revealed after they choose some principle for the distribution of resources, and that the reward that they will take home will vary in accordance with the principle that they have chosen as well as the identity that will ultimately be revealed (see, for example, Frohlich and Oppenheimer 1992). It might suggest that an experimental setup is preferable when a Rawlsian thought experiment is used. However, we do not believe that this is necessarily the case. In an experimental situation, individuals typically interact with strangers on a one-off (or limited repeat) basis and the sums at stake are normally trivial. But distributive justice is concerned with repeated interactions between the members of a group or society and the distribution of large sums that determine how individuals’ lives are going. It would be perfectly consistent in an experimental setup to express a “winner-takes-it-all” mentality while arguing for a safety net at the level of society.

In light of these considerations, we believe that a survey compares favourably with an experimental approach in discovering distributive justice beliefs. We included a small test to check whether our thought experiment worked. For each principle of justice, we asked the participants to what extent they endorsed it as just, and to what extent they preferred it as being in their interest. If the thought experiment worked, there should be no difference between the two answers. The answers of our participants were, indeed, highly correlated (r = 0.56 at society level of analysis and r = 0.84 at company level of analysis).

Findings

Our research is exploratory, meaning that we touch on many issues in order to investigate existing theories, to seek new hypotheses, and to see the overall shape of the phenomena (Bewley 1999). We carried out a battery of statistical tests looking for patterns in the data, which we then compared with narrative responses to the survey. We debated our results with practitioners and other academics in a variety of forums, and took account of these discussions in developing our conclusions. We report below the main results of our analysis.

Descriptive statistics are set out in Tables 3 and 4. Table 3 provides a detailed overview of the agreement and disagreement that the participants expressed. Table 4 sets out average levels of agreement by principle at society and company levels. We coded participants’ level of agreement as “ + 1” (strongly agree with a principle), “ + 0.5” (agree with the principle), “0” (neither agree nor disagree), “ − 0.5” (disagree), and “ − 1” (strongly disagree). Thus, the higher the average for a principle, the more agreement exists among the respondents that the principle expresses a distributive justice ideal.

Table 3 Descriptive statistics—preferred (normative) and perceived (actual) principles of justice

Table 4 Empirical results—preferred (normative) and perceived (actual) principles of justice

Preferred Principles of Justice at Society and at Company Level

The approval ratings for the different principles set out in rows (1) and (2) of Table 4 show that at society level the two merit-based principles of equality of opportunity and desert have the highest ratings. These are followed by sufficiency and efficiency. Less than half of all respondents endorse maximin or entitlement. At company level, as at society level, the two merit-based principles have the highest levels of approval, although the order is reversed. Sufficiency and efficiency again follow, while entitlement and maximin lag behind. In summary, while the picture at company level is similar to the picture at society level, equality of opportunity has a higher approval rating at society level, desert at company level, and sufficiency and efficiency are both endorsed more strongly by participants at company level. Row (3) in Table 4 reports the Pearson correlation coefficients between the endorsement of any given principle at society and at company level.

Perceived Justice Shortfall

Rows (4) and (5) in Table 4 provide corresponding ratings for perceived principles of justice. The gap between the extent to which principles of distributive justice are endorsed as ideals and the extent to which they are perceived to govern the actual distribution of income is evident by comparing the ratings in rows (1) with (4) and (2) with (5). We can see that the principles that are considered to be most morally relevant (i.e. desert, equality of opportunity, and sufficiency) are not considered to be strong governing principles in the societies in which the respondents live or the companies for which they work. In particular, while most respondents agree that a society governed by equality of opportunity or desert would be distributively just, less than half believe that the societies in which they live are governed by either of these principles. Conversely, for the principles that are generally considered to be less relevant (maximin and entitlement) the distribution of responses is much more similar in the normative versus actual cases at society level. Much the same pattern of results appears at company level.

Pluralism and Clustering

Table 5 shows the extent to which respondents’ conceptualisations of distributive justice are pluralistic. Specifically, we count how many principles are strongly or very strongly endorsed by a respondent. At both society and at company levels, there is pronounced pluralism, with respondents showing agreement with multiple principles.

Table 5 Pluralism in views about distributive justice

To understand the plurality in respondents’ views in more detail, we asked respondents to rank order their preferences and performed a cluster analysis of the ranked data. The forced ranking of preferences was highly correlated with the approval ratings (r = 0.93). The cluster analysis involved partitioning the observations with the aim of uncovering groups of homogeneous observations. To determine the clusters, we used a k-mean algorithm, which involves the iterative calculation of the mean of each cluster (Everitt et al. 2001). As each new observation is allocated to a cluster, the cluster’s mean is recomputed. Multiple passes are made through a dataset, allowing observations to change cluster membership based on their distance from the recomputed cluster mean. The algorithm stops when no observations change clusters. To determine the optimum number of clusters, we first ran the algorithm with multiple numbers of clusters (1 to 20) and, for each cluster solution, calculated the within-cluster sum of squares (WSS) and the corresponding proportional reduction of error (PRE). The plotting of WSS and PRE revealed a solution of four clusters. The clusters are presented in Table 6. As the cluster analysis is based on ranked data (1 = most preferred, 6 = least preferred principle), the lower the value corresponding to a particular principle, the higher the relative preference towards that principle.

Table 6 Cluster analysis

Cluster 1, a group of respondents we label as “relational egalitarians”, is characterised by a relatively high level of endorsement of equality of opportunity and sufficiency. Cluster 2, which we call “meritocrats”, is characterised by a relatively high preference for desert, as well as for equality of opportunity and sufficiency. Cluster 3, labelled “welfare liberals”, had a relatively high level of endorsement of the two needs-based principles, sufficiency and maximin, in addition to a high preference for desert. Cluster 4, labelled “free-marketeers”, emphasises desert, economic efficiency, and equality of opportunity. Interestingly, entitlement did not feature significantly in any of the clusters. Cluster 1 (relational egalitarians) was noticeably smaller than the other three clusters, which are broadly of equal size.

We also investigated the extent to which the distribution of socio-demographic characteristics varies across clusters. Specifically, we calculated χ²s to compare observed age, gender, income, nationality, industry sector, and type of company for each cluster with expected data for the whole sample. The χ² tests in Table 7 show the statistically significant associations which we found.

Table 7 Comparison of cluster demographics

Qualitative Data

Of 1123 participants, 413 provided narrative comments on their answers to the survey. We were unable to reliably interpret 115 of these comments, and a further 70 were not relevant to considerations of distributive justice, leaving 228 useable comments. Of these, 161 refer exclusively to the society level of analysis, 23 refer exclusively to the company level of analysis, and 44 make reference to both. These data were analysed and assigned to one of the six principles of justice. 65 comments are primarily supportive of desert, 61 of equality of opportunity, 53 of sufficiency, 32 of efficiency, 25 of entitlement, and 14 of maximin. These proportions are broadly in line with the quantitative data results. Exemplary quotes are provided in the “Discussion” section.

Discussion

We structure the discussion of our findings around the four sets of research questions set out in the introduction.

Preferred Principles of Justice at Society and at Company Level

Both at society and at company level, the three principles that are most widely endorsed are the merit-based principles of desert and equality of opportunity, combined with the needs-based principle of sufficiency. Of the remaining principles, only efficiency has significant support from participants. The majority view of justice that emerges, both at society and at company level, is that differences in income should reflect differential effort and individual productivity, with the proviso that everyone ought to receive an income that is sufficiently high to live a dignified life.

Desert and equality of opportunity are natural allies—if some individuals are discriminated against because they are members of a particular social group, it is unlikely that differential earnings will reliably track deserved contributions. The third most highly endorsed principle of sufficiency can be interpreted either as a constraint on merit-based principles or as consistent with a particular interpretation of what constitutes “merit”. If the former, sufficiency becomes relevant whenever an individual’s contribution is too small to allow him or her to receive an income that is high enough to live a decent life. In these circumstances, considerations of need trump considerations of desert as a matter of justice. If the latter, sufficiency is consistent with a merit-based formulation according to which everyone who contributes to society has a right to receive an income that would allow them to live a dignified life. Our data suggest that many executives favour this second interpretation of sufficiency. Note, in particular, that the approval index for sufficiency is significantly higher at the company level than at the level of society. One explanation for this difference is that executives think of all company employees, but not of all members of society, as socially productive.

Pluralism in Executives’ Conceptions of Distributive Justice

Both at society and at company level, a large percentage of executives agree or strongly agree with three or more principles of distributive justice. Around 20% of all executives endorse all six principles of justice. Only a small percentage agree with only one principle. Participants’ conceptualisations of distributive justice are therefore notably pluralistic.

At first sight, this pronounced pluralism in executives’ views is rather surprising. One possible explanation is that our respondents expressed incoherent views, ticking boxes more or less at random.⁷ While we cannot exclude this possibility, we are not inclined to give it much weight. After all, our respondents were all highly motivated—they participated voluntarily, were prepared to complete a long and complex questionnaire which required a great deal of thought, and were free to exit the survey without completing it. There was, moreover, a high level of consistency between the approval ratings and forced rankings, and both the cluster analysis and our analysis of the qualitative responses suggest a good deal of coherence in the views being expressed. Nevertheless, while it seems relatively unproblematic to combine merit- with either needs- or free market-based principles to form coherent pluralistic conceptions of distributive justice, it is less clear how more than a fifth of all respondents felt that they could consistently endorse all six principles. To do so is necessarily to endorse a laissez-faire approach to distributive justice, while at the same time recognising that considerations of need have validity as well. Prima facie, this seems contradictory. When defending his entitlement theory of distributive justice, Nozick (1974) contrasts the domain of justice with the domain of charity, and argues that considerations of need fall into the charitable domain. Thus, as far as Nozick is concerned, entitlement theory leaves no room for needs-based considerations. Why might so many respondents to our survey have felt otherwise?

We make progress in resolving this puzzle if we accept that respecting Nozickian entitlements might, to some extent, be necessary to incentivise people to develop their talents and to exert effort in a socially productive way, thus contributing to the economic output of their society. A higher economic output will, in turn, allow a society to improve the fate of its worse-off members. It follows that, even if someone thinks that distributive justice should aim exclusively at meeting human needs, they might nevertheless believe that respecting Nozickian entitlements can be instrumentally important in the pursuit of justice.⁸ A similar argument applies a fortiori to the principle of efficiency. Rawls recognised this fact when he devised the difference principle: he explicitly accepts that some inequalities are likely be to the benefit of the worst-off members of a society. But considerations of Nozickian entitlements need not be put at the service of needs-based considerations in a pluralistic conception of justice that combines both. Despite appearances to the contrary, it is perfectly coherent to independently value both the satisfaction of human need as well as to the non-interference with outcomes that are the result of voluntary transactions. To give an example, someone might believe that a society is distributively just as long as (1) everyone in this society has enough to live a dignified life, and (2) everyone in this society is entitled to what others have voluntarily transferred to them provided that everyone has enough to live a dignified life. In this example, sufficiency takes priority over entitlement, but the good protected by entitlement—a person’s claims to what others have voluntarily transferred to them—nevertheless has independent value.

As a further alternative, and moving beyond the specific concern of combining needs-based principles with those advocating non-interference in market outcomes, it is conceptually coherent to defend a view of distributive justice according to which seemingly contradictory principles all matter independently to justice, and where no principle always takes priority over others. In such a pluralistic conception of justice, as espoused, for example, by Walzer (1983), which principle takes precedence depends on the particular circumstances of the case. While it might be possible to capture the relevant justice trade-offs through a simple formula, it is equally possible that no such simple formula exists. Here it may help to consider an analogy with deontological theories of normative ethics. Many deontological theorists claim that it is important to protect rights to non-interference while at the same time recognising the importance of promoting the general welfare. The proper trade-off between these considerations might be quite straightforward: for example, it might be permissible to infringe a right and harm one individual if this is necessary to promote the more significant welfare interests of, say, five other individuals. Alternatively, whether it is permissible to infringe someone’s rights might depend on a host of different factors that interact in potentially complex ways, such that no simple formula can be constructed to specify the proper trade-off between rights to non-interference and the promotion of welfare under all possible circumstances.

In deontological normative ethics, it is common to assume that comprehensive moral theories are too complex to be easily attainable, and that moral theorising should therefore best proceed in a contextualised and piecemeal manner. In contrast, it tends to be an accepted background assumption among distributive justice theorists that the theories that they put forward should be both simple and comprehensive. As Walzer puts it: “The deepest assumption of most philosophers who have written about [distributive] justice, from Plato onwards, is that there is one, and only one, distributive system that philosophy can rightly encompass” (Walzer 1983, p. 5). That is, it is unusual for distributive justice theorists to assume that needs-based, merit-based, and free market-based principles might all protect values that matter independently in ways that are difficult to incorporate into a handy formula. However, while unusual, such views are not entirely absent from philosophical discussion. Besides Walzer’s influential view, pluralistic theories of distributive justice are put forward also by Rescher (1966), Galston (1980), and Miller (1999).

The overt pluralism in our respondents’ distributive justice views suggests that business executives are drawn to rather complex conceptualisations of distributive justice. They appear to believe that distributive justice is governed by many different considerations that are not easily distilled into a single instructive formula, in much the same way that some deontologists think about normative ethics in a pluralistic manner. This is a most interesting outcome, and one that has not been widely explored by distributive justice theorists.

If the pluralism endorsed by business executives fits the subject matter of distributive justice, this raises the question of what is the most appropriate analytical framework for organising and comparing different views of distributive justice. In her study of American beliefs about distributive justice, Hochschild (1981, p. 52) arranges the five norms of justice that she identifies along a continuum, from what she describes as the “principle of equality” at one end of the scale to the “principle of differentiation” at the other end. Schminke et al. (1997) draw a binary distinction between what they call “formalist approaches” (associated with Kantian ethics) and “utilitarian approaches” (associated with Jeremy Bentham and John Stuart Mill).

Based on our findings, we propose a conceptualisation of the distributive justice space as a “field” (after Lewin 1951; see also Mey 1972). A field is characterised by the “operant factors” that are at work in it. For our purposes, the relevant operant factors are captured by different principles of distributive justice, and include considerations such as an individual’s effort, their contribution, or their need. The field is constructed by imagining that our data are projected into what is mathematically speaking a six-dimensional space, with each dimension or axis corresponding to one of the six principles of justice. Each of the operant factors exerts a force that might vary depending on the context to help determine what is distributively just in a given situation. Conceptualising distributive justice as a field allows a more sophisticated analysis and categorisation of distributive justice ideals than the two-dimensional representations of Schminke et al. (1997) and Hochschild (1981).

The findings from the cluster analysis that we performed on the ranked principles fits well with the proposed field-theory framework. It groups the respondents of our survey into four clusters that assign differential strengths to different operant factors such as merit or need. We summarise the main characteristics of these four clusters in Table 8, along with exemplary quotes drawn from the narrative comments in the qualitative component of our study.

Table 8 Characteristics of the four clusters

Two of the clusters, “relational egalitarians” (Cluster 1) and “meritocrats” (Cluster 2), emphasise the importance of merit-based principles (equality of opportunity in the case of relational egalitarians, and both equality of opportunity and desert in the case of meritocrats), but combine this with a sufficientarian floor constraint. “Welfare liberals” (Cluster 3) combine desert, sufficiency, and the maximin principle. “Free-marketeers” (Cluster 4) combine equal opportunities, desert, and efficiency. Interestingly, entitlement does not play a significant role in any of the four clusters, indicative, perhaps, that notwithstanding our technical analysis set out above, business executives find Nozickian entitlement difficult to reconcile with merit or needs.

For relational egalitarians (Cluster 1), needs and the provision of equal opportunities trump all other considerations, including the consideration that differential effort and contribution should be rewarded differentially. In their narrative comments, relational egalitarians say things like: “Everyone should have the same benefits and same opportunities – people should have the same educational opportunities that allow them to access jobs with fair payments according to their abilities and desires”; “A society in which wealth inflation can be greater than savings potential on minimum wages will never be just – we need a method that can address this, which makes property and wealth accumulation more accessible objectives for all”; and “Markets do not work – either in terms of labour or trade – skills in high demand are more easily developed or bought through high cost education or work experience by those with existing wealth or social status”.

Meritocrats (Cluster 2) are more concerned with rewarding productivity, which they combine with a strong focus on equality of opportunity and sufficiency. In their narrative comments, they say things like: “In a society which values contribution, effort, skill and experience, the major focus should be on impact and contribution”; “All people should have opportunities in the job market equally, but their appointment should depend solely on their effort and not on external influences… People should only be promoted on merit”; and “Ideally, remuneration should be based on merit (e.g., qualifications, special ability, skill shortage, special responsibility, commitment, entrepreneurial spirit, etc.)… People who simply do not make an effort should also only receive the absolute subsistence level support”. Their thought seems to be that, while we should try to reward differential contributions differentially, we also have some justice obligations based solely on need.

Welfare liberals (Cluster 3) share Rawls’s belief that rewards for contribution are desirable as long as the benefits spill over to others. They say things like: “I agree with the principle that guarantees the welfare of all society without exception, always thinking about the dignity of all people, including the most disadvantaged, but also recognising the importance of freedom of choice, equal opportunities, and encouraging talent – we need both”; “Those who have more must contribute more to society”; and “There must always be the principle of collective responsibility for those who have less: we must act with social responsibility towards our neighbour who suffers a greater degree of poverty, under the principle of ‘love for your neighbour’”.

Finally, “free-marketeers” (Cluster 4) reject the force of needs-based principles, instead conceptualising distributive justice as based entirely upon merit and economic efficiency. As they see it, distributive justice is about rewarding talent and ensuring that a functioning price mechanism leads to an efficient allocation of resources. They say things like: “Without the rich, there is no wealth creation”; “I strongly believe members of society should be motivated to work and that income redistribution removes this motivation”; and “Let free markets reign! I want a society where people are free to win according to their skills, abilities, efforts and contributions – I accept that in such a society, there will be some losers”. These narrative comments appear to reflect a sympathy with the principle of entitlement, even if the quantitative analysis does not provide equivalent statistical support.

Is Distributive Justice Viewed Differently at Society and Company levels?

Respondents to our survey see distributive justice as both a societal and an organisational issue. Results at society and company levels of analysis are highly correlated. Participants endorsed similar principles at both levels, and the vast majority agree or strongly agree with the statement that “corporations should take into account principles of fairness and distributive justice when determining their pay strategies”. Many economists argue that there should be a division of labour between market participants and governments, where market participants ought to ensure efficient outcomes, and where governments are charged with redistributing wealth to turn efficient outcomes into fair ones (Heath 2014). Our data suggest most executives believe that companies cannot simply outsource matters of distributive justice to governments to address, for example via the tax system.

One of the main differences between views expressed at society and company levels is the higher approval of efficiency at company level. As a group, respondents seem to recognise that businesses have an important role to play in ensuring that prices paid for labour reflect relative scarcity. However, crucially, this is seen as just one consideration among others, with sufficiency also receiving a higher approval rate at company level.

Do Executives Believe that Societies and Companies are Predominantly Just?

By and large, executives do not believe that they live in just societies or work for companies that provide just incomes according to the principles most highly endorsed. Both at company and society levels, the extent to which the widely endorsed principles of desert, equality of opportunity, and sufficiency are perceived to govern the distribution of income falls significantly short of the extent to which this is deemed desirable and morally appropriate. At company level, the worst perceived shortfall is with respect to sufficiency, where the vast majority of respondents agree that the principle should govern the design of pay packages, but less than a third believe that it actually does. At society level, the most significant perceived shortfall is with respect to equality of opportunity—the majority thinks an ideal society would be governed by this principle, but less than half agree that their society is.

Interestingly, these findings suggest that the role played by biases in executives’ views of distributive justice is more limited than suggested by Miller (1992). Miller draws attention to the presence of self-serving biases in perceptions of distributive justice in his survey of the empirical literature. He reports that how someone assesses the justice of the economic system depends on their socioeconomic status. Those in lower income brackets tend to think that existing inequalities in income are due primarily to unequal opportunities and are therefore unjust, whereas those in higher income brackets are more likely to think that existing inequalities are due to differential effort and ambition, and are therefore justified (Miller 1992, pp. 585–586). Our findings indicate that many business executives are able to recognise that a person of lower socioeconomic status might be stuck as a result of unequal opportunities.

Lerner (1980) draws attention to another type of bias. He shows that what he calls “just world beliefs” tend to distort our views about the fairness of the world that we live in. According to Lerner, we have a deep-seated need to perceive our social environment as largely just and, therefore, controllable. Through a series of experimental studies, he shows that this need leads us to explain away injustices, for example by unfairly blaming victims, or by fabricating stories that explain why lucky winners supposedly deserve their good luck. In a more recent study, Frank et al. (2015) show that just world beliefs affect individual and societal support for redistributive principles. They begin with the observation that individuals and societies that are more sceptical of redistribution have much higher faith in the just world notion that effort and success are tightly linked in their society. They then show that just world beliefs have a causal effect: if an individual or a society becomes more confident that making an effort will lead to success, this causes the individual or the society to become less supportive of redistributive principles.

In our study, we found that the vast majority of respondents endorse the redistributive principle of Sufficiency. If Frank et al. (2015) are correct, this endorsement is at least partly linked to the fact that many respondents do not feel that merit-based principles govern the distribution of income in the societies that they live in or in the companies that they work for (see also Krawczyk 2010).

Do Conceptualisations of Distributive Justice Vary by Socio-demographic Markers?

While some heterogeneity in distributive justice views is apparent in the data, there are not many large differences in views across demographic markers such as age, gender, income, nationality, company type, or industry sector. While the views of a free-marketeer differ markedly from the views of, say, a welfare liberal, the prevalence of free-marketeers and welfare liberals is roughly equal among different age groups, income brackets, genders, industry sectors, etc. This finding is consistent with earlier empirical findings reported by Marshall et al. (1999), who find that distributive justice views do not differ significantly between socioeconomic classes and different cultures. Having said that, we did identify at least some noticeable differences across the four clusters (see Table 7). Here, we discuss only the highly statistical significant differences in the distribution (p < 0.01), and comment on the magnitude of these differences.

Among welfare liberals, there are a disproportionally large number of Chinese nationals—indeed, on further investigation, we discovered that 60% of Chinese participants are characterised as welfare liberals. Conversely, only 10% of Swiss nationals fit this category. The ages and industry sectors of welfare liberals are also interesting. They are younger than expected—33% are under the age of 35 years of age, and a further 30% are between 35 and 40. They also have comparatively lower incomes on average, in part likely due to their relatively younger age. Welfare liberals are also highly represented in the technology sector—39% of technology executives are found in this cluster. The flipside of the high number of Chinese nationals who are welfare liberals is that only a small number (less than 10%) are free-marketeers. Lastly, the age profile of meritocrats is skewed towards older executives—this cluster is under-represented in executives under the age of 40.

Summary and Conclusions

We report four main findings. Our first finding is that the distributive justice views of business executives are pluralistic, and that they tend to become more pluralistic as an executive’s seniority increases. At both society and company levels of analysis, the majority of executives endorse three or more principles of justice. At first sight, this pronounced pluralism seems rather puzzling. However, we have advanced a number of possible explanations for this phenomenon, and we have argued that a novel field-theory framework can be used to systematise and compare pronouncedly pluralistic views of justice.

Our second finding is that executives endorse similar principles at societal and company levels of analysis. Business executives do not appear to believe that companies should simply leave matters of distributive justice for governments to deal with via the tax system. Some economists argue for a separation of responsibilities, such that market participants are responsible for ensuring efficient economic outcomes through their self-interested behaviour, and governments are responsible for redistributing wealth to turn efficient outcomes into fair ones (Heath 2014). Our research suggests that the majority of executives reject this approach, believing instead that companies have a more direct role to play in bringing about distributive justice. This has important practical implications. It suggests that, while senior executives may be happy to receive a significant element of highly geared performance-based pay for themselves, they would also, if pushed, express a belief that all employees in their companies should be allowed to make a decent living.

Our third finding is that most business executives live in societies and work for companies that they perceive to be significantly unjust. Most strikingly, the vast majority of executives think that equality of opportunity is an important principle of distributive justice at the society level of analysis, but less than half believe that the society that they live in is characterised by equal opportunities. At the company level, many executives endorse a desert-based principle of distributive justice, but less than half think that such a principle is representative of existing remuneration policies within their organisations. This finding is of significant practical interest as well. It suggests executives should be prepared to accept a greater focus on distributive justice in company-wide pay policies.

Our fourth finding is that, while distributive justice views among executives are relatively homogenous across demographic markers, some significant variations can nevertheless be identified. Most prominently, respondents who are either younger or who live in China are more likely to endorse the idea that we should combine a sufficientarian floor constraint with desert, thus falling into the cluster of welfare liberals.

Managerial and Public Policy Implications

Our research suggests a potential gap between the beliefs and behaviours of senior executives when it comes to distributive justice: folk opinions about what people think in the C-suite, derived not unreasonably from behaviours exhibited by some prominent senior executives as well as extensive media coverage, differ significantly from managers’ actual beliefs. Companies should be encouraged to think of ways of closing this gap by requiring senior executives to think about their own pay in the context of their firms’ overall pay policies, particularly how the least well-off employees are compensated in comparison with the most well-off.

In the UK, for example, it used to be the case that company pay policies considered the pay of workers, middle managers, and executives as part of a continuum, with grade bands, pay spines, and job-evaluation applying across the whole spectrum (for further details see Willman and Pepper 2020). Executive pay was largely determined by reference to internal labour markets (and hence internal relativities) rather than external labour markets. Today executive and all-employee pay is generally determined quite separately, paid from different expense budgets, in accordance with different sets of principles, and using different financial instruments (with a significant weighting of variable stock-awards in the case of executives compared with fixed cash payments in the case of other employees). Determining pay by reference to internal labour market relativities meant that company top-to-bottom pay ratios were generally smaller, and possible more distributively just, than they are today. Governments have the ability to nudge companies to change their behaviour in this regard. Recent moves in the US, UK, and other developed economies requiring large companies to publish pay ratios are a step in this direction, but an increased public policy focus on distributive justice should be encouraged.

Contribution

We add to the empirical literature on distributive justice and executive compensation in various ways. By employing a unique set of primary data about the beliefs of business executives from around the world, we shed light on how senior executives think about distributive justice; as far as we know, our study is the first to explore the views of high-level business executives in this way. We also propose a new field-theory framework for conceptualising beliefs about distributive justice. We argue that, unlike other approaches found in the extant literature, this framework can be used to capture the pluralistic views about distributive justice which our respondents expressed. A field-theory framework might be used more widely to represent the complex and context-dependent theories of justice which we have argued underpin such pluralistic views.

Limitations and Implications for Further Research

A key limitation of our study is that we obtained our data not by random sampling, but by quota sampling. While only random sampling can guarantee that the insights obtained are representative of the underlying population, feasibility constraints make it unlikely that future research will be able easily to address this limitation. Further research might, however, deal with other limitations of our study. As we argue above, the pluralism in executives’ justice views suggests that they endorse complex theories of distributive justice that incorporate seemingly contradictory principles whose force might vary depending on context. Our study is unable to shed much light on what executives think about the relationship between the different principles that they endorse: the only question in our survey that pertains to this issue asks respondents, rather crudely, to rank the questionnaire’s six principles of distributive justice from most preferred to least preferred. While forced ranking tells us something about which principles are perceived to be most central to distributive justice, future empirical research—conducted, for instance, in the form of interviews or using focus groups—might aim to capture more nuances in the perceived relationship between the different principles. If executives’ views are as pluralistic as our study suggests, further research might help give shape to the pluralism which is at work, which could in turn be captured by the field-theory framework we have proposed.

Future research might also investigate how the distributive justice views of business executives compare to those of the general public, perhaps by using the same questionnaire we used for our study. In the “Discussion” section, we make some limited comparisons between the views of business executive and those of the general public based on secondary literature. Less guarded comparisons are possible only if the data collection instruments are matched.

Conclusion

In conclusion, business executives are frequently depicted as tough-minded individuals with a winner-takes-all mentality. Because they have worked hard and achieved professional success, business elites assume, or so the conventional thinking goes, that everyone else can achieve equivalent success if they work hard as well. Our findings indicate that this picture of executives’ beliefs is too simplistic. Many executives are not convinced that they live in just societies or work for just companies. Most endorse pluralistic views of justice that balance considerations of merit with considerations of need. The majority believe that matters of distributive justice should be dealt with both at a societal and company levels. This suggests that executives are more sensitive to issues of distributive justice than is frequently assumed, which in turn has important practical implications, some of which have been sketched-out in this paper. We hope that our work will encourage further discussion about distributive justice among business executives, and spur companies into thinking more about distributive justice when determining corporate pay policies.