In the spirit of fair discourse, I sent to Eugene Soltes at HBS a link to my previous post criticising his HBR online column "Why It’s So Hard to Train Someone to Make an Ethical Decision." I did not expect a response. I was pleasantly surprised to receive a polite and gracious response from Dr Soltes. Naturally enough, he referred me to a longer treatment of the issues upon which he had blogged in a recently-published book Why They Do It: Inside the Mind of the White-Collar Criminal. I've not read Dr Soltes' book.
To his considerable credit, in the personal email he stated:
[E]ven if it's to disagree with me, I'm happy to see this spurring more discussion since I believe that is critical.
The crux of his argument was encapsulated in the following statement.
"We know that ethical, well-intentioned people under pressure or poor incentives often do make unethical decisions and these are "normal" people."
But I remained troubled; ethics is a troubling field; at least I find it so. It is also a topical one. But we will return to that theme shortly.
Harvard Business Review has, over many years, promoted the issue of business ethics; many of its pages have been filled with commentary on ethics and it has sought to position itself as the purveyor of leading thinking in the field. One of my personal favourites from their extensive canon is the 1986 article by Saul Gellerman titled "Why 'good' managers make bad ethical choices." While I had studied ethics and morality as an undergraduate, this was the first article I encountered on the topic of ethics in the professional workplace. Gellerman argued that individuals use four rationalisations to justify unethical behaviour:
- A belief that the activity is within reasonable ethical and legal limits—that is, that it is not “really” illegal or immoral.
- A belief that the activity is in the individual’s or the corporation’s best interests—that the individual would somehow be expected to undertake the activity.
- A belief that the activity is “safe” because it will never be found out or publicized; the classic crime-and-punishment issue of discovery.
- A belief that because the activity helps the company, the company will condone it and even protect the person who engages in it.
"Put enough people in an ambiguous, ill-defined situation, and some will conclude that whatever hasn’t been labeled specifically wrong must be OK—especially if they are rewarded for certain acts."
This presents a very different line of argument from Soltes' "ethical, well-intentioned people." This is not about failure properly to evaluate inter-temporal trade-offs or factor in social costs or other externalities or unintended or unanticipated consequences; it is usually classic displacement – "someone else must have judged the action to be acceptable, so I don't need to think about it."
Gellerman goes much further than Soltes:
"We know that conscience alone does not deter everyone."
Those undeterred by conscience cannot claim to be "ethical, well-intentioned people;" at best, they would be amoral.
Perhaps the most frequently cited discussion of morality and business is Milton Friedman's 1970 editorial in the NY Times Magazine "The Social Responsibility of Business is to Increase its Profits." Frequently argued against as polemical, Friedman states at one point:
"In a free-enterprise, private-property system, a corporate executive is an employee of the owners of the business. He has direct responsibility to his employers. That responsibility is to conduct the business in accordance with their desires, which generally will be to make as much money as possible while conforming to the basic rules of the society, both those embodied in law and those embodied in ethical custom." (emphasis added)
In his 1986 article, Gellerman addresses the problem also:
"Executives have a right to expect loyalty from employees against competitors and detractors, but not loyalty against the law, or against common morality, or against society itself."
The practical problems are (i) how to operate a system of control in the firm to deter from illegal, immoral, anti-social or other detrimental acts even those whose consciences are not engaged, and (ii) how to educate people to do so.
In an article in the Washington Post almost exactly three years ago, Jonathan Haidt, whom I quoted in the earlier blog responding to Dr Soltes' HBR column, considers explicitly how to do the latter. One of his suggestions is to "add a course on ethical systems design." But we have these already. And we have published guidance on how to design such a system. COSO, originally published in 1992 (and on which I have previously blogged extensively) and enshrined in US law by the PCAOB's interpretation of s.404 of the Sarbanes Oxley Act purports to apply just such principles, enshrining the phrase 'tone at the top' in the management lexicon. However, 'tone at the top' doesn't mean anything, precisely because it is so open to variable and subjective interpretation. Far better is Kellerman's phrase:
"Top management has a responsibility to exert a moral force within the company."
That places responsibility for action squarely where it belongs. Haidt is wary of teaching business students Kant. Yet, still, Kant's philosophic construct of the categorical imperative is unsurpassed in secular thinking as a precept for ethical and moral reasoning and action. And Haidt is perfectly willing to reflect on Stanley Milgram's experiments of subjects administering (supposedly) lethal electric shocks because it is directed by the study; Soltes also claims to place considerable store in "behavioural ethics." We need to be careful that we do not substitute the easy and appealing for the important, in a business education no less than in life. Just because there is a video record of it does not make it intrinsically more worthwhile.
Clearly, one (but only one) element of design of an ethical system for business is the search for objective information. In the 'information age', businesses rely increasingly heavily on data as an input to managerial decision-making. This trend has increased enormously the allocative and productive efficiency of capital and businesses in my lifetime (which coincides neatly with the duration of Gordon Moore's famous article about cramming circuits on a micro-chip, giving us 'Moore's Law').
All information systems are subject to error. The first and most obvious is 'basis error': that what is being measured is actually not what was intended to be observed. But there are many other errors that arise. Among complex social phenomena (such as most ambiguous areas of business), categorisation and classification errors are common problems. All too often, phenomena are measured because it is convenient to do so. Complex social feedback systems are minimised or ignored and non-linearity in systems is overlooked. As Angell and Straub (1999) stated:
The existence of positive feedback implies that there can be no permanent control over any society that is continuously evolving and emerging, or over any of its institutions; the imposition of standardised and standardising categories comes with the inevitability of long-term damage. The perversity of consequence turns design logic on its head. There can be no expectation of ‘being in control’. Control only exists in the sense of purposefully formulating and precipitating actions or intentions; but this is not being in control of consequences. In learning to live with this uncertainty, societies cannot deny inevitable ambiguity and vagueness.
When applied to the logic of audit and verification, as advocated actively by Gellerman, these phenomena result in what Mike Power of LSE has famously called the illusion of control, consciously using a phrase from Ellen Langer, whose work Dr Soltes also quoted in his blog. We must take care when thinking about 'objective information'. Such care would form an essential part of any Haidt-inspired "course on ethical systems design."
The last few days have seen another phenomenon emerge. The unedifying spectacle of leading figures in the new US administration inflating the number of attendees at the Presidential Inauguration last Friday and subsequently arguing for "alternative facts" introduces a new, more spectacular form of error: conscious, deliberate falsehood. While ludicrous and laughable, the positing of 'alternative facts' introduces a considerably more pernicious problem: 'tone at the very top', if you will. If senior US administration figures feel they can lie with impunity, this creates an environment that devalues objective information. In such an environment, it becomes problematic to expect rational debate based on a shared understanding of reality; ethics, be damned. When you can dismiss objective information and make up 'facts', ethical judgment is a minor problem.
Ultimately, Eugene Soltes' position and mine are remarkably similar, we just disagree on the specific cases he used (as he noted in correspondence). We concur that it is a difficult and complex exercise to establish effective and meaningful systems of corporate control of which the ethics of managerial judgment play a role. I simply feel he lets transgressors off too lightly. In the cases he cited, they were not ethical people doing unethical things; they are unethical people doing unethical things. There are shades of grey, but not that involve illegality from people who know and understand the law and their obligations under the law; there is nothing penumbral about that. Milton Friedman would allow no such quarter.
If the disregard for objective fact observed in the last few days is sustained, if we have growing disdain for what political pundit and comedian Stephen Colbert called "truthiness" under the last Republican administration, ethical judgment based on truth will suffer. If the disdain for truth is sustained, ethical violations and corporate illegality in the US will grow alongside a further deterioration in respect for politicians and the political process. And we will all suffer as the more egregious of these violations inevitably come to light; to quote The Merchant of Venice, "at length truth will out." What price will we need to pay this time?