Imagine a world in which medical researchers did experiments on rats, but never on people. Furthermore, suppose that doctors ignored the rat literature entirely. Instead, they talked to each other and swapped tips, based on their own clinical experience. In such a world medicine would not be the cumulative science that we know today.
That fanciful clinical world is the world of business ethics research. University researchers do experiments, mostly on students who come into the lab for pay or course credit. Experiments are run carefully, social and cognitive processes are elucidated, and articles get published in academic journals. But business leaders do not read these journals, and rarely even read about the studies second-hand. Instead, when they think and talk about ethics, they rely on their own experience, and the experience of their friends. CEOs share their insights on ethical leadership. Ethics and compliance officers meet at conferences to swap 'best practices' that haven't been research-tested. There are fads, but there is no clear progress.
What it should look like
Free-market societies — indeed all societies — would be vastly better off if we could improve business ethics. Efficiency would improve, enlarging the pie, and workers would be treated better, removing some of the animus directed toward capitalism and free trade in recent years.
But improving business ethics is not a one-time achievement. The world of business is ever-changing. We need to connect business leaders, researchers, and regulators in ways that create feedback loops that keep the improvements going. To put this in evolutionary terms, we need to generate a constant stream of innovations in business ethics, which are then tested empirically and selected so that the most effective ones become more common throughout the business world, while those that are ineffective or counter-productive fade away.
Why it's not there yet
There are at least three reasons why business ethics is not yet a cumulative science.
The hyper-complexity of business ethics. The scientific method and scientific mind generally favour creating simple versions of phenomena, sometimes artificial versions in a metaphorical test tube, which allows researchers to manipulate one element and infer causal processes. But a real business is extraordinarily complex, with variables interacting in ways that cannot be modelled in the behavioural lab. Businesses are composed of individuals who make choices. Individuals are placed in groups, each group developing its own emergent properties, sometimes known as culture. Businesses are also embedded in the legal and cultural ecosystems of the nations within which they operate.
Current research in behavioural ethics almost exclusively addresses individual choice. For example, is dishonest behaviour reduced if people are asked to sign a form as proof of honest intent at the beginning rather than at the end?
However, to understand what drives good and bad ethical behaviour in a modern corporation, one must understand what is going on at all three levels (the individual, the group, and the legal and cultural ecosystem), and one must examine the alignment (or misalignment) across levels, and within each level. Sometimes a reform at one level will cause adverse effects at another level, which are not visible for several years. For example, in 1993, the US Congress mandated that executive pay higher than US$1 million a year is not a fully deductible expense unless the company puts in place a 'pay for performance' plan tying the executive's compensation to objective measures such as the company's stock price. But this new law had the unintended consequence of shifting executive behaviour and corporate culture toward a focus on raising stock price — via accounting tricks when necessary — rather than on creating underlying long-term value. These sorts of cross-level unintended effects cannot be effectively studied in the lab.
The risk aversion of firms. Because of the hyper-complexity of business ethics, it is crucial that researchers gain access to real companies, but how? Put yourself in the shoes of the firm's top lawyer — the general counsel. An inquiry comes in from some university professors who want to study ethical and unethical behaviour among your employees. What possible upside is there for the company? Even if the researchers tell you that their findings might help you improve the company's ethical culture, that's not your job. You're paid to assess and protect against risk, particularly litigation risks. Why take unnecessary risks by inviting strangers in to poke around and ask questions, knowing that these strangers will then publish their findings, even if they say they will hide the name of your company? Request denied.
The siloed data problem. The academy is mostly siloed. Researchers publish in their fields' journals and rarely venture outside disciplinary boundaries. Of course, researchers who study accounting or finance are welcome to read the journals of researchers studying management or negotiation and occasionally they do. Yet when it comes to sharing data, the walls are higher. Data is normally kept private and only shared with other researchers under limited conditions, particularly when it benefits the researchers in some way. When those data are (rarely) gathered from real companies, which are concerned about downside risk, the walls are even higher. It sometimes happens that even measurement tools are made proprietary and not offered freely to the entire research community.
When we look on the business side, siloing of data is almost universal. Most companies collect little or no data about their ethical culture, and if they do, they most likely would not share it with anyone. Many hire consultants to perform assessments using proprietary instruments. Companies such as Gallup and Kenexa can compare data across companies they have surveyed (benchmarking), but they rarely, if ever, share these data with researchers, nor do they allow researchers to use their proprietary survey questions.
The net result is that a large amount of data is collected every year on ethics in business, but it is done using such a welter of instruments, and with so little sharing of data, that there is little forward momentum.
How to fix it
What can we do to get the business and research communities together, and to establish the sorts of long-standing trusting relationships that can lead to longitudinal studies in which data is gathered using the same instruments over the course of several years and many companies, while various interventions are tested? This is the holy grail of business ethics research.
In 2014, a group of ethics researchers from many subdisciplines came together to form a research collaborative called Ethical Systems (see EthicalSystems.org). We were formed to address the hyper-complexity problem. We began by summarizing the existing research on topics as varied as accounting, fairness, business law, human rights, conflicts of interest, ethical culture, and whistle-blowing. Our initial goal was to aggregate the vast and varied research literature and make it accessible — always for free — to business leaders and especially to ethics and compliance officers. (Because culture and regulatory frameworks differ by country, we have limited our work to the United States so far, but we plan to expand globally in the future.)
Our second goal was to bring researchers together from multiple subfields and link them to the many business leaders who have begun to realize that they can't just focus on compliance with regulations; they must invest in improving their ethical cultures. We have found a great deal of interest in working together from all the relevant groups — including federal regulators.
Here is what we believe must happen next:
- Researchers and business people must work together to create a common set of open-source assessment tools that are rigorous enough for academic work, and that measure the constructs that businesses most want to know about (such as what gets employees to report misconduct). These tools must go far beyond employee surveys, offering measurement of implicit attitudes, corporate reputation, and other relevant constructs. We have already begun to create this set of tools with an 'ethical culture measurement project' at Ethical Systems. The collaboration will find ways of sharing or pooling data that are stripped of identifying information, however, this task will be a major challenge.
- Government regulators must join the collaboration and assure business leaders that they will be more flexible and lenient with companies that are making a sincere effort to improve their ethical cultures. In the United States, regulators have said that they would take such a stance since the federal sentencing guidelines were revised in 2004, but it's not clear whether or how they have done so, in part because there is not yet an agreed upon way to define, measure, and implement ethical culture. Defining legal compliance is much easier. Now is the time to reward companies that answer the call. As a first step, prosecutors and regulators (such as the Department of Justice's compliance counsel) could state clearly that they will include a company's contributions to ethics research as a factor to be weighed in any enforcement decision involving the company.
- Researchers must make the research they carry out with these open-source tools and improved access useful to companies; they must make the collaboration mutually beneficial. They and their graduate students must spend time inside companies, talking to leaders, listening to employees, and developing interventions that help both groups reap the benefits of creating a high-trust high-ethics environment.