AMERICAN ACCOUNTING ASSOCIATION
http://aaahq.org
CONTACT: Ben Hamowitz (718-398-7642)

How a simple, trusting practice fosters worker honesty

To stop a thief can be a formidable challenge. But keeping honest people honest doesn’t seem easy either – or so the CEO of Wells Fargo appeared recently to suggest in defending the firing of 5,300 employees for deceptive practices.

How to keep the honest majority honest in the face of temptations to be unethical? New research in the scholarly journal Behavioral Research in Accounting tests a simple, trusting approach and finds it yields promising results.

The new study examines the practice of self-certification, defined as "an informal, trust-based control mechanism whereby managers sign their names to accept responsibility for a decision but know that there are no penalties associated with the decision because the information on which it is based is private and not monitored."

Asked how common it is for companies to be this trusting, the study's authors, Nicole Ang and Mandy M. Cheng of the university UNSW Australia, acknowledge that formal monitoring of employees continues to predominate accounting and business practice. Still, self-certification is often used in activities that, while by no means trivial, are less than critical to organizational functioning, such as submitting claims for remuneration of discretionary expenses or making absence-from-work requests. In universities, the professors add, self-certification has been used with conflict-of-interest disclosures and teaching-assessment reviews.

And this trust-based mechanism can work quite well: the mere signing of one's initials to a request or statement considerably improves truthfulness, the study finds.

Promising though it is, the practice also has its limits. It increases the chance that an individual who violates trust on one occasion – for example, by claiming reimbursement for inappropriate or misstated expenses – will do the same in the future. It also has little effect if combined with some kind of formal monitoring, even monitoring conducted randomly and rarely. In the words of the study, "this suggests that a control mechanism that appeals to managers' inherent moral tendencies is fragile and can be overwhelmed by traditional control systems."

The professors add: "Organizations should not simply impose self-certification requirements on top of existing monitoring-based controls [but rather combine them] with other involvement-oriented controls (i.e., control mechanisms that encourage participation and empowerment) to promote pro-organizational behavior.”

In other words, for trust to be effective requires a departure from traditional control systems that are common in today's business world, the study suggests.

The paper's findings derive from experiments in which graduate business students were exposed to ethically-challenging hypothetical situations. Subjects were directed to assume the roles of corporate managers tasked with deciding whether to continue projects they were responsible for and had promoted extensively. The principal interest of the research was cases where projects were under-performing and unlikely to become profitable, projects whose termination would represent a personal setback for a manager but would most probably be advantageous for the company. In about half these cases, subjects were asked to initial a brief certification form acknowledging responsibility for their decision; in the remaining cases there was no certification form, and participants decided the matter by simply checking a box.

In the paper's main experiment, 103 participants were informed that they alone had access to information about the project, and those who made certification statements were assured the statements would not be reviewed by others but merely filed away with other project information. Among self-certifiers, 53% decided to terminate under-performing projects, compared to only about 29% of those who were not asked to self-certify. This statistically significant result leads the authors to observe that "a relatively simple and non-intrusive control mechanism – self-certification – can promote honesty and goal-congruent decisions in a setting that prior research has demonstrated often results in opportunistic behavior."

Then the experiment was extended into a second stage, and participants were informed that a year had passed since their initial decision and that projects which previously were under-performing were still doing so. Now 14 of 15 self-certifiers who had earlier opted to continue their projects decided again to do the same, once again favoring their personal interests over the company's. In the words of the study, "managers who choose to continue an under-performing project in stage one are significantly more likely to continue that project in stage two if they have a self-certification requirement than if they do not."

Why did self-certification cease to foster good citizenship in stage two? The authors cite the much-studied psychological phenomenon of cognitive dissonance, mental stress caused by contradictory thoughts or actions. As they write, "Our finding is consistent with prior psychology research that shows that the desire to maintain commitment to a previous unethical act in order to reduce dissonance can lead to even more unethical acts."

In a second experiment, 49 different participants were also asked to decide the fate of their under-performing projects. But now subjects were informed there was a five-percent chance their decisions would be reviewed by internal auditors who would provide a report to the participant's superior. In this circumstance, self-certification lacked the salutary effect it had in the main experiment where confidentiality was guaranteed. In short, as the study notes, "self-certification is not effective when combined with formal monitoring."

The new paper, entitled “Does Self-Certification Encourage or Reduce Opportunistic Behavior?" is in the fall issue of Behavioral Research in Accounting, published twice-yearly by the American Accounting Association, a worldwide organization devoted to excellence in accounting education, research, and practice. Other journals published by the AAA and its specialty sections include The Accounting Review, Auditing: A Journal of Practice and Theory, Accounting Horizons, Issues in Accounting Education, Journal of Management Accounting Research, Journal of Information Systems, and The Journal of the American Taxation Association. In addition, the AAA is inaugurating two new publications, the Journal of Financial Reporting and Journal of Forensic Accounting Research.