Tuesday, March 18, 2008

Curing the Moral Hazard : By Josh Hess

This post reflects the opinions of Josh Hess only.

Critics of the Honor System frequently insist that the inflexibility of its sanction creates perverse incentives for guilty students to lie once accused. At the heart of this contention is the realization that the system actually rewards those who decide to contest their guilt by allowing them to continue to accrue course credit until the date of their dismissal at trial.

In recognition of this state of affairs, critics of the system propose replacing the single sanction with an “informed retraction.” Under this new regime, students who admit their guilt within a certain period of time after accusation would be permitted to return to the University at some point in the future. Students who contest their guilt and request a trial would still face only permanent dismissal.

One problem with this approach is that it creates perverse incentives of its own. The potential disparity in outcomes for students deciding whether to “leave admitting guilt” or contest their guilt at trial is substantial. Students who are genuinely unsure as to whether they have committed an Honor offense confront a choice between admitting their “guilt” and facing mere suspension and contesting it and facing permanent dismissal. There is a very real risk here that not guilty students would subject themselves to suspension because it is the “safer” option.

But the realization that the informed retraction is undesirable (as students did in 2002) leaves the problems it sought to address untouched.

This past Sunday, the Honor Committee reviewed three proposals to address the “moral hazard” without stirring up new complications. These ideas represent a moderate, creative alternative to the informed retraction and my hope is that the Committee will adopt them.

The first proposal offers to change the transcript notation for students who LAG (“leave admitting guilt”) rather than request a trial from “enrollment discontinued” to “withdrawn.” “Enrollment Discontinued” is already meant to offer dismissed students a certain measure of confidentiality post-dismissal. The switch to “withdrawn” would simply offer one more degree of anonymity.

The differentiating degree is not large enough, though, to create the perverse incentives that one finds with the informed retraction. The shift simply makes the decision by knowingly guilty students to more clearly rational. The shift in transcript notation, unlike the informed retraction, does not present a stand-alone incentive to LAG for not-guilty students.

The other two proposals seek to place reasonable limits on accused students’ capacity to “game” the trial scheduling process to obtain more credit. One proposes giving accused students who request a trial a pre-determined set of trial dates in the near future from which to choose, shifting the burden on them to make a showing as to why none of the dates work and delay is justified. The other attempts to mitigate the incentive to delay in dishonest pursuit of credit accrual by giving the Honor Committee’s Executive Committee the ability to, with a unanimous finding of system-gaming, request that the Faculty Senate deprive delaying students of the credit they acquired post-accusation.

Animating these two proposals is an impulse to minimize the breadth of the moral hazard by making the “reward” of continued dishonesty post-accusation less certain.

My firm belief is that these proposals individually will not have a material capacity to address the moral hazard complications. But, passed together, I do think they would represent a substantial contribution.