Talk 1:

Informal Sanctions on Prosecutors and Defendants and the Disposition of Criminal Cases
Jennifer Reinganum (Vanderbilt University, United States)

We model the strategic interaction between a prosecutor and a defendant when informal sanctions by self-interested outside observers may be imposed on either or both.  Non-strategic outside observers rationally use the disposition of the case (plea bargain, case drop, acquittal, or conviction) to impose these sanctions, but also recognize that errors in the legal process (as well as hidden information) means they may misclassify defendants and thereby erroneously impose sanctions on both defendants and prosecutors.  There is a unique equilibrium wherein a guilty defendant accepts the prosecutor’s proposed plea offer with positive (but fractional) probability, an innocent defendant rejects the proposed offer, and the prosecutor takes a defendant who rejects the offer to trial.  Furthermore, we show that: 1) changes in the level of the formal sanction affect the level of informal sanctions imposed by outsiders on defendants and prosecutors; and 2) increases in the informal sanction rate imposed on prosecutors results in changes in the level of informal sanctions imposed on defendants.  The latter case is particularly noteworthy, as (for example) an increase in the rate associated with informally sanctioning prosecutors for convicting the innocent can result in an increase in the level of informal sanctions by third parties on innocent defendants.

We use the base model to examine two extensions of the analysis.  In the first extension, we assume that a fraction of defendants are risk and/or ambiguity averse.  If this response is strong enough, then some innocent defendants accept the prosecutor’s plea offer.  In the second extension, we consider the effect of increasing the informativeness of the jury’s decision by extending the model to allow for a three-outcome verdict (not guilty, not proven, and guilty), sometimes referred to as the “Scottish” verdict.  We find that:

1. guilty defendants are worse off, as plea bargains get tougher but the plea acceptance rate does not change;

2. innocent defendants are better off; 3) the prosecutor’s overall payoff goes up; and 4) the outside observers’ concern over possible misapplication of informal sanctions is reduced.  In this sense, the Scottish verdict is justice-improving when compared with the standard (two-outcome) verdict. Joint work with Andrew F. Daughety.

*****

Talk 2:

Selecting Among Acquitted Defendants: Procedural Choice vs. Selective Compensation
Andrew F. Daughety (Vanderbilt University, United States)

In a previous paper we show how social pressure, in the form of informal sanctions by outside observers, affects choices made by defendants and prosecutors, both with respect to plea offers and to decisions to proceed to trial.  In that paper, we assume that the defendant’s guilt or innocence is his private information, and that evidence at trial is imperfect and may result in errors in conviction or acquittal.  We consider two types of trial verdict: the standard, two-outcome (acquit/convict) verdict and a three-outcome verdict (usually referred to as the Scottish verdict, wherein acquittal is subdivided into not guilty and not proven).  We show that the Scottish verdict raises the expected cost to truly guilty defendants and lowers it for truly innocent defendants.  This, in turn, means that (relative to the standard verdict) the equilibrium plea offer is higher without lowering the likelihood of plea acceptance.  Significantly, the Scottish verdict generates more information for outside observers than occurs under the standard verdict, thereby leading to reduced misapplication of informal sanctions; overall justice is improved.

In this paper we consider two means for implementing the informational benefits of the Scottish verdict.  First, we consider a proposal by Leipold to allow defendants to choose whether they will be tried under the standard or the Scottish verdict.  Regardless of whether this choice is made prior to plea bargaining or just prior to trial, we find that defendants of both types will choose the Scottish verdict, because the choice of the standard verdict leads to an adverse inference of guilt. Thus, this would require a wholesale shift to the Scottish verdict, something that has been previously resisted when proposed in the U.S.  Second, and alternatively, we consider a policy of selective compensation wherein the jury further refines the set of acquitted defendants by designating some, but not all, of them as deserving of compensation.  This is accomplished within the standard verdict institution, but replicates the informational advantages of the Scottish verdict, as there are now three outcomes of the trial:  conviction; acquittal with no compensation; and acquittal with compensation. Joint work with Jennifer Reinganum.