Core 1: Prosecutorial Decision-Making

Prosecutorial Decision-Making: Modeling the Process that Generates Plea Bargains 

Summary. The Prosecutorial Decision Making: Modeling the Process that Generates Plea Bargains core, led by Dr. Anne Piehl, grapples with one of the central puzzles of plea bargaining in the criminal justice context – what motivates prosecutors and judges. There is a rich literature in law and economics on bargaining in a civil context, where both parties are bargaining over money. However, the concerns of the prosecutor and judge are not as obvious in the criminal case, since these actors do not gain anything directly from the defendant. This core focuses on the purposes and institutional factors that drive prosecution using insight from economics, behavioral economics/psychology, sociology and political science. This core also focuses on the difficulties of empirical modeling for a process where samples are always being selected non-randomly from a population as part of the decision process. 

Full Description. This core will emphasize the decision-making of the government, focusing on prosecutors and judges. The theoretical modeling of this core of the RCN will draw together the various approaches that have been taken to examine the motives and behavior of the prosecutors and judges, which has primarily occurred within the fields of law, economics, criminology, and political science. There has been a fairly steady flow of rich formal models of plea bargaining in law and economics (see Polinsky & Shavell 2000 for a review), but this literature can be enriched to reflect the institutional variation analyzed in the more applied literatures, both theoretical and empirical, including legal limits (Stith 2008, Kessler & Piehl 1998), career incentives (Glaeser et al. 2000, Bandyopadhyay & McCannon 2011a, Huber & Gordon 2004), as well as internal policies (such as supervision and promotion policies) within prosecutors’ offices (Bandyopadhyay & McCannon 2011c). In addition, this core will draw upon insights from extensive law and economics literature on civil cases, including the question of why cases ever go to trial. Explaining this “market failure” is important given the near ubiquity of settlements in a world where the rational- choice expected utility model predicts outcomes reasonably well. Concepts of uncertainty and asymmetric information should translate well, but to this point the standard assumption has been that parties are motivated in similar ways as in the civil case. This “prosecution” core will work closely with the other cores on this set of theoretical concepts, as there is clearly overlap, but will take the lead in bringing this particular set of concepts to the larger group (including, for example, reviewing them for the white paper).

Developing creative strategies for testing prosecutorial incentives, as well as building some degree of consensus about what the existing evidence tells us, is a second primary goal of this core of the RCN. Recently, the creation and availability of large, new conviction-only data from jurisdictions with guidelines has created a productive line of research on the behavior of judges and prosecutors, but one that ignores many steps in criminal justice processing (see, among many others, Bushway & Piehl 2001, Payne 1997, Schanzenbach 2005). But to push the research forward, this core will encourage the rigorous empirical modeling of the process of plea bargaining using data that links, and distinguishes between the many stages in the process leading to conviction (plea). While bringing a general appreciation and understanding of identification issues in empirical research, we again will work closely with members of other cores to develop solutions to some of the key methodological challenges facing the empirical examination of plea bargaining using administrative data. One particular substantive goal is to consider how to model and analyze the proliferation of new sentencing options like drug courts, and a rise of collateral consequences like sex offender lists, work restrictions, and deportation in addition to a renewed emphasis on non-incarceration sentences like fines. This range of sentencing options and consequences creates a rich, multi-dimensional bargaining environment that is completely missed by models that emphasize only sentence length. Thus, the concerns of broadening empirical specification and developing new data sources that will allow the linking of arrest and indictment data to final conviction outcome data are additional linkages to facilitate productive engagement across the cores.

For a full list of the RCN steering committee and network members, please visit the Network Members page.