ECONOMIC ANALYSIS GROUP
DISCUSSION PAPER
Criminal Plea Bargaining with Uncertainty:
An Empirical Model of Sentencing
by
Luke Froeb*
BAG 88-4 March 14, 1988
‘The Economic Analysis Group (FAG), formerly the Economic Policy Office,
consists of the Economic Litigation and Regulatory Economic Sections of the
Antitrust Division. EAG Discussion Papers are the primary vehicle used to
disseminate research and comment from both sections. These papers are
intended to inform interested individuals and institutions of EAG's research
program and to stimulate comment and criticism on economic issues related to
antitrust policy.
This series of papers is open to a wide range of material including
applied theoretical work, empirical research, policy analysis, and reviews of
contemporary publications. In addition to the broad scope of material, titles
in the EAG Discussion Papers will include both working papers and completed
research.
The analysis and conclusions of BAG Discussion Papers represent those of
Lae quuuuis end dy wut uecessarily seflect Lue views uf Ube Depa: tment uf
Justice. Information on the BAG research program and single copies of EAG
Discussion Papers may be obtained by writing to Albert Smiley, Director
ff Research. FAG. Antitrust Division. U.S. Department. of Justice. Judiciary
Center Building, Room 11120, 555 4th Street, NW, Washington, D.C. 20001.
Comments on specific papers may be addressed directly to the authors at the
same address. If a published version of the Discussion Paper is available, it
is noted on the list at the end of this report so that the final version of
the paper can be more easily obtained.Abstract:
Plea bargaining is modelled as a game between a prosecutor and a defendant. The
prosecutor makes a take-it-or-leave-it plea offer to a defendant but does not know
whietler the uffer will be avvepteu. In making an offer, the prosecutor faces a tradeott
between a higher plea offer and a greater likelihood that it will be rejected. All rejected
plea offers result in trials so that the prosecutor must try to avoid being caught with too
many trials and not enough attorneys. The model implies that smaller plea discounts are
offered to defendants that commit more serious crimes and that more of these defendants
reject the plea offer. This implication is supported by individual sentencing data.
However, there is no evidence that plea discounts are more likely to be rejected in
districts with lighter caseloads, also as predicted by the model.The importance and prevalence of plea baraaining to the processing of criminal
cases has motivated a number of empirical studies. All of the studies report findings
consistent with a model of plea bargaining where trials result when tha prosecutor and
defendant cannot agree on a sentence. For example, each could be overly optimistic
about the chances of winning at trial ar the defendant could be less risk averse (more
prone to "roll the dice") than the prosecutor. In these cases, the smallest plea offer that a
Prosecutor would accept is above the largest plea offer that a defendant would auvept.1
The model that follows takes a different view of the cause of trials, Trials result
due to uncertainty on the part of the pruseuutu! about the size of a plea otter that a
particular defendant would be willing to accept. Occasionally, the prosecutor makes
offers thal are rejected because they are too high. Then a trial occurs. Following earlier
approaches, the prosecutor is modelled with a caseload large enough and a staff small
enough, so that all cases could not be tried in court. Consequently the size of a plea offer
serves as a rationing device for trials (the "price" of a trial) that equates the demand for
trials to the supply of trials. The demand for trials is determined by the size of the offered
pleas and the supply of trials is determined by the size of the prosecutor's staff
In making plea offers to a group of defendants, the prosecutor has two concerns:
first, to put as many defendants in prison for as long as the law allows; and second, to
avoid being caught with too many trials and not enough attorneys. If the prosecutor could
observe the highest plea that a defendant would accept (the "reservation" plea), then it
would be possible to capture all of the bargaining rents from the defendant by making a
take-it-or-leave-it offer just below the reservation plea. In this world all trials would bo
“Twish to acknowledge excellent research assistance from Rob French and useful
discussions with Michaol Blook, Jon Joyce, Bruce Kobayashi, Preston MvAlee, Bill
Rhodes, Steve Salop, Marilyn Simon, Bert Smiley, Rick Warren-Boulton and Greg
Werden. In addition to the usual disclaimer, the views expressed in this paper do not
necessarily reflect the views of tle U.S. Department of Justice or U.S. Sentencing
Commission.
1. Soc for example, Brian Forst and Kallierine Brosi, A Theoretical and Empirical
Analysis of the Prosecutor, 6 The Journal of Legal Studies, 488 (1977). William Landes,
‘An Economic Analysis of the Courts, 14 Journal of Law and Economies, 61 (1971).
William Rhodes, The Economics of Criminal Courts: A Theoretical and Empirical
Investigation, 5 The Journal of Legal Studies, 311 (1976). David Weimer, Plea
bargaining and the Decision to go to Trial: the Application of a Rational Choice Model.10
Policy Sciences, 1 (19/8).
Ralston and Simmons, Attorneys at Law, Jurisdiction of The Court of Claims: Considered With Relation To Its Departmental, Congressional, and Special Powers (1905) .
Henry John Stephen, A Treatise On The Principles of Pleading in Civil Actions: Comprising A Summary View of The Whole Proceedings in A Suit at Law (1867)