Stare decisis and judicial log‐rolls: a gains‐from‐trade model

• Author: Giri Parameswaran,Lewis A. Kornhauser,Charles M. Cameron
• Date: 01 September 2019
• Published date: 01 September 2019
• DOI: http://doi.org/10.1111/1756-2171.12288

RAND Journal of Economics

Vol.50, No. 3, Fall 2019

pp. 505–531

Stare decisis and judicial log-rolls: a

gains-from-trade model

Charles M. Cameron∗

Lewis A. Kornhauser∗∗

and

Giri Parameswaran∗∗∗

The practice of horizontal stare decisis requires that judges occasionally decide cases “incor-

rectly.” What sustains this practice? Given a heterogeneous bench, we show that the increasing

differences in dispositional value property of preferences generates gains when judges trade

dispositions over the case-space. These gains are fully realized by implementing a compromise

rule—stare decisis. Absent commitment, we provide conditions that sustain the compromise in

a repeated game. When complete compromises become unsustainable, partial compromises still

avail. Moreover, judges may prefer to implement partial compromises even when perfect ones

are sustainable. Thus, stare decisis is consistent with a partially settled, partially contested

legal doctrine.

1. Introduction

A desideratum of a well-functioning judiciary is that courts decide like cases similarly. In

the common law tradition, the body of existing case law (i.e., “precedent”) constitutes a source

of law, and under the doctrine of stare decisis,1courts should decide instant cases following the

results or principles laid down in prior cases. Indeed, many areas of law are considered “settled”

and cases are decided consistently, notwithstanding the heterogeneity in judges’ beliefs about

how they ought to decide cases. However, in other areas of law, there is less agreement between

courts, and the same case may be decided differently, depending on which judge is presiding. To

give just a few examples, courts continue to disagree about what constitutes an “offer” or how

∗Princeton University; ccameron@princeton.edu.

∗∗New York University School of Law;lewis.kornhauser@nyu.edu.

∗∗∗Haverford College; gparames@haverford.edu.

Wethank three anonymous referees and participants at workshops at ToulouseSchool of Economics, New YorkUniversity

Law School, the 2015 Conference on Institutions and Law-Makingat Emor y University, and the 2015 American Lawand

Economics Association Meetings.

1The term “stare decisis” derives from the phrase “Stare decisis et non quieta movere,”an injunction “to stand by

decisions and not to disturb the calm.” For a brief discussion of stare decisis, see Kornhauser (1998). Formore extended

discussions, see Cross and Harris (1991), Duxbury (2008), and Levi (2013).

C2019, The RAND Corporation. 505

506 / THE RAND JOURNAL OF ECONOMICS

to determine “unconscionability” in contract law, how to assess “materiality” in securities law,

or what constitutes “probable cause” in the context of search and seizure.2These disagreements

persist even after hundreds of cases have been decided. In these contexts, the law resembles a

“standard” in which judges exercise discretion over how cases are decided within some bounds,

rather than a “rule” which prescribes how the case should be decided.

In this article, we ask why heterogeneous judges agree to decide cases according to a

consistent rule in some areas of law, but leave the law unsettled in others. Our analysis begins

with the recognition that, even when a norm of horizontal stare decisis3exists, nothing compels

courts to follow precedent. The US judiciary is a heterogeneous body,comprised of judges having

different interpretations and conceptions of the law.In principle, each judge may dispose of cases

as she sees fit. The practice of horizontal stare decisis, then, is voluntary; it reflects a tacit

agreement among judges to decide cases consistently,even if doing so requires them occasionally

to decide cases contrary to their ideal disposition.

Wedevelop a theory that explains why judges may voluntarily apply a common compromise

rule when they are free to decide cases by their own lights. We make twoimportant contrib utions.

First, we showthat cooperation between judges is a natural consequence of a simple and plausible

feature of judicial preferences, which we call the increasing differences in dispositional value

(IDID) property. We demonstrate that such preferences create opportunities for gains-from-trade,

and that these gains are maximized when judges settle on applying a common legal rule to decide

cases. Second, weintroduce the notion of par tial staredecisis, which is novel in the literature. We

show that commitment problems may preclude judges from fully realizing all gains-from-trade,

but that partial gains may nevertheless be realized. This providesa novel account for the sustained

existence of areas of law that are partially settled and partially contested.

To study the political logic of stare decisis, we analyze a model in which a heterogeneous

court decides an infinite stream of cases that present a common issue. A case is summarized by a

number on the real line that captures salient facts, for example, the speed at which the defendant

was driving. Each case is decided by a single judge selected at random from the bench.4Judges

havepreferences over case dispositions characterized by a threshold that separates cases that would

ideally be decided for the plaintiff from those that would ideally be decided for the defendant.

Importantly, we assume that these preferences satisfy the increasing differences in dispositional

value (IDID) property, which states that the utility differential from deciding cases “correctly”

rather than not, increases the further the case is from the threshold; it is more costly to incorrectly

decide “clear-cut” cases than “contestable” ones. We introduce heterogeneity into the model by

assuming that judges are drawn from one of two factions (“L” and “R”), which are distinguished

by the location of the threshold case. Although judges may only decide the cases before them,

they receive utility from the disposition of every case, regardless of which judge decided it.

These simple and reasonable assumptions on judicial preferences are sufficient to provide

incentives for judges to adhere to compromise rules. Given preference heterogeneity, there will

be a set of cases (the “conflict zone”) in which, no matter how the case is decided, some judge(s)

will be disappointed. The IDID property implies that the relative cost of an adverse decision

varies overthe set; the cost to L-faction judges of a wrongly decided case close to the L-threshold

is small relative to the cost of a wrongly decided case close to the R-threshold (far from the

L-threshold). The opposite is true for R-faction judges. This creates an oppor tunity for gains-

from-trade. Although each judge could dispose of cases according to her ideal, there are mutual

gains from judges instead implicitly agreeing to dispose of cases in a way that assigns the losses

2Wethank an anonymous referee for suggesting these examples.

3Horizontal stare decisis refers to a court’spractice of “respecting” its prior decisions. Vertical stare decisis,by

contrast, refers to the practice of following the decisions of superior courts in a hierarchical judicature. In this article, we

focus on horizontal stare decisis, where such a hierarchy is absent.

4Wethus offer a stylized model of a court in which a panel of judges is drawn from a wider bench. The practice of

drawing a panel from a bench is common on intermediate courts of appeal in most legal systems. Many supreme courts

and many international courts have a similar practice.

C

The RAND Corporation 2019.