Law, biology, and property: a new theory of the endowment effect.

• Author: Jones, Owen D.

ABSTRACT

Recent work at the intersection of law and behavioral biology has suggested numerous contexts in which legal thinking could benefit by integrating knowledge from behavioral biology. In one of those contexts, behavioral biology may help to provide theoretical foundation for, and potentially increased predictive power concerning, various psychological traits relevant to law. This Article describes an experiment that explores that context.

The paradoxical psychological bias known as the "endowment effect" puzzles economists, skews market behavior, impedes efficient exchange of goods and rights, and thereby poses important problems for law. Although the effect is known to vary widely, there are at present no satisfying explanations for why it manifests when and how it does. Drawing on evolutionary biology, this Article provides a new theory of the endowment effect. Briefly, we hypothesize that the endowment effect is an evolved propensity of humans and, further, that the degree to which an item is evolutionarily relevant will affect the strength of the endowment effect. The theory generates a novel combination of three predictions. These are: (1) the effect is likely to be observable in many other species, including close primate relatives; (2) the prevalence of the effect in other species is likely to vary across items; and (3) the prevalence of the endowment effect will increase or decrease, respectively, with the increasing or decreasing evolutionary salience of the item in question.

The authors tested these predictions in a chimpanzee (Pan troglodytes) experiment, recently published in Current Biology. The data, further explored here, are consistent with each of the three predictions. Consequently, this theory may explain why the endowment effect exists in humans and other species. It may also help both to predict and to explain some of the variability in the effect when it does manifest. And, more broadly, the results of the experiment suggest that combining life science and social science perspectives could lead to a more coherent framework for understanding the wider variety of other cognitive heuristics and biases relevant to law.

TABLE OF CONTENTS INTRODUCTION I. THE ENDOWMENT EFFECT A. What It Is and Why It Matters B. Causes of the Endowment Effect II. AN EVOLUTIONARY HYPOTHESIS A. Law, Behavior, and Behavioral Models B. Applying Behavioral Biology Perspectives to Endowment Effects C. Predictions III. AN EXPERIMENT WITH CHIMPANZEES A. Experimental Design B. Results 1. Group Level Analyses a. Prediction 1 b. Prediction 2 c. Prediction 3 2. Individual Level Analyses a. Prediction 1 b. Prediction 2 c. Prediction 3 C. Discussion of Results IV. POTENTIAL CONFOUNDS, LIMITATIONS, AND RESEARCH QUESTIONS A. Control Conditions 1. Stability of Preferences 2. Effect of Interaction with Experimenter 3. Willingness to Exchange: Potential Role of Distrust B. Some Cautions V. CONNECTIONS A. Connections to Other Endowment Effect Experiments B. Connections to Research on Brain Structure and Function CONCLUSION INTRODUCTION

In a 2005 article, Jones and Goldsmith proposed and illustrated twelve broad contexts in which legal thinking could benefit by integrating knowledge from behavioral biology. (1) In one of those contexts, they argued, behavioral biology can help to provide theoretical foundation for, and potentially increased predictive power concerning, various psychological traits relevant to law. This Article describes the results of one effort to explore that context empirically. Along the way, it attempts to pull various threads of cognitive psychology, neoclassical economics, behavioral economics, evolutionary biology, primatology, and brain imaging into the same analytic space.

Specifically, this Article investigates whether insights from behavioral biology might usefully join with insights from other disciplines to help improve the theoretical foundation for, and predictive power concerning, the "endowment effect." The endowment effect is within a suite of related behavioral phenomena often referred to as reflecting "cognitive heuristics and biases." (2) Not only has the effect proven important to behavioral economists, (3) it also plays a key role in a large and interdisciplinary research program known to many as "behavioral law and economics." (4) That research program adds to (and in some cases challenges) neoclassical law and economics reasoning with various behavioral findings, most of which trace to work by cognitive psychologists.

As described more fully below, the endowment effect is a psychological phenomenon that appears to underlie some seemingly irrational pricing of property and to thereby impede efficient exchange. Because the effect seems inconsistent with standard neoclassical, rational actor, expected-utility theory economic assumptions, a large and growing swarm of articles and working papers--by both traditional and behavioral economists, among others--has explored and often contested its existence, causes, boundaries, and implications. (5) And, because the existence of the effect would suggest that one of the assumptions inherent in the Coase Theorem is rarely if ever true in practice, many legal scholars have raised concerns about the potential implications of the endowment effect for a wide variety of legal approaches to distributing goods and rights. (6)

Part I of this Article provides an overview of the central problem. It provides necessary foundation concerning the endowment effect, describing what it is, how it manifests, why it matters to law, and why there is much at stake. Part I also summarizes some prevailing social science theories for what may cause the effect. It then explains why these theories generally fail to provide a satisfying theoretical foundation for the phenomenon and explains why, as a consequence, we still lack any meaningful ability to predict or explain either when the endowment effect will appear or its relative magnitudes when it does appear.

Part II provides brief background on how life science perspectives might provide useful insight into the endowment effect. It describes why behavioral biology perspectives, in particular (and within that set, a subset of evolutionary biology perspectives), may contribute to a deeper and more scientifically coherent foundation for understanding the phenomenon, as well as provide some increased or expanded predictive power about when and how it may manifest. The Part then describes three novel predictions that arise in combination from a behavioral biology perspective and from no other known causal theory.

Part III then describes an experiment we and several colleagues conducted with chimpanzees (Pan troglodytes) and recently published in a leading biology journal, Current Biology, (7) to investigate those three predictions. We chose chimpanzees as our subjects because our first prediction concerns a possible evolutionary root to the endowment effect, which suggests that our closest primate relatives--with bonobos (Pan paniscus)--might also exhibit it. This Part discusses how the data currently provide evidence consistent with each of the three predictions.

Part IV then reflects on some limitations of the experiment. It considers a few possible confounds that could potentially be inconsistent with our interpretation of the data. And it describes how subsequent research might clarify these issues and extend our understanding of the endowment effect.

Part V considers how the hypotheses explored here, and the underlying theoretical framework they reflect, connect with the existing endowment effect literature, as well as with developments in brain-scanning.

Throughout, it should be remembered that our narrow focus on testing life science perspectives about the endowment effect in chimpanzees is only partly about the endowment effect itself, and the relevance of deeper understanding of that effect to the legal system. Although our work has potential implications for scholars in behavioral economics, primatology, and law, the entirety of our approach here is the exploratory front edge of a potentially promising research program that would--through synthesis of life science and social science perspectives--aid law's efficiency and effectiveness by increasing the accuracy and robustness of the behavioral models on which it necessarily relies.

1. THE ENDOWMENT EFFECT

   1. What It Is and Why It Matters

      As is well known, the Coase Theorem describes how---if several assumptions are true and the initial distribution of property rights is taken as a given--the end distribution in society of goods (and other tradable things, such as options and permits) will be Pareto efficient, regardless of the initial distribution. (8) That is, people will simply buy and sell goods and rights until they end up in the hands of the people who value them most.

      The assumptions include the absence of wealth effects (by which one's demand for a good may vary with personal wealth), the absence of significant transaction costs (which could impede the ability of those who most value something from actually finding, negotiating for, and acquiring it), and the rational valuation of goods and rights.

      Of these three assumptions, we are here concerned with the third. In this context, "rational valuation" means that the value a person ascribes to a good or right will be stable and insensitive to logically irrelevant circumstances. And the word "rational" bears its economic meaning (referring to a substantively efficient choice, regardless of the process deployed to achieve it) rather than to its common, psychological meaning (which often suggests, in contrast, that one is procedurally deploying self-conscious, analytic deliberation). (9)

      In recent years, experiments in cognitive psychology and behavioral economics have given good reason to question whether in fact people always, or even generally, value goods and rights as rationally as both law...