Property and liability rules in the light of bargaining under asymmetric information

Abstract

My thesis examines the efficiency and welfare implications of property rules and liability rules as legal mechanisms for protecting entitlements in scenarios of bargaining under asymmetric information. These rules, which are key in many legal areas like torts, contracts, intellectual property or antitrust law, differ in that property rules require consent of the entitlement holder for transfer, while liability rules allow non-consensual transfers subject to compensation. The choice between these two types of rules has been extensively debated in the law and economics literature. The conventional belief suggests that liability rules are superior when transaction costs are high, and property rules when costs are low. Others even argue that liability rules always enhance welfare by enabling efficient takings, regardless of transaction costs. The novelty of my thesis lies in considering asymmetric information, a significant source of transaction costs that alters bargaining behavior and can lead to inefficient or no trade. My thesis analyzes whether this effect of asymmetric information on bargaining and efficiency differs between property and liability rules and identifies the factors influencing the optimal rule choice. My thesis combines theoretical models and empirical analysis, covering both one-sided and two-sided asymmetric information. It comprises three main chapters, each based on a distinct paper. Chapter 2 begins with a game-theoretic approach, modeling the interaction between an entitlement owner and a potential taker as a bargaining game with one-sided asymmetric information, the owner‘s valuation being private information. Chapter 3 extends this analysis to two-sided asymmetric information, the owner‘s and the taker’s valuation being private information. In both scenarios, I apply two bargaining models where either the potential taker or the owner makes a take-it-or-leave-it offer. This allows for an examination of how the distribution of bar-gaining power influences the outcomes. These theoretical models are then paired with a laboratory experiment involving two-sided asymmetric information. Unlike the formal theoretical analysis, which necessitates restrictive assumptions on the bargaining process, the experimental approach implements a more realistic bargaining protocol permitting players to freely and simultaneously make and accept offers. This experiment, focusing on contract remedies, provides insights into how real-world players would renegotiate a contract under two-sided asymmetric information. Such unstructured bargaining approaches may not be as popular, but they have the potential to inform future theories and test broader predictions that stem from theoretical analysis. Our approach follows this line of thought, aiming to test a more general prediction. We argue that, compared to expectation damages (liability rule), specific performance (property rule) increases the surplus from an agreement when the contract performance has become inefficient. This leads us to the hypothesis that specific performance improves the likelihood of reaching an agreement. The experimental design focuses on individual incentives, removing the contractual context. This is not to deny the importance of normative preconceptions and preferences for promise keeping or contract breaching with compensation. However, it’s beneficial to distinguish various factors influencing the optimal remedy choice. An important assumption throughout my thesis is that the court compensates the owner based on his true valuation if his entitlement is taken. This “variable” liability rule leaves the owner indifferent between retaining the entitlement or being compensated. In contrast, “fixed” liability rules, where damages are set in advance or based on market value, do not make the owner indifferent. While fixed rules are argued to favor bargaining, real-world liability rules are typically variable. For example, in the US, liability for torts follows the make-whole principle, extends to consequential damages, and covers even unforeseeable harm rooted in the plaintiff’s conditions (American Law Institute 2010, §§ 2, 4, 6, 7). My thesis models the law based on its actual content, not its “presumable” ideal form or judicial shortcomings.

The main findings of my thesis are as follows: First, liability rules allowing for unilateral takings against monetary compensation might give them an efficiency advantage over property rules when transaction costs impede consensual transfer of an entitlement. My thesis shows that transaction costs themselves depend on the mode of entitlement protection. The game theoretic model with one-sided asymmetric information, the owner’s valuation being private information, and costly enforcement of compensation, reveals, that only a property rule achieves the first best when the owner of the entitlement has all the bargaining power. The property rule is more efficient than the liability rule when the informed party makes the take-it-or-leave-it offer, as it induces truthful revelation and trade in all efficient cases. In the opposite case with a take-it-or-leave-it offer from the potential taker as the uninformed party, a property rule is more efficient than a liability rule for most parameter values. In that regard it is important to note that my model of one-sided asymmetric information assumes that liability rule administration involves costs. Damage assessment replaces agreed transfers with court-designed transactions. These costs reflect the time, effort, and expense devoted to determining damages, including evidence collection and settlement bargaining to avoid higher litigation costs. Conflict costs burden the liability rule, but the competition between property and liability entitlement protection involves more than a simple tradeoff between these costs and allocative efficiency. Welfare losses result from both misallocation of the entitlement and conflict costs. While liability rules can overcome bargaining impasse, they hamper the exchange of information and raise the cost of voluntary transactions. The liability rule often loses on both counts. Second, the game theoretic model with two-sided asymmetric information reveals that even with high transaction costs due to two-sided asymmetric information, the liability is not superior. This contradicts the conventional belief that liability rules are ideal in scenarios characterized by high transaction costs. Conventional theory recommends liability rules as they allow for unilateral takings of entitlements when parties fail to trade under a property rule. However, the analysis reveals that asymmetric information complicates the efficiency of the rules in a nuanced manner. It shows that liability rules entail negative side-effects, as it results in inefficient takings and less effective bargaining. The model demonstrates that neither rule achieves the first best outcome, as both rules suffer from inefficiencies due to either non-trade or over-trade. The relative performance of the rules depends on the distribution of player types. The liability rule works better than the property rule when the expected benefit of transferring the entitlement is high and certain. Conversely, the property rule works better than the liability rule when the expected benefit of transferring the entitlement is low and uncertain. This is because the property rule allows parties to reveal their type through their offers, which helps to determine if the transfer is efficient. Third, the bargaining experiment with two-sided asymmetric information confirms the theory that both property and liability rules have their benefits. In particular, the experiment supports the theoretical prediction, that property rules facilitate the bargaining process. The analysis reveals that giving the buyer a right to specific performance, which is the equivalent to a property rule, promotes efficient bargaining: The parties more often agree on non-performance, or continue to execute the contract, when it is optimal to do so. The seller’s ability to breach the contract, which is the equivalent to take unilaterally, can correct bargaining failure. But in the experiment the benefits are small and outweighed by losses from inefficient breach and costly conflict. Conflict costs arise if the parties do not find an agreement as both property and liability rule involve equal administration costs in the experiment. The seller is emboldened to take a tougher bargaining position and thereby prevents agreements for which a subsequent breach is only a poor substitute. At the general level, this finding supports the key advantage of property over liability rules: making an agreement strictly necessary for value creation can reduce the transaction cost of reaching it.

The findings of my thesis have wide-ranging implications. They offer valuable perspectives for lawmakers designing remedies in various legal fields, including patent law, contract law, and antitrust law. These insights are also useful for parties negotiating a contract and deciding on appropriate remedies. Beyond conventional legal environments, my results are relevant to other regulatory bodies, such as FIFA, deciding on the penalties to be imposed when a player breaches a contract to move to a different club. This underscores the thesis’s broad relevance and applicability across diverse legal and contractual scenarios.

It is important to note that the analysis is subject to some limitations. The thesis is gradually moving towards more realistic assumptions, from one-sided asymmetric information to two-sided asymmetric information and from extreme bargaining situations in the models to a more open bargaining protocol in the experiment. However, lab negotiation is still far from real high-stakes contracting. Future research should include the normative preferences to the game. Furthermore, the experiment could be replicated with different distributions of types, to test whether the distribution influences the efficiency of the rules as predicted by the game theoretic model of two-sided asymmetric information. Another interesting research question stipulated by this thesis will be to analyze whether private information is related to the variation of remedies in real-world contracts.

Meine Dissertation untersucht die Effizienz und die Wohlfahrtseffekte von Eigentumsregeln („Property rules“) und Haftungsregeln („Liability rules“) als Regelungsmechanismen zum Schutz von Rechten in Verhandlungssituationen unter asymmetrischer Information. Diese Regelungstypen, die in vielen Rechtsbereichen wie dem Deliktsrecht, Vertragsrecht, geistigem Eigentum oder Kartellrecht von zentraler Bedeutung sind, unterscheiden sich darin, dass Eigentumsregeln die Zustimmung des Rechteinhabers zur Übertragung des Rechts erfordern, während Haftungsregeln einseitige Übertragungen bei Zahlung von Schadensersatz zulassen. Die Wahl zwischen diesen beiden Arten von Regeln wurde in der Literatur zu Recht und Ökonomik ausführlich diskutiert. Nach herrschender Ansicht sind Haftungsregeln überlegen, wenn die Transaktionskosten hoch sind, und Eigentumsregeln, wenn die Kosten niedrig sind. Andere argumentieren sogar, dass Haftungsregeln immer die Effizienz fördern, indem sie effiziente Übernahmen ermöglichen, unabhängig von den Transaktionskosten. Die Neuheit meiner Arbeit liegt in der Berücksichtigung asymmetrischer Informationsbeziehungen, einer bedeutenden Quelle von Transaktionskosten, die das Verhandlungsverhalten verändern und zu ineffizienten Rechteübertragungen führen oder diese gänzlich verhindern. In meiner Arbeit untersuche ich, ob sich die Auswirkungen asymmetrischer Informationsbeziehungen auf das Verhandlungsverhalten und das Wohlfahrtsergebnis zwischen Eigentums- und Haftungsregeln unterscheiden und es werden die Faktoren ermittelt, die die Wahl der optimalen Regel zum Schutz von Rechten beeinflussen. Meine Arbeit kombiniert theoretische Modelle und empirische Methoden, die sowohl einseitige als auch zweiseitige Informationsasymmetrien analysieren. Meine Arbeit besteht aus drei Hauptkapiteln, die jeweils auf einem eigenen Aufsatz beruhen. Kapitel 2 beginnt mit einem spieltheoretischen Ansatz, der die Interaktion zwischen einem Rechteinhaber und einem Interessenten als Verhandlungsspiel mit einseitig asymmetrischer Informationsbeziehung modelliert, wobei Wert für den Inhaber eine persönliche Information ist. In Kapitel 3 wird diese Analyse auf eine zweiseitig asymmetrische Informationsbeziehung erweitert, wobei der Wert für den Rechteinhaber und den Interessenten persönliche Informationen sind. In beiden Szenarien wende ich zwei Verhandlungsmodelle an, bei denen entweder der Interessent oder der Rechteinhaber ein Angebot macht, das er annehmen oder ablehnen kann. Auf diese Weise lässt sich untersuchen, wie die Verteilung der Verhandlungsmacht die Ergebnisse beeinflusst.

Die theoretischen Modelle werden mit einem Laborexperiment in Kombination gebracht, bei dem es um zweiseitig asymmetrische Informationsbeziehungen geht. Im Gegensatz zur formalen theoretischen Analyse, die restriktive Annahmen über den Verhandlungsprozess erfordert, ermöglicht der experimentelle Ansatz ein realistischeres Verhandlungsprotokoll, das es den Spielern erlaubt, frei und gleichzeitig Angebote zu machen oder zu akzeptieren. Dieses Experiment, das sich auf vertragliche Rechtsmittel zum Schutz von vertraglichen Ansprüchen konzentriert, bietet Einblicke, wie Parteien in der realen Welt einen Vertrag bei zweiseitig asymmetrischer Informationsbeziehung nachverhandeln würden. Solche unstrukturierten Verhandlungsansätze sind nicht populär, aber sie haben das Potenzial, künftige Theorien zu untermauern und allgemeinere Vorhersagen zu testen, die sich aus der theoretischen Analyse ergeben. Unser Ansatz folgt diesem Gedankengang und zielt darauf ab, eine allgemeinere Vorhersage zu testen. Wir argumentieren, dass im Vergleich zum Schadenersatzanspruch (entspricht einer Haftungsregel) der Naturalerfüllungsanspruch (entspricht einer Eigentumsregel) den Mehrwert einer Vereinbarung erhöht, wenn die Vertragserfüllung ineffizient geworden ist. Dies führt uns zu der Hypothese, dass der Anspruch auf Naturalerfüllung die Wahrscheinlichkeit einer Einigung erhöht. Das Design des Experiments setzt den Fokus auf individuelle Anreize und lässt den vertraglichen Kontext außer Acht. Damit soll nicht zum Ausdruck gebracht werden, dass es keine normativen Meinungen und Präferenzen über die Einhaltung von Versprechen oder Vertragsbrüche gäbe. Es ist jedoch von Vorteil, zwischen verschiedenen Faktoren, die die Wahl der optimalen Regelung zum Schutz von Rechten beeinflussen, zu unterscheiden. Eine wichtige Annahme, die sich durch meine gesamte Arbeit zieht, ist, dass das Gericht dem Rechtinhaber Schadensersatz in Höhe seiner persönlichen Bewertung des Rechts zuspricht, wenn ihm sein Recht entzogen wird. Bei dieser "variablen" Haftungsregel ist der Inhaber indifferent zwischen der Inhaberschaft des Rechts oder des Schadensersatzes. Im Gegensatz dazu wird der Rechteinhaber bei "fixierten" Haftungsregeln, das sind solche, bei denen der Schadenersatz beispielsweise im Voraus festgelegt wurde oder auf dem Marktwert basiert, nicht vollständig für den Entzug des Rechts entschädigt. Es wird behauptet, dass feste Regeln Verhandlungen begünstigen, aber in der Praxis sind Haftungsregeln in den meisten Fällen variabel gestaltet. In den USA beispielsweise folgt die Haftung für unerlaubte Handlungen dem Grundsatz der Wiedergutmachung, erstreckt sich auf Folgeschäden und deckt sogar unvorhersehbare Schäden ab, die in den Umständen des Klägers begründet sind (American Law Institute 2010, §§ 2, 4, 6, 7). Meine Arbeit modelliert das Recht auf der Grundlage seines tatsächlichen Inhalts, nicht auf der Grundlage einer "mutmaßlichen" idealen Form oder Grenzen der Judikatur.