Bias in International Law

Abstract

This special issue looks at how cognitive bias matters to international law. We wish to shed light on the legal frames, labels, and cognitive biases that shape our understanding of international rules, the application of these rules, and outcomes of international adjudicatory processes. Adopting the behavioral approach to international law, we focus on actual behavior rather than assumed behavior of actors taking part in the international legal process. The central idea of this approach is that human cognitive capacities are limited—or bounded—by a variety of cognitive, emotional and social, or group-based biases. Our aim is to explore how these biases operate on the individual, group, and state level in various spheres of international law. This Symposium therefore looks beyond the traditional understanding of international law as applying between states, and focuses on how individuals, as actors in the international sphere, use international law language to influence other people, to create communities, and to shape identities.

This Introduction first serves to explain the type of shortcuts we make in our decision-making. This description of biases is followed by an overview of behavioral literature in international law that has thus far examined how bias operates in different aspects of international law—in relation to sources, to compliance, and individuals taking part in the international legal process. We then turn to introduce the Symposium and explain its contribution to the existing literature.

A. Introduction

The “behavioural movement has become one of the most influential developments in legal scholarship in general“ in recent years.¹ Instead of assuming how actors taking part in the international legal process behave, the focus of inquiry has shifted to analyze their actual behavior. The central idea of this approach is that human cognitive capacities are limited—or bounded—by a variety of cognitive, emotional and social, or group-based biases.² These shape our understanding of international rules, affect the manner in which we apply these rules, and influence outcomes of international adjudicatory processes. In this Symposium, our aim is to flesh out how these biases operate on the individual, group, and state level in various spheres of international law. In this regard, the issue builds on the extensive work in behavioural analysis of two psychologists, Tversky and Kahneman, who radically revolutionized cognitive and social psychology by revealing the common human errors that arise from human decision-making.³ Tversky and Kahneman found that human behaviour is not always purely rational, but could be influenced by cognitive biases—shortcuts made by humans in judgment and decision-making that lead to suboptimal decisions. Since then, a growing legal scholarship of experimental and empirical studies has shown that people’s decisions are influenced by irrelevant information, by how the information is presented, and who is involved in the process.⁴ Scholars have also shown that people make common sense moral decisions,⁵ and that they are motivated by fairness and altruism.⁶ Throughout the world, the behavioral movement and the insights it has generated have influenced legal policymaking and regulation.

This Symposium therefore looks beyond the traditional understanding of international law as applying between states, but instead focus on individuals as actors in the international sphere. Specifically, the Issue explores several aspects in which international law frames our decisions and triggers individual cognitive biases and emotional and moral responses. From international trade law to human rights, our authors look at how legal language potentially affects social attitudes, preferences, and policy. Through large computational studies and experiments, as well as theoretical interventions, we see the multitude of ways in which international law is infused in our lives and how it affects our decisions and behaviour. The Symposium shows that apparently neutral international norms may not be neutral in their application, that cognitive errors affect our understanding and drafting of these norms, and that this influences how we understand and apply international rules. In addition, the Symposium looks closely at how different international law actors operate in their communities and the specific heuristics and biases that are born out of operation in those communities. In this regard, the articles in this Symposium seek to broaden the scope of addressees of international law.

In many respects, the Symposium is a call to a different study of international law. It wishes to draw attention to the hidden and the unsaid, but the ever present–to our bias. It seeks to lift the veil on how we—individuals, groups and states—actually behave in response to international law norms and how we use international law to influence other people, create communities, and shape identities.

This Introduction proceeds in three parts: In the next section, we explain the type of shortcuts we make in our decision-making. This description of biases is followed by an overview of behavioural literature in international law that has thus far examined how bias operates in different aspects of international law–in relation to sources, to compliance, to individuals taking part in the international legal process. We then turn to introduce the Symposium and explain its contribution to the existing literature.

B. What is Bias?

Traditional approaches to law, so-called rational choice theory, assume that people are rational and know what they want, and are able to rank different choices according to the utility they derive from them.⁷ But as early as in the mid-1950s, Herbert Simon introduced the concept of bounded rationality. Bounded rationality recognizes that we have limited cognitive capacities. As such, we do not always behave fully rationally.⁸ According to Simon, our bounded rationality shows through when a problem is sufficiently difficult,⁹ leading us to choose solutions that are “good enough.”¹⁰ It was in the early 1970s that Kahneman and Tversky pushed the boundaries of bounded rationality further by showing that even simple problems can induce suboptimal choices.¹¹ They argued that in addition to making “good enough” decisions, people predominantly rely on simple, intuitive, rule of thumb, decision-making. When making decisions, Kahneman and Tversky showed that people rely on a limited number of mental shortcuts—also known as heuristics—which reduce the complex tasks of assessing probabilities and predicting values to simpler judgmental operations.¹² In general, these shortcuts are quite useful: because they come about automatically, they allow people to allocate their mental energy elsewhere. However, Kahneman and Tversky showed that the use of heuristics may also lead to severe and systematic errors, known as motivational and cognitive biases.¹³ In the next two sections, we introduce these two types of errors and explain how they affect our decision-making.

I. Motivational Biases

Motivational biases reveal a tendency to form and hold beliefs that serve the individual’s needs and desires.¹⁴ On the one hand, people essentially read situations and information through the lens of what suits them by avoiding drawing conclusions they would find uncomfortable, and in contrast, accepting those conclusions that they would find pleasing. Individuals are, for example, motivated to enhance and protect their egos.¹⁵ In this context, they will readily accept credit for their own success, attributing achievements to their internal factors such as intelligence, ability, and determination to succeed. On the other hand, individuals often refuse to accept blame for failure and will assign responsibility for any disappointments to factors that are external to them.¹⁶ These self-serving biases reveal that we often perceive reality in an ego-protecting and ego-enhancing manner, interpreting our own behavior, judgments, and choices as relatively common and appropriate and attributing—positive—characteristic and traits to ourselves, our friends and our reference group, while avoiding attributing negative characteristics to these actors.¹⁷ In contrast, we view alternative responses as uncommon, deviant, and inappropriate.¹⁸ In many ways, therefore, motivational biases help us maintain a favorable image of ourselves as “wise and judicious, as persons whose judgments and behaviors are likely to be met by wide approval.”¹⁹ Other explanations of motivational biases move away from ego-centric interpretations and underline that we emphasize internal factors because these help us explain events in our lives through the lens of effective control. That is, they help us view decisions and events as being a result of elements over which we exercise “effective control,” like work, ambition, and effort.²⁰ According to Kelley, our inherent desire to exercise effective control over our lives may bias individuals to attribute events to controllable factors, rather than to factors over which we have no control.

II. Cognitive Biases

While motivational biases arise from “wishful thinking” and serve to accommodate our self-serving motives such as a need for self-enhancement or effective control, cognitive biases produce systematic errors of judgment without serving any hidden needs or desires. Instead, they reflect deviations from rational decisions that occur because of our limited capacity to understand and properly process all the information that is potentially available to us. Scholars have written how we frequently make quick and intuitive decisions, using shortcuts or judgmental strategies, which many consider suboptimal.²¹ Generally, these strategies are assumed to direct one’s attention to certain types of irrelevant information and lead them to ignore other types of relevant information.

According to Hastie and Rasinski, cognitive biases can be categorized into three broad classes of systematic errors in judgment. These are: sins of imprecision, sins of commission, and sins of omission.²² The first group relates to the discrepancy between criteria we are supposed to apply and the reality of human judgment. Studies have shown that as individuals we rarely alter our own subjective probability judgments in response to new diagnostic information. Instead, we remain over-confident in the accuracy of our own assessment, making shortcuts that overestimate the likelihood of an event or its frequency not by investigating statistics but based on how many similar instances are brought to mind—availability heuristic—²³or focusing on items that are more prominent or emotionally striking and ignoring those that are unremarkable—salience bias—and believing that multiple events can co-occur at the same time—conjunction fallacy—even though this is unlikely or irrelevant by objective standards. Similarly, we adjust out decisions depending on how outcomes are framed. Kahneman and Tversky have shown that we have a clear preference for uncertain loss—compared to certain loss—but also a preference for certain gain—compared to uncertain gain. How things are framed or ‘labelled’ therefore shapes our perception of them and how we react to them. In the context of prospect theory, we are risk-seeking when outcomes are framed as losses but risk-averse when outcomes are frames as gains.

The second group – the sins of commission – note that although we are required to treat certain information as irrelevant for judgment—such as race of a victim or attractiveness of a defendant—we nevertheless take it into account and use it to affect our assessment of guilt or innocence or other decision-making. Similarly, we remain overreliant on information that is or ought to be irrelevant to our judgments, including for example the belief that past investments justify further expenditures—sunk cost fallacy—or we rely on reference points as “anchors”—anchoring effect. We also tend to interpret, favor or recall information in a manner that confirms our beliefs or values and thus expresses preference when none should be had—confirmation bias.

The third group – the sins of omission – reveal that biases can also have the opposite effect. In certain situations, we fail to use information that is relevant to inform our judgments, for example, by sticking to the status quo and perceiving any change from this baseline as a loss—status quo bias—or we underemphasize situational and environmental explanations for an individual’s observed behaviour while over-emphasizing personality-based explanations—correspondence bias. In short, we tend to ‘believe that what people do reflects who they are’.

C. Bias in International Law

While biases are referred to as “errors” that we make in our judgments and are thus labelled as deviations from rational behavior, the terminology conceals the fact that these errors influence our behavior and decision-making and often affect our treatment of others. From a legal perspective, therefore, biases are not only relevant in terms of how we perceive, interpret, and approach the law but also in how they affect our treatment of other actors in the legal system. In this section we therefore focus on how cognitive biases matter to international law. Specifically, we look at how biases shape our decisions pertaining to the legal process, how they affect the choices that we make and thus potentially influence outcomes, how they affect our reading of legal norms and practices, and then finally we turn to how biases affect our application of the law and how this might affect our treatment of others.

II. Studying Bias in International Law – Specifying the Unit of Analysis

One of the main questions that arises in the context of studying bias in international law is—what is our unit of analysis? Or in other words, whose behavior do we study? Is the center of our inquiry the state as a unitary subject and thus do we assign biases to a “black box,” regardless of how it is set up and what its internal workings are? Or do we study the bias of individual actors who appear in the international legal process, such as judges, governmental officials, trade negotiators, other “elite” decision-makers or are we concerned with the general public and for example, study the bias of victims of human rights abuses, etc.? As other scholars have shown,“there is no methodological challenge if individual behavior is attributed to the state under international law.”²⁴ Because bias in individuals has been extensively studied, attributing such bias to individuals appearing in the international legal process is not problematic. But in international law, decisions are often taken at other levels. Often decisions are a result of small group decision-making and group dynamics. But group psychology “is often different from individual decision-making.”²⁵ As studies have suggested, individual biases within groups can become enhanced or diluted and so the picture becomes more complex and complicated.²⁶ In this context, a careful empirical reconstruction and analysis of networks and communities that practitioners and officials form can illuminate the sociology of specific institutions and even areas of law, ²⁷ while experiments can help confirm the operation of individual bias at group level as well as the formation of a separate “group think” bias, which pressures group members towards conformity and cohesiveness.²⁸

Beyond studying individuals and groups, international law tends to focus on the study of behavior of the state, as the unit of analysis. This, however, presents specific methodological challenges. As Anne van Aaken argues,²⁹ there are three possible approaches to studying state behaviour: (1) the study of the state as an organization, where psychological and behavioural insights have already been successfully applied;³⁰ (2) the study of the relationship between the individual—citizen—and the politicians or officials; this relationship focuses on how power is exercised—“the two level game”—³¹but thus far only a few studies have looked at the operation of cognitive biases in this context;³² and (3) the study of the state as analogous to the individual, where in order to reduce complexity cognitive biases, preferences and beliefs assigned to the individual are assigned to the state and the state is assumed to succumb to the same biases as individuals.³³

Each of the choices comes with its own set of methodological challenges and limitations. We have explored these in a separate Symposium on the Limitations of the Behavioural Approach in the American Journal of International Law Unbound.³⁴ In this Symposium, Yildiz and Yüksel expand on these limitations. In their contribution, they draw on rational and behavioral approaches to formulate different expectations about the process of policymaking and updating in the context of maritime delimitation. Undertaking a quantitative analysis of how states make and update their policies in light of the decisions of the International Court of Justice, they highlight how, when studying the behavior of states, it is not always possible to conclude whether certain behaviour of the ‘unit’ is motivated by rational choice behavior—self-interest—or is due to cognitive bias. Their contribution reconfirms the need for behavioral scholars to make use of additional experimental research methods to be able to establish whether—in a given situation—the behavior of the state as a unit is indeed biased.

II. The Operation of Bias in Selected International Law Topics

In this section, we provide an overview of existing studies in international law, which have revealed the operation of cognitive bias in decision-making at individual, group, and state level.³⁵ Within each topic, we explain how the contributions in this Symposium contribute to the existing literature.

D. Sources of International Law

International scholars and political scientists have for decades sought to answer the question of why states agree to be bound by certain sources of international law or indeed why they sign treaties that might be hurtful to them. In her seminal article on Behavioural International Law, Anne van Aaken maps out how the the sources of international law are set up in a manner that provides a very specific choice architecture if observed from a behavioral perspective.³⁶ Specifically, she notes that treaties which are written and impose specific international rules on states act as an opt-in mechanism, with countries being required to provide their consent explicitly to be considered bound by the rules. In contrast, customary international law are unwritten and binding, but allow states to opt-out from them through the mechanism of a persistent objector. Only jus cogens norms do not allow for exceptions and can thus not be opted-out of. This different perspective of the basic sources of international law shifts our view by underlining the behavioral, psychological impact that the sources may have on the addressee.

Rational choice scholars argue that when states choose between the different forms of legal obligations, they make their choices between the different types of binding obligations—soft v. Hard law—or by adjusting the degree of vagueness—unwritten to written—depending on the flexibility they require. The choice, they argue, is therefore based on state’s interests. But these perspectives fail to take into account the operation of the default rule mechanisms. From a behavioral perspective, the default bias has considerable impact on the behavior of actors pushing states to stay with the default choice and choose inaction over action. This could suggest why there are so few persistent objectors to customary rules and why states rarely violate jus cogens—not only would it be costly for them, but it would go against the current choice architecture. The contrary is the case for treaties, which are opt-in mechanisms. For these, states will often negotiate at length and may also—especially when text is fixed and non-negotiable as in the case of boiler-plate treaties³⁷—resort to reservations to actively shape obligations they are to be bound by.

Jean Galbraith adds to this discussion by studying the behavior of states concerning treaty reservations.³⁸ Drawing on a treaty database, she reveals a high correlation between framing and state consent: where treaties allow states to “reserve out of” ICJ jurisdiction, 95% continue to accept the said jurisdiction; and where treaties give states the explicit right to opt out of ICJ jurisdiction, 80% continue to accept it. In contrast, where states have to opt into ICJ jurisdiction, only a mere 5% of state parties do so on average.³⁹ This result confirms the important effect of default rules, even when states’ behavior is at play.

Galbraith also shows that the operation of the default rule applies in case of optional compliance mechanisms in human rights treaties—for example the jurisdiction of UN bodies under the International Covenants. These can either be formed as optional opt-in clauses in the main text or instead offered through optional protocols. Galbraith shows that depending on how the option is framed, states will react in different manners. She confirms again that states have proved much more willing to ratify commitments when they are presented in optional protocols—that is, documents separate from the main treaty—than when these are presented as opt-in commitments in the main document. She argues that states sign the optional protocol due to salience bias, in other words, the tendency to focus on information that is more noteworthy—for example the signing of the main treaty—while ignoring that which does not grab our attention.

Galbraith’s findings have important implications for the negotiation of any new international agreement, including climate change rules. As van Aaken argues: “using opt-out framing in treaties is preferable to opt-in if universality and large participation is desired. If there is a need to use an opt-in form it should be framed as an optional protocol, whenever possible.”⁴⁰

Another important question that is addressed in the literature concerns not so much how treaties are framed, but rather why they are signed and ratified by states in the first place. In his book Bounded Rationality, Lauge Poulsen asks the question why developing countries tend to sign and ratify modern bilateral investment treaties, although this is not always in their interest.⁴¹ Rational choice theorists argue that the reason for doing this lies in the wish to make oneself more attractive for investment relative to other host countries.⁴² Building on extensive quantitative and qualitative research, Poulsen finds that the bounded rationality framework has considerable potential to better explain the actions of these states. Poulsen suggests that states are “constrained, not just by the complexity of their environment but also by limitations to their own problem-solving capabilities.”⁴³ He suggests that developing countries were encouraged to believe in the inflated promises about the benefits of signing investment agreements. State officials and lawyers were“motivated optimists,”⁴⁴ believing that signing the treaties would have a positive effect on investment. But even if they believed that they were acting rationally in pursuit of investments, the behaviour they exhibited was “predictably irrational”⁴⁵ Many interviewees reveal “they did not have a clue“ and did not “really know that treaties had any bite.”⁴⁶ Had they sought more expertise or more information, or indeed looked to disputes arising under the North American Free Trade Agreement, the decisions coming out of those dispute settlement bodies could have served as a warning. Instead, state officials remained ignorant of the low probability, high risk potential that an investment treaty claim might hit them, until it did.⁴⁷ The salience bias and neglect of probability, a cognitive error where individuals systematically neglect events with low probabilities—especially if they are not salient—therefore heavily influenced state’s‘ decision-making.

Stepping away from treaties, behavioralists have also explored why states may be actively contributing to the formation of opinio juris in the development of customary law. In this regard, the traditional argument made by those espousing the rational choice theory is that states will participate in the formation of opinio juris when it is in their interest. In contrast, behavioral scholars have shown that states may make their decisions based on what they think they are expected to contribute and thus may be acting in response to their prior experiences: “If they had contributed less than the average and if they react by increasing their contributions, we can interpret this as an adjustment to the perceived expectation of others to make a higher contribution.”⁴⁸ Engel concludes that states may interpret the contribution level as a normative expectation.⁴⁹ This motivational bias—or how states want to be seen—potentially motivates state behaviour and goes against rational choice view that states will not contribute to the public good.

In the context of looking at sources of international law, international scholars have also attempted to study how we “read” international law norms. International law, like other areas of law, may appear neutral, but our biases may affect our evaluation, interpretation and understanding of these norms. The framing of international law rules—how a norm is phrased or how the information is presented to us—may affect our perception of norms, or indeed of groups and individuals to whom they are applied. In this Symposium, Benedikt Pirker and Iza Skoczen investigate to what extent a neutral looking provision of an international treaty can imply more than its ordinary meaning. Following the customary rule of interpretation found in the Vienna Convention on the Law of Treaties, which states that treaties shall be interpreted according to the ordinary meaning given to the terms of the treaty, Pirker and Skozen explore what this means in practice. They uncover our need to fill gaps in interpretation and to do so by having recourse to moral principles.

E. State Compliance

The next area in which behavioral approaches are being increasingly applied is state compliance.⁵⁰ Here, scholars are addressing two questions which have kept both international lawyers and international relations scholars busy for decades: why do states comply with international law? And can international norms be designed more efficiently to change states’ behavior?

As in other areas, behavioral scholars underline the gap between what is said to motivate states and what actually influences their behavior. For example, rational choice scholars argue that reciprocity acts a central mechanism of compliance.⁵¹ According to this approach, states decide whether to comply or not depending on whether the benefits for future cooperation outweigh the costs of reciprocal non-compliance by others.⁵² Yet, such reciprocal action depends on those states’ perception of the initial deviation from the norm. As van Aaken argues, “reciprocity as a reaction to the violation of other actors is often based on whether an action is perceived as kind or unkind.”⁵³ When there are strong links between countries and the perception of the other state is as a beneficent actor, non-compliant behavior—such as in cases of technical or capacity problems—is unlikely to be viewed as intentional. In contrast, if violation of international rules is seen as intentional, unfair and “unkind,” then “stronger reactions are expected.”⁵⁴

Similarly, reputation has long been characterized as the “most talked-about, yet least understood strategic phenomenon in international relations.”⁵⁵ Reputation is important for compliance scholars across the spectrum, from those that adopt a rational choice approach to constructivists. The former argue that the violation of international law results in a bad reputation for a state, which in turn leads other states to exclude it from cooperation within their own group. Accordingly, states comply with international law—at least partially—in order to be able to benefit from the gains of future cooperative activity. From a constructivist perspective, reputation as a law-abiding state may be important because it represents membership of a group of like-minded actors.⁵⁶ But the operation of reputation depends on a state’s perception of how others view it. It is in this context that bias could be in operation. Loss aversion, for example, could explain why states “are more concerned with preventing a decline in their reputation or credibility than increasing it . . . [In addition,] prospect theory also suggests that those effects would be even stronger if the threat of loss of reputation were perceived to be certain.”⁵⁷

The operation of biases occurs not only at the international level, but also at the individual level. Experiments with elite decision-makers and with the general population have, for example, shown that individuals are far more likely to oppose policies that would violate international law than otherwise identical policies that would not.⁵⁸ In this Symposium, Jonathan Kolieb contributes to this literature by investigating what consumer attitudes and perception can tell us about efficiency of international law and therefore indirectly about how states are perceived at the level of individual.

The importance of studying compliance from a psychological perspective also comes to light when we consider what type of measures would be efficient in encouraging states towards better compliance. In their seminal article Rewarding in International Law,⁵⁹ van Aaken and Simsek address the presence and operation of sanctions in international law and argue that from a psychological perspective, we should consider also whether there is room for rewards. Their main argument is that penalties and rewards produce different emotions. Threats of penalties trigger emotions such as fear, anxiety, anger, and stress and potentially lead to suspicion, hostility, aggression and loss of trust. These emotions are likely to trigger hawkish biases, with actors “more likely to see threats as more severe than an objective observer would perceive them.” As a consequence, they are more likely “to act in a way that will lead to unnecessary conflict,”⁶⁰ leading to non-compliant behavior.

In contrast, in the case of rewards, the emotions triggered are normally positive, they encourage cooperation and trust and decrease the potential for conflicts. They are also more likely to be reciprocated;⁶¹ “Rewards incorporate an important signalling function that increases the perception of respecting law in the international arena.”⁶² In this regard, it leads to de-escalatory behaviour. Van Aaken cites evidence from tax compliance, suggesting that we adjust our own compliance with taxes on the basis of what we perceive others—especially our neighbors—have done.⁶³ In a similar manner, states may be mirroring other states in their own behavior. It is clear, therefore, that states’ responses to penalties and rewards differ and that from a psychological perspective, there are no major disadvantages in using rewards rather than penalties.

Other studies have investigated also other remedies from a behavioural perspective and taking into account individuals’ and states’ cognitive biases. Scholars have, for example, found that punitive sanctions do not always lead to improved state behavior⁶⁴ and that rankings and other “nudges” can cause states to improve their actions.⁶⁵ Further research will look at how the remedies architecture in international law can be redesigned to maximise state compliance.

F. Individuals in the International Legal Process

In this section, we turn to how individuals who appear in different spheres of international law make decisions and how their cognitive biases shape these decisions and ultimately affect the application or interpretation of international law.

I. Litigants

One of the most frequent decisions that individuals make when entering the legal process is deciding whether or not to settle.⁶⁶ The choice they are making is therefore between litigating their claim or agreeing on an amicable solution with the other side so that there is no need to proceed to litigation. Both practitioners and scholars agree that settlements allow greater party participation and are more responsive to the needs and underlying preferences of the parties. The two opposing sides can save money and time by avoiding litigation, they can be inventive in the range of outcomes, flexible with regards to solutions and creative when it comes to remedies. In the end, the compromise that is reached can work for both sides and studies have shown that often such decisions lead to increased party satisfaction and better enforcement than decisions of a court. Participants often speak of having experienced the process of negotiation as a transformation, the process of back and forth allowing both sides to air their grievances and outline expectations while seeking to understand the other side.⁶⁷ This enthusiasm for settlement has been reflected in practice, with courts encouraging settlement of cases as a way of enabling court efficiency, in other words a tool to save courts time and resources and conserving judicial attention for only the hard cases. As a consequence, in certain areas of domestic law settlement rates currently stand at 90% of all cases brought.⁶⁸

The push towards settlement is felt not only in domestic law. At the international level too, settlement is encouraged. Both the Inter-American Court of Human Rights and the European Court of Human Rights have instituted procedures to encourage settlement, similar efforts are also being explored by the African human rights system.⁶⁹ A system of settlement also exists for WTO disputes.⁷⁰

But decisions to proceed to settlement are fraught with traps and probability assessments about the likelihood of success at litigation and about the perceived fairness of the offer made by the other side. Domestic studies have shown for example that in majority of cases rather than making rational decisions about minimizing litigation costs and maximizing benefits of proposed settlement offers, individuals appear to be seeking ways to re-establish equity in their relationship with the other party. While rational models of settlement assume that people decide on settlement based on comparison of expected monetary values of trial and settlement and speed of proceedings, often for individuals the process of settlement is not only about money. Interpersonal comparisons influence individual behaviour.⁷¹ People want to be treated justly and they become distressed in an “unjust relationship”⁷² Often, this means that victims appear more concerned with “issues of vindication and with obtaining an adequate hearing of their dispute than with the actual award that they obtain.”⁷³

In addition, litigants entertaining settlement are likely to automatically devalue an offer of settlement because it is made by the party that wronged them. The sole fact that the offer originates from an adversary means that respondents treat it less favorably and that they give way to personal feelings of vindication and retaliation over economically rational calculations—“reactive devaluation” bias. The offer will appear to them “disadvantageous” or even disingenuous.⁷⁴ This cognitive bias often comes out of spite or because of a fear of the adversary having access to more information or ill feeling towards them. In the end, the offer made will be “unpalatable”⁷⁵ and applicants will want to reject it, even if the proposal would make economic sense.

These biases are important because they influence procedural decisions we make at the very beginning of a potential legal process. In many ways, they should be taken into account when we seek to make decisions about settlement and if we are aware of them, they should steer us towards solutions that could help us address them—such as engagement of a mediator or a third party, who would help reduce devaluation bias and help us repackage and reframe offers made by the other side into neutral proposals. This proposal is explored in this Symposium by Eva Nissioti, who looks at the role of mediator in international sphere.

II. Adjudicating Panels

We turn now to how our biases affect others, how once we belong to a social group, we have a tendency to differentiate ourselves by group membership, how we use our biases to help us maintain a favorable image of ourselves and to attribute positive traits to our friends and our reference group, and ultimately how this leads to discrimination. As studies have shown, it is not the biases of one individual that are necessarily key to discriminatory behavior. Instead, it is the formation of an “ingroup” whose members seek to provide each other preferential treatment that leads to discrimination. Once we identify as part of an ingroup, we see other members of the ingroup as “trustworthy, cooperative, peaceful, and honest.” On the contrary, members of the “outgroup” are believed to be “untrustworthy, competitive, quarrelsome, and dishonest.”⁷⁶ In this context, although intended to merely provide positive favoritism towards ingroup members, our behavior is accompanied by a number of biases that sustain and reinforce discriminatory behavior and potentially lead to “outgroup” negativity.⁷⁷

Studies have revealed such ingroup bias in the appointment of arbitrators,⁷⁸ including a study uncovering a tendency of arbitrators in international law to wear two hats—both as counsel and arbitrator.⁷⁹ Some studies have noted the lack of women judges and arbitrators at international level, while others have asked how the implicit gender bias could be affecting outcomes or determination of damages at international level.⁸⁰ Other studies have focused on the question whether national judges on international courts are more likely to vote with their country out of loyalty.⁸¹ In this volume, Moshe Hirsch builds on his work of the sociology of international tribunals⁸² and looks at how the WTO Dispute Settlement Body should investigate the intent of the domestic regulator, their social identity, as well as inter-group relations which could affect their behaviour. Along similar lines and building on a large dataset of judgments, Runar Lie shows how language can be used as a way to tie together a legal community.

III. Negotiators, Investigators, Assessors

Many international law decisions are made by individuals, often possessed with expertise, legal or otherwise. Recent international scholarship focuses on these decision-makers by conducting experiments on how they make their assessments. For example, in their work, Broude and Levy “examine individual international humanitarian law decision-making on two levels: military decisions made ex ante regarding real-time operational questions under conditions of uncertainty and imperfect information, and subsequent ex post evaluations of the propriety of military decisions in the context of military investigations regarding legal responsibility with respect to proportionality and reasonableness.”⁸³ As they show, international humanitarian law requires investigators to consider only information available at the time decisions were made. Through an experimental vignette study conducted with laypersons, legal experts and people with field experience, they uncover that negotiators are susceptible to cognitive “outcome bias, ” specifically the extent to which the knowledge of operational outcomes, especially regarding incidental civilian harm, influences ex post normative evaluations. Revealing a general tendency towards outcome bias, they do show that bias is somewhat tempered by expertise in that “individuals with operational decision-making experience may be less prone to outcome bias than legal experts.”⁸⁴

In a similar manner, Shiri Krebs analyses the working of an Israeli investigatory commission tasked with the process of establishing what happened in the case of a targeted killing of Hamas Leader Salah Shehadeh in Gaza in 2002.⁸⁵ The operation killed fifteen people, including Shehade, his family, and a family that lived next door. Eight children were among the dead and another 150 were injured as a result of the attack. The commission was asked to determine what happened on the ground and to draw lessons for future operations. Yet, as Krebs shows, while investigating the “fog of war,” the investigation “embrace[d] uncertainty as a justification for gaps and inconsistencies in the factual framework.”⁸⁶ The Commission said that the “unfortunate harm” caused by the attack was “unintentional and unpredictable” and resulted from an intelligence failure caused by “incorrect assessments and mistaken judgments.” ⁸⁷ The Commission accepted that individuals acted under uncertain circumstances, but took these for granted, without looking back to understand what information could have been missed so that lessons could be learned for the future. In fact, the Commission accepted that there was no positive information about the presence of civilians and that special forces acted on the basis of absence of such information. In doing so, Krebs argues, the Commission reinforced the salience and availability bias with which decisions in the security agencies were made—fixation on Shahadeh’s future activities and active ignorance of collateral damage risks, overconfidence and refusal to consider other feasible alternatives—and accepted that the assessment or anticipation of the potential for collateral damage should be based on positive information about the presence of civilians, rather than the reverse—that their presence should be assumed and that it is their absence that needs to be proven. The example demonstrates “how an incomplete factual framework—justified and presented as inevitable through the ‘fog of war’ metaphor– renders otherwise unjustified killings lawful and acceptable. . . The outcome is legitimisation of otherwise unlawful actions, and perpetuation of faulty processes and human insecurity.”⁸⁸ The commission concluded that Shehadeh was a legitimate target and found that although the death of the civilians was disproportionate and excessive, it accepted the security agencies’ claim that it was not anticipated. The conclusion followed that no international or Israeli law had been violated and no criminal proceedings were necessary.

Building on this focus on individual actors in the international legal process, Eva van der Zee’s contribution in this Symposium looks at the role of experts in the field of environmental impact assessments. These are notoriously subject to biases because they require individuals to make risk assessment and probability estimations based on limited information. Van der Zee explores how cognitive biases distort experts’ decision-making and investigates how these biases could be addressed. In this context, her contribution adds to the existing research by revealing how these biases can significantly distort the effectiveness of impact assessments in bringing environmental concerns to the forefront.

G. This Symposium

In looking at how cognitive bias matters to international law, this special issue builds on the existing literature and seeks to add to it. By focusing on legal frames, labels, and cognitive biases, we reveal how these shape our understanding of international rules, our application of these rules and the outcomes of international adjudicatory processes. Most of all, we wish to open up the black box of the state and reveal how individuals on their own and in groups make decisions that are systematically imperfect and that in order to understand how international law works, we have to understand how they—the actors in international law—behave.

This issue proceeds in three parts. In Part I – Uncovering Bias, we look at different ways in which we can identify bias in the application and interpretation of international norms. Moshe Hirsch in his article looks at the requirement of Article III of the GATT that prohibits discrimination against imported products or internal regulations applied to imported or domestic products that could affect internal sale “so as to afford protection to domestic product.” He notes that regardless of the provision preferential treatment for domestic products is often pervasive, with domestic regulators intentionally or unintentionally discriminating against foreign products. While rational choice analysis rejects interventions such as pressure exerted by domestic interest groups and response measures favoring domestic industry, discrimination of foreign products can also occur due to cognitive biases, which often discriminate against foreigners and foreign products/services. In this context, Hirsch insists that those who assess the compatibility of domestic regulation with GATT need to go beyond the traditional blunt assessment and remain sensitive to the intent of the domestic regulator, their social identity, as well as inter-group relations which could affect their behavior. In fact, numerous cognitive studies have revealed that once people identify with a particular social group, they tend to provide in-group members with better treatment. Such biases—involving attention, interpretation, or memory—feed and reinforce discriminatory behavior towards people belonging to other groups. This ingroup favoritism can go undetected if cognitive literature is disregarded. The article therefore argues that social cognitive literature can contribute to the way we conceive trade discrimination, suggesting several legal and policy strategies aimed at uncovering and mitigating discrimination in the international trading system.

In the next contribution, Benedikt Pirker and Iza Skoczen investigate to what extent a neutral looking provision of an international treaty can imply more than its ordinary meaning. Following the customary rule of interpretation found in the Vienna Convention on the Law of Treaties, which states that treaties shall be interpreted according to the ordinary meaning given to the terms of the treaty, Pirker and Skozen explore what this means in practice. By providing individuals with different formulations of the Treaty on the Functioning of the European Union and the Convention on the Prevention and Punishment of Genocide, they experiment to what extent participants are certain of the meaning of the norms, and to what extent they “read in” a moral—or morally informed—interpretation of treaty provisions. Revealing our need to fill in gaps in interpretation, the authors ultimately challenge the assumption that norms are neutral and conclude that “ordinary meaning” from the Vienna Convention of Law of Treaties often appears to be a moral concept.

Along the same lines, the next article by Jonathan Kolieb pursues the question of how allegations of “war crimes” violations affect consumer buying behavior. In traditional economics literature, the study of consumer behaviour is a frequent phenomenon and is often used to determine the efficiency of a certain norm, in other words, whether the norm is able to trigger a change in consumer behaviour. In international law, when we ask the question “does international law work?”, we do not really consider international norms’ impact on “consumers.” Kolieb’s contribution is a call to a different understanding of efficiency of international law. By looking at how lay individuals purchasing everyday products react to messaging around countries’ or companies’ complicity in war-crimes, he argues that we can determine whether these labels lead to a change in consumers’ choices in response to information about international humanitarian law being violated by the country or the company. Kolieb shows how the use of international humanitarian terminology can play on consumers different biases—from the anchoring effect to the representativeness bias, as well as the role of social context and peer pressure. These non-typical or anticipated consequences of the use of labels in everyday life therefore ought to be explored through experiments.

The articles thus far have focused on identifying cognitive biases expressed by individuals—regulators, consumers, everyday people. The final article in Part I looks more closely at international actors and the community they belong to and explores the specific heuristics and biases that are born out of operation in those communities. In this context, Runar Lie shows how we use language as a way to define legal communities. He maps out the move in international arbitration from referring to multiple international treaties in plural to a reference to international arbitration law in singular. This change in language serves as a convergence tool with which actors are helping create and unify a community of actors working in the same area. The article maps out changes in actors’ terminology, linguistic features or citational patterns, and argues that such changes may indicate that the actors themselves are creating a common, self-constituting society of law. This not only serves in the identification of a specific area of law, but also leads to consistency and predictability in a specific area of law. It also sets the ground for ingroup favoritism raised in Hirsch’s article and biases and discrimination brought on by it.

In Part II—Addressing Bias, we turn to two articles that investigate how our biases can be mitigated and addressed. Eva Nissoti examines how in a situation where two parties are pursuing a settlement, the presence of a mediator in an international dispute can mitigate their biases. Negotiation and settlement procedures are widely acknowledged for fostering behavioral biases and heuristics, thus leading to suboptimal results. The article explores which biases are most prevalent in the process of negotiations, including anchoring, overconfidence, framing, status quo, and self-serving bias. Nissoti then discusses how the presence of a skilled negotiator can decrease the operation of these biases and lead to a better outcome. Specifically, she engages with the issue of reactive devaluation bias, according to which a party typically devalues the offer made the respondent—in other words, the other party—and then analyzes how the presence of a skilled mediator can redress this bias and reduce its operation and impact in the dispute at hand.

Along similar lines, Eva van der Zee looks at the role of experts in the field of environmental impact assessments. These are notoriously subject to biases because they require individuals to make risk assessment and probability estimations based on limited information. Van der Zee explores how cognitive biases distort experts’ decision-making and investigates how these biases could be addressed. Specifically, she focuses on biases that tend to be resistant to logic or the use of training tools. These biases can significantly distort the effectiveness of impact assessments in bringing environmental concerns to the forefront. She explores how these biases could be corrected through procedural regulatory tools, such as requiring or recommending multiple experts with different points of view, the inclusion of more alternatives early on in the impact assessment process, and the incorporation of multiple stakeholders to provide different value perspectives. The article concludes by underlining how insights from behavioral science could further improve international guidance towards environmental impact assessments.

While Parts I and II discussed topics from trade to arbitration, consumers, environment, and negotiations, they focused on how biases might affect individuals’ decisions to litigate or settle, their risk assessments, their reading of legal provisions, or indeed their responses to international law labels. In contrast, in Part III we turn our attention to studying the behavior of states. Recent international law scholarship has applied behavioral analysis and cognitive biases to states as unitary entities. The last contribution in this Symposium explores The Limits of Behavioural Analysis in International Law by uncovering methodological issues in adopting a behavioral approach to analysing state behaviour. Ezgi Yildiz and Umut Yüksel draw on rational and behavioral approaches to formulate different expectations about the process of policymaking and updating in the context of maritime delimitation. While self-interest and incentives are the usual elements involved in a rational choice explanation of policymaking, behavioralist scholarship casts doubt on whether decisionmakers are able to identify and pursue their interests in a rational manner. Yildiz and Yüksel focus on the ways in which states make and update their policies about the appropriate method of maritime boundary delimitation in light of the decisions of the International Court of Justice. Using a data set of continental shelf delimitation policies of states between 1958 and 2019, the authors find evidence that at least some states take court decisions into account. However, they are unable to distinguish between mechanisms consistent with rational choice and those suggested by behavioralism—also known as the anchoring effect. They show how and why behavioralist explanations of policymaking processes are difficult to test in a large-N setting. Moreover, they reveal that additional evidence from interviews also proves insufficient, notably due to the tendency of actors to rationalize state policies ex post facto. In the end, they provide suggestions about how these problems could be addressed in future research. Their contribution is proof that articles that are not able to reject the null hypothesis deserve equal attention.

H. Conclusion

This Symposium is a call to a different study of international law. It wishes to draw attention to the hidden and the unsaid, but the ever-present—to our bias. It seeks to lift the veil in how international law rules and applied, perceived and how they influence our behavior. It does this by first showing that apparently neutral international norms may not be neutral in their application and that regulators, legislators and others should think of our cognitive biases and heuristic errors when drafting such norms; secondly, by revealing that international law labels such as “human rights,” “refugees,” ‘international crimes” may have an impact that goes beyond the envisaged or intended by the law; thirdly, by underlining that when we study international law, this is not an abstract, theoretical discussion of judgments, laws, and scholarly writing. Instead, it is a domain filled with people – judges, arbitrators, scholars, practitioners, but also individual human rights victims, everyday consumers, and others. By looking at how bias in international law is exhibited and expressed, the articles in this Symposium seek to broaden the scope of addressees of international law. In the end, we argue that when we study international law, we should always think about who we study.