Kevin A. McCabe - US grants

This project uses functional Magnetic Resonance Imaging to investigate biologically plausible brain mechanisms that produce economic behavior. Traditionally experiments in economics have taken place either in the field, where individual behavior can be observed before and after a treatment variable has been manipulated, or in behavioral laboratories where, in addition, there is greater control over incentives, information, and the rules of the game. The purpose of such experimentation is to better understand both the behavioral mechanisms that produce particular types of behavior, such as investor overconfidence in 'bull' markets, and the resulting behavior of markets, such as price bubbles. While a great deal of progress has been made in understanding market behavior very little progress has been made in understanding individual behavior. Recent advances in functional brain imaging now makes it possible for economists and neuroscientists to collaborate to study how brain activity is linked to economic decision making. This project studies how the human brain is genetically designed, and environmentally influenced, to solve the problem of mutually advantageous cooperation. This research is likely to have a significant impact on economic thought, and institutional design, by linking observed economic behavior to the biological mechanisms that produce this behavior.

Field observations have established the ubiquity of mutually beneficial, reciprocity based economic behavior, and experimental research has demonstrated that such behavior is robust to controlled laboratory environments. Unfortunately, non-cooperative behavior founded in individual self-interest can corrupt trust-based exchange. Consequently, humans continually face the problem of finding credible procedures to commit to reciprocal, mutually beneficial exchanges. An important question is what kinds of social institutions enhance trust-based exchange.

The project examines the implications of trust relationships for market performance
in both personal and impersonal exchange settings. In these experiments, two treatments are considered in each of two contracting institutions. The first institution, bilateral bargaining, is relatively personal. The second institution, a double auction, is more impersonal and involves multilateral negotiation rules that allow buying and selling through a publicly observed bid-ask spread. The two treatments are manual contract execution, where a subject must choose to complete their contract, and automatic execution, where contracts are executed immediately through the experimental software. The goal is to investigate whether and how breakdowns in trust relationships can lead to market failure, and whether this failure is related to contracting institutions.

Over the last ten years laboratory experiments, informed by the work of game theorists, experimental economists, economic historians, evolutionary psychologists, and sociologists, have helped to develop theories that take account of the ecological rationality of economic behavior; and particularly the way incentives have changed as humans moved from small group personal exchange to large group impersonal exchange. This proposal describes a number of new experiments that examine how the way in which human performance in markets can fail during a transition to impersonal exchange.

Understanding how human economic performance relies on trust relationships, and finding procedures that help to facilitate and sustain this trust-based performance, is increasingly important to society at large. One important reason is that online markets are expected by many to be an engine of economic growth over the next few decades, but as the research suggested herein will make clear, these markets can potentially fail because of failures of trust. In particular, buyers and sellers in online markets must transact with counterparts about whom they know very little, and, for example, an untrustworthy buyer might not deliver payment, or an untrustworthy seller might not deliver an expected product. These possibilities are real: complaints relating to online markets accounted for 78 percent of all Internet fraud complaints received by the National Consumers League in 2000. Research presented in this proposal represents a step towards better understanding the link between human performance in markets and trust, and provides insights on how to design market institutions that help to prevent market failures due to trust failures, and that consequently act as catalysts for future economic growth.

Legislators, bureaucrats and others frequently make economic decisions that affect others. A judge may require a defendant to pay damages to a litigant. Congress may pass legislation that taxes one group of citizens to finance spending on another group. Bureaucrats may impose regulations at a cost to some people to benefit others. While economists, political scientists, psychologists, and decision scientists have studied group decision making, we do not know much about how individuals make decisions in these situations.

This research uses economic experiments to understand this kind of decision making. The goal is to test implications of the equity theory of fairness, which predicts that a third party is less like to redistribute income and assets if he or she believes that these funds have been earned by the owner, rather than simply a result of fortune or luck.

The results add to our scientific knowledge in behavioral economics, political science, law and social psychology.

In this research controlled economics experiments are used to test the predictions of economic theories that hypothesize effective cybersecurity tradeoffs within an organization depends on both worker incentives and the structure of job duties. In these experiments a team of economists and cybersecurity experts are working together to design virtual world experiments that measure the impact of different incentive arrangements and job design on operational cybersecurity risks. This is particularly interesting in computer intensive environments where the mix of competing tasks include easily monitored tasks with easily verified impact on organizational goals, such as tasks performed by application programming teams, and tasks that have a more ambiguous impact on organizational performance such as tasks associated with cybersecurity practices. In the former case high powered incentives often compete against the relatively low powered incentives associated resulting in the inefficient management of cybersecurity risks. The goal of this research is to better understand how management practice interacts with available cybersecurity technologies and cybersecurity threats to enable better risk management strategies within organizations. Such research provides evidence based findings that will make organizations more aware of their cybersecurity/operational-efficiency tradeoffs and thus allow them to improve their organizational practices and consequently their risk management strategies against cybersecurity threats.

In the absence of strong institutional rules within a group governing the usage of shared resources, the decisions of individuals to engage in resource extraction may create suboptimal outcomes for the entire group -- this outcome, when it occurs, is commonly called the "tragedy of the commons." Forests, fisheries and water represent some of the best-known examples of resources whose management is vulnerable to this dynamic, wherein an individual's choice to extract a given amount of resource may be made without internalizing the negative impacts that this decision will have on the resource's ability to replenish itself in order to facilitate sustainable extraction in the future.

The establishment of private property rights over the common resource is a classic mechanism by which the tragedy of the commons may be avoided. Property rights guarantee private resource owners the rents accrued from future extraction of the resource, and thus these owners do not face strong incentives to immediately exploit the resource out of fear that someone else will if they do not. However, in many cases "privatizing the commons" may be infeasible because of difficulties in defining or enforcing formal property rights. Nonetheless, even when resources are not formally privatized groups are often able to effectively define rules governing a commons.

The exploration of the effectiveness of different institutional rules and organization forms in different commons environments has generated a very important literature in the past half-century which has generated valuable advice pertaining to how policymakers should both identify and respond to the tragedy of the commons in its various manifestations. These insights have been applied to deal with the management of resources such water in southern California, crab populations from the Chesapeake Bay to the Aleutian Islands, and most sorts of game wildlife.

The design in this proposal will shed light on how different levels of survival pressure may affect a group's ability to establish rules for commons management. A controlled laboratory experiment will be implemented wherein subjects will be given control of avatars in a naturalistic online environment. Subjects will acquire income by moving their avatars around within this environment and gathering resources; however, if resources are overharvested then they will replenish more-slowly and the group's overall income will be reduced.

In a previous experiment implemented in a similar environment, the establishment of institutions such as informal property rights and territoriality proved essential to effective commons management. However, this design introduces the additional notion that poor resource management may lead to the disruption of homeostatic equilibrium within avatars -- in essence, a failure to gather resources may lead to avatars becoming weak or dying. This process may also be hastened via avatars being able to directly attack one another.

Survival pressures will take the form of a requirement that avatars must gather resources at different rates in order to survive each period of the experiment. By examining multiple levels of pressure -- ranging from very low (avatars need to do very little gathering to survive) to very high (even under optimal management, not enough resources will be produced for all avatars to survive) -- data will be gathered concerning how this threat of avatar mortality directly impacts the institutions that groups implement in order to govern the common resource. It is hypothesized that increased pressure will make informal institutions much more fragile since tactical failures may lead to the deaths of avatars. In turn, this will lead to worse commons management and increased violence in this environment.

It is undoubtedly the case that for many individuals, ineffective commons management may lead not simply to marginally-diminished incomes but also to severe impoverishment and death. In the real world, high survival pressures may result in the collapse of informal institutions and outbreaks of violence. This design brings to bear the advantages of the controlled experiments by examining the direct impacts of survival pressure on commons management. The naturalistic environment being used allows for the collection of much-richer institutional data compared to previous designs that have examined similar issues. In doing so, this design aims to glean new insights into the conditions under which groups can effectively self-govern.