Experiments on Motivation, Incentives and Rationality – EMIR

This program is based on two sets of theoretical and experimental findings: - On the one hand, experimental economics has confirmed that economic agents react to incentives and competition, and agency theory shows that it is possible to design contracts that lead these agents to behave as efficiently as possible in a situation where there is imperfect or incomplete information about abilities or effort; - On the other hand, the introduction of monetary incentives can sometimes exert a negligible or even a negative influence on motivation and performance. This is due to crowding-out effects, income targeting, choking under pressure, or social comparisons. On the basis of these findings, this program seeks to contribute to an understanding of the relationships between incentives, motivation and performance under imperfect information. This will be done by examining some monetary and non-monetary dimensions of motivation that have not been explored so far. We will investigate how the agents solve the tension between what rationality suggests for the maximization of individual payoffs and the social comparisons and moral aspects of decision-making. The program will produce new experimental evidence that in turn will contribute to the development of behavioral economics. The program will investigate 5 perspectives and sets of hypotheses: - Incentives and competition. If the agents have a better knowledge of their abilities than the principals, the incentives provided by the contracts offered on the market must be designed so that these contracts solve the adverse selection problem. Unbalanced competition (with excess agents or excess principals) affects the nature of incentives provided to the agents on the market and influences the level of efficiency. - Incentives, principles of action and social preferences. The agents, whether selfish or motivated by social preferences, self-select into different types of contracts (and incentives). Therefore, the efficiency of a specific incentive scheme depends on these orientations and on the sorting process. But we will also test the hypothesis that when agents can hide their true motivation when making decisions, there may be what we shall call a false consciousness or self-serving justification at work. Rather than people following principles that guide their actions, the agents make up principles to justify their actions. Therefore, monetary incentives guide the actions and the latter ex post influence moral principles and social justification. - Incentives and comparisons. A first aspect will focus on the measure of the contagion goal effect between groups of agents. A second aspect will focus on performance comparisons in terms of rank. We plan to measure the willingness to pay to improve own rank. Costly investment in rank-improving activity may be all the more strong when incentive schemes are flat. It may be less important if it requires malevolent actions (sabotage). - Intermittent reinforcement and behavior persistence. When the reinforcement (reward/punishment) schedule is random, people persist in their behavior for much longer. If employers are inconsistent across time in giving rewards or punishments, good or bad behavior will tend to persist. - Evaluation, incentives and motivation. If agents have less precise knowledge of their abilities than the principals, then feedback about intermediate performance can be biased and manipulated strategically by the principals to influence motivation and future performance. The relationship between the evaluation of performance and the agents' motivation depends on the compensation scheme in use. The program will conduct laboratory and field experiments. This will be supported by the cooperation with researchers from the University of California at Santa Barbara, the University of Minnesota, the University Pompeu Fabra, and the University of East Anglia. These researchers are famous international experts in behavioral and experimental economics, and especially in the analysis of monetary and non-monetary incentives. The program will enable invitations and the design of joint experiments. It includes the organization of a workshop and it will give rise to the submission of research papers to top-ranked journals.