Academics and Faculty
Zell CenterKellogg School of Management
AcademicsFacultyResearchKellogg Homepage
Email Sapienza

Sapienza's bio

Sapienza's CV (PDF)

Zell Center

Site Index Search Internal Site Northwestern University
The Zell Study on the Determinants of Risk Attitudes

Paola Sapienza and Luigi Zingales

Preferences towards risk are an important determinant of economic behavior such as the likelihood of investing in a risky asset or the willingness of individuals to start their own businesses. Despite a large body of theoretical research on the impact of risk attitudes on individual decisions, we have a limited understanding of why individuals have different attitudes to risk and how these attitudes are affected by the individuals’ background and education.

The Zell Study on the Determinants of Risk Attitudes aims at measuring risk attitudes and understanding their determinants and the consequences on individual behavior and success. For a very large sample of MBA students, we collected risk attitudes, other behavioral traits, and certain biological and psychometric characteristics. These measures were collected as part of The Chicago MBA Longitudinal Panel Study (supported by a grant of the Templeton Foundation).

Volunteers for the study are drawn from first year MBAs participation in the Leadership Effectiveness and Development (LEAD) Program at University of Chicago, Graduate School of Business. LEAD is a laboratory class where students practice and perfect key communication skills such as negotiation, team-building, and giving feedback. All first year MBAs, as part of their CORE program, participate in LEAD to understand about business success. The LEAD program started in September 2006. First, as assignment in the LEAD class, the students were asked to fill a survey containing several psychological tests and eliciting economic preferences and attitudes. The survey contains standard psychological tools to measure optimism, self esteem, emotional intelligence measures, empathy, ability to influence others, attitudes with respect to collectivism and individualism, and several other measures. It also contains tools to measure economic preferences and attitudes (eg. risk aversion, inequality aversion, lopsided view of gains and losses, altruistic preferences). Finally, it contains some demographic characteristics and a parenting style questionnaire that measure the educational and family background. These data have been merged with demographic information collected by the admission office at the University of Chicago, Graduate School of Business (ranging from undergraduate institutions and GPA to parental demographic information).

Welcome screen for the game session during LEAD.
Photo © P. Ledesma

In addition, to these data during class time ("LEADERSHIP GAME") students were asked to play six standard games aimed at identifying their willingness to interact and cooperate with others, to invest time and money in risky endeavors, their competitiveness, and their ability to trade. Subjects earned up to $300 and on average $95 dollars for participating in these games, and the payoff from the games was correlated to their ability and luck.

Finally, the students will be followed over time while at the University of Chicago and after graduation, by the career development to measure their success in the job market.

To investigate the determinants of risk attitudes and the consequences on individual behavior, during the experimental game session, we collected saliva from the students, before and after they played the games. The saliva samples are important in our research because biologists have found that in populations of extreme risk lovers (e.g. extreme gamblers) students have extremely high testosterone levels. One hypothesis we want to investigate is whether testosterone level in normal population is correlated with risk attitudes in games, in the survey, and in future career’s decisions.

Students participating in a trading game.
Photo © B. Riley

In the Zell Study on the Determinants of Risk Attitudes we analyze what characteristics are correlated with risk attitudes distinguishing between biological and education factors. We link the subjects’ attitudes to his or her backgrounds, including education, economic conditions, religion, and place of origin, family circumstances, and ethnicity. For example, we have found some preliminary evidence that parenting styles (classified as authoritarian, permissive, and authoritative) are correlated with risk attitudes. These data will then be compiled in order to evaluate the relative importance of biological (for example, birth order, testosterone levels) and educational factors in determining levels of optimism, risk attitudes, self esteem, and trust. With two measures of performance, in the games and in the job market, we are planning to study also the effect of risk attitudes, combined with other characteristics, on career choices decisions and individual performance.

Paola Sapienza and Luigi Zingales monitoring the progress
of the games.
Photo © P. Ledesma


©2001 Kellogg School of Management, Northwestern University