Principal Investigator: Bruce Weinberg, Department of Economics

Why do some people become leaders?  Do group members see leaders as the same or different from themselves?  Are leaders chosen because they are representative of the group, or do the actions of the group reflect the will of the leader? 

Bruce Weinberg tackles these questions by examining the effect of leaders on group behavior.  To measure this, Weinberg used the National Longitudinal Study of Adolescent Health, a data set covering more than 90,000 students in grades 7 though 12 in 132 schools nationwide. Schools make a great laboratory to study social interactions because the information is well defined and consistent from one school to the next.

Weinberg's research team identified leaders by finding out who each student listed as friends; those nominated by the most students were classified as leaders.  Then the team tried to quantify the influence these leaders had on their peers.  Such quantification has been challenging because leaders can affect their peers in two ways: by influencing a large number of people, and by influencing each person intensively.

Another challenge has been to disentangle the process by which someone becomes a leader from the effects the leader has on the group.  Most leaders emerge because their characteristics and behavior are highly representative of the group.  Yet they also influence the group toward these same characteristics and behavior.  Any research into the effect of group leaders must identify and separate these two variables.

Weinberg's research team is using two strategies to address these issues.  First, they are comparing leaders in the 9th grade to leaders in the 12th grade.  If leaders are representative of their groups, the behavior of both sets of leaders should be similar because they come from the same school.  However, the 12th-grade leaders should have greater influence on students in the lower grades than the 9th-grade leaders have on upperclassmen. This difference lets researchers quantify the affect of leaders on the group.

Second, the research team looks at the characteristics of people most and least often identified as friends.  If people derive utility from associating with leaders, they should be willing to trade off similarity in a friend for status.  This is because people usually associate with someone of low status if that person is similar to them, but they are willing to associate with someone of high status even if that person is quite different. 

Weinberg's research findings could help policy-makers gain a better understanding of societies with strong leaders such as Osama bin Laden or Kim Jong Il.  If a leader who operates against the interests of the United States is highly representative of his people, it may be difficult to dislodge him from power.  But if the leader does not represent the people and rules only through force, his hold on power might be more tenuous.