In studies of employment outcomes, we usually relate outcomes to an individual's personal attributes, like education, ability, race and gender; less often do we consider the attributes or the productivity of the group in which the person works or socializes. Yet interactions between group members and group leaders may have powerful effects on an individual's job performance, especially for a young adult who is new to the labor force. In this research program, we will investigate several questions about the role of peer groups and group leaders in determining the employment outcomes of young adults. Peer groups and superiors may be especially important for young adults, since they bring little in the way of prior workforce experience and may more readily turn to coworkers and superiors for feedback, guidance, and informal on-the-job training. There might also be peer pressure and social norms regarding work-related performance, and new employees may learn about these social norms from coworkers and mentors. Furthermore, early labor market experiences may be particularly important in the formation of a work ethic, initiation of a career path, setting of promotion prospects, and the determination of future earnings growth. Although we might believe groups matter, it is difficult to credibly measure group effects because individuals usually self-select into groups or are assigned to them on the basis of unobservable characteristics. In this project, we avoid the problem of self-selection by exploiting the random assignment of newly enlisted young adults by the U.S. Army to their first work units. Although the Army amounts to just a single employer, it is also the nation's largest employer. We will use confidential Army personnel records containing a panel of approximately 70,000 new enlistees during 2002, with 2 million person-month observations up to 2005. Our data include every enlisted individual in a given unit, as well as officer personnel. In our first specific aim, we will study the causal relationship between job performance of newly enlisted young adults and the demographic composition of their initial assigned unit, including racial, gender, educational and ability composition. In our second aim, we will study the causal relationship between the job performance of newly enlisted young adults and the race, gender, and rank composition of unit leadership. In both aims, job performance is measured by time to promotion to the levels of specialist and sergeant. In this research we will study how the demographic composition of peer groups affects the employment outcomes of young adults. Early market labor market experiences may have strong effects on the wellbeing of young adults, and may also be an important determinant of future earnings growth. ? ? ?