What Can Managers Learn From College Basketball?

Successful job hunting usually requires active networking that reaches beyond our immediate circle of contacts. According to past research, we are less likely to find jobs through our close friends and family than through an outer circle of acquaintances. And there’s a simple reason for that: Those individuals have information that we don’t already know. But social networks aren’t just important for information access. In fact, a recent study shows that they might also play other key roles, for example, by helping to shape employer perceptions of job candidates.

The research was conducted by Daniel Halgin, a doctoral student in the organization studies department of Boston College. Halgin investigated a particular kind of social network—professional networks in which people have a shared sense of belonging and identity that they retain through different career moves. Such networks often develop around shared workplace experiences. For instance, consultants who have worked for Bain & Co. Inc. often think of themselves as “Bainies” for life, and former high-level managers of General Electric Co. are frequently referred to as “graduates of Welch U.” (referring to ex-CEO Jack Welch) even after they’ve left GE to run other businesses. These types of affiliations certainly provide opportunities to network and exchange information, but do they also play other important roles in the labor market?

To answer that question, Halgin studied the movement of head coaches for the top U.S. college level of men’s basketball teams—the National Collegiate Athletic Association (NCAA) Division I. That sample might be somewhat unorthodox, but it’s a veritable petri dish for investigating job movement. The 341 colleges in the study tend to switch head coaches frequently: Between 2001 and 2007, more than 280 changes occurred, including more than 150 firings. The sample also allows for certain effects to be studied in relative isolation because the data could be controlled for various factors, including employee performance, as reflected by the win-loss records of the coaches. Moreover, because of the frequent job changes and relative insularity of college basketball, the study sample was a highly interconnected network: At the start of the 2007 season, 90% of the head coaches had worked with one of the other head coaches previously in their careers.

But perhaps the most intriguing aspect of the study sample is that many of the coaches were affiliated with one of eight distinct “families,” each associated with a legendary coach (for example, the “John Calipari” family, named after the current head coach of the University of Memphis, who has won more than 400 games throughout his career). Membership in those groups was based primarily on past relationships (someone who had once been Calipari’s assistant coach, for instance) and shared characteristics. For example, the “Rick Pitino” family, named after the current head coach of the University of Louisville, is known for a certain style of basketball play that encourages three-point shots. (Details of the research are contained in Halgin’s paper “All in the Family: Network Ties as Determinants of Reputation and Identity in NCAA Basketball,” Proceedings of the Academy of Management 2008 Annual Meeting.)

Not surprisingly, Halgin found that coaches who were well-connected were more likely to find jobs after being fired. This was true even after controlling the data for the win-loss performance of those individuals. But Halgin also discovered that those coaches who were members of one of the eight coaching families he identified had even greater career resilience and were able to obtain more prestigious positions than other coaches who had comparable win-loss records and similar network connectivity. In other words, the advantages of “family” membership were above and beyond the benefits of being well-connected.

One obvious explanation is that the coaching families confer a strong sense of identity that encourages members to help each other, similar to the way biological family members support one another, particularly in times of need. But in addition to that, another subtle factor might be at play. Family membership could signal to prospective employers a level of competence and ability beyond a person’s actual level of past performance. In other words, prospective employers (namely, the athletic departments of colleges) might base their hiring decisions partly on the assumption, whether justified or not, that “Calipari” or “Pitino” coaches are better choices (perhaps because they are “known” quantities) than unaffiliated coaches.

Halgin’s study has its limitations. Although cleverly conceived and conducted, it investigates a sample that is both small and insular. Those qualities enabled him to conduct a tightly controlled experiment, but they also raise questions about the applicability of his results. How similar is the microcosm of college basketball coaches to the far larger world of corporate executives, marketing managers or IT professionals?

That said, it doesn’t take much imagination to envision how, for instance, a Fortune 500 corporation might favor a “Welch U.” manager over an executive who was similarly qualified but lacked a GE pedigree. On the one hand, companies might do well to be aware of how such biases could affect their hiring decisions—and they should question the validity of such biases. Those issues might well be on the minds of those who actively recruited Robert Nardelli, a “Welch U.” executive, for the top position at The Home Depot Inc., the home-improvement retailer. During Nardelli’s rocky tenure there, the value of Home Depot’s stock stagnated, while competitor Lowe’s Companies Inc. saw its stock double.

On the other hand, companies should also consider how the “family” effect could be used to their advantage. Specifically, could a management consulting company like McKinsey & Co. that has a strong alumni network (and a shared sense of family) have a slight edge in hiring top-notch talent partly because prospective candidates might recognize that working there will enhance their future career opportunities? In other words, the social identity that comes with membership in a professional “family” could be a job perk, increasing the resilience of a person’s overall career. In particular, it could enable individuals to recover more quickly after losing their jobs, helping them to shake off any stigma from being fired. Former GE executive Nardelli, for example, was able to land the top job at Chrysler LLC just seven months after The Home Depot ousted him.

Of course, the “family” effect can cut both ways, sometimes impeding a person’s attempts to find work. Consider, for example, ex-executives of Enron Corp. and other corporate disasters. Those individuals are more likely to be eager to shed—instead of tout—their associations with their former employers.

—Alden M. Hayashi