Click for Academy of Management home page

Academy of Management

Fat pay for top executives spurs high managerial turnover, researchers find

February 1, 2002

For more information, contact: Benjamin Haimowitz,

And that's not good for companies, study of NBA teams suggests

With signs of an economic upturn now in evidence, research in the February/March 2002 issue of the Academy of Management Journal suggests a lesson for the next wave of corporate growth: Resentment in company ranks over fat paychecks for top executives causes more problems than just grumbling.

In the largest study to date to gauge the effect of corporate pay disparities on firms' employment patterns, researchers find large spreads to be associated with lower job tenure and higher turnover among company managers.

Dispersed pay structures promote the survival and retention of star managers, but at the cost of increasing workforce instability and turnover among remaining managers, write the study's authors, Matt Bloom and John G. Michel of the University of Notre Dame. In contrast, minimizing pay differentials creates a more egalitarian environment, which tends to reduce both the competition among managers and the impetus for them to leave creating greater workforce stability.

Meanwhile, another study in the same issue of the journal finds in the world of pro basketball evidence of the harmful effect high turnover of personnel has on teamwork. Surveying the records of 23 teams in the NBA from 1980 through 1994, researchers uncover a strong link between teams' won-loss records and the amount of shared experience among players. Even on losing teams, the researchers find, keeping teammates together resulted in a significant gain in the following year compared to teams with more turnover.

Shared experience is the mechanism by which a stock of tacit knowledge is accumulated over time, conclude the authors, Shawn Berman of Santa Clara University, Jonathan Down of Oregon State University, and Charles W.L. Hill of the University of Washington. Developing an internal tacit-knowledge stock may be more important than renting the services of top talent for a brief period.

Greater pay spreads at the top mean greater management turnover

Previous research has yielded mixed findings about the optimal degree of pay dispersion in an organization, Bloom and Michel observe at the outset of their study. The reason? Up until now, researchers have focused on particular groups, like auto racers and golfers (where greater pay spreads are associated with better performance) or college faculty and major league ballplayers (where greater spreads are associated with worse performance).

To get a more definitive answer, the management professors analyze data from two broad samples of organizations and employees. One set of data, drawn from 460 publicly owned organizations in 173 industries over the years 1992-97, consists of information on total compensation for the five highest-paid managers in each firm as well as records on the tenure of this group. The second set, drawn from 274 publicly owned organizations in 127 industries over the years 1981-88, consists of information on salaries, bonuses, and tenures of all managers within five or ten reporting levels of the board of directors (CEO=1).

Thus, the first analysis is based on total compensation among a small set of very senior managers during a recent period. The second involves a much larger group of managers during an earlier period and is based only on their salary and bonus but not other forms of compensation, such as stock options (which, in any event, were not as important in the 1980s as they were in the 1990s).

In both instances, the researchers find, greater pay disparities are associated with shorter tenures and higher turnover in management ranks.

The findings, Bloom and Michel add, "are robust across different samples, at different periods of time, at different managerial levels and after accounting for external labor market effects. When added to earlier evidence about the effects of pay dispersion on individual and team performance, we believe these findings support the assertion that pay dispersion per se has important implications for employee outcomes and is of strategic importance for organizational decision-makers."

Considering that large pay disparities have such negative outcomes, why do companies maintain them? The authors probe this issue but find no clear answer.

In certain circumstances, they reason, it makes sense to have the kind of star performers at the top who create big pay spreads. This should be the case, for example, in high-growth fields with big investment opportunities or in highly competitive environments in which there is a large degree of uncertainty. And, in fact, when they investigate these hypotheses, they find a significant association between investment opportunity and pay dispersion and a somewhat weaker association between environmental uncertainty and dispersion.

But Bloom and Michel also looked at industries that were enjoying a steady and widespread increase in sales, a condition that they term munificence. They expected to find, under favorable conditions, that the drawbacks of high pay dispersion (managers leaving the company quickly and often) would outweigh the benefits (star-quality top executives) in determining pay structure. In other words, they expected that favorable conditions would be related to a more egalitarian pay structure.

But Bloom and Michel could not find any evidence for this. Big salary gaps are associated with uncertain times, but more equal pay is not associated with more certain times.

Trying to account for this, Bloom comments, "what we saw as munificence is not necessarily the safe environment we thought it was. Even when times are good, there is still pressure on top executives to keep performance up, and to take full advantage of a bountiful market. There's never any down-time."

Or, perhaps, to put it less politely, top management is simply adept at justifying extravagant pay for itself whatever the competitive or investment environment.

If large pay spreads are the norm, as the study suggests, what can companies do to counteract their negative effects? Bloom's advice: Pay managers a little above market rates, so they'll feel satisfied with their salary relative to their industry counterparts even as they grumble about the pay of top executives in their own company. Also, send a message by investing in training and development for employees. Fold in the kind of rewards that are tied to overall group performance.

Low Turnover as a Source of Competitive Advantage in the NBA

Although Bloom and Michel do not assess the effect of high turnover on company performance, a second study in the current Academy of Management Journal does focus on the performance issue, but in a quite different context -- that of professional basketball.

Surveying the records of 23 teams in the NBA from 1980 to 1994, Profs. Berman, Down, and Hill find a strong link between won-lost records and the amount of time teammates have played together over the years. As far as professional basketball is concerned, lower turnover turns out to be associated with higher performance.

To measure shared experience for each of the 23 teams in each season of the 15-year period, the researchers first multiplied the number of on-court minutes per player in a season by the number of years spent with a team through that season's end. Then they added the results for the individual players to get a total figure for each of the 23 teams in each of the 15 years. Finally, they divided the total for each team-year by the number of players to get a measure of shared experience for each of the 345 team-years (23 teams multiplied by 15 years).

In gauging the relationship of shared experience to team performance, researchers controlled for three other major factors that contributed to a team's won-loss record -- the quality of its players, as measured by their positions in the NBA draft; players' ages; and the coach's years of experience with the team.

Controlling for these other factors, Berman, Down, and Hill found a highly significant association between the number of games won by a team and the measures of shared experience.

Since winning teams obviously tend to keep their players together, the investigators sought to sort out cause and effect. Does shared experience per se help produce a winning team? Or is it just that winning teams tend to be kept intact, engendering high levels of shared experience, and losing teams are broken up, engendering low levels of shared experience?

One way the authors sought to distinguish cause and effect was by focusing on franchises that didn't have any special incentive to keep their teams intact -- namely, those with losing records. They found that losing teams with an increase in experience in the following year won an average of 5.7 more games than in the previous year. In a season consisting of 82 games, this would mean roughly that a team with a record of 40-42 would improve to 46-36, probably enough improvement to earn the team a playoff berth.

In contrast, losing teams with a decrease in team experience the following year won, on average, only 1.2 more games in that year.

In other words, losing teams that were kept relatively intact improved their records significantly, while losers that shuffled their rosters stayed in about the same losing rut. These results, the authors conclude, support a reciprocal relationship between shared experience and performance, clearly showing that increasing shared organizational experience has a positive effect on increasing performance.

How are these findings from the rarefied world of pro basketball relevant to corporate management? The key connection is what the authors call tacit group knowledge, knowledge that arises when individuals work together for an extended time and thereby develop a sixth sense with respect to co-workers.

Tacit knowledge is knowledge that is not codified, the authors write. If it could be codified, then it would no longer be tacit knowledge; it would have become explicit knowledge. Tacit knowledge cannot be taught by reading manuals or listening to lectures; it must be learned through experience.

Being all but impossible to codify, tacit knowledge, the authors observe, is all but impossible to imitate. Thus, tacit knowledge constitutes the basis for sustained competitive advantage, even though in time the things that make organizations great -- their shared skills that are accumulated over time might also lead to rigidity and decline.

Accordingly, the decision to alter work groups within organizations should be carefully considered with regard to how those changes will impact the group's collective mind.

The Academy of Management Journal, a peer-reviewed publication now in its 45th year, is published every other month by the academy, an organization with some 12,000 members in 60 countries that seeks to foster the advancement of research, education, and practice in the management field.

Media Coverage:
Business Week. Chalk it up to teamwork: stable rosters are a plus in the NBA. (Monday, March 11, 2002).
The Indianapolis Star. Huge executive pay has its merits, but can't guarantee success. (Thursday, February 14, 2002).
The Washington Post. Turnover is for losers. (Thursday, February 14, 2002).

Academy of Management
Member Services
Academy of Management
Online Opportunities
Academy of Management