Managers Hoard Talent to the Detriment of Worker Development

Main article

Firms must continuously determine how to promote or move their workers to roles in the company that will maximize their talents and contributions to company productivity and output (Rosen, 1982; Holmström and Tirole, 1989). However, measuring a worker’s productivity is difficult and includes observing intangibles and capabilities that aren’t necessarily tracked through recorded contributions or statistics. As such, companies rely on managers to identify talented employees and support their progression to more appropriate or higher-level positions.

This creates a potential conflict of interest. Managers are often rewarded based on team performance. Thus, while the company benefits when a talented worker is promoted or transferred to a position that maximizes their capabilities, the worker’s current team loses out if that position is on a new team. Thus, managers have an incentive to convey inaccurate information—or suppress positive information—to the firm that discourages the promotion or transfer of their team’s talented workers (Holmström, 1979).

Managers also have an incentive to prevent their workers from applying for or obtaining positions outside their current teams. These actions may be explicit, such as discouraging applications by suggesting that it would be disloyal. Tactics can be more subtle, such as withholding career advice, restricting access to high-visibility training and projects, or assigning lower assessments of an employee’s potential which human resources and other decision-makers see. Collectively, these actions are called talent hoarding.

Uncovering talent hoarding

Talent hoarding is difficult to measure because it often takes place through hidden managerial actions and because workers may refrain from applying for many reasons unrelated to managerial behavior. My paper provides the first empirical evidence on the existence and consequences of talent hoarding by combining survey evidence, personnel records, internal application data, and a quasi-experiment based on workers’ responses to manager rotations from a manufacturing firm employing more than 200,000 workers worldwide.

The main analysis focuses on the firm’s largest internal job market: the more than 30,000 white-collar and managerial roles based in Germany. At the firm, employees seeking a promotion or a different type of work generally need to apply for a formally advertised position, and nearly all applications are to jobs outside the employee’s current team. Applications, callbacks, and hiring decisions are recorded in a centralized portal, which allows for the observation of all internal applications and their outcomes.

I link these application data to rich personnel records containing employees’ demographic and educational characteristics, job assignments, earnings, working hours, team membership, and direct managers. The records also include assessments of their performance and potential, as well as participation in internal training. I supplement these records with two surveys. An employee survey received more than 15,000 responses, representing a response rate of approximately 50%. A separate survey of managers received more than 3,000 responses, or 62% of the firm’s managers.

The employee survey illustrates how managers typically rely on two tactics to hoard talent. The first is to suppress employees’ public visibility, including through potential ratings and access to visible projects or training. Second, managers discourage workers from seeking positions elsewhere. Workers’ survey responses also highlight that managerial influence shapes application decisions even when employees do not need permission to apply. Forty-one percent of respondents report being afraid of negative repercussions if their manager discovers an internal application, while 25% feel they need their manager’s permission before applying, although company policy does not require it.

Managers’ incentives to hoard talent are common

The manager survey provides direct evidence that a conflict between team and firm-level interests shapes how they develop talent. Managers overwhelmingly believe that they can influence employees’ careers: 96% say their intervention matters substantially. However, only 36% of managers think that employee career development is valued as highly as the manager’s effect on team performance, and 55% agree that helping workers develop will increase the likelihood that they leave the team. These responses imply that managers are disincentivized to develop talent or prevent developed talent from leaving the team.

Seventy-five percent of managers report that managers at least sometimes discourage employees from exploring opportunities in other departments because of immediate team needs or performance goals.

Indeed, 75% of managers report that they at least sometimes discourage employees from exploring opportunities in other departments because of immediate team needs or performance goals. Difficulties replacing workers appear to be an important part of the problem: 68% say they would be more supportive of employees’ career development if departing workers were easier to replace. Importantly, I find that managers are more likely to report having engaged in talent hoarding if the incentives to do so, such as compensation linked to team performance, are stronger.

Once I collected managers’ responses, I then examined whether managers’ actions actually observed these revealed incentives. The firm’s personnel records make it possible to compare two assessments assigned by the same manager: a relatively private rating of an employee’s current performance and a more widely visible rating of the employee’s potential for higher-level positions. In theory, the visible rating should closely track the private rating. I used any difference between the two ratings as a measure of managers’ propensities to suppress talent visibility.

Managers are more likely to suppress employees’ visibility when the incentive to retain them is stronger: when a larger share of managerial compensation is linked to performance, when teams are smaller… and when employees’ abilities are less visible to other parts of the firm.

The evidence from this comparison closely mirrors the behavior we would expect from the survey findings. Managers are more likely to suppress employees’ visibility when the incentive to retain them is stronger: when a larger share of managerial compensation is linked to performance, when teams are smaller and the loss of one worker is more consequential, and when employees’ abilities are less visible to other parts of the firm. The same behavior appears in alternative measures. These include managers’ decisions to restrict worker access to high-visibility training and their own survey responses reporting hoarding behavior.

Together, the survey and administrative records establish a systematic link between the incentives managers face and the actions they take. Talent hoarding does not appear to be characteristic of particular managers. Instead, it becomes more prevalent when organizational conditions make retaining talent more valuable or suppressing it easier.

What happens when hoarding incentives disappear?

It is clear that managers hoard talent, but the question remains how this impacts workers’ career decisions. To answer this, I study how managers’ talent-hoarding changes when they are rotated to another team. Managers typically learn in advance that they will move to another team, and once that move is known, they no longer bear the cost of their team losing a strong employee. Further, per company policy, managers must generally inform their staff as soon as they learn they will be rotated to another team. The time between a manager learning they will be rotated and a new manager taking their place creates a temporary window of time during which the managerial pressure on talented workers not to leave the team is reduced.

Using administrative records, I find that applications rise sharply in the fiscal quarter during which the manager’s rotation occurs. During this temporary window, workers are 78% more likely to apply for another internal position. The response is larger among highly productive workers, among employees who are more difficult to replace, and among workers whose managers appear more inclined to hoard talent. I find little support for alternative explanations for these patterns, such as a worker leaving because their manager is leaving, finite project timelines, or other team dynamics.

Talent hoarding does not just reduce the number of applications but also lowers the quality of the applicant pool.

Talent hoarding deters strong applicants

The next question is whether talent hoarding genuinely prevents qualified workers from successfully moving to new roles, or whether affected workers would have been unlikely to secure new roles regardless. To assess this, I looked at employees who only applied because the manager was leaving the team.

These applicants have an estimated 49.1% chance of being hired, compared to just 27.6% for all applicants. The result implies that talent hoarding deters a substantial number of strong applications that would likely have succeeded.

The applicants held back are also highly qualified. Compared to the average applicant, they are more likely to have a graduate degree (59.4% vs. 48.9%) and to have received high performance ratings (70.4% vs. 61.8%). Talent hoarding does not just reduce the number of applications but also lowers the quality of the applicant pool.

These findings suggest that talent hoarding may constrain highly qualified women to a higher degree than men with comparable qualifications.

Talent hoarding may also contribute to inequality in the firm. In the employee survey, women reported placing a higher value on preserving a good relationship with their manager and relying on managers for career guidance than men. In the administrative data, female workers who applied for new positions due to a manager rotation appeared particularly qualified in terms of education, prior performance, and their chances of being hired into higher-level positions. These findings suggest that talent hoarding may constrain highly qualified women to a higher degree than men with comparable qualifications.

Together, these findings show how managers shape workers’ careers by hoarding talent and contribute to a growing literature on how middle managers impact their workers (Lazear, Shaw, and Stanton, 2015; Frederiksen, Kahn, and Lange, 2020; Fenizia, 2022; Hoffman and Tadelis, 2021; Cullen and Perez-Truglia, 2023; Benson, Li, and Shue, 2026).

What can organizations do about talent hoarding?

This paper provides the first empirical evidence on the existence of talent hoarding and its consequences for workers’ careers and firms’ internal allocation of talent. The findings also point to several ways organizations can mitigate the problem.

A first policy response is to reduce the cost that the current team—and thus the manager—bears when a worker leaves. For instance, HR can offer more support to expedite processes to replace workers…

A first policy response is to reduce the cost that the current team—and thus the manager—bears when a worker leaves. For instance, HR can offer more support to expedite processes to replace workers and invest more in succession planning. In theory, financially rewarding managers who promote talent would offset the financial disincentives of losing talented workers. However, this could create its own problems. Instead, firms could recognize talent development through non-financial rewards or by incorporating it into managers’ own performance evaluations.

Second, firms can give employees access to career support outside their relationship with their direct manager. HR, mentors, or managers in other units could help to independently identify promising workers, provide career advice, and invite them to apply to new roles. More access to high-visibility projects and training could likewise reduce the ability of a single manager to control a worker’s reputation and opportunities.

Third, firms could make internal applications genuinely confidential and establish clear protections against retaliation. Confidentiality would not eliminate hoarding that occurs earlier through restricted training, dishonest ratings, or career guidance, but it could reduce workers’ fears of initiating an application.

Ultimately, addressing talent hoarding requires firms to align managers’ incentives with the organization-wide benefits of developing and allocating talent effectively.

References 

Benson, Alan, Danielle Li, and Kelly Shue. 2026. “‘Potential’ and the Gender Promotion Gap.” American Economic Review 116 (2): 375–417. 

Cullen, Zoë, and Ricardo Perez-Truglia. 2023. “The Old Boys’ Club: Schmoozing and the Gender Gap.” American Economic Review 113 (7): 1703–40. 

Fenizia, Alessandra. 2022. “Managers and Productivity in the Public Sector.” Econometrica 90 (3): 1063–84. 

Frederiksen, Anders, Lisa B. Kahn, and Fabian Lange. 2020. “Supervisors and Performance Management Systems.” Journal of Political Economy 128 (6): 2123–87. 

Hoffman, Mitchell, and Steven Tadelis. 2021. “People Management Skills, Employee Attrition, and Manager Rewards: An Empirical Analysis.” Journal of Political Economy 129 (1): 243–285. 

Holmström, Bengt. 1979. “Moral Hazard and Observability.” Bell Journal of Economics 10 (1): 74–91. 

Holmström, Bengt, and Jean Tirole. 1989. “The Theory of The Firm.” In Handbook of Industrial Organization, Vol. 1, edited by Richard Schmalensee and Robert Willig, 61–133. Elsevier. 

Lazear, Edward P., Kathryn L. Shaw, and Christopher T. Stanton. 2015. “The Value of Bosses.” Journal of Labor Economics 33 (4): 823–61. 

Rosen, Sherwin. 1982. “Authority, Control, and the Distribution of Earnings.” Bell Journal of Economics 13 (2): 311–23.