By Dave DeFusco

When people think about how work gets done inside a company, they often imagine job titles and reporting lines that stay fixed. Behind the scenes, however, organizations constantly face a tricky problem: deciding who should work on which projects, with which managers and alongside which co-workers. Those decisions matter deeply, not just for productivity, but for whether workers feel satisfied, motivated or ready to leave.

That tension is at the heart of a Management Science paper, “Stable Matching on the Job? Theory and Evidence on Internal Talent Markets,” co-authored by Dr. Patryk Perkowski, an assistant professor of strategy and entrepreneurship at Yeshiva University’s Sy Syms School of Business, with collaborators from Columbia Business School, the University of Oregon and Emory University.

“At its core, this paper is about a trade-off that almost every organization faces,” said Perkowski. “Leaders often prioritize placing people in roles where they can deliver results quickly, but workers want roles that align with their interests, relationships and long-term growth.”

The research focuses on two common ways organizations match workers to jobs. In the first, leaders assign people to roles from the top down. Executives or managers decide who goes where, based on what they believe will best serve the organization’s goals. The second approach relies on internal talent markets, where workers and managers express preferences for roles or teammates, and an algorithm, often borrowed from economics, matches them.

Internal talent markets have grown rapidly over the past two decades. Companies like Google, Walmart, Accenture and even the U.S. Army use them, supported by a growing industry of HR technology platforms. The promise is appealing: if people can choose work they like, they may be happier and less likely to quit.

“Happiness, however, isn’t the same thing as productivity,” said Perkowski. “That’s where the real tension comes in.”

To study this tension, the researchers combined theory with real-world data from a large organization that used an internal talent market. The company had detailed information on worker skills, job requirements and preferences submitted by workers and managers.

The researchers compared two outcomes. One reflected what would happen if executives assigned workers to jobs to maximize what the firm considered “match quality.” The other reflected the matches created by a preference-based system using a well-known method called deferred acceptance.

The results were striking. From the firm’s perspective, the executive-assigned matches were 33% more valuable in match quality than randomly assigning people within job categories. That improvement was so large that Perkowski said it would be extremely hard to achieve through hiring or training alone.

“Matching better turned out to be more powerful than upgrading the workforce,” he said. “To get the same gains another way, you’d need a huge jump in average worker quality.”

By contrast, the matches created through worker preferences were only about 5% better than random by the firm’s own metric. At first glance, that might seem disappointing, but that picture changes when worker satisfaction is considered.

When workers and managers ranked their assignments, the preference-based matches performed far better. On average, people ranked their internal-market assignments 38 percent higher than the firm-dictated ones. They were also far less likely to end up in roles they viewed as their worst possible option.

“In terms of how people felt about their assignments, the difference was enormous,” said Perkowski. “From the workforce’s perspective, firm-dictated matches felt almost as bad as random ones.”

The study helps explain why. Workers tended to favor what economists call “positive assortative matching,” meaning strong workers preferred to work with other strong workers and strong managers. They also prioritized roles that helped them build new skills, especially skills that could be useful beyond their current employer.

Executives, however, preferred the opposite in many cases. From their viewpoint, pairing strong managers with weaker workers helped spread expertise across the organization and delivered better immediate performance. “Workers are often thinking about growth and future opportunities,” said Perkowski. “Executives are thinking about what gets results right now.”

One of the paper’s most important conclusions is that neither approach is simply right or wrong. Top-down assignment delivers clear productivity benefits, while internal markets offer real advantages in retention and morale. This insight helps explain why internal talent markets continue to spread, even when they don’t maximize short-term output. In industries facing high turnover, giving workers more say may be worth the cost.

“If you’re losing your best people, the retention benefits can easily outweigh some productivity loss,” said Perkowski. “That calculation, however, depends on your organization’s situation.”

The researchers hope their work will encourage leaders to think more carefully about how assignments are made, and to move beyond one-size-fits-all solutions.

“The extremes rarely work perfectly,” said Perkowski. “The real opportunity is in designing smarter systems that respect worker preferences while still advancing organizational goals.”