Every organization has one: the person everyone routes questions to. The one who knows why the system was built that way, which client relationship is fragile, what actually happens when the standard process breaks. They’re usually praised for it. They’re also, often without meaning to, the reason knowledge doesn’t move.
The instinct is to call this a character problem. Someone is being territorial, or precious about their expertise, or quietly building leverage. Sometimes that’s true. But the research on why people withhold knowledge at work points somewhere less flattering to the “just a bad actor” explanation, and somewhere more useful for anyone trying to actually fix it: knowledge hoarding is, in the majority of cases, a rational response to conditions the organization itself created.
The behavior has a name, and a mechanism
Researchers distinguish between knowledge hiding, the deliberate withholding of information when someone specifically asks for it, and knowledge hoarding, the broader pattern of retaining knowledge that could be shared but isn’t proactively offered. Both behaviors show up for similar reasons, and the strongest predictor in the literature isn’t personality. It’s job insecurity.
A study of Korean employees tracked across three time points found that when people perceive their job as insecure, their sense of psychological safety drops, and that drop is what drives the increase in knowledge hiding, not the insecurity directly. Psychological safety is the mechanism in between. Take it away, and withholding knowledge becomes a defensible, even sensible, response to a threatening environment.
A separate study of knowledge workers in China found the inverse relationship holds too: organizational inclusion reduces knowledge hoarding, and it does so specifically by increasing psychological security and a sense of cohesion with the team. Belonging predicts sharing. Threat predicts withholding. Neither has much to do with whether someone is a generous or selfish colleague by nature.
Why the incentive quietly rewards the opposite of sharing
There’s a second, more structural layer underneath the psychological one. In a lot of organizations, being the only person who knows something functions as informal job security. If a colleague is consistently the one others go to for answers, that dependency is protective. Sharing that knowledge widely can feel, correctly in some cases, like giving away the one thing that makes them hard to replace.
This isn’t paranoia. It’s a reasonably accurate read of how many organizations actually evaluate people. Performance reviews reward individual output and visible expertise far more often than they reward the unglamorous work of documenting, teaching, or making a colleague’s job easier. McKinsey’s research on knowledge worker productivity points to this directly: rigid hierarchy and weak incentives are named as core barriers to knowledge sharing, and the fix McKinsey recommends is structural, not motivational. Companies that want more sharing need to build knowledge contribution into performance reviews and make expectations around response times explicit, rather than hoping goodwill fills the gap.
Put simply: if the system quietly rewards being indispensable and never rewards being generous with what you know, most people will optimize for the system, not against it.
Trust runs in both directions
There’s a third factor that gets less attention than job security but matters just as much: whether employees trust the organization itself. Survey data from the American Psychological Association found that nearly one in four workers say they don’t trust their employer, and one in three believe their employer isn’t always honest with them. Knowledge sharing asks employees to make themselves more replaceable in service of the organization’s benefit. That’s a reasonable ask only where there’s enough trust to believe the organization won’t use it against them.
This is why knowledge-sharing initiatives that lean entirely on culture messaging (posters, values statements, “we’re a collaborative team” language) tend to underperform. They ask for a behavior change without addressing the trust or security conditions that are actually driving the withholding in the first place.
What actually reduces hoarding
The research points toward a small number of levers that address the real cause rather than the symptom:
- Build psychological safety deliberately, not aspirationally. Studies on servant leadership found that leaders who prioritize their team’s needs and demonstrate consistent, low-ego support measurably increase psychological safety, which in turn reduces knowledge hoarding. This is a leadership behavior, not an HR policy.
- Make knowledge contribution visible in how people are evaluated. If documenting a process, mentoring a junior colleague, or answering questions in a shared channel never shows up in a performance conversation, it will always lose out to work that does.
- Separate expertise from irreplaceability. The goal isn’t to make experts interchangeable. It’s to make sure their value is recognized for the judgment they bring, not just for the information only they hold, so that sharing doesn’t feel like self-erasure.
- Build structured spaces for knowledge to move. Communities of practice, McKinsey notes, help because employees are far more likely to give timely, useful answers to people already inside their network. A shared repository nobody feels ownership of does less than a working group people actually belong to.
- Address organizational trust directly, not just team-level trust. A generous manager can’t fully offset an employee’s reasonable doubt about how leadership treats people during layoffs, restructuring, or performance cycles. That trust has to be earned at the organizational level too.
The reframe that matters for KM leaders
Treating knowledge hoarding as a discipline problem, something to be corrected through mandates or guilt, tends to fail because it misdiagnoses the cause. The employees holding onto knowledge are usually responding rationally to an environment that has, in ways big or small, taught them that sharing costs more than it pays. Change the incentives, build the safety, and give expertise somewhere safe to move, and the hoarding tends to resolve on its own, without anyone needing to be persuaded to become a more generous person than they already are.