Every team has someone who consumes more than they contribute and has learned to make that invisible. The problem is not the person. It is the system that makes it possible.

In my content design course, student teams work on a real campaign brief over the course of a semester. At the end of every semester, without exception, at least one team has had a version of the same conversation with me. Someone was not pulling their weight. Everyone knew it. Nobody said anything. The final pitch happened anyway, the team got a collective grade, and the person who contributed least walked away with the same outcome as the people who drove the work. The donut was eaten. Nobody named the donut eater.

What the Donut Eater Actually Looks Like

The donut eater is not always the most obviously disengaged person on the team. The more costly version is the person who has learned to maintain the appearance of contribution while systematically avoiding the substance of it. They attend every meeting. They speak in a way that sounds engaged without committing to anything specific. They volunteer for tasks that are visible but low risk. They are reliably unavailable when the difficult, unrecognised work needs doing. They have developed, often without conscious intention, a very efficient system for consuming the team’s output while contributing minimally to its production.

The scale of the problem is measurable. In Gallup’s State of the Global Workplace 2026, South Korea data, 13% of Korean employees count as engaged at work, against 18% in East Asia and 20% worldwide, and the figure has barely moved since 2012. A team where one person in eight is pulling is a team where the donut eater problem is the norm, not the exception.

Why Nobody Says Anything

The silence around the donut eater problem is not primarily about conflict avoidance. It is about the structural ambiguity that makes the problem difficult to address without appearing unfair. In a collaborative team, contribution is genuinely hard to measure. Different people contribute in different ways at different times. But there is a difference between genuine ambiguity and manufactured ambiguity. The donut eater problem persists not because contribution is impossible to observe but because the systems most teams use, collective deadlines, shared deliverables, group grades, make individual contribution invisible by design.

The two get mixed up all the time, and the confusion protects the donut eater. Quiet quitting is common and mostly honest. In an Incruit survey from March 2024, cited in the Korea Herald, 51.7 percent of respondents said they were in quiet quitting mode (The Korea Herald, July 2025). Someone doing exactly their job, openly, is not eating anyone’s donut. The person who attends every meeting, says the right things and leaves the real work to others is doing something different, and it is far harder to see because it looks like engagement.

Korean teams have extra reasons to stay quiet. Raising it means judging a colleague in front of others, and in a team built on 화합 that can feel like a bigger breach than the free riding itself. If the donut eater is a 선배, or sits a grade above you, the hierarchy makes it close to impossible to say directly, as I explain in how Korean workplace hierarchy actually works. Most Korean professionals first met the problem at university, in 조별과제, where the 무임승차 classmate is almost a stock character. Many then carried the same silence into their first team.

The Framework That Helps

The framework I use in my courses has three components. Individual task ownership: every piece of work has a named person who is accountable for it. Progress visibility: the team checks in on individual tasks at a defined cadence to surface blockers early. And contribution reflection: at the end of each phase, each team member writes two sentences about what they contributed and what they will contribute next. These three components do not eliminate the donut eater problem. They make the contribution pattern visible enough that it can be addressed with evidence rather than feeling.

A Worked Example: Five People and a Chuseok Promotion

Here is the framework applied to one team. Park Junho is invented. He is 34 and leads a five person team at a food delivery app in Seoul, planning the Chuseok promotion. Four weeks before launch, everyone agreed the work was behind, and everyone privately suspected the same colleague: present at every meeting, keen to present the deck, hard to pin down on anything that had to be built. Junho did not confront anyone. He changed the system for the last four weeks.

Nothing in the table accuses anyone. That is why it works in a Korean team: the system does the naming, so nobody has to lose face. By the second Monday the pattern was visible to everyone, including the colleague concerned, who quietly picked up a real task. Some donut eaters stop as soon as the work becomes visible. The ones who do not have now given you evidence instead of a feeling.

This matches what Google found when it studied 180 of its own teams. What really mattered was “less about who is on the team, and more about how the team worked together”, and one of the five dynamics that set effective teams apart was dependability: “members reliably complete quality work on time” (Google re:Work, Understand team effectiveness). Dependability is not a personality trait you can hire for. A team can only see it when the work is visible enough to check.

The Organisational Dimension

The donut eater problem scales up. In organisations, the equivalent is the team or department that consumes resources without producing proportionate output, and has learned to make that invisible through reporting and narrative. The Korean concept of 눈치 (nunchi) is relevant here. A team or organisation with good collective nunchi can feel when something is not contributing its share, even before the evidence is explicit. When the room feels like something is off, it is usually right.

Nunchi tells you something is off. It does not tell you what to do next, and in many Korean organisations the feeling is shared by everyone and acted on by nobody. A department that reports busily and delivers little survives because its reports are about activity, not outcomes. The fix at that scale is the same as in a team of five: named ownership of outcomes, a regular look at progress, and a record of what each unit actually delivered. There is more on reading the room in 눈치, the communication skill the world is noticing.

How to Name It Without Making It Personal

When the evidence is there and you do have to talk about it, start with the written reflection, not the person. The two sentences each member writes at the end of a phase are the best tool you have, because a vague reflection is easy to spot next to four specific ones.

Ask everyone for the second kind, in the same format, every week. When one person keeps writing the first kind, the conversation is no longer about character. It is about a sentence on a page: what did you produce, and what will you own next? To see the roles people have settled into, Group Dynamics maps them, and Trace Your Impact helps anyone, the donut eater included, turn what they did into a specific statement. If the hiding happens mostly in meetings, the meeting that should have been an email is worth a read too.

→ The Strategy Desk covers team performance, communication culture, and organisational effectiveness in depth. If you want to bring these frameworks to your team through a workshop, the Work With Us page covers what that engagement looks like.

---