Culture Is Revealed When Values Become Expensive

Culture is learned when values become expensive

Few lines are repeated more often in boardrooms than “culture eats strategy for breakfast”, usually with Peter Drucker’s name attached. There is no reliable evidence that he said it. The attribution appeared long after the phrase had entered circulation.

The mistake is small, but revealing. Much of the corporate conversation about culture takes place at the level of the slogan: a memorable line, a new set of values, a campaign with polished language. Employees have a more dependable source of information about what their organisation actually values.

They watch what happens when doing the stated thing carries a cost.

The late Edgar Schein described culture as a pattern of assumptions learned by a group and reinforced over time. Leaders embed those assumptions through what they pay attention to, how they react to critical incidents and whom they reward, promote or protect.

That makes two moments unusually instructive. The first is what happens to the person who brings bad news. The second is what happens to the high performer who breaks the rules.

Values statements describe the culture an organisation wants. These two moments reveal the one it has.

Moment one: the bad news

Amy Edmondson began her research on teamwork with a result that initially appeared backwards. Studying hospital units, she expected better-led teams with stronger working relationships to make fewer medication errors. The data seemed to show that they recorded more.

The explanation was not lower clinical competence. It was greater candour. In teams where people felt safer admitting mistakes, errors were more likely to be raised, recorded and examined. Elsewhere, problems remained hidden.

That finding became part of the foundation for Edmondson’s work on psychological safety: a shared belief that a team is safe for interpersonal risk-taking, including asking a question, acknowledging an error or challenging a decision.

The leadership implication is uncomfortable. A healthy culture can look worse on the dashboard than an unhealthy one. A rise in reported incidents may mean performance deteriorated. It may also mean reality has begun to travel. A fall may mean improvement — or silence.

The first response matters because it becomes a story. If the messenger is thanked, the issue examined and action taken, others learn that speaking up is useful. If the response is defensiveness, blame or an inquiry into why the messenger did not fix the problem alone, they learn something else.

Every organisation says it wants problems early. Culture determines whether it receives them early enough to act.

Why leaders are often the last to know

If silence is costly, why do so many leaders believe their organisations are more open than employees experience them?

Megan Reitz’s research with John Higgins points to power. The people most responsible for creating the conditions for candour are also poorly placed to judge whether those conditions exist. A senior executive may experience themselves as approachable. A junior colleague experiences the title, reputation, access to resources and memory of how previous dissent was received.

That asymmetry means “speak up” is an incomplete instruction. It transfers the burden to the person with the greatest risk and the least authority. Reitz and Higgins argue that leaders must also learn to “listen up”: noticing how their own status shapes a conversation, seeking voices unlikely to volunteer and responding in ways that make the next contribution more likely.

An open-door policy is not proof of an open culture. It is an invitation people can rationally decline.

This is where culture becomes a form of distributed decision architecture. When inconvenient information cannot move, judgment remains concentrated at the top while reality sits elsewhere. Leaders may believe they are delegating decisions even as the organisation filters the information those decisions require.

Moment two: the star who breaks the rules

The second moment is often more visible. It is the decision about what happens when a high performer violates the standards everyone else is expected to uphold.

Donald Sull, Charles Sull and Ben Zweig analysed more than 1.4 million employee reviews alongside attrition data from large US companies during the Great Resignation. Toxic corporate culture was a far stronger predictor of industry-adjusted turnover than compensation. Their related work made “toxic” concrete: employees described environments as disrespectful, non-inclusive, unethical, cut-throat or abusive.

Such cultures rarely begin with a formal decision to abandon standards. They are built one exception at a time. A leader concludes that a particular rainmaker, technical expert or executive is too valuable to confront. The individual keeps producing. The organisation pays elsewhere — through silence, lost trust, avoidable departures and the quiet conclusion that values apply only when performance is ordinary.

Each exception teaches two lessons. It tells the star that results purchase immunity. And it tells everyone watching that the official rules are not the real rules.

Frances Frei’s work on trust helps explain why reversing that lesson is difficult. Trust depends on authenticity, logic and empathy; it can be rebuilt, but only through behaviour others can observe. During Uber’s 2017 leadership and culture crisis, Frei took leave from Harvard Business School to serve as the company’s first Senior Vice President of Leadership and Strategy. Her later work with Anne Morriss defines leadership as creating the conditions for other people to excel — including when the leader is absent.

That standard makes the high-status exception untenable. A culture cannot guide behaviour without supervision if the organisation visibly suspends its principles for the people it values most.

The strategic consequence is distributed judgment

The two moments do more than reveal whether an organisation is pleasant to work in. They determine how much judgment can be distributed without every ambiguous decision travelling up the hierarchy.

Global organisations cannot write a rule for every market, customer or operating condition. Human-AI workflows are adding new decisions about disclosure, review, accountability and acceptable automation faster than policy can settle them. In both cases, people need to interpret principles under conditions leaders did not fully anticipate.

Erin Meyer’s work on cross-cultural management adds an essential qualification. Strong culture cannot mean identical behaviour everywhere. Expectations about hierarchy, disagreement, trust and decision-making vary across national contexts. What scales is not behavioural uniformity but context: clarity about what matters, which trade-offs are acceptable and where local judgment begins.

A culture capable of guiding decisions without constant supervision therefore requires four things:

Remove any one of these and distributed judgment becomes fragile. Context without authority creates informed permission-seeking. Authority without signals creates ambiguity about the real rules. Signals without learning reward confidence over truth. Learning without standards turns candour into noise.

The purpose of culture is not to make every employee think alike. It is to make better decentralised decisions more likely.

What this changes for leaders

Culture work should begin less often with a new statement and more often with a review of consequential decisions. Four disciplines follow:

None of these choices is costless. Welcoming bad news means hearing things that are unpleasant and sometimes unfair. Removing the performance exemption can mean losing a commercially important person. Giving real authority means accepting that others will not always decide exactly as the leader would.

That cost is precisely why these choices carry cultural weight. Employees know the difference between values that are cheap to declare and values that survive when they become expensive.

Four lenses on culture when it counts

Leadership and culture is not one keynote brief. The relevant perspective depends on whether an audience needs to examine candour, power, trust or judgment across different contexts.

Voice & learning

Amy Edmondson

Why psychological safety allows errors, questions and inconvenient information to surface while there is still time to learn.↗

Power & silence

Megan Reitz

How status shapes what people say, what leaders hear and whether speaking up creates organisational value.↗

Trust under pressure

Frances Frei

How leaders build and rebuild trust — and enable other people to perform when the leader is no longer in the room.↗

The +1 · Context

Erin Meyer

How leaders create alignment across national cultures without mistaking consistency for behavioural uniformity.↗

These are complementary conversations, not interchangeable speaker choices. For adjacent decisions, explore why change stalls under accumulated load, how the future of work changes work design, and why digital transformation is an operating-model question.