Too Much Trust

The Trust Default

Management literature is almost uniformly pro-trust. High-trust organizations outperform low-trust organizations. Psychological safety enables innovation. Trust reduces coordination costs. These claims are broadly accurate.

What the literature underemphasizes is that trust is not a pure good. Under certain conditions, too much trust produces worse outcomes than appropriately calibrated skepticism.

When Trust Becomes a Liability

The most common liability form of trust is deferential trust - trusting someone else's judgment on a domain in which that person's incentives, expertise, or information are not adequately aligned with good outcomes.

Consider a manager who trusts their team's self-reported project status because confronting discrepancies feels like distrust. The team learns that optimistic reporting is rewarded with continued autonomy while accurate reporting of problems produces friction. Deferential trust in this context selects for flattering information rather than accurate information.

Or consider an organization that trusts its senior leaders' strategic judgment based on their track record of past successes. The past successes were achieved in a competitive environment that no longer exists. The leaders' patterns of judgment were calibrated to conditions that have changed. Trust in the past-validated judgment produces decisions calibrated for the wrong environment.

In both cases, the trust is not misplaced in the abstract. The team is not lying; they are reporting what they believe will be well-received. The senior leaders are not incompetent; they are applying genuinely good judgment to the wrong conditions. The problem is that trust without verification has removed the feedback mechanism that would reveal the misalignment.

Trust and Verification

The cliche is "trust but verify." The actual implication is more nuanced. Verification is not the opposite of trust. It is what makes trust sustainable.

Trust without verification is not really trust - it is hope. Hope that incentives are aligned. Hope that information is accurate. Hope that judgment is well-calibrated. These hopes may be warranted. But without verification, you cannot know, and the asymmetry of outcomes under misaligned trust is significant.

The function of verification is not to catch liars. It is to create the feedback loops that keep trust calibrated. A team that knows their project status will be examined has incentives to report accurately. A leader whose strategic judgment is tested against outcomes has incentives to update their mental models when conditions change.

Verification done well does not undermine trust. It anchors trust in something real rather than in hope, which makes it more robust and more durable.

The Calibration Problem

Calibrating trust to context is the core practical challenge. Too little trust is as dysfunctional as too much - it destroys the autonomy, initiative, and candor that trust enables.

The calibration factors include: the cost of being wrong, the degree to which incentives are aligned, the quality of available information, and the track record of judgment in this specific context.

High-stakes irreversible decisions warrant more skepticism than low-stakes reversible ones. Situations where incentives are clearly misaligned warrant more skepticism than situations where they are clearly aligned. Novel situations where no one has a reliable track record warrant more skepticism than familiar situations where track records are available.

Trust should increase gradually as track records are established and should be revised when conditions change enough that existing track records no longer predict current performance.

The Organization's Version

Organizations that have figured this out tend to have explicit norms around verification that are not treated as expressions of distrust. Reviews, audits, post-mortems, and challenge processes are valued because they keep the feedback loops functioning, not because anyone assumes bad faith.

The organizations that have not figured it out tend to treat verification as offensive - as implying that the person being verified cannot be trusted. This confuses the social relationship of trust with the informational function of verification. You can trust a person completely while still wanting to check the outputs of their work, because the verification is about maintaining the quality of information, not about assessing the character of the individual.

Too much trust, paradoxically, can be an expression of avoidance rather than generosity. It is easier to trust and not verify than to engage with the uncomfortable reality that trust needs calibration.