Trust in the Team: Why Control Replaces It and Rarely Improves It

Why does more control so often lead to less effort?

An owner notices that orders are going out without consultation and introduces an approval for every order. Another notices that an employee working from home is hard to reach and demands daily reports. Both measures are understandable. And yet many owners report that something strange happens afterward: employees do exactly what is required, and not a stroke more.

This essay pursues one question: Why does control in the team replace trust, and when does it actually improve collaboration? To answer it, I first clarify what trust and control are, then present the trust model of Mayer, Davis, and Schoorman, summarize what studies show about trust and about the costs of control, weigh two views, and close with a check every control rule should pass.

Trust and control: two answers to the same uncertainty

Trust is the confidence that the other party will not exploit a vulnerability. If you leave a customer, a budget, or a decision to an employee, you make yourself vulnerable, because the employee could do it worse or pursue their own goals. Control here means all arrangements that restrict the scope for action or monitor behavior: approvals, reports, minimum requirements, surveillance.

Both are answers to the same uncertainty. The article Principal-agent theory in the business describes it in more detail: the owner knows less about the work than the employee who does it, and the two do not automatically pursue the same goals. Trust accepts this uncertainty; control tries to eliminate it.

The difference shows in a simple question: What does the owner learn about the employee? If you control, you only see that the requirements are met. If you trust, you see what the employee does with the latitude.

The thinking model: how trust develops according to Mayer, Davis, and Schoorman

The best-known model in trust research comes from Mayer, Davis, and Schoorman (1995). The article Trust as an economic factor presents it for business relationships. According to it, the willingness to trust someone depends on how one assesses three characteristics: their ability, whether they can deliver what is required; their benevolence, whether they want what is best for the other; and their integrity, whether they keep to shared principles.

For the team, another part of the model matters. Trust leads someone to take a risk in the relationship, for example handing over a task without control. The outcome of this risk then changes the assessment of ability, benevolence, and integrity. Trust therefore grows in a loop of advance and experience.

1. Assessment

   The owner assesses the employee's ability, benevolence, and integrity, initially with little knowledge.

2. Advance

   The owner takes a limited risk and hands over a task with latitude.

3. Experience

   The employee uses the latitude, well or badly. Both sides see the result.

4. New assessment

   The experience changes the assessment. If it is good, the next advance is larger.

This leads to the thesis of this essay. Control interrupts the loop at the second step. Where every order is approved, there is no advance, therefore no experience, therefore no reason to improve the assessment. Control replaces trust, but it cannot improve it, because it prevents the experiences from which trust develops.

An everyday analogy: if you never lend a teenager the car because they have not yet shown that they drive safely, you will never find out. At some point there has to be a first drive, perhaps with someone in the passenger seat and on a short route, but without a drive there is no experience.

Control answers the question of whether someone follows the rules. Whether you can trust them, it never answers.

What trust achieves in a team

De Jong, Dirks, and Gillespie (2016) summarized research on trust within teams in a meta-analysis. They evaluated 112 independent studies covering a total of 7,763 teams. Trust within the team was positively related to team performance, with an effect the authors classify as above average. The relationship held when trust in the leader and the team's past performance were taken into account.

For small businesses, another finding is revealing. Trust was especially important when team members depended heavily on each other and when their skills differed greatly. That describes many small businesses quite precisely: few people, each with their own specialty, and hardly anyone who could fully check another's work.

The hidden costs of control

In an experiment, Armin Falk and Michael Kosfeld (2006) examined how people react to control. A participant in the role of employer could decide whether to prescribe a minimum performance to the other or leave the choice entirely to them. Performing cost the employee something and benefited the employer. According to classic economic expectations, every employee would have performed nothing without a requirement and exactly the minimum with one.

The result was different. Most employees performed less when the employer set a minimum requirement than when the employer trusted them. Asked about their reasons, most said they had experienced the control as a sign of distrust and a restriction of their freedom. The majority of employers seemed to anticipate this and refrained from control. Employers who controlled expected less from the outset, and their expectation was confirmed. With the sociologist Niklas Luhmann, the authors speak of a self-fulfilling prophecy of distrust.

The authors themselves name two qualifications. With a low minimum requirement, control yielded clearly less than trust; with a higher requirement, the two evened out. And when the requirement was not chosen by the employer but set externally, performance did not drop. So it is not the limit itself that demotivates, but the message contained in choosing it.

Falk and Kosfeld quote a recollection by David Packard, one of the founders of Hewlett-Packard. As a young engineer at General Electric, he saw how the company strictly guarded its tool and parts bins. Many employees, he wrote, took this open distrust as a challenge and walked off with tools whenever they could. At Hewlett-Packard he therefore kept the storerooms open, as a symbol of trust.

More recent studies on electronic monitoring fit this picture. In a field study with supervisors and employees from various organizations, Thiel, Prince, and Sahatjian (2022) found that electronic monitoring weakens the trust that ethical leaders otherwise build. In a survey of 633 employees in the United States, respondents valued working from home and considered some monitoring a necessary part of it. But the study also shows that perceived risks of monitoring are linked to privacy concerns, which in turn affect trust and satisfaction (Brinson et al. 2024).

Two views: trust is good, control is better?

The first view defends the proverb. Where money, safety, or customer data are at stake, it would be negligent to rely on trust alone. The article Corporate governance for small businesses gives the number: in the ACFE's 2024 fraud report, a lack of internal controls was the primary weakness in 32 percent of all cases. If you never check, you invite errors and abuse.

The second view emphasizes the costs. Control signals distrust, lowers many people's willingness to do more than required, and prevents the experiences from which trust grows. Stewardship theory, presented in the article on principal-agent theory, describes employees who want to act in the business's interest and lose precisely this attitude through too much control.

I consider the two views compatible if one distinguishes between two kinds of control. The Falk and Kosfeld experiment itself provides the key: a limit that applies to everyone and was not chosen out of distrust of a person did not lower performance. Control as a general rule, such as the four-eyes principle for payments, protects without offending. Control as a reaction to a person, such as daily reports for a single employee, tells them they are not trusted and thereby easily produces the behavior one fears.

The Control Check: four questions for every rule

Before you introduce a new control or keep an existing one, a short check is worthwhile. It takes a few minutes and prevents most rules that cost more than they protect.

• What does it protect? Name the specific damage the rule is meant to prevent and estimate its size. Without nameable damage, the rule is superfluous.
• Does it apply to everyone? A rule that applies to everyone, including the owner, is understood as order. A rule for one person is understood as distrust.
• Is it explained? Say in advance why the rule applies and what it protects. Unexplained control invites the worst interpretations.
• Does it check results or people? Key figures, spot checks, and agreed results leave latitude. Checking attendance and activities takes it away.

How much time controls cost is often underestimated. A worked example from a business with eight employees shows the order of magnitude.

In the example, the two person-related controls together cost 18 hours a month, the result-related ones three. The difference in their effect on trust is likely to be even larger than the difference in effort.

What follows beyond the individual business

Trust in the team is not a feeling that can be produced on demand. It is the result of many small advances that turned out well. For the owner this means: they have to take the first step, with tasks where a mistake is bearable. What such handovers look like is described in the guide Delegating tasks with a clear handover. This becomes especially important when the owner is no longer on site every day. The guide Leading from a distance shows which structures then replace presence.

The limits of this assessment are clear. The Falk and Kosfeld experiment took place in the lab, with students and small amounts of money. The meta-analysis by de Jong and colleagues shows relationships, not causes: it is also possible that good performance creates trust rather than the other way round. And there are employees who really do exploit an advance. They need consequences, but not a rule that punishes everyone else too.

Conclusion: why does control replace trust, and when does it help?

The question was
control in the team replaces trust and when it improves collaboration. According to the model of Mayer, Davis, and Schoorman, trust develops from experiences that are only possible when someone takes a risk. Control eliminates this risk and with it the experiences. The Falk and Kosfeld experiment also shows that control understood as distrust lowers voluntary effort.

Control can improve collaboration where it applies as a general rule, is explained in advance, and checks results rather than people. In this form it protects the business without telling employees anything about their worth. In any other form, the owner gets exactly what they feared and takes it as confirmation of their caution.