The Art of Delegation

Delegation of tasks is completely different to delegation of authority.

UPDATE: I have started sharing short notes on LinkedIn each day. If you find my content useful, let’s connect there too. Here is the link.

This article is part of The Art of Leadership series, where I cover twelve principles of leadership that my research found underpin superior company performance. Read the full series here.

Two senior people disagree about something that matters. Instead of settling it between themselves, each goes separately to the chief executive to argue their case. The chief executive hired both of them so that exactly this kind of thing would not need their attention, and now finds themselves carrying messages between two adults who ought to be talking to each other.

The leader in this situation has usually done everything the standard advice prescribes. And yet every decision of consequence still comes back to the leader, because the people below pass the hard decisions upward the moment they become difficult.

This is the distinction most discussions of delegation miss. Delegation is commonly understood as the distribution of work. Its harder and more consequential form is handing over authority, and the two are not the same thing.

A leader who hands over the execution but retains every decision has not delegated in any meaningful sense.

They have relocated the labour and kept the power, and in doing so they have often made themselves busier rather than less busy, because they must now supervise the work as well as approve its every turning.

Why the distinction matters

The task view of delegation is appealing because it feels safe. Handing someone a defined piece of work, with a deadline and a review at the end, leaves the essential control undisturbed. The leader still decides; the subordinate simply executes. Nothing of real consequence has been surrendered, which is exactly why this form of delegation so rarely relieves the pressure it was meant to address.

Authority is different because it transfers the right to be wrong. To delegate a decision is to accept that it may be made differently from how the leader would have made it, and occasionally made badly, without the matter returning for correction. That is a materially harder thing to concede, and the evidence suggests that many leaders do not concede it.

The consequences of withholding that authority compound as an organisation grows. In the earliest phase of a company, centralised decision-making is not a flaw but an advantage; the founder can take every decision, and this is the right call. The difficulty is that organisational demand does not grow in a straight line. Each new hire creates coordination, each new customer creates exceptions, each new product line creates decisions, while a single person’s capacity is capped by a fixed ceiling of hours and attention. At some point, usually reached sooner than expected, the leader who was the engine of the company becomes the brake on it.

A question of temperament, not technique

Much of the writing on delegation treats the failure to hand over authority as a skills deficit, correctable with the right process or the right template. A less comfortable explanation is that it reflects something about the person at the top.

A 2021 study published in the Journal of Organization Design examined the relationship between a chief executive’s personal dominance, a stable character trait marked by a strong desire for influence and control, and the way that executive designed the organisation beneath them. Linking measures of dominance drawn from earnings calls to data on reporting structures across large US corporations, the authors found that more dominant chief executives maintained a significantly wider personal span of control and delegated fewer decision rights than their less dominant counterparts. The reluctance to distribute authority, in other words, is visible in the shape of the organisation chart, and it tracks a trait that tends to intensify rather than soften as a person reaches positions of ultimate authority.

This finding reframes the problem. If the concentration of decision rights were a matter of technique, better technique would resolve it. If it is partly a matter of temperament, then the leaders most in need of distributing authority are, by disposition, the ones least inclined to do so, and the correction has to begin with self-awareness rather than with a new process.

What it costs the people below

When leaders hold on to authority, the problem is usually examined from their own side, as a drain on their time. Its more serious cost may fall on the people denied it.

Capable employees do not remain for long in roles that carry responsibility without authority. Decades of research into psychological ownership indicate that control over one’s work, the sense that a domain is genuinely one’s own to shape, is what produces the engagement and commitment that organisations depend on. Where decisions are participative and authority is real, ownership follows; where every judgement must be ratified above, it does not.

The reason is not hard to see. A talented person given a title but not the authority that should accompany it learns quickly that their judgment is not trusted, and that the interesting decisions will always be made elsewhere. Some tolerate this for a time. The best rarely do, because they can find, without much difficulty, an organisation prepared to let them decide. A leader who hoards authority not only slows the company but also steadily selects for people willing to work without it, which is not the population any organisation wishes to retain.

Distributing authority without losing it

Distributing authority well is not the same as abandoning it. The fear that the two are the same is exactly what stops leaders from doing it.

The most effective approach observed across studies of delegation begins not with tasks but with decisions, and specifically with the decisions an organisation makes repeatedly: pricing exceptions, hiring approvals, refunds, procurement, and the allocation of marketing spend. For each recurring decision, a single owner can be named, together with an explicit statement of the boundaries within which they may act alone, the threshold beyond which they must consult, and the point at which a matter escalates. What this framework provides is not a loosening of standards but a clarification of them, converting authority from something granted case by case, and therefore constantly renegotiated, into something structural.

The harder discipline lies in what happens next, at the first moment the newly delegated authority produces a decision the leader would not have made. Delegation fails when leaders reclaim authority at the first sign of discomfort. The instinct to intervene is strongest precisely when intervention is most damaging, because a reasonable decision taken independently is often worth more to an organisation than a marginally better one that reinforces its dependence on a single person. The aim is not to eliminate error, which is unattainable, but to build an organisation capable of acting and learning without waiting for one individual to be available.

What the argument is not

None of this amounts to a case for the leader’s disengagement, for the wholesale surrender of judgment in the name of empowerment. Some decisions properly belong at the top and should never be delegated: the direction of the enterprise, the most consequential appointments, and the few choices that guide the entire strategy. Distinguishing these from the far larger set of decisions that have merely accumulated at the top through habit is itself part of the skill.

Nor is the argument that any distribution of authority is better than none. Authority handed to the unprepared, without the boundaries and standards that make its exercise safe, is not delegation but abdication. The claim is narrower and more specific: that handing over decision rights, bounded and deliberate, is a different and more demanding act than handing over tasks, and that it is the one on which an organisation’s ability to outgrow its founder ultimately depends.

Implications

The most revealing exercise available to a leader who suspects they have become the bottleneck is not to audit their tasks but to audit their decisions. Over the course of a fortnight, the decisions that reach them can be sorted into those that genuinely needed them and those that came up only because no one below had been given the authority, or the confidence, to make them without permission.

For most leaders the second category is far larger than the first, and its size is a measure of authority withheld rather than work undone. The organisations that outgrow their founders are not those whose leaders learned to distribute tasks more efficiently. They are those whose leaders learned to distribute the right to decide, and then, at the first uncomfortable test of that decision, declined to take it back.