Doing it yourself is a tax you pay in the future. Delegating is a loan you take out against the present. The whole art of managing is knowing which one you can afford right now.

Every manager I have ever met has a story about the task they should have handed off and didn't. They kept it because they were faster at it, because it felt safer in their own hands, because explaining it would have taken longer than just doing it. And for one afternoon, all of that was true. The trouble is that the same logic, repeated for months, quietly turns a manager into the most overqualified individual contributor on the team, while everyone around them stays exactly as junior as the day they were hired.

The dilemma is not really about trust. It is about economics, and most of us run the numbers badly. We compare the cost of delegating today against the cost of doing it today, notice that doing it is cheaper, and stop there. We forget that "today" is a sample size of one and that almost nothing we do at work happens only once. So let me lay out a way to think about this that holds up over time, built on four variables you already have a feel for: your skill advantage, the time you have, the stakes, and who grows from the work.

The four variables

Start by naming the things that actually move the decision. When people say "it depends," this is what it depends on.

• Skill advantage. How much better are you at this than the person you would hand it to? Not in the abstract; for this specific task, today.
• Time available. How much slack do you have? A task you could happily teach on a calm Tuesday becomes undelegatable the night before a launch.
• Stakes. What happens if it goes wrong? A typo in an internal note and a number in a board deck are not the same risk class.
• Growth value. Who becomes more capable by doing this? Sometimes the most valuable output of a task is not the deliverable at all; it is the person who learned to produce it.

The mistake is to treat the first variable as the whole equation. "I'm better at it" feels like a complete argument, and in a vacuum it is. But skill advantage decays the moment you account for the other three. Your edge might be real and still not worth claiming, because the time cost is trivial, the stakes are low, and someone on your team would grow more from the doing than you would lose from the handing off.

The skill-versus-time grid

The fastest first cut is to plot just two of those variables against each other: how big is your skill advantage, and how much time do you have. It will not settle every case, but it sorts the obvious ones in about ten seconds and tells you where to think harder.

The only quadrant that is unambiguously yours is the top left. Big skill advantage, no time to teach. The plane is boarding; you fly it. Everything else has an escape hatch. When you have a real edge and some breathing room, the right move is usually to teach once and delegate forever, paying a known cost now to delete a recurring one. When your edge is small, the question stops being whether to delegate and becomes whether the task should exist at all; a task you are not especially good at and have no time for is often a task nobody should be doing.

The grid is a starting point, not a verdict, because it deliberately ignores stakes and growth. A board presentation might sit in "delegate" on skill and time, yet you keep a firm hand on it because the stakes are high. That is fine. The grid tells you the default; the other two variables tell you when to override it.

Why repetition changes everything

Here is the piece almost everyone misses. The delegate-or-do calculation is not a single decision; it is the same decision made over and over, and the math flips as the count climbs.

Doing it yourself looks cheap because the cost is flat and immediate: an hour each time, forever. Delegating looks expensive because it front-loads. You pay a teaching cost up front, the first few attempts are slower than yours and need correcting, and only after that does the line drop below your own. The two cost curves cross somewhere, and where they cross decides the right answer. If a task recurs only once or twice, doing it yourself genuinely wins. If it recurs weekly for a year, doing it yourself is one of the most expensive habits you have, and it never shows up on any budget.

Notice what the chart quietly assumes: that you will still be doing this task in three months. Most of the time you will be, and you discount that future to zero because the present is loud and the future is quiet. This is not a character flaw; it is the same present bias Kahneman and Tversky spent careers documenting. We are wired to overweight the cost in front of us and underweight the identical cost arriving repeatedly down the line. The fix is not willpower. It is to ask one boring question before you grab a recurring task: how many more times will this happen this year? If the answer is more than a handful, the chart has already made your decision.

There is a related trap worth naming. Parkinson's Law says work expands to fill the time available, and a manager who hoards tasks gives that law a generous home. The work you keep "because it is quick" swells to occupy the slack you were going to use for the things only you can do. Delegation is, among other things, a way of refusing that expansion.

Don't think delegate-or-not; think how much

The word "delegate" hides a false binary. It sounds like a switch: either you own the task or you throw it over the wall and hope. In practice delegation is a dial with many settings, and choosing the right setting is most of the skill. You can hand off the doing while keeping the deciding. You can hand off both but ask to see the result. You can hand off everything and only learn it happened when it is already done. Each rung up the ladder buys you back more time and costs you a little more control, and the right rung depends entirely on the stakes.

Start a person low on the ladder for high-stakes work and let them earn the climb. A new hire handling something that touches a customer might begin at "recommend, I approve," and after a dozen clean recommendations move to "decide, then tell me." The error managers make is treating the ladder as fixed: they pick one rung for a person and never revisit it, so a capable teammate stays on a leash for years and a green one gets thrown off the top. The rung is a function of stakes and demonstrated trust, and both of those change. Revisit them.

This is also where the growth variable earns its keep. If you only ever delegate at the bottom rung, doing exactly as you specify, you get your time back but the person learns nothing except how to follow instructions. The judgment, the part that actually compounds into seniority, lives in the higher rungs. Sometimes the right call is to deliberately delegate a notch higher than the stakes strictly require, accept a slightly bumpier result, and bank the growth. That is not inefficiency. That is you building the team that makes next year's grid lopsided in your favor.

A short checklist for the moment of choice

When a task lands and you feel the reflex to keep it, run it past four quick questions before you do.

• Will this happen again? If it recurs, lean hard toward delegating, even when doing it yourself is cheaper today.
• Is my edge real and large? Be honest. "I'd do it slightly better" is not a skill advantage worth a manager's hour.
• What truly breaks if it goes wrong? Low stakes mean you can delegate high on the ladder and stop hovering.
• Who grows from this? If the answer is someone other than you, that is often the deciding vote.

None of these needs a spreadsheet. They need about thirty seconds of honesty, which is roughly thirty seconds more than most of us give the decision in the moment.

The work only you can do

Here is the quiet promise hidden inside all of this. Every task you successfully push down the ladder is an hour returned to the small set of things that genuinely require you, and that set is where your actual value lives. The strategy nobody else can set. The relationship only you hold. The judgment call that needs your particular scars to make well.

Doing it yourself will always feel responsible, even noble, in the moment. But a manager measured over a year is not judged by how much they personally produced; they are judged by what their team became capable of while they were paying attention to the right things. Hand off the loan you can afford. Pay the teaching tax early. And spend the time you buy back on the work that, when you finally look up, only you were ever going to do.