Decision Rights: Who Can Decide What Without Asking

Why so many questions land with the owner

A customer wants a three percent discount. A technician needs a spare part that is not in stock. A colleague asks whether she may move a delivery date by a week. Each of these questions is small. Together, in many companies, they add up to a day in which the owner mainly does one thing: grant approvals that others could have made just as well.

The reason is rarely distrust. Usually nobody has defined who may decide what. People who are unsure would rather ask, and whoever has asked once will ask again next time. This grows into a habit that slows the company down and ties up the owner.

This article answers one question: how does an owner define who may decide what without asking, so that the company becomes faster without the owner losing control over the important decisions? To answer it, we first clarify the terms, then present a thinking model from organizational economics, examine the findings, weigh two perspectives, and finally build a decision map.

Owners who want to make every decision themselves end up making mostly the small ones.

What is a decision path?

A decision path defines, for a certain kind of decision, who makes it, who has to be consulted beforehand, and who is informed afterwards. A decision right is the authority to make such a decision bindingly. A value limit is an amount up to which this right applies. Above it, the decision goes to the next level.

A decision path works like the right-of-way rule at an intersection. It does not determine where someone drives, but who may go first. Without it, every intersection would require new negotiation, and traffic would stand still. With it, everyone moves along, and accidents happen where the rule is missing or ignored.

The decision path complements two tools already described on Owners Council. The RACI matrix clarifies for individual workflows who performs, who is accountable, who is consulted, and who is informed. The approval rules in purchasing set value limits for orders. The decision path extends this idea to all recurring decisions: prices, dates, goodwill, personnel, and spending.

The thinking model: formal and real authority

In 1997, the economists Philippe Aghion and Jean Tirole introduced a distinction that describes the problem precisely. Formal authority is the right to decide. Real authority is effective control over the decision. The two often diverge. Whoever decides formally but does not know the facts will usually follow the proposal of whoever does. Real authority then lies with that person (Aghion and Tirole 1997).

The authors also show that more real authority for the employee promotes initiative but means a loss of control for the superior. They name circumstances under which employees' real authority grows, even if formally everything stays with the boss.

A sober insight follows for the owner. The question is not whether to give away decisions. The owner already does so as soon as overload or time pressure reduces approval to a formality. The question is whether this happens deliberately and with clear limits, or by chance and without feedback.

What studies show about distributed decisions

Bloom, Sadun, and Van Reenen (2012) surveyed almost 4,000 companies in the United States, Europe, and Asia on how much decision authority headquarters leaves to the managers of individual sites, measured by investment, hiring, production, and sales. Companies from regions with high trust between people delegated significantly more decisions. The authors see this as a channel through which trust raises productivity: where owners can delegate more, more efficient companies grow more easily. The study concerns larger companies with several sites. The mechanism, that trust is the precondition for delegated decisions, holds regardless of size.

Filippi, Peters, and Suitner (2024) examined what distributed decision power means for employees. Via an online platform, they surveyed a total of 1,960 employees in the United Kingdom, the United States, and Italy in four samples of about 500 people each, deliberately spread across micro, small and medium-sized, and large companies. They measured how strongly respondents perceive decision power in their organization as distributed, how much they identify with it, and how satisfied, engaged, and exhausted they are. In all four samples, employees identified more strongly with their organization when they perceived decisions as distributed, and through this identification, higher satisfaction, more engagement, and less exhaustion were also associated. The authors themselves stress that a survey at a single point in time does not allow statements about cause and effect.

A meta-analysis of 55 samples, described in detail in the article on delegating tasks, paints a similar picture: giving leeway mainly improves attitudes and motivation, while performance rises measurably but to a lesser degree (Kim, Beehr, and Prewett 2018).

Control or speed? Two perspectives

The first perspective emphasizes risk. The owner is liable, bears the loss, and knows the connections best. An overly generous discount or a wrong hire costs more than any follow-up question. From this view, it is reasonable to concentrate important decisions in one person.

The second perspective emphasizes the cost of waiting. Every follow-up question delays the process, and the waiting almost always hits the customer. Moreover, nobody learns to decide who is never allowed to decide. Following Aghion and Tirole, control fades anyway as soon as the owner is overloaded.

I consider both perspectives justified, but for different kinds of decisions. The dividing line is drawn with two questions. How large is the impact, meaning amount and risk? And how easily can the decision be reversed? Jeff Bezos described this idea in his letter to Amazon shareholders for the year 2015 with the image of doors: some you can walk through in both directions, others only in one. A postponed delivery date is a two-way door. A hire or a three-year framework contract is not.

The decision map: four levels on one page

The decision map is the tool of this article. It assigns the most frequent decisions in the company to one of four levels. The smaller the impact and the easier the reversal, the lower in the table a decision sits.

The value limits for materials follow the example from the article on purchasing and approvals so that both rules fit together. More important than the exact amounts is that each level has a deadline. A proposal level without a deadline becomes a waiting loop. And every limit needs a deputy, otherwise the owner's vacation becomes a standstill, as the article on arranging deputies shows.

How to introduce decision rights in three weeks

1. Collect questions for one week

   Note every question that lands with you, with its type and amount. After a week you will see which decisions recur.

2. Sort by impact and reversibility

   Assign each type of decision to one of the four levels. When in doubt, start one level higher and lower it after a few weeks.

3. Discuss the map with the team

   Explain the limits and the standard. Ask where the team needs more leeway to serve customers faster.

4. Agree on feedback

   Define how level 2 informs, for example with one line in the shared channel. The report replaces the question, not the control.

5. Review after four weeks

   Look at a sample of the decisions. Where the line is right, the limit can rise. Where it is not, the reason usually lies in a missing standard.

What matters is what the owner does after the introduction. Criticizing level 1 and 2 decisions after the fact takes them back without saying so. A mistake within the limit is a reason to sharpen the standard, not to lower the limit. Anyone who wants to record larger decisions in order to learn from them later will find a tool in the article on the decision journal.

Conclusion: who may decide what without asking?

The opening question was
an owner defines who may decide what without asking, without losing control over what matters. The answer begins with an insight from Aghion and Tirole: an overloaded owner gives away real authority anyway, only in a disorderly way. It is better to distribute it deliberately. Impact and reversibility draw the line. Small, reversible decisions belong to those who work close to the customer. Large and final ones stay with the owner, after consultation. The studies suggest that distributed decisions go along with more identification and engagement, and they show that trust is the precondition. A one-page decision map turns this trust into a rule everyone knows.

How to get started

• For one week, note every follow-up question with its type and amount.
• Assign the five most frequent types of decisions to one of the four levels.
• Set a deadline and a deputy for each level.
• Write down a one-sentence standard for each limit.
• Discuss the map with the team and schedule a review in four weeks.

The first week of note-taking usually shows that a large share of the questions need no decision from the owner, only a rule that nobody has written down so far.