Why Decisions Escalate When Uncertainty Has Nowhere To Go
- Andrew Pigott
- Jul 3
- 6 min read
The decision rights work is done. The governance paper has been signed off. Team leads now hold clear authority to approve a defined category of cases without going anywhere near the senior table. The message went out weeks ago: decisions need to happen closer to the work.
And yet the escalations keep coming.
Service leads bring back decisions they were technically allowed to make. Case managers ask for a second opinion on matters that sit well within their remit. Files that should have been closed at the first desk keep finding their way to the top of the pile.
The obvious diagnosis is familiar. People lack confidence. Managers will not take ownership. Senior leaders say they want delegation but keep pulling decisions back. Decision rights, despite the paperwork, are still not clear enough.
Some of that may be true. But look more closely at what is actually being escalated, and a different pattern often appears. People are not always escalating the decision itself. They are escalating the exposure that sits around it.
They know who is supposed to decide. What they do not know is how exposed they will be if their judgement turns out to be wrong.
The tempting diagnosis
Delegation and decision-rights work usually gets the visible part right. It names who can approve what, and it draws a cleaner line through the org chart. That is genuine progress, and it is worth defending when someone suggests scrapping it and starting again.
But a decision right answers one question: who decides? It rarely answers the question that actually determines whether someone will use that authority: what am I allowed to be uncertain about?
A team lead may know, correctly, that a decision sits within their delegated threshold. What they often do not know is what level of incomplete evidence is acceptable, what assumptions they are allowed to make, what happens if new information appears after the decision is made, or whether reasonable judgement will be protected if the outcome turns out imperfect.
When that is missing, delegation exists on paper without existing in practice. The signature has moved. The uncertainty has not.
Decision rights are not uncertainty rights
This is the distinction worth sitting with. Decision rights tell people who holds the pen. They do not tell people what risk they are allowed to carry while they hold it.
A decision may have a clearly named owner. The uncertainty around that decision often does not. Nobody has said what assumptions are reasonable, what should trigger a second look, what counts as good enough evidence to proceed, or how the decision will be reviewed if it turns out to have been the wrong call given what was known at the time.
Without that, escalation becomes the rational choice. Not because people are weak or the delegation was badly designed, but because the organisation has, without quite meaning to, made being wrong more visible than being slow. A system that treats a delayed decision as invisible and a flawed one as a story that reaches the board will train people to wait, however clearly the delegation paper is written.
Some escalations are genuine requests for authority. Others are requests for protection. Treating both as the same problem, and trying to fix both with a decision-rights map, rarely closes the gap.
What public-sector risk guidance already tells us
This distinction is not a new discovery. UK public-sector risk guidance has been pointing at it for some time, even if the language used is different from the one above.
HM Treasury's guidance on risk management treats risk as part of ordinary governance and decision-making, not as a separate compliance exercise sitting alongside the real work. It expects organisations to be clear not only about who decides, but about roles, escalation routes and how significant issues get raised quickly and appropriately.
Its guidance on risk appetite goes further. It asks organisations to define what level of risk is genuinely tolerable, not only in principle but in practice, and to record the justification, evidence and conditions when a decision sits outside that appetite. The implication for everyday work is clear. The task is not to remove uncertainty before anyone is allowed to act. It is to make that uncertainty bounded enough for a reasonable person to act on it, and reviewable enough that judgement can be checked fairly afterwards.
That is a governance idea. It is also, in practice, an organisational design problem, and one that sits well below board level.
A pattern from a housing association
A national housing association is redesigning how it handles repairs and tenancy casework. The intent is sound. Routine cases should be resolved by neighbourhood teams. Complex, high-risk or safeguarding-related cases should reach specialists faster, because that is where real expertise is most needed.
The policy is approved. Neighbourhood managers now hold delegated authority to close a defined category of case without reference upward.
Six months on, the queue at the centre has not shortened. Borderline repairs, cases with incomplete history, anything with a hint of vulnerability or reputational sensitivity, still travel to the top. Neighbourhood managers are not confused about their authority. They know they are allowed to decide. What they do not know is what happens to them if a case they closed locally later turns into a complaint, an ombudsman referral or a story that reaches the board.
Nobody has defined what counts as sufficient evidence to close a routine case. Nobody has said which assumptions are safe to make when a tenant's file is incomplete. Nobody has described what should trigger a second look, or how a reasonable decision, made with the information available at the time, will be treated if it later looks wrong with hindsight.
The association has delegated the signature. It has not delegated the uncertainty. Senior leaders are frustrated that the bottleneck persists. Neighbourhood teams say the new authority feels rhetorical. Both are describing the same missing piece from different ends of it.
Four practical moves
Separate the decision from the uncertainty around it
For any category of decision being delegated, ask what is actually being decided, what remains genuinely unknown, what assumptions are being made to proceed anyway, and what would need to be true for this to be the wrong call. Most decision maps show who owns the decision. Fewer show who owns the uncertainty sitting around it. That second map is usually the more useful one.
Define what is good enough to decide
Be specific about the evidence threshold, not just the sign-off threshold. A reversible, low-stakes decision should not need the same evidential bar as an irreversible one. Naming this distinction explicitly, rather than leaving it to individual judgement under pressure, does more to unblock decisions than another round of confidence-building.
Make risk appetite operational
A risk appetite statement that lives in a board paper and never reaches an actual case is not doing its job. Translate it into working examples: what is acceptable to decide locally without a second opinion, what should be tested first, what should be monitored after the fact, and what genuinely needs to come back. This is what turns a principle into something a manager can actually use on a Tuesday afternoon.
Treat recurring escalation as design information
When the same category of case keeps escalating, resist the urge to simply remind people they have the authority to decide. Ask what uncertainty they are actually trying to move upward, and why the current design has not made it safe enough to hold locally. Repeated escalation from the same place is rarely a discipline problem. It is usually a sign of an undefined threshold, an unresolved policy question, or a decision owner without adequate protection.
What escalation is actually telling you
Escalation is not, in itself, a problem to be designed out. It is genuinely useful when it moves the right issue to the right level, at the right time, with the right people around it. The difficulty is escalation that moves uncertainty upward simply because nowhere else has been made safe enough to hold it.
People can only make good decisions at the right level when they know three things: what uncertainty they are allowed to carry, what risk is genuinely acceptable, and when a decision truly needs to travel further. Decision rights answer the first question people usually ask, which is who decides. They rarely answer the harder one, which is what that person is allowed to not yet know.
At Common Path Connection, we help leaders see where decisions are really getting stuck. Often it is not only in unclear ownership, but in the uncertainty, exposure and unwritten rules that shape what people feel able to hold at their own level.
Our Role Clarity & Decision Rights work helps organisations clarify not just who decides, but what risk and uncertainty they are equipped, and protected, to carry.
Related reading
Escalation is often a signal that uncertainty, authority or protection has not been designed clearly enough. These insights explore the wider decision system:
Decision Rights at Work: Why Your Meetings Keep Repeating Themselves — why authority needs to be explicit before meetings can stop repeating.
Why the Same Decisions Keep Coming Back — why decisions return when the consequence has not been properly held.
Role Clarity After Restructure: Why the New Boxes Do Not Settle the Work — why structures only settle work when decision paths and hand-offs change in practice.
Comments