Programme debt

A decision was carried forward from the last steering committee. And the one before that. The risk register hasn't been updated since the gate review. The team is building scope that was never approved through change control. No change request, no revised baseline, no sponsor approval. The programme is moving yet nobody has named what's accumulating.

What's accumulating is programme debt. Programme debt is the accumulated liability from decisions, approvals and validations that remain unresolved while delivery continues. Some deferral is deliberate and managed. A deferred decision with an owner, a review date and a stated consequence is a governance choice. The dangerous kind is the debt nobody has acknowledged, where the programme keeps moving but the authority, baseline or rationale has not caught up. The principal is the unresolved governance obligation and the interest is the shrinking set of choices. The programme later repays it through negotiation, rework, scope reduction, delayed approval or a sponsor being forced to ratify a decision after the team has already built on it.

Programme debt moves governance obligations from the people with delegated authority, when evidence is fresh and options are still open, to a later meeting where evidence is stale, options are fewer and the accountable people have changed or lost context. Left long enough, the obligation becomes a fait accompli the sponsor inherits rather than a decision they can make. Programme debt is not a new control. Change control, issue logs, risk registers, decision logs and benefits reviews each catch part of this. The missing view is aggregation. No single artefact shows the sponsor the accumulated liability created when unresolved outputs do not get turned back into a decision.

The same debt can appear at programme, project or portfolio level. The artefacts differ, but the mechanism is the same. Unresolved obligations continue while delivery builds on top of them.

How programme debt accumulates

Programme debt accumulates in recognisable forms and each one leaves a later decision, dispute or rework cost.

When steering decisions are carried forward without being made, they become scope disputes that can't be resolved cleanly, because nobody documented the intent at the time. Resolving that later costs negotiation time, trust and often rework.

When scope changes are implemented pending formal approval and never baselined, they produce unbudgeted work. If the team absorbs them quietly, they widen the gap between what was approved and what is being delivered.

When risk treatments approved at the gate review are never refreshed, the programme has no approved contingency when the risk materialises. The register looks functional, but the treatment is stale and the risk is live.

When engagement with users, service owners and operations is deferred, those groups usually enter the conversation at user acceptance testing, when changes are most expensive and hardest to make. By then, there is often no funded design option left. The people who would have shaped the design earlier are now being asked to validate something they had no hand in building.

When benefits assumptions are not revalidated after the delivery model changes, the programme is delivering against a business case written for a different approach. Nobody has re-justified the investment.

In MSP terms, much of this should be caught at tranche boundaries. A tranche-end review is supposed to revalidate the business case, refresh benefits, revisit risk and confirm the next commitment. Programme debt builds when that boundary becomes a reporting checkpoint rather than a re-authorisation gate.

When governance decisions are made below the authority level required and kept moving by the PM to avoid a steering escalation, they sit unresolved. The decision is still there, one level down, with no more authority and more consequence. In PRINCE2 language, this is often a manage-by-exception failure. The tolerance has effectively been breached, but the exception is being absorbed inside delivery rather than escalated to the people authorised to decide.

Programmes operate on deadlines, so the compounding is fast. You can defer the governance decision, but you can't defer the consequence. Not every deferred obligation creates the same consequence. The debt worth escalating first is the kind that, left for another quarter, forecloses options for the steering committee rather than just makes them harder.

The test

The test is not whether governance looks untidy. It is whether delay is consuming choices the sponsor may later need.

The smell of programme debt is repetition without decision. The same scope item appears in three steering packs. The same risk treatment survives three reporting cycles unchanged. The same benefits assumption carries forward after the delivery model has changed. The registers may look tidy, but the programme is building on unresolved obligations. That's how you detect the debt. The debt is the liability underneath.

A scope change implemented pending approval and never baselined narrows your options before the programme can recover. The vendor says it was always in scope and you have nothing that says otherwise. A governance decision made below the authority level required has the same effect, because the contractual position can be challenged at any point and everything built on it is contingent on the outcome.

A programme carrying two or three items across these categories has a liability view whether it's named one or not. No new methodology is needed. The view only needs to say who owns each item, what happens if it is not resolved and the next decision point. Unlike an issue or action register, the liability view groups unresolved obligations by the choices that will be lost if they remain unresolved. Take an unapproved vendor scope item owned by the sponsor. If unresolved by the next steering committee, the delivery baseline and contract position diverge before the next release. That turns a governance complaint into a sponsor choice, because it attaches cost to delay and puts a decision back where it belongs.

What it looks like in the audit record

In Land 400 Phase 3, Defence's original planning pointed towards a mature, proven solution, but the procurement evolved into a more developmental pathway. The Australian National Audit Office (ANAO) found that the risks of the developmental approach were not clearly communicated to government and that Defence did not adjust its overall risk posture when all tendered platforms involved developmental risk. Later, Defence managed progress through remediation, which masked underlying schedule pressures. Naming this as programme debt would not have fixed Land 400, but it describes the mechanism clearly. Obligations left unresolved during procurement narrowed the options available during delivery.

In Queensland Health payroll, the scope shifted from a whole-of-government implementation to a Queensland Health interim system in 2008. The 2009 governance revision did not assign a clear Accountable Officer for overall delivery. Testing for casuals and overtime was not completed. The decision to proceed to go-live in March 2010 followed a final user acceptance testing phase that had identified 14 “showstoppers” among 1,007 defects, with outstanding defects then managed through a post-go-live defect plan. The Queensland Auditor-General found it was not clear which Accountable Officer had responsibility for the overall governance and successful completion of the project. The Chesterman inquiry later found unwarranted urgency and lack of diligence on the part of state officials were principal causes. Chesterman's point was not that Queensland needed a new rulebook. Existing policies and standards were adequate if followed. Calling this programme debt would not have solved the payroll failure. It would have made the accumulation easier to name earlier, before those obligations came due at go-live across a workforce of about 78,000.

What the vocabulary makes possible

"We have some governance gaps" doesn't describe an obligation. It doesn't name the cost of delay. It doesn't tell a sponsor what choices are being spent to keep the programme moving.

"We're carrying three items of programme debt. An unapproved scope change, a risk treatment awaiting review and a benefits assumption from the old delivery model. If we wait another quarter, these are the options we lose." The facts are the same, but the vocabulary gives the sponsor a decision to make.

The debt doesn't disappear when you defer the decision. It just gets more expensive.