Loading...
Loading...
Governing means setting boundaries. Managing means operating within them. When the distinction collapses — and it collapses in most institutions — board intent degrades from 100% to 30% by the time it reaches operations. The problem is not bad people. It is the absence of constraint enforcement.
A board's job is to govern: set the boundaries within which the institution operates. Management's job is to manage: operate freely and effectively within those boundaries.
In practice, most institutions blur this distinction. Boards either micro-manage (doing management's job) or rubber-stamp (abandoning their own). In both cases, management ends up interpreting board intent without constraint, and board resolutions degrade through what we call the resolution decay chain.
A board resolution passes through four layers: Board, Executive, Management, Operations. At each layer, fidelity drops. By operations, only 30% of the original intent survives. Not because anyone acted in bad faith, but because no mechanism preserves intent across organisational boundaries.
Constitutional governance solves this by making resolutions into enforceable constraints. Management retains full autonomy — within boundaries that actually hold.
The distinction is conceptually simple. Governance sets the rules of the game. Management plays the game. But in institutional practice, the boundary between these two activities is almost never clear — and institutions pay an enormous hidden cost when it blurs.
Setting boundaries, constraints, and conditions under which the institution operates. Defining what must happen, what must not happen, and what requires explicit authorisation.
Operating within the boundaries governance has set. Making day-to-day decisions, allocating resources, executing programs — with full autonomy inside the governed space.
Most institutions operate in one of two failure modes: the board micro-manages (usurping management authority, creating bottlenecks, and neglecting actual governance work) or the board rubber-stamps (abdicating governance responsibility, leaving management to govern itself). Both failures have the same root cause: the absence of an enforceable boundary between the two domains. When there is no constraint architecture, the boundary is maintained only by convention — and convention erodes under pressure.
Boards do not simply "govern well" or "govern badly." There is a spectrum of governance modes, each with distinct failure characteristics. Most institutions oscillate between the first three. The fourth requires architectural investment that few have made.
Board approves everything management proposes. Governance exists on paper only.
Board asks questions in meetings and receives reassuring answers. Creates the appearance of governance.
Board involves itself in operational decisions, undermining management authority and effectiveness.
Board sets enforceable constraints. Management has full autonomy within those boundaries.
When governance and management are not architecturally separated, board resolutions pass through a predictable decay chain. Each organisational layer introduces reinterpretation that erodes the original intent. This is the resolution decay chain.
The chart below measures intent fidelity — the degree to which the operational action matches what the board actually decided — at each stage.
A board resolution starts at 100% fidelity. By operational reality, approximately 30% of the original intent survives.
Board: 100% | Executive: 75% | Management: 50% | Operations: 30%
The decay is not random. At each transition between organisational layers, specific structural mechanisms erode intent. Understanding these mechanisms is prerequisite to preventing them.
Example
Board resolves to 'prioritise impact measurement.' CEO interprets this as 'add a section to the quarterly report' rather than 'restructure program evaluation.'
Example
CEO directs 'impact measurement across all programs.' Management implements it for 3 of 14 programs — the ones already performing well.
Example
'Measure impact' becomes a checkbox exercise. Staff tick boxes on a form designed to confirm existing practice, not evaluate it.
The common thread: At every transition, intent is transmitted through human interpretation rather than architectural enforcement. Each interpreter has legitimate authority, competing priorities, and no mechanism to check whether their interpretation preserves the original intent. The result is operational drift — not through malice but through structure.
Perhaps the most insidious governance failure is not non-compliance but false compliance: technically following the letter of a board resolution while completely violating its spirit. This is performative governance at its most dangerous — because the board believes its intent has been fulfilled.
The "compliant but wrong" pattern arises because board resolutions are expressed in natural language, which is inherently ambiguous. When the resolution says "diversify," "strengthen," or "evidence-based," it creates a space of interpretation wide enough to drive any operational outcome through it.
"Diversify revenue sources"
Relabel existing grants under new category names
Revenue structure unchanged; reporting shows 'diversification'
"Strengthen risk management"
Produce a risk register that is reviewed once and filed
Risk register exists as compliance artefact; no operational change
"Improve board diversity"
Add one additional director from a different professional background
Board composition unchanged in substance; one seat rotated
"Evidence-based program delivery"
Commission an evaluation that confirms existing programs work
Evaluator incentivised to deliver positive findings; no programs changed
"Compliant but wrong" is worse than outright non-compliance because it is invisible. Non-compliance triggers review. False compliance satisfies the governance system while undermining it. The board receives reports confirming that its resolutions have been implemented. The management dashboard shows green. The only thing that hasn't changed is reality.
The difference is not incremental improvement. It is architectural. Traditional governance records intent and hopes it survives. Constitutional governance encodes intent and ensures it persists.
| Dimension | Traditional Pipeline | Constitutional Pipeline |
|---|---|---|
| How board decisions are recorded | Minutes, memos, prose resolutions | Machine-readable constraints with enforcement rules |
| How decisions are transmitted | CEO interprets and delegates verbally or via email | Constraints propagate automatically to relevant decision points |
| How compliance is verified | Management self-reports to board quarterly | Constraints are checked at point of action, violations flagged in real time |
| How exceptions are handled | Management makes exceptions unilaterally, reports post-hoc | Exceptions require explicit authorisation, creating audit trail |
| What survives leadership change | Nothing — new CEO reinterprets from minutes | Everything — constraints persist independently of people |
| Management autonomy | Unlimited within vague boundaries (feels free, is ungoverned) | Full within defined boundaries (genuinely free, demonstrably governed) |
Minutes, memos, prose resolutions
Machine-readable constraints with enforcement rules
CEO interprets and delegates verbally or via email
Constraints propagate automatically to relevant decision points
Management self-reports to board quarterly
Constraints are checked at point of action, violations flagged in real time
Management makes exceptions unilaterally, reports post-hoc
Exceptions require explicit authorisation, creating audit trail
Nothing — new CEO reinterprets from minutes
Everything — constraints persist independently of people
Unlimited within vague boundaries (feels free, is ungoverned)
Full within defined boundaries (genuinely free, demonstrably governed)
The constitutional approach does not require better people, more meetings, or stricter oversight. It changes the architecture of how board decisions interact with operational reality. Resolutions become machine-readable constraints that enforce themselves at the point of action.
The board resolves to prioritise evidence-based programs. This resolution is encoded as a constraint: 'No program renewal without independently verified impact data meeting threshold T within the preceding 12 months.'
The constraint is attached to every program renewal decision point, every budget allocation involving those programs, and every management report on program status. No human transmission required.
When a program manager submits a renewal for a program without qualifying impact data, the system flags the violation. The renewal is not blocked — but it requires explicit board authorisation to proceed, creating accountability.
If the board authorises an exception (e.g., 'continue Program X for 6 months pending evaluation'), that exception is itself a constraint with an expiry. When it expires, the original constraint reasserts.
Instead of reviewing management narratives about progress, the board reviews constraint fidelity data: how many decisions were made within constraints, how many required exceptions, which constraints are under pressure.
In traditional governance, the board asks: "Did management do what we decided?" Management says yes. The board has no way to verify.
In constitutional governance, the board reviews: "How many operational decisions were made within our constraints? How many exceptions were authorised? Where are constraints under pressure?" The data is generated by the system, not narrated by management. Fidelity is measured, not reported.
The word "constitutional" is deliberate. A constitution does not govern through instruction — it governs through constraint. It does not tell the executive what to do; it defines what the executive cannot do without authorisation. This is the same principle applied to institutional governance: the board defines constraints, and management operates freely within them.
"The board resolves to prioritise evidence-based programs and wind down programs without demonstrated impact within 12 months."
Ambiguous. What counts as "evidence-based"? What is "demonstrated impact"? Who decides? By when? Every term is open to interpretation.
CONSTRAINT: No program renewal without independently verified impact data (threshold: p < 0.05 or equivalent qualitative standard, verified by external evaluator)
SCOPE: All programs with annual budget > $50K
DEADLINE: 12 months from ratification (15 March 2027)
EXCEPTION: Requires board vote (simple majority) with documented rationale
Unambiguous. Every term is operationally defined. Compliance is verifiable. Exceptions are governed.
A common objection is that constraints reduce management freedom. The opposite is true. When constraints are explicit, management knows exactly where the boundaries are and can move fast within them. When constraints are implicit (hidden in board expectations, past discussions, and unwritten norms), management must constantly seek permission, second-guess, and self-censor. Paradoxically, well-defined constraints create more operational freedom, not less.
It is well understood conceptually. Every governance textbook teaches it. The problem is that most institutions have no architecturalmechanism to enforce it. Knowing that the board should set strategy and management should execute does not prevent the board from micro-managing or the CEO from interpreting board intent without constraint. The distinction needs to be encoded in governance infrastructure, not just governance training.
No. Boards should be involved in operational decisions when those decisions hit a constraint boundary — when management needs an exception, when a decision has strategic implications the board has reserved for itself, or when constraint fidelity data shows drift. What boards should not do is routinely make operational decisions that management is better positioned to make. Constitutional governance makes this distinction self-enforcing: the system routes decisions to the board when governance authority is required and keeps them at management level when it is not.
The principle — expressing board decisions as enforceable constraints rather than aspirational prose — can be applied manually. A small organisation can write resolutions as constraints with defined terms, verification criteria, and exception procedures. Technology makes enforcement automatic and continuous, but the conceptual shift from "record intent" to "encode constraint" is valuable at any scale.
Constraints should have review dates. When a constraint is due for review, the board can reaffirm, modify, or retire it — with a documented rationale. This is governance working as intended: the board maintains the constraint architecture over time, adapting it to changed circumstances rather than letting it drift through unacknowledged reinterpretation at the management level.
What happens when those with authority to govern lack the capacity to do so — and those with capacity lack the authority.
Read ExplainerWhen governance rituals produce compliance reports but not governance outcomes — the organisational theatre that resolution decay enables.
Read ExplainerThe same structural mechanism that drives resolution decay also explains why institutional commitments and pledges reliably degrade over time.
Read ExplainerHow institutional authority survives leadership turnover — and why constitutional constraints outlast the people who set them.
Read Explainer