Part 9 of 15 Construction Contingency Management: Why One Budget Line Invites Disaster (And the Three-Layer Fix)

Stop treating contingency like a slush fund—here's the discipline that turns budget reserves into actual risk management

If your contingency is just one mysterious line item that everyone raids without rules, you're not managing risk—you're hiding problems until they explode at the worst possible moment.

---

Contingency isn't a rainy-day piggy bank or a political slush fund to make uncomfortable numbers look better. It's your planned fuel for managing inevitable project risks.

I've watched too many facilities directors discover—way too late—that their contingency got drained on finish upgrades and "small improvements" that had nothing to do with actual risk. Then a real problem hits (like discovering a buried fuel tank no one knew about), and suddenly there's no money left to handle it.

Treat contingency with discipline and you'll keep scope intact, absorb genuine shocks, and finish projects on time. Treat it like a mystery fund that "we'll figure out later," and you'll hide problems until they explode in front of the board—usually about three weeks before you promised to open the building.

This post gives you a clear model to properly size, assign ownership, and control contingency spending—so your audit reads like a victory lap instead of a cautionary tale.

The win (why this matters right now)

• No surprise overruns: Project risks get paid from the appropriate contingency layer with clear justification.

• Faster decisions: Pre-established rules mean approvals happen in minutes, not endless meetings where everyone argues.

• Clean audits: Every dollar drawn has a documented trigger, a named approver, and a clear audit trail.

When contingency management is working, it's nearly invisible. When it's broken, it's all anyone talks about—usually while pointing fingers.

Define three distinct layers (never use one lump line)

This is the foundation that makes everything else work:

1. Construction Contingency (CC)

Purpose: Cover unknowns within the defined scope of work—differing site conditions, coordination conflicts discovered during construction, buyout gaps between estimates and actual market bids.

Owner: Project Manager in coordination with GC/CM, governed by contract-defined release rules.

2. Project Contingency (PC)

Purpose: Handle owner-driven changes within the project's core promise—scope clarifications after award, code or regulatory agency requirement changes, minor program adjustments that don't fundamentally alter the deliverable.

Owner: Project Sponsor (typically the facilities director or assigned project champion).

3. Program Contingency (ProgC)

Purpose: Absorb portfolio-level risks—market escalation beyond baseline assumptions, district-wide funding or policy shifts, systemic estimating variance across multiple projects.

Owner: Program Director and/or Chief Business Officer.

One contingency line creates temptation and confusion. Three lines create control and accountability.

I can't count how many times I've seen "we need to tap contingency" turn into a 45-minute argument because nobody knew which bucket to use or who had approval authority. Split it clearly and those arguments disappear.

How much contingency? Use planning bands, then refine with data

Planning starting points:

• New construction projects: Start around ~10% of total project budget (not just construction cost—total project including soft costs)

• Heavy modernization or occupied site work: 15–20% (significantly more unknowns and coordination complexity)

• Portfolio reserve (ProgC): 2–5% of total remaining program work; scale up when market conditions are volatile

Then sharpen these numbers using a proper risk register—identify your top 10 risks, assign probability and impact to each, and calculate your actual exposure. If the risk register math says you're under-funded, move the contingency number now before bid, not after you discover the problem.

I've watched districts set contingency at "feels about right" levels, then act shocked when predictable risks actually happen. Do the math. Fund the real risks.

Release rules (write these once, use them forever)

Put these rules in writing and actually follow them:

Construction Contingency (CC) CAN fund:

• Genuinely unforeseen field conditions (buried utilities, contaminated soil, unexpected structural conditions)

• Buyout shortfalls when market bids exceed reasonable estimates

• Design coordination conflicts discovered during construction

Construction Contingency (CC) CANNOT fund:

• New scope additions, betterments, or finish upgrades

• Late owner-requested changes

• Items that should have been in the base scope from day one

Project Contingency (PC) CAN fund:

• Owner-requested changes that maintain the core project promise

• Code or regulatory agency requirement changes issued after permit

• Missing program elements that should have been caught during Design Development (yes, that's your miss and your bill)

Project Contingency (PC) CANNOT fund:

• Fundamental scope expansions beyond the original project definition

• Items properly belonging in Program Contingency

Program Contingency (ProgC) CAN fund:

• Market escalation significantly exceeding baseline assumptions

• District-level policy changes affecting multiple projects

• Portfolio-wide funding or revenue adjustments

Every contingency draw requires documentation of:

• Clear trigger (what specific event or condition caused this need)

• Amount and assigned category (CC/PC/ProgC)

• Permanent budget trade if scope is growing (what are you cutting to pay for additions)

• Named approver (specific person, not "management")

• Effect on remaining contingency (update the glidepath projection)

These aren't suggestions. These are the rules that prevent your contingency from disappearing mysteriously.

The Contingency Glidepath (your single source of truth)

Create and maintain three tracking lines from project kickoff through closeout:

• Starting balance (CC, PC, ProgC)—set at budget approval

• Approved draws (running log with date, amount, reason, and approver)

• Remaining balance versus project timeline (should decline steadily as construction risks burn down)

If your glidepath line is flat or rising late in the construction phase, you're either hiding risk that should have been addressed, or you're dramatically under-estimating closeout and punch list costs.

A healthy contingency glidepath starts high, declines steadily through construction, and ends near zero at project acceptance. Any other pattern is a warning sign.

Governance that speeds decisions instead of blocking them

Weekly routine: Project Manager publishes contingency draw candidates showing proposed layer, documented trigger, and budget impact.

Bi-weekly board reporting (one slide): Show beginning balance → approved draws with one-line descriptions → ending balance for each contingency layer.

Approval thresholds by dollar amount:

• CC draws ≤ $X: PM approval authority; above $X requires PM + Sponsor

• PC draws ≤ $Y: Sponsor approval authority; above $Y requires Sponsor + CBO sign-off

• ProgC draws: CBO and Program Director approval required (no delegation)

Set your dollar thresholds based on project size and organizational norms. Publish them once. Follow them always.

When everyone knows the rules and thresholds, approval takes hours instead of weeks. When the rules are fuzzy, every single draw becomes a negotiation.

Communication: Use numbers, not vague adjectives

Stop saying vague things like "minor change" or "we need a little more contingency."

Instead, say this:

• "CC draw: $62,400 – Unforeseen utility conflict discovered at grid line C/5; resolution requires reroute but creates no scope change and maintains schedule."

• "PC draw: $120,000 – Add shade structures per community request documented in March board meeting; funded by executing deduct alternate #3 (decorative paving)."

• "ProgC release: $250,000 – Steel price escalation Round 2 exceeds baseline by 18%; affects three active projects; funded from program portfolio reserve."

Specificity kills confusion. Vagueness breeds suspicion and endless questions.

Common failure modes (and what to do instead)

Failure: You maintain one pooled contingency without clear categories or ownership.

Fix: Split immediately into CC/PC/ProgC with distinct owners and documented approval thresholds. Do this today.

Failure: You use Construction Contingency to quietly fund owner add-ons and scope improvements.

Fix: Route those requests through Project Contingency where they belong, and log the program decision transparently. Stop hiding scope growth.

Failure: Your team treats contingency as sacred reserve funds that should never be touched.

Fix: Contingency exists for planned, appropriate use. The goal is smart spending with declining risk exposure over time—not zero spending that leaves risks unmanaged.

Failure: Every request for "just a small upgrade" gets approved without identifying a trade-off.

Fix: Enforce the iron rule: Every addition requires a deduction (scope, quality, or schedule). Document the trade or deny the request. No exceptions.

Failure: You discover major estimate gaps after bids come in, then drain contingency to cover them.

Fix: Run thorough Value Engineering at Design Development, not after bidding. Reserve

Construction Contingency for genuine unknowns, not predictable estimation failures.

I've personally seen every one of these mistakes cost programs hundreds of thousands of dollars and months of delays. Learn from their expensive lessons.

Real example (how your log should read)

Date	Layer	Amount	Trigger & Decision	Remaining
02/10/2025	CC	$48,900	Hidden water main at Building 2; PM approved (≤$50K)	$1,351,100
03/01/2025	PC	$120,000	Add shade structures; Sponsor + CBO approved; trade: delete skylights	$680,000
04/12/2025	ProgC	$250,000	Steel escalation round 2; Program Director approved	$1,050,000

One simple table. Complete transparency. No room for debate about what happened or why.

Quick implementation (one focused week)

Day 1: Set your percentage bands for CC/PC/ProgC (new construction versus modernization) plus portfolio reserve level.

Day 2: Publish the complete Release Rules document—what each contingency layer can and cannot fund.

Day 3: Stand up your Glidepath tracking dashboard showing starting balance, approved draws, and remaining balance by month.

Day 4: Adopt formal approval thresholds by dollar amount and role, plus the one-slide board reporting format.

Day 5: Train all Project Managers on "add equals deduct" discipline and proper draw log documentation hygiene.

Five days of focused work creates a system that runs itself for years.

What "good" looks like (the reality check)

• Contingency draw logs are concise, specific, and properly mapped to the correct layer.

• Remaining contingency balances decline steadily over project timeline—no mysterious sudden drops or "where did it go?" questions.

• Leadership teams approve legitimate draws in hours, not weeks of meetings, because rules are crystal clear.

• Auditors can trace every contingency dollar without requesting clarification meetings or additional documentation.

When contingency management works, it's boring. When it's broken, it's dramatic and expensive.

The takeaway

Contingency management isn't magic or art. It's disciplined management with clear rules.

Split your contingency into three layers with distinct purposes and owners. Size it properly using actual risk analysis. Govern it with simple, published rules that everyone understands and follows.

When every contingency dollar has a clear purpose and approval path, you maintain project momentum, tell clean budget stories to stakeholders, and finish with confidence—not excuses about "unexpected costs" that were entirely predictable.

Good contingency management is invisible until you need it. Bad contingency management is visible to everyone, all the time.