Claims, disputes and alternative dispute resolution

When a risk occurs and the parties disagree about who should pay for it, the result is a claim. Claims are a normal part of construction; on large projects there may be dozens. They become a problem when they are poorly documented, left unresolved until the end of the job, or escalated into disputes that cost more to fight than the amount at stake. CON-220 covers the law of changes, delay and termination and the dispute resolution spectrum from negotiation to litigation, and CON-310 lesson 10 covers how changes and claims are tracked in the cost report. This lesson takes a risk management view: how to treat claims and disputes as risks to be identified, valued, reserved for and resolved at the lowest total cost, and how disputes with insurers and sureties fit into the picture.

Why claims arise

Most claims trace back to the risks discussed in this course: unexpected site conditions, design errors and late information, owner changes, delays and disruption, escalation, and disagreements about scope. The contract allocates the risk, but the parties often disagree about what the contract says, about the facts, or about the amount. Studies of construction disputes in North America regularly list the same causes: incomplete or unclear contract documents, failure to understand or comply with contract obligations, poorly administered changes, and unrealistic schedules.

The subject of notice deserves repeating. Many claims fail not because the contractor was wrong on the merits but because it did not give the written notice the contract required within the time allowed (CON-220 lesson 9). A notice calendar, listing every notice period in the contract and assigning responsibility, is one of the cheapest risk controls available.

Valuing a claim as a risk

For both parties, an unresolved claim is an uncertain future cost or recovery, which makes it a risk that belongs in the register. The owner should estimate what it may have to pay and hold contingency for it; the contractor should estimate what it is likely to recover and avoid booking unlikely recoveries as profit (CON-320 lesson 4 covers the accounting rules for unapproved change orders and claims). The same expected value logic as lesson 3 applies, adjusted for the costs of pursuing or defending the claim.

Owners use the same analysis from the other side. If the city believes the claim has a 55 % chance of success, its expected cost of defending is 0.55 × 540,000 plus its own legal costs of perhaps $140,000, or about $437,000. Settling at $250,000 is cheaper for the city too. When both parties do this arithmetic honestly, a zone of possible agreement usually appears, which is why most construction claims settle.

# Settlement zone for the dewatering claim (illustrative)
p_win, award = 0.55, 540_000
contractor_cost, owner_cost = 160_000, 140_000
contractor_walkaway = p_win * award - contractor_cost      # minimum it should accept
owner_walkaway = p_win * award + owner_cost                # maximum it should pay
print(f"Contractor's minimum acceptable settlement: {contractor_walkaway:,.0f}")
print(f"Owner's maximum sensible settlement:        {owner_walkaway:,.0f}")
print(f"Zone of possible agreement:                 {owner_walkaway - contractor_walkaway:,.0f} wide")
for p in (0.35, 0.55, 0.75):
    print(f"If success probability were {p:.0%}: zone {p * award - contractor_cost:,.0f} to {p * award + owner_cost:,.0f}")

Contractor's minimum acceptable settlement: 137,000
Owner's maximum sensible settlement:        437,000
Zone of possible agreement:                 300,000 wide
If success probability were 35%: zone 29,000 to 329,000
If success probability were 55%: zone 137,000 to 437,000
If success probability were 75%: zone 245,000 to 545,000

The width of the zone equals the two parties' combined costs of fighting, $300,000 here. The more expensive the dispute process, the wider the zone and the stronger the incentive to settle. Disagreement about the probability of success is what most often prevents settlement: if each side is overconfident, the zones may not overlap. Independent evaluation, discussed below, is a way of narrowing that disagreement.

Delay claims and time impact analysis

Delay claims are among the largest and most contested claims, because their value depends on proving what caused the delay and how much of the time-related cost flows from it. CON-220 lesson 9 covers the legal framework, including concurrent delay. The analytical method most commonly used during a project is time impact analysis: a fragment network representing the delay event is inserted into the current updated schedule, the schedule is recalculated, and the change in the completion date is the measured delay. Retrospective methods, such as comparing as-planned and as-built schedules or analyzing windows of time, are used after the fact. The Association for the Advancement of Cost Engineering's recommended practice on forensic schedule analysis is widely cited for these methods.

From a risk point of view, the key controls are a realistic, updated schedule and contemporaneous records. A delay claim built on a schedule that was never updated, or on records reconstructed months later, is weak whatever the merits. Contractors should treat schedule updating and daily records as risk controls, not paperwork.

Dispute resolution as a risk choice

The contract's dispute resolution clause decides how disagreements will be settled, and choosing it is itself a risk decision. Faster, cheaper processes reduce the cost of disputes and the time cash is tied up; more formal processes give more thorough decisions but at higher cost and delay.

In the United States there is generally no statutory construction adjudication of the kind used in the United Kingdom and several Canadian provinces, although some contracts create contractual adjudication or initial decision makers. AIA A201-2017 uses an Initial Decision Maker followed by mediation and then arbitration or litigation, as the parties choose. Many construction contracts specify arbitration under the American Arbitration Association's Construction Industry Arbitration Rules, which were updated effective 1 March 2024: the fast-track ceiling rose to $150,000 and the threshold for a three-arbitrator panel in large, complex cases rose to $3 million. Dispute review boards, promoted by the Dispute Resolution Board Foundation, are common on large public infrastructure projects.

Dispute boards illustrate the preventive value of a good process. A board of experienced neutrals is appointed at the start of a large project, visits the site regularly, and hears disagreements as they arise, issuing recommendations within weeks. Because the board knows the project and the parties know it will hear their arguments, many disputes are settled before they are ever referred. Board costs, often a small fraction of a percent of project value, are typically shared equally.

Disputes with insurers and sureties

Some disputes are with the parties who were supposed to absorb the risk. Insurers may deny claims or accept them under a reservation of rights, meaning they will defend or investigate while keeping the right to deny coverage later. Coverage disputes commonly turn on exclusions such as defects (LEG wording, lesson 7), late notice, breach of conditions such as hot work rules, or whether a loss falls within the policy period. Sureties may deny performance bond claims on the grounds that there was no default, that the obligee did not follow the bond's conditions, or that the obligee itself breached the contract (lesson 10).

The best protection against coverage disputes is preparation before any loss: reading policies and bond forms when they are bought, keeping certificates and endorsements organized, training site staff on notice and loss prevention conditions, and involving the broker early when something happens.

Reserving for claims

Owners and contractors both need to reflect open claims in their forecasts. The owner's forecast at completion should include an allowance for the expected cost of open contractor claims, usually from owner contingency (lesson 4), and the contractor's forecast should include only recoveries it is likely to achieve, with legal costs budgeted. Reporting claims to senior management with their expected value, range and next steps lets the organization see the whole exposure rather than just the optimistic view of the project team.

# Claims reserve for the owner: expected cost of open claims (illustrative)
claims = [   # (description, amount claimed, probability owner pays, expected share of amount if it pays)
    ("Dewatering and delay",        620_000, 0.55, 0.87),
    ("Mechanical coordination",     310_000, 0.40, 0.70),
    ("Rowing club access changes",   95_000, 0.80, 0.90),
    ("Pool tile substitution",      140_000, 0.20, 0.60),
]
total_claimed = sum(c[1] for c in claims)
reserve = sum(amt * p * share for _, amt, p, share in claims)
for name, amt, p, share in claims:
    print(f"{name:<28} claimed {amt:>8,}  expected {amt * p * share:>9,.0f}")
print(f"Total claimed {total_claimed:,}; expected cost {reserve:,.0f} ({reserve / total_claimed:.0%} of claimed)")

Dewatering and delay         claimed  620,000  expected   296,670
Mechanical coordination      claimed  310,000  expected    86,800
Rowing club access changes   claimed   95,000  expected    68,400
Pool tile substitution       claimed  140,000  expected    16,800
Total claimed 1,165,000; expected cost 468,670 (40% of claimed)