Contingency drawdown rules are the provisions that decide what the money may be spent on, who decides, what the owner sees and what happens to the balance. Most contracts answer the first question in general terms and leave the other three entirely open.
The four questions
Every contingency on every project has four questions attached to it and they are independent of each other.
What may it be spent on. The permitted purposes. Usually stated as something like costs within the scope of the work that were not foreseen, which is broad enough to cover almost anything.
Who decides. Whether the contractor draws it unilaterally, notifies the owner, or requires consent. This is the question that most changes behavior and the one most often left silent.
What does the owner see. Whether drawdowns are reported individually with a reason, reported as a total, or not reported at all until the end.
What happens to what is left. Whether the balance returns to the owner, falls into the shared savings pool, or simply disappears into the final cost of the work.
A contract that answers all four is unusual. A contract that answers only the first is normal, and on that contract the contingency behaves as a fund the contractor manages and the owner learns about afterwards.
The distinction between the two funds that exist on most projects is set out in contractor contingency and owner contingency, and these four questions apply separately to each.
Why the permitted purpose clause does so little
The standard formulation covers unforeseen costs within the scope of the work. Read carefully, that phrase excludes two things and includes everything else.
It excludes changes in scope, which are change orders. And it excludes costs outside the work, which are the owner’s problem.
What it includes is coordination errors, estimating shortfalls, trade contractor difficulties, rework, acceleration to recover contractor caused delay, buyout overruns, and the ordinary friction of building something.
Several of those are things the contractor would otherwise carry out of its fee or its margin. Where the contingency can absorb them, it does, and the guarantee is never tested.
That is precisely what a contractor contingency is for and no owner should object to the principle. The objection is to the absence of visibility, because a fund that absorbs both genuine unforeseen conditions and ordinary performance shortfalls tells the owner nothing about which it was spent on.
A drawdown log with one line per event and a cause code costs nothing to produce and changes that entirely.
The contingency drawdown rules worth negotiating
Notification above a threshold. Any single drawdown above a stated figure is notified with a description and a cause within a stated period. Below the threshold, monthly totals are enough. This is the lightest possible requirement and it is usually agreed without argument.
Cause coding. Each drawdown is coded: unforeseen condition, coordination, trade contractor default, estimating variance, acceleration, rework. The codes matter more than the amounts, because a contingency spent mostly on coordination is telling you something about the design that will continue for the rest of the job.
A consent threshold on the largest items. Not on everything, which would be unworkable. On single events above perhaps half a percent of contract value.
A statement of what happens at exhaustion. The most important and the most often missing. When the contractor contingency reaches zero in month fourteen, what then. Does the guarantee still hold in full. Does the owner contingency become the next line of defense automatically.
That fourth provision is worth more than the other three combined and is the one nobody asks about, because at signature the contingency is full and the question feels remote.
A worked example
Illustrative figures. Not taken from any client project and not a quotation.
A 71 million dollar project with a contractor contingency of $2.8 million, roughly four percent, and an owner contingency of $3.5 million held outside the contract.
The contract states the contingency covers unforeseen costs within the scope of the work, requires no notification and provides for the balance to fall into the shared savings calculation.
By month thirteen the contractor contingency is exhausted. The owner learns this from a single line in a monthly report.
Reconstructed afterwards, the drawdowns break down as $610,000 on genuinely unforeseen conditions, $1.4 million on coordination between trades, $520,000 on rework and $270,000 on acceleration to recover slippage.
The $1.4 million coordination figure is the finding. It was visible from month six as a pattern, it predicted the rest of the job, and it would have supported a very different conversation with the design team while the interiors packages were still unbought.
Instead the owner contingency absorbs $1.9 million over the following ten months, on the same coordination issues, with nobody having connected the two.
Reading the drawdown log as a forecast
The single number, contingency remaining, is almost useless on its own. The pattern is not.
Three ratios do most of the work. Contingency drawn against work complete, which says whether the burn rate is sustainable. Drawdowns by cause, which says what kind of project this is. And drawdowns by month, which says whether the rate is accelerating.
A contingency 60 percent drawn at 35 percent complete is a forecast, and the arithmetic is straightforward enough to do in a meeting.
A contingency drawn mostly on coordination is a design problem that will continue, because the same documents govern the remaining packages.
A contingency drawn mostly on unforeseen conditions in the early months is usually a ground story that is now behind you, and the remaining balance may be adequate.
Those are three very different projects with the same headline number, and distinguishing them requires the cause codes rather than the total. Which is why the coding requirement is worth more than the reporting requirement.
Sizing follows the rules, not the other way round
Owners usually negotiate the percentage first and the rules afterwards, if at all. The order should be reversed.
A four percent contingency with clear drawdown rules, cause coding and a stated position at exhaustion is worth more than a six percent contingency with none of those, because the owner can see what is happening and can act while acting is still cheap.
The percentage is also a poor instrument on its own, since it bears no relationship to what the project is actually exposed to. Sizing against named events is the argument in how much contingency is enough, and the drawdown rules are what make that sizing meaningful afterwards.
Put differently: the percentage decides how much protection exists, and the drawdown rules decide whether the owner finds out it has been used.
There is a further reason to settle the rules first. A contractor asked to accept notification, cause coding and a consent threshold will sometimes ask for a slightly larger contingency in exchange, and that is usually a good trade. Visibility on a larger fund is worth more than blindness on a smaller one.
It is also a trade that can be made at signature and never afterwards, because once the contract is executed the owner has nothing left to offer for it.
What we do
We read the four questions against your contract and say which are answered, which are silent, and what each silence is likely to cost. The provision at exhaustion gets particular attention because it is the one that decides whether the guarantee means anything in month fourteen. That work is part of the readiness review. When a lender or an investment committee has to approve the position, the lender and investment committee memo sets it out in two pages.
Questions people ask
Should an owner approve every drawdown?
No, and a contract requiring it would not function. The contingency exists to let the contractor solve problems without a negotiation each time. A notification threshold with cause coding gives the owner visibility without inserting them into daily decisions, which is the balance most projects need.
What if the contractor refuses cause coding?
It is worth understanding why, because the information already exists in their cost system and producing it is a report rather than an analysis. Refusal usually reflects a concern about how the pattern will read rather than about the effort, and that concern is itself the answer to the owner question.
Does an owner contingency need the same rules?
It needs different ones, held internally. The useful discipline is deciding in advance what the owner contingency is for and refusing to spend it on things that belong to somebody else. Owner contingency spent on contractor coordination problems is the commonest way a project loses its last line of defense.
This is general information about construction contracts and is not legal advice.