Electrical
Construction estimating contingency and risk pricing field guide
Price the unknown on purpose: what contingency is, how it differs from profit, allowance, and management reserve, how design completeness and the AACE class size it, and how escalation and drawdown fit.
Direct answer
Contingency is money carried in an estimate to cover costs you know will occur but cannot yet quantify: the gaps in an incomplete design, the unforeseen, and estimating uncertainty. It is not profit, padding, or a slush fund. Size it to the project's risk and design completeness, not a habit percentage.
Key takeaways
- Contingency is money carried for known-unknowns: costs that will occur but cannot yet be priced line by line. It is not profit, padding, or forgotten scope.
- Size contingency to the project's risk and design completeness, not a habit 10 percent flat percentage applied to every bid.
- Keep contingency, profit, allowances, and escalation on separate lines so each can be tracked and drawn down independently.
- AACE estimate classes run Class 5 (conceptual, accuracy about -20 to -50% low and +30 to +100% high) to Class 1 (definitive, about -3 to -10% low and +3 to +15% high).
- Owner contingency and management reserve do not cover the contractor's risk; price your own scope as if no one will bail you out.
What contingency is, and why pricing the unknown is not padding
Contingency is the money you carry in an estimate to cover costs you know will happen but cannot yet quantify. The design is not finished, the site holds surprises, and your own takeoff has error in it. None of that is free, so you price it. The number is real cost waiting to be spent, not a cushion you add for comfort.
The mistake that costs the most is treating contingency as padding or as a second profit line. It is neither. Padding is fear with no basis behind it. Profit is your margin, and it is not there to absorb the misses. Contingency sits between them: a priced estimate of the work the documents cannot yet describe. Get the distinction wrong and you either eat every surprise out of margin, or you pad the bid until a sharper competitor takes the job.
This guide covers the pricing of risk: what contingency is, how it differs from profit, allowance, and management reserve, how design completeness and the AACE estimate class set the size, and how escalation and drawdown fit. Counting and pricing the base work is the electrical estimating takeoff and bidding guide. Recovering overhead and setting profit is the overhead recovery and bid markup guide. This one is about the number you add for what you cannot yet see, and contingency is not profit.
Why pricing risk on purpose keeps a tight bid from becoming a loss
Every estimate carries uncertainty, and the unknowns are real costs whether or not you priced them. That is the whole case for contingency. The job will cost what it costs. The only question is whether you collected money for the part you could not pin down, or whether it comes out of profit when it lands.
Two failures sit on either side. Too little contingency and you eat the surprises: the unforeseen condition, the design gap, the scope the drawings implied but never detailed. Too much and the bid is heavy, and you lose to a contractor who priced the same risk tighter. Neither is bad luck. Both are pricing decisions you made, or failed to make, at the bid table.
So price the risk on purpose. A tight bid that ignores the unknowns is not competitive, it is exposed, and the exposure shows up as a loss on the surprises after you have already won. The contractor who knows what risk they carried, and charged for it, is the one still standing when the job turns out harder than the drawings promised. The discipline is to put a defensible number on the uncertainty, not to hope it stays small.
What is contingency in construction estimating?
Contingency is money in the estimate for the known-unknowns: costs that will occur but are not yet defined well enough to price line by line. You know the design will develop and add scope. You know the site will hold something the drawings missed. You know your own quantities carry error. The specific dollar of each is unknown. That some of them will land is not.
What separates contingency from a plug number is that it answers identified uncertainty, not a single forgotten item. It is not the cost of a particular panel you left off the takeoff. That is an error you fix. Contingency covers the general gap between an incomplete set of documents and the finished job, sized to how incomplete the documents are and how much risk the work carries.
Carry it as a defined number with a basis behind it. A contingency you cannot explain is indistinguishable from padding, and the first owner who asks why it is there will treat it as negotiable. A contingency tied to the design completeness and a risk assessment is a number you can defend and hold when the pressure comes to cut it.
What contingency is not
Contingency gets blamed for four things it is not, and each confusion costs money. It is not profit. Profit is the margin you intend to keep, set in the overhead and markup work, and spending it on misses means you worked for nothing. It is not padding, which is fear with no basis and prices you out of the bid. It is not scope you forgot, which is an estimating error to correct, not a risk to carry. And it is not a management reserve, which sits above the estimate for the truly unforeseeable.
The slush-fund version is the most dangerous, because it feels prudent. A round number added to every bid out of habit, with no risk behind it, is not contingency. It is a guess wearing the word. On a clean job it pads you out of the win. On a rough one it runs dry early because nobody sized it to the actual exposure.
Keep contingency as its own line with its own basis. The moment it blends into profit or hides inside the markup, you lose the ability to tell a pricing problem from a risk that simply landed, and you cannot fix what you cannot see.
What is the difference between contingency and profit?
Contingency covers cost risk and is expected to be spent on the job. Profit is your margin and is not there to cover misses. They answer different questions and belong on different lines, and blending them is the most expensive habit in risk pricing.
Run the two apart. Contingency is your priced estimate of the unknowns: spend it as the risks land, and release what you do not spend. Profit is the earn that is yours to keep after every cost, including the spent contingency, is paid. If you bury risk money in the profit line, a rough job that burns the risk allowance reads as lost profit, and you cannot tell whether you mispriced the work or simply caught a risk that was always priced in.
The discipline pays off when a job runs clean. Unspent contingency, kept on its own line, converts to margin you can see and account for. Unspent padding hidden in profit just means you were lucky and never knew by how much. Separate them and every job teaches you something about your pricing. The overhead recovery and bid markup guide covers setting the profit itself. The point here is that the two numbers do not mix.
What is the difference between contingency and allowance?
An allowance is a placeholder dollar amount for a specific item that is not yet defined. Contingency is general money for uncertainty across the whole estimate. The allowance names the thing. The contingency does not.
Use an allowance when you know an item is in the scope but cannot price it yet, because the selection is open. Light fixtures still to be chosen, a finish not yet specified, a piece of equipment still in design: you carry an allowance, a dollar figure that stands in until the real number arrives, and the contract trues it up against actual cost. The allowance belongs to one identified item with a known function and an unknown price.
Contingency answers a different question. It is not attached to any single item. It is the money for the gaps you cannot enumerate: the conditions, the design development, the estimating error spread across the job. An allowance gets reconciled to the actual cost of its item. Contingency gets drawn down as risks materialize and released if they do not. Treat them as the same line and you will double-count some risks and miss others.
Contingency vs management reserve
Contingency lives inside the estimate, in the baseline you bid and build to. A management reserve sits above the baseline, held separately for the unknown-unknowns: the events nobody could identify in advance. The split is who owns the money and which kind of risk it answers.
Contingency covers the known-unknowns, the risks you can name even if you cannot price each one exactly, and the estimator carries it. A management reserve covers the unforeseeable, the risks outside any risk register, and it is usually held by the owner or senior management rather than priced into the working estimate. On the owner's side, the reserve is the cushion above the project budget. On the contractor's side it rarely exists as a separate pot, so the contractor's contingency has to do that work within the bid.
The practical point for a contractor is not to assume a reserve is backing you. If the owner holds a management reserve, that money answers the owner's surprises, not yours, and it is theirs to release. Your contingency is the only risk money inside your price. Size it as if no one is going to bail you out, because usually no one is.
Design completeness is the biggest driver
The single biggest driver of how much contingency a job needs is how complete the design is. A concept on a napkin needs far more than a finished construction set, because the unpriced gap between what is drawn and what gets built is large early and shrinks as the documents mature. Price the maturity, not a habit.
Walk it forward. At a rough concept, most of the scope is assumption, so the contingency is large because almost everything can still move. At schematic, the shape is set but the details are not, so it comes down. At design development, the systems are defined and the count is closer, so it comes down again. At a complete, coordinated construction set, the remaining uncertainty is mostly conditions and execution, and the contingency is at its lowest. The percentage falls as the design fills in, every stage.
The error is carrying the same contingency regardless of where the design sits. A flat number on a conceptual estimate is almost always too thin, and the same flat number on a finished set is almost always too fat. Read the completeness of the documents in front of you and set the contingency to it. This is also why the AACE estimate class, which is defined by design maturity, maps so directly onto the contingency you should carry.
What are the AACE estimate classes?
AACE International classifies estimates from Class 5, an early conceptual estimate, to Class 1, a near-complete definitive estimate, by how mature the project definition is. The class sets the expected accuracy range, and the accuracy range tells you how much uncertainty, and therefore how much contingency, the estimate has to carry.
The classes track design completeness directly. A Class 5 estimate is built on minimal definition and carries a wide expected accuracy range, commonly cited around minus 20 to minus 50 percent on the low side and plus 30 to as much as plus 100 percent on the high side. As definition grows, the range tightens: Class 3 is often around minus 20 to plus 30 percent, and Class 1 around minus 3 to minus 10 percent low and plus 3 to plus 15 percent high. Those ranges come from AACE recommended practice and the process-industry guidance in 18R-97, and they vary by industry and source, so treat them as the published framework, not a guarantee.
The class is a starting point for the conversation about contingency, not the contingency itself. A Class 5 estimate carries more contingency than a Class 1 because it is less defined, but the actual number still comes from a risk analysis of the specific project. Use the class to set expectations with the owner about how firm the number is, and verify the ranges against the current AACE practice and the estimate basis you are working to.
| AACE class | Design maturity | Typical purpose | Expected accuracy range (approx.) |
|---|---|---|---|
| Class 5 | 0 to 2 percent defined | Concept screening, order of magnitude | About -20 to -50% low, +30 to +100% high |
| Class 4 | 1 to 15 percent | Feasibility, study | About -15 to -30% low, +20 to +50% high |
| Class 3 | 10 to 40 percent | Budget authorization, control | About -10 to -20% low, +10 to +30% high |
| Class 2 | 30 to 75 percent | Detailed control estimate | About -5 to -15% low, +5 to +20% high |
| Class 1 | 65 to 100 percent | Definitive, bid or check estimate | About -3 to -10% low, +3 to +15% high |
How much contingency should you add?
There is no universal contingency percentage, and any number handed to you as the right one is wrong somewhere. The right amount comes from the specific risk of the job and how complete the design is, not from a habit 10 percent applied to everything. Size it to the project in front of you.
Start with the design completeness and the estimate class, which set the floor for how much uncertainty is already baked in. Then read the specific risks: the site, the schedule, the scope, the client, the method. A finished set for a repeat building type you have built before carries little. A conceptual estimate for an unfamiliar scope on a tight schedule in an occupied building carries a great deal more. The same percentage on both is wrong on both.
The trap is the comfortable round number. A flat 10 percent feels disciplined and is the opposite, because it ignores the one thing contingency is supposed to answer, which is the actual risk of this job. Use the design maturity and a real look at the risks to set the size, and be ready to defend the number with the basis behind it rather than the habit behind it.
Judgment percentage or a quantitative risk model?
How you set the contingency scales with the size and stakes of the job. On a small, familiar job, a judgment percentage informed by your history is fine and faster than the analysis would be worth. On a large or complex one, a structured risk method earns its cost by putting a defensible number behind the contingency instead of a feel.
The structured approach starts with a risk register and ranges. You list the identified risks, estimate each one's likelihood and cost impact, and build the contingency from the exposure rather than a flat add. For big jobs, this goes quantitative: a Monte Carlo simulation runs the cost ranges thousands of times to produce a probability distribution, and you set the contingency at a confidence level, often the point where there is about an 80 percent chance the cost lands at or below the number. Expected-value math, likelihood times impact summed across the register, is the simpler cousin and works without the simulation.
Match the method to the stakes. A full risk model on a small tenant fit-out is wasted effort. A judgment percentage on a hundred-million-dollar program is negligence. The principle is the same either way: the contingency should trace to identified risk, whether that trace is a quick informed estimate or a full simulation.
The risk register behind the number
A risk register is the list of identified risks behind the contingency, each scored by likelihood and impact so the ones that actually drive the number stand out. Without it, the contingency is a number with no story, and a number with no story is the first thing an owner negotiates away.
Build it plainly. Name each risk, rate how likely it is, estimate what it costs if it lands, and multiply to rank the exposure. A few risks usually dominate: the unverified existing conditions, the long-lead item that might slip, the design area still in flux, the scope interface with another trade. Those top risks are where most of the contingency belongs, and pricing them specifically beats spreading a flat percentage across everything.
The register also tells you what to watch once the job starts. The risks you priced are the risks you track, and as each one resolves, you know whether to spend the contingency you held against it or release it. A register built at bid time and never opened again is half the value. Kept live, it ties the contingency you carried to the risks as they actually play out.
What drives the risk
Risk on a job comes from a short list of usual sources, and naming them is how you size the contingency to the project instead of to a habit. The drivers compound: a conceptual design on a tight schedule in an occupied building stacks three of them at once, and the contingency should reflect the stack.
Incomplete design is the largest and most common. The further the documents are from finished, the more scope can still appear, and the more the estimate is guessing. Site unknowns come next: existing conditions, soils, what is actually behind the wall on a renovation. A long schedule brings escalation, the cost of material and labor rising before you buy and build. New or unfamiliar scope and method add uncertainty because your own labor data does not cover them. An aggressive schedule raises the odds of overtime, acceleration, and rework. Single-source material or a sole supplier concentrates the risk of a price spike or a delay. And the client matters: a slow decider, a known change generator, or a thin-funded owner is a risk you price like any other.
| Risk driver | What it is | Why it raises contingency |
|---|---|---|
| Incomplete design | Documents short of a finished, coordinated set | More scope can still appear; the estimate is partly assumption |
| Site unknowns | Existing conditions, soils, hidden work on a renovation | What you cannot see, you cannot price exactly |
| Long schedule | A job that runs many months to years | Material and labor escalate before buyout and install |
| New scope or method | Work or means your crews have not done | Your own labor and cost history does not cover it |
| Aggressive schedule | A timeline tighter than the work wants | Overtime, acceleration, out-of-sequence work, and rework |
| Single-source supply | One supplier or a sole-sourced item | A price spike or delay has no fallback |
| The client | Slow deciders, frequent changers, thin funding | Changes, delays, and payment risk you carry |
Escalation: pricing cost increases on a long schedule
Escalation is the rise in material and labor cost over the life of a long job, and it is priced separately from contingency. The two answer different questions. Contingency covers uncertainty about scope and conditions. Escalation covers a cost you can see coming, the predictable drift of prices between the bid and the buy. Fold them together and you cannot tell whether you ran over on risk or on inflation.
Price it on anything that runs long. A job under a few months usually holds its bid pricing. A multi-year job can see material move enough to erase the margin if you priced everything at bid-day cost. The method is to carry an escalation assumption forward to the midpoint of the relevant work, supported by published indices rather than a guess. Cost engineers lean on indices like the Producer Price Index for materials and the ENR Construction Cost Index, or a cost-based escalation tied to actual incurred cost, to set the rate.
On volatile material, an escalation clause in the contract is cleaner than burying the risk in the price. A clause lets you bid lower because you are not carrying a large hedge for a price move that may not happen, and it shifts the actual movement to a documented index or actual cost. Either way, name the escalation, support it, and keep it off the contingency line.
Bid-day uncertainty and the estimating contingency
Part of every contingency answers the estimate itself, not the design. The takeoff has error in it, some quotes came in as plug numbers against a deadline, and assumptions stand in for facts you did not have time to confirm. That estimating uncertainty is real cost risk, and it deserves a piece of the contingency.
Name the soft spots in your own number. The quantities you took off fast and did not reconcile, the sub bid that arrived minutes before the deadline, the gear you carried at a budget price because the real quote never came, the area you priced on an assumption because the documents were unclear. Each is a place the estimate could be light, and the contingency is what covers the aggregate of those soft spots until they firm up. The electrical estimating takeoff and bidding guide covers tightening the count and the quotes themselves. This is the money you carry for the count you could not fully tighten in time.
The honest move is to track which assumptions drove the bid and revisit them after award. The plug numbers get replaced with real quotes, the assumed quantities get reconciled, and the part of the contingency that was covering bid-day uncertainty either gets spent on the truing-up or released. An estimating contingency you never reconcile is padding by another name.
Should you show contingency as a line item?
Whether to show contingency as a visible line or carry it inside the price depends on the contract and the relationship. On a negotiated or cost-plus job with an open-book owner, a transparent contingency line builds trust and gets managed jointly. On a hard-bid lump sum, a visible contingency line is usually the first thing an owner asks you to cut, so it is commonly carried inside the unit prices and the markup instead.
Read the delivery method. Negotiated, design-build, and construction-management work often expects the contingency to be disclosed, tracked, and drawn down in the open, with unspent money sometimes shared or returned per the contract. Hard-bid public and private work rarely rewards disclosure, because the low number wins and a visible risk line reads as soft. There the contingency still exists. It is just distributed into the bid rather than sitting on its own line for a competitor to undercut.
The decision is a contract decision, not a moral one. What does not change is that you know your own contingency and its basis, whether or not the owner sees the line. Carry it knowingly. The danger is the contractor who hides the contingency from the owner and from themselves, and then cannot say what risk the price was supposed to cover.
Owner contingency does not cover your risk
The owner carries their own contingency, and it is not there to cover yours. The owner's contingency answers the owner's risks: scope changes they initiate, design development on their account, the management reserve above the project budget. Your contingency answers your risks: your means and methods, your productivity, the conditions you took on in the contract. Do not assume one backs the other.
This bites contractors who see a healthy owner contingency in the budget and price their own work thin, expecting the owner's money to absorb a rough job. It will not. When your labor runs over because the conditions were harder than you priced, that is your contingency that should have covered it, not the owner's. The owner's contingency is released for the owner's purposes, on the owner's decision, and a contractor's overrun is rarely one of them.
Price your scope as if the owner's contingency does not exist, because for your risk it effectively does not. The two contingencies sit in the same budget and answer to different masters. Knowing which risks are yours under the contract, and pricing only those, is the line between a contingency that protects you and a hope that someone else's money will.
Managing contingency drawdown on the job
Contingency is not money you set and forget. You track it down as the risks it covered resolve, spending it when a risk lands and releasing it when one passes without cost. That tracking is drawdown management, and it is how the contingency does its job instead of quietly disappearing into the general cost of the work.
Run it against the risk register. Each risk you priced either materializes, in which case you draw the contingency you held against it, or it clears, in which case that money is freed. Released contingency is not found profit to spend elsewhere on the job. It is risk that retired, and it should show up as recovered margin you can see, not as budget that gets absorbed into other overruns without anyone noticing. Know at any point how much contingency you have drawn, why, and whether what remains still covers the risk ahead.
This only works if the field data is captured as the work happens, not reconstructed at closeout. Costs and hours logged against the job and tied to the risks they relate to, in a field tool like FieldOS, are what let you see the drawdown in real time and decide whether the remaining contingency is enough. Manage it actively and unused contingency converts to margin. Ignore it and it leaks away, and you never learn whether you priced the risk right.
The trap of too little contingency
A thin or zero contingency on an incomplete design is the fastest way to turn a won bid into a loss. The design will develop, the site will surprise you, and the estimate had error in it. With no contingency, every one of those costs comes straight out of profit, and on a conceptual or schematic estimate there are enough of them to wipe the margin out.
This is the trap of the aggressive bidder who treats contingency as fat to be trimmed. Cutting it makes the number look sharper and the bid more likely to win, which is exactly the problem. You win the job by underpricing the risk, and then the risk lands and you own it. A clean takeoff and a tight markup do not save a bid that carried no money for the unknowns, because the unknowns were never in question. Only the timing was.
Match the contingency to the design completeness and the risk, and hold it. The pressure to cut it is real, especially on a competitive bid, but a zero-contingency bid on an unfinished design is not a competitive bid. It is a bet that nothing you could not see will cost anything, and that bet loses on most real jobs.
The trap of too much contingency
Too much contingency loses the job. Pile risk money on every bid out of fear and your number is heavy, and a contractor who priced the same risk with discipline wins the work at a price that still pays. The padding that feels safe is its own slow way of going out of business, by never winning.
The cure is discipline, not fear. A contingency built from the actual risk of the job is defensible and usually leaner than a blanket add, because most risks do not justify the round number people reach for. When you size to the design completeness and a real risk read, some jobs carry less than your habit would, and those are the ones you win without giving away margin. Padding treats every job as the worst case. Pricing treats each job as the job it is.
The discipline also protects the relationship. An owner who sees a fat, unexplained contingency reads it as either soft pricing or a contractor hedging against their own incompetence, and neither wins the next job. A contingency you can explain, sized to named risks, is a number an owner respects even when it is not the lowest. Fear pads. Expertise prices.
Feedback: calibrate the next contingency
The only way to know whether your contingency was right is to track what it actually got used for, job after job, and feed that back into the next estimate. A contingency you never check against the outcome is a guess that never improves. Closed jobs are where the calibration lives.
Compare by job type. After a job closes, look at how much contingency you carried, how much you actually drew, and what you drew it for. Do that across enough jobs and the pattern appears: renovations run hotter than you priced, conceptual estimates needed more than the habit number, a certain client generates the changes you always underprice. That history is what turns the next contingency from a feel into a calibrated number, the same way job-cost feedback tunes labor units in the takeoff and bidding work.
The catch is that this only works if the drawdown was tracked while the job ran. Released and spent contingency reconstructed from memory at year-end is worthless. Recorded against the risks as it happened, it is the most useful pricing data you own. Bid the contingency, track the drawdown, and let the difference set the next number.
What to document
A contingency you cannot explain is a contingency you will lose at the negotiating table and learn nothing from at closeout. Write down the basis: the risks behind the number, how you sized it, and how it was drawn down. The record is what makes the contingency defensible during the job and useful after it.
Keep the risk register with its likelihood and impact scores, the basis for the contingency size and the design completeness or estimate class it was set against, the escalation assumption and its index, whether the contingency is shown or carried, and the drawdown log as risks resolve. When the job closes, record what the contingency was actually spent on, so the next estimate inherits what this one learned. A field tool like FieldOS that holds the bid, the risks, and the actual costs in one place is what keeps that record defensible instead of a memory.
| What to record | Basis | Note |
|---|---|---|
| Risk register | Likelihood times impact per risk | Names the risks the contingency answers |
| Contingency size and basis | Design completeness or AACE class plus risk read | Why the number is what it is, not a habit percent |
| Escalation assumption | Index used and the period covered | Kept separate from the contingency line |
| Shown or carried | The contract and delivery method | Disclosure is a contract decision, recorded |
| Drawdown log | Spend and release against each risk | Tracks the contingency as risks resolve |
| Closeout actuals | What contingency was spent on | Feeds the next estimate's calibration |
Common mistakes
- Carrying zero or near-zero contingency on an incomplete design and eating every surprise out of profit.
- Blending contingency with profit, so a rough job that burns the risk allowance looks like lost margin.
- Applying a habit flat percentage to every bid regardless of design completeness or the actual risk.
- Ignoring escalation on a long schedule and pricing multi-year work at bid-day costs.
- Setting a contingency with no risk register or basis behind the number.
- Assuming the owner's contingency or a management reserve covers your risk under the contract.
- Treating contingency as a slush fund and absorbing released money into other overruns instead of recovering it.
- Never tracking the drawdown, so the next contingency is the same uncalibrated guess as the last.
Field checklist
Want this checklist to run itself on every job — with photo proof and a signed record crews can hand the customer? That's FieldOS.
Standards and references
There is no code that sets your contingency the way the electrical code sets a clearance. Contingency and risk pricing are governed by sound cost-estimating practice and the contract, with the actual numbers coming from the specific project. Use the references for the method and your own risk read for the size.
AACE International is the main reference. Its recommended practices cover the cost estimate classification system, Class 5 through Class 1, the expected accuracy ranges by class, and the principle that contingency is determined from a risk analysis of the project rather than picked as a flat percentage. AACE practice also covers risk-based contingency methods, including range estimating and Monte Carlo simulation, and contingency drawdown as a project proceeds. For escalation, published indices such as the Producer Price Index and the ENR Construction Cost Index give a supportable basis, and a contract escalation clause can shift the risk to a documented index or actual cost.
Treat every percentage, class range, and method in this guide as framework, not a quoted figure to copy. The AACE accuracy ranges vary by industry and edition, so verify them against the current recommended practice. The contingency itself comes from the design completeness, the risk register, and the contract terms of the job in front of you. Above all, keep contingency for the known-unknowns and out of profit, size it to design completeness and risk, and price escalation and track the drawdown on their own.
Terms and definitions
The vocabulary of risk pricing gets used loosely, and loose definitions are how contingency, allowance, and reserve get blended at the bid table. These are the terms that have to be exact for the pricing to be right.
- Contingency
- Money in the estimate for known-unknowns: costs that will occur but are not yet defined, sized to the project's risk and design completeness. Not profit, padding, or forgotten scope.
- Allowance
- A placeholder dollar amount for a specific, identified item not yet priced, such as fixtures still to be selected, reconciled to actual cost later.
- Management reserve
- Money held above the estimate baseline for unknown-unknowns, the unforeseeable, usually controlled by the owner or senior management rather than the estimator.
- Known-unknown
- A risk you can identify but cannot yet price exactly. Contingency covers these; the unforeseeable unknown-unknowns are covered by a management reserve.
- Escalation
- The rise in material and labor cost over a long schedule, priced separately from contingency and supported by a published index or actual cost.
- Estimate class
- AACE's classification of an estimate by design maturity, Class 5 conceptual to Class 1 definitive, which sets the expected accuracy range.
- Risk register
- The list of identified risks behind the contingency, each scored by likelihood and impact, used to size and then draw down the contingency.
FAQ
What is contingency in construction estimating?
Contingency is money in the estimate for known-unknowns: costs that will occur but cannot yet be priced line by line, from design gaps to site conditions to estimating error. It is sized to the project's risk and design completeness, and it is not profit, padding, or scope you forgot to count.
What is the difference between contingency and profit?
Contingency covers cost risk and is spent on the job as risks land. Profit is your margin and is not there to absorb misses. Keep them on separate lines. Blend them and a rough job that burns the risk money reads as lost profit, so you cannot tell a pricing problem from a risk that landed.
How much contingency should you add?
There is no universal percentage. The right amount comes from how complete the design is and the specific risk of the job, not a habit 10 percent. A finished set for a familiar building carries little; a conceptual estimate on a tight schedule in an occupied building carries much more. Size it to the project and the basis behind it.
What is the difference between contingency and allowance?
An allowance is a placeholder dollar amount for a specific identified item not yet priced, like fixtures still to be selected, reconciled to actual cost later. Contingency is general money for uncertainty across the whole estimate, not tied to any one item. The allowance names the thing; the contingency answers the gaps you cannot enumerate.
What is the difference between contingency and management reserve?
Contingency sits inside the estimate baseline and covers known-unknowns the estimator carries. A management reserve is held above the baseline for unknown-unknowns, the unforeseeable, usually controlled by the owner or senior management. Do not assume an owner's reserve covers a contractor's overrun; it answers the owner's surprises, not yours.
What are the AACE estimate classes?
AACE classifies estimates from Class 5, an early conceptual estimate with a wide accuracy range, to Class 1, a near-complete definitive estimate with a tight one. The class tracks design maturity and sets how much uncertainty, and therefore contingency, the estimate carries. Verify the ranges against current AACE recommended practice.
Is escalation the same as contingency?
No. Escalation is the predictable rise in material and labor cost over a long schedule, priced on its own line and supported by an index like the PPI or the ENR Construction Cost Index. Contingency covers uncertainty about scope and conditions. Fold them together and you cannot tell an inflation overrun from a risk that landed.
Does the owner's contingency cover the contractor's risk?
No. The owner's contingency answers the owner's risks, like scope changes they initiate, and it is released on the owner's decision. The contractor's contingency answers the contractor's risks, like productivity and conditions taken on in the contract. Price your own scope as if the owner's contingency does not exist, because for your risk it does not.
What is contingency drawdown?
Drawdown is tracking the contingency as the risks it covered resolve: spending it when a risk lands and releasing it when one passes without cost. Released contingency is recovered margin, not found money to absorb other overruns. Run the drawdown against the risk register so you always know how much remains and whether it still covers the risk ahead.
Should contingency be shown as a line item?
It depends on the contract. On negotiated or open-book work, a visible contingency line builds trust and gets drawn down jointly. On hard-bid lump sum, it is usually carried inside the prices, because a low number wins and a visible risk line gets cut first. Either way, know your own contingency and its basis.