Chapter 3 — Quiz

B — the contract type. A GMP is a pricing mechanism. It can sit inside CM at risk, design-build, or even a negotiated design-bid-build arrangement, and it can be arrived at competitively or by negotiation. You still do not know who holds the design contract, when the builder was engaged, or how they were selected. Ask all three questions every time.

C. This is the principle usually called the Spearin doctrine: an owner that furnishes plans and directs a contractor to build to them impliedly warrants their adequacy. The architect may ultimately owe the owner something if the owner can prove a breach of the professional standard of care — but that is a separate claim between owner and architect, and it is not automatic. Note also that this default can be modified by contract, with enforceability varying by jurisdiction, and that it does not excuse a contractor from reporting errors it should reasonably have caught.

C. Design-bid-build is sequential by definition: the documents must be complete before they can be bid, and the contract must be awarded before construction starts. This is not a matter of management skill; it is the shape of the method. The other three all permit or actively require overlap.

C — $1,320,000, roughly 3% of the direct cost of work plus general conditions plus insurance and bonds. ($575,200 is the escalation allowance, $900,000 is insurance and bonds, and $1,804,800 is the CM fee at 4.0% of the $45,120,000 subtotal.)

B. Each method has its own signature failure: A is design-build's, C is CM-as-agent/multiple-prime's, and D is design-bid-build's. The CMAR failure is the premature GMP — a number set under schedule pressure on a 45% or 50% set, defended afterward by a thin assumptions list that nobody read carefully.

B — a professional standard of care: the care and skill ordinarily exercised by design professionals on similar projects. This matters commercially as well as legally, because professional liability insurance responds to negligence, not to a contractual guarantee of outcome. An owner who negotiates a hard fitness-for-purpose warranty may create an obligation the design-builder cannot insure.

B. When the duct and the joist conflict, there is nobody to send an RFI to — the problem gets solved inside one organization and inside a price already agreed. A is false (design-build is not always cheaper). C is false and dangerously so; the owner still owns every decision it makes, and DB actually demands more owner discipline at the front end. D is false: the owner still owns its criteria, its furnished data, and its own changes.

C — the owner. The owner signed every trade contract, so the space between them belongs to the owner. This is the structural cost of the method, and it is the reason the "savings" from eliminating a general contractor's fee often evaporate. Under a GC or CMAR arrangement, the same gap is the contractor's problem because it signed for the whole building.

C — 75/25. On a $400,000 unused balance that is $300,000 back to Meridian and $100,000 to Kestrel. The point of the split is not the money to the owner; it is that the contractor's project manager has a personal, dollar-denominated reason to manage the contingency rather than consume it.

B. A tight spread means everybody priced the same building. A wide spread means they did not — and the bidder furthest from the pack is the one most likely to have missed something. The owner's move on seeing a spread like this is to conduct a scope-review meeting with the low bidder before award, not to celebrate the savings.

False. IPD's core features — a multi-party agreement, a shared profit pool, and a mutual waiver of claims among participants — do not fit most competitive sealed-bidding frameworks. This is one of the main practical reasons IPD remains rare, alongside the requirement for a sophisticated repeat owner and the insurance complications.

False. The owner still holds the design contract and still furnishes the documents, so the implied warranty still generally runs from the owner. What CM at risk changes is volume: preconstruction constructability review catches many errors before they are built, and the CM's GMP assumptions and qualifications list allocate some document-completion risk to the CM. Fewer errors reach the field, but the underlying allocation has not flipped.

False. It contains contingency; it is simply buried inside the number rather than disclosed on a line. The differences that matter are that nobody but the estimator knows how much it is, that it is not returned to the owner if unused, and that competitive pressure tends to shave it below what a prudent estimator would carry — with the shortfall reappearing later as change orders and claims.

True. That is the defining feature. The owner selects on qualifications, works open-book through design, and negotiates a price at a defined milestone — typically with an "off-ramp" allowing the owner to pay for design to date, take the documents, and bid the work if the negotiated price is unacceptable.

False. JOC is built for high-volume, small, repetitive, well-understood work priced from a unit price book times a bid coefficient — roof replacements, classroom renovations, accessibility upgrades, deferred maintenance. Design complexity and novel scope defeat the pricing mechanism entirely.

Most of the $1,650,000 gap is risk changing hands rather than work changing scope: the hard bidder carries less contingency, no escalation allowance, and thinner overhead and profit, while carrying more direct cost for pricing without constructability input and more general conditions for a longer construction period. The only permanently real saving is the roughly $856,550 of fee compression; the rest is risk that either comes back through change orders and claims or was never removed at all. Modeled to completion, the DBB path produces an owner outlay of about $48,640,250 against roughly $47,755,000 under CM at risk — and that ignores the seven-month schedule difference entirely.

1. Set the GMP at a defensible design percentage (usually 60–90% construction documents) — below about 50% you are guessing rather than pricing.
2. Write the assumptions and qualifications list as if it will be litigated — because it will be read that way, and quantified assumptions ("priced 18,600 LF of partitions") are enforceable while adjectives are not.
3. Use named, priced allowances for genuinely undefined scope — an honest placeholder with a stated reconciliation mechanism, rather than a guess buried in the base number.
4. Formally reconcile the GMP to the completed documents — an agreed, documented comparison at 100%, rather than a two-year argument about what was included.

Under a general contractor or CM at risk, one entity holds contractual authority over every trade on site and can therefore set rules, control access, run orientation, and stop work. Under multiple prime, that authority is fragmented across eight or thirty separate contracts, so no single party has the practical means to enforce a site-wide program. Meanwhile duties can still attach under OSHA's multi-employer worksite approach based on an employer's actual role — creating, exposing, correcting, or controlling — which means an owner or agency CM that in practice directs sequence and controls access can find itself in the controlling-employer position regardless of what its contracts say. The fix is not a clause; it is one site safety plan, one orientation, one named competent person, and participation made a condition of every trade contract.

Leverage comes from the ability to change something before money is committed to it; once concrete is placed, steel is fabricated, or a subcontract is signed, every option costs more and several options disappear. Under design-bid-build, maximum leverage is the pre-bid question period and the first thirty days of the job, when a document conflict can still be raised without a change order attached to it. Under design-build, it is the criteria review before signature, because after that every requirement you missed becomes a negotiation with a party holding price, schedule, and design. Under CM at risk, it is preconstruction — every dollar you influence while the drawings are still moving is a dollar you do not have to fight for later. This is the contractual expression of the idea that the project is built twice, and the first build determines the second.

Recommend CM at risk, not design-build, despite the schedule pressure — and the reason is the sheriff's office.

The schedule constraint (24 months from schematic) rules out design-bid-build, which would consume roughly a third of that window before construction started. But design-build's price advantage depends on stable, measurable criteria, and the dispatch console and server room requirements — the two most technically demanding and most expensive elements in the building — are explicitly undecided. Signing a design-build contract now means negotiating every one of those decisions with a party holding all the leverage. CM at risk gets the schedule overlap (early site and foundation packages, GMP conversion later) while keeping the design team working for the county.

Risk 1 accepted: the GMP will be set before the sheriff's requirements are final. Mitigate with named, priced allowances for the dispatch console rough-in and the server room MEP, valued off real vendor budget quotes rather than square-foot factors, with the reconciliation mechanism written into the GMP amendment.

Risk 2 accepted: a two-person facilities office cannot staff preconstruction oversight. Mitigate by budgeting for an owner's representative and an independent cost estimate — the county cannot run open-book CMAR without somebody reading the open book, and that person cannot also be the person running every other county facility.

The factor that would flip it: if the sheriff's office would commit, in writing, to a frozen console and server-room specification before the RFP goes out, design-build becomes the better answer — it is faster, it needs less county staff, and the criteria risk would be resolved. The recommendation is not really about the building; it is about whether the user can be pinned down.

What is being proposed is to convert disclosed, administered, returnable money into buried, unadministered, non-returnable money. The GMP total is identical, so on the surface nothing changes — which is exactly what makes it seductive.

What it costs the owner: the $400,000 leaves the contingency line, so it is no longer reported monthly, no longer drawn down against named risks, and — critically — no longer part of the balance subject to the 75/25 savings split at closeout. Under the split the owner would have expected $300,000 of any unused portion back. Buried in the direct cost of work, it simply disappears into the buyout, and the owner loses visibility into whether it was needed.

What it costs the CM: less, which is the problem. The CM gives up $100,000 of potential savings-split upside but gains a cushion inside the cost of work that nobody is watching. That asymmetry is precisely why this proposal is a governance failure even when both parties are acting in good faith.

What you say: something close to — "I can do that, but I want to be clear about what it does. That $400,000 stops being reported to you every month and stops being subject to the savings split, so if we don't need it, you don't get 75% of it back — it just goes away inside the buyout. If the concern is how the contingency reads to the board, I'd rather fix the presentation than the accounting: let me bring the risk register that sits behind the $1,320,000 so the board can see it isn't padding, it's twelve named risks with dollar values attached."

The general principle: when someone asks you to move money between lines without changing the total, ask what reporting, control, or entitlement is attached to each line. That is always where the real transaction is.