Progressive and collaborative models

The traditional methods share one feature that troubled projects have exposed: the price is fixed early, often before the design is developed and before the risks are understood. On a straightforward project that is efficient. On a complex project, a hospital in a dense city, a subway under existing buildings, a treatment plant expansion while the old plant keeps running, it forces bidders to price risks they cannot see, encourages either inflated bids or under-priced bids followed by claims, and makes it hard to change course when new information emerges. Over the last decade owners in both countries have adopted a family of progressive and collaborative models that fix the price later, after the parties have developed the design and investigated the risks together.

This lesson draws together the progressive variants of the methods already covered and adds a few others: early contractor involvement, progressive P3s, target cost contracts and two-stage approaches. What they share is a development phase in which the owner pays for design and pricing work open book, followed by a commercial close at which a price or target is agreed, with an off-ramp if it is not.

The common structure

The trade-off is clear. The owner gives up competition on the construction price, which is the main protection design-bid-build provides, in exchange for a price based on much better information, fewer claims, and influence over how risks are managed. To protect itself, the owner competes the elements that can be competed at selection, such as fees, markups and labor rates, and benchmarks the eventual price.

Early contractor involvement

Early contractor involvement, or ECI, is the general idea of bringing the contractor in during design under a services agreement so that it can advise on constructability, sequencing, temporary works, staging and cost. CMAR and CM/GC are forms of ECI with a defined route to a construction price. In highway, rail and utility work, owners also use ECI agreements in which the contractor helps develop the design with the owner's engineer and then has the first opportunity to price the construction. If the price is acceptable, the contractor builds; if not, the owner can tender the work using the improved design.

Progressive P3s

Progressive P3s are less common in the United States, but the idea has appeared in some transit and social infrastructure procurements, where a private partner is selected on qualifications and fees, develops the design and price with the public owner, and only then signs the full project agreement and financing. More often, US owners combine progressive design-build for construction with separate operations and maintenance contracts, or use predevelopment agreements, in which a private developer studies and designs a project with exclusive rights to negotiate the concession.

A progressive P3 keeps the long-term maintenance and finance features of a P3 while moving the commercial close later, after the development phase. That reduces the risk premium bidders charge for uncertain scope and the risk of a consortium bidding low and failing later. It also reduces competition and makes value for money harder to demonstrate, so owners rely on benchmarking and independent reviews at commercial close.

Target cost contracts

In a target cost contract the parties agree a target for the cost of the work. The contractor is paid its actual costs plus a fee, and the difference between actual cost and target is shared according to agreed ratios, often in bands. Target cost contracts are common in the United Kingdom under the NEC Engineering and Construction Contract options C and D and in alliance agreements, and they appear in North American progressive and alliance models.

# Banded pain/gain share on a target cost contract (illustrative)
target = 200_000_000
fee_rate = 0.06
bands = [  # (upper limit of over/under-run as share of target, contractor share in that band)
    (0.05, 0.50),   # first 5 % either side: shared 50/50
    (0.10, 0.30),   # next 5 %: contractor 30 %
    (9.99, 1.00),   # beyond 10 % overrun: contractor bears all (savings beyond 10 % to owner only)
]
def contractor_share(diff):          # diff = actual - target (positive = overrun)
    share, remaining, lo = 0.0, abs(diff), 0.0
    for hi, s in bands:
        width = (hi - lo) * target
        part = min(remaining, width)
        if diff < 0 and s == 1.00:
            s = 0.0                    # savings beyond the last band go entirely to the owner
        share += part * s
        remaining -= part
        lo = hi
        if remaining <= 0:
            break
    return share if diff > 0 else -share
for actual in (180e6, 192e6, 200e6, 208e6, 216e6, 230e6):
    fee = fee_rate * actual
    pain_gain = contractor_share(actual - target) or 0.0
    print(f"actual {actual/1e6:5.0f}M: contractor pain(+)/gain(-) {pain_gain/1e6:+6.2f}M, "
          f"contractor margin {(fee - pain_gain)/1e6:5.2f}M, owner pays {(actual + fee - pain_gain)/1e6:6.2f}M")

actual   180M: contractor pain(+)/gain(-)  -8.00M, contractor margin 18.80M, owner pays 198.80M
actual   192M: contractor pain(+)/gain(-)  -4.00M, contractor margin 15.52M, owner pays 207.52M
actual   200M: contractor pain(+)/gain(-)  +0.00M, contractor margin 12.00M, owner pays 212.00M
actual   208M: contractor pain(+)/gain(-)  +4.00M, contractor margin  8.48M, owner pays 216.48M
actual   216M: contractor pain(+)/gain(-)  +6.80M, contractor margin  6.16M, owner pays 222.16M
actual   230M: contractor pain(+)/gain(-) +18.00M, contractor margin -4.20M, owner pays 225.80M

The bands concentrate the shared incentive near the target, where outcomes are most likely, while limiting the owner's exposure to a large overrun. At $230 million the contractor's share of the overrun exceeds its fee and it loses $4.2 million. A contractor facing a large overrun under the last band bears all of it, which protects the owner but can push the contractor toward claims, so some contracts instead cap the contractor's share at its fee, much like IPD. Getting the shares right is as much about behavior as arithmetic.

Incentives and key performance indicators

Progressive and alliance models often adjust the contractor's fee for performance on key results areas beyond cost: safety, quality, schedule milestones, environmental performance, community impact, Indigenous and local participation. Each indicator has a scale from poor to outstanding, and the fee is adjusted up or down according to a weighted score.

# Fee adjustment from key performance indicators in an alliance (illustrative)
base_fee = 9_000_000
kpis = {  # indicator: (weight, score from -1 poor to +1 outstanding)
    "Safety (recordable injury rate)":       (0.25, +0.6),
    "Quality (non-conformances closed)":     (0.20, +0.2),
    "Milestones achieved on time":           (0.25, -0.4),
    "Environmental incidents":               (0.10, +1.0),
    "Local and Indigenous participation":    (0.20, +0.5),
}
weighted = sum(w * s for w, s in kpis.values())
adjustment = 0.20 * weighted * base_fee          # fee can move up to +/- 20 %
for k, (w, s) in kpis.items():
    print(f"{k:38} weight {w:.2f} score {s:+.1f}")
print(f"Weighted score {weighted:+.3f}; fee adjustment {adjustment:+,.0f}; adjusted fee {base_fee + adjustment:,.0f}")

Safety (recordable injury rate)        weight 0.25 score +0.6
Quality (non-conformances closed)      weight 0.20 score +0.2
Milestones achieved on time            weight 0.25 score -0.4
Environmental incidents                weight 0.10 score +1.0
Local and Indigenous participation     weight 0.20 score +0.5
Weighted score +0.290; fee adjustment +522,000; adjusted fee 9,522,000

Indicators must be measurable, within the parties' control and balanced, so that chasing one does not damage another. A schedule incentive that encourages unsafe acceleration, for example, defeats its purpose.

Protecting the owner without price competition

Because construction prices in progressive models are negotiated, owners use several tools to make sure they are fair.

When progressive models are the right choice

Progressive and collaborative models suit projects with high technical complexity, significant unknowns such as ground conditions or existing facilities, many stakeholders and interfaces, or a market in which bidders will not accept fixed-price risk on reasonable terms. They are not a cure for every problem. They demand owner staff who can engage in joint design and commercial negotiation, they need strong governance to keep development phases from drifting, and they can be criticized for weaker competition. For simpler projects with clear requirements and competitive markets, conventional fixed-price methods remain efficient.