For thirty years British infrastructure has told itself that target cost contracts deliver better outcomes than the lump-sum contracting still standard across much of Europe. The comparative evidence does not support that confidence. Britain is now the most expensive place in Europe to build - and the collaborative model adopted in the name of efficiency demonstrably failed to prevent it. The contract form did not cause that gap, and lump-sum would not have closed it; but target cost was sold as the cure, and the evidence that it is one has never been assembled.
The sharper contention is not that target cost is a bad instrument - it is that Britain over-claimed for it, and reached for it as a substitute for the disciplines that actually determine whether projects succeed, disciplines that the cheaper European builders never gave up. The charge in this piece is therefore against the orthodoxy, not the tool: against the belief that choosing a collaborative contract is itself a delivery strategy. It is diagnostic, not a call to abandon NEC or revert to lowest-price lump-sum.
The one clean comparison
The foundations are familiar. Sir Michael Latham's 1994 report Constructing the Team recommended the New Engineering Contract (NEC) as a less confrontational standard form, setting the totemic target of a 30% real reduction in construction costs by 2000 [1]. The NEC became the default for UK public infrastructure. Even Latham's admirers accept the 30% pledge was missed [2]. None of this disproves the model. But it should make us cautious about treating three decades of advocacy as three decades of evidence.
There is one metric on which UK and continental delivery can be compared with some rigour: the capital cost of building a kilometre of railway. The Transit Costs Project finds that HS2 is the single most expensive project in its entire database, at roughly $232m per kilometre - and that its cost is not explained away by its share of tunnelling or elevated structure. Germany's Wendlingen-Ulm high-speed line came in at about $71m/km, under a third of HS2 [3].
The unit-cost gap has many causes that have nothing to do with how the work was contracted: land, consenting, specification, geology and a stop-start funding pipeline. And HS2 is a single, extreme data point, not a population. But neither caveat rescues the orthodoxy. The honest reading is not that target cost caused the gap - it is that British practice reorganised itself around target cost in the name of closing a gap that target cost did nothing to close.
Nor is lump-sum Europe a panacea. Finland built Olkiluoto 3 under a fixed-price €3.2bn turnkey contract under which the contractor was to absorb any overrun; it came in at around €11bn and roughly 14 years late [4]. Bent Flyvbjerg's data show nine in ten megaprojects exceed budget, a pattern constant for seventy years across every geography and ownership model [5]. On one-off megaprojects, the contract form is a second-order variable.
The European builders who deliver cheaply share a profile, and it is not a contractual one. Spain, the standout [6]:
- builds fast, so less time is exposed to inflation and political change
- standardises designs across a long pipeline
- streamlines consenting
- sustains a capable, continuous public client.
The lesson is repetition, standardisation, speed and client capability, not the colour of the contract. The UK's own genuine successes fit the same pattern. The A14 Cambridge to Huntingdon scheme, on NEC3 Option C, opened around eight months ahead of schedule and on budget [7], with a permanent, expert public client that builds roads continuously, a mature project type and an integrated delivery team. It is the exception that states the rule.
Target cost works where the client and the work are already disciplined, and fails where it is asked to substitute for that discipline.
Why Britain reaches for target cost
There is a serious structural defence of the UK's preference. For nationally significant infrastructure, consent comes through the Development Consent Order regime under the Planning Act 2008 - and a DCO is not granted on a fixed design. It is granted on a parameters envelope, the so-called Rochdale Envelope, assessed on a worst-case basis, with detailed design deliberately left to vary post-consent [8]. In that world, appointing a contractor early to help define what is buildable, under a target cost contract that can flex as the design firms up, is not laziness; it is a rational adaptation.
But that defence justifies target cost for the genuinely undefinable residue. It does not justify using target cost as a blanket substitute for defining the parts that were always within the client's gift. HM Treasury and IPA (now NISTA) work on major projects records the chronic pattern: schemes moving forward before scope and benefits are properly defined, reliance on immature early estimates, blurred accountability and systematic optimism bias [9]. The genuine uncertainty created by the consenting regime supplies respectable cover for the self-inflicted uncertainty created by weak client definition.
HS2 shows where that leads. Its Main Works Civils Contracts were, on the government's own description, based on the NEC3 Option C target cost model [10]. Yet the incoming chief executive, Sir Jon Thompson, told MPs in 2023 that they had been let on a cost-plus basis with "very few incentives or penalties", noting that a contractor spending 100% more than originally agreed would suffer only a 1% reduction in fee [11]. In June 2026 the National Audit Office reported the reset programme's cost at £87.7bn to £102.7bn, roughly double the 2020 estimate [12]. Strip the pain share down to a rounding error and a target cost contract is simply cost reimbursement in collaborative costume. The objection is not that the label was misapplied; it is that the form permits this.
Open-book transparency also assumes a client that can read the book. Contractors field deep benches of commercial managers, quantity surveyors and claims specialists whose full-time business is the operation of these contracts; the client rarely matches them in-house. A client that cannot read the book is not collaborating; it is being shown what it is given.
The real charge
A target cost contract lets a client start before the design is mature, because the price is provisional by design. It defers cost certainty, which is convenient for a sponsor who wants to announce a project before it is properly defined. And it appears to outsource cost discipline to a pain-share mechanism - which, as HS2 demonstrates, is too often set so soft that it disciplines no one. The collaborative rhetoric supplied a respectable language for proceeding without the things that make projects cheap: completed design, standardisation, a continuous and capable client, and a programme rather than a parade of bespoke one-offs.
Continental lump-sum practice, for all its adversarial reputation, forces the client to finish the design before committing and to confront a real price up front. It is no guarantee against failure on one-off megaprojects. But on the routine, repeatable majority of work it imposes a discipline that Britain has allowed its collaborative model to relax.
An honest scorecard
Clean, like-for-like outcome comparisons between UK and European delivery barely exist outside the rail-cost data, and the academic literature on the NEC, after thirty years, still rests largely on practitioner-satisfaction surveys rather than systematic outcome evidence [13]. NEC4 is the current edition, yet essentially all the evidence, good and bad, concerns NEC3. "NEC4 has been good for clients" is, at present, an extrapolation and a marketing proposition, not a finding.
So the scorecard reads as follows. Target cost contracting is a sound tool in its proper place: genuinely uncertain scope, real early buildability value, a capable client, and a pain share hard enough to bite. Used that way - on programmatic, repeatable work by a strong client - it can help, as Anglian Water's @one Alliance shows [14]. Adopted as a general philosophy, and as a substitute for design maturity and client capability on one-off megaprojects, it has coincided with the worst infrastructure cost performance in Europe. The contract form is downstream of capability, and Britain spent thirty years optimising the contract while letting the capability decay.
Thirty years after Latham, the most useful thing the industry could do is stop confusing collaboration with proof, rebuild client capability and design discipline, and choose contract forms on the evidence rather than the orthodoxy. A word on standpoint: that front-end definition and client capability is the work Synaform's Advice and Process services do, so the recommendation is not disinterested and should be weighed with that in view. The evidence it rests on stands on its own terms.
An earlier version of this article was first published by Effectus Construction Services at [effectuscs.com](https://www.effectuscs.com/insights/target-practice).
Sources
- [1] Latham, M. (1994) Constructing the Team; the 30% cost-reduction target. Designing Buildings Wiki, "Latham Report". https://www.designingbuildings.co.uk/wiki/Latham_Report
- [2] Building, "Latham's report: Did it change us?" (missed 30% target). https://www.building.co.uk/focus/lathams-report-did-it-change-us/5069333.article
- [3] Transit Costs Project (NYU Marron Institute), high-speed rail preliminary data analysis (HS2 ~$232m/km as the most expensive in the database; Wendlingen-Ulm ~$71m/km). https://transitcosts.com/high-speed-rail-preliminary-data-analysis/ .
- [4] World Nuclear News, "Olkiluoto 3 EPR parties agree settlement", https://www.world-nuclear-news.org/Articles/Olkiluoto-3-EPR-parties-agree-settlement ; Olkiluoto Nuclear Power Plant, Wikipedia (fixed-price €3.2bn turnkey contract signed Dec 2003; overruns to ~€11bn; ~14 years late), https://en.wikipedia.org/wiki/Olkiluoto_Nuclear_Power_Plant ; World Nuclear Report, "Europe's First EPR: 13 Years Behind Schedule", https://www.worldnuclearreport.org/Europe-s-First-EPR-13-Years-Behind-Schedule-Olkiluoto-3-in-Finland-Starts-Up
- [5] Flyvbjerg, B., "The Iron Law of Megaprojects" (nine in ten over budget; constant over 70 years; all countries, public and private), https://medium.com/data-science/the-iron-law-of-megaprojects-18b886590f0b ; Flyvbjerg, B. (2014) "What You Should Know About Megaprojects and Why: An Overview", Project Management Journal, https://arxiv.org/abs/1409.0003
- [6] Works in Progress, "How Madrid built its metro cheaply" (speed, standardisation, streamlined consenting, capable continuous client). https://worksinprogress.co/issue/how-madrid-built-its-metro-cheaply/
- [7] NEC, "A14 Cambridge to Huntingdon improvement scheme, UK" (NEC3 ECC Option C; completed on budget and ahead of programme), https://www.neccontract.com/projects/a14-cambridge-to-huntingdon-improvement-scheme-uk ; New Civil Engineer, "How the A14 upgrade was completed eight months early amid a global pandemic" (12 Jan 2021), https://www.newcivilengineer.com/latest/how-the-a14-upgrade-was-complete-eight-months-early-amid-a-global-pandemic-12-01-2021/
- [8] Planning Inspectorate / NSIP guidance on the Rochdale Envelope and post-consent design flexibility, https://www.gov.uk/guidance/nationally-significant-infrastructure-projects-advice-on-preparing-applications-for-linear-projects ; Morphet and Clifford (2017), "Infrastructure Delivery: the DCO process in context", NIPA/UCL, https://nipa-uk.org/wp-content/uploads/2023/08/UCL_Morphet_and_Clifford_-_NIPA_Main_Report_-_June_2017.pdf
- [9] HM Treasury, The Green Book and major-projects guidance on front-end definition and optimism bias, https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-government ; Lords Library briefing on the Green Book review, https://lordslibrary.parliament.uk/governments-green-book-review-and-reforms/
- [10] UK Parliament, Written Question HL8035, 3 June 2025 (HS2 Main Works Civils based on the NEC3 Option C target cost model). https://questions-statements.parliament.uk/written-questions/detail/2025-06-03/HL8035/
- [11] House of Commons Public Accounts Committee, oral evidence on HS2 (Sir Jon Thompson, 2023): the "very few incentives or penalties" / cost-plus criticism and the 1% fee reduction for a 100%-over-budget contractor. Reported in Construction Enquirer, "New HS2 boss blasts cost-plus contracts for £6bn of overspend" (17 Nov 2023), https://www.constructionenquirer.com/2023/11/17/new-hs2-boss-blasts-cost-plus-jobs-for-6bn-of-overspend/ ; New Civil Engineer, 17 Nov 2023, https://www.newcivilengineer.com/latest/hs2-phase-1-cost-close-to-exceeding-benefit-as-officials-blame-civils-contracts-17-11-2023/
- [12] National Audit Office, High Speed Two reset, HC 52, 29 June 2026, para 8 (£87.7bn-£102.7bn, about double the 2020 estimate; mixed price base). https://www.nao.org.uk/wp-content/uploads/2026/06/high-speed-two-reset-summary.pdf
- [13] Rooke/Wright, NZ ECC case study (ScienceDirect), https://www.sciencedirect.com/science/article/abs/pii/S0263786308000483 ; Hong Kong NEC performance-index Delphi study, International Journal of Construction Management, https://www.tandfonline.com/doi/abs/10.1080/15623599.2022.2100058
- [14] Binnies, "Anglian Water's @one Alliance", https://binnies.com/frameworks/anglian-waters-one-alliance/ ; New Civil Engineer, "How @one Alliance's model and lessons learned could improve delivery of its £2.6bn AMP8 portfolio" (30 Jan 2025), https://www.newcivilengineer.com/in-depth/how-one-alliances-model-and-lessons-learned-could-improve-its-amp8-2-6bn-portfolio-delivery-30-01-2025/