1. Optimism bias: documented reality, not an excuse
Optimism bias is one of the most robust findings of behavioural psychology applied to economics. Individuals, and teams, systematically tend to under-estimate the costs, durations and risks of the projects they run. It was formalised as the planning fallacy: we project our future undertakings as if everything will go to plan, without allowing for the statistical overruns observed on comparable projects.
In construction the bias is amplified by dynamics particular to the sector. The architect presenting a first cost band at concept stage is often in a selling posture, seeking to enthuse the client rather than cool them with prudent figures. The client, for their part, often has a target envelope in mind, and a low band reinforces their sense that the project is feasible. Those two dynamics converge on the same result: an initial estimate too low, which will become the psychological reference budget for the rest of the project.
ℹ Psychological anchoring: why the first figure sticks Once a client has heard "around three million euros" for their project, that figure anchors in their reference frame. When the economist announces 3.8 M€ at outline stage, the natural reaction is "why has it gone up?", when the reality is that the first figure was not reliable. That is why the quality of the first estimate presented is decisive: it need not be precise to 5 %, impossible at concept stage, but it must be honest about its uncertainties, its assumptions and its contingencies. A figure presented without a contingency is not a professional estimate.
2. The eight most frequent causes of under-estimation
| Type of bias or cause | Frequency | Typical scale | Who is responsible | The remedy |
|---|---|---|---|---|
| Optimism bias | ★★★★★ | 5 – 15 % | Everyone: architect, client, economist | A reference-class estimate by an independent economist |
| Client pressure on the figures | ★★★★ | 5 – 20 % | The economist under commercial pressure | An external economist with no interest in the design appointment |
| Incomplete brief data | ★★★★★ | 3 – 12 % | A brief badly defined by the client | A detailed brief approved before any estimate |
| Forgetting the invisible items | ★★★★★ | 15 – 40 % | Everyone; items outside the works are forgotten | A systematic checklist: land, fees, tax, external works |
| Obsolete or unsuitable ratios | ★★★★ | 5 – 20 % | An economist with a dated database | Ratios less than 12 months old, adapted to the area |
| Under-provided contingency | ★★★★★ | 5 – 20 % | An economist or client who compresses it | An explicit 10–20 % by phase, non-negotiable |
| Regulatory premiums not anticipated | ★★★★ | 3 – 15 % | Architect and engineers during design | A checklist of accessibility, environmental, seismic and acoustic requirements from concept stage |
| Regional factor not applied | ★★★ | 10 – 40 % | An economist using provincial ratios on a Paris or overseas project | Apply the right factor before communicating anything |
Two causes stand out for their combined frequency and scale: forgetting the invisible items, land, fees, tax, external works, which represent 30 to 60 % of the total cost, and under-providing the contingency. Taken together, those two omissions can generate a gap of 30 to 50 % between the works estimate presented at concept stage and the complete cost of the operation.
3. The invisible items
The most common and most expensive error is confusing the construction cost with the total cost of the operation. Construction ratios per square metre cover the works: foundations, structure, finishing trades, technical equipment, internal external works. They do not cover the following, which are systematically added.
- Land: variable with the area, it can represent 10 to 60 % of the total cost in tight markets. Analysing a Paris dwelling without the land is an economic fiction.
- Design and engineering fees: 12 to 18 % of the works cost according to complexity. Architect, structural engineer, services engineer, economist, site coordinator, technical controller, health and safety coordinator.
- VAT: 5.5 %, 10 % or 20 % according to the programme and the client's status. On a 5 M€ site, forgetting VAT at 20 % is a 1 M€ gap.
- External works and connections: public networks, water, drainage, electricity, gas, telecoms, access roads, external landscaping. 5 to 15 % of the building works according to the configuration of the site.
- Insurance and finance costs: 1 to 3 % of the total cost according to the structure and the holding period.
- Marketing and management costs: for developers and landlords, sales costs, initial letting management and the completion guarantee are added to the appraisal.
ℹ Refurbishment: the uncertainty doubled On refurbishment every under-estimation bias is present, and structural uncertainty is added. Discoveries in the existing fabric, asbestos, lead, degraded structure, non-compliant services, are statistically very frequent and can represent 20 to 60 % above a prudent outline estimate. The contingency on a heavy refurbishment must be 15 to 25 % at outline stage, not the 10 % used for new build. Any refurbishment estimate presented with a contingency below 15 % is suspect.
4. Gaps observed between concept estimate and reality
To move beyond generalities, here are the gaps observed between first concept bands and developed-design estimates on common programmes. They are median values; projects with substantial scope creep can exceed them.
| Programme | Typical concept ratio | Real developed-design ratio | Average gap | Principal cause |
|---|---|---|---|---|
| Standard apartments | 1,800 – 2,000 €/m² | 2,000 – 2,300 €/m² | + 10 – 15 % | Accessibility, acoustics and external works under-estimated |
| Social housing | 1,700 – 1,900 €/m² | 1,900 – 2,200 €/m² | + 10 – 15 % | Full accessibility, signage, insulation to the current regulations |
| Offices, category A | 1,600 – 1,800 €/m² usable | 1,800 – 2,200 €/m² usable | + 12 – 22 % | Technical equipment, cabling, urban external works |
| Three-star hotel | 130 – 160 k€ a room | 160 – 220 k€ a room | + 15 – 30 % | Operating equipment, technical plant and finishes under-valued |
| Primary school | 1,800 – 2,000 €/m² | 2,200 – 2,800 €/m² | + 20 – 40 % | Acoustics, ventilation, catering, accessibility: all forgotten |
| Ambient warehouse | 400 – 500 €/m² | 430 – 580 €/m² | + 8 – 16 % | Offices, sprinklers and internal external works under-provided |
| Heavy refurbishment | 800 – 1,000 €/m² habitable | 1,100 – 1,600 €/m² habitable | + 20 – 60 % | Discoveries in the existing fabric, compliance works, risk doubled |
Heavy refurbishment shows the largest and most variable gap, from 20 to 60 % according to the quality of the prior investigation of the existing fabric. That is precisely why the surveys, asbestos, structure, services, must be carried out before any serious outline estimate on an existing building.
5. Commercial pressure on the economist: a systemic risk
Under-estimation is not always an unconscious bias. It can result from explicit or implicit commercial pressure on the economist, which takes several forms.
- Direct pressure from the client: "Your colleague gave me 2 M€, you are at 2.8 M€, you must be wrong." An economist who yields produces a pleasing but irresponsible estimate.
- Indirect pressure from the lead architect: in a combined design appointment, the architect sometimes has an interest in the project being judged feasible in order to secure the commission. An economist inside the team can face implicit pressure not to spoil the mood.
- Alignment on a predefined envelope: the client states their budget before the estimate. The economist unconsciously produces a figure that fits the envelope rather than one that reflects the market.
The remedy to all three is the same: an independent external economist, instructed directly by the client, with no commercial interest in the project proceeding. Their estimate may be unwelcome, which is precisely why it is valuable.
6. The checklist for a reliable outline estimate
A professional outline estimate is not a black box: it documents its assumptions, its sources and its provisions. The following checklist lets any client verify the reliability of an estimate received.
| Control point | Responsible | Warning signal if absent |
|---|---|---|
| A ratio less than 12 months old, adapted to the area | Economist | A conference ratio or a national ratio on an overseas or Paris project |
| The reference area identified and coherent | Economist | Confusion between floor, habitable and usable area, an apparent gap of 15–25 % |
| A detailed brief approved by the client before estimating | Client | An estimate on an incomplete brief, a whole package forgotten |
| Items outside the works listed explicitly | Client and economist | Works budget mistaken for total cost, an error of 30–60 % |
| An explicit contingency of at least 10 % at outline | Economist | No contingency, or one melted into the ratios: total opacity |
| A price revision provision where more than 12 months separate estimate and contracts | Economist | Offers above the design budget in a period of inflation |
| Regulatory premiums verified | Architect and economist | Discovery at developed design, a painful budget revision |
| The regional factor applied | Economist | A provincial budget applied to a Paris project, 30 % adrift |
| Assumptions documented and communicable | Economist | A black-box estimate, impossible to defend or verify |
| Specification levels stated | Client and architect | Comparison impossible, ratios not comparable between projects |
7. What the economist actually does to prevent it
An economist intervening at outline stage brings three things neither the architect nor the client can produce alone: a database of current ratios from recent comparable projects; a costing method by trade that breaks the brief into measurable elements; and an independence of judgement not compromised by the architect's creative enthusiasm nor by the client's budget constraints.
On a project with 3 M€ of works, the outline estimate costs 1,000 to 3,000 €. If it reveals that the real project costs 3.6 M€ rather than the 2.8 M€ assumed, it lets the client readjust the funding before the planning application rather than after. That early intervention avoids the escalating-commitment syndrome: continuing with an unviable project because too much has already been invested to turn back.
In summary
Under-estimation at concept and outline stage results from eight cumulative mechanisms: optimism bias, commercial pressure, an incomplete brief, forgetting the invisible items, obsolete ratios, an absent contingency, unanticipated regulatory premiums, and the regional factor not applied. The gaps observed between concept estimate and developed-design reality run from 10 % on standard apartments to 60 % on heavy refurbishment. The systematic protection is an outline estimate by an independent economist, with an explicit contingency of at least 10 % and a checklist of the items outside the works. An outline estimate at 1,500 € that prevents a 300,000 € overrun returns two hundredfold.
The nature of the figures quoted
The percentages and amounts quoted on this page are orders of magnitude for framing, not measurements. They illustrate mechanisms and proportions, and they substitute neither for a tender exercise nor for an estimate prepared on drawings.
Articles liés
Sources: the BT construction cost indices published monthly by INSEE on its 2010 base, used to update the figures · construction cost statistics from SDES, the statistical service of the ministry responsible for construction · regulatory texts published in the Journal officiel and consolidated on Légifrance for the requirements cited.