IndependentBudgetOverrun.com is an independent reference site. No PM software vendor owns, sponsors or reviews it; one PM tool link is a labelled partner link.
Reference / Trust surface

Methodology and sources

BudgetOverrun.com compiles project cost-overrun figures that other people have published. It holds no dataset of its own. This page lists every document behind the headline figures, the exact link, what we take from it, and the date we last opened it and checked the figure.

Data verified October 2026

Sources ledger

19 documents. We opened 17 of them on or before 8 October 2026 and checked each figure against the text, table or exhibit named in its entry. The rest are marked: a paid report we have not read, or a book we could not check page by page; their figures are labelled the same way wherever they appear on the site.

Pulse of the Profession 2018: Success in Disruptive Times
Project Management Institute (PMI), 2018.
Read 8 October 2026 · Free to read

Appendix (report p. 25), global mean of 4,455 practitioners' estimates for projects completed in the prior 12 months: 57% finished within their initial budgets (so 43% did not), 52% experienced scope creep, 69% met original goals, 15% were deemed failures, 32% of budget lost when a project fails.

Pulse of the Profession 2017: Success Rates Rise
Project Management Institute (PMI), 2017.
Read 8 October 2026 · Free to read

Appendix: 57% of projects finished within their initial budgets (3,234 respondents). Cover: the survey has been conducted since 2006.

Pulse of the Profession 2021: Beyond Agility
Project Management Institute (PMI), 2021.
Read 8 October 2026 · Free to read

Figure 2, Better Project Outcomes: projects completed within original budget 57% (2019), 59% (2020), 62% (2021).

The High Cost of Low Performance: The Essential Role of Communications
Project Management Institute (PMI), Pulse of the Profession in-depth report, May 2013.
Read 8 October 2026 · Free to read

US$135 million at risk for every US$1 billion spent on projects; 56% of that (US$75 million) at risk due to ineffective communications.

CHAOS Report 2015
The Standish Group International, 2015.
Read 8 October 2026 · Free to read

Copy of the full report hosted by Info-Tech Research Group. Modern Resolution, all projects, FY2015: 29% successful, 52% challenged, 19% failed (25,000-plus projects, FY2011-2015). Page 3: project size by CHAOS resolution (Small 61/32/7 through Grand 6/51/43). Agile versus waterfall table (39% vs 11% successful).

CHAOS Report: Beyond Infinity (CHAOS 2020)
Jim Johnson. The Standish Group International, 2020.
Not read by us (listing checked 8 October 2026) · Paid report

Sold by the Standish Group (digital edition, US$450). We have not read the report. The split it is quoted for (31% successful, 50% challenged, 19% failed) comes from secondary write-ups, so we label it as reported. The figures we have checked ourselves are the CHAOS 2015 ones above.

Delivering large-scale IT projects on time, on budget, and on value
Michael Bloch, Sven Blumberg, Jürgen Laartz. McKinsey & Company, with the University of Oxford, 1 October 2012.
Read 8 October 2026 · Free to read

More than 5,400 IT projects; large IT projects (initial budget above $15 million) run 45% over budget and 7% over time on average and deliver 56% less value than predicted; total cost overrun of $66 billion across the sample; 17% go so badly they can threaten the company (black swans: budget overruns above 200%).

Unlocking the potential of public-sector IT projects
Kerstin Balka, Breanna Heslin, Sina Risse-Tenk. McKinsey & Company, with Oxford Global Projects, 5 July 2022.
Read 8 October 2026 · Free to read

6,003 IT projects completed 2001-2017: cost overruns occur in nearly one in two public-sector projects against about one in three private-sector; the average public-sector IT cost overrun is nearly three times the private-sector average; 14% of public-sector projects exceeded budget by at least 100% (7% private); 59% of all projects completed within budget; one in 200 met budget, schedule and benefits.

Why Your IT Project May Be Riskier Than You Think
Bent Flyvbjerg, Alexander Budzier. Harvard Business Review, vol. 89 no. 9, September 2011.
Read 8 October 2026 · Metered (free author copy linked) · free author copy (arXiv 1304.0265)

1,471 IT projects: average cost overrun 27%; one in six was a black swan with a cost overrun of 200% on average and a schedule overrun of almost 70%.

The Empirical Reality of IT Project Cost Overruns: Discovering a Power-Law Distribution
Bent Flyvbjerg, Alexander Budzier, Jong Seok Lee, Mark Keil, Daniel Lunn, Dirk W. Bester. Journal of Management Information Systems 39(3), 607-639, 2022.
Read 8 October 2026 · Free to read · free author copy (arXiv 2210.01573)

5,392 IT projects completed 2002-2014: cost overruns follow a power-law (fat-tailed) distribution, so averages understate the risk.

The construction productivity imperative
Sriram Changali, Azam Mohammad, Mark van Nieuwland. McKinsey & Company, 1 July 2015.
Read 8 October 2026 · Free to read · PDF version with Exhibit 2

Text: 98% of megaprojects suffer cost overruns of more than 30%; 77% are at least 40% late. Exhibit 2 (megaprojects in mining, oil and gas and infrastructure; sources: company annual reports, IHS Herold Global Projects Database, November 2013): the average cost increase is 80% of original value and the average slippage is 20 months.

Megaprojects: The good, the bad, and the better
Nicklas Garemo, Stefan Matzinger, Robert Palter. McKinsey & Company, 1 July 2015.
Read 8 October 2026 · Free to read

Context only. It cites Flyvbjerg's nine-in-ten finding and gives McKinsey estimates of 35% average overrun for bridges and tunnels and 20% for roads. It does not contain the 98% or 80% figures; this site previously attributed them to it in error.

Underestimating Costs in Public Works Projects: Error or Lie?
Bent Flyvbjerg, Mette K. Skamris Holm, Søren L. Buhl. Journal of the American Planning Association 68(3), 279-295, Summer 2002.
Read 8 October 2026 · Free to read

258 transport infrastructure projects in 20 countries on 5 continents, completed 1927-1998: costs underestimated in almost 9 out of 10 projects (86% likelihood that actual cost exceeds the estimate); actual costs on average 28% higher than estimated; rail 44.7%, fixed links (bridges and tunnels) 33.8%, roads 20.4% (Table 1).

What You Should Know About Megaprojects and Why: An Overview
Bent Flyvbjerg. Project Management Journal 45(2), 6-19, 2014.
Read 8 October 2026 · Free to read

The iron law of megaprojects (over budget, over time, over and over again); nine out of ten megaprojects have cost overruns; megaproject spending of USD 6-9 trillion a year, about 8% of global GDP; real-terms overruns of 80% for the Channel Tunnel, 200% for Denver International Airport, 220% for Boston's Big Dig, 1,400% for the Sydney Opera House.

How Big Things Get Done
Bent Flyvbjerg, Dan Gardner. Currency (Penguin Random House), ISBN 9780593239513, 2023.
Not read by us (listing checked 8 October 2026) · Book

Book; we have not checked these against the printed page. As quoted consistently: 16,000-plus projects, 47.9% on budget, 8.5% on budget and on time, 0.5% on budget, on time and on benefits (chapter 1); appendix base-rate table (around p. 192): IT mean overrun 73%, 18% of IT projects in the 50%-plus tail with a tail mean of 447%, nuclear waste storage mean 238%. The 2026 Public Management Review paper below cites the same appendix page for the nuclear-storage ranking.

Regression to the Tail: Why the Olympics Blow Up
Bent Flyvbjerg, Alexander Budzier, Daniel Lunn. Environment and Planning A: Economy and Space 53(2), 2021 (online September 2020).
Read 8 October 2026 · Free to read · free author copy (arXiv 2009.14682)

Every Olympics 1960-2016 with data ran over budget, at an average of 172% in real terms. Table 3: per-Games sports-related cost overrun, real terms, for the 19 Games with valid data.

Do projects learn across space and time? Evidence from the Olympics
Atif Ansar, Bent Flyvbjerg, Alexander Budzier. Public Management Review (open access), 2026.
Read 8 October 2026 · Free to read · arXiv 2604.17970

Table 1: 23 of 23 Olympic Games 1960-2024 with available data overran, mean cost overrun 159.3% in real terms (standard deviation 160.3%); Montreal 1976 overran by 720%.

Supplementary Green Book Guidance: Optimism Bias
HM Treasury (prepared by Mott MacDonald), 2003 (study 2002).
Read 8 October 2026 · Free to read

Table 1 upper-bound capital expenditure uplifts: standard buildings 24%, non-standard buildings 51%, standard civil engineering 44%, non-standard civil engineering 66%, equipment and development 200%.

Cost Performance Index Stability
David S. Christensen, Scott R. Heise. National Contract Management Journal 25(1), 1993.
Read 8 October 2026 · Free to read

155 US Department of Defense contracts: the cumulative CPI was stable (range under 0.2) from the 20% completion point for 134 of the 155, and it usually got worse, not better, as the contract proceeded.

Case-study pages (HS2, Crossrail, Big Dig, Berlin Brandenburg, Sentinel and others) and the defence pages cite their audit reports inline, next to the figures they support. Corrections to this ledger are dated in the page history of the Digital Signet repository.

In scope

  • ●Published research statistics on project budget overrun rates, each linked to the document that states it.
  • ●Sector-specific overrun rates (construction, IT, government, infrastructure, defence) from the documents in the ledger.
  • ●Earned Value Management formula reference (PMI / AACE standard notation) and worked examples.
  • ●Historic case-study overrun figures from audit reports, official statements and peer-reviewed papers.
  • ●Free EVM calculator with illustrative presets that runs entirely in the browser.
  • ●Contract-type allocation of overrun risk under industry-standard contract families.

Out of scope

  • ●Original data. This site compiles published figures; it does not run surveys, hold a project database, or publish a downloadable dataset.
  • ●Private or confidential project cost data from any individual organisation.
  • ●Predictions about whether a specific named project will overrun, by how much, or when.
  • ●Vendor-by-vendor PM software comparisons or rankings. The short PM software lists on /prevention and /templates name three widely used tools with budget tracking, in the order they had before any partner programme existed; one link (monday.com) is a labelled partner link, and none of the three is ranked or endorsed.
  • ●Country-specific legal advice on contract disputes, variation claims, or adjudication procedure.

Calculation framework

EVM formula stack

PMI / AACE standard notation: PV (Planned Value), EV (Earned Value), AC (Actual Cost). CPI = EV/AC. SPI = EV/PV. Three primary EAC methods: EAC = BAC/CPI (assumes current efficiency continues), EAC = AC + (BAC - EV) (optimistic), and EAC = AC + ETC (bottom-up re-estimate). A fourth: EAC = AC + [(BAC - EV) / (CPI x SPI)]. VAC = BAC - EAC. TCPI = (BAC - EV) / (BAC - AC) when staying on BAC, or (BAC - EV) / (EAC - AC) when staying on the new EAC.

Overrun percentage math

Overrun% = (AC - BAC) / BAC x 100, where BAC is the originally approved budget and AC is the final actual cost. Worked example: Sydney Opera House (AUD 102M - AUD 7M) / AUD 7M x 100 = 1,357%. Case-study percentages on /famous-examples use this formula against the original estimate named in the row, in the currency and price basis the source gives (nominal unless the row says real terms).

CPI stability

Christensen and Heise (1993) studied 155 US Department of Defense contracts: the cumulative CPI was stable (a range under 0.2) from the 20% completion point for 134 of the 155, and it usually got worse rather than better as work continued. The CPI bands on the calculator (0.95 to 1.1 on track, 0.8 to 0.95 action required, below 0.8 severe) are our own working thresholds, not figures from that study.

Reference class forecasting

Flyvbjerg's outside-view method: take the actual cost outcomes of comparable completed projects as the anchor, then adjust for project-specific risk. The published uplift tables we use are HM Treasury's optimism-bias guidance and the UK Department for Transport's 2004 percentile table, both shown with their sources on /reference-class-forecasting.

Contingency

We do not publish a house contingency percentage. Where a page discusses contingency it points to a published benchmark: AACE International RP 18R-97 estimate-class accuracy ranges, or the HM Treasury optimism-bias uplifts (24% standard buildings, 44% standard civil engineering, up to 200% for equipment and development).

Rounding and derived figures

Where we derive a number from a published one, the page shows the arithmetic. Example: PMI's 2018 survey reports 57% of projects finishing within their initial budgets; the 43% we quote is 100% minus 57%. Book figures we could not check page by page are labelled as quoted.

Review dates

The site-wide "data verified" date is held in one constant and moves only after a source-by-source review. The most recent review (8 October 2026) re-read every document marked with that date in the ledger above and corrected what did not match: the McKinsey megaproject figures now cite the article that states them, PMI's 43% is shown as the remainder of the 57% that PMI reports, the public-sector IT study is dated 2022, and the Olympic Games figures follow the peer-reviewed papers. New editions of the surveys (PMI Pulse, Standish CHAOS) and new GAO or NAO audits of the case-study projects are the usual reasons for an update between reviews.

Limitations

  • ●This is a compilation. The figures belong to the studies that published them, with those studies' definitions and samples; read the linked document before reusing a figure in research.
  • ●The Standish CHAOS reports and PMI Pulse surveys rest on respondent-reported outcomes and samples whose frames are not fully published. The Standish 2020 report is sold, and we have not read it.
  • ●Megaproject figures depend on each author's selection criteria; those criteria are described on the relevant study page.
  • ●Historic case-study figures shift as final audits land (HS2 is an in-flight programme; published cost figures lag the latest committed-cost position).
  • ●Calculator outputs are working bands, not predictions for any specific project.

Editorial position and commercial relationships

This site is operated by Digital Signet, an independent reference-content studio. Digital Signet does not sell project-management software, does not act as a project-rescue consultancy, and does not run a programme-governance practice. The site's commercial elements today are one labelled partner link (monday.com, on /prevention and /templates) that may pay a commission, and a sponsorship slot that is currently unsold. No figure or ranking on the site is paid for. See /about for the operator and the wider portfolio context.

Editorial direction is set by Digital Signet's editor. Drafts are produced with AI assistance and checked against the documents in the ledger before publication. Where sources disagree the page presents the range with the methodology context, not a single point estimate.

Corrections

For methodology questions, stale figures, or a better primary source, email [email protected] with the page URL and the document you would like cited. Substantive corrections are typically actioned within five business days.

Read next

Didn't find your answer?

Ask us. A real person reads every question and we answer the ones we can, with sources. If your question would help other readers, we may publish an anonymised version, with your permission. General reference, not financial advice.

Updated 2026-10-08