Bent Flyvbjerg on Megaprojects
What this covers
Bent Flyvbjerg of Oxford University joins Russ Roberts to dissect why billion-dollar infrastructure and technology projects fail so reliably. The conversation centers on a stark pattern: nine out of ten megaprojects underestimate costs, overestimate benefits, and miss schedules. Flyvbjerg argues this is not accident or incompetence but the predictable result of structural incentives. He identifies four sublimes—political, technological, economic, and aesthetic—that push decision-makers toward ever-larger projects regardless of true value, from politicians seeking visible legacies to engineers chasing records to contractors and unions seeking lucrative contracts. The conversation does not propose accepting these failures as inevitable; instead, Flyvbjerg presents reference-class forecasting, a de-biasing technique rooted in Danny Kahneman's behavioral economics that abandons planners' optimistic projections in favor of actual historical performance data from similar projects.
The discussion moves across specific failures—the Channel Tunnel's 80% construction overrun, rail systems forecasting twice as many passengers as appear—alongside the Megaproject Paradox: societies keep building these projects at growing cost despite persistent underperformance. Flyvbjerg touches on the Hiding Hand Principle, the practice of low-balling initial budgets so projects become too big to cancel, and the Bootlegger and Baptist pattern where self-interested groups cloak their push for public funding in civic ideals. A notable turn addresses democratic legitimacy: if citizens knowingly accept true costs, a project is legitimate; if they decide based on false forecasts, the decision is misinformed. The exchange also surfaces counterexamples—the Bilbao Guggenheim delivered on budget despite its complexity—and identifies conditions for success: experienced project managers, strong ownership with continuity beyond political cycles, and breaking large projects into smaller units. A recurring contested point is whether private megaprojects perform better than public ones; preliminary data suggest they do not.
Flyvbjerg argues that megaprojects systematically overrun costs and underdeliver benefits not by accident but because of a predictable incentive structure ('sublimes') and optimistic forecasting, and that realistic 'de-biasing' based on past project track records can produce far more accurate estimates.
- Nine out of ten megaprojects underestimate costs, overestimate benefits, and underestimate schedules.
- Political, technological, economic, and aesthetic 'sublimes' drive more and bigger projects regardless of true value.
- Reference-class forecasting / de-biasing using actual past performance corrects systematic optimism.
Politicians deliberately underestimate initial project costs to secure approval, counting on lock-in.
- The hiding hand operates in practice as politicians' deliberate low-balling: former San Francisco Mayor Willie Brown publicly stated that the first budget for civic projects is 'really just a down payment,' that if people knew the real cost nothing would ever be approved, and that the idea is to start digging a hole so big there's no alternative to funding it.
“In the world of civic projects, the first budget is really just a down payment. If people knew the real cost from the start, nothing would ever be approved.”
Self-interested groups leverage civic ideals to justify public funding that lacks broad cost-benefit support.
- Megaproject support follows the bootlegger-and-baptist pattern: a self-interested group (e.g., contractors, team owners, fans) allies with a high-minded justification (e.g., civic pride or 'being a Major League city') to win public funding, even though the broad benefits don't justify making taxpayers pay.
“the bootleggers are the people--they like banning liquor sales on Sunday because then people will buy from them... So you have an alliance between people who have a self-interested motive and the people who have a more high-minded motive.”