Planning Fallacy
Planning Fallacy: Plans are almost always optimistic. The Sydney Opera House was budgeted at $7M and cost $102M. Software projects consistently run 2β3Γ over. Your next project will likely take longer and cost more than you plan. The fix is simple but underused: ignore your plan and look at how long similar projects actually took. Base your estimate on that β it's almost always higher than your inside view.
What Is the Planning Fallacy?β
Daniel Kahneman and Amos Tversky described the Planning Fallacy in 1979. The bias: when planning, people focus on the specific details of their plan (the "inside view") while ignoring the base rate of similar projects (the "outside view"). Their plan has specific advantages that explain why it will succeed; they overlook the class of projects like theirs, most of which ran over.
The fallacy has three components: underestimating time, underestimating costs, and overestimating benefits. All three contribute to systematically over-optimistic plans that lead to budget overruns, delayed deliveries, and disappointment.
How It Worksβ
Inside view (produces planning fallacy):
β Focus on this specific project
β Generate a narrative of steps and how each will go
β Estimate based on best-case scenario
β Result: optimistic estimates
Outside view (corrects planning fallacy):
β Find the reference class: similar projects of this type
β What's the distribution of outcomes for those projects?
β How did they compare to initial estimates?
β Anchor estimate in that distribution
Reference Class Forecasting process:
1. Choose reference class (similar projects)
2. Find the distribution of outcomes (time, cost, benefit)
3. Make initial estimate using the reference class as anchor
4. Adjust for specific features of your project
5. Your estimate should still be well within the distribution
Three Real-World Examplesβ
Sydney Opera Houseβ
Budget: A$7 million. Projected completion: 1963. Actual cost: A$102 million. Actual completion: 1973. The overrun β 14Γ cost, 10 years late β is extreme but not unusual for major public infrastructure projects. Bent Flyvbjerg's analysis of 258 infrastructure projects found that 90% came in over budget, by an average of 28%, with large projects averaging much higher overruns.
Software Developmentβ
A widely cited study of software projects found that the average project takes 2Γ the estimated time and exceeds budget by 45%. The CHAOS Report (Standish Group) finds only 29% of software projects are completed on time and on budget. Despite decades of evidence, software estimates remain optimistic because each team believes their specific project has differentiating advantages. The fix β examining the distribution of similar past projects β is rarely applied.
Individual Life Projectsβ
Research by Roger Buehler, Dale Griffin, and colleagues found that students completing academic projects consistently underestimated completion time by 50β100%, even when asked to consider "worst case" scenarios. When explicitly asked to recall how long similar past projects took, their estimates improved substantially. The same pattern applies to renovation projects, book writing, and business plan execution.
When to Recognise Itβ
π¨ Planning Fallacy is likely operating when:
- Your estimate assumes everything goes to plan with no major obstacles
- You haven't examined how long similar projects historically took
- The plan was developed entirely from the inside view
- You're more optimistic than the base rate of similar projects warrants
β Countermeasures:
- Reference Class Forecasting: find similar projects and base estimate in their distribution
- Add a "planning buffer" β Kahneman recommends adding 50% to typical project estimates
- Pre-mortem: assume the project is late and over budget; ask what happened
- Track predictions: review how your past estimates compared to actuals
| Pairs well with | Why |
|---|---|
| Reference Class Forecasting | RCF is the direct antidote to the planning fallacy |
| Optimism Bias | Planning fallacy is partly driven by optimism bias about project-specific factors |
| Pre-mortem | Pre-mortems counteract planning fallacy by surfacing risks before commitment |
Common Misuses and Limitationsβ
Applying it to all estimates equally. The planning fallacy is strongest for novel, complex, long-duration projects. Routine, well-understood tasks with good historical data show less planning fallacy. The bias is largest when the project is unique enough that historical reference classes are hard to identify.
Using it as an excuse for not planning. The planning fallacy doesn't mean plans are useless β it means estimates should be calibrated against historical base rates. Better planning means: plan the steps (inside view) AND anchor estimates to historical distributions (outside view).
Related Modelsβ
| Model | Relationship |
|---|---|
| Reference Class Forecasting | The direct antidote to planning fallacy |
| Optimism Bias | Planning fallacy is driven by optimism about project-specific factors |
| Pre-mortem | Pre-mortems surface planning fallacy risks before commitment |
Frequently Asked Questionsβ
Why does the planning fallacy persist even when people know about it?
Because the inside view is cognitively compelling: your project has specific features that make the past distribution seem less applicable. "Yes, most software projects run over, but our team is experienced, the requirements are clear, and we've scoped it carefully." This reasoning is exactly the planning fallacy β each of these differentiators was also present in the projects that ran over. The discipline of using the outside view requires deliberate override of the compelling inside narrative.
What is Reference Class Forecasting and how does it work?
Reference Class Forecasting (developed by Bent Flyvbjerg, based on Kahneman's outside view concept) involves: (1) selecting a reference class of projects similar to yours; (2) finding the distribution of outcomes (time, cost, benefit) for that class; (3) using the distribution β specifically the 75th percentile outcome β as your base estimate; (4) adjusting up or down for specific features of your project. Research shows this dramatically improves forecast accuracy for infrastructure projects, software, and large organisational initiatives.
Does the planning fallacy apply to individuals as well as organisations?
Yes, consistently. Roger Buehler's research showed individuals planning personal projects (assignments, tax returns, renovations) exhibit the same pattern: focusing on how the specific project will go rather than how similar past projects went. Individual planners underestimate completion time by 30β70% on average. The correction β "how long did similar projects take me in the past?" β works for individuals as well as organisations.
Further Readingβ
- Kahneman, D. & Tversky, A. (1979). "Intuitive Prediction: Biases and Corrective Procedures." TIMS Studies in Management Science
- Flyvbjerg, B. (2008). "Curbing Optimism Bias and Strategic Misrepresentation in Planning." European Planning Studies
- Buehler, R., Griffin, D. & Ross, M. (1994). "Exploring the 'Planning Fallacy.'" Journal of Personality and Social Psychology
Apply with AIβ
π Calibrate your project estimates with MindMax β
This page is part of the MindMax Mental Models Knowledge Base.