Planning Fallacy

Examples

Tomas: “The kitchen renovation? Three weeks, tops.” Ira: “You said that about the bathroom. It took nine.” Tomas: “The bathroom was different — we hit problems.”

Every project “hits problems”. That’s not bad luck breaking a good estimate; it’s what projects are made of, and the estimate ignored it — again.

The pattern scales up brutally. Students asked to predict their thesis timelines miss even their own “worst case” estimates. Infrastructure megaprojects — stadiums, rail lines, software rollouts — run over budget and schedule with such regularity, across countries and decades, that the overrun is the norm, not the exception.

Why it happens

When you plan, you imagine the work happening — step one, step two, done. That mental movie contains no flu, no supplier delay, no discovering that step two was really four steps. Each specific interruption is unlikely, so none makes it into the plan; but some interruption is near-certain, so the plan is wrong on arrival. Meanwhile, the evidence that would fix the estimate — how long such things usually take — feels irrelevant, because “this time is different”.

The scope matters: this is about substantial, multi-step projects. For tiny tasks, people often over-pad. And once a late project is running, the sunk cost fallacy stands ready to keep it running.

How to counter it

• Use the outside view: ignore your movie, and ask how long comparable projects actually took — including yours. That reference is your best single predictor.
• Track actuals. A note of estimate vs. reality for a few months is humbling and permanently useful.
• Ask a non-involved person — the bias is strongest for your own projects; observers estimating your work are notably harsher and more accurate.
• Plan the interruptions generically: you can’t predict which problem, but you can budget for a problem. Multiply first, apologize never.