What is the Planning Fallacy? Why Everything Takes Longer Than Planned

The planning fallacy is our systematic tendency to underestimate how long tasks will take — even when we know that similar tasks have always taken longer before. The term comes from psychologists Daniel Kahneman and Amos Tversky (1979), and it's one of the most replicated findings in behavioral science: students predicting thesis timelines, engineers estimating projects, governments budgeting infrastructure — all reliably optimistic, rarely by accident twice in the same direction.

Why It Happens

• Inside view. When estimating, we mentally simulate the task going as planned — the simulation contains no sick days, no scope creep, no waiting on other people. Reality always contains them.
• Best-case anchoring. The first number that comes to mind is the best case; adjustments away from it are always too small.
• Motivated optimism. We often estimate what we hope (or what the boss wants to hear), then believe our own number.
• Forgetting the past. Memory smooths over how long things actually took — last quarter's overrun becomes "an exception."

Four Fixes That Actually Work

1. Take the outside view. Ignore this task's details; ask how long similar tasks actually took. Reference-class data beats simulation. If you track your time for even a few weeks, you own that data.
2. Multiply, don't add. Overruns are proportional. A practical rule: your honest estimate × 1.5 for familiar work, × 2 for novel work.
3. Plan the week with buffers. Book only ~70% of available hours; the rest absorbs reality (see schedule buffer time). A weekly planner with 3–5 big rocks forces this honesty better than a bottomless to-do list.
4. Run a weekly estimate audit. In your weekly review, compare planned vs actual for the week's big tasks. Two months of that feedback loop recalibrates you more than any technique.

The planning fallacy never fully disappears — Kahneman himself admitted falling for it while writing the book about it. The goal isn't perfect estimates; it's a system that expects imperfect ones.

Related terms: Schedule Buffer Time · Parkinson's Law · Weekly Review