The Planning Fallacy in Personal Finance: Why a $30,000 Kitchen Closes at $41,000

Thirty-seven psychology students were asked how long their senior thesis would take. Their average estimate was 33.9 days. The average actual completion time was 55.5 days — and only about 30% of them finished by the date they themselves had predicted. That 64% miss is the cleanest experimental demonstration of the planning fallacy ever run, and the same arithmetic is quietly wrecking household budgets right now.

This post walks through the planning fallacy in personal finance as a case study: one household project priced two different ways, the research on how far off our cost estimates actually run, and a five-step method for producing a number you can defend. You will finish with a repeatable way to price any project — a renovation, a wedding, a move, a career gap — that does not depend on you being an unusually honest estimator.

The case: a $30,000 kitchen that closed at $41,000

Take a household planning a kitchen remodel. They get three contractor quotes, land on $30,000, and add what feels like a generous cushion — $2,000, call it 7% — because they are sensible people who know things come up. Total planned outlay: $32,000. They have $34,000 in a dedicated account, so they feel comfortable.

The project closes at $41,000. Nothing catastrophic happened. The subfloor under the old dishwasher had rot, which added a few thousand. A backordered range pushed the timeline three weeks, which added two more weeks of eating out. The countertop fabricator’s measurement came back over the slab allowance. The electrician found aluminum wiring behind one wall. Each item was individually reasonable and individually unforeseeable. Collectively, they were entirely predictable.

The $9,000 gap does not come out of nowhere. It comes off a credit card at whatever the going rate is, or out of the emergency fund, or out of the retirement contribution for the next eight months. And the household concludes it was unlucky, which guarantees the same thing happens on the next project.

Why the planning fallacy in personal finance hits harder than it does in project management

Daniel Kahneman and Amos Tversky named the planning fallacy in 1979: the tendency to underestimate the time and cost of a task even when you know that similar tasks have historically run long. The critical word is even. This is not an information problem. It is a framing problem.

Kahneman’s own favorite illustration involved a textbook project he was part of. The team estimated a year and a half to two and a half years. A curriculum expert on that same team, when pressed, said roughly 40% of comparable projects were never finished at all, and he could not recall one that took less than eight years. The team heard this, absorbed nothing, and pressed on. The book took eight years and was never used.

Two things make households more exposed than institutions here:

You have no portfolio to average across. A construction firm runs forty projects a year; the overruns and the underruns partly cancel. You renovate a kitchen once a decade. Your single draw from a right-skewed distribution has no smoothing mechanism, and the distribution of cost outcomes is emphatically right-skewed — costs can run 100% over, but they essentially never come in 100% under.

Your buffer is your safety net. A firm covers an overrun from a credit line. Most households cover it from the same pool of money that exists to handle a job loss or a medical bill. The Federal Reserve’s 2025 Survey of Household Economics and Decisionmaking found that 63% of adults could cover a $400 emergency expense with cash or its equivalent — a figure that has been flat for four years and sits below the 2021 high of 68%. A $9,000 planning miss does not just cost $9,000. It converts a funded emergency fund into an unfunded one, and it does so at exactly the moment you are least able to notice.

What the research says about how far off estimates run

The most useful data comes from infrastructure, because someone tracks it. Bent Flyvbjerg’s study with Holm and Buhl examined 258 transportation projects across 20 countries and found that nine out of ten ran over budget. Rail projects averaged a 44.7% overrun. Roads averaged 20.4%. The overall average was 28%. Critically, the overrun rate showed no improvement across the seventy years of projects in the sample — better software and better methodology did not fix it, because the bias is not in the tools.

Consumer-side numbers are messier but point the same direction. Houzz’s 2025 renovation research found 37% of homeowners exceeded their set budget, slightly more than the 35% who came in on target. The Knot’s study of more than 10,000 U.S. couples married in 2025 put the average wedding at $34,000 and found that nearly half of couples discovered their initial budget fell short of actual costs.

The right-hand column is the whole point. You do not need a better estimate. You need a multiplier applied to the estimate you already have.

Inside view versus outside view on the same kitchen

Kahneman’s fix is called the outside view, or reference class forecasting. Instead of building a forecast from the details of your specific plan, you ask what happened to the last hundred projects that looked like yours, and you anchor there.

Run the same kitchen both ways:

The outside view did not predict the rotted subfloor. It did not need to. It predicted that something in that category would appear, because something always does, and it sized the reserve accordingly. Note also the second-order effect: at $39,000 planned, this household might have decided the project was not worth it. That is not a failure of the method. That is the method working.

Five steps to price a project against the planning fallacy

1. Write down your estimate first, and date it. Before you do anything else, record the number your gut produced and the completion date you have in mind. You will need it later, and you will not remember it accurately if you do not write it down — hindsight quietly rewrites what you originally thought.

2. Define the reference class, then find three real outcomes. Not “kitchen remodels.” Something like “full kitchen remodels in a house built before 1980, mid-range finishes, contractor-managed.” Then find three people or three documented projects that fit it and ask what the final invoice was, not the quote. Three data points beat zero by an enormous margin.

3. Take the median of the reference class, not your quote. If the three actuals came in at 1.2×, 1.35× and 1.6× the original quote, your planning number is 1.35×. Do not average in your own optimism. Do not adjust downward because your contractor seems unusually organized — every household believes that about their contractor.

4. Fund the reserve before the project starts, in a separate account. A reserve you intend to fund is not a reserve. Park the full 1.3× amount somewhere before the first invoice lands. This is exactly the mechanism behind our five-bucket sinking fund system — separate accounts stop project overruns from silently borrowing against your emergency savings.

5. Set a kill point in advance. Decide now, in writing, what number makes you stop or descope. “If we cross $44,000 we skip the island.” Written before the work begins, that is a rational limit. Decided halfway through, when you have already paid $28,000, it becomes a fight with the sunk cost fallacy that you will probably lose.

Where the planning fallacy in personal finance shows up beyond renovations

Renovations are the obvious case. The subtler ones cost more over a lifetime.

Debt payoff timelines. “We’ll have the card cleared in eight months” assumes eight months with no car repair, no vet bill, no dental work. Price the payoff at your realistic surplus, not your best-month surplus.

Retirement contribution plans. “I’ll raise my rate to 12% after the promotion” is a forecast about a future self, and it fails for the same structural reason — the plan lives in an imagined frictionless month. This overlaps heavily with present bias in retirement contributions: the automatic escalation you set up today works precisely because it does not require the future you to act.

Side income ramps. Time-to-first-revenue on a new venture is a textbook Buehler-style estimate, and it is routinely off by more than 64%. Budget the household as if the income arrives twelve months after you think it will.

Active investing plans. “I’ll rebalance quarterly and review each holding” is a time estimate dressed as a strategy, and it interacts badly with overconfidence in stock picking — you overestimate both the hours you will put in and the edge those hours will buy.

Moving costs. The truck is the quoted number. Deposits, overlap rent, utility connection fees, replacement furniture and two weeks of takeout are the actual number. Apply the multiplier.

What happened when I ran the multiplier on my own numbers

I started applying a flat 1.3× to every household project a few years ago, mostly because the engineering version of this problem is so familiar — software estimates are famously optimistic, and every team I have worked on eventually adopts some version of “take the estimate and multiply it.” I was curious whether the same crude correction held up in personal finance, where I do everything myself without an advisor and there is nobody to catch the error.

It held up better than I expected, with one caveat. The 1.3× was roughly right in aggregate, but it was too high for small, well-defined purchases and much too low for anything involving other people’s schedules. The version I use now is closer to 1.15× for a single-vendor purchase and 1.4× for anything with three or more parties involved. The genuinely useful part was not the accuracy. It was that pre-funding the reserve meant I stopped raiding the emergency fund for things that were not emergencies — which is a behavioral fix, not a forecasting one.

The related habit worth building alongside it: recording the original estimate. Once you have five or six projects logged with estimate-versus-actual, you have your own reference class, and it beats any generic multiplier. It is the same discipline that makes anchoring bias when buying a house manageable — you cannot beat a bias by resolving to be less biased, only by installing a number that was set before the bias had a chance to operate.

Photo by immo RENOVATION on
Unsplash

Written by Chris Steve

Chris Steve is a software engineer with a deep interest in personal finance, behavioral economics, and AI. He started Money & Planet to share clear, research-backed money guides — the kind that explain the math instead of pushing products. His writing focuses on long-term wealth building, the psychology behind spending and investing decisions, and the practical tools regular people can use to make smarter financial choices.