Sunk Cost Fallacy in Personal Finance Decisions: A Deep Dive
A $2,400 car repair on a car worth $3,000 feels like a fair price if you already spent $1,800 on the last repair. It isn’t, and that gap between how it feels and what the math says is the core of the sunk cost fallacy in personal finance decisions. This guide covers what the fallacy is, where it shows up in money choices (cars, investments, subscriptions, degrees, houses), and a simple decision test you can run to stop paying for yesterday’s mistakes.
What the Sunk Cost Fallacy Actually Is
A sunk cost is money, time, or effort you have already spent and cannot get back. Rational decision-making says those costs should be ignored: only future costs and future benefits should matter. The fallacy is letting the past spend steer the next choice anyway.
The classic research is a 1985 study by Hal Arkes and Catherine Blumer in Organizational Behavior and Human Decision Processes. Theater-season ticket buyers were randomly given discounts of different sizes. The people who paid full price attended significantly more plays in the first half of the season than those who got discounts, even though their enjoyment of the shows should have been identical. Paying more made them feel obligated to show up. Barry Staw described a related pattern in 1976, which he called escalation of commitment: decision-makers who had already invested in a course of action tended to put in more resources after bad news, not fewer.
The behavior is not a sign of low intelligence. It is closely tied to loss aversion, the tendency to feel losses more sharply than equal gains. Quitting makes the loss real, so we keep going to avoid “realizing” it. If that sounds familiar, our piece on loss aversion and why reframing your budget often fails covers the underlying mechanism.
Where Sunk Costs Quietly Drain Your Money
The fallacy hides in ordinary decisions. Here are the places it does the most damage, and the question that cuts through each one.
| Situation | What sunk cost whispers | The forward-looking question |
|---|---|---|
| Old car needing repairs | “I just put $1,800 into it.” | What is the cheapest reliable way to get transportation over the next 3 years? |
| Losing stock or fund | “I’ll sell when it gets back to even.” | Would I buy this today at this price with this cash? |
| Unused gym or app membership | “I already paid for the year.” | Will I use it enough next month to justify the next payment? |
| Degree or career path | “I’ve already done three years.” | What do the next years cost vs. what they return? |
| Home renovation gone over budget | “We can’t stop now.” | Does finishing cost less than the value it adds? |
| Side business that never earns | “I’ve put in so much.” | If I started today with what I know, would I do this? |
Sunk Cost in Investing: The “Get Back to Even” Trap
Investing is where this bias gets expensive, because the market has no idea what you paid. A stock trading at $60 that you bought at $100 has the same future prospects for you as for a stranger who just bought it at $60. Refusing to sell because you would “lock in” a loss confuses the accounting record with the economic reality: the loss already happened when the price fell.
There is also a tax angle most people miss. Under IRS rules (Topic 409, Capital Gains and Losses), realized capital losses can offset capital gains, and if losses exceed gains you can deduct up to $3,000 of the excess against ordinary income per year ($1,500 if married filing separately), carrying the rest forward. In other words, selling a loser is sometimes worth money on its own. Watch the wash-sale rule, which disallows the loss if you buy a substantially identical security within 30 days before or after the sale. We walk through the numbers in whether tax-loss harvesting is worth it for small portfolios.
A useful reframe: if you had cash instead of this position today, would you buy it? If the honest answer is no, holding is effectively a decision to buy it again. This is also why the endowment effect often travels with sunk cost thinking: we overvalue what we already own and what we already paid for.
Subscriptions, Memberships and the “I Already Paid” Problem
Research on gym memberships shows how sticky this is. In a widely cited 2006 study, “Paying Not to Go,” Stefano DellaVigna and Ulrike Malmendier found that members on monthly contracts paid an average of roughly $17 per visit, when a 10-visit pass at about $10 per visit was available to them. Many people kept paying for the contract they were not using, hoping to justify it later. The fee was gone either way. Only the next payment was still a choice.
Apply the same logic to streaming services, software, meal kits, and course subscriptions. Cancel based on expected future use, not on how much you have paid so far. If you want a systematic way to do this, build a recurring-expense review into your budget; our budgeting guide is a good starting point.
Big-Ticket Decisions: Cars, Renovations and Degrees
The larger the sunk cost, the stronger the pull, which is why the biggest purchases produce the worst examples. With a car, compare the repair bill plus remaining risk against the monthly cost of a replacement, not against what you have already spent keeping the old one alive. A useful rule of thumb many mechanics and consumer guides cite is to think twice when a single repair approaches half of the vehicle’s value, but the real test is the total cost of ownership over your planned holding period.
With renovations, get a fresh contractor quote for the remaining work and compare it with the value the finished project adds. With a degree or certification, compare the remaining tuition and lost income against the expected salary gain. Sometimes finishing wins by a wide margin. Sometimes the best move is a pivot. In both cases, the decision is cleaner once you refuse to count what is already spent. Setting an explicit spending ceiling before you start, and treating it as a hard stop, is the cheapest protection against escalation of commitment.
A Personal Note on Testing This Myself
As a software engineer, I run into the same bias in code: teams keep a failing architecture alive because “we’ve put a year into it.” I started applying the same forward-looking test to my own finances a few years back, mostly out of curiosity about behavioral economics, and I manage my own accounts in index funds and tax-advantaged accounts without an advisor. The honest result: the test rarely tells me to make dramatic moves. Most of the time it just stops me from doing something expensive out of stubbornness, like holding a single stock purely because I hated admitting the mistake. I now keep a short written note for each “should I quit this” decision, and the act of writing the forward costs down is usually enough to expose the bias.
A Five-Question Decision Test to Ignore Sunk Costs
You cannot switch off the bias, but you can build a routine that works around it. Use these five steps whenever you catch yourself saying “but I’ve already put so much into this.”
- Write down the sunk amount, then set it aside. Naming it in writing reduces its pull. Then label it “gone regardless.”
- List only future costs. Money, time, and opportunity cost over the next 6–36 months.
- List only future benefits. Be realistic and use ranges, not best-case guesses.
- Run the fresh-start test. If you were starting from zero today, with no history, would you choose this option?
- Set a kill criterion in advance. For example, “If this side project earns under $200 a month by March, I stop.” Pre-commitment removes the emotion later.
Notice that this test does not tell you to quit. Sometimes the future benefits genuinely outweigh the future costs, and continuing is correct. The point is that the reason to continue must come from the future, not the past.
Not sure which recurring expenses are worth keeping? See where your money goes each month.
When Continuing Is the Right Call
Avoiding the sunk cost fallacy does not mean abandoning everything at the first setback. Two situations are worth separating from the bias. First, when past investment has created real future value: a half-built deck that is worth more finished, or a professional license that is one exam away. The past spend is not the reason; the remaining cost is small relative to the remaining payoff. Second, when a commitment protects you from your own present bias, as with automated retirement contributions. If you tend to under-save, a contribution you would be tempted to skip is a good thing to keep, and the same forward-looking test should confirm that. See our take on status quo bias in financial decisions for the flip side, where inertia (not sunk cost) is what keeps money stuck.
Finally, watch for sunk-cost thinking in how you treat windfalls too. Our article on why we treat bonus money differently shows how mental labels can push you to spend or hold money based on where it came from rather than what it could do next.
Key Takeaways
- Sunk costs are gone either way. Only future costs and benefits should influence your next money decision.
- Classic evidence is robust. From theater tickets to gym contracts, people act to justify past spending rather than to maximize future value.
- In investing, ask “would I buy this today?” If not, holding is a purchase decision. Also check the tax benefit of realized losses (up to $3,000 against ordinary income per year, per IRS rules) and the 30-day wash-sale rule.
- Use a written fresh-start test and a pre-set kill criterion to take emotion out of quit-or-continue choices.
- Continuing can still be right, as long as the reason is the future payoff and not the past price.
Photo by Jens Lelie on
Unsplash
