Why Sunk Costs Make It So Hard to Quit
A couple has spent four years and most of their savings renovating a house that keeps revealing new problems. Every contractor tells them the same thing: it would be cheaper to sell the property as is and buy something else. They keep going anyway. Ask them why, and they don’t talk about the house’s future value. They talk about everything they’ve already put into it.
This is the sunk cost effect, and it shapes decisions far beyond home renovations. It explains why people stay in failing relationships, why companies keep funding products that were never going to work, and why governments continue wars long after the original justification has collapsed. The money, time, or effort already spent is gone no matter what happens next. Economically, it should have no bearing on the decision ahead. Psychologically, it often decides everything.
The puzzle is not that people make this mistake occasionally. It’s that the mistake is so consistent, so widely documented, and so resistant to correction, even among people who understand the logic perfectly well. Understanding why requires looking past the simple explanation of “loss aversion” and into how the mind actually processes commitment, identity, and regret.
The Basic Logic Economics Says We Should Follow
Standard economic theory offers a simple rule for decisions involving past investment: ignore it. A sunk cost is any resource already spent that cannot be recovered regardless of what choice is made next. Because it cannot be recovered either way, it shouldn’t factor into a rational comparison between options.
Consider a business that has spent $2 million developing a product, only to discover a competitor has released something better and cheaper. The rational question is not “how do we justify the $2 million we already spent?” It’s “given where things stand now, is continuing this project the best use of the next dollar?” If the honest answer is no, the $2 million is irrelevant to that decision. It’s already lost whether the company presses on or stops today.
This principle, sometimes called the sunk cost fallacy when violated, has been documented across contexts ranging from consumer purchases to military strategy. People routinely let past investment influence choices about the future, even when they can articulate the rule that says it shouldn’t.
Why the Mind Refuses to Let Go
If the logic is so simple, why is it so hard to follow? Researchers have proposed several interacting explanations, and most of the evidence suggests it isn’t one single mechanism but several reinforcing each other.
Loss Aversion and the Fear of Wasting
One of the most influential explanations comes from behavioral economics research on loss aversion, the finding that losses tend to feel more painful than equivalent gains feel pleasurable. Abandoning a project after investing heavily in it forces a person to register that investment as a loss immediately and completely. Continuing, by contrast, keeps the loss uncertain. There’s always a chance the investment will still pay off, and that sliver of hope can feel more tolerable than a guaranteed loss today.
This creates an asymmetry: quitting produces a certain, immediate cost, while continuing produces only the possibility of an eventual, larger cost. Many people will choose uncertain future pain over certain present pain, even when the math points the other way.
The Desire to Avoid Admitting a Mistake
A second explanation centers on self-justification. Walking away from a heavily invested project often means acknowledging, at least privately, that the original decision was wrong. Continuing lets a person postpone that admission indefinitely. Each additional dollar or hour spent can be framed as “almost there” rather than “further evidence this was a mistake.”
This dynamic becomes especially powerful when the original decision was made publicly. Executives who championed a failing initiative in front of their board, or political leaders who committed troops to a conflict in a televised address, face a cost that has nothing to do with money: the reputational cost of reversing course. Quitting isn’t just financially costly in that moment. It’s an admission that earlier confidence was misplaced.
Mental Accounting and the Search for Closure
A third factor involves how people mentally categorize money and effort. Behavioral economists have described a tendency to place resources into separate mental “accounts,” such as “money spent on this car” or “time invested in this degree.” Once resources are assigned to an account, people often feel a need to see that account “closed out” successfully rather than left as a loss.
This helps explain why sunk cost effects appear even in situations involving no reputational risk and no public scrutiny at all. Someone finishing a boring novel purely because they’re forty pages from the end isn’t protecting their image. They’re trying to close an account they’ve already opened.
What the Research Actually Shows
The sunk cost effect has been studied extensively since the late 1970s, when economists and psychologists began documenting it experimentally. Early studies presented people with hypothetical scenarios: imagine you’ve paid for a ski trip and a separate, better ski trip, then discover the two trips are on the same weekend and only one ticket can be used. Which do you choose? A striking number of participants chose the trip they’d paid more for, even when they rated it as the less enjoyable option.
Subsequent research extended these findings well beyond hypothetical vacations. Studies of NBA draft picks have found that teams tend to give more playing time to players drafted earlier in the draft, independent of how those players actually perform once on the court, suggesting that the size of the initial investment shapes later decisions long after performance data should have taken over. Laboratory experiments involving real money, not just hypothetical choices, have shown similar patterns, indicating the effect isn’t simply an artifact of imagining costless scenarios.
At the same time, researchers have found the effect is not universal or automatic. It weakens under certain conditions: when the decision is made by someone other than the person who made the original investment, when the sunk cost is made highly explicit and separated from the decision at hand, or when people are specifically trained to recognize the pattern. This variability is itself informative. If the effect were a fixed feature of how the brain processes numbers, training and framing shouldn’t change it. Instead, the evidence suggests it’s a psychological habit, powerful but not fixed, shaped by attention and self-awareness as much as by underlying cognitive wiring.
A Genuine Debate: Is It Always Irrational?
Not every researcher accepts that continuing after a sunk cost is automatically a mistake. Some economists have pointed out that past investment can carry legitimate information, even if the cost itself is sunk. If a company has spent two years developing a product, that history may reveal something real about the team’s capability, the technical difficulty of the remaining work, or the market’s likely reception, information that a decision-maker starting from scratch wouldn’t have.
There’s also an argument from reputation and incentives that isn’t purely psychological. In organizations, sticking with a costly commitment can sometimes preserve trust with partners, investors, or employees, even if abandoning it would be more efficient in a narrow financial sense. Quitting projects at the first sign of trouble can create a culture where nobody wants to commit to anything ambitious in the first place, since ambitious things almost always look shaky partway through.
This is a genuinely useful complication. The sunk cost fallacy describes letting an unrecoverable cost distort a decision that should rest entirely on future consequences. It does not mean that persistence is always foolish or that quitting is always wise. The mistake lies specifically in treating the past expense itself as a reason to continue, rather than treating it as one data point among many about what the future is actually likely to hold.
What People Commonly Get Wrong
The most common misunderstanding is treating the sunk cost fallacy as a simple failure of willpower, something that only affects people who are bad at math or bad at making decisions. The evidence points the other way. The effect appears reliably among trained economists, experienced executives, and professional gamblers, groups whose expertise should, in theory, make them immune. Understanding the rule intellectually does very little to prevent falling for it in the moment, because the pull comes from emotional accounting, not a lack of arithmetic.
A second misconception is assuming the fallacy is mainly about money. The evidence suggests time, effort, and emotional investment trigger the same pattern, sometimes even more strongly than cash. People report staying in unsatisfying careers, unfinished dissertations, and long relationships specifically because of how much of themselves they’ve already put in, independent of whether continuing serves their interests going forward.
A third misconception is that recognizing the fallacy in others is the same as recognizing it in yourself. Studies on this kind of bias blindness have consistently found that people are far quicker to spot sunk cost reasoning in someone else’s decision than in their own, which is part of why the effect persists even in highly analytical fields like finance and engineering.
Why It Still Matters
The sunk cost effect matters because the stakes of the decisions it distorts are rarely trivial. It shapes whether companies discontinue products that have already burned through their budgets, whether governments extend military or infrastructure commitments that have stopped serving their original purpose, and whether individuals leave relationships, careers, or projects that no longer make sense for them.
The most useful response isn’t to demand that people become emotionless calculators. It’s to build in structures that separate the decision from the history behind it: asking “knowing what we know now, would we start this today?” rather than “how do we justify what we’ve already spent?” That single reframe, simple as it sounds, is one of the few interventions that has consistently reduced the effect in controlled studies.
The house renovation, the failing product, the unfinished degree: none of these situations are irrational to have started. What makes them costly is treating the decision to continue as though the past investment were still on the table, when in fact it disappeared the moment it was spent. The only real question left is what the next dollar, the next year, or the next commitment is actually worth, considered entirely on its own.
