The Sunk Cost Trap: Why We Stay Too Long, and Why Age May Help

Checked against primary sources by the Editorial Team. Every figure in this article was traced to the study it came from, including the ones that contradict what we previously published. How we review.

You have probably stayed somewhere longer than you wanted to. A job, a project, a house you kept fixing up, a relationship that stopped working, a membership you never used. The reason you gave yourself sounded sensible: "I have put too much into this to quit now."

Researchers call that reasoning the sunk cost effect. Below we go through what the studies found, what they did not test, and what the age findings mean if you are somewhere between 35 and 65.

What the early studies found

Arkes and Blumer (1985) ran a field study with people who bought season tickets to a university theater. Buyers were randomly assigned to pay the normal price of $15, a $2 discount, or a $7 discount. After removing couples who received different discounts, the authors analyzed 54 people. In the first half of the season, the full price group used more tickets (an average of 4.11) than the two discount groups (3.32 and 3.29). In the second half of the season, the groups did not differ significantly. The authors describe the effect as a greater tendency to continue an endeavor once an investment of money, effort or time has been made.

An earlier version of this article said the people who paid more "attended more shows they disliked." The study does not say that. It measured how many tickets people used, not how much they enjoyed the plays, and the gap showed up only in the first half of the season.

Staw (1976) looked at a related question: why do people put more resources into a course of action that is going badly? In a business investment simulation with 240 business students, participants who were personally responsible for the earlier, poor result committed the most resources to the same course. Staw calls this escalating commitment. For your own life, the lesson is about ownership. The decisions you made yourself are the ones that are hardest to walk away from.

How much does it hold up?

Roth, Robbert and Straus (2015) pooled 98 effect sizes on monetary sunk costs. The authors report clear evidence that the effect emerges, but its size depends on the type of decision. They separate decisions about using something you already paid for from decisions about whether to keep funding a project. They also report that the effect weakens as time passes in decisions about using something you paid for.

Two limits matter for you. First, this literature is about money, and mostly about scenarios researchers wrote. We did not find, in the sources we read, a study that measured sunk cost reasoning in marriages or friendships. When people apply the idea to relationships, they are extending it. It may fit, and it is still an extension. Second, a meta analysis describes averages. It cannot tell you why you in particular stayed.

What the age findings say

This is the part most articles skip, and it matters if you are between 35 and 65.

Strough, Mehta, McFall and Schuller (2008) compared 75 college students (ages 18 to 27) with 73 older adults living in the community (ages 58 to 91). Each person read vignettes in which a prior investment of money or time was present or absent, and chose how much more to invest. The authors report that older adults showed significantly less sunk cost fallacy and made more normatively correct decisions. Controlling for education and cognitive tests did not substantially change the result. The authors point to age differences in positivity and negativity biases in how people process information.

A follow up by Strough, Schlosnagle and DiDonato (2011) asked 41 people (21 older adults and 20 college students) to explain their decision goals. The older adults were less likely to commit the fallacy, and they were less likely to focus on whether a prior investment existed when they described their reasoning. The authors report that this difference in what stood out to people mediated the age gap. The sample is small, so we would not lean on it for anything beyond that mechanism.

The most relevant study for midlife readers is Strough and colleagues (2016). They used a national panel of 397 U.S. adults ages 18 to 85 (average age about 47.5) and a vacation scenario with a failing plan. Older adults were more willing to cancel the failing plan, and they were less likely to interpret the scenario in terms of past losses.

What to be careful about

• The older groups in the first two studies were 58 and up. A 40 year old cannot assume they already have the advantage, and a 63 year old should not assume it either.
• These are vignettes, not records of what people did with their own marriages, careers or savings.
• Age is not the whole story. A 2025 study by Hamzagic and colleagues covered ages 2 to 97 and reports that adolescents and adults consistently made sunk cost judgments while children did not. The authors report a slight decline in some measures in older adulthood, with patterns that varied by task.
• Feldman and Wong (2018) are often cited in this area, but their studies were about how framing a choice as action or inaction changes escalation. They did not study age, so they do not support any claim about older adults.

From findings to a decision

Strough and colleagues (2011) found that older adults focused less on the prior investment. Based on that, one way to act on it is to describe your goals for the next year in writing without mentioning what you have already put in. If the investment keeps creeping into the paragraph, the research suggests that is the pull to watch for.

Strough and colleagues (2016) tested a thoughts and feelings instruction against an improvement focused one and against no instruction. The thoughts and feelings instruction reduced the bias most. Their instruction was to describe the thoughts and feelings you would have in the situation. Based on that, a second way to act is to write out what you expect to think and feel one year from now if you stay, and one year from now if you leave. The authors tested this on a hypothetical vacation, so it is a reasonable thing to try and not a proven fix for a marriage or a mortgage.

Money tied to the decision still deserves a plain look. If a choice involves a large sum, retirement savings or a shared home, a licensed professional can see your whole situation, and a study of vignettes cannot.