Denise Cummins Ph.D.

Good Thinking

I Can’t Give Up Now, After Everything I’ve Put Into It

Be smart, don’t fall for the Sunk Cost Fallacy.

Posted Jul 10, 2012

You wait in line for four hours and pay $300 for a ticket to see your favorite music star in a beautiful outdoor concert. But on the day of the event, it turns out to be cold and rainy. If you go, you’ll feel miserable the entire time. But you tell yourself “If I don’t go, then the $300 and four hours I waited in line will have been wasted.” So you go anyway, and sit miserably through the entire concert—plus encores!

 Was this a smart thing to do? Nope. You have just committed the Sunk Cost Fallacy: You made a decision about a current situation based on the amount you have already invested in the situation.

Think of it this way: You had two choices: 1) spend $300 and wait in line four hours to stay comfortably at home, or 2) spend $300 and wait in line for four hours to sit uncomfortably in the rain. Seen this way, it is a no-brainer; the money’s gone, so you might as well be comfortable at home.

If you fell for this psychological lure, don’t feel bad. You’re not alone. Examples of this fallacy can be found in the annals of “dumb decisions made by really smart people”. Iridium World Communications, a provider of satellite-based wireless telephone services, filed for Chapter 11 bankruptcy in 1999 because the company fell prey to the sunk cost fallacy. As reported by the Wall Street Journal: "Having already spent billions of dollars and years of effort on the project, they were loath to abandon it. Instead, they redoubled their efforts and spent more billions." The rest, as they say, was history.

In 1995, Staw and Hoang published a paper showing that the most expensive players in the NBA are given more court time and are kept on the team longer than cheaper players, even when the expensive players are not performing as well as the rest of the team. Having invested a lot of money in these players, coaches feel compelled to continue playing them, even if it means losing!

Why do we behave this way? Short answer: Because we are wired this way. The anguish you feel over losing $100 is about twice as strong as the happiness you feel when you win $100. Daniel Kahneman and Amos Tversky incorporated this “kink” in the value function people use when evaluating decision outcomes. The dropoff in value is steep for losses, but more shallow for gains of the same size.

The additional danger lurking behind this bias is that people tend to make risky choices in order to avoid losses. Suppose I give you $1,000. Then I tell you to choose between receiving another $500 straight out, or flipping a coin and getting another $1,000 if it lands heads up and $0 if it lands tails up. Most people choose to take the $500 and avoid the risk.

But suppose instead, I give you $2,000 and tell you to choose between losing $500 straight out or flipping a coin and losing $1,000 if it lands heads up and losing nothing if it lands tails up? Now most people choose to take the bet. In order to avoid a sure loss, people become risky. But these are exactly the same bet: $1,500 straight up or a 50-50 chance of walking away with $2,000 instead of $1,000. Because we are trying to avoid the anguish we will feel when we incur a loss, we make risky choices. And so we will hold onto bad investments hoping to avoid the sure loss we’ve just incurred rather than selling the duds and possibly buying back in if the data seem to indicate they will actually recover.

It turns out that we are not the only species who act this way. Laurie Santos has found that capuchin monkeys behave exactly the same way: Two researchers offer the monkey grapes, and the monkey has to choose which one to accept. If one researcher is a “safe bet”, that is, always adds another grape while the other sometimes adds another grape and sometimes adds nothing, they will choose the “safe bet”. But if the “safe bet” researcher always takes away one grape while the “risky” researcher sometimes takes away two and sometimes takes away nothing, they will opt for the “risky” guy. Given our shared evolutionary history, Santos estimates that this bias is at least 35 million years old. (Santos TED talk on this topic; move timer to 14:35)

But now that you’ve been informed, hopefully you will be smarter than the capuchins. Don’t become risky when potential losses are involved. Don’t make decisions based on money or effort you’ve already sunk into an endeavor. When the data change, you have to change course as well.

More Posts