The Sunk Cost Fallacy: Why “We’ve Already Paid for It” Is Bad Economics

 

One of my favorite ways to start the semester is by introducing a concept many students are at least tangentially familiar with: opportunity cost.

Closely tied to opportunity cost, and to a core economic assumption about marginal decision-making to maximize well-being, is the sunk cost fallacy. While this fallacy shows up frequently in large public and private projects, I find it most effective to bring it down to a scale we’re all familiar with (and probably guilty of): sitting through a bad movie simply because “we’ve already paid for it.”

Before working through an example, one quick definition. In this post, I’ll refer to economic rent, which, following the CORE Microeconomics framework, means the extra benefit you receive from an action after accounting for all relevant economic costs, including opportunity costs. A positive economic rent means the choice improves your well-being relative to the next-best alternative.

The Initial Choice

Suppose you have two groups of friends asking you to hang out on the same evening:

• One group wants to go to the movie theatre
• The other wants to go to the pub

You can only attend one.

To keep things simple, let’s assume:

• Your time is the scarce resource, not money (Lucky you)
• You enjoy both groups of friends
• The movie starts earlier, and the pub hangout begins later

Economists often represent enjoyment or satisfaction in dollar terms, not because it’s literally money, but because it gives us a common unit to compare options.

Option A: The Movie Theatre

You’ve heard the movie is good, so you expect:

• Benefit: $50 (movie + time with friends)
• Explicit cost: $35 (ticket, popcorn, etc.)

That leaves you with a net explicit benefit of:

$50 − $35 = $15

Option B: The Pub

You know from experience the pub is always a good time:

• Benefit: $60
• Explicit cost: $50 (food, drinks, tip)

So, your net explicit benefit here is:

$60 − $50 = $10

Choosing Using Economic Rent

Using economic rent as the decision rule:

If you choose the movie:

• Net explicit benefit = $15
• Opportunity cost = forgone $10 from the pub
• Economic rent = +$5

If you choose the pub:

• Net explicit benefit = $10
• Opportunity cost = forgone $15 from the movie
• Economic rent = −$5

To maximize well-being, you choose the option with positive economic rent, in this case, the movie.

When Expectations Turn Out to Be Wrong

Now suppose you get into the theatre and quickly realize the movie is terrible.

Your original expectation of a $50 benefit was wrong. Going forward, the only enjoyment you expect to receive is about $5, purely from being with friends, not from the movie itself.

At this point, you face a new decision:

• Do you stay and finish the movie because you already paid for it?
• Or do you leave and head to the pub, which is just getting started?

This is where many of us fall into the sunk cost fallacy.

Why the Sunk Cost Shouldn’t Matter

The $35 you paid for the movie is sunk:

• It’s already been paid
• It’s irrecoverable
• It’s in the past

Because of that, it should not influence your decision going forward.

What matters now are the costs and benefits from this point onward.

The Decision Going Forward

Option A: Stay at the Theatre

• Benefit: $5 (time with friends)
• Explicit cost: $0 (the ticket is already paid)

Net explicit benefit = $5

The opportunity cost of staying is the forgone pub experience worth $10.

Economic rent = $5 − $10 = −$5

Option B: Go to the Pub

• Benefit: $60
• Explicit cost: $50

Net explicit benefit = $10

The opportunity cost is the $5 you would have received by staying.

Economic rent = $10 − $5 = +$5

The Rational Choice

The economically rational choice is to leave the movie and go to the pub.

Importantly, notice what doesn’t enter the calculation: The $35 already spent as that cost is sunk. It cannot be recovered, and including it in your decision would only lead you to make a worse choice going forward.

A Hiking Analogy

I find this analogy really helps drive the point home.

Suppose you’re on a hike and have already walked 5,000 steps. Ahead of you is a river you planned to cross, but when you arrive, the river is raging and the bridge is washed out. Crossing now would risk serious injury or death.

Should you continue just because you’ve already spent 5,000 steps getting there?

Of course not.

Those steps are sunk. The only relevant question is:

• Do you take a dangerous risk?
• Or do you turn back and choose a safer alternative for the day?

When the stakes are life and limb, the correct choice is obvious. The sunk cost fallacy feels tempting mainly when the costs are monetary or effort-based, but the logic is exactly the same.

Final Thought

What are your thoughts on the sunk cost fallacy?
Have you caught yourself thinking “I might as well finish, I’ve already paid for it or invested so much into it”?
Have you seen examples where sunk costs led to clearly worse outcomes?

Let me know in the comments.