Sunk Cost Fallacy
Helios - we once debated whether it was a sunk cost fallacy for a college student to say something along the lines of, "I'll finish college, after all, I've already spent 3 years here."
I had been thinking about that. I still stand my by claim that this is not a sunk cost fallacy, my argument:
A sunk cost fallacy is a fallacy because you cannot justify further losses with previous losses that you cannot recoup. You can't say, "I'll gamble another $10, after all, I already gambled and lost $100." Because the $100 you spent is already gone. Any future action you take because of that $100 loss is illogical. There's no connection between that $100 and you anymore.
The reason the college statement is different is because you are NOT justifying future loses (one more year in college) with previous losses (spent 3 years in college already). Instead, you're recognizing that the degree is worth all 4 years. The previous 3 years have gotten you to the point where you are just one more year away.
Those 3 years are a "sunk cost" in terms of that you'll never get them back. However, you are still going to recoup that loss - and they helped you get there.
You can imagine looking back at it like this, "I spent 4 years in college to get a degree, the degree made me X money, it was worth it." You look at the totality to see if the degree was worth it.
The gambler might say, "I gambled $2500 over the past year and won $50,000 today!". It sounds similar, but losing the $2500 was never a condition for winning the $50,000. Only the money they gambled right before winning the $50,000 actually directly impacted them winning that money.
With college, those 3 years were necessary, and now you need one more. The degree will more than recoup you for all 4 years spent, so its not a sunk cost fallacy to say, "I invested 3 years, I need to invest 1 more year, then the payoff will make up for all my investments".
Because you're not justifying further losses with previous losses, you're justifying further gains because they will offset all your losses.
Now, you can say this is unnecessary - because you only have to analyze that 1 remaining year to conclude that investing 1 year and getting a college degree is worth it, therefore you didn't have to include the 3 years spent into the equation.
That may be logical in terms of what decision to make right now, but its not logical when analyzing your investment and reward. You wouldn't look back on that and say, "I spent 1 year and got a college degree!", you'd say you spent 4 years.
A big point: Let's say the logic does say 1 more year of college is well worth it to get a college degree, therefore, the sunk cost of 3 previous years doesn't matter.
Sure. But let's say you think, "That one year will pay off big time to get that degree, but I'm kinda lazy, let's not." That's perfectly reasonable logic if you don't account for the sunk cost. Because saying 1 year = college degree separates you from the losses you have already inccurred. It isolates you from the investments you already made to get the degree, saying that it'll only cost you 1 year of your time.
However, the reality is that you DID incur loses. You're not justifying future losses with previous losses, your justifying further investment to recoup ALL of your losses.
This is why I think its fundamentally different to say, "I've already spent 3 years to get this degree, I may as well finish" (assuming you're going to benefit from this degree and you're saying this because you're tired of the effort or something similar) as compared to saying, "I gambled away $1000 so I have to keep gambling."
I had been thinking about that. I still stand my by claim that this is not a sunk cost fallacy, my argument:
A sunk cost fallacy is a fallacy because you cannot justify further losses with previous losses that you cannot recoup. You can't say, "I'll gamble another $10, after all, I already gambled and lost $100." Because the $100 you spent is already gone. Any future action you take because of that $100 loss is illogical. There's no connection between that $100 and you anymore.
The reason the college statement is different is because you are NOT justifying future loses (one more year in college) with previous losses (spent 3 years in college already). Instead, you're recognizing that the degree is worth all 4 years. The previous 3 years have gotten you to the point where you are just one more year away.
Those 3 years are a "sunk cost" in terms of that you'll never get them back. However, you are still going to recoup that loss - and they helped you get there.
You can imagine looking back at it like this, "I spent 4 years in college to get a degree, the degree made me X money, it was worth it." You look at the totality to see if the degree was worth it.
The gambler might say, "I gambled $2500 over the past year and won $50,000 today!". It sounds similar, but losing the $2500 was never a condition for winning the $50,000. Only the money they gambled right before winning the $50,000 actually directly impacted them winning that money.
With college, those 3 years were necessary, and now you need one more. The degree will more than recoup you for all 4 years spent, so its not a sunk cost fallacy to say, "I invested 3 years, I need to invest 1 more year, then the payoff will make up for all my investments".
Because you're not justifying further losses with previous losses, you're justifying further gains because they will offset all your losses.
Now, you can say this is unnecessary - because you only have to analyze that 1 remaining year to conclude that investing 1 year and getting a college degree is worth it, therefore you didn't have to include the 3 years spent into the equation.
That may be logical in terms of what decision to make right now, but its not logical when analyzing your investment and reward. You wouldn't look back on that and say, "I spent 1 year and got a college degree!", you'd say you spent 4 years.
A big point: Let's say the logic does say 1 more year of college is well worth it to get a college degree, therefore, the sunk cost of 3 previous years doesn't matter.
Sure. But let's say you think, "That one year will pay off big time to get that degree, but I'm kinda lazy, let's not." That's perfectly reasonable logic if you don't account for the sunk cost. Because saying 1 year = college degree separates you from the losses you have already inccurred. It isolates you from the investments you already made to get the degree, saying that it'll only cost you 1 year of your time.
However, the reality is that you DID incur loses. You're not justifying future losses with previous losses, your justifying further investment to recoup ALL of your losses.
This is why I think its fundamentally different to say, "I've already spent 3 years to get this degree, I may as well finish" (assuming you're going to benefit from this degree and you're saying this because you're tired of the effort or something similar) as compared to saying, "I gambled away $1000 so I have to keep gambling."
I agree, but did you consider whether it was a sunk cost fallacy to keep writing past the 11th paragraph? :)
But I also think I would be susceptible to other forms of sunk cost fallacy that aren't as obvious. The gambling example is too trivial, because you know the probabilities are independent. Other situations are more complex.
Oh, and even gambling, I can think of situation where you'd keep gambling. Perhaps it's a silly scenario, but let's say you're $100'000 in the hole, and the mafia will "take care" of you if you don't pay them back. If you do nothing, you're dead. So you have no choice but to take a loan and potentially dig yourself a bigger hole -- either you lose and you now owe $200'000 before dying, or you win and you live. Maybe that's apples and oranges to your original example. Hence why I think I would totally be susceptible to sunk cost fallacy ways of thinking.
But I also think I would be susceptible to other forms of sunk cost fallacy that aren't as obvious. The gambling example is too trivial, because you know the probabilities are independent. Other situations are more complex.
Oh, and even gambling, I can think of situation where you'd keep gambling. Perhaps it's a silly scenario, but let's say you're $100'000 in the hole, and the mafia will "take care" of you if you don't pay them back. If you do nothing, you're dead. So you have no choice but to take a loan and potentially dig yourself a bigger hole -- either you lose and you now owe $200'000 before dying, or you win and you live. Maybe that's apples and oranges to your original example. Hence why I think I would totally be susceptible to sunk cost fallacy ways of thinking.
Last edited on
| I agree, but did you consider whether it was a sunk cost fallacy to keep writing past the 11th paragraph? :) |
No, I should actually type more.
| So you have no choice but to take a loan and potentially dig yourself a bigger hole -- either you lose and you now owe $200'000 before dying, or you win and you live |
Well that's not really sunk cost fallacy. Your only chance at life is winning, so the better option (no matter any previous sunk cost) is to keep gambling.
It's only a sunk cost fallacy if your life is already forfeit and you kept gambling with no possible hope of regaining what you've lost. In that case, better spend that money making your last time alive comfortable/happy.
From a purely economic (money in money out) perspective, any sunk cost taken into account is a sunk cost fallacy. My argument is that when you have other factors that are not purely economical, then it can be logical to use sunk costs without it being fallacious.
I gave the example of the the college degree, but here's another:
Imagine a business man plans to invest $1 million, and expects to get $10 million dollars back. He invests this million over the span of 10 years, so $100k every year.
At the last year, this man need only invest the last $100k, then await his $10M return. From a pure economic perspective, the previous $900k doesn't matter, the question is, "Will this $100k that I'm going to invest RIGHT NOW be worth it?"
But here's the issue, let's introduce another variable - you hate the business owner. In the past couple years, you've started hating the guy. At this point, you actually don't want to keep investing, because he's such a prick.
Let's analyze the situation again:
Without Sunk Cost: $100K to prick -> $10M profit for me, but I kinda don't want to give this guy $100k. If I don't give him the $100k, then I miss out on $9.9M of net profit but don't lose anything.
With Sunk Cost: $100K to prick -> $10M profit for me. I don't want to keep giving this prick money, but I've already given him $900k. So if I don't give him the $100k, then I lost $900k for nothing - I miss out on $9M net profit and I lost $900k overall.
These situations are VASTLY different. The analysis without sunk cost, we don't account for the fact that we've lost money for no reason if we don't fork over the last $100k.
Now someone may argue that its still a fallacy, because you've changed your mind over a sunk cost - you should spend that $100k according to how it'll best serve you in that moment.
Sure, but notice that the total cost for not spending that $100k is $900k. That's not the loss you were willing to accept when you rejected to give the $100k in the non-sunk cost fallacy world, you were only willing to lose a potential profit.
As soon as you step outside the realm of pure economics, sunk costs aren't guaranteed to be fallacious.
Last edited on
Registered users can post here. Sign in or register to post.