Got my van checked out today from a check engine light. Long story short, part of the transmission has gone bad, and given the lower quality of this particular transmission, rather than diagnose the specific root cause issue, it's better to replace/overhaul the whole transmission.

We're still waiting on a quote, but I know from previous research that this repair would've cost $3000 two and a half years ago; I bet it's likely $4000-$5000 these days.

We bought the van two-and-a-quarter years ago for $6700, it currently has about 115,000 miles on it. Kelley Blue Book private party value for "fair" condition is $5700, "good" condition $6400. I bet these are actually low values; on Craigslist right now, for example, I see a same make, model, and year van at 103,000 miles for $13,000. Though I also see one two model years newer at 130,000 miles for $8000.

My spouse thinks that since the repair cost is likely to be most of the current value of our van, that's an argument against undertaking the repair and instead applying that money to the purchase of a different car. To me, that sounds like some form of a sunk-cost fallacy, but I can't quite tell, can someone help me understand on this?

My thoughts are that, ignoring all previous costs, we basically have a choice between "buying" our current van (with a new transmission) for the repair cost, or buying a different vehicle for the price of the repair + whatever we can get out of selling our unrepaired van. Is this a rational approach to this decision? If not, where is my logic going wrong?

(There's also a third choice of spending even more money and buying a more expensive vehicle, but I list the two above as equivalent choices since in terms of dollars spent they are equal.)

Update: got the repair quote, with tax it will be about $4400.

  • 1
    “Sell and buy a new one” seems the opposite of the sunk cost fallacy. Given the shortage of used cars, it might be best to get that one repaired (even though that runs risks, too).
    – RonJohn
    Commented Apr 13, 2022 at 11:49
  • 2
    Yeah, it's not a sunk cost fallacy. That would apply if you wanted to keep the current car because you recently spent money fixing something else. Or if you paid a fee to test drive a bunch of cars so you feel you need to buy one of those cars or the fee is wasted.
    – Eric Nolan
    Commented Apr 13, 2022 at 12:47
  • 4
    Arguing that you should never spend a lot of money on something that isn't worth much money is fallacious but it's not a sunk cost fallacy. For instance my shoes have a resale value of approximately 0. They need to be resoled for $50. Paying that is not necessarily right or wrong. If I could replace them for $50 or less then it's likely sub-optimal to pay for the repair. If it would cost me $100 to buy new ones then the repair may be the best thing to do. Other factors should be taken in to account. Perhaps the repair won't last as long as new shoes, perhaps I don't want to break in new shoes.
    – Eric Nolan
    Commented Apr 13, 2022 at 12:52
  • @EricNolan, is there a name for that type of fallacy then? I should've included that question in my post. Commented Apr 13, 2022 at 14:00
  • 1
    Here's the thing. A fallacy is a error in a formal logical argument. I can believe what my doctor says about my cough and ignore what my barman says and the "Appeal to Authority" fallacy has nothing to do with it unless I write a paper claiming to prove that. Not every poorly thought out decision is based on a fallacy. Not every decision that is based on logic that would be fallacious in a formal argument is bad. People tend to value things they already have more than the replacement cost. Look up the Endowment Effect.
    – Eric Nolan
    Commented Apr 14, 2022 at 9:34

5 Answers 5


Neither of them use the sunk cost fallacy.

Your spouse is making an implicit argument that there's little chance that a repair that's near the cost of the same used van is worth it. That's more of a natural intuition based approach, ie when the van was bought, we weren't paying ~60% of the cost for the transmission, so we might as well buy a new vehicle and we'll get more $/value. It's a reasonable rule of thumb, but I think the main problem with that is that a used transmission with 100K miles is not very comparable to a new transmission.

Your logic is a bit more sound in that it clearly lays out the options rather than defaulting to a single rule of thumb. If you're planning on buying the exact same used van at the exact same mileage, etc, the main difference here here'll be benefits of having a new transmission vs another used transmission and the $ comparison between the 2 options. (WARNING: I know very little about cars, so this is just illustrative) If you buy a new transmission, you might get another 200k miles out of your existing van. If you buy another used van with the same mileage, you might only get another 100K out of it given it'll come with a used transmission. So if you are calculating buying the new vehicle and selling the scrap of the current vehicle has more $ value, I would consider whether it's worth the potential extra mileage out of your current vehicle. And then there's a multitude of different factors to consider depending on the different characteristics of the alternative car you could practically buy at the moment.

  • 1
    Good answer but there's another factor that comes to my mind. When buying another used van for a price that's a bit above the repair costs of the transmission, you'll have an additional vehicle full with spare parts (besides the transmission and also given that you have space to store it) that might/could be used for future repairs. If one is technicallly versatile you could also take the old car apart and sell everything as spares - besides the transmission obviously.
    – iLuvLogix
    Commented Jul 18, 2022 at 16:41
  • I once had an offer for an Audi TT 2010 (high milage, engine broken) for 8k - I took everything apart and sold almost all the parts (took 1.5 years) for about 14k. So I think that scenario might be an alternative if OP has storage-space.
    – iLuvLogix
    Commented Jul 18, 2022 at 16:41

From my reading the difference between your spouse's appraisal and your is that you would include "whatever we can get out of selling our unrepaired van"? I think it would be better to include this price in the calculation, but there is more uncertainty that should be included in the calculation that would swamp the magnitude of this difference, under the assumption that the value of a broken van is low. These include:

  • True replacement cost
    • As you note, there is a massive difference in the estimates you get of the cars value. I suspect even the asking prices on Craigslist are on the high side, you need to know the price these are selling for, not the price people are asking for them. Two sources of this information I use are ebay (select "Show completed listings only" and note the price cars actually sold for) and online auction listings, in the UK I use John Pye Auctions. I suspect if you look at these sources you will get a significantly lower estimate of the value of your vehicle. Another solution would be to attempt to buy a replacement for the value of the repair (perhaps + broken car value), and if you succeed then replace the vehicle, otherwise repair it, but this costs significant time.
  • Quality of the vehicle
    • With any used vehicle the quality of the mechanics is the most important factor in determining its value. However this is always unknown. With your own vehicle you know that at least in the last 2 1/4 years it have been serviced and not abused. You know that you are not intentionally hiding something that will become expensive in the near future. However you also know the transmission has just gone. This may be a sign of previous mistreatment, or that it is coming to the point that things will start going wrong. With a new second hand vehicle these are not known. The relative magnitude of these effects will always be unknown, and you must make a decision as the the relative value of your car to an identical one of unknown provenance. This calculation is likely to have a greater magnitude than the value of an unrepaired van.
  • For the transmission specifically, for my make, model, and year, they are notoriously bad, so in my mind it's probably not a sign of previous mistreatment or other things about to go bad. Commented Apr 13, 2022 at 14:03
  • And I should say, my wife would include the resale price of the van in any consideration of a follow-on vehicle to buy, I'm just saying that her main argument considers only the fact that the repair cost is a large (likely majority) fraction of the current value of the (if in good condition) van. Commented Apr 13, 2022 at 14:05

As others have already answered, no this is not a sunk cost phallacy.

On the question on what to do with your van, from an economical point of view I would say:

  • Consider the sale value of the unrepaired van, i.e. what you could get it sold for. Call this "VU".
  • Consider the purchase value of the repaired van, i.e. what it would cost you to buy such a van. Call this "VR".
  • Call the repair costs (i.e. $4400) "RC".

You own a van of value VU. You want a repaired van that will have value VR. Now there are two paths to go there:

  • spend RC on repairs
  • sell your van for VU and buy another one for VR.

If RC is greater than VR - VU you should repair, else buy a new van.

The trick is in getting reliable estimates for VU and VR.

Note that VR should be higher than the cost of an equivalent van with a gearbox that is not broken, as a new gearbox represents more value than a gearbox with some miles on it that might break any day.


This is not an example of sunk cost fallacy, sunk cost fallacy is the tendency to continue doing something because you've invested time/money in it even when it makes more sense to just bail. Say you spent on the transmission and then the engine needed a full rebuild a week later, you might feel compelled to repair the engine because you just spent a bundle on the transmission. That's not a perfect example, because some portion of the repair cost is recoverable through sale of the vehicle (not truly sunk cost), but I think it helps convey the idea.

Ultimately you're both trying to figure whether the repair is worth it and factoring in the value of the vehicle into your calculation. The big issue is that you are not likely able to get accurate valuations on either side of the repair without going through a fair bit of effort to try to sell it.

To your spouse's point, if a functioning version of your vehicle costs as much as the repair to make your vehicle function again, then clearly it would be better to buy a functioning version of your vehicle and get what you can for salvage on the old one. So, repair cost vs value is meaningful, but it's difficult since to really math it out you'd need to assess value without the repair and value with the repair.

It's a complex decision, but I think you're both looking at it rationally.


Your car repair is $4,400. If you are willing to pay $4,400 right now, you have two choices:

  1. You pay $4,400 to have the transmission repaired. You then have on old(ish) car with a new transmission.

  2. You sell your car as it is for as much as you can get, probably to someone who thinks they can repair the transmission for cheaper. Then you add $4,400 and buy the best car you can get for the total.

So your decision is now: Each costs $4,400, but which gives you the better car? If you look at it this way, you can compare the two decisions easily.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .