That's in reference to the desirability, not the availability. When I say "all else being equal" that is in reference to the future version of the same comparison. If one option is cheaper now, it will most likely be cheaper in the future too, and if it is cheaper then that means it is less demanding in terms of resources to acquire, and if it is less demanding in resources it is "easier". You might have a special case where the less expensive option will suddenly become much more difficult than the expensive option to acquire, but there is no reason to assume in any particular instance that's true.

For a more concrete example - if you have bought a $10 and a $100 ticket, buying a second $10 ticket is, all else being equal, preferable to buying a second $100 ticket, regardless of how desirable either option is.

> if you have bought a $10 and a $100 ticket, buying a second $10 ticket is, all else being equal, preferable to buying a second $100 ticket, regardless of how desirable either option is

Not if you won't desire it anymore (or desire less) at later date. Or if you won't be able to take it at a later date at all (which influences desirability at present). Or when the price doesn't really matter, you'll be able to afford it anyway. Or in multitude of other circumstances that may be relevant to the given example that you may think about once you decide to not miss the point of that example anymore, since it's very far from implying that "all else is being equal" ;)

Again, when I say "all else being equal" that is the explicit assumption that the circumstances in the future are essentially the same (ie equal) as they are now. We are assuming all else is equal because we are given no reason to believe otherwise. If something were, for example, a once in a lifetime opportunity, we'd refer to it as the once in a lifetime option, not as the less expensive option. The fact that we are talking about this decision in terms of price implies these are normal, purchasable commodities whose relative prices are reasonably stable, and that monetary expense is a significant concern.

Of course there are situations where you can't or wouldn't want to buy the same thing at a later point in time, but in general you can. Starting with the assumption that you purchased both a $10 and a $100 ticket to two different events because you wish to experience both and further assuming that you can still experience both by purchasing a duplicate of one ticket which is available at the same price, as is generally the case, keeping the expensive ticket and buying a duplicate of the less expensive ticket is the rational choice. Without an explicit good reason to do otherwise, it is not an example of the sunk cost fallacy to reduce your future expenses by keeping the more expensive ticket.

> The fact that we are talking about this decision in terms of price implies these are normal, purchasable commodities whose relative prices are reasonably stable, and that monetary expense is a significant concern.

Are trip prices "reasonably stable"? Isn't there actually a huge seasonal variance, last-minute offers and so on?

> Of course there are situations where you can't or wouldn't want to buy the same thing at a later point in time, but in general you can.

The example is about trips, which - at least in my experience - are usually chosen based on a huge set of variable incentives to go at specific time to a specific place.

Anyway, the article gives no reason to assume that "the circumstances in the future are essentially the same (ie equal) as they are now" - it actually gives a reason to assume that there are other incentives than the price and doesn't mention whether they change or not in the future - so we can't assume that they won't (and even if we could, it would still be irrelevant to the point it's trying to convey - all it talks about is that autistic people are apparently more likely to take those other circumstances into account, which can lead to a different outcome).

For the third time, we are told that there are two trips, one of which is "significantly more costly". It doesn't matter if the prices fluctuate, the fact that one is the "significantly more expensive" one, as opposed to any other description which does not utilize price, indicates that the relative difference in price is useful for distinguishing them, and indeed more useful than some alternative method of distinguishing them. You would not refer to something as the "significantly more expensive option" if you reasonably expect it to be cheaper than the "cheap option". It would be one thing if we were told that there was a last minute offer or a seasonal variation at play, but we were explicitly not. The question is what is the logical choice given this information, and that logical choice is to go on the more expensive trip. It doesn't matter if you'd make a different decision under different circumstances, it's fine for something to be logical under some circumstances but not under others.

The whole purpose of this discussion is that the point that the article is trying to convey, that autistic people are more likely to consider other circumstances and are thus more logical because they avoid the sunk cost bias is incorrect, because this is not an example of the sunk cost bias. The non-refundable cost you sunk into two things in the past is relevant to the decision of which one to keep because you can use it to predict future prices. What the article claims most people do is in fact the rational approach, and what the article claims autistic people do is generally irrational, thus working against the thesis of their article.

> The question is what is the logical choice given this information

Not enough information is given to make any choice, and there's no intention to give it, since this is not a puzzle. It has no correct answer. It just mentions that some non-financial motivations are at play, but does not specify them (because it's not needed to make a point). You can't make a decision, because you don't have the data.

> The non-refundable cost you sunk into two things in the past is relevant to the decision of which one to keep because you can use it to predict future prices.

You don't have enough information to assume that, or that any future prices will be relevant at all. In fact, the article explicitly says that this cost is "irrelevant" - which makes sense, since you may have no intention to rebook that cancelled trip at all!