Alchian–Allen effect
en.wikipedia.org
en.wikipedia.org
Today: Gold $10; Silver $0.10
Tomorrow: Gold $1010; Silver $1000.10
By adding a constant to both prices, the price ratio no longer remains aligned with the underlying value ratio. Assuming you have to get one necklace each day, there is less of a decision tomorrow than there is today, because silver is overpriced relative to gold. The correct prices for tomorrow should have been "Gold: $1010, Silver: $10.10", so you're effectively getting more scammed by going with silver.
> Another example is that Australians drink higher-quality Californian wine than Californians, and vice versa, because it is only worth the transportation costs for the most expensive wine.
Maybe it's a branding thing, and French wine's got enough power in that department that they don't need to send us good stuff to move product, at least at the lower end. Maybe with some French writing on the label low-price-shopping consumers are equally likely to buy it whether or not it tastes good, so there's no reason to send anything but the worst they have. Other European countries seem to have no trouble shipping us nice wine in the lower price tiers, but then they're not France.
Especially with wine, where it is not actually so clear that a true quality ordering exists, all sorts of other effects are at play. For example, we know how the signaling value of price overrides some true quality variables.
In economics, its important to know these simple models and their outcomes, but it is equally important to know when other effects are at play, or when assumptions do not hold
[1]: https://en.wikipedia.org/wiki/Sunk_cost#Loss_aversion_and_th...
Uhm, I'm not an economist but how is this a "theorem"? What is the theory, what are the axioms, where is the proof? This seems like an empirically observed trend.
"the harder the punishment, the harder the drug" is another very important non-trivial observation, missed by most people.
Did you read my comment? That's exactly what I wrote. This is not a "theorem". Theorems live inside theories, they're logically implied by axioms of the theories. That basic arithmetic proves nothing of the sort.
axioms: * given same price, consumer goes for higher grade product. * as relative price difference decreases, some consumers start going for higher grade product. * rules of arithmetic.
in any case, it seems that we have appreciation for different things. you prefer detailed verbosity, i prefer a simple idea that can be easily understood.
great and profound ideas are not necessarily complex, many of them are quite simple. let's keep them that way.
Since the whole theory won't fit in one wikipedia article, I suggest you read the first chapters or either a) Mas-Collel, Whinston and Green or b) Kreps, Microeconomic Foundations I (not the old undergrad book from him)
This would provide you with the theory to understand this result in a more mathematical correct way.
[0] https://mpra.ub.uni-muenchen.de/901/1/MPRA_paper_901.pdf
As a person who loves Belgian beer, I came to this realization when I was on a bit of a beercation there. For the most part, the good Belgian beer makes it here. When I'd try beer from breweries I hadn't heard of while I was there, it was mostly worse than the Belgian beer I had tried at home.
"The effect has been studied as it applies to illegal drugs and it has been shown that the potency of marijuana increased in response to higher enforcement budgets,[2] and there was a similar effect for alcohol in the U.S. during Prohibition.[3] This effect is called Iron law of prohibition or Cardinal rule of prohibition.
Another example is that Australians drink higher-quality Californian wine than Californians, and vice versa, because it is only worth the transportation costs for the most expensive wine.[4]"
To clarify: I just don't see where the evidence for any of these claims is accessible. Four of the six references don't appear to be available on-line. Of the two that are, one is a brief summary that leads to a paywalled FT article that I can't read, and the other is a ~200 pdf economics textbook.
I'm happy to accept that a price differential changes when a constant is added to both prices. But I do question whether people in the real world will switch to a different type of coffee bean because the differential goes from x2 to x1.6. And I wanted to know more about the real-world circumstances in which it had been observed.
I might be pursuaded that the market for illegal drugs/alcohol could exhibit this effect, but it would be interesting to know how how, given that illegal products are probably going to be subject to other effects not shared by legal products, biases and selectivities were compensated for.
Edit: Made first paragraph clearer.
> Thanks, I also read the article.
I'm not sure the extent to which you considered this, but in this context it's hard to read that sentence without making you sound like a jerk.
> I'm not sure the extent to which you considered this, but in this context it's hard to read that sentence without making you sound like a jerk.
Did you consider the possibility that I was simply thanking dyeje for their reply, and confirming that I had noticed the examples that they referred to? Assumption of good faith, etc? Nevertheless I've edited my reply to hopefully make my intended meaning clear.
But, the thing you think is a confounder here "prohibition makes a product hard to get in desired volume" is pretty much equivalent to an economist as saying "the price went up". There is no really meaningful difference. I have to expend more - either currency (to some bootlegger or intermediary, to cover their risk), time/effort (to find and get to a seller), or risk (I might get caught and go to jail - we can put a price on imprisonment).
Modern economic theory is not some abstraction that doesn't resemble the real world, even in cases where the empirical evidence is as-yet weak (which it's not in this case, as is apparent to anyone who reads the whole entry).
Rationality is an observation of aggregate behavior, rather than individual. This is pretty robustly observed. It is interesting because this is in spite of the glaringly obvious observation that most people are not rational at all times (though many people are sometimes rational). Thus it is a useful abstraction to reason about rational individuals, though I have never been exposed to a serious claim in academic economics that individual human beings are rational.
The interesting results of behavioral economics are many, but they primarily give us proof of the already-known fact that individual humans are often irrational. The interesting part is not that humans are individually irrational, but how.
Behavioral economics makes the observation of aggregate rationality even more interesting, because many irrational actors can cause aggregate rationality.
If you would like my musing about how it is that what I have said above is so often misunderstood, then I have some thoughts below.
In my academic experience, all of my intro level professors took pains to ensure we understood that we were learning useful models, and that rationality is an observed phenomenon in aggregate. My intermediate level professors often didn't take as many pains to make this clear, but did make similar explanations. For anything 300-level or above, we never really discussed this topic, as the professors assumed we had already learned such.
Based on my observations, I would guess that most people who do not take a full course load of intro and intermediate economics do not have drilled into them that rationality is not something we expect of individuals, despite it having a lot of explanatory power in aggregate observations.
Additionally, as with all fields, those who practice often forget what is taken for granted in the field. Among economists, unless one has significant experience teaching economics to beginners, I expect it is not a common thing to preface every statement with the type of 101-level introductory remarks that would help to remove such confusion.
Additionally, economics seems to me to be plagued with jargon that are homonyms with common words in the vernacular. E.g. "cost" to a layperson often means "value denominated in currency, which I might pay to someone to acquire a specific good or service" or "the amount my supplier paid for the thing I am purchasing, such that their cost is less than the price that supplier charges me", whereas to an economist, "cost" is pretty much always going to mean "opportunity cost"[0], which is the value (to the entity making the choice) of the next best alternative. This sort of homonymity is rampant in economics jargon.
Finally, there is a pervasive assumption in economic thinking of "ceteris paribus", or "all other factors being held constant" (technically "all else equal", but give me some leeway in the translation to capture intent a bit better). Most theories, models, and statements in economics are going to be statements about a single factor/observation with a magical "nothing else can or will change" condition applied to the entire universe. This is, obviously, a huge over-simplification, but one that is useful for discussion.
In summary, all models are wrong, but some are useful. Economists often discuss in terms of models and the way they discuss uses words you think you understand but you don't, unless you study economics (not a dig, just a truthful observation of laypeople using jargon - see a recent SMBC for non-economics examples[1]).
[0]: https://en.wikipedia.org/wiki/Opportunity_cost - Key excerpt:
> Thus, opportunity costs are not restricted to monetary or financial costs: the real cost of output forgone, lost time, pleasure or any other benefit that provides utility should also be considered an opportunity cost.
If the price of all marijuana goes up, then I bet overall usage of it goes down even if the individual potency goes up.
>If the price of all marijuana goes up, then I bet overall usage of it goes down
in high margin (and supply limited) goods i think it works in the opposite direction - the price goes up in response to the demand going up.
Let's say your "strength" is simply the sum of the value of your pieces. The losing side has a strength of l and the winning side has a strength of L.
The losing side loses by l-L in absolute terms, or (l-L)/(l+L) in relative terms.
An even exchange of value k makes it go to (l-k)-(L-k) = l-k in absolute terms (no change); and to (l-L)/(l+L-2k) in relative terms. (l-L)/(l+L-2k) < (l-L)/(l+L).
To take an example, if the situation is white : two pawns and black: one pawn, going to white: one pawn and black: nothing is a bad deal for black (the losing side)!
Am I missing some reason why it would be the opposite?
They were saying they didn't realize it had a name.
I'm not sure the effect is logically sound -- before the price hikes you weren't willing to pay a dollar more for name brand (or premium drugs etc), but after the price hike you are? I'm not sure if the authors have a mathematical model supporting the decision, or are just observing illegal drug markets and generalizing.
Imagine that 10 normal people and Bill Gates want to buy bottles of wine. Bill goes for a bottle of $6000 Richebourg Grand Cru, and the 10 normal people always go for Shitscreek Red, sold for $4. So about 9% of the buyers are willing to pay extra for the name brand.
After the Great Wine Tax Bill of 2021, the prices are both increased by a fixed per-bottle amount of $6000. So now the Richebourg costs $12000, and the Shitscreek costs $6004. The 10 normal people migrate to beer, and Bill Gates still buys Richebourg. So about 100% of the buyers are willing to pay extra for the name brand.
A similar irrationality is when people will do an extra hour's work to save $50 off a $100 purchase, but not to save $50 off a $10000 purchase.
Now, imagine that someone puts a tariff in place so that the low quality product now costs ten dollars, and the higher quality product costs eleven dollars. Sure, you might still buy the lower cost product, but if you are going to buy the product at all, the incremental cost of spending eleven dollars instead of ten dollars seems much more worth it. As long as you have to spend ten bucks anyway, might as well spend eleven and get a higher quality good.
It's not that people always buy the more expensive product.