He writes,
I have not been able to find a causal account as to why information failures (particularly with regards to quality) lead to market failures
In textbook economics, a market failure is when the private incentives lead to either too little or too much of a good being produced.
In terms of information, consumers make purchases based on what they can observe. If what they observe is highly correlated with quality, they should do well. But not necessarily otherwise.
Consider a high school student making a college visit. The appearance of the facilities can be observed relatively accurately, but it is not very highly correlated with quality. The quality of classroom instruction cannot be observed so accurately, because the high school student will not sit in on very many classes. But suppose that the quality of classroom instruction is highly correlated with the value that the student gets out of college.
We can predict that colleges will over-spend on the appearance of facilities, because that factors heavily into the decision of the high school student. We can predict that colleges will under-spend on classroom instruction. Market failure.
The public policy response should be to tax college facilities and/or subsidize quality classroom instruction.
I am not offering this as a realistic picture of a market failure in the market for higher education. My point is to answer the reader’s question about connecting information failure to textbook market failure.
Should we mention “The market for lemons” about used cars? George Akerlof’s article.
Akerlof’s article is right on point, but is not without its critics. For example:
https://fee.org/articles/uneven-information-causes-market-failure/
Watch the facility for a day, especially the front entrance. Look for smooth consistent student and faculty flow that matches the entry door. That is what catches the eye of students and parents, activity in and out. If the student observes the school building to be standard, then a well balanced flow means the class rooms uses to optimum capacity. The students usingthe facility to scale.
That is information, in theory and practice. Information is a feature selection that maximally uses the carrying capacity of a fixed channel. That is literally an equivalent definition to the theory and practice of Shannon encoding and school building watching.. Observing parents are Shannon encoders, they get a Shannon information map of two universities, then literally overlap the structures, like overlapping architectural plans.
I should mention, this is the hologram effect.
When Walmart or a school facility can stablize the queues in and out, then the concept of student/square foot work, the building can be considered a bath tub with flow in and out, this is linearity. Walmart manages flow in and out at the checkout stand, causing the hologram effect. Thus Walmart can scale stores up an down using money velocity equations.
This is a basic of quantizing systems, quantizing classrooms, floors and entry doors. It is equivalent to information flow over constricted channels it isphy6sical Shannon machine in operation.
Before we leave the subject, Arnold, I claim my description of events, as above, is exactly the macro math of PSST, the one you claim cannot exist. It exists for one reason, the theory above conserves chaos energy, you think it indecipherable. True, but it can be dispersed, in the sense that we find the optimum imprecision everywhere, the inventory slop needed for sustainability.
Speaking as a Canadian where we spend about 10-11% of GDP on health care versus the roughly 18% of GDP the US spends, what would US health care look like if hospitals had to post a price list for their procedures.
I’m given to understand there is a new law that requires them to do so but this article states that they’re hiding their price lists.
https://qz.com/1518545/price-lists-for-the-115-biggest-us-hospitals-new-transparency-law/
In situations where you can’t price shop (emergencies), a price list would be meaningless.
In situations were you can shop, it’s not that hard to get prices.
Of course the price you get would only be the point of sale price for you. In healthcare the same procedure can have many price points for many different contracts, and the net price may greatly differ from the sticker price.
So let’s say procedure X costs man on the street $10,000. It might cost an insurer $2,000. But the insurer pays $5,000k upfront and that gets a $3k from the hospital months later (which it doesn’t use to reduce your cost sharing payments retroactively). Or maybe it doesn’t get anything from the hospital but another procedure done for another patient that the hospital should charge $20k for then instead charge $17k, so it balances out to that insurer in aggregate but the costs and prices have little to do with one another.
Oh any every insurer has a different deal, and they are all confidential.
So what exactly would a “price list” mean in this instance.
I think you would have to add another pre-condition: the insured bears more than a paltry amount of the the “now discovered” cost. Otherwise, little will happen. The vast majority of US health care consumers bear one some of the lowest out-of-pocket costs in the OECD.
I would be tempted to drive home the point about the market failure(s): The high school student makes a sub-optimal allocation of scarce resources (time and money) because he is poorly informed about the quality of the colleges he can choose from. The colleges make sub-optimal allocations of scarce resources (e.g. too much groundskeeping, not enough instructor pay) because the market rewards them better for looking like good colleges than for being good colleges. After many iterations, all surviving colleges look better and teach worse than they should because of this market failure.
Tempted, but think twice.
My claim is that an AI algorithm can watch the main college building doors, analyze the flow and tell you which build is the most productive, to within 10%, easily. Never know what the building was used for, all I need is the assumption that it standard construction, and sustainably used.
The key here is standard construction, the engineer knew the use of the building and designed rooms and the number of rooms to match the doorway entry and keep flow smooth. That is all I need to know, and call that standard construction. The key knowledge is that human do not like being crowded. Any building being used efficiently, for its intended purpose, will set the crowding level to human comfort. I can measure that directly from the the entrance.
I was talking about the main post, not your claim. I get that you can measure optimum facility usage; I simply have no idea whether or how I should expect that to correlate to the quality of the instruction. Given that facilities are long-lasting and enrollments are variable, I would expect facility usage to vary over time for unrelated reasons. For example, if your method tells me that the philosophy building is underutilized, that might be because philosophy is less popular than it used to be. Another building may have low utilization because a lot of its classes are highly specialized and have small classes in standard-sized classrooms.
If you are looking for signs of over-spending on facilities, I would think visible construction (i.e. present spending on facilities) would be a better signal because it includes less time-averaging. If the amount of current construction is not in line with enrollment trends, that’s a much less ambiguous signal about current facilities spending relative to needs. But still kinda ambiguous, because some classes may be online.