Letter from an “Anti-School Teacher” | Hacker News Reader

Letter from an “Anti-School Teacher”

econlib.org

51 pointsjseliger96 comments

Game theory is a useful framework for this problem in the sense that almost any double-sided market can be trivially formulated into a game-theoretic framework. But viewing college admissions as a Nash equilibrium doesn't meaningfully clarify the problem.

That said, I generally agree with the article. America's elite private universities have almost always been about signalling over substance, an orientation that can be traced back to their origins as bastions of elite privilege. Our recent attempts at reform, via affirmative action and more recently, the SAT adversity score, have been nothing more than attempts to spread the benefits of signalling to new, heretofore underrepresented, groups.

The real solution, of course, is to tear this reputation market down. We need to work to remove the signalling benefits of so-called "elite" institutions, and in doing so, remove their stranglehold on access to a number of prestigious jobs and career paths.

How do we do this? I'm not sure, exactly. The radical solution is to nationalize all universities, redistribute their endowments more equitably, and enforce a singular admissions and educational process on all of them. This enforced uniformity would destroy "elite" signalling.

There are incremental steps to this, though. Implement new rules to force colleges to expand their enrollment to maintain non-profit status (and thus retain untaxed earnings from their endowments). Pour more money in flagship public universities to create strong alternatives to private education (i.e. make a UC Berkeley for every state). And so on. Reduce the distinctiveness of the elite privates, and their value as signalling institutions decreases.

"Tearing the reputation market down" is not generally a good solution, people want to acquire reputation for sensible reasons and tearing one market down just leads to it being replaced by other, generally suboptimal markets. What you actually want to do is (1) stop subsidizing reputational games heavily, so that agents who engage in them don't get to externalize their cost to third parties, and (2) try to outcompete the standard reputation game by signaling the same desirable qualities more cheaply.

The latter is sometimes difficult to do in the case of education, because conformity is itself something that people might want to signal, and it's hard to outcompete prevailing standards while conforming to them. But providing alternatives is still valuable, because sooner or later you'll get a positive unraveling cascade where people realize that the advantages of conforming to the existing standard don't justify the cost, and it's better to switch to a cheaper alternative. (It's something that's mostly unpredictable however, since it works on the principle of a popularity contest where the prediction itself is what people are constantly optimizing for!)

What would (1) look like in the context of education? What exactly do you mean by externalities?

Otherwise yes, I agree that the ideal solution would be to provide compelling alternatives at lower cost. The issue is, we've seen this fail already in education. MOOCs attempted to democratize elite education, but instead increased the entrenchment of those institutions. In software, bootcamps and the like have also failed to dislodge the prominence of signalling. Indeed, if there's anything the past decade has shown, it's the increased availability of information, and the increased democratization of resources, have only increased the value of signalling. The market itself has become so skewed, so incredibly inelastic (to cost in general), that competition is becoming nearly impossible.

There is no one solution... I think that we need to start accepting that, and acknowledge that there are overlaps in terms of ability and the course one takes to learn and inject themselves into a profession.

Like bootcamps... from my experience the best of the bootcamp grads will outperform and do better than the vast majority of developers from traditional computer science backgrounds for most work. The bottom of the bootcamps are worse than those from traditional education. Self-taught tend to be towards the top as well. The rest is a deck of cards and some roles lend themselves to those traditionally trained, or the few that go very far above and beyond what they have been required to do in terms of learning.

This also overlaps with regards to natural aptitude as well. There are people that do not, and likely will never get it. There are those with personality issues that will take years to understand how to interact with others (if ever). It's a mixed bag.

The "pre-interview homework" doesn't really help much. Depending on that process, as an interviewee it comes so early that you aren't even sure that you're interested enough in the company and have enough other places to talk to. It will depend.

That said, it isn't that hard to compete. Show competence, drive and a desire to learn. If you are Junior level, work to level up, make something, throw it up on github and interact with a broader community. Read up on the things you want to learn and use them. Write about what you are learning, and get feedback.

If you're mid level, get to know your design patterns better. Don't assume you should apply them everywhere, but be aware of them.

If you're senior level, branch out and learn new things. Yeah, you've used C# and SQL Server the past decade, have you done .Net Core? Have you looked at other modern languages lately? Challenge yourself.

Competing means dedicating time and effort, and it doesn't take too much. It does take more at the beginning of your career. I have been at this for almost 25 years now. I spend 2 hours a day on average looking at technical articles and commenting. I spent my recent vacation reading about Rust, usually do road trips. I am passionate about my own career, and still enjoy learning above anything.

Above all else, be honest about what you know, and what you don't.

You're talking about changing the supply side of the equation, which isn't inherently a bad idea, but the demand side of the equation is driven by a job market that often doesn't have time to properly vet candidates and would RATHER use signals because they're faster. So you'd have to find a way not to eliminate the signaling, but rather to more closely align the signals with reality.

This works in some fields, like software, which are based on measurable skills that can be directly taught. And we see it already with bootcamps and MOOCs (though these will need reinterpretation for real success).

However, in many other fields, like consulting, or investment banking, or law school admissions, there isn't a similar quality of measurability. Indeed, for many of these fields, the signalling is the end in and of itself, where the job itself doesn't require specific skills, but it's a benefit to the hiring institution to hire from signalling institutions). These sort of problems are uniquely resistant to market mechanisms and need top-down reform.

I don't know about "the job itself doesn't require specific skills," I think it's more that the skills are "soft" and thus much harder to measure. A consultant's job often involves listening to a lot of people and collecting a lot of data and then synthesizing all of that into actionable recommendations... but then again, consultants are often hired just to confirm whatever a manager already believes, so their perceived utility isn't actually proportional to their ability.

Yeah, these are tough problems.

I'm not sure that this is true even in software. "Professional" skills, interactions, leadership and other skills can indeed be taught, but are difficult to actually test for. There are a lot of skills in software that are invaluable and come only with time, experience and self-reflection.

Also, I find a significant divide in working with people who have come out of bootcamps, etc at a junior level. The divide is really down to those who did work outside/beyond their assigned work, and those that didn't. It's an indicator of passion and aptitude that isn't really testable because then that would become the new baseline.

I'm not saying that reform isn't needed, I think there's a significant space for it. For that matter, changing perceptions of what higher education actually is and isn't good for would be a start. In my own experience, the top performers out of a vocational program or self-taught will outperform and do better than the vast majority of those traditionally educated. The rest will stack in the middle, and those that didn't do well in the vocational setting will be at the bottom.

I think the start is to recognize that procedural differences can and do have overlapping outcomes, and to look for those secondary indicators and try to acknowledge them.

> This enforced uniformity would destroy "elite" signalling.

It would definitely destroy the supply. What it wouldn't do is destroy the utility of elite signaling (humans love tribal reasoning, it's cognitively easy) and thus the desire for elite signaling would continue. Much like even in schools with a strict uniform dress code, you can still tell who has money and who doesn't.

Removing the supply of something rarely eliminates its demand. Thus, removing the supply into the reputation market is unlikely to eliminate its demand. People will find another way to signal.

Also, it might be considered that many states already have what could be considered excellent, high-quality, strong flagship public universities. Michigan, Ohio, Illinois, and Georgia all spring to mind as states with excellent public universities, and I know there are more. Some of them are larger than UC Berkeley.

One thing an elite school is signaling is that it’s at the very least not among the worst.

I went to an elite school and the thing I’m most proud is that we learned knowledge that doesn’t decay as quickly (math, oh so much math). If I’d taken a less elite school I’d probably have some less than useful bazaar, ruby on rails, software architecture of the month thingy

Anybody up for a grand bargin? I'll trade you Medicare-for-All in exchange for a complete privatization (w/ vouchers) of education.

Privatization would make these sorts of problems worse, as schools compete more vigorously to sell signalling value to parents.

The fix is to monopolize signalling value through legal fiat - make it illegal to ask job applicants or allow applicants to disclose the school they attended, and replace it with an examination and/or interview based certification monopoly.

And, indeed, what studies are available seem to confirm this. Not just for private schools - charter schools also tend to underperform the public schools.

Granted, kids who go to them do tend to get better test scores, go to better universities, have more successful careers, etc. But that's because the whole education industry is a massive exercise in selection bias. The most careful studies I've seen, though, seem to indicate that, all other factors being equal, a typical kid is going to learn more in public school than in private or charter school.

This rather mirrors my own experience. I did 2 years of public high school, and 2 years of college prep. I can say that it's absolutely true that my classmates were generally smarter and more studious in the prep school. But the prep school was also easy. It really didn't particularly matter how much effort I put into my homework. As long as I didn't blatantly blow it off, my teachers would dutifully send my parents a report card full of the As and Bs they were paying for.

By contrast, it was the public schools where I had the teachers who'd keep ratcheting up their expectations until they could sense that I was actually working to try and get good grades on my papers.

In terms of actual student performance (as opposed to grades), I'd expect parents to be a major factor. They're going to build support networks-- running the fundraisers, supporting extracurriculars, keeping an eye on student performance and scandals.

When we provide "options" -- charters, private schooling, even public magnets-- the parents who are motivated and concerned enough to build those networks tend to pounce on them. The "default" public school loses its champions and ends up producing worse outcomes for everyone whose parents aren't tuned in.

> Privatization would make these sorts of problems worse, as schools compete more vigorously to sell signalling value to parents.

That's one of the most compelling arguments against school vouchers I've encountered. Especially as that is exactly what has happened in higher education where consumer choice is higher.

Living in Japan, I am compelled to agree.

Most universities here are private, and everybody knows that you have to get "a college education".

The end result being that, outside of a select few national institutions that appear to have some degree of academic rigor, university in Japan is referred to as a "four-year vacation", and graduation is effectively automatic.

Reflecting back on the original post, I'm not sure that nationalization of universities in the US would help. We have plenty of high-quality state schools as-is, but Harvard and friends are prestigious because their brands come with deeply powerful social networks.

The same effect applies to making it illegal to put university names on resumes.

Sure, there is definitely a halo effect in being able to put Harvard or Stanford on your resume, but I would wager that the vast majority of top-tier university students don't apply by (metaphorically) filling out a job application and handing it to the shift manager.

Instead, your work comes through your network. Alice introduces you to Bob who knows that Charlene has invested in TinyCo, which is poised to be the next unicorn, and so you get in on the "ground floor" in this fashion.

That's not the kind of thing you can really address via legislation.

(Note that there is an assumption here that you can Do The Thing. Charlene probably doesn't want you working at TinyCo and squandering her investment if you aren't going to help make it the next unicorn.)

I discovered that it’s even worse. Alice’s mom sends an email to Bob who tells Charlene that his friend’s kid is super smart and of course from Harvard, and then when Alice calls she can ask for Charlene directly and see if she wants to go for a glass of wine.

Competition should be a good thing. The real issue is that colleges and employers are looking at the wrong signals. This leads the existing competition.

GPA for example is arbitrary yet colleges and employers are basically asking schools to 'optimize' / maximize it. Standardized testing (ACT/SAT) is much better, but still misses important aspects like leadership ability. Even the prestige of the school is a poor signal.

Give employers / colleges a better objective measure and they will gladly use it. Perhaps we should have a post-college standardized test.

Privatization shifts the signaling game theory from pandering to government funding to pandering to consumers. Given that consumers don’t have nearly as much time to research options individually as governments do collectively, this can and likely will create far worse outcomes on average.

edit: This comment has gone from 8 points to 4 points in about 10 minutes, so I can only assume this thread has become the target of a downvote brigade from some other web site. The joys of discussing politics in the age of 4chan!

By this logic, shouldn’t we nationalize most major industries? oil/energy, motor vehicles, the list is endless.

There might be a better way now than what was practical in 1900.

At an abstract level, one of the major goals of regulation is to stop companies from doing things that consumers can only figure out are bad on a timescale long enough for them to have already done damage and made enormous profits. There are a ton of regulations around motor vehicles, like disclosing MPG, maintaining safety standards, and consistent odometer readings - things that consumers wouldn't necessarily be able to figure out on their own until it was too late. If a car just plain doesn't work, or craps out in less than a thousand miles, that's easy to remedy.

The problem with applying these standards to education is that the timeframe to assess damage is very long, and it's more or less irreversible.

If you spend a week driving a car and the engine dies, you can take it back and get a refund and then go get a new one.

If you put your kid in a charter school and they fail to learn math for a year, you can't get that time back. It's gone forever, and your child will be behind for anywhere from months to years to forever.

This is why we need to treat certain specific sectors differently. Consumer goods, sure, whatever, the government can make sure that toys don't contain lead and that our TVs wont randomly explode, and after that it's up to us to decide.

But education, healthcare, etc., the consequences of giving consumers "choice" and allowing those sectors to advertise freely and self-regulate are severe and often irreversible.

My middle school and high school career was upended by an experimental math program that wasted an enormous amount of time forcing children to answer questions like "What is 2+3?" with a minimum of three sentences. The head of the department was later demoted for injuring a child when he sat on him as a form of punishment. He was later acquitted of assault charges. Did regulation work there? Could it have? Do government schools not need to be regulated by virtue of the fact that they're already promulgated by the government?

As an aside, while looking this up I found news stories that demonstrated the rumors about sex abuse were true and a few coaches-turned-principals got nailed. While your theory about regulation seems well-intentioned, I'm just not sure it is worth anything in practice. Lots of governmental theory seems revolutionary until it ends in failure (and, at worst, bloodshed).

The plural of anecdote is not data. And taking the furthest logical conclusion ("Government regulation ends in violence!") is utterly ridiculous and unnecessary in this context.

To be clear, the United States' public education system is BAD. It is sometimes VERY BAD. But it's not consistently bad; there are public schools that are quite good.

There are other countries with heavily regulated public education systems that are so much better than the United States - read up on Finland's remarkable public education system if you don't believe me. (And don't get tripped up on inconsistent usage of "public" and "private" to mean opposites in different countries!)

So logically, the problem isn't that government regulation of education is inherently bad; Rather, the problem is that we (in the United States) are doing a bad job of it. Splitting public education funding between public schools and charter schools and forcing public schools to compete with test scores and other proxy metrics is demonstrably making things worse, not better.

"We" are not doing a very good job of it. I agree with you there. Both my parents (and their parents) are or were educators in public schools, from teachers to instructors to principals and coaches. By all accounts it is getting worse, not better.

Anecdotes are not data, but it would be foolish to throw them away blindly. Anecdotes become data when they can be used to answer questions that can inform future decision making processes. For example, we could ask, "is more politics and political process positively impacting education?" If the answer is that no, one hundred years of American politics hasn't improved the situation [0] despite it costing far more than it used to per-student [1], then we should probably conclude that said process is a dead end and look elsewhere.

I never claimed that the anecdote I provided was data; I would rather like to know the answers to the questions I asked. Handwaving them away seems closer to arguing in bad faith than what you accused me of doing elsewhere. Further, I would suggest that the increased presence of School Resource Officers in schools, the increased usage of courts for school-related punishment, and the continued emotional and physical violence experienced by students from their peers show that, indeed, forced participation in government schools does indeed involve violence. That your experts have been unable to solve this problem (and that you hand wave it away) does not generate confidence in your position.

[0] https://www.nationsreportcard.gov/ltt_2012/summary.aspx [1] https://www.cato.org/blog/public-school-spending-theres-char...

"Further, I would suggest that the increased presence of School Resource Officers in schools, the increased usage of courts for school-related punishment, and the continued emotional and physical violence experienced by students from their peers show that, indeed, forced participation in government schools does indeed involve violence."

This is a pretty bad correlation/causation argument.

You're making arguments that only make sense if: 1: The United States is the only country in the entire world with a government-run education system and there are no other examples to look to 2: Every decision made that has negatively impacted our education is the only possible decision that could have been made under the circumstances

Neither of these are true.

Just because our government has made some really bad decisions doesn't mean that this is the only possible outcome of government-run education. If you look at education rankings globally, the United States is pretty far down the list, and most if not all of the countries above us use government-run education systems. The number one in the world, Finland, prohibits private schools all together.

It simply does. not. logically. follow. that the overreach of our criminal justice system is the inevitable result of a government-run education system.

You're ignoring so many political realities that are unique to this country.

There is nothing hand-waving about saying that your anecdotes are not data. I went to a school that had exactly zero teacher/child sex scandals, can I use that to "prove" that the system is working? No, I can't.

Our public education system delegates a LOT of responsibility down the line. The federal government offers some funding that is contingent on certain requirements being met, but the states are responsible for the execution. The states have elected officials and appointed boards of education, but they delegate a lot of decision-making to individual districts, which also have elected and appointed officials.

Finally, just because I am defending the idea of government-run education doesn't mean I am defending the implementation of it that we have. Please do not confuse those two distinct positions.

The role of private industry is to efficiently produce commodities for individuals or other companies - minus externalities, autos, chips or airplanes are great in private hands.

The role of the state is to regulate positive and negative externalities. And students learning something is actually a (positive) externality. Neither the student nor the parent or even the district directly care about everyone learning basic skills - that provides no relative advantage for them. Signals of being superior are worthwhile to the individual. Those might actually result in the individual being smarter but if an individual is "selfish" or thinking in pure "homo-economicus" terms, they are just happy being able to always seem competent rather than being competent.

So the state (or whatever institution can represent "all of society") is what is needed to increase the base competency of a society.

Of course.

It's a lie perpetuated by capitalists that the government can't do things efficiently. The government can absolutely do things efficiently, the costs are usually by having more protections for the employees, which is good. Additionally, we remove the need for a profit layer that would go to owners (the nation is the owner in this case) so to achieve the same results as nation owned companies

Not to mention that there is no exit to look for, nation owned companies cannot be stripped of assets and they cannot claw and scrape the consumer, which is how for-profit-private companies would operate.

Now am I saying that all things should be nation owned? Of course not. I'm saying that monopolies, infrastructure and huge old government companies should be and should stay as nation owned. Private companies will work with providing for these companies resources that they don't have and would pay more to develop themselves.

While I will conceptually agree, that optimizing some things via the government can work better. In practice, however, it often leads to cronyism, corruption, bureaucracy and/or stagnation.

On the flip side, I'm also against certain things being incentivized by the government. I'm against private prisons on pure premise.

In the end, I'm pragmatic and prefer to optimize for those solutions which require the least oversight and ensure for the best competition practical. I'd also like to see an "open body" of course content made more widely available, so that districts and teachers may pull in said content for individualized courses. Educational publishing is about as corrupt as it gets IMHO.

I do think that for most things, there are ways to incentivize appropriate outcome from private businesses and initiatives with the government. This includes more appropriate provisioning and budgeting practices.

consumers have skin in the game

If consumers having skin in the game were a sufficient condition for them to make good choices, we wouldn't have needed the Federal Food, Drug and Cosmetic act to ban radium water in order to get people to stop buying it.

Radium water has a tendency to infect the water supply, so that even people who don't buy it are affected by it. The principle of intolerance applies the other way for schools: people who are picky about schools will do lots of research, whereas people who aren't picky will just go where their friends go.

What you're suggesting is building a system that actively punishes parents who are too busy to a ton of research because they're working multiple jobs, taking care of sick family members, etc. Characterizing people who don't have time to do this as merely "not picky" is ignoring and punishing a huge segment of the population.

The principle of skin in the game doesn't merely improve the situation for picky people, it brings up the average improving the situation for everyone. If picky people at large boycott a particular school, everyone who is friends with these people will boycott it by association, and it will lose too many students to remain solvent. If picky people prefer giving their business to certain proprietors, those proprietors will open more schools, increasing the likelihood that a person choosing schools using a dart board will strike such a school.

Your characterization of the difficulty of choosing a school is overdramatic. A person could devote fifteen minutes to googling "best schools in {area}", maybe even turning up a government sponsored list. If they ever notice the chosen school is bad, say from complaints of their ward or from bad grades, they would switch schools, improving the situation via negativa.

Googling it? The whole point of this thread is that that can be gamed. Do you want to reward schools for having the best SEO? Or for doing a better job of incentivizing people to review their school on whatever web site?

Google "best restaurants" and you'll find a list of decent restaurants. They may not be the best restaurants, but they also won't suck. That brings us par with the status quo, from 15 minutes of work.

It only takes one meal to know if a restaurant is good. It takes a lot more than that to know if a school is good.

I've often thought it would be a great idea to allow for parents to work one day a month at their child's school. This would put more parents in the circle. I think it would take time and some training to get to a good place on this, but would allow for a lot more insight and probably increase interaction dramatically.

One or two more sets of eyes in most classrooms each day, more people to help answer questions, etc. Far fewer WTF moments with some schools.

This is a very obvious argument for privatization of any industry, but check out my response in this same comment thread. Consumers OBVIOUSLY have skin in the game, but asking a consumer to figure out which charter schools are great and which are frauds and which are merely mediocre requires far more research than the average parent has time to do. The incentive for a charter school to misrepresent its achievements is very high, and once a parent has committed to putting their child in that school, the cost of moving them to another if/when they discover problems is very high. By that point, the charter school will have already taken the voucher funding and can use it to advertise to the next sucker.

AFAICT, a public school has the same incentives and has even higher switching costs (approximately 6% of the value of your house, whatever transaction fees your local government tacks on, plus whatever it costs to pack up and move).

Charter schools can be for-profit enterprises, or can be structured in such a way that executives can make a lot of money (even if they are technically non-profits). This is what skews the incentives for charter schools versus public schools.

There a phds in education who are trying to figure out which schools do the best job teaching using more information and statistical information than a parent will have in their entire lives. And they have trouble figuring this out, how the hell is some random parent supposed to figure this out?

Because “consumer choice” is one of those libertarian ideas that’s fueled entirely by theory and ideology and not by data.

An individual parent doesn't figure this out on their own. The market, composed of all parents, determines it. The market has access to more information than any phd will ever be privy, because it knows the individual feelings of every parent, child, teacher, and administrator. The market is unique in being able to perform randomized, controlled, double-blinded studies on the whole population, whereas phds are stuck looking at correlational data of small samples. The market performs continuous replication of every result ever identified, such that subgroups can be made arbitrarily small while maintaining significant p-values (and has no publication bias). And the market has access to the intellect of every phd, bachelors degree, or idiot that read a book once who ever had children or who wants to make a quick buck, conspiring in the greatest peer-review in history.

Skin in the game doesn't make complex and often temporally distant utilities from sui generis non-repeated transactions any less opaque.

Why would mere time to research options lead to better research being done? My five year old has effectively infinite time to research, say, efficacy of cancer treatments compared to an overworked researcher in a clinical setting. In fact, let's just shift all science to people in nursing homes who need to fill their time anyway. Utopia awaits?

This sounds like a complete nonsense argument. Can you explain what your point is? Are you saying that it's impossible to accurately research anything? Or that, given infinite time, parents would be so bad at researching charter schools that they would never be able to divine the best choice anyway?

What do you think MY argument was? I suspect you misread.

I believe the gp is engaged in the bad-faith approach of taking your comment about the state having more time as being literally the only argument you're making, where it's of course obvious the state has more advantages than simple time (including trained personnel, institutional practices, etc, etc).

Sounds like it.

I really miss the good old days of Hacker News when there was literally none of that kind of nonsense.

Why is that a bad-faith approach to response? That's the argument being presented, and there are myriad reasons it's entirely without merit. If the argument is that the state has more advantages, they can also be presented one-by-one and weighed. Politics often fails precisely because it smashes far too many decisions into one do-or-die package. Why should we repeat that unproductive pattern here?

It's a bad-faith fallacy because the OP's point holds even if the playing field is level and the state doesn't have an advantage in any other component than the stated one (time). The bad-faith comment (yours) instead reversed it, implicitly claiming the state was negatively positioned with respect to factors other than time, and pretended that was the OP's argument.

The state is negatively positioned; it has poor-to-no feedback with regard to the personal outcomes of its implementations. Parents and students do, on an ongoing and daily basis. The idea that I "pretended" that this was anyone else's argument but my own seems completely absurd.

While I'm glad you're able to infer implications, it would be wonderful if you could take it a step further to find that an argument cannot be true if a necessary premises is false. That's not bad faith, that's logic.

You're not doing the bare minimum of even attempting to understand the purpose of my argument and instead taking the dumbest possible literal interpretation and then responding to that. That is the very definition of a bad-faith argument.

Governments can hire experts to determine standards by doing research, and then can apply those standards over time to the schools they regulate. They can also collect massive amounts of data and compare performance between schools, across time, determine means and check for outliers, etc. etc. etc.

My argument is that consumers don't have time to do any of that. They have time to do what they do with colleges: Read a limited number pamphlets, thumb through Newsweek's annual "best colleges" report, maybe spend a few minutes here and there checking data online.

We can, as a society, devote more resources to making sure the government regulation side of things improves. What we can't do is collectively give OURSELVES more time to do that research individually.

Oh no, I understand your argument full well. I just think it's extremely shallow. Parents and students have plenty of time to research important aspects of their lives, starting with the experiential reality of the school they're subjected to for several years. One of your argument's shortcomings is the implication that only an a priori decision making process is possible for the over-a-decade process of going to school.

I would argue that these experiential data are more valid than metrics like standardized test scores by ethnicity or "letter grades" that the government gives itself for its school. Ultimately I question the motives behind many of these testing methodologies, because when they're not meant to satisfy the motivation of the student and/or their parents, the student (and their parents) become the product.

But even if parents could choose between multiple public school pedagogical methodologies, many can't simply because it requires selling the family's house and relocating to a different district. It's hard to have time to impact a situation when the situation is already fixed before the fact, but of course then it's not a time issue at all.

Ultimately I was responding to a two-sentence comment from you. I don't think complaining that my response was "bare minimum" is an appropriate way to handle the challenge, and I find the resort to name-calling about the "age of 4chan" particularly off putting.

"Oh no, I understand your argument full well. I just think it's extremely shallow."

Maybe you should poke around the rest of the comments for the article, because I've gone into a lot of detail about it.

"Parents and students have plenty of time to research important aspects of their lives"

I've addressed this elsewhere in the thread, but this assumption penalizes the children of parents who are poor, working multiple jobs, or dealing with health issues. It penalizes single parents as well. Basically, if you put the onus on parents to do the research, parents with more disposable time will make better decisions for their children, thus exacerbating the existing inequalities in the system. The whole point of the public education system is to level the playing field; what you're suggesting will increase the gap between the haves and the have-nots.

"...starting with the experiential reality of the school they're subjected to for several years. One of your argument's shortcomings is the implication that only an a priori decision making process is possible for the over-a-decade process of going to school."

What you're suggesting is that it's okay to have a system where parents and students make the decision about whether or not schools are good or bad by spending time at them. That is a ludicrously impractical suggestion. What happens if your kid hates their first day of school? Their first week? Their first month? How do you decide when they can make decent decision about it? What if they just hate going to school? Are you going to go sit in the classroom and watch the teacher for hours or days at a time?

Experiential data is not particularly useful on an individual basis; There are so many kids who just have a bad time at school no matter what, often for reasons that don't implicate the school directly as good or bad. And there are plenty of kids who have a great time at school precisely because they're not getting a very good education. (IE they're ignored, the work is too easy, etc.) Their experience is going to be fantastic, and they will enter college completely unprepared.

I'm not arguing that the EXISTING methods for measuring school performance are perfect; I am arguing that theoretically, the best possible version is mass-collection and analysis of data. Relying on individual decisions based purely on experience is a recipe for chaos; Kids would be begging their parents to change teachers and schools all the time, and how would parents have any idea if the kids were right?

I didn't complain that your response was the bare minimum, I complained that it was LESS than the bare minimum. And my 4chan comment was about the downvoting, not about your comment. Please read more carefully.

As these things tend to do, the issues balloon and it becomes impractical to respond to every point.

The idea that parents who don't care are penalized is simply a fact of life. Should we make the same argument about diet— that the state has the resources to appoint Top Dietitians and mandate a uniform State Diet? In fact, the experts that the state has hired have been corrupted by players like the corn and dairy lobbies and the nutritional advice they've given in the past has been roundly derided as terrible. Further, obesity has gotten worse since that advice has been given. Why would we think that the same state mandating institutional assignment would produce better results? Apparently, according to you, because other State education departments in other cultures and parts of the world have had comparatively more success by state-reported metrics? Again, I'm unimpressed at this justification for property taxes.

"What you're suggesting is that it's okay to have a system where parents and students make the decision about whether or not schools are good or bad by spending time at them"

In part, Yes! The way every other cooperative voluntary arrangement works! Do you even have children? I do. I have had babysitters that have made me uncomfortable on the first day. Yes, I have fired them and made the necessary decisions for my children. Why you think I should be legally prevented from doing that in the case of decades of dubious-quality education is a complete mystery to me at this point.

Experiential data are extremely useful on individual bases. Children who enjoy school because they don't have to do anything have been taught that education is painful. Anyone without that supposition would simply be bored, which would be a negative complaint rather than a positive review. The fact that we culturally accept common utterances like "school sucks" should tell you everything you need to know about a one-size-fits-all approach to educating a population. The argument here is to walk back the do-this-or-lose-your-house force-based approach and move toward a competitive and innovative approach toward improving individuals' quality of life. Your argument against this seems to be that, while it hasn't worked here, it theoretically (under some theory you've yet to elucidate) could, because infinite resources by fiat magically translate to infinite productivity and efficiency in real life. I don't think you've just done a poor job at defending that claim, I think you've simply done the best job you could of defending a bad idea.

Now what's that they say about ideas being so good the have to be mandatory?

Look, I can see that this is coming down more to an ideological and theoretical argument than a fact-based one. You dismissed any evidence that other countries have done this better out of hand because in theory those statistics can be corrupted, and you've dismissed the idea that state experts could make the system better because they could also be corrupted.

This is the core conservative argument for abolishing just about every government service: "We screwed it up in the past. Rather than trying to fix our past mistakes, we should instead give up."

And I am well aware of the problems with the "food pyramid." Your example here ignores the fact that the government has since acknowledged that the original recommendations were wrong, and has been updating them over time to reflect the best research. The argument that "the process can be corrupted therefore we should give up" ignores that it can also be improved over time, and that governments are accountable.

"Your argument against this seems to be that, while it hasn't worked here, it theoretically (under some theory you've yet to elucidate) could, because infinite resources by fiat magically translate to infinite productivity and efficiency in real life."

I did provide evidence that it could work, which was that it has worked in numerous other countries. You just chose to dismiss that evidence out of hand because you'd rather come up with an excuse to disbelieve it rather than accept that a system you fundamentally disagree with can work. Whether you mean it to or not, it comes off as utterly and completely bad-faith argument to dismiss the evidence I provided out of hand and then tell me several paragraphs later that I haven't adequately described how it can be done. I pointed you in the direction of a perfect example, and yet apparently you haven't even checked to see if it's a valid example.

The short version is this: Finland had a crappy education system, and they made a decision as a country to fix it. They approached it with open minds, willing to throw out any ideas that were outdated and not backed up with evidence. What they ended up with was little to no homework, fresh-made school lunches, and the abolition of private schools. (The reason for the last bit is that as long as wealthy people can use private schools as an escape hatch for their own kids, they have zero political incentive to fix the public schools.)

You mentioned in another comment that you would absolutely accept a system that rewarded parents who spent more time researching schools, and I think this is where we fundamentally disagree. You want a system that allows people succeed if they have the time and resources to do so. I want a system that does its best for all people, even though that takes a lot more work and coordination.

Now, I'm not exactly sure what your motivation for this is. Pessimistically, it looks like a very selfish line of reasoning: Since you have the time and resources and think you're smart enough to choose wisely, you want the choice to do so, and if other people don't have the same opportunities then too bad for them. Or, more optimistically, you just can't imagine a system that works adequately the way I describe, so you just think your way is safer.

Frankly I think you're either misunderstanding or misrepresenting my positions. We have been arguing from efficiency right along, but you're now shifting to an argument from morality. You posit that it's acceptable for the state to take ownership of its citizens (subjects at that point?) Ultimately that's a position I have to entirely reject. There comes a point where the removal of autonomy to improve a person's life deprives them of what it means to be alive. Your elitist position that you and your experts know so much more than the little people that you can't "allow" them to make their own choices comes off rather poorly. It also comes off as rather short sighted, as politics is often inherently unstable, and all it takes is a "bad politician" to take your well-meaning but authoritarian system and take it full-blown authoritarian.

I appreciate your extrapolation on your ideas. While I'm sure neither of us have changed our minds, I find your position interesting and I'm sure others reading have been able to see a small amount of the breadth of the topic at hand.

You originally make a completely bad-faith that compares the state to a two year old. Showing a more developed argument now, regardless of its truth, doesn't make the original argument better.

I don't agree with your developed argument either but that's a different matter.

Although "Governments can hire experts to determine standards by doing research, and then can apply those standards over time to the schools they regulate.", those supposed experts can be pretty bad.

Richard Feynman relates a fascinating account of this in his book "Surely You're Joking, Mr. Feynman!". Aside from him, every "expert" gave good scores to books that were entirely blank.

Some quotes from that: https://fs.blog/2016/07/richard-feynman-teaching-math-kids/ http://whyhomeschool.blogspot.com/2006/08/richard-feynman-on...

The argument that “we could screw this up” is not the same as “we will screw this up and therefore should not do it.”

Sounds like they're saying that poor parents are too stupid to know the difference.

Consumers are the ones paying for it. So I see nothing wrong with providing what they want.

M4A is a 10~15% GDP expenditure. Way more than vouchers.

> Anybody up for a grand bargin? I'll trade you Medicare-for-All in exchange for a complete privatization (w/ vouchers) of education.

It's not only a bad deal, but you don't have the power to deliver your end of the bargain, and the people that would have the power are at least as opposed to M4A as Conventional public schools, so no, it's not worth considet on many levels.

And, anyway vouchers aren't really privatization, it is public with either reduced accountability or increased overhead or both, depending on the conditions attached to the vouchers, which are just a mechanism for outsourcing delivery of a public function not actually privatizing the function so that government is not involved.

> I'll trade you Medicare-for-All in exchange for a complete privatization (w/ vouchers) of education.

OK, how fucked up do you want to allow the unlucky kids to become?

I mean, it's a given that in a Pure True Private Education System, anti-vaccination propaganda will be unchallenged dogma in some schools. It's a big deal now and this will just remove the few impediments it has to becoming the majority opinion in some regions. Medicare-For-All means precisely shit if the people in that region think modern medicine is pharmashill toxins gunna make you an NWO Zionist zombie-slave.

But that isn't even the worst of it. Public school is, at the very least, a way to get kids out of the house and into an environment where it isn't normal for parents to fuck their children. Teachers are mandatory reporters for a reason. I'm not saying that public schools are perfect at detecting child abuse, but completely privatizing the whole thing will shove that effectiveness right off a cliff. Don't piss off the people paying your bills, eh?

We already have segregation academies, but complete privatization will remove all choice in some regions, in that only the segregation academies will capture enough vouchers to be effective educators. Well, if you consider "The David Duke School of Racial Tolerance" to be effective education.

Would it still actually be "private" with government vouchers in the mix?

If it's delivered by private companies, yes. In the same way, General Motors is still a private entity even if it sells cars to the government.

The point is that you have a bunch of independent entities competing with each other. You don't have that with a government school district.

That is a somewhat unrepresentative example. There is a big difference between a company that sells a few things to the government vs a company where 90%+ of revenue ultimately comes from the government.

Given the way most things seem to eventually consolidate, the latter scenario would be the more probable one--a WalMart of education so to speak.

In that case, if the government is the one signing the checks, it is also in a very strong position to dictate how the whole thing is operated. Not that it will--see WalMart again--but it could.

So, the worst impulses of both extremes? Where do I sign?

It's not even clear what this modest proposal has to do with the article, which is written from the premise that the whole enterprise of college-oriented secondary schooling is misguided, that no matter how well-taught the subjects, secondary school educational objectives are about signaling --- to colleges --- and not human capital, and that there's even more broad acceptance of the notion that college is itself about signaling to the labor market.

How would privatizing schools address this person's concerns?

You would get charter schools that emphasized signalling. You might get some that emphasized human capital. That would leave parents able to choose - if they actually perceived what was going on.

As it is, though, if I want actual human capital in my child's education, it's either private school or homeschooling. To the degree (if any) that private schools emphasize human capital over signalling, I'm not sure that privatizing schools plus vouchers would make them change to emphasizing signalling. So I don't think the situation would become worse, and it might become better.

The biggest problem with education is that who the kid is(smart, hard working, resilient) has 20x the impact of his education. So it's much easier to select 5% better students than it is to teach them 100% better. And if teams of phd researchers are arguing over whether colleges in general are signalling versus human capital, how are some random parents going to be to solve the much harder problem of if a single school's results are due to signalling or human capital with less time, training, and information?

This letter is making some pretty superficial observations that I don't think hold up.

The background you need for the letter is the debate between "human capital" (H.C.) and "signaling" in education; H.C. is a longhand for "intrinsically valuable", and signaling is a shorthand for "indirectly valuable by improving the student's reputation".

First, the author (correctly, I think? I suck at math but this sounds correct) believes that AP Stat would be better for H.C. than AP Calc, which is mostly about signaling. School districts have to decide which courses to carry, and the author presumes they carry AP Calc to minmax for prestige college admissions.

Second, the author observes that schools have varying levels of rigor; a better, rigorous school might give a C where a weaker, more generous school would give a B. The weaker school produces graduates with higher GPAs, again minmaxing for college admissions over H.C.

Problems with both observations:

1. It's not clear that AP Stat disadvantages students versus AP Calc. The author didn't make the argument that AP classes in general optimize for signaling --- that argument seems harder to defend. So why don't high schools sub out Calc for Stat? "Because colleges would have a problem with that" doesn't seem like the answer. So what are we left with? "There are better high school math curricula than the ones we have now?" Lots of people already agree with that.

2. Selective colleges already account for the differing rigor of different high schools. A 3.4 at one school is not necessarily as valuable as the 3.4 of every other school.

I think you've misunderstood. I believe the H.C. argument is that statistics are far more valuable in real life than Calculus, but Calculus is more heavily valued as an "advanced class" for signaling purposes.

> First, the author (correctly, I think? I suck at math but this sounds correct) believes that AP Stat would be better for H.C. than AP Calc, which is mostly about signaling.

It seems you agree with 'tptacek

> 1. It's not clear that AP Stat disadvantages students versus AP Calc.

This is what stuck out to me.

Isn't AP Calculus the obvious choice if you have to choose between AP Stats and AP Calculus?

1. Calculus is really, really useful. Maybe AP Stats is also useful, but singling out Calculus as an example of useless signalling sets off really loud alarm bells.

2. If you want a STEM/Engineering degree, at least one Calculus course is required. More importantly, often a long sequence of 2-4 courses (Calculus I, II, III and ODEs) are required. Because of that long list of sequentially dependent required courses, getting one or two calculus courses out of the way in high school is enormously useful (like, "graduate a semester earlier for each course" useful). AP Stats is not at the head of this sort of long sequential course dependency.

3. The AP Stats course has a major disadvantage: lots of colleges/majors that require a stats course don't accept AP Stats as credit because they require a calculus-based statistics course.

It really depends on the field. In many branches of science, statistics is really important and calculus not really at all. I'm a biomedical scientist and I use statistics every day to interpret the results of experiments. Calculus, not so much. And yet I was required to take three semesters of calculus as an undergrad, and while I took a statistics course then (and more in grad school), that was optional. It really should be the other way around.

> 1. Calculus is really, really useful. Maybe AP Stats is also useful, but singling out Calculus as an example of useless signalling sets off really loud alarm bells.

34 years old. Been writing software since I was 14. Used to know some calculus and sometimes poke at picking it back up because I feel like I "ought to". Have usually been the one to tackle tough or odd problems where I've worked.

Haven't once managed to find a reason to use calculus for anything whatsoever. Not a damn thing to differentiate, not a damn thing to integrate. I think the need for it is in a very, very narrow slice of all jobs, even in "STEM".

Statistics is 100% for sure more useful to me, in everyday life and at work. And I've not even worked on anything especially stats-ish, it just happens to come up a lot. That's the thing I really ought to work to get better at.

While people often don't have to do the explicit differentiation and integration taught it calculs that doesn't mean you aren't using knowledge learned in a calculus class.

I often think about rates flows and how they relate to each other. For example to do monitoring we count the total number of messages received in each task at a given time, then use the diffrence to get a rate. That is basically differentiation and understanding calculus makes it easier.

The other thing you acquire from doing Calculus is problem solving experience, figuring out how to apply various approaches to a problem to get to a desired solution. In this way calculus strengthens your brain similar to how a sports player might do weight lifting even if their sport doesn't involve lifting heavy objects.

I think lots of people understand rates without doing calculus. For instance distance vs velocity vs acceleration is a pretty simple and intuitive idea to understand.

There is an enormous amount of evidence that weight lifting transfers to sports. But I don't think there is much if any evidence that calculus transfers to general problem solving.

> 34 years old. Been writing software since I was 14.

We have very similar profiles.

I use calculus every day.

> Statistics is 100% for sure more useful to me, in everyday life and at work.

I use statistics every day too, working on ML systems.

> That's the thing I really ought to work to get better at.

The calculus comes in pretty damn quick. See also the third point in my original post; concretely, every ML course I've ever seen requires Calculus as a pre-requisite.

The Library of Economics, funded by the Liberty Fund[1], is part of the right-wing Atlas Network[2] of highly ideological and often flat out disinformational propaganda institutions.[3]

Take anything written with tons of salt.

________________________________

Notes:

1. https://www.econlib.org/library/About.html

2. https://www.atlasnetwork.org/partners/global-directory/liber...

3. https://www.sourcewatch.org/index.php?title=Atlas_Network

I don't disagree, but this is a simple point and all the math language doesn't really add anything to it.

One solution to the problem of schools ranking kids with a B average versus a C average is simply to not let individuals set the grades at all. Have a national set of standard exams (as is done in the UK and many other countries).

This doesn’t solve the issue completely - schools turn into “SAT prep centers” to an extent - but it definitely helps compare students and teaching across multiple schools.

This is already happening in the US... many schools are teaching for the testing. Those that come up with the standards are making bank providing course materials and publishing. They publish disposable workbooks instead of textbooks to ensure they change each year and require new purchases.

I'm surprised more districts, teachers and schools haven't worked to create more open curriculums.

Or even better have colleges administer their own exams. Oral ones are best ;) this is how it worked in USSR. Nobody cared what marks you had in high school as village school and good city schools were completely different.

You came to exam and had 3 hours and 5 problems(example of physics entrance exam to KGU) you then went and presented your solution if you had any to examiners. They could grill you to explain your solution

Another solution is no traditional grades at all. Introduce standards based grading (tldr: grading if a student met the learning standard or not).

The most obvious example of this game theoretic problem in teaching is the phenomena of "teaching to the test". Focusing on preparing students to generate the right signal, rather than laying the foundation for long-term comprehension of the material.

From the article: "Schools of standard C produce graduates who are more ready for college. Schools of standard B produce graduates who appear more college ready." I just left a job teaching community college in Iowa, who is now boasting the nation's highest high school graduation rate. Having taught community college in a different state, I can honestly say that many, if not most, community college students I saw were just not ready for college. Doesn't matter what the graduation rate, or even GPA was. They were just not ready for the responsibility. Far too many could not even be bothered to read directions.

Doesn't this overlook the fact that there are standardized tests, which allow colleges to ascertain the relative grade inflation at various high schools? Also, they are repeat players, which further mitigates the information asymmetry.

Then they add offsets based on race and income disparity.

The title is misleading (it was meant to be intriguing, but doesn't convey the summary of the article well).

The letter, in fact, is a well-balanced rebuke to Caplan's treatise "The Case Against Education".

The tl;dr of the treatise is that schools manufacture signaling instead of value, with the implication that they inherently cannot produce value.

This rebuke is that the schools produce signaling because they are incentives are aligned this way. Producing signal vs. value becomes a prisoner's dilemma, where it's rational to focus on manufacturing signal.

Ergo, the schools and educators are not to blame. Hate the game, don't hate the player.

I agree with the rebuke.

This is a rebuke only to the extent that "the emperor enjoys the breeze" is a rebuke to someone pointing out that the emperor is naked.

Not quite.

It's like someone someone is pointing out that the emperor is naked, and the rebuke is "Have you considered that we are all at a beach, and invited the Emperor to go skinny dipping with us".

Merely pointing out that the emperor is naked, without sufficient consideration for context, implies that there is something terribly wrong with the emperor. You'd think that the guy just walks around naked!

Now with context, you understand that if you expect the Emperor to be clothed, then perhaps you shouldn't have spent the whole week convincing him to go skinny-dipping with you, and then act surprised when you are both at a beach. The fault, if any, is with what you asked of the Emperor.

So here we are with education. We tell our kids to do well in school so that they could go to a good college and get a good job (a chain that revolves around signaling!), measure the education by signaling it provides (acceptance rates, employment rates out of college), and then insert surprised Pikachu write books about how education is a waste because it's about signaling!

I think many if not most people involved in academia would strongly disagree that it teaches you nothing. Caplan's thesis was certainly not widely accepted prior to publication.

> with the implication that they inherently cannot produce value.

Where in Caplan's book does it say this? To the contrary, I believe he points to specific cases where we know schools could produce more value (e.g., spaced repetition) but do not.

Reply on news.ycombinator.com