Open threads at the Open Thread tab every Sunday and Wednesday

Against Tulip Subsidies


Imagine a little kingdom with a quaint custom: when a man likes a woman, he offers her a tulip; if she accepts, they are married shortly thereafter. A couple who marries sans tulip is considered to be living in sin; no other form of proposal is appropriate or accepted.

One day, a Dutch trader comes to the little kingdom. He explains that his homeland also has a quaint custom involving tulips: they speculate on them, bidding the price up to stratospheric levels. Why, in the Netherlands, a tulip can go for ten times more than the average worker earns in a year! The trader is pleased to find a new source of bulbs, and offers the people of the kingdom a few guilders per tulip, which they happily accept.

Soon other Dutch traders show up and start a bidding war. The price of tulips goes up, and up, and up; first dozens of guilders, then hundreds. Tulip-growers make a fortune, but everyone else is less pleased. Suitors wishing to give a token of their love find themselves having to invest their entire life savings – with no guarantee that the woman will even say yes! Soon, some of the poorest people are locked out of marriage and family-raising entirely.

Some of the members of Parliament are outraged. Marriage is, they say, a human right, and to see it forcibly denied the poor by foreign speculators is nothing less than an abomination. They demand that the King provide every man enough money to guarantee he can buy a tulip. Some objections are raised: won’t it deplete the Treasury? Are we obligated to buy everyone a beautiful flawless bulb, or just the sickliest, grungiest plant that will technically satisfy the requirements of the ritual? If some man continuously proposes to women who reject him, are we obligated to pay for a new bulb each time, thus subsidizing his stupidity?

The pro-subsidy faction declares that the people asking these question are well-off, and can probably afford tulips of their own, and so from their place of privilege they are trying to raise pointless objections to other people being able to obtain the connubial happiness they themselves enjoy. After the doubters are tarred and feathered and thrown in the river, Parliament votes that the public purse pay for as many tulips as the poor need, whatever the price.

A few years later, another Dutch trader comes to the little kingdom. Everyone asks if he is there to buy tulips, and he says no, the Netherlands’ tulip bubble has long since collapsed, and the price is down to a guilder or two. The people of the kingdom are very surprised to hear that, since the price of their own tulips has never stopped going up, and is now in the range of tens of thousands of guilders. Nevertheless, they are glad that, however high tulip prices may be for them, they know the government is always there to help. Sure, the roads are falling apart and the army is going hungry for lack of rations, but at least everyone who wants to marry is able to do so.

Meanwhile, across the river is another little kingdom that had the same tulip-related marriage custom. They also had a crisis when the Dutch merchants started making the prices go up. But they didn’t have enough money to afford universal tulip subsidies. It was pretty touch-and-go for a while, and a lot of poor people were very unhappy.

But nowadays they use daffodils to mark engagements, and their economy has never been better.


In America, aspiring doctors do four years of undergrad in whatever area they want (I did Philosophy), then four more years of medical school, for a total of eight years post-high school education. In Ireland, aspiring doctors go straight from high school to medical school and finish after five years.

I’ve done medicine in both America and Ireland. The doctors in both countries are about equally good. When Irish doctors take the American standardized tests, they usually do pretty well. Ireland is one of the approximately 100% of First World countries that gets better health outcomes than the United States. There’s no evidence whatsoever that American doctors gain anything from those three extra years of undergrad. And why would they? Why is having a philosophy degree under my belt supposed to make me any better at medicine?

(I guess I might have acquired a talent for colorectal surgery through long practice pulling things out of my ass, but it hardly seems worth it.)

I’ll make another confession. Ireland’s medical school is five years as opposed to America’s four because the Irish spend their first year teaching the basic sciences – biology, organic chemistry, physics, calculus. When I applied to medical school in Ireland, they offered me an accelerated four year program on the grounds that I had surely gotten all of those in my American undergraduate work. I hadn’t. I read some books about them over the summer and did just fine.

Americans take eight years to become doctors. Irishmen can do it in four, and achieve the same result. Each year of higher education at a good school – let’s say an Ivy, doctors don’t study at Podunk Community College – costs about $50,000. So American medical students are paying an extra $200,000 for…what?

Remember, a modest amount of the current health care crisis is caused by doctors’ crippling level of debt. Socially responsible doctors often consider less lucrative careers helping the needy, right up until the bill comes due from their education and they realize they have to make a lot of money right now. We took one look at that problem and said “You know, let’s make doctors pay an extra $200,000 for no reason.”

And to paraphrase Dirkson, $200,000 here, $200,000 there, and pretty soon it adds up to real money. 20,000 doctors graduate in the United States each year; that means the total yearly cost of requiring doctors to have undergraduate degrees is $4 billion. That’s most of the amount of money you’d need to house every homeless person in the country ($10,000 to house one homeless x 600,000 homeless).

I want to be able to say people have noticed the Irish/American discrepancy and are thinking hard about it. I can say that. Just not in the way I would like. Many of the elder doctors I talked to in Ireland wanted to switch to the American system. Not because they thought it would give them better doctors. Just because they said it was more fun working with medical students like myself who were older and a little wiser. The Irish medical students were just out of high school and hard to relate to – us foreigners were four years older than that and had one or another undergraduate subject under our belts. One of my attendings said that it was nice having me around because I’d studied Philosophy in college and that gave our team a touch of class. A touch of class!

This is why, despite my reservations about libertarianism, it’s not-libertarianism that really scares me. Whenever some people without skin in the game are allowed to make decisions for other people, you end up with a bunch of elderly doctors getting together, think “Yeah, things do seem a little classier around here if we make people who are not us pay $200,000, make it so,” and then there goes the money that should have housed all the homeless people in the country.

But more important, it also destroyed my last shred of hope that the current mania for requiring college degrees for everything had a good reason behind it.


The only reason I’m picking on medicine is that it’s so clear. You have your experimental group in the United States, your control group in Ireland, you can see the lack of difference. You can take an American doctor and an Irish doctor, watch them prescribe the same medication in the same situation, and have a visceral feel for “Wait, we just spent $200,000 for no reason.”

But it’s not just medicine. Let me tell you about my family.

There’s my cousin. He wants to be a firefighter. He’s wanted to be a firefighter ever since he was young, and he’s done volunteer work for his local fire department, who have promised him a job. But in order to get it, he has to go do four years of college. You can’t be a firefighter without a college degree. That would be ridiculous. Back in the old days, when people were allowed to become firefighters after getting only thirteen measly years of book learning, I have it on good authority that several major states burnt to the ground.

My mother is a Spanish teacher. After twenty years teaching, with excellent reviews by her students, she pursued a Masters’ in Education because her school was going to pay her more money if she had it. She told me that her professors were incompetent, had never actually taught real students, and spent the entire course pushing whatever was the latest educational fad; however, after paying them thousands of dollars, she got the degree and her school dutifully increased her salary. She is lucky. In several states, teachers are required by law to pursue a Masters’ degree to be allowed to continue teaching. Oddly enough, these states have no better student outcomes than states without this requirement, but this does not seem to affect their zeal for this requirement. Even though many rigorous well-controlled studies have found that presence of absence of a Masters’ degree explains approximately zero percent of variance in teacher quality, many states continue to require it if you want to keep your license, and almost every state will pay you more for having it.

Before taking my current job, I taught English in Japan. I had no Japanese language experience and no teaching experience, but the company I interviewed with asked if I had an undergraduate degree in some subject or other, and that was good enough for them. Meanwhile, I knew people who were fluent in Japanese and who had high-level TOEFL certification. They did not have a college degree so they were not considered.

My ex-girlfriend majored in Gender Studies, but it turned out all of the high-paying gender factories had relocated to China. They solved this problem by going to App Academy, a three month long, $15,000 course that taught programming. App Academy graduates compete for the same jobs as people who have taken computer science in college, a four year long, $200,000 undertaking.

I see no reason to think my family and friends are unique. The overall picture seems to be one of people paying hundreds of thousands of dollars to get a degree in Art History to pursue a job in Sales, or a degree in Spanish Literature to get a job as a middle manager. Or not paying hundreds of thousands of dollars, if they happen to be poor, and so being permanently locked out of jobs as a firefighter or salesman.


So presidential candidate Bernie Sanders has proposed universal free college tuition.

On the one hand, I sympathize with his goals. If you can’t get any job better than ‘fast food worker’ without a college degree, and poor people can’t afford college degrees, that’s a pretty grim situation, and obviously unfair to the poor.

On the other hand, if can’t you get married without a tulip, and poor people can’t afford tulips, that’s also a pretty grim situation, and obviously unfair to the poor.

But the solution isn’t universal tulip subsidies.

Higher education is in a bubble much like the old tulip bubble. In the past forty years, the price of college has dectupled (quadrupled when adjusting for inflation). It used to be easy to pay for college with a summer job; now it is impossible. At the same time, the unemployment rate of people without college degrees is twice that of people who have them. Things are clearly very bad and Senator Sanders is right to be concerned.

But, well, when we require doctors to get a college degree before they can go to medical school, we’re throwing out a mere $5 billion, barely enough to house all the homeless people in the country. But Senator Sanders admits that his plan would cost $70 billion per year. That’s about the size of the entire economy of Hawaii. It’s enough to give $2000 every year to every American in poverty.

At what point do we say “Actually, no, let’s not do that, and just let people hold basic jobs even if they don’t cough up a a hundred thousand dollars from somewhere to get a degree in Medieval History”?

I’m afraid that Sanders’ plan is a lot like the tulip subsidy idea that started off this post. It would subsidize the continuation of a useless tradition that has turned into a speculation bubble, prevent the bubble from ever popping, and disincentivize people from figuring out a way to route around the problem, eg replacing the tulips with daffodils.

(yes, it is nice to have college for non-economic reasons too, but let’s be honest – if there were no such institution as college, would you, totally for non-economic reasons, suggest the government pay poor people $100,000 to get a degree in Medieval History? Also, anything not related to job-getting can be done three times as quickly by just reading a book.)

If I were Sanders, I’d propose a different strategy. Make “college degree” a protected characteristic, like race and religion and sexuality. If you’re not allowed to ask a job candidate whether they’re gay, you’re not allowed to ask them whether they’re a college graduate or not. You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail. I realize this is a blatant violation of my usual semi-libertarian principles, but at this point I don’t care.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

748 Responses to Against Tulip Subsidies

  1. Chris Billington says:

    I totally thought that this was going to be about housing.

    So a bit off topic, but it’s a strange situation that everyone who has a house wants the prices to go up, and everyone who doesn’t wants them to go down. And because everyone needs to live somewhere, it’s in the government’s interests to give tax breaks to first home buyers and such. But because so many people have a big chunk of their net worth tied up in their houses, it’s in the government’s interest to keep prices from falling much, lest people are poorer.

    I’ve only recently realised how bizarre this is. It’s terrible that we have a speculative market based around something that everyone actually needs.

    At least with the tulips/degrees you can change the culture – you can’t just stop people needing houses.

    • Pku says:

      You can (under some systems, at least) get around this by living in trailers and such, which (I don’t think) are subject to the same speculative market forces. Hence poor people living in trailer parks? (This sounds right, but I don’t actually know enough economics to be sure).

      • Trailers lose out in the long run because they’ve got horrid maintenance and utility costs (barring really high-end trailers, but if you’re looking for a trailer in order to save on a home then those aren’t the kind that you’re looking at).

        • f says:

          What Gauvreau says is not true according to many people who live in trailers to save money. At least that’s what I learned reading frugality websites.

          • kernly says:

            It sounds utterly nonsensical. “Maintenance and utility” costs are obviously dwarfed by mortage/rent costs…

          • Gbdub says:

            “Smaller than”? Probably. “Dwarfed by”? Not at all.

            I live in an average sized, site built, 15 yr old single family home in Arizona, and my utilities (gas, electric, water/waste, HOA fees, not including cable/phone/data) are typically about 30% of my 30-yr fixed rate mortgage payment. Maintenance probably averages out to about another 10-15%, but often comes in big chunks (e.g. had to replace a blown water heater, which was >$500 plus another $1000 insurance deductible for the water damage, though I hopefully won’t need to do that again for 15 years).

            Now some of that would obviously be cheaper in a mobile home – less house to cool, less landscape to water – but a lot of those costs don’t scale linearly. Mobile homes have more surface area per volume and are generally less well insulated, so there’s a higher heating/cooling cost per sq. ft. Household water and electric is driven more by how many people live there than the size. Waste removal is a fixed rate per house. Plumbers, electricians, and pest control services charge the same rate per hour regardless of house size.

            So I can easily see maintenance and utilities approaching or exceeding costs of rent/mortgage on a mobile home, even more so if it’s shoddy. A renter might not see that cost directly if all or part of maintenance and utilities are included, but the cost is still there.

          • Steve says:

            Gbdub, “dwarfed by” is accurate in the sense that matters. Maintenance, utility, *and* mortgage/rent for a trailer, all put together, are dwarfed by the mortgage/rent for a small single-family house.

          • Gbdub says:

            Since there’s been some confusion in this thread, I’ll clarify that I’m referring to “mobile home” in the American sense of a usually cheap, manufactured home constructed offsite and moved to a basically permanent site at a “trailer park”, not the European sense of a Traveler’s trailer or RV.

            For those, I think my point still stands, and I should clarify my logic:
            1) maintenance and utilities make up a substantial fraction of my cost to own my site-built single family home.
            2) a single family manufactured home will have many of the same maintenance and utility costs.
            3) thus, unless rent is negative, utilities and maintenance set a “floor” for the cost of a single family dwelling that is still a substantial fraction of the cost of my site built home , not “dwarfed by”.

            Anecdotally, this seems to bear out – folks in trailer parks aren’t so much killed by the rent, it’s the unplanned $200 plumber or roofer or car repair or whatever bill that gets you, since their cash reserve is usually low.

        • Megafire says:

          Sounds like a case of Vimes’ Boots Theory of Economics.

          • William O. B'Livion says:

            Which, BTW may have been true up through the early 1980s, but modern manufacturing techniques and materials are such that PayLess shoes are often just as sturdy and long lasting as those purchased at much higher prices. Sometimes even sturdier.

          • Tom West says:

            I don’t know. My cheap PayLess running shoes usually last only 12 months or so, while on the occasion that I’ve bought expensive running shoes, they’ve often lasted a whole year.

        • William O. B'Livion says:

          Generally contemporary “trailers” lose out in the long run because in most trailer parks you do not own the land.

          Modern “trailers” are more correctly called “manufactured homes”, and are built according to federal, rather than local building codes[generalization]. This doesn’t make them any better, or worse, just some minor differences.

          Modern manufactured homes are generally more sturdy than traditional stick built homes because they have to withstand being moved over highways.

          They have generally about the same utility costs as a traditional home *at the time they were built*, and maintenance costs are generally in favor of the manufactured home because they’re 2 feet off the ground and easier to get under.

          Now, these statements are about contemporary “mobile homes”, and not older stuff (technically a “mobile home” was made prior to 1976 and you can’t (or have a hard time) getting a home mortgage on them (often you would have to get a vehicle loan (???) on them as they were titled as trailers). Homes built after 1976 meet FHA standards and you can get mortgages on them at relatively normal rates (another reason for the historic cost difference leaning in favor of a stick built home over time).

          As in anything there are a range of options, but having to meet base federal means that many newer manufactured homes are *more* energy efficient than stick built, especially older stick built homes.

          In some parts of the US and Canada these buildings are treated differently for tax purposes depending on whether they’re on a permanent foundation, still on the chassis they were shipped on, or still have the wheels on them (no, SRSLY. In some places local/county taxes are levied on them as if they were *vehicles* if they still have the wheels on, and as homes if they do not. At least that was what was explained to me when I lived in one in Sunnyvale).

          In some parts of the county (especially down around San Jose) many people don’t have the option of buying a small house on a small piece of land, and buying a manufactured home on rented land might not be an investment (IIRC you generally don’t *make* money on the transaction), but it’s a way of having a little bit nicer home that *you* own, and breaking even or losing less over the long haul.

          On the “high end” of manufactured homes you can get net-zero homes or even net-positive buildings assembled with *very* nice features and furnishings in very custom layouts and dropped on a “real” foundation, or even basement. Things like Structured Insulated Panels (SIPs) and factory building (with assembly on site) reduces (or so the industry claims) a lot of waste and time lost to weather delays.

          If I was in a position to buy land and build a house it’s something I would seriously look at.

          Back in the 80s my father arranged to have a factory built home (not like todays “manufactured homes”, but sort of half way between a manufactured home and a traditional stick built structure). It was a sturdy building and the people that bought it from my family are still (at least as of a few years ago) living in it.

        • asdfoij says:

          Utility cost wise it should be really cheap due to the small spaces and the energy efficiency affordances for being on the road. And you have to have a direct electric hookup for that to work.

          Now insulation is something else on the other hand.

      • Deiseach says:

        Trailer parks are not ideal living situations. Trailers need to be, at least in theory, capable of being driven away so it’s not a fixed structure. And it depends on the park – if you get good landlords who provide decent facilities and keep it up, you’re okay, but if the landlord doesn’t give a toss, only cares about getting the rent, and will kick you out if you complain, where do you go?

        You may not own the trailer (you’ve been renting it from the landlord/owner) and even if you do, if you try parking it on unapproved sites, or by the side of the road, you get moved on – “illegal encampment” is an offence.

        Our council offers people in desperate need a demountable (basically, a trailer) but (a) this is not for homelessness which is a different matter, it’s for people whose own homes are uninhabitable (b) the costs of making the house habitable are prohibitively expensive or the person can’t afford to do it (c) there is a public health issue involved.

        Right now, we’re trying to find private rented accommodation for clients on our social housing list and who are also applying under the new HAP scheme. In a town of around ten thousand people, there is ONE private rented advertised. Literally one. There just is not the rental accommodation available for various reasons, and for various reasons we haven’t got social housing free – people come in assuming we can provide a house with a snap of the fingers and that is not how it works. With our existing stock, we only have vacancies when an existing tenant dies (and you wouldn’t believe how many phone calls we get quite literally the same day someone dies, asking if the house is available and can they have it please?) or when we build new housing stock, and we’re only getting funding and permission this year to start new builds.

        In the meantime – the homeless services have to take up the slack (insert hollow laughter here).

        Rent supplement allowances and similar schemes operate on a rent cap; that is, the state only pays a certain amount towards rent. Now that the economy is improving a little, those levels are completely unrealistic – when banks were foreclosing on houses because people couldn’t pay mortgages, landlords were happy to take tenants at any price in order to generate income to pay their mortgage.

        Now, though, once leases run out, the landlords are upping the monthly rent by a hundred euro or more, or not renewing leases, because the property market means they can sell the house now and it won’t be at a loss. People are not, as in the boom years, buying their own houses and so freeing up rental properties that way. People in rented accommodation are staying there and there is no spare capacity at the moment.

        So it gets thrown onto the local government and as I’ve said, we don’t have the spare capacity either as yet. Now, if development is stimulated again and all the properties and landbanks (the ‘ghost estates’) that crashed during the boom are developed again, that will mean more housing capacity. But that depends on people getting the mortgages to buy new houses, and the banks have been heavily warned and censured not to go crazy on lending as that caused the crash, and the employment situation while getting better is not yet at the Celtic Tiger levels where there was plenty of good-paying employment and money galore to sustain large mortgages.

        • This isn’t the sort of “trailer” people are talking about when they talk about poor people living in trailer parks, at least in the US. There is no question of driving a trailer park trailer away or parking it by the side of the road unless you have a semi truck. The trailer home industry now refers to their products as “manufactured homes”, and has for quite a while, but people still seem to call them trailers or mobile homes.

          What you’re talking about in your comment applies much more to recreational vehicle (RV) parks. Of course, some trailer parks have some RVs in them, so there’s overlap.

          Anyway, the upshot is that mobile homes are really, really cheap on the low end, but don’t appreciate in value, and have a lower-class vibe.

          • Deiseach says:

            Yes, I am aware of the difference. That’s why I said “in theory”.

            Anything which is a permanent fixture requires planning permission; a mobile home which is fixed in place, or a “manufactured home”, will both be considered permanent residential structures and will come under the law.

            Even if it’s your cousin Ted letting you put up that modular dwelling in a field he owns, it will still require planning permission.

            Things may very well be different in America, but I’m fairly sure if there is any kind of park or other business where these kinds of modular buildings are set up and rented to tenants and have water/sewerage in place, then there are regulations to follow and taxes to be paid.

            And there are landlords who don’t want to declare their property as rental property because they’re not declaring rental income for tax purposes. We see that when tenants come in to us and can’t provide the rent book, tenancy agreement, etc. that by law they are entitled to have: “Oh, I pay that straight into the landlord’s account” or “No, he didn’t give me a contract, he doesn’t want to write anything down”.

          • Mary says:

            Fun fact: The “mobile” in “mobile home” is like the “French” in “French toast”, only more accurate, since they really did first come from Mobile Alabama.

            That is, it doesn’t mean “moveable.”

        • William O. B'Livion says:


          The older ones *are* mobile in that all of them were towed to the site, parked on a concrete slab and then raised/stabilized. Some older ones I’ve seen still have the towing hitch attached.

          There are also multiple cases of people moving them to new pieces of land or rented plots.

          As they’ve increased in size and complexity this has become less true as contemporary buildings of this type are generally assembled out of two or more pieces, and even in some cases more than one floor/storey.

      • You can (under some systems, at least) get around this by living in trailers and such, which (I don’t think) are subject to the same speculative market forces. Hence poor people living in trailer parks? (This sounds right, but I don’t actually know enough economics to be sure).

        We call them “mobile homes” or “manufactured homes” these days in the U.S. They don’t have wheels, and most of them are moved only once ever, from the factory to a site in a mobile home park.

        Here in Michigan, we have many mobile home parks, and by and large they are full of poor and working class people. Yet they are paying MORE per month for the privilege of living there than nearby homeowners pay on their mortgages.

        If you calculate it in dollars per square foot (since mobile homes are much smaller than even very modest houses), the difference is even greater. You’d think flimsy, temporary abodes would cost residents less than solid, permanent, site-built houses, but that isn’t how it works.

        Part of the reason for this is that the park owners exploit their power. A resident will buy a unit for top dollar, but is only allowed to sell to the park owner — for a very low price set unilaterally by the park owner. Hence, once you move in, it’s not possible to move out without taking a huge loss. Depending on the owner, there can be other scams going as well.

        But a more important reason is that, even though the residents could afford the monthly payments for a mortgage on an actual house, their credit is not good enough.

        People get bad credit histories for many reasons, but often it’s because they have had problems organizing their own lives. It means they have had trouble staying employed, staying healthy, paying bills on time, making rational decisions between time and money, etc., etc.

        Hence, each mobile home park is a dense concentration of people who, on the average, have some history of bad choices and/or bad luck.

        • houseboatonstyxFor says:

          In the US, for very low cost to rent or buy, a class or two down from A)”mobile/manufactured homes in a travel park” is B)”cheap old travel trailers/RVs in a campground”. In the US, B) really can be a better situation than the A “trailer park”.

          Largely because the B) units do not need sewers or water supply or even constant grid electricity; they have a tank that carries their fresh water and another tank that carries their sewage, Some have a generator to make their own electricity, and all can use a long ac extension cord to the nearest ac outlet. Thus someone who wants to offer non-utility ‘camping’ space, or let a friend ‘camp’ free, does not have any of the water/sewer costs and regulations to worry about.
          B) units can be moved, sometimes every day, by an ordinary car with an ordinary trailer hitch, or by an internal motor (RVs).

          Of course as the unit deterates and becomes harder to move, the dweller becomes more and more vulnerable to the park owner practices you describe. Otoh the dweller usually owns zis own unit so can sell it to an outsider, or can walk away and buy a similar unit for little cost.

          • Deiseach says:

            Okay, can I ask: are there such things as halting sites in the U.S.A.?

            I’m curious, because I’m pretty sure you lot have vagrancy laws, and I’d imagine things such as travelling circuses in the past would be exactly the kind of “move in with caravans and park in a field” activity that would only be tolerated for a short time; that is, if a bunch of carnies tried living in a field near a small town, they would be moved on as undesirables.

            Now, my curiosity is this: do you have the same problems or same attitudes in America towards people moving caravans/mobile homes onto sites, or empty fields, or disused parking lots, or the side of the road, etc. and setting up to live there?

            Also, re: what houseboat says about letting a friend “camp” for free on your patch of land if they’ve got an actual trailer/caravan (not a “modular dwelling”) – that depends. Out the country, I’m sure it’s doable. In a town or suburb or on the outskirts, it would be a different thing if someone was living in a caravan in a back garden. The local authority (whatever your town council) might have to get involved for health reasons – are they dumping their waste tank down the public sewer? Are the neighbours complaining?

          • houseboatonstyx says:

            Basically, everything you mention is worse for poor people in the US, as we don’t have any such recognized demographic as your “Travellers”. Our poor people who live in units that can actually be “highway legal”, are called “homeless” (unless paying rent in an approved “trailer park” or “RV park” or “campground”), do not travel as a group, and have not organized themselves as a group. I’ve never heard of a travelling circus/carny setting up without lots of permissions; certainly there are no official “halting sites” for such groups. Highway “Rest Areas” have strict time limits, sometimes less than one day.

            Everything else you mention is probably technically illegal and has the same stigma. Except new shiny units with well-to-do people living in them; those people are called “full-timers” and/or “snowbirds”, and often seen as spending tourist and/or pension money in the local area.

            Enforcement varies greatly — by neighbor complaint only, unless you’re parked well within police view. These “self-contained” units carry their own sewage tanks and can dump safely into any sewer cleanout hole, which most houses have.

        • Scott Alexander says:

          I’m moving soon, and so far I’ve come across two types of houses for rent:

          1. Houses with huge yards and many bedrooms that are much bigger than I, as a single guy living alone, need.

          2. Mobile homes, which seem much more appropriate and the price is right and they’re in pretty much the same areas.

          So far my stereotypes have led me to not consider the mobile homes (stereotype = full of people who will probably be loud and disruptive), but this seems kind of dumb. Why can’t there be a nicer version of trailer parks?

          I’m thinking back to Nydwracu’s thesis (I’m sure it’s not original to him) that the whole point of having overpriced homes is to price poor people out as a filter mechanism to make neighborhoods “safe”. It’s the only thing I can think of to explain why there aren’t better trailer options available.

          • bartlebyshop says:

            > Houses with huge yards and many bedrooms that are much bigger than I, as a single guy living alone, need.

            Growth mindset. Why not start a grow-op, or acquire a menagerie of pets (do you like snakes?), or put a small bounce castle in each spare bedroom?

          • Matt C says:

            I lived in a mobile home park for a while. It wasn’t bad. I’m sure some trailer parks are unpleasant, but I wouldn’t assume out the gate that one was worse than, say, a lower middle class residential neighborhood or apartment block.

            (Er, this is probably overstating it. But still, if I were a single dude again I would consider a trailer park a reasonable temporary place to live.)

            If you have the time you might spend an hour walking around the place a little after quitting time or on a Saturday afternoon, that would give you an idea of what kind of neighbors you’d have.

          • Gbdub says:

            Your problem is that you live in a not-so-dense place with few single guys living alone, so there are few quality apartment blocks for young professional types.

            Out here in Phoenix we’ve got a ton of very nice apartments/condos/town houses at pretty reasonable rates, and after a long pause from the housing crash, now there are more being built all the time.

            The “downside” is that those large single family homes aren’t much more expensive, as long as you can qualify for a decent mortgage. But apartments do save certain headaches and provide more on site amenities, so it’s a trade off. But at least you can make a choice.

            Mobile homes seem like the worst of both worlds, with a lot of the downsides of both single family homes and apartments.

          • houseboatonstyx says:


            The “travel trailer/RV” option I described in the post above yours, was the worst case low end. The high end is a rather luxurious RV in a high end “RV park”, with “full hookups” ie water, sewage, electricity, wifi, etc. These “parks” can be like a resort, with swimming pool, clubhouse, etc — with nice neighbors.

            The financial downside to either end is, you have to buy an RV (from any dealer, or any individual selling theirs), and hope to get a good price when you are through with it. Depreciation is quick — as with a regular automobile. The comfort downside is cramped living space, though new RVs are coming out with “slide-outs” that can make some areas up to 12′ wide.

          • A good friend of mine used to sell mobile homes. He told me there is a huge markup on them, which probably means the market is not very efficient. This might translate into the rental market as well, so you might be able to get a much better price if you negotiate.

          • William O. B'Livion says:

            There are nicer mobile home parks, but they’re generally in places where “real” homes are extremely expensive.

            In places where land is more available buying a house is cheaper and people further down the socio-economic scale will do things to purchase or rent something with a bit of land around it.

            We are in the position of looking to buy our first house as soon as my employment situation settles down, and here in Denver I can get a (rounding/roughly) 1500 square foot 3 bed 2 bath in a modest neighborhood for 200 to 300k, which (with our available downpayment) on a 30 year note is about what we’re renting the same thing for.

            On a similar salary/economic situation in Silicon Valley there simply are *NO* houses available (at least on zillow) under about 400k (until you get into San Jose/Fremont). Thus nicer mobile home parks (and prices on those homes being higher.

          • Eli says:

            I’m thinking back to Nydwracu’s thesis (I’m sure it’s not original to him) that the whole point of having overpriced homes is to price poor people out as a filter mechanism to make neighborhoods “safe”. It’s the only thing I can think of to explain why there aren’t better trailer options available.

            The truth, as I know it, is somewhat even simpler: Americans aren’t trying to make housing be overpriced, but they have acquired a prejudice that high population density is for poor people. Thus, everything built for the middle classes is designed to have a low population density, and thus a high ratio of land/floor area to inhabitants, and is thus fucking expensive. The long-term effect, once a few first-generation suburbanites benefit, is that folks in our set (youngish, professionalish, looking for housing) can’t find decently-priced places to live: we either take a shithole that’s not oversized and overpriced, or an oversized, overpriced “nice” place.

            My fiancee, my other friend, and I all observed more-or-less exactly the same problem apartment-hunting in Boston. Another friend of mine blogs on housing issues and has explained all this stuff to me.

          • Devilbunny says:

            Why can’t there be a nicer version of trailer parks?

            Because they are subject to Gresham’s Law: bad tenants drive out good.

          • nydwracu says:

            I’m not so sure about it these days. There’s something else going on, at least.

            If the point of restricting supply is to drive up prices to filter out demographics unlikely to make good neighbors, supply should be less restricted in places with fewer such people. That is: if the problem with cities is diversity, less diverse countries should have bigger cities.

            Just based on eyeballing, it doesn’t look like that’s the case: except for NYC, you just don’t get large, dense cities in the West. (The only Western city more populous than NYC is Moscow.) And NYC is… you know. (If you want statistics: 36% foreign-born, 22% Jewish, and 44% white, but I think that includes Hispanics.

            White people don’t seem to like large, dense cities, just in general. What’s going on here? Or is the eyeballing misleading?

          • Deiseach says:

            Why can’t there be a nicer version of trailer parks?

            There are; they’re called “recreational vehicle parks” (as in the comment above) in America, “caravan parks” and “campsites” over here, or in France.

            They are meant as temporary holiday accommodation for families or retired people who have disposable incomes to spend on local attractions and who will be moving on after a period of time.

            Places for permanent living, on the other hand, don’t really exist over here. Halting sites are the nearest thing, and those are for Travellers.

            Why you don’t have nicer ones in America? I imagine it’s the whole idea of vagrancy and transience; mobile homes (despite what Mary says) were intended to be exactly that: caravans or units that could be towed by a car or trailer. People who moved from town to town would be travelling circuses, carnivals, tramps and hoboes, possibly petty criminals, and the unsuccessful who weren’t able or didn’t want to settle down. Solid citizens didn’t want to live in caravans, they’d be renting until they could buy a house of their own.

            So trailer parks were places of last resort for some, and were associated with the poor, the unsuccessful, and the undesirable. You don’t sink money into providing good services for people who don’t have money to pay accordingly for what they’re charged.

            Maybe it’ll change. Maybe, as the switch from ‘trailers’ to ‘manufactured dwellings’ happens, more upscale parks for medium to long term accommodation will be built.

          • eddie says:

            White people don’t seem to like large, dense cities, just in general. What’s going on here?

            What’s going on is that white people don’t like large, dense cities. I see nothing mysterious about this. Yards are nice.

            I’m betting that black, brown, and yellow people don’t like them all that much either.

          • onyomi says:

            Someone mentioned this in a different comment section, but America is bad at cities. There are a number of cities in Asia (Taipei comes to mind) which are dense yet super comfortable and convenient to live in, and not even that expensive (though they probably seem so to the rural population of those countries). Most American cities have sub-par public transportation and few pleasant walkable areas.

          • nydwracu says:

            Someone mentioned this in a different comment section, but America is bad at cities.

            That it is. Why is it? Is Europe bad at cities? If so, why? If not, why aren’t there many big cities in Europe?

          • onyomi says:

            I have not spent enough time in Europe to say, but my impression is that, on average, European cities are more walkable and comfortable to live in on a modest salary. There does seem to be some correlation between age of a city and livability, since cities built up before cars are necessarily going to be more walkable. But by that logic Boston should be easy to live in, which I do not find to be the case at all, partially because the roads are narrow and illogical, having been designed for horses rather than cars.

            The problem with cars is they encourage sprawl, which encourages people to get out of the city altogether and just live in the suburbs, but on the other hand, some cities are much more drivable than others, which is why, for example, I like Chicago much better than Boston.

            I have similar feelings about Tokyo and Kyoto. Tokyo’s roads are a spaghetti string mishmash by virtue of being a fishing village grown to massive proportions. Kyoto, by contrast was intentionally designed on the model of a Tang Dynasty Chinese capital, and so is a very logical grid (also not trying to conform to the shape of a bay, though the site for Kyoto was also chosen for its fengshui).

            There may also be a cultural thing: I know in a lot of non-US cities there is a culture of “going to market,” where you buy the food for tonight’s dinner today at a neighboring stall and there are a bunch of different stalls for a wide variety of products. Americans now drive their SUVs to Costco and stock up on all the food they need for a week or even a month. Makes you wonder how much the freezer has changed our urban culture.

          • nydwracu says:

            That’s possible in Bremen — there are grocery stores everywhere. I’d assume it exists in at least some parts of Japan: the writers of Detective Conan thought nothing of having their characters go to the market, call home, and ask what they should pick up for the night’s dinner. But there still aren’t many large cities in Europe. Berlin (a godawful mishmash of American and Soviet building plans — million-lane roads meeting immense slabs of Stalinist concrete — and it fucking sprawls) is the largest city in Germany.

            Here is a list of countries that have cities larger than any in Germany: China, Japan, South Korea, Singapore, the US, the UK, Russia, Indonesia, India, Bangladesh, Brazil, Chile, Mexico, Thailand, Vietnam, Turkey, Burma, Colombia, Saudi Arabia, Egypt, Iran, Iraq, Pakistan, South Africa, Tanzania, Nigeria, the Ivory Coast, and the Democratic Republic of the Congo.

            It could be different definitions of ‘city’. The big African cities aren’t very dense. Then again, can they build 30-story apartment buildings in the Democratic Republic of the Congo?

            Come to think of it, Christopher Alexander apparently hated high-rise apartments — he said they imposed isolation, since it’s harder to go outside or maintain contact with the outside world. So that would impose an upper density limit. But the same logic imposes a lower density limit, even more so: if you live in a suburb and have to get in your car to go anywhere…

            But this is about size, not density.

          • Deiseach says:

            There really is not a lot of single person accommodation out there, even over here (I’ll spare you all the rant about how society is set up for coupledom).

            I don’t know how much room you’re looking for, Scott, or if a single bedroom unit would do you.

            But think about it; unless it’s very basic, you can’t make much off a single-person flat. Even if you have a nice professional type like a hard-working young doctor such as our host, you can only charge so much. Put two people who are working full-time into an apartment or house, and you can charge higher rent. If they’re a couple, you assume they’re cohabiting and sharing the one bedroom. Two people in your one-bedroom unit can pay more rent than one person.

            So for a landlord, a larger property is a better bet (depending of course on local demand and population density). Trailer parks are for short-term, transient, temporary accommodation; retirees; or, if you’re living there long-term, for the poor and the losers.

          • Shenpen says:

            How comes apartments, flats, condos just don’t exist there?

            As for Nydwracu’s theory, I have heard the same reasoning in rural Hungary, so true or not it must be pretty global. The idea is that when there are houses for sale in a village, the locals try to bid up the prices or something because they don’t want a poor gypsy family to buy it. If they do, then soon enough 18 people will live in that, which is illegal but unenforecable (not really officially living there, just being there all the time) and according to rural people will fuck up the neighborhood, like throwing litter into other people’s gardens, and vandalising stuff, and this means the prices dropping further and this invites more poor gypsy families. I mean, this is how the rural people think, I don’t know if it is actually true or not.

            However, there is a fun aspect. There is more to it that standard racism and classism. If ethnic group A has hardly any kids and ethnic group B has kids, group B will per definition make worse neighbors. Regardless of class and ethnicity, it is typically teenagers who fuck shit up in a neighborhood. Have a hood where everybody is 40 and it will be tranquil enough, no matter what the class or race is.

            My point is, they may be better off simply keeping out larger families. For example, have a small local school and refuse to enlarge it

          • eddie says:

            If ethnic group A has hardly any kids and ethnic group B has kids, group B will per definition make worse neighbors.

            In the US, there is an ethnic group B that has kids and actively seeks to live next to others who also have kids, considering them much better neighbors than those of ethnic group A who do not.

            We call this ethnic group B “families”. They tend to like suburbs.

          • Mary says:

            “mobile homes (despite what Mary says) were intended to be exactly that”



          • ADifferentAnonymous says:

            An extra minute of elevator ride imposes isolation?

        • William O. B'Livion says:

          “You’d think flimsy, temporary abodes would cost residents less than solid, permanent, site-built houses”

          Modern homes are generally *at least* as sturdy as site built homes.

          • William O. B'Livion says:

            (the post this was in response to seems to have disappeared).

            It’s not a matter of “plastic box” v.s. “wooden box”. Many of the same materials used, and though the standards/codes come from different bodes (USG v.s. Local) they are generally very similar.

            There are more and more “manufactured” homes being built dropped on foundations on private (as opposed to commercial park) sites, and in those cases they generally are on price parity with stick built homes of the same size and appearance.

            From the pricing that I can see it does not look to me like manufactured homes in my area are on a higher monthly cost than more traditional homes, other than generally the manufactured homes *in parks* do not appreciate in value, and you do not own the land.

            But that’s not what you seem to be talking about, you’re talking about *renting* a manufactured home v.s. buying a traditional home.

            The question you are asking is “why do people live in trailer parks when they could buy a house and land for the same money”.

            This is the same as asking “Why do people rent when they could buy a home”, and there are LOTS of answers.

            For some people they have sh*t credit and can’t get a loan *at all*, much less at a reasonable rate. For some they’re in the process of saving up for a home. Some people are fairly transient, and don’t live in one place long enough (The place I currently live is the longest I’ve spent in one house since 1984).

            I don’t know the Michigan market *at all*, and haven’t really looked into trailer parks here in the Denver Area much so I don’t know how comparable rents are, but when it comes to housing purchases I’ve noticed people don’t really rely strictly on economic calculus.

    • Wulfrickson says:

      And this dynamic is of course at the core of housing-unaffordability crises in San Francisco and a few other cities – zoning is decided locally, so the only people with political power over a town’s zoning are its current residents, mostly homeowners, who inflate their own house values by enacting artificial supply restrictions. The “tulip-subsidy” solution of rent controls are also globally negative-sum, helping current residents at the expense of prospective ones; Alon Levy explains it pretty well here.

      • Jbay says:

        While that may be true for San Francisco, it is not, as far as I know, the reason for the housing affordability crisis currently going on in Vancouver BC. That said, we have very little data here on what’s actually causing it.

        Recently, when asked if she would consider implementing measures to reduce property speculation in order to lower our housing prices, our Premier said of course she wouldn’t do that; it would cause homeowners’ property values to go down.

        (reference below)

        • Eli says:

          The core problem is that a durable good (built housing) is being treated as a safety asset (land is being used to dump excess value for “investment”, as if it was gold bricks).

        • Deiseach says:

          A lot of people have borrowed using their houses or property as the collateral; if you reduce the value of housing, then you’re automatically bumping up their indebtedness. People were targeted and encouraged to borrow against the value of their house, as the idea was sold to release the value of the money they had ‘tied-up’ in property. Since house prices were booming, the idea was that you could always sell your house, pay off your loan/second mortgage, and still have enough to buy a smaller house more suited to your needs as empty nesters/retirees.

          No politician is going to want to be blamed for ‘hard-working families being driven out of their homes because the bank considered their debts greater than their assets’.

          It’s also an easy way to create feel-good sentiment; if your house is going up in value, you feel better-off or you have the idea that you have a valuable asset that you can cash in on should you need to do so; governments reap the reward of that sentiment. Conversely, if your house loses value, you feel poorer and the sense of consumer confidence (which economists rely on to drive domestic spending) plummets, which has knock-on effects on the economy.

          Consumer spending is an important element of economic growth. … Trends in this component are therefore very important for forecasting and planning. …In the US measures of consumer confidence by the University of Michigan and the Conference Board receive much attention, both domestically within the US but also internationally.

          Such indicators have a broader use than solely as an input to model based forecasts. They provide some barometer of consumer sentiment and thus are an additional piece of information that may be used by those analysing, or interested in, the health of the economy.

      • Deiseach says:

        But without rent control or some kind of rent cap, landlords will charge what the market will bear – we’re seeing that right now in Ireland, where rents are going up relatively hugely.

        People on social welfare payments or in low-paying jobs are being caught two-fold; firstly, rent supplement is being phased out so they no longer get help towards paying rent. Secondly, the levels of rent supplement are much lower than the true market value of rents. Thirdly, if the government raises the rent supplement, landlords (not all, but enough) will and do raise the rent accordingly – “now you still have to pay me a hundred a month extra on top of the rent allowance”, and many people don’t have the money to do that.

        There’s no one simple solution. The favourite of the libertarians – leave it to the market – is unrealistic. Well, landlords in city Y raise the rent beyond what you can pay? Rent from someone cheaper! (Nobody is cheaper) Someone must be cheaper, there’s plenty of rental properties out there (No, there’s not) Go elsewhere! (Umm – where elsewhere? a whole different city!) Buy a house! (I can’t get a mortgage, I don’t earn enough) Well, it’s your own fault for being that poor!

        The places with the most rental properties available will be the big cities, but if everyone goes to The Big City for lack of accommodation elsewhere, that means that demand will outstrip supply, and when we got a building and construction and development boom the last time to feed that demand, you see what happened with the crash 🙂

        • Harald K says:

          How are tenants taxed versus owners in Ireland? In Norway, interest payments are deductible on your taxes, so owners/landlords are subsidized at the expense of tenants.

          That was a deliberate policy once upon a time, pushed by the socialist government – the idea was to get as many people as possible owning their own home so they wouldn’t be at the mercy of landlords. But it really doesn’t seem to have helped much, especially not us who for various reasons can’t buy. And on the whole it really doesn’t seem such a great idea.

          • Deiseach says:

            Far as I can make out, no bloody government in Ireland of whatever party or political persuasion wants to get into the messy ground of landlords versus tenants and holding people accountable. They’re much happier letting the courts settle any disputes. Because they’re afraid that if they enforce regulations, landlords will give up renting properties, and then the government will be left having to house all the suddenly homeless people.

            From a historical point of view, Irish people have always wanted to own their own home rather than rent; I hate to drag Our Unfortunate History into it, but it really is a hangover from the days when the landlord could evict you out onto the side of the road for any or no reason and then it was either the workhouse, starvation or emigration. Irish people have this feeling in their bones that if they own their own house, at least no-one can turn them out of it (the bank repossessing it for not paying the mortgage is a different matter).

            That’s why the rental market is undeveloped compared to other countries, and why landlords can get away with blue murder in some instances (e.g. I’ve read a policy document by a housing charity and the submissions by interested parties about regulations for rented accommodation, and landlords were arguing that bedsits should be exempted from these regulations because a lot of them were built before current building regulations came in, and doing them up would be too expensive, so landlords wouldn’t rent them out, and people would then be deprived of cheap single room accommodation as their first step in renting. I rented a bedsit years back for a short time, and my brother was in one as well, and they were in dreadful condition; where my brother was living, all the tenants were warned not to use the upstairs shower because of the danger of electrocution. The landlady was in no rush to fix this, and I’m pretty sure was not declaring her rental income for tax purposes).

            Re: interest relief – a lot of people during the boom years bought second and more properties and rented them out, and the government was only too happy to encourage them to do so (it fed the construction industry which was driving the money-making good times) with tax breaks and the like.

            All the chickens came home to roost, of course, during the crash when people could not keep up the mortgage repayments on these second houses.

        • Cerebral Paul Z. says:

          From an an economic viewpoint this is quite a puzzle. If landlords are getting such a windfall, you’d expect to see new supply crowding into the market to get in on the action. It wouldn’t all have to be new construction either: where are the homeowners renting out rooms for some extra income? The solo renters advertising for roommates?

          You’d also expect, in the immediate wake of a housing boom-and-bust, to see excess capacity. Where’d it all disappear to?

          • Deiseach says:

            Okay: excess capacity was often in the form of landbanks (there was huge speculation and churn with buying land, getting planning permission for a development, and then selling on the land with the permission for a profit; often plots changed hands several times with never a spadeful of earth turned to start building) and what are called ghost estates.

            Where a builder or developer went bust and left housing estates unfinished, there’s a tangle of legal issues about who owns them, who has rights to them, can they be sold to pay the creditors, what about estates where half the houses are occupied and half aren’t, etc. etc. etc.

            There’s not agreement on how many empty houses are out there, and part of the problem is that there may be housing capacity available in one part of the country – but nobody wants to live there as it is too far away from the centres of employment (e.g. according to the Wikipedia article linked above, “Leitrim, for example, has a housing surplus of 401%” but Leitrim is perhaps uncharitably considered the arse-end of nowhere).

            We see that as well: people who apply for social housing can put down their “areas of choice”; some areas are very over-subscribed, while others where we might have houses available have little demand (often because they’re in isolated rural areas and people need to be near schools, shops, etc.)

            Also, the problem is sometimes not so much availability of housing as inability to pay the increased rents in a more buoyant property market, where wages have not correspondingly increased (and we’ve been hit and are continuing to be hit with new taxes and charges to pay off all the bailout loans and bondholders that started the whole austerity economy bit).

            Or properties which were rental properties were repossessed by the banks and haven’t been sold, or the buyers don’t want to rent them out – we’re seeing that in my workplace (though not as much recently): a lot of people coming in with letters from banks saying the landlord couldn’t keep up mortgage repayments, the bank is repossessing the house, and they have to vacate the property by such-and-such a date.

            It’s complicated. There’s no one simple soluition. For instance, central government has loosened the purse strings and there will be more social housing built, but it’ll still be two-three years before those houses are available, which is not much good to someone who needs a house right now.

          • Jacob Steinhardt says:

            In San Francisco, there was until recently an “AirBNB” law making such short-term rentals illegal in many situations. (1 guess on which group was most in favor of such a law.)

          • Nornagest says:

            From an an economic viewpoint this is quite a puzzle. If landlords are getting such a windfall, you’d expect to see new supply crowding into the market to get in on the action.

            If you’re feeling puzzled, look up how hard it is to get housing built in San Francisco. (Hint: the answer is “near-impossible”.)

            Market-based solutions only work when there’s a reasonably fluid market.

          • Cerebral Paul Z. says:

            My hat’s off to the Irish government, then. Our own government here in the US has proven itself no slouch when it comes to creating either surpluses or shortages, but I can think of no case where it’s managed to do both at the same time.

          • Matt says:

            It’s not housing that boom-busted: it’s locations, land. Its supply is fixed. That’s the cause of this phenomenon.

        • Glen Raphael says:

          The libertarian solution is to make it LEGAL to build and rent housing that poor people can afford. Which is currently not the case in most cities due to the whole “propping up housing values” dynamic. Zoning laws and building codes are the enemy of affordability.

          At the very lowest end, consider that if there weren’t laws prohibiting it, a guy who runs a “storage space” facility could LEGALLY let people live in their lockers – he’d just need to install a communal bathroom/shower facility somewhere in the building to make it practical. Or a bunch of “single family dwellings” could LEGALLY subdivide into multifamily apartment spaces and people with a large backyard could build a shed or small apartment back there or just let people pitch a tent back there. They can’t do this because of laws you take for granted. But don’t put that on the libertarians. “Leave it to the market” might work pretty well if it were actually TRIED. But “leave it to the market…while keeping in place all the policies that prevent the market from clearing properly” is indeed a nonstarter.

          • Deiseach says:

            The libertarian solution is to make it LEGAL to build and rent housing that poor people can afford.

            Social and affordable housing! In practice, there were some loopholes

            (a) Builders/developers could opt-out of building 20% social/affordable housing in their developments by a monetary contribution. It made more economic sense for some to pay the ‘levy’ because they could sell the new houses for full price, instead of at the reduced social/affordable price, and so still make a profit after paying the charge. Local authorities were often happy to make such deals because this was money in hand which was needed for running services.

            (b) Builders put as few of these in as they could get away with, and there was a noticeable difference between them and the other houses in quality and size (and that’s saying something as they were slapping up cheap’n’cheerful housing during the boom and charging crazy money for them; the amount of new-build – that is, built within the past ten years – houses I’ve seen BER certs for that are woefully underinsulated, for one thing, is amazing)

            (c) People didn’t and don’t want to live where social houses are included on their estates (we’ve recently had a local representative complaining on behalf of a particular area about plans to build too many social houses there and forcing people to take more than their fair share. We’re talking a development of 20 social houses in an area of a couple of hundred others). People complain about plans to do this. People get on to local politicians to complain about this. People go to local radio stations and newspapers to complain about this. NIMBY is alive and kicking.

          • Jiro says:

            Deseach: What you describe is not “making it legal”. What you describe is mandating it. Mandating it and making it legal are not the same thing.

          • Xplo says:

            “At the very lowest end, consider that if there weren’t laws prohibiting it, a guy who runs a “storage space” facility could LEGALLY let people live in their lockers – he’d just need to install a communal bathroom/shower facility somewhere in the building to make it practical.”

            Of course, that leads to a country where the poor are living in storage spaces, sheds, and tents, which makes all those run down inner-city ghettos seem downright homey. I would suggest that any libertarian who thinks this is an acceptable solution be required to live in the least desirable housing thus produced by the free market.

          • Lupis42 says:

            While any progressives will be required to be living in the least desirable housing provided/subsidized by government?

            Seriously, I’d much rather a typical storage locker than some of the apartments I’ve lived in/visited as a student, and those were better than the section 8 housing.

          • Jacob Steinhardt says:

            Typically even when it is legal to make affordable housing, no one does it, because it isn’t as profitable as building more expensive houses. As Deiseach notes above, even when such housing is legally mandated, many companies attempt to find ways to get around it because it brings in so much less money.

          • The Anonymouse says:

            Seems unlikely. Under that idea, you would predict that no manufacturer of goods would ever build or sell anything except the most ultra-lux goods.

            Which we know isn’t true: there tend to be working markets at every point on the spectrum. The problem–demonstrated here in Portland with a lack of developers wanting to build “affordable” housing–is that when you mandate that new construction have all of the amenities of high-end housing, you can’t be surprised when developers are only selling high-end housing.

          • Glen Raphael says:

            xplo wrote:

            “Of course, that leads to a country where the poor are living in storage spaces, sheds, and tents…”

            Er, no. You’re missing my meaning. When I suggest that I would wish to ALLOW the poor to legally live in such places, you seem to be reading that I would REQUIRE them to do so. Whereas as a libertarian, I merely want that sort of choice to be available alongside whatever other options there are. Adding new options is good; I seek the staid, moderate, middle ground between prohibition and compulsion. I’m saying that if somebody WANTS to live in a storage locker or someone’s backyard or to split a house into high-density condos or any of a zillion other options we haven’t thought of yet and that’s okay with the owners, the government shouldn’t toss anyone in jail or fine them for making that choice.

            I am “pro-choice” on housing.

            If we merely ALLOW people to select from a wider variety of super-cheap currently-prohibited housing options, which people are likely to voluntarily avail themselves of such options? Only the ones for whom such options are a step UP from their next-best available option.

            For instance: if your next-best option is sleeping under a bridge or on a park bench or in jail, that shed or storage locker is an improvement.

            Or if your next-best option is a government low-income housing project BUT there is a 6-month waiting list and you need a temporary waystation. Or if you just really really value SAVING MONEY with the lowest possible housing payment so you can spend more money on health or food or education or supporting your kids or entertainment.

            I would love to have the OPTION to live in super-cheap minimalist housing – something like the Tokyo “capsule hotels”. I don’t have that option now any more than poorer people do, and it’s because building codes and zoning laws and fire codes keep ratcheting up to make the housing options poor people would have picked in prior decades illegal today.

          • Jacob Steinhardt says:

            > Seems unlikely. Under that idea, you would predict that no manufacturer of goods would ever build or sell anything except the most ultra-lux goods.

            I don’t think so; just because one market incentivizes luxury goods does not mean that all must.

            In particular, in most cases, the land itself is the largest cost to the landlord, in the sense that it’s in pretty finite supply. On the other hand, the amount you can make off the land mostly just scales with the niceness of the stuff built on it. Therefore high-end housing ends up more profitable.

          • Deiseach says:

            Jiro, as it stands, there is nothing saying you can’t build or rent housing that poor people can afford. We have housing organisations and charities doing just that.

            The problem is with private landlords. Why rent housing to someone who can only pay you €425 per month when you can rent the same property to someone who can pay €600 or €650 per month? We have landlords who will NOT rent to people on rent supplement precisely because they can only charge up to a certain limit; the solution there is that we’re encouraged to ‘turn a blind eye’ to payments under the table – i.e. the tenant and landlord both sign an agreement saying “yes, only charged €425 per month”, then the tenant makes an undocumented payment of €150-200 extra per month. Illegal as hell, the landlord has no obligation to declare this extra income for tax, the tenant has no rights, how people who are living on the single person social welfare payment of €188 per week are supposed to afford it who knows, but hey – at least people are not out on the side of the streets, right?
            Or the government can relax “fit for habitation” regulations (as that landlord’s’ group submission I mentioned requested) so that the properties can be rented more cheaply.

            Which is fine, if you think poor people are only entitled to live in places with no heating, mould everywhere, leaks, dangers of electrocution, etc. Because if the landlord is charging cheap rent, the property is run-down and they’re not going to do it up – why do it up unless it’s to get a higher price? Then the poorer tenant is moved out and someone who can pay the higher price gets the benefit of the newly-refurbished property.

            I’d love to think there are legions of public-spirited private landlords who would provide decent, habitable accommodation at a reasonable price for poorer tenants out there, and they’re only held back by meaningless red tape – we don’t have rent control in Ireland, for one thing, contracts are only for one year and renewed every year, not for longer periods – but what I’m seeing in my limited slice of reality is not that happy scenario.

            After the crash, there were landlords happy to take lower rents (because that was all they could get). And there were and are decent landlords who wanted to give tenants as much help as they could. But now that the economy is slowly recovering, rents are going up. Landlords in some cases need to get as much return as they can on a property; in other cases, it is simply charging more because they can charge more. I am not seeing anyone charging less.

          • notes says:

            Deiseach –

            Density is the missing piece of the puzzle. There are private landlords who would indeed be happy to rent to lower income individuals at affordable rents – sometimes by renting smaller efficiencies, more often by adding roommates until the math works.

            This is, in most places, prohibited by zoning laws, usually by occupancy restrictions.

            The only major exception of which I am where is around universities: it’s quite common to find ‘by the bed’ rental properties nearby, which are offering dormitory-style accomodations, with several tiny rooms sharing a common kitchen and bathroom. The rents are cheaper for each individual resident – but in total, are enough to encourage a profit-minded landlord to rent by the bed rather than by the apartment.

            There are other ways to slice it up than by the bed, but the broader point stands: often, those unable to afford an apartment face this difficulty because the minimum amount of housing legally available for rental is too large for their means.

            Social and Affordable Housing, in the vast majority of cases, results in nothing more than a lottery windfall for those chosen to receive housing at a subsidized price, without material impact on the vast majority of those who cannot afford housing.

          • cypher says:

            Are you forgetting why we have fire codes and building codes in the first place?

          • notes says:

            Cypher –

            This isn’t about fire and building codes for safety; this is about fire and building codes used as a vehicle for NIMBY.

            University dormitories are very rarely the kind of housing that people with other choices choose; they’re also not more prone to fire or collapse. But that model of rental housing almost doesn’t exist, outside of university neighbourhoods.

            Similarly, boarding houses were almost entirely eliminated from the US by zoning.

            When the building code mandates, as it often does, occupancy limits on the number of ‘non blood-related adults’ resident on a property, and gives as its explicit justification ‘to preserve the character of the neighbourhood’, you know that this isn’t about safety… it’s about not having apartment buildings full of poor people next door. A dozen people in a house are fine… if they’re family. Try and rent that house out to a dozen roommates, and it becomes illegal.

            (Then again, I may have misread you: perhaps you meant that fire and building codes have always been primarily vehicles for zoning the poor out of the way? It’s uncomfortably, though not wholly, true: ‘the law, in its majestic equality, forbids rich and poor alike’ to build or rent housing the poor can afford.)

          • Jiro says:

            Having apartments with large numbers of people imposes an externality on the rest of the neighborhood. Zoning laws are a way to prevent such externalities.

            (The externality is imposed on a Bayseian level. Not every such apartment imposes the externality, but they are statistically more likely to do so.)

          • notes says:

            They do impose externalities – this is an argument for correctly pricing those externalities, not all of which are negative, not for banning that use.

            NIMBY isn’t called that because it’s generally based on a lie; it’s called that in the belief that naming one of the faces of Moloch will motivate people to act otherwise.

          • Jiro says:

            notes: How does not having a building code prohibiting many occupants “correctly price those externalities”? The whole idea of an externality is that since the price is paid by other people, a free market price won’t take the externality into consideration. Not having a building code means that those externalities aren’t considered in the price at all.

            (A libertarian answer might be to have the multiple occupants prohibited by a homeowner’s association instead of by the law but I suspect that you would not approve of that, since you seem to oppose such prohibitions, period.)

          • notes says:

            Prohibition is not a statement of price.

            Prohibition and price are different ways of handling problems. Prohibition is the statement that there is no price high enough where someone could legally consent to do this, the statement that this choice is not within the realm of acceptable consideration at any time or for any price.

            Prohibition is an acceptable and appropriate tool to deal with acts so heinous that we would prefer to set the full force of the state against those who commit such. As applied to people who would like to live in the circumstances of a college dormitory, it’s an overreaction both comic and tragic.

            ‘Externality’ is often used as shorthand for negative externality; there are positive externalities also. In either case, they are costs or benefits that impact those who did not choose to participate in the action causing those results.

            Not all of the externalities caused by larger groups of people are negative – for a simple example, look to Eddie’s post about how families with children prefer to live in neighbourhoods with other families with children. (Contrast it with Shenpen’s post immediately preceding: this provides a convenient example of how the decision of whether an externality is positive or negative isn’t always straightforward).

            There’s also the issue of deciding which externalities matter. It is not difficult to find examples of people who are mortally offended by the existence of differing opinions, religions, or lifestyles. What weight should be given to those externalities, if any? Particularly when it’s also not difficult to find offsetting pairs of opposed opinions?

            Point being, it is not enough to say that there are externalities, therefore this must be prohibited. That is literally stating that an action has consequences for others, and therefore must be prohibited. This avoids the actual questions: what consequences? How shall we account for them? Why shall we choose this method of counting?

            If the argument was that multiple inhabitants use more public utilities, such as water or power… well, utilities are generally metered by residence, so that is likely priced correctly (or priced no less correctly than for the neighbours). If the argument is that multiple inhabitants require more infrastructure, which must be built out – usually something like that gets handled by charging a developer an impact fee which covers the increased road/sewer/etc. construction. If the argument is that they’ll consume more services and disproportionately burden their neighbours (to take a poverty-neutral example, this is exactly the argument often raised in dealing with families with small children – schools can be expensive), then there’s usually some way to attribute the costs more specifically… unless the law forbids that too, which it often does (specifically in the case of families with small children, not that this stops municipalities from trying alternatives).

            I don’t think zoning codes prohibiting multiple non-blood-related occupancy are illegitimate. I think they are often unwise, but I may be mistaken… and they are, in any case, almost always popular because they are believed to (and usually do) increase local property values.

            I do think that prohibiting housing that the poor could afford and then complaining that the poor can’t afford housing because not enough people will donate it to them at a loss is… well, is a pretty typical outcome of the political process. But it doesn’t reflect a good understanding of what’s going on.

            (Technically, I suppose you could argue that prohibition is a statement of price… it’s just that the price in this case is not denominated in currency, but rather in becoming subject to imprisonment, confiscation, or worse. It’s not common for someone to argue that position.)

          • Jiro says:

            notes: The externality is that places that house many non-related people are disproportionately likely to be nuisances.

            (And if you reply “they’re not all like that”, remember what I said aboyt Bayseianism._

          • notes says:

            Jiro –

            Perhaps there was an omitted argument which I did not follow that bridged the distance between ‘denser occupancy is disproportionately likely to cause nuisance’, with nuisance undefined, and ‘therefore, denser occupancy should be prohibited by law.’

            If there were, it likely would have addressed the tradeoff between ‘nuisance’ and ‘inability to afford housing due to prohibition’, as well as the choice between prohibition of the action producing the externality or pricing that externality in via some Pigouvian adjustment. It would also, necessarily, involve considering the other externalities in play – and not all externalities from denser occupancy are negative.

            Or were you just noting that actual zoning laws often don’t bother to address those issues, and do leap directly from ‘this offends someone’ to ‘ban it’?

        • Winter Shaker says:

          It seems to me, in a pulling-ideas-out-of-thin-air way, that you could have progressive taxation on number of tenants – a landlord pays nil, or a trivial amount, in tax out of the rental income they derive from one tenant (or one family of tenants), but as you acquire more properties, you pay a progressively higher percentage of tax on the rental income for each successive tenant; perhaps on an asymptote approaching a high enough number that nobody sane would bother to own an extra property as a buy-to-let if they were only going to get to keep such a small fraction of the rent. Make sufficient effort to avoid creating easily-exploited loopholes, so that there really is no way to avoid diminishing returns the more people you are leasing to.

          That way, at no point is anyone technically legally prohibited from buying another house to rent out to people, and people who have saved up for one or two houses to fund their retirement aren’t going to suffer noticeably, but at some point it becomes not worth anyone’s bother to add any more to their portfolio, thus reducing demand for buy-to-let houses, and reducing costs for buy-to-inhabit purchasers, or indeed social landlords.

          I can’t help but feel there must be a) some obvious reason why this wouldn’t work, otherwise some places would be doing it already, or b) some places that actually are doing it, but I haven’t heard about it. Can anyone enlighten me?

          • Deiseach says:

            You put a ban in place on building rental properties, they won’t get built. A lot of builders/developers were the ones putting up apartment blocks, etc. If they get hit harder the more they rent out, they are not going to build these, or buy the already built developments.

            Builders can’t sell their new apartment block, they’ll (often) go bankrupt; construction companies often run on amazingly tight margins and the difference between what looks like a multi-million euro going concern and a guy going to court to plead bankruptcy can quite literally be “am I going to get paid for this one particular project that is nearly done?”

            Depending on the government, e.g. if they’re right-wing “business friendly, small government, light touch regulation”, they may prefer to let the market handle the problem of supply and demand. Large scale social housing building by local authorities/national government will be perceived as interfering with the market which is a mortal sin.

            If the government, on the other hand, commits to large scale social housing construction, the money has to come from somewhere. Often by borrowing for capital investment projects on the international financial markets, which is a whole other headache because you are putting the nation in debt for years and the repayments come out of tax surpluses which is fine when everyone is working and there’s loads of tax revenue coming in, but not so fine when employment dips.

            Raising money by taxing large corporations? If they’re overseas investors, they can take their ball and go funnel their revenue stream through Lichtenstein instead. If you want to build up a reputation as a centre for financial investment, you go very light on the taxes and costs:

            Ireland is absolutely committed to
            maintaining its 12.5% corporation tax rate.
            This commitment is protected in an EU
            context by the principle of unanimity in
            taxation matters, and is accepted as part
            of the EU/IMF Memorandum of

          • Winter Shaker says:

            So where you are, a lot of homes are actually build-to-let, ie the company that builds them expects to be the landlord long-term? In which case, fair point.

            But in places/times where that’s not the case, I’d still expect such a scheme to have some positive impact in deflating the buy-to-let market without disproportionately reducing the number of homes actually available.

        • Mary says:

          ” Well, landlords in city Y raise the rent beyond what you can pay? Rent from someone cheaper! (Nobody is cheaper) ”

          Then what you do is figure out why people aren’t moving into the landlord business. If the rents are that high and there is no competition, there must be either formidable barriers or horrific problems.

          • Creutzer says:

            Perhaps because the supply for flats or houses to buy is just as small as the supply of flats or houses to rent (and there is no space to build new ones).

          • Jacob Steinhardt says:

            “Then what you do is figure out why people aren’t moving into the landlord business. If the rents are that high and there is no competition, there must be either formidable barriers or horrific problems.”

            Possibly because land is almost by definition a fixed/limited resource? I think this is almost the origin of “rent-seeking” as a term:


          • Alex Godofsky says:

            It turns out in the recent past (so you may not have heard of it yet) we invented technologies that allow us to build buildings more than a few stories tall. As a result suddenly it is possible to house more people on the same amount of land.

          • Jacob Steinhardt says:

            > It turns out in the recent past (so you may not have heard of it yet) we invented technologies that allow us to build buildings more than a few stories tall. As a result suddenly it is possible to house more people on the same amount of land.

            In San Francisco (and I would guess other cities) there are legal limits on the heights of buildings.

            ETA: Though your point is interesting in cities where such laws don’t apply.

          • Samuel Skinner says:

            And since it is an earthquake zone we can’t dig far into the crust either.

            Yes, this means our only hope is zeppelins and floating cities like in the Jetsons. Or live somewhere that isn’t a narrow rocky peninsula.

          • Dude Man says:

            or live somewhere that isn’t a narrow rocky peninsula.

            I seem to recall several tech companies promising to eliminate the importance of geography with their products. The dearth of successful tech startups outside of Silicon Valley seems to suggest that network effects were harder to get rid of than they imagined.

          • Nathan says:

            To the people saying there is no room to build – there’s tons. Unless every building is already at the height of the burj Khalifa or you’ve run out of sky you can always build up. Unless of course there’s some legal restriction on doing so (there always is). Where I live there is a ban on any building above a certain altitude (in a very hilly place) so there are big empty hills surrounded by massively overpriced houses.

            The vast majority of the problem is over regulation. Liberalise building restrictions and watch as affordable housing springs into existence.

          • Adam says:

            I don’t know about San Francisco, but in Los Angeles, it was illegal to build higher than 12 stories until something like the late 70s because the technology didn’t exist to build that high and withstand earthquakes. There isn’t exactly a ton of unused land sitting around, so to build up now, they need to first tear down existing housing stock to rebuild it taller, which is a much harder problem than just building high to begin with. Those buildings are occupied, so they’re either going to have to kick people out or buy them out to do it.

          • Nornagest says:

            I don’t know about San Francisco, but in Los Angeles, it was illegal to build higher than 12 stories until something like the late 70s because the technology didn’t exist to build that high and withstand earthquakes.

            If the average residential unit height in San Francisco was 12 stories, it could support the population of Los Angeles.

            (Not for very long, since the infrastructure isn’t there, but you take my meaning.)

          • Mary says:

            If there is not enough land, living in the location desired is a luxury good.

          • Adam says:

            Sure, but expecting the average to equal the max is a little extreme, isn’t it? That would imply zero single-family homes.

          • Nornagest says:

            Yeah, but I don’t expect SF to attract the population of LA anytime soon, either. The point I was making is that height limits aren’t adequate to explain the problems we observe, though they certainly don’t help as regards building densely in an affordable way.

          • Adam says:

            Fair enough. My point was more that if you combine the historical height restriction with car culture and the fact California was sold as the Italy of American with open land and large estates, and I think that goes some way to explaining why California cities have always tended to grow outward, not upward. San Francisco is a particularly odd case where so many people seem to insist on actually living in the city. Everywhere else, Californians seem perfectly content to commute. Heck, people even seem to insist on living there when they don’t work there.

        • BillG says:

          So this is a supply problem, though, right? There are going to be a limited number of potential tenants who can afford the higher rents. If it’s possible to rent out an apartment profitably to the lower income individuals, why wouldn’t someone be doing it, since it’s apparently not a market that’s currently being exploited?

          The best answer I can come up with for that question is that regulations/space/etc prevent the creation of enough supply to fill all of the markets. That doesn’t seem like a libertarian problem to me….it seems like either a spacial or regulatory problem. Maybe I’m just seeing it incorrectly though?

          • Cerebral Paul Z. says:

            In the Irish case it certainly appears to be regulation holding down supply: not only are they apparently echoing San Francisco’s policy of punishing builders with a “social and affordable housing” mandate, but they expect any landlord renting to a rent-supplement recipient to settle for a below-market rent.

          • cypher says:

            Also consider, there may be tenants that are simply unprofitably risky at the rate that they can afford to pay. The market can’t really solve that.

          • Harald K says:

            If it’s possible to rent out an apartment profitably to the lower income individuals, why wouldn’t someone be doing it, since it’s apparently not a market that’s currently being exploited?

            Not sure what’s the case in Ireland, but here in Norway I see one trend that may contribute: A significant part of the housing stock is owned by private landlords who are simply rich, and can afford to let their units stand empty. You rent it out, fine, but if it’s too much hassle you just don’t. A significant share of them are not really very rational profit-optimizers.

            And of course because they’re already rich, they want to charge a rent that makes a difference for them.

            Housing just isn’t a very liquid market, and you get these sorts of problems in not very liquid markets.

        • Anthony says:

          So what’s happened to all those buildings that got built during the boom? Are people living in them?

          As I understand it, San Francisco and New York approach most European countries in the difficulty of building new anything anywhere that’s not a government project. San Francisco’s population has gone up by only about 60% since 1925, and it’s 40% the density of Paris. But in the midst of the housing crunch, it’s considered a bold political move to advocate allowing 3,000 new housing units of which only 1,000 will be subsidized. San Francisco needs three hundred thousand new units, not three thousand.

          Meanwhile, ad-industry html writers need to get over themselves and go live in the ‘burbs where they belong, and where they can afford twice as much space and a car.

          • Gbdub says:

            So this seems to be the key – no one who can build anything is going to voluntarily build for the low rent market, because there’s still a ton of pent up demand in the luxury market! NOBODY is able to get the housing they really want. Combine that with legal controls on density, and building anything but the swankiest digs you can is basically charity. Why build cheap flats when you can build luxury condos that go for twice the price and sell just as fast?

          • notes says:


            Throw the gates wide open to development (which has its own issues), and you’re likely to see people building both luxury apartments/condos and middle income apartments and low income apartments – (some) developers are savvy enough to avoid head to head competition by going for underserved markets.

            The main issue is that it’s not just new build competing against new build, but also against all of the existing apartments in the area… and uses of existing apartments can change.

            So the reason that the development mix is tilted to luxury is partly profit margin, and mostly the fact that brand new middle/low income apartments are competing with yesteryear’s ill-planned luxury apartments, now being operated as middle/low income apartments.

          • Anthony says:

            Build 3000 units in San Francisco, and you’ll get 3000 new high-income households, even if all the new units are for the poor. Because if the units are subsidized, they’ll get rented to current poor residents of San Francisco, and the landlords of the lucky 3000 aren’t going to re-rent those units to more poor people – they’re going to jack up the rent as much as they can. And because the tech boom creates lots of young single men with lots of money, there will be takers for even the shittiest situations at high prices.

            Build 300,000 more units, even if they’re all overpriced luxury units, and some of the new people in the City will be middle-income, or even poor. Because lots of those new units will be taken by high-income people currently living in flimsy rotting firetraps above sketchy restaurants, and eventually, landlords of existing properties will have to settle for less high-income tenants.

        • Careless says:

          So there’s a housing shortage and you think the libertarian suggestion would be to let them move downmarket? Ideological Turing test fail.

        • ChristianKl says:

          What the market will bear depends on supply and demand. If you allow denser housing you will get more supply and thus lower rent.

        • Dain says:

          Libertarians have actually been screaming for some time about limits on building new units. So they have an idea, but I sense that even those being squeezed out by prices here in SF don’t want to see a sea of cheap high-rises go up.

          It’s not just a NIMBY issue. It’s simply the aesthetic preference of everyone, despite libertarians’ attempt to focus on the political economy of some class being pitted against another.

          • Nornagest says:

            There’s a lot of everyone in SF. I suspect you’re overestimating their homogeneity.

          • Anthony says:

            Everyone’s aesthetic preference is for lower density that has enough spaces for them to afford one. This is not achievable in the real world. So some part of “everyone’s” preferences has to be sacrificed – either the lower density and stock of historic buildings, or the affordability for anyone who can’t pay $3,000/month for a 1-bedroom apartment.

            Currently, San Francisco’s progressive politicians are choosing to sacrifice affordability. The result is an increasingly richer and whiter (and more Asian) city, because that’s what progressives really want.

    • Salem says:

      Yeah, thought for sure this was about housing, with foreigners bidding up the prime housing in Central London and the government scrambling to improve affordability with Help To Buy, housing ISAs, etc.

    • Cerebral Paul Z. says:

      It is about housing, in a sense. The higher-education bubble is replicated to a large extent in primary- and secondary-school education: something which, for some reason, comes bundled with every house.

      • Jbay says:

        In terms of living near desirable school zones, as I understand that’s a fairly uniquely American issue, isn’t it?

        • Deiseach says:

          Any Australians here want to comment, as I am going to traduce your education system? 😀

          My immediate boss has a married daughter with two very small kids living in Australia, and (from the granny’s side) the idea is that (amongst other reasons) maybe they’ll come home in a few years as soon as the kids are old enough to go to ‘big’ school, because the public schools in Australia are terrible.

          Part of this comes from an Australian neighbour who, although he’s agnostic/atheist or somesuch, recommending she send the kids to parochial school because the public schools are terrible.

          I have no idea what Australian schools are like, and I certainly wouldn’t tout the Irish educational system as being anything special. So – are Australian schools dreadful, or is it like everything else: it depends on what school where?

          • James Picone says:

            I went to a private school for all of my education (different ones, but still). I’ve certainly heard that a number of public schools are hives of scum and villany.

            As of recently I’ve been volunteering to talk to some kids at a local public school (primary) about maths, though, and it’s very definitely not awful. The suburb is moderately gentrified and generally quite a nice one, I imagine that’s relevant.

          • Daniel says:

            My primary school in a southern Adelaide suburb sucked. The suburb I live in is one that’s some blocks over from some notorious trouble suburbs like Hallett Cove, Elizabeth, Hackham West, etc. There were quite a few asshole kids, including me (I got better). The teachers didn’t have good methods of dealing with kids who bullied, and class sizes were in their 30s. I don’t know how much drug stuff went on because I was never a part of that. My brother went to the high school associated with that primary school and he was bullied especially in year 12.

            I went to a completely different high school in years 8 and 9, and I get the impression it was roughly the same atmosphere, but more students. In years 10 to 12 I went to a different, more well-off high school about 10 kilometres north of my home and it was pretty great.

            If I was raising kids I dunno where I’d send them, I’d definitely send them to that last high school but I have no clue what schools would be good before that.

            However, I don’t know how this all compares to American schools—I’ve never even been to America. Neither do I have a point of comparison between public and private schools here. I think my family belongs to the working class or middle class? Dunno where one label ends and the other begins tbh.

        • DavidS says:

          Massive issues in the UK too. I assumed that most countries would have
          a) state schools of varying quality
          b) admissions to said schools at least somewhat based on how near you live

    • eqdw says:

      I develop a curious mix of fury and mirth when I read San Francisco urban planning documents that make reference to “maintaining/increasing property values”.

      • The Anonymouse says:

        Allow me to recommend mirth. There are, after all, about 3000 miles of habitable land between the Bay Area and the next most outrageously unaffordable city.

    • I think the clearest example of the pattern you are describing is not the federal subsidies, primarily the deductibility of mortgage interest, but state and local land use controls. If I drive from San Jose to San Francisco in the late afternoon, I pass a massive traffic jam of people trying to go the other way. The reason is that land use controls at various levels sharply reduce the amount of housing in or near San Francisco, forcing many people who work in San Francisco to spend several hours every time commuting.

      I’ve seen the claim that between eighty and ninety percent of the land in the Bay Area cannot be built on due to a variety of restrictions, and since I have seen the same number both from people who thought it a bad thing and from people who thought it a good thing, my guess is that it’s correct. Where it is legal to build, most obviously SF itself, there are restrictions on what you can build that hold down the number of people per acre.

    • Esquire says:

      The major problem with saving money by living in a trailer park is that your neighbors will all be people who live in trailer parks. In a world where there were a ton of “Mr. Money Mustache” types, this would not be so bad… but, we live in the world we live in, and the current equilibrium dictates that you will not be around a lot of folks whose sons you don’t want your daughter marrying.

    • tb says:

      It is not at all unusual that owners of assets wish for asset prices to go up and that would-be owners of assets wish for asset prices to decline. Housing is an asset like any other – cars, stocks, 1986-1987 Fleer #57 in mint condition.

    • Why would people speculate on things nobody needs? I don’t think you understand the social utility of speculation.

    • Matt says:

      I thought it was going to be about housing as well.

      But actually, what you’re seeing is not a paradox with housing per se, but land. Our system of private, commodified land is inherently unfair and unworkable. Think about it: the land that comprises our country is privately owned by some fraction of the population. It’s crazy when you stop and think about it. How can we be equally free when some citizens literally own the space that constitutes the country, and the others are compelled to pay them for the right to be here?

      The politics of this are obvious: those who own the earth want to be able to charge those who do not more money. This is the true source of NIMBYism and all the manifold restrictions you inevitably get on building in dense cities. Who owns Manhattan? A handful of wildly wealthy individuals. Of course you end up with a byzantine set of rules regarding the construction of housing, and resulting rents that charge tenants crazy amounts for apartments that are little more than glorified closets.

      The solution is simple, but very painful to the ultra-wealthy: we socialize those land values, and make them our primary tax base.

  2. Pku says:

    Wait, I have an issue with the tulip thing: When the country was spending all its money on tulips, where was all the money actually going, in terms of actual utility? I mean, possibly the tulip farmers all got mega-rich and this caused a high level of income inequality, but that still seems like a second-order effect and not something that should bankrupt the country, especially if you adjust tax rates appropriately (If you start taxing tulips proportionately to how much they cost, the government shouldn’t even be losing much money, aside from what they spend on implementing a more convuluted tax system).
    (Granted, this kinda breaks the metaphor since giving people a college education actually does take work, but still. The tulip kingdom SHALL NOT FALL.)

    • DanielLC says:

      They’re not just buying tulips anymore. They’re buying the best tulips. Any tulip that’s not perfect gets thrown away. There may be many expensive things that can help improve a tulip. It’s not that the farmers are getting mega-rich for the same effort. To some extent they are, but they’re also putting large amounts of resources in order to produce sufficiently many perfect tulips.

      • pku says:

        Why would they be buying the same tulips? Presumably a mediocre tulip also works for proposals, and a rival farmer could just undercut their prices with them. (Because tulips have (somewhat) material value in enabling marriage, it seems like the dutch tulip mania might not apply).

        • Deiseach says:

          Look how much people are willing to pay for tiny chips of diamond in engagement rings. Some manky tulip with half its petals falling off just won’t cut it!

          And why on earth would I sell a tulip for fifty baxoons when I could sell the same tulip for sixty, eighty or ninety baxoons? It’s cutting off my nose to spite my face! Sure, I can outcompete my rival, but unless I drive him out of business, buy his land, start growing my tulips there and then hike up the price per tulip to eighty baxoons, then both of us charging sixty baxoons per tulip instead of him charging sixty and me charging fifty is a nice, cosy agreement.

        • Daniel says:

          A mediocre tulip works, but nobody wants to be the one guy that proposes with a mediocre tulip when everyone else has a perfect one. And the government is paying for it, so they have no incentive to pick a cheap one.

    • Steve Johnson says:

      Economics predicts that the money would flow to the owners of land suitable for tulip cultivation.

      Tulip farmers who aren’t also land owners’ rents would go up until the tulip farmers were making normal farm returns – otherwise someone else would bid more for the right to cultivate tulips on that land.

      Presumably the tulip farm owners would generously pay out bribes – reflecting the fact that they don’t actually fully own the land and that the value of the land is dependent on the continuation of the subsidies.

      I suppose you could calculate the ownership percentage based on the bribe budget.

    • Sherkaner Underhill says:

      Growing tulips isn’t free; land and labor must be allocated to tulip-growing from other industries at massive opportunity cost.

    • Deiseach says:

      It’s like diamonds; there is plenty of capacity for mining diamonds, but if you mine and sell all the diamonds you can mine, you reduce the price. Diamonds are kept expensive by artificial scarcity and controlling how many get onto the market to maintain prices. That’s why stories like H.G. Wells’ “The Diamond Maker” are not realistic in a market sense; if you can churn out diamonds as plentifully as paste jewellery, they will cost the same and be as valuable as paste jewellery.

      First when tulips get expensive, you raise and sell as many tulips as you can. But everyone else is doing that too. So you get smart and you breed special varieties of tulip and charge extra, and you don’t sell every tulip you harvest, and you form a Tulip Growers’ Guild that agree on who can and can’t grow tulips and all the rest of it.

      Secondly, the money is being made not so much by the growers as by the speculators. Borrowing money to purchase future crops of tulips, and repaying that money by your profits, then getting into more and more convoluted financial instruments where you’re robbing Peter to pay Paul, and all the time the whole edifice is resting on a demand that is unsustainable (tulips get so expensive nobody can afford them, or only the select very rich few, which is not enough to keep the wheel turning). The government is not interfering (light touch regulation) because hey, the country is (at present) awash with money from all this speculation, and you’re a business-friendly government, and this boom proves the wisdom of the market, right?

      My cynicism is in part derived from the past few days of the news in Ireland, listening to the bank regulator explaining how he knew nothing of what was going on and never asked questions because it wasn’t his job to interfere in how the banks ran their business. Because, you know, he was only the guy in charge of checking how they ran things to make sure they weren’t doing something stupid to crash the economy. He’s shoving the blame on the government of the day which strongly encouraged him not to poke his nose into things when everything was going so swimmingly.

      And that’s how we ended up with billions of debt and no money in the kitty to pay it.

      • Richard Gadsden says:

        The Tulip® Grower’s Guild registers a trademark in the Tulip® and then everyone else that grows Tulipa gesneriana can’t call them Tulip®s.

        If they can convince the people of the kingdom that only a real Tulip® is good enough to marry a woman, then they make out like bandits.

  3. Madplatypus says:

    I’m somewhat confused as to why the Government can, in your model, realistically control the contents of an interview – but not realistically implement any sort of price control or other anti-bubble mechanism on the bubble itself.

    • Chris H says:

      I was going to say that the reason is price controls restrict supply, something straight out of Econ 101 (and a contributing factor to the expense of housing in cities like San Francisco and New York). But then I realized, that’s EXACTLY what we want in this situation! Higher education is now mostly worthless signaling that makes everyone worse off but no one has incentive to escape it. Basic Moloch stuff. But what if we put a draconian price control with no subsidies to cushion the blow? Tomorrow we decree no college may charge more than $5,000 per year and no one gets a federal or state subsidy. What happens? A shit ton of universities have to close their doors, only a handful make it out to the other side and because they’re the only game in town they get to be super exclusive. College becomes rare again and employers can’t demand sheepskins because otherwise there’s not enough people to hire.

      And think of how easy this would be to sell to the American people! Most of them don’t have a good conception about how price controls actually impact the economy so they would just be thinking “yes I/my child gets college for cheap!” You can even convince the conservatives because it’s a) less money subsidizing liberal academia and b) a generally cheap policy. There will be a nostalgic generation of college goers who are going to eventually be pissed when they realize their kids won’t go to college, but fuck them because by that time the equilibrium will reset and college won’t be needed for a half way decent job anymore!

      I didn’t think a situation would come up where I’d say this but…let’s do price controls!

      • Halfwitz says:

        This is a brilliant judo move. How does one get this meme started?

        This whole line of thought of applying bad policy to bad policy seems very promising.

      • Cauê says:

        You’re missing the part where the public demands the government to provide the “college for cheap” they think they were promised, when those greedy capitalists refuse to do it.

        • Here in New Zealand, all the Universities (our word for “college”) are Government-run. [PS – that’s over-simplified, but the subtleties aren’t relevant in this context.]

          Our economy hasn’t suffered for it, or at least we’re still standing.

          • And we have basically-guaranteed govt loans, and still have relatively cheap tuition fees. My university isn’t Harvard or anything, but it’s somewhere in the top 250 in the world overall, and even doing a supposed-to-be five-year double degree there’s no way it’ll cost me more than 50K on my loan. Likely closer to half that, (and that’s including the optional ~1K a year for ‘course costs’ that I’m mostly just putting in a savings account and accumulating interest.)

          • RCF says:

            Why are there no private universities? Is there no market for them, or are there laws against them? PS they’re called “universities” here, too. People go to college at a university. The nomenclature is a bit strange (technically, a college is a subdivision of a university).

          • Jesht says:

            Australia is similar. There are private universities around, but they’re generally not as big or prestigious. Our model includes a lot of ‘full-fee’ paying foreign students, who will pay ~$40,000 p.a for a degree vs a publicly subsidized local student who will pay ~$10,000 p.a. I expect that our system is propped-up by these foreign students.

          • @RCF: for context and the sake of completeness, there *are* private tertiary institutions, i.e., polytechs, language schools and the like, eligible (subject to conditions) for the same per-student subsidies as the public universities, and at least one of which ( is in fact officially a university – though IMO it doesn’t really qualify by the traditional (as opposed to the legal) definition.

            In order for a new privately-run university to be eligible for the per-student subsidies, it would have to justify to the Government why a new university was needed – what unoccupied niche of the tertiary ecosystem it would fill, so to speak. Also I believe it would be subject to the limits imposed by the Government on how much they could charge the students. It would be somewhere between challenging and impossible to make any sort of profit under those conditions; the public universities are always struggling to make ends meet as it is, and if I understand rightly they get grants as well as the per-student fee subsidies. Most of them don’t have to pay a lease, either, they’re located on land the government gave them ages ago.

            As far as I know, a private university that was *not* government subsidized would be legal (though in order to call itself a university it would have to meet quality standards) but I don’t think there would be enough of a market. Perhaps if a world-famous university, Harvard, Oxford, or whatever, opened a branch … but then our public universities aren’t exactly unknowns in the academic world either.

            That’s the status quo as I understand it; it has of course changed over the years. When I went to university, for example, I don’t think private institutions were eligible for any sort of government funding or subsidies.

          • RCF says:

            I find it odd that there are private universities in the US, receiving no subsidies (other than indirect through student loans) and charging more than public universities, and yet whatever market forces produce them in the US don’t produce them there. Besides elite universities, there are also religious ones. Do religious universities qualify for subsidies? Are there limits to how fundamentalist they can be?

          • BBA says:

            Those indirect subsidies through student loans can make or break a school. Southeastern University in Washington DC, for instance, shut down when it lost its accreditation, because unaccredited schools are ineligible for federal student loans.

            There are a handful of entirely unsubsidized colleges in America, which for reasons of religion or ideology (but I repeat myself) have opted out of the accreditation/student loan system. Their impact is negligible.

          • @RCF: I didn’t know the US had public universities; if they’re cheaper, why doesn’t everybody go to them? (The answer to that question will very likely help answer the question of why we don’t have them.) In the meantime, I have a few thoughts.

            (1) Our public universities are probably already pretty much as elite as any NZ university is capable of being, given our size.

            (2) – cancelled – I was going to say that we probably don’t have enough people rich enough to send their children to a completely unsubsidized university, but if I’m reading the statistics correctly as many as 20% of households are earning $130k or more and might therefore be able to afford $50k/year. That could in principle support one private university of roughly the same size as the existing public ones.

            (3) Since we don’t have a history of private universities that are better (or thought of as better) than public ones, employers have no reason to prefer applicants from private universities, so there is no reason to think you would gain any advantage from going to one.

            (4) We *do* have a history of what you might call anti-elitism (or “tall poppy syndrome”) so it is even possible that some employers would be prejudiced *against* people who went to an “elite” private university, if there were any.

            I have no idea whether a proposal to open a (subsidized) university specifically for students belonging to a particular religious sect would be allowed; probably not, is my guess, on human rights grounds. (That is, I would imagine that in order to qualify for funding, tertiary institutions have to select amongst applicants in a non-discriminatory way.)

            Even if that isn’t true, I would imagine there would be considerable public resistance to the idea. I’m not sure there would be enough potential students, either. Even among those who identify as religious, there aren’t many whose religious belief is that extreme.

          • Gbdub says:

            @Harry – most public univerisities in the US are funded by states, and are substantially cheaper, but only for residents of that state.

            Some of them are regarded as elite – Berkeley is a California public school. The Universities of Virginia and Michigan, U. Texas at Austin, and a few others are also all considered very good, probably the match of any private school other than the best of the Ivy League.

            But as tuition costs continue to rise and state budgets get strained, the fraction of funding coming to these universities from their states is getting smaller. I know at Michigan their were rumblings about “going private” if the state support got too small, and the state occasionally gripes about how many out of state and foreign students the university accepts while turning away over half the in-staters who apply (the “full cost” out of state and foreign students help shore up the budget).

            However many states don’t have schools at that level, and their best students will try to attend an Ivy or one of the top publics at “out of state” prices that more or less match a private school tuition.

      • Madplatypus says:

        As tempting as your idea certainly is, the methods for determining which students get to have a college education will quickly become extremely controversial.

        However, not all controls on price need take the form of a strict cap. I do not believe that the levels of tuition charged today are necessary to achieve the level of education provided, and that the system for doing so can be simplified.

        Limited college education, and the vast status that would be increasingly attached to it, might also exacerbate standing problems in wealth distribution.

        • Paul the Apostate says:

          Which, for some political sectors, might be a definite plus.

        • Careless says:

          If the colleges can’t spend and/or charge a lot of money, they don’t care about the wealth of prospective students, so the only wealth distribution problem I’m seeing is a heriditary meritocracy

      • Lambert says:

        I know higher education in the uk was price controlled. It was a huge fiasco for Nick Clegg when he allowed the price ceiling to rise.

        • Peter says:

          It’s more complicated than that. Basically what you’ve got is a system partly funded by the state and partly funded by fees. The Lib Dems had a manifesto commitment not to raise the level of the fees – this got broken, and large parts of the anger are about the broken promise. At least that’s the view in Lib Dem circles – there’s a leadership election on and one concern is picking someone with the integrity not to do something like that, and to be seen to be the sort of person not to do something like that again.

          Their Conservative coalition partners had no such commitment, and they don’t seem to have had any collapse in their support, alas.

          • aguycalledjohn says:

            To further complicate things, you don’t pay at the point of entry but take out a government subsidised loan, that you only have to pay back once you are earning above a certain amount. The liberal democrats raised that threshold significantly, so would argue the policy was actually more progressive, but by then the damage was done.

        • AlphaGamma says:

          It is not entirely price-controlled. There are private universities that are allowed to charge what they like but receive no government funding (from the education budget, the one private research university may receive government research funding).

          Occasionally one of the better-known universities makes noises about “going private” if the government cuts the education budget (and hence the subsidies). Note that universities in the UK, while “part of the state system” in terms of receiving government subsidies, are de jure entirely independent bodies not branches of the government.

      • Deiseach says:

        college won’t be needed for a half way decent job anymore

        Yes, it will. Oh, maybe not the “four year basic undergraduate degree” but the days of “high school qualifications got you into an entry level office job and you worked your way up from there” are gone.

        Manufacturing industry is overseas or automated, and getting increasingly so. Joe the Ford Assembly Line worker who made a decent living isn’t there anymore, his job is taken over by robots. Joe the Plumber is doing a bit better, but that’s an apprenticeship system (and vocational qualifications are necessary there). There is still room for unskilled labour in construction, where you walk on site and get hired, but that’s equally being paid “on the lump” and it’s transient, temporary, often done by illegals, and paid under the counter with no taxes and (if you get hurt on the job and you have a good chance of getting hurt on the job) no social insurance or health insurance payments to help you live on while you can’t work or need hospital care.

        Employment is shifting to services industries, which traditionally have been less well-paid. If you want anything better, you need a college qualification.

        Or some kind of qualification. The future is in tech, and not just “working on the assembly line soldering transistors”, it’s programming and development. You may not need a degree, but you need something like App Academy, or some way of proving “I can code, I can debug, I can run your social media presence and create value and make profits for you and invent the new must-have products that will make the iPhone look like Granny’s windup phonograph and even better write the apps that will run on these”.

        And right now, walking in off the street with a portfolio of “Here’s what I did last week that’s up on Google Play” will get you there, but even so – that kind of approach relies heavily on networking, contacts, who do you know who works there and can get you a leg in the door; eventually it will go back to “what college or academy did you go to? who did you work with?” and qualifications of some kind will be needed.

        And not everyone can programme. Sorry, lads, I know ye love to say “Oh, you can learn [language] from a few lessons on the Internet in no time!” but not everyone can do that.

        • Surlie says:

          That’s why I think things like App Academy are the future. It just makes more sense to pay less for a quick-paced, tightly-focused trade school that teaches you exactly what you need to know than to spread that out over a 4+ year vacation from the real world at a purpose-built miniature city that includes unnecessary classes, extracurricular activities, dormitory housing, etc.

          Maybe the problem is that the expansion of college was only enabled by the unsustainable postwar economic boom, and maybe trade schools are a better solution.

          • Daniel Armak says:

            In the defense of college, most would-be programmers study computer science as undergrads, which is not at all intended to teach programming, and only does so under protest. A good 3-year software engineering program may well out-perform the 3-month App Academy course by a factor of 9 in terms of knowledge and skills imparted. Whether people should start programming without that much knowledge is a different question: as an employer of programmers, it depends on what I want them to write.

          • Deiseach says:

            That’s the American model. I can’t speak for universities elsewhere, but “purpose-built miniature cities with dormitory housing” weren’t part of Irish universities. Even when they did build on-campus housing, it’s often for first year students only and you have to find your own digs after your first year.

            There are a lot of charges and costs besides fees that go along with going to college in Ireland.

            I do agree, though, that vocational training is under-valued (it’s seen as the inferior sibling to academic qualifications).

        • JDG1980 says:

          The future is in tech, and not just “working on the assembly line soldering transistors”, it’s programming and development. You may not need a degree, but you need something like App Academy, or some way of proving “I can code, I can debug, I can run your social media presence and create value and make profits for you and invent the new must-have products that will make the iPhone look like Granny’s windup phonograph and even better write the apps that will run on these”.

          OK, so what are all the people with IQs of 100 or less supposed to do? Should we just acknowledge that society can’t provide decent jobs for them any more, and establish a Universal Basic Income?

          • Deiseach says:

            JDG1980, it’s going to have to happen. I’m probably one of those “IQ 100 or less people” and I have some kind of clerical training which helps, but even though you still need to be able to use an old electric typewriter for certain tasks in the job, it’s all moving more and more to computerisation.

            Which is why, on my last upskilling (that’s what it’s called nowadays) course during a period of unemployment, there was a guy there doing an IT course who had been a butcher, lost the job because of Repetitive Strain Injury, and was now being sent on an employment training/upskilling course.

            He had no interest in office work, didn’t see what use this course was to him, but that’s where the perceived job openings were and that’s what the training was for.

            Services industries will continue to operate at least until they invent robot waiting staff and hairdressers and so forth, which I don’t see happening for a good while yet, but the kind of manufacturing jobs where you got a job on an assembly line at the automobile plant is gone, it’s going in the pharmaceutical industries (that’s where my brother works) and it’ll go in the tech industries as well.

            Being both mathematically gifted and high IQ enough to teach yourself to code is great, but not everyone can do it (I certainly can’t). So what is the future of employment going to be like? I can’t forecast. But I do think governments will encourage people to go for college, because businesses need college grads and the economy will depend on high-value, high-skill industries, and education is not going to be about “cultivate the individual”, it’s “what are the needs of business and industry” (I’ve said it on here before about the press releases from Irish governments about “education and the needs of the economy”).

            If you want your kid to have any chance of a job, never mind a good job, you’ll try and send them to some kind of third level college, even the University of Podunk (formerly Podunk Polytechnic). Even that may not be enough; from this article about what new graduates in Ireland need, from the employers’ viewpoint:

            So, if you come out top of your class, have multiple languages and extensive IT skills, you’re a shoe-in.

          • Deiseach wrote:

            I’m probably one of those “IQ 100 or less people”

            No, you’re not.

          • Careless says:

            I’m probably one of those “IQ 100 or less people”


            Go to popular youtube videos. compare your writing with the average comments there.

            No, you do not have a 100 IQ

          • jaimeastorga2000 says:

            I’m probably one of those “IQ 100 or less people”

            You know this is not so. Kindly stop fishing for compliments.

      • chris says:

        What would stop immigrants with College Degrees coming in from other countries and soaking up all the good jobs?

        If College Degree>No College Degree


        Foreigner (say Englishman) with College Degree be > than American without College Degree?

        • aguycalledjohn says:

          Nothing in theory. In practice getting a work permit for America, even as a person with a degree, is ridiculously hard, and many overseas qualifications are not recognised.

          [Whether this is a deliberate strategy to keep americans in jobs depends what you think of the competence of the govenment.]

        • Careless says:

          There’s no way our current immigration system could remain in that situation, when even the children of the elite were being shut out of university educations.

      • maxikov says:

        I actually suspect that in case of universities, price control may not be as detrimental as it would be in other areas. This is because they tend to have a crapload of luxuries that are completely irrelevant to the ability to provide even good education, leave alone just a degree – gyms, fraternities, huge campuses with vast empty spaces and old buildings, tenured professors teaching 101 classes, instead of giving a link to a MOOC, and one TA to answer questions, etc.

        It should be possible to massively reduce the cost of a college degree, but only if everyone does that at the same time – otherwise that could compromise the ability of a single university sacrificing those to compete with others.

      • Izaak Weiss says:

        But I want people who want to go to college to be able to go to college, not prevent anyone from going to college.

        • Gbdub says:

          I want a Ferrari. Why do you prevent me from having a Ferrari? Transportation is a human right!

          Oh, you mean to say a Ferrari isn’t any more useful than a Camry for getting to work, so you don’t see why we should subsidize a $100,000 thing that only provides $20,000 of practical value? Carry on then.

          • Careless says:

            I don’t see him buying you a Camry, either. Even used.

          • Gbdub says:

            Good point. So the problem isn’t lack of Ferrari subsidies, it’s a lack of anyone selling used Camrys, coupled with a failure to acknowledge that a Camry would be just as good for getting one’s arse to work.

      • Gbdub says:

        You don’t even need price controls. Just cap government subsidized loans, and put colleges on the hook for repaying them if students can’t get a job and default. Then, nobody goes to college (and colleges don’t accept them into a program) if it’s not reasonably going to help them get a career.

        Right now there’s no downside for colleges jacking up their prices (often for nice-to-have but totally impractical/unnecessary amenities and departments) to the stratosphere , and we’re subsidizing the hell out of it publicly with loans that are unlikely to be paid back. Give the colleges some skin in the game, and they’ll be incentivized to trim the fat.

      • Anthony says:

        and no one gets a federal or state subsidy.

        This is what takes your proposal from briliant to sheer fantasy.

      • Steven says:

        But what if we put a draconian price control with no subsidies to cushion the blow? Tomorrow we decree no college may charge more than $5,000 per year and no one gets a federal or state subsidy.

        You’d need to pass laws in 51 jurisdictions to do that thanks to federalism.

        However, a “soft” Federal price control whose enforcement mechanism was preventing Federal funding/student loan guarantees/student loan non-dischargability to institutions charging more than $5,000/year would have a lot of the same effect.

        • Careless says:

          Leaving the multi-billion dollar endowment schools, the state flagship schools, and not much else standing, I suppose

          • Subbak says:

            If there is a demand for degrees, people will get them at institutions that offer them for cheaper. Institutions without huge-ass campuses, prestige buildings, or absurd salary for administration.

    • Sherkaner Underhill says:

      Employers aren’t interested in whether or not you have a college degree because it’s a good indicator of job performance. (And it *isn’t* a good indicator of job performance.) They’re interested in whether or not you have a college degree because having a college degree is high-status. And generally speaking, being an entity that discriminates against protected classes is very low-status. So the ban is almost self-enforcing: status-concerned people won’t even *want* to discriminate against non-degree-holders any more.

      • Will says:

        Having a college degree is actually pretty highly indicative of performance. It’s not as indicative as previous work history, but it’s still a decent indicator.

      • Paul the Apostate says:

        There’s nothing low-status about high-status people discriminating against low-status people. That right there will militate against lack of college degrees – currently a low status – becoming a protected status.

        Besides, every other protected class is something you’re born into or thrown into by Act of God. The great bugaboo these days is transgender people, who have opted to change, being granted any protected status. Never mind that they’re discriminated against. They CHOSE it, how dare they arrogate benefits to themselves, end of discussion. Expect the same reaction re the non-college-educated.

        • Daniel Armak says:

          Going to college costs a huge amount of money, so poor people naturally attend college at lower rates. (Even if college were free, it has the opportunity cost of not earning money for several years.) And being poor is already a protected status. On this view, ‘choosing’ not to go to college is like ‘choosing’ not to earn more money and staying poor.

        • Taradino C. says:

          Religion is a protected class, even for people who clearly chose their religion (e.g. people who converted as adults so they could get married in their spouse’s church).

      • JDG1980 says:

        Employers aren’t interested in whether or not you have a college degree because it’s a good indicator of job performance. (And it *isn’t* a good indicator of job performance.)

        A college degree indicates that the degree-holder likely has:
        * A moderately high IQ (105+) – maybe higher depending on the college and major
        * Reasonable ability to conform to a moderately structured environment

        The reason employers don’t just use IQ tests for the first requirement is that IQ tests are considered racial discrimination via disparate impact, because on average, African Americans do about one standard deviation worse than white Americans. The reason employers don’t use high school diplomas for the second requirement is that high school diplomas don’t mean anything any more; we used to flunk people from high school and allow dropouts for those who clearly weren’t interested, but too high a proportion of the flunkees and dropouts were African American, so we started giving high school diplomas to anyone who could fog a mirror.

        • Ptoliporthos says:

          So why not just let employers use IQ tests but do what colleges do with SAT scores, and spot African American applicants an extra standard deviation? They could re normalize East Asian scores down too, while they’re at it, just like colleges! Then private industry won’t have to outsource discrimination to colleges anymore.

        • Is there any reason given here why looking at college degrees doesn’t count as discrimination via disparate impact?

          • Anthony says:

            Because lawyers all have better college degrees than you do. And they wouldn’t want to reduce the value of their fancy degrees, or give jobs to people who went to the wrong colleges. And if the lawyers won’t get on board, you’re not going to win that battle.

            Besides, people without college degrees aren’t very likely to come up with the idea, and/or have the fortitude to push through such a campaign. If they did, they’d have a degree.

          • Anonymous says:

            …because no one has put a case through, yet.

    • Adam says:

      In all likelihood, the price of college would stay mostly the same, but only $5000 of it would be directly monetary. Students would have to complete some kind of signaling to get into college that on average would end of costing them the same amount as the part of tuition that was made illegal.

      • Honeas says:

        Now that’s just non-sense. And if it ever got to that (which it wouldn’t) then one could do to the colleges the same thing that was done to the employers.

    • Scott Alexander says:

      This seems to work in real life – I know many interviewers, and they all talk about how strictly they are told not to ask any questions about protected classes (race, religion, etc) in interviews. Presumably they know that if they ask them, and the person doesn’t get the job, they can sue.

      • aguycalledjohn says:

        How likely are interveiwers to switch to better models of performance for distinguishing between otherwise equally qualified candidates, or will they instead switch to other signalling mechanisms? (E.g. race, accent, unpaid internships).

        • Anthony says:

          Hiring managers would love to switch to better ways of distinguishing actually qualified candidates from those who only look good on paper. If something actually worked, and wouldn’t get them in trouble, they’d switch.

          I’m not sure HR people would actually switch, even if they could be convinced that a more accurate evaluation procedure wouldn’t create liability for them. I’d expect a lot of HR types to do everything they could to block such a change, unless it was one which could be done by computer more easily than filtering on lousy keywords.

        • Adam says:

          I’d like to know what these models are. When I first got out of college, before I ended up joining the Army, I applied to be a prison guard. They gave us a standardized intelligence test, a psych profile test, a physical test, and conducted a year-long background investigation. Does this actually result in prison guards being better at their jobs than in industries that just screen resumes and conduct interviews?

  4. Halfwitz says:

    Is anyone makeing a killing underpaying smart people who lack college degrees?

    • DanielLC says:

      I suspect small businesses are more equipped for using non-standard methods to gauge how good employees are. I’ve been told the company my brother works at (Parallax) has virtually no correlation between education level and pay. The CEO didn’t go to college. My mom who did go to college worked there as a receptionist.

      By and large, this isn’t caused by people being stupid. It’s signalling. People who aren’t willing to go to highschool even though they don’t have to pay for it are genuinely people who you probably shouldn’t hire. It’s just that it has nothing to do with their book learning, and paying for highschool for everyone probably didn’t do much to help.

      • Error says:

        Datapoint: My first IT job was at a small computer repair shop. The boss’s method of gauging potential employees was to ignore their educational background (he never even asked for mine), throw them at a few broken customer machines, and see how well they did.

        Best boss I ever had. Didn’t give a fuck what I did as long as stuff that came in not-working left working. The pay was crap, but the work experience seemed to be a semi-acceptable substitute for a degree when I moved on a few years later.

        This is interesting because it probably *does* qualify as “underpaying smart people who don’t have college degrees”. But I don’t feel like I was taken advantage of; I got the resume-equivalent of the degree and got paid for it instead of paying an institution. Up until that point I had found nothing, precisely because all entry-level jobs required a degree (and, usually, two years experience, but that’s a different rant).

        This probably doesn’t generalize outside of industries where the main difficulty is getting in in the first place, and once you have a history life gets much easier.

        • AlexanderRM says:

          I think most people on places like SlateStarCodex (since rational consequentialism and understanding basic economics are both common here), when they say “underpaying smart people who don’t have college degrees”, don’t generally think of it as “taking advantage of them”. I read Halfwitz’ idea as being… more like the way you regarded it: Nobody else would hire you and they would.

          That’s good for both the smart people who lack a college degree, and for the consumers. The fact that the company makes money by doing so is an incentive for companies to use ways of finding the smart people besides college degrees.

      • Brett says:

        Google confuses me, because having just interviewed with them they’re manifestly trying to give candidates an intelligence test (5 interviews, each of which went directly into a hard problem, with absolutely no discussion of anything except the problem), except that their methods of doing intelligence testing are not very good. But then, good intelligence tests are, as I said below, illegal for hiring purposes.

        • HlynkaCG says:

          Cynical bastard mode:

          Intelligence tests illegal for hiring because if they weren’t companies might use them instead of sheepskins.

          Teachers unions professors et al. know this and spend a lot of time pestering congressmen and other officials to make sure that intelligence tests STAY illegal and also ensure that as many jobs as possible require degrees.

          If anyone resists they are labeled as stupid / ant-intellectual, hating children, and denying young people the same opportunities that they had.

          /End Cynical bastard mode

          • suntzuanime says:

            It’s nicely cynical, but my understanding is that countries other than the US that don’t have the same sorts of disparate impact laws still give college degrees a lot of weight when it comes to hiring and don’t widely intelligence-test job candidates in lieu of degrees. It’s possible this is because people actually learn something in college, or because what college tests for is more complex than straight intelligence and hard to design a good test for.

          • The actual reason IQ tests for employment are effectively illegal in the U.S. is not the teacher’s unions, though they’d be plenty evil/statist enough to get them (effectively) banned if another interest group with clout hadn’t gotten there first.

            No, the people we have to thank for this are the racial-grievance peddlers. IQ testing has a disparate statistical impact on blacks, you see.

            Therefore they are Not Allowed, and the predictive utility of IQ must become a taboo subject, and we all have to pay huge costs for galloping credentialitis and an inefficient vetting system that does at least as much injury to black people as the IQ tests putatively did.

            Isn’t social engineering wonderful?

          • Sol says:

            Pretty sure you’ve got cause and effect backwards here. Back when intelligence tests for hiring were legal, college degrees were not required for basically every decent job. When they were made illegal, companies decided a college degree was a pretty good still-legal proxy.

          • Edward Scizorhands says:

            The ruling of Griggs v Duke bans the use of high school degrees the same way it bans the use of IQ tests. Very few people have internalized this because there haven’t been lawsuits about it. The same logic presumably bans college degrees, too.

            So instead of waiting for Bernie Sanders to make education a protected class, all we need is a few aggressive trial lawyers to start filing lawsuits against fire departments that require college degrees.

            Even if the fire department wins, it will be expensive, because it will be the fire department’s job to prove that the college degree actually matters. (IANAL but this is my understanding of how disparate impact works from talking with lawyer-friends.)

            Once a few of these lawsuits catch fire in the news, HR departments will notice and stop requiring college degrees for office managers.

            If someone wants to crowdfund the first of these lawsuits, I’d consider it part of my effective altruism to donate.

          • Foo Quuxman says:

            Isn’t social engineering wonderful?

            “We meant it for the best, to make people safer, Oh GOD!” *screaming*


            Indeed, though now that the system works this way there is a big fat interest group to push things the way HlynkaCG speculated. That many of those people are also part of the Grievance Maximiser Club just sweetens the deal.

          • Daniel Ford says:

            I teach community college and every time I try to explain the signalling model of higher education to my colleagues they never believe me. At this point I’ve given up on talking about it with them.

            So no, we’re not driving this system. Doing that would require us to accept the signalling model, which is never going to happen.

          • mico says:

            There’s a good explanation why IQ tests aren’t used in the US, but in many other countries IQ tests are legal and yet they have not replaced college.

            One explanation is that employers appreciate being able to select on the basis of IQ and “hard work”/compliance, rather than IQ alone. But at least in my country employers seem to make little attempt to discriminate between majors and colleges based on the (often significant) differences in IQ and hard work required.

            I suspect that most companies actually do not have that much incentive to choose the best graduate employees. People reach peak earnings in their 40s and 50s, so with a handful of exceptions it’s these experienced workers who will be making most companies the most money. If you hire the wrong 20-somethings, well, you can just up your bid for people who have a track record rather than just a college degree, which is exactly what they do. Degrees are a cheap (for them) filter that is not totally worthless.

          • Schmendrick says:

            @ Edward

            Griggs v. Duke may not be applicable here, because that case is primarily concerned with disparate impact among minority groups. I don’t really see what grounds a court could justifiably point to if they were to try and throw out college-degree requirements for jobs. After all, the Griggs test is whether or not the required qualification bears “a demonstrable relationship to successful performance of the jobs for which it was used.” Assuming that businesses aren’t actually putting “must have a college diploma or equivalent” in classifieds for barista positions, there’s a plausible argument that requiring a college diploma to be, say, a clerk at a steel company, could in fact bear just such a demonstrable relationship. There are lots of holes in Griggs.

          • Edward Scizorhands says:

            Here’s one of many news articles about college degrees being required where traditionally they weren’t, often explicitly, even for the company runner for a law firm:


            So the pattern to stop this:

            1. Recruit an employment lawyer.
            2. Recruit a high school graduate to apply at that firm for the runner position.
            3. Have the candidate get rejected.
            4. Sue.

            This is (vaguely) how Rosa Parks did her bus boycott: she wasn’t a random person who just decided to take a stand one day; she was an employee of the NAACP who planned to create the best possible test case.

        • sh says:

          AIUI, the idea of those interviews is not just to test for intelligence. In the case of engineers, the tests are for a mix between general intelligence, domain-specific cognitive abilities, and a high level of interest in software systems.
          And all of these are relevant. The tests are highly loaded towards specific skills, but the idea is less to test those skills because that’s exactly what will be needed, than it is to test for them as a proxy for the candidate’s ability and inclination to pick skills of this sort up.

          This is important. You don’t want to hire highly intelligent people for a job where they’ll need to do work that they just have no interest in. You don’t want to hire natively highly intelligent people that have sufficiently biased minds that they’re unable to pick up a basic working understanding of engineering, either. So the interviews are for a mix.

          • Paul the Apostate says:

            I know a lot of shops that wouldn’t want to hire people of high native intelligence, period. They don’t do well with organizational chickens**t, they’re too good at spotting managerial incompetence, and they tend to have, and value, lives that extend too far outside the office.

          • sh says:

            Google is definitely not in that category – the interviews do strongly test for intelligence, and I suspect candidates for engineering positions have no chance at all unless they’re significantly above average. But even so, there are other things that are relevant to test for.

        • Eric Rall says:

          Former Googler here (Test Engineer).

          Google interviews work the way they do largely because they’re done by a broad sampling of engineers working at Google, and most engineers are bad at interviewing. We tend to gravitate towards asking hard coding problems because asking hard coding problems is the easiest and most obvious way for engineers to interview potential engineering hires.

          When I was there (2010-2013), Google was actively trying to get its engineers to use interviewing techniques which have been empirically found to be better predictors of job performance (mostly Behavioral Interviewing, which features questions of the forms “Tell me about a time when…” and “What would you do if …”), but weren’t getting much traction.

    • speedwell says:

      A data point, but I know I am far from uncommon.

      I am an American citizen living in Ireland with my Irish husband. I was laid off from a US-headquartered multinational firm in 2014, just nine months after my husband joined me in the US with his green card, partly because I was the “least-qualified” member of my team (no college degree). I was just offered basically the same job in a multinational firm in Ireland, headquartered in the UK. Still no college degree. After reviewing my CV and references list at a lengthy interview, one of the interviewers literally said, “I can’t see how they could afford to lose you”.

      To answer your actual question, though, yes, I was formerly paid less and had less job security than the members of my team with the same job title and duties who had degrees, despite having at least equivalent experience and performance, or better.

      Edit: Just found this, apropos:

      • Richard Gadsden says:

        I was formerly paid less and had less job security than the members of my team with the same job title and duties who had degrees

        I appreciate that this is in the US where you don’t have useful sex discrimination laws, but…

        UK equal pay law (and my understanding of the EU equal pay directive from which it is derived) is that is that if you can demonstrate that your work is of at least equal value to the company as another employee (the “comparator”) and you’re being paid less than the comparator and you’re of a different sex to the comparator, then you win, and you’re entitled to back-pay of the difference in pay, back to when you were first employed or a maximum of seven years with interest plus your legal fees have to be paid by your (ex) employer. And (unusually in the UK) class-action suits are allowed. If, in the court’s view, the employer’s behaviour was egregious (e.g. they should have known there was a breach, there’s evidence of actual sex discrimination, etc), then they can double or triple the compensation.

        Note that this goes much further than disparate impact – if job A is 80% male and job B is 80% female, then a man doing job B can sue for sex discrimination because there is a woman doing job A who gets paid more than him.

        Quite why they don’t just turn this into a general fair pay law, where you don’t have to demonstrate that the comparator is of the other sex, I don’t know – the effect would be practically the same.

        • mico says:

          Possibly because the government does not actually want to create some kind of Gosplan to work out what people “should” be being paid in any arbitrary comparison. I don’t understand how that can be knowable even in principle in the (more limited) example you gave.

          The effect of such a law with unlimited scope and rigorously applied would essentially be to nationalise the entire economy.

    • It’s not just about intelligence. Having a college degree also shows you’re able to show up to classes everyday and willing to conform to what society expects of you.

      • Sol says:

        You showed up to class every day?!

        • I went to MIT so for most classes I needed to. I’m given to understand that certain other schools like the one up the river don’t require as much but managing to graduate still shows you’ve got it together to a certain extent.

      • Harald K says:

        A good deal of what IQ tests measure is also willingness to conform. They’re usually of the underspecified puzzle type, where you must figure out what’s expected of you, what the test maker wants you to do. That’s a skill you’re probably not likely to have developed much unless you care what people expect of you.

        • Nathan Cook says:

          Conformity is almost as easy to test as IQ, so it should be rather easy to create an “IQ corrected for conformity” measure.

          • Paul the Apostate says:

            Not if no one will come clean about the nature, extent, or importance of conformity. Everyone wants it imposed on others, no one wants it imposed on themselves, and no one wants to be pinned down about where or why it might be needed (or not needed).

    • Lambert says:

      Self-employment? Tech startups? Other jobs whose continued existances are predicated on competance of the worker(s).

    • Berry says:

      The IDF.

    • Deiseach says:

      I think they are, but I can only go by anecdotal evidence: person of my acquaintance who was self-taught, could use the systems, was capable of learning on the fly, but because they didn’t have the Magic Certification, was not even considered for the job interview or when employed, was paid less for doing the work because “Well, you’re unqualified, aren’t you?”

      College degrees are supposed to get you better jobs with better money. The system is self-perpetuating; if you know you can do the job, but you also know you’ll be paid less for lacking the qualification, of course you go out and get the qualification so you can ask for the higher salary. And a company can justify paying Bob more than Jim by saying “Yes, but Bob has a BSc or MSc or PhD and Jim doesn’t; all his learning was done on the job”.

    • Yes, many startups and other poorly funded tech companies do exactly this. It’s not super difficult to gauge whether someone can code or not based on interviews/code samples/etc, so why bother with degree?

      I personally scan resumes for a github link, ignore everything else, then I give a relatively standardized coding and statistics test.

      The only time I actually pay attention to job history/education is my “are you lying” question – I pick a technology off the resume that the candidate claims to know well and grill them on it.

    • Is anyone makeing a killing underpaying smart people who lack college degrees?

      Yes! The U.S. Postal Service. And it’s one of the reasons why the United States has the best and cheapest* and most reliable postal service in the world.

      * cheapest, I mean, adjusted for national per capita income, and taking account of subsidies. And if you dispute “best” and “most reliable”, then you’re probably an American with zero experience of other countries’ postal services.

      The Postal Service is practically unique among employers. All kinds of personal characteristics that would cause a person to be rejected in a job interview are completely ignored.

      Something like a bad facial tic could be a steep disadvantage in most hiring situations, but it makes zero difference to the Postal Service.

      This happened by completely by accident. When the Post Office was de-politicized, it was necessary to take away any discretion in choosing who to hire, because at the time, all of the incumbent managers were political appointees, and it was assumed that if they were allowed any discretion, they would continue to hire their political friends.

      Hence, all hiring is based on competitive exams.

      Why is the Postal Service allowed to use intelligence tests (very lightly disguised) for hiring, when nobody else can?

      Probably because IQ-related skills are BFOQ’s: bona fide occupational qualifications. Mail sorters need to have enough brainspace to memorize thousands of street names and address ranges.

      Most people who are capable of doing this kind of work have many other, more lucrative options. The Postal Service has prospered by finding people who don’t have other options.

      To get hired by the Postal Service, you have to (1) do well on an exam very much like the quantitative SAT, (2) really want the job, (3) be willing to wait months or years to get through the process, and (4) be available at a moment’s notice.

      In other words, if you are already employed or have other things going on in your life, you’re not likely to get all the way through the process.

      • Foo Quuxman says:

        And if you dispute “best” and “most reliable”, then you’re probably an American with zero experience of other countries’ postal services.

        Annnnd there is my dose of horror for the month.

      • JerhumeBrunnenG says:

        Most people who are capable of doing this kind of work have many other, more lucrative options. The Postal Service has prospered by finding people who don’t have other options.

        Hmm. This reminds me of various organizations whose names begin with “Night” and end with “Watch”.

        • Hmm. This reminds me of various organizations whose names begin with “Night” and end with “Watch”.

          Confession of ignorance: I have no idea what this is a reference to.

          • jaimeastorga2000 says:

            Game of Thrones.

          • Lupis42 says:

            Discworld – specifically, the Vimes books.

            Eta:should have thought of Game of Thrones, that’s an even better fit. Discworld will always own the “Night watch” name in my headpiece though.

          • Linch says:

            Is GoT more popular than Discworld among nerds now?

            …Annnnd there is *my* dose of horror for the month.

          • Nornagest says:

            TV is a tremendously bigger market than books. I’ll bet even a relatively unsuccessful nerdy TV show, like… I’m tempted to say Firefly, but Terminator: The Sarah Conner Chronicles is probably a better example… has a bigger fandom than even a relatively successful nerdy book series.

            That being said, I gather Discworld is something of a cultural institution across the pond. Here it’s more the sort of thing you’re likely to have read if you’re specifically into literary nerdery.

          • AlexC says:

            Huh, and here I thought it was a reference to the fantasy film series actually titled Night Watch.

            It is indeed distressing to me if Game of Thrones is more popular among nerds than Terry Pratchett, but that might at least be a pond-side difference.

        • JadeNekotenshi says:

          Is it weird that my first thought here was of Babylon 5’s Nightwatch? I was a little puzzled how it pertained to a domestic spying organization, though.

      • Bruno says:

        I am a German who lived in the US for a while and I strongly dispute “best”. Otherwise, really interesting insight, thx.

    • Saint_Fiasco says:

      Everyone who outsources. Hiring people with no college degrees is low status, but when you outsource it somehow does not count because they are not really your employees.

  5. Noumenon72 says:

    The repeated references to Spanish Literature, Art History, and Gender Studies make it sound as if those were common signalling degrees. According to a comment on Marginal Revolution I read once, most degrees are kind of vocational:

    The overwhelmingly most popular major is business, which constitutes 21.7% of all degrees awarded. When you add education (6%), communications (4.9%), health professions (7.5%), parks, recreation, leisure, and fitness studies (2%), psychology (5.7%) and security and protective services (2.6%), you will see that a majority of degrees are awarded in subjects not recognizable within the traditional conception of academe. If we add up all of the traditional humanities and social science subjects, we see that they constitute about 19% of all degrees awarded.

    I guess the 30% he didn’t mention are the not-useless STEM degrees.

    • Baisius says:

      As a STEM graduate (Chemical Engineering), I use approximately 0% of my education in the field. Most STEM degrees are worthless too.

      • Bugmaster says:

        FWIW, one of the better programmers I’ve worked with was a Ph.D. in Chemistry. He originally learned how to program because, if you want to be a Ph.D. in Chemistry, you inevitably end up needing to process terabytes of data in a reasonable amount of time, and at this point neither a slide rule nor your trusty TI-85 will quite do the job.

      • Jesse says:

        Also a chem eng, but i use very significant chunks of my undergraduate material routinely. I would not want a practicing engineering who did not have a comprehensive background in the material, as many disasters are the result of not recognizing interactions between things that one may naively think were unrelated.

        • Baisius says:

          Was a four year degree the best way to learn that material though? Or would you have been better served by a four year internship?

          • Anonymous says:

            Aerospace engineer here. I think there can be value in internships, but they’re far from an across-the-board solution. The biggest issue is that there is a large culture gap between universities and companies/organizations that offer internships.

            Universities are preparing a number of students for a variety of different jobs, even when they’re nominally the same ‘field’. They know this, so they’re concerned with ensuring that their students have an appreciably wide knowledge base. Most organizations which offer internships aren’t interested in providing a wide educational experience. At most, in the big companies, you’ll experience two or three different groups on a short rotation program.

            Secondly, it’s hard to imagine developing an internship program which takes an individual from high school all the way to practicing engineer. I have quite a bit of teacher in me, so I could maybe see myself doing it, but I imagine it would look quite a bit like a facsimile of a university. I’d pick the list of topics they need to know, give them books, and the first several years would basically be me giving them homework and teaching them out of those books until they became competent enough to do things. I have no idea how or why a company or gov’t organization would allocate funds for this process. I suppose it’s possible to pattern it after something like the SMART grant, where you’d require them to work for you afterwards… but again, we’d pretty much just be recreating the university model in a distributed fashion… and only the biggest employers would really be able to dedicate man hours to do this.

            If universities have a failure mode, it’s that they’re also training their own replacements. Professors want students to become professors, because to a professor, professing is the best thing you could possibly do. This causes the curriculum to be theory-heavy and taught at a high-level. Since there are a lot of more practical jobs that don’t require this, some people think that it’s totally useless. If I’m honest, I don’t care. If the biggest failure mode is that people might learn too many things at too high of a level, I’m not all that sad. Nevertheless, my experience is that universities have started to adjust for this. There are a variety of more hands-on, practical courses that you can take at good universities. They still need to teach high-level theory for those who continue to do research, but they’re getting decently good at identifying those who aren’t cut out for it and pointing them in the direction of more practical experiences that will just result in a job.

            Of course, all of this is in STEM, where we have some (but vanishingly little) non-technical requirements. Here, the complaint is usually, “Why are STEM majors so bad at writing,” rather than, “Why are we wasting so much time making STEM majors too good at things they don’t need?”

    • Besserwisser says:

      Note that he never made any objection to medical degrees. The 4 year part studying something else is the problem, not medschool. Scott doesn’t want to get rid of college degrees as a means to determine employability, he wants to get rid of unneeded college degrees. Which opens up a whole new can of worms.

      • Paul the Apostate says:

        The punch Scott pulled in his explanation is as follows: college degrees are a product there is demand for, and which benefits two groups, employers who don’t question their usefulness and provide the demand, and educational institutions who meet the demand. Between them, what incentive do they have to care about how useful or needed a degree actually is for the applicant? I would say damn little. The applicant is a pawn here. He meets increasing numbers of arbitrary criteria or he’s f***ed. And that is very much the way employers want it today.

        • Paul Brinkley says:

          I’d imagine the loop here could be broken by an employer deciding he can still produce the widgets by hiring people without degrees.

          There’s a nasty catch here; in some cases, the employer’s customers could threaten to sue the employer if they find out the employer hasn’t been hiring “the best”.

  6. zz says:

    I actually wrote a thing a few weeks back that complements this and, I think, provides a better solution than Scott’s (for 2–5%ish of students, at least).

  7. Wesley says:

    I suspect if Ireland did go to a 4-year undergrad degree -> medical school system, it’d be much more similar to Canada, not to the United States. You can get an undergraduate degree in Canada for $0, if you get some scholarships and work hard during the summer. And that’s at one of the top 5 schools in the country. It’s less than $25,000 per year, including living expenses, for pretty much any program in the country, excepting deregulated Commerce/Business degrees. You simply can’t spend the ludicrous amounts of money that you do in the US.

    And if you’re not eligible for scholarships (merit), how exactly will you qualify to get into medical school at the end, given the competition?

    • Nathan says:

      This is basically saying that subsidized tulips are fine because they are affordable for the tulip-buyers.

    • drethelin says:

      Sure, who cares if young people have to throw away THREE YEARS OF THEIR LIVES as long as it doesn’t cost them any money?

      • merzbot says:

        I don’t think many people consider college to have been years of their life thrown away. But you’ll probably find a lot of Americans who didn’t think it was worth the $100k+ they paid for it.

        • Lee says:

          I’m not sure people always fully realize opportunity costs. A lot of great things probably happened during their time in college—but it is hard to say whether these things would not have happened if they’d spent those years working, traveling, or studying on their own.

          For myself, not going to college might have meant extra years of /earning/ income rather than paying. So the loss is greater than just tuition.

    • Deiseach says:

      What Scott did not mention was the points system which regulates entry to third level education in Ireland; probably something along the lines of SAT scores but not quite. It’s based on scores for the grades you achieve in your Leaving Certificate, the State examination taken by secondary students when finishing second-level education.

      Medicine was (and still is) traditionally one of the choices that required high points (last year required a minimum of 480 points but in reality, you need to score in the 700s to be considered).

      They’ve also introduced a parallel test – the HPAT – to help people get in who might not have the high academic score but do have the aptitude for medicine. At least, that was the theory, but as ever: people immediately starting doing grinds and resits to score as high as possible on the HPAT to bump up their chances, and the points race goes on.

      It’s not just a matter of applying for a college place after you do your Leaving and they’ll take you in. Scott’s philosophy degree probably did help enormously, and the swank level didn’t hurt either 🙂

  8. Evan Þ says:

    Off-topic – I see you mentioned your “ex-girlfriend”; I hope you’re both doing fine? Did it have anything to do with either of your blogging?

  9. Leif says:

    This is well-put. A couple of things:

    1) I wonder if the higher education bubble might already be partly the result of subsidies. I haven’t researched this in depth, but this graph really looks like something happened around 1980, which would line up with the Middle Income Student Assistance Act of 1978.

    2) Or maybe the higher education bubble is the result of hiring discrimination laws; if that is the case, your conclusion, to make hiring discrimination laws even stricter, sounds potentially hazardous. Employers aren’t currently allowed to use IQ tests, for example, because those are seen as potentially discriminatory. Any creative testing is considered legally dicey, because the law is vague enough that it’s hard to be sure what someone might try to sue for. So maybe employers feel like their safest option is to look for college degrees.

    • J says:

      Used to be you could get a job at a big company without a degree if you could do well on their aptitude test. But during the civil rights era Griggs v. Duke power killed that because it held that the tests were discriminatory. So employers required college degrees instead, which end up making things much harder for people without the time and money to go to college for 4 years:

      • Peter says:

        The suggestion at the end of it seems like a neater version of Scott’s suggestion making “college degree holders” a protected category – because having a college degree correlates with protected characteristics, discriminating based on college degrees is indirect discrimination and thus requires stringent justification.

        That said, people who like thinking about indirect discrimination should think about Meehl’s observation that at least in “soft” domains everything correlates with everything else.

        • Deiseach says:

          The trouble is though that basically this calls for “cannot discriminate on basis of level of education”, and you might end up with “Okay, so you want us to hire someone who can’t read to run the machine making your cancer medicine? Where pressing the wrong button means wasting hundreds of thousands of pounds worth of production per day? Where they can’t tell if that’s the right ingredient in the barrel being fed into the line? Really?”

          • Edward Scizorhands says:

            IANAL, but my understanding is this:

            1. You can require IQ tests or degrees,

            2. but if you are sued,

            3. and if the candidate can show disparate impact (which means it’s something that minorities do less well on)

            4. then the company needs to show that the requirement was reasonably related to the job

            #4 can be quite difficult and/or expensive.

          • Paul the Apostate says:

            Too bad IQ and its white middle-class cultural assumptions are set in stone. (And they are. I once did a report on readability indices; it turned out the most used are not the most accurate, only the most established and widely understood.)

          • Kevin C. says:

            @Edward Scizorhands

            Relevant to the difficulty of your #4:

            Judge Rules Second Version of New York Teachers’ Exam Is Also Racially Biased.

            In particular from the article:

            Neither version of the exams is still in use in New York. Instead the state administers a new test called the Academic Literacy Skills Test, or the ALST, along with a slate of other assessments. The fate of the ALST, however, was recently called into question as well. This spring, Judge Wood began questioning whether that test, too, was racially discriminatory. A hearing is scheduled on the issue for later this month.

    • Alex Zavoluk says:

      “I wonder if the higher education bubble might already be partly the result of subsidies. I haven’t researched this in depth, but this graph really looks like something happened around 1980, which would line up with the Middle Income Student Assistance Act of 1978.”

      Yep. It’s fairly basic economics–if you subsidize college, the demand curve shifts up. Thus the supply curve must shift (difficult, time-intensive, and many schools like the size they’re at) or prices must rise.

      “Or maybe the higher education bubble is the result of hiring discrimination laws; if that is the case, your conclusion, to make hiring discrimination laws even stricter, sounds potentially hazardous. ”

      This is also the case, and it extends to all forms of credentialism, not just college.

      • Deiseach says:

        The idea behind subsidising college education was (a) those who could afford to go to college were a self-perpetuating elite (see the British cabinet, stuffed with public-school educated Oxbridge graduates) and so (b) opening entrance to those from less well-off backgrounds would help overcome poverty by lifting them out of dead-end jobs. If they went to college and got good jobs, their children were likely to go to college too, and the virtuous cycle of improvement would continue down the generations. Instead of Joe the Ford assembly line worker (particularly as the markets for those kinds of jobs were shrinking), or Joey the unemployed guy, you would have Joseph the white-collar office worker or engineer or working in a managerial capacity or businessman.

        And colleges responded to market demand: more people want places? We’ll open up extra places! And offer all kinds of courses! And rake in those sweet, sweet tuition fees!

        • Alex Zavoluk says:

          Unfortunately, I can’t find a link to it; maybe it’s not published. But recently I was at a panel where James Heckman presented a variation on the classic “Gatsby” curve, where he looked at education instead of income, and he found that intergenerational elasticity of schooling was lower in the US than in Denmark, where college is basically free (that is, whether you go to college is more strongly related to whether your parents went to college in Denmark than it is in the US).

        • Gbdub says:

          Of course the problem was there was never a glut of white collar jobs for Joey in the first place, so either Joey stayed unemployed (now with a fancy sheepskin and associated debt) or Joey turned out to be pretty good, and displaced Elderly Ed, who was forced into retirement 10 years early and couldn’t find another job, so now lives on a minimal investment income and government assistance.

    • Nicholas says:

      Can’t speak for anywhere else, but in my home state the problem is that state licensing boards keep adding more advanced degrees to the licensing requirements (if you want medicaid to pay for the babysitter who watches your kid with a lung deformity to make sure he doesn’t die overnight, that babysitter must be, or must be the employee of, a licensed home healthcare professional [even if all he can do for the problem is call 911 and begin chest compression]) so even though my office manager at a home healthcare job was just an office manager who handled sensitive paperwork but didn’t do any hands on medically stuff, he still had to have a license that required a four year degree.
      So the issue is often less one of hiring choice as it is regulatory capture.

      • Deiseach says:

        Part of that though is that people will and do sue. The babysitter the service provided for your child with a lung problem had no first aid or medical qualification (even though, as you say, all they could realistically do is call the ambulance and start CPR)?

        Unacceptable! If they had proper qualifications, my child might be alive now because recognition of the problem and speedy treatment might somehow have helped! The agency knew my child had medical needs, why did they send out an unqualified sitter? Somebody should do something!

        And so politicians get guilt-tripped (also with an eye on votes) into passing Precious Little Blossom’s Law where everyone has to have licensed professional qualifications even if they’re not technically necessary for that position.

        • Gbdub says:

          Eh, I think it really is regulatory capture – the health care workers with fancy degrees want to close off their guild to keep salaries high, so they lobby hard for more regulation (that’s also how you end up with $1000,0000 taxi medallions and still not enough cabs). Your story is just how they sell it to the uninformed voter “We’re protecting you!” (Collects campaign donation from health worker union).

    • Boris says:

      An alternate hypothesis is that nominal tuition growth was high in the 70s, along with inflation… and just stayed high even when inflation was broken in the early 80s, because people were used to it.

      But yes, it’s easier to explain a few years’ inertia that way than to explain the entire ensuing 35 years.

      • Paul the Apostate says:

        It’s a lot more profitable for any institution to serve other institutions, with their massive money and influence, than it is individuals, who have less money, little or no influence, and a lot of frustratingly diverse needs. Thus the current situation where academia meets industry’s needs first, and the “needs” of individuals that it meets are mostly imposed requirements.

    • Svejk says:

      IQ testing for purposes of employment is not illegal in most of Europe, and in fact certain EU civil service positions require candidates to sit a test much like the SAT + an abstract reasoning component + segments related to the specific position. Private employers may also set exams, which are not required to be knowledge tests, and many do. These tests often include a short abstract/numerical/verbal reasoning test + a short personality inventory. Yet there still exists a college premium in most of these countries.

      • Jiro says:

        It may exist there, but it it as high? Not being able to test IQ is only one of the factors involved. I’d expect that if you get rid of that one factor and keep others it would be less but it would still exist to some extent.

        • Svejk says:

          If we approximate the wage premium by using the 2013 US numbers from the Economic Policy Institute, which gives a ~48% increase over secondary school finishers for men and a 50% increase for women =~ 1.49X US premium, and compare them to Eurostat data reported in “College wage premium over time: trends in Europe in the last 15 years” by E. Crivellaro, which estimates a premium of roughly 1.45-1.64X over all genders and age groups under 50 [these Eurostat data are from 2001-2009; the US premium was between 1.44X-1.47X during this period], the premia appear to be similar. The degree-holding proportion of the European sample was approximately 30%, which is comparable to the US proportion at that time reported by the American Community survey. However, I’m not certain the US and EU populations of tertiary degree holders are completely equivalent, as the rise in tertiary education was much more recent in Europe, which I suspect affects the age composition of the sample. Additionally, the Eurostat data truncates the data at 50 years of age ), and as far as I can tell the US dataset includes all respondents over 25. A finer parsing could tease out relevant details.

          • Jiro says:

            If Europe lacks one factor that causes it, but has other factors, could it be that it has risen at a slower rate in Europe, but still eventually caught up with the US because rising at a slower rate still stops at the same plateau? That could explain why the rise was more recent in Europe.

  10. DanielLC says:

    > Also, anything not related to job-getting can be done three times as quickly by just reading a book.

    That may be true for you, and I know it’s true for me, but I’m not sure it’s true for everyone. The thing to remember here is that you’re smarter than most people. You’re probably smarter than most college graduates. Not everyone can learn the material just be reading a book.

    • sweeneyrod says:

      Yes, and I expect ability to learn things from books varies a lot with the subject and the person. Also, college provides you with access to intelligent people studying the same subject to talk to.

    • Tangent says:

      Also learning anything maths-heavy from a book is pretty hard going for most people, you can get better at it but it takes time. I can’t imagine learning my maths degree from books.

      • RKN says:

        Yes. It’s well known among pre-meds in this country that acing two semesters of Organic Chemistry is a necessary (not sufficient) requirement to be accepted into a good medical program. Along with good scores on the MCAT. If “reading some books” over summer truly was sufficient to pass the academic rigors required to get into (or continue) the medical program in Ireland, that’s impressive — either for what it says about Scott, or the academic standards in Ireland, possibly both.

    • Scott Alexander says:

      1. I am better than the average person at learning some things, but I predict worse than the average person at learning some other things, like chemistry (I got Cs in my high school chemistry classes). Nevertheless, I did okay learning college chemistry out of a book.

      2. Even if only 25% of people are smart enough to learn something out of a book, that’s still a couple billions of dollars we’re wasting by removing that option.

      • Paul the Apostate says:

        It is highly politic to quietly squander a few billion untracked dollars per annum in support of a profitable, time-saving status quo – or even one perceived to be profitable or save time.

  11. Moshe Zadka says:

    Umm….making race and gender protected categories has done little to solve gender and race discrimination, I am pretty sure your solution wouldn’t work either 🙂

    I suggest a simple solution: there are three distinct services colleges offer: teaching, keeping exams “honest”, and grading exams. Force a “chinese wall” between them: separate organizations for teaching, for monitoring, and for grading. Ideally, the “grading” part would only be done by government-sponsored organizations (state universities and colleges). You can study at Harvard, and you can choose which state will grade you (you might need to pay more for CA grading, if CA is known for having tougher graders and you want to show your mettle to prospective employers). Monitoring we can do with private institutions being accredited, because I’m assuming that this is something private industry can provide.

    Of course, universities will lobby against this. This is something both proposed solutions have in common (as opposed to Sanders’ “let’s throw money at universities” bill).

    • You can only make grading separate from teaching if what is being taught is fairly standard or generic. That works fine at high school level… schools general don’t grade certificates tests themselves, and school curricula are generally tightly controlled…but how is a university supposed to get external people to grade an innovative course it has designed itself?

      • AlphaGamma says:

        A version of grading-separate-from-teaching is essentially universal in British universities, in the form of the external examiner. Academics from other universities are brought in both to approve the questions on the final exam, and again after the exam to review student’s scripts and the marking. They will also be involved in the meeting to set grade boundaries. The academics who teach the course will have set and marked the exam.

        This does not mean that the same grade in the same subject is equally respectable regardless of which university it is from, but external oversight of university-level grading is certainly possible.

        Regarding the “innovative course” argument, does any university teach a course where no-one who isn’t connected to the university has enough expertise to oversee exams? After all, a PhD is much more innovative than an undergraduate degree, and people find external examiners for thesis defenses…

        Another note: you don’t have to be entirely unconnected to the university to serve as an external examiner. Both of the two for my degree had studied at that university- one at undergrad, one at PhD.

    • imuli says:

      One gets much more information about somebody’s race and gender just by looking at them. A closer parallel might be religion, you can make some guesses, but unless they’re wearing a ring or symbol or something…

    • Paul the Apostate says:

      A simple solution in concept; monumentally complex in application.

    • The Unloginable says:

      Speaking as someone who spent time in the bond pits, my rule of thumb is that any proposal requiring a “Chinese Wall” should be assumed to fail before it is even launched.

  12. Sniffnoy says:

    You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail.

    Note of course that in the US, at the moment, giving an examination that’s sufficiently like an IQ test is legally risky.

    • grendelkhan says:

      Data point here: I was given what was very obviously an IQ test as part of a job interview a few years ago; they told me specifically that it was not an IQ test. I was also given a programming puzzle, which, I was told, I was the only candidate to actually complete, as in, only one other person’s code compiled, and that didn’t work right. I have no idea what the value of that IQ test was, and I’m pretty sure it wasn’t legal to give.

      • Paul the Apostate says:

        I’d guess the loopholes about what can be called “Not An IQ Test” are pretty few, and limited only to certain lines of work (such as the postal service, which has peculiar requirements mentioned upthread).

        The anti-IQ-test laws are one of the few things that civil rights activists, industry, and educational institutions can get together on, all for their own, distinct and mostly unrelated reasons.

      • Careless says:

        As long as they’re sufficiently obscure about how the test is used, they’re probably not going to get in trouble for it.
        “You scored high enough, we’ll hire you” – bad
        “ahh, I see your results. Very interesting…” – possibly ok

  13. The Do-Operator says:

    I have vague memories of an old SSC discussion about a magical house where only the smartest were allowed to enter. The only effect of the house was to reduce your life expectancy by four years, but employers started insisting on proof that employees had entered the house, as a signal of intelligence

    Does anyone have a link to this discussion? My google-fu is failing me

  14. Alex Zavoluk says:

    Two comments:

    I think a good liberal arts curriculum can teach you (or at least help you to think about) a lot of things that are very good to know, especially for those who are going to be in high-paying and/or influential positions to think about. I’m thinking here of ethics, logic, writing, political history and power structures, as well as a basic understanding of the important facts and principles of the natural and social sciences and mathematics. I think a world in which all or most members of the most influential class (including doctors, but also lawyers, businesspeople, engineers, assorted other professionals, politicians, the heads of major nonprofits, leading academics, and major media and cultural perosonalities) have thought about and understood these things is preferable to one in which they have not. My college has a strong Core program for all undergraduates, and I’m glad we have it.

    Maybe it’s not worth 200 grand per student, but maybe we should lobby colleges to stop wasting money on stupid bureaucrats instead of skipping all that.

    The rise of credentialism is (partially) a result of the existence of the protected classes you want to expand. When employers are prevented from using their own personal judgement, they turn to paper credentials so that they cannot be accused to readily of discrimination. Maybe it’s an improvement if all employers are the KKK, but I don’t think in today’s world employers are so discriminatory that credentialism (which favors the already wealthy and entrenched, as they have the ability to control the credentials being required, and the money to pay for the classes) is really an improvement. Most likely “college degree” will be replaced with some other weeding-out measure, which may or may not be any better now, and will certainly get worse as everyone tries to pile on to the new measures. Goodhart’s law strikes again.

    • zz says:

      Dunning-Kruger is what happens when ignoramuses know so little about something that they fail to comprehend the depth of their ignorance and wind up believing they’re waaaaaay more skilled than they actually are. Witness, for instance, 90% of things written about economics by non-economists.

      The solution is, of course, to get everyone to a level where, maybe they don’t understand economics well, but they understand economics enough to shut up and listen to the economic consensus rather than try to reduce housing prices by issuing a moratorium on construction. I picture a decreasing-then-increasing graph of perceived ability as a function of actual ability: our goal is to get students to the bottom of this graph.

      (nb, gen eds fail to do this. I’m familiar with results where students who have taken an introductory economics course do about as poorly in a 6-month follow up as students who never took an economics course and that students who take introductory mechanics courses fail to improve their scores on the force-concept inventory (even if they took an Ivy League physics course). This is obvious if you understand forgetting curves, and, even though I’m not aware of empirical results for other introductory college courses, I expect much the same lack-of-improvement. Really, they make the problem worse, because not only is Dunning-Kruger not mitigated, since students are equally unskilled, but they now have a credential saying they know about something they don’t know about.)

      Anyway, the clever thing here is the name of whatever trains students to get them to the bottom of that graph: Dunning-Krugtonite.

      • Winter Shaker says:

        I’m glad I’m not the only one who is disappointed at the continuing non-existence of this.

      • Douglas Knight says:

        The failure of physics to improve performance on the force concept inventory has nothing to do with forgetting curves because it fails even during the class.

        • zz says:

          …Huh. I should’ve known that without your comment. Thanks for helping me squash some of my idiocy, though.

          Right now, I suspect if we can figure out a way of improving FCI scores, the learning may wind up durable, since physical situations which can act as reviews crop up in everyday life often enough.

          Similar arguments can be made for other disciplines.

    • LTP says:

      I agree with respect to the general liberal arts education and the types of people who should have that. But, that really only means that 10%-20% of the population should be getting a post-secondary liberal arts education at all, and the other 40%-50% of students attending 4-year universities shouldn’t be there.

      • Alex Zavoluk says:

        I agree that most people shouldn’t go to college, but I also think we can probably have some of that liberal arts I mentioned to high school and replace some of the useless nonsense.

  15. James Picone says:

    In Australia the government subsidises education on the basis of ranking in scores for the last year of school, which have some standardisation. Wiki description. It’s treated as a loan, you pay back some of it eventually.

    This hasn’t produced the same bubble in higher-education costs that there is in the US, it makes higher education affordable for everyone on the basis of merit, there’s still a partial incentive to take something that will pay money, because you need to pay off the loan. Frankly I think it works pretty well.

    I’m not sure I agree with the implication that tertiary education is entirely useless. I think a computer science degree, for example, does produce a better programmer than a 12 week course. Probably not $200,000 to $15,000 better, but that’s not the payoff I’m looking at where I live, so that might be a local concern.

    Twelve weeks is certainly enough to produce someone who can hack together whatever mix of technologies they were taught, and it’s enough to demonstrate that they’re someone who wants to learn and is able to learn, so they’ll probably learn well on the job. But it’s not enough time to pick up all of the more abstract important stuff, like complexity theory, formal grammar, a variety of data structures and algorithms on those data structures, noncomputable functions, Turing machines and lambda calculus, etc..

    Maybe the problem is that ‘programmer’ is too big a role. I recall you arguing that ‘doctor’, as a role, could be split up into ‘someone who has done some basic courses on identifying and treating easy conditions that are most patients, and who can identify complex cases and bump them up the chain’ and ‘someone who has done n years of medical school and is an expert at diagnosis + treatment of complex conditions’ (or splitting up further). I think there’s a similar thing going on here – App Academy is great for someone who can write most web code given a spec, but I don’t think I’d trust a recent graduate of theirs to design software.

    • Buck says:

      I studied and taught at both App Academy and ANU, and now I’m a software engineer at PayPal making more than 90% of American software engineers do.

      But it’s not enough time to pick up all of the more abstract important stuff, like complexity theory, formal grammar, a variety of data structures and algorithms on those data structures, noncomputable functions, Turing machines and lambda calculus, etc..

      All of the things you’ve mentioned there (except for the algorithms and data strutures) are part of a theoretical computer science course. At ANU at least, only about 10% of students do that course, and most of them forget most of it. So it’s not like ANU students would do very well on questions on those topics.

      When I was at App Academy, I occasionally told students to learn a bit of lambda calculus, and they immediately understood it: they pretty much instantly figured out how to define linked lists in the lambda calculus, which I didn’t understand for ages after it was first explained to me. They were able to understand this so quickly because they’d gotten really really good at dealing with manipulating the abstractions which are useful in programming: in this case, it turns out that building complex UIs in Javascript gives you a great intuition for lambda calculus.

      I also don’t buy that those topics are actually very useful most of the time. So if knowing the actual content doesn’t matter, but getting the intuition which would let you learn it does, then a/A is way ahead.

      And when it comes to algorithms and data structures: App Academy teaches students about that, and I promise that a/A students would win hands-down in a quiz on it.

      And when it comes to designing software: are you fucking kidding me? At the end of App Academy, students have been writing code for many hours a day, every day for a few months. This is code where they were given a problem and had to design the program to solve it. Whereas ANU students have perhaps one or two occasions where they actually started a project from scratch. And they’ve certainly spent less time building software than their a/A counterparts. a/A students are much more qualified to design software than university students.

      Part of this is that a/A students are smarter than university students, but a lot of it is that the course is so much better.

      (There are a few things which ANU students know more about, like computer systems. But a/A students can learn that really fast when they actually need to.)

      • Will says:

        In a social setting, I asked some former app academy students about functional programming, some numerical methods stuff, and some complexity stuff and they all were pretty clueless.

      • James Picone says:

        My degree in Software Engineering wasn’t much like that, and I have several friends who did a Computer Science degree at the same university and they also learned both theoretical computer science stuff and actual programming.

        Every one of the examples I mentioned was covered both in my degree and the CS degree. Both courses also did a lot of programming in Java, C, C++, and a simplified version of MIPS assembly.

        The software engineering degree did involve more from-the-start writing software, but there was a lot of programming in both courses. Software Engineering had two year-long group projects that were completely from nothing, CS students did a six-month one. Obviously there was a lot of mickey-mouse ‘write quicksort’ stuff, but there were several other assignments with design aspects – “here’s the spec for a microwave controller, write the code to implement the spec”, “write the AI for a Robocode bot”, “build a neural-network thing for recognising digits” (for that one part of the grade was your class ranking for accuracy). These were all subjects in both SEng and CS.

        Someone below claimed that the most complicated algorithm A/a students learn is sorting (presumably quicksort or mergesort or something). If that’s the case, any decent CS course should be teaching substantially more – graph search, for example, or various tree things.

        Given how many mistakes people make in the real world that they wouldn’t make if they knew some of this stuff, I think it’s more relevant than you. For example, how many times have you run into people trying to parse HTML with regular expressions? Or implementing dumb linear-search-over-a-list algorithms when they could be doing it in constant time? Or not understanding FSM-based approaches to lexing or protocols or event handling? Or not using functional techniques where they’d be perfect? This stuff really does happen. I must admit I don’t often see people trying to compute noncomputables, though. Still worth knowing.

        • Niklas says:

          I would add that in software engineering, there’s a lot of stuff you don’t absolutely need to know to solve a problem, but will help you figure out a better solution quicker – and you won’t even realise this if you haven’t studied the related subjects.

          Sure, if you’re good, you might independently reinvent half of graph theory, but it probably won’t be as efficient, will take you more time and you won’t understand what you just did as well as if you’d had a more formal education, whether through university or a book on it.

          And worse, you won’t even realise. You might have a sneaking suspicion that someone out there has thought about this before, but no idea how to look for more information.

          • Adam says:

            The other really easy trap to fall into if you just teach yourself to write software without learning about algorithm analysis, computer organization, cache hierarchies, network design, etc. first, is you can write something that works, but it won’t scale for shit and someone else has to completely rewrite it down the line.

            Also, just as another point of data, I don’t know what ANU is, but I studied CS at Georgia Tech, and my first class there we wrote a threading library, proxy server with web cache, barrier synchronization framework for scientific computing that had to work under both OpenMP and OpenMPI, and a virtual memory pager with automatic persistence management, all in one class.

            Also worth noting that even if you’re building complex UIs in JavaScript “from scratch,” someone with a BS and MS in both CS and Math did the heavy lifting for you creating JavaScript in the first place.

    • Max says:

      And I do wonder how with all those magical things taught in college a lot of CS/SE people can barely code their way out of paperbag when it comes to actual work

      Don’t get me wrong – all this stuff can be is extremely useful. However nothing precludes you learning it on your own and not every code monkey needs to deal with it on day to day basis

      but I don’t think I’d trust a recent graduate of theirs to design software.

      I don’t think corporation would trust a recent graduate of anything to design their sensitive software.

      And mavericks dont give a f$ck – they quit college and start their own companies

      • James Picone says:

        And I do wonder how with all those magical things taught in college a lot of CS/SE people can barely code their way out of paperbag when it comes to actual work

        I believe they would do better than something like App Academy scaled up so that selection effects weren’t a thing. And from the sounds of it, things are Different in the US compared to my tertiary education, and that might be driving some of this disagreement.

        I don’t think corporation would trust a recent graduate of anything to design their sensitive software.

        I had a role that entailed software design my first year out of university.

  16. I think its a mistake to assume that because sometimes degrees are required needlessly, that they are always required needlessly. Employers need something to do go.

    I think the problem is orthagonal to that – both tertiary education and HR is broken. There not many quick reliable heuristics available on whether a person is good to hire or not, and HR-folks are rarely skilled enough to find a way around that. Unemployment training programs teach people to lie better on their resumes, and almost all jobs and all resumes put forward the same generic rubbish about working in a team or independently, team player etc etc. A tertiary qualification of some kind is the obvious indication that a person might not be a useless idiot, but these days the standards have dropped, learning is secondary to getting good marks, and there’s endless cheating facilitated by ghost writers and the internet. A lot of people emerge from college/uni without a clue. Employers want non-idiots, but there’s not any real way to select for that, so they shout at HR who cling to the only heuristics they have even tighter. Thus, idiotic job requirements.

    I think some of the problems of the non-libertarian approach are overcome by not having idiots make the decisions about tulip subsidies. For a start, a cost-benefit analysis plus a look at alternative policies would knock both the tulip problem on the head right away. Any first-year economist would recognise those problems. Probably a balance between libertarian and non-libertarian allows you to avoid the worst of each approach.

    On the education subsidy plan, I agree, its like handing the keys to the treasury to the tulip traders. But don’t just hang people out to dry instead. Take that money and put it into researching a new education system that actually works, so that it is a good heuristic once again.

    If it were me, I’d be researching:
    (1) Using the internet to reduce cost
    (2) Ways to filter cheaters and people that don’t really learn the material so you can trust that a basic degree is a reliable heuristic
    (3) Basic citizenship training (maybe 1yr) were they were required to learn very basic logic, math, stats, social science, psych, and what the major political groups are, so education doesn’t become just vocational and hollow-out the country.

    I’d also only provide 3-4 years of free tertiary and make it market price after that.

    • ” For a start, a cost-benefit analysis plus a look at alternative policies would knock both the tulip problem on the head right away.”

      A survey of the free experiments that have been performed in other countries would be helpful to, but we live in a world where where politicians proudly operate on Not Invented Here bias,

    • Scott Alexander says:

      I’m not saying they’re always required needlessly. If they’re only required needlessly half of the time, that’s still a $35 billion waste, not counting the damage done to the poor, etc.

      • Sure, that seems correct. But you have to subtract the cost of employers losing that heuristic when hiring from the benefit you get from eliminating the waste. Also, if I’m right about there be a shortage of reliable heuristics, people will probably switch to other semi-broken heuristics, because they’re desperately looking for anything that will let them hire less useless people (which is hugely costly). So you don’t eliminate the whole $35b, just a much smaller amount. So the subsidy debate is a choice between two bad options.

        In the case of tulips, rather than subsidise or don’t subsidise, why not invest the money on reducing the market distortions (punish speculation, develop tulip quality regs, R&D a cheap tulip production system that anyone can use in their backyard)? For education, why not take the money and R&D a better, cheaper tertiary education system, so that there is a easy, fair heuristic employers and employees can rely on.

        I think this dichotomy actually typifies a lot of left-right debate – either throw money at inefficient crappy subsidies/programs, or leave innocent people to rot. For me, that debate crowds out any effort to discuss mechanisms that work.

    • Anthony says:

      not having idiots make the decisions about

      This is where most ideas about having government do it in the U.S. fail.

      Government can be (not necessarily is) competent at things which the general public doesn’t care about except when it breaks – sewage plants, water supply, bridges – but once the voters start paying attention, they always want the wrong thing.

  17. Steve Johnson says:

    On the other hand, if can’t you get married without a tulip, and poor people can’t afford tulips, that’s also a pretty grim situation, and obviously unfair to the poor. But the solution isn’t universal tulip subsidies.

    It is if the people who make your propaganda are the tulip growers and if all people of any consequence are expected to work as interns for four years for tulip growers so they can be enlightened about something or other (propagandized in American communism).

    If I were Sanders, I’d propose a different strategy. Make “college degree” a protected characteristic, like race and religion and sexuality. If you’re not allowed to ask a job candidate whether they’re gay, you’re not allowed to ask them whether they’re a college graduate or not. You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail. I realize this is a blatant violation of my usual semi-libertarian principles, but at this point I don’t care.

    This is even more ludicrous. This would have disparate impact across races and sexes. It would put professors out of work. The money is incidental.

    • Will says:

      Dude, professors are ALREADY mostly out of work. Most higher ed profs are adjuncts, and most adjuncts make basically nothing.

      • Steve Johnson says:

        So the fact that people are willing to endure horrible conditions to get a professor position shows that academia is powerless?

        Better interpretation – it’s a far above market compensating position for people who are fairly useless except at making propaganda so people line up and suffer to get one of the available positions.

        • Anonymous says:

          So are all professors propaganda creators or just some of them? How do math professors make communist propaganda? What about the Republican professors (there’s not many of them, but they exist)?

        • Anonymous says:

          This actually varies a lot by field. In aerospace, for example, a driving factor for the desirability of academia is immigration. I’ve met many foreign-born students who come here to get their PhD. Often times, when they start, they say that they plan on going back afterwards. Then, at the end, they sometimes change their mind – they like it here and want to stay. However, it’s pretty difficult to get a job in industry or a gov’t lab if you’re not a US citizen. Thus: academia. This causes the academic job market to be artificially inflated, which has a whole host of consequences.

        • Professor Frink says:

          Your moving the argument. You said making life harder for professors isn’t something that would be done- but it already has been done, and has been for at least a decade, probably more.

          Then you toss out a new theory- “sure, professor is a shitty job, but these people are SO awful that they can’t do anything else.” But again, you don’t know anything about what you are talking about – hard science jobs (bio, physics,etc) are harder to come by!

          Your model of the world is broken. Time to update.

  18. Not interested says:

    One missing piece at the end here and of Sander’s plan is that in many countries with larger education subsidies it is not required to have a university degree for many many jobs, including teaching, firefighting etc. These jobs do get some training, but is focused on the specific needs of the job, much like the description of the doctors in Ireland.

    • In Australia there is some subsidises for other forms of tertiary, such a technical diplomas and apprenticeships, so that a much wider range of jobs are addressed. They get somewhat bipartisan support based on the argument that they’re required to address skill shortages. The left also defends them on the social good/externality argument.

  19. Orb says:

    Just not in the I would like.

    Typo? Delete me if you want.

  20. Orb says:

    You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail. I realize this is a blatant violation of my usual semi-libertarian principles, but at this point I don’t care.

    This won’t work because people would still volunteer their college grad status on their resume. If you left off college info on your resume, it’d be assumed that you’re not a college graduate. Just because it’s illegal to ask about doesn’t make it illegal to share.

    Ideally you want to engineer things so some kind of new certification economy develops in the absence of college degrees. One that has private companies measuring whatever human capital benefits college degrees are supposed to confer, so hotshot research labs can continue hiring hotshot researchers, law firms can continue hiring qualified lawyers, construction firms can continue hiring people who know civil engineering, etc. etc.

    Another way to get to this point would be to promote a norm that it’s OK to lie about whether or not you have a college degree. Or promote the flourishing of “diploma mills” that give you accreditations for nothing.

    Another way to get to this point is to continue on our current path: once everyone goes to college, it stops being as meaningful a signal and we’ll need some better signal. (Unfortunately college prestige is a meaningful signal because top colleges remain selective… Maybe the end result will be that no-name colleges will eat themselves and we’ll go back to an equilibrium where only a few people go to college, and they’re the ones who get in to the prestigious colleges.)

    A tricky thing here is that from the perspective of any given employer, asking for a college degree makes sense because it *is* a signal of quality. It’s only the counterfactual effects of this behavior on a large scale that cause trouble. The cycle is self-reinforcing: the more employers require college degrees, the more people feel the need to get one, the more employers feel they can require college degrees.

    Another idea is to try to create some interesting way to countersignal that you’re too cool for college. Minerva University is an interesting step in this direction. The best way to do this might be to get a single prestigious person to create a really difficult test and then let people take it to show off their aptitude. If you only have a high school degree, but you scored a 94 out of 100 on the Elon Musk Fuck College Test, there’s a certain kind of Silicon Valley startup that would be very interested in hiring you. Unfortunately this doesn’t let you signal conscientious conformity the way mainstream college does. But you might have a self-reinforcing spiral where the more people start doing it, the more conformist it is to do.

    An interesting way to measure conscientiousness that I read about somewhere is to tell people to do a lot of work and tell them you are only going to randomly look at a subset of it. On a System 1 level, this causes people to become less motivated, and only conscientious people are able to overcome this. Maybe the way this would work in practice would be that the Elon Musk Fuck College Test would make you study the contents of 30 of Elon’s favorite books (please include Thinking Fast and Slow, Elon; it’s the best book ever) but you knew that 10 of those books would be randomly left out of your test. Also no retests (enforced with toe tattoos).

    • Jiro says:

      This won’t work because people would still volunteer their college grad status on their resume. If you left off college info on your resume, it’d be assumed that you’re not a college graduate. Just because it’s illegal to ask about doesn’t make it illegal to share.

      People still put their gender on their resume (at least implicitly, since most names indicate gender), but it’s illegal to discriminate based on gender. And that’s enforceable.

      • Daniel Armak says:

        But it’s very hard to prove that in a large firm with many applicants, a particular candidate’s success wasn’t influenced by their gender among other factors. It would be much easier for the firm if the gender could be hidden from it, but it can’t be – at least not at the interview stage.

    • Deiseach says:

      For a start, if there’s a four year gap on their CV between “finished high school” and “started my first job”, they’ll be asked “what did you do in that time?”

      And they’ll say “Oh, that’s when I got my Degree In Octopus Wrestling”.

      Even if they say “I’m sorry, I can’t answer that because it’s an illegal question”, the employer can infer “Ah, they did a degree in octopus wrestling during that time!”

      So there’s no real way to avoid finding out if someone went to college – or was in prison for those four years, could happen too 🙂

  21. Brett says:

    Because it’s illegal to a first-, second-, and third-degree approximation to give job candidates examinations that might tell you whether they can do the job. So nobody does, and instead they just check your college credentials and work history. Well, OK, it’s not technically illegal to give any examination, only those that produce racially disparate results, but since every (useful) test does that, then yes, practically all tests are illegal.

    • Bugmaster says:

      Wait, it’s illegal ? How ? I don’t want to go to jail for interviewing people ! Help !

      • Nicholas says:

        The joke is that Brett thinks that some race of people have no noteworthy ability, and thus if you never hire incompetents you will only hire people of Brett’s favorite race. He will probably respond to this with a claim along the lines of “It’s not racist when you’re right” or “It’s okay to be racist if you’re right” or “It’s not morally wrong when I do it.” But we all know that’s just how people who are Brett’s race act.

        • Bugmaster says:

          *slow clap* You, sir, are a Formidable Opponent.

          But still, on a less social justice-y note: is there any reason why I can’t seat a job applicant in front of a computer, give him/her/etc. some problems to solve, and then come back in two hours to see how much of his/her/etc. code compiles, runs, and prints out the correct answers ? Is that illegal now, or what ?

          I understand that there may be some issues with IQ tests or whatever, but I don’t care about IQ by itself, I want to know if people can actually program…

          • With the thoughts you'd be thinkin says:

            Griggs v Duke Power is the supreme court case that bans testing not related to a career. Plus look at the antidiscrimantion cases against fire departments and police departments based on disparate impact. Also civil service exams were gotten rid of disparate impact by the Carter administration not deciding to fight the case.

          • A lot of jobs aren’t as straightforwardly based on problem solving as IT.

          • Brett says:

            If you can prove in court that the testing procedures you’re using are narrowly tailored to the job (i.e., you’re testing what the applicants will actually do) and there’s no other way to test candidates that isn’t facially discriminatory, then you’re probably OK as long as you’ve got records.

          • Randy M says:

            “there’s no other way to test candidates that isn’t facially discriminatory”

            Well, it shouldn’t be too hard to prove a negative, so you’re safe.

          • Mary says:

            Big IF there, especially considering whom you’ve got to prove it to.

    • Emily says:

      To teach in public school, you have to pass various examinations. Yes, there is disparate racial impact. To enlist in the military, you need certain scores on the ASVAB. Yes, there is disparate racial impact. To get hired as a programmer, you will probably have to pass some sort of technical test. Probably there is disparate racial impact in those, too, but no one has bothered figuring that out because no one cares.

      No one will come after you, as an employer, for giving tests with content knowledge that applies to the job. This stuff is only an issue in the hiring of cops and firefighters.

      • Bugmaster says:

        On the one hand, I am somewhat comforted. On the other hand, I am now somewhat terrified, knowing that — apparently — aptitude plays very little role in the hiring of cops and firefighters.

        • Emily says:

          It’s not ideal. I’m not sure what would be, though. Like, there’s been some rolling-back of How Much We Care (actually, how much the DoJ cares/can get judges to care) about disparate racial impact in the hiring of police and firefighters, and already we’ve seen department shenanigans involving getting people with department connections hired. (Look up LAFD firefighting hiring – the LA Times did some really good work on this.)

          And I’d actually vote for more caring about disparate racial impact in the hiring of teachers.

          Edit: What I meant by the first paragraph is this: when you give departments a longer leash in terms of hiring practices, it’s not clear that they use that to make a more difficult test. Maybe they just use it to get their buddies hired.

        • Nicholas says:

          Hilariously, at least one Police Department will not hire new officers with an IQ above a certain threshold, based on the claim that intelligent beat cops become bored and corrupt, but do not have the practical experience to make detective.

      • Leif says:

        No one will come after you, as an employer, for giving tests with content knowledge that applies to the job.

        Well, you’re right, that’s basically what the law says: tests must only be for skills the job requires. On paper that might sound kind of harmless, but I don’t think it is.

        Sometimes an employer doesn’t want to test someone for specific skills; it just wants to make sure they have general competence. Maybe the job doesn’t require any specific skills; or maybe the skills correlate heavily with general competence, and developing a test for the skills would be costly.

        The current law makes a college degree the safest way to filter for general competence.

        • Emily says:

          I agree that using the college degree as a filter is the safest way. But I would argue that it’s only marginally less safe for most employers to give an exam or ask for exam results. Which is why many do.

          • Leif says:

            If every question on the exam is directly related to the job, and the employer can prove that every question is directly related to the job (the burden of proof is on the employer!), then yeah, it’s not risky. But creating a different exam for each position in your company, and getting legal advice about whether the exams are legal, is costly; and some of the positions may not really have specific skill requirements, yet you still want someone competent.

            Also, in principle, degree requirements can be ruled discriminatory right now if they aren’t directly related to the job. But it’s common practice for applicants to list their degree on a resume, so employers have that information by default. Even if a degree requirement and an exam are both theoretically discriminatory under the law, the latter is more likely to go to court, because there would be evidence in the form of the employer requesting exam results.

          • Emily says:

            That cites only Griggs and no cases from the past 40 years. Private sector employers are not successfully getting sued for using either college degree or tests in hiring. No one is scared because it’s just not happening.

          • ” But creating a different exam for each position in your company, and getting legal advice about whether the exams are legal, is costly”

            Getting useless degrees is costly…the questions is, for whom.

    • Will says:

      Every single job I’ve applied to required me to take a test of technical ability that strongly resembled an IQ test. You claim “it is illegal” but it is a law that no one enforces.

      • Mary says:

        It’s been enforced. The fun part is that it can be ruinous if you hit the jackpot.

  22. Andrew Swift says:

    I agree with everything you write. I think it’s due to signaling. People need a way to signal that they are willing to work hard, able to concentrate and are intelligent.

    In the US this is done by getting a degree at a “hard” college.

    I have heard from multiple people in the tech industry that while a college degree is not in itself meaningful, it is a great signal for knowing who will actually be able to get meaningful amounts of technical work done.

    How are students in Ireland selected? Are there competitive exams? Does everyone who wants to become a doctor? How does an excellent candidate signal that she is worth hiring?

    I live in France (I’m American), and one of the hallmarks of the French system is that they have gone to great lengths to get rid of signaling. College is essentially free etc.

    It’s hard to argue with their medical outcomes, but there are two néfaste consequences:

    – too many job applicants. When everyone who wants to can get a Ph.D. in chemistry you end up with many more aspiring scientists than there are places for them to work (like gender studies, I suppose). Looking for a job becomes nightmarish and well-qualified candidates have to find another signaling method (like number of publications).

    – a glut of people who studied in fields where they really shouldn’t be. My PhD’d wife works in a scientific field and she’s surrounded by people that would be completely unwilling to do/incapable of doing the work to graduate from a “hard” college. It’s a huge drag just trying to get basic things done.

    So my basic questions to you are:
    – how much of the extra college required serves as signaling
    – is this necessary or useful to sort people
    – if so is there an alternative (and difficult to fake) way that people could signal their qualities?

    Thanks for a great article.

    • zz says:

      Is there an alternative (and difficult to fake) way that people could signal their qualities?

      The Royal Statistical Society offers exams equivalent to a degree for about 800 USD. Having taken them, I can attest that they are not easy to fake (the questions are fairly straightforward if you understand the material, but I can’t imagine a way of training students to guess the teacher’s password.)

      See also: Yudkowsky on education

      • Please note that The Royal Statistical Society has decided to withdraw from offering its own professional examinations after the May 2017 session. This decision has been made to allow the Society to focus its resources on setting and maintaining professional standards in statistics and allied disciplines, in the UK and Internationally.

    • Arcaseus says:

      I live in France (I’m American), and one of the hallmarks of the French system is that they have gone to great lengths to get rid of signaling. College is essentially free etc.

      I’m French and I would partially disagree with this point. There are (mostly) two paths through tertiary education in France:
      – You can go to “Université”, which indeed are quite easy to get in (although not always easy to graduate once you’re in) for 3, 5 or 8 years (approximately equivalent to BS/MS/PhD). This is what most people do.
      – Or you can go to “Classes préparatoires” (a.k.a. “prepas”) , followed by “Grandes Ecoles” (2 years + 3 years, delivering the equivalent of a MS, you can optionnaly do a PhD afterwards). The “prepas” are somewhat hard to get in (about 10% of high school graduates are accepted) and are fairly intense (60 hrs/week is the average, 80hrs/week is common especially in the hardest ones). At the end of the two years, you take competitive examinations, and the rank you achieve decide to which “Grande Ecole” you go to.

      Having a degree from a prestigious “Grande Ecole” is entirely about signalling that you are reasonably smart and were willing to work hard for two years.. and this pretty much guarantees you a job.. it is definitely at least as important a signal in France as a Harvard degree is in the US.

      Compared to the US, the system has some benefits and advantages:
      + No one cares about your extra-curricular activities in high-school
      + No possible cheating (the competitive examinations are closely watched), and no possible way to buy the entrance to a “Grande Ecole” even if you are a billionaire/alumni.
      + Money is not a limiting factor (2 years of prepa, including textbooks, food, tuition, health insurance and a room in the dorms in Paris was about 2k$ a few years go, it would have been free if my family was on welfare).
      – An astronomically high rate of burnout/depression/suicide in the “prepas”
      – Difficulties for the 90% not going through prepas to signal smarts/conscientousness as you said.
      +/- (depending on your opinion) No affirmative action, leading to very homogeneous populations for the “Grandes Ecoles”
      I am not sure whether it is better or worse than the american system overall.

      • Ever An Anon says:

        Well the suicide rates sound about right for an American engineering school or top tier university so while it’s a downside in the absolute sense it isn’t as much of one in the comparison. And I suspect the Big Schools won’t be homogeneous much longer given the way French politics have been going lately.

        But anyway, if we’re going to copy anyone’s system it ought to be the Germans. It’s faster, cheaper, highly focused on vocational skills (i.e. actual education), and the testing process gives you your credential information right upfront. Besides stealing from the Germans is a proud tradition in the history of education.

  23. Bugmaster says:

    Also, anything not related to job-getting can be done three times as quickly by just reading a book.

    I often hear very smart people say this; but, for people of average intelligence such as myself, this is often not the case. Reading books on quantum physics, or even something relatively simple like machine learning, can be very difficult when you’ve got no one to help you through all the tricky points; no structured regimen of tasks designed to increase your understanding at a manageable pace; and no community of like-minded people whose shared experience you can draw upon. College provides all of these things (or, at least, some colleges do).

  24. suntzuanime says:

    Ireland is one of the approximately 100% of First World countries that gets better health outcomes than the United States.

    I thought I had read somewhere that once you controlled for how fat, lazy, and stupid we are, the US health outcomes picture looked a lot better, but I’m searching and I can’t find anything on either side except something saying that our life expectancy is artificially lowered by our high rate of car accidents.

    Do you have a source for this claim that controls for the various factors it’s not fair to blame doctors for?

    • speedwell says:

      Once you control for my diabetes and high blood pressure, and control for the approximately ninety percent of the population who test lower than me on IQ tests, I am approximately as healthy and intelligent as the average person.

      • suntzuanime says:

        You’re missing the point.

      • Mary says:

        So what?

        If we were testing school systems, we should certainly control for intelligence, because “turning out well educated students” qualities of the school system can only be determined by how it improves on the raw material.

        By the same token, you can only judge the health care system if you control for the raw material it has to work with.

    • Deiseach says:

      Hey, I’m fat, lazy and stupid, and I resent the implication that any American can outdo me on these! 🙂

    • hawkice says:

      What you really want to do is subtract gun-related and vehicle-related deaths and injuries. We have a lot more of both in America, and if you look at health care outcomes without those (which is reasonable, unless a doctor is shooting you so he can get more work), then America actually has the best outcomes. The issue is, the edge America has is extremely small, and is priced at a fairly outrageous premium.

      • malpollyon says:

        And how do you explain America’s infant mortality statistics? More special pleading?

        • Anonymous says:

          The line between infant mortality and stillbirth is arbitrary.

          • Richard Gadsden says:

            Very premature or very disabled births will be reported as stillbirths in other countries?

            In which case, the ratio you want to measure is:

            Babies that reach their first birthday
            Pregnancies (that reach some gestational age at which national differences in pregnancy identification are minimal) – abortions (excluding abortions done where the foetus or was going to be stillborn or die shortly after birth)

            But I doubt that enough countries have those figures.

  25. zz says:

    Also, I’m surprised there’s no mention of Bryan Caplan’s arguments against education, which address the signalling-vs-adding-value aspect much more fully.

    • Bugmaster says:

      I watched a little of the video, and I think that he commits the same mistake most people do: confusing education (especially college education) with vocational training.

      It is absolutely true that most topics you will study will not directly translate into marketable job skills. Most people will never need to apply Calculus, or Trigonometry, or English Literature, or, indeed History, to any task they will directly need to perform on the job. Vocational training, on the other hand, is immediately applicable. If you take a course in automotive repair, you can immediately start working as a junior car mechanic. If you take that App Academy course, you can immediately start cranking out Web applications. If all you care about is finding a job — any job ! — quickly, then vocational training is the way to go.

      The problem, though, is that in order to become really good at your job, you will need to achieve most of the following:
      * Have a job that you actually enjoy doing, at least to some extent
      * Understand what you are actually doing, and have the ability to synthesize diverse knowledge into a solution to some novel problem.
      * Know how to learn, and recognize where more learning is needed.
      This is where all those useless History, Computer Science, and Algebra classes come in pretty handy. They are not your tools; they are building blocks from which you can construct any tool you need.

      The problem with being the kind of person who only knows which button to push, which bolt to unscrew, or which checkbox to tick on a questionnaire is that — in our modern world — you are no longer competing with other people just like yourself. You are competing with robots. You cannot out-compete robots on their home turf. Until the Singularity hits, you need to outsmart them, and you can’t do that without an education.

      • Randy M says:

        “This is where all those useless History, Computer Science, and Algebra classes come in pretty handy.”

        Do they in practice though? How specifically? And is the actual results on the median student worth the benefits? Are the benefits rising commensurate with the costs?

        • Bugmaster says:

          In my experience, the answer is yes. Most of the programmers I have ever worked with fall into two distinct group. Group A is very good at producing well-defined applications (usually, Web applications) very quickly. They are fully trained on which button to click, and what code snippet to copy/paste, in order to achieve one of the pre-determined desired results. Group B, on the other hand, understands what the code actually does, and is able to produce code to handle situations that were not fully enumerated beforehand.

          If you want a standard webapp to be produced very quickly, Group A is the way to go. If you want anything else, then Group A is completely useless; worse than useless, in fact, because you have to somehow drag them along with you, and they are incapable of learning.

      • Tom West says:

        I’m pretty certain that Bryan has addressed the “learning to learn” and “transferable learning” aspect of a college education as irrelevant as far as job outcomes go.

        They may make you feel happier or more fulfilled as a person, but they don’t boost your productivity over what it would be if you only got the skills training needed to do the job.

        Personally, I can’t figure out how you’d tease apart “college teaches these useful soft skills” and “people who naturally have these soft skills usually go to college”, so I can’t argue either way.

      • zz says:

        I play ultimate [frisbee] with a guy who programs professionally and, near as I can tell, is good at it. He’s written a program to form fair teams for hat tournaments, which apparently takes a really long time to do by hand. (You can, of course, throw together teams quickly if you’re less worried about the teams having skill parity.) It takes a list of players and their self-rankings. As we’re driving to the event and he’s describing the algorithm to me, I ask if it’s commutative.

        “I don’t know what that means.”

        “It means that if you change the order the list of players is in, it changes the teams that it generates.”

        “Oh, my program doesn’t do that.”

        His program does that. (I know because he spent 5 minutes at that tournament putting players in a different order so he could play on the team with the most players, since he wouldn’t always be able to sub in because he was organizing and had responsibilities.)

        I agree that, even if you forget a bunch of abstract knowledge and technical information, putting generally useful patterns into your brain is helpful. I (and Caplan)) agree that college does this, but (a) not very efficiently (Caplan estimates 20%) and (b) not necessary, since a lot of these generally useful patterns can/are picked up on the job. See also Scott’s thoughts on unschooling.

        • Bugmaster says:

          As I said above, my personal experience (both as a student and as an employer, albeit by proxy) is quite the opposite. People who went to college and obtained a CS degree have a higher probability of knowing basic concepts, and being able to use them to solve problems.

          For example, many people know how to write a basic loop; however, very few people can answer the question, “In big-O notation, what is the time complexity of your solution [featuring said loop]”, or the easier question, “Let’s say that your program takes 1 second to go through 100 items; if we give it 200 items instead, how long will it take to run ?”.

          You might argue that modern computers are fast enough, and thus the answer doesn’t matter, and most of the time you would actually be right. But not all of the time; and a programmer who actually understands why someone might want to ask that question can be a valuable programmer indeed.

  26. dlr says:

    Actually the situation is even worse than you describe. How much of those 4 years of medical school were spent learning the skills you actually need to practice as a psychiatrist? 1/5th? 1/10th? 0?

    Medicine should parse people out into their specialties on day one. Dermatologists should be trained to be Dermatologists. Psychiatrists should be trained to be Psychiatrists, etc, etc, etc. I bet it would take a year tops for any of them to have learned what they need to know to be ready to begin their residencies.

    • Nicholas says:

      According to the American Bureau of Labor Statistics, more than 1/4 of all doctors are general practitioners without any particular specialty. If we assume that the largest single group of graduates is also the source of the majority of dropouts, then it is likely a waste of time starting with specialties when so few of your students have them.

    • Scott Alexander says:

      There are some complications here (a lot of medical issues can cause psychiatric issues, and vice versa, and medical medications and psych medications interact with each other and cause side effects to each other) but overall you’re not wrong.

      My medical school spent the first two years doing what it called “self-directed learning”, which is where they make you spend a few hours a day doing bullshytte classes on Communication Skills and Ethics and The True Meaning Of Medicine, and tell you that you can learn all of the “rote memorization stuff” (read: medicine) on your own time. If I’d known that was their philosophy I’d never have applied, but they changed to it the first year I got in. The last two years, with actual clinical rotations, were pretty good, though.

      I stick to my claim that given a week and a group of very intelligent people, I could train them in psychiatry well enough to place them in the 25th percentile of psychiatrists in terms of patient outcomes.

      • Anon256 says:

        Is there some way you could actually do that sometime? It sounds more fun and useful to me than CFAR.

      • kieran M says:

        I think this is a view from the top mistake. That is, if you’ve been studying mathematics for a decent amount of time, simple algebraic rearrangement seems like the easiest thing in the world to you. In fact, by your fourth year of study you may find that most of your first year is embarrassingly easy.

      • Deiseach says:

        the first two years doing what it called “self-directed learning”, which is where they make you spend a few hours a day doing bullshytte classes on Communication Skills and Ethics and The True Meaning Of Medicine

        Having seen letters (both hand-written and self-word processed*) from doctors and consultants in my work, I think at least a portion of those two years on Handwriting Practice and Spelling, probably in those Communication Skills classes, would be no harm at all 🙂

        Doctors who cannot spell the names of the conditions they are assuring us their patients suffer from, doctors whose handwriting is illegible – there was one case where I had to show the letter around to the entire office to ask “What the hell is that saying?” and nobody could make it out. After a good twenty minutes of puzzling, I assumed that if the second word was “carcinoma” (and not something beginning with “z” as it looked)… then Google came to my aid.

        Dear, dear Google – it has often ensured that when we’re putting down the medical needs of clients for social housing, we’re putting down what they actually suffer from and not what the best guess from the doctor’s letter says.

        *I know they’re self-typed because nobody on a secretarial course who made those kinds of mistakes in basic English grammar and spelling of common words would be permitted to pass the exam and get the qualification for the job of medical secretary.

      • Gram Stone says:

        Second Anon256.

    • Douglas Knight says:

      Atul Gawande has a nice essay about this, but I cannot find the text freely available (in this book). Here is a summary and gated link. The best hernia surgeons in the world are made by a single year of surgical training, compared to normal surgeons who receive a decade (post-MD).

  27. slutmonkey says:

    I agree but my question has always been what’s been causing tulip inflation?
    Why do Japanese ESL teacher companies (or anyone else) demand degrees? I could be convinced that it continues because high requirements are sticky, but what made anyone think it was a good req to add in the first place?

    • Nicholas says:

      In many places the process goes: A government institutes a law that all companies of [industry] must have work forces that either consist entirely of, consist [some %] of, or be supervised by the holders of some government sponsored license. In a bid to demonstrate that the license works to keep out undesirables (often after an abuse scandal) the license is made harder to get.

    • zz says:

      Relevant smbc. In particular,

      The easier college gets, the dumber you look for not having a degree

      So, to answer your question, Japanese ESL looks at people without degrees very suspiciously. Degrees are easy to get! Why don’t you have one? Are you lazy, stupid, or anti-authority? Because we don’t want any of those.

      (I’m given to understand that Bay Area folk have realized this, and not having a degree is currently a working countersignal there.)

    • Deiseach says:

      To be blunt, my impression was the Japanese are racist in their attitudes, and these kinds of requirements for high qualifcations are a filtering mechanism to keep out undesirables – the same attitude as the “Daily Mail” diatribes about foreigners swarming into England and driving down property prices, leeching off social welfare, and sponging off the taxpayers.

      • onyomi says:

        Yes, both Chinese and Japanese are quite racist, though they tend to like white people with college degrees. You can even get a job in China just being a white guy in a suit who sits in business meetings and doesn’t do anything. I had a friend who was Polish but got jobs teaching English in Taiwan with a backstory about being from Minnesota. By contrast, Asian american friends whose first language was English had trouble finding English teaching jobs in Taiwan. The “hip hop” fashion stores in Tokyo are also staffed by West Africans, so it’s all about looking the part, whoever you are.

        Asian people=normal people
        Asian Americans=retarded normal people
        White people=”Foreigners,” prone to be loud and obnoxious, but also glamorous because looks like people in movies and pop music. Speak English
        Black people=gangsters, rappers; speak English, possibly dangerous
        All others=somebody’s housekeeper; may need to use hand gestures

        They do have a not entirely unjustified notion that foreigners don’t understand how to behave in appropriately face-giving ways. I saw a funny blog post wherein a Chinese commenter said of Uighurs in Beijing: “why can’t these foreigners go back to their own country?!”

  28. Emily says:

    We have a subreddit now! Come hang out!

    • Noumenon72 says:

      This is awesome because Scott’s comment threads are so long that only upvotes can make them worth reading.

    • Aw, congratulations, it’s an upboat baby. 😀

      *…seriously considers unhostblocking reddit (for me, the ultimate murderer of free time) for a moment*

      • Emily says:

        Thanks! I can promise it won’t murder much of your free time because at the moment there are only four posts and 14 comments.

        • I’m not concerned about the SSC subreddit, it’s all the other subreddits I don’t trust myself to stay away from (especially the ancap one). There’s a longer story here, but it’s not very exciting, so I’ll spare you the specifics. Let’s just use the useful shorthand ‘I had a (silly) addiction problem’ and leave it at that, even if that’s grossly oversimplifying.

          reddit is pretty awesome, though, so I definitely hope the SSC subreddit picks up, regardless whether I decide to trust myself to go back to reddit or not. 🙂

    • Mark says:

      Is there anyway to import the comments made on the main site?

      • Emily says:

        I very much doubt it. Does anyone else know?

        • ThrustVectoring says:

          Reddit has good support for APIs and general programmatic interaction with their platform. I’m not familiar enough with the blog side of things, but it’s likely doable. Maybe a weekend project for a programmer with more job dissatisfaction than me.

    • What does Scott think of this? I imagine he would be a little bit concerned about dilluting/splitting the main discussion of recent posts, especially to an area where he can’t moderate?

      I do like the idea of creating longer term discussions about specific SSC issues though. For example I’d like to see further discussion to develop Archipelago related ideas.

      • Emily says:

        I e-mailed him about it and I’ve posted a few times about it and he hasn’t objected. I’m not sure beyond that.

        • houseboatonstyx says:

          You could clone and re-brand it, or keep the SSC but say it stands for something else.

  29. EricSlusser says:

    Oh, please be careful about using the word “bubble” to describe the education situation. “Bubble” is a metaphor twice over that is iffy in both layers. As you use the word “bubble,” you’re comparing episodes of asset prices to physical bubbles like soap bubbles or bubblegum bubbles, and you’re comparing the financial return to education to these episodes of asset prices. The bubble metaphor is that asset prices will suddenly and inevitably revert to fundamental values like physical bubbles suddenly and inevitably pop. But it’s questionable in different examples of asset bubbles whether their sudden price declines were inevitable or whether they had really diverged from ‘fundamental’ values. For higher education, signalling models of education don’t (necessarily) have bubbles, they have equilibria, and students are being individually rational in acquiring their degrees . Both your descriptions of the imaginary kingdoms and of the current situation of higher education seem to describe equilibria rather than bubbles.

    One could argue that “equilibrium” is also an iffy metaphor, but I think it has much less baggage and has a more pinned down meaning in an economic context.

    So, I make zero substantive criticisms of your analysis. “Bubble” is just a peeve of mine. I think “bubble” implies some wrong ideas and “equilibrium” would do as a better word.

    Bubble, bubble, bubble, bubble, bubble, bubble, bubble, bubble, bubble, bubble, bubble, bubble, bubble time to go to bed.

  30. soru says:

    One thing that works perfectly well in the tulip case is the government stops giving people money to buy tulips retail. Instead, it just buys a tranche of tulip-growing land, pays some tulip fans to farm it, and gives the crop away for free.

    That perhaps produces mediocre tulips, or at least less well-marketed ones. And you need some rationing system, so no-one gets more than say 6 tulips a year. But it does the job at a cost to society that probably doesn’t even end up as a line item in the national budget.

    The analogy with housing or health is straightforward; the smallest government is the one that does things, not pays for things to be done. Builds houses, not subsidies rents and mortgages, provides universal free health care, not insurance schemes, trains soldiers not buys F-35s. Every international comparison between western democratic governments that do things differently for different sectors shows that effect clearly.

    You’d think that would be obvious; capitalism is all about growth, finding new markets, expanding existing ones. Government service is all about making do on a budget. If you shift some area of the economy from one to the other, the results should be entirely predictable.

    Slightly weird that the opposite prediction is so often made, even by people without a financial incentive to be wrong…

    • Daniel Armak says:

      > trains soldiers not buys F-35s

      I don’t understand that one. Shouldn’t it be either “trains soldiers, not pays mercs” or “builds F-35s, not buys them”?

      • soru says:

        Perhaps a bit of a cheat, you can’t do a more direct comparison given I don’t think any western government has ever done military jet design in-house, or used mercenaries as a primary form of defense.

  31. maxikov says:

    I’m so glad that you have international experience, and are able to discuss the problems with higher education without defaulting to OMG MURICA SO STOOOPID. This problem is pretty much universal to the entire First World, but every country talks about it like they’re alone with it. In the Second World things are quite different though – they’re worse. Most of the ex-USSR countries retain conscription for all men, and most waive it for full-time students and either college or PhD graduates. Thus, while girls graduating from the high school are under pressure to get some degree to get a job, boys are under pressure to get into a university ASAP after school, and stay there at all costs, under the threat of learning about dedovshchina instead.

    So yeah, it looks like any somewhat industrialized country is experiencing the similar problem. Although it’s a bit easier for the ex-USSR, since there you can get away with paying professors $200/mo, and having tables falling apart in classrooms, so it doesn’t cost governments nearly as much to maintain it free.

    App Academy graduates compete for the same jobs as people who have taken computer science in college, a four year long, $200,000 undertaking.

    I wouldn’t be so sure that it’s true. Scientific achievements like research projects and publications seem to be valued when applying to jobs involving data science, artificial intelligence, etc., rather than pure software engineering. That said, CS is one of the areas best suited for MOOCs, so we may see some changes here soon.

    But, well, when we require doctors to get a college degree before they can go to medical school, we’re throwing out a mere $5 billion, barely enough to house all the homeless people in the country. But Senator Sanders admits that his plan would cost $70 billion per year. That’s about the size of the entire economy of Hawaii. It’s enough to give $2000 every year to every American in poverty.

    How much would it cost to send everyone who wants to countries that charge considerably less for higher education? Alternatively, while we’re on topic of libertarian nightmares, why not introduce price regulation? Applicants already have a mechanism for competing against each other for the limited number of positions – test scores, essays, and other aspects of the application quality. It’s one the rare cases where the price competition appears to be not that much relevant.

    • Daniel Armak says:

      FWIW, waiving conscription for higher-ed students and graduates was also the policy in Russia during the 70s and 80s, when my parents were studying. So it’s not a new post-Soviet invention.

      > CS is one of the areas best suited for MOOCs

      Is this different from math in general? Or do you mean specifically programming and the necessary CS basics for that?

      • maxikov says:

        More programming – automatically testing code snippets, for example, is one of the easiest things ever. Combined with static code analysis, it can provide feedback on programming assignments that’s almost as good as a human instructor could do. Plus, the ability to generate machine-readable output that won’t be accepted unless it adheres to the standard exactly is an entirely realistic skill for a programmer, and incorporating it into the curriculum doesn’t compromise the quality of teaching.

        • Daniel Armak says:

          Testing if a code snippet or assignment compiles and produces the right output is easy. But helping the student understand why it fails the tests when it does is very hard.

          Also, automated testing of code behavior without a person ever reading and commenting on the code means students will learn the most atrocious coding practices they can get away with.

          • maxikov says:

            Sure, but most assignments aren’t meant to be tricky, so the majority of students will get them right. And automatically marking correct answers removes a huge burden from the human graders.

            I don’t have a lot of experience with static code analysis (aside from what Eclipse does out of the box), but I was under impression that it’s not that bad in detecting poor style and errors that don’t always lead to failing tests:

          • Saint_Fiasco says:

            In addition to the automated testing, most MOOCs make you evaluate the code of your classmates. So there is an incentive to be as clear as possible. After all, your code will be read by someone who is not an expert!

  32. Jiro says:

    Diamond rings actually work somewhat like tulips as described here. They are often given in the process of marriage, and their cost is highly inflated as a result. And they serve as signalling.

    Yet somehow I don’t think the proposed education solutions would work for diamonds. And part of the reason they won’t work is that diamonds *are* signalling, and you can’t just get rid of them unless you can get rid of the need for signalling.

  33. Ilya Shpitser says:

    Wouldn’t getting doctors straight out of high school result in extremely innumerate doctors?

    “But is that different from now?”

    Actually, I have a question for you Scott, since you have gone through doctor training: how much formal statistics do doctors get in medical school? Do they get taught e.g. correlation/causation, Bayes theorem, base rates, false positives, etc.

    • speedwell says:

      That getting doctors out of college instead doesn’t necessarily result in particularly numerate doctors, is, I think, Scott’s point.

      • Ilya Shpitser says:

        Sure, hence my self-reply.

        I think there are unrelated issues here:

        (a) There is a higher edu bubble, and

        (b) Doctors need to know some stat and probability to not commit scary errors.

        We need to solve (a) and (b) jointly. If we do things the Irish way, we bypass (a), but not (b). If we do things the American way, we don’t solve either (a) or (b). But there exist separate moves to solve (a) and (b) independently of each other. Solving (a) entails higher edu reform, solving (b) entails forcing doctors to take more math, as part of their prep.

    • Elissa says:

      I’m not Scott, but I’m in med school in the US. The statistics content we cover is mainly specified in the Biostatistics section here: a smattering of sensitivity and specificity, odds ratios and relative risk, interpeting p-values, etc. Sounds ok, but you have to keep in mind that this is presented in one lecture per month or so, and each week includes about 20 other lectures full of information we’ll be tested on, so the average physician doesn’t necessarily retain much of it.

      (Also, I totally went to Podunk Community College, and it doesn’t seem to have hurt me any.)

    • Glen Raphael says:

      A friend of mine became a doctor (in the US) after getting an undergraduate math degree. He said that in med school being a guy in the room who could actually do math was “like having a superpower”.

      • Devilbunny says:

        Out of my (American, public, not prominent) medical school class of 100, perhaps two of us understood statistics in any meaningful sense. When I was a resident, I produced a one-page outline that covered every topic that would appear on the anesthesiology boards. It was simple pattern matching, and it worked so well that I was able to raise the entire program’s stats scores.

        Your anecdote does not surprise me.

    • The Do-Operator says:

      I also attended medical school in Ireland (not at the same school at Scott). If I remember correctly, I think we had at most a single lecture on sensitivity and specificity. In my case, the minimal quantitative training was probably a good thing, it meant I didn’t have to “unlearn” classical epidemiology when I started my PhD

      I have since been involved in teaching several courses on quantitative methods for clinicians at Harvard. These are summer courses in basic statistics for clinical research, were physicians attend three summers in a row and get a master’s degree. Even among this group of students, which is selected both by Harvard’s admissions office and self-selected for motivation, the students simply don’t get it. You can spend hours trying to teach them Bayes theorem, and at best they will be able to memorize the teacher’s password.

      Of course, it doesn’t help that we teach courses that make every mistake in the book. This includes pseudo-Bayesian interpretations of frequentist confidence intervals, incoherent definitions of confounding based on change-in-estimate criteria etc. I’ve even seen instructors teach students to interpret p-values as posterior probabilities.

    • Scott Alexander says:

      You know, all doctors have passed Calculus and Calculus II before entering medical school. And yet when a doctor trying to figure out a way to calculate insulin levels accidentally re-invented calculus and published her “discovery” in a medical journal under her own name, everybody was totally okay with it and in fact it got cited a hundred fifty times.

      This should give you an idea how I feel about the utility of all the undergraduate math classes doctors are required to take, and how well the benefits do or don’t stick around.

      • Alexander Stanislaw says:

        She wasn’t a doctor.

      • Ilya Shpitser says:

        I am not saying practicing doctors need to be intimately familiar with like .. Riemann integrals, etc.

        But it seems to me doctors ought to know certain types of math (mostly from the stats end of the pool) that seem to me to be too advanced for high school.

        I certainly agree we need to reform a lot of things, but in my ideal world, we don’t train doctors out of high school unless medical school in that ideal world has a ton of quantitative classes of certain types.

        • ad says:

          If doctors need to understand something, you should teach it to them in medical school, not require them to complete an undergraduate degree which may or may not include that thing.

          • Ilya Shpitser says:

            Maybe you are right. Are you sure you calculated the best minimal path from our world to utopia, though?

  34. Tim Martin says:

    “Before taking my current job, I taught English in Japan.”

    Oh really! Where did you teach? I did the JET Programme for 2 years after college in good ole Fukui Prefecture.

    • Scott Alexander says:

      NOVA in Nagoya for a year, then an ALT in elementary schools via a private company in Ichinomiya in Aichi.

  35. Kaj Sotala says:

    An alternative approach – nationalize universities and then remove tuition.

    That’s basically the situation in Finland, where universities are theoretically autonomous entities but in practice receive a large part of their funding from the state. A quick Googling found a report from some years back by our national statistical institution mentioning that study expenses in Finland are about 6000 euros per student per year, which is roughly the OECD average. In the United States they’re the highest in the OECD region, 16,000 euros per student per year, almost two times as much as the runner-up of Switzerland that spends 9000 euros per student per year. ( cites OECD Education at a Glance 2009 as their source.) So it sounds like there should be a lot of room to cut down on unnecessary expenses, which you could do by nationalizing the universities and then just cutting their budget back to, say, the OECD average.

    Easy, no? 😉

    (Yes, this may mean that the government needs to pay more than it used to. But previously ordinary families would have been paying those expenses anyway, so if you increase the taxation somewhat to compensate, you’re just shuffling the source of money around – while reducing the net expense.)

    • anon says:

      You still have your entire society wasting 3-5 years per person. And in the case of doctors, limiting the amount of graduates instead of getting as many as possible, because you’re forcing them to spend an arbitrarily long time in university instead of whatever smaller period also seems to work.

      • Kaj Sotala says:

        That’s assuming that higher education genuinely produces zero value.

        It’s been a while since I looked at the economics literature on that claim, but at least back then, that didn’t seem to be the consensus. Rather there was a combination of signaling effects and genuine value.

        • Kaj Sotala says:

          Also, I would argue that having the opportunity to spend four years studying whatever it is that happens to interest you, as well as getting that time to think about what you’d like to do with your life, is hardly “wasted”.

          • Mary says:

            Depends on whether you exercise it, and what happens to interest you, and what you think about it.

      • myself says:

        Yes, you still have a lot of students wasting years of their lives, but AT LEAST society doesn’t spend 200.000 dollars per student. Since it is state owned there is much less interest in increasing the expenditures of the university.

        I live in another continental European country and it’s the same. I believe it’s the same all over continental Europe. Universities have little money compared to their Anglo counterparts and it’s a good thing.

        • Ptoliporthos says:

          Less incentive to increase expenditures, you say? Truly your government must be unique. Are your public sector workers not unionized? Do they not vote for and donate to the Labour party?

          • Kaj Sotala says:

            Related to your question, the recently-elected new government of Finland has announced its intention to carry out budgetary cuts to higher education.

          • At my university (in a different but probably similar country) a large minority of employees belong to the union. But the people in charge don’t.

    • Anon says:

      “An alternative approach – nationalize universities and then remove tuition.”

      In Slovenia this results in everyone wasting 4 years of their life. But not 200 000 USD on top of that.

      Alternative proposal: Outlaw universities.

  36. lmm says:

    But the result of your experiment is the other way around. Isn’t Ireland the country with subsidised degrees and the US the country without? And Ireland does better?

    Maybe in America a degree is more of a maker of social status, so employers hire for degrees as a way to get that “touch of class”. Subsidised degrees eliminate that.

    • HeelBearCub says:

      This story is like the Wizard of Oz, but every character is Scarecrow.

      No mention, at all, of the fact that the health care industry in the US is far and away less regulated and more private market than all of the other industrialized nations, with some similar (but still less restrictive) broad based regulation just now coming in with the ACA.

      And no mention that costs are rising as subsidies are falling.

      And of course, no discussion of why college education is different than primary or secondary education. Perhaps we should eliminate those as public goods as well and make all schools private schools?

      There is a certainly a point to be made in there somewhere. But if he thinks the people coming out of App Academy are as well prepared for the job market as those coming out of traditional CS degrees … lets see how they do in a slack market. App Academy and others like them work well now because the market is tight, tight, tight, with unemployment below natural levels.

      I wonder how Scott feels about the Swiss system, where even the College level is public.

      • Psmith says:

        > Perhaps we should eliminate those as public goods as well and make all schools private schools?

        Yes, absolutely.

        • BBA says:

          How would the poor send their children to school then? Vouchers from the dreaded government? Or do you propose that the poor children just sit idle, or that we bring back child labor?

          • Taradino C. says:

            Sending children to school may not be all that necessary or beneficial:

          • BBA says:

            Many (most?) kids live in households where homeschooling, “unschooling” or otherwise, is impossible because both parents work full-time and don’t earn enough to afford a tutor.

            Compulsory education took root at around the same time child labor was banned, not coincidentally. The alternative to school isn’t “unschool”, it’s the factory, or street gangs.

          • Lupis42 says:


            While I broadly agree, a quibble:

            Compulsory education took root at around the same time child labor was banned, not coincidentally.
            1) Since that was still firmly the era of the stay-at-home parent, and child labor was in major decline, I’m not sure that the causal link there is all that strong.

            I think the reverse causal link is much stronger – i.e. most families are structured around two incomes now, and taking away the free daycare would be a huge hit, so the idea is effectively DOA. Of course, that need could probably be satisfied at around 20% of what we currently spend…

          • You might want to take a look at _Education and the Industrial Revolution_ by E.G. West if you can locate a copy. Working class children in early 19th century English cities, at a point when schooling was entirely private, got about as much schooling as similar children in Prussia at the time, where schooling was governmental and compulsory. It was very low cost schooling—but then, the parents were a lot poorer than the modern poor.

      • Lupis42 says:

        No mention, at all, of the fact that the health care industry in the US is far and away less regulated and more private market than all of the other industrialized nations, with some similar (but still less restrictive) broad based regulation just now coming in with the ACA.

        I’m struggling to find a way to define “less regulated” that makes this plausible. Can you unpack that?

        And no mention that costs are rising as subsidies are falling.

        Are we still on health care, or back to education? Because the main driver of rising cost to education is rising number of administrators and facilities for a given number of educators. Which suggests that the primary source of rising costs is that overhead has risen to fill the budget – the expected behavior of any bureaucratic institution.

        And of course, no discussion of why college education is different than primary or secondary education. Perhaps we should eliminate those as public goods as well and make all schools private schools?

        Yes! Odds are that good private schools would be much cheaper than the price differentials associated with good school districts now, and of course, you wouldn’t even have to move. I’d happily accept a voucher plan, but I’m enough of a Caplanian on this that I would also expect ‘no vouchers’ to yield a net improvement.

        • HeelBearCub says:

          I’m struggling to find a way to define “less regulated” that makes this plausible. Can you unpack that?

          Would you like single provider (complete government ownership of healthcare), single payer, or universal insurance? Those are the models followed in all the the other 1st world countries. In what model are those options less regulated than our system?

          • Lupis42 says:

            I would like a Singapore model – as described here:

            But that’s not my point. Lack of market low regulation. Single payer systems almost certainly involve less regulation than the US has now, because the payment end is done through the tax code, which could only be made more complicated than the current insurance market through a malignity that would boggle the mind.

            Healthcare in the US is (mostly) privately paid for and privately supplied, but it’s still incredibly heavily regulated. We limit the supply of doctors and hospitals through regulation, we restrict who can supply basic medical services to doctors even where there’s no obvious reason to do so, doctors have all sorts of restrictions on their interaction with patients, and on how much information they can share with other doctors. Hospitals have a variety of regulations on what they have to provide, and how they can charge for it.
            Maker’s of pharmaceuticals face a huge regulatory apparatus that restricts what they can make, who it can be prescribed to and what for, etc.
            Providers of health insurance face approximately 50 different regulatory regimes, one for each state they want to do business in, and each one of those has it’s own laundry list of requirements.
            Then, there’s all of the other tools that exist to help people deal with the cost of the care that insurance won’t cover.
            In 2008, operating an FSA came with a huge amount of regulations on how the payment providers could store, accept, and transmit data, how long it had to be retained, &c even relative to other, similar, tax advantage payment setups.

            I would bet that the amount of time, effort, and money spent on making sure that regulations are complied with in the US is more than what’s spent in Britain as percentage of total medical spending.

            So how are you defining “less regulated”?

          • HeelBearCub says:

            Wikipedia says that
            “Singapore has a non-modified universal healthcare system where the government ensures affordability of healthcare within the public health system, largely through a system of compulsory savings, subsidies, and price controls.

            The government regularly adjusts policies to actively regulate “the supply and prices of healthcare services in the country” in an attempt to keep costs in check. ”

            Price controls, single payer and universal health-care? Why yes, a free market Nirvana. 70% to 80% of all care in the country is provided through the public system, so it’s not as if the the private system that exists to supplement the public system can take credit for the great outcomes.

            If you want to claim that is NOT more regulation than in the US, I’m all ears. But typically if someone proposes single-employer or single-payer in the US, this will be objected to as an increase in regulation, not welcomed as a decrease.

            Do I think our existing system is horrible? Yes. Would I like to see a model that is more like Singapore (or Germany or France or England or Cuba or Switzerland or Canada or any other country that has far better outcomes for far less money)? Yes.

            But all of those models go towards a single-unified market where almost all health care is provided directly or indirectly by the federal government. Even the Swiss market (which consists of relatively few private insurers) consists of health plans where the covered treatments are mandated by federal law.

          • Lupis42 says:

            And the sentence after your quote in that paragraph:

            However, for the most part the government does not directly regulate the costs of private medical care. These costs are largely subject to market forces, and vary enormously within the private sector, depending on the medical specialty and service provided.

            Also, from further down:

            Many Singaporeans also have supplemental private health insurance (often provided by employers) for services not covered by the government’s programmes.

            Patients are free to choose the providers within the government or private healthcare delivery system and can walk in for a consultation at any private clinic or any government polyclinic.

            The market is a lot more free along a great many axes than it is in the US. This is a perfect example of where a lot of the talking past each other happens when people talk about free markets and regulation.
            Singapore has a market for health care, in which the state is a player. This means that the state competes on price, which keeps prices down in the private sector, but it also means that the state faces price pressure to adopt cost saving techniques when the private sector comes up with them.
            The state helps ensure that the market is a market by requiring private medical providers to tell you what the price is before you agree to pay it, which is technically a regulation, but which I classify in the same general space as “No defrauding your customers”.
            The state also provides emergency medicine – which, I’m quite happy to concede if it gets a functioning market to the *rest* of healthcare.
            And the state subsidizes the poor, and mandates targeted saving for the rest. It’s not the best system I can envision, but it’s a hell of a lot more market-friendly than most other things people are doing, and it’s proven to work in at least one country, so sure.

            I’m not sure where you get single payer though, because Singapore is more or less the opposite of that – it’s completely user-pays. Hence the price competition, means-tested assistance, and government mandated private savings accounts.

            We’re now comfortably far afield of the definition of “less regulation” in a private market. I would suggest a simple measure: the amount of money a private institution has to spend determining whether or not it is compliant. If the answer is automatically “no” because there are not/cannot be private profit-making firms in the industry, than there is no market.

          • HeelBearCub says:


            Yes we are “talking past each other” because you ignored almost everything I wrote.

            The government enacts price controls. 70 to 80% of the care is provided directly by the government.

            If, in the US, the federal government proposed to own 4 of 5 hospitals, pay 4 of 4 doctors’ salaries, etc. and force every single person into the the federal insurance plan that payed for those, very few would be saying this amounted to the government NOT regulating healthcare or reducing regulation. They would be concentrating on the regulation of doctors, hospitals and individuals this would be constitute, as well as the regulation out-of-existence of the vast bulk of the private insurance market.

            Heck, we already have supplemental Medicare plans, an analogous situation with one insurer for those over 65 but with supplemental private insurance possible. Is this less regulation than the ACA, where (almost) all the insurance is private market (which must comply with laws)? Do you argue for keeping Medicare as is, or privatizing it?

            Now if you want to argue that it is more efficient for the federal government to regulate the market by directly owning the supply side, thereby turning those laws into policies of the health department, I agree with this. It would be more efficient.

            Whether that constitutes less regulation is where I think you are engaging in bait and switch.

          • Lupis42 says:


            Yes we are “talking past each other” because you ignored almost everything I wrote.

            You described Singapore is single payer, which it isn’t, and cited Wikipedia, which never describes it as single payer, but does talk about mandated savings accounts, private providers selling at market prices, and public providers competing on price – all features that would be nonsensical in a single payer system.

            The government enacts price controls.

            Citation needed. The government acts as a supplier, and competes on price. The government requires publishing of prices. Neither of these is a price control.

            70 to 80% of the care is provided directly by the government.

            If, in the US, the federal government proposed to own 4 of 5 hospitals, pay 4 of 4 doctors’ salaries, etc. and force every single person into the the federal insurance plan that payed for those

            If the government set itself up as a supplier of healthcare, competing on price with other providers on the market, (but restricting medicare/medicaid payments to government medical facilities, that would be an enormous pro-market reform on it’s own. If the government was sufficiently successful at this to become the dominant provider with 70-80% of the market, that wouldn’t bother me much – as long as new entrants were still permitted.

            They would be concentrating on the regulation of doctors, hospitals and individuals this would be constitute, as well as the regulation out-of-existence of the vast bulk of the private insurance market.

            The how matters here. Entering a market, competing on price, and winning is not a regulatory strategy. Changing the rules for private providers until they fold is a regulatory strategy.

            Heck, we already have supplemental Medicare plans, an analogous situation with one insurer for those over 65 but with supplemental private insurance possible. Is this less regulation than the ACA, where (almost) all the insurance is private market (which must comply with laws)? Do you argue for keeping Medicare as is, or privatizing it?


            Now if you want to argue that it is more efficient for the federal government to regulate the market by directly owning the supply side, thereby turning those laws into policies of the health department, I agree with this. It would be more efficient.

            I argue that it would be a lot more efficient for the government compete with the private sector to supply care than the for the government to subsidize the purchase of insurance policies, mandate that those policies cover the sort of care least efficient to insure against, and prevent the providers of those policies from using information about risk to make them work properly.

          • HeelBearCub says:

            The government enacts price controls.

            Citation needed

            It’s literally the last two words of the first sentence I quoted from the wikipedia article.

            Do you think England is single-payer? Do you think Medicare is single payer?

          • HeelBearCub says:

            More on cost controls in Singapore

            The main point of that blog post is debating whether HSAs “work” on their own. This is not my point, rather in it you can see several sources cited for how Singapore imposed controls.

            Specifically an eBook by William Haseltine which is cited as providing the following:
            “They limited the number of physicians. They controlled the number of specialists. They limited the numbers and kinds of expensive technology that could be purchased. The government bought and ran all public hospitals, which provide 80% of care in Singapore. Some kinds of care were limited to just a few hospitals. ” (This is a quote from the blog post, not a direct quote from Haseltine.)

          • Lupis42 says:

            Do you think England is single-payer?

            As far as I know it’s single provider, and possibly also single payer, but only thing I know about the NHS is that I recently saw an article saying it was overwhelmed and struggling to keep costs down.

            Do you think Medicare is single payer?
            Medicare is not a complete system. It’s single payer only in the sense that Blue Cross is single payer, or for that matter in that my household is single payer.

            More on cost controls in Singapore
            Thank you. I’ve been trying to do some research of my own, but I’m short on time and finding a lot of conflicting information, and very little in actual detail, so I’m giving up on trying to figure out how Singapore actually works for now, and I’ll just lay out my perspective on this:

            To the extent that a government restricts private supply, enacts direct controls on private prices, etc, we’re talking about more “regulations”. My prior is that these tend to be unhelpful, but there’s a huge amount of variability here.
            To the extent that the government competes on price as a supplier, we’re not talking about regulation – there’s nothing stopping a private provider from competing, or using new or different models.
            When the government subsidizes some care or some providers, and doesn’t subsidize other care/providers, that’s “regulation”, but it’s impact is usually going to be substantially less than the first example.

            So the US, in the pre-ACA world, was a very very very heavily regulated market – possibly even more that it would have been were it single-payer but still privately provided. That regulatory structure, among other things, put the incentives in place for an insulating layer (insurance companies) hiding the costs from the consumers.
            That produced a bit of a worst-features-of-both outcome, in that we got neither the price competition of a proper market nor the bureaucratic pressure of a top down system.

          • HeelBearCub says:

            Do you think Medicare is single payer?
            Medicare is not a complete system.

            That’s the second time you have referenced that particular Greek letter in a way I am not familiar with. I’m not sure what you mean.

            In any case, I think you are using a non-standard definition of single payer. Medicare is the ONLY allowed primary insurance for everyone over 65. You can have supplemental policies, you can pay completely private market (but I believe those doctors can’t take medicare payments at all then). Medicare is the single payer for all essential medical services for everyone over 65.

            If you don’t think that is single-payer, then you need to either support through citation that this is not commonly recognized as single payer, or recognize that you are trying to use a non-standard definition (and then not straw-man arguments by substituting your definition when someone is using the term intending the common one.)

            As to England, it is single-provider (with some ancillary private practices allowed) and single payer (the government pays for it all). Single-provider is de-facto single-payer. Theoretically, you could have a single mandated provider and then many private insurers, but I’m not aware of anywhere that has such a byzantine setup.

            Incidentally, single payer does not rule out insurance premiums, co-pays and deductibles. If you are trying to shoe-horn that into your definition of single-payer, then I would again want a citation.

          • notes says:

            Ran into the spam filter trying to link a definition of mu – let’s see if this works better.

          • HeelBearCub says:

            Ah, I see. You think referencing Medicare is a “gotcha” question.

            No, I’m really try to understand what you think single-payer actually is. My contention is that you aren’t using the definition of single-payer that is standard, as you don’t think Singapore is single-payer.

            Singapore makes public providers available to everyone where they means test subsidize up to 65%-80% of the cost of visits. They force everyone into a plan called Medisave that saves money for their care. They then allow people to buy extra coverage (Medishield and perhaps private) and also have an equivalent to program to Medicare/Medicaid (ElderShield/MediFund)

            But you can’t have additional private insurance unless you first buy MediShield. So every healthcare transaction involves the government paying for healthcare directly (subsidy, MediShield, etc.) and indirectly (MediSave, which is compulsory). There is no competing private market of insurance, it’s all additional coverage.

            More from Singapore’s Ministry of Health

          • Lupis42 says:

            Ah, I see. You think referencing Medicare is a “gotcha” question.

            I don’t think you intended it as a ‘gotcha’, but I think it contains an assumption that is invalid – I don’t think calling Medicare single payer or not makes sense. Just trying to be concise, since I’m generating more than enough words anyway.

            Single-payer describes things from the providers perspective. If providers are billing one entity, presumably a government entity, then you have single payer. (Even, as you say, if that entity passes copays or deductibles down to the consumers) That’s single payer. If providers bill consumers, or a variety of insurance companies, or a mix, that’s not single payer (even if they also bill the government).
            Theoretically, a country could even set up a single-provider system where the country ran the facilities, but those facilities billed users, which would be single-provider without being single-payer, although there wouldn’t be much point.

            WRT Singapore, I’m interested in the market in care, not the market in insurance. AFAIK, public providers bill consumers for some or all of their bill. Private providers do likewise. That means it’s not single payer.

        • Deiseach says:

          Private schools are fee-paying. For those who can’t afford fees, you have to fall back on public schools. And a two-tier education system gets you “self-perpetuating elite” and “everyone else, consigned to be drones”.

          Scholarships are one means attempting to bridge the gap between “smart but poor” and “what school you go to”, but even there, as in the U.K. with its grammar schools, there has been some necessity for “social engineering”; removing class-based questions from the early versions (questions on the role of household servants? really?) and apparently, since girls do better in exams than boys, sitting the “Eleven-Plus” meant that if they took the results strictly on grades achieved alone, the girls would crowd out the boys for places in the schools, so they had to weight the figures to let equal numbers of boys in.

          • Julie K says:

            > two-tier education system

            We already have that in America; the tiers are inner-city public schools and middle-class-suburb public schools.

  37. Linch says:

    I don’t think your proposed solution will have significant utility.

    Q: “So I see here that you did not have a job/only had a part-time job doing “IT” at the “University of Elbonia Computer Science department.”

    A (idealized): “I do not wish to disclose. This is protected information.”

    A (any rational job candidate): “Oh, I attended the University of Elbonia, ranked 33rd in the nation. I graduated cum laude with a Underwater basketweaving degree and learned important skills like X, Y, Z.

    The point is, even if candidates are not required to reveal information, any beneficial information WILL be revealed. “Protected class” doesn’t help when refusing to reveal is implicitly a bad signal.

  38. OK, anyone who writes on this topic without mentioning Griggs is signaling that he doesn’t know what the hell he’s talking about. I see others have already mentioned this fact more politely. I just googled his entire blog; only commenters mention it, from what I can see. That’s just absurd.

    The single biggest reason US employers started using college degrees as a proxy is the Griggs case. Basically, they off-loaded their intelligence discrimination to colleges, who could reasonably require a certain level of cognitive ability back in the 60s and 70s, when this began.

    Then the country proudly demonstrated that cargo cult mentalities aren’t unique to primitive cultures. “Hey, people with college degrees make more money. We need more black/Hispanic/poor people in college!” And of course, once they start focusing on it, they realize “Hey, people who take algebra I are more likely to go to college. Let’s start requiring everyone to take Algebra I!” and once we successfully degraded algebra I, it turned out that the *real* marker for college success was algebra II, so the requirements were upgraded once again.

    High schools couldn’t set standards without being fined for EEOC or facing a class action lawsuit on behalf of the Hispanic/black/poor kids of the district, which they mostly settled.

    Colleges, whose contempt for high school teachers knows no bounds, started remedial courses, thinking hey, the kids just don’t know math/reading/writing because they had incompetent high school teachers, so we can just fix that right up. They’ve taken close to 30 years to figure out otherwise, and when they did figure this out, they realized that if they started demanding a certain level of achievement, they’d be the ones getting sued for discrimination or disparate impact. So instead, state colleges are now declaring an end to remediation entirely, saying kids go straight into college level courses regardless of ability. College professors will thus start to see more scrutiny of their pass rates for racial discrimination, and that’s in a world where they are already under considerable pressure to pass everyone. Community college courses are only as rigorous as their average IQ can tolerate.

    Meanwhile, colleges need more money, so as long as they are accepting totally unqualified students, the thinking goes, they may as well accept Chinese kids with rich parents willing to pay full freight for both the fraudulent records and the actual university fees.

    These days, a college degree is as useless a predictor as a high school diploma was 30 years ago. Today, a high school diploma means nothing and will soon mean less than that if more states don’t abandon the new, harder GED. Because that GED will put even more pressure on high schools to graduate kids who can’t read or write, since they won’t have a plan B anymore. Failing to graduate kids of color will, soon enough, be another area for lawsuits.

    At the time Europe set up its educational system, it was still homogenous and tracked ruthlessly. I don’t think any European country has the equivalent of a Civil Rights Act or the 14th amendment, so they don’t face lawsuits–just social pressure. Which isn’t the same thing, historically. America never tracked ruthlessly, although it did track more in the past.

    So comparing the US to European schools is just tragically stupid.

    This comment is long enough, so I’ll do testing in a new one.

    • Randy M says:

      “Meanwhile, colleges need more money” I suspect this is true only in the sense that “businesses need more money” is also true.

    • Scott Alexander says:

      I’m familiar with Griggs, but I decided not to go off onto a tangent on exactly why credentialism happens here.

      PS: Your tone annoys me. First warning.

    • Psmith says:

      Yep, agreed. This also suggests that the appropriate solution is to overturn Griggs and bring back aptitude testing in hiring, not to make education level a protected class.

  39. Daniel Armak says:

    Without detracting from the rest of your examples, the one about programming is wrong. A Computer Science degree does not teach programming, it teaches computer science – math, algorithms, etc. Most programmers, and certainly entry-level programmers with no programming experience, don’t need to know comp sci; and there are other degrees (usually called something like software engineering) that actually spend several years teaching programming (or at least trying to).

    ETA: Because lots of people who study comp sci as undergraduates only want to learn programming and get a degree paper, modern degree programs certainly do have courses on programming, the amount varying a lot between universities, but they are still a minority of the course load – or supposed to be, as long a it’s called a comp sci degree and not a software eng one.

    Also, one would have to be talented to the tune of several sigmas to learn programming from scratch in 3 months. But with the kind of informal exposure many tech-savvy people get for years beforehand, it’s plausible to learn a few specific subjects well enough to get an entry level position.

  40. Mary says:

    ” Ireland is one of the approximately 100% of First World countries that gets better health outcomes than the United States.”

    ha. ha. ha.

    The United States get the best healthcare outcomes in the world. Whenever we look at studies that find things like that, we discover the “health outcomes” are either consequences of things outside the control of the healthcare system (such as car accidents, bad diets, mothers who do drugs and drink during pregnancy) or not health outcomes at all but the failure of the healthcare system to adhere to leftist policy prescriptions.

    • Edward Scizorhands says:

      I’ve heard this argument from smart people and it seems decent. I’m not convinced, but it definitely merits discussion.

      What we can say is that the US spends a lot more for not-noticeably-better results. That’s not necessarily wrong, because maybe we want to spend more money on health deliberately, except we never really voted or shopped for it, so it’s like Moloch chose the price for us.

  41. onyomi says:

    If the diploma were a pure proxy for ability then surely a test would be an even better indicator. But what if the diploma isn’t so much about signalling ability as signalling membership in a certain social class, membership in which is associated with other positive things, like understanding of the need to show up on time, or ability to engage in good water cooler banter?

    Also, I think an interesting twist on the tulip story would be to make the tulip-using population 80% male. This would reflect the fact that, right now, the economy in the US is so weak and so unfavorable to employers that there is a huge mismatch between number of job seekers and number of jobs offered. The result of this is employers needing to winnow applicant pools by, for example, immediately throwing out all non-degree holders.

    If each woman has three suitable suitors, she can afford to dismiss those offering the cheaper daffodils out of hand. So long as this mismatch exists, there will be a strong incentive to find some criteria other than strict suitability to aid in decision making.

  42. Mary says:

    “At what point do we say “Actually, no, let’s not do that, and just let people hold basic jobs even if they don’t cough up a a hundred thousand dollars from somewhere to get a degree in Medieval History””

    When we force all the lawyers, judges, legislators, governors, presidents, and bureaucrats to go back to school.

    Namely, Statistics 101.

    Then we will laugh disparate impact suits out of court, and employers can go back to using tests that WORK, not that won’t get them sued, such as degrees.

    So we will get your ” You can give them all sorts of examinations,” which is effectively illegal, and not even have to add a law.

    • Peter Scott says:

      What makes you think that taking Statistics 101 will give people a basic understanding of statistics? As far as I can tell, its main effect is to give people vague memories of plugging numbers into formulas. Some of the better students will grasp what those formulas are for; of those, most will forget it a year later. If I had to make a list of ongoing educational disasters, the standard introductory statistics classes for non-majors would be on the first page.

  43. On employer testing: It’s simply not true that private employers can test without risking disparate impact lawsuits. While Google et al don’t face much pressure, it’s largely because they can point to the lack of applications.

    Back in the days of corporate IT, when many decent-paying jobs didn’t require college degrees or programming skills, most large companies had a program to ensure they were moving blacks and Hispanics through IT to ensure they didn’t face lawsuits. Within IT, you had lots of blacks and Hispanics in operations jobs, and the strongest of these would move into low-level programming jobs. These days, corporations are specialized, operations as a department doesn’t exist (we call it network administration instead, and it’s often outsourced to contractors). So Google/Facebook can say, truthfully, that they aren’t turning down blacks and Hispanics–because they aren’t applying. If they handled a large number of blacks/Hispanic applications, they’d quickly be under pressure to hire more of them.

    Few employers can pass the test to require test scores. Kaplan, the test prep company (it doesn’t do much of that any more), could legitimately require people to score in the 98th percentile, thus eliminating it as an option for blacks and Hispanics, because they were selling high test scores. They could ignore the fact that the test was a proxy for cognitive ability, because “we need teachers with high test scores to teach others how to get high test scores” was directly on point.

    Teaching, oddly enough, is another area that can demonstrate legitimate need for demonstrated knowledge through tests or college proxy. High school teachers have had credential tests for 30 years or more; when states began requiring elementary school teachers to pass tests in the late 80s, they were sued. And they won, because they could legitimately argue that demonstrated knowledge on elementary school subjects was an essential component of a job. (Note: people who think credential tests are easy are usually thinking of the early tests. Since 2002, elementary school teachers in most states have to demonstrate 10th grade level ability in all four subjects, which may sound easy but it’s a pretty high standard when you consider how many teachers we need.)

    I wrote on ed schools and affirmative action, explaining why teaching, the field routinely dismissed as low-achieving, actually has the most stringent credential exams. The high school content knowledge exams in math and science in the tough credential states probably require more cognitive ability than the bar exams in some easy states. In Alabama, for example, a person scoring 3 points below the black average LSAT score still has a 1 in 4 chance of passing the bar. I’ve taken the high school tests in 3 subjects in a tough state, and no one with SAT section scores in the low 400s has a shot at passing the tests. My guess, without knowing cut scores, is that they’d be hard for most people who didn’t have 1500 (2005) SAT scores or higher, and that leaves out a whole bunch of blacks and HIspanics.

    But a teacher exam can legitimately require demonstrated high school knowledge for a high school teacher. So you see this constant tension between the education reform folks, who are less likely these days to tag people as stupid but still think it makes sense to “raise standards” by raising credential test cut scores, and the reality that higher cut scores will simply destroy the base of black and Hispanic teachers. That’s why the head of CAEP (the ed school credentialing agency) just got fired, because he’d instituted credentialing demands for ed schools to require scores of 1800 or higher (current SAT) OR grades in the top 3rd of their graduating class. Elite ed schools can thread that needle by the OR and get white candidates with high test scores, URM candidates with low scores but good grades at low achieving schools. State schools, the ones pushing out most teachers, can’t. HBCUs are completely screwed. And so he was fired, and my guess is that new requirements will come out soon.

    Likewise, NY is currently pushing back its more stringent tests because it’s clear they would otherwise face a disparate impact lawsuit. In fact, a judge just ruled its credential test racially biased. Why? Because the state had basically designed the test to identify cognitive intelligence rather than directly on point teaching knowledge.

    While these lawsuits lost back in the 80s and 90s, it’s a different story these days, because the research is *at best* ambiguous proving any link between smarter teachers and better academic outcomes. At worst, there’s no link—and there’s plenty of research showing that black kids do better with black teachers. Follow the history of any attempt to raise teacher standards past the 2002 effort, and it follows a similar standard. I would argue that our standards for high school teachers, which have been consistent for 30 years or more, are about right. The standards for ES teachers are probably higher than they should be. I’d say we could create new standards for K-3, 4-6, to let in more URMs.

    So look at any company that requires tests, and they’ve got one of two scenarios: a legitimate reason to require high scores, and that reason isn’t “they get better results” but “the knowledge on the test is *directly* related to the job” and even then, they’ll face trouble. Ask the states trying to come up with fire fighter/cop tests.

    The second scenario is a company that doesn’t get many URM applicants, and that’s what is happening with tech.

    Okay, sorry for the long second comment. Last one should be short.

    • Murphy says:

      Huh, it sounds like the book Incompetence ended up based on the wrong continent.

      The book is about a world where discrimination based on competence was banned hence everyone the main character interacts with is spectacularly inept and unsuited to their job. Think blind nightclub bouncers, octogenarian male lap-dancers etc.

      (“No person shall be prejudiced from employment in any capacity, at any level, by reason of age, race, creed or incompitence,”)

  44. Hemid says:

    This seems simple.

    “Free” means “paid for by people who don’t receive it.”

    It’s traditional for the peasantry to be forced to fund the work and leisure of the gentry and to receive, on average, far worse than nothing for it.

    Does that tradition need reinforcement? Is it in danger of fading? Would its death be a loss?

    For whom?

  45. Chen Dason says:

    “You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail. I realize this is a blatant violation of my usual semi-libertarian principles, but at this point I don’t care.”

    This was unusually dense – the reason many jobs require degrees is because due to current law, an employer cannot give people all sorts of examinations when they apply for a job without quickly running afoul of discrimination law, and have trouble getting accurate work histories and recommendations due to fear of lawsuits. It’s perfectly reasonable that they would then outsource this portion of their HR function to universities – anyone who graduates underwent the necessary examinations and has at least 4 years of reliable history.

    That said, I don’t think this is as big a problem as you’re making it out to be. If as you say there’s this large pool of untapped potential in the poor classes, why isn’t there a company out there hiring them and beating the wasteful incumbents?

    The degree requirements for medical school are just another way to artificially hold down the supply of doctors – if the AMA couldn’t do this, they’d lobby for something else. Much more important is just to drastically increase the number of medical schools or expand the privileges of healthcare professionals like nurses.

  46. BBA says:

    At elite levels, getting a job is more about who you know than what you know, and elite universities are a big boon in the “who you know” department.

    Lower-tier universities don’t and can’t work like Ivies, but everyone involved either doesn’t understand this or doesn’t want to admit it, yadda yadda yadda, crippling student debt for no good reason.

  47. Finally, on why Ireland doesn’t require a college degree for med school, and the US does. Whenever you see lower standards, you can expect that it has something to do with competition. When I read this, I instantly figured that Ireland wants to increase its desirability to students from other countries and lo, look what I found: Shut out at home, Canadians flocking to Ireland’s medical schools — and to an uncertain future . It’s a 15 year old article, but this one, from 2010, says that the practice is increasing. All the people involved are college graduates who are on waiting list for Canadian med school–which does require a college degree. I couldn’t find anything saying that Irish medical schools accepted high school graduates. I’m not doubting the claim, but I can’t find the specifics.

    The point is, Ireland is clearly looking for people from countries with very stringent med school requirements who still want to be doctors. They’re the new Caribbean med school, I guess, with colder winters but more museums.

    And that’s what you’ll usually find if you scratch under the surface of any unusual admissions policy. It’s not set up to increase access from local students, but to increase attractiveness to wealthy students from other countries.

    In the US, calculus and even pre-calc have only been high school options since Sputnik (it’s one of the “new math” innovations everyone forgets about). Until the 50s, high school math was algebra–and not the algebra I of today, but real algebra that included much of what we call algebra II–two years of geometry, which was mostly about memorizing formulas, and really advanced math included trigonometry. And relatively few high school grads had even made it to geometry, because they failed out of algebra.

    Calculus was an upper-level college course for most of our history. That probably has something to do with why we require college first. Of course, today, it’s ludicrous to even think about letting high school graduates go to med school without college first, because it would subject med schools to the same disparate impact scrutiny that high schools and colleges now get, and we Americans, we like our doctors to be competent. As it is, black doctors have higher malpractice insurance rates. But med schools keep their stats firmly under wraps; it’s hard to find out the black pass rates on the boards, for example. I imagine it’s the same for Hispanics.

    So that ain’t happening here. What we’ve done instead is put a lot of lower level medical training in college: OT, RN, PA, and so on. I imagine the five year program Ireland has for high school graduates is much the same thing.

    Oh, one more thing: Only two states require MAs for teachers these days, and they are backing off of it. But in most states, the MA is simply a way to get more money. And the reason teachers get paid by getting more education is not because anyone believes that it will actually improve their teaching, but because it’s an unambiguous way to pay teachers more, rather than get into the ultimately doomed area of merit pay.

    • Deiseach says:

      I couldn’t find anything saying that Irish medical schools accepted high school graduates.

      From the Irish Central Applications Office online site (this is where you apply for third level education after you finish secondary school); courses for Medicine and Scott’s alma mater, U.C.C. and its requirements for medicine:

      Course Title: Medicine

      College: Medicine and Health
      Duration: 5 Years

      Teaching Mode: Full-time

      Qualifications: MB, BCh, BAO (Hons)

      NFQ Level: Level 8

      Costs: Full-time EU/EEA/Swiss State undergraduate students may be exempt from paying tuition fees. The State will pay the tuition fees for students who satisfy the Free Fees Criteria. In 2015/16 the Student Contribution Charge will be €3,000 and the Capitation Fee will be €165.

      Entry Requirements: HC3 in Chemistry and either Physics or Biology and passes at H or O level in the Leaving Certificate from Irish, English, another language and Mathematics. A minimum of 480 points at Leaving Certificate is required, and students must also sit the HPAT-Ireland (Health Professionals Admissions Test – Ireland) exam and combine the results of this with the Leaving Certificate.

      Entry Points: 2014: 724. Points may vary from year to year

      Course Code: CK701

      “H” and “O” are the Higher and Ordinary/Pass levels in subjects for the Leaving Certificate, our national examination for the final year of secondary school. “HC3” is “Grade C3 (marks 55-59% in test) at Higher level”. Points are based on the grades you get in that examination and the level at which you achieve that grade (thus the entry requirement in 2014 for “724 points” which would be calculated from your Leaving Cert and the score of the HPAT exam). To give you an idea, I’m so old that in my day we didn’t have the C1, C2, C3 etc. divisions when I did the Leaving, so calculating my points on seven subjects, I would be in the range 300-370 (definitely too stupid for medicine) 🙂

      I’ll throw in the Royal College of Surgeons in Ireland as well; you can see their CAO course code RC001.

      • Murphy says:

        the 724 points is unusual in the irish system.

        The maximum score you can get on the leaving is 625. (6 subjects 0-100 for each, they recently added a change where higher math gets a bonus of 25 possible points for many courses)

        They added the hpat for medicine since for years you pretty reliably had to get 580 or 590 out of 600 and they wanted to test for some of the more med specific stuff to distinguish between people more.

    • Nita says:

      And that’s what you’ll usually find if you scratch under the surface of any unusual admissions policy.

      The admissions policy in the USA and Canada is the unusual one. Ireland just does the same thing as the rest of the world.

    • Murphy says:

      Highschool students absolutely can and do go into med in ireland. One of my close friends went into med from highschool and did extremely well winning lots of first-in-year awards thrashing most of the foreign students.

      The uni does make a premium from foreign students but about 2/3rd of the year is normally irish/EU students fresh out of second level school.

  48. AcidDC says:

    I’m down with your plan.

    I think the Sanders plan is also a bit better than you give it credit for though. The key difference between it and the tulip subsidy is that Sanders isn’t recommending that we pay for people to go to college no matter how much it costs, he’s proposing that we cut tuition to zero or almost zero at public schools. This means the government’s costs are going to be based on how much it actually costs to run those institutions, not endlessly upward spiraling tuition bills. Presumably having a free education on the table will also have a positive effect on the price of education at private schools.

    So basically it’s not so much subsidizing tulip purchases as it is having government controlled tulip farms that ration out enough low-priced tulips that no one has trouble getting married, while letting the private tulip market continue to do its thing.

    • Mary says:

      Of course, what that’s doing is putting more control over what is taught in government hands. That would be unwise in many ways.

      • Taradino C. says:

        How would it put more control in government hands? These public institutions already exist.

        • Mary says:

          It’s funneling more students toward them.

          • Taradino C. says:

            That presumes the institutions are running below capacity today, and they’d be teaching more students after eliminating tuition. I’m not sure that’s true in most cases: how would tuition have been bid up so high?

            I suspect there’d just be more students competing for the same number of seats, with the unlucky ones having to pay for private schools or go without.

          • Mary says:

            You assume their capacity is a fixed thing.

  49. Jonathan Paulson says:

    My college degree (CS/Math) has provided good value: in addition to the signaling you’re concerned about (which, to be clear, was definitely valuable), I actually learned a lot. App Academy would not have been a good replacement: it only teaches web development and a little bit of theory (e.g. the most advanced algorithm they cover is sorting). As a direct consequence, I have been able to get more-interesting higher-paying jobs than I would have through App Academy.

    Your solution doesn’t address the thing that companies need: a high-signal way to screen through resumes. “College” isn’t great, but it’s better than nothing.

    • Ilya Shpitser says:

      I find it extremely amusing that some people think App Academy is a replacement for a proper computer science education. That App Academy graduates get jobs is more a testament to a desperate labor shortage than the merits of that type of education for creating good programmers.

      A lot of “coders” that float around in the Valley are _ruinously_ bad.

      • Jones says:

        Are you assuming that only universities can provide “a proper computer science education”? A list of exceptionally skilled programmers and coders who have never attended University courses, or have dropped-out prior to completing any sort of degree, would be astonishingly long.

        The resources are out there. For the most part, they’re free or nearly free. Interested people can learn on their own, at their own pace.

        (…It’s what I’ve done; I’m a first-semester dropout, but I now own and manage several highly successful tech businesses and have become quite wealthy. When hiring, my organizations don’t assign any importance to a degree, but only to skills and attitude.)

        • HeelBearCub says:


          What he is saying is you don’t get to know what you actually need to know by taking at a 3-month course and paying $15K for it. Given their model (you only pay if you get a job), this looks like more of a pump-primer for tech-hub companies hungry for talent. Much like paying a recruiter if you hire their candidate.

          The tech-hubs are happy enough to pay people to learn on the job, once they have been satisfied that the candidate has the aptitude for doing so.

          The model falls down completely once the labor market loosens.

          • Jones says:

            “What he is saying is you don’t get to know what you actually need to know by taking at a 3-month course and paying $15K for it. ”

            What is it that they “need to know” that they’re not getting? If that course suffices to find them a job decent enough to cover the $15k fee, if it gives them a foot in the door in the industry, and if it can help signal to future employers that the applicant has the innate aptitude to take more advanced training… isn’t that enough? A 4-year CS degree would cost vastly more, take much longer to complete, wouldn’t guarantee you a job, and wouldn’t teach you anything that you can’t easily, easily learn on your own time for next to nothing.

            I know quite a few really decent coders, and all of them are self-taught to a significant extent. Pursuing a bachelor’s-level CS degree would be a disgusting waste of their time and money. I also believe, wholeheartedly, that these people are not unusual — that they’re a dime a dozen in the tech industry.

          • Mary says:

            Some of you may have heard of the term “Suzie COBOL” meaning, “A coder straight out of training school who knows everything except the value of comments in plain English.”

            (Training school meaning a 12-week course that IBM would send secretaries to; it was, at one time, their biggest supply of female programmers.)

          • Daniel Armak says:

            Jones, I think you’re confounding CS degrees with programming. CS isn’t primarily about programming; it’s about lots of delicious math with programming thrown in as a bonus. CS to programming is as theoretical physics to electrical engineering.

            Some people can learn the content of a 4-year CS degree, or any other BSc degree, on their own, without teachers. Most people do much better in a good university, with fellow students, helpful teachers and a regimented program. And even among those who can learn it on their own, people who could learn it all in 3 months with no prior knowledge are extremely rare.

            A typical 3-year CS degree usually covers these core subjects, most of them at a 1-semester level and a few, like calculus, at a 2-semester level:

            – Math: calculus, linear algebra, combinatorics, intro to probability & statistics, intro to formal logic, set theory & algebraic structures, graph theory, intro to cryptography.
            – Basic programming: typically C, one mainstream higher-level language (Java or C# or even C++), assembler, SQL. These days probably also something Web-related.
            – Applied comp sci: algorithms, data structures (at least 1 course each), operating system design, compilation theory.
            – Theoretical comp sci: computation theory, complexity theory, programming language theory.

            This typically takes 2.5 years out of 3; to this add 1.5 years of other courses, probably mostly electives, in a 4-year degree.

            It’s perfectly true that if you want to become a great programmer, this is far from the optimal way to spend 3 years. Graduates with a comp sci degree and zero programming experience are notorious in the industry for being, on average, not at all useful as programmers and needing a lot of extra training. In line with Scott’s original point, it’s a problem that wannabe programmers feel they need to study comp sci and not, say, software engineering. But it’s simply not true that anyone can learn the contents of a CS degree easily on their own or that they have no value to programmers.

          • HeelBearCub says:


            I know quite a few really decent coders

            I am a really decent coder, with 20+ years of programming experience. I have interviewed and hired many programmers.

            Most people who are self taught have absolutely no idea how to write a decent, efficient algorithm. They have a tendency to use “brute force”, are less likely to write elegant code and typically have trouble knowing how to debug a problem, especially someone else’s code.

            But they can get a piece of code to do what it needs to do when looking at it from the outside (well at least if it doesn’t come under load).

            Is every self-taught or quick-taught programmer like this? No, of course not. But good lord I have seen plenty.

            Now, I happen to live smack dab in the middle of three damn good comp-sci universities, so there is that. It could just be availability bias. But it’s not as if there is no empirical basis for my biases.

          • Ilya Shpitser says:

            Yeah, I concur with above. Being a good programmer is sort of like being a good mathematician (proving theorems, and writing programs are very similar activities, after all).

            I have personally never seen a good autodidact programmer (I am sure they exist, but they seem to be rare).

            I am sure autodidacts can and have become rich in the Valley. But that is not what we are talking about, is it?

          • Mary says:

            “But they can get a piece of code to do what it needs to do when looking at it from the outside (well at least if it doesn’t come under load).”

            Had a job where we had a programmer whose code would always work if you used his test case.

            Wasn’t self-taught, either.

            sigh. I was the go-to girl for “Hey, make this actually work,” that job.

  50. Matt C says:

    My wife had a class on health economics and her professor said “first thing we need to do is bomb all the hospitals.”

    Maybe we can bomb the universities next.

    (I guess we’ll need to bomb the defense contractors last.)

  51. “A billion here, a billion there, pretty soon, you’re talking real money.”

    Although often quoted, it seems Sen. Everett Dirksen (1896-1969) never actually said this.

    Dirksen’s comment: “Oh, I never said that. A newspaper fella misquoted me once, and I thought it sounded so good that I never bothered to deny it.”

  52. Allison says:

    The high costs of credentialing are exacerbated by today’s need for continuous lifelong learning. MOOCs have been pretty underwhelming so far, but I think they can be fixed: check out

  53. The App Academy example is a bit of an exaggeration. The kind of jobs App Academy grads get are mostly different from the kind of jobs people with a CS degree from Stanford get. App Academy rarely gets people jobs at Google straight out of the bootcamp, and getting a job at Google potentially means $50,000 in compensation more a year (once you factor in bonuses and equity). That said, if you went to App Academy straight out of high school, you could well be making more money after App Academy + 3-4 years of work experience than the new Stanford grad.

  54. Alexander says:

    Would you be able to write as insightful a blog if you hadn’t studied philosophy? If you hadn’t debated it with live professors and classmates?

    Also, you are a psychiatrist. Doesn’t philosophy enter into your work?

    • Murphy says:

      right. here’s the thing. this is going to meander off into the vague “self fulfillment” stuff which isn’t totally invalid but it is of very low value in economic terms. If the choice is between 2 money holes: give person A a roof over their head and treatment for their diseased leg or funding person B for a degree in philosophy there’s a big opportunity cost there.

      Even in terms of self fulfillment we give college a privileged place.

      How many high school students would be given 100K by the state to “find themselves” by traveling around the world and trying every drug and discount hooker they can find? it might be much more fulfilling than a degree in art history with similar economic returns.

  55. John says:

    When you started talking about tulips, my immediate first response was: why doesn’t the government grow its own damn tulips rather than paying other people’s inflated prices? That way the poor people are able to marry and have working roads. And it seems to me that the same holds for education.

    My modest proposal would be this: free government-run, government-examined MOOCs in all the major courses. So no buildings, no student accommodation, no sports teams, no pastoral care – just a bunch of people working from home to stream courses out to the students, who are living with their parents.

    The obvious advantage is that poor people get their tulips. On top of that, there are quite a few careers where a degree isn’t a tulip at all, but legitimately essential. If I’m looking for a research mathematician, I can’t force them to sit twelve hours of exams to see how good their overall grasp of the subject is – I have to rely on a university to do that for me. And a well-educated population has benefits beyond people being able to get jobs.

    Oh, and since students cost us money rather than earning us money, we can apply actual academic standards. No more intentionally easy exams to keep the tuition money flowing, and no more turning a blind eye to cheating and plagiarism. So after a little while, a degree from this place should start being more respected by employers than a degree from a conventional university.

    Let’s do a Fermi estimate of the cost per student. Each course needs lecturers, but given that the number of lecturers needed doesn’t scale with the number of students that should be insignificant when it’s been amortised. (This is true even if you splash out and get *good* lecturers, who significantly outperform textbooks – another benefit of the system.)

    I’m not sure what TA ratios are like in the US, but let’s say each piece of homework needs 20 minutes to mark and a student gets three pieces of homework a week. From my time TAing in the UK, this is an overestimate for most universities. Taking on-line recitals and such into account, that means you need roughly one full-time TA for every 35 students. Marking is a deeply unpleasant but relatively skilled job, so finding markers will be difficult – normal universities get around this by using grad students as low-paid or unpaid labour, but since we don’t have a normal college structure we would probably have to offer high salaries instead. (Working from home would be a perk as well.) Let’s say we offer £30,000 per year, which is definitely on the generous side. Then we’re looking at a bit under £1,000 per student per year.

    On top of that, you’re going to have costs from exams (you’ll need to have physical test centers across the US) and administrative overhead. Let’s be generous and double the overall cost to £2,000 (roughly $3,000) per student per year. That seems a whole hell of a lot more manageable than $50,000 per year.

    But let’s say you want to cut costs further. There are two easy ways of doing that. The first is that tuition only needs to be free for US citizens – we can soak everyone else for as much as we like. (This is a major source of income for UK universities.) Second, we can compromise on “free”. Instead of charging money up front, driving people into debt or locking them out altogether, we garnish people’s post-university wages until they’ve paid some or all of the cost. (Again, this is done in the UK. I have a lot of issues with the way it’s done, but at the very least it’s better than charging people up front.)

    It won’t outperform a really good university. Really good universities do things like supervisions/tutorials, in which students are divided into pairs or threes and spend an hour talking with a TA after each piece of homework, and that’s expensive. But it will outperform most universities, and that’s all it needs to do.

    The one downside I can see is that it’s limited to people who have and know how to use computers.

    • Murphy says:

      Ireland does have free gov funded 3rd level degrees for everyone.

      The cost is really quite low. I worked out the cost per citizen based on the governments education budget and it’s still lower than what the US spends on the same problem.

      The US tends to select the worst and most expensive of all possible models when faced with choices like that.

  56. Bryan Willman says:

    First, there is an argument that health outcomes in the US are in fact the best in the world – but you don’t see in general statistics because of the murder and accident rates. Note that in any case the differences are not all that large.

    Second, people need to understand two other deep forces.
    1. Employers in general and all of them of any size require some fast cheap filter function. It’s NOT that a degree in English Lit makes you competent to manage to the local branch office. It’s that imposing that requirement quickly and cheaply filters out vast numbers of people who are not competent to do anything. The fact that many perfectly competent people will be filtered out by the same rule is irrevelent. Saying “they should interview every applicant” is nonsense – I worked a long time a hot company where interviewing every applicant would have required more time than the entire working lives of the existing staff. You HAVE to have a filter.

    2. There’s quite a lot of status and class jostling in the human world, and especially in the US. If say Bernie Sanders waves his hands and everybody get to do a 4 year degree, what will follow is that many jobs require masters degrees, just to get the filtering of #1 and to hold the status.

    In both cases, the point of an undergrad degree is “not everybody has one” – if you change the world so everybody does have one, some NEW demand will be made for people to have some credential that everybody else doesn’t.

  57. I often ask my Smith College students which they would prefer: (1) They attend Harvard for four years but forever after everyone will think they worked at Burger King for these four years, or (2) They work at Burger King for four years but forever after everyone will think that for these four years they attended and then graduated from Harvard. Most pick (2).

    • onyomi says:

      This is very telling, and the answer unsurprising.

      As Bryan Caplan points out, there’s also the fact that if you just moved to Boston and started showing up in some of the bigger classes at Harvard and MIT, it’s not likely anyone would kick you out. Even if you asked the professor “hey, I’m just a local community member interested in the subject, mind if I sit in?” The reaction in most cases would not be “how dare you try to reap the benefits of our classes without paying tuition?” it would be more like “somebody is interested in my class for its own sake??!! *small tear*”

      • As it happens, my wife has done what you describe at a pretty good university in the Bay Area (the one I teach at). Our home unschooled daughter took a year of Italian there before she went to college, taking advantage of a program for letting high school students take a course in the summer plus being the daughter of a faculty member. Her mother thought it would be fun to learn Italian. She asked a professor our daughter thought well of if she was willing to have an auditor, got a positive response, and audited a year or so of Italian.

        • onyomi says:

          I actually had the parents of a high school student approach me earlier this year to ask if their son could audit my Chinese language class next semester. I inquired with the university as to their policy on this, and it amounts to: if the auditors don’t mind there being no record of their having been there, I am free to let whomever I wish sit in on the class. If someone wants credit, however, they must pay a fairly steep $3,000 per class, per semester, or so.

          In other words, not only learning (which you can do on your own nowadays with most subjects), but even instruction is basically free if you’re willing to do a little legwork; it’s the credentialing you have to pay for.

    • HeelBearCub says:

      A) interesting that you don’t ask about Smith, but rather Harvard.

      B) Are they paying the Harvard tuition? Because it doesn’t matter how great your education was if you can’t pay your student loans. Ivy League tuitions are palatable if you expect to get the kind of job that only an Ivy League degree can get you. Otherwise, they aren’t worth it.

  58. Daniel Speyer says:

    Not sure I buy the whole “bubble” thing.

    The field I know is C.S. I looked at the App Academy curriculum and I wouldn’t want a co-worker for whom that was the total of their knowledge. Someone who’d spent years studying interesting software-related stuff independently and then AAed to fill the gaps, sure. But someone who knows MySQL and Ruby only in terms of how to use them and has no idea what’s going on inside or why? No. You can’t squeeze a Bachelors’ of CS into three months. And while there are probably jobs for which a shallow knowledge is fine, they’re not ones I’ve had contact with.

    The pure-signalling theory for other fields sounds more plausible, but that’s disturbingly Gell-Mannish.

    Also the classic concept of a “bubble” (such as tulips) is that people buy something at a high price specifically so that they can resell it at a higher price to someone basically like them. This is not happening in education. The signaling race is somewhat similar, but it’s not the same and we can’t safely apply cached thinking.

    Would prices drop if demand dropped? Maybe.

    As I understand it, tuition *still* won’t pay the expenses of getting a degree, except maybe “full” tuition at colleges where most students are on “financial aid”. Everything’s subsidized, one way or another. Tuition is going up because subsidies are going down.

    Standard economics predicts that a drop in demand will lead to a drop in prices because standard economics assumes a monotonically upward supply curve. That seems pretty unlikely here.

  59. Tom says:

    One hundred years ago, William James wrote an essay about fetishizing Ph.Ds (Google The Ph.D. Octopus).

    Once a certain piece of paper is shown to improve outcomes, more people want it. Several decades ago, it was the high school diploma. Then there came a push to get everyone through high school, because people with high school diplomas got better jobs. Well, getting everyone to meet the old standards was tough, so the standards got progressively lowered to the point that anyone could pass.

    I surmise that one reason American medical schools want a college degree is because they don’t trust American high schools. I’ve taught introductory calculus a few times at a big state university. The fail rate is appallingly high (20-30%) considering that most of these students were getting As and Bs in high school and many had passed high school calculus. I understand that other introductory science courses have high failure rates as well. Let me emphasize: these students did rather well at the high school level. I don’t know that the Irish system would be suitable without major changes to US K-12 education.

    I’d expect similar reasoning from employers. They don’t care what your degree is in, but getting one presumably means you’re functionally literate.

    Switching from tulips to daffodils is a great solution if you’re the Marriage Czar, but not if you’re the prospective groom.

    • LTP says:

      “The fail rate is appallingly high (20-30%) considering that most of these students were getting As and Bs in high school and many had passed high school calculus. I understand that other introductory science courses have high failure rates as well. Let me emphasize: these students did rather well at the high school level. ”

      I think this mostly has to do with the difference in the structure of a high school vs. college class, even if the high school class is at the AP level.

      In high school, you spend much more time in class and you meet up for class 4-5 times a week instead of 2-3 times. Additionally, a semester’s worth of content is stretched to be over two semesters of school. This means there’s more time for lecture, examples, Q&A, etc. There is also often time to do at least part of your homework assignments and essays/projects in class or in study halls. Homework is also, as a consequence of the slower pace, split into smaller and more manageable chunks.

      What I’m getting at is that college classes require much more self-discipline, self-direction, focus, punctuality (WRT homework) and motivation to succeed in relative to high school courses of even the same content due to the faster pace, less time in class, and less time to work with the professor.

      High school doesn’t teach people the skills needed to *learn* at the college level, even if they impart the content well. The high fail-rate could be mostly or entirely by these structural differences rather than an inability to teach the content.

      • “What I’m getting at is that college classes require much more self-discipline, self-direction, focus, punctuality (WRT homework) and motivation to succeed in relative to high school courses of even the same content due to the faster pace, less time in class, and less time to work with the professor.”

        As a college professor myself, I totally agree.

        • Julie K says:

          BTW, MOOCs would require even more of these traits than typical college classes. I suspect that the people saying “let everyone learn in MOOCs” happen to be the type of people who would do well in such classes, but they don’t realize that not everyone is like that.

          • Agree to some extent, although not the part about having to show up on time. Perhaps some future MOOC should require students to take tests at 8:00 am.

      • Tom says:

        My point was that performance in high school is only loosely correlated with performance in college science courses. The students who fail those in college can switch major to Underwater Basket Weaving and still get their degree.

        On the other hand, if one were to accept students directly into medical school, we’d still expect a high fail rate in basic science for the same reasons. What would become of those students? A school that kicks out ~30% of the incoming class each year and still charges medical school tuition would meet with a lot of backlash.

        • LTP says:

          Sure, but you implied the reason was that American high schools don’t impart knowledge well to students, while I’m saying they (high schoolers who do well enough to get into college in the first place) often do have the knowledge, they are literate, but they don’t have the study habits, at least initially.

          • Lupis42 says:

            I can confirm that – I had a lot of difficulty with college at first, because my study habits were attuned to a wholly different type of pedagogy/work/testing/grading model than I was now facing.

          • Tom says:

            I implied that high schools dumbed down the curriculum. They CAN impart knowledge well, but since much of it is not required to graduate a lot of students don’t learn it.

            My own observations suggest that lack of knowledge is a big problem (in math at least). Many students are uncomfortable with sines or logarithms and have difficulty manipulating fractions. I get the distinct impression that they learned a “cookbook” approach to math in lieu of reasoning skills.

  60. Deiseach says:

    After listening to this, I have to ask – Scott, did you pick up any trace of a Cork accent after your four years in the People’s Republic? 😀

    • Scott Alexander says:

      No. I have a mysterious accent that nobody can place. I think it’s probably from speaking very slowly and exaggerating my enunciation during my time as an English teacher.

  61. Anonymoose says:

    The Walden study you mention in part III is not horrible but does not deserve the inclusion in “rigorous well-controlled studies” that you give it. The study takes its data on student outcomes and on Master’s degree status and directly regresses that, attempting to use a series of controls to protect it from confounders. This is not an abnormal thing to do, but is also pretty bad practice when you’re studying education because 1) there are loads and loads of variables that you cannot easily measure that are may be important in determining both whether or not someone pursues a Master’s degree and how effective of a teacher someone is (a whole pile of personality measures, for instance) and more importantly 2) it’s typically very easy to pick solid instruments to use for an instrumental variable regression when it comes to education level, so there’s little reason not to, and doing this immediately wipes away the endogeneity problems I mention above. I’ve never had to look into instruments for Master’s degrees specifically, but typically proximity to nearest educational institution is a fantastic instrument, having a pretty strong effect and also being quite easy to measure, and rarely having anything to do with your error term.* If our only takeaway from this study is that the effect sizes are hella small (although note to those of you who didn’t read it, it did find a significant and positive result, just a small one) and so Master’s degrees are not very important for teachers, I think we should be okay–most confounders I can imagine would make Master’s degrees look more important than they really are (e.g. taste for working hard). I would definitely be wary trying to get more specific than this, though.

    *I would be a little more worried about this last bit when it comes to school with graduate programs in education than, say, high schools, as if there are few enough of these schools then proximity to one might covary notably with one of your confounders–but I don’t really know enough about these schools or about the layout of whatever area in Georgia this was conducted in to even speculate.

    **Brief copypasta explanation of instrumental variables regression for those who haven’t heard of it, because last time I posted about this people hadn’t: You go find some thing (your instrument) that determines part of your independent variable but has nothing to do with your confounders (think of things determined by nature for the best examples of these, generally, like amount of rainfall in a location). Then, regress your independent variable on this thing so that you can pick out only the variation that this instrument produces, and use the results of that in your main regression. It’s very nice because you essentially are letting nature (or whatever exogenous process it is that controls your instrument) run a randomized control trial for you, and as such you get all of the upsides of an RCT (yes, seriously!)

  62. Will says:

    You are missing the REAL problem here, which is scarcity of good work. Because competition for good jobs is incredibly fierce, what is happening is a signalling arms race.

    On average, college grads REALLY ARE better hires- they have a longer track record on which to judge them, and more data is always better than less data. In my state, it isn’t REQUIRED to have a college degree for most jobs but it’s defacto required because 3/4 of the applicants have the degree.

    What is happening is that good jobs are becoming harder to find, so more educated people are applying. This causes the previously less educated to either signal harder or get left behind.

    • onyomi says:

      Agreed. I described in

      A way to possibly make the hypothetical more reflective of the US job market: make the tulip-giving nation an 80% male populace.

    • BBA says:

      I’m a fan of David Graeber’s theory of “bullshit jobs”: there isn’t enough meaningful work to go around, so to avoid the stigma and hardship of unemployment more and more people end up doing meaningless work that adds nothing of value to society. Now I disagree with Graeber’s leftist assumption that this was all orchestrated by the evil upper classes to maintain their system of societal control – it seems to me more like a gradual, unintentional phenomenon – but clearly this is something that’s taking place. And if the work is meaningless, a meaningless signifier like a college degree is a perfectly fine criterion for it.

      Of course the universities themselves are full of bullshit jobs like deputy associate deans for student affairs, or whatever. And this leads me to speculate about how this all will turn out.

      The job market for lawyers collapsed in the aftermath of the 2008 bust, leaving a glut of ABA-approved law schools. (Unlike the AMA, the ABA wants to increase the supply of law students. Law firm partners see new lawyers as potential hires rather than potential competition.) To improve their post-graduation employment numbers many law schools have started to hire their own graduates to short-term “fellowships” – funded, of course, by the tuition of the next group of suckers students.

      Now lots of people see through this charade and law school enrollment numbers are plummeting. Meanwhile there are inklings of the same model spreading to the undergraduate level. Which, if it becomes widespread and the conventional wisdom of college = jobs goes unchallenged, leads to the ultimate cycle of bullshit – paying money to a university for a meaningless credential so the same university can pay you money to work at a meaningless job.

      • Drew says:

        My sticking point with that article is here:

        The ruling class has figured out that a happy and productive population with free time on their hands is a mortal danger (think of what started to happen when this even began to be approximated in the ‘60s).

        He’s describing the ‘ruling class’ like it’s some unified block that can come together and easily solve coordination & free rider problems.

        That would be ok if this were just a poetic license. But the whole premise is that major corporations are literally spending money to create jobs that they could easily — and profitably — drop.

        If only the country’s elite were that coordinated and altruistic.

    • Dude Man says:

      This isn’t a new problem, our standards have just gotten higher. The idea that works needs to be meaningful or fulfilling is a relatively new one. It used to be understood that most people would spend their working hours as cogs in the machine and you would get fulfillment outside of your job. The future everyone imagined was not one where everyone loved what they did, but one where working hours would plummet.

      Maybe we just need to accept that 70% of people will have to dread going into work Monday morning.

      • Jiro says:

        Unfortunately, employers like to demand “passion”. In other words, if you actually are willing to become a cog in the machine, you don’t get a job at all, which is even worse.

        And a lot of the demand for “passion” is like the demand for college degrees; it’s become a signalling race where the employers ask people for something and then everyone has to try to get it, leaving everyone worse off.

        Incidentally, this also affects applying to elite colleges in the first place, where in order to get admitted, you have to show you’re special. You can’t actually want to go to college because you need it to get work, even though for most people that’s the only reason that makes any sense. As a result, everyone has to become special, which leads to doing useless signalling things to pad your college admissions form (that incidentally make it even harder for poor people to get into an elite college).

        • onyomi says:

          This reminds me of my most dreaded questions in every job interview: “why do you want to work here?” and “do you have any questions for us?” The only honest answers are “because there’s a strong possibility no one else will hire me” and “are you going to offer me this job?” but obviously you can’t say that.

          You have to signal that you are really passionate about this particular job and that you are uniquely suited to the place in some vague sense of “fit” by the serious questions you ask. You also get bonus points for asking questions which make it seem like you probably have other options.

          • Not sure about ‘can’t’. It does seem like sometimes it’s fine to give honest answers.

            When I was somewhat desperately looking into a job offer in Hamburg so I could get out of Bonn (which is a lovely place, but my lungs hated it there), when they asked me why I wanted to work with them, I told them it was for health reasons (allergies I didn’t even know I had before I moved down to Bonn). I had no questions for them, either. And I do work there now.

            Granted, the whole atmosphere of the interview had the vibe of ‘we’re just going through our checklist of questions, you’re already hired’, which is not true for most scenarios. And I don’t even know how else to be but embarrassingly honest about these things, so if I do slip out of whatever ridiculous wave of good luck I’m riding, I’m probably in for a rough time.

          • onyomi says:

            Well, the pollution answered the question of why you were leaving your old job, but not the question of why you wanted to work at the new job. Ideally you would have a reason to especially want to work at the new job as well.

            Now, if you happen to be the most qualified candidate by a reasonable margin, failure to evince great passion for working at that particular place will not decisively hurt your chances. However, in my field, at least, there are usually many very qualified people applying for every job opening, many of them desperate for any tenure-track job and very likely to take it if offered. In such circumstances, committees look for anything to help distinguish candidates, and so vague and subjective qualities take on more importance.

            That said, I have also experienced that bluntness can signal confidence, which can be good.

        • Deiseach says:

          I have never believed that employers ask for “passion” because they want you to be fulfilled and thrilled by your work; it means “Are you willing to put the job first before personal life, family, everything? You will still come in early and stay late even if you’re not paid for it? You will come in even if your child is in hospital gravely ill or your house is burning down?”

          It’s a way of demanding commitment and loyalty, even if both parties know (a) as soon as it is economically more beneficial to lay you off than to keep you in the job, you’re gone and to hell with your passion for the job (b) if the person can get a better job – be that better paid or work they prefer or more work-life balance – they will leave.

          It’s another way of asking “How will you make more money for us?” You have to sell yourself and “I have a real passion for excellence/ever since I was three years old I have dreamed of working in the paperclip counting field” is the acceptable current buzzword for doing so.

  63. HeelBearCub says:

    The Higher Education Authority (HEA) is the statutory planning and development body for higher education and research in Ireland. The HEA has wide advisory powers throughout the whole of the third-level education sector. In addition it is the funding authority for the universities, institutes of technology and other designated higher education institutions.

    Scott, is the higher-education system in Ireland really where you want to hang your hat in this debate, given your position? It does not appear to be private at all and is completely controlled and funded by the federal government.

    Honestly, that seems like either a tremendous oversight, a deliberate self-deception, or a situation where I am completely missing the point you are trying to make.

  64. Pingback: 67 – Against Tulip Subsidies | Offer Your

  65. Douglas Knight says:

    Make “college degree” a protected characteristic, like race and religion and sexuality.

    That already is the law. But laws don’t enforce themselves.

    Are you married?

    You are not allowed to ask me that.

    I know.

    (Education isn’t quite that protected. It’s OK to ask about and use, but if that creates a racial disparity, that is illegal. This the explicit law, passed by Congress. But no one enforces it.)

    • Daniel Armak says:

      I really don’t understand how that’s supposed to work. You can ask, but if after asking it turns out the answers are correlated with race, that means asking was illegal in the first place? Or is acting on the answer illegal?

      Is the employer obligated to calculated the correlation with race to know if they can act on the knowledge? If they do calculate it and find out they’ve been hiring more people from group 1 because that group happens to have a higher rate of diplomas, are they now legally obligated to hire people from group 2, or people without diplomas, or people from group 2 without diplomas?

      • Douglas Knight says:

        Why do you think that there is a way things are supposed to work, let alone that it is related to the law?

        • Daniel Armak says:

          I’m trying to understand your sentence: “It’s OK to ask about and use, but if that creates a racial disparity, that is illegal”. Can you explain?

          • meyerkev248 says:

            Exactly what it says on the tin.

            You’re allowed to have a test. But you’d better make sure that:

            a) All races/genders/orientations/etc score equally.
            b) You can prove that every single question on the thing is a job requirement and you have a lot of money for lawyers.

            /See the fate of the US Civil Service exam, for example.

  66. Pingback: 68 – Against Tulip Subsidies | Exploding Ads

  67. Xplo says:

    You do realize you’re pushing rope here, don’t you?

    The basic problem is that employers don’t want to hire people without college degrees, so in order to get a job, you have to have a college degree (and pay whatever that costs). I think, and it seems you would agree, that employers place too much value in college degrees.


    I think your “make college degree a protected status” notion is farcially naive. Maybe, MAYBE that would work for the lowest-end jobs. But many jobs out there require, or at least benefit from, some relevant (I said relevant!) training. If I’m going to hire an engineer, I damn well want to make sure that he has an engineering degree. Are you going to maintain a great master list of employers who are allowed to ask for degrees? And college has replaced high school as the basic proof that its product is at least minimally able to read, write, think, show up on time, and not wear their pants on their head. (Which may merely indicate that many employers no longer remember college.) No matter what you do about the “some jobs need college” problem, employers who still want that reassurance that the candidate they’re interviewing isn’t a pantshead will find ways to get around it or encourage voluntary disclosure to an extent that makes the ban worthless. Like this:

    “So, what are your qualifications for this position at Impressive, Inc?”
    “I see from your employment history that you were minimally employed during a characteristic 4-year-period. What were you doing at that time?”

    Then there’s the other half of the problem: to actually realize the cost savings from reduced education, you have to stop kids from going to college. And of course by reducing attendance you’re reducing class choice and flexibility because colleges have to drop their less popular classes to save money, etc, etc. – but never mind those because they were obviously artifacts of inefficient spending anyway. (No one needs a whole class on the cultural influence of Asian-American beatnik poets.) The real question is if we still have a society where employers want college graduates and people stop going to college, HOW WILL THOSE PEOPLE GET JOBS?

    Yes, the reduction in available graduates will have some effect on employer expectations, eventually. But in the meantime you have people who are unemployed, careening uncontrollably toward extreme poverty and homelessness, and still strongly incentivised to get a college education by any means possible (because those who have them will escape this trap). Will you be providing these heroic non-degreed folks with free housing, food, utilities, clothing, retirement funds, and spending money so they can actually go out and participate in society as though they were really functional members of the economy instead of disenfranchised outsiders?

    • onyomi says:

      I think it would be interesting if, when applying for an academic job requiring a PhD, for example, people could write down the fact *that* they held a PhD in the relevant subject matter, but not which institution granted it. As a practical matter, most people could figure out where you got your PhD by looking at who writes your letters of recommendation, though it’s not uncommon to have one or more outside recommenders.

      I think this would have a positive impact on the academic job market, though it would negatively impact me personally, and would be viewed as a horrible inconvenience by search committee members. The beauty, however, is that the search committee members couldn’t complain too loudly because they don’t generally admit that, when winnowing applications, they look first at the letterhead and often reject applicants from less prestigious institutions out of hand. It would be funny for them to have to actually attempt to evaluate the academic value of the writing sample, etc. and to see how “Emperor’s New Clothes” everyone got in being afraid they’d love the candidate from PodunkU while rejecting the candidate from Ivy League.

      Not that this will actually happen, and the real problem in the academic job market, as in the job market more generally, is just too many qualified applicants for available positions incentivizing arbitrary ways to winnow the pool, but it’s fun (for me, being in academia) to imagine, at least.

  68. Albatross says:

    Let me split this baby in half:

    Free two year degrees at affordable colleges. Let everyone learn brewing or accounting or nursing or truck driving. My state pays for this already if you complete it before graduating high school.

    Second, end all education subsidies beyond that for everyone who is not poor.

    Free two year. Free market four year.

    My wife has her Masters and has taught college for ten years. She would like to switch to K-12, but you can’t take the teaching license tests in my state without an undergrad teaching program. Her double major undergrad with honors and her Masters in teaching English as a second language and her experience mean nothing. She must complete a third undergrad. But if she does, her pay scale will start higher, because she already has her Masters. Sigh.

    • Paul the Apostate says:

      How would you feel as a K-12 teacher knowing that non-unionized, possibly very bright people from outside your discipline could earn your license and snap up jobs that should go to your own kind? You’d want them to jump thru an unnecessary, time-and-money-consuming hoop first. And you’d especially want to let those with high achievements in their prior discipline know how little you value them.

      • Albatross says:

        I switched industries after a year of unemployment, so I am of course delighted my field lacks protectionism. My wife is forced to jump through the hoops, and as commenters below have mentioned, she will be discriminated against because she would automatically earn higher pay. And I have to pay taxes to fund this system which is both unfair to my wife and obviously intends less qualified people to teach my kids.

        Lots of teachers in my family, more than a dozen. And I know the history of unions in the US. Schools used to put restrictions on teachers dating or getting married.

        However, having been unemployed whenever I hear about requirements that hair dressers complete 1,000 hours of training or food trucks get restaurant licenses in every town they visit I immediately think of all of the poor people who are never going to be a hair dresser, food truck cook, teacher, or doctor because they are too poor.

        How many more poor people could become doctors or teachers if we didn’t tack a few more years of education on to it?

        These hoops are preventing income mobility and wasting billions of dollars a year.

    • Steve Sailer says:

      Accounting is a famously practical subject to study in college. The accounting profession has responded by getting many states to mandate that you can’t take the CPA exam unless you’ve taken not four but five years of higher education.

      • Deiseach says:

        Interesting – I’ve just looked at the CPA Ireland page, and there are progression routes other than “you have to go to university”.

        My old workplace runs a 2-year Accounting Technician certification course (and this is not easy, it is tough) and once you have that qualification, you can go on to do more training/education for the CPA exams (and you’re exempted from some of them because of your existing qualification). A lot of people get the IATI qualification, get a job, and do part-time study (often funded by employers) to go on to become a CPA.

    • Anonymous says:

      I’ve heard that a master’s is a strike against hiring teachers, especially English teachers, because of that higher starting salary.

      • Randy M says:

        Arrrgh! What is the purpose of paying more for it if you don’t want the increase in performance it putatively brings?

        • bartlebyshop says:

          You’re assuming a unity of purpose here that doesn’t exist. Group A decides to pay teachers with a master’s more, because they know more about the subject and will teach it better. Many years later, Group B has a pool of money to hire with and can hire more teachers if they only hire people without master’s degrees. Since student:teacher ratios predict performance, it’s easy to argue that they’re actually making the best decision!

          • Randy M says:

            I was more or less just expressing frustration and realize it is a coordination problem, but is your group B there a successor to A, or are different groups involved in setting the starting salaries of differing levels of qualifications versus deciding how best to allocate the budget?

          • bartlebyshop says:

            It probably depends on your location, but I think it’s a bit of both. Local to me I think the starting salaries are set by the school board/district and then each school is responsible for hiring with their fixed blob of funds, so if they hire a teacher with an MFA to teach art they may not have money to pay a teacher to supervise the after school underpriviledged students express themselves through painting club.

          • Glen Raphael says:

            Since student:teacher ratios predict performance, it’s easy to argue that they’re actually making the best decision!

            IIRC it’s not true in general that student:teacher ratios predict performance. What is true is that EXTREMELY SMALL classes (less than a dozen students) enable different kinds of teaching techniques and improve performance. But once the class size is much more than about a dozen, there’s no correlation between class size and student learning. Tweaking the ratio still does reduce teacher workload (so the teachers are *happier*) and increases the number of teaching slots (so the unions like it) but the positive impact on student achievement of cutting class sizes from 40 to 35 or 30 or even 25 is not measurably significant.

          • bartlebyshop says:

            Sorry, you’re right. I should have said “since we very strongly believe that the ratio increases performance…”

    • Deiseach says:

      Yes, but primary school teaching and secondary/third level teaching are different specialities (unless American schools are run differently, I have no idea).

      Your wife has been teaching one subject in depth. Does she think she could cover a broad curriculum of everything from English to Music to Nature Study to the other subjects for a class of six or nine or twelve year olds? Granted, trying to hold the attention of a bunch of bored eighteen year olds may be like minding a class full of six year olds, but some kind of general proficiency “Are you up to scratch in teaching Geography and finger painting and playing the recorder and English grammar” course is no harm.

      Assuming you can just walk in the door of the classroom and immediately switch from college level to primary level may not be so!

  69. William A. says:

    Taranto in the WSJ in 2007 wrote:

    > What most professional jobs require is basic intellectual aptitude. And what has changed since the 1970s is that the court has developed a body of law that prevents employers from directly screening for such aptitude. The landmark case was Griggs v. Duke Power Co. (1971). A black coal miner claimed discrimination because his employer required a high-school diploma and an intelligence test as prerequisites for promotion to a more skilled position. The court ruled 8-0 in the miner’s favor. “Good intent or absence of discriminatory intent does not redeem employment procedures or testing mechanisms that operate as ‘built-in headwinds’ for minority groups,” Chief Justice Warren Burger wrote.

    > This became known as the “disparate impact” test, and it applies only in employment law. Colleges and universities remain free to use aptitude tests, and elite institutions in particular lean heavily on exams such as the SAT in deciding whom to admit. For a prospective employee, obtaining a college degree is a very expensive way of showing that he has, in effect, passed an IQ test.

    > But why are employers able to get away with requiring a degree without running afoul of Griggs? Because colleges and universities — again, especially elite ones — go out of their way to discriminate in favor of minorities. By admitting blacks and Hispanics with much lower SAT scores than their white and Asian classmates, purportedly in order to promote “diversity,” these institutions launder the exam of its disparity.

    > Thus the higher-education industry and corporate employers have formed a symbiotic relationship in which the former profits by acting as the latter’s gatekeeper and shield against civil-rights lawsuits. Little wonder that in 2003, when the Supreme Court considered the constitutionality of discriminatory admissions policies at the University of Michigan, 65 Fortune 500 companies filed a friend-of-the-court brief urging that they be upheld.

    With degrees being subsidized, and the bars for entry, grades and graduation (i.e. net accrued skills) lowered, the signalling value of a degree is diminished. Still, if employers continue to be constrained by Griggs v. Duke Power Co., how does an employer screen candidates?

    • Matt M says:

      Wasn’t there a recent case where a police department was allowed to continue using an intelligence test to filter out candidates… who scored too HIGH?

      • houseboatonstyx says:

        I’ve seen more than one credible account of this. Somewhere in or near NY state, and the applicant sued but the judge ruled that the department had the right to do this.

  70. Eli says:

    This is why, despite my reservations about libertarianism, it’s not-libertarianism that really scares me.

    Sorry, Scott, but I have to call you out on this one. You’re not wrong: you’re not even wrong. This question (higher-ed prices in the US versus elsewhere) simply is not about a “libertarian” option versus a “not-libertarian” option. You have failed to cut the socioeconomic and political reality at the joints.

    Really, the following systems are more like what reality puts on our table:

    1) “Full” libertarianism, pure as the white snow: all universities are privately chartered, and state subsidies don’t exist, not on the demand end nor on the supply end, nor even in the form of research grants. Student loans exist, but are entirely private, unsubsidized, and dischargeable in bankruptcy (along with all other loans). Market dynamics are allowed to operate unfettered. It’s worth noting that the custom of making American doctors obtain a 4-year undergraduate degree before medical school came from this system: private medical schools invented that requirement as their own privately chosen admissions criterion.

    1A) The fully private, libertarian system, plus the availability of public research grants. This is roughly what libertarians are actually proposing right now, when they don’t go full batshit crazy and declare that “taxation is theft” or “patent [trolling] can fund science research just fine”, or other ridiculous nonsense.

    2) A truly mixed system: some universities are private and some are state-owned. The state-owned universities are subsidized on the supply end: they receive a sufficiently high percentage of their per-student educational expenses in consistent, year-to-year state funding that they can charge a low sticker-price tuition to everyone. Scholarships exist at both private and public institutions, but loans are entirely private, and financial aid of any form is uncommon — it may even be reserved for a small subset of applicants deemed too poor or too brilliant for the sticker price. The supermajority of professors are tenured or tenure-track, and something like 2/3 or 75% of teaching is done by permanent, tenured-or-tenure-track professors, with very little foisted onto adjuncts and grad-students. This is roughly what the United States used to have in its “glory days”, and its legacy continues to structure how Americans expect higher education to work, even when the reality is nothing like that anymore.

    3) A Malignantly Mixed system: state block-grants have been cut to the bare bone, and to make up for it, huge amounts of personal grants and publicly-subsidized student loans have been added to the demand end of the transaction. Despite all this money flowing around, 75% of professors are adjuncts, with no job security, no research support, and too high a workload necessary to achieve survivable wages to put in high-quality teaching — in fact, many academics are on Food Stamps. Higher-ed prices skyrocket, quality plummets due to the deskilling of university-level teaching, and a debt bubble blows up, up, and away. This is what the United States actually has right now.

    4) The Slightly-Less Malignantly Mixed system: the weird dystopia libertarians think leftists are proposing, as represented in your story. The debt-bubble has been moved from students to states by eliminating student loans in favor of tax-funded demand-side subsidies. That did nothing to stop the bubble, since it didn’t turn off the ever-growing supply of easy money to university administrations, but instead promised to fund it even more directly. The states are now on the hook for a massive, ever-growing obligation.

    5) The Benevolently Mixed State-Ownership System: separate, lower academic statuses are given to private colleges over public universities. Subject to a few scant licensing rules, private colleges are allowed to set themselves up, teach, and charge arbitrary prices as they please. State-owned, state-run universities have their tuition and expenses deliberately kept low-enough that students can work summers or part-time to pay for their education. The universities are also solely empowered to grant doctorates, and sometimes master’s degrees, which helps ensure that all the really top-quality work goes on at the publicly owned institutions. Price and quality competition from the universities thus helps keep the private colleges in-line. Quality of students is kept high, and their numbers low enough to fund, through noxiously difficult entrance examinations. This is what I lived with at the Technion where I did my MSc — and it seems to work decently-enough, once you account for the Technion being pathologically, oppressively hard on its undergrads.

    6) The Fully State-Owned System: all higher-ed institutions are nationalized, and funded with block-grants so that tuition and fees can be low or nothing. Students may even receive a small monthly grant for living expenses, if you’re going Full Scandinavian. This is what many American leftists are actually proposing.

    I think that (2), (5), and (6) all work fine, and you can easily go with (2) or (5) if you’re worried about the expense to the state of funding (6).

    • The debt-bubble has been moved from students to states by eliminating student loans in favor of tax-funded demand-side subsidies. That did nothing to stop the bubble, since it didn’t turn off the ever-growing supply of easy money to university administrations, but instead promised to fund it even more directly. The states are now on the hook for a massive, ever-growing obligation. (emphasis added)

      But when the state directly funds something, then it becomes subject to budget constraints and trade-offs.

      Universities have got things figured out to keep the money faucets wide-open in their direction. Nobody but the beneficiaries would have done that intentionally, rather, it happened as the accidental byproduct of other policies.

      But just as nationalized health care or a single-payer system is envisioned to entail stringent cost controls, nationalized universities would no longer be free to spend unlimited money on anything they wanted.

      I have seen this happen on a small scale. Direct funding gets more scrutiny.

      • Eli says:

        But when the state directly funds something, then it becomes subject to budget constraints and trade-offs.

        Yes. I consider that a feature, not a bug. Universities are not supposed to have luxury-quality dining halls or gymnasiums or spas. They’re also not supposed to have mediocre-quality gymnasiums pretending to be luxury ones.

        In fact, I’d really much rather get universities out of the business of feeding and gyming people entirely. I can see the plain argument for their housing people, so that their admissions policies don’t impose a negative externality on surrounding communities, but I think they should build apartment-style housing rather than the current form of dormitory that then requires separate feeding and exercise facilities. Technion did this, and functions just fine.

        Actually, Technion has a gym, but they have no incentive to go wild with it, because they can only admit the number of students the State allows at the tuition level the State allows, and aren’t mandated to make any profits by drawing exercisers from outside Technion to use their/our gym.

        (Although actually, doing these things collectively when you expect to have a mass of 20,000 people with roughly the same needs may well be more efficient, in which case the issue is to make sure that state funding isn’t allocated with perverse incentives, like dumping huge sums of money into the construction budget while cutting the teaching budget. Massachusetts did this when I was in undergrad, and still does, but more so.)

        Since we’ve seen, from experiment that 2 decays into 3, I think claiming it works is a little strong.

        Lupis, this is one of those cases where you have to admit malice – or rather, deliberate action directed by a combination of incentive and ideology – into your hypothesis space. America’s state university system didn’t decay, it was neglected and destroyed.

        Neglected in the sense that once the Baby Boomers aged up, the birth rate for Gen X declined, and the 30s-50s set thus became the unstoppably dominant voting demographic, there was a radical shift in USA federal-and-state public spending towards services for grown-ups and the elderly rather than services for the young. Taxes were cut, affordable housing programs were cut, state universities were cut, and pensions were overpromised. Policy got made on the presumption that society already had enough trained-and-employed whatever-the-hells, and certainly didn’t need to invest state dollars producing more.

        There was also a matter of right-wing ideology. Basically, dirty hippies pissed off the voting public, especially the blue-collar voting public (both whites and people-of-color) who saw a large class gap between themselves and the hippies. Dirty hippies were known to come from universities, universities which had also granted the privileged exemption from the Vietnam War’s draft.

        As a result, Nixon, Reagan, and the Southern Strategy-using Republican Party have spent the entire subsequent era of politics demonizing universities, and especially public universities. Private university was expensive, they thought, so it had a good incentive for the student to shut the fuck up and behave themselves. Public university was so cheap, and so damnably democratic in its ideals, that those damn dirty hippie students thought they could get away with anything!

        If you think about it for five minutes on the clock, this is actually very counterintuitive behavior. Consider those older days of the Republican Party, especially pre-Reagan, and especially in the Eisenhower and Goldwater days: the party of the advantaged and well-heeled, yes, but also thusly the party of exactly the sort of cultured, educated, professional person that one becomes by successfully going through university! Insofar as the university-educated, and especially postgraduate-educated, make more money, they ought, intuitively, to have a dual class interest in voting Republican! Hell, there was a time when, IIRC, about 50% of professional scientists voted Republican! Until the “dirty hippie effect” convinced the entire American right-wing that publicly-sponsored university education in specific and university education in general were factories that took swing voters in and pumped out devoted liberals (which isn’t really even true!), there was every reason for the Republican Party to appeal to the educated and expect to find the educated receptive to voting Republican!

        (This is an instance of a repeating historical pattern, in which “extreme” behavior on the part of a tiny portion of liberal (in the classical sense) or left-wing activists causes a backlash many times larger and longer-lasting than the actual behavior! For example, the German upper classes originally lent support to the Nazi Party as a bulwark against Communist insurrection in the Weimar Republic… and today who even remembers the German Communists of the 1920s, except as footnotes to the rise of the Nazi “backlash”?)

        Also, I will happily come out and say that I think neither 5 nor 6 ‘works’, and prefer 1B – like 1A, but where we don’t fund research in advance, but rather with prizes for replicable results.

        I will gladly leave you with a burden of proof for showing how countries like Japan, Israel, and almost the entire European Union are all failures, in terms of resource-relative research and teaching accomplishments, relative to a hypothetical system that has only ever existed in the few years of the 19th century, in America, between the establishment of colleges and the establishment of land-grant colleges.

        • Lupis42 says:

          Lupis, this is one of those cases where you have to admit malice – or rather, deliberate action directed by a combination of incentive and ideology – into your hypothesis space. America’s state university system didn’t decay, it was neglected and destroyed.

          I buy neglect, although I think you might overestimate it’s significance. The malice part, though, seems to me like it wouldn’t pass an ideological turing test.
          Although you mention it, I think you’re giving short shrift to the nature of incentives for the people running a college or university. First, tenured professors are really really expensive per unit of teaching, and lots of people want to get that kind of a gig, and are willing to work for peanuts and a shot at it. So naturally, you start shifting your labor pool from “expensive” to “cheap and begging for the job” – they’re fully qualified, and room in the budget allows you a lot more leeway as a political tool. Once you do that, it becomes apparant that tuition fees can actually be a source of income, at least up to the point of your existing physical capacity.

          Fundamentally, bureaucracies behave like a sort of highly adapted cancer – they keep fulfilling their nominal function, but they grow to consume as many resources as they can, and the useful functions get none of those resources, and instead get squeezed. What keeps them in check is threat of the organization becoming too disfunctional, and either losing it’s resources or cutting the bureaucracy back. That pressure is still nowhere on the horizon for public universities in the US.

          I will gladly leave you with a burden of proof for showing how countries like Japan, Israel, and almost the entire European Union are all failures, in terms of resource-relative research and teaching accomplishments, relative to a hypothetical system that has only ever existed in the few years of the 19th century, in America, between the establishment of colleges and the establishment of land-grant colleges.

          In order to do this any more justice, we’re going to have to agree on a metric to judge by, where I expect to hit some irreducible differences. To a first order approximation, I’d look at useful* ideas generated/percent of populations time/resources poured in.
          *Useful here is a very squishy word, which I’m using because it would take a ton of time and space to elaborate further.

          The system in use during the heyday of the “Great Enrichment” beats the system in use during the “Great Stagnation”?
          One problem that occurs to me with #5 is that, if you’re thinking in terms of research and teaching achievements, the top thirty or so universities in the US are private universities and military academies. There are ~four public schools in the top fifty. Depending on who makes the list, around seven of the top ten in the world are American Private universities. A model that says those need to stop doing research seems pretty radical, and I’d be interested to hear your criteria for ranking it.

        • 27chaos says:

          Why would a group of college students living somewhere be a negative externality?

          • Deiseach says:

            You seriously have to ask that? For a start: drunkenness, late night parties, noise, not bothered in upkeep of the property (because feck it, it’s not theirs, they’re only staying there for nine months of the year, and what 18-22 year old is voluntarily going to spend free time painting or gardening when they’re not being paid for it?)

            Traditionally, students stayed in “digs” (that is, they were lodgers in a family home). That, or they got cheap bedsits. It’s more usual now for people to rent a house and house-share. See above re: anti-social behaviour. Pubs in our county capital (where the institute of technology is located and which has a large student population during the academic year) got hammered for running cheap drink promotions specially to appeal to students.

            Though I keep forgetting in America you’re probably not legal to drink until you’re at least 21 🙂

    • Lupis42 says:

      Since we’ve seen, from experiment that 2 decays into 3, I think claiming it works is a little strong.

      Also, I will happily come out and say that I think neither 5 nor 6 ‘works’, and prefer 1B – like 1A, but where we don’t fund research in advance, but rather with prizes for replicable results. Preferably, prizes would also be available for negative results.

      There’s a lot of weeds we could get into on this topic if I can find the time, but I think most people overvalue education’s public benefits way too much, and probably even it’s private benefits.

      • gbdub says:

        Not only does 2 decay into 3 (because that’s exactly what happened), but the universities brought it on themselves because, while they had 2, they demanded self-governance for the sake of “academic freedom”. Now you can argue one way or the other on the true value of that, but the fact remains that public universities are less and less “public” at least in part by choice. State subsidies per student have gone down, yes, but that’s been more than matched by skyrocketing per student gross expenditures created by the policies of the self-governed universities.

    • Anonymous says:

      I really don’t see how you’ve gotten an implication for, “Despite all this money flowing around, 75% of professors are adjuncts, with no job security, no research support, and too high a workload necessary to achieve survivable wages to put in high-quality teaching”.

      Your phrasing even implies that this is counter-intuitive, so it’s something that really ought to be explained if it follows necessarily from the funding mechanism for universities. I’d also argue that your claim, “This is what the United States actually has right now,” is incredibly field-dependent. There are many areas where tenure-track positions are still the vast majority, and they are well-compensated and highly-coveted.

      I haven’t yet seen anyone actually explain why this phenomenon occurs, and I don’t think you have, either.

    • ” It’s worth noting that the custom of making American doctors obtain a 4-year undergraduate degree before medical school came from this system: private medical schools invented that requirement as their own privately chosen admissions criterion.”

      Before or after the institution of medical licensing enforced by the states and effectively controlled by the AMA? That’s a non-libertarian element of the system which creates an incentive for the medical profession to create entry barriers in order to hold up the incomes of those already licensed. As you may be aware, during the Depression the AMA sent a letter to the medical schools telling them that they were graduating too many doctors, and the medical schools cut back—prudent, given that the states got their list of approved medical schools from the AMA.

  71. Eli says:

    My ex-girlfriend majored in Gender Studies, but it turned out all of the high-paying gender factories had relocated to China. They solved this problem by going to App Academy, a three month long, $15,000 course that taught programming. App Academy graduates compete for the same jobs as people who have taken computer science in college, a four year long, $200,000 undertaking.

    I would like to see App Academy graduates write a System F typechecking algorithm (ie: for Haskell), code a microkernel from scratch, or design a superintelligent AI.

    I’m joking about the last one, of course. Proving, or even explaining, the basic theorems behind Support Vector Machines, or implementing a probabilistic programming inference method on their own, would be just fine.

    But, in this list of tasks, I’ve given one thing that I did in high school, on my own, due to my being unusually geeked-out on the subject matter, and which required me to buy and read a standard uni-level textbook on the subject. It’s the microkernel.

    The type-checking algorithm I could more-or-less do by senior year of undergrad, precisely because I was made to go through theoretical coursework in my undergrad curriculum, and really, you need at least a double-major in maths or beginning grad-level coursework in CS (thorough groundings in type-theory and category-theory) to design such an algorithm and get it right. I know, because I kept failing at it until I decided to buckle down and learn the appropriate academic material.

    The machine-learning and statistics tasks require something much more like a double-major math+CS degree, with the math focusing specifically on probability, statistics, and learning theory. The superintelligent AI thing is an open research problem :-p.

    I’m currently employed doing embedded firmware development, which mostly uses architecture, embedded, and systems knowledge that I learned on my own or got from one undercooked undergrad-make-up course at Technion. In fact, I occasionally find something I don’t know and ought to at work, because I’m relying on self-training.

    Now, I do think you could give people a much sparser dependency tree of subjects to take in order to obtain a “complete” education for whatever they want to do, including science or engineering subjects. This has its own trade-off, though: you’ll need to make each course pack in either more instruction-hours or more material in the same instruction-hours in order to provide a complete education in fewer total courses. In my experience as a TA, more material in the same hours doesn’t really work, even when dealing with the little geniuses who get into Technion. You need to make the dependency tree sparser by just adding instruction hours to the core courses that will allow professors to cover more material with more in-depth instruction. There’s no free-lunch for a better education system: once you’ve cut away the “fat” of unneeded, unwanted coursework that makes people hate school, you still have a shit-ton of work to do training them for whatever task they intend to take on!

  72. Doug S. says:

    My father is a professor of electrical engineering at City University of New York. I tried to get him to read the blog post. His first response was “this is too long to read at the moment” so I summarized, leaving out the tulip bubble metaphor.

    He responded with two critiques, which I’ve paraphrased and elaborated upon:

    As a professor, I’ve taught incoming freshmen and soon-to-graduate seniors, and the difference in general capability is extreme. As part of my Engineering 101 class, I’ve had students build a simple electrical buzzer, and then write a one or two paragraph explanation of how it worked. The incoming freshmen are invariably horrible writers that can’t put together a coherent explanation of anything – it was clear that they did indeed know how it worked, but nobody who didn’t already know how the buzzer worked could possibly understand what they wrote. One time, there was a student that stood out from the rest because she could actually write, and it turned out that she had already earned an English degree. Something happens during those four years people spend getting a humanities degree. An office manager who can’t write a coherent memo shouldn’t be an office manager, and the firefighter with just a high school diploma might be able to fight fires but he won’t be able to do the administrative work of running a fire department.

    Colleges are not vocational schools. The purpose of college is to educate people to be citizens and scholars, not to be tradespeople; to understand the world, not to be able to do a specific job. You don’t need to know about the laws of thermodynamics in order to be a plumber or programmer, but if you know about them, when someone tries to scam you by selling you a device that claims to improve your car’s fuel economy by using your car battery to extract hydrogen from water, you know it can’t possibly work. We want people going to college for the same reason that we want people to go to elementary school and high school – so they learn to understand the world they live in and can make reasonable decisions. Do you really want court cases involving complicated “scientific” evidence being decided by juries that consist of people who never went to college?

    There do exist institutions that are supposed to train high school graduates for specific vocations, and they’re called technical schools or junior colleges. (And, yes, App Academy is one of them.) They make you good at a single, narrow, marketable skill, but don’t attempt to teach you things beyond that. Universities are not technical schools and should not be expected to become technical schools.

    • Psmith says:

      Mm but how much of that is just due to seniors being older on average than freshmen?

    • Eli says:

      The second critique gets to the heart of the matter: what people really want is a vocational school, and they currently have no option to attend one, so universities end up substituting, badly, at a heavy deadweight loss.

      • Randy M says:

        College is sold to prospective students as an income boost; “Life time earnings is X million more with a degree than without one!” so if people think if them as primarily an vehicle for economic advancement the blame rests first upon the educational guild itself.

        Also, I think personal advancement and broadening one’s skills is great, but no good can be considered outside the costs (opportunity and direct). A lot of what we expect of college graduates in terms of general education and the non-vocatioal skills and knowledge used to be obtained in high school, or at least that’s my impression. I suspect a lot of the diminishing in incoming college student skills and knowledge is due to more enrollment and graduation from high school in the past few generations, due to changing native intelligence of the students (from compulsory education and changing demographics), immigrants being added to the pool who need a few years learning to be fluent in the language, perhaps (certainly in CA), various k-12 educational experiments or fads, and so forth.

      • Anonymous says:

        they currently have no option to attend one

        I don’t believe you. I’ll admit that there are industries that are new enough that vocational schools don’t yet exist for them, but that’s why places like AppAcademy are being created.

        Let’s ask the question this way: what job would you like to have that doesn’t currently have a vocational school option available?

    • Deiseach says:

      Do you really want court cases involving complicated “scientific” evidence being decided by juries that consist of people who never went to college?

      Given that, unless the particular college course was mathematical and you are proposing we have only juries composed of statisticians (that is what “complicated scientific evidence” boils down to: prosecution claims probability of match between DNA sample and someone else other than the defendant is in the zillions, defence rebuts this with more statistics), hell yes I do!

      I’ve been called for jury service twice (though never selected). I never went to university. So I am probably too stupid to give an opinion on “Did Micky Murphy rob the local Spar when he was off his face on prescription tranquilisers”? Thank you for that, Doug’s da!

      Also, he contradicts himself re: scientific ability and the student who was able to write a comprehensible piece; she wasn’t qualified in a science subject but in a humanities subject. The others understood the principles but couldn’t write well (and yes, I believe it, I’ve had to translate Techie to English before). That has nothing to do with being able to parse “complicated scientific evidence”; his English scholar might or might not be just as capable as a juror as Bob who left school at 15 and now owns his own chain of garages.

    • 27chaos says:

      The idea that people who don’t go to college aren’t capable of being functional citizens is pretty elitist. I think most high school graduates today would be amazingly well educated for laypersons in the eyes of democracy’s founders, and I think democracy ought to work about as well today as it did back then.

      Even if I were an elitist who thought most people weren’t fit to rule themselves without special training, the best answer to that situation would seem to be to improve or possibly extend the duration of high school, not to create separate expensive institutions that don’t educate the poor.

      Vocational training should be decoupled from mandatory citizenship training, not intertwined with it. Mixing the two different things together causes unnecessary compromises to be made.

      • Kevin C. says:

        “I think most high school graduates today would be amazingly well educated for laypersons in the eyes of democracy’s founders…”

        I’m not sure what Solon or Cleisthenes would think of our educational system.

        • gbdub says:

          Solon and Cleisthenes were part of an hereditary elite of aristocrats who had sufficient wealth to support sitting around thinking deeply all day. Today’s public school grad may not be their match, but would be an intellectual god to the average Athenian.

    • gbdub says:

      Honestly the whole “we don’t want to be JUST a vocational school” thing stinks of classist elitism and I don’t know why more profs and educators don’t get called on it.

      Some people couldn’t give a damn about the various self-actualizing “good of mankind” stuff that humanities professors peddle, and they certainly can’t afford to spend $100k on it. They just want the skills to get out of dead end jobs, and the humanities are a luxury good (at luxury prices).

      I don’t want humanities departments to go away, but this idea that a vocationally focused college is somehow “lesser” because it focuses on practical skills is irritating, especially when its locking people into lifetime debt to give them something they don’t want and are ill-suited for.

      To quote the great Judge Smails, the world needs ditchdiggers too, and too many college profs think training ditchdiggers is beneath them.

      • Matt M says:

        Not only that, they tell all of their students that ditchdigging is beneath THEM (the students) as well. So the students graduate with humanities degrees, and there is only one humanities-related job opening for every ten graduates, and they refuse to dig ditches because they spent four years being taught that ditch digging was for simpletons and they’re too brilliant for it.

        This is how occupy wall street happens.

  73. a.morphous says:

    I suppose the position that an undergraduate education has value in itself, and college is not supposed to be a god damned trade school, is just laughably naive, or it would have been floated in an earlier comment.

    Seriously, for a bunch of extremely educated people, the idea that education might be an end in itself and not primarily an intake channel for something else doesn’t really seem to have crossed your minds.

    • Susebron says:

      Is it worth $200,000, though? And, if it’s an end in itself, is there any reason that jobs should require it?

    • Anonymous says:

      Actually, for me it actually *was* an end in itself, and I resent the people who can never acknowledge that. (My parents both have elementary school educations, which was all that was required where they are from.)

      I’d bring it up here, but every time I have done so in the past, it was quickly dismissed. So, if you don’t see educated people bringing up the obvious, there may be a reason . . .

    • whateverfor says:

      Yes, it is naive. I went at least partially for that value in itself, as I’m sure you did, but that can’t explain the sheer size of the current higher education behemoth.

      Quick Google suggests over 40% of high school graduates in 2012 enrolled in a four year college. It’s over 66% if you count two year schools. What percentage of those students are there for the value of education in and of itself?

      If the pizza place sells 250,000 a year of pizza, but launders 2.5 million for the Mafia, does it matter if the pizza tastes good?

    • Randy M says:

      I addressed this in a reply to a comment directly above (which is chronologically after this, though), but stated differently, intellectual improvement is a valid good, but most people are a bit further down the maslow hierarchy of needs than that. If you are idle rich, go learn the glories of thought for its own sake, but the average citizen needs to learn marketable skills first. And given the price, the idle rich is the only group college should be marketed towards, anyway.

      Maybe college isn’t designed/optimized for imparting vocational training, but you can’t blame applicants for having the misconception when that is how it is explicitly sold to them as.

    • nydwracu says:

      Yeah, college isn’t particularly good at not bombing that idea out of the mind of anyone who goes through it.

      ~80% of the educational value I got from college came from the fact that I took classes with the professors whose lectures were least likely to follow the curriculum. I could’ve gotten that from having a social group that includes some interesting people, or following Handle or Michael Vassar around with a notebook for a month.

      The other ~20% is that they made me read bits of Nietzsche, acquire a basic familiarity with constitutional law, and take music classes, only two of which turned out to be at all memorable. I took two conlaw classes and one Nietzsche class. So that’s five classes out of… I don’t know. I took about 25 credits most semesters.

    • Matt C says:

      Lots of people value education for itself, but hardly anybody values the big smelly bundle of carrots, sticks, bullshit, and gold stars that is a university undergraduate degree in and of itself. Everyone is there for the credentials. Everyone knows it. See James Miller’s comment about Burger King and Harvard elsewhere in the thread.

      You can watch people pursue education where they do value the learning for itself, and there are no external rewards: arts, sports, niche interests, Khan Academy and similar. Looks and works a little different than your typical college classroom.

      • Randy M says:

        “but hardly anybody values the big smelly bundle of carrots, sticks, bullshit, and gold stars that is a university undergraduate degree in and of itself. Everyone is there for the credentials.”

        I’m not sure this is true, although it might be of Gen-X and later. But you’ll often hear talk of “the first in our family to graduate college” etc., from people who are quite proud to have the gold star, rather than “the most knowledgeable in our family about philosophical traditions” or what have you.

        Apart from the professional signalling, a lot of what college is (or was) selling was the social validation as a smart person.

        • Matt C says:

          Well, yes. A lot of bright people who don’t finish college feel vaguely guilty about it, and some people do see finishing their degree as a valuable personal accomplishment in itself.

          When I wrote my comment I was thinking of college having “value in itself” as meaning the actual knowledge you got from the classes you attended. I wasn’t thinking of this other (reasonable) interpretation of the phrase.

    • Education or schooling?

    • gbdub says:

      Naive? No. Smacking of privilege and elitism? Absolutely. Not everybody wants, needs, can afford, or is suited for “an education as an end in itself” and requiring 4 years of adult life directed exclusively to it as an entry point for decent jobs that don’t really need it has all sorts of negative externalities that hit hardest those who can least afford it.

  74. Pingback: Against Tulip Subsidies | Official site of DJ Michael Heath

  75. Magus says:

    “You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail.”

    Yeah, good luck with that. The second everyone starts seeing Ashkenazi Jews scoring two standard deviations above black americans on average, RACISM. Non stop RACISM.

    The second the West went down the social engineering blank slate race-denying “IQ is just a number for your ability to score on IQ test” path it basically eliminated your admittedly more rational option.

    THAT’S why Sanders can’t propose the obvious solution. Education Realist (and others) have addressed it in more detail above. What I don’t understand is why someone as bright as you Scott (who obviously knows the score) would pretend to not notice the elephant in the room. Likely to comfort your more, how do we put this gently, “naive” readers no doubt.

    • “You can give them all sorts of examinations, you can ask them their high school grades and SAT scores, you can ask their work history, but if you ask them if they have a degree then that’s illegal class-based discrimination and you’re going to jail.”

      Yeah, good luck with that.

      Did nobody notice that Scott was being tongue-in-cheek here?

    • Svejk says:

      IQ testing for purposes of employment is legal and commonly employed in many of the more ethnically homogenous nations of Europe, and yet there is still a strong revealed preference for college degrees.

      • Matt M says:

        My working hypothesis on all of this is that when it comes to jobs that require a very specific skillset, the preference towards the college degree gets smaller and smaller. In other words – if you need someone to write code in a specific obscure programming language, the question you ask is “Can you program in this language?” If the language is obscure, you might have a potential applicant pool small enough to legitimately test and evaluate candidates solely on their programming ability and make your hiring decision based on that. Whether they learned that language at MIT or at some coding bootcamp or from a hypnotist at a carnival isn’t really relevant to you if they outperform their competition.

        Now consider the less specific, more “entry level” (in a post-college sense) jobs. You need someone who is generally intelligent, capable, able to learn, easy to get along with. Try putting stuff like that on a list of requirements and see how many applicants you get. Turns out EVERYONE considers themself to be all of these things. (In Europe, with higher unemployment rates, this is probably an even worse problem).

        So you now get 1000 applicants for every job. Ideally you’d give an IQ test and an ethics test and a subject-relevant test and have a personal interview with each one. But that’s not very efficient, you don’t have the time or money to do it. You COULD just randomly throw out 95% of the resumes and test the 50 people who are left, but that seems rather un-scientific and your boss would never go for it.

        But wait. You recall the existence of a third-party who does all of these things. They administer an intelligence test, they research the person’s background, they interview, they observe and record their ability to learn over a four year period. So you agree to hire this firm to help pre-screen your candidates and filter out a certain number who didn’t meet expectations of the firm.

        Except that it turns out the firm is a university and they do all that all the time for free. So by instituting a college requirement, you’ve dramatically reduced the amount of resumes you have to search through and you’ve done it in a way that seems to be based on something reasonable and not just random arbitrary chance. Your boss is happy, the public won’t get mad at you because they themselves respect the university system generally. You’ll still likely end up with a capable person. You win on all fronts. Just like the doctors in Ireland by getting to have someone in the office capable of medicine and also bringing “a touch of class.”

  76. Max says:

    Hmmm is it really impossible to get a job without degree? You already gave an example of your ex-gf getting software developer without relevant degree

    I got several IT jobs almost without a degree. Almost. Basically I mentioned in my resume that I have bachelor from in CS from a state or udjistania, and produced when required Russian transcript (which might as well be my birth certificate – since no HR people could actually read it)). Most employers actually didn’t care about it – claiming it on my resume was enough (they did drill my relevant skills though)

    Sure it probably helps, but in my experience what helps more is relevant skills and relevant previous jobs experience

    Otherwise I agree – college education in US is ridiculous bubble. Its not bad per se, but its overpriced due to it being basically a mandatory “resume tax”

    • meyerkev248 says:

      So the value of an American college program is that it will:

      * Place you in a great environment for networking
      * Provide lots of opportunities to develop those skills (though your best bet is to use that networking to go out and get an unpaid position working for a professor or start doing the whole open-source thing).
      * Leave you reasonably certain that you have at least a basic grounding in your course of study so that there’s no obvious major holes. (IE: I had implemented pretty much the entire STL at least once before leaving college)

      And most importantly:
      * Upon having done the first 3, place you in a room with the Google (and hundreds of other) recruiter(s) in order to get paying internships, even more experience, and eventually a job using the recruiters that they brought to you and the signalling resume that comes from having worked for these professors on these projects and at these companies on these projects.

  77. Droque says:

    Given that most working visas require a degree (or several) (or being really rich or honestly marry a citizen), how would that protected characteristic work with them? In particular the H1B that requires, aside from the degree, companies to submit proof that they actually need a foreigner because they can’t find an American with the characteristics they need.

  78. James Kabala says:

    The undergraduate degree of a future M.D. can theoretically be in anything, but in practice, it usually is in biology, or perhaps even in a program explicitly labeled “pre-med,” right? Or am I out of it?

    • ahd says:

      in australia, it can usually be anything, provided you score highly enough on the standardised entrance examination, the GAMSAT.

      so, the impatient go straight for the biomedical science degrees, aye. but there are many paths, and some of them trade off efficiency for other considerations.

    • Devilbunny says:

      Many have biology degrees, but unless you are certain you want a healthcare career, it’s a terrible major except at elite schools. Almost anything else would provide better alternative career paths if you don’t get into medical school.

      The pre-med curriculum is surprisingly small – one year each of biology, physics, and English, plus two years of chemistry, plus some school-specific requirements (generally a year of calculus, although AP credit applies).

  79. rsj says:

    I’m also critical of the subsidies, but the proposed solution of disregarding the years of regulated training and relying on “examinations” or certifications to award medical licenses is impractical.

    You would end up with many more unqualified doctors. The license exam is just the last verification that you have skills. Taking and passing four years of classes is a much more accurate and important verification. You would need to expand the license examination phase to something like a costly year long series of tests that cover those tests you took in medical school. And if someone gets halfway through but struggles in one area, you want them to be able to learn more and then retake it, no? And it’s certainly efficient to provide some training centers alonside of the examination centers to help them, right? So you’ve recreated medical school.

    It’s just more efficient, from the point of view of the licensing agency, to have a smaller test and only allow highly qualified people to take it who have passed other tests. As it’s more efficient from the point of view of the medical school to only allow highly qualified candidates to enter, etc.

    For the same reason, I would not want someone to have an engineering license just because they did well on a single exam. I want them to have taken classes, done homework, and taken many exams on many related disciplines, and to have worked on their own engineering projects over the course of a couple of years, before I give them a license.

  80. Hari Seldon says:

    Higher education suffers from a lot of the same issues as healthcare in the United States. The left cries “see what happens with unfettered capitalism.” The right cries “see what government interference does.” But we don’t really have a left or a right leaning system; we have the worst of both worlds.

    Education and healthcare are mainly private, profit-driven sectors with tremendous government subsidies and huge regulatory protections. The federal government has made education loans a no-risk proposition for lenders. The lenders are backstopped by the government and the debt can’t be erased in a bankruptcy.

    In turn, this makes it possible for the universities to raise tuition pretty much as high as it wants. Even the dumbest kids can get $50,000 in student loans as long they have an acceptance letter.

    Take away the government subsidy and suddenly the lenders are going to start looking a little more closely at who they are giving money to. “You got a 2.3 GPA in highschool and want to study French Lit? Ummm… yeah…. I think we’re gonna pass on this.”

    With a little skin in the game, lenders would actually develop actuarial tables for all the different degree programs and lend accordingly. Programs with high utility would easily get loans. “Recreational degrees” like philosophy (no offense, Scott) would be harder to get loans for. The schools would start skewing their offerings towards programs that actually provide paying jobs and career paths for their graduates. The recreational degrees wouldn’t disappear altogether, but they would certainly shrink in size and budget.

    The solution to education is the same as the solution to healthcare. Choose one system and stick to it. Either the government essentially takes over or it gets the heck out of the way and lets the free market actually function. Subsidizing a profit-driven entity is a recipe for perverse incentives and skyrocketing prices.

    While I tend to favor the free market solution, at this point I don’t know that it can happen. Any politician who talks about reducing any kind of subsidy for healthcare or education is committing career suicide. Even hinting at something like that is enough to be immediately branded as hating education, children and fuzzy ducklings. Moloch wins again. Stupid Moloch.

    • Tom says:

      I don’t expect education loans to be erased in a bankruptcy any time soon. It’s an insane risk to lend $100k to someone with no assets, no income, and no credit. Students would have a strong incentive to just declare bankruptcy upon graduation, dispose of the debt, then get on with their lives. Take away that protection and lenders are going to require students to find someone to co-sign the loan.

      • Deiseach says:

        Don’t parents co-sign as guarantors on loans? I thought it was illegal to lend money to minors, and in law “under the age of 21” is a minor?

        • Devilbunny says:

          The American age of majority is 18, although you can’t purchase alcohol legally at that age. The loans are issued in the student’s name. No co-signer.

          • Deiseach says:

            May I ask: is the American financial institutions system insane? Who the hell cheerily says “Why yes, young person not legally entitled to buy a pint of beer in a pub, I will lend you a couple of hundred thousand on your bare name and the chance that in four years’ time you will be in a job making $$$$$$$$ instead of, you know, working the tills at Lidl or Aldi! Collateral? Co-signers? No no no, why would I need that?”

            Though I suppose it is something better than threatening to sue the parents out of their socks and forcing them to sell their house when the repayment of the loan comes due.

          • onyomi says:

            It’s not the financial institutions that are insane–they would never offer the loans if the federal government weren’t backstopping them. As for whether the government is insane… well, don’t European politicians trade short-term happiness of voters for long-term insolvency too?

          • notes says:

            The American financial system looks at that young person, shrugs, and thinks ‘not dischargeable in bankruptcy… they’ll probably make that kind of money over their lifetime, even if they are working tills somewhere. Not a bad risk!’

          • HeelBearCub says:

            Point to those above, I believe it’s actually the Federal government itself, or the school, that makes the loans in most cases, not independent financial institutions.


            It’s a quirk of the American educational system (and the American system in general). I guarantee that if Ireland were to impose the kinds of costs on students that the American system does some violence would be done in the name of access to higher education.

            For whatever reason (I have my thoughts on why, but it isn’t material to this discussion) tertiary/college education isn’t seen as a public service open to all citizens in America. Therefore, you end up with the big kluge that we have.

            It’s either loans or go without college education for a large bulk of the population.

            Basically, in the US at this time, more and more of the costs of education are being pushed on to individual students, rather than being born by the state as a whole.

            Most (decent) colleges aren’t for-profit, by the way. The for profit offerings mostly generate notoriously bad outcomes at high cost. You would think that good for profits would out compete the bad ones, but that doesn’t seem to have been the case so far.

          • Gbdub says:

            You say “more of the costs are being pushed on the students”, but the costs are going up (fourfold relative to inflation) in the past 30 years, independent of who is paying it.

            If college only cost what it did in the 70s, adjusted for inflation, most people could afford it even without subsidies, and this whole thing would be a non-issue. I know my parents were able to graduate from a non-flagship state school in the early 80s almost entirely on part-time income from after school/summer jobs.

            And that should be the lesson of the tulip metaphor – the problem isn’t that the government isn’t providing tulips, it’s that tulips cost $100k in the first place, never mind who ultimately foots the bill.

          • HeelBearCub says:


            You can’t look at sticker price. Tuition at State schools wasn’t set so that it covered the actual cost of the college. Less.and less of a universities funds are coming from state budgets and more is being covered by tuition.

            Does the current untested loan program make much sense? No. But it is hardly the key problem.

          • Gbdub says:

            HeelBearCub, you missed my point. The difference in TOTAL cost (state and other subsidies + tuition) per student has grown way faster than inflation. And that’s on the universities, not the legislators.

            I’m not just looking at “sticker price”. Yes, that has been exacerbated by lower state subsidy as a fraction of the cost, but to place the whole or even most of the blame on the states is to miss the explosion in total cost.

            Tulips used to cost $100, of which $50 was covered by the state. Now they cost $500, of which $100 is covered by the state. Yeah, the state subsidy has “reduced” as a fraction of total cost, but to call that “pushing the costs on students” is disingenuous.

      • Edward Scizorhands says:

        The theory is that if student loans are dischargeable in bankruptcy then schools will stop letting students take out $50,000 in debt and schools will be forced to reduce their prices (either their sticker price or the effective price after they factor in financial aid).

        • houseboatonstyx says:

          Actually there are several recent periods that could be examined.

          According to the non-partisan Congressional Research Service, until 1976, all student loans could be discharged in bankruptcy. Up until 1998, student loans could be discharged after a waiting period (of initially five and later seven years after repayment was scheduled to begin). In 1998, Congress made federal student loans nondischargeable in bankruptcy, and, in 2005, it similarly extended nodischargeability to private student loans.

          Ordinarily I wouldn’t cite Huffington Post, but when I Googled for { student loansdischargeable in bankruptcy } this was the only snippet in the first ten to mention that it was not a new idea.

          • Anthony says:

            If HuffPo were a real news source, they would have asked (and reported) who lobbied for those changes. Though Edward Scizorhands above isn’t entirely correct, because the colleges aren’t impacted by a student defaulting on student loans. They would only care if problematic default rates got lenders to restrict student loans in some way.

    • “Education and healthcare are mainly private, profit-driven sectors with tremendous government subsidies and huge regulatory protections.”

      Higher education. K-12 is overwhelming run by governments.

      • HeelBearCub says:

        What percent of higher ed students go to a for profit school? Nothing close to 50+%, I don’t think.

  81. meyerkev248 says:

    So the other subtle issue here is that college is still totally worth it.

    Could it be cheaper? Hell yes
    Could we do things that replace college and provide better value? Maybe. (Though I do share a lot of the “App Academy is not a perfect replacement for a CS degree” reservations albeit with the “Check for $185K, 3.75 years of your life and only an app academy degree might be worth more than a CS degree” caveat).

    But at my interest rates $1000 in debt == $11/month for 10 years * 12 months/2000 hours == 6.6c/hour per $1000 in debt for 10 years. And a career lasts 40.

    Assuming that out-of-state tuition costs are reasonably representative of actual costs, college costs you $200K at a good university. Or $13.20/hour for 10 years. That’s $26,400/year.

    Which ya know, I hate to say “Only $26,000/year”, but comparing 40 years of a post-college career to 40 years of working at the sandwich shop… ONLY $26,400/year.

    This isn’t the equivalent of one must have a tulip to get married, it’s one must send the duchess a really nice tulip as part of asking her hand in marriage at which point you inherit all of her lands. At which point of course tulips will get crazy expensive.

    • Randy M says:

      Hmm, are you saying college costs about $26k per year for the rest of your working life? And you didn’t even factor in the 4 years of foregone wages, which sure is minimum wage at the start, but not necessarily for long. I think lots of non-college requiring work pays more than $26k less than a lot of work people with degrees are doing (especially as the supply of college educated job seekers is expanded!).

      But never mind. Scott isn’t saying here, “Don’t go to college!” He’s saying that a heck of a lot of the jobs that are requiring this enormous investment of people don’t put much of what is purchased with that sum to use, and what they do could probably be taught in a quicker and cheaper course over a year or so, and still we require this sort-of ritual sacrifice on an enormous scale.

      It is like if one of the lesser nobles will only marry people first burn their families estate to the ground. Those who do so are better off than those who don’t, sure (except for those on the bottom end of the average, who are lured to college and flunk out or whatever) but why should society set in place a system for the destruction of so much wealth* merely so that the duchess will have an impeccable way of knowing who is really committed and who isn’t?

      *I don’t mean to say that college is worthless here, but it is certainly of less practical value than the cost.

      • meyerkev248 says:

        I don’t mean to say that college is worthless here, but it is certainly of less practical value than the cost.

        But that’s the thing. Right now, it totally is.

        It might be pure credentialism to varying degrees (see: “literally this entire thread”), but most people who go to college don’t lose money on it. They might be better off doing some other cheaper way of getting that knowledge (IE: Would you rather have a CS degree or App Academy, 3.75 years of your life, and a check for $185K?), but college still has positive utility. You put $200K in and get more than $200K out.

        The $26,400 is $26,400 for 10 years at which point you get off. Assuming 6% interest, your loan payment on $1000 goes from $11.20 to $5.50 if you extend it out from 10 years to 40, so call it $13,000/year for the rest of your career.

        Anecdotal, but my father is one of 13 children, currently ranging from 71 to 54. And the college-bound kids made WAY more than $13K/year more on average over their career. Enough that I’m not even worried about those 4 years of lost income (And honestly, part-time jobs and summer internships will make up a surprising amount of that difference).

        Mind you, that average hides a lot of variation, but yeah, college is still a good deal. It’s just that it could be a way better deal.

        • Leif says:

          On an individual level, college is often worth its costs. On a societal level, it isn’t. The comment you replied to seems to have been talking about the societal level, so I think what it said was correct.