Where did Swedish schooling go wrong?

Some parts of this paper seem a priori implausible to me, and I don’t think the abstract puts the best foot forward for the paper, but these are such important issues I wanted to pass along the new piece by Magnus Henrekson and Johan Wennström.  Here is the opener:

The Swedish school system suffers from profound problems with teacher recruitment and retention, knowledge decline,and grade inflation. Absenteeism is high, and psychiatric disorders have risen sharply among Swedish pupils in the last ten years. In this pioneering analysis of the consequences of combining institutionalized social constructivism with extensive marketization of education, we suggest that these problems regarding school quality are to no small extent a result of the Swedish school system’s unlikely combination of a postmodern view of truth and knowledge, the ensuing pedagogy of child-centered discovery, and market principles. Our study adds to the findings from previous attempts to study the effects of social-constructivist pedagogy in nonmarket contexts and yields the implication that caution is necessary for countries, notably the U.S., that have a tradition of social-constructivist practices in their education systems and are considering implementing or expanding market-based school reforms.

At the risk of sounding like Bryan Caplan, is schooling even effective enough for mistakes in method to be so fatal?

For the pointer to the paper I thank Daniel Klein.


It is really hard to make someone smarter but really easy to make them stupider.

My great grandfather, also my great granduncle, told my great grandmother, he would push just the tip. Now look what happened? I ended up with mongoloid features due to a trisomy in my chromosomes. Tough luck, wouldn't you say?

"Absenteeism is high, and psychiatric disorders have risen sharply among Swedish pupils in the last ten years"

Hmm, could it be that the student population has changed in Sweden over the last 10 years? /s

As a concerned swedish parent who has been following the evolution of our school system for the last eight years, my conclusion is that the constant lowering of the age when pupils get grades is at fault.

When I was in school or wasn't until 7th grade, now it's in fourth.

I assure you that letting a 10-year-old know how he is doing in school is not responsible for the collapse of your educational system.

It's funny, I am also a concerned parent leaving in Moldova, but I am concerned about exactly the opposite, recently the system was changed and grades are assigned since grade 5. When I was a kid we got grades from the very start in grade 1 and I believe it was a good system.

So I guess Sweden and Moldova converged here but we came from opposite poles.

Did you look at the gap in advanced students in the graph on pg. 9 of the report?

Less than 5% of Swedish 8th graders were advanced in math, and that decline happened in the 1990s when the curriculum changes were implemented. The US and Russia have double Sweden's percentage of advanced 8th graders. East Asian countries have up to 10 times Sweden's percentage of advanced learners.

Sweden's advanced learners are not being handicapped that much by the 8% of Sweden's population that are Muslim.

+1, this isn't an immigration issue. This is evidence of a rapid decline in the quality of the primary school system. Personally, I would have thought a drop this quickly in a relatively short period (12 years) to be implausible. But the evidence seems to be pretty good.

The article does not comment on the administration of the schools or on the incentives for the teachers. It would be possible for a generation of teachers who hoped to make a difference and be rewarded for their skills to be de-motivated by a system which denied any objective reality and devalued all individuals in the name of equality, even if the skills of those teachers were in fact minimal.

HELLO! It's the Muslims. They made the Swedish dumb at math and science. I mean come on its so obvious what's happening. They need to build that wall!

I've heard that in Swedish schools even Christian teachers are promoting Al-Gebra .

Speaking of mental illness... my doppleganger is back.

Sweden still outperforms the US on the PISA (admittedly a very low bar): https://factsmaps.com/pisa-worldwide-ranking-average-score-of-math-science-reading/

And the highest performing nations the article highlights, Singapore, Japan, Hong Kong, South Korea, and Taiwan, all have significantly great school choice than in the US. And all have large majorities of students co-enrolled in private supplemental schools such as cram schools and the like. It is pretty funny to blame marketization for Swedens fall. The leading performers also incorporate other forms of school choice.

Singapore has a balloting system to keep public schools open to students outside of the principal catchment area thus providing a measure of school choice. In the US of course, open borders applies to everything except for public schools. Nothing is more zealously guarded and enforced than attendance zones to maintain rich school district attendance boundaries. Send us your poor, diseased non-English speaking children, but just not to wealthy school districts. The primary purpose of public schools being to prevent mixing of aristocrat blood with commoners.

Hong Kong has direct-subsidy private schools.

About 30% of high school students in Japan are in private schools.

Private English immersion schools are extremely popular in Taiwan.

And South Korea also has elementary admission lotteries to equalize student access across schools. It also features a significant role for special-purpose, autonomous private and international schools and schools for gifted students.

When it comes down to it, very few education systems feature the strict social segmentation and authoritarianism that the authors purport to find to be an advantage in the US.

I found myself trying to unmetal the thinking in the article in question and as I was doing so I stumbled onto your comment. Thank you very much for your detailed response as I was afraid that the libertarian / authoritarian stuff was slipping by the regular readers of the stuff.

In the US of course, open borders applies to everything except for public schools. Nothing is more zealously guarded and enforced than attendance zones to maintain rich school district attendance boundaries. Send us your poor, diseased non-English speaking children, but just not to wealthy school districts.

Simply not true. First, there is no 'in the U.S.' for these issues -- they are controlled at a state level. But at this point, nearly all states do have some form of interdistrict and intradistrict enrollment:


Here in Michigan, not only is open enrollment across district lines common, but funding has been centralized, so funding follows the student when they choose to enroll in a neighboring district.

Now, for extra credit, can you guess why lefties don't like letting kids from poor areas enroll in wealthy neighboring districts?

Because they live in the wealthy neighboring districts?

Nope. Because when kids leave poor districts, so does the money, and the poor districts end up shrinking having to close schools and lay off teachers -- just like a business that's losing customers would. Lefties, unsurprisingly, turn out not to be big fans of this kind of market competition. Here's a local story along those lines, where the poorer district is begging the wealthier one NOT to provide bus access to cross district students:


The latest evidence seems to show that the overwhelming majority of benefits from schools choice stems from the closing of bad schools and expansion of goods one.

>Nope. Because when kids leave poor districts, so does the money, and the poor districts end up shrinking having to close schools and lay off teachers

Great point.

Hey, by pure coincidence, does this explain any of the lefty love for open borders and illegal immigration? By any chance, does this flood of children end up in hopeless crappy urban school systems, increasing the money flow to the unionized teachers at said school systems?

There is a nationwide war on charter schools and voucher programs in the US with a never ending stream of union-funded propaganda research trying to outlaw homeschooling and eliminate charters and vouchers. Aided by the media, it has been a highly successful campaign with only about 5% of students in a charter school and only 10% in private school. The defining characteristic of the US education system is the neighborhood school with closed attendance boundaries strictly mapped to segregate students by real estate value.

Swedish scholar Tino Sanandaji did work on the US Pisa showing that national or ethnic subgroups of Americans usually did better than the home country. Hispanics do better on PISA than South Americans, Blacks than Africa or even Jamaica I believe. Scandinavians outperform Scandinavia, Western Europeans outperform W. Europe, Japanese Americans outperform Japan; Asian Americans outperform their home countries. So the low US score seems to reflect composition effects.

An acquaintance of mine who lives in the Beverly Hills Unified School District receives the report cards of several students who clearly do not live at his house. The BHUSD claims to hire a private detective to figure out who lives or does not live in the district (although clearly not as effective as you may think).

'a prior implausible'

While an interesting suggestion for a succeeding user name, one can safely, a priori, assume a simple typo.

I would assume a mental issue.

Psychiatric disorders are strongly correlated with low IQ (more schizophrenia, depression, anxiety, mania, etc.). Sweden's recent influx of immigrants largely hail from low-IQ nations (the average IQ in the Middle East is about 83, if you're being generous), so I'd think a big chunk of the increase in disorders could be explained by that, rather than by a "postmodern view of truth and knowledge." Sweden's new demographics could also explain the increasing absenteeism. The country now has a few hundred thousand new problem cases that it didn't have before, and gradually the consequences of that shift are becoming more and more visible in all aspects of daily life there.

Wut so low abut an 83 IQ?
Skul is 4 fukers anyway. Swedin sux.

But Australians are truly stupid and Schizophrenia has the lowest impact there in the world. Australia is also one of the least depressed countries. However, they are the most anxious people. Probably comes from the stress of being forced to maintain a laid back attitude at all times. After Australia, New Zealand is the next most anxious, followed by the United States.

I think we should get the low IQ white rural population out of the US. What a much better country we'd be!

Second, what's the population of Sweden? Just under 10 mil. Good thing we are always holding it up as an example that might or might not work for the US (e.g., but the Nordic countries have such good universal healthcare!). Or, it works for Sweden/Norway, it should work for us! Stupid.

"low IQ white rural population"

Is low IQ correlated with rural? I doubt there's much connection. Anyway they regress to the European mean.

Honestly, I don't know. Is there a definite source for IQ data? I'd love to actually look at a dataset. Can anyone provide a pointer to some good data?

If you were actually looking for the low IQ groups, rural whites would not be the first group, or even the second, that you would go to.

the US NAEP test results showed the rural was doing better than city, town and suburban, table 6b. https://nces.ed.gov/programs/coe/pdf/coe_tla.pdf


< Rural rulez!! >
(\__/) ||
(•ㅅ•) ||
/   づ

Thanks. Interesting. As loathsome as some of the commenters here are, the statistics (here for public education only) do suggest that rural whites aren't the dumbf*cks I thought they were.

"rural whites aren't the dumbf*cks I thought they were."

No, the evidence would indicate the opposite. Apparently those residing in cities are the " dumbf*cks".

Maybe the swedish (or nordic) utopia envisioned by some (perhaps bernie s) is really Coming True (some book predicted utopia will come back or rise again).

One has 'markets in everything'---a combination of profit and market based education with postmodernism / social constructivism. Possibly Liberty University, Praeger College, and Texas public schools (and curriculum) might be best US analogs (eg laws of USA and evolution don't apply to Texas).
While postmodernism and constructivism are sometimes associated with 'leftism' they are no more ideological than a smart phone.
You just 'learn by doing' (economics), 'make it up and pay as you go if you can', and combine it with the market using the rule 'whatever sells'.

Perhaps test score declines show (following caplan and andrew hacker before that) that education is obsolete--you can just have an app for education on your phone. Who needs arithmatic when you have a calculator. Perhaps in future people will be born with a smartphone glued to their face, mouth and ears .

I wonder how sweden compares with finland.

Interestingly the sequences 'immi' 'migr' and 'refug' are not found which makes me think that.... the paper is not very comprehensive....

And I didn't know "social constructivism" is a thing. Donald Trump may be our first post-truth president, but it's not his fault. Truth is a social construct and is a function of one's socialization. Truth to you and me may be as different as the color of our hair. Or more accurately, the people with whom we socialize. Or attend school in this case. I've commented before about the socialization function of the public schools I attended in the 1950s in my small Southern town. All the students looked and acted alike: we were all white, Christian (with a few exceptions), and middle class as the schools (and everything else) were segregated by race and inequality wasn't a thing. Today, are any two students alike? The warning about market-based school reforms is that it would reinstate the segregation of my childhood, except the segregation would not be limited to race but all manner of differences, from ethnicity to religion to social and economic class, even political preferences. In such a world, there would be as many truths as there are differences among us. Tribalism on steroids.

WKPD: 'Strong social constructivism as a philosophical approach tends to suggest that "the natural world has a small or non-existent role in the construction of scientific knowledge".'

So, shite then.

I don't know what to make of this manuscript. The ideas are intriguing and important and the premise may even be plausible. But the over-hyped conclusions and lack of sound methodology leaves me wondering what the contribution is. The view that all knowledge is socially constructed and the downplaying of "facts" is shown to be correlated with all sorts of changes in academic performance, psychological problems, etc. But so is the growth of video games, immigration (please don't start on a digression about this), growing wealth disparities, etc. The lack of any methodology that could distinguish between these myriad developments leaves me thinking this essay is just a thought-provoking conjecture. Fine - but perhaps all of the declarative hyped conclusions could be stated more accurately as such.

Video games?

The report did mention Sweden's high absenteeism. Video games are often a proximal cause of absenteeism.

I assume they are not trying to argue that the schools are causing the rise in psych disorders, so it’s not really surprising that conditions are more difficult with a more difficult population, or that recruitment is more difficult in conjunction with an increased market orientation. Market reforms of course an euphemism for lower pay and less benefits.

"Market reforms of course an euphemism for lower pay and less benefits."

Please look up the word "euphemism" in the dictionary, thanks

I looked up troll instead.

There was a picture of you

Is he handsome?

Pro tip: it doesn't mean what you think it means

Here's me on the topic:


Here is related article by me, rendered into Swedish by others:

psychiatric disorders have risen sharply among Swedish pupils in the last ten years

If Jean Twenge's data in iGen is correct, then psychiatric disorders (is that equivalent to mental health disorders, or are psychiatric disorders a subset of mental health disorders?) are also rising in the United States and perhaps the world.

Amazingly the modern focus on obsessive parenting is itself destroying our children. Imagine all the countless hours spent on helicopter parenting only to discovery it gave your child mental illness

Much of the article is dubious, but this line, in the conclusion, is worth a ponder: "...the psychosocial environment in the classroom plays an important role. A lack of structure and peace causes pupils’ survival instincts to react to perceived dangers and crowds out cognitive capacity for knowledge acquisition"

Well security and etc are certainly important for an education environment, but lets not oversteer and start arguing to bring back the shame and cane method

Johan can't learn because he's too preoccupied with not having a paperclip jammed in his eye? Seems plausible but also not unfixable.

I think that's a pretty well-known and non-controversial statement in educational circles. They teach it in Ed school in the US.

Well, the data backing Caplan's arguments don't support your conclusion at all. Just because there isn't any good markers within the US parenting population (and not cofounded with heritable properties) that can statistically indicate parenting performance leading to better child achievement doesn't mean it doesn't exist. Certainly, a large portion of outcome cannot be explained by heritable properties. So we can easily image that broad changes in parental behavior or educational practices could have substantial impacts on student performance. Nothing in the data suggest otherwise.

I don’t know why you’re misrepresenting Caplan.

“Large portion of outcome that cannot be explained by heritable properties.”

This may or may not be true, dependent on one’s definition of “large.”

Regardless, the bucket isn’t parents versus genes.

It’s peer group versus genes. Or more likely, peer group x genes in some weird interaction term.

A smart bet would be peer effects, IQ, and low neuroticism.

Which is what we see in the real world.

Apologies, meant in reply to mpowell.

I'm inclined to agree with Tyler that higher education is meaningless, and that most of what he and Bryan Caplan do is a waste of their time and society's resources--though the null hyothesis, that there is no more productive use of their time, cannot be rejected--but I don't think that applies to elementary school. There are big productivity differences between populations with high literacy and those with low, so elementary education can clearly make a difference.

"There are big productivity differences between populations with high literacy and those with low, so elementary education can clearly make a difference."

+1. I'm not sure where the break comes in, but at the very least literate populations seem to have a distinct advantage. That's not signalling.

At some point we started thinking the purpose of schooling was to control student outcomes, as opposed to offering knowledge or providing a useful filter to employers. But there's not much evidence they actually can.

People go crazy in Sweden because there is very little sunshine in winter I think.

That paper does not contain the words "immigrant" or "foreign-language", and there is only one footnote on "mother language."

There are population composition issues here; quite simply, the native-Swedish to non-native Swedish composition of children has changed dramatically, which is highly correlated with school outcomes. And I would say the language competency differences between native (multiple generations) born and immigrants (first- and second-generation) is large.

The effect seems to be very pronounced and wide spread. I wonder if the stats have been broken down by ethnic group?

Comments for this post are closed