Join 3,562 readers in helping fund MetaFilter (Hide)


Sweden’s School Choice Disaster
July 27, 2014 4:56 AM   Subscribe

It’s the darker side of competition that Milton Friedman and his free-market disciples tend to downplay: If parents value high test scores, you can compete for voucher dollars by hiring better teachers and providing a better education—or by going easy in grading national tests. Competition was also meant to discipline government schools by forcing them to up their game to maintain their enrollments, but it may have instead led to a race to the bottom as they too started grading generously to keep their students.
So it turns out that the good results of the Swedish school voucher system of "free" school choice, long the benchmark for those wanting to disrupt public schooling were created by, well, cheating.
posted by MartinWisse (41 comments total) 24 users marked this as a favorite

 
The article touches on one of the problems for marketizing education: the difficulty of knowledge. The consumers of education are mostly young with little understanding of what makes good or bad education. Furthermore, even if they are aware of the bad education they are receiving, they're not the ones making the decisions. The decision makers are parents whose engagement with the schools their children attend may be very shallow, and they certainly don't sit in on lessons or take the course of study. Hence they lean on test results, which are only an index of education and not the thing itself, and are too often gamed.

When the buyer is not the consumer, markets suffer a failure of knowledge.
posted by Thing at 5:35 AM on July 27 [13 favorites]


So it turns out that the good results of the Swedish school voucher system of "free" school choice, long the benchmark for those wanting to disrupt public schooling were created by, well, cheating.

Shocked. Shocked I am, I tell you...
posted by Thorzdad at 5:55 AM on July 27


Well now we know why the hand of the market is invisible. Makes it a lot easier to cheat that way.
posted by localroger at 6:03 AM on July 27 [4 favorites]


Even if the "buyer" here (the student) were the consumer, that would presume that they're both able to properly assess the quality of their education and that they're primarily interested in getting a quality education rather than just a good grade.

It's the same problem that you run into when teachers are accountable for student assessments. It turns out that in many, if not most, cases that students aren't so much interested in learning as they are with passing and oddly aren't fans of difficult work.
posted by CaffinatedOne at 6:05 AM on July 27 [4 favorites]


The consumers of education are mostly young with little understanding of what makes good or bad education.

Because they think it is all about getting an A when it really is all about learning. A grade can easily be turned into window dressing -- but not knowledge. When I taught communications at the college level, I used to have students demands A's for C papers and I would simply tell them that education was not for them per se -- it was to let employers knows that in lieu of real-life work experience, they had the foundation and a level of competency through education because that's why they were there, and if they went in as ignorant as they were in front me of -- that diploma would rapidly lose value until it became meaningless.

Because education is usually a sign of an area's economic health -- if it is red hot -- they'll let ten years perform surgery to keep up with demand...
posted by Alexandra Kitty at 6:30 AM on July 27 [1 favorite]


It turns out that in many, if not most, cases that students aren't so much interested in learning as they are with passing and oddly aren't fans of difficult work.

... and given that teachers and schools are being made responsible for something they have relatively little control over, it's hard to imagine another outcome.
posted by mhoye at 6:33 AM on July 27 [2 favorites]


It turns out that in many, if not most, cases that students aren't so much interested in learning as they are with passing and oddly aren't fans of difficult work.

There is conflicting research on the relationship between expected final grade and teacher evaluation. Here, for example, is a large recent study (pdf) which found no significant correlation.
posted by yoink at 6:44 AM on July 27 [1 favorite]


I remember reading something from one of the Freaknomics guys I think where he said his goal was for his children to love him once they were adults.

I know the teachers I really liked as a child are not ones that I am grateful I had now that I am in my forties.
posted by srboisvert at 8:05 AM on July 27 [2 favorites]


Wouldn't students at different ages presumably expect different things out of their teachers? In first grade I might have preferred the teacher who let us eat chocolate cake and send us to recess 10 minutes early, but in college I was well aware that I'd much prefer an instructor who was intellectually stimulating and pushed me to succeed.
posted by (Arsenio) Hall and (Warren) Oates at 8:15 AM on July 27


Swedish universities are strongly constrained in how they can assess applicants. In particular, they cannot take the applicant's school into account when setting what grades they'll accept. This creates terrible incentives. If universities could adjust grades based on school, there'd be no lasting advantage in soft grading, and schools would not be able to build a reputation for soft grading. If universities were forbidden from asking for school grades entirely, and forced to rely on national standardized testing results and applicant-provided supporting material, that would also solve this problem.

Yet here we are in the dumb middle ground where universities have to use broken metrics they can't adjust.

Milton Friedman, whatever you think of him, would have seen this coming, and would not have recommended a system with this failing. So what's the basis of the political attack, exactly - libertarians bad? Please.
posted by topynate at 8:37 AM on July 27 [5 favorites]


Teachers can be measured. Just make their evaluations be based on their actual job performance (as reported qualitatively by peers, students and administrators the way most professions are evaluated) not on indirect proxy measures that don't take into account the individual circumstances and difficulties faced by students taking test instruments well known not to be good measures for lots of well understood reasons.
posted by saulgoodman at 8:47 AM on July 27 [3 favorites]


Swedish universities are strongly constrained in how they can assess applicants. In particular, they cannot take the applicant's school into account when setting what grades they'll accept. This creates terrible incentives. If universities could adjust grades based on school, there'd be no lasting advantage in soft grading, and schools would not be able to build a reputation for soft grading. If universities were forbidden from asking for school grades entirely, and forced to rely on national standardized testing results and applicant-provided supporting material, that would also solve this problem.

It would only solve the specific way in which test results are gamed, not the overall gaming of test results. Although you could then compare how one school performed in those tests with another, it still wouldn't needfully be a good measure of the quality of education a child will receive. At best it would ensure a lowest standard of "passing the test" without showing anything more than that. The disconnect between the buyer and the consumer would remain because, fundamentally, test results are not education.
posted by Thing at 8:58 AM on July 27 [1 favorite]


In grade one, we handed our worksheets to someone else to grade, usually forward or back, but sometimes left or right. If schools self-grading is an issue, have them swap test booklets (without school or student identifiers) with another school, and vary the rotation to keep grading relationships from building up.
posted by fatbird at 8:59 AM on July 27 [2 favorites]


You're right, Thing, mandatory use of standardized testing wouldn't be a comprehensive solution. My point is that what Sweden is using is worse.
posted by topynate at 9:01 AM on July 27 [1 favorite]


My apologies if I took your criticism to be broader than it should have been, topynate.
posted by Thing at 9:05 AM on July 27 [1 favorite]


Here's the bigger concern with privatizing education, IMO:

It’s also a reminder that the cold, hard calculations of markets aren’t necessarily suited to the realm of education. Governments don’t shut schools because they fail to turn a profit. Private equity firms do. The parents of more than 10,000 students learned this difference the hard way last year, when the Danish private equity group Axcel abruptly announced its exit from the Swedish school market, stating that it could no longer cover the continued losses.

as reported qualitatively by peers, students and administrators the way most professions are evaluated

Many people, including Bill Gates, have suggested this. The problem is, what are the evaluators looking for? After 15 years I feel like I can walk into a classroom and know whether the teacher is good or not, but am I always going to be right? Teachers are all very different.

Remember, you're not looking at the students at all, just the teacher, or we get right back into the problem of proxy measures... if you're looking at student engagement on the day after Halloween, or in a neighborhood where there was a shooting last night, things are going to look bad.

So the Bill Gates plan is to train veteran teachers to recognize good teaching and send them to schools a few times a year, which is going to cost a hell of a lot of money... where will it come from when some schools can't even afford the textbooks they need to teach to the test?

Letting an administrator evaluate us is how it's done now, so that's not going to solve any problems. Contrary to common belief, an administrator can absolutely get a teacher fired (at least in my district), but it is a long and involved process.

As far as student evaluations go, I'm fine with my first graders evaluating me, but I'm not sure middle and high school teachers are going to be really happy with that. I certainly wouldn't want my job to depend on a 15 year old thinking I'm a great teacher.
posted by Huck500 at 9:13 AM on July 27 [5 favorites]


If the problem is that doing evaluation the right way is too hard, how is it any kind of solution to just do it in a very wrong way that isn't even pointing in the same direction as the right way?
posted by saulgoodman at 9:27 AM on July 27


I have little expectation of serious reform from a system which still bases its calendar on getting the harvest in.
posted by Devonian at 9:42 AM on July 27 [4 favorites]


Thank goodness that there's never been any instances of district-wide systemic cheating in America's public schools.
posted by DWRoelands at 9:55 AM on July 27 [2 favorites]


If the problem is that doing evaluation the right way is too hard

I'm not saying it would be TOO hard, only that it would be hard, and in a cost-reward analysis I'm not sure it would make sense. If it's true that the problems in American education are caused significantly by ineffective teachers, then yes, it would make sense. However, I have experience in both low-scoring (poor) and high-scoring (affluent) areas, and have only run into a few teachers who would likely be let go after evaluation, and maybe half of those were teaching in high-scoring areas anyway.
posted by Huck500 at 9:59 AM on July 27 [1 favorite]


Because the main problem isn't teacher ineffectiveness and never has been, bad teacher's are just the wedge used in the political game.
posted by saulgoodman at 10:01 AM on July 27 [6 favorites]


("bad teachers," natch.)
posted by saulgoodman at 10:08 AM on July 27


If a peer-evaluation system does come to pass, I'll sign up to be an evaluator for sure... for one, it would mean some extra money, and I'd want to be sure to know exactly what the evaluators are trained to look for for my own evaluation.

One more problem with evaluations: Some horrible teachers are great at pulling it together and doing a fantastic lesson for their evaluation, and some wonderful teachers fall apart when they know someone is there judging them. This is similar to other jobs, of course, but if you're a poor coder, for example, you can't fake it three times and produce good product for the whole year, whereas a poor teacher in a high-scoring area can, because your students do well despite you.

Because the main problem isn't teacher ineffectiveness and never has been, bad teachers are just the wedge used in the political game.

/agree
posted by Huck500 at 10:10 AM on July 27


Milton Friedman, whatever you think of him, would have seen this coming, and would not have recommended a system with this failing. So what's the basis of the political attack, exactly - libertarians bad? Please.

the undead spirit of milton friedman doesn't actually care whether the system works or doesn't, as long as social democracy socialism has been dismantled. The market will, of course, punish failure...

The article touches on one of the problems for marketizing education: the difficulty of knowledge. The consumers of education...

the undead spirit of milton friedman is glad that you can't actually conceive of a situation which isn't dominated by market behavior.
posted by ennui.bz at 10:43 AM on July 27 [5 favorites]


I'm a little surprised at how credulous and incurious the general public is about trends in public policy: suddenly everyone has an opinion on teachers and how to evaluate them, on how to conduct education itself, but rarely does anyone ask how the evaluative mode became the default or normal orientation. Or in other words, what circumstances obliged us to put education and teachers on trial? It's not like other areas of human affairs have been presenting too few problems to merit consideration for reform in the past decade or so. Most other useful professions do just fine by having specialized education, a professional interest group, high practical standards, and a good mean salary which rewards all of that dedication.

There have always been issues to improve in education, as there are in literally every human endeavor, but there was no precipitating incident, no damning report that came out, no shameful scandal that prompted the logic of this era of "reform". It seems more likely that the sharks of big business and finance smelled blood in the water and got into the game, starting I suppose with the No Child Left Behind act.

It just seems deeply odd to me that the conversation about "reforming" education has become so normalized as to make the necessity of doing so beyond question. Most troubling is that the model of reform which emphasizes the centrality of privatising control and building in evaluative systems which are not really open to discussion has also become utterly normalized. In short, the problems that education has had in America at least are not even being addressed by this "reform" scheme, and in fact the reformation is both creating new problems and exacerbating old ones in new ways.
posted by clockzero at 11:02 AM on July 27 [4 favorites]


While there are many problems with Swedish friskolor (charter schools), blaming the secular decline in quality of Swedish K-12 education over the past 15-20 years on them is quite naive. Over the same period, there were a number of other huge and structural shifts, including:

1. A huge decline in social and economic status for the teaching profession
2. Pitiful growth in teacher salaries compared to those of other professions
3. A plunge in the quality of teacher training programs (and lower-than-ever admissions requirements)
4. Complete decentralization of the school system about 20 years ago (responsibility shifted from central government to counties/municipalities)
5. Complete curriculum revamp in the mid-90s

It seems to me that some of these are more likely to have driven the decline than shifting something like 15-20% of students to friskolor.

There is a lot of research showing that the success of charter schools in the US depends very strongly on the type of regulatory regime they are in (how are students selected? can schools be run by for-profits? how are schools evaluated/held accountable? are under-performing schools shut down on a regular basis?). While there is little evidence for charter schools "as a whole" outperforming non-charter schools in the US, there are definitely success stories which (surprise!) almost always fall in high-regulation/high-accountability regulatory environments. If we can't even compare charter schools within the US, why do we think that Swedish friskolor (who operate in a completely different regime/environment) make an appropriate comparison?

(Moreover, it may be worth pointing out that most Swedish friskolor serve middle/upper-middle class families looking for an "alternative" education experience. Most US charter schools target high-poverty/high-minority communities that have traditionally been underserved by their traditional public schools. Seems to me that these are quite different...)
posted by yonglin at 11:14 AM on July 27 [8 favorites]


the undead spirit of milton friedman is glad that you can't actually conceive of a situation which isn't dominated by market behavior.

Are you always this persuasive?
posted by Thing at 11:15 AM on July 27 [1 favorite]


Akerlof's The Market for Lemons: Quality Uncertainty and the Market Mechanism is a paper published 15 years after Milton's “The Role of the Government in Education. It's pretty well known among economists.

The Akerlof paper makes a few predictions about what happens in the presence of information asymmetry:

1. Buyers, having no ability to distinguish one seller's goods from another, value all goods at an estimated average quality
2. Sellers with above average goods leave the marketplace
3. A downward spiral follows; each wave of sellers leaving the market reduces the average quality, reducing the price buyers are willing to pay for goods, and lowering the incentive to sell into the market.

In order to apply Akerlof's thesis to the education market, you have to believe that teachers know the average teacher quality and where they rank on that scale. They need to have an incentive to lie about it, and they need to have no credible way of establishing the truth. Does any of the following sound familiar?

A. Neighbors (and NYT trend pieces) saying they homeschool their children because the local options aren't good
B. Furloughs and declining teacher wages
C. A great teacher in your child's school quitting to take a better paying job in the tech sector
posted by pwnguin at 11:55 AM on July 27 [4 favorites]


I found this Slate article very difficult to parse, mostly because the first 500-600 words barely mentions Sweden at all, but, as usual, features a lengthy preamble that uses a foreign factoid to make ideological arguments about the American system (rather than actually talking about the subject of the headline).

There's also the problem with using PISA scores as some sort benchmark for determining the success of the Swedish school system. Of course, PISA is a valuable benchmark, but it's not the only one, and at least the Slate article could have mentioned that PISA is sometimes considered to be flawed.

The Slate article briefly mentions the researchers, but I couldn't find an actual link to the study (maybe there was one in the Slate article, but if so where???) and the only thing, at least in the first 700 words or so of the article that I could find that was actually written by a Swedish person about the testing results was this.

Local insights provided by local subject matter experts (rather than a Slate writer with an obvious agenda) matter.

Anyway, the article continues by comparing the Swedish approach to grading (done by teachers at the schools themselves, which seems quite reasonable to me) to the New York Regents exam system.

We have no way to determine how closely the Swedish evaluation model is similar to the NY Regent model. We'll just have to take the writer's word for it.

Finally, after that, about 800 words into a story that is supposed to be about the study, a link to the damn study.

It's in Swedish, so unless you're a Swedish speaker, you're out of luck in terms of analyzing the finer details.

Anyway, while I understand the argument about charter schools is a valid one to have, I just these this Slate article as a one-sided hack job. It's not real journalism.

For whatever it's worth, I'm a former teacher (BEd, social studies) that care very much about assessment, and I have kids in school right now, and I am very aware of how aging demographics, combined with ideology etc are putting pressure on the traditional school system (which hasn't much changed in over a hundred years). I also think there should be a better way to do things than to shove kids in an institutional box with little bells going off every hour.

I'm also quite conscious of how the foreign/American media distorts "foreign" news stories. Local context and insights are important. This article has none of them. If I wanted to read a jeremiad about Milton Friedman I would go to the anti-Milton Friedman website. If I wanted to read a critique of the New York charter schools movement, I would go look for info about that.

I think using this study as a tool to talk about something almost totally unrelated is cheap and dishonest and is bad journalism.
posted by KokuRyu at 12:07 PM on July 27 [3 favorites]


(Moreover, it may be worth pointing out that most Swedish friskolor serve middle/upper-middle class families looking for an "alternative" education experience. Most US charter schools target high-poverty/high-minority communities that have traditionally been underserved by their traditional public schools. Seems to me that these are quite different...)

Is that true? My experience in the US is that many charter schools are operating as "alternative" schools for the wealthy and well-connected, i.e., people who would have otherwise sent their kids to private schools. Those are the well-funded ones with the cool programs run as non-profits with some parents who are really good at writing grants. In contrast, we have the for-profit schools, located in high-poverty/high-minority communities that are now once again being underserved by public schools, but with a bunch of shareholders now making money off of them.
posted by hydropsyche at 2:10 PM on July 27


I teach in a private school in Japan. Over the past few decades, and especially in the past few years, enrollment is dropping. Partly because of increased competition with other schools but mostly because Japan's population is simply declining. There aren't as many kids to fill schools as there used to be.

My school has an entrance exam prospective students have to take, but they have to address this low enrollment problem, and they lose money when they reject a kid for a low score on the test.

Solution? Go easy on the entrance exam grading. Lower the minimum test score.

Which results in some (not all, not even most, but some) students whose parents have money, but the students are not all there academically. Or, often, emotionally. Since my school has lowered the test score requirements, all the teachers have noticed increased discipline and academic problems with students. Of course, we're expected to give them the same (easy) grades as always, even though those decisions are rather out of my hands, thank god.
posted by zardoz at 3:19 PM on July 27


There are a few types of friskolor in Sweden, the alternative education such as montessori, waldorf or IB (English) education style, or the religious schools with Islamic teaching or branches of christianity as the star of the curriculum. Recently the Danish schools corporations have arrived - and left - after trying their hand at an alternative education.

Another thing to keep in mind is that kids in Swedish schools don't get graded until 8th grade. Without grades, it's difficult to know how you score. Once the grades begin they are 3 - failed, passed, or passed with honors.
posted by dabitch at 3:28 PM on July 27


Thank goodness that there's never been any instances of district-wide systemic cheating in America's public schools.

If anyone is claiming that America's system is a model for the world, they are lying to you. Still, the problems persist in both of them, and in the U.S. probably the biggest hurdle is the tie between property values and school funding. NCLB was just shit icing on top of all that, and yeah, schools cheated because their survival was on the line, because NCLB was (is) an awful incentive structure and doesn't exist to accomplish anything worthwhile in terms of furthering education opportunities for all children.
posted by Navelgazer at 3:43 PM on July 27


My experience in the US is that many charter schools are operating as "alternative" schools for the wealthy and well-connected

Very much not true in my school district. Our charters are overwhelmingly Hispanic and majority English Learners. I very much support their efforts to educate these kids, who often come from economically challenged backgrounds. Don't paint all charter schools with the same brush!
posted by SPrintF at 4:33 PM on July 27


I invite you to also not assume that all charter schools are like the ones in your district. Since they aren't that way in my district.
posted by hydropsyche at 5:05 PM on July 27


There have always been issues to improve in education, as there are in literally every human endeavor, but there was no precipitating incident, no damning report that came out, no shameful scandal that prompted the logic of this era of "reform". It seems more likely that the sharks of big business and finance smelled blood in the water and got into the game, starting I suppose with the No Child Left Behind act.

The report that came out was called A Nation At Risk. It came out under Reagan and was not at all objective. It was politically motivated and designed to do just what is coming to pass -- start a discourse of the failures if schools. What Reagan really wanted was to dismantle the Department if Education. Dianne Ravitch stood behind the report, and has since become a vocal critic of everything that has come to pass. She did an about face.

The thing we heard about education back in the 80's, that required the strategy you see today, was that "research shows that Americans believe education is getting worse, but they believe their own schools are doing well." Deploying national standardized tests, and making them high stakes, without making any effort to ameliorate gross economic inequalities is an excellent "scientific" way to demonstrate to parents that their own school districts are failing and should be turned over to private interests -- there just happen to be some for-profit charter school chains hoping to educate your children on the cheap! (But sorry -- we'll be discouraging kids with special needs from attending -- they'll cut into our profit margins.)

You're right about the sharks of big business and finance circling the blood in the water. The acceleration of technology and the parallel religious zealotry of Big Data and the blind faith people have in the ability of quantitative measures to reveal truths that we can't already see and to tell us how we should behave have lent an urgency to the movement to open markets in education.

Reagan, Bush, Clinton, Bush, and Obama have all done their parts. ALEC, Pearson, Gates....Arnie Duncan pissing of suburban moms...I can't tell which ones believe data is the True Answer and which ones know how data lies and are just deploying it as a cynical strategy.
posted by vitabellosi at 7:37 PM on July 27 [1 favorite]


Vouchers in Sweden made it possible for people of low income to put their children in the private schools that used to be reserved for the people who had deep enough pockets to pay for that education.

This also opened up for the possibility of starting schools as a for-profit business. The deregulation of who and how a school could be run went pretty far, at the same time as grades changed, the national curriculum changed and everything else that yonglin mentioned above.
posted by dabitch at 8:42 PM on July 27


Our charters are overwhelmingly Hispanic and majority English Learners.

Charter schools require parents to apply for admission, through lottery or other means, and also require heavy parent involvement, so therefore generally get the cream of the crop in a community, whether it be rich white folks or poor Latino folks (or vice-versa).
posted by Huck500 at 9:40 PM on July 27 [1 favorite]


Well, no. That's only true for one of our charter schools. The others have neighborhoods that they MUST serve. Perhaps this is unusual, I don't know. I just don't think it fair to assume that all charter schools are elitist.
posted by SPrintF at 6:44 AM on July 28


The Lottery, (available here on Hulu), is a great documentary about public and charter schools in the United States.
posted by dabitch at 8:18 AM on July 28


Campbell's Law: "The more any quantitative social indicator (or even some qualitative indicator) is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor."

There's a tremendous amount of resources being put into a technocratic educational solution, where testing measures the effectiveness of teachers. Which, honestly, I think is a good thing — my regular gripes are about how it gets used. Leaving aside the idea that curriculum-based tests are only a proxy for some of the teacher's skill and the bias toward quantification, I don't believe that teaching accounts for the majority of a student's success at school. (But those things are harder to quantify, hence the monomania about what we can reasonably track and have a public responsibility for: the quality of teachers.)

I do still believe that we should have standardized tests, just that it's incredibly poor public policy to base funding decisions on them — they're poor metrics for allocating funds, but they're really valuable if you want to do something like research what pedagogical techniques work best. By removing the funding tie we'd see less teaching to the test and all the limitation that entails, but also capture better data about just what superior teachers are doing and how we're able to turn those successes into learnable skills for other teachers. But that goal is compromised by the corruption of the scores as they're used as a proxy for school and teacher quality.
posted by klangklangston at 2:46 PM on July 28 [1 favorite]


« Older Anil Dash reflects on the changing meaning of publ...  |  Originally published in 1962 a... Newer »


This thread has been archived and is closed to new comments