Featured Post

Sweden’s Education Privatization Failure

I have spent countless hours writing about and discussing Common Core and the international benchmark for measuring national education systems (PISA, or Programme for International Student Assessment). Some of those pieces can be found here and here. In sum, I find that most discussions about the competitiveness of America schools that emerge whenever new PISA scores arrive to be simplistic, and often ignore the class and cultural dynamic that leads East Asian and Scandinavian countries to excel on such markers.

But not all Scandinavian countries are equal in the realm of educational competence. Sweden, a country that shares a great deal culturally with the much vaunted Finland, has stumbled significantly since the privatization of their education system. Sally Weale, writing in the Guardian, summarizes Sweden’s current predicament.

Sweden, once regarded as a byword for high-quality education – free preschool, formal school at seven, no fee-paying private schools, no selection – has seen its scores in Programme for International Student Assessment (Pisa) assessments plummet in recent years.

Fridolin acknowledges the sense of shame and embarrassment felt in Sweden. “The problem is that this embarrassment is carried by the teachers. But this embarrassment should be carried by us politicians. We were the ones who created the system. It’s a political failure,” he says.

Observers in the UK may well be vexed upon reading about Sweden’s problems, since its friskolor policy – privately run schools funded by public money – was one of the key inspirations behind the introduction of free schools in England under the coalition government.”

Some conservatives have come to the defense of Sweden’s voucher program. Others have detailed that the model pursued in the country doesn’t quite fit free-market ideals in regards to school choice. These arguments remind me of days in far-left political parties, where members would often proclaim that communism would have worked if the states practicing it had only gone further in implementing their aims. It is never the failure of your politics, just a lack of vigor in furthering your ideals.

It is not at all surprising that a corporate driven education system practiced by a quarter of Swedish students failed to produce results. Much like any corporate entity, they cut corners by employing cheaper teachers that were not qualified while spending lavishly on outside consultants. They also employed a competitive scheme where teachers were financially rewarded for garnering better test scores. I can’t imagine a sadder principle to run a school. Katharine Birbalsingh, writing in the conservative British magazine Standpoint, explained how performance pay in our schools undermines their very foundation.

Private schools would use PRP [Performance Based Pay] if it worked. I have never heard of any that do. If you ask some of these old headmasters why, they won’t tell you it is because teachers are above the grubbiness of money. They’ll say that PRP would destroy the ethos of their school.”

She goes on to clearly explain the obvious: schools cannot run as businesses in the corporate world. Good schools are inherently collaborative. If a teacher is struggling, they work with more experienced practitioners to improve. Collaborating effectively means you share good ideas and discuss how to improve your school’s curriculum as a whole. What happens under PBP is that the best teachers end up invariably with the best students at the school, leaving the weaker teachers to flounder on their own. Why support your compatriot by taking on challenging students if it means you will be less likely to receive merit pay? Why distribute worthy lessons and methods if the individuals you are sharing them with are your direct competitors? The minute each individual’s pay reflects student test scores, you will begin to see nefarious horse-trading of students, with senior or competent staff taking on those more likely to guarantee “merit” pay for said educators.

A few years back, I taught at a struggling rural school in California. The 8th grade class had a large number of difficult boys that were often disruptive. I volunteered to put most of those students in my classes, as I had better classroom management than some of my compatriots. As many of these boys struggled academically, I am sure my overall test scores were lower as a result. If I knew I would not receive merit pay at the end of the year for this decision, I am unsure I would have done so. PBP would generate antagonism between teachers at any one site that would hurt the school overall.

PBP will also have an adverse affect on teacher retention. Strong teachers will be inclined to move to schools that have stronger test scores (and thus more likely to receive higher pay) than stay at a struggling school. Schools that already have the deck stacked against them will be caught in a perpetual spiral, seeing their best and brightest educators quickly move to better performing schools to maintain quality pay.

But I digress. The lesson I took from the Sweden’s privatization failure is its contrast with its neighbor Finland. Finland has long been celebrated for its successful schools, with countries around the world attempting to borrow what works from their system. The Finnish system is centralized yet community centered; only the best university students are recruited into the profession, and are paid well for their services. But more importantly, trust and autonomy is placed in the hands of each school’s administrators and educators. Outside corporate and political figures are not given reign to muck with the workings of the school as they see fit (Joanne Barkan has an excellent piece about how philanthropists in America, while celebrated, are doing harm to our schools). Their schools do not shift direction aimlessly as educational fads come or political figures look to make a point.

Good national school systems serve a purpose that exceeds efficiency concerns. They help unite a people and a nation around a community framework, both implicitly and explicitly. While distinct, the Korean and Finnish education systems demonstrate a united vision for a society that the individuated corporate schools attempted in Sweden cannot achieve.

(Image: Nikolaos Gyzis, “To krifó scholió”, Oil painting, 1885/86.)

Please do be so kind as to share this post.
TwitterFacebookRedditEmailPrintFriendlyMore options

79 thoughts on “Sweden’s Education Privatization Failure

  1. Any system involving performance monitoring is highly contingent on the performance metric being used. I note that the Slate article you linked to mentions that teachers got to grade their own class’s tests. This creates a blatant conflict of interest problem that is easily corrected. Teachers shouldn’t be grading their own classes on standardised tests anyway.

    On a similar note, the incentive to cherry pick a community can be eliminated by using a performance metric that adjusts for student demographics. This is eminently possible – Raj Chetty has done some very good work on this.

    Report

    • the incentive to cherry pick a community can be eliminated by using a performance metric that adjusts for student demographics

      This might be true on the macro level, though I see much potential for a political morass in the details of that adjustment. But it doesn’t help in the micro level. Students vary within a demographic–or even within a single family. A teacher who gets to choose students will still be at an advantage over a teacher who gets the kids the others don’t want.

      Report

  2. I do think that however schools are funded, the people doing the funding shouldn’t be mucking about with the pedagogy (unless of course you are a trained expert in pedagogy).

    Report

    • Except that the funders, the tax payers, are the next level of beneficiaries below the students themselves. Parents and payers of the schools have no say in how their money is spent on how their kids are educated?

      Report

      • Parents do. They just have to send a kid to a school which they think does a good job of educating their kid. If they don’t like it they can pull the kid out and homeschool him/her or hire private tutors or something.

        Report

        • How do you square your comment “the people doing the funding shouldn’t be mucking about with the pedagogy” with your replay to me that “Parents do”

          Are you saying that parents shouldn’t have a say is how the school is run and what is taught and that their only option is to remove their kid from a school in which they disagree with and place them somewhere else? So you’re saying that regardless of all the funds they provide via taxes to the school, parents should have say in how / what the schools teach, that they have no voice in that at all. So, shut up and enroll your kids or GTFO?

          Report

          • 1. Local Property taxes are a crappy way to fund public schools. Fund it using federal taxes either directly or by vouchers.

            2. Yes. Most people know shit all about pedagogy. That is to say, their preferences over pedagogy may not (and probably will not) provide the results they want for their kids. Given that this is the case, exit is more important than voice. If you don’t like the results of the school’s pedagogy whatever it may be, move the kid to a school whose results you approve of. Stop trying to fuck it up for everyone else’s kids as well.

            Report

  3. All things said though, there is nevertheless something deeply illiberal about wanting a school system to “demonstrate a united vision for a society”. Or at least it would end up being relatively more unworkable in more pluralistic societies. South Korea is still fairly monolithic culturally. There are no significant groups without an ingrained Confucian heritage. Finland is still sufficiently monolithic that they can get away with laws that require people to take Finnish names. That said, there are some systems that work and some that do not. Respect for the teaching profession seems to be a thing that is common to all the countries that do well. It is, however, unclear whether this is merely symptomatic of other things or if there is something deeper. Trying to higher the cheapest teachers just seems to be burning your seed corn.

    Report

  4. One thing that seems odd is that if corporate schools encouraged teaching to the test, why did they perform badly in tests? Shouldn’t the students be performing fantastically in tests (but still failing to obtain real knowledge and skills)

    Report

    • Probably, some percentage (the market-sorted good-schools/good-teachers/good-students) did very well on the test.

      The problem would be the percentage of bad-schools/bad-teachers/bad-students, who would perform worse because they were the low end of the incentive structure.

      Report

  5. The kind of PBP you’re describing is very obviously the wrong way to do it. Merit pay should be based on added value, not on the quality of the students assigned to the teacher. Ideally you would create a model to predict students’ test scores based on demographics, family income, their past test scores, etc., and then base merit pay on deviation from their students’ predicted scores.

    Report

    • Or you might just give schools as a whole more money based on how much value they add to the student on average. It can be difficult to measure outcomes over such a short time period (of over 1 year). Let the principal manage her students as she sees fit. Large numbers over a whole school over their whole secondary education (4-5 years) provide better statistical power (to calculate value added-ness) than the numbers you get over the few classes a given teacher would take.

      Basically, I’m betting that the data over the smaller sample size and shorter time horizons is too noisy to be useful. Over longer time horizons and larger sample sizes, the question of how much students get out of a school becomes clearer.

      Report

    • This was also my initial reaction: why not reward teachers based on the change in scores rather than the mean? But given that PBP has been shown to fail, the onus is now on the person designing the metric to demonstrate that similar capture will not occur. Off the top of my head, I can think of a few ways one could skew an improvement metric by e.g. starting with difficult assignments and moving towards easier ones. A prediction model would alleviate some of that, but do you honestly think something like that could be implemented?

      Report

      • To me, most of the value in the metrics is to make note of consistently high scorers and the consistently low ones (accounting for demographics of course). See what we can learn from the former, if anything, and take action on the latter. For teacher pay, maybe use it in conjunction with other things as a minor part of the whole (unless an outlier).

        It just strikes me as potentially perilous to do much more than that.

        Report

        • See what we can learn from the former, if anything,

          “Have involved parents”.

          That’s assuming you’ve accounted for poverty, medical issues, and learning disabilities.

          Teacher skills is quite important, but what people don’t really like to think about is that students are not identical blank slates. What a teacher can do with a student most often depends on the quality of that student, and the quality of that student is generally set before he or she walks through the door for their first day of kindergarten.

          Just as a similar anecdote: My wife’s a teacher, and last year (and probably again this year) she will have a classroom of juniors and seniors who have not passed the standardized (required for graduation) tests for her subject. Her pass rate was close to 85% — these are students who had been failing these tests for years. Now the overall pass rate for her subject is like 98% or so. She’s taking four years of that 2% and pushing them forward in a single year. Most forms of merit pay would see her get marked down for her low pass rate.

          Most good teachers do something like that. They take troublemakers, discipline problems, kids with issues at home. It’s the competent but not exceptional teachers that get the classrooms full of solid kids that respond well to standard techniques.

          There’s forms of ‘value added’ metrics that track that, but they’re expensive as all get-out, often subjective, and hard to use. It certainly doesn’t fit onto a power-point slide and can’t be sold to voters and honestly is rarely worth the price — because 90% of what you find is “crappy home life = crappy education results” and the other 10% could have been done by having people observe the class.

          Report

      • I could definitely get behind PBP based on a change in individual student scores from year to year. As it is often done now, we take an entire class scores as the measurement for whether a teacher is doing well. Thus, the results vary yearly.

        Report

        • Would it be better if it was based on the performance climb of the whole school? In other words, encourage teachers to help with difficult students in ways that the teacher benefits the entire school?

          Report

          • Yeah, it would be better if they used the whole school. At the class and individual levels, it would be too noisy. There are so many other factors which could influence the performance. If the school as a whole is doing better than expected, then it might be doing something well.

            Report

      • Here is a crude way to implement it (I think Singapore uses something like this, but I haven’t been able to find anything that relates to how they measure value-added-ness):

        Institute a nationwide (or statewide if you want to be all federalist about this) standardised test at the end of elementary school. Follow each student and see how they do in their SATs. You can now draw a scatterplot correlating the elementary school leaving examination and the SAT score of each student. (There is likely to be a trend). Thus, for a given student who is just entering high school, there is a certain score he can be expected to get on the SATs. If in some school, the students consistently outperform their expected score by some significant margin, then that school is doing something right. Give more money to that school and let the principle decide how to spend it.

        Report

        • I think this is useful as a diagnostic, but I see a few potential pitfalls:

          * Accurately quantifying the relationship between socioeconomics and elementary -> high-school scores. A low-scoring student in a wealthy neighborhood is not the same as a low-scoring student in a poor one; likewise for parental involvement, race, etc. Who do you train the model on? How do you design a model that will capture non-linear relationships and do you really think you’ll be able to convince voters and teachers to then use it.

          * School districts gaming the system by investing vastly more in higher education than lower education to artificially deflate the initial scores.

          * Penalizing schools that prioritize lowering variance (i.e. everyone graduates by there’s no gifted program) over increasing the mean.

          I think the main flaws with trying to apply market-based solutions to schools is that you have a mix of (a) bad teachers that are doing badly because they’re bad teachers and (b) good teachers that are doing badly because they don’t have enough money. So you need a metric that rewards good bad teachers without rewarding bad bad teachers. On top of this you have huge variability in student aptitude that’s beyond the teacher’s control.

          Report

      • I’m curious about the logic model behind the PRP concept. The symptom is that students are not performing terribly well, and not learning as much as we wish. The diagnosis is that this reflects teachers not trying enough to help those students learn. The cure is new pay structures that incentivize teachers more to improve student outcomes, with some debate over which metrics best capture that.

        My question is the diagnosis: that students not performing well is because teachers are not motivated to try hard enough to teach well. Is there any evidence that teacher motivation has correctly been diagnosed as the issue? I’m cautious to endorse a cure based on that logic model without better evidence that that’s the core challenge requiring a response.

        Report

        • This is a fair point, it’s entirely possible that our resources should be spent on some other root cause that has nothing to do with teachers and schools. Schools happen to be deeply intertwined with government in a way that encourages tinkering, but that is not justification alone especially given the relatively uneven success of such tinkering. That said, I think it’s generally important for public sector work to have performance-based metrics because the invisible hand is not as strong. This is especially true for education where many people argue that poor performance is an indication that *more* resources are needed; the typical market incentives are totally flipped. Even if PBP is not itself a panacea, it’s an important part of convincing voters to give teachers the money they say they need to improve in other ways.

          Report

          • There’s a pattern here:

            Deal with child poverty through the school system because that’s where government consistently interacts with children.

            Same with dealing with the rape crisis through college campuses. There is not doubt that more women are assaulted off campus then on; but again, this is where the biggest cluster of women are interfacing with an institution consistently, and one that sets social standards and influences culture. Determining the acceptable standards of women who are able to participate in college — the whole Title IX point — will become more of the accepted social standard of behavior.

            Sadly, for some, the common interfaces are the civil and criminal justice system and prisons.

            Report

            • I’ve said before that I wonder if, at times, it would be worth the additional cost to have some schools run a bit more like boarding schools, where the kids stay at school 24-5 (home on the weekends, maybe)? I mean, if the home life is not involved, or is actively harmful, perhaps the kids should not be there so much.

              Report

              • It absolutely would.

                Some kids need to get out of their homes and neighborhoods. Others have parents who’s lively-hood depends on travel (military, sales, trucking, construction jobs etc.), and so some of the stability of a two-parent family is missing.

                Report

          • Please explain this more; it’s an interesting critique. I read it as wondering if the lack of profit motive in teaching (particularly at the low, inner-city and rural ends,) create haphazard, low-quality teaching? I’d say it goes hand-in-hand with the lack of professional respect.

            Report

            • I’m working off memory, but my takeaway was that the modest salary increases were insufficient to motivate the unmotivated. My response was that you are better off just hiring more motivated people from the getgo with a higher salary than trying to incentivize the folks at the bottom of the barrel.

              Report

          • That should be entirely unsurprising. People are motivated by money of course, but it is far from the only motivator. People who think more or less money is the best and only motivator are projecting their own views. Most people have other and often more powerful motivators then cash money. The view that giving financial incentives will improve schools is itself pretty questionable.

            Report

            • Given the low pay and crappy benefits of being a teacher (I’m speaking of Texas here, where the pay is indeed very low and the benefits so bad that you’re better off working as an administrative assistant, because you’d take home more money a year AND have an actual health care plan) the reason pay doesn’t heavily motivate teachers is because you don’t get into teaching for the money.

              Teachers would LOVE to be paid more, don’t get me wrong, and it’d attract more people — my wife couldn’t afford to teach until we got married (the aforementioned crappy healthcare and low pay). She’s not in teaching to get rich, but she does actually need to be able to live off it.

              Unfortunately people often take this as a reason to pay teachers crap (and private schools do this quite often too), paying peanuts and saying “You should be in it for the kids, not a paycheck!”.

              Report

              • It’s often used as a hammer against teachers.

                “Negotiating for higher pay? What about your passion? Don’t you love the kids?”

                Passion and love don’t put food on the table.

                Report

  6. Regarding PBP, are you opposed to the particular system as outlined (specifically, the reliance on test scores) or any system that seeks to reward teachers for effort and success?

    I take the former position (for many of the exact reasons offered) but not the latter. I do think schools should have the ability to reward teachers who excel at their job. The difficulty is in identifying those who do that. I sincerely believe that a talented administrator who has an active presence and close relationship with his faculty, study body, and parents can figure out which teachers are really bring extra value to the table.

    No system will be fool proof. The more subjective a system is, there greater chance there is for bias (conscious or otherwise) or other factors to creep in. However, I do not think there yet exists an objective system for determining teacher quality. And yet, not all teachers are created equal and treating them as if they are — as if they are cogs in a machine easily replaced with one another — can be very harmful to morale and can disincentivize teachers who can provide additional value to their students or school from doing so.

    If one teacher is constantly offering support via collaborative meetings, has a strong relationship with students and parents that allow him to reach his constituents more deeply, gives up additional time and energy to see his students through to success, and models professional behaviors for others, I see no reason not to pay him more than the teacher who punches the clock, requires more support during collaborative meetings, and meets but does not exceed expectations.

    Report

    • I agree with your assessment of the difficulty of assessing teachers. I also agree that an honest disinterested administrator can tell you.

      For that matter, the kids know, at least at the older levels. Ask a bright kid who is genuinely interested in learning, and you will get a pretty good assessment. The problem is that there are any number of other kids with different priorities, which lead to perverse incentives if you listen to their assessments.

      The problem is that just as asking the kids risks skewed answers in response to perverse incentives, so it is with administrators. Some administrators correspond to the bright kid interested in learning. Others correspond to the various other groups of kids: slackers interested in sliding by with the least effort, bullies interested in power games, etc.

      So if the answer to how to assess teachers is to have good administrators, this merely pushes the question back a step: how to assess administrators?

      Report

    • Kazzy:
      If one teacher is constantly offering support via collaborative meetings, has a strong relationship with students and parents that allow him to reach his constituents more deeply, gives up additional time and energy to see his students through to success, and models professional behaviors for others, I see no reason not to pay him more than the teacher who punches the clock, requires more support during collaborative meetings, and meets but does not exceed expectations.

      I agree completely. I don’t believe the alternative to the way PBP has been implemented is to give up any form of compensation based on performance. The hypothetical teacher you just described deserves praise (and pay in my opinion) for the work you describe. Obviously, you will have teachers doing just those things in low performing schools that would likely not receive PBP using the models discussed in my piece.

      Report

      • Then we are in agreement! I abhor test-based teacher assessment (and, really, standardized test-based student assessment… At least in the younger grades…). Some systems I’ve seen have at least moved toward measuring progress as opposed to absolute scores… Which is better but, yea, still awful.

        Report

  7. Your description of the Finnish system (which I don’t know anything about beyond the accolades heaped upon it) is exactly what I think we need to look at, though I think it would require a grand cultural shift that can’t happen simply via legislation.

    I do no and have never taught in the public system. This is primarily because my preferred age range (PreK) has not consistently been a part of the public system and its recent adoption in my area (NY) follows the general public system of top down micromanagement. And I can speak to this personally despite being an “outsider” because I am part of the group that trains folks teaching in these programs and I can see the myriad hoops they have to jump through. And while many of these hoops are genuinely intended to promote best practice, I would personally find the boxes I need to check stifling and ultimately undermining my ability to deliver a quality education.

    And yet, I think we currently need those hoops. Why? Because starting salary for those positions is currently $40K and most of these teachers are being thrown into very demanding situations which they are simply unqualified for.

    Working in independent schools, all of my bosses have more or less taken an approach in which they said, “We hired you to be the expert on PreK. Go to work.” I wasn’t given complete autonomy but I was given a great deal of it. This would be very unlikely to happen in public schools because most of our current crop of teachers would flounder without the external structure they are given from the top. And that is because we do not do what you describe Finland as doing in terms of recruiting top minds and great talents to work in our schools. There is no doubt a great number of amazing teachers in our public system, but many of them are hamstrung by all the “rules”. And far, far too many of our teachers are simply there because we set a low bar and then agreed to hold their hand throughout the process.

    Unfortunately, as I said at the onset, I don’t know how we change this until we change the very idea of who and what a teacher is. There are many folks who genuinely believe the old adage that those who cannot do, teach. We need to change the culture around education and educators. And, to be honest, I think much of this needs to come from teachers themselves. Strong, talented teachers need to stand up and say, “I don’t want to be lumped in with these folks any more. I want higher standards, for myself and them.” This is why I had loose support for Michelle Rhee’s efforts… or at least the ideas behind them. I am in no way opposed to unions but by giving teachers the opportunity to excel and be rewarded for that (see my above comment), we could have begun to potentially shift the paradigm.

    Report

  8. First, a disclaimer: I am really strongly in the School Choice camp. Not a radical in the sense that I want to remake our entire education system into a market enterprise, but in the sense that I am rather unyielding in my opposition to the notion “If you can’t homeschool and cannot afford private school, you will go to this school over here and if it fails well tough.” I’m open to means-testing, or applying choice only to students of public schools that underperform, as well as an entire menu of regulations that would make it very much not a “free market” in the general sense. I also consider the Finnish model completely non-viable as applied to the USA (though at a state level, let’s see what happens) and I consider arguments that we have to fix poverty before we can fix schools or a top-down “We’ll be like the Finns” or “Unless we’re talking about spending more more more the system is good enough as is” to be somethings of a dodge (albeit not always an intentional one). Like I said, rather unyielding on this, and deeply mistrustful of the other side, which is why I don’t talk too much about it (I prefer issues on which I have more of an open mind).

    Also, I don’t think your characterization of the National Review article is particular fair. Her argument isn’t that “This isn’t a true free market system and a true free market system would work” (as was the argument with “The Soviet Union isn’t real communism”) but rather an implicit agreement that regulation is very necessary. I also believe that article, as well as this one actually do a pretty good job of undercutting the argument that the problem in Sweden is the profiteers. In a system where only 14% of students actually attend private schools, I don’t think we can really blame the private schools. Nor, if Coulson is correct, can we blame the private schools for the deteriorating function of the public schools (by drawing away their best students, for example) because the public schools that compete with private schools seem to be doing well compared to those that don’t.

    Having said all that, the combination of the Swedish experience as well as our own experience with “school choice” at the collegiate level do give reason for pause. Not pause from going forward, but being careful about how we do so. Which is why I look askance at what Nevada’s doing right now. Which is an advantage to having a decentralized system. We can try different things in different places, and in some places nothing at all. And school choice, for me, allows for a degree of experimentation (and teacher autonomy) that makes me more comfortable when the students are there by choice (albeit their parents) rather than strictly geography. And absent school choice, I am drawn to mechanical methodologies and metrics that people who oppose school choice also tend to oppose (and that I myself view as the lesser of evils rather than a positive good).

    Report

    • First, a disclaimer: I am really strongly in the School Choice camp. Not a radical in the sense that I want to remake our entire education system into a market enterprise, but in the sense that I am rather unyielding in my opposition to the notion “If you can’t homeschool and cannot afford private school, you will go to this school over here and if it fails well tough.”

      This is a legitimate complaint; but the answer isn’t simply to allow flight from a school without actually working to address the problems that cause people to flee.

      Personally, the whole framing bothers me; thinking of schools as good/bad versions of the same learning expectations seems rather narrow. There are other ways to present choice in education.

      I’d like to see regional school networks where different schools focus on different things — arts at one, industrial arts at another, drama, music, foreign languages, winter sports, agriculture, math and science, literary, business, etc. — particularly at the high school level. Had I my druthers, I’d encourage merging of the high-school/community college systems in each state. Both of these things, it seems to me, are more expressions of the Finnish system of strong local control and direction. Select a school because of some program you value, not because the neighborhood school sucks or you might rub shoulders with someone your mommy and daddy don’t like.

      The right to opt-out of public schools produces some weird stuff. I’ve already expressed my concerns that a lot of Christian home-schooling might be abusive and deny women even knowledge of their civil rights and (potentially) train men to be abusive of women’s civil rights; the option to opt-out is only presented in terms of eternal damnation and not contrasted with existing legal structure (which includes the right to opt in). White flight is a well-studied problem. I’m not a fan of the details of No Child Left Behind, but the name’s certainly catchy.

      Report

      • I’d like to see regional school networks where different schools focus on different things — arts at one, industrial arts at another, drama, music, foreign languages, winter sports, agriculture, math and science, literary, business, etc. — particularly at the high school level.

        I’m curious about your thoughts on how to make transportation work for that. I could see it working in a large urban area with good safe public transportation. Making it work in the suburbs is much more difficult. And in some rural areas, any kind of choice gets very difficult to implement.

        Colorado’s constitution requires the state to guarantee a uniform system of public education. To meet that, more and more of the K-12 funding comes through the state’s General Fund in order to offset shortfalls in property taxes in poor and rural areas. An urban (and possibly suburban) system of specialized schools providing better opportunities in different fields seems like an invitation to ongoing lawsuits here.

        Report

        • We have a somewhat limited regional system now; they run it with smaller busses; and students attend two schools — their regular high school and a technical school. Their regular core requirements in math, English, science, and social studies at the local high school. At the technical schools, kids can study careers like truck driving, nursing/home-health/medi-tech, culinary arts and hotel management, machine tool and automotives, forestry.

          I’d like to see it expanded to include art and sport programs.

          Report

          • That’s not necessarily true in many rural places, though it might be true for really sparse places.

            Here, it’s a regional effort; and towns within a 50 to 75-mile radius participate.

            Report

        • Sacramento has this (SacUnified) at the high school level, but transportation is handled at the “it’s up to you” level. Sac also has quite a few charter schools, giving even more choice to families.

          Report

    • I actually agree with your instinct towards giving some choice to families and students. Having taught in a bad school for a number of years, you saw a number of good students and caring parents that were doing everything in their power to make sure their kids received a decent education, even if the school spent most of its time dealing with crime and order. If there was an alternative in that district, those parents who cared would have taken their students out of the public school in a heartbeat.

      Now that I am a parent as well, I would have done the same.

      Thus, I am constantly considering what type of choice, and yes competition, can be used in our education system to better allow students to find a place that works for them. I have yet to find a model that works.

      Report

    • I think the fact that our response to “failing” schools (based solely on test scores) being a full shut down (and sometimes replacement with charter(s)) tells us much about what is wrong with our system. School choice (as outlines it) would be less of a need if we understood why schools fail and work to fix them.

      I’d rather choice exist around learning needs and teaching models.

      Report

      • I am writing a piece in my mind now about how most conversations about education are actually conversations about race/culture. But at the moment, I don’t have any specific data on the matter. Mostly conjecture and observation at this point.

        Report

        • I failed to include a comment about lacking the political will to fix failing schools because of the types of people whose children tend to attend them. But… Yes. Is it about race, insofar as we treat kids of different races vastly different in our school system, both individually and collectively.

          Report

        • I’d like a little more pushback on the “failure” of our public schools. Colleges and the workforce are full of the graduates of public schools.

          The vast majority of which are doing just fine, thank you.

          Those that aren’t — well, given how schools are run almost entirely by the state and local governments, how on earth could there be one “cause” unless that cause was something like “poverty”? Yeah, Common Core is the current buzzword, but anyone who has ever worked in education knows that even two districts side-by-side are often run entirely different, and state curriculum and testing standards vary widely.

          Report

          • Fwiw, I’m referring to individual schools that the state defines as failing (some of which undoubtedly are and some of which are not). The way NY defines it is so many years of test scores below a certain threshold. I consider that a pisspoor measure.

            Report

            • Big local layoffs, recession — lots of things can produce a year filled with students who are struggling more than normal, which can drop test scores and get you labelled failing.

              Then the way demographics are sliced and diced — it’s entirely possible that a weird demographic outlier might result in your school’s status depending on a specific handful of kids.

              Report

      • I’d rather choice exist around learning needs and teaching models.

        That’s definitely where I would like to see things end up. It’s one of the reason that I find the phrase “competition” (mentioned by Roland above, but introduce by school choice proponents) off-putting. It doesn’t need to be a competition where the “best model” wins because I have my doubts that there is a “best model.”

        In the meantime, though, allowing parents of kids who got to schools that perform poorly seems like a reasonable compromise between universal school choice and no school choice. Ideally I’d prefer see it used in an advisory capacity (parents can look up the scores and decide whether or not they matter), but that might be expecting too much of parents (as much as I hate to say that) so I would use it mostly as a floor for both assigned and charter schools. And charter take-overs seem to be the main (maybe only) way to do an institutional reboot and/or cleaning of house. This seemed to be the case with the local high school where I lived in Estacado.

        Report

        • Perhaps not “competition” as defined by a corporate structure, but actually having schools compare their teaching philosophies with qualitative and quantitative data they believe fair in a way that lets the better programs rise. This all comes back to our measurement tools for teachers/schools, and that is a difficult process. As noted in this conversation, it may not be that there is the “right” school for every student. But students who share similar issues would benefit from a closer collaboration between schools in similar situations.

          This does happen on a limited scale, but it often fails for reasons that really require a whole other post….

          Report

      • Kazzy:
        School choice (as Truman outlines it) would be less of a need if we understood why schools fail and work to fix them.

        Incidentally, has anyone on this thread listened to this week’s episode of This American Life? It’s the first of a two-parter which posits that desegregation is actually one of the best tools for dropping the achievement gap between black students and white students, but that we stopped doing it because it was too hard. It’s not exactly on point here since it’s primarily concerned with race, but I suspect that if its insights are correct, the lessons might be more broadly applicable to “bad” schools vs “good” schools.

        Report

        • Yeah. I’ve listened to it and was considering writing about it. I think desegregation needs to be tried more. Racism is an obviously problem as the town hall clip made clear. Though I think a more pressing but related issue is the distance/geography one. The narrator said she needed to do a four-hour commute to school everyday. I think this is a cruel thing to do especially to young children. Maybe it is more okay for high school students but still rough. Perhaps we would just need to do it though.

          Report

  9. The actual (limited) math stats that are presented in the Guardian pieces don’t paint quite a dire quantitative decline as the Guardian’s breathless qualitative assessment of the decline.

    Report

    • There is a lot going on in Sweden that trend towards unrest in the social order. Much like in the US, when people talk about education, I feel they are actually discussing a number of fears regarding race/changes in the economy/fear of change/etc. The comments from the pieces notes all trend towards similar comments I have heard from American parents. There is this overwhelming sense that things are changing for the worse (whether its true or not) and they are beginning to flail about to find some solution in their schools, without talking about those aforementioned underlying fears.

      Report

  10. Pingback: Sweden’s Education Privatization Failure | in hope and darkness

  11. Why are you using a 19th century Greek nationalist propaganda painting about the days of the Turkish occupation for an article about school privatization in Sweden? The painting used is called the Hidden School and it depicts a Greek Orthodox priest teaching Greek children how to read and write Greek in secret while under the careful guard of a watchmen. The entire point was to depict how the Turks suppressed Greek culture during the centuries of Ottoman occupation, which isn’t actually true.

    Report

  12. Why did Sweden experiment with school choice/privatization/vouchers in the first place? This is what I am trying to figure out and there does not seem to be that much information.

    Report

  13. I think it’s hard to talk about why Finland has had success without reading this article by Pasi Sahlberg:

    http://www.washingtonpost.com/blogs/answer-sheet/wp/2013/05/15/what-if-finlands-great-teachers-taught-in-u-s-schools-not-what-you-think/

    I find his conclusions a lot more compelling than proposals for school choice or merit pay or making the education model more corporate. Finland’s education system is good because it doesn’t face the same socio-economic problems (some) US school districts have. Here we expect teachers to solve problems our entire government struggles with.

    Report

  14. Pingback: Angry Bear » Swedish Privatization of Education Fails

  15. No discussion like this should leave out Massachusetts where minority kids score as well on PISA as do kids in Finland. (Granted, white kids in MA do better, but ….) The magic of Massachusetts is the “cherry sheets” which tell each township how much education equalization money they are getting so that education spending isn’t completely linked to property taxes as it is in so many other places. Mike the Mad Biologist (http://mikethemadbiologist.com/) is always foaming on about this, and he’s right. Better education across the board does not require Nordic or east Asia solidarity. It just requires a better attitude and more money.

    Report

  16. Pingback: Links and Tweets for the Week of August 3, 2015

Comments are closed.