• Uncategorized

Experimentation and Policy


James K

James is a government policy analyst, and lives in Wellington, New Zealand. His interests including wargaming, computer gaming (especially RPGs and strategy games), Dungeons & Dragons and scepticism. No part of any of his posts or comments should be construed as the position of any part of the New Zealand government, or indeed any agency he may be associated with.

Related Post Roulette

53 Responses

  1. Avatar Alan Scott says:

    The problem is, when 80% of your ideas for making toasters fail, then you just shrug your shoulders and go with the ones that work. When 80% of your ideas about education fail, can you really justify subjecting children to them? It’s okay to have some broken toasters in the trash pile. It’s not okay to have broken kids.Report

    • Avatar Murali says:

      When 80% of your ideas about education fail, can you really justify subjecting children to them?

      When the status quo looks like 100% failure rate. Improving your success rate to 20% is a good idea.

      One thing Singapore sometimes does is try something out with one school and see what happens then.

      Since there are about maybe 400 students in a single cohort, it provides a decent samoke size. Yet, if things don’t improve very much, there are 100s of other schools which chug along. Also, since there is ofen frequent testing in many schools, (most schools have at least 4 tests a year) performance is monitored fairly closely.

      One thing the ministry of education in Singapore has done is to develop the notion of value added-ness. Given previous performance on a nationwide standardised test (say the Primary School Leaving Examination), a baseline of what the expected performance in a future standardized test ( for example, the GCE N, O or A levesl) is constructed. Schools’ which tend to add more value to the students are noted (this is true because My secondary school received the people’s developer standard among other awars), and their practices are emulated (hopefully. I mean it would be great to do so and there wouldnt be any point to giving us a nice award if the best practices were then not shared)Report

      • Avatar Freddie says:

        When the status quo looks like 100% failure rate. Improving your success rate to 20% is a good idea.

        This is absolutely absurd. The vast majority of American schools produce educational outcomes that are highly competitive with global competitors when correcting for SES status and racial composition of classrooms. Our public school problems come from a numerically small number of dismal performers.

        The problem, James, is shit like this. We do know many things about education, actually. For example: school vouchers don’t work. Sorry. But libertarian ideologues are too invested in the idea that they do to be willing to respond to that failure. Your entire thesis is undermined by libertarian fixation on school vouchers. Been tried. Failed. Advocacy continues.Report

        • Avatar Jaybird says:

          For example: school vouchers don’t work.

          It depends on how much of a premium one puts on “choice”. If you see “choice” as a positive good in its own right, school vouchers do, in fact, work.

          If you don’t, they don’t.Report

          • Avatar Freddie says:


            Incidentally, the endgame for libertarians is less choice, because vouchers are the first salvo in destroying public education and government schools.Report

            • Avatar Jaybird says:

              While I appreciate being told what the “endgame” is for Libertarians, I will, again, point out that if you see “choice” as a positive good in its own right, vouchers work.

              If you don’t, they don’t.

              Here, let’s look at each of the articles you linked to in your essay and examine what they say.

              Your first link says “Here’s bad news from Ohio”. Open up the article and see: “Voucher results mixed”.

              Mixed? I thought it was bad! Get into the article and you see such things as: “On the whole, Ohio students who used tax-funded vouchers to attend private schools last school year did no better on state tests than public-school students.”

              Did no better? Why didn’t they say “they did worse”?

              Then you read the article and you see that the results were mixed. Here’s a couple of paragraphs from the middle:

              That was true in the Columbus City Schools, too, where district students outperformed voucher students on seven of 12 standardized tests. The public-school students bested their private-school counterparts most often in math.

              But middle-school voucher students had better passing rates, particularly on last school year’s state reading exams. Eighth-graders outscored Columbus district eighth-graders on all three tested subjects – math, reading and science.

              Huh. Mixed.

              I guess it depends on how much of a positive good you see “choice” as being.

              Let’s look at your next article. “Here’s bad news from Milwaukee.”


              Nah, I’m just kidding. (I see choice as a positive good, you see.)

              Let’s go down to page 10. (I think it’s 10. There’s a 10 in the upper left hand corner.)

              Of 42 statistical comparisons made between similar MPCP and MPS students, no statistically
              significant differences in student achievement growth were reported in 36 cases (86%).

              Huh. Let’s see where the statistically significant difference were and what they were.

              3. Three statistically significant differences in achievement growth favored the sample of MPCP
              students. All three involved the sample of seventh graders in 2008, who demonstrated
              significantly higher growth in math achievement if they were in the MPCP.
              4. Three statistically significant differences in achievement growth favored the matched sample
              of MPS students. Two of those advantages involved achievement growth that was higher than
              MPCP students after one year but comparable to them after two years. The third statistically
              significant result favoring the MPS students was an additional overall gain of 3.4 scale score
              points in math after two years. This estimate came from a regression model that included a
              control variable for the effect of school-switching run only on the subgroup of students who
              remained in their original school sector from 2006 to 2008.

              Golly. Is there a conclusion?

              In sum, the evidence in the LEGS report suggests that students in the Choice program generally are
              experiencing achievement growth rates that are comparable to similar MPS students.

              That’s *BAD* news?

              Well, there’s another report. Maybe you were referring to that one.

              Report 28! The summary!

              Let’s look at page 3 and see “Major Findings To Date”.

              The MPCP remains popular among Milwaukee families, as evidenced by consistent and at times dramatic growth in MPCP enrollments over the past 12 years.


              Students in the MPCP appear to be performing at lower levels than MPS students in the younger grades
              but somewhat higher levels than MPS students in the older grades. When similar MPCP and MPS students are tracked carefully over time, however, their rates of achievement growth are statistically
              similar after three years.


              MPS students themselves are performing at somewhat higher levels as a result of competitive
              pressure from the school voucher program.


              Independent public charter schools are generating signi cantly higher rates of achievement growth
              for their students compared to similar students in MPS.


              (I ask the readers of this comment to read the PDFs themselves to read the stuff that I didn’t exerpt and confirm for themselves whether I left out damning evidence and engaged in selective quotation to make the “bad news” look like “not bad news”. I don’t think I did.)

              So, so far the two articles you pointed to that you called “bad news” strike me, at worst, as “mixed news”.

              Your next link is, by your own admission, “controversial”. Given the last two articles that you gave, I’m going to assume that this means that it was fairly successful. Let’s see for ourselves.

              Overall, low-income students who were awarded vouchers to attend private schools through the DC Opportunity Scholarship Program (OSP) were performing at similar levels in reading and math 4 to 5 years later as students who also applied to the OSP, but were not awarded scholarships. However, students awarded vouchers (and old enough to have graduated from high school) graduated at significantly higher rates than did their counterparts, according to parent reports. The OSP also had a positive impact on parents’ satisfaction with their child’s school and their perceptions of the school’s safety, but students themselves rated school satisfaction and safety the same whether they received a voucher or not.

              Yeah, I pretty much called it.

              If you see the point of vouchers as solely being making sure that students out-achieve students in public schools, then results are, as has been pointed out, mixed.

              If, however, you see choice as a positive good in and of itself, then vouchers work.

              Your definition of “bad news” is not mine.Report

              • Avatar Jesse Ewiak says:

                Choice is fine. Don’t ask me to pay my taxes for your choice. I want to pay taxes for well-funded public institutions with teachers that are well compensated with unionzied protections. If you want to send your kid to a for-profit charter school or a religious school, great. Pay for it or get a scholarship.Report

              • Avatar Jaybird says:

                Don’t ask me to pay my taxes for your choice.

                Welcome to my world.

                Oh, wait. Having kids isn’t a choice, is it. It’s a part of society and we, as a society, have decided that it’s better to provide something something to prevent the underclass from burning our houses.

                Which means that you are asking me to pay taxes for your choice of the institution of education paying for union protection for teachers. No, wait, forcing.

                And when I try to talk about alternatives suddenly “choice” becomes something that you don’t want to have to pay for.Report

            • Avatar James K says:

              The trouble when you say things like that Freddie is that it sounds like you care more about who owns the schools than how well they perform.

              I’m a big fan of specificity, so rather than talking at the high level, lets get down to details a bit more. What is it about government schools that you desire? You can’t run a good experiment without defining your success criteria, so by all means let’s evaluate any future experiments (and not just with vouchers, I’d be really interested to see what a school run by a teachers’ co-op could do) against the criteria that matter to you and other government schooling advocates.

              My point in writing this piece was to go beyond the status quo vs. vouchers debate in education. Education is an extremely complicated good and there are any number of variables one could alter about it. I believe there are answers out there, but to find them we actually have to look. My point was that if privately operated providers are beyond the pale, then government will have to do that experimentation itself.Report

              • Avatar E.D. Kain says:

                This makes sense, James. I think the dividing lines between vouchers and traditional public schools, edu-traditionalists and “corporate reformers” are all sort of silly and bizarre. Here in my hometown the charter movement is dominated by progressives, for instance, making use of the very libertarian approach to school choice that Arizona has taken. The only way we’ll know what works is through experimentation – school itself, education itself, is just part of a long evolutionary experiment.

                Freddie – I just don’t think it’s useful, at a certain point, to cast the sides in such stark terms. Obviously many schools in America do just fine – but what about the ones that don’t? There’s more to that than simply poverty, and even though poverty is a driving factor, well poverty isn’t going anywhere any time soon. Why not experiment?

                But yes, to the point, a good discussion of the nitty gritty is in order.Report

        • Avatar Murali says:

          This is absolutely absurd. The vast majority of American schools produce educational outcomes that are highly competitive with global competitors when correcting for SES status and racial composition of classrooms. Our public school problems come from a numerically small number of dismal performers.

          I dont know. It certaomly doesnt compare to performace by Singapore and South Korea. Even so, I was talking about inner city schools, which certainly aren’t just a numerically small number. Inner city schools are failing inner city kids and I’m surprised that you think Freddie that they aren’t.

          Also, I wasn’t just talking about school vouchers, but was talking about experimenting within the public school system. Your problem Freddie, is that you jump too quickly to conclusions.

          Funding schools from local taxes is a terrible idea. (inner city kids need more help in order to compensate for bad environments at home and the neighbourhood)

          School vouchers do not spell the death knell for public education and government schools. Singapore has edusave accounts which pay for a lot of school fees and costs and is topped up by bursaries etc for lower income kids. Yet, more than 90% of the schools are government schools.

          I’m for standardised testing, but not american style standardised testing. The test should be set at a federal level (private schools have the option of getting tests designed by private third parties). I would recommend the international baccalaureate or the GCE as models to be used. The SATs (both I and II) I took seven years ago (I studied for them while I was in the army) were ridiculously easy.Report

          • Avatar Barry says:

            “Even so, I was talking about inner city schools, which certainly aren’t just a numerically small number. Inner city schools are failing inner city kids and I’m surprised that you think Freddie that they aren’t.”

            (1) Please don’t put words in peoples’ mouths, and (2) the major problem with inner city schools is poverty; it’s hard for a school to deal with that situation.Report

      • Avatar Alan Scott says:

        I’m not sure how you can reasonably suggest that all of our schools are failures, Murali. The vast majority of them serve their students well.

        My point about the 80% failure rate is that reforms need to be limited in scope. They should target schools that already perform poorly, and they should be implemented on a broader scale only when they have an established record of success.

        I’ll not go so far as to Freddie to say that charter schools are failures, but in my experience, they’re usually full of students who were well served by their existing schools.Report

        • Avatar Murali says:

          Refer to above, I was looking at inner city schools as I assumed that that was what James K was talking about. I should have been more explicit.Report

    • Avatar Mike Schilling says:

      Does “fail” mean “make things worse” or “not make things better”? If it’s the latter, than, sure, you try the ones that seem most promising because it beats the hell out of the alternative.Report

    • Avatar James K says:

      A couple of points:
      1) You’re assuming there are no broken kids now. That’s a fairly heroic assumption.
      2) There are three ways to approach experimentation on kids: A – never change anything ever. B – change thins without properly testing them C – changes things as part of a rigorous experiment. I suspect we can agree that C is preferable to B, but are you suggesting A is preferable to C? If so, are you at all uncomfortable with a norm that states the more important something is the less effort you should put into making it work better?Report

      • Avatar Alan Scott says:

        Really, I’m suggesting that A is preferable to B, at least when we’re talking about schools in general as opposed to schools which we know to be failing.

        I was subjected to a number of experimental programs throughout my k-12 education. Several of them were harmful to my education, and none seemed to help me better than a traditional classroom would have.

        And I don’t think I’m an aberration in that regard.Report

        • Avatar James K says:

          I can believe A is better than B, that’s a perfectly reasonable position. What I want is less B and more C.Report

          • Avatar Murali says:

            A is really scary as a position. If the case of Singapore and South Korea demonstrates that there are better ways of educating our young, American’s intransigience about changing their education system seems profoundly unreasonable.Report

            • Avatar Jaybird says:

              You think that because you think that the point of educating our young is to educate our young.

              Once you realize that the point of educating our young is protecting unions and providing pensions to the middle class, you’ll change your tune.Report

              • Avatar Murali says:

                Its kind of early for me and I am enough like Sheldon Cooper that I need sarcasm tags but….

                Once you realize that the point of educating our young is protecting unions and providing pensions to the middle class, you’ll change your tune.


              • Avatar Jaybird says:

                I see myself as a Leonard.

                One of the things that I encounter more often than I think is reasonable in any given discussion of education policy/reform is that the fundamental assumption of the point of education is to provide pensions and protect unions. It’s, like, students are an afterthought.

                In any discussion of education, watch: you will see people who immediately start talking about teachers and administrators and unions without talking about students and will *WAVE AWAY* discussions about students. In a discussion about education!

                Anyway, I was making an oblique reference/criticism to/of those folks.Report

              • Avatar Jeff says:

                I think it’s fair to say that no one legitimately participating in dialogue about education really believes that the purpose of public education is to support unions. You’ll see just as many people who pretend that “education” exists in some sort of platonic form, without teachers to do the hard work of educating. They treat the people who actually teach students as inconvenient road blocks at best, and malicious saboteurs at worst. At the end of the day, both sides are often guilty of using the fates of children as a cudgel in public debate: “If you really cared about children, you would [dissolve teacher’s unions|abolish tenure|triple funding for inner city schools|support vouchers|oppose vouchers|etc].

                I try not to base my arguments around the silliness of the hypothetical extremes — it breeds laziness, and tends to distract from real problem solving. Pointing out why ridiculous people are wrong is much easier, and much more fun, than trying to figure out how to improve systemic problems in public education.Report

              • Avatar Jaybird says:

                Pointing out why ridiculous people are wrong is much easier, and much more fun, than trying to figure out how to improve systemic problems in public education.

                My main argument (that, seriously, predates the comment you’re responding to!) is that we really ought to come out and say what the point of education is.

                What is the point of having an educational system? Is there an essential point where, if it is not achieving this goal, we can say, without qualification, that it is failing?

                It seems to me that we can.

                Additionally, there are issues of culture that we have to carve our way through. There are subcultures out there that act at odds with the point of education (that, granted, we haven’t yet defined for the sake of this particular argument… but I think you can grant me this point). To address these cultural issues would require a degree of Paternalism that I think would make most technocrats blanch.

                Additionally, there is the problem of the top third of schools really serving its students well, the middle third of schools doing an adequate job of serving its students, and the bottom third of schools doing an absolutely atrocious job… and all of us here today went to schools one of the top two thirds and none of us went to the schools in the bottom third… so when we think about “school reform”, odds are, our thoughts are drawn to schools that fall somewhere on the spectrum between Not Bad and Pretty Good so discussions of reform feels like an attack on the schools that work and do a good job than on the schools that are failing their students (no pun intended).

                Folks generally don’t like talking about that, though.

                It’s easier to talk about people who hate unions and how they want to destroy our schools.Report

      • Avatar Larry says:

        I think your point 2, James, undermines the more general point you make in the post itself. There you say:

        It seems that to a large extent the private sector is just trying things more or less at random – the differences between businesses and government is what happens next. Effective organisations learn from their failures – they try many things, work out what’s working and what isn’t, and shut down the failures. This combination of variation, selection and persistence is exactly what drives evolution.

        This is a useful observation, though it should be broadened to include individuals and households as well as business organizations. The general point is that entities engaged in a market are not pursuing a policy of deliberate or conscious experimentation, and certainly not a “rigorous” one (not generally, though of course experiments of all kinds can be an aspect of their behavior). They are, as you said originally, “just trying things more or less at random”. That’s already different from government, which imposes things, more or less monolithically. The whole secret of market success lies not in its experimentation but simply in its enormous diversity, and any notion that such diversity can be emulated by “scientific” experiment by the state falls into the technocratic illusion. That’s an illusion, by the way, with some horrifying history, and is a reason people are right to be concerned about references to state experimentation upon children.Report

        • Avatar James K says:

          I’m not suggesting markets can be replaced by experimentation in their entirety, I still like markets a lot. But experimentation can produce some of the benefits produced by markets.Report

          • Avatar Larry says:

            I’ll agree that at least trials are an improvement over the more common practice of simply imposing a policy on an entire society. But, in addition to its inability to replicate the often surprising results of real market diversity, experimentation in areas of social/political dispute is always going to be plagued by political bias — not only on the part of the researchers themselves, but bias inherent in the institutions that decide which researchers to hire and to offer tenure, in the agencies that decided which research projects to fund, in the journals that filter which results to publish, and in the peers that advise those journals. All too often, then, as a result of this bias, “experiments” in these areas become merely a cover for long-standing political agendas. It’s not to say that such experiment is worthless, but that people looking for honest guidance from such research, policy-makers and citizens alike, need to be especially alert and critical when looking at results that appear to support or flatter their own political bent.Report

            • Avatar tom van dyke says:

              +1, Larry. How many studies “prove” liberals are smarter, better, faster, better-looking, and kinder to puppies and children? Ugh. Who pays for this shit?

              [That last one was a rhetorical question.]Report

  2. Avatar tom van dyke says:

    JamesK, Milton Friedman’s pencil and the sign in Clarence Thomas’ office anticipated all this.

    I appreciate your essay, but have great trepidation about experimenting or our children’s lives. Neither am I convinced that the “old way,” phonics, learning to read and do math, etc., even “rote learning,” were ever proven to be insufficient to the task of educating the young skulls full of mush. 😉

    I think it was the teachers who got bored, not the students. Students are always bored: Getting an education is work, not play. It’s a job, as is educating.

    As far as Google having a 20% success rate in trying to innovate, that’s much higher than I would have predicted or expected. Props.Report

    • Avatar James K says:

      Variation in preferences is a whole different issue, and if there is a sufficiently large variation in what people consider to be a good school then any form of government schooling is likely going to be inadequate.Report

    • Avatar Rufus F. says:

      Much as it pains me, I agree with Tom here. 😉
      Even more than bored teachers, the problem I see is that there’s an army of experts and consultants in most school districts trying to build a better educational experience and there is absolutely no incentive for trying something that’s worked in the past. The point is to ceaselessly innovate, since kids today are radically different from kids last year, allegedly, regardless of whether the innovations work or not. Just keep trying new things. In my experience, the “2.0 digital natives” still can handle the older methods and appreciate being pushed to do something hard that has a big payoff.

      I mean, it’s not much different from the corporate world paying ridiculous fees for “consultants” to keep cranking out bullshit year after year. Education has a lot of advocates of experimentation, a few of whom have even taught a class before.Report

    • Avatar Barry says:

      “As far as Google having a 20% success rate in trying to innovate, that’s much higher than I would have predicted or expected. Props.”

      I’d like to point out that Google is a cash-rich business; IMHO that’s the sort which seriously experiments, because they have the luxury (see Bell Labs, IBM research center, etc.).Report

  3. Avatar Larry says:

    The notion that markets make use of a kind of Darwinian natural selection is a good one (though it’s not in conflict with the idea of the far-sighted innovator, either). But the other key ingredient to market effectiveness is individual choice, which implies freedom to choose — that’s how we find out what works. And that’s precisely the ingredient that is in conflict with state policies, by their nature, which, at a minimum, rely on compulsory payments for their funding. This is to say, in other words, that governments are very different kinds of things from markets, and that often the best lesson state policy-makers can learn from markets is simply to get out of their way. The second best lesson, when it’s not so easy to follow the best, is not to experiment directly, but rather to offer people choice — with obvious implications for education policy.Report

    • Avatar Alan Scott says:

      The problem of choice is exactly why it’s so hard to improve upon our schools. They are almost uniquely unfit to respond to capitalist pressures.

      When I go to the store and buy a can of tomatoes, I choose the brand I want based on the price (because, after all, I’m paying for them), and the quality (I’m also eating them). The invisible hand is at work–if a brand is to expensive, I won’t buy it. if a brand is of poor quality, I might buy it once in ignorance, but I won’t buy it again.

      Contrast with education. With the tomatoes, I pay, choose, and consume. With a public k-12 education, the state pays and chooses, but the child consumes. With vouchers or improved school choice programs, the parents choose instead of the state, but the child still consumes.

      The basic alignment by which capitalism functions is fractured by these arrangements. The pressures that select towards the best quality at the best price don’t work. They can’t work. It’s a quality inherent to the education system as it exists today, and it’s inherent (in greater or lesser degrees) to every system that’s been proposed to replace it.Report

      • Avatar Larry says:

        Alan: With vouchers or improved school choice programs, the parents choose instead of the state, but the child still consumes.

        I suppose the logic of this, then, would suggest that children should make their own decisions about their schooling — they’re the consumers, after all. But, in case the absurdity of that isn’t self-evident, I’ll just say that these consumers, being by definition immature, aren’t competent to make such decisions. So your comment seems to express a kind of resignation or even, for many, despair that the education system could ever be improved. But that doesn’t follow either — parents, as the people closest to, and most intimately engaged with, these not-yet-mature consumers, are in general the people best able to make choices about the education of their children, and judge the results. Of course those choices won’t always be “correct” — that’s inherent in the nature of the evolutionary, adaptive process that choice and markets enable. But their aggregate outcome can show us the way toward improvement, in the process amending and defining the very meaning of “correct”.Report

      • Avatar Rufus F. says:

        Tomatoes is an interesting example. Canning tomatoes generally are bred to be flavorful without an emphasis on how strong they are. The tomatoes you buy in the grocery store are generally bred to be stronger for shipping than your average garden tomato, which will barely make it out of the garden without a few breaking, while flavor is of secondary importance. Even less than that usually. Compare a tomato you’ve grown in the garden with one grown on an industrial farm in the eastern states for flavor and see how well the invisible hand is maximizing quality. Quite often the grocery store tomato is nearly flavorless. The qualities that are maximized are convenience- you can buy the tomatoes all year round- stability for transport, and price. Not that any of these are bad qualities- it’s just that the invisible hand often maximizes one quality or another over other qualities that might be there. Another example would be MP3s- they’re incredible portable and that’s what’s being selected. You can carry hundreds of songs in a little iPod. It’s fantastic and I love it. In terms of sound quality, they’re shit compared to CDs, which are actually shit compared to records. But nobody’s going to carry a record player in their car on a trip to the beach!Report

        • Avatar Larry says:

          The “invisible hand” isn’t a magic hand, as James notes in his post — it can’t deliver incompatibles, such as low cost and high taste in tomatoes, or portability and high quality in sound recordings. But it can and does service niches of every imaginable kind, and some that were unimaginable. For those willing to pay the real costs of flavorful tomatoes, for example, we see businesses springing up to deliver goods to the local and/or organic food markets. It’s just that no market, however niche, can simultaneously provide both sides of a trade-off.Report

          • Avatar dhex says:

            “In terms of sound quality, they’re shit compared to CDs, which are actually shit compared to records.”

            something of a tangent, but i think at 320 most people would be hard-pressed to tell the difference. if nothing else it’d be fun to bet on.

            that said, there’s flac, shorten, ape, tta, aac, and so on and so on and so on.Report

            • Avatar Mike Schilling says:

              As you point out, the quality of mp3 depends on the compression rate. I’m actually unconvinced that CDs sound noticeably worse than records, especially considering how records degrade over time.Report

              • Avatar Rufus F. says:

                It’s pretty impossible to compare what I hear to what you hear, of course, but I’m pretty sure I could tell the difference right off. To me, even at 320 sounds like music coming over a radio, while the analog signal sounds like music being played in the general vicinity. But I have really weird hearing. I’ve freaked out my wife before by describing conversations that were happening at the other end of a crowded auditorium that she couldn’t hear at all. It was one of the reasons she wanted me tested for autism- there’s some connection with ultra-sensitive hearing.Report

              • Avatar Jaybird says:

                It actually has to do with the artists embracing the medium as well.

                When MP3s came out, artists (generally) sucked. The artists who did their thing when CDs were big were better by comparison but not as good as the artists who worked primarily with cassette tapes for walkmen or car players. Those artists weren’t as good as the artists who had to work with 8 tracks and *THOSE* guys weren’t as good as the artists who knew they’d be working with vinyl. You want real music, you need to listen to 78s. Those guys knew how to play.Report

              • Avatar Mike Schilling says:

                Don’t leave out player piano rolls.Report

              • Avatar Jaybird says:

                Kids messing with the tempo of the music ruined them!

                Playing the William Tell Overture as if it were dirge. Playing Rock of Ages as if it were a common rag.Report

        • Avatar Will Truman says:

          Given what I do to CDs once they are in my possession, a 96kbs recording will sound better without all the skipping due to scratches…Report

  4. Avatar BSK says:

    “Which brings me back to education policy. We’d all like to know how to give a good education to children in adverse circumstances, but that’s not something we know how to do. So the only way to find out is to experiment. Let people set up new schools and new ways of doing things. Closely monitor their performance and learn.”

    This is theoretically what the charter system is designed to do. The problem is that too many are designed for fast and vast expansion and branding rather than actually developing a workable method and slowly exporting it and adapting it as necessary. Some schools will run their course for 3 years and immediately go into a plan to open 3 new schools every 2 years. How can you know that a method worked until you see the long term effects? The problem is that no one has a guaranteed shelf-life long enough to say, “We won’t know if this works until these kids are 18 or 20 or 25.” Certainly no politician and even most school administrators lack this type of job security, so they do what they can to improve education during their tenure so that they can stay on the job, sacrificing long term benefits for short term gains. And, generally, we end up with the one-size-fits-none model that we have today.

    Any long-term solution would basically require an 12+ year model, taking kids when they enter school and seeing how they leave it on the other end. You’d risk ending up with a “lost generation” of kids who might have gotten part of the benefits of reform without the entirety of it, but we have multiple lost generations now. That is why I believe in starting with a bottom up. If you say, “We’re going to institute reform for these high school seniors!” you may have some impact on some of them, but for most of them, they are likely too far gone to really be reached by these remedies (which is not to say that every child will fail, only that they are more likely to succeed in spite of the education system they are in than because of it). I realize this is a very pessimistic view, which I don’t intend it to be. Overall, the education system works in that we generally pump out at least semi-literate functioning people. But I don’t think that should be the standard by which we measure ourselves and think we can do a whole hell of a lot better. There just isn’t much incentive out there for people to really work for that type of change.Report

    • Avatar Jeff says:

      Any long-term solution would basically require an 12+ year model, taking kids when they enter school and seeing how they leave it on the other end. You’d risk ending up with a “lost generation” of kids who might have gotten part of the benefits of reform without the entirety of it, but we have multiple lost generations now.

      One of the difficulties of using some of these metaphors is that students — at least in the traditional language of process and production and investment — aren’t consumers of education. Well-educated students are the product of a production process, and the ongoing debate is how to improve the production process.

      In some ways, radical innovation in education is like radical innovation in whiskey-making: there are interesting ideas, but none of them are easy to judge quickly. And, as you pointed out, every bad barrel is a human life.Report

    • Avatar Barry says:

      “And, generally, we end up with the one-size-fits-none model that we have today.”

      That statement puts your credibility into question.Report

  5. Avatar RTod says:

    I like the idea in principle, but have great doubts about it’s practicality. Here, as I see it, is the big difference between Google and, say, my sons’ public school district: With a Google “experiment” there is a clearly defined goal and benchmarks for “success;” if a product/service/ management structure does not hit those benchmarks there is a person who can and is willing to pull the plug or make instant changes. Government has none of those things; and it also has the PR issue that those that are most likely to criticize that it *doesn’t* operate that way are the ones that get most bent out of shape when a bureaucrat *does* try to operate that way.Report

    • Avatar James K says:

      In his talk Harford spent a decent amount of time discussing the difference between experimentation and mere tinkering. You need clear criteria, ongoing monitoring and some form of controls for it to be an experiment.Report

  6. Avatar mac says:

    80% of google’s projects fail, but that’s at the techie level. The actual capitalization of those projects is tiny ~$1M compared with Google’s revenue. The government already does have organizations that work this way.
    * NIST
    * CDC
    * NIH
    * Government fundent university grants.

    But again, they are running at tiny capitalization. When really big ideas go bad–in both the government and business–you end up with crap like bank bailouts or nuclear meltdowns.
    I agree experimentation at the small scale is good, but you need a successful pilot before you scale it up to industrial–or governmental–size.Report