As Conor says, there is a decent level of agreement between us on certain things. Imagine this classical liberal’s joy when his progressive interlocutor states that “Our current educational market has such an extraordinarily high resource barrier to entry that it essentially closes off school choice to the upper- and upper-middle classes.” (And readers who did not read the article Conor linked to in that statement should do so.) As a former public school teacher who remembers students being asked to leave our Baltimore County high school and return to their zoned-for–Baltimore City school, I find it illustrates the travesty of our current system.) I still hesitate to call it a market, because it is the house that is chosen, and the school comes along for the ride. But we’ve had that discussion before, so rehearsing it only spins our wheels. And it is a small disagreement.
Also, I couldn’t be more delighted when Conor writes that “it’s critical that we establish a choice market that extends as much choice as possible to as many parents as possible while also maximizing the good choices available to less-wealthy parents.” Well, at least we know that Conor and I seemingly have the same goals!
But the Cato Institute didn’t invite us to agree on things, and there are some important points of disagreement to discuss. Conor rightly notes that, when discussing statistics that the government might mandate that schools make easily available to consumers, I only included data on “output” – graduation rates, teacher to student ratios, and the like. He suggests that while this data has some use, he’d push me “on the question of the state mandating that schools take common assessments and share those data with the public.”
I was very careful only to include “output” data on my list, because I am reluctant to give governments power to mandate common assessments that all schools must have students take. To Conor’s point, though, I do understand the potential benefit of this approach. Markets function best when people are able to make well-informed choices, and this is most likely to happen when consumers are able, as they say, to compare “apples to apples” by seeing different schools’ results on common metrics. (For this reason, Myron Lieberman, a pro-market writer on education whom I much admire, even suggested that choice systems should encourage schools to sign on to a “nationally accepted” (but not mandatory) curriculum on which standardized tests would presumably be built. (Public Education, an Autopsy, 275).
Yet, when considering the potential costs and benefits of such a device (state-mandated standardized testing), I still see the costs as outweighing the benefits. In brief, I believe that enforced standardized testing dampens two of the best aspects of markets – innovation and diversity. But to the degree that standardization becomes necessary, there is still good reason to think that markets are more than capable of developing the kind of standardization that – because ultimately voluntary – doesn’t endanger these two features of markets.
First, the perils of standardized testing. Mandating that all students take state-mandated standardized tests allows schools to have some flexibility – flexibility in how they teach and run their day-to-day operations. But those who have taught in k-12 settings know that standardized (summative) testing has a deceptively strong impact on how schools do everything else. To put it simply, those who control the tests indirectly control the curriculum: I still recall being a high school teacher planning my science courses, where the first thing we did each year was look at past editions of Maryland’s High School Assessments, which greatly impacted what we did during the school year.
As long as a test is “high stakes” (and the idea that a school’s results on this test will impact the way your school compares to other schools makes it so), that test will limit what schools feel they are able to do within the curriculum. The implications for education design are often greater than for other products and services. Standardized tests that essentially tell schools what students should know at certain ages are mandating how they should do a large part of their job.
Also, as I’ve argued elsewhere, state mandated tests allow one group to hold a monopoly on telling us what students should know at what age, which is not only staggeringly difficult to do in a world of diverse learners, but especially dangerous to have monopolized. If we are worried that schools, left to their own, might get bad standards or create bad assessments, imagine what would happen if a state or a nation were to get it wrong. Even my hypothetical makes a big assumption: It assumes that in a world of diverse learners, we can still know what “wrong” is regarding standards and assessments. It may be that different standards or assessments work well for some kids, but are poor matches for others.
Just as Hayek argued that no person or group could ever have all the knowledge, including local knowledge held by economic actors, to plan an economy, I argue the same for education: no person or group can possibly hold enough knowledge to know what assessments will appropriately gauge whether we are “properly educated.” So, while I understand the potential for standardized testing to allow for “apples to apples” comparisons among schools in a market, I worry that doing that will force all schools to be apple orchards, when there may be benefit to allowing some to be orange groves or grape vineyards. (Sorry, couldn’t resist.)
What are my solutions? Well, the classical liberal in me prefers voluntary solutions whenever possible. Therefore, I’d first like to suggest that a great many industries have – without government help – standardized their services in order that customers can shop among interchangeable services. (See here, specifically section 1.2 and 1.5.) The internet has standardized HTML as its dominant language without government assistance. CDs, DVDs, and MP3s became standard containers for media without government assistance, as did QWERTY keyboards become the default style of keyboard and (at least for now) HDMI cables become the standard multimedia cable. What I suspect could happen – if inter-school comparison does become the problem for consumers that we suppose it will – is that organizations like the nonprofit Educational Testing Service or for-profit Pearson Assessments will introduce their own tests, allowing schools to choose among competing alternatives. Schools may have incentive to choose into a particular testing service because doing so will provide parents with a ready way to compare the school’s data with other schools that use that test. It will also signal to parents the school’s confidence that their school will yield good results on such tests.
If something like this did not happen, I might be willing to consider a role for the state in creating a particular testing system that it believes would provide the most useful gauge on the quality of education schools are providing. But from there, I’d prefer that states only make the test available for schools to use voluntarily.
Both methods, of course, are speculative. And neither, I’m afraid, affords any sort of full guarantee that all schools will have common data to share. But both solutions are closer to what I’d be comfortable with, largely because they achieve at least a level of standardization of data, while allowing schools the flexibility to innovate and use such diverse approaches to education as they choose. That way, a Montessori school doesn’t have to change the way it educates because state tests tell it that it must prove that all xth graders know y in z detail. It can choose not to use any standardized testing, or perhaps use one developed by the American Montessori Society, and allow parents shopping for a school to decide it that is good enough for them.
I should also address Conor’s last paragraph, where he suggests that “Advocates for more school choice ought to be able to show that they can improve on the existing system beyond a general faith that markets always ‘raise the quality and/or lower the cost of all groups’ products over time.”
Fair enough, but I hope Conor and others understand why this is a tough one in the current political climate, where suggestions that we run the experiment are either met with a flat “No!” or, at very most, “Okay, but we’ll only let you try it under conditions limited enough that your experiment will bear no real resemblance to an actual market.”
When the market is limited to a small sliver of the population, when prices are capped or set by the state, and when for-profit companies may not enter the market – these among other features of current voucher plans – it is unfair to look at these as indicative of how markets work. I don’t want to overstate this, of course, because good data exists showing markets do work well in education, even in the very limited and constrained spaces where we allow them. Even most studies showing markets not leading to better educational results seem to show that, at worst, they do no worse than public schools.
So, alas, at this point, the case must be largely theoretical. But that case has been made in spades, from folks on all sides of the ideological spectrum. Of course, the strongest theoretical case for markets in education is that we use markets to great effect in just about every other area of our lives. Going back to my original post, there not terribly strong reason to think that the market forces that work well in so many other areas will stop working in education. Beyond that, there are very good reasons to think that government bureaucracies will manage schools much worse than private actors in markets. Markets provide both a profit motive and a competitive process that spurs innovation and efficiency in production, as well as a decentralized system that allows producers and consumers to adjust to changing circumstance better than a centralized and quasi-monopolized system. I am sure that Conor is familiar with this literature.
Is this proof that markets will give better results than a public system? As long as proof means empirical evidence, not exactly. But it is not exactly a faith, as Conor suggests, either. It is pretty strong economic reasoning, and the kind of strong economic reasoning that seems to bear out in other industries. It does seem to me that the burden of proof really should be on those who suggest that what works in almost every other area just couldn’t work in this one.