Matt Yglesias doesn’t agree with Arnold Kling and Nick Schulz on the impact of undermining public sector cartels in K-12 education. Because I’m sympathetic to the Kling-Schulz argument, and I’ve argued along similar lines in the past, I thought I’d highlight Matt’s objection:
“Imagine,” ask Arnold Kling and Nick Schulz, “what might happen if government involvement in education were restricted to giving school vouchers to households below the median income.”
I think it’s pretty obvious what would happen. K-12 education would start looking a lot more like college education, where by far the biggest federal involvement is semi-targeting tuition subsidies for people in need. What would happen is that is that out of pocket spending by higher income families on schooling would explode. Right now across the vast majority of the country, the convention is that moving to a place with a decent public school system and sending your kids to it constitutes doing the right thing by your kids. But in vouchertopia, there will be massive status competition. Parents who can afford to foot the bill to send their kids to the Ivy League of high school will do so. And, like in higher education, the “quality” of a given school will be determined primarily by the quality of the inputs. Consequently, schools will compete to try to be as selective as possible and will offer targeted tuition discounts to kids who enter school with unusually high ability levels. High school graduates will earn a substantial premium over dropouts, but we’ll have endless disputes over how much of this earnings premium is signaling/selection and how much reflects actual learning.
There has been an explosion in private tuition services, and I have no reason to believe that this has been a bad thing. Among affluent parents in cities like New York and Los Angeles, it is common to hire private tutors, even for — particularly for — strong students. The idea is that a private tutor can help keep bright students engaged and on track. It should go without saying that this is a luxury service, and it is not uncommon for private tutors to have attended highly selective colleges and universities. Some tutors are veteran teachers with graduate degrees. Now, should we be alarmed by this particular kind of consumption?
I’d suggest that the answer is no. The private tuition boom has created lucrative employment opportunities for a small but growing number of young people who are interested in pursuing creative work. Creative work is a tournament, in which one must invest long periods of time in building a skill set and networking in the hope, often unrealized, of securing lucrative employment more closely aligned with one’s creative ambitions. Private tutoring can serve as a kind of subsidy for this effort. The money that is now spent on private tuition would presumably be spent on some other form of consumption — luxury goods? a slightly more expensive vacation — if there were a widespread cultural stigma associated with the practice.
As affluent parents in elite cities have spent more of private tutors, middle-class parents have increasingly embraced services like Kumon and Sylvan Learning Centers. One assumes that more entrepreneurial schools would seek to capture some of the revenue that would otherwise go to Kumon et al. by embracing new forms of personalized instruction. Matt is of the view that this will mean a continued drive towards smaller class sizes, which is certainly possible. But if cost-conscious parents are offered alternatives that deliver higher quality at lower cost, e.g., by allowing the most effective teachers to take on more students, it is not obvious to me that they wouldn’t embrace said alternatives in large numbers.
Moreover, an explosion in out of pocket spending doesn’t strike me as a particularly bad thing particularly because it is out of pocket spending. I’d be much less concerned about people spending large sums on their homes if we didn’t have a mortgage interest deduction and severe restrictions on the supply of housing in desirable neighborhoods. (A digression: One of the constraints on the supply of housing in desirable neighborhoods is the structure of school financing arrangements. Voucherization could mitigate this problem by decoupling one’s place of residence from one’s school.) This isn’t quite Matt’s objection — I’ll get to that in a moment.
Would perceived quality necessarily track the quality of inputs, i.e., the selectivity of schools? That is certainly possible. Yet it is also possible that parents will track other metrics, e.g., metrics relating to the quality of outcomes for different kinds of students. I suspect that some parents will want to send their kids to schools that specialize in the arts while others will want to send their kids to schools that specialize in science and math. Better still, we may well see a great deal of “unbundling.” Students will assemble a portfolio of courses drawn from a variety of providers — traditional and online will co-exist in blended learning environments, and cost-consciousness will increase if parents are allowed to retain some share of the money they save.
To be sure, not all parents will be in a position to navigate all of these choices. Students from less affluent or more chaotic backgrounds might benefit from attending KIPP-like schools, which might also proliferate in this environment.
Matt suggests that the federal government’s main involvement in the higher education space is in tuition subsidies for less-affluent students. This is a debatable proposition. As Kling and Schulz explain in an earlier essay, the federal government provides a wide array of direct subsidies to colleges and universities and governments at all levels create a number of arbitrary degree requirements that constitute an implicit subsidy for higher education.
[S]ince the vouchers would need to be pegged to some kind of objective measure of school costs, the vouchers would increase in size as tuition costs skyrocket, driven by status competition. Libertarians wouldn’t like this, and they’d write articles about how the problem could be solved if we would just get rid of subsidies and accreditation.
Here’s the thing: Matt is suggesting that an explosion in out of pocket spending will be a bad thing because it will drive up school costs across the board and “vouchers would increase in size as tuition costs skyrocket.” This is a real danger. The key is pegging vouchers to a defined benefit, per the Vance Fried approach. As you might recall, Vance Fried, a professor at Oklahoma State University, engaged in a provocative thought experiment: how much would it really cost to offer a world-class undergraduate education? He concluded that the number was under $8,000:
[A] CELS with 3200 students would have a total operating cost (without room and board) of under $8,000 per student. This is the cost to the school, not the cost to the student. Price (i.e. tuition) is the cost to the student. Because most colleges are heavily subsidized by a state and or/private philanthropy, they are able to charge tuition well below their actual cost. CELS’s cost of under $8,000 is drastically below the cost of “top” liberal arts schools ($25,000 to $62,000) that cater to prestige-oriented customers. But it is also well below the $12,000 cost of public regional colleges who have many price driven customers and a less academically selective student body. So, if a CELS received the same level of subsidy as a public regional college, it could charge students about $4,000 less than the regional, even though the product was competitive with the “top” liberal arts schools.
To arrive at this low cost position, I didn’t cut corners on anything that was important to the CELS value proposition. CELS doesn’t use many adjuncts, faculty salaries are competitive with those at research universities’, a laptop is included in tuition, the Division III football stadium has a Jumbotron, etc. As the CELS example illustrates, a college using a value designed model could deliver a prestige quality product to its target market and yet have vastly lower costs.
One would have to use a similar method to determine the size of vouchers. This premium support model (aha!) might then involve restraining the growth rate of premium support to encourage cost control.
Now, Matt doesn’t believe that this would work well in the health space, and it’s a safe bet that he won’t think it will work well in K-12. My own view is that it has a far better shot at working in K-12 than it does in medical care (though I’m optimistic about its prospects in that domain), in part because we’ve seen a great deal of labor-saving innovation in education that hasn’t been effectively deployed in part due to political, institutional, and cultural resistance. A more diverse educational marketplace would allow for more experimentation, and for more “disruptive innovation,” i.e., for the proliferation of low-cost, lower-quality models designed to meet the needs of the underserved that get better faster than traditional high-cost options.