In Friedrich von Hayek’s Nobel Prize acceptance speech, he urged increased attention to
‘gardening’ – effort towards improving underlying conditions – and less focus on engineering
better solutions from existing incentive/constraint regimes. This paper explains what ignoring
Hayek’s advice has meant in K-12 education, and what increased attention to ‘gardening’ might
mean for K-12 education; that is, how might we expand the scope for much-needed improvement
through creative introduction of price signals as a cornerstone for deciding what is taught, where,
how, and to whom.
In the first week of a ‘Principles’ course we talk about ‘isms;’ the macro policy choices for
determining what is produced, where, how, and for whom. We tell our students that while a ‘mixed
economy’ is the norm, that there are only two basic ways to ‘determine’ what an industry produces,
where, how, and for whom. 1.) The first is a central planning approach wherein political pressures
and administrative processes allocate resources. It is essentially a price-less process in that planners
set priorities, and any explicit prices (for example, teacher salaries) arise from planner choices, not
from just market pressure to balance the quantities supplied and demanded. Price-lessness has a
terrible track record, especially when applied economy-wide. But even when applied to just
individual industries, severe inefficiencies have typically, eventually, prompted privatization of
nationalized industries – with 2 exceptions. Added to inertia, equity aspects of health care and
schooling for children (K-12 = primary and secondary) have kept those two industries price-less,1
despite obvious major inefficiencies virtually everywhere. The significant inefficiencies are widely
seen as unavoidable trade-offs.
2.) The second basic way to determine what an industry produces, where, how, and for
whom is, like day is the counter-part to night, decentralized planning. For it, property right
definition and enforcement is the only indispensable government role. Independent producers,
potential producers, consumers, and potential consumers react to prevailing and predicted prices
(for example, based on outlook or potential earnings through innovation) to decide whether to
participate in a market, or through entrepreneurship consider creating a new market. So, while
central planning is price-less, impersonal price formation and adjustment is a key determinant of
decentralized planning outcomes. Since decentralized planning through the price signal-incentive
process has a track record of high efficiency (key conditions such as good information, competition,
and insignificant spillovers are the norm, but not ubiquitous), it is fortunate that it is the default
circumstance; the natural state unless competition fizzles or collective action creates different
circumstances. So, most industry outcomes in most countries arise from independent responses to
prices, often distorted some by taxes and regulations.
Though price-less schooling is a global norm, I’ll focus on the specific shortcomings of the
51 US school systems (51 because of state [+federal for DC] primacy in dictating K-12 funding and
governance). Despite the readily apparent inefficiencies, alongside breathtaking failure on equity
grounds, not nearly enough of the American electorate has connected the typically visceral, but
poorly grounded widespread disdain for ‘central planning’ with the terrible and deteriorating
outcomes of the 51 US school systems. Worse still, our leaders do not believe that decentralized
planning can be harnessed, and needs to be, without the politically difficult step of abandoning
central plan optimization for the public school system. Indeed, the best possible reform for K-12
schooling – and healthcare – may be to create a level playing field between the incumbent centrallyplanned
options (you can change attendance areas) and the options created by entrepreneurship that
sovereign consumers patronize in sufficient numbers to determine which options stay in business.
The paper proceeds as follows. In the next section, I describe the specific circumstances of
K-12 central planning common to the 51 US school systems, and two significant specific instances,
and consequences, of price control and price-lessness. The third section critiques the terrible
assumptions underlying the troublesome policies common to the 51 US school systems. The fourth
section describes the essential elements of a system that can yield significant efficiency and equity
gains starting from where we are and the different ways – policy options – to achieve those essential
elements. To fit all of that in a paper, I will make frequent referrals to my new book, School System
Reform: Why and How is a Price-less Tale (e-book, $10!!).
Price-lessness and Centrally Planned Schooling American-Style
The political process has minimized the probability of being a successful public school
teacher. And American-style politics has made the central planning process – terribly inefficient
under the best of circumstances – even less efficient. That may be why the outcomes of the US
central plans average out below the central plan outcomes of many other countries; though not
Decentralized Central Planning, plus Political Process Shortcomings
The 51 US school systems do central planning without a solid center. We have fragmented
– decentralized (!!) – central planning. Because states can directly and indirectly specify school
system governance and funding policies, the US has 51 school systems. However, for a long time
all of the central planning was ceded to school districts, which does NOT establish decentralized
planning. Local primacy just creates more central plans.
Inevitable failure to achieve efficiency through central planning (widespread, persistent low
performance), plus corruption, led first to decreased political accountability, without market
accountability, to empower education experts; to shield them from nasty local politics; the ultimate in central planning thought. Trust the experts. Then, when it became widely true that expert local
control through weak political accountability was producing mostly ‘Nation at Risk’ outcomes, state
governments became more actively involved in the central planning process. When the
combination of state and local central planners failed to yield the hoped-for school system
performance gains, periodic revisions of the Federal Elementary and Secondary Education Act of
19653 added federal legislators and administrators to the central planning process.
We’ve added weak political accountability, fragmentation, and distraction to a process that
is fatally flawed under the best of circumstances with central planning’s always weak information
generation process and perverse incentives. I say weak accountability because the central planners
are only strongly accountable to a fraction of the public, especially the local central planners,
because of circumstances that typically leave voter turnout for school board and bond elections in
the single digits. Low turnout is partly by design; to empower the experts; insulate them from nasty
local politics. On top of the turnout-depressing effects of off-cycle elections and awkward polling
times and locations, we have an over-extended electorate. Though the sum total of what we expect
in terms of voter attentiveness varies from place to place, it is mostly WAY MORE than a typical
voter CAN study, and exceeds voter willingness to devote effort to candidate assessment by an even
wider margin. For example, in my Texas county, an election cycle can produce over 240 candidates
to assess. Distraction arises from the fact that K-12 policy choices are just a small part of what the
political accountability of the elected state and federal central planners depends on where, again,
low information voting and low turnout are major factors.
Two Examples of Price-lessness in Action
1.) The appearance of fairness political imperative yields a lot of equal treatment of unequals. In
the nearly always prescient words of Thomas Jefferson, there is “nothing is more unequal than
that.” The nearly universal ‘single teacher salary schedule’ is a great example of how the political process yields the equal treatment of unequal teachers through price-lessness. The politicallydetermined
schedule predictably yields shortages of teachers with higher opportunity costs (STEM
fields), or more difficult work, such as with special needs children and children from low income
families. That predictably yields a lot of out-of-field teaching (weak content, reduced student
engagement) and inexperience where experience is needed most (where low income families
concentrate). 2.) The price control underlying the delusion of free schooling, and the delusion that
zero tuition at all schools receiving public funds is an equity imperative, yields disengagement,
shortages, virtual economic segregation, uniformity and excess need for differentiated instruction.
That delusion often persists even for non-public schools that receive government funds, indirectly
through tuition vouchers when copayment is illegal, or through full tuition direct payments in many
other countries. In truth, only an area’s worst public schooling is ‘free’ in the sense that the full
tuition is pre-paid through direct and indirect taxation that everyone must pay. For all but the worst
public school choices made by deciding where to live, you pay a housing premium on top of the
taxes allocated to K-12 education. Housing premiums that vary by relative perceived school quality
yield economic stratification (virtual segregation) by attendance zone and school district.
There is a big difference between the effects of rationing school quality through housing
cost differences and the effects of co-payment (public-private shared financing) of market-set
tuition. The former reinforces/widens qualitative differences in the uniform comprehensiveness of
public schools that must attempt to fit nearly everyone that could be assigned to the school, but can
only offer instructional approaches that can be provided everywhere.
Payment through tuition would likely yield specialized schooling options as diverse as the
engagement factors of an area’s schoolchildren. Motivated by the potential to temporarily reap a
large difference between per-pupil cost and tuition price, schooling option entrepreneurs will
constantly find and then quickly fill high-value schooling niches. Through imitation of the popular schooling innovations, a shared financing approach can actually yield the truly ‘free’ high-quality
schooling options the current system lacks. Some of the schools, under competitive pressure, might
charge no more than the government funding share of allowed shared financing, and/or charities
will finance most or all the private share of a non-zero co-payment on a means-tested basis. And
choice will likely reduce the housing premium to be paid to attend a better public school.
So, mandating a tuition price of zero does NOT yield the equity benefit still widely seen as
significant. Quite the contrary. We have an equity disaster, including the concentration of poverty
around the worst schooling options. Note, BTW, that we get the latter effect even if the worst
options still effectively deliver comprehensiveness as defined by what will be provided everywhere.
Families with the means to pay the housing premiums will prefer best to good and okay. Sadly, the
actual worst of the current system are not nearly okay. Stories of terrible assigned schools abound,
and persist for nearly all of them. And the turnaround record is abysmal.
The most heroic assumption – the central assumption of the world’s prominent school
systems – is that politicians and their appointees can succeed at something fatally flawed; to be the
first to achieve acceptable, sustainable results through central planning. And the flawed equity
assumption – that schools need to be 100% taxpayer funded – did not require a commitment to the
central planning of schooling. The initial proponents of the system – for example, Horace Mann –
can be perhaps forgiven for not recognizing the fatal flaws of price-less determination of what is
taught, how, where, and to whom.
We know much more about the central planning process, now, than we did when
government-run schooling was promoted as the answer to the real and imagined schooling
challenges of the 19th century. As early as the 1920’s the opposing sides of the ‘calculation debate’
agreed that price signals were essential to efficient outcomes.4 But inertia and special interest politics – somewhat redundant – plus mounting social science illiteracy, prevented connecting the
mounting disappointments with our traditional public schooling-dominated school systems with the
key elements of those systems.
As noted above, the flawed and still widely held equity assumption was that ‘free’ schooling
was achievable and the best guarantor of a high level of opportunity for all. It survives even the
current terrible equity outcomes, which are partly due to the virtual economic segregation promoted
by the current system that reinforces the inefficiency of the US systems. Part of the problem is the
obsession with equality of opportunity which is unattainable, and with its pursuit that weakens the
strong without any widespread benefit to the disadvantaged. For example, when we ban topping off
of tuition vouchers (Milwaukee), so that we supposedly get “school choice not limited by family
income,” we virtually prohibit schooling options that cost more than the voucher amount, even just
temporarily in the just-introduced phase. That harms the disadvantaged that might access those
options immediately through means-test top off assistance, and/or eventually as experience and
competition drives the top off amount down to negligible or zero. And we know that through
housing markets, public school choice is limited by family income. It is likely true, also, for private
school choice. We need studies to document the extent to which proximity to different degrees of
private school quality impacts housing values. We know that school choice expansion increases
property values (increases housing costs) relative to places with less school choice.
Another terrible and breathtakingly audacious and heroic assumption is that monopoly –
high barriers to public school system alternatives – enhances school system (all of the schooling
options, public and private) performance. Compared to what? We know the status quo is awful, so
are the public finance monopoly proponents arguing that without high entry barriers to private
schools, we’d have even worse school system outcomes? And except in the sparsely-populated areas where you might have a credible natural monopoly argument, there are no standard theoretical
grounds for arguing that lower entry barriers would yield reduced efficiency.
Related to that are the twin heroic assumptions that justified subsidies are best channeled
through the producers of goods and services, rather than through the consumers, and that, indeed the
public dollars generated to support schooling belonged to particular producer; the monopoly is best
argument. As Milton Friedman noted, ranking producer merit is not easy. Indeed, as Andrew
Coulson demonstrated, the measured performance of California chartered public schools did not
explain donor support for CPS. Since, likewise, withdrawing support from producers that have lost
their way – perhaps only because their once pathbreaking ideas had become obsolete – isn’t easy,
rare subsidies, public or private, should only be allocated to production through consumer choice.
Another incredibly heroic assumption is that even acceptable performance, much less great
and relentlessly improving performance, can be expected from a system virtually incapable of
rewarding merit, or punishing low, even counter-productive performance. Public school systems
struggle to make revenue reflect performance, and to make revenues to flow to the agents of extraordinary
performance. The one possible connection between public school performance and
revenue per pupil is increased housing values in the attendance areas of the better schools. Even if
that more than offsets lower housing values within the district, there is no guarantee that the
additional dollars will find their way to better schools, much less the educators most responsible for
the extra-ordinary performance. In a system that does not have to reward merit to compete for top
talent, additional dollars are just as likely to find their way to the struggling schools. As seen
through the price-lessness theme, the politically correct price of effectiveness is zero. Of course the
implied incentives don’t matter assumption is hope triumphing over mountains of experience.
The appearance of fairness imperative of political accountability forces public school
uniformity. Only instruction that can be offered to everyone, except for extreme special needs cases, will be offered to anyone. That, and the diversity of children, has led to uniform significant
specialization within mega-campuses; despite much evidence that schools are too big. The
economy-wide recipe for efficiency is specialization by firms, not within multi-product firms. To
achieve the former within a school system, you need choice. You can’t have attendance zones for
Another implicit heroic assumption is that there is one set of best practices for the vast
majority of schoolchildren5 and the political process will discover them, adopt them, and optimally
deploy them; something the U.S. systems have yet to manage. As Eric Hanushek pointed out, there
are no universally-best instructional practices/content. A closely related, implied huge heroic
assumption is one dimensional student ability, school quality, and educator talent. The current US
systems contains some utterly dysfunctional schools. They can persist because of attendance zones,
housing cost differentials, and high informal barriers to private schools. But there are no public
schools that are great, or even good, for everyone assigned to them. So, what does ‘this public
school is great’ mean? It means that the school is very good, at least in comparison to other public
schools, at what it is tasked to do, which means it is an okay fit for most children. And that means it
will cost a lot more to live in its attendance zone.
Comparison of ‘vehicles’ – policy tactics
Since decentralized planning means determination of what is taught, how, where, and to
whom through price system orchestration of independent producer/consumer pursuit of self-interest,
we have to determine the best ways to introduce that combination into the 51 US school systems.
To sufficiently erode the entry barrier to entrepreneurial participation in the production of schooling
options, we have to substantially level the playing field between the public schools – traditional and
charter – and non-public schools. Ideally, from both an efficiency and equity perspective, we would
implement a non-discrimination policy, which means that the public funding support for a particular child does not depend on the ownership of the lawful schooling option preferred by a child’s
parents. Because of inertia – the need to repay debt and fund pension liabilities – and since it is
unclear exactly which expenditures should be included in a non-discrimination calculation, it is
fortunate that near non-discrimination is a very adequate degree of leveling.
Regardless of the basis for determining the public funding that supports a particular child –
probably a Weighted Student Formula – the absolute NEED to avoid having the government
implicitly decide what it should cost to educate certain categories of children (price control), we
have to allow shared financing of private school tuition. The shared financing term is better than its
many synonyms (co-payment, topping off, adding on, etc.) because it makes clear that when the
public funding is not enough to provide access to a preferred schooling option, that the needed
supplementary funding can come from sources other than the child’s family. I can state as a fact
that charities will share in the financing of private school tuition on a means-tested basis because
they already do. And with tuition price decontrol, charities will have additional means and
motivation to share the tuition costs of low income families. With only a fraction of the tuition to
fund (just the top off; the copayment), existing charity funding can support many more children.
And price decontrol will free up all of the donor funding now paid directly to school operators that
have made a compelling case that they cannot provide a particular instructional approach for just the
per-pupil funding amount.
Having the probably-WSF-based, per-pupil public funding directly follow the child seems obvious, and many countries do that. But in the USA, some indirectness through tuition vouchers,
tuition tax credits, or Education Savings Account (ESA) deposits seems necessary to keep within
the US Supreme Court’s guidelines for keeping church-run schooling on the level playing field
between public and private schooling options. Doing so is not quite as important to decentralized
planning-based transformation of school systems as many people might assume because of the
existing, church-run school dominance of the private schooling options. That dominance is likely
an artifact of having to pay twice for privately-provided schooling in the current 51 US systems.
Only private schools can deliver religious course content, and that content is among the only
schooling features that a lot of people will pay for even when a no-additional-charge schooling
Because tuition vouchers and tax credit scholarships, especially the former, are the most
vulnerable to being on the wrong side of the controversial shared financing issue, the absolute need
for tuition price formation and market-based change says that tuition tax credits and education
savings accounts (ESA) are the best way to level the playing field between public and private
schools. ESAs are the new kid on the block, with still great room for improved enforcement of the
education-only restrictions on the spending of ESA balances. In the few states that allow ESAs,
eligible families get an annual ESA deposit for each child they opt out of the assigned school. The
potential to supplement the ESA deposit, or carry an unused balance into the next year virtually
assures that shared financing of education-related purchases, including private school tuition, will
not be prohibited. And thus decentralized planning of alternatives to the assigned public school can
proceed on a much more level playing field.
Refundable tuition tax credits – you are paid the difference between the targeted tax liability
and the credit amount – are virtually identical to tuition vouchers; perhaps with a slightly lower
political risk of a shared financing ban. But nonrefundable tuition tax credits have great advantages
and disadvantages over the other ways to level the playing field between the public and private
schooling options. Because courts have ruled that the dollars leveraged by a nonrefundable credit
are not public funds, a nonrefundable credit policy is less likely to be a basis for increased
regulation of private schools than any of the other ways to pursue near nondiscrimination.
Considering what is happening to chartered public schools in many states, that’s a huge selling
point. But because the public money that funds schooling comes from many taxes paid over entire
lifetimes, not just when a family’s children are in school, we need some creativity for a nonrefundable
credit to achieve significant playing field levelling even for a single student per family,
much less several at once. Extending the credit to businesses can provide tuition assistance for
families with little of the credit-targeted tax liability, typically a state income tax. For example, a
refundable state income tax credit, by itself, has virtually no value to low income families. The
business tax credit-motivated tuition assistance, plus credit bankability, can elevate the nonrefundable
credit to a near non-discrimination basis. Bankability means that households can
continue to defray tax liability with a credit long after the tuition payment basis for the credit was
made, often long after the children are out of school. That is how we pay for public schools; with
taxes long after our children are beyond school-age.
So, the ESA and nonrefundable credit approaches are much superior to tuition vouchers or
direct payment, regardless of constitutional issues regarding direct payment to churches. ESA
strategies are less complicated than nonrefundable tuition tax credit approaches that can achieve
near nondiscrimination, but ESAs can be abused – spent on non-education purposes – and an ESA
policy is more likely to yield some additional private school regulation.
Summary and Concluding Remarks
Central planning malaise besets primary and secondary education, globally. The US version
is especially bad because we have fragmented central planning. We need to introduce decentralized
planning of schooling options; that is, eliminate price-lessness. The worst of the existing price
controls implicitly imagines that eliminating tuition – 100% taxpayer funded schooling – yields
schooling, and perhaps “school choice not limited by family income” at an acceptable efficiency
cost. Indeed, we need price decontrol to realize equity improvements and efficiency gains through
relentlessly-improving schooling options able to engage a diverse population of schoolchildren.
Political inertia will likely continue to make it impossible to replace the current school
systems with a fully decentralized planning-based system. Because central optimization has a very
low upside, we need to eliminate the entry barriers to schooling options created by decentralized
planning which is the result of entrepreneurship and parental choice informed and motivated by
market-clearing prices. We can make equity gains without foregoing efficiency maximization.
Through the price decontrol that will result from allowing shared financing of tuition, we can
meaningfully signal full social marginal cost of different instructional approaches without making
best-fit schooling options unaffordable to the least advantaged families. They will have access to
means-tested copayment assistance.
Why and How is a Price-less Tale.
2 See chapter 2 of School System Reform: Why and How is a Price-less Tale
3 https://en.wikipedia.org/wiki/Elementary_and_Secondary_Education_Act; Major Amendments: Improving
America’s Schools Act of 1994 (‘Goals 2000’), No Child Left Behind Act, and Every Student Succeeds Act.
5 All but the most severe special needs children are thrust into regular classrooms; “mainstreamed.”