Schooling Is Not Education

Ivan Illich succinctly described the problem with schooling in the sentence cited by Kerry McDonald: “School prepares for the institutionalization of life by teaching the need to be taught.” But when people are eager to learn something, it is hard to stop them from doing so.[1] The importance of the unschooling movement is that it reminds us that education does not require schooling.

As an increasing number of families realize that schooling is not only unnecessary, but also often actively harmful, then we will see a tipping point in which millions of families will escape schools as we know them. This does not imply that we won’t have learning centers in which children learn together and at which academic courses are taught. We most certainly will. But we will gradually develop a much more nuanced understanding of when it is valuable to consent to being taught versus when it is more worthwhile to learn autodidactically.

Schooling creates a dependency mindset while providing the illusion of substance. In the absence of compulsory government schooling and other associated ways of penalizing youth (e.g. occupational licensing laws, minimum wages, and other obstacles to adolescent work), we would have seen a large, diverse set of approaches to learning and human capital development. Today we see teens learning to code, produce videos, and sell products on their own via various online communities and resources. We see bootcamps offering similar skills that typically take only a few weeks. Some of them are free up front and then take a modest percentage of earnings after the fact. In a world where unschooling had greater influence, we would almost certainly see more teens working and learning valuable skills, including product design, UX design, project management, sales, entrepreneurship, and more. The standard, mandatory high school curriculum completely ignores these key 21st century skills, often while teaching hostility to business.

Meanwhile, compulsory schooling is an evolutionary mismatch that is a causal factor for adolescent dysfunction and mental illness.[2] There is solid data on the increase of teen suicides during the school year and a corresponding reduction in suicides during summers and holiday vacations.[3] With the wealth and diversity of learning and networking options available today, a significant percentage of those who currently become engaged in dysfunctional behaviors, experience mental illness, or kill themselves could instead be on their way to satisfying, productive lives.

For these reasons I’m an enthusiastic supporter of the unschooling movement.

But the challenge with simply describing the alternative to schooling as “unschooling” is that by focusing on what not to do, unschooling advocates fail to provide guidance about what might be valuable alternatives. Kevin Currie-Knight points in the right direction when he notes that unschooling should include “a supportive and resource-rich environment.” But that formulation does not adequately differentiate successful from unsuccessful educational pathways. If we want to liberate millions of teens from the toxic aspects of schooling, we need to be clearer about how to provide successful educational experiences beyond traditional schooling.

I’ve spent most of my life creating innovative high schools that often attract students who have been unschooled or homeschooled up until the high school level. Some seek out a flexible high school environment because they long for more social connection and more comprehensive and rigorous academic experiences. Most are exceptionally mature, thoughtful, and responsible. A few are weak in some academic area, most often mathematics. A small subset have been addicted to video games (but that is also occurs with teens who have been schooled).[4]. Thus the positive outcomes described by McDonald and Currie-Knight represent the majority of the cases I’ve seen.

But until and unless we can reassure parents and the public that there are ways to ensure positive outcomes beyond conventional schooling environments, unschooling will remain a tiny niche movement. And simply endorsing a “supportive and resource-rich environment” does not provide enough guidance.

One of the most amazing young people I’ve ever met, Laura Deming, was unschooled.[5] She was working in a lab at UCSF at 12, went to MIT at 14, dropped out to accept a Thiel Fellowship at 16, and now in her early 20s is a leading venture capitalist in the anti-aging technology sector. I have no doubt that schooling would have dramatically slowed Deming’s intellectual progress and stunted her sense of initiative. When I asked John Deming, Laura’s father, how he educated her, he simply said, “I just let her do whatever she wanted.” But he is an extremely curious, highly intellectual, highly verbal entrepreneur. Of course he can “unschool” her and obtain spectacular results. What can we learn from his style of verbal interaction with her?

I’ve come to see the essence of valuable education, as opposed to schooling, as enculturation. My impetus for doing so came initially from my work introducing Socratic intellectuality into public school classrooms through Mortimer Adler’s Paideia Program in the late 1980s and early 1990s. Most cohorts of public school students were not capable of serious intellectual dialogue. But I gradually learned how to train groups of students to take ideas seriously and to work to understand difficult texts on their own.

At one point, an erudite rabbi observed the Socratic classes at one of my schools. He noted, “This is what we did at Talmudic school. The students would spend the week arguing with each other about the meaning of passages in the Talmud. Then once each week they would peel back the thick red velvet curtains of the rabbi’s study and ask if they could discuss their interpretations with him.” What I found instructive about this comparison was that within the context of students committed to studying the Talmud, a largely student-driven activity of textual interpretation was regarded as the core educational experience.

Thomas Sowell has documented the cultural foundations for disparate outcomes thoroughly, yet his evidence has been largely been ignored by the schooling establishment.[6] I interpret the discrepancy in outcomes among different ethnic groups as due to the fact that cultural capital is often more significant for success than is schooling. But as long as the mainstream regards schooling as equivalent to education, we will not see significant progress in reducing discrepancies in ethnic and racial outcomes.

While certainly there are those who regard the exceptional intellectual and professional success of the Ashkenazi as due to genetics, relying on a genetic explanation alone neglects the obvious role of living within a deeply intellectual verbal culture from birth. While the “30 million word gap” is the popular headline version of the distinction between children from educated families versus uneducated families, there is a very active research program examining the various distinctions arising from differing early childhood verbal environments,

“There is some evidence that the sheer amount of language input affects language growth (Huttenlocher et al. 1991), whereas other studies suggest that the quality of language input, such as the diversity and complexity of vocabulary and grammar (Huttenlocher et al. 2010, Rowe 2012), the contingency of language addressed to children (Bornstein et al.

2008), the use of questions (Aram et al. 2013), and language that goes beyond the here-and-now (decontextualized language; Rowe 2012), is also important. Recent research examining both quantity and quality simultaneously suggested that quality might be the primary predictor of language outcome (Rowe 2012, Hirsh-Pasek et al. 2015a), and different qualitative characteristics might play a role in different developmental periods (Rowe 2012, Tamis-LeMonda et al. 2014). For example, the diversity and sophistication of vocabulary facilitate toddlers’ lexical growth, whereas decontextualized language is more beneficial for later vocabulary growth in preschool (Rowe 2012).” [7]

The early childhood environment is the prototypical “unschooling” experience. I look forward to the unschooling movement moving well beyond, “a supportive and resource-rich environment,” and toward a deeper analysis of the particular characteristics of those homes and environments within which unschooling is likely to result in outcomes superior to those of schooling.

We will need greater specificity with respect to the critical elements of an environment in which freedom in education is more effective than is schooling before less structured forms of education become widespread. Maria Montessori famously described her approach as involving a “prepared environment,” by which she meant not only the physical environment, but very much the human environment within which children were immersed.

My main interest in unschooling is not “libertarianism” or “rights” per se, but rather universal human flourishing. Insofar as a belief that schooling equals education is supporting ever greater “investments” in schooling, I see the entire schooling establishment as an ever more expensive cargo cult. Public school expenditures have gone up threefold since the 1970s with no improvement in outcomes.[8] If we increase them another threefold will we have better outcomes? Utah schools have the lowest per pupil expenditure in the United States yet Utah has the highest rates of social mobility in the country.[9] What evidence is there to believe that significantly more conventional schooling will improve human lives?

I see the real problem with schooling is that it misleads parents, policymakers, and the public with respect to what is most important in getting an education. With the right enculturation, schooling may be superfluous, and unschooling works brilliantly. With the wrong enculturation, schooling seems to be at best limited with respect to its impact. Thus the entire existence of schooling as the dominant form of education has distorted the conversation around education. This is why I say that Bryan Caplan’s book should have been called The Case Against Schooling, not The Case Against Education.[10]

From this perspective, the schooling system is committing fraud by encouraging people to believe that learning the content of schooling is a necessary or sufficient condition for success. I see the next frontier in learning and human capital development to be a focus on cultivating specific cultural elements: intellectual dialogue, entrepreneurial initiative, self-discipline, and others. We need to create new and better subcultures of learning. But we are not likely to do so as long as schooling is the dominant paradigm. Simply adding Angela Duckworth’s “Character Playbooks” to a schooling curriculum will not result in the development of key character traits among populations in which they are absent.

Thus rather than advocating for “unschooling” per se, I’m very much in favor of a much broader and deeper dialogue along with greater social experimentation on exactly what education is. Gradually I expect we will find that “schooling,” in the conventional sense, becomes a marginalized element of learning and human capital development.


[1]. Frederick Douglass’s account of how he learned to write is a stunning example of learning despite immense obstacles.

[2]. Michael Strong, “Evolutionary Mismatch as a Causal Factor in Adolescent Dysfunction and Mental Illness,”….

[3]. Peter Gray summarizes the evidence the evidence in “Child & Teen Suicides Related to the School Calendar,”….

[4]. For a striking personal account of how schooling contributed to a video game addiction, see Cade Summers, “The Virtual World,” Original Path, June 17, 2018,….

[5]. John Deming, Laura Deming’s father, describes her education in a Quora post here,…

[6]. See, for example, Thomas Sowell, Race and Culture: A World View, Basic Books, 1995.

[7]. Amy Pace, Rufan Luo, Kathy Hirsh-Pasek, and Roberta Michnick Golinkoff, “Identifying Pathways Between

Socioeconomic Status and Language Development,” Annu. Rev. Linguist. 2017. 3:10.1–10.2.

[8]. Andrew J. Coulson, “School Funding System Isn’t Broken … It Just Doesn’t Work,”, May 13, 2013,…

[9]. Megan McArdle, “How Utah Keeps the American Dream Alive,” The Atlantic, March 28, 2017,….

[10]. Bryan Caplan, The Case Against Education: Why the Education System Is a Waste of Time and Money, Princeton University Press, 2018.

Also from this issue

Lead Essay

  • Kerry McDonald says that those who care about liberty should give unschooling a look. Structured education with a fixed curriculum and standardized testing is the product of a bureaucratized system. But children are spontaneous learners, says McDonald, and they do best when adults give opportunities and support rather than structure and evaluation.

Response Essays

  • Kevin Currie-Knight is a libertarian. He also likes unschooling. But, he says, it’s a mistake to conflate them; either view should stand or fall on its own—political liberty for adults might be the best choice, but this doesn’t imply that unstructured learning is best for children. And unstructured learning may be best for children without implying much of anything about the adult world of politics. While his personal answer to each of these is “yes,” the two are independent questions in his view.

  • Corey DeAngelis discusses the complicated relationship between unschooling and school choice. While he finds the evidence for school choice persuasive, and while he is also open to unschooling, he sets out several reasons why the two initiatives may not be wholly compatible. He recommends Educational Savings Accounts (ESAs) as potentially flexible enough to offer meaningful help for unschoolers.

  • Michael Strong considers what society would look like if we schooled less and educated more. While he is supportive of unschooling in principle, not everything of that name is equal, and he cautions that we still have much to learn about learning itself. One thing that seems unlikely to work, though, is spending more and more money on traditional schooling.