Part of my How Homeschool? Series.
Art Robinson came up with a rigorous back-to-the-basics homeschooling program that could be self-taught. Today, 5/6 of his children have doctorates in nuclear engineering, veterinary medicine, or chemistry. You can read all about it in our previous correspondence.
But a standard objection is that it’s impossible to know if the Robinson program “worked” because his kids inherited the genetics of a CalTech trained chemist and his biochemist/computer system programmer wife. Robinson himself waves away this concern saying that all but “institutionalized children” can run his program and by insisting that his children were “average,” who were totally capable of being distracted but for his regime. But there was no randomized control trial of assigning each kid to a different homeschooling regime - and it’s worth noting that the older kids, including the one who got his PhD at CalTech, had more of Laurelee’s original program (though even the oldest still had six years with just Art).
Robinson outlined some very big (and admirable) ambitions in the late 1990s:
“Our children must be able to think – and to think so much more effectively than their opponents that they are able, in one generation, to become such a superior force in science and engineering and in industry and government that they dominate American society. Our children must be such shining examples for the home school movement, that the majority of American families demand the same quality for their children – a quality that can only be obtained by becoming Christian families who take responsibility for themselves.”
Yet Robinson has claimed that over 60,000 children have used his curriculum and we don’t often hear about Robinson alumni. These emails may be the very first you’ve heard of Art Robinson, so there’s no general outcry to adopt his methods. Very probably many more people buy the curriculum than actually implement it in all of its glory, but that still begs the question: does the Robinson method succeed in its goals? 60,000 sounds like a lot but it’s admittedly a very small share of the total number of American children since the curriculum was introduced. If they didn’t master calculus by 15 (or even 18), was it a failure to execute the method (especially when it required an epic battle of wills) or are only very few children actually capable of it? Robinson is on stronger ground by insisting that the standards of public schools in the past were much higher - but high schools with those standards also had much lower graduation rates. Still, Robinson personally knew very well world-renowned scientists and concluded “The top performers in any human activity usually achieved their positions by very hard work. Moreover, they are seldom the most talented. The most talented often have a tendency to become lazy as a result of their talents and therefore to be surpassed by those who really work.” Almost certainly it’s a combination of genetics and hard work - who knows exactly what dosage of each, but if your family may have competitive genetics, why satisfy yourself with mediocrity - why not try the hard work?
Figure 1. On the other hand, as the old joke goes, they say hard work never killed anybody - but why take a chance?
This brings us to Bryan Caplan, a Berkeley undergrad and Princeton economics PhD dad who has famously written a book about how you should have more kids because you overestimate how much your interventions will contribute to their success (since genetics is so prominent) and underestimate how much you’ll want more time with your kids (and grandkids) when you’re old. Despite this, Caplan did the work to homeschool his kids in middle and high school, partially because he wanted to give them an enjoyable childhood free of the arts and crafts they hated in their prestigious but anti-intellectual institution, but also because he gambled that his particular interventions might yield results. Remarkably, despite no knowledge of Robinson, Caplan independently created something very similar!
Every morning of the Caplan regime began with 2 hours of math - justified because “most good jobs in the modern world require strong math skills, and very few kids like math enough to learn it on their own” and his sons were clearly on the college track. Art and music could be avoided for a better life but math drills couldn’t. Early on, Caplan tested (but did not grade) his kids every Friday to ascertain what needed work but subsequently adopted a different program in which his kids worked on their math on their own for 2 hours every morning and then, very contrary to Robinson, got direct feedback from dad. But the general instruction was to try their best, then ask each other for help, then google for help, and only ask dad as a last resort. This probably sounds familiar: “If your students do 95%+ of their problems correctly, they are ready to advance.”
Figure 2. But if you never take art, how can you prompt AI to make art for you?
The Robinsons skipped lunch; Caplan would take his boys out to a long, healthy lunch with his colleagues at George Mason University, arguably (I would argue) the best economics department in the world, “plausibly four standard deviations above whatever peer group they’d have in a conventional middle school.” Robinson’s kids had considerably less social interaction outside the family - but did include regular visits with Nobel science prize winners.
And then, in the afternoon, the Caplan twins would get assigned reading - but it was based on their interests, in this case, less literature and more history and economics. Indeed, the boys wanted assignments because they were interested in subjects but didn’t know where to start. In a variant of Robinson, the boys wrote about two essays a week, which got line-by-line feedback from Dad. Partially to prove to outsiders (especially college admissions departments) that the homeschooling regime was rigorous, the boys took lots of APs. Caplan would generally have his boys read the best books about the subject over the course of a year (and interestingly, also had a fondness for earlier texts, using Carlton Hayes’ 1916 book to study most of what they needed for AP European History) and then do more specific cramming ahead of each one. Notably, the Caplan boys took calculus before physics (or any other science AP). The approach took Caplan 50 hours a year to design the curriculum and time each week to grade, especially the line by line essay comments. His twin boys got full-merit scholarships to Vanderbilt and plan to become college professors.
You can surely see why I group these two together, though there are real differences. Caplan insisted that “Every student gets personal feedback with their math every day.” Robinson insisted: “The student should never be helped with any problem. He should learn that he must always study the book by himself until he can work the problems.” But both would agree with Caplan: “If they’re struggling, they get extra practice. If they’re excelling, they move on to the next level.” Unlike Robinson, who insisted that he trained the character of his “average” kids, Caplan agreed to homeschool because his sons were naturally well-behaved intellectual introverts - but warned in advance, if there was misbehavior, they’d be sent back to an institution. Caplan was not a sugar nor television authoritarian (indeed, while not quite the opposite, he certainly shared his love of ice cream and certain pop culture with his kids - so far, this has not led to problems!). Both PhD Dads mandated reading and writing but Robinson eventually assigned a specific book list while Caplan said “I don’t especially care what they read or what they write about. Indeed, the best route is if they read and write whatever excites them most.” Indeed, Caplan remarked “While I think history is a waste of time for 99% of people, I think my sons are in the other 1%.” Caplan very reluctantly immersed his kids in Spanish - but just because it was basically a college requirement everywhere (I am not sure how Robinson’s kids beat this, because Robinson was clearly very sensitive to the opportunity cost of learning a foreign language. All the Robinsons went to state school undergrad, clearly negotiated certain things that I would have thought were non-negotiable, and zipped through in just a couple of years). Although Caplan is not religious, he cares deeply about character about his sons being “great husbands and great fathers” - though he certainly suspects there’s more genetics to this than what he can intervene in than Robinson.
The “PhD Dad” approach I’ve described is of course notable for the fact that it was innovated by two highly-educated parents with highly-educated spouses - but an underappreciated aspect is also that, unlike most homeschooling families, the dads did all of the homeschooling. You can draw what you will from potential gender differences in choices they’d make - but the more important aspect is that these dads also sought to work full-time (though with considerably more flexibility than the average job). If they had been stay-at-home teachers, would they have designed something different? Would it be better? Robinson insists that his family would not have pursued what they did had his wife survived, but that they, by necessity, innovated something that he now believes to be superior. Somewhat amusingly, given how rigorous the regime is, the Robinson curriculum now advertises itself as a way for moms to avoid the burnout of having to teach all the subjects themselves. I should also note that both Caplan and Robinson in their main careers are famous for being contrarians and embracing it. Robinson’s version: “We want our children to be different. We want them to be different spiritually, academically, socially, mentally and physically from the norms that are currently established in the secular world.”
When you consider the output of this approach, one major argument is that it is a more precise hack to get into the best college than the Traditional school at home curriculum. But Caplan would argue that numeracy and literacy are the keys to success in modern society and the deeper you go, the more opportunities. Robinson would add that scientific literacy (which is only possible with deep numeracy) is necessary so as to minimize your dependence on bad experts. “Our society is filled to the brim with public school graduates who imagine that they are independent thinkers when they actually are programmed to believe anything they perceive as fashionable.”
You can be well within the PhD Dad framework with a program that focuses on just math, reading, and writing even if you don’t pursue an exact replica of Robinson or Caplan. If your child really hates math or has little facility for it, you might abandon the harshness of the two hour a day regime - but it’s almost certainly still worthwhile to have some substantial daily math requirement at some point until (a) your child covers everything on the SAT or (b) you very intentionally decide to shut off college options. Statistics is also especially helpful to understand. On the other hand, if your child is really good at math, the Saxon method in particular could involve lots of busywork (90% every day is review!) and they may need something much more accelerated. One online program called Math Academy seems to be responsible for about a fifth of kids taking AP Calculus BC before 16. I don’t know what Robinson would think about that, but the approach certainly seems to be within the spirit.
But how might other homeschooling approaches critique and comment on PhD Dad? (Note that this is imaginative speculation, not quotation, from each point of view)
Evidence-based Traditional: Genetics aside, spending more time on math will result in better numeracy - but it’s difficult to evaluate the scalability of something that worked for two very unusual families. We only really have a broad overview, anyway - Robinson’s method is described in a janky website from 20 years ago, Caplan’s method in a few scattered blog posts. Yes, Robinson claims 60,000 users but they haven’t been studied and even his old forums suggest much deviation and thus it’s impossible to determine base rates of success. At best, we have a couple of very interesting case studies - and they are admirably focused on mastery, which is very well-established. But Robinson’s extremity on self-teaching, while provocative, is not well-established - indeed, what we know from much wider analysis is that the most powerful intervention discovered is 1:1 tutoring by an adult working through direct instruction. Robinson’s math-before-formal science approach is probably sufficiently radical to be unstudied - but his ideas about initially learning science informally from sources prior to 1960 are decidedly non-mainstream and ideological (though, yes, untested). Admittedly, Traditional school standards have dropped over time, but it still seems unlikely that any “typical” child can master calculus as fast as Robinson projects - and also be ready to consume formal science from CalTech, one of the most selective institutions on the planet. Neither Robinson’s reading list nor Caplan’s haphazard reading approach will give your kids the modern and broad cultural literacy they need to truly thrive. But if we are evaluating this strictly on the basis of trying to get into the best college, this program is fairly well-suited, if only because it focuses on what standardized tests capture (in particular, it’s more sensitive to math than any other approach). Calling Robinson “back to the basics” can also be a bit of a misnomer since he basically does wind up covering all the subjects except foreign language and, ironically given his name, art. Ultimately, avoid what a couple of legendary contrarians proffer and instead lean on what has demonstrably worked.
Classical: PhD Dad fails to pass along Western civilization. Caplan is a barbarian who puts some guardrails on unschooling but freely admits that 99% of kids could avoid history - yet history is the proper backbone of a curriculum, revealing not just the how but the why of humanity. Robinson is only a little better: his formal reading list may be heavy on 19th century books but it’s light on Great Books, which he inaccurately thinks can just be replaced with math and then science to train minds - but anyone can easily see that wrestling with a difficult text in the humanities engages a different part of the mind than with a math problem. When your father dies and you are struggling with your emotions, calculus won’t help you. They frankly overemphasize math, which is just one part of a broad curriculum for a fully educated person, and Robinson fails to understand how children can learn science in different stages of Trivium development. Both PhD Dads fail to appreciate the immense benefits of seeing the genius in art, which is not just elementary classroom “arts and crafts” - it can uplift the soul. PhD Dad shapes contrarian cranks when you should be shaping virtuous citizens. Don’t just tell your children to build, on their own, a narrow tower with a few required support beams; instead, work with them closely to establish a firm foundation for a magnificent cathedral that will stand the test of time.
Figure 3. Some of the other homeschooling approaches think of PhD Dad as the equivalent of an educational conspiracy theory they’d prefer not to deal with. But what if the earth really does revolve around the sun? Or washing hands really does reduce hospital fatalities?
Charlotte Mason: Robinson is properly invested in the most significant habits (obedience, attention, and truthfulness) but there are lots more and he leaves too much for other parents to figure out on their own. Overall, “there are more things in heaven and earth, [PhD Dad], than are dreamt of in your philosophy.” PhD is devoted to the drudgery of decimals unnecessarily - orderly, accurate, and precise thinking can and should be trained through the humanities and, while some math is necessary, PhD completely overdoes it, especially in requiring so much of every child. Relatedly, PhD Dad has absurd focus expectations; children should be switching up topics relatively frequently (depending on their age) and expecting a 7 year old to work on math for two hours is deadening. PhD Dad fails to set out a gentle feast that cultivates in children a love for many things - and Caplan is especially dangerous insofar as he advises following a child’s interest rather than training a child’s interest on the best literature. While you should aspire to a child drawing their own conclusions, PhD Dad goes about this the wrong way by almost removing adults from the picture; instead, adults must work closely with children to get them to narrate significant works in their own words. This method is far better than just reading on one’s own and taking an SAT-style test - what if he fails? Does he have to read again? Importantly, Robinson’s selections are old enough that Mason herself would recognize most of them and yet there is probably low overlap between her original list and his. Don’t just dispatch your children to slog through each day; draw out from them great stories in the morning and then take them out into the nature every afternoon.
Unschooling: Let’s be crystal clear - forcing your child to sit in a chair for 12 hours a day, 6 days a week, for 2 months, as Robinson did to his child resisting his program, is abusive. Robinson and Caplan have a ridiculous preoccupation with compelling math, much more math than anyone other than a university professor uses in their life. The Robinson math curriculum is many children’s version of hell. When math is of interest, if it is useful, children will naturally want to learn it and will. Though reading should not be mandatory, Robinson’s original program of having his children just select books from the family library was the right approach; Caplan’s method of finding the best books for his kids’ interests is great (and Robinson should lean more into what happened with his ant example). Traditional education gives every child the same sized shoe and calls that efficient - PhD Dad has the same problem, the shoe is just less decorated.
Mentors: Taken too literally, PhD Dad is for parents who are trying to minimize the time they have to give to their kids - and that is the wrong thing to optimize. The Robinson curriculum requires no more than a proctor; adult contribution almost ends as soon as it starts. Instead, you should embrace the adaptability available in formal tutoring, where an expert adult can make judgments about a child’s learning instantaneously and continuously. (Caplan comes closer, but still expected most learning to be done alone.) Of course, both Robinson and Caplan did in fact ensure that their kids were informally exposed to truly top-notch adults (including themselves!) - Caplan every single day at the GMU economics faculty lounge and his twins also started a podcast where they conducted interviews with world renowned experts. Robinson socialized his kids to some of the world’s leading scientists, including Nobel winners. And that’s the point: don’t think you can just do the three Rs in the privacy of your home and call that the end. Admittedly if this accelerated program works, your kids may be more interesting to geniuses in those fields - but that doesn’t change the fact that genius exposure should be your goal.
Montessori: Robinson’s method of self-correction is actually a core Montessori principle but Robinson only believes in children’s capability, not their dignity or autonomy. Robinson demands focus rather than curates it, comes up with a conclusion rather than following the child. Caplan’s variant of PhD Dad is considerably better by giving children high quality choice based on their interest. And help me do it myself does still imply some help, and Caplan better accomplishes that. Robinson started much earlier than Caplan and yet still starts too late (around 7 years old) and too roughly dismisses hands-on learning as a crutch when it can be crucial to actual understanding - Robinson has far too much faith in learning on paper. Relatedly, Robinson dismisses too easily early science that can captivate.
Specialists: Most charitably, this is a specialized math regime - and if you want your kid to get a PhD in nuclear engineering, have at it. But if your kid doesn’t have the knack for math, this is still way too much time spent on things that are not your specialization - multiple hours a day, six days a week! The self-help ideas are probably not applicable in most specializations, but Robinson is essentially correct in his approach to getting kids to concentrate. Overall, the Robinson curriculum admirably and aggressively cuts the fluff out of the curriculum but still promises to serve as a base to do pretty much anything rather than specializing in something very particular. Caplan did tailor for his boys’ interest, ambition, and talent in a valuable way and, insofar as his boys considered being economists, the math may have been appropriate - but it may have been excessive if they were determined to become historians.
Figure 4. Might we say that a variant of Amy Chua’s “tiger mom” is Art Robinson’s “mountain lion dad”?
Unit studies: Maybe you do have to separate math into its own silo but it does not need to be nearly as intense or the overall focus that PhD Dad makes it out to be. “Pretend science” is way too dismissive - you’re telling me that you don’t understand the basic idea of survival of the fittest because you haven’t mastered calculus, physics, and chemistry? PhD Dad is a narrow tunnel to knowledge that misses out on interdisciplinary insights that promote cognitive flexibility. It is way too abstract, relying on pure math and dated reading to understand the real world. And children need adults to be partners in drawing connections, not just telling them to hit the books.
Pragmatists: Robinson’s kids at least grew up on a farm, but the accelerated math program accelerates into total impracticality. Admittedly, the “trust and parrot” problem is real - but Robinson’s solution requires an intense amount of work to avoid it; it may be better to figure out some rules of thumb about whom to trust. Reading about the Civil War is ridiculous as a requirement - that is deep in extracurricular interest category - and so is much of Robinson’s reading list (though Nuclear War Survival Skills could arguably be useful way down the line in a curriculum for a low-probability event - it’s actually his extracurricular recommendations where the value is at - though perhaps not the 1911 Encyclopedia). There’s already too many academic days - you don’t need to add 50% more! Robinson is admirably attentive to self-help, but probably overdoes it; not every child is going to just claim chores for himself. Robinson certainly has an admirable work ethic - and his kids really did get the most practical of intellectual degrees - but, again, where’s the budgeting and marriage curriculum? (He does manage to give a bit of dieting advice!)
Pioneers: Admirably math focused, but needs to be open to a lot more hands-on experiences. Understand the argument for delaying formal science but should have lots of it as an extracurricular (e.g. blowing things up, rocketry, etc). The Robinson reading list is actually fairly boy-oriented, but Caplan is probably better to follow interest. Both Caplan and Robinson mention that their kids had time outside, but would be helpful to make it an explicit part of PhD Dad. Robinson is too skeptical of outside organizations, though his point that contributing to the well-being of the family on the farm was better than competitions may still be true for boys. Robinson is admirably attentive to the problems with diet and TV and the overdiagnosis of disability (AR: “They would have said Einstein had a learning disability”) - but there’s very obviously some genetic component here, not everyone can focus on math every day for 2 hours. Still, the best part about PhD Dad is that Dads are doing it! Boys need male role models.
Bible First: Robinson admirably features Bible study every single day - even to the point where he would call home from business travel to participate and, later, kids at college would do the same. Math and science are beautiful further revelations of God’s handiwork in constructing the universe - but they also need that context. Robinson seems to have been attentive to this in his own home, but it could be made more clear. But, ultimately, you can’t just have the Three Rs (arithmetic, writing, reading); the Amish wisely add another three: religion, respect, and responsibility.
Cyberschoolers: Art Robinson’s curriculum wouldn’t be meaningfully different in 1949, which still leaves a lot to go. It’s wild to deny children in the computer age access to computers until they’ve mastered calculus - which may not happen till they’re out of the household. This is especially problematic because computers can better manage a personalized algorithm than anyone just working with flashcards. Math Academy is a great example of an online math program that has real demonstrable results. And why shouldn’t a child be able to work through MIT’s Courseware, all available online? If you like out-of-copyright books because they’re free, let me tell you about YouTube. Don't navigate the information superhighway with a horse and buggy (or even a 1949 OIdsmobile!)
My own takeaway: while I intend to steal my favorite parts of other approaches, this is the best base model to go “from phonics to physics.” In fact, I essentially described it before knowing about it and then was directed by an experienced homeschooling mom to look into Art Robinson. It has its quirks but you don’t need to pursue every aspect of the Robinson regime to recognize that the three Rs are the essential aspect of education and a foundation upon which all else can be built. It’s intuitive to me, at least, that math can be a slog where expertise doesn’t just naturally happen due to interest. On the other hand, although I find Robinson’s reading list interesting, I do think you can follow interest better with reading, so long as you maintain quality, and Caplan has the better approach: you pick the topic, I pick the book. Indeed, I suspect that a lot of the right answers in education can be found between these two. We’ll see if genetics cooperate with these ambitions to be a “JD Dad”, though I am counting on my wife’s 98th percentile math scores being passed along!
Figure 5. ChatGPT’s impression of the most stereotypical students of PhD Dad education.
Further reading:
To read Art’s (excellent) book and get access to all his materials, you have to subscribe to his curriculum, but you can get a substantial flavor for it by working your way through his website page by page on the left. Check out, in particular his rules to achieve remarkable results
Here’s a Youtube video of Art explaining his approach to a homeschooling group. Also includes some personal details about the family, like how although he strongly believed his kids should not have sugar, he and his wife couldn’t give it up. They’d send the kids to bed and then bake cookies for themselves and eat them all before morning. After his wife died, Art would have a cookie stash at his desk and noticed that cookies were disappearing. He lamented that his failure to give up his bad habit had now led at least one of his kids to develop it too. He investigated and discovered that indeed one of his kids was stealing the cookies…. and throwing them away because he thought dad was more irritable when he had too many cookies.
The McGuffey Readers taught millions of Americans to read and are a good flavor for the kind of books Art includes.
Bryan Caplan’s homeschooling odyssey
Bryan has not as systematically described his homeschooling approach as Art, but you can find much of his material by searching for “homeschool” on his blog, Bet on It. Some highlights include the above linked odyssey overview, a portrait of his school, the Caplan family school attendance contract, why he decided to homeschool, a (covid) emergency guide to homeschooling.
When his twin boys “graduated” home middle school, he interviewed them. They lament that their mother is forcing them to go back to regular school and even the normal subjects seem all to be arts and crafts. They are not happy about being force-marched back into school spirit, either. Luckily, after this podcast aired, within a couple of weeks of regular school, their mom relented and they homeschooled until getting into Vanderbilt.
This blog covers a third highly-educated Dad’s journey in homeschooling his children to success. This entrepreneur met his wife at Oxford (he studied physics, she classics). Their regime was focused on math and reading and foreign languages (French, Arabic, Chinese, Urdu, and Malay), a mix of intense self-teaching and 1:1 tutoring. Once his children had mastered enough math, he put them on a five month crash course in physics, which they had never studied before - and then had them take the British standardized test for physics intended for those aged 16 and up who had taken two years of physics classes. His son became “the youngest person in the 30-year history of the” test to have achieved the highest score. He was 10 years old. They had a series of these type accomplishments on British standardized tests and his kids all got into the two best schools in the UK, one daughter into Cambridge at 15 years old. The program has a bit of variance from the PhD Dads I describe but I think still is comfortably within the overall category of cutting the extraneous, really mastering the basics, and proving the concept on standardized tests. Of course, this is also yet another example of highly educated parents - but if those kids had gone to a typical school (or been subject even to a typical homeschooling regime’s pacing), would they have had the same results?
Thanks for reading! If you enjoyed this, forward it to a friend: Know anyone who is thinking about homeschooling their kids? How about anyone with kids? Or do you know anybody who was once a kid?
If you liked this, be sure to check out my other newsletters, especially the others in this series, which started with this introduction that will be updated with links as they are released: How Homeschool?
If you’ve received this email from a friend and would like more, sign up at www.grantreadsbooks.com. I read over 100 non-fiction books a year (history, business, self-management) and share a review (and terrible cartoons) every couple weeks with my friends. Really, it’s all about how to be a better American and how America can be better. Look forward to having you on board!