Learning, education, and school are tricky. We definitely want more learning to happen, but figuring out what it should be and look like is harder.
What We Want It to Do
More learning is good for both society and the individual.
From a societal standpoint, the reason we definitely want more learning is because it accounts for 70 to 80 percent of improved living standards over the last 200 years. That is according to Nobel Laureate Joseph Stiglitz who believes we should aim to create a “learning society” as a result. He explains,
most of the increases in standards of living are a result of increases in productivity—learning how to do things better. And if it is true that productivity is the result of learning and that productivity increases (learning) are endogenous, then a focal point of policy ought to be increasing learning within the economy; that is, increasing the ability and the incentives to learn, and learning how to learn, and then closing the knowledge gaps that separate the most productive firms in the economy from the rest. Therefore, creating a learning society should be one of the major objectives of economic policy. If a learning society is created, a more productive economy will emerge and standards of living will increase. (Kindle Locations 313-318)
Much of the learning Stiglitz discusses relates to “learning by doing” or on the job learning. He points out that the industrialization process itself is responsible for much of the “learning how to learn” that societies need. That makes it harder for nations to innovate when they skip over the industrialization process and go from agricultural to service economies.
From an individual standpoint, psychology research has repeatedly included mastery, competence, and accomplishment as factors that consistently contribute to people’s happiness and drive. From this viewpoint, people have a natural tendency and joy to learn when not being coerced to do so and more learning therefore results in more overall peak experiences, or what psychologists call flow.
Besides living in a richer society with happier people, we probably also hope that education decreases criminal behavior and increases overall health and longevity. We’ll see if any of these wishes are realized, but let’s first look at some costs of providing education.
What It Costs Us
Public education in the United States is expensive. According to the National Center for Education Statistics, “Total expenditures for public elementary and secondary schools in the United States amounted to $620 billion in 2012-13.” Looking at calendar year 2013-14, “postsecondary institutions in the United States spent approximately $517 billion (in current dollars). Total expenses were nearly $324 billion at public institutions, $173 billion at private nonprofit institutions, and $21 billion at private for-profit institutions.”
If we only look at the public institution expenditures, then elementary through postsecondary institutions spent a total of $944 billion. The numbers for elementary/secondary and postsecondary are one year apart, but the general expenditures don’t swing wildly. The United States spends roughly a trillion dollars on education each year, equivalent to over five percent of total GDP.
Using the same sources above, these numbers amount to $12,296 per public school student in elementary and secondary and $30,502 per public university student. These figures are roughly 30 and 80 percent higher than those of countries in the OECD. This shows these numbers are not just absolutely high, but also comparatively so. Imagine what a family could potentially do with that type of money each year if it was given directly to them by the government and not spent on their behalf.
Two Theories of Education: Human Capital and Signalling
There are two primary theories of the effects of education in economics, although some economists do have their own pet theories. The theories of human capital development or screening and signaling theory are summarized below,
The relationship between education and earnings has long intrigued economists and in recent years two contrasting views have emerged. The theory of human capital holds that education directly augments individual productivity and, therefore, earnings (Schultz, 1961; Mincer, 1974; Becker, 1975). By forgoing current earnings and acquiring, or more precisely, investing in, education, individuals can improve the quality of their labour services in such a way as to raise their future market value. Human capital, according to this view, is akin to physical capital, the acquisition of which entails a present cost but a future benefit. Thus education may be regarded as an investment good, and should be acquired until the point at which the marginal productivity gain equals the marginal opportunity cost.
Most people are shocked the first time they learn of the signalling theory of education and immediately dismiss it as ridiculous. How could it be possible that school doesn’t develop human capital in some capacity? Instead of rejecting it out of hand, keep an open mind as some evidence is presented below.
Behavioral Genetics, or a Look at Twin Studies
Bryan Caplan has written a couple fantastic books, but the one that concerns us here was an attempt to summarize the state of knowledge as gleaned from twin studies. His point was to emphasize that parents should chill out regarding their self-induced stress over parenting, but the information could just as easily be applied to education and students; it had little to do with parenting per se and more to do with the effects of genetics and environment on life outcomes.
He makes sure to state a clear caveat,
What you find depends on where you look. Almost all twin and adoption studies are set in developed countries. Families that adopt are usually middle class, and always want children. It would be irresponsible to read these studies, then tell the world that child abuse does no lasting damage or that your child will turn out equally well if he grows up in the Third World. (p. 165)
I’ve summarized his points below. Caplan writes that,
Environment has an effect on:
Environment doesn’t affect:
Environment has little or no effect on:
As one can see, environment has zero to little effect on many of the things we find most important, including income, happiness, criminal behavior, and health. As “school” is a part of the environment, we really would hope for the opposite of these findings if we want to endorse the human capital model of education explained above. Let’s look a bit closer at a few of these points.
According to economist Branko Milanovic, we also know,
a person’s income in the world [can be explained] by only two factors, both of which are given at birth: his citizenship and the income class of his parents. These two factors explain more than 80 percent of a person’s income. The remaining 20 percent or less is therefore due to other factors over which individuals have no control (gender, age, race, luck) and to the factors over which they do have control (effort or hard work).
So learning is important for innovation, which increases living standards as noted by Stiglitz above, but income is mostly accounted for by geographical location and parents’ income class, with very little left over for one’s personal effort, in school or otherwise.
In addition, Bryan Caplan elaborates that,
Yes, wealth and poverty run in families. According to twin and adoption studies, however, the main reason is not upbringing, but heredity. Nurture has even less effect on income than on education.
None of the above is good if we are hoping for education and school to have large impacts on incomes. It gets even worse for the education optimists though,
Dale and Krueger concluded that students, who were accepted into elite schools, but went to less selective institutions, earned salaries just as high as Ivy League grads. For instance, if a teenager gained entry to Harvard, but ended up attending Penn State, his or her salary prospects would be the same.
It therefore seems that specific colleges aren’t doing much in particular to increase incomes for people, something both Milanovic and Caplan would have assumed. Instead, it is likely that general intelligence as estimated by the SAT and the ability to be selected in the first place is what matters. This would give credence to the idea that schooling is simply a screening and signalling device for talent and other socially desirable characteristics like conformity.
Perhaps rather than increasing income, school increases intelligence. We saw from Caplan’s list above that environment had zero effect on IQ. Again, this seems very counterintuitive for most. What can we find if we look?
If we take a look at critical thinking, reasoning, and communication as skills that would assumably help future citizens deliberate and decide, we find that,
Richard Arum and Josipa Roksa reported that most students didn't improve after four years of college, and a full third didn't improve at all. Now they've written a follow-up, which concludes, unsurprisingly, that students with high Collegiate Learning Assessment (CLA) scores do better in the job market than students with low scores.
This means that students with better thinking and reasoning scores on the CLA do better in the job market, but that those abilities are hardly or not at all improved after four years of college.
To help explain intelligence, Daniel Willingham writes in The Atlantic,
Intelligence has two components. One is akin to mental horsepower—how many pieces of information a person can keep in mind simultaneously, and how efficiently that person can use it. Researchers measure this component with simple tasks like comparing the lengths of two lines as quickly as possible, or reciting a list of digits backwards. The other component of intelligence is like a database: It entails the facts someone knows and the skills he or she has acquired—skills like reading and calculating. That’s measured with tests of vocabulary and world knowledge.
The above may tempt us to infer that while we don’t seem to be very good at teaching critical thinking or increasing mental horsepower, content matters enough for increasing intelligence that we should focus even more on the content we deliver. That’s not the right conclusion, though. Knowing more is good, but facts don’t have to come from teachers and many of the facts that do aren’t remembered. All of this fits with findings on expertise, where factual knowledge is known to be the bottleneck.
Daniel Willingham also tells us that retention of content is also quite low, leveling off around 30 percent of content learned in class for A students and around 20 percent of content for B students. In Why Don’t Students Like School? he writes,
In one study, researchers tracked down students who had taken a one-semester, college-level course in developmental psychology between three and sixteen years earlier. The students took a test on the course material. Figure 5 shows the results, graphed separately for students who initially got an A in the course and students who got a B or lower. Overall, retention was not terrific. Just three years after the course, students remembered half or less of what they learned, and that percentage dropped until year seven, when it leveled off. The A students remembered more overall, which is not that surprising—they knew more to start with. But they forgot just like the other students did, and at the same rate. (Kindle Locations 1955-1961)
If most of the content teachers are delivering is forgotten, where should students get their knowledge?
Sugata Mitra’s research informs us that kids can learn quite a bit on their own when given access to a computer and the internet. He writes,
When working in groups, children do not need to be "taught" how to use computers. They can teach themselves. Their ability to do so seems to be independent of educational background, literacy level, social or economic status, ethnicity and place of origin, gender, geographic location (i.e., city, town or village), or intelligence.
In addition, Stephen Krashen’s research on language acquisition and literacy shows that those that read more:
The “more reading” Krashen refers to is “free voluntary reading” where students can select their own texts and difficulty level freely. He also states that results do not improve if teachers use rewards, lexiles, error correction, or supplement with writing, although less writing apprehension is experienced by the free voluntary readers. Finally, he also notes that surfing the internet can increase reading and that greater access to books of all types does so as well. Many of these findings helped me explain why reading 100 books in a year was the fastest and most effective way for me to learn more than any other period in my own life.
The reason most teachers attempt to handpick and deliver content is the hope of learning transfer. It is somewhat the holy grail of school. However, learning transfer is hard and not very reliable. Furthermore, we don’t know what content will be particularly useful in the future as technologies change and some jobs disappear, while entirely new ones appear.
According to the Committee on Developments in the Science of Learning,
the ultimate goal of schooling is to help students transfer what they have learned in school to everyday settings of home, community, and workplace. Since transfer between tasks is a function of the similarity by transfer tasks and learning experiences, an important strategy for enhancing transfer from schools to other settings may be to better understand the nonschool environments in which students must function. Since these environments change rapidly, it is also important to explore ways to help students develop the characteristics of adaptive expertise. (p. 73)
When research is conducted on actual transfer of learning, David Perkins summarizes,
a superficial look at how research on transfer casts its vote is discouraging. The preponderance of studies suggest that transfer comes hard. However, a closer examination of the conditions under which transfer does and does not occur and the mechanisms at work presents a more positive picture. (p. 10)
Perkins also writes,
In many situations, transfer will indeed take care of itself - situations where the conditions of reflexive transfer are met more or less automatically. For example, instruction in reading normally involves extensive practice with diverse materials to the point of considerable automaticity. Moreover, when students face occasions of reading outside of school - newspapers, books, assembly directions, and so on - the printed page provides a blatant stimulus to evoke reading skills.
This description sounds much closer to the type of learning that Stiglitz and other economists have in mind when talking about increased productivity via learning by doing. “Hugging” could be summarized as learning by doing the thing you want to get better at, while “bridging” could then be thought of as actively reflecting and seeking out ways to improve that thing. A teacher would then be acting as a coach and the model would look nearly identical to that of deliberate practice within the field of expertise studies.
We know from John Hattie that almost anything a teacher does “works” for enhancing academic “achievement”, which is another way of saying student grades. He then moves on to make the further point that since everything works for enhancing achievement, asking what works is the wrong question. We should then ask about what works more than the average intervention.
Hattie measures interventions using effect sizes and finds that 0.4 is the average effect size in educational settings and encourages schools and teachers to then focus on effect sizes greater than this, which he kindly lists based on his research.
To give the reader an idea of the educational influences with the largest effect sizes, the top 30 as of 2015 are:
Don’t worry too much about what all the influences mean. It’s not always clear that some of these influences are causal in any way and critiques of Hattie’s entire method have been made.
However, what I’d like to point out by referring to the bolded items in the list above is that many of the strongest influences on achievement overlap greatly with the ideas mentioned in the section on intelligence and transfer of learning. We want students spending a lot of time doing the thing they want to get better at with feedback and assessment from clear teacher instructions, much like coaches in the deliberate practice model or hugging and bridging from Perkins recommendations for transfer.
These ideas seem routine, but they’re also somewhat of a tacit admittance that teachers can’t really force change. In fact, some of the most progressive learning theories are those of constructivism, which all but admit that students must learn for themselves and cannot be “taught” in the traditional sense. These are strange angles to approach school from in its traditional role as developer of human capital or increaser of student ability.
GPA and Tests
So we do have some sense of what is important if our goal is to increase grades, but that doesn’t necessarily mean that grades predict anything of value or matter for long-run life outcomes such as happiness, income, and health. As we’ll see below, it is not obvious that grades tell us much of anything about a person’s future prospects and life outcomes.
While all of the effect sizes mentioned by Hattie are essentially measurements related to grades, it is important to keep grades in their proper place. In an interview with Laszlo Bock, senior vice president of people operations at Google, Adam Bryant reports,
One of the things we’ve seen from all our data crunching is that G.P.A.’s are worthless as a criteria for hiring, and test scores are worthless — no correlation at all except for brand-new college grads, where there’s a slight correlation. Google famously used to ask everyone for a transcript and G.P.A.’s and test scores, but we don’t anymore, unless you’re just a few years out of school. We found that they don’t predict anything.
So even if we sort of know what works for increasing GPAs and test scores when looking at research, employers are finding those numbers less and less relevant.
School doesn’t do much.
It doesn’t increase income. It doesn’t increase intelligence. The content it does deliver is mostly forgotten. Most of it could be learned independently anyhow. School may be entirely signalling. Teachers can affect achievement as measured by grades with the biggest effects coming from a coaching, deliberate practice type model. However, grades aren’t too important as employers are becoming more and more aware that they have no predictive ability for job performance and students can always opt to go the community college route without stressing over them.
So how should we think about all of this? For now, I think it’s best to relax and lighten up about school and grades. Learn for learning’s sake. It makes you happier as stated at the beginning of this article, and if you have the talent and drive to learn at the cutting edge of STEM fields, you can also make the world better through your innovations. If you don’t quite have those abilities, perhaps Charles Murray’s ending paragraph from Real Education can help redefine aims,
Educational success needs to be redefined accordingly. The goal of education is to bring children into adulthood having discovered things they enjoy doing and doing them at the outermost limits of their potential. The goal applies equally to every child, across the entire range of every ability. There are no first-class and second-class ways to enjoy the exercise of our realized capacities. It is a quintessentially human satisfaction, and its universality can connect us all. Opening the door to that satisfaction is what real education does. (Kindle Locations 2095-2099)
Perhaps more than anything else, we should think really hard about the expenditures per pupil summarized above. What are we getting for $12,296 per public school student in elementary and secondary? Would that money be better utilized if simply given to students to do as they wished as a sort of universal basic income for minors? Would we get all the same results if we made use of credentials, certifications, and online curricula instead of diplomas? Would the money be better spent entirely focused on preventive health in order to make a sizeable dent in the chronic diseases that make up roughly 85 percent of the $3 trillion spent on healthcare each year? These questions are difficult, but looking at the evidence is important if we wish to begin answering them with the hope of improvement.