Bryan Caplan’s The Case Against Education argues that employers value college degrees not because college imparts skills and knowledge, but because a degree is a good signal that someone has a pretty high IQ, is conscientious, and mostly conforms to social norms.

Simply put: going to college helps you get a job not because of what you learn, but because it tells employers about who you are.

I found Caplan’s arguments both persuasive and a bit depressing. I get that from an employer’s perspective, it’s rational to focus on prospects with college degrees. I get that on average this sort of statistical discrimination is a smarter bet.

But even if this decision is rational, it still sucks for the many people who didn’t complete college but would do just as well in a given job.

I’ve had a few non-academic jobs since I graduated college over 15 years ago. My big take-away is that job success is multi-dimensional. I think the major dimensions are IQ, skills, knowledge, personality and talent. This metric is admittedly heuristic. But even if you quibble with the components, it captures the idea that job success depends on many factors.

Note that IQ is simply one of the components. It may be the most important. But it’s not everything. The smartest people aren’t always the most successful at a given job. (I describe a variation of this rubric in the clip above.)

Here’s the rub: IQ is probably the most important factor for college. The smartest people do tend to be the best students. I suspect that the academic approach to learning works for only the top ~20% of IQs, and so Caplan understates IQ as an initial filter. Only after eliminating the bottom 80% of IQs does college add in conscientiousness and conformity as filters.

But again…that really sucks for a lot of people. Many jobs don’t honestly require that much smarts. Even jobs that allegedly do (e.g. engineering) make coursework harder than it needs to be.

During college I tutored in the Penn State math center. I remember working with a forestry major who had to take calculus. Not the class for engineering or science students, but a much easier version.

He was a good kid who really wanted to do well. He came in several times for help, and even paid me to work with him one-on-one. But he still struggled, and didn’t pass the mid-term.

I’ve been thinking about this student a lot. What I keep coming back to is: why on earth do forestry majors need to take calculus?

Unless a significant number of forestry majors actually have to use calculus during their job (I’m doubtful), it shouldn’t be a requirement. Making it so is simply an unfair barrier to entry.

Many colleges have a co-op program for business and engineering majors. It allows students to get much more work experience than they typically would. Starting in their junior year, co-op kids will work every for 6 months instead of taking classes, ultimately delaying their graduation by a year.

I love the concept–anything that helps prepare students to work is a good idea in my book. But I would push it even further. As it stands, co-op programs still fit within the four years of school ideal. Co-ops simply tweak that model to add work experience on the margins.

Why not flip the basic approach? Instead of starting with two or three years of classes, you could begin with 3 to 6 months of work, and add in just as many classes as needed to be more productive in the next co-op. The ideal could be a minimum of three years of work, which could then be tweaked to add education on the margins.

How much school would students need? It would ultimately depend on the particular jobs, majors, and even geographies at play. And so I don’t have a definite answer. I can, however, offer a framework for how to approach the problem, from my time as a software product manager.

When you’re taking a new product to market, you have an almost infinite list of features that can be in the first release. How do you decide what should be included? A helpful starting point is to recognize that you don’t actually know. You can and should analyze the customer base, similar products, etc.

But there’s a good chance that what you think customers want won’t be what they actually want. Too much up-front planning is thus a waste of time.

For this reason, modern software engineering increasingly favors creating the Minimum Viable Product (MVP). To get the MVP, you often think in terms of use cases. Some feature will meet 50% of the use cases, others 75%, and so on. You continually ask yourself: is this feature absolutely necessary for the first release? You eliminate as much as you possibly can.

You might end up with a release that only meets 50% of the use cases. That’s fine. You will also knowingly release something with bugs. That’s also fine. The goal isn’t perfection. It’s to get your product in front of actual users ASAP. They will tell you what features to develop next and which bugs to fix first. Done right, this approach minimizes waste and maximizes customer value.

While education is  not a for-profit business and shouldn’t be run like one, aspects of this mindset can be helpful. Many college majors seem to have adopted the opposite of the MVP approach, including much more knowledge than most students will ever use.

In terms of electrical engineering, I’d describe the approach as: if there’s a 10% chance any EE anywhere might use a certain snippet of knowledge, teach it to everyone without exception. I can think of no other reason to force EEs to take, e.g., intro chemistry.

A better approach would be to get students into the workforce as soon as possible, and for the education system to adapt, learn and change in response to what students actually want and need.

With all that, here’s a first pass on how I’d structure the MVP degree for electrical engineers:

1. Definitely include: calculus 1, physics I and II (classical mechanics and E&M), basic circuits, intro circuits lab, intro programming course, and some sort of project-based design course.
2. Definitely drop: any abstract math class like matrices, Laplace transforms, advanced circuits, microcontrollers, chemistry, chemistry lab, all general education requirements except maybe writing.
3. Probably should drop but I can’t bring myself to do so: advanced E&M (my area of research!), solid state physics, all advanced senior level courses.
4. Can’t make up my mind (which almost always means you can drop it): calculus II, differential equations, writing.

Note that the above is only what we’d require of students. If schmucks like me want to take plasma physics and quantum mechanics, we should be free to. And obviously the above list could change as we learn more. That’s the point actually.

But as a general goal, reducing course requirements is a good first step. I find it grotesque to force people to sit in a classroom for any longer than necessary. Degree requirements should be closer to an associates degree than what it is now.

Which brings me to my closing thought: If a BS or BA has already become the new high school degree (i.e. you need college for a middle class life), a useful corrective for us reformers is to reverse that trend. We should want to make the associates the new bachelors.

In an over-qualified labor market, employers will fill the “highest” jobs with those who have the “highest” credentials. Since over-schooling means there are too many workers who are highly educated, some of these workers are necessarily allocated to “mid-level” jobs. This process is repeated  for those with mid-level qualifications, where, since there are not enough mid-level jobs, many are forced to compete for low-level jobs. – Quoted in Bryan Caplan’s The Case Against Education 

An email exchange with Oren Cass made me want to answer a few questions for myself: Why should we even try to reform how college works? Won’t focusing on college detract attention from the “other half” students who most need a vocational path?

I’ve hinted at the answers to these questions in some of my other posts. But, mostly to help me clarify own thinking, I thought I’d try to summarize. In order of decreasing importance, here’s why we should include college in these discussions:

1. Credential inflation: As the quote at the top illustrates, credential inflation will eventually infect all jobs. We’ve already reached the point where file clerks need a college degree. I worry that at some point we’ll require everyone have some college education and there will be no true blue collar jobs left, a situation both Cass and I want to avoid. And so we have to fight the education obsession head on.
2. Higher ed is ignored: I’ve been working through (and enjoying!) Oren Cass’s book. While I find his overal framing and analysis novel, the notion of a vocational track itself isn’t that new.  Presidential candidates routinely call for more of it. But all these essays and public exhortations imply that the academic approach to learning (aka college) is perfectly fine for white collar workers. I disagree. Many college students would benefit from a greater work emphasis, and it’s a neglected line of analysis.
3. Forming a broad coalition: From a purely strategic standpoint, we have a better chance of rethinking education with as broad a coalition as possible. Since most kids in college want a job at the end of the day, it would be politically wise to include them. A work track does a better job on that point, and I suspect it would get more buy-in than a CTE track. I concede, however, that I am making an empirical claim that might be wrong. While my personal bias says a work track would be more effective in helping push the reforms both Cass and I support, it might not be. It’s possible that a large coalition would only detract attention from the CTE students who need the most help.
4. Devaluing the idea that education is intrinsically good: Regardless of whether a work frame is a more effective political strategy, I think it’s important to support it for philosophical reasons. There seems to be this widely held belief that education has some intrinsic worth. Beyond ~10th grade, I’m not sure it does. Work is more valuable for both individuals and society, and that’s what we should collectively promote.

Electrical engineers at Penn State have to take EE350, a course on Laplace transforms. It’s notoriously difficult–abstract and math intensive with tough assignments and tests. People often fail multiple times, and many drop out of electrical engineering because of the class.

The professor once explained why he made the class so hard. He said something like: “One day someone in this class will have to design air traffic control systems using Laplace transforms. If I pass you when you can’t do the work, then you’ll cause a lot of harm.” I remember agreeing with that basic logic.

I now see that my agreement was self-serving. I’m just lucky that I was able to do well in classes like that, and I supported a stance that privileged people like me.

The fact is that the EE350 coursework was much more difficult than what you’d encounter in most jobs. I’m guessing that in most cases you could get away with some pre-packaged software that requires only basic knowledge of LT. You definitely wouldn’t need to solve problems under time pressure.

The disconnect between EE350 as a course and its practical use isn’t surprising because it wasn’t designed for students to use the content. If that were the goal, we would have focused much less on theoretical concepts and mathematical wizardry.

EE350–like all of higher education–was designed by and for academics. They like the problem sets and theorems approach to learning, which is why the course exists as it does. Students who might make great engineers but don’t have the skill-set to pass EE350, or who don’t do well with the academic approach to learning, are simply out of luck.

For my senior design project, I worked with someone I’ll call J. He was one of the best engineers I’ve ever met. He had a special knack for debugging what was wrong: more capacitance in one part of the circuit, overheating in another, etc. In one of the few classes that attempted to replicate on-the-job engineering, he excelled.

But he also failed EE350 multiple times, delaying his graduation. Although I suspect he eventually passed, I don’t know what he, or anyone else, gained by forcing him to take the class. I’m not sure why should have had to.

There are many people like J–those who would succeed in white-collar careers but are hampered by the academic approach to learning. They too would benefit from rethinking college and work pathways, and reform efforts shouldn’t ignore them to focus solely on vocational students.

A genuine commitment to [an] inclusive society requires a willingness to shape institutions to that end, even when doing so is for the benefit of others or when it creates tension with other values – Oren Cass

As a conservative who’s written extensively on diversity, I appreciated Cass’s emphasis on creating an inclusive society. In higher education, diversity and inclusion understandably focuses race, gender, sexual identity, etc. It’s common to hear phrases like: “An institution built by and for white, cis-gendered males has to be redesigned if we want it to be welcoming to women, minorities, etc.”

Although I may quibble with how diversity programs are implemented (intellectual and viewpoint diversity should be a bigger component), in general I sympathize with those arguments. I’m married to a professor of engineering, and so I appreciate how corrosive sexism in academia can be.

But at least with gender, many administrators recognize changes need to be made and have some ideas on what to do. Not so with Cass’s “other half”–students who might be better off on a vocational track. They are rarely, if ever, included in diversity dialogues. Unfortunately, they probably won’t be anytime soon.

For all the changes in higher education over the centuries, it still has the same basic premise: the only way students learn is for a professor to teach them in a classroom setting. I call it the academic approach to learning.

Sure now we may have the occasional project-based course, and many professors also do try to use practical case studies in their teaching. And some majors may require a few semesters of work to graduate.  But all of these examples still exist within the paradigm that learning primarily occurs in a classroom. Even majors like engineering or accounting rely on the academic approach.

This approach, needless to say, works best for those who enjoy formal learning, advancing scholarship, and pursuing knowledge for its own sake: i.e., academics. More than a particular race, gender, or sexual orientation, academia was designed by and for academics. The neglect of non-academics is the major inclusivity deficit colleges face.

But meaningfully changing this property would negate colleges’ reason for being and undermine their own existence. If learning can meaningfully happen outside a classroom, then why would we need colleges?

Which is why higher higher education will never be able to truly welcome vocational students. They almost always need the opposite of what colleges offer–practical experience rather than academic knowledge. Institutions that catered to them wouldn’t look anything like higher education as we know it.

Even stronger, institutions that catered to students who just want a job wouldn’t look anything like higher education as we know it. It’s not just vocational students who are short-changed by the academic approach to learning. The overwhelming majority of students who simply want to work are poorly served by this model.

Which brings me back to the idea of a working, rather than just vocational, track. Almost all students need practical experience over academic knowledge. That’s what we need to be pushing. My next few blog posts will go into a few concrete examples of how the academic approach fails students even in engineering and business.

Before I get into more details of the working track, I want to lay out my core premise when it comes to college education: college is for jobs, and work is better than education. Put another way: mass college education exists to help people become more productive workers, and thus working is better than education for its own sake.

I know us over-educated types often believe that education is intrinsically worthwhile, and that college is to learn how to think or expand your horizons or study the classics or whatever. Maybe you could have made that argument a hundred years ago when higher education was a bunch of white dudes studying gibberish like plasma physics. But not anymore.

Don’t believe me? Just ask the people in college why they are there. Better yet, ask their parents why they want their children in college. Also ask college administrators and department heads why more people should attend college.

I’d bet that almost everyone’s response ultimately connects to getting a job. Whatever the historical reasons higher education exists, in the 21st century it’s about work. I might actually even go a little further and say that after ~10th grade, formal education of any kind is to help people become more productive workers.

I get that there are philosophical, non-job related reasons for primary school education (basic literacy and numeracy, social cohesion, a civic culture, patriotism, etc.). But beyond a certain point, formal education should primarily be about helping people succeed in the labor force.

I suspect many academics, and especially those in the pure sciences and humanities, will protest because many of them believe their jobs are special. They’re not. Knowledge-producing jobs are just another type of job, albeit ones that happen to require lots of formal education.

So whether we’re talking about construction, mechanical engineering, cancer research, or philosophy, we should judge college education on how it prepares people for work. It is a brutally utilitarian calculation.

Again: we would be having a different discussion if higher education were still reserved for the elite. But if we’re asking basically everyone to go to college, and allocating hundreds of billions of public dollars to it, then we have no choice but to be utilitarian. And we should do so for everyone, working in all types of jobs.

Oren Cass’s How the Other Half Learns almost gets there. But since his analysis sets up a dichotomy between vocational and college tracks, it doesn’t rethink education as much as it could have. If the central focus of public policy is a strong labor market for all Americans, and if education is to play a role in that goal, then we need to focus on how both halves learn.

Oren Cass has a couple essays out that previews arguments he’ll expand on in an upcoming book. He discusses many of the themes I’ve touched on: the purposes of education, skills and job success, education standards, and the fact public education should cater to different types of students.

I sum up Cass’s core argument as: the US education system’s strong bias to college attendance does a profound disservice to the overwhelming majority of students. We thus need a vocational / Career and Technical Education (CTE) track: “We need a pathway to prepare young men and women for productive participation in the labor force that relies less on academics and more on concrete skills and real-world experience.”

As much as I love this point, my main criticism is that Cass doesn’t take this line of reasoning far enough. It’s not just CTE students that need less academics and more concrete skills. All students do. I’ve made similar points myself in the context of science literacy and education (see here, here, here, here, here and here).

A separate CTE track doesn’t address the disturbing trend where many white-collar jobs–from receptionists to doctors–require more education than needed. That’s the problem we need to fix–we’re overeducating across the board, not just the kids who ultimately end up as welders or hospital techs.

A vocational-only track will also probably worsen the “cultural imperative to push more people into the college pipeline” because–regardless of pay–the CTE track is where blue-collar workers would end up. And however much my peers praised Geoffrey “All Jobs are Worthwhile” Owens for working at Trader Joe’s, the ugly fact is that we in the educated elite don’t want our children to end up in blue-collar jobs. We’d ensure that CTE never becomes “co-equal.”

So instead of a CTE/college binary, I propose the working track and the college track. The working track would still offer “close partnerships between school systems and employers that get students in the workplace, earning money and industry credentials, while they are still completing their formal education.” But these partnerships and credentials would be for jobs in sales, marketing, engineering, as well as CTE.

Any non-college educational track would enjoy much more support if it included white-collar work and appealed to children with a range of career interests and cognitive abilities. Since people like Cass (and me!) are ultimately talking about overhauling how we think about education, we should aim for as broad a coalition as possible.

I’m going to flesh out this argument over several blog posts. But I’ll basically be expanding on what Cass wrote, and trying to explain why I think almost all jobs ultimately come down to “concrete skills and real-world experience.” That is: most jobs have a very strong vocational component, a fact our educational system should acknowledge.