Measuring 'Return on Investment' of Various College Degrees | Hacker News Reader

Measuring 'Return on Investment' of Various College Degrees

nytimes.com

34 pointslxm54 comments

This whole discussion is based on wrong assumptions.

You don't go to uni, learn about thermodynamics, and then go and build turbines. It's certainly possible to have such a career path, but you won't have actually learned the thing you need to know in uni and then applied it at work for the next 40 years.

What you do in uni is demonstrate that you're interested in certain things and you have enough work ethic to learn the exam answers. That's all.

I had a guy with a biology degree at an HFT debug a threading issue. Really smart guy. He didn't learn how CPP works in his PhD, he learned it in his own time because he had an interest.

I actually did study thermodynamics as part of my engineering course, and guess what, it's not in itself useful in finance. It's just a demonstration that I might be able to learn certain mathy things, not even the same things.

So coming back to this idea of ROI on a degree, there's only broadly two things you can demonstrate. I'm more of a lumper than a splitter, so you might have more. But the two things for me are whether you can do technical cave diving like figuring out what happened to that threading issue, and whether you have the capacity to read and understand a large amount of text, like being a lawyer or political science student.

The thing to remember is you don't just gain some skill in uni, chances are you mostly have personal preferences that are revealed, and that that is far more important than what you learned.

ROI in this case would seem to just mean "the kids who studied business did so because the are interested in business and so went and got high paying MBA jobs". Similarly for kids who liked building robots or kids who read a lot about history.

Finally I'd like to say we all need to learn broadly about the world, I don't condemn people who happened to not be technical. But education itself is worth pursuing just for it's own merits.

So college degree is like POW (proof-of-work) in blockchains.

And, estimating the cost... I see some site say $27k for tuition per year. Assume four years per person; assume people live for 80 years; assume half the population goes to college; then 1/40 of the population is attending college in any given year. If we say the U.S. has 300M people, then that's about $200 billion per year, for this proof of work.

Meanwhile, this site estimates Bitcoin's annual electricity consumption at $7 billion worth: https://digiconomist.net/bitcoin-energy-consumption/

Bitcoin manages to perform ~7 financial transactions a second, whereas this 'proof of work' manages to assign jobs for ~300 million people. Some of those people even manage to learn something useful, and go on on to design things like Bitcoin and the power plants that produce electricity that is wasted by it.

> Bitcoin manages to perform ~7 financial transactions a second, whereas this 'proof of work' manages to assign jobs for ~300 million people.

First, obviously, no it doesn't.

But second, even if it did, would that be more or less than 7 transactions / second?

> the two things for me are whether you can do technical cave diving like figuring out what happened to that threading issue, and whether you have the capacity to read and understand a large amount of text, like being a lawyer or political science student.

Aaaand this is the moment where I worry that I'm a decade or two into the wrong track, because while my math degree says I'm not chopped liver at the former, I'm probably better at the latter.

But actual work experience is an even better hint that someone can do a job...

Is it really? I doubt I am the only one who has worked with a highly regarded colleague, who when you actually do the work on the ground with, you discover they are incredibly unimpressive.

Promotions for incompetence, serial job hoppers who move on before being discovered as being shit ... etc

I've been in that situation as well. It's not like signals can't be wrong. Question is who do you want to hire, someone with no experience or someone with some years doing the job?

Naturally experience, ultimately hiring managers have never seen a cv without experience, the choice is how you differentiate different experiences. Ultimately it all comes down to 'which candidate do I want to (or can tolerate) spending more time with than my own family'.

Yes but in this given day and age, maybe a high school graduate can take a year or two off just to focus on creating open source projects or contributing to open source projects. Heck they can even build some hacky basic hardware + software systems using Pi or Arduino. That will include learning across disciplines beyond computer science like thermodynamics if they work on measuring and controlling thermal outputs.

A candidate whose github project is highly starred and forked is a lot more valuable to employer than someone with a paper showing the degree. It will cost less than one fourth of what it will cost to a semester in top US school.

Those people are also the ones who would benefit most from higher education. They are learning and doing new things because they are genuinely interested in it. A good university would give them a chance to spend a few years exploring the possibilities without the pressure to do something immediately productive. They would learn from experts, find new interests, and work with peers with similar interests.

Any job a bright self-taught recent high school graduate can get is likely a job they should not take. Many bright young people turn bitter and disillusioned, because they made the easy choice when they got an offer for a real job. The job may very well be a local optimum with no good way forward. Because they are too focused on short-term goals, they may never get a chance to stop and think what to do with their lives.

The academia has a similar trap further on the road: grad school. If you are a good student who is genuinely interested in learning for the sake of learning, grad school looks like the obvious step forward. For many, it will be the wrong choice.

> Any job a bright self-taught recent high school graduate can get is likely a job they should not take. Many bright young people turn bitter and disillusioned, because they made the easy choice when they got an offer for a real job.

Any examples of this happening? I haven’t seen this to be the case.

> What you do in uni is demonstrate that you're interested in certain things and you have enough work ethic to learn the exam answers. That's all.

I think you’re right, but the next question would be: how is the hefty price tag for American universities justified, if your statement is true?

It's a guild fee is what I'm saying. If the guild is making a lot of money it's worth it to pay to join. See MBAs. I can attest as a graduate of a world famous business school (ok it's the the uni and not the biz school) you won't learn how to run a business. But you do get that "ambitious hard worker who is somewhat social" label.

Satement of intent is how I think of a MBA - there is nothing you couldnt learn from a book and $1.50 in late fees from a public library.

At the end of the day it is all theatre to satisfy the hiring managers. Who do you select from; the candidate who says they read the book or the one that ivy league school says read the book?

Take the same person with the inartistic desire to learn. Offer them an excellent library with vetted books. Give them 2 years to focus on this learning full-time. Surround them with like-minded peers so they can share ideas and grow together. Add world-level experts to stimulate and group discussions. Guarantee them a job at a company they respect, starting at $150k, if they do well at the end.

Will these things make a difference to their self-study course? Is this difference in trajectory worth $100k? Enough people say yes, and very few of those who actually did it regretted their choice.

I see it like you. So I guess it makes sense to evaluate the ROI of American universities based on the guild you’re joining. Kinda sad.

What you do in uni is demonstrate that you're interested in certain things and you have enough work ethic to learn the exam answers. That's all.

That might be true for you—and others—but it’s not always that way. I use the knowledge I learned in my degree all the time in my work - sometimes directly, sometimes as the foundation on which I can build my understanding of a more domain-specific problem. That knowledge is augmented by the extracurricular I did and my experience working in various roles over the years. I’m pretty sure this isn’t uncommon.

I agree that the softer skills of self-directed education, research, writing etc. are important and widely transferable. Plenty of people I know have been successful in careers distinct from their higher education. But I do think it’s a bit reductive to dismiss the value of a deep technical education so easily.

IMO: economics is just heat transfer with extra steps.

> What you do in uni is demonstrate that you're interested in certain things and you have enough work ethic to learn the exam answers. That's all.

Wrong. Why? There are many counter examples. But perhaps you meant that it tends to be true, if so then I read your comment a bit too black and white. That would be a lot tougher to see whether that's true or not. I'm purely going to keep this comment on a content level (e.g. I'm not concerned with the fact that you also make lifelong friends at uni, etc.).

I've done quite different study programs, so let's see what happened there (I'm only talking about the ones I've finished since I know those well).

I studied:

- Psychology (bachelor)

- Information Science (20% business, 80% computer science, bachelor)

- Game-design (master)

- Computer science with a specialty in multimedia and security (master)

For my psychology degree, I really agree with your general message. When I look at where my classmates are today, I feel that content-wise university was a bit worthless. However, of those psychology students there are about 5 to 20 percent of people that are using their knowledge. One does a lot of statistics at work and did a lot of statistics in a psychology minor (structurual equational modelling). Another one did research for the government and wrote fairly similar-ish research reports compared to psychology papers. Finally, a large part of this group becomes a clinical psychologist and then a large part of your study is relevant because most of the program is about interpreting research.

So... 80% agree, 20% disagree.

Information Science: 20% of the people were already working in IT. For us other eighty-percenters a lot of us went into consulting and a few of us into hardcore computer science. Information science definitely teaches things about requirements gathering that I've used as a freelancer later on. I wouldn't say that the whole program is useful, mostly because it's split between two worlds and you end up picking one of them. But for the world that you pick, it had a lot of useful courses. For me, I gained the ability to learn anything in the IT world. Before I went to uni I've tried it but I failed. I tried learning PHP, I failed. I tried learning Objective-C/iPhone apps but I failed. But at uni, I learned programming, networking and computer systems (among other things) and suddenly revisiting those topics 3 years laters I felt learning Objective-C was easy!

So... I tend to disagree completely here.

Game-design: people went to a lot of different fields after this, I think most of us aren't using any of the theory/practice that we've learned in this degree. However, some of us are! I'd estimate about 10% is still heavily using their game-design skills (e.g. one is a online curriculum-designer and the whole thing about user feedback shares a lot with game-design). In my case, it's mixed. I made one game Unity3D and it was the first time I was introduced to C#. I did a few things in JavaScript. But mostly I learned about player behavior and I've noticed a lot of game-design knowledge is a specialized version of UX. So when I tend to develop an app, I tend to see "users" as "gamers" (if applicable) and utilize my intuition that I've built up with this program.

So... I tend to agree, but there are exceptions.

Computer Science: in one course I made a computer graphics engine that had 10K lines of code. That was very educational. In another I made and sold an iPhone app (rewarding and educational). And I've done a lot of hacking courses that now allows me to pentest production applications while I'm a full-stack engineer.

So... I tend to disagree. Yes, the skills I've learned here are foundational and not fully on point to what I do now (full-stack dev, currently interviewing with a FAANG company, searching for referrals to Google still :P) but it definitely helps with what I do today. At the very least, I tend to catch critical security issues in code review (and hack into them as a PoC when no one believes me) and code (hopefully) secure code myself.

Undergraduate degree is ~90% generic learning ~10% specific learning. These are lego bricks in ones personality and thinking process.

> But at uni, I learned programming, networking, and computer systems

You also learned logic, planning, typing, troubleshooting, project management, etc. You could be thrown into many situations and would have to learn a few extra specifics associated with the task and attach those to your existing knowledge. You would (should) outperform a history major in many of these tasks, and should be at a disadvantage in many others.

Consciously or not you will be applying your psychology learnings to interviews along with your failed PHP dreams. Naturally having that psychology background you don't need me to tell you that acknowledging and understanding that will be beneficial to you in that process.

> Undergraduate degree is ~90% generic learning ~10% specific learning.

If you mean generic learning about a specific field then I tend to agree. If you mean "general learning" at all then I disagree. It's because one learns generically about a specific field that it allows them to learn quicker later on in life about various topics within that field. I'm proof of this given that after my CS degree, I wanted to do a coding bootcamp and was immediately hired to be an instructor, I self-taught myself that program within 2 weeks and then taught it to others (with a non-CS background) for 3 months. Moreover, the quality of learning is different since I noticed I was way quicker at building stuff out or diagnosing bugs compared to the best students in their final week. The reason why was simple: I've written a lot more code and experienced a lot more bugs than they had.

> You also learned logic, planning, typing, troubleshooting, project management

Partially agree, partially disagree, in many cases I knew those things and merely applied it. I've learned logic, planning, typing and some forms of project management at elementary school (e.g. a touch typing course, basic if this then that logic, etc.) and high school (calculus, fairly advanced statistics etc.). There were more advanced versions in some cases (e.g. logic, certain forms of planning), but in most cases I relied on skills I've learned during elementary/high school. There are a lot of presentations given before someone enters at university.

> You would (should) outperform a history major in many of these tasks, and should be at a disadvantage in many others.

Don't know, I think some might be pretty good at typing, certain forms of logic and perhaps project management as well.

It's a very superficial take of what a university course gives to you; sure, it doesn't give immediate work skill, but the shared "vocabulary" of theorems, formulas, notations and conventions you build if your curricula matches your job goals and your team formation gives you tremendous synergy both in learning on the job and in functioning as a team; the understanding of principles behind the workings enables you to learn faster, to more aptly move your knowledge into similar area interdisciplinarily, and to be more flexible in modifying existing solution to novel issues within your field.

But you'll end up getting a whole new vocabulary when you join a new industry. What you did at uni was practicing learning new vocabulary and models.

I would say as far as something like math appears to be foundational, what it really makes you practice is precision thinking: this thing matters, this other thing doesn't.

One thing that might convince you is that when you look at math the subject itself, how many people who were good at applied math like calculus and PDEs actually know anything about mathematical foundations like ZFC, Gödel, and that type of thing?

no lol job providing theory learning are exceedingly scarce, at most you get some practical level stuff and off you go. all the learning you do on the side do help, but formal learning is a completely different thing than running two courses off udemy

Well said. In Liar's Poker, the author has an Art History degree. That book inspired many young people at the time to work on Wall Street. Several of my friends who went to work at various trading desks did not study business. One of the most successful (highest "ROI") majored in Philosophy.

The idea of an uneducated electorate that increasingly gets its (mis)information from social media and aims to "earn" its money only from venture capital and online advertisers is, IMHO, frightening. However, the bigger picture is not something these "ROI" analysts ever seem to consider.

The NYTimes barely mentions the data. Here's the original article that talks about the ROI:

https://www.thirdway.org/report/which-college-programs-give-...

I think the most alarming thing is the "For Profit" section of the charts. 40% (of students) don't recoup their cost ever vs 13% in Private and 8% in public.

"For Profit" schools prey on the venerable to take out huge loans for a degree from "City University" or ITT Tech (defunct) with the promise that you could make a lot of money "just like our 1 person we put in all our commercials."

I'm glad a number were shut down, but there are still too many that operate.

For-profit schools also take on more marginal students, who may be less likely to prosper in any case.

My father, admittedly an outlier, went from growing up in poverty, to ITT Tech, to Marquette (a Catholic University), to a Ph.D. from U of I Urbana-Champaign in one long continuous path of transfers.

It’s not clear to me that he would have achieved what he did without the initial venue to acquire and demonstrate skills, but I could see how many might not travel that path.

To what extent can we really weigh the experiences of those who do net benefit (presumably the 60%) against those who do not? Perhaps one man’s preying is the answer to another man’s prayers.

Not to diminish what your father has accomplished because it is rightfully amazing, but a possible question to whether your father needed ITT there? Could there be another avenue through a community college or other non-profit system?

> Perhaps one man’s preying is the answer to another man’s prayers.

I reject this premise. This is a really poor justification for preying. So many lives were destroyed that ITT was forcibly shut down.

I don’t like NYTimes tone about this. If you decide to major in Cinematography and then blame the college for negative ROI, the fault is entirely on you. No amount of fudging will bend the market forces for employment in your favor. Supply far exceeds the demand for such roles (similar video game software devs).

We need more community and polytechnic institutes. Trade craft schools and apprenticeships like they do in Switzerland. I don’t understand why people (mostly progressives but also conservatives) in USA continue to double down on degree programs.

Right, but:

- a child makes the decision

- the main influences on the child probably don't impress on the child the need to analyse the decision rationally

- even if the child tried to analyse it rationally, what % of prospective college students have the financial literacy to do that successfully?

- even if they had the financial literacy, it's not like colleges provide useful data. Even this article talks about earnings uplift without isolating the impact of the degree (vs. existing increased prospects compared with typical HS grad).

Unemplyment rate by qualificaion (ignoring participation rate which would skew these numbers far more to the side of getting qualified) https://www.bls.gov/news.release/empsit.t04.htm:

- HS dropout 9.5%

- HS only 6.3%

- Some college or associate degree 5.0%

- Bachelor's degree and higher 3.1 %

What conclusion do you draw from these data?

Go to university, even better if you can get your degree.

The data are not sufficient to support your conclusion. The data say nothing about:

* The opportunity cost (of the time to get a degree)

* Tuition costs

* The $ returns in higher income

Maybe getting a degree (any degree) reduces the chance of being unemployed, because Starbucks prefers to hire people with degrees than people without. But that doesn't mean getting a degree is worth the cost for everyone in every situation.

> - even if the child tried to analyse it rationally, what % of prospective college students have the financial literacy to do that successfully?

One of the MANY reasons I find that federally backed student loans should be stopped asap.

The dominant culture in the United States is one of "following your passion." This is the advice that college students get from college advisors and the way colleges advertise themselves to potential students. In light of that, I think it's unfair to place the blame entirely on students.

Is there a billionaire yet who hasn't written on this in their 'top 10 pieces of advice to make it big like me';

Get up every morning at 4am.

Choose a job you love, and you will never have to work a day in your life.

How is informing people about report where they can find information about this exactly thing "bad tone"? Kids should decide rationally and in the same sentece you criticize article for telling them some degrees have bad ROI. They should guess and make up this info on their own, I guess.

Also, in fact, if you look at where people apply, they go mostly for business sounding degrees. Cinematography and what not are not where majority of students go.

Students and families in fact are trying to guess which schools or degrees are good for them. Informing them about reports that actually measure them is better then having to guess.

NYT says the authors define the earnings premium by reference to a typical HS grad. But folks who can gain acceptance into a good program at a good school would (after 4 years) earn more than a typical HS grad, even if they skipped college.

Why attribute 100% of the uplift to the degree?

Is there another thing you would attribute it to?

It's not hard to simply ask "of this profession, what is the salary of those w/ a degree, and those w/o" at the national level

Imagine there are two groups of 30-year-olds:

A) Those who were offered university places, but decided not to go.

B) Those who were not offered university places.

I would expect group A to have a significant 'earnings premium' over group B. Even though neither group attended university, group A are more likely to be smart, hard working, lucky, or whatever drives career success.

If you want to estimate the earnings premium of a university degree, you have to compare the folks who went to university with other folks just like them, i.e. with folks in group A. Not with a mixture of groups A and B.

> Is there another thing you would attribute it to?

Yes, some of the earnings premium is due to the students pre-university chances of success.

> It's not hard to simply ask "of this profession, > what is the salary of those w/ a degree, and those w/o" at the national level

This wouldn't measure the thing we're trying to measure, because it suffers from selection bias.

On one hand, with the cost of a university education in the US, doing this analysis is rational. On the other, as a society and an economy, we give up so many indirect benefits by filtering out degrees that don’t have a high ROI or even a first order negative ROI.

> we give up so many indirect benefits ...

So if you make those benefits direct, or have those who obtain value from these benefits pay for them, then you can obtain the true cost of a degree, and the true ROI.

Wow, US degrees aren't as expensive as I thought, it comes out pretty comparable to the UK. And then US tech salaries are 3x higher so it's still a good investment overall.

What a wise man does in the beginning, a fool does in the end.

I graduated in the 80's in UK, and it was the best time of my life. I had no debt. Fantastic! If I was asked to make the same decision again, I'd probably decline. The idea that I'd start my young adult life tens of grand in the hole - a mini-mortage - strikes me as highly unpleasant. As far as I'm concerned, these guys take their "return on investment" calculations and shove it where the sun doesn't shine.

"But blippage," I hear you enquire, "aren't you being irrational?" To which I respond: "I'm not the one who's insane here."

Of course, it's not "wrong" to want to get an education especially when the job prospects are good at the end of it. Still, if you couch things in terms of "returns on investment", then I'll figure that either you've a bridge that you're trying to sell me, or you're severely lacking in common sense.

Yeah, sure, maybe at the end of the day it's just me that doesn't "get" it, but still, I don't like debt, and I suggest that you don't, either.

University fees escalate ever-higher, and people keep coughing up based on Groupthink. In my not-at-all-humble opinion, of course.

The cited report includes an extract of all programmes in .xlsx format [1]. There are some surprising majors towards the bottom of the <= 10 year ROI:

* Neuroscience - 21% proportion w/ ROI

* Biology - 27% proportion w/ ROI

* Anatomical sciences - 45% proportion w/ ROI

* Biochemistry, Biophysics and Molecular Biology - 46% proportion w/ ROI

Why might this be?

[1] https://thirdway.imgix.net/Program_Level_PEP.xlsx

My initial guess would be because those are degrees where higher earnings and upward mobility tend to require higher level degrees.

https://ghostarchive.org/archive/DeNGv

https://archive.is/PMu54

"If their degree cost them $60,000 to obtain, it would take them four years to recoup their education costs ($60,000 net cost / $15,000 earnings premium)."

They completely omit the time spent getting the degree and associated loss of earning during that period

Can anyone offer a mirror?

The wayback machine is a good resource to avoid paywalls

https://web.archive.org/web/20210813130806/https://www.nytim...

Brilliant, thanks for this.

Reply on news.ycombinator.com