back
93 comments
This kind of BS analyses are exactly the root of the current problems with the higher education system (insane tuition, CS majors who can't do fizbuzz etc.) - because they are self-fulfilling prophecies.

For comparison, let's imagine that some evil marketing genius decides to examine the corresponding statistical impacts of wearing a business suit. "Our data indicate that people who wear a business suit to work earn a career average of $1M extra over people who don't. Considering many quality suits cost under $2000, this results into a higher than 100:1 return on investment." Before long, tailor shops are doing a booming business, wearing a $400 suit becomes a faux pas, and candidates get dismissed early from interviews because they showed up with an Armani at a Dolce&Gabbana shop. People no longer care whether a suit is useful or even comfortable in their line of work, and prices shoot through the roof for apparel that offers little beyond a famous label.

I don't know who started the trend of promoting flawed statistical studies [1] in this regard, but if they worked for higher education they must have been a very evil marketing genius indeed.

[1] Why are people who couldn't go to college lumped to bring down those who wouldn't? Is "college" a monolithical category where e.g. differences between majors matter much less than four years's experience in specific industries?

You did not read the article all the way through, nor did the people who voted you up.

If you had, you would have encountered references to a variety of studies that attempt to control for going to college as opposed to other predictive factors using things like twin studies. They have found that the impact of college is less than simple analyses show, but the largest contribution to the difference between college graduates and non-graduates is college itself.

There are many students who have a choice between a state school and a private school, or an Ivy League school and a top school near their home, or an engineering degree and an arts degree.

I think a detailed study of career outcomes and cost of schooling is incredibly valuable. Your suit analogy is flawed. Students invest in a program and later earn money in their chosen career. In your suit example, the ignoramus already has a job and then chooses to buy an unnecessary suit.

There are many subtleties to an analysis like this. You mention choice of work experience. I would add that the calculation of lifetime earnings is also suspect since far fewer people went to college 30 years ago than today. But please don't throw the baby out with the bathwater. We should look at studies like this.

Somehow I doubt Priceonomics is going to move the needle on anything one way or the other. I really doubt many people would forgo getting a degree based on any kind of statistical evidence. Going to college is often an emotional choice and not a logical choice.

Regardless, their conclusion seemed to me fairly tempered. They acknowledged learning through MOOCs, but talked about the positive externalities of physically being at college. They also said the price of college was far too high, so I'm not sure how your comment relates to the article.

I'm interested in your critique, but I don't see any research-based evidence in your comment that disproves the claims made in the article.

Will you elaborate on your comment and provide links to show why the author's statistical analysis is flawed?

anyone can buy a suit.

not anyone can complete a 4 year degree from MIT; employers know this, and use it as a signal to show intelligence.

I have a problem with this method:

"Payscale draws on its information about alumni salaries. It calculates the expected earnings of a newly minted graduate over a 30 year career by summing the median incomes of alumni from his or her school who graduated between one and 30 years ago."

Yes, I know that college was worth it 30 years ago, and even 15 years ago, and even 10 years ago. That is not the real question. The question is will this continue to be true in the future? Those who are challenging the value of a college degree are in essence arguing "We may be at an inflection point where the value of a college degree will, from this point forward, decrease."

You can not prove or disprove the point by pointing out that trend lines from the past, if projected into the future, show that college remains a great investment.

That line of reasoning is irrelevant if we are in fact at an inflection point. What is relevant is any argument for or against the idea that we are at an inflection point. Arguments based on trend lines from 30 years ago miss the heart of this subject.

The strongest argument to be drawn from this chart is one based on the 5 year trend line: if a college degree was worth a great deal during the Great Recession, then surely that proves the enduring value of a college degree? However, there is a counter argument: what is being measured here is the income premium beyond what someone with a high school degree makes, and the Great Recession wiped out millions of jobs in industries that hired large numbers of people who only had high school degrees (for instance, the construction industry).

Again, the question to be answered is are we at an inflection point? When the economy recovers, and jobs in the construction industry have returned, will a college degree still offer the large income premium that it offered in the past?

Another problem with a pure financial projection is that it does not fully consider opportunity cost. The cost/benefit is not simply the difference in earnings, minus tuition, interest, missed earnings and investment on earnings.

The real cost is that being in debt limits your options to pursue possibly more lucrative opportunities, or to enjoy your life outside the system. The better term would be liquidity cost. Every one of us will at some point come across a great investment opportunity. Every one of us can have an idea we would like to pursue. But an employee fully loaded with college debt, added with mortgage and kids, can't afford to do anything else but add to a 401k each paycheck.

Furthermore, every financial investment should have a healthy margin of error to compensate you for risk. This should be at even higher margins when the more leveraged you are. Things like college and a mortgage are a total leverage of your earnings potential as an employee. Thus the question isn't whether college is merely worth it. The question should be if college is totally a no-brainer as an investment.

And it looks like it's not anymore.

Yes, I know that college was worth it 30 years ago, and even 15 years ago. That is not the real question. The question is will this continue to be true in the future?

I know this gets brought up all the time on here but HN is not a good indicator of college worth. One of the few (if not the only) disciplines where college might not be necessary is software and this board is primarily software. Even then looking at the difference between the definitions of software developer and a software engineer may make this moot.

That line of reasoning is irrelevant if we are in fact at an inflection point. ... Arguments based on trend lines from 30 years ago miss the heart of this subject.

Not really. Until there is a real, tangible way for the average individual to get an accepted post-secondary education, we aren't at an inflection point. And other than software development, I personally can't think of another engineering or science profession that could be replaced by self-study and/or experience.

"Yes, I know that college was worth it 30 years ago, and even 15 years ago, and even 10 years ago. That is not the real question. The question is will this continue to be true in the future?"

I think that's a very good point, considering that (1) the cost of college education has been rising very steeply in recent years and (2) unemployment and underemployment among college grads is currently very high (there are tons of college grads who would be overjoyed if they could only get an unpaid internship).

If you really want to treat a college education as an investment, you have to keep in mind the disclaimer that anyone selling an investment has to make:

"Past performance is not a guarantee of future results."

A somewhat more formal argument is the first and/or second derivatives are probably significantly large negative numbers so numerically integrating the past implies very little about the future.

Or if you like an analogy, gravity makes a negative 2nd deriv of height. So if you toss an apple into the air, you can numerically integrate the "foot seconds" underneath the graph of height for the first two seconds. That doesn't mean the number of "foot seconds" area of the graph for time interval 20 seconds to 22 seconds will be identical or even have a linear relationship with "foot seconds" time interval 0 to 2 seconds.

(edited to remove the asterisks in foot seconds which makes the HN software somewhat unhappy)

Of course it isn't perfect information, and of course things are changing. But they can't give you the numbers from the future because nobody has them! So projecting forwards based on past trends is just one of the ways we can speculate.

It can still be useful data, too. Assume we are at an inflection point. Concoct numbers, and tack speculative movement on to the end of the graph. See where that puts you.

Yes, it's speculative, but that's all we've got right now.

Here's the other problem with this method of asking, "Is college worth it?": as far as I know, Alex Mayyasi doesn't collect data on dropouts.

Starting and finishing at a public school in four years is almost certainly worth it for most people. Starting and dropping out, especially when you're more than one semester in, is a disaster, and the dropout numbers for many schools are staggering. I was a grad student at the University of Arizona in English Lit and routinely saw that something like half of those who enrolled as 18-year-old freshmen matriculated within six years.

The rest? Some presumably had parents to pay, but many were stuck with loans that couldn't be effectively discharged through bankruptcy and other normal means. The U of A, and other universities, pay no price for this. They have no skin in the game.

I've written more about this here: http://jseliger.wordpress.com/2012/10/22/the-problem-with-ju... and elsewhere.

The article does discuss dropouts (although the word "dropout" is not used). Emphasis mine.

"A recent Washington Post article covers some of the studies - investigations of the increased earnings of those who attend college without graduating - that suggest a college education itself leads to increased earnings."

"To accurately measure the earnings bump graduates receive thanks to their bachelor’s degree, Payscale takes this expected income of a college graduate over his or her career and subtracts the total income they would have made during that time if they did not attend college. To do so, Payscale subtracts the median wage of a high school graduate over those 30 years as well as over the four years a college graduate spent attending college instead of working."

"Accurately"? Rubbish. I don't see them controlling for a single confounding factor here.

I'd love to see a study comparing college attendees with those who were accepted into college and did not attend (e.g. for financial or personal reasons). At least then you are controlling somewhat for ability, ambition and circumstance. My suspicion is that being "good enough" to get into a good college is sufficient to wipe out the wage differential for many fields of study, independent of the fact of attendance.

Or, as Groucho might say, "I wouldn't go to a college that would have me as a student."

Payscale subtracts the median wage of a high school graduate...

This is a common mistake, and one which completely invalidates the process. The median high school graduate is someone who wasn't able to get into college, and many of the factors associated with that -- starting with low high school grades -- are also associated with lower earnings. You might as well say that purring results in pets gaining legs, because pets which purr have on average more legs than pets which don't purr.

To get any accurate measure of the value of going to college, you need to compare students who are similar in all ways except college attendance. Of course, this is a difficult thing to do, since it involves looking at high school grades, SAT scores, parental income, parental education, and gender at a minimum -- all of these factors are known to correlate with both individual educational achievement and individual earnings.

The article actually deals with this. "One recent study compared the earnings of students who just made the academic cutoff to attend the Florida State University System with those of students who fell just below the cutoff (and mostly did not attend college as a result). We might expect college not to be worth it for these students on the margins of qualifying, yet they reaped returns of 11%."

While this does not look at high school grades, SAT scores, etc., I think comparing just above the cutoff and just below the cutoff is a valid methodology to counter the effect. What do you think?

The graph titled "Best and Worst Median Starting Salaries Ranked by College Major" is mislabelled.

Looking at the source material[1], the table it's based on is titled, "[TABLE] 12: TOP 10 MAJORS WITH THE HIGHEST MEDIAN EARNINGS" and is data on "Full-time, full-year workers with a terminal Bachelor’s."

So it's not about starting salaries at all.

[1]http://www9.georgetown.edu/grad/gppi/hpi/cew/pdfs/whatsitwor...

Isn't a $98K median starting salary for math and CS grads a bit high? Yes, the few lucky ones who get a job at Google or Facebook or an investment bank might make six figures right out of school, but it doesn't seem like there would be enough of those jobs to pull the median up that high. The average CS grad probably won't end up working for an internet company; there are probably many more programming jobs in corporate IT departments.
Their data is incorrect. If you pull up the linked article, the table they sourced is #12 on page 15, titled "Top 10 Majors with the highest median earnings".

That's the median of all salaries for all people in that job description where the highest level of education is a Bachelor's degree, NOT the starting salary as indicated.

Investment banks pay alot for both CS and Math grads... A programmer will make a lot more at a Wall Street firm than at Google or Facebook...
This data rewards schools that reject more applicants and/or fail more students (prevent them from graduating). Because that increases the average & median of their graduates. Charter schools and private schools do the same thing to increase test scores, but when controlling for other factors they are no better than public schools.

In order words, the data shows which schools do the best job of attracting and filtering out the top students. Many folks including some faculty think that IS their job - to set a high bar and see which few students can make it over, but from my point of view that is only half of the job - the other part, the hard part, is to actually educate students - to help more students get over that high bar.

Wait, 98k starting salary for Math/Computer Science is median? In the US?

What?!?! This does not mesh with my experiences or the experiences of my peers at all. If you want to include the financial industry, you're talking about the top 1% of salaries, if that. Are they so disparate that they drag the median up that much?

Edit: Looking at the source data[0], they've separated "Computer Science" and "Mathematics" from "Math and Computer Science". What is the purpose of this? Is "Math and Computer Science" a specific major?

[0] http://www9.georgetown.edu/grad/gppi/hpi/cew/pdfs/whatsitwor...

That's incorrect data. That's the median for all salaries in the field, not starting salaries.
This article, like nearly all on the topic, ignores the distinction between correlation and causality. Of course those who attend college, particularly a prestigious one, earn more. How much of this is due to college as a symptom of their competence and/or privileged lot in life, and how much of this is due to college as a cause, is rarely discussed.
Looks to me like the distribution for the more expensive schools on the Y axis is nearly identical to that of the inexpensive state schools. So no, paying a lot for college is not worth it, when you could get the same result for $50k at a state school.

Unless, of course, you go to one of the top four or five private engineering schools in the country.

The question that really jumps out at me is "What the hell are you paying $150k extra for at a private school?"

Edit to add: If I wrote an article that said, "Buy US Savings Bonds - it's a great investment! You are guaranteed at least to get more money back than you put in!" I'd probably get criticized, because the interest rate of a savings bond is fairly low compared to most other investments. But apparently by the same logic I can feel great about saddling myself with $200k worth of student debt, because hey, eventually I'll make it back and then some.

Given the costliness of most colleges in the US - why not consider getting your degree outside of the country? There are many great universities in Asia, for example, that have a much cheaper price tag. I have not seen very many (if any) US students consider this option. I guess prestige and name-brand recognition is still very important.

But really, I am on a semester abroad in Taiwan at the moment, and I have calculated tuition and airfare combined at the school here is cheaper than just tuition at my school in the US (for 1 semester). Aggregate tuition and costs of living over 4 years and I am sure you could save a ton of money compared to staying in the US. You could then go on to do a master's degree in the States. Indeed, many Taiwanese here do just that - at places like Stanford, Caltech, etc...

My student loan debt is completely oppressive. Educating myself was the stupidest thing I've ever done. Oh the horrible irony.
My wife is an economist and one of her biggest pet peeves is that the idea that a child "must" get a college education has had a dramatic impact on how we perceive trade school or apprenticeship educations.

Not everybody needs college.

College is a pretty damn good resource if you choose to utilize it. Personally, I haven't felt that higher education has actively blocked me from pursuing the knowledge I needed or making resources available to me. The problem is with the mindset of kids going into college thinking it's a mandatory hoop they have to jump through instead of a resource they have to actively utilize.

Ultimately, what you get out of higher education is what you choose put into it.

From the article kindly submitted here: "The data comes from an analysis by Payscale, a salary data and software company for individuals and businesses."

In other words, the data is correlational, and not based on random assignment to an experimental condition and control condition. So we have no idea what the causation is here. By my observation, the kind of young people who get into private engineering colleges are different already, at high school age, from the kind of young people who don't go to college at all, so it's by no means clear that attending college causes any part of their difference in earnings in adult life.

For someone going to college with someone else's money (whether the someone else is a parent, a government agency, or a fee waiver by the college), sure, go to as much college as you can eat. Being paid to be a student is a very happy condition of life. But for someone who is paying for college out of pocket, the economic calculations become different, because college has to be paid for with present dollars, but its claimed economic return comes in future dollars, if at all. Economists, of course, have finance calculations for figuring out the approximate present value of a future income stream, and quite a few economists think (if only we could resolve that issue of causation) that what colleges charge for tuition is a bargain compared to the economic value of gaining a college degree.

A young person of college-applying age will, according to research,[1] be very strongly dissuaded from applying to college if poor, even if also smart. But dumb high school students from rich families go to college in large numbers. Only some of the Ivy Plus research universities and liberal arts colleges make much of an effort at all to identify high-ability students from low-income families and recruit those. Right now college mostly advantages the economically advantaged.

[1] Partial list of links to commentary and research on low-income, high-ability students:

http://www.nytimes.com/2011/05/25/business/economy/25leonhar...

http://ideas.repec.org/p/nbr/nberwo/18586.html

http://siepr.stanford.edu/publicationsprofile/2555

https://www.amherst.edu/aboutamherst/magazine/issues/2011sum...

http://www.jkcf.org/assets/files/0000/0084/Achievement_Trap....

As I always used to tell my students, your degree is just a piece of paper that helps you get your foot in the door. Its worth is equivalent to the effort and energy you put into it. Those that work hard and are passionate about their major, whether it be computer science or whether it be philosophy -- they'll end up doing well in their chosen fields -- or at the very least grow as people. If you're really into what you study, spend a lot of your extracurricular activities in related areas (e.g. being in debate club if you're a political science major), and are excited about working in a field pertaining to your major, you will do well. If you came just to party, socialize, and not major in something that you didn't care about or find challenging...you'll likely not get your return on the investment. I majored with plenty of people in computer science, some were even able to cheat their way to decent grades. However, we all know how well that works when you have to sit down for an interview at Google or wherever and you have a blank stare when you can't tell the difference between a binary tree and an apple tree. Your computer science degree in that case is almost worthless.

So it's worth it if you know what you're getting yourself into. Some people think it's just a rubber stamp to a job that pays a minimum of $50K a year somewhere in the city where they can continue partying. Not true unfortunately. I knew since I was a child what I wanted to study...for most the answer is not that easy. I always suggest that if you don't know what you want to do, don't go into debt by going to college without a plan. Do something else -- join peace corps, volunteer, get an apprenticeship...maybe join a community college first. Getting a college education is one of the most fantastic and redefining experiences a person can have -- you may only have one shot and should feel blessed you have that opportunity. Don't squander it.

It's great to see some research that shows a college education is still worth it. And I certainly believe that, although I'm admittedly part of the academically trained elite.

That said, the work to "democratize post-secondary education" is still an exciting area that can co-exist with traditional higher ed. There's a whole segment of the population for whom higher ed is not appropriate (for a variety of reasons). See what the Gates foundation is doing: http://www.gatesfoundation.org/What-We-Do/US-Program/Postsec...

The schools doing well didn't surprise me. $120K for a Petroleum Engineer - wow!
I make six figures with just an AA degree that has nothing to do with technology.

When I'm interviewed it is clear I know what I'm doing and have the experience and passion to back it up. Maybe I'm lucky, or maybe I don't think or act like the rest of the "real" college grads.

At some point you have to really be honest with yourself and with others about what it is you're looking for and are willing to work for in life. The rest will fall into place for you.

College is probably worth it but that's not what this graph is really showing. It really should be entitled 'Is Private College Worth It?' because it appears that the only significant cost:wage value is to 'Private Engineering' and even then arguments could be made that state engineering is good enough.
It's not about whether or not college is worth it, but rather what you want to get out of it and your expectations... You can't expect every degree to optain you a good paying job, certainly not without some creativity and a little extra curricular effort anyway...
Off topic: How is it that such a rad concept never made it to P/M fit and is now a, albeit a highly analytical and well written, blog?
Misses the part where University of Pennsylvania, Princeton, and others have no-loan financial policies and need-blind admissions.
I really like the visualization component of this analysis. It's really neat, but I wish I could search by school.
if you are talking about the knowledge and degree you got, the answer is no. but college also provie you human relationships, this seems very important for most common people, so the answer is yes
Why aren't the UC schools mapped?
If you go to an ivy league, the answer is yes

I went to Princeton, and the degree paid for itself in 2 years

YES
No.
NO