A First Try at ROI: Ranking 4,500 Colleges [pdf] | Hacker News Reader

A First Try at ROI: Ranking 4,500 Colleges [pdf]

1gyhoq479ufd3yna29x7ubjn-wpengine.netdna-ssl.com

53 pointsSQL221949 comments

So apparently this is using data from the Dept of Education, but I wasn't able to find anything about their methodology on the site that the paper referenced[1] (footnote #22 says that's the data they're using).

I am curious to know: how are earnings calculated? Are these numbers coming from voluntary surveys or something more reliable?

In any case, this data is kind of interesting but it's certainly not anything that would lead me to conclude that college is worth it. At the very least, I'd want to see some attempt to control for what I'd assume are the most powerful confounding variables:

* controlling for industry (in which industries to college grads make more than non-attendees?)

* controlling for socioeconomic background (do rich kids who don't go to school make less than those who do?)

* and controlling for opportunity (do students who are accepted into top schools but choose not to go make as much money as those who do?)

[1] https://collegescorecard.ed.gov/data/

While it's not perfect, it's surely far superior to counting the number of citations of research papers published by the academic staff, who might not even be teachers. Isn't that roughly how the popular rankings which prospective students use to compare universities work?

Because number of citations equates to success how again? The university obsession with number of citations is just as dubious as GDP

A few potential issues I see:

1) Adjustments for cost of living geographic distribution of a college's graduates when earnings is measured. This distribution would not be the same for all colleges. E.g. Texas vs California - 100K/year in SF is not the same standard of living as 100K in Dallas.

2) Is there any adjustment for the fact that for some colleges a high percentage of 4 year degree holders go on to earn a graduate level degree. That could certainly be the case for the Pharmacy schools where I'd bet that many BS graduates go on to earn a Doctor of Pharmacy. You would end up comparing the NPV of a pharmacy degree with a BS from say CSU, San Jose.

NPV is more sensitive to major than college. A Berkeley liberal arts degree has lower lifetime earnings than a Cal State nursing degree.

This is an important point. Also this list by it’s very nature favors schools with less overall majors and more majors of a technical nature. Not sure how useful this data really is considering the number of variables.

> Also this list by it’s very nature favors schools with less overall majors and more majors of a technical nature.

That's true, but if your model of using the list is "someone is looking for advice on how to make money by going to college", it is a desirable feature. This person doesn't know what they want to do, and funneling them to a school with only good options will guard them against choosing a bad major later, when they've stopped looking at the list. Adjusting ROI to major availability would make the list less accurate.

True - at the same time its easier to switch majors at a school than to switch schools for a given major. So there’s a slightly different risk profile if one decides by major primarily.

Is it? One of my professors kept a table, for a few decades, of his students’ “intended major”, “actual major”, and “job”, and found essentially no correlation among any of these. I’m certainly not doing anything for work that one might guess from my major or minor/concentration.

Here is a link to a summary, and a tool to search for colleges: https://cew.georgetown.edu/cew-reports/collegeroi/

or just download the data table in csv format since they wanna load all kinds of social media tracking crap along with the tool: https://cewgeorgetown.github.io/collegeROI/ROIforWeb.csv

Okay, what I'm getting from this is that pharmacists make lots of money, and it's a really good bang-for-your-buck proposition.

Other professional schools are not listed or they’d be much higher

It's like 6 years to get a degree that AI should replace though.

Physician assistant might be the next best thing in that space, also 6 years, and a pretty high median salary.

Not to undervalue a pharmacist's other functions, but you don't really need AI... you just need a bot to pull orders off a shelf and file online insurance forms.

I'm thinking more for smart insight into possible drug interactions. Since they may not know all the drugs you are on. My parents had doctors that both prescribed things and neither we're aware of it. So if you just pull orders you are doing the bulk of the busy work but not fully replacing the pharmacist.

With approaches like Amazon's PillPack and CVS Multidose, it's not clear why either AI or a pharmacist is needed.

Pharmacist jobs at a drugstore, grocery store, Walmart, Costco are probably not the road to riches. Managing the pharmacy at a major hospital with lots of specialized meds is probably a good gig.

The major matters a lot.

It used to be. There are now far more pharmacists than jobs so their wages and quality of life at work are plummeting. Read basically any thread at pharmacy subreddit or sdnforum. They barely start out at $80k to $100k, with over $200k of debt, terrible work hours and scheduling, and no future increases in income or job security (actually decreases).

The excess of pharmacy schools looking to make an easy buck on tuition also makes it so they let anyone into pharmacy school. There is literally no entry exam or requirement to get in other than ability to pay, i.e. take on taxpayer funded student loans.

The whole business itself got squeezed by the handful of large health insurers left, who control the money, such that even grocery stores are shutting down their pharmacies.

This is false. The PCAT is required by most schools.

Yes, technically many schools require it, but the requirement is being dropped by more and more schools every year, and employers don’t care which school you come out of, so your competition is greatly increased.

And it’s not like med school or many other grad school options where you have to take MCAT, which is far more rigorous than PCAT anyway, but since the graduate of a pharmacy school that doesn’t require it is the same as one that does in the job market, it effectively nullifies the bar an entry exam might set.

https://www.studentloanplanner.com/getting-into-pharmacy-sch...

Good stuff. It should be noted though that it's not really calculating NPV for the students themselves, as it can't measure what the students' earnings would have been had they not gone to college (or gone to a different college). Much of the difference in measured NPV across colleges will be due to the quality of students, not the quality of education / opportunities provided by the college.

One problem with this analysis is that it mixes the numbers for schools that have fairly siloed online and traditional programs. For instance, one school I am associated with is listed as having an average start age of 27. However, it has completely separate online and traditional degree programs. The average start age is 18 in the traditional program and probably in the 30s in the online program. This averages to 27, but I imagine the ROI is quite different for the traditional program versus the online program. Traditional students take no online classes there.

There are so many potential confounding variables that these sort of comparisons are meaningless. A big one would be your confidence. Being able to walk up to a recruiters table, feel comfortable, and make an impression could get you your first internship and snowball your career from that brief moment. Or making friends with a professor who encourages you to make a collaboration with someone who one day hires you, or at the least inspires you.

The best thing a college applicant can do to better their odds is to make sure they are going to a school where these types of events happen regularly. It doesn't have to be berkeley; most state universities are just as excellent even if us news says otherwise.

Are professors publishing/going to conferences/collaborating with the field/spinning off companies? How many companies are present at career fairs? Would you be able to do internships nearby? Does the department have a seminar series with speakers from around the country? Do you have many opportunities to do independent projects or research? Would you have a chance to do outside research at another local institution? Are upper division classes covering fundamentals as well as new technology, or has the course been the same for 25 years? These are questions potential students should be asking if they are looking to maximize ROI, and that is ROI beyond just getting a healthcare related degree and collecting the same 80k with the same title from when you were hired to when you retire.

That's how it ought to be, but in practice at your average college the best-paying major is the nursing degree. That tells volumes about how far the deindustrialization of the country has gone.

Was expecting to see Georgia Tech. $12k/year tuition for in-state, but typically top 10 rated for their BSCS degree program.

They rank pretty well, #27. 1.7M 40-year NPV.

I wonder if it is possible to run a similar calculation for graduate degrees. I'd like to see it for OMSCS. Probably need a lot more history before that becomes plausible.

I'm surprised "free" colleges don't dominate the short term ROI.

https://www.usnews.com/education/best-colleges/paying-for-co...

Four years from date of enrollment, with negligible earnings, counts for a lot.

You would think the various military academies would be pretty high on the list with free tuition and an officer paycheck immediately following graduation.

This isn't necessarily a problem with the analysis, but one thing to keep in mind that the type of student that a college accepts might have a large influence over their graduates income.

So, it's difficult to say whether or not selective schools (ivy league and whatnot) actually prepare students better, or if they just tend to admit people who have high earnings potential in the first place.

Might sound like a nitpick, but it's pretty relevant if you're comparing a low-cost public school to a selective private one.

I should've gone to The Creative Circus, GA instead of Georgia Tech

What's with the super-high graduation rates for supposedly "hard" colleges like MIT and Yale? In a "hard" college I went to the graduation rate was _at best_ two thirds.

It's possibly a factor of how selective they can be by reputation. If the number of candidates is a few times higher than your college (which to be clear I have no idea about), then it's possible their selection criteria also selects for those likely to graduate, whether on purpose or as a side-effect of other things they attempt to select for.

they're hyper-selective, but there's no particular reason to arbitrarily fail 1/3 of their students. you could just make the classes harder until this happens, but employers/grad schools seem to be pretty happy the way things are, so why do it? Yale students even mostly get As in most classes and it works out fine for everyone.

MIT graduate here. It’s a hard school to get into. It’s not especially hard to graduate from (given that you were admitted), as long as you’re willing to grind out the work.

That's the thing though, few people are willing to "grind out the work" over any extended period of time.

Well, being notoriously hard to get into, to the point that you have had to grind out the work/studying and also be talented to get in, probably means just keep on doing what you were.

Agree. I went somewhere easy to get into and very very hard to get out of with a degree. You had to know your stuff. All of it.

Also, if this is coming from DoE data, medical withdrawals are subtracted from both the numerator and denominator—on the idea they would have happened anywhere. It’s definitely not the stress, the climate, or the concrete doing it.

IHTFP.

As someone who has never been to grad school, I assume this is how most grad programs work too.

Not sure about MIT, but Ivy League schools have always inflated grades.

That’s not necessarily a bad thing. I went to a state school that had bell curve grading. Brutal.

These kids mostly got straight As and perfect SAT scores in high school, so I don't think it should be that surprising that it continues in their college studies also. I think it kind of depends on whether you think students should be evaluated on their own or in comparison to the rest of the class.

Lots of schools that are hard to get into try to make it (relatively!) easy to graduate. The average GPA at Harvard is like 3.7. I do wonder if part of it is to prevent students that were really successful in high school from having imposter syndrome when they don't get straight As in college.

I went to MIT for grad school. At the prospective student orientation, the admissions committee stood up in front of us and said (paraphrasing since this was a decade ago)

>after careful deliberation we have concluded everyone standing in the room has what it takes to graduate, and that we consider it our duty as a department to ensure that everyone who is admitted does graduate. We are committed to provide you whatever resources--be it academic, financial, emotional, etc--you need to graduate, and we want you to focus on doing the best possible possible research.

Needless to say I picked them over Stanford, where it was a battle royale to see what 2/3 of the class would graduate. Got out in 4 years for my masters/PhD and loved every minute of it.

Yep, 70% percent of my freshmen engineering cohorts had flunked out by 4th year.

Is there a way to get the table as csv? This data is worth some charts and analysis.

https://collegescorecard.ed.gov/data/

Reply on news.ycombinator.com