Is Paying $100,000 for College Really Worth It?

Parents Forum
141

College Scorecard database has earnings taken from tax returns at the 4, 6, 8, 10 year marks. There are tons of caveats to this data, starting with the fact it only includes those who took federal financial aid (including only those who took the federal student loan.)

One has to take a view on whether or not this cohort would be similar to the non-federal financial aid group, @data10 may have some thoughts on that.

Here is the dashboard, which only has 4 year median earnings. You can search by school and by major within a school:

For earnings from years 6, 8, 10 you need to download the full data set:

Georgetown’s ROI analysis tool goes a step further and calculates a school’s ROI (using College Scorecard’s data on earnings and a school’s average net price.)

Here’s their methodology: https://cew.georgetown.edu/wp-content/uploads/cew-college_roi_faqs-2025.pdf

2 Likes
142

This information is pulled directly from federal databases. I have concerns about the relevance of the data, but it is “real” … they have access to earnings information for those who received federal loans. I agree that this subset of students may or may not represent the whole, but it’s the best available data.

2 Likes
143

Exactly…They link federal student aid data with Department of Treasury records:

Administrative Earnings Data from Tax Records
To gain insight into the labor market outcomes of individuals attending institutions of higher education, data on federally aided students have been linked to earnings data from administrative tax records maintained by the Department of the Treasury. These linked data are used to produce aggregated and de-identified estimates of institution-level statistics, such as the median of the earnings distribution of
federally-aided students in a year. Earnings are defined as the sum of wages and deferre compensation 39 from all W-2 forms received for each individual, plus self-employment earnings from Schedule SE. Importantly, because W-2 forms are filed by employers, the estimates of labor-market outcomes cover the population of employees in the Social Security system and the self-employed individuals who file tax returns. Moreover, the earnings information offers national coverage, in contrast to some commonly used information from statewide programs. For example, state unemployment insurance administrative data is limited to the subset of students who work in the same state after leaving college.

Full, long, painful document here:

1 Like
144

Seems like the students not using federal financial aid would be on average from higher SES, and therefore more likely to be connected or advantaged when it comes to job seeking.

1 Like
145

Rather than federal aid sample vs non-federal aid sample, I think 2 bigger caveats are:

Caveat 1 – Most of the analyses using this information including Georgetown’s linked above do not control’s for student’s major, which is far more correlated with future earnings than is name of college. For example, the Georgetown link shows the following colleges have the highest 10year ROI for bachelor’s degree. An obvious pattern is all of these colleges have a large % of students choosing majors associated with higher earnings – engineering and pharmacy.

Highest ROI Colleges: Georgetown Analysis

  1. MIT
  2. Caltech
  3. Albany Pharmacy
  4. MA Pharmacy
  5. Olin Engineering
  6. St Louis Pharmacy

While not commonly used, College Scorecard does list earnings by major, and these are all over the map for HYPS… type colleges rather than most majors being near the median. I’ll use Cornell for this example since it is the largest Ivy and may have a greater sample size, but same type of pattern occurs for all Ivy+ colleges. The overall median is going to vary tremendously depending on how large a % of students choose higher earning fields like is common for CS, engineering, and econ majors.

Median Earnings by Major at Cornell (4 years out)
Computer Science – $223k per year
Elec Eng – $147k per year
Economics – $138k per year
> Median = $113k <
Political Science – $95k per year
Psychology – $78k per year
English – $65k per year

Caveat 2 – The College Scorecard sample, Georgetown analysis and similar do not consider individual student differences and instead assume difference in earnings entirely relate to name of college. Kids at highly selective colleges are not just a randomly selected group of HS students. They instead tend to be gifted, high achieving, motivated, well informed, and interested in pursuing higher earning fields. These types of individual differences from the average HS student influences their future outcome. As such it’s important to have some type of control for this type of difference.

For example, how do outcomes compare for high achieving kids at college A vs high achieving kids at college B? Rather than how do outcomes compare for mostly high achieving kids at highly selective college A vs outcomes for mostly not high achieving kids at less selective college B?

Analyses that do have an adequate control for individual students consistently find much smaller differences in median earnings for different colleges, often negligible difference for highly selective vs not highly selective. However, there are often specific subgroups that differ from overall median, such as the Chetty analysis that found a different rate of top 1% earnings at 10+ years out.

3 Likes
146

I strongly suspect this is operating in the background of different perspectives on “worth it” as applied to something like college.

If you are a big experiences over things person, you get pretty used to the idea you will sometimes spend a bunch of money in a relatively short period of time and then afterwards only have things like memories, photos, stories, new perspectives, new or deepened relationships, and maybe a t-shirt.

I think some other people struggle with spending a lot of money and then not having something more tangible at the end.

OK, so they then might focus on the degree you get as the most tangible thing, and further try to in some way evaluate the tangible value of this degree.

But for experiences over things people–eh, a useful degree is nice enough, but it isn’t where all the value needs to be found.

And then college isn’t just like a week or two, it is something like a 120-130 weeks total, even more if you do anything directly related in the summers or other breaks.

So that’s a lot of experiences, and again if you value experiences enough, you might not balk at the idea that much in experiences could cost quite a bit.

Of course if you can’t afford it, then you have alternatives. But if you can? Honestly, I don’t know a lot of people who, say, who could afford to start flying business class for international vacations and then went back to basic economy. Doesn’t get you there any faster, but worth it? Of course, if you can afford it and you are the sort of person who thinks like this about experiences.

2 Likes
147

“Only”…?! As in, “You only had a life-enriching experience!” Kind of like college, ideally, should be. Goes back to what I said before that I don’t look at the ROI on college as just future income vs cost of the degree.

4 Likes
148

Yes, that was a little rhetorical device on my part. I also value such things highly.

1 Like
149

I figured, but sometimes you have to spell it out….!

1 Like
150

You could still guesstimate :wink: The top 1% is in the 450k total comp range. Out of undergrad there are only two paths - AI labs, Quant finance. Big tech tops out in the 250k range max with a few candidate specific outliers. In some sense you can just add # of students hired at each of the firms in this small universe.

2 Likes
151

I was half joking. The people that are not interested to begin with will not bother counting, often because they have no exposure – you don’t know what you don’t know. Those decisions are often made early in high school, not in college – at least the decision to be prepared in terms of rigor of high school coursework. The kids that are not interested are not going to socialize with the kids that are interested and knowledgeable. If they become interested some time in college, and did not have the general math prep whether they are interested in these careers or not, it is often late for some of these careers. There is some overlap in skills between big tech and the AI labs. I think some of the IB→ PE / HF paths will earn you high income by year 5 etc, even if they start slower. Revenue per employee at some of the top PE shops is often competitive with the quant firms.

2 Likes
152

“The kids that are not interested are not going to socialize with the kids that are interested and knowledgeable. “

You are not describing any reality that I have seen. The kids who end up doing Quant play Ultimate, play chess, play in a jazz band, work out, volunteer, take swing dancing lessons and do all the other social things their peers do. On which college campuses do the kids segregate themselves based on their career goals???

And every quant shop has a couple of hires who had NO idea what they wanted to do with their degree. Their friends ran the gamut of interests! Like every other college kid!

9 Likes
153

I am closer to the college dynamic than you are because I have seen my kids, their friends etc move through both high school and college more recently :slight_smile: . Whether they are in quant or not is not the matter. But they hang out with kids that have like interests. The hanging out is not based on career interests, but often about academic interests. There are always exceptions about hanging out with the theater kids etc, but the theater kids are not changing majors just because they are hanging out with the math kids. And vice versa.

2 Likes
154

My nephew (3 years so far as a quant) played saxophone, was in a drumming circle, and trained for two marathons in college. I stored stuff for his friends every summer and post-grad (I have a big basement and attic) and as they show up to finally collect their stuff i get to hear about what they do.

Teach history at a prep school. Med school. Law school. Finishing grad program in architecture. Major auction house as a specialist in Asian Art. Develops dynamic pricing models for a theatrical/concert ticket reseller.

Do you really think that math kids only hang out with math kids? What a sad world you are describing. And you do realize that quants read Tolstoy, visit botanical gardens, compose music, collect civil war ephemera, right? All sorts of interests….which allows them to maintain relationships even with friends who don’t make as much money as they do.

9 Likes
155

You are missing my point. I am going to leave it alone.

156

I get your point. I just don’t agree with you.

6 Likes
157

Love your lottery ticket metaphor.

I would add a “you must end up in the top X percent of your class” dimension.

Going to Harvard versus ASU won’t help you much if you’re towards the bottom at Harvard in comparison to being top 10% at ASU.

158

Yes but being at the bottom of your class at Harvard is orders of magnitude different, with WAY better opportunities, than if you are at the bottom of your class at ASU.

6 Likes
159

I have an S24 who just finished his second year in college, and I agree with blossom.

Like, he is on two club sports teams, and those kids are doing a wide variety of different things academically. Those kids have then become most of his core friend group. A lot of those people also have been figuring out what they wanted to do during their first couple years, and talking to other people on the teams including older students has been part of that process for them.

I do understand there are certain sequences of prereqs where it may be more difficult to jump in later (although not necessarily impossible). But that certainly isn’t stopping these kids from interacting, and of course it is often easy to jump OFF those tracks even when it is harder to jump on.

5 Likes
160

I find that some of the prep is set in place sometime in high school and hard to fix. I also think nerdy kids hang together. There are always exceptions. I’ve seen many years of school dynamics. And some years of college dynamics. And anyway, even if you become aware of opportunities through friends sometime in college, it is very hard to fix gaps. I am not saying everything is hard to fix. You can go down the consulting path with 3-6 months of prep. Medicine usually requires long and patient prep. Banking needs networking prep from early sophomore year. Quant needs a mindset that is prepped in high school. A software engg prep can start pretty much anytime up until junior year maybe. You need like minded company at the right time in life for some path to be interesting and seem viable to you. Or you need very determined help to bridge gaps in an accelerated manner. Everybody has their life experience and experience watching other people. This is my experience.

1 Like