A redditor analyzed the earning potential of many top schools based on CollegeScoreCard Data, comparing the average earnings by major among students at the school to average earnings by majors across all schools.
Very-High Premium Schools (+$30,000 Starting Salary Premium) Harvey Mudd CalTech MIT UPenn Stanford Harvard High Premium Schools (+$15,000 Starting Salary Premium) Dartmouth Duke Johns Hopkins CMU Yale Claremont McKenna Georgetown UChicago Columbia Northwestern Vanderbilt Princeton Rice Cornell UC Berkeley NYU Moderate Premium Schools (+$10,000 Starting Salary Premium) Washington & Lee Bowdoin Georgia Tech Northeastern Notre Dame BC Pomona Amherst Villanova USC Emory Williams Swarthmore Barnard Colgate Wake Forest Middlebury BU UVA Tufts WashU St. Louis Wellesley Low Premium Schools (+$0 Starting Salary Premium) Trinity (Texas) Bucknell Wesleyan Brandeis Lehigh UMichigan UT Austin Colby Brown UCLA Davidson URochester UW Madison Haverford Case Western Bates UNC Chapel Hill Bryn Mawr UIUC UC San Diego Hamilton URichmond UMiami UFlorida William & Mary Kenyon UGeorgia Vassar Disutility Colleges (Less than $0 Starting Salary Premium) Tulane Macalester Carleton Grinnell Smith Colorado |
Averaging earning across major when you include some schools with multiple types of engineering majors and some without any seems disingenuous. |
How do you know this is accurate? Some of the results are surprising to say the least. |
It factors that in. For example- school A heavily provides engineering degrees. the average engineering major earns 130K there. the average engineering major across all schools earns 125K. So school A has a 2.5K earning difference. School B heavily produces psychology degrees. the average psychology major earns 50K there. the average psychology major overall earns 35K. so school B has a 15K earning difference. By this measure, school B has better outcomes, even though school A has students earning more money. Basically, it's equalizing across majors and trying to assess if attending a particular school has a difference in the amount you earn. |
They posted their dataset online and the numbers seem to add up. One of the big limitations though is that this doesn't account for geographic discrepancy. Coastal school grads tend to stay on the coast where average salaries are higher. |
Yeah, mostly this is a map of which schools send kids to NYC/LA/SV, without controlling for COLA. |
There are some issues with College Scorecard. For starters, as noted by others, it doesn't take cost of living into account. College alumni tend to be more concentrated around the college itself, both in terms of state and region. Second, College Scorecard is sparsely populated. The number of students indicated is not the number of salary data points but the number of students graduating with a particular major. Third, College Scorecard only gets data on those taking federal loans. This may not be a representative group, depending on the college and your own student. Fourth, the site gathers data for students only four years after college. Colleges which send a higher proportion into grad school or into nonprofit services shortly after graduation could be more penalized, especially if those tend to be the higher performing students. There is also an argument that some schools teach skills that take longer to manifest in terms of salary, ie, developing unrelated skills (ie writing may not be important for an entry level tech worker, but it can become important as they try to break into management later on), how to learn brand new things (ie colleges with broader distribution requirements might teach how to confront one's own learning weaknesses in different academic contexts), or how to be a leader in a community (which could be influenced by the breadth of campus opportunities typically experienced by an undergraduate.)
I have always found PayScale more useful. It doesn't account for cost of living differences either, but it provides data over a broader economic group and career stages (anyone who contributes data) and is transparent about the number of data points. As an example, if I go to the latest version (posted last week I believe) of PayScale's College Salary Report for four year colleges and select "All Alumni" and look at the list sorted by "Mid Career Pay," I see Santa Clara at 16th (omitted from OP's list of "top schools.") I think it's is a good school but one that benefits greatly from its central location in Silicon Valley, so I am more likely to compare it to, say, Stanford, which is 2nd. If I look at a college like Carleton in Minnesota, it's ranked 67th. I'm more inclined to compare it to the top schools in the Midwest, like UChicago, Northwestern, and Vanderbilt. Carleton is in line with those (actually higher), even though the list quoted by OP has Carleton as a "disutility college" and the others as "high premium." This is all the more interesting given two of them offer engineering and Carleton doesn't. In fact, Carleton just edges NYU, which surprised me given that NYU not only offers engineering but is in the most expensive city and many grads work there after, so has a cost of living benefit. When looking through the info for Carleton on College Scorecard, most majors have no salary data at all, and the ones that do have an unknown number of source salary points. The bottom line is College Scorecard is a good idea but has too little useful data to be of value with this kind of comparison. https://www.payscale.com/college-salary-report/all-bachelors |
Haven't seen this list, interesting indeed. |
Kruger and Dale long ago put to rest any claims that the specific college attended has any effect on the earnings of graduates. Yes, grads of certain colleges earn more, but it's because of the individual's abilities, not anything the college did. This essay explains a bit and has a link to the original study, which is definitely worth a read.
https://lesshighschoolstress.com/page/6/ |
Doesn't control for student quality. Selective colleges graduate students who get paid more because the kids are smarter (or better connected), not just due to the college. Less selective colleges have more less smart students. A smart student could opt for a lowe tier college and still succeed well above average for that school. |
I think it has less to do with selectivity than school culture. Many of these schools are similarly selective. Some schools have capitalistic go-getter types and some have more earnest intellectual types. Some produce a lot of quant-y types across majors that tend to earn more right out of college. Others are more writing focused that tend to build their earnings over time (and I mean this distinction within majors--a linguistics major at MIT is more likely to go towards computational linguistics, at a liberal arts college more towards foreign languages). |
Did you not see Carleton and other top SLACs listed under “disutility”? |