My university (Ohio State) did absolutely terrible on this ranking. That’s good to know.
I’m suddenly very glad that I’m applying to washington and lee
I find this “ranking” system more confusing than most. I’m not even sure what the point of it is because I doubt most kids looking at colleges are going to look at it.
Why do JHU and Rice perform so poorly?
Johns Hopkins and Rice produce an above average number of pre-med students. It may be that their higher incomes do not kick in by the time this study looks at the data. Additionally, pre-med students who do not make it to med schools sometimes get stuck in fields like biology that are over-supplied and have depressed wages.
@zinhead The large number of pre-meds at those schools was the reason of my surprise at their poor performance. But I just learned that the averages are for 28 year olds ( according to the OP). In that age, most med students are still doing their residencies, which basically means that the list is completely inaccurate for schools with large pre-professional populations. In other words, the inaccuracy is not limited to schools whose students have a tendency to pursue careers in the academia. I feel that the data will show a high co-relation with the percentage of CS, IB, Econ and etc. students. Correct me if I am wrong.
Thanks for the answer.
They adjusted expectations based on the mix of majors. However, there was no adjustment for other factors like “fast track to Wall Street” recruiting paths (described in the article in relation to W&L).
Also remember the most important piece of data, which is the graduate earnings, in this analysis is collected by matching up students who revived federal aid and their tax returns. Which means for some schools that subset could include very few of the graduates and others a much bigger proportion.
My nephew graduated from Rice, he’s at MIT getting his PhD. Not earning big bucks either.
This is my point : The article fails to note that certain majors need more training before they are marketable.
Thank you for answering my question.
BTW : sorry for posting in the parents forum. I am not sure whether this is acceptable or not.
I agree that grad/professional school differences between comparable schools is a contributing factor, but the average age is not 28* and varies for different colleges. For example, the source data lists a mean age is 20.5+10 = 30.5 for Rice and 22.9+10 = 32.9 for Johns Hopkins. I suspect age of “entry” means age of starting the fed financial aid programs, rather than age of starting college.
*Average age is only 28 for Bard College and some seminary schools, many LACs are 29, Harvard and Columbia are 34, SUNY Empire State is 44, …
Those bizarre average ages for “entry” plus 10 years may be consistent with the tiny n for Harvard and Columbia, but I can’t imagine how Hopkins gets an average “entry” age of 22.9.
Thanks for posting this discussion. I was curious about the dichotomy between CalTech (near the bottom) and MIT (near the top). It doesn’t make much sense to me. Apparently the Economist has data that show that CalTech students make only 74,000 median income. Is this because an inordinate number of CalTech grads are in grad school or take government jobs (at JPL) or is it just bad government data? In contrast, other rankings/studies I have seen (Forbes) have said that Cal Tech students have some of the highest median incomes.
Some other of the earlier posts about investment banking and management consulting as explaining the difference between Harvard and Yale is ignoring the reality of the college job market today. The hottest employers on campus are high tech–the Googles, Facebooks and Amazons. Their summer internships are paying pretty competitively now with the banks, so I would attribute Yale’s relative underperformance on not just the fewer number of graduates interested in finance careers but also its lower percentage of STEM and engineering graduates. When you talk to employers at technology firms, they are doing much more recruiting at Harvard, Stanford, Illinois and Berkeley than they are at Yale.
(Another possible factor about income disparity: lawyers, which Yale seems to have an overabundance of, have stagnating pay.)
Wow, what a surprise. I never would have thought STEM majors might be more in demand than non-STEM majors. 
Out of curiosity, how is it that “the effect of SAT scores” parameter is so insignificant in LACs, compared with universities? Compare Amherst with Rice to see what I mean. Even Harvey Mudd’s effect of SAT parameter is lower than universities with lower average scores. What does the effect of SAT signify?
“Yale’s not in the top! There must be something wrong with the methodology.”
Oh CC.
It seems that the Economist included being a high-end LAC as an independent variable. Since high-end LACs tend to have high SAT scores, the impact of being considered a LAC might have “washed out” the SAT score impact. Without seeing the regression formula and how they structured the independent variables, it is hard to tell.
The Economist didn’t publish the final regression formula, but by looking at different colleges, you can see the impact of various independent variables on expected earnings. I went though the results and pulled out some conclusions:
Marx and Marley - Reed -$8,605 = This suggest that colleges that are very liberal/pot smoking have worse income outcomes. Other high Marx and Marley universities like Smith and Oberlin are also dinged in expected earnings.
Business College - Ohio State +$2,243 = This suggests that having a business college adds to the expected earnings.
College Size - Central Florida +$3,488 = UCF is the largest college in the US. The positive college size indicator suggest that larger schools have higher income outcomes. This is backed up by similar adjustments for other large schools like Ohio State and Texas A&M. Oddly enough, Liberty University with 13,800 residential students and 100,000+ online students gets a large bump due to college size. Small colleges like Sweet Briar get dinged.
Sex Ratio - Otis Art and Design -$1,770 = Otis has a ratio of 33%M/67%F which is very lopsided. It seems a high ratio of women drive down earning potential. This may be related to the lower percentage of women in STEM but other schools with high female ratios like Howard have the same issues. However, all female colleges like Wellesley do not seem to be impacted for this. Also, some schools with more men than women, like Harvey Mudd and Missouri Valley, have a positive impact on earnings due to sex ratio. All male colleges like Moorehouse and Wabash College also had large positive impacts on earnings due to sex ratio.
Religion - Holy Family University +$2,394 = This category was surprising. It seems that schools with a Catholic background like Villanova get a bump for religion. However, schools of other faiths gets dinged. For instance, if you look up all of the schools with Baptist in their name, several of them have negative adjustments for religion. I checked some predominantly Jewish schools and did not see religion show up as a major independent variable.
Racial Diversity - This was confusing as it is hard to correlate the racial diversity of various campuses with how the IV were coded by the Economist. For example, several HBCUs like Bethune-Cookman and Dillard had positive impact due to diversity, which others like Clark Atlanta had negative impacts due to diversity. For non-HBCU’s, it seemed that having a low diversity hurt expected earnings (Tulane has a -$1,427 for diversity and has a somewhat high percentage of Caucasians compared to other schools).
Public/Private - FSU -$2,210 = Being a public school dinged the expected earnings compared to private schools.
I’m a little late to the game. Just reading this thread for the first time. I was curious about this and just checked. I’m not sure if the web site I’m on is accurate, but if it is, then 39% of the recently admitted Harvard class was from the East coast, from Maryland up to Maine, including Pennsylvania, at 3%. 7% were from the southern East coast, from Virginia down to Florida. California comes in at 14% and Texas at 5%. Some midwest states and Colorado, and that’s it.
That’s surprising, especially the number of states that aren’t represented at all.
In the vast majority of recent classes, Harvard has had students from 49+ states. The most commonly missed state is Wyoming. Harvard has never had more than 2 students in their entering class from Wyoming in any year available in IPEDS, and they usually only have 1. I haven’t seen stats for the current entering class, but it would be really shocking if a large number of states were not represented.