Its true that companies may not pay a Brown/MIT bonus - but the way I see these salary surveys is that: with a Brown/MIT degree you have a better probability of getting into those companies that pay top dollars.
My evidence is purely anecdotal - worked in high-tech industry for 20 years from mediocre companies to top names. The companies with good but not spectacular pay - I used to see new-grads from mostly local tech-colleges or state univ. Then I got into one of FAANG which pays new-grads really well - I see lot more of new-grads from top colleges (there are still good number of students from local colleges). I think there is a bit of selection bias when it comes to new grad recruitment: it is easy to get recruited if you know someone in the company. top colleges have the advantage of more alumni there and with a good name recognition its easy to score an interview (than a student from a not-so-well-known college). It may sound like not a big diff - but it is - that maybe a reason for better salaries from top colleges. After 3-4 years experience - no one really cares about college but your first few years of exp carries forward.
The selection bias toward more selective colleges at some of these big computing employers is more likely due to the selection of strong students at those colleges to begin with, where strong students are more likely to pass the technical interview gateways to get hired at those employers. Big companies are more likely to recruit very widely, as well as being so well known that anyone can just apply to them (no connections needed to do so).
The advantages of being local, being at a better known school (overall or for the major), having connections (even if the employer has strict anti-nepotism rules, connections can help inform a job seeker that the employer and job opening exists to apply to), etc. are likely to be more applicable to smaller companies that are less well known and do not recruit so widely. They may also apply more strongly to college graduates seeking major-agnostic entry level jobs (probably lots of liberal arts major graduates), where testing on specific skills typically learned in school and directly applicable to the job (as for those seeking computing and engineering jobs, for example) may not be the main focus of interviews.
Based on the data, there could be a Ivy to FAANG pipeline for Computer Science or a geographic preference in recruiting. The top private colleges almost all have starting salaries of over $100k while the large public colleges have starting salaries of $20k or $30k less except for a couple of well-regarded public universities in California.
@hs2020dad The number of graduates of a college will weigh heavily in favor of the very large public engineering programs. So there are much more likely to be UIUC, Berkeley, Michigan, etc alumni in positions of hiring than Brown or Upenn.
When talking about engineering, Brown isn’t really one of the “top colleges”.
The salary results do suggest CS grads from certain colleges have substantially typical starting salaries than others, but the reasons for that difference is not clear. I would not assume the primary reason is employer preference. Instead I’d expect differences at the student level have a greater contribution to salary differences.
One student level difference is the general quality of the student body. Highly selective colleges have their choice of students and tend admit a larger portion of students who excel in the criteria CS employers value and are more likely to ace interview questions. This can lead to CS employers being more likely to hire grads for selective colleges, even if the employers hired college blind without being aware of the name of college students attended. Using IPEDS reported test scores for the specified years as a proxy for general quality of student body, a list of the reported median salary for the highest test score colleges is below.
The typical CS salary for this high scoring college group looks like ~$120k, with all high scoring colleges having median starting salaries far above the national average, but there is a good variation. Chicago, Johns Hopkins, Northwestern, and a few others are ~20% lower than expected based on test score; while Brown and a few others are ~20% higher than expected base on test score. This +/- 20% could be primarily error due to a small sample size. However, the lower salaries seem to be especially concentrated at colleges that are not known for CS, rather than just a random distribution. The relatively lower salaries also tend to fit well with other sources. For example, WUSTL reports a salary of $96.7k (CS is 2nd highest salary major) at https://engineering.wustl.edu/our-school/why-wash-u-engineering/Documents/AnnualReport18.pdf compared to $96.2k in the ScoreCard database. So I’d expect that there really are significant differences in median salaries between different colleges among similar score range students.
CS Salaries For Highest 25th/75th ACT/SAT Score Colleges
Caltech – $153k
Rice – $112k
MIT – $118k CS / $117k EECS
Mudd – $125k
Olin – NA
Johns Hopkins – $93k
Chicago – $87k
Harvard – $129k
Yale – $120k
Vanderbuilt – $98k
WUSTL – $96k
CMU* – $139k
Penn – $135k
Princeton – NA
Northwestern – $89k
Brown – $141k
Stanford – $126k
*CMU SCS averages far higher than overall school average above
This still does not necessarily mean that CS employers prefer grads from Caltech/CMU/Brown/Penn… over Chicago/Hopkins/Northwestern/WUSTL/… Another key factor is portion of grads who are interested in and apply for jobs in high salary + high cost of living areas such as Silicon Valley. It’s not a given that every CS grad applies to Google in Silicon Valley or similar, and such a position is their 1st choice. Many would prefer to live in other areas, including areas where cost of living is a small fraction of cost of living in SV, making a 20% lower salary go further in quality of life. This is particularly true for graduates from public schools. In contrast, I’d expect few Brown grads would choose local Rhode Island employers. Along the same lines, students who are really concerned with the prestige and reputation of their college are probably far more likely to be really concerned about the prestige and reputation of their CS employer, leading to an much higher rate of applying for high salary SV type jobs and a high rate of choosing them upon other alternatives when there are multiple job offers.
There is not enough information to estimate student preferences for particular employer like this. We can see some trends in which companies colleges with higher and lower salaries choose. For example, for the sample years, Brown’s OIR reports the following most common employers among CS grads. Note that almost all Brown CS grads seem to chose high cost of living areas, and extremely few choose local employers.
Here is another comparison of the NCES numbers for federal FA recipients to the college’s own general career survey. This is for Virginia Polytechnic Institute and State University (Virginia Tech).
I think a lot of that difference may be selection bias. For example, as summarized below, the majority of persons with a high salary major seem to list their salary in the UCB post grad salary survey. However, the rate is roughly half that UCB grads in lower salary fields. It’s likely that persons with higher salaries are more likely to list their salary in the UCB salary survey than persons with lower salaries. The CollegeScorecard salaries are based on tax returns, so they don’t have this type of selection bias towards higher salaries.
The VT survey shows a similar pattern, but not to the same extent as UCB. For example, the engineering and liberal arts schools have the same survey response rate of 70%. However, among those who responded there was a notable difference in the portion listing salaries. 90% listed salaries for College of Engineering vs 70% for College of Liberal Arts & Human Sciences. It’s likely the VT grads with lower salaries are less likely to list their salary.
Another factor may be how the post grad questionnaire is phrased and grouped. Persons who are working part time in an unrelated field while looking for a long term more relevant job may be grouped as “seeking employment” in the UCB salary survey, but they would be part of the median in the CollegeScorecard tax return analysis… Similarly the VT survey seems to exclude short term work, which would be included in the CollegeScorecard tax return analysis…