Why the College You Go to Doesn’t Matter

This article is about a thorny issue in tech--compensation. Websites like Blind, Paysa, and levels.fyi have revealed numerous issues with how people in tech are paid, with the gender pay gap being the most well known. Employers are cagey about disclosing their reasons behind deciding how much employees get paid, and they encourage their workers to treat salary as a taboo subject that mustn’t be broached. This behavior helps the employers’ bottom line by keeping wages depressed.

It also leads to pay inequity. As new graduates entering the workforce, compensation has (unfortunately) become an important issue in our lives. College was expensive, and we each took out student loans to pay for our education. Now, as we are reviewing job offers and deciding where to begin our careers, we’ve begun to wonder whether or not the cost of our education was worth it when faced with the prospect of having to pay these loans off. This leads to an important question: Does the college you go to affect how much money you can make in industry?  Folks in other fields have found something they call the “Ivy League Pay Gap” -- where Ivy League graduates’ median salaries 10 years after starting school is over $36,000 more than their non-Ivy League counterparts. We wanted to find out whether something like this exists specifically in the tech industry. Levels.fyi and Glassdoor have lots of information on compensation based on experience in tech, but nothing connects one’s university background to how much employers decide to pay. 

The question of connecting college to compensation is critical now because of the rising cost of tuition. Tuition at public 4-year universities has increased 26% over the last 10 years, and has risen over 35% in private schools. Every prospective college student must consider difficult questions of how they are going to pay for their education. And the problem is worse in our field of computer science. Many schools are charging “differential tuition” — increased cost of education for those pursuing technical degrees, with the assumption that a career in technology pays more so it will be easier to handle a greater load of debt. 

We set up an anonymous survey to answer this important question. We asked employees of tech companies about their alma mater, their total compensation for 2019, the years of experience they have on the job, and their current location. We then posted our survey on several tech industry-focused tech Reddit and Facebook groups. This resulted in a total of 60 responses that represented 25 schools, from “top-ranked” (and often the most expensive) schools like MIT, to large state schools such University of Michigan and UIUC, to smaller lesser known schools including Truman State University. Critically, we found that the university you go to has little to no predictive power of your earning potential, with experience and internships having the most correlation to higher earnings.

Our methodology for arriving at this conclusion was as follows: We first eliminated all the data points from people that were not working in software. This left a total of 50 responses. We then used data here to apply a cost-of-living adjustment to the salary data based on someone’s location. The place you live has a high impact on the amount employers are willing to pay you, so we can’t compare the amount of money you earn without taking into account how much it costs to live where you work. After doing this we found the median salary of folks that went to name brand private schools was only $8,000 higher than those with cheaper degrees (as compared to the $36,000 dollar “Ivy League” pay gap). To be statistically rigorous, we ran a T-test* between these two groups of data and found that the small differences between them were statistically insignificant (p-value that they were different was 0.17). This indicates a good sort of meritocracy in the tech industry that isn’t present in other lines of work. The graph below illustrates our data:

There are some limitations to the data set we collected. Like many other surveys there is certainly going to be a self-report bias, and our sample size also has to be kept in mind when drawing any conclusions. Despite these issues we think our survey begins to indicate that the college you go to matters less than people think. At least in technology there seems to be little correlation between the university you go to and your earning power. With this being said, the tech industry is by no means perfect. It suffers from widespread gender discrimination and underrepresentation of minority groups. However, unlike the American job market as a whole, it looks like your Alma mater is not a big deciding factor in the shape of your career.

This has important implications not only for tech employees but also college students. For tech-interested high school students deciding what college they want to attend, it may not be worth it to attend expensive private schools when their earning power will be the same if they went to a cheaper public university. University administrators have two tough questions to answer: why has the cost of education risen so rapidly over the last few decades for a service that hasn’t materially changed in years, and why does it look as if this rise in cost may not even have a justifiable return on investment? 

* A T-test is a statistical tool that measures the likelihood that the differences observed between two sets of data are the result of an external force, or merely random chance.