I'm on record arguing that it's a good idea to rank colleges, and hold them accountable, for what their students earn after graduation. But salary data can be tricky to interpret in a useful way. And this week, in a piece titled, "Where to Go to College if You Want the Highest Starting Salary," the Washington Post offered up an excellent example of exactly how not to do it.
Using salary data collected by the website Payscale, Roberto Ferdman breaks out the 25 schools that produce the best-paid grads during the early years of their careers. This is not a particularly useful bit of information. First, early-career salary probably isn't as important as what grads make later on in their lives. Second, most of the schools rank highly on the list because they're heavy on engineering students (or produce military officers), who tend to earn solid salaries out of the gate. I've reproduced the list below, with schools that Payscale says emphasize engineering in orange, and the rest in blue.
Surveying these results, Ferdman concludes that "the nation's most elite universities—which rank highly on U.S. News and World Report's college rankings list—aren't necessarily places to go if you want the best starting salary out of school."
That's not really correct. If you were actually interested in studying engineering, most of the schools up above would not be among your top choices, based solely on starting salaries (more on that in moment). As shown on the color-coded graph below, of the 18 engineering-heavy schools that made the Post's top 25, only seven also rank in the top 25 for the salaries that their engineering grads earn.
So who ranks highly on that list? Surprise: It's Ivy League universities like Columbia, Cornell, and the University of Pennsylvania, along with other elite institutions such as Northwestern, Johns Hopkins University, Rice, the University of California–Berkeley, and the University of Michigan. Sometimes, the conventional wisdom is pretty much right.1
I'm not pointing this out just to nitpick the work of Ferdman, who typically writes great stuff. Rather, I think it's an illustration of why it's difficult to suss out which schools are really a good value based on post-graduation pay. It depends on the mix of majors at the school. It depends on the socio-economic background of the students. And it depends on whether you look purely at salary, or at salary compared with the cost of tuition and the average debt kids take on. Even if it's worth comparing schools this way, its easy to get it wrong.
Footnote1: As an aside, I don't know whether I entirely believe Payscale's rankings, which are submitted by the site's users and thus suffer from some self-selection bias.