You have /5 articles left.
Sign up for a free account or log in.
I remember the first time I saw a car ad that referred to J. D. Power’s “initial quality” ratings. It struck me immediately as missing the point; years later, it still does. Initial quality measures the number of identified problems that emerged within the first 90 days of owning a car. Any car owner can tell you that most issues like those are trivial; what matters is how the car lasts after five or 10 years. For most people, cars are long-term purchases. A transmission that dies after 50,000 miles is deeply unsatisfying, but it won’t show up in the initial quality ratings. Put bluntly, initial quality is measuring the wrong thing.
Starting salaries out of college strike me as similar. A colleague recently retrieved this piece from The New York Times in 2019 detailing the relative salaries over time between four-year graduates in STEM majors and four-year graduates in the humanities and social sciences. The short version is that the STEM majors are way ahead in the first few years, but the humanities and social science majors catch up over time. They reach parity by age 40, and some may even come out a bit ahead after that. Even if they struggle in their 20s, they tend to find their way to materially satisfying careers.
Asterisks abound, of course. Some of the higher earnings that come later are a function of graduate training, whether in law, medicine, business or something else. It’s hard to attribute that to their humanities training, exactly, but bachelor’s degrees in English and political science have long been understood to work well as pre-law. Attrition from STEM-based fields is relatively high over time, for reasons I’ll leave to people who know those areas better than I do. Parenthood has drastic effects, as can the timing of graduation: people who graduate into recessions can take years to catch up. And as with any aggregate data, there’s no shortage of exceptions, some of which are quite vivid.
It’s also possible that the world has changed in such fundamental ways in the last few years that the past is no longer prologue. I tend to discount arguments based on the premise that “this time is different,” though. Patterns have a way of reasserting themselves. And while past performance is no guarantee of future results, neither is it guaranteed to be irrelevant.
But the asterisks don’t override the basic message. Choosing a major in a field like history is not a form of economic suicide. The skills honed through historical study—combined in uncertain proportions with the social signaling function of a degree—count.
If cars should last years, careers should last decades. Let’s use the right data.
