I want to cross-pollinate this essay with a dialogue of sorts that I had on Facebook with a reader who argued that the current data shows that college graduates make more money than non-college graduates.
This is actually not something about which I disagree. I think the data does indeed show that college graduates make more money, but I think the interpretation of that data and the methodology of the studies is so flawed that we cannot conclude at all that graduating college generally causes you to make more money, only that it currently correlates to it.
What follows is a fairly short dialogue in which I make the case that claiming the data proves college helps you make more money is like claiming basketball helps you grow taller.
Derek: I’m not convinced there are wage premiums in general anyways. What they’re comparing now is apples to oranges. But regardless, yes only 54 percent graduate in 6 years or less, which is staggeringly dismal.
Protagoras: The literature among labor economics has been pretty clear about there being premiums. The the issues is that they have been flattening.
Derek: I think the data shows something else and their methodology is skewed.
Everybody who can read today is encouraged from a very early age to go to college. Over the last 20 or so years, most of the smartest people went to college whereas not going was often a pretty bad sign you had some issues.
It’s a difference of input and not output. People going to college were more likely to be successful before starting than people who didn’t.
An accurate study would have parallel universes where you take Derek A and put him through college and Derek B and have him not go. Then see what happens.
Protagoras: In other words “labor economics is wrong because I only accept facts in absolutes”.
Derek: It’s the total opposite. Data is open to interpretation. They’re making an absolutist claim from data that can have multiple valid interpretations.
All they’re measuring is “college grads vs non-college grads” If I strictly measured “basketball players vs non players” I could easily conclude basketball gives you a premium on height even though it doesn’t.
I also filmed a video version of this argument while I was in Poland last year.
Data-ism is not science
I think one of the unfortunate products of the Information Age is a sort of worship of data.
While data is indeed important, we fall into silly and often dangerous data-ism when we attempt to use data to make our case without reference to the methodology behind the collection of that data or an understanding of the important of the interpretive process. “Look at that data” becomes a cudgel to win an argument and not one tool among many to inform our understanding of the world around us.
Though my example in the dialogue — playing basketball makes you taller — is ridiculous, the data could suggest that until you interpret it for what it actually means, that taller people play basketball to begin with.
The studies about college graduate incomes that claim to “prove” they make more money are done with the exact same methodology as my fictional basketball study. We should be wary about concluding anything definitive that could inform our decisions about whether to go to college or not.
Does college cause you to make more money?
I don’t know whether college causes you to make more money or not, and nobody else does either. I do however think we can look at several points that suggest college does not have a serious impact on earnings, including the following:
- Smarter people tend to go to college. Dumber people tend to not go to college. We all know this from experience.
- It does not appear that most students show many measurable increases in learning or knowledge from their time college.
- Surveys suggest the majority of employers regard their new hires as under qualified and in need of remedial training.
- The data doesn’t take into account college debt and lost earnings of time in college (at least 6 years on average)
- Skipping college gives you a 4-6 year head start on working, assuming other traits are equalized (a college-capable optout as opposed to a slacker or dummy who simply couldn’t get in)
Of course, I don’t think we can make a definitive conclusion, but I don’t think we need to.
Individuals, not averages
One of the things I realized when I was in college was that I didn’t care one way or another what the data ended up showing.
If you base your important life decisions around what “average people with average results” are doing, you’re probably going to end up average.
What proud defenders of the college income data are really doing is reducing individuals to statistic and arguing for being part of the middle, which is certainly better than being at the bottom, but it wasn’t enough for me.
And if being average isn’t enough for you, I suggest you stop obsessing over the data, stop following what everyone else is doing, and create your own path. Life is much more fun that way.