Transferring to a better college leads to lower income
Striving to get into the fanciest college might not be the best choice
I’ll send a newsletter about every two weeks.
I was there in the room when one of my favorite high school students got the rejection e-mail from UNC-Chapel Hill. She melted down. Sobbing. Shivering. She was going to be the first person in her extended family to go to college somewhere other than UNC.
“It’s ok,” I told her, “you can go to a different college, and then transfer to UNC after two years.” She was mollified, and I felt good about my advice.
I was probably wrong.
New research from economist (and soon-to-be “Dr.”) Lois Miller shows that, among students who apply to transfer to a more selective college, those who are narrowly accepted eventually earn lower incomes than those who are narrowly rejected. Transfer students earn between $7,000 and $11,000 less per year. For students at a community college or lower-tier four-year college, staying put and completing their degree where they started would be better.
How could this be? Here are the reasons Lois considers in the paper: “The mechanisms include transfer students’ substitution out of high-paying majors into lower-paying majors, reduced employment and labor market experience, and potential loss of support networks.”
However, the main reason was that men who transferred worked less after college. 11 years after transferring, men who transferred worked a job an average of 2.5 years less than those who didn’t transfer. This seemed to be driven by transferring men taking longer to graduate. Strangely, these spells of unemployment didn’t occur right after college, but were likely to happen several years later. Women didn’t have this problem. Lois speculates that this may be because men are overconfident about their ability to succeed at a harder school.
I asked Lois what she thought the takeaway from her paper was. She replied, “If your kid can get into a flagship college as a freshman, they will benefit from it, but if they don’t get in, they should go to a less selective four-year college and stay there.”
By the way, this is not a “correlation isn’t causation” situation. These are causal effects. The paper studied students who applied for transfer who were just-a-tiny-bit below an arbitrary GPA cutoff vs. those who were just-a-tiny-bit above the GPA cutoff. They all wanted to transfer. The students were statistically identical in every way except for a fraction of GPA. But their school experience led them in different directions.
Because this study is only about marginal students, it didn’t look at students who were very far above the GPA cutoff. If a particular student is getting really good grades in community college, transferring still might be the right choice.
As parents, teachers, and students, what do we do with this information?
First, let’s not get so hung up on a school’s brand. There are a lot of great reasons to get a flagship college education, but to be able to go around saying you went to so-and-so school isn’t one of them.
Secondly, students should evaluate how they are doing in their current college, and be honest with themselves about whether they are ready for an extra challenge at their new school.
Thirdly, students should take their support network seriously. The relationships they build over the first two years of college may be crucial for getting them through the last two years.
Fourthly, your college major matters. It may be better to get an in-demand degree at a lesser school than a different degree at a more prestigious school. I recommend you major in econ!
Fifthly, employers are looking for skills, not just a big-time college name on a diploma.
A word for parents of younger children (pre-K through elementary grades): you don’t need to be fanatically dedicated to getting Junior into that flagship school. Maybe extra tutoring will get your child into a top college, but he might not succeed there. This is especially important for students who are at the margin between one school and a (purportedly) better one. While your child grows up, help him be successful where he is, rather than where he isn’t.
My student who didn’t get into UNC turned out fine. When I caught up with her during her freshman year at UNC-Greensboro (a great school), she gushed about how much she loved it and didn’t want to leave. She didn’t transfer, even though she could have. I’d say she made the right choice.
Comment below to keep the conversation going. Coming soon on Paper Robots: how to get your kids to stop begging for things in a store, and family economics in the TV show “Bluey.”
Interesting read. I never thought about the time addition to my transferring from Purdue to Indiana U of PA. But there was delay, and that was lost opportunity to work. Students transferring may not have all credits accepted, etc. So yes, there would be a cost benefit from on staying, getting the education, and entering the work force.
I'd like to clarify something. If the average annual income gap is driven largely by transferring men earning $0 for 2.5 years out of 11, wouldn't we do better at predicting lifetime earnings by looking at the different groups' incomes *at the end* of the 11-year period, along with the differing rates of increase leading up to that point? We might even be better off looking at where income lands after a certain number of *working* years, discounting the unemployment period, unless there's reason to believe unemployment will continue to be higher for upper-tier graduates. I admit, though, I may be misunderstanding the scenario: you said the 2.5 year gap was largely due to delayed graduation, yet also that it tended to take place several years after graduation. By definition, a delay in graduation occurs before graduation. Can you tell me what I'm missing?
Setting aside these questions, I wonder about the reason for a pay gap. For one, perhaps instead of suggesting the problem is male applicants being overconfident (or ambitious, aspirational, or determined, depending on how one wants to frame it), we might find the problem in schools that admit students who aren't likely to succeed. (Didn't we take issue with mortgage companies in the mid 2000s giving out loans that information-deprived buyers were unable to pay back?) Maybe the schools using the wrong techniques for predicting success. Or is school success even a predictor of income here? I'm not sure if Miller's study compared the *post-transfer* GPAs of the students who ended up with lower vs. higher incomes in the years that followed. If the workers who ended up making less money did not have particularly low GPAs *at the schools they transferred to,* we may need another explanation. (We might also wonder how students thinking of transferring can know if their GPAs are marginal; there's an information asymmetry here.)
Standardized test scores could be another way to go here: the literature is pretty clear that the SAT and ACT are better predictors of college success than grades -- a fact that's been shared a lot lately given the SCUSA affirmative action case and how schools can achieve diversity.