When It Helps to Have Rich Parents

It always helps to have rich parents, of course! But do rich parents specifically make it easier to get a four-year degree? That's the topic of a new National Center for Education Statistics report and a New York Times writeup from Susan Dynarski.

Here's the key breakdown from Dynarski, showing that, even after taking math ability into account, richer kids are more likely to get degrees. According to the NCES report, the patterns are similar for reading.

Glancing this over, I was curious to see what would happen if we looked at ability differently — taking into account the whole spectrum instead of grouping kids into four chunks, and not limiting our measure to math or reading alone. Fortunately, the National Longitudinal Survey of Youth has combined math/verbal test data on thousands of kids from 1997, along with parental-income figures, as well as follow-up information from the 2011-12 academic year on whether they had a degree.

Here's a rough chart I put together; the shaded areas are the confidence intervals:

As I suspected, something interesting happens out toward the extremes: At very high or low levels of ability, parental income appears to exert somewhat less force. For kids in the broad middle when it comes to smarts, however, income seems to have a lot of sway, often more than doubling a kid's chances of getting a four-year degree — it can boost the likelihood more than 30 percentage points. No doubt much or most of this is simply the power of money; other factors could include culture and personality traits, which also vary between socioeconomic groups.

I suppose we can take some comfort in the extremes: Geniuses graduate from college even if they're poor; those with very low ability rarely get degrees even if they're rich. But it's obviously troubling when the fortunes of those in the middle depend on how much money their parents have.

(Standard disclaimer: I'm a journalist, not a social scientist, so this graph is a very simple analysis of the data. There are no weights or adjustments for age, for example, and unlike the NCES report I used parental income alone to indicate socioeconomic status, rather than factoring in other things as well. I did restrict the data to the nationally representative 1997 sample and exclude cases with no follow-up information, of course. You can see my R code here and download NLSY data yourself here.)

Robert VerBruggen is editor of RealClearPolicy. Twitter: @RAVerBruggen