I thought I read that the correlation between your income and your parent's income was stronger than the correlation between your education and your income. I can't find a reference, though, so it's possible I'm making it up.
"I find that family income remains an important positive predictor of eventual adult outcomes. The effects persist even when many characteristics that are related to income, such as parents’ education, home environment characteristics, parental involvement, school characteristics and student ability, are controlled in a regression framework. "
...
"Even conditional on a host of other characteristics such as grades, test scores, and parents’ education and level of involvement, a student from a family earning $50,000 a year can still expect to earn about 10% less in the future than an otherwise identical student from a family earning $150,000 a year. These relationships are all statistically significant at the 1% level."
The parent was responding to a quote about successful people, so there would be no reason to look at all college graduates. A significant number of them will not be successful (as measured by income).
As an aside, while trying to find income data by education that excludes outliers (sadly, I failed to find anything) as prompted by your comment, I noticed that college graduates with only a bachelors degree represent roughly 20% of every single income group, except for incomes below $30K, where they only represent ~10%.
I haven't had much time to digest the data, so I am quite possibly overlooking something obvious, but wouldn't you expect that number to rise with incomes if simply having a degree provides an income advantage?
