PEW Research on the "Rising Cost of Not Going to College"

This week PEW Research published survey results showing the effect of a 4 year degree on earnings unemployment and poverty.  They found that the median income of respondents with a 4 year degree was $45,500 while the median income of respondents with only a high school degree was $28,000. Respondents with a 2 year degree did not fair much better with a median income of $30,000. So a Bachelors degree (or more) produces a 62.5% wage premium over a HS diploma and a 51.6% wage premium over an Associates degree or some college.

Comparing their current results to earlier PEW studies, they found that the gap between the earnings of college grads and non-grads has been growing since 1979. The results are best viewed in their graph of earnings from 1965 to 2013. The authors note that, even though the percentage of college graduates in their sample has been increasing (and the percentage of people with only a HS degree has been shrinking), median income in their surveys has remained flat since 1965.

Their report goes on to consider many more aspects of the situation and is definitely a must read. Of particular interest to college students and their parents is the section on regrets among college graduates. While one might expect that many would regret their choice of major, only 29% listed that as a major regret. The most commonly mentioned regret (50%) was not gaining more work related experience while in school.

While on the subject of income, the subject of assortative mating is getting some play due to an NBER paper by Greenwood, Gunar, Kocharkov, and Santos. Assortative mating refers to the tendency of people to marry other people with similar education/skill levels which results in people with high (low) earning potential forming households with someone with a high (low) earning potential. The prevalence of high-high and low-low pairings in comparison to low-high pairings exacerbates measures of income inequality where the unit of analysis is households, not individuals. Looking at 2005 data on incomes, the authors find that, if people were randomly paired with others, the gini coefficent of household income would decrease from .43 among actual households to .34 among randomly paired households (a 25% reduction).

So what do you do with this. Matthew Yglesias at Slate's Moneybox gives it a whirl.