After reading the last couple of pages, you might suppose that anyone with a college degree has it made. College degrees help many thousands of people earn higher incomes. But they do not ensure success or affluence, and many college graduates remain poor. Nearly 2 million college graduates nationwide fall below the poverty threshold. They make up 9.5 percent of all adults in poverty. Of all poor adults in Indianapolis, 7.6 percent have college degrees. For Central Indiana, it’s 8.7 percent.  

Source: American Community Survey

The “scatterplot” chart at right illustrates a relationship between college attainment and college-educated poverty. Each dot represents a city. Its position on the horizontal axis shows the share of all adults in that city who have college degrees. The vertical position of the dot, meanwhile, shows the share of college-educated adults in that city who are poor. The chart positions 62 major U.S. cities according to the two variables. There is a clear positive correlation (r2=.48).⁵

The upward-sweeping line running through the chart indicates the trend. Simply, communities with higher levels of college attainment also have more college-educated poverty.

The cities with the highest incidence of college-educated poverty in the sample are San Francisco (22.1 percent); New Haven, Conn. (20 percent); Boston (19.7 percent); and Austin, Texas (19.2 percent). Among Midwestern cities, Minneapolis is the highest, with 15.6 percent. By contrast, the rate of 8.7 percent for the Indianapolis metropolitan statistical area, illustrated in the graph, is low.

This chart will perplex many people, as it seems to go against the accepted wisdom that college graduates earn more. That is still true, but there are only so many college-wage jobs. The staffing pattern in each community simply cannot put more than 20 percent to 30 percent of its workforce to work in high-wage jobs. The demand for high skills and higher education is greater than in the past and growing, but it is not infinite. Consequently, any community that succeeds too well at getting kids through college faces the possibility that some of those graduates will be waiting tables and dispensing coffee.

The U.S. Department of Labor Web site CareerOneStop: Pathways to Career Success provides a table for each occupation, showing what percentage of people doing that job have a diploma or less, some college and a bachelor’s degree or more. It may be the best tool available to find out more about the issue of overstocked college degrees. The information, however, is not broken down by state but rather is for the entire country.

5 The r-squared value is the statistical measure of co-variation of two variables. It measures whether a change in one variable is likely to happen along with – or independent of – a change in the other variable. A value of zero would mean the two variables do not vary together at all, but were independent. A value of 1.0 would mean perfect correlation. It would appear on the chart as a solid line of dots.