Putting technology to work for inclusive prosperity: Challenges for public policy


Putting technology to work for inclusive prosperity: Challenges for public policy


…And justice for all: Community colleges serving the middle class

A woman walks along the campus of DeVry University in Chicago, Illinois, U.S., September 20, 2017. Picture taken September 20, 2017.  REUTERS/Joshua Lott - RC158CCCFBA0

Where Metallica leads, the government should follow. While the rest of us were distracted by stories of America’s aristocracy fake-rowing their way into elite colleges, the heavy metal band just made the right kind of donation: $1 million in grants to 10 community colleges ($100,000 per college) to fund career and technical education programs. Much larger gifts to elite colleges are a dime a dozen, of course, but donations to community colleges are all too rare, considering their greater need.

Community colleges serve America’s middle class. They educate at least as many middle- and low-income college students as public four-year colleges do, but receive much less funding and less media attention. For students who would not otherwise have attended college, they are vital engines for upward mobility and preparation for the labor market.

Where the middle class goes to college  

About 60% of middle-class students born in 1991 attended college at some point between the ages of 19 and 22. Among those who attended college, two in five went to nonprofit two-year schools, according to our analysis of data from Chetty et al. (We define the “middle class” as the middle three income quintiles.) This exceeds the percentage who attended public four-year schools (36%):[1]

By contrast, fewer than one in four students (23%) from households in the most affluent quintile attend community colleges, while 26% attend an elite or private four-year college.[2] 

A few details about the data. Students are assigned to the institution they attended most frequently between the ages of 19 and 22, so students who transfer from two- to four-year schools will be assigned to the institution in which they spent the most time. If students split their time equally across schools, they will be assigned to the first institution attended.

Given the age cutoffs, the analysis does not include the many adult learners in the higher education system, who primarily attend community colleges.

Additionally, institutions are classified as two- or four-year based on the highest degree offered. This means that some schools that are primarily two-year institutions may nevertheless be classified as four-year institutions if they offer any four-year degrees.

Alternative estimates 

To see how these numbers line up with another data source, we re-estimated the distribution of students across college tiers using the High School Longitudinal Study of 2009 (HSLS09) conducted by the National Center for Education Statistics. A few key differences between the two sources:

  • Chetty et al. have data for nearly the full population of students in the U.S.; HSLS09 is a nationally representative sample of around 23,000.
  • Chetty et al. divide students into quintiles based on their parents’ pre-tax, pre-transfer household income; HSLS09 divides students into quintiles based on a composite measure of socioeconomic status, which incorporates parents’ educational attainment and occupational prestige, as well as family income and school urbanicity.
  • Chetty et al. assign students to the institution they attended most frequently between ages 19 and 22; for the HSLS09 analysis, we looked at the most recent institution attended by 2016 (3 years after expected high school graduation).
  • Above, we used data for the 1991 birth cohort in the Chetty data; HSLS09 tracks students who were in 9th grade in 2009, most of whom were born in 1994-1995.

Despite these methodological differences, the two datasets provide fairly similar pictures of undergraduate enrollment. The HSLS data shows that middle class students attend community colleges and four-year publics in roughly equal proportions (37% and 40%)[3]. The full quintile breakdown is below.  

The takeaway here is that despite differences in how college enrollment by income is measured across data sources, it is clear that middle-class students are about as likely to attend community college as they are to attend public four-year institutions, if not more so.  

Middle class kids increasingly heading to community college 

We next pool the middle class (middle 60%) in the Chetty data for each birth cohort to identify trends in where the middle class enrolls in college. The previous section showed that levels will differ slightly under different methodological choices, but a consistent methodology should provide insight into trends over time.

The share of middle-class college students enrolled in public four-year institutions declined by 9 percentage points between the 1980 and 1991 birth cohorts, offset by increases in enrollment in community colleges and the for-profit sector:

Some of these trends may be due to a change in composition rather than a change in the likely destination for a particular student. The share of middle-class students in the dataset who attended any college increased from 49% to 60% over this period. If students on the margin of college attendance are more likely to go to two-year schools than those who have traditionally enrolled in college, we may see a shift toward even greater two-year enrollment as overall college enrollment increases. But it is clear that the increased demand for a college education for those from a middle-class background is being met in large part by community colleges (and to some extent, for-profit colleges too).

Community colleges: neglected but vital  

Middle-class students are at least as likely to attend two-year colleges as they are to attend public four-year colleges. Our analysis suggests that this is even more true now than in the past. Whether these findings are “good” or “bad” is a matter of interpretation: some will rightly raise concerns that students are enrolling in institutions with low completion rates where they will have more trouble succeeding.

The challenges facing community colleges are severe, as Harry Holzer and Zeyu Xu show in a recent report. But, they argue, low completion rates at community colleges are at least partly the result of funding constraints. As Glenn Hubbard, Austen Goolsbee, Amy Ganz write in a paper for the Aspen Institute Economic Strategy Group, “despite their promise and potential, community colleges are under intense resource pressures that constrain the educational and labor market outcomes of their students.” In particular, investments in career counselors, mental health resources, academic coaching and supplemental instruction look promising. Interventions in New York City and Texas, for example, have managed to raise persistence and completion by providing students with greater support while pursuing associate degrees.

One thing is clear: community colleges have a vital role to play in any serious effort to strengthen the American middle class.

[1] This figure excludes “elite” public institutions, which earn the top competitive ranking on the Barron’s index and are grouped with elite private institutions, including Ivy Leagues. Since only 1.6% of middle-class students attend elite institutions (public and private combined), adding the students who attend elite public colleges to those who attend other public four-year colleges would not make up the difference between two- and four-year public schools. 

[2] While the data do not separate public two-years from private non-profit two-years, the overwhelming majority of two-year colleges are public institutions. We refer to these interchangeably as two-year and community colleges.

[3] The difference between these two estimates is statistically significant at the 95% level. Standard errors and confidence intervals available upon request.