news analysis

Article: Leonhardt, David. 2013. “Better Colleges Failing to Lure Talented Poor.” New York Times, March 16. http://www.nytimes.com/2013/03/17/education/scholarly-poor-often-overlook-better- colleges.html

Original Source: Hoxby, Caroline and Avery, Christopher. 2013. “The Missing “One-Offs”: The Hidden Supply of High-Achieving, Low-Income Students.” Brookings Papers on Economic Activity, Spring:1-50.

The main argument in the article is that low-income students are not applying to and

attending the country’s more selective colleges or universities, even if they are eligible to attend.

This has ramifications for income inequality because college graduates earn more, on average,

than their counterparts who do not graduate from college. Exacerbating this problem is the low

graduation rate of low-income students even from less selective colleges and universities. The

article is based on a research study published by the Brookings Institute, cited above.

The research uses data on every student who took the SAT or ACT in 2008, followed

through to their presumed date of college graduation. These are quantitative data. The data are

based on a population, because they include the universe of students who took the SAT or ACT

in that year, and not a sample of test-takers. However, many students do not take these

standardized tests and low-income students who did not take the test are not included in the

study.

The findings show that among high-achieving low-income students, only 34% go to the

most selective 238 colleges and universities in the U.S. In contrast, among high-income

students, 78% of high-achieving students go to these colleges. Instead, many low-income

students attend community college or attend a local college closer to their homes. This is

! 1

especially the case for students who live outside the largest metropolitan areas. Because less

selective colleges have fewer resources and lower graduation rates than the more selective

colleges, the author concludes that these students will be missing out on career selection and

advancement (although the data cannot tell the reporter that information because they do not

follow students through their careers). The data do follow students through graduation and show

that graduation rates are higher among low-income students attending more selective colleges

than those attending less selective colleges.

There is a quote in the article from a dean of admissions at Amherst College that says he

believes the findings. I don’t have any reason to dispute the statistical findings, but I don’t know

if there is enough information to support all the conclusions drawn. For instance, the author

Caroline Hoxby says that these students, “lack exposure to people who say there is a

difference among colleges.” It’s unclear if that finding comes from the data or is her own

opinion.

I did find the actual research report and it is more than 50 pages long so I did not read the

entire report. I skimmed through it and I read through an executive summary and it appears that

the author of the NYT article was correct in his reporting. The NYT article does not discuss the

actual quantitative methodology, but the findings are descriptive and the study used regression

analysis. The authors could not have used an experiment or a quasi-experiment because they are

looking at an entire population of students taking the SAT or ACT. However, the article says,

“The authors emphasized that their data did not prove that students not applying to top colleges

would apply and excel if colleges recruited them more heavily. Ms. Hoxby and Sarah Turner, a

University of Virginia professor, are conducting follow-up research in which they perform

! 2

random trials to evaluate which recruiting techniques work and how the students subsequently

do.” This means that it’s possible that the students who opt to apply to more selective colleges

may have some underlying characteristics that make them more likely to succeed in any college.

Recruiting more low-income students does not necessarily ensure their success at the same rate.

If the article wanted to be more comprehensive, it could have asked low-income students

their preferences for college-going. As a companion to the future experiment, the authors could

include a qualitative component that includes student voice instead of just administrators

speaking about how they believe low-income students act with regard to college application and

attendance. This could be through interviews or focus groups with low-income students in high

schools both in and out of major metropolitan areas.

! 3