In a nutshell
- What is the problem? Attending a selective college may carry large benefits, but high-achieving, low-income students tend not to apply to or to attend such colleges at the same rate as higher-income peers.
- What are possible interventions? A randomized controlled trial found that sending application fee waivers and information about selective colleges to high-achieving, low-income students increased their likelihood of applying to and enrolling in selective colleges. A funder may be able to extend the intervention to other students or support further research on the topic.
- Who else is working on it? The program that has been studied to date has received funding from a number of high-profile sources, including the Gates Foundation and the US Department of Education. The College Board has recently agreed to undertake the intervention for all high-achieving, low-income students who take College Board tests.
1. What is the problem?
High-achieving, low-income high school students typically do not apply to or attend selective colleges at the same rates as other high-achieving students, even though selective colleges may be cheaper than non-selective colleges for such low-income students (due to financial aid resources and scholarships).1 Because attending a selective college may substantially increase future earning potential—though we have not thoroughly investigated the evidence on this question2— low-cost interventions which increase the likelihood of high-achieving, low-income students attending such colleges could feasibly have a very large return on investment and may increase social mobility.3 We are aware of one randomized controlled trial which found that sending information on college quality and cost as well as no-paperwork application fee waivers to high-achieving, low-income students increased their rates of application and enrollment at selective colleges.
1.1 A randomized controlled trial of sending college information and application fee waivers to high-achieving, low-income students
In a recent randomized controlled trial, Hoxby and Turner sent a variety of information on college quality and cost and no-paperwork application fee waivers to 12,000 high school seniors who:4
Another 6,000 students who were in the top decile but either had higher family income or did attend feeder schools were included to evaluate whether the intervention had different effects on students outside of the target group.6 The 18,000 students were randomly divided into a total of six groups to receive four separate information or fee waiver treatments, a combination of all the treatments, or no treatment at all (control), to test a variety of theories that might explain their college application behavior: lack of awareness of quality differences between colleges, lack of awareness that the cost of more selective colleges is likely to be lower than the “sticker price”, or being deterred by the costs of applying.7 Versions of the intervention were run over multiple years, but the main results from the study focus on the comparison between the combination intervention and the control group in 2011-12 (the first year the combination intervention was tested) in the target group.8 Based on the results reported in the paper, we do not perceive selective outcome reporting to be a significant problem.9 The authors find, based on an intention to treat analysis, that receiving the combined intervention leads students to attend statistically significantly more selective schools (a 15 point higher median SAT score), with higher student and instructional spending and a higher graduation rate (3 percentage points).10 Application and admissions outcomes show similar changes.11 The authors do not report a result for overall college attendance rates. Because the intervention was extremely cheap—roughly $6 per mailing—these figures suggest a potentially very high return on investment, though the authors do not account for the possibility that admissions decisions may be zero-sum (i.e. that the enrollment of members of the treatment group may offset enrollment by other students).12
2. What are possible interventions?
There are several ways a funder could get involved in this area:
- Attempt to scale up the intervention described above to all high-achieving, low-income students (though major scale-up is already under way; see below).
- Fund further research. The researchers discuss their desire to undertake similar research projects for other target demographics (e.g., low-income mid-achieving students and mid-income high-achieving students, students at different stages in their education).13 A funder could also potentially provide resources to continue to follow the treatment group over a longer period of time, to determine the magnitude of the benefits they receive from attending more selective colleges.
- Lobby the federal government to share more extensive data with researchers in an appropriate fashion. The researchers note that interventions like ECO are only possible with ample data from the federal government, and that such data is currently hard to access.14
3. Who else is working on this?
The research project described above has received over $6 million in funding from the U.S. Department of Education, the Bill & Melinda Gates Foundation, the Spencer Foundation, the Andrew W. Mellon Foundation, and the Smith Richardson Foundation, as well as a number of other non-profit organizations, colleges and universities.15 The College Board (the organization behind the SAT, among other exams) has committed to carry out the intervention for every high-achieving, low-income student who takes a College Board test.16 This covers roughly half of all students that take college-admission tests nationwide. The state of Delaware has also recently committed to undertaking the intervention for every high-achieving, low-income high school senior in the state.17 Including high-achieving ACT takers in the mailing would ensure that most high-achieving, low-income students in the country would be reached by the intervention.18 We have not investigated the question, but given the considerable amount of funding and attention that this intervention has received, it seems unlikely that the limiting factor with respect to the ACT’s involvement is a lack of funding.
4. Questions for future investigation
Our research in this area has been relatively limited, and we have yet to answer many important questions. Amongst other topics, our further research on this cause might address:
- What are the benefits to a student of attending a more selective college? Our current impression is that the evidence on this question is fairly limited, due to the difficulty of controlling for the selection effect of talented people attending the most selective schools, but we have not investigated thoroughly.
- What are the obstacles to ACT scaling up the intervention?
- What related research is currently ongoing? Is sufficient funding for follow-up research already available? Does ECO plan to follow up the participants to track long-term effects?
- What are the broader social effects of the intervention? Presumably students who attend more selective colleges because of ECO displace other students who would otherwise be admitted. What is the magnitude of the (presumably negative) effect on those students?
5. Our process
We decided to look into this area after reading about the RCT described above in the popular press. We prioritized it more highly than we otherwise would have (relative to other potential shallow investigations) because we felt it was a useful training exercise for new staff members. Our investigation consisted of some limited desk research.
|Expanding College Opportunities website||Source (archive)|
|Gates Foundation Grants Database||Source (archive)|
|Hamilton Project Policy Brief 2013-03||Source (archive)|
|Hoxby and Turner 2013||Source (archive)|
|Mellon Foundation Summary of Grants and Contributions, 2008||Source (archive)|
|New York Times: “Delaware Seeks to Steer the Poor to Top Colleges”||Source (archive)|
|Smith Richardson Foundation Annual Report 2009||Source (archive)|
|Spencer Foundation Grant Summaries||Source (archive)|
|US Department of Education website||Source (archive)|