The Case of College Coaching in the United States

Fieldwork by:
New Hampshire, United States of America
2,624 students in 20 high schools
2009 - 2014
Target group:
  • Higher education and universities
  • Students
  • Rural population
Outcome of interest:
  • Enrollment and attendance
Intervention type:
  • Information
  • Social networks
AEA RCT registration number:

A number of programs in the United States seek to encourage young people to attend college. Researchers examined whether in-person mentoring or the provision of information increased rates of enrollment among high school seniors on the margin of going to college. Mentoring increased enrollment rates among women but had no significant effect among men. There was no significant effect associated with mailing students additional information about the benefits of college and letters encouraging them to apply.

Policy issue

The United States ranks 12th in the world in terms of the fraction of 25-65 year olds who have completed four years of college (32-35 percent). With evidence suggesting high rates of return to college education, policymakers in the United States have sought to encourage more young people to attend college. Programs offering high school seniors a few hours of counselling, information packages, or application fee waivers have increased college enrollment among high school seniors at relatively low cost. Would these programs help students on the margin of going to college, and if not, would more comprehensive mentoring succeed? Furthermore, how do modest interventions so late in a student’s academic career have an impact on such an important decision as attending college?

Context of the evaluation

In New Hampshire, there are 79 high schools as well as a number of public and private post-secondary institutions. While 64 percent1 of New Hampshire high school graduates enroll in college, many of their peers who are similarly qualified for college do not enroll. Of the 14,000 students in the class of 2010, more than 1,000 students with math test scores higher than the median score did not apply to any colleges. These students illustrate a key research question: why do qualified high school seniors fail to apply to and enroll in college?

Two women having a conversation
College mentoring increased enrollment rates among high school women in New Hampshire, but no effect was seen for high school men. Photo:

Details of the intervention

Researchers partnered with twenty of the largest high schools in New Hampshire to examine whether in-person mentoring or mailings influence college enrollment decisions. 2,624 high school seniors were identified by guidance counselors at each school as having expressed an interest in attending college while making little to no progress in applying. Depending on the year of the study, roughly half of the identified students in each school were randomly assigned to one of two treatment programs while the remaining half were assigned to the control group.

In the mentoring program, Dartmouth College undergraduates met weekly with groups of high school seniors to provide advice and assistance in applying to universities. Application and test fees were paid upfront by the program, and participating high school students were offered a $100 cash bonus for completing the application process. In the mailing program, students could opt to have their transcripts sent to colleges in New Hampshire. Roughly twenty-five percent of these students (based on the quality of their transcript) received a letter from one of four selective four-year colleges in the state encouraging them to apply. In addition, all students in the mailing program were sent a letter from the New Hampshire Community College System highlighting the benefits of college and providing a URL to enable the student to apply.

Results and policy lessons

The offer of mentoring increased enrollment among women by 14.6 percentage points from an initial rate of 41.1 percent but had no significant impact on men. Women who worked with a mentor and enrolled in college were also no more likely to drop out before their second year than students who enrolled without being mentored. This suggests that mentoring did not encourage students who were unprepared for college to enroll.

Follow-up surveys offered possible explanations for why the program was successful for women but not men. Men had lower perceived returns to college than women and forecasted higher hourly wages at age thirty with only a high school degree. American Community Survey data support these forecasts. In New Hampshire, there is no apparent wage premium for men aged 22-30 with one to three years of college compared to men who only have a high school education; wage premiums only become evident among older men with some college education. For women 22-30 years old, however, there is a wage premium associated with having some college education.

Survey responses suggest that the mentoring program may have increased women's perceived returns to college education and had assisted students who could not rely on help from parents when preparing their applications.

Offering to send students’ transcripts to in-state universities had no significant effect on college enrollment. Only 14 percent of high school students accepted the offer to have their transcript sent to colleges. Of those that did, very few responded to letters of encouragement by submitting an application. Students who have made little progress applying on their own may require more hands-on support than was offered through the mailing program.

Carrell, Scott E. and Bruce I. Sacerdote. "Late Interventions Matter Too: The Case of College Coaching New Hampshire." NBER Working Paper No. 19031, May 2013.
“College-Going Rates of High School Graduates - Directly from High School,” National Center for Higher Education Management Systems, accessed June 1, 2015,