Letters of recommendation are an important addition to UC’s existing holistic review admissions processes, which aim to look beyond grades and scores to determine academic potential, drive, and leadership abilities.
Background
Unlike other highly selective universities, UC Berkeley did not ask applicants to submit letters of recommendation prior to 2015. This may have limited the information available for use in holistic review, and some at Berkeley believed that as the university became more selective it was getting harder to make informed admission decisions with the evidence available. Others, however, were concerned that students from disadvantaged backgrounds may not have had access to adults who could write strong letters, and that the use of letters would further disadvantage these students in the admissions process.
Research Project
CPL was asked to evaluate the second year of a pilot experiment in which a subset of applicants to UC Berkeley (approximately 30%) were invited to submit letters of recommendation if they wished. Any submitted letters were incorporated into a “second read” evaluation of their applications.
CPL evaluated the impact on the admissions outcomes of applicants from four groups typically underrepresented among successful applicants to Berkeley: students from families with low incomes, students whose parents did not attend college, students from low-scoring high schools, and students from underrepresented racial and ethnic groups.
A second phase of the project examined the actual content of the letters for further information about the role that letters play in admissions. We examined whether letter writers introduce bias by writing different or more positive letters for students from advantaged backgrounds, and whether readers equally reward strong letters for students from underrepresented groups.
Research Team
Professor Jesse Rothstein (Principal Investigator), Patrick Kennedy, Charles Davis, Elsa Augustine
Results
CPL found that the option to submit letters of recommendation led to a more diverse class of admitted students overall, though many students did not submit letters when given the opportunity to do so. There was little evidence that letters submitted on behalf of underrepresented applicants were distinctive in their language or style. Application readers evaluated underrepresented applicants’ portfolios more positively when letters were available, though this effect diminished as the letters grew stronger. The effect was to raise the admissions chances of underrepresented applicants, concentrated among those in the middle of the distribution of traditional academic credentials rather than among those with the strongest credentials.
Berkeley is now asking for letters of recommendation from students who are at the cusp of admission and are subject to a round of augmented review.
Read the reports:
Report: The Impact of Letters of Recommendation on UC Berkeley Admissions in the 2016-17 Cycle Jesse Rothstein (2017)
Working paper: “Varying impacts of letters of recommendation on college admissions: Approximate balancing weights for subgroup effects in observational studies” Eli Ben-Michael, Avi Feller, and Jesse Rothstein (August 2021)
JOURNAL PUBLICATION: “Qualitative information in undergraduate admissions: A pilot study of letters of recommendation”
(Forthcoming in Economics of Education Review)
Pre-publication version 