In pursuit of prestige? Preliminary evidence of the effects of need-blind admissions policies for international students

• In Event: Education, prestige and funding policies in education

Proposal

Recently, several selective private institutions in the United States have adopted need-blind admission policies for international students, including Bowdoin College and Dartmouth College, both implementing these policies in the 2022-23 admissions cycle. In the press release announcing its policy, Dartmouth attested that: “For any international applicant for whom affordability is an essential factor of a U.S. college search, our adoption of universal need-blind admissions means financial need will never limit access to a Dartmouth education” (Coffin, 2022). Prior research indicates that domestically-focused need-blind admissions policies may help selective institutions increase socioeconomic diversity through the recruitment of lower-income, first-generation, and working-class students (Gunther & Benson, 2021; Monks, 2018).

Notably, the institutions that recently implemented need-blind admissions policies for international students are prestigious, private institutions. Toma (2012) argues that higher education institutions, even already-prestigious ones, adopt common policies aimed at attracting more accomplished students to maximize prestige. Zerquera (2023) identifies prestige as “a strong driving force in higher education” wherein “colleges and universities prioritize prestige-boosting behaviors to enhance their perceptions in the public eye and in rankings” (p. 268). Prestige-seeking behaviors often include enrollment management strategies that shift an institution’s enrollment towards affluent and high-achieving students (Brewer et al., 2001; Hossler, 2000; Jaquette et al., 2016). In a news article describing its 2022-23 entering class, Dartmouth noted the near-record number of applications it received, resulting in a selective acceptance rate of only 6.2% (Dartmouth, 2022). While this article also highlights diversity among the institution’s new class, nods to institutional prestige were foregrounded.

Brewer et al. (2001) suggest that prestige-seeking institutions may pursue applications from students they do not perceive as admissible to decrease acceptance rate and increase selectivity. O’Meara (2007) concurs and indicates that institutions in pursuit of prestige “actively solicit applications from lesser qualified students to make the admissions process more selective” (p. 146), behavior that can also impact offers of admission that are accepted, or yield rate. This practice is thought to make the institution appear more desirable to external observers. Prestige-seeking behaviors have been associated with significant equity concerns (Zerquera, 2023), and the long-term consequences of policies that support prestige-seeking include reduced access, particularly for marginalized student populations (Doran, 2015; O’Meara, 2007).

The purpose of this study is to explore the impact of need-blind admission policies for international students at Bowdoin and Dartmouth Colleges on three outcomes: number of applications, acceptance rate, and yield rate. The effect of these policies may align with their stated purpose, such as an increase in applicants, and less obvious effects, such as increases in institutional selectivity and yield rate. These latter two outcomes are often associated with prestige-seeking, behavior that may have long-term negative consequences for prospective students, domestic and international.

Data

The data used for this study derive from the U.S. National Center for Education Statistics’ Integrated Postsecondary Education Data System (IPEDS) for academic years 2001-02 through 2022-23. Institutions were included if they had data available across all 22 years. Given that the two institutions that implemented need-blind admissions policies for international students in the 2022-23 admissions cycle, Dartmouth and Bowdoin, are highly-selective, private, not-for-profit institutions located in the US Northeast, the dataset was further limited to institutions in these same categories. These delimitations resulted in a total of 26 institutions, Dartmouth, Bowdoin, and 24 institutions without need-blind admissions policies for international students.

Analysis

A synthetic control analytic approach was used to explore the potential impact of need-blind admissions for international students on this study’s three outcomes (Yan & Chen, 2023). This approach is useful for causal inference in instances where the treatment unit is a specific case, such as a country, region, or institution (Abadie et al., 2010). Synthetic control constructs a counterfactual unit that is as comparable as possible to a treatment unit by drawing from donor unit pool (Abadie et al., 2015). In this study, treatment institutions are Dartmouth and Bowdoin Colleges, while all other institutions in the dataset comprise the donor pool.

Synthetic control assumes that a weighted combination of units from the donor pool are more effective in reproducing treatment unit characteristics rather than a single comparison unit (Abadie et al. 2015). This approach exploits pre-treatment data (academic years 2001-02 through 2021-22) over time to construct counterfactuals using a weighted average of donor pool units to best resemble the treatment unit (Abadie et al., 2015). In this study, pre-treatment indicators of non-US resident enrollment, total enrollment, full-time tuition/fees, and all three outcome variables contributed to counterfactual construction. The first three variables were used because this study explores a policy intended to impact enrollment generally and international student enrollment specifically among students who tend to pay full-time tuition/fees. That is, a reasonable counterfactual would be similar to the treatment unit in terms of overall enrollment, international student enrollment, and tuition/fee charges in the pre-treatment period. Additionally, reasonable counterfactuals would exhibit similar pre-treatment time trends across all three outcomes.

Counterfactuals were constructed individually for each treatment unit and for each outcome, for a total six counterfactuals. Comparing outcomes of the treatment unit and its counterfactual in the post-treatment period provides an estimation of the effect of policy implementation (Abadie et al. 2015). Given the limited post-treatment data currently available, the results outlined here are preliminary. An additional year of post-treatment data will be available and included in this study prior to the CIES conference.

Results

Preliminary results indicate that both treatment institutions exhibit a decrease in the number of applicants and an increase in acceptance rate in the year after need-blind policy implementation, as compared to their counterfactuals. Regarding yield, Dartmouth exhibits a decrease while Bowdoin exhibits an increase in the post-treatment period, as compared to their respective counterfactuals.

Contribution

This study contributes knowledge about how U.S. higher education institutions may exploit international students to pursue objectives that benefit the institution in ways that do not necessarily benefit students, both domestic and international. Although these results are necessarily preliminary, the inclusion of additional data prior to the CIES conference will shed further light on the research questions explored here.

Author

• Melissa Whatley, William and Mary