Open the gates wider: affirmative action and dropping out

Nir Rotem*, Gad Yair, Elad Shustak

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Affirmative action policies are oftentimes pitted against the need of universities to maintain meritocratic standards in enrollment. The current study tackles this institutional dilemma against the standard of student attrition. It does so by analyzing records of 41,483 undergraduate students who attended the Hebrew University of Jerusalem (2003–2015). Approximately 5% of the students were marked eligible by an affirmative action policy that is unique to Israel. Specifically, a non-governmental organization oversees the Israeli indirect affirmative action policy and grants applicants with the status largely based on class. Descriptive statistics reveal no significant differences in dropout rates between affirmative action students and their normative peers. To verify those patterns, we test hierarchical logistic models alongside advanced decision tree models. The findings show the superiority of first-year grade point average and other academic indicators in predicting dropouts. They also confirm that students who are eligible for affirmative action depart at virtually the same rates as normative students and other risk groups. We conclude by suggesting that under certain conditions, universities do not pay any price by admitting students through the backdoor of affirmative action. Therefore, universities can and should open their gates wider.

Original languageAmerican English
Pages (from-to)551-566
Number of pages16
JournalHigher Education
Issue number3
StatePublished - Mar 2021

Bibliographical note

Publisher Copyright:
© 2020, Springer Nature B.V.


  • Affirmative action
  • Dropout
  • Institutional admission policy
  • Israel
  • Meritocracy
  • Student attrition


Dive into the research topics of 'Open the gates wider: affirmative action and dropping out'. Together they form a unique fingerprint.

Cite this