Cristhian Molina

College admissions formulas determine which student characteristics translate into admission chances, and when they change, they can interact with differences along dimensions the designer did not target. I study Chile's 2013 ranking reform, which required programs to place at least 10 percent of admissions weight on a student's class rank within their own school. Because age at school entry affects academic performance, the reform raised the weight on a dimension correlated with the school-entry cutoff. Using a difference-in-discontinuities design, I find that the reform widened the application gap at the July 1 cutoff by 3.8 percentage points (12 percent) and the GPA gap by 0.082 standard deviations. The response concentrates in secondary-educated, higher-income families. Marginal applicants are less likely to be admitted or to enroll, producing no net gain in enrollment. Formula design can interact with age-at-entry differences to produce distributional effects in college access.

Employers cannot directly observe program quality, and institutional reputation provides reliable signals for only a small fraction of graduates. We study whether program-level accreditation in Chile affects graduate wages, using a staggered difference-in-differences design applied to administrative earnings records for approximately 2,000 programs and exploiting first-time accreditation decisions between 2007 and 2018. After accreditation, the ability composition of graduating cohorts shifts upward while the formal employment rate falls. The average wage effect is approximately 3 percent, concentrated at longer horizons when higher-ability cohorts begin graduating. The wage response varies systematically with institutional reputation and accreditation intensity, increasing by 9.1 percent at lower-reputation institutions and decreasing by 4.8 percent at elite institutions. The wage response is largest where prior information about the program was scarce, and the accreditation term was short.

This paper examines program-level accreditation in higher education, analyzing student responses and institutional adjustments to this quality signal. Using administrative data from Chile (2007-2018) and a difference-in-differences approach exploiting staggered adoption, we estimate the effects of this information shock. First-time accreditation increases student demand (applications +10.2%, enrollment +7.3%), attracts stronger students, and reduces transfers, suggesting improved match quality. These effects are consistent across socioeconomic backgrounds. Investigating the signal's construction, we find no evidence of pre-accreditation increases in long-term structural inputs but document a significant rise in on-time graduation rates concentrated immediately before evaluation and not sustained post-accreditation. Effects are amplified by institutional reputation and weaker under mandatory regimes. Our findings reveal a policy trade-off: accreditation signals can enhance student sorting, even when institutional responses focus on evaluation metrics rather than long-term, structural improvements.