This dissertation comprises three essays that empirically examine the educational outcomes of for-profit college students, military enlistees and immigrant youth. All of these are groups of "non-average" students that, in different contexts, pose challenges to the traditional provision of education. Therefore, their outcomes need to be studied in order to assess the need and room for public policy measures to intervene. The first essay, Academic and Early Labor Market Outcomes of For-Profit College Students, employs a selection on observables framework on a nationally representative longitudinal study and finds that, compared to their peers in the public sector, for-profit students experience higher debt; when starting at 4-year institutions, lower 4-year degree completion rates; when starting at 2-year institutions, higher 2-year degree completion rates, but higher unemployment and lower earnings. Results are robust to departures from a selection-on-observables-only assumption. The second essay, The Effect of Military Enlistment on Education, aims at identifying causal effects by comparing veterans to non-veterans who applied to enlist and are similar in the characteristics that the military uses to screen applicants. The results indicate that enlistees delay college but eventually enroll at comparable rates to similar non-enlistees; furthermore, enlistment positively impacts degree attainment at two-year institutions but negatively impacts degree attainment at four-year ones. The third essay is entitled Home-Country Academic Quality, Time Spent in the U.S., and the Math Achievement of Immigrant High School Students. By virtue of augmenting survey data with scores from international education assessments, it shows that home-country academic quality has a positive and significant relationship with mathematics achievement in the U.S., and that such relationship tends to decrease in size as a function of time since migration. This evidence suggests that one reason for the segmentation of immigrant assimilation along national lines, a phenomenon documented in the literature, is the diversity in academic background. The following are appended: (1) Countries of Origin with Less than 20 Students; (2) States with Favorable Tuition and Financial Aid State Policies for Undocumented Immigrants; and (3) Detailed Regression Results.