↵Notes: The estimates are based on the sample of students who graduate from compulsory schooling, 1978–2001. The schooling outcome is defined as years of educational attainment, log-earnings is the log of (positive) yearly earnings, and earnings is the yearly earnings relative to the average earnings of the cohort of compulsory school graduates. The first three columns report results from regressions of these outcomes on student background characteristics. The four last columns report results from regressions of the instrument (above maximum class-size threshold in primary school [PS] or middle school [MS]) on student characteristics. Column 4 reports these estimates for all students, Columns 5 and 6 report the estimates for the sample of students who graduated from a combined school, while the final column reports the results for the baseline sample of students who graduated from combined schools without takeup from other primary schools. The regressions control for segment and segment-specific linear splines in both primary-school (Columns 5, 6, and 7) and middle-school (Columns 4, 5, and 6) enrollment, as well as interactions with three maximum-class-size regimes dummies for the primary school controls. All regressions include year of graduation fixed effects, individual background controls for gender, parental education, immigrant background, month of birth, and age at graduation. F-tests (and corresponding p-values) are for the joint significance of the variables reported in each column. Standard errors clustered at the school level are reported in parentheses. *p < 0.10, **p < 0.05, ***p < 0.01.