| Schooling | Log-Earnings | Earnings | |
|---|---|---|---|
| Panel A: Baseline Schools | |||
| Grades 1–9 | –0.0080 (0.0093) | 0.0004 (0.0018) | 0.0020 (0.0024) |
| Grades 1–6 | –0.0089 (0.0150) | 0.00002 (0.0025) | 0.0023 (0.0035) |
| Grades 7–9 | –0.0006 (0.0115) | 0.0007 (0.0017) | –0.000004 (0.0024) |
| Panel B: Combined Schools | |||
| Grades 1–9 | –0.0026 (0.0082) | 0.0021 (0.0016) | 0.0038* (0.0021) |
| Grades 1–6 | –0.0017 (0.0068) | 0.0015 (0.0014) | 0.0029 (0.0019) |
| Grades 7–9 | 0.0003 (0.0039) | –0.0002 (0.0007) | –0.0006 (0.0009) |
| Panel C: All Schools | |||
| Grades 1–6 | –0.0014 (0.0070) | 0.0016 (0.0013) | 0.0020 (0.0018) |
| Grades 7–9 | –0.0025 (0.0026) | –0.0003 (0.0004) | –0.0004 (0.0005) |
↵Notes: The table summarizes class-size effects across the three different samples (all schools, combined schools, and the baseline sample of combined schools without takeup from other primary schools), and all three labor market outcomes. The estimates for the effect of average class size in compulsory school are based on the specification in Table 3, and the estimates for the separate effects of class size in primary and middle school are based on the specification in Table 4. Standard errors clustered at the school level are reported in parentheses. *p < 0.10, **p < 0.05.