| Schooling | Log-Earnings | Earnings | |
|---|---|---|---|
| Panel A: School Size | |||
| First segment | –0.0025 (0.0086) | 0.0009 (0.0016) | 0.0017 (0.0022) |
| Second segment | –0.0178 (0.0206) | 0.0029 (0.0037) | 0.0021 (0.0044) |
| Panel B: Single School Municipality | |||
| Average class size, Grades 1–9 | 0.0070 (0.0280) | –0.0032 (0.0040) | –0.0052 (0.0056) |
Notes: This table shows IV estimates using the baseline specification in Table 3 for the first and second maximum-class-size thresholds separately (Panel A) and for the subsample of single school municipalities (Panel B). The results in Panel A use all available data around each discontinuity to estimate the class-size effect at each threshold, not only the observations with enrollment within ±15. The results in Panel B are based on 32 municipalities with only one school and use a sample of 4,134 individuals. Standard errors clustered at the school level are reported in parentheses.