Table 12: Comparable ITT treatment effects by contractor
Panel A: Student test scores English (standard deviations) Math (standard deviations) Composite (standard deviations) Panel B: Changes to the pool of teachers % teachers dismissed % new teachers Age in years (teachers)
48
Test score in standard deviations (teachers) Panel C: Enrollment and access ∆ Enrollment ∆ Enrollment (constrained grades) Student attendance (%) % of students still in any school % of students still in same school Panel D: Satisfaction % satisfied with school (parents) % students that think school is fun Observations
(1) BRAC
(2) Bridge
(3) YMCA
(4) MtM
(5) Omega
(6) Rising
(7) St. Child
(8) Stella M
(9) p-value
0.16* (0.09) 0.07 (0.10) 0.10 (0.10)
0.25*** (0.09) 0.33*** (0.10) 0.31*** (0.10)
0.25 (0.16) 0.13 (0.18) 0.19 (0.17)
0.20 (0.15) 0.13 (0.16) 0.16 (0.16)
0.05 (0.11) -0.01 (0.11) -0.00 (0.12)
0.25 (0.16) 0.24 (0.18) 0.26 (0.17)
0.23* (0.12) 0.22* (0.13) 0.23* (0.13)
0.07 (0.17) -0.00 (0.19) 0.02 (0.18)
0.089
-8.84 (6.35) 38.60*** (11.04) -5.53*** (1.70) 0.12 (0.13)
50.36*** (7.01) 70.87*** (12.89) -9.10*** (2.17) 0.25* (0.14)
12.37 (12.87) 35.93* (20.77) -3.46 (3.56) 0.06 (0.23)
14.01 (11.28) 48.36*** (18.74) -7.63*** (2.55) 0.23 (0.18)
-5.22 (6.65) 23.44** (11.77) -5.79*** (1.72) 0.18 (0.13)
1.43 (11.91) 20.68 (20.28) -7.99*** (2.74) 0.17 (0.18)
-2.21 (8.98) 37.05** (14.95) -6.53*** (2.08) 0.23 (0.16)
-7.02 (12.80) -8.55 (25.84) -5.92** (2.70) 0.17 (0.18)
<0.001
26.54 (25.45) 52.93 (39.15) 18.04*** (5.43) -1.23 (3.38) 0.52 (1.79)
5.92 (26.65) -45.78*** (11.55) 10.75* (5.69) 1.26 (3.52) 2.48 (1.94)
22.38 (33.05) – (–) 15.86** (8.08) -1.99 (5.21) 0.14 (2.85)
11.56 (32.22) – (–) 21.78** (8.81) -3.33 (5.61) 0.40 (2.59)
25.13 (25.70) -22.11 (39.47) 11.07* (5.74) -0.98 (3.56) 0.72 (1.88)
16.94 (31.95) 52.21 (39.25) 19.04** (8.32) -2.51 (5.22) 0.76 (2.58)
22.28 (28.74) 18.40 (51.82) 18.21*** (6.81) -1.02 (4.20) 0.29 (2.26)
19.78 (32.76) – (–) 13.39 (8.16) -1.96 (5.05) 0.30 (2.64)
0.91
11.90* (6.34) 4.17 (3.89) 40
11.60* (6.47) 2.83 (3.59) 45
8.36 (9.38) 4.80 (5.95) 8
3.10 (8.72) 2.74 (5.43) 12
1.66 (6.28) 3.40 (3.99) 38
2.02 (9.29) 3.57 (5.50) 10
-0.84 (8.49) 2.84 (4.60) 24
9.82 (9.41) 0.24 (6.51) 8
0.025 0.033
0.0027 0.12 0.47
0.0031 0.22 0.80 0.82
0.19 0.58
This table presents the ITT treatment effect for each contractor, after adjusting for differences in baseline school characteristics, based on a Bayesian hierarchical model. Thus, this number should be interpreted as the difference between treatment and control schools, not as the mean in treatment schools. Column 9 shows the p-value for testing H0 : β BRAC = β Bridge = βYMCA = β MtM = βOmega = β Rising = β St.Child = β StellaM . Some operators had no schools with class sizes above the caps. Standard errors are shown in parentheses. Estimation is conducted on collapsed, school-level data. ∗ p < 0.10, ∗∗ p < 0.05, ∗∗∗ p < 0.01