Table 4.

Effects of College Completion on Proportion of Time in Low-Wage Work: Recursive Partitioning Causal Tree

	IPW	NN Matching	Causal Forest (grf)	n
L1: mothers’ education < 12	−.225*** (.041)	−.261*** (.034)	−.220*** (.042)	1,832
L2: L1 & number of siblings ≥ 2	−.264*** (.041)	−.291*** (.034)	−.251*** (.045)	1,645
L3: L2 & low control ≥ 10	−.343*** (.045)	−.372*** (.056)	−.318*** (.063)	800
L4: L2 & low control < 10	−.189** (.068)	−.099 (.150)	−.176*** (.064)	845
L5: L4 & female	−.255** (.079)	−.263** (.091)	−.179** (.065)	425
L6: L4 & male	−.133 (.099)	−.130 (.142)	−.170** (.098)	420
L7: L1 & number of siblings < 2	−.037 (.111)	−.110 (.070)	−.043 (.123)	187
L8: mother’s education ≥ 12	−.124*** (.021)	−.140*** (.023)	−.150** (.029)	2,550
L9: L8 & ASVAB scale < −.44	−.347*** (.049)	−.381*** (.050)	−.355* (.102)	490
L10: L8 & ASVAB scale ≥−.44	−.086*** (.022)	−.089*** (.018)	−.100* (.026)	2,060

Note: Data are from the National Longitudinal Survey of Youth 1979 cohort. The sample is restricted to individuals who were 14 to 17 years old at the baseline survey in 1979 (n = 5,582), who had completed at least the 12th grade (n = 4,548), and who had no missing data on the outcome (n = 4,382). College completion is measured as a 4-year degree completed by age 25. Estimates are based on IPW, NN matching with four control units per treated unit on the linearized propensity score, and causal forest (grf) estimates applied to each partition. ASVAB = Armed Services Vocational Aptitude Battery; grf = generalized random forest; IPW = inverse propensity weighting; L = leaf; NN = nearest neighbor. Shading indicates instability in the partitions.

^*p ≤ .05

^**p ≤ .01, and

^***p ≤ .001 (two-tailedtests).