TABLE VII.
Baseline characteristic | Control | Treatment |
---|---|---|
No. of spells = 1,322 | No. of spells = 1,371 | |
Demographics | ||
Age | 16.13 | 16.15 |
Black | 0.83 | 0.84 |
Hispanic | 0.14 | 0.12 |
White | 0.03 | 0.03 |
Other | 0.01 | 0.01 |
No. of prior JTDC spells including focal spell | 3.35 | 3.19 |
Reason for JTDC admission | ||
Violent crime | 0.17 | 0.18 |
Property crime | 0.10 | 0.09 |
Drug crime | 0.08 | 0.07 |
Other crime | 0.35 | 0.35 |
Directly admitted (no police arrest recorded) | 0.30 | 0.30 |
Number of prior arrests | ||
All offenses | 8.25 | 7.85* |
Violent offenses | 2.11 | 2.00 |
Property offenses | 1.55 | 1.49 |
Drug offenses | 1.34 | 1.25 |
Other offenses | 3.22 | 3.08 |
Neighborhood characteristics | ||
Percent adults 25+ with at least HS diploma | 72.80 | 72.75 |
Percent black | 69.13 | 68.12 |
Percent Hispanic | 17.71 | 18.44 |
Percent below poverty | 34.54 | 34.58 |
Unemployment rate | 18.59 | 18.58 |
p-value on F-test of treatment-controlcomparison for all baseline characteristics | .65 |
Notes. Asterisks indicate statistical significance of pairwise treatment-control comparison for a given baseline characteristic, clustering standard errors on individual and clustering on census tract for the neighborhood characteristics. Neighborhood characteristics come from geocoding address information and linking to tract level data from the American Community Survey. Joint significance test for equality of all baseline characteristics uses only nonmissing data (n = 2,643) and clusters standard errors on individual. * p < .10, ** p < .05, *** p < .01.