Skip to main content
. Author manuscript; available in PMC: 2022 Mar 22.
Published in final edited form as: J Hum Resour. 2019 Nov 13;56(4):997–1030. doi: 10.3368/jhr.56.4.1115-7543r5

Table 4:

Health, Violence and Drug Use, Marginal Effects

Domestic Violence Cocaine Use Heroin Use
[1] [2] [3] [4] [5] [6]

Full Sample
Treatment × HAART −0.015* −0.017* −0.029** −0.022 −0.022*** −0.019**
(0.008) (0.009) (0.013) (0.014) (0.007) (0.008)
0.068 0.063 0.027 0.109 0.003 0.011
Obs. 6669 6669 16265 16265 16261 16261

Pre-HAART
treatment group mean 0.103 0.103 0.178 0.178 0.087 0.087
Mean 0.063 0.063 0.121 0.121 0.052 0.052
Pseudo R2 0.026 0.059 0.006 0.065 0.009 0.090

Black Sample
Treatment × HAART −0.021** −0.024** −0.034 −0.028 −0.022** −0.016
(0.011) (0.011) (0.021) (0.022) (0.010) (0.011)
0.044 0.034 0.103 0.201 0.037 0.151
Obs. 4280 4280 9355 9355 9352 9352

Pre-HAART
treatment group mean 0.130 0.130 0.228 0.228 0.089 0.089
Mean 0.069 0.069 0.145 0.145 0.053 0.053
Pseudo R2 0.033 0.057 0.007 0.058 0.009 0.103

Basic Controls N Y N Y N Y

This table shows the marginal effects of the interaction term from the difference in difference probit models. Standard errors are presented in parenthesis, and p-values are found below. Basic controls include age at visit, age squared, age cubed, race (Caucasian omitted), and site of visit (Chicago omitted). In all specifications, errors are clustered at the individual level.