Table 3:

Health, Violence and Drug Use, Probit Coefficients

	Domestic Violence		Cocaine Use		Heroin Use
	[1]	[2]	[3]	[4]	[5]	[6]
Full Sample
HAART available	−0.365***	−0.235***	−0.203***	−0.304***	−0.163***	−0.273***
	(0.075)	(0.085)	(0.052)	(0.058)	(0.058)	(0.068)
Treatment Group	−0.061	−0.023	0.043	0.071	0.091	0.122
	(0.116)	(0.116)	(0.098)	(0.099)	(0.113)	(0.114)
Treatment × HAART	−0.202*	−0.213*	−0.171**	−0.129	−0.276***	−0.242**
	(0.117)	(0.121)	(0.080)	(0.084)	(0.098)	(0.104)
Obs.	6669	6669	16265	16265	16261	16261
Black Sample
HAART available	−0.411 ***	−0.315***	−0.246***	−0.361***	−0.177**	−0.396***
	(0.092)	(0.098)	(0.068)	(0.073)	(0.081)	(0.085)
Treatment Group	−0.004	0.041	0.085	0.150	0.086	0.126
	(0.140)	(0.141)	(0.126)	(0.129)	(0.147)	(0.155)
Treatment × HAART	−0.273*	−0.291**	−0.165	−0.135	−0.267*	−0.193
	(0.145)	(0.148)	(0.105)	(0.109)	(0.139)	(0.147)
Obs.	4280	4280	9355	9355	9352	9352
Basic Controls	N	Y	N	Y	N	Y

This table shows difference-in-differences probit model estimates where the outcome variable is having experienced domestic violence, used cocaine, or used heroin since the last visit. 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.