Table 2.

Association of Medicaid enrollment at T1 with substance use disorder treatment between T1 and T2 in participants of National Epidemiologic Survey on Alcohol and Related Conditions.^a

	Bivariate probit model			Naïve probit model
	B	SE	p	B	SE	p
Health insurance (Ref: uninsured)
Medicaid	.200	.638	.755	.246	.121	.046
Medicare	.044	.106	.680	.044	.106	.680
Private insurance	−.201	.061	.002	−.201	.061	.002
Other insurance	.163	.133	.227	.163	.133	.228
Female sex	−.180	.052	.001	−.180	.052	.001
Age, years (Ref:18–29)
30–39	−.023	.075	.761	−.023	.075	.761
40–54	.020	.069	.776	.020	.069	.776
55+	−.555	.108	<.001	−.555	.108	<.001
Race/ethnicity (Ref.: Non-Hispanic white)
Non-Hispanic black	−.050	.089	.575	−.050	.089	.575
Hispanic	−.038	.085	.654	−.038	.085	.654
Other	−.126	.157	.423	−.126	.157	.423
Personal income in $1,000 (Ref: <20)
20-<35	−.113	.071	.120	−.113	.071	.120
35-<60	−.290	.087	.001	−.290	.087	.001
60+	−.361	.101	.001	−.361	.101	.001
SUD in past year (at T1)	.256	.052	<.001	.256	.052	<.001
Constant	−1.266			−1.268
Model predicting individual Medicaid enrollment
Instrumental variables
% of low-income covered by Medicaid	.027	.010	.009
% of low-income covered by Medicaid among those with Medicaid or uninsured	.004	.007	.586
Constant	−2.672
rho	.021	.282	.941

Abbreviations: T1 represents baseline interview and T2, the follow-up interview. B stands for the regression coefficient, SE, for standard error, and SSI for supplemental security income.

^aThe models also adjusted for the state fixed effects. The full analyses results are presented in Appendix Table B.