Skip to main content

Table 2 Estimates for two-level generalized linear models of Marketplace enrollment

From: Variability of assister availability in health insurance marketplace in the U.S.

 

Model 1

Model 2

Model 3a

β

S.E.

β

S.E.

β

S.E.

Fixed effects

Intercept

0.23

0.17

0.68*

0.39

0.79

0.50

County Rurality

 Metropolitan

  

0.71***

0.13

0.75***

0.15

 Micropolitan

  

0.34**

0.13

0.31*

0.16

 Rural (reference)

  

–

–

–

–

% uninsured among adults

  

0.17***

0.06

0.16**

0.06

% uninsured among people with below or at 138% FPL

  

−0.25***

0.03

−0.25***

0.03

% uninsured among people with 138–400% FPL

  

0.52***

0.06

0.48**

0.07

% uninsured among people with 400% FPL or higher

  

−0.75***

0.09

−0.62***

0.10

% uninsured among female

  

−0.11

0.08

−0.11

0.09

Medicaid expansion

    

−0.08

0.80

Medicaid * 138% FPL

    

−0.09

0.06

Medicaid * 138–400% FPL

    

0.39**

0.16

Medicaid * 400% FPL

    

−0.73***

0.22

Medicaid * metropolitan

    

−0.25

0.28

Medicaid * micropolitan

    

0.05

0.29

-2LL

2899.71

2646.81

2629.37

Pearson Chi-square/DF

0.97

0.99

0.99

Level 2 intercept (covariance parameter)

0.87***

0.28

0.55***

0.18

0.53***

0.18

ICC

0.21

0.14***

0.14***

  1. Note. *: p<0.01; **:p<0.05; ***:p<0.001= likelihood ratio test significant; ICC = 0.21; values based on SAS PROC GLIMMIX. References of rurality and Medicaid expansion are rural area and no expansion, respectively. % potential enrollees, % uninsured, and % female are the share of non-elderly individuals with 138 to 400% FPL, uninsured, female among the non-elderly individuals, respectively
  2. aBest fitting model