Skip to main content
. 2022 Jul 6;23(1):1–26. doi: 10.1007/s10754-022-09336-7

Table 3.

Estimation Results – Healthcare Utilization (Realized Access)

No fixed effects, no chronic condition (1) (2) (3) (4)
Ln(HC Exp) Share Self pay Office visits ER visits
Union Member

0.469***

(0.0211)

− 0.113***

(0.00233)

0.271***

(0.0179)

0.168***

(0.0252)

N 256,310 202,827 256,310 256,310
Controls Y Y Y Y
Chronic Condition N N N N
Year FE Y Y Y Y
Individual FE N N N N
No fixed effects, chronic condition (1) (2) (3) (4)
Ln(HC Exp) Share Self pay Office visits ER visits
Union Member

0.450***

(0.0295)

− 0.103***

(0.00322)

0.268***

(0.0251)

0.142***

(0.0373)

N 139,349 108,756 139,349 139,349
Controls Y Y Y Y
Chronic Condition Y Y Y Y
Year FE Y Y Y Y
Individual FE N N N N
Fixed effects, no chronic condition (1) (2) (3) (4)
Ln(HC Exp) Share Self pay Office visits ER visits
Union Member

0.131*

(0.0792)

− 0.0473***

(0.0122)

0.104**

(0.0520)

0.312***

(0.0956)

N 256,310 202,827 256,310 256,310
Controls Y Y Y Y
Chronic Condition N N N N
Year FE Y Y Y Y
Individual FE Y Y Y Y
Fixed effects, chronic condition (1) (2) (3) (4)
Ln(HC Exp) Share Self pay Office visits ER visits
Union Member

0.153

(0.119)

− 0.0422**

(0.0196)

0.132

(0.0804)

0.486***

(0.134)

N 139,349 108,756 139,349 139,349
Controls Y Y Y Y
Chronic Condition Y Y Y Y
Year FE Y Y Y Y
Individual FE Y Y Y Y

Table presents regression results from estimating Eq. (1). Estimates for continuous variables obtained using ordinary least squares. Estimates for count variables obtained from either negative binomial regression or Poisson pseudo-maximum likelihood (when individual fixed-effects are included). Standard errors in parenthesis, clustered at the individual level.

*p < 0.10, **p < 0.05, ***p < 0.01. Observations weighted using IPUMS-MEPS sample weights. Each regression includes controls for age, age-squared, sex, race, educational attainment, marital status, income, and year-fixed effects. Table reports regression coefficients, such that—for count data models—one can obtain the incidence rate ratio by exponentiating the parameter estimate