Skip to main content

Table 5 Addressing Endogeneity and Omitted Variables Bias, Multinomial Logistic Regression Analysis

From: Effects of health insurance on patient demand for physician services

  Routine
Visits Only a
Emergency
Visits Only b
Mixture
of Visits c
Routine
Visits Only a
Emergency
Visits Only b
Mixture
of Visits c
Coefficient (1) (2) (3) (4) (5) (6)
  A. Acute Event Occurred Within One Year of the Survey (N = 1274) B. Regression Discontinuity at Medicare Eligibility Threshold (N = 2203)
Private insurance 12.41**
(6.03)
−3.98***
(1.45)
−1.00
(5.63)
   
Public insurance 2.66
(6.31)
−2.85**
(1.46)
13.76***
(5.51)
   
RD Term (≥ 65)     37.96*
(20.73)
−61.43**
(27.61)
6.44
(7.52)
  C. Managed Care Effects (N = 36,424) D. Detailed Coverage Type (N = 36,424)
Private insurance 24.24***
(1.13)
−2.16***
(0.29)
1.33*
(0.75)
   
Public insurance 17.03***
(1.50)
0.20
(0.30)
8.17***
(0.81)
   
Private insurance × managed care 0.28
(0.80)
0.11
(0.29)
−0.20
(0.48)
   
Public insurance × managed care 2.20
(1.64)
0.12
(0.35)
0.96
(0.69)
   
Job-Based Group Insurance     22.22***
(0.94)
−2.16***
(0.26)
−0.24
(0.55)
Other Group Insurance     17.24***
(4.55)
−1.97
(2.38)
−1.76
(2.55)
Individual
(Non-Group)
    14.17***
(1.21)
−1.48
(0.33)
0.91
(0.67)
Medicaid     12.56***
(1.08)
0.48**
(0.24)
7.89***
(0.54)
Tricare/CHAMPVA     1.01
(2.22)
1.05*
(0.60)
2.58**
(1.22)
Other Public     14.49***
(5.21)
−5.26**
(2.39)
−0.30
(2.49)
Uninsured 1 1 1 1 1 1
  1. Notes: The multinomial logistic regression coefficients have been converted into marginal effects to improve interpretability. Standard deviations are reported in parentheses. All regressions include individual controls and time trends. The reference category is uninsured respondents
  2. a Routine visits only, 2+ routine office based visits only
  3. b Emergency visits only, 2+ emergency room visits only
  4. c Mixture of visits, 2+ routine and emergency room only
  5. *Significant at 10% confidence level
  6. **Significant at 5% confidence level
  7. ***Significant at 1% confidence level
\