Analysis of Economic Effect and Mechanism of Basic Medical Insurance for Urban Employees in China
Table 4
Empirical results of Heckman sample selection model.
Variable
Logarithm of total medical expenses
Logarithm of self-paid medical expenses
(1)
(2)
(3)
(4)
Selection equation
Expenditure equation
Selection equation
Expenditure equation
Interaction effect
−0.0209
1.206
−0.0209
0.745
(Treat × time)
[0.0206]
[0.6120]
[0.0206]
[0.5823]
Logarithm of income
0.00143
−0.252
0.00143
−0.228
[0.0053]
[0.1296]
[0.0053]
[0.1234]
Previous medical history
0.0373
−0.0759
0.0373
0.0961
[0.0126]
[0.4803]
[0.0126]
[0.4622]
Illness or injury
0.255
−5.599
0.255
−4.275
[0.0143]
[4.3940]
[0.0143]
[4.2515]
Any other relatives
−0.0326
−0.0326
[0.0259]
[0.0259]
Inverse Mills ratio
−15.51
−12.08
[10.9397]
[10.5715]
Constant
19.29
16.15
[9.9237]
[9.6066]
Provinces
Control
Control
Control
Control
Sample size
1836
213
1836
213
Note:the selection equations of models (1) and (3) are estimated by probit model, which gives marginal effect. The standard errors are shown in brackets, and the standard errors of models (2) and (4) are obtained by bootstrap method. The model also controls the variables of age, gender, nationality, marital status, education level, occupation, permanent address, smoking, and drinking. ,, and .
