Table 5.

Estimated stochastic frontiers and technical (In)efficiency effects (dependent variable—healthcare expenditure)

Variable	Cobb-Douglas		Translog
	Coefficient	t statistics	Coefficient	t statistics
Frontier model
Constant	4.195***	(166.57)	0.901	(0.40)
ln(Patient)	0.951***	(359.16)	1.454***	(3.84)
Comp air pollutants index	0.000664***	(6.40)	0.00119	(0.97)
T	0.000091***	(17.48)	0.00042	(0.89)
$\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.ln{\left(\text{Patient}\right)}^2$	–	–	− 0.0371	(− 1.14)
$\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$2$}\right.{\left(\text{Comp}\right)}^2$	–	–	− 0.000022***	(− 4.11)
1/2 T2	–	–	0.0000015***	(18.74)
ln(Patient)*(Comp)	–	–	0.0000627	(0.48)
ln(Patient)*T	–	–	− 0.0000883*	(− 2.24)
(Comp)*T	–	–	0.0000001	(0.41)
D1 Spring	0.0208***	(5.77)	0.00167	(0.23)
D2 Summer	0.0351***	(10.29)	0.0174***	(3.43)
D3 Autumn	0.0416***	(10.19)	0.0264***	(6.89)
Policy factors of inefficiency
Constant	–	–	− 1.544	(− 1.24)
ln(Patient)	–	–	0.455***	(6.02)
Ln(CARS)	–	–	− 0.157*	(− 2.40)
Ln(INDU)	–	–	0.00927	(1.73)
ELDER	–	–	− 0.299	(− 1.12)
T	–	–	0.00447***	(4.02)
Non-policy factors of inefficiency
TEMP	–	–	− 0.0242*	(− 2.31)
WIND	–	–	0.0518*	(2.21)
RAIN	–	–	− 0.0145	(− 1.39)
Adj R2	0.9825	–	0.9913	–
Observations	2400	–	2400	–

t statistics in parentheses

^*p < 0.05

^**p < 0.01

^***p < 0.001