Table 3.

Random-effects estimation of policy attributes, plot features, and household characteristics on PES enrollment

Independent variables	Coefficientsa (SE)	Marginal effects
Payment (in 1000 yuan)	0.421*** (0.008)	0.006
Duration	0.151*** (0.006)	0.002
Wildlife raid rate	3.074*** (0.154)	0.047
Wildlife raid rate * GTGP participation	− 1.105** (0.407)	− 0.017
Area	1.481*** (0.450)	0.022
Walking distance	0.086*** (0.003)	0.001
Elevation	0.127*** (0.033)	0.002
Slope	0.090*** (0.008)	0.001
Aspect	0.001 (0.001)	0.00002
GTGP participation	0.714*** (0.099)	0.021
Household size	− 0.244*** (0.024)	− 0.004
Farming income	− 0.026*** (0.006)	− 0.0004
Off-farm income	0.014*** (0.001)	0.0002
Cropland	− 1.025*** (0.185)	− 0.016
Age	0.060*** (0.004)	0.001
Education	0.065*** (0.014)	0.001
Constant	− 10.998*** (0.354)
${\sigma }_{\mu }$ b	3.387*** (0.034)
$\rho$ b	0.920*** (0.001)
χ 2 c	20.363***
Observations	2907
Number of plots	969

^aSignificance: ** $p\le$ 0.01; *** $p\le$ 0.001

^bThe random-effects probit parameters for ${\sigma }_{\mu }$ and $\rho$ were statistically significant, suggesting that the random-effects model is appropriate

^cThe test statistic for a ${\chi }^2$ test of the random-effects model versus the pooled model is 19456 with one degree of freedom, which is statistically significant (p value < 0.001)