Algorithm of Ecocompensation in Sloping Land Conversion Program Based on Heckman’s Two-Step Model

Table 3

The estimated results of Heckman sample selection model.

	Probit regression			OLS regression
	Coefficient	Standard error	z-score	Coefficient	Standard error	t-value
Core explanatory variables
Risk preference
Planting risk	0.096	0.087	1.11	15.251	5.168	2.95
Investment risk	0.214	0.108	1.97	20.489	5.769	3.55
Social capital
Social prestige	1.526	0.543	2.81	30.771	19.256	1.60
Social network	−0.055	0.061	−0.91	1.583	3.479	0.45
Social participation	0.139	0.092	1.51	9.837	4.914	2.00
Cognitive preference
Perception of the environment	0.071	0.106	0.67	16.958	5.982	2.83
Perception of income	−0.015	0.082	−0.18	8.040	4.824	1.67
Perception of property rights	0.075	0.089	0.84	29.016	4.304	6.74
Land plot characteristics
Slope	0.037	0.076	0.49	−10.550	4.795	−2.20
Control variables
Low-income households	−0.150	0.191	−0.79	−4.367	12.808	−0.34
Age of the head of household	−0.043	0.013	−3.27	−4.628	0.668	−6.93
Health of the head of household	−0.246	0.117	−2.09	−6.780	8.182	−0.83
Education of the head of household	−0.085	0.123	−0.69	−3.217	7.043	−0.46
Political status of the head of household	0.767	0.296	−2.59	−6.502	18.263	−0.36
Household labour	0.035	0.082	0.43	−4.351	4.978	−0.87
Proportion of SLCP income	0.083	0.097	0.86	0.856	4.789	0.18
Area of arable land	−0.015	0.007	−2.25	0.800	0.343	2.34
Area of converted land	0.002	0.023	0.09	0.932	1.005	0.93
Identification variables
Proportion of SLCP peasant households	31.794	4.832	6.58
Inverse Mills ratio ()				35.716	21.327	1.67
Constant variable	−10.625	2.368	−4.49	321.800	82.631	3.89

, , and represent the estimated coefficient is statistically significant at the 1% level, at the 5% level, and at the 10% level, respectively.