Table 2.
Estimated impacts of the Grassland Ecological Compensation Program (GECP) on herder income: household level, FE model.
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| Panel A. Overall impacts on income | ||||
| Dependent variable | Household income per capita | Net pastoral income per capita | Non-pastoral income per capita | Non-GECP income per capita |
| Annual GECP payment (yuan) | 0.366*** | 0.132 | 0.048 | 0.144 |
| (0.004) | (0.369) | (0.366) | (0.332) | |
| Control variables | Yes | Yes | Yes | Yes |
| Household fixed effects | Yes | Yes | Yes | Yes |
| Year fixed effects | Yes | Yes | Yes | Yes |
| No. of observations | 3,474 | 3,474 | 3,474 | 3,474 |
| R2 | 0.731 | 0.749 | 0.863 | 0.744 |
| Panel B. Heterogeneous impacts by grassland types | ||||
| Grassland | 0.405* | 0.221 | 0.025 | 0.179 |
| (0.051) | (0.347) | (0.746) | (0.445) | |
| Meadow | 0.232* | −0.243 | 0.122** | −0.043 |
| (0.057) | (0.131) | (0.030) | (0.777) | |
| Desert | 0.274 | 0.309 | 0.014 | 0.300 |
| (0.414) | (0.334) | (0.873) | (0.345) | |
| Panel C. Heterogeneous impacts by socioeconomic variables | ||||
| Education of the labors (years) | −0.004 | −0.002 | 0.020*** | −0.001 |
| (0.183) | (0.473) | (0.000) | (0.852) | |
| Distance to the closest township-level road (km) | −0.0002 | −0.0000 | −0.001** | −0.0001 |
| (0.654) | (0.978) | (0.011) | (0.742) | |
| Grassland area per capita (hundred ha in log) | 0.009 | 0.016 | −0.035*** | 0.010 |
| (0.438) | (0.344) | (0.000) | (0.467) | |
Note. Panel A provides the estimated overall impacts on income from the fixed-effect (FE) model using household-level data (Eq. (3)). All outcome variables and the key explanatory variable, annual GECP payment, are transformed using an inverse hyperbolic sine transformation to avoid taking logarithm of zero, following ln(y+(y2+1)1/2). Household income includes net pastoral income, non-pastoral income, and GECP payment. Both household and year-fixed effects are controlled for. Household-, village-, and township-level time-variant variables are also controlled for. Household-level controls include the quantity of labor used in raising livestock, operated farm size, share of the joint operated area, the total number of different plots, a dummy variable for grassland harvesting, and a dummy variable for planting crop/fodder. Village-level controls include an indicator of whether a village has local grassroots measures in place to limit grazing intensity, an indicator whether a village has a formal government-run monitoring system, and climate variables (the cumulative rainfall and the mean temperature from May to October in each year). Township-level controls include farm-gate livestock prices, hay prices, wages for non-pastoral employment, and grassland rental prices. Standard errors are clustered at the village-by-year level. The results are consistent between an inverse hyperbolic sine transformation and a log transformation of the dependent variable in each column. The log transformation results are available upon request. Panel B provides the heterogeneous impacts on income by grassland type. We use the grassland type with the largest area in a county as its major grassland type. Each set of coefficients and corresponding p-values are from one single regression model. All model specifications are the same as in Panel A. Panel C provides the heterogeneous impacts on income by different socioeconomic variables. A two-sided t test is performed for each coefficient. The exact p-values are in parentheses. *p < 0.10, **p < 0.05, ***p < 0.01.