Table 3.
Estimation results of Spatial Durbin Model (SDM) under different weight matrices.
| Weight matrix | Spatial adjacency matrix | Geographic distance matrix | ||
|---|---|---|---|---|
| Econometric model | FE | RE | FE | RE |
| RTI | −0.055*** (0.014) | −0.047** (0.020) | −0.038*** (0.010) | −0.032*** (0.008) |
| PS | −0.021* (0.013) | −0.023* (0.014) | −0.019** (0.008) | −0.015* (0.009) |
| RH | 0.142** (0.058) | 0.116*** (0.031) | 0.127** (0.052) | 0.082** (0.033) |
| PEF | −0.230 (0.197) | −0.192 (0.166) | 0.207 (0.375) | 0.126* (0.077) |
| RMI | 0.107*** (0.028) | 0.098** (0.040) | 0.092** (0.037) | 0.084* (0.050) |
| RLG | 0.034** (0.013) | 0.045** (0.019) | 0.040*** (0.011) | 0.019** (0.008) |
| RPC | 0.056** (0.024) | 0.072* (0.043) | 0.063** (0.026) | 0.081* (0.049) |
| W*RTI | 0.023*** (0.006) | 0.029** (0.012) | 0.025*** (0.007) | 0.020** (0.008) |
| W*PS | −0.029** (0.012) | −0.024** (0.010) | −0.031** (0.013) | −0.028* (0.007) |
| W*RH | 0.218** (0.089) | 0.206*** (0.054) | −0.178** (0.072) | −0.193*** (0.051) |
| W*PEF | −0.192*** (0.051) | −0.138** (0.057) | 0.154 (0.132) | 0.0137 (0.116) |
| W*RMI | 0.091* (0.056) | 0.057** (0.023) | 0.072** (0.030) | 0.065*** (0.017) |
| W*RLG | −0.045** (0.018) | −0.063*** (0.016) | 0.058** (0.024) | 0.074** (0.030) |
| W*RPC | −0.036 (0.044) | −0.031 (0.039) | 0.047 (0.058) | 0.039 (0.052) |
| ρ | −0.493*** (0.127) | −0.451*** (0.118) | −0.624*** (0.163) | −0.572*** (0.151) |
| Sigma2 | 0.004*** (0.001) | 0.001*** (0.000) | 0.003*** (0.001) | 0.002*** (0.000) |
| R2 | 0.893 | 0.824 | 0.925 | 0.906 |
| Log-likelihood | 385.072 | 406.238 | 431.752 | 415.028 |
| Wald-err | 46.58*** | 42.95*** | 50.31*** | 47.62*** |
| Wald-lag | 81.37*** | 73.52*** | 92.64*** | 79.48*** |
| LR-err | 54.19*** | 48.71*** | 60.28*** | 53.93*** |
| LR-lag | 62.64*** | 54.06*** | 70.59*** | 65.87*** |
| Hausman test | 22.75 | P = 0.023 | 23.72 | P = 0.000 |
represent the significance level of 10, 5, and 1%, respectively, and the data in brackets is standard deviation.