Results of conditional β-convergence modelling in a spatial approach with the use of an SDM form_
| Weights matrix → | W1 (contiguity) | W2 (distance) | W3 (flow) | W4 (block) | ||||
|---|---|---|---|---|---|---|---|---|
| Variable ↓ | Coeff. | Prob. | Coeff. | Prob. | Coeff. | Prob. | Coeff. | Prob. |
| LAG_LN_Y | 0.500 | 0.0000 | −0.903 | 0.0000 | 0.278 | 0.0035 | 0.651 | 0.0000 |
| CONSTANT | 0.826 | 0.0001 | 1.930 | 0.0004 | 1.877 | 0.0000 | 0.562 | 0.0086 |
| LN_EU_FUNDS | −0.006 | 0.3050 | 0.003 | 0.5801 | 0.007 | 0.2838 | 0.006 | 0.1957 |
| LN_GDPpc07 | −0.006 | 0.6663 | −0.043 | 0.0048 | −0.028 | 0.0716 | −0.005 | 0.6844 |
| LN_INVEST | 0.126 | 0.0000 | 0.155 | 0.0000 | 0.128 | 0.0000 | 0.098 | 0.0000 |
| LN_LABOR | 0.606 | 0.0013 | 0.516 | 0.0153 | 0.495 | 0.0286 | 0.515 | 0.0018 |
| LN_INNOV | 0.102 | 0.0028 | 0.102 | 0.0054 | 0.065 | 0.1355 | 0.076 | 0.0505 |
| LAG_LN_GDPpc07 | −0.074 | 0.0003 | −0.170 | 0.0007 | −0.156 | 0.0000 | −0.049 | 0.0188 |
| LAG_LN_EU_FUNDS | 0.013 | 0.0696 | 0.071 | 0.0000 | 0.008 | 0.4339 | −0.002 | 0.8243 |
| LAG_LN_INVEST | 0.030 | 0.2533 | 0.539 | 0.0000 | 0.167 | 0.0000 | 0.018 | 0.5068 |
| LAG_LN_LABOR | −0.387 | 0.0699 | −0.099 | 0.9104 | 0.283 | 0.5247 | −0.618 | 0.0058 |
| LAG_LN_INNOV | −0.091 | 0.0520 | 0.200 | 0.1080 | −0.003 | 0.9701 | −0.079 | 0.0923 |
| Pseudo R2 | 0.828 | 0.781 | 0.795 | 0.865 | ||||
| Log likelihood | 367.079 | 348.556 | 357.995 | 394.95 | ||||
| AIC | −710.159 | −673.112 | −691.99 | −765.901 | ||||
Results of conditional β-convergence modelling in a spatial approach with the use of a SAR form_
| Weights matrix → | W1 (contiguity) | W2 (distance) | W3 (flow) | W4 (block) | ||||
|---|---|---|---|---|---|---|---|---|
| Variable ↓ | Coeff. | Prob. | Coeff. | Prob. | Coeff. | Prob. | Coeff. | Prob. |
| LAG_LN_Y | 0.495 | 0.0000 | 0.622 | 0.0000 | 0.573 | 0.0000 | 0.582 | 0.0000 |
| CONSTANT | 0.508 | 0.0009 | 0.804 | 0.0000 | 0.307 | 0.0331 | 0.286 | 0.0456 |
| LN_GDPpc07 | −0.049 | 0.0002 | −0.085 | 0.0000 | −0.033 | 0.0097 | −0.030 | 0.0154 |
| LN_EU_FUNDS | 0.009 | 0.0423 | 0.015 | 0.0024 | 0.011 | 0.0055 | 0.013 | 0.0008 |
| LN_INVEST | 0.162 | 0.0000 | 0.209 | 0.0000 | 0.114 | 0.0000 | 0.130 | 0.0000 |
| LN_LABOR | 0.400 | 0.0253 | 0.414 | 0.0424 | 0.522 | 0.0014 | 0.401 | 0.0155 |
| LN_INNOV | 0.076 | 0.0010 | 0.122 | 0.0003 | 0.044 | 0.0108 | 0.055 | 0.0499 |
| Pseudo R2 | 0.788 | 0.722 | 0.802 | 0.818 | ||||
| Log likelihood | 344.738 | 317.553 | 354.265 | 367.000 | ||||
| AIC | −675.475 | −621.107 | −694.53 | −720.001 | ||||
Results of conditional β-convergence modelling with the classic OLS method_
| Variable | Coefficient | Std. Error | t-Statistic | Probability |
|---|---|---|---|---|
| CONSTANT | 1.068 | 0.183 | 5.825 | 0.0000 |
| LN_GDPpc07 | −0.102 | 0.016 | −6.403 | 0.0000 |
| LN_EU_FUNDS | 0.014 | 0.005 | 2.675 | 0.0080 |
| LN_INVEST | 0.263 | 0.015 | 17.017 | 0.0000 |
| LN_LABOR | 0.483 | 0.221 | 2.179 | 0.0302 |
| LN_INNOV | 0.097 | 0.036 | 2.659 | 0.0083 |
| Regression diagnostics | ||||
| R2 = 0.681 | JB test = 56.92 ( p = 0.0000) | |||
| Log likelihood = 300.256 | BP test = 10.07 ( p = 0.0773) | |||
| AIC = -588.513 | KB test = 4.23 ( p = 0.5160) | |||
| Diagnostics for spatial dependence: | Moran’s I | Lagrange Multiplier tests | ||
| W1 matrix (contiguity) | 0.2740 (p = 0.0000) | LM(SAR) > LM(SEM) | RLM(SAR) > RLM(SEM) | |
| W2 matrix (distance) | 0.1201 (p = 0.0000) | LM(SAR) > LM(SEM) | RLM(SAR) > RLM(SEM) | |
| W3 matrix (flows) | 0.2557 (p = 0.0000) | LM(SAR) > LM(SEM) | RLM(SAR) > RLM(SEM) | |
| W4 matrix (block) | 0.3451 (p = 0.0000) | LM(SAR) > LM(SEM) | RLM(SAR) > RLM(SEM) | |