Table 2.
Testing for convergence and convergence clubs.
| Panel I. Testing for Convergence | ||||||
|---|---|---|---|---|---|---|
| Provinces | PS | |||||
| Full sample | −0.809 (−218.01) |
|||||
| Panel II. Convergence Clubs | ||||||
| Panel A. Initial Estimation | Panel B. Adjacent Analysis | Panel C. Final Estimation | ||||
| Initial Clubs | Provinces | PS | Merging | PS | Final Clubs | Provinces |
| C1 | Asturias, Córdoba, Granada, Huelva, Jaén, Sevilla, Ceuta | 0.211 (10.56) |
C1 + C2 | −0.46 (−412.57) |
Club 1 | Asturias, Córdoba, Granada, Huelva, Jaén, Sevilla, Ceuta |
| C2 | Alicante, Almería, Badajoz, Burgos, Cádiz, Cantabria, Castellón, Cuenca, Gerona, Guipúzcoa, Huesca, León, Murcia, Pontevedra, La Rioja, Tarragona, Teruel, Valencia, Valladolid, Zamora, Melilla | 0.019 (1.30) |
C2 + C3 | −0.1173 (−8.99) |
Club 2 | Alicante, Almería, Badajoz, Burgos, Cádiz, Cantabria, Castellón, Cuenca, Gerona, Guipúzcoa, Huesca, León, Murcia, Pontevedra, La Rioja, Tarragona, Teruel, Valencia, Valladolid, Zamora, Melilla |
| C3 | Barcelona, Cáceres, La Coruña, Lérida, Lugo, Málaga, Navarra, Orense, Palencia, Santa Cruz de Tenerife, Zaragoza | 0.173 (11.86) |
C3 + C4 | 0.023 (1.63) |
Club 3 | Álava, Albacete, Ávila, Barcelona, Cáceres, La Coruña, Lérida, Lugo, Málaga, Navarra, Orense, Palencia, Salamanca, Santa Cruz de Tenerife, Soria, Toledo, Vizcaya, Zaragoza |
| C4 | Álava, Albacete, Ávila, Salamanca, Soria, Toledo, Vizcaya | 0.261 (17.40) |
C4 + C5 | 0.145 (2.58) |
Club 4 | Islas Baleares, Ciudad Real, Guadalajara, Las Palmas, Segovia |
| C5 | Islas Baleares, Las Palmas, Segovia | 0.613 (2.77) |
C5 + C6 | 0.445 (14.15) |
Divergent | Madrid |
| C6 | Ciudad Real, Guadalajara | 0.327 (23.75) |
C6 + Divergent | −1.819 (−40.13) |
||
| Divergent | Madrid | |||||
This table presents the results of the PS methodology. Panel I includes the analysis of the null hypothesis of convergence. The value in parentheses is the log-t ratio, and the value above it corresponds to the estimation of the parameter β in (3). The null hypothesis is rejected if the log-t ratio is lower than −1.65. For the sake of clarity, Spanish names are used for all provinces. Panel II presents the results of applying the PS clustering algorithm. Panel A shows the initial results, Panel B presents the merging analyses of the adjacent clubs, while Panel C shows the final results. The “PS” column values are the results of the estimation of Equation (3) for the different combinations of provinces, with the values in parentheses reflecting the log-ratios and the values above them to estimate the parameter β in (3).