ABSTRACT
The depopulation process of rural areas in Europe during the last decades is widespread. However, in many regions, this performance is largely an inheritance from past migration and the drag it causes in terms of an ageing population structure. Removing the inertial effects of past migration, we may find a much greater diversity in the current demographic dynamics of rural areas. We analyse the case of Galicia, a NUTS-2 region in Spain that represents a paradigmatic case of traditional emigration. We remove the drag from past migration to obtain the population performance of rural counties once the effects of that inertia were removed, and interpret the results. The analysis works as a descriptive tool: refined data help to identify which rural areas, however they continue to lose population, are being able to slow down or even revert the processes of depopulation inherited from the past. Finally, we explore possible explanations for such better performance, verifying that this is mainly related to key factors in regional planning and development such as the proximity to urban nodes and the quality of connections.
Acknowledgements
The authors wish to thank suggestions from Jose M. Andrade (Fundación Juana de Vega), Rubén Lado-Sestayo, Paulino Montes-Solla and Jorge Rodríguez-Álvarez (UDC), and the two anonymous reviewers. IBM SPSS Statistics version 21 and R Project (Packages rriskDistributions and zipfR) were used for statistical analysis.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. In the absence of a literal translation, we use the name ‘county’ to refer to Galician ‘comarca’ – though both words may refer to different administrative realities in different countries. Thus, the lowest administrative level in Spain are councils, headed by mayors elected every four years, while in Galicia different councils are grouped in 53 areas known as ‘comarcas’ (‘shires’).
2. Positive (negative) values of EMIG.50-91 mean a positive (negative) population growth from 1950 to 1991 – which, to some extent, is indicative of migration inflows (outflows). Correlation signs in Table A2 are coherent, as they imply that the higher the inflows, the lower the average age (d1), the percentage of population aged 65 years and more (d3), the elderly dependency (d4) and ‘85+ over 65+’ (d5) ratios, the demographic burden (d6), and the labour force structure (d7) and replacement (d8) indicators, while the percentage of young population (d2) would be higher. The only low correlation, with indicator ‘85+ over 65+’, is coherent as well. Indeed, indicator d5 uses data in year 1991, but people aged above 85 years in 1991 would be above 45 years in 1950. Hence, higher levels for indicator d5 would be related to migration processes before 1950.
3. We also checked that similar regressions for the subperiods 1991–2001 and 2001–2011 yield similar results. Results available upon request to the authors.
4. Residuals are not heteroscedastic nor autocorrelated, and are normally distributed. Results are provided in the SM.