Surpassing the administrative division limits on regional analysisThree essays on urban and regional economics

Resumen

The concept of Region is one of the elements which has differentiated Regional Economics from other fields of Applied Economics. In spite of this, however, researchers in this field of economic analysis have not paid a great deal of attention to this concept. In this research we have proposed a concept of Region which goes beyond the administrative division of territory. Our regional aggregation has been based on agglomeration economies, one of the fundamental concepts in the fields of Economic Geography and Urban and Regional Economics. In accordance with the work of Polèse et al. (2007), the territory has been classified into analytical regions which take into account the size of the population and the distance from the main urban areas. In doing so, we achieve an aggregation which corresponds with the differences in agglomeration economies across space. However, their robustness in comparison with the administrative units commonly used has -to date - not been evaluated. The objective of the first chapter of this thesis was to prove that the functional regions defined under such economic criteria provide better defined regions -in terms of greater compactness and separation - than the administrative ones commonly used to carry out labour market studies at sub-national level. Using micro data from the last available Spanish Census, the functional and administrative regions are evaluated using the Theil index and the Davies-Bouldin Validation index. Applied to employment (by gender, industry and level of qualification and occupation), both indexes show better results for the analytical regions than for any of the ordinary administrative ones (NUTS I, II or III regions). In other words, the analytical classification generates areas where the distribution of employment is more homogeneous within and more heterogeneous between the regions. The following chapters have provided two applications of this analytical division of the territory to Labour Economics issues: the factors affecting the probability of being employed (Chapter 2) and the effects that labour mobility and commuting have on the central regions (Chapter 3). In the second chapter, we presented a spatial analysis of employment at local level where, among other factors, the demographic and geographical characteristics can and do affect the outcome. The empirical results support the hypothesis that size -in terms of population- and location -in terms of distance to a metropolis - are explanatory variables for the probability of being employed. In other words, employment depends not only on the personal characteristics of the individuals (level of education, age, sex, etc.) but also on the type of analytical region -as defined in Chapter 1- where they live. However, the natural behavior of the workers with problems of employability is moving from their regions to others more adequate to their characteristics. In the third chapter we complete the analysis studying these migrations but using our alternative functional regions Using the last available Census, the estimations for Spain of an input-output multi-regional model that includes the possibility of commuting show that the arrival of in- and im-migration to the core generates a set of effects induced by the redistribution of population among other regions. The arrival of workers from the periphery to the core provokes reallocations of residence in all cases (displacement effect). However, the intensity of these reallocations increases with size, which shows the existence of some agglomeration diseconomies associated with big cities. When the possibility of commuting is also considered, the arrival of workers from the periphery to the core generates the reallocation of both jobs (economic activity) and residences. The larger cities are the ones pushing out more residents to other areas, while keeping most of the jobs. In other words, they are becoming attractive areas to work in, but not to live in (due to, among other reasons, high housing costs, congestion or other negative externalities). The opposite is true for the smaller cities, which are attractive for residing in but for working in. The distributional pattern of residences proves to be different to the distributional pattern of jobs.