**5. Discussion**

The research results confirm the article's hypothesis: During the post-crisis economic recovery period, the spatial distribution of the population in Spanish urban areas was reshaped. Population growth patterns were more diverse than they were during the economic crisis years. Indeed, when calculating population growth rates for city cores and peripheries and for each of the urban areas as a whole, significant differences appear not only among them but also within them. Cores and peripheries show different behaviors in diverse urban areas, ring municipalities differ from one another, and metropolises vary among each other. Regarding this last point, the results show that from 2011 to 2015, most of the 69 examined FUAs lost population, whereas from 2015 to 2019, population growth was positive in most of them. Although both phases showed an enormous variety of situations in which urban cores and peripheries gained and lost population, it can be concluded that this heterogeneity increased in the second period under analysis. For example, some of the urban areas that always either gained or lost population experienced new realities, such as their cores growing more than their peripheries. Large cities such as Madrid and Bilbao, for instance, increased more than their rings. The post-crisis reactivation of foreign immigration (in 2018, there were 760,804 entries from abroad, very similar to the highest figures registered in first decade of the 21st century) could be behind this renewed attractiveness of core cities.

As previously mentioned, there seem to be no obvious geographical patterns. Perhaps the FUAs that have continuously lost population since 2011 show the clearest one. These are mostly either inland provincial capitals or cities located in declining industrial areas. Burgos, Cuenca, Jaén, León, Ourense, Palencia, Salamanca, Valladolid, and Zamora are the first type, and Alcoy, Avilés, Cádiz, Ferrol, Ponferrada, and Talavera are the second. Both groups of cities share a low capacity to attract international immigration. At the other end of the spectrum, FUAs that gained population throughout both analyzed phases seem to have a greater diversity of features: They are predominantly located on the coast or near a large Spanish metropolitan area and mainly have a diversified or an industrially based economy. This is the case for Almería, Ciudad Real, Donostia, Girona, Granada, Guadalajara, Huelva, Lleida, Málaga, Marbella, Murcia, Pamplona, Seville, and Toledo.

This increasing variety of situations (or heterogeneity of population dynamics) among urban areas is also verifiable within the five large metropolises analyzed, which addresses the second objective of the article. In other words, both city cores and ring municipalities show examples of population growth and decline. Although suburbanization was reactivated in the post-crisis period, urban core attractiveness increased during those years, as, except for Seville, their population grew. Many peripheral municipalities also experienced accelerated population growth during post-crisis years, although some of them (still) lost inhabitants. These uneven population dynamics act similarly to what, we suppose, occurs inside large cities among neighborhoods. Therefore, further analysis focusing on census tracts and additional research on migratory flows, as well as natural growth, will provide answers to new questions raised by the present study. Nevertheless, what this paper does show is that the stages defined by the urban development model are no longer useful for analyzing and interpreting present urban dynamics in Spanish FUAs.

In any case, urban core and periphery growth between 2015 and 2019 seems to have been intimately related to international migration and internal residential mobility increases in an improving economic context. Indeed, in Spain, the economic crisis was extremely intense, and residential insecurity rose more than in other European countries, as vulnerable groups were overrepresented [96]. The recession entailed an interruption of international immigration entries and a modification of internal flows, though these did not decrease as much as expected in metropolitan areas. Furthermore, after the real estate bubble burst, there was a reduction in the vast suburbanization flows that characterized previous decades [11,97,98] not only in large urban areas but also in intermediate-size ones [15]. This decrease in decentralization was even interpreted by some authors as evidence of an emerging urban population recentralization process [6,99]. However, as residential mobility patterns vary with economic cycles, these rates started to grow in 2015, reflecting diversification of socio-residential strategies and the transformation of interactions between intra-metropolitan spaces [20,100]. This recent enhancement in residential mobility and its increasingly complex patterns of origin, destination, age, socioeconomic category, and life stage, among others, has had pronounced demographic and socioeconomic impacts. Moreover, it has had spatial effects on urban areas, as many suburban municipalities grew again at the same time as their cores.

Though it is difficult to establish cause–effect relationships with this largely descriptive research, Spain's rising residential mobility during the post-crisis period runs parallel to several socioeconomic processes mentioned in the theoretical framework chapter: housing and labor market modifications, family and household transformations, and spatial socioresidential pattern changes. In our view, all of these processes would lead to growing and more diverse residential mobility.

Housing market transformations, for instance, would have intensified multiple flows between (and within) cores and rings through diverse mechanisms, such as gentrification, touristification, and in general, problems faced by young people and immigrants in affording a dwelling. On the other hand, the labor market also recovered between 2014 and early 2020, creating thousands of jobs that, once again, attracted foreign immigrants and reactivated internal and intra-metropolitan flows. Population and household transformations, which are together known as the Second Demographic Transition, are apparently less related to economic circumstances. Single and divorced people seem to prefer living in central areas, whereas family households tend to live in peripheral regions during both expansive and recessive economic phases. However, the post-crisis era may have accelerated life cycle transitions: For example, having a child or divorcing is expensive, and in consequence, so is residential mobility. Therefore, during recent recovery years, all of these family transformations would have also had an impact on urban demographic and spatial structures.

Finally, was the suburbanization of poverty occurring in the post-crisis period? Maintaining that poverty was expelled to the peripheries of Spanish cities can be considered an oversimplification. However, a global review of recent theoretical and empirical con-

tributions shows that residential mobility does play an essential but extremely complex role in the social reconfiguration of metropolitan spatial interactions [87]. In the absence of more detailed studies, we believe that rather than a simplistic dichotomy between affluent cores and poor rings, residential mobility generates new "fragmented" spaces in a similar sense to that applied by Borsdorf and Hidalgo for Latin American urban areas [101], although with significant differences. Indeed, Latin American cities had a historical urban structure characterized by affluent people living in city cores and poor residents inhabiting peripheries [102]. As gated condominiums for the middle and upper classes have been built in peripheries, these privileged social groups are partly suburbanizing in those countries [103]. However, the situation is different in Europe and the U.S., where, as previously noted, middle to high classes are moving downtown. Nevertheless, from a spatial socio-demographic point of view, the results obtained are similar: Both cores and peripheries are increasingly complex, with affluent, middle-class, and deprived groups, as well as aged and rejuvenated populations, living in adjacent neighborhoods (within the same city) or nearby residential developments (within the same suburban municipalities). As city cores and rings are progressively more mixed and diversified, we can maintain that urban areas are increasingly demographically complex, diversified, or "fragmented." This paper's results for Spanish urban areas during recovery years are in line with this urban tendency.

Of course, these research results reflect a post-crisis reality that was abruptly interrupted by the COVID-19 pandemic and the state of alarm in mid-March 2020 (Royal Decree 463/2020, issued on 14 March 2020). Mandatory confinement decreed by the government to stop the spread of the virus had a deep economic and social impact that will modify, at least in the short and medium term, the described urban population dynamics. The closure of Spain's and other countries' borders, breaking global circulation, has also had a great impact on the main component of metropolitan population growth: migration. This prevents new entries while fixing the population to the territory. Therefore, we are facing a new phase, and its uncertain future developments, depending on how the pandemic continues to unfold, will require new research.

**Author Contributions:** Conceptualization, F.G.-A., C.L.-V., J.B.-i.-C., and I.P.; data curation, J.B.-i.-C.; formal analysis, F.G.-A. and J.B.-i.-C.; funding acquisition, C.L.-V. and F.G.-A.; investigation, F.G.-A., C.L.-V., J.B.-i.-C., and I.P.; methodology, F.G.-A., J.B.-i.-C., and I.P.; project administration, C.L.-V.; supervision, F.G.-A. and I.P.; writing—original draft, F.G.-A. All authors have read and agreed to the published version of the manuscript.

**Funding:** Research for the present paper was conducted within the following I + D + i project: Nuevas movilidades y reconfiguración sociorresidencial en la poscrisis: consecuencias socioeconómicas y demográficas en las áreas urbanas españolas (New mobilities and socio-residential reconfiguration in the post-crisis: socio-economic and demographic consequences in Spanish urban areas, ref. RTI2018- 095667-B-I00), directed by Cristina López Villanueva and Fernando Gil Alonso. It was funded by the Ministry of Science and Innovation, the Spanish Research Agency, and the European Regional Development Fund (AEI/ERDF, EU). Some of the authors are members of the Territori, Població i Ciutadania (Territory, Population, and Citizenship) consolidated research group, recognized by the Catalan government (ref. GRC\_2017SGR1298) and coordinated by Fernando Gil Alonso.

**Data Availability Statement:** Publicly available datasets were analyzed in this study. This data can be found here: https://www.ine.es/.

**Acknowledgments:** A very preliminary version of this research was presented at the XV Coloquio de Geografía Urbana (2020). The authors are grateful for the comments received from the organizers and participants, but also from the *Urban Science* editorial team and the three reviewers, all of them contributing to improvements in the text.

**Conflicts of Interest:** The authors declare no conflict of interest.
