Internal migration of a highly skilled population is a key factor in a current knowledge-driven labour market. However, there are uncertainties regarding their migration patterns and how they differ from those of the general population. This paper utilises an innovative methodology based on network theory to identify functional regions (FRs) and describe internal migration using as a case study Spanish workers over the 2005–2023 period. The results show that migration trends differ between high-skilled labour and the general population. Moreover, the analysis of the main attractors shows a specialisation, with Barcelona and Madrid focusing mainly on educated workers. We observe variations in the attractive power related to changes in macroeconomic variables, although the time series in the Barcelona province from 2018 shows a discrepancy, with migration from trained workers decreasing, although economic factors point to a healthy economy. In conclusion, this study demonstrates how network-based methods can assist in understanding the migration of workers, underlying the differences between demographic groups and how these patterns are affected by macroeconomic variables in combination with sociopolitical aspects.
The research was based on the database collected by the Spanish National Institute of Statistics on their employment survey over an extensive period (2005–2023). This database includes the province of employment of workers, as well as their province of birth and their level of study. A network was constructed based on this data to shed light onto the patterns of worker migration. Using community detection algorithms, FRs were identified for both the general population and highly educated workers, analysing how these varied over time. Furthermore, by analyzing the hub and authority scores, we observed dynamics in the attractive power of different provinces, relating them to socioeconomic variables also provided by the Spanish National Institute of Statistics.
Results underline how methods from network theory can help in the identification of FRs, shedding light on the migration patterns of populations. Through this novel methodology, we observe that highly skilled labour shows distinct migration patterns. Also, we observe a degree of specialisation between Spanish provinces, with some being more attractive for highly skilled workers. This attractive power shows a temporal dynamic, which could be associated with economic factors and other socio-political variables. Although the conclusions are limited to Spain, this novel approach based on network theory could be used as a blueprint for similar studies in other territories.
One limitation of the research is that its conclusions are based on methodologies from network theory that are not specific for this field, such as those for identifying the FRs and for estimating the attractive power from each province using the authority score. Nonetheless, these methodologies are broadly used in other fields to analyse networks, so they should also be applicable for this study. Furthermore, the general approach is highly innovative, so it could be adapted in the future once network methodologies specific for this field are available.
This research shows that migration patterns of highly educated workers differ from that of the general population. It also provides evidence for the specialisation of certain provinces towards highly skilled workers and how these variations vary over time. These changes seem to be related to economic variables, such as the attractive power of Barcelona and Madrid decreasing after the 2008 economic crisis. It may also be related to other political variables, as exemplified by the attractive power of Barcelona decreasing after 2018, coinciding with a situation of political turmoil.
The results of this research can aid in the development of government policies. Migrations by highly educated workers are closely related to the economic development of regions. Understanding their migration patterns and how they are related to economic variables will allow the development of more effective incentives to encourage or prevent it.
The methodology proposed within the article is of high originality. It demonstrates that methods from network theory can go beyond delineating FRs for manpower mobility, also studying their properties and how they vary through time. Although this research is limited to the case study of Spanish workers, it can be used as a guideline for future studies with similar research questions.
