Per Kropp, Barbara Schwengler (p)

Alexandre N Almeida

In this study, we revise and improve our previous delineation of labour market regions (Kropp/Schwengler 2016) by balancing density and self-containment as recommended by Martinez-Bernabéu and Casado-Díaz (2021) and by combining the results of different clustering methods. Based on current commuting data at the level of associations of municipalities in Germany, we try to optimise the quality of a functional delineation of labour market regions.
Our approach follows three steps. First, we use various methods (a graph-theoretic approach, the Travel-To-Work-Areas method, and clustering algorithms from network analysis) to generate an extensive set of meaningful delineations. Second, we select a small set of "best" delineations that balance self-containment (modularity) and a new density indicator. Third, we combine these delineations by identifying cores (regions that are equal in all selected delineations), parts (regions that sometimes belong to only one core region and in other cases are independent) and other regions (belonging to different or no core regions). The latter are finally redistributed to cores or parts according to dominant commuting flows.
This procedure results in the delineation of 36 main labour market regions, which include 117 sub-regions. Our result is a robust delineation because it combines main features of several clustering results. Using density and modularity indicators ensures that both self-containment and density are taken into account. The hierarchical result of main and sub-regions meets the diverse needs of researchers and policy makers.

Laura Helena Kivi (p), Tiiu Paas

Barbara Schwengler

The objective of the paper is to investigate neighbourhood effects as spatial interactions in European regional labour markets in terms of domestic and cross-border regions looking for answers to the following two research questions. First, whether the evidence of neigbourhood effects in labour markets is also valid in the case of European Union countries’ cross-border regions. Second, whether the existing neigbourhood effects are different in the sense of geographical and cultural distances.
The analysis relies on the data of European NUTS-2 regions’ unemployment and employment indicators as well as the characteristics of regions’ human capital, demography, industrial composition, and country specific conditions such as institutional setting. Spatial econometrics methods like spatial lag model (SLM), spatial error model (SEM), spatial autoregressive model with spatial autoregressive disturbances (SARAR) are used to analyse regional labour markets. Geographical and cultural distances are used to construct weight matrices to conduct the analysis based on the implementation of spatial econometrics methods.
The estimation results show that both substantive and nuisance forms of spatial dependence exist for regional labour market indicators. That confirms the existence of neighbourhood effects in European regional labour markets. The results also indicate that cooperation effects dominate between the regions. The existing neighbourhood effects are different between the labour markets of domestic and cross-border regions. Neighbourhood effects within domestic regions lie on both geographical and cultural proximity and are much stronger compared to cross-border interactions. Spatial interactions between cross-border regions are statistically significant only when cultural distance is considered when analysing neighbourhood effects. Thus, while elaborating regional and labour policies, attention should be paid not only to the labour market developments of culturally and geographically close domestic neighbours, but also to the developments in culturally similar regions across international borders.

Narae Lee (p)

Laura Helena Kivi

Does the railway increase jobs for rural labor markets in developing economies? Literature suggests that the modal addition resolves resource misallocation and labor market information asymmetry. At the same time, the connectedness causes the sorting of productive sectors into urban regions. By taking advantage of the weather-induced natural experiment in Colombia, I estimated the railway effects on the jobs in rural endpoints. Two rail lines damage each in inland and coastal regions enable the comparison of export and domestic sectors. The primary estimates of the Multi-valued Difference-in-Differences (DID) show that the infrastructure had gradient job effects for counties adjacent to the rail lines. While inland job effects enhanced as the infrastructure construction went by, the port-side job effects disappeared when the Design & Build (DB) finished.

Marcello L Souza Junior, Alexandre N Almeida (p), Carlos R Azzoni

Narae Lee

This paper uses non-parametric smoothing methods to identify subcenters (SBD) of four labor market areas (LMA) inland Brazil in 2006 and 2016. We use a two-stage method based on locally weighted regressions, with an endogenously determined cut-off critical value. We use matched employee/employer geocoded data from the Ministry of Labor. Given the labor market dynamics in the period, we found new subcenters in some cases, a stable situation in others, and the suppression of a subcenter in one case. Even in cases with just one center, we found growth in the area and sectoral diversification. The results reveal important structural changes in the sectoral composition of jobs in the dynamic centers. These changes constitute relevant inputs for public policies, particularly in planning the supply of public transportation and creating new interactive economic centers in these regions.