Commuting times and the mobilisation of skills in emergent cities

Abstract

Labour mobility within a large city or metropolitan area is a necessary condition for the optimal exploitation of agglomeration economies. We propose a method to establish which municipalities should be considered part of a metropolitan area based on labour market integration. In order to aggregate geographically proximate urban municipalities, we develop a network-based model that makes industry productivity (manifesting in formal employment creation) dependent on firms' ability to find, within city limits, the diversity of skills they need to move to new complex industries. In this way, we uncover the natural city scale at which firms optimally exploit the range of skills available to them within a broad catchment area. Considering Colombian cities, we find that commuting times between 45 and 75 min, corresponding to between 43 and 62 distinct cities or integrated labour markets, allow firms to maximise formal employment creation (between 2008 and 2013). This result supports the development of passenger transport limiting commuting times within cities, connecting small-and mid-size cities to nearby large cities, and coordinating transportation investments across traditional administrative boundaries.

Administrative boundaries and urban areas in Italy: A perspective from scaling laws

“…The aim of this paper is to propose a method for the fusion of administrative boundaries in Italy at municipality level, driven by area-population scaling relations. The proposed procedure is grounded on our view of the empirical area-population scaling relation as a “performance” assessment of each individual datum in the data set (municipalities, in this case) as compared to expectations generated by the whole data set based on the individuals’ size (population, in this case) ( Bettencourt, 2013 , West, 2017 , van Raan, 2019 , O’Clery et al, 2019 ).…”

Connecting up embedded knowledge across Northern Powerhouse cities

“…While a large variety of empirical metrics have been proposed to proxy for inter-industry relatedness (Hidalgo et al, 2007;Frenken et al, 2007;Neffke et al, 2011), we will use inter-industry job switches to measure industry 'skill relatedness' (Neffke and Henning, 2013). Similar approaches have previously been used to explain employment growth and industry diversification in a range of contexts (Neffke et al, 2011;O'Clery et al, 2016O'Clery et al, , 2019Landman et al, 2020).…”

“…For a more detailed review of historical approaches to defining cities see Duranton (2015). Here, we develop an algorithm related to that of O’Clery et al (2019) and Dijkstra and Poelman (2012) and use travel time between neighbourhoods (MSOA) in order to vary our city radius. Importantly, when including new neighbourhoods into a city, we are capturing the presence of industries (and skills) in those neighbourhoods as recorded by firms – and not the residence of commuters.…”

“…We develop a model based on one proposed by O’Clery et al (2019) that estimates the effect of labour pooling on employment growth in cities for varying city size. Specifically, we adapt a model for industry employment growth based on the number of workers present in ‘skill-related’ industries in a city proposed by Neffke and Henning (2013).…”

A bi-directional approach to comparing the modular structure of networks

“…Skill-relatedness networks have been applied to model and predict the diversification paths of regions in a large number of contexts [35,37,46]. They have also been used in various other applications, such as models for urban formality growth [45,47], employment resilience and adaptability [10] and knowledge spillovers from MNEs [6].…”