Inessa Tregubova (p)

Thomas de Graaff

The COVID-19 pandemic has fundamentally reshaped urban dynamics, primarily through the accelerated adoption of remote and hybrid work models. These shifts challenge long-standing urban equilibrium frameworks, such as the Alonso-Muth-Mills model, which has traditionally guided understanding of residential choices in urban economics. Previously, proximity to the central business district (CBD) served as a primary factor influencing residential desirability and housing prices. However, as remote work increases, proximity to the CBD has become less relevant, while other factors of the residential choice model like housing affordability and access to local amenities have gained high importance in the residential decision-making process.

Given that, the research primarily develops a theoretical model that explains the consequences of remote work on urban equilibrium, including the emergence of so-called ‘donut effect’. Overall, these modeled shifts explain why central areas may experience reduced residential demand, while suburban neighborhoods with attractive amenities see population growth.
The empirical testing of the model is done on the Tel Aviv metropolitan area in Israel, a region characterized by monocentric structure. The analysis mainly relies on a dataset of mobile GPS-signals collected between January 2019 and September 2023 which after removing noise represents approximately 5% of the area’s population.

The high spatial granularity and accuracy of these data allow for precise estimation of remote work patterns at the neighborhood level. Specifically, we identify individuals' home and work locations and measure their time spent at home during their individual working hours, applying Bayesian conditional probability.

To establish the existence of new urban equilibrium, we integrate neighborhood-level insights on remote work and new commuting frequency, and examine its impact on key component of urban equilibrium which is housing prices. To do this, we build spatial panel regression. This approach allows us to account for spatial dependencies and temporal variations, ensuring a robust analysis of the factors influencing urban equilibrium.

The findings of this research have direct policty implications. Municipalities are currently working to undertake appropriate actions to enrich the resilience of cities, and our empirical findings are likely to help prioritize these actions.

Kristýna Sauerová (p)

Thomas de Graaff

Unemployment is a socio-economic phenomenon that negatively affects not only the economic but also the social environment of territorial units. If a region is affected by high unemployment over a long period of time, eliminating the effects of this phenomenon imposes significant costs on national and regional policies. The effects of unemployment have made this topic a traditional subject for a wide range of scientific disciplines, including social geography. At the same time, it has already been empirically demonstrated that regional unemployment depends on a variety of factors that give rise to it, but these factors vary from place to place and cannot be defined as universally valid. Similarly, the unemployment path is closely linked to the evolution of economy, as the labour market responds to global, national and regional trends. Countries are tackling regional differences in unemployment rates through employment policy instruments. In this context, Czechia is characterised by very stable spatial patterns of low and high unemployment. Especially, structurally affected regions characterised by high unemployment rates present a challenge that cannot be effectively solved. The question is how effective national employment policy instruments are. This paper examines the relationship between the resources spent on active employment policy and labour market outcomes in Czechia. The study uses quantitative methods to uncover the relationship between the funds spent and the evolution of the unemployment rate. Specifically, the relationship is measured using data for the unemployment rate and the amount of funding provided for employment policy, namely the data for active employment policy from the European Social Fund for the period 2014-2020. In order to reveal the causal relationship, cross-regional sequencing is used, which allows to better separate the actual impact of the selected phenomenon on the outcome, and the impact that spontaneous change has. Furthermore, the spatial regression method is used in order to reveal other variables (population structure, geographical location, economic structure of the region) that affect the value of unemployment in the regions. The analyses are performed on detailed units of administrative districts of municipalities with extended powers, which makes it possible to analyse a micro-regional perspective that would remain hidden when performing analyses on higher administrative units. The analyses have shown that, surprisingly, there is no relationship between the funding provided and the unemployment rate; the influence and significance of the individual factors studied on the long-term stability of regional unemployment is then examined using regression methods.

Jacopo Tramontano (p), Valeria Cirillo, Dario Guarascio

Thomas de Graaff

Amazon is the dominant digital platform in e-commerce. Its diffusion in Italy began in 2011, with a sudden expansion of coverage between 2018 and 2021, in which the company has built 35 new facilities. Amazon's local impact on wages and employment is an empirical question of relevance for local decision-makers. E-commerce can be beneficial for wages and employment in core and complementary industries, such as logistics and transportation. It can however decrease them in brick-and-mortar retail and in trade-exposed goods producers. Further, buyer market power effects can be present in subcontracting, increasing employment yet decreasing wages. This paper, using income tax data at the municipality level coming from the Italian Ministry of Economics and Finance, estimates the net effect of two different types of Amazon facility, Fulfillment Centers (FCs) and Delivery Stations (DSs), on Italian local labor markets. FCs are located farther away from urban centers, and employ most logistics jobs. DSs are smaller facilities located in cities, where most consumers are and retail is more present. The centers are first pooled together to estimate a general "Amazon effect". An analysis differentiated between types of Centers then sheds light on potential heterogeneities between different typologies of local labor market. The paper relies on a staggered difference-in-differences methodology allowing for cohort-time heterogeneous effects, using not-yet-treated controls. Several robustness checks are performed. Results indicate that the impact of Amazon on per capita wages is negative, while no impact is detected on unemployment rates. Amazon Delivery Stations decrease wages in local labor markets, while Amazon Fulfillment Centers increase unemployment. The results are interpreted as supporting the hypothesis of a negative net impact of Amazon Centers driven by losses in retail jobs.