Raul Silveira Neto (p), Flávio Rodrigues , Roberta Rocha

Although its average income stage, the urbanization process in Brazil is extremely advanced, being the percentage of population living in cities around 85%. Different from the situation of developed countries, this very quick urbanization leaded to the presence of housing informality and lacking provision of public transport and household infrastructure. These factors appear fundamental for understanding the urban structure of the cities and may imply distinctive characteristics. Very few researches, however, have explored the characteristics of Brazilian cities and metropolitan regions, being the available evidence focused on specific cities. The objective of this work is, thus, to characterize the urban structure of the most important Brazilian metropolitan regions of the most populous macro regions of the country (São Paulo, Rio de Janeiro, Porto Alegre, Salvador, and Recife) by identifying the presence of sub-centers and pointing out their distinctive aspects related to the main CBDs. The research applies the methodology proposed by McMillen (2001), based on semi-parametric estimation, to an unique dataset of geocoded information of the precise locations of firms and employment. The set of evidence indicates that, in general, Brazilian metropolitan regions present urban structures more similar to those of US cities than to the ones present in Europeans urban centers, with the employment more concentrated relative to the population but also with important presence of important number of sub-centers (McMillen, 2001; Khrel, 2018). The results also indicate regional important differences: while the decentralization in the sub-centers of employment is stronger in the richest metropolitan regions of São Paulo and Rio de Janeiro, in the northeastern metropolitan regions of Recife and Salvador the CBDs are relatively more important and the number of sub-centers is smaller. These results are consistent with both the higher commuting time of the two first biggest metropolitan centers and with the more deficient provision of public transport and household infrastructure in the last two metropolitan regions.

Keisuke Takano (p)

This paper investigates empirically the interrelationship between the perception of a forthcoming disaster risk based on the actual disaster damage and the change in the spatial distribution of firm transactional networks (supply chains) by examining the Nankai Trough area after the Great East Japan Earthquake in the period 2009 to 2017. By adopting the propensity score matching and the difference-in-difference method, the study estimates the effects of tsunami damage on the magnitude of the spatial concentration or dispersion of the supply chain network stemmed from risk perception, measured by the inverse of the Herfindahl-Hirschman Index after 2011. The results show that the magnitude of the spatial concentration or dispersion of suppliers changes heterogeneously after 2011 largely dependent on firm size, which can indicate the barrier of the capacity constraint toward the pre-disaster preparation.

Carles Méndez-Ortega (p)

This paper aims to determine which reasons lead Software and Video games firms (SVE hereafter) to locate in certain areas of Barcelona. This high-tech industry is a key industry in developed economies mainly located in urban areas. To carry out this analysis, we use SVE firm entries at neighbourhood level between 2011 and 2013 and a set of covariates that capture neighbourhood characteristics (localization and agglomeration economies, high-tech amenities, diversity, human capital and crime). Our results show that i) SVE firms tend to choose locations with a high diversity and good high-tech amenities (e.g. 22@ district), ii) the importance of the localization and agglomeration economies, since spatial spillovers are a key factor for this type of firms and iii) the role of the diversity in the location process of these firms, since SVE firms choose places with a high diversity of cultural and creative activity.

Alexandre Alves (p), Jackelline Favro

The industrial production structure is a fundamental component for understanding the dynamics of local employment. Thus, this study aims to investigate the relationship between the local economic structure and the level of agroindustrial employment for Brazilian microregions from 2006 to 2016, considering both the sectoral dimension, where the degree of externality of specialization and diversification affect employment, as well as the geographic dimension, which identifies the spatial externalities from agglomeration economies, where the performance of a region is affected by the neighboring regions due to the spatial dependence on the concentration of economic activities. For doing so, the Exploratory Spatial Data Analysis (ESDA) and static spatial panel data analysis with fixed effects were used as econometric tools, allowing capturing possible regional differences invariant over time. The ESDA indicated the existence of positive spatial autocorrelation in the agroindustry employment level in all years under analysis, evidencing High-High clusters predominantly in the South and Southeast regions and Low-Low clusters in the North and Northeast regions. In the Central-West region, a High-High cluster was observed in 2011 in Goiás state, as well as a Low-Low cluster in Mato Grosso state during the whole period. The results obtained by the spatial panel revealed that the level of employment in the agroindustry is positively associated with the level of specialization in the same region and that the level of employment in the neighboring regions is also relevant for the growth of the local agroindustrial employment. Regarding the diversified agglomerations, no statistically significant relationship was identified with the level of employment in the agroindustry. Another result of the model was to identify that the agroindustry contributed to the dynamism of the Brazilian economy in the crisis periods, in 2009 and 2015, as it contributed to the positive aggregate performance of the manufacturing industry as a whole. In this sense, the present study provides elements that can aid in the design of regional development policies by indicating that the segments that make up the agroindustry are more productive when they operate close to one another and are important for the generation of employment and income for the Brazilian economy.