Library Header Image
LSE Theses Online London School of Economics web site

Location of economic agents in Brazil: an empirical investigation

Lage de Sousa, Filipe (2009) Location of economic agents in Brazil: an empirical investigation. PhD thesis, London School of Economics and Political Science.

PDF - Submitted Version
Download (1MB) | Preview


This study focuses on regional aspects of the Brazilian economy. Three avenues are explored: two related to individuals’ and firms’ location decision and the last regarding public policy. In general, firms and individuals seem to be moving away from the main economic centre in Brazil, Sao Paulo. It seems nevertheless that public policy is not very related to these movements. In other words, government interventions to accelerate growth in less developed regions have not achieved their goal. Chapter 2 addresses the issue of internal migration in Brazil. This chapter investigates the influence of amenities and/or disamenities on migration flows, which is an issue not yet fully covered by the literature. It investigates whether changing dwellings across cities is associated specifically with violence using urban-urban migration data at municipality level. Results show that migration is affected by violence not only locally, but also when neighbouring effects are taken into account. These findings back up previous research which evidenced an inverse relationship between city size and violence. Turning to firms, Chapter 3 explores the role of geography in the location of manufacturing and of regional disparities in wages. According to theoretical models, employment concentrates closer to the market when increasing returns to scale are taken into account. As a consequence, regional wages are a decreasing function of transport costs to markets, since firms tend to compensate for these costs by paying less to their employees. Trade shocks may impact these regional wage disparities by making foreign markets relatively more attractive for firms than internal markets, or vice-versa. This chapter tested these hypotheses using Brazilian regional data. Having two isolated trade shocks, Brazil provides an excellent case for testing which shock was more effective in reducing regional disparities. Results show that regions with higher transport costs tend to have lower wages and a reduction in this cost through trade shocks has affected these regional disparities. However, it is not possible to distinguish which trade shock was more efficient to impact these regional unbalances. Chapter 4 evaluates the effects of the Brazilian Development Bank (BNDES) loans on firms’ productivity. The importance of BNDES in the Brazilian economy is quite sizeable, reaching over 10% of aggregate investment. Using micro level data, it was possible to investigate the impact on productivity, but also distinguish its effects between large and small projects as well as between rich and poor regions, since regional development is one of its statutory goals. Results suggest BNDES loans have no effect on firms’ productivity, even though some association was found without controlling for all firms’ characteristics. Overall, some lessons may be learned after this work. Not only are economic reasons key determinants for individuals’ and firms’ location decision as shown in Chapters 2 and 3 but also some other factors seem to be important as well. Social amenities, locally and in surrounding areas, are highly correlated to individuals’ migration decisions in the Brazilian case, especially violence. For firms, economic reasons prevail since trade shocks appear to change regions’ attractiveness between internal and external market. Last, but not least, government intervention does not seem to be associated to firms’ productivity after BNDES loans.

Item Type: Thesis (PhD)
Additional Information: © 2009 Filipe Lage de Sousa
Library of Congress subject classification: H Social Sciences > HC Economic History and Conditions
Sets: Departments > Geography and Environment

Actions (login required)

Record administration - authorised staff only Record administration - authorised staff only


Downloads per month over past year

View more statistics