Regional Imbalances and Market Potential in Brazil

Abstract

This paper estimates the effects of market potential on regional wage imbalances between Brazilian municipalities. Using data from the 1980, 1991 and 2000 Brazilian Census, which render 3,630 comparable municipality areas, we estimate the NEG's wage equation using a spatial panel data model accounting for endogeneity. Our results show new evidence of a strong relationship between market potential and wages, indicating that regional attributes, as well as individual, are correlated with wages and their regional disparities.

Inégalités régionales et potentiel commercial au Brésil

RÉSUMÉ La présente communication évalue les effets du potentiel commercial des inégalités salariales régionales entre différentes municipalités au Brésil. Sur la base de données extraites des recensements effectués en 1980, 1991 et 2000 au Brésil, permettant de comparer 3,630 municipalités, nous sommes en mesure d'estimer l’équation salariale de NEG en utilisant un modèle de données de panel spatial, en tenant compte de l'endogénéité. Nos résultats fournissent de nouvelles informations sur l'existence de rapports étroits entre le potentiel commercial et les salaires, indiquant la corrélation entre les attributs, tant régionaux qu'individuels, et les salaires ainsi que leurs disparités régionales.

Desequilibrios regionales y potencial de mercado en Brasil

EXTRACTO Este trabajo estima los efectos del potencial de mercado sobre los desequilibrios salariales regionales entre municipalidades brasileñas. Utilizando datos extraídos de los censos brasileños de 1980, 1991 y 2000, que incluyen 3630 áreas municipales comparables, estimamos la ecuación salarial de NEG empleando un modelo espacial de datos de panel que tiene en cuenta la endogeneidad. Nuestros resultados muestran nueva evidencia de una fuerte relación entre el potencial de mercado y los salarios, indicando que los atributos regionales, así como los individuales, se correlacionan con los salarios y sus disparidades regionales.

Keywords:

• Regional imbalances
• market potential
• New Economic Geography
• Brazil

JEL classification :

• J31
• R12

Acknowledgements

The authors are grateful to Danilo Igliori and Sueli Moro for their contributions to a previous version of this paper. The authors would also like to thank the grant of a Travel Award by the Regional Studies Association, which allowed the presentation of this paper at the II World Conference of the Spatial Econometric Association in New York, 2008, as well as the helpful comments received therein by Nancy Lozano-Gracia and other participants. The usual disclaimers apply.

Notes

1. In his review of the existing empirical literature on the predictions of New Economic Geography models, Redding (2010) also points out the relative imbalance between the development of theoretical and empirical NEG research.

2. The municipalities that have been excluded from the sample are: Altamira, Aripuanã, Atalaia do Norte, Barcelos, Barra do Garças, Boa Vista, Caracaraí, Carauari, Chapada dos Guimarães, Fernando de Noronha, Itaituba, Japurá, Lábrea, Luciara, Nobres, Santa Isabel do Rio Negro (Ilha Grande), Santo Antônio do Içá, São Félix do Xingu, São Gabriel da Cachoeira, Tapauá e Vila Bela da Santíssima Trindade. Those municipalities’ areass, except for the islands, have suffered major changes by the methodology used in order to make the municipal areas compatible betwen 1980 and 2000. The resultant area would have great heterogeneity, which justifies its elimination from the sample. Besides, most of those municipalities are part of the Amazon Rainforest, where the main transportation routes are rivers and the average municipalities’ area is huge. An example is Altamira, the world's biggest municipality, with an area of 161,445.9 km2, which is equivalent to the sum of the areas of the Netherlands, Belgium and Portugal. These municipalities represent 0.5% of the total and since they are concentrated in the Northwest region of Brazil, their removal has only a minor impact on the spatial structure of the other municipalities.

3. As pointed out by the anonymous referee, this pattern can be explained by the high share of public servants in the North region. Together, public administration, defence and social security accounted for 7.45% of the employment in the region in 2000. This share is 50% higher than the average for other Brazilian regions.