Family Life Cycle and Deforestation in Amazonia: Combining Remotely Sensed Information with Primary Data

This paper examines the relationships between the socio-demographic characteristics of small settlers in the Brazilian Amazon and the life cycle hypothesis in the process of deforestation. The analysis was conducted combining remote sensing and geographic data with primary data of 153 small settlers along the TransAmazon Highway. Regression analyses and spatial autocorrelation tests were conducted. The results from the empirical model indicate that socio-demographic characteristics of households as well as institutional and market factors, affect the land use decision. Although remotely sensed information is not very popular among Brazilian social scientists, these results confirm that they can be very useful for this kind of study. Furthermore, the research presented by this paper strongly indicates that family and socio-demographic data, as well as market data, may result in misspecification problems. The same applies to models that do not incorporate spatial analysis