Estimação de curvas de indiferença de acessibilidade urbana via modelos ordenados e krigagem

Abstract

The process of urbanization, growth of cities and urban structuring in recent decades among large Brazilian cities revealed the issue of accessibility as a relevant factor in quality of life. In this sense, analyzing the level of accessibility and welfare of individuals from where they leave their homes up to the point of execution of activities or consumer satisfaction becomes a matter of great scientific importance, yet to be explored in rigorously way. Thus, in this dissertation we analyze and model urban accessibility considering a theoretical perspective based on the methodology of utility maximization and estimation of econometric models. Therefore, this study is divided into two lines of research. The first one analyzes the accessibility using generalized ordered models through a new geo-referenced micro data set collected in the city of Fortaleza, Brazil. Our results show that variables such as income, car ownership, distance, and others are important for explaining accessibility of individuals. The second line of inquiry proposes and develops, in a pioneering way, a surface of spatial utility by means of Kriging techniques. The results point to the fact that the distance between home and destination has a very heterogeneous relationship with accessibility, revealing a spatial pattern greatly influenced by the prevailing economic inequalities all over the city. This result puts into question simplistic traditional assumptions that assume a linear or polynomial relation between distance and accessibility.