Abstract :
Social science literature has long discussed the importance of so-called “social norms”, “neighborhood effects” or “peer influences” on the decision making process of individuals. However, traditional econometric techniques usually rely on the assumption that observetations are independent from each other, and therefore cannot reflect these effects, and often leads to incorrect inferences. Spatial econometrics can be considered as the modeling techniques that account for the peculiarities caused by the space component. Here, one of the most critical points of spatial models is the definition of the neighborhood, in other words, the location of observations. The proximity among locations can be defined based on the geography as in the regional studies or economic distances. Even abstract concepts of proximity, such as the inter-personal distance, can be used in such techniques. Hence, the distance definition that is appropriate to the notion of the study plays an important role. This study attempts to classify the spatial regressions, in which each one of which has a different interpretation, and tries to guide the researches while selecting the correct modeling technique. Another major point of this study is that it presents examples of studies in the field according to their distance definitions. These papers are grouped on the base of the geographical or non-geographical distance concepts employed.
Keywords :
Location , Spatial Regression , Geographical Distance , Economical Distance , Neighborhood Effect