A NEW DEFINITION OF ‘RURAL AREAS’ FOR THE METROPOLITAN CITY OF ANKARA, TURKEY

Abstract

The aim of this study is to create a new definition of ‘rural area’ for the metropolitan city of Ankara in Turkey by developing a rural index. Two models are employed in the paper. One of them is a logit model which identifies the factors affecting population density, and the second one is principal component analysis. The variables used in this study are population density, the number of businesses, the number of summer cottages, the proportion of the population with university degrees, total asphalt roads, the distance to the nearest administrative centre, the number of agricultural holdings, total agricultural land and the number of points of interests (banks, pharmacies, schools, etc.) for each neighbourhood. According to the principal component analysis, classifications are made at three levels: urban quarter, intermediate quarter and rural quarter. This analysis is made on the basis of existing data at the level of the neighbourhood. There is an important drawback of this study: it does not include data regarding income levels of the people since there was no data at the neighbourhood level. But with 10 variables, this study is quite sufficient for defining the rural and urban Ankara, including socio-economic and spatial characters of neighbourhoods.