Forced moves and chosen moves: residential mobility in Ankara, Turkey

This thesis presents an analysis of residential mobility in Ankara, Turkey. The principal question posed is whether the household adjustment model, in which residential mobility is defined as a mechanism enabling households to meet their housing needs, is applicable.

The thesis examines the previous literature on residential mobility, describes the economic and social context of housing decisions in Ankara, and then presents a detailed analysis of a survey of a representative sample of Ankara households. The critical review of previous writing on residential mobility leads us to set out a conceptual framework which includes household decision making and the context in which such decisions are made. It is shown that in Ankara this context includes economic liberalisation policy, declining average real wages and housing costs rising faster than inflation. The scene is set for an examination of the relative value of the household adjustment model and a 'forced mobility' model in which households are strongly constrained by land and housing market actors.

The analyses of the survey distinguishes four types of area (high, medium and low income authorised and unauthorised) and two types of tenure category (owners and tenants). Unlike most of the few previous residential mobility studies in third world cities our sample covers the whole population rather than migrants only. Path models and logit models are developed of past residential mobility, planned residential mobility and housing satisfaction. It is shown that owner-occupiers are highly immobile compared with tenants.

These analyses reveal that the household adjustment model has only limited value in explaining residential mobility. It is relevant in the high income areas and t osome extent in the middle income areas but has only limited relevance in the low income area types. The models of satisfaction are used to show that residential mobility in low income areas does not lead to greater housing satisfaction, or to better housing as measured by an objective index. Further support for the applicability of the forced mobility model was provided by qualitative follow-up interviews with tenants who had moved. These revealed the importance of landlords pressures on their housing decisions. An exception to the above statements concerns owner-occupiers in unauthorised areas who are relatively happy with their situation, possibly due to the speculative potential of land in many such areas.

It is therefore concluded that the household adjustment model cannot be applied as a general model of residential mobility in Ankara. Rather, the further one moves from the high-income type of area to the low-income and unauthorised type of area, the greater the relevance of the forced mobility model. It is thus argued that the main difference between third world cities and advanced capitalist cities affecting the character of residential mobility lies in the contexts within which mobility decisions are made rather than in households' housing needs.