This thesis is a study of the conduct of local authority possession proceedings and the relationships of the parties involved in them: the courts, local authority housing departments and their tenants. On a more general level the thesis is concerned \\ ith the process of change and barriers to change in legal proceedings. The issue underlying these relationships is the nature of the security of tenure given to council tenants by the Housing Act 1980 (consolidated by the Housing Act 1985). On the face of it, discretion as to recovering possession of their properties was removed from local authorities and given to the courts through that legislation. Under the 1985 Act. local authorities are required to show that it is reasonable for a possession order to be made each time they make an application for possession on the ground of rent arrears. Various studies have been carried out about what happens in practice in possession proceedings, both independently and for the Civil Justice Review in 1986, which reported that the courts were more or less "rubber-stamping" local authorities' applications. As a result, the Lord Chancellor's Department instituted some procedural reforms in 1993 which were intended to ensure that courts properly exercise their discretion. This thesis looks at whether the procedural reforms have been effective in changing practice, and having found that they were not, then addresses the question of why that is the case. In seeking to answer the latter point, the author uses Giddens' theory of structuration to analyse the research data. Layder's 'realist' approach to social research is adopted, and provides a framework for discussion of the material by considering it in its micro, macro and historical contexts. The relationships between the agents and the structures pertinent to possession proceedings are analysed, and consideration is gi\en to where power is held and how it is used. By adopting such an approach, it is intended to make a contribution to the social theory oflaw