The HOPE VI programme in the US displaces tens of thousands of low-income households to disperse pockets of poverty and transform sites of `severely distressed' public housing into mixed-income housing. A complete evaluation of this programme's impacts on residents must examine the meanings and functions of these communities before they are dismantled. Therefore, this paper examines residents' lived experiences of place in one site before redevelopment. This socially well-functioning community allowed residents to lay down roots, form place attachments and create bonds of mutual support with neighbours, contrary to typical depictions of severely distressed housing. Implications for US public housing policy and parallels with the discourse on social housing and social inclusion in western Europe illuminate overarching trends in housing policy for the poor.

To what extent are people of different incomes and housing tenures engaged in social relationships in new mixed-income, New Urbanist HOPE VI communities? In Seattle's NewHolly Phase I, neighboring relationships are generally more frequent than in other mixed-income situations. Yet systematic differences among housing tenures by language, family composition, and patterns of local facility use and community involvement curtail social interactions. Most important, lack of proximity curtails relations among public housing residents and others on site, implying that the level of physical integration of housing units for the various tenures and incomes in a mixed-income development has repercussions for social interactions.

As the HOPE VI (Housing Opportunities for People Everywhere) program redevelops public housing, residents must relocate. Little is known about how they make the choice to stay or to go, if they are given one. Survey interviews with 200 residents of Seattle's High Point HOPE VI project provide the data to address four questions about such moves. First, what factors predict residents’ initial choice to stay on site during redevelopment or to move permanently away? Second, how does the initial choice predict actual behavior? Third, what is the role of place attachment and place dependence on residents’ relocation choices? Fourth, what is the role of other trade‐offs in decision making? Findings suggest that family situations and place‐dependent considerations shape initial relocation preferences of public housing residents and that their family situations may be a more important influence on the actual move. Implications for the HOPE VI program are discussed.

In theory, housing poor families in the suburbs among those who are not poor can provide better housing options and help families connect with economic and social opportunities. Social networks are vital links to larger social systems and the neighborhood networks of low‐income people may thus influence their access to opportunity. Does living in small clusters of public housing in a nonpoor area instead of in a dispersed housing pattern influence the types of social ties poor people use when they look for jobs? Dispersed residents have neighborhood social networks that contain greater diversity, so therefore they have greater access to diverse sources of information. Dispersed residents, however, use their neighbors less frequently when they look for a job than clustered residents do. Implications for theory, policy, and practice are discussed.