Homelessness is commonly perceived as an urban problem. Lee and Price-Spratlen (2004) City & Commy 3, 1 observe that, in the USA, the homeless, and services for the urban homeless, are mostly located relatively close together, usually in more deprived central city areas, and almost forming a service ghetto. Fagan (2003, San Francisco Chronicle) identifies Main Street as the centre of welfare and medical services in San Francisco. For the homeless, such a concentration renders services more accessible, but has a number of disadvantages: ‘For many, simply getting to the local night shelter is a frightening experience [and this fear has a] gender dimension, with women especially tending to prefer to use services located away from the main “service hub” ’ (homeless-research.org.uk). Informal areas—‘interstitial niches’—develop in streets, parks, and public car parks, where people find places to set up ‘home’, sleep, eat, make friends, and look for security, money, and entertainment. These niches are colonized by the homeless at particular times: ‘at dusk, the street families lay out boxes, build plastic houses, warm themselves, and cook on small fires. The children rarely have anything. Public space is their space only at night’ (Morelle and Fournet-Guérin (2006) Cybergeo, politique, culture, représentations 342). See also P. Cloke, J. May, and S. Johnsen (2010), which is both informative and well-reviewed.

Differentiation is not merely temporal; it is also social. Homeless Research record a division between ‘straightheads’, ‘smackheads’, and ‘pissheads’: ‘what for one homeless person might be experienced as a space of care, might for another become a space of fear because of the dominant presence of a particular homeless sub-group…thus embodied and performative geographies can become underpinned by specific spatial logic.’ See Cloke, Milbourne, and Widdowfield (2003) Geoforum 34, 1 on rural homelessness.

http://www.geog.qmul.ac.uk/homeless/ Research into homelessness.