Mean Streets is the first study of its kind, a careful examination of samples of homeless children, in situ – that is, on the streets of two Canadian cities, Toronto and Vancouver. Although “runaways” have been much publicized – even advertised – no previous study has systematically located children on the street and looked at their social worlds, comparing them with samples of in-home and in-school children, from the perspective of the children.

Mean Streets is also an all-too-rare combination of rigorous theoretical and empirical inquiry applied to a significant research problem. The problem is more complex than is implied in the book's title – that is, the relationship between youth homelessness and crime – though that surely is complex enough. John Hagan, Bill McCarthy, and their colleagues Patricia Parker and Jo-Ann Climenhage must first convince us that homeless youth are a sufficiently important population to warrant serious attention. They do that convincingly in Chapter One and in the book's Appendix, which I recommend reading in tandem.

Importantly, the two cities differ significantly in their approaches to handling homeless youth, with Toronto providing a more fully developed safety net for them and Vancouver relying to a greater extent on a crimecontrol model. Although law enforcement plays an important role in both cities, Vancouver youth, without the support services and shelters of Toronto's social welfare model, experience greater exposure to the traumas of street life and to criminal opportunities.