labor force participation (see recent reviews by Gephart, 1997; Leventhal and Brooks-Gunn, 2000; Sampson, Morenoff, and Gannon-Rowley, 2002). Until recently, health outcomes had been noticeably absent from this list, but multilevel studies are becoming increasingly popular in health research. This new multilevel research on local social context and health has garnered wide attention in social epidemiology, as evidenced by the publication of four reviews of this literature in the past 3 years (Diez-Roux, 2002; Ellen, Mijanovich, and Dillman, 2001; Pickett and Pearl, 2001; Robert, 1999).
One of the reasons that public health scientists have become so interested in local context is to better understand the striking and persistent racial and ethnic differences across a range of health outcomes that have eluded most efforts to explain them using data at the individual level (Krieger, 1994; Williams and Collins, 1995a). There are large racial/ethnic differences across many causes of morbidity and mortality, and even through casual observation it seems obvious that perhaps some of this health inequality is related to the different types of contexts, or “ecological niches,” into which different racial/ethnic groups are born, and within which they grow up, live, and work. As research on residential segregation demonstrates (Acevedo-Garcia, 2000; Cooper et al., 2001; Ellen, 2000; LaVeist and Wallace, 2000; Massey and Denton, 1993; Polednak, 1996) place-based disparities are of central importance to understanding race-based health disparities in the United States. Moreover, a better understanding of why place and context matter also promises to yield new insights and intervention strategies for addressing racial/ethnic inequalities in health. Thus, the potential of place-based research to inform health intervention strategies that target places as well as people has given further impetus to this research in public health (Macintyre, MacIver, and Sooman, 1993; Sooman, Macintyre, and Anderson, 1993).
At this point, research on local context and health remains somewhat disengaged from recent theoretical and methodological developments in the sociological literature on neighborhoods. For example, whereas the sociological literature on neighborhood effects has taken a “process turn” in recent years and begun to focus more on the mechanisms that explain why neighborhoods matter (Sampson et al., 2002), most research on the neighborhood context of health is still attempting to establish that context matters. This is partly because most health research is framed in a paradigm where individual-level proximal influences—such as behaviors or biomarkers of pathogenic processes—take precedence over contextual factors (Krieger, 1994; Palloni and Morenoff, 2001; Schwartz, Susser, and Susser, 1999; Susser, 1998).1
To some extent neighborhood effects research on health remains mired in a “poverty paradigm” (Rowley et al., 1993) focusing mostly on the