Urban scientists, policy makers and public administrators have long been aware of the issue of concentrated poverty as the root of many social problems. The attention of most empirical studies has been concentrated on the impoverished urban areas where poverty concentration is present already.

This study provides an alternative perspective on addressing concentrated poverty problems by exploring poverty dynamics in moderate-poverty neighborhoods in US metropolitan areas 1990–2000. The theoretical basis of this study is that neighborhood economic status filters down over time and the current moderate-poverty neighborhoods can be seen as potential future high-poverty neighborhoods. Therefore, the best way to fight concentrated poverty is to discover the poverty change patterns of moderately poor neighborhoods in order to prevent them from degrading into higher-poverty categories.

Using the Neighborhood Change Database, this study starts with a longitudinal descriptive analysis of the general poverty trends of different poverty-level neighborhoods, and finds sufficient evidence to support the proposition that moderate-poverty neighborhoods are a major source of future high-poverty neighborhoods. A carefully designed quantitative model is then used to explore poverty change patterns of moderate-poverty neighborhoods 1990–2000. The model sets itself apart by explaining poverty transition from a competitive perspective. Several interesting findings are discovered: (1) maintaining a healthy regional economy is the most reliable way of addressing concentrated poverty problems, (2) neighborhood poverty succession process can be modeled from a competitive perspective as those neighborhoods with advantaged positions are less likely to experience poverty growth and those neighborhoods with disadvantaged positions face continuous declines, (3) large portions of both old and new housing stock are negatively associated with future poverty growth, (4) presence of racial minorities is predictive of future poverty increase but the effect gets weaker as the neighborhood poverty category rises, (5) more homeownership and presence of high-status households help to resist future decline, (6) significant effects of non-market forces is identified and the effect gets stronger in higher poverty-level neighborhoods.

This study contributes to our understanding of poverty succession mechanisms in moderate-poverty neighborhoods. From a practical perspective, a set of poverty change predictors are identified including female-headed families, initial poverty level, homeownership rate, and so on, which can serve as the basis for formulating anti-poverty policies targeting moderately poor neighborhoods to prevent them from degrading into higher-poverty areas. From a theoretical perspective, it enriches the existing urban poverty literature with its unique model explaining neighborhood poverty dynamics through inter-neighborhood competition. In addition, the spatial and multi-level techniques applied in this study improve the accuracy of the analysis results and contribute to richness of research tools applied in this area.