Concentrated Poverty: Components of Change

Presented at the Urban Affairs Asssciation (UAA) on April 5, 2013

While a strong economy reduced the concentration of poverty in American cities in the 1990s,recent
studies have shown that the extent of concentrated poverty increased again after 2000. All of this
research, however, presents data only on the net change in this measure, yet policy would be better
informed by understanding the shifts for the three component categories of neighborhoods that yield
the overall result: (1) census tracts that had high poverty at the start of a period but then improved (i.e.,
poverty rate fell below the threshold needed to quality as "concentrated");(2)tracts that had low
poverty at the start, but moved into the concentrated group by the end; and (3)tracts that remained in
the concentrated poverty group throughout. This paper first examines the changes in the numbers of
tracts(and their population sizes)in each of these categories in the 2000s (comparing data from the
2005/09 ACS to that forthe 2000 census, in the nation's 100 largest metropolitan areas). We look at
differences by region and, where confidence intervals allow, contrast patterns across metros and for
cities versus suburbs within metros.One interest is learning how the spatial pattern of those that
improved compared with those that worsened. Another is learning how the characteristics of places
where the patterns are volatile (high proportions of tracts moving in an out of concentrated poverty)
differ from those for places where the patterns are more stable. The following section examines how
conditions within the three categories changed over the period. For example, we expect that indicators
of well being generally improved for tracts that moved out of the high poverty group and worsened for
those that moved into it, but we want to find out about the magnitude of those changes and whether
there were different trajectories for different indicators. For this purpose we use a set of measures
selected by the National Neighborhood Indicators Partnership for its Shared Indicators program. These
include measures related to well-being (e.g., income, homeownership rates, education levels) but
explanatory factors(e.g.,shifts in racial composition and age structure) are included as well.