Data tabulated by the U.S. Census shows that poverty rates have jumped by 2.6 percentage points since the onset of the recession, rising from 12.5 percent in 2007 to 15.1 percent in 2010. Although the annual rate of 15.1 percent is the highest rate since 1983, growth in the U.S. population means that the 46 million people in poverty at the end of 2010 is the highest number in poverty since the official poverty line was first established in the early 1960s. Poverty has also been changing in ways other than the simple increases in numbers and rates. The common perception is that poverty is largely a phenomenon of metropolitan central cities. Trends in the last two decades however, have dramatically shifted the poverty landscape. Over the last decade alone, the number of poor in suburban areas grew 53 percent, almost double the 23 percent growth in central cities. The total number of poor persons living in the suburbs of the largest 95 metropolitan areas now exceeds the number of poor persons living in the central cities of those metropolitan areas by several million. Recent research has paid little attention to these trends, and even less is known about how the recession and its aftermath have affected the geography of poverty. How has the spatial distribution and concentration of poverty shifted? Has the recession accelerated the suburbanization of poverty? Is the social safety net structured to respond to the suburbanization of poverty, or do changes need to be implemented to better serve the increasing geographic diversity of poverty?
Professor Scott Allard proposes to address these questions in a study designed to explore the changing spatial distribution of poverty, the causes of these changes, and their implications for safety net programs. He will begin his study by presenting descriptive trends of the changes in the spatial distribution and concentration of poverty in the 100 largest metropolitan areas between 1990 and 2010, with primary attention to trends over the last decade. He will focus his analysis on these metropolitan areas because they house two-thirds of the nation’s population and capture the greatest suburban expansion in the last decade. These analyses, based on data from the Census and the Current Population Survey, and carried out at the level of Census tracts, will document the substantive changes in the geography of poverty and set the stage for the causal analyses.
The second phase of the project will focus on evaluating the many different hypotheses commonly offered to explain the shifts in the geography of poverty. The primary focus will be on structural changes in the economy over the last two decades and the job loss following the Great Recession, although other demographic patterns and economic trends thought to be responsible, such as increases in the numbers of female-headed households, population aging, and population changes in racial and ethnic composition, will also be assessed. For example, incorporating the Longitudinal Employer-Household Dynamics (LEHD) data compiled by the Census Bureau will allow him to assess the hypothesis that changes in the location and concentration of poverty will be particularly sensitive to the loss of low-skilled or lower-paying jobs. He will estimate a series of multivariate regression models to understand how demographic and economic conditions are related to changes in poverty at county and census-tract levels. These analyses will also include spatial regression models to explicitly account for the relationship between change in poverty in a given place and changes in poverty, economic conditions, and demographics in neighboring places. Case studies from Los Angeles, Chicago, and Washington, D.C. will be used to help situate and explain the findings from the quantitative analyses.