13.07.2015 Views

The Spatial Concentration of Subsidized Housing - Poverty & Race ...

The Spatial Concentration of Subsidized Housing - Poverty & Race ...

The Spatial Concentration of Subsidized Housing - Poverty & Race ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

scale at which the impacts occur; 2) the level <strong>of</strong> subsidized housing that constitutesover-concentration; and 3) the level <strong>of</strong> poverty that constitutes over-concentration.<strong>The</strong> first is the definition <strong>of</strong> the geographic areas <strong>of</strong> concentration. Moststudies (including this one) use the census tract based primarily on both the availability<strong>of</strong> data at this scale but also the presumption that this is the scale at whichneighborhood impacts occur. Perhaps an impact area is one that includes multiplecensus tracts, or it could be an area smaller than a census tract such as a block group oreven a street. Impacts may not be aligned with census tracts at all. We do know fromprior impact studies that different impacts have different scales so it is important todistinguish the impacts that are significant to the policy decisions being made. In thiscase, spatial statistics could be helpful in measuring the scale <strong>of</strong> impact areasspecifically for the joint impacts <strong>of</strong> concentration <strong>of</strong> subsidized housing and poverty.<strong>The</strong> second aspect <strong>of</strong> concentration that needs attention is the definition <strong>of</strong>what constitutes “too high” a level <strong>of</strong> poverty. Research studies <strong>of</strong>ten use 40 percentor 30 percent poverty but these have been set by the researcher to categorize the datarather than developed as a result <strong>of</strong> the study. It is possible that the actual thresholdcould be as low as 20 percent for potential negative threshold effects. This studyrelied on evidence in Galster and Quercia (2000) that there was a potential negativeimpact above 20% poverty in order to develop policy guidance on limiting additionalsubsidized housing units. This level <strong>of</strong> poverty may not be the appropriate one. Asubstantial number <strong>of</strong> subsidized units are located in census tracts with between 20126

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!