Introduction

The following is an exploration of comparative economic geography. I will deliver two maps and two Moran’s I autocorrelation plots, showing the clustering of families on SNAP benefits in Detroit and the median household income in Portland. I found significant clustering in both areas, although the hot and cold spots were inverted — Portland had hotspots of wealth in the inner-city while Detroit had hot spots of poverty in the inner-city. In the final analysis, I posit that this might be the result of more advanced stages of gentrification in Portland.

Detroit, Urban and Suburban Clustering

variable code
Total Households B22010_001E
Number of Houses on Food Assistance B22010_002E

Analysis

An spatial analysis of the local Gettis-Ord calculation for the Detroit metro area paints a picture of inequality between urban and suburban areas. The center of the City of Detroit in Wayne County has a high percentage of citizens receiving SNAP benefits. Most of the area consists of tracts with high numbers of people receiving SNAP and living next to people who receive SNAP (ie, a “hotspot” of high values and high neighboring values). Around the major hotspot lie tracts that have high values but are surrounded by tracts with insignificant values (ie, high-low values).

Beyond the buffer of insignificant values lies a blue “coldspot” with low percentages of SNAP recipients surrounded by tracts with similarly low percentages (ie, low-low and low-high). This would be a map indicating county inequalities, since Wayne County contains the largest number of SNAP recipients as a percentage of the population; however, looking at Pontiac in Oakland County gives us another interesting hotspot, again in the downtown area. Given this spatial autocorrelation, we are forced to conclude that urban areas are more likely to have high-high clusters, while the outlying suburbs (eg, Beverly Hills and Farmington) are more affluent. We often view this kind of spatial autocorrelation in areas of the Rust Belt and Northeast that experienced a large amount of white flight during the 1970s, as white populations moved to suburban areas to avoid integration brought on by the successes of the Civil Rights movement.

Some areas, like the tract west of John C. Lodge Expressway and east of Rosa Parks between West Euclid and Clairmont, have upwards of 60% of people on food stamps, and a tour of Google Street Maps shows that many of the properties are abandoned and condemned – many properties have been totally demolished. Looking further at the population of Detroit, one can see a drastic decline after the 2008 housing crisis, which hit these neighborhoods the hardest. Many of these people probably had sub-prime mortgages, were hit by the decline of employment due to transition to post-industrial economies, and faced foreclosure during the crisis. Some city council proposals have included literally depopulating some of these areas in order to save the city money on distributing electricity to areas that no longer pay enough into the system to make service worthwhile. Others, like James and Grace Lee Boggs have offered a post-work economy as a solution to some of the dire problems facing Detroit.

Comparing Detroit and Portland

I developed a Local Moran’s Eye for Detroit (above) and the PDX-Salem-Corvallis commutershed (below). As you can see, there is clustering in both studies. There is more analysis following the below webmap.

Final Analysis

The structure of census-grouped Combined Statistical Areas, or commutersheds draws together a space through patterns of movement from residence to employment. The Portland-Vancouver CSA has grown in recent years to encompass Salem and then Corvallis. I mapped the local Moran’s I statistic for spatial autocorrelation of MHI in the PDX-Salem-Corvallis commutershed, revealing fascinating trends that invert the model describing the Detroit/Rust Belt spacio-economic clustering, while maintaining discernable autocorrelation.

Instead of the population core in the center of the main city containing greater poverty, Portland’s more impoverished communities appear to live outside of the inner-city, its wealthy live in the hills closer to the downtown area. There is a small exception, in the form of a small section of the inner west side where I live, typified by mixed-income spatial variation of apartment blocks and single-family houses. However, the urban centers of Salem and Corvallis show the same trends as the Rust Belt, with cold spots in the central urban area.