Exploratory Analysis of the Homeless Shelter System in Columbus, Ohio

Bibliographic information

Lobao, E. G., & Murray, A. T. (2005). Exploratory analysis of the homeless shelter system in columbus, ohio. Geografiska Annaler Series B: Human Geography, 87(1), 61-73.

Glossary

ESDA

Exploratory Spatial Data Analysis

CSB

Community Shelter Bureau

Outline

• Introduction
• Homelessness is a widespread problem. Cities in North America often provide shelters and services to the homeless. An understanding of spatial issues associated with homelessness can enhance these services.
• Previous research has not examined important spatial questions. Quantitative information is necessary for planners to make informed decisions.
• GIS-based research is ideal for examining spatial issues because it facilitates exploratory spatial data analysis (ESDA). It also provides understandable visual output.
• Columbus is a growing urban region with a well-developed shelter system. Two main types of shelters: emergency (short-term) and transitional (long-term, stable housing).
• The Community Shelter Bureau (CSB) was established in 1986 to oversee the management of the city's shelters. The shelter system is comprised of fifteen different organizations.
• Background (Literature Review)
• The goal of homelessness research is to understand who the homeless are and the process by which people become homeless.
• Two main schools of thought: 'personal pathology' (homelessness as an individual problem) and 'structural' (homelessness as the result of systematic forces). Both schools of thought have influenced public policy at different times.
• Decisions about where homeless shelters will be placed require neighborhood involvement. Neighborhood characteristics are difficult to define; boundaries are unclear and official data collection may not match such boundaries.
• Decision-making is often based off of perceptions of shelter operators. Perceptions are shaped by personal experience and may not reflect actual realities. In order to improve access to services, planners must have access to quantitative spatial data.
• Research Question: Do perceptions of service providers reflect the observed conditions of neighborhood characteristics?
• Methods
• Interviews with 2 CSB staff and 7 shelter organization members. 'Creative interviewing' technique (open-ended discussion to identify perceptions).
• Participant observation (author worked alongside CSB for 6 months)
• Exploratory spatial data analysis (ESDA) involves a wide variety of quantitative methods including GIS.
  • The ESDA approach is that "data should be approached with few assumptions and research questions do not need to be heavily structured at the onset of analysis... instead, one should allow the patterns and relationships within data to present themselves." (65)
  • GIS is ideal for this particular research because it enables a uniform definition of 'neighborhoods', minimizing spatial variability. (Neighborhood = the area within 800m of a particular shelter)
• GIS software: ArcView.
• Results: Interviews & Perceptions
• Shelters offer specialized services to individuals with specific problems. Individuals are refered to to other shelters on the basis of services provides without awareness of neighborhood characteristics.
• Barriers to individual independence or development of new programs: inadequate transportation, lack of employment, lack of affordable housing.
• All shelters are located in areas with high concentration of minority populations, low incomes, low education, and no safe, affordable housing.
• Shelter operators felt that neighborhood residents were 'powerless' in the location of shelters. CSB members felt that since development of shelters is rarely challenged, neighborhood residents recognize and support homeless shetler placement decisions.
• Results: ESDA & GIS data
• Figure 1: Spatial distribution of shelters. Shows that emergency shelters are closer to downtown while transitional shelters are further away.
• Figure 2: Number of beds at each location (differentiated by size). While more shelters are further from downtown, the number of beds in each shelter shows that transitional services are also concentrated downtown.
• Figure 3: Section 8 subsidized housing & shelter locations.
• Figure 4: Zip code of last residence for all homeless individuals in Franklin County.
• Figure 5: 800m buffer neighborhood example. By comparing the area 800m in all directions from all shelters to a similar area from randomized points, a comparison of the average characteristics of shelter neighborhoods vs. typical neighborhoods was conducted.
• Significant average characteristics of shelter neighborhoods (shetler : non-shelter)
• More employment centers (358.9 : 77.1)
• Lower land value ($198,353 : $254,457)
• Higher concentration of commercial land (22% : 8%)
• Fewer people live near shelters (950 : 1600)
• Fewer households/families (183 : 404)
• Higher concentrations of minority population (51% : 22%)
• Nearly all 36 variables showed differences between sheltered and typical neighborhoods.
• Discussion/Conclusions
• First step in improvement: Visualizing the shelter system. This will enable organizations involved in management to develop a system-wide perspective rather than one grounded in their own perceptions.
• By comparing concentrations of last ZIP code of homeless individuals, one can identify areas to target for homelessness prevention.
• Most of the perceptions associated with homeless shelter neighborhoods match the reality: low income, high minority concentration, historically economic transition zones.
• Differences in perception and findings:
• More commercial land in shelter areas. Conflicts with perceptions of a lack of employment. May be explained because data deals only with the number of employees & employment centers, not the type of jobs available.
• Lower number of residents & families in shelter areas. Conflicts with CSB perception that shelter placement considers the community's opinion. The lack of protest may be a result of the lower number of individuals in the area, not community acceptance.
• By exploring shelter systems spatially, a more nuanced understanding of the system can be reached that distinguishes between reality and perception.

• Click here to return to my homepage.

• Click here to return to the class homepage.