Yu Zhou and Bruce W. Smith

Evaluating the Location Efficiency of Congregate Meal Sites Using Network Analyst

The spatial distribution of the elderly has experienced a significant change in the past decades. In Lucas County, Ohio, for example, the elderly were concentrated in the downtown area in 1950s. By 2000, the highest percentages of elderly population were mostly distributed in suburban areas. Even though the spatial distribution of older Americans has changed, the location pattern of services has not shifted substantially. In Lucas County, the congregate meal sites are mostly located in the same places as 30 years ago. In this paper, the strategies of relocating such sites with the use of ArcView Network Analyst are demonstrated.


INTRODUCTION

The number and percentage of older individuals in our society has risen dramatically during the past century. Today more thirteen percent of the total population in the United States is aged 65 and older, compared to just over four percent in 1900. This increase has led to more attention being given to the situation of older individuals, and strategies for improving the quality of life for individuals in later life.

Maintaining one’s independence as one grows older is a concern for the individual, their family, and our society. Congregate meal sites, or nutrition sites, are important components in the mix of social and medical services that help older people maintain their independence. The federal government under Title III of the Older Americans Act primarily funds these facilities. Specific goals of the program are: 1) to provide elderly persons and their spouses with low-cost nutritionally sound meals in strategically located centers, where they can obtain other social and rehabilitative services, 2) to promote better health among senior citizens through better nutrition, 3) to increase the social interaction of the elderly, and 4) to allow seniors an opportunity to live in dignity (Wacker, Roberto, and Piper, 1998). In addition to meals, these facilities provide nutrition education programs, screening, and information about other elderly services in the community. Because of their information and referral services, congregate meal sites play a significant role in coordinating and delivering other services to older persons, and therefore their access to the elderly is important.

Access to services has long been a research theme in geography. Geographers have examined the accessibility of a variety of services, including recreation, shopping, police protection, and education (Pacione, 2001; Talen, 2001). However, the accessibility of congregate meal sites to the elderly has not been investigated. The purpose of this study is to analyze the location pattern of congregate meal sites in Lucas County, Ohio. More specifically, it will illustrate the use of ArcVew Network Analyst to evaluate their locational efficiency.

DATA

The study area is Lucas County, Ohio, which is the central county of the Toledo, Ohio Metropolitan Statistical Area (Figure 1). In 2000, Lucas County had a population of approximately 455,000 people, of whom 59,441, or 13.1 percent, were aged 65 and older. Data describing the spatial distribution of various groups of older people were collected for block groups in Lucas County (http://www.geographynetwork.com/). Block groups were used instead of census tracts because they provide a finer resolution. On the other hand, census blocks would provide an even finer resolution but their greater number, 8138 census blocks as opposed to 434 block groups, creates presentation problems when one is confined to paper output.

The location of meal sites in Lucas County were obtained from the the Area Office on Aging (http://www.agingnorthwestohio.org).The Area Office on Aging is responsible for planning and funding a multitude of services for the elderly in a ten-county area in Northwest Ohio, including Lucas County.

DISTRIBUTION OF CLIENTS AND SERVICE PROVIDERS

The highest concentrations of persons aged 65 and older were in the suburban areas of Toledo (Figure 2). One type of area showing a concentration of senior citizens is the older suburban communities, such as Sylvania and Ottawa Hills, in the northwest suburbs, as well as Waterville in the southwest. Undoubtedly many people moved into these communities during middle age and subsequently have “aged-in-place.” In other block groups, the concentration of older persons may be attributable to the location of nursing and retirement housing. In a few block groups on the edge of the urban area there has been substantial population growth due to the comparatively recent development of multifamily housing. These developments have attracted large numbers of people, both young and old. Because of the block groups’ larger geographic extent, they contain large numbers of older persons even though the percentage of elderly in these areas is below the county average. One notable pattern on Figure 2 is that there are not large concentrations of elderly in the inner city. This is a departure from past patterns. In their review of the intraurban distribution of the elderly in Toledo between 1940 and 1970, Hiltner and Smith reported a concentration of older people in the CBD and the inner city of Toledo (Hiltner and Smith, 1974). It should be pointed out that the percentage of persons aged 65 and older is high in some of the downtown block groups. For example, senior citizens constitute 22.4 percent of the population in the block group adjacent to the Central Business District. However, this high percentage is attributable more to the absence of younger persons than to the presence of large numbers of elderly. When looking at numbers of persons, it is apparent that senior citizens, similar to all age groups, have become more suburbanized over the years.

Lucas County contained 24 meal sites located throughout the county in 2000. The pattern exhibited a distinct central city orientation, with 50 percent of the facilities being located within 2 miles of the Central Business District (Figure 3). In part, the central city orientation is the result of historical development. In the 1970’s, six meal sites opened and they were all located in close proximity to the Central Business District. Their location in the CBD was partially a result of the concentration of senior citizens in that area. Moreover the need for services was greater in that area because of lower incomes and a greater concentration of the “old old”, or persons over the age of 75 (Smith and Hiltner, 1988). After the 1970’s 18 additional meal sites were established. Several were opened on the outer edges of the city as well as in the eastern and western suburbs of Toledo (Figure 3). This growth in the outlying areas is probably the result of urban spatial growth as well as growing numbers of older persons residing outside the central city due to “aging in place”. Despite the decentralization of the meal sites, it is notable that six of the newer facilities were opened in or close to the Central Business District even though the six original meal sites were continuing to operate in that general area.

NETWORK ANALYSIS

ArcView GIS and the Network Analyst extension were used to evaluate the accessibility of older persons to the network of meal sites by establishing service areas for each facility. There are few guidelines in the literature as to the appropriate size of service areas. For example, Wacker et. al. suggested that congregate meal sites should be located strategically, but added no explanation as to what constitutes a strategic location (Wacker, Roberto, and Piper, 1998). In contrast to some other public services, such as fire stations, there are no prescribed standards for the size of the service areas of meal sites.

Given the absence of guidelines in the literature, the service area of the meal sites was arbitrarily set at two miles, as an approximation of the maximum walking distance. Service areas with a radius of two miles were created around each congregate meal site (Figure 4). It is important to emphasize that this is not straight-line distance, but instead is distance measured along the local street network, which provides a more accurate measure of distance (Chang, 2002). As one would expect given the numerous facilities in that area, there is substantial overlap in the service areas in downtown Toledo. In contrast many block groups outside the city are several miles from a facility. For example the areas in the southwestern corner of the county are as much as 15 miles from the nearest facility.

Using GIS all the census block groups lying totally within the service areas were identified and the number of persons aged 65 and older residing in those block groups were calculated. If any part of a block groups was outside the service area, it was excluded. Because block groups not totally within the service area were excluded from the analysis by the software, some people living within the specified distance of a facility were excluded from the final estimates. Thus the number of persons and corresponding percentage should be considered as only an approximation.

In Lucas County in 2000, 23,300 persons aged 65 and older lived in block groups located within the two-mile service area of meal sites. Those elderly living within the specified distance constituted 39.2 percent of the total elderly population (Figure 5).

The percentage of persons living within any given distance of a set of service locations may be of interest in and of itself. However, more importantly, it can also be interpreted as a index of the locational efficiency of a given network of service facilities, or the extent to which a given location pattern provides access to a group of clients distributed throughout a service area. When applied in this manner, it can be used to compare alternative location patterns. For example, if one of the meal sites in the inner part of Toledo were relocated to an outlying area as shown in Figure 8, the percentage of persons aged 65 and older residing within two miles of a facility would increase from 39.2 percent to 47.3 percent (Figure 6). In fact, due to the overlap of service areas in the inner city, more than one facility could be closed down without impacting the percentage of the elderly living within two miles of a facility. Moreover one could use a similar approach if one were strictly concerned with the best location for a new meal site without closing one of the existing facilities. In general, this type of analysis helps identify a location that will improve the access for the entire network. Applied in this way ArcView Network Analyst can help decision-makers determine the spatial allocation of community resources to best fulfill the needs of older persons.

SUMMARY

The congregate meal sites in Lucas County show a distinct inner city orientation. This pattern has developed because in earlier decades the elderly, particularly those ostensibly in greatest need, were most prevalent in the downtown area. Over the years, the elderly population has become more suburbanized largely due to “aging-in-place”, but the location pattern of congregate meal sites has not shifted to the same extent. Consequently, the elderly in the inner city have greater access to congregate meal sites than do those persons living in the periphery of the urban area.

One reason meal sites have not shifted commensurately with the geographic shift in elderly is that congregate meal sites often become a fixture within the local community. While overseen by the Area Agency on Aging, congregate meal sites are established through a partnership with non-profit organizations in the area, including senior apartments, churches, temples and community centers. If no non-profit organization is willing to undertake the project, then it is likely no service will be opened in that area regardless of need. Funding constraints make expansion to new sites less attractive than expanding service at existing sites. The location pattern of congregate meal sites thus evolves over time in a series of autonomous decisions that result in one facility being established at a time by independent decision-makers dealing with constraints which limit concern with the overall locational efficiency of the network of facilities.

In closing it must be noted that this use of Network Analyst  is of interest to social service practitioners who focus on the planning and evaluation of elderly services, at least those persons working in the Ohio’s state and regional agencies. They may consider urban and rural differences in service delivery but seldom have used areal units as detailed as census block groups or census blocks. Moreover they are interested in this approach as a method of increasing the “efficiency” of the network of service locations.

REFERENCES

Area Office on Aging of Northwestern Ohio, Inc. 2002. Congregate Meal Sites. http://www.agingnorthwestohio.org/.

Chang, K. 2002. Introduction to Geographic Information Systems. New York: McGraw-Hill Higher Education.

Geography Network. http://www.geographynetwork.com/.

Hiltner, J., and B. Smith. 1974. Intraurban Residential Location of the Elderly. Journal of Geography. 73(4):23-33.

Pacione, M. 2001. Urban Geography, A Global Perspective. New York: Routledge.

Smith, B., and J. Hiltner. 1988. Changing Intraurban Location of the Elderly and Access to Nutrition Services: A Case Study of Toledo, Ohio. Ohio Journal of Science. 88(5):189-191.

Talen, E. 2001. Schools, Community, and Spatial Equity: An Empirical Investigation of Access to Elementary Schools in West Virginia. Annals, Association of American Geographers. 91(3):465-486.

Wacker, R., K. Roberto, and L. Piper. 1998. Community Resources For Older Adults. Thousand Oaks: Pine Forge Press.


Yu Zhou
Associate Professor of Geography, Bowling Green State University, Bowling Green, OH 43403.

Bruce W. Smith
Professor of Geography, Bowling Green State University,Bowling Green, OH 43403.