«Urban Problems and sPatial methods VolUme 17, nUmber 1 • 2015 U.S. Department of Housing and Urban Development | Office of Policy Development and ...»
Discussion and Conclusion The title of this article poses the question of which scale (the house or the neighborhood) “matters more” when predicting abandonment. This question can be answered using multilevel modeling techniques because multilevel models allow for the same variables to appear at different scales in the same model. Furthermore, because odds ratios are an indication of effect size, the variables (and scale) with the largest odds ratios can be thought of as the ones that are most important when predicting abandonment. The large odds ratios for the house-level variables initially seem to indicate that the characteristics of a house matter more than the characteristics of a neighborhood. The answer to this question about scale, however, is more nuanced because the odds ratios for neighborhoodlevel variables are based on a 1.00-percent increase. This interpretation means that the effects of neighborhood-level variables are greater if one considers a threshold higher than 1 percent. Take, for example, property values, which has an odds ratio of 1.642 at the house level. At the neighborhood level, a 10-percent increase in the number of properties whose values are less than the median property value yields an odds ratio of 1.185 (0.017 x 10 = 0.17; e0.17 = 1.185), which is less than
the odds ratio of that variable at the house level. If 50 percent of properties in a neighborhood are valued at less than the median value, however, the odds ratio increases to 2.3 (0.017 x 50 = 0.85;
e0.85 = 2.340); and if all the properties in the neighborhood are valued at less than the median property value, the odds ratio increases to 5.5 (0.017 x 100 = 1.7; e1.7 = 5.474), which is much higher than the odds ratio at the house level.
Exhibit 3 emphasizes this point by comparing the odds ratios at the house level with the odds ratios obtained with different percentages of the neighborhood-level variables for all variables in the study.
As one can see from the exhibit, some of the odds ratios increase rapidly due to the exponentiation that occurs with logistic regression. Many quickly exceed the odds ratios at the house level, suggesting that, for stable neighborhoods (neighborhoods with low levels of the independent variables), the characteristics of a house matter more; however, for more distressed neighborhoods, neighborhood characteristics have a greater influence. There thus appears to be a tipping point after which neighborhood characteristics become more important when predicting the probability of a house being abandoned, although the exact point differs by variable. In addition, the ICC indicates that more than one-half of the variability is attributable to neighborhoods, further emphasizing that neighborhood characteristics are important to consider when addressing abandonment.
In sum, to create effective policies, the scale or scales at which the problem of interest operates should be identified. Although it is useful to create separate models to examine scale, this article demonstrates that true multilevel models are the preferred method. Failure to use multilevel models when the data are nested propagates the notion that the process of interest works the same way in different contexts—in this case neighborhoods—which is likely not true (Luke, 2004). In addition, the multilevel models in this article identify cross-level interaction effects. This shows how, had the nested nature of the data been ignored, different conclusions would have been reached. If a problem is ultimately a neighborhood-level problem, but policies are implemented at the house level, for example, it seems likely that the impact of the policies would be diluted at best. Further research is necessary to confirm this theory of spatial mismatch and policy ineffectiveness.
Author Victoria C. Morckel is an assistant professor of urban planning in the Earth and Resource Science Department at the University of Michigan-Flint.
References Field, Andy P. 2009. Discovering Statistics Using SPSS: (And Sex and Drugs and Rock ‘n’ Roll), 3rd ed.
London, United Kingdom: SAGE Publications.
Luke, Douglas A. 2004. Multilevel Modeling. Thousand Oaks, CA: SAGE Publications.
Morckel, Victoria C. 2014. “Spatial Characteristics of Housing Abandonment,” Applied Geography 48: 8–16. DOI: 10.1016/j.apgeog.2014.01.001.
———. 2013. “Empty Neighborhoods: Using Constructs To Predict the Probability of Housing Abandonment,” Housing Policy Debate 23 (3): 469–496. DOI: 10.1080/10511482.2013.788051.
O’Connell, Ann, Jessica Goldstein, H. Jane Rogers, and C.Y. Joanne Peng. 2008. Multilevel Logistic
Models for Dichotomous and Ordinal Data: Multilevel Modeling of Educational Data. Charlotte, NC:
Information Age Publishing.
Abstract The need for downtown revitalization is a growing concern for community stakeholders who are attempting to make their communities more sustainable and minimize urban sprawl. One strategy to make the downtown more active is to increase the attractiveness of the downtown for street-level customers and residential development. Success in this strategy attracts more people to the downtown; however, the challenge is to provide adequate parking. This study examines parking and its spatial dimensions in downtown Laramie, Wyoming. A parking inventory of both on- and off-street parking revealed the uneven spatial distribution of parking in the downtown area. Street interviews provided information on length of parking, purposes for coming downtown, and the location of destinations once downtown. A three-dimensional land use inventory supplied detailed locations of all activities in each building and floor for the 28 blocks of downtown Laramie. A bubble analysis of each parking space identified the spatial dynamics of the downtown parking demand and its distributional inadequacy for downtown residents.
Introduction The downtown of most cities is considered the heart of the community. Not only does the downtown have a substantial concentration of businesses and employment, it is also the cultural and social center of the community, with museums; historic sites; theatres; and social events such as festivals, parades, and ceremonies. The involvement of the people makes the downtown area a thriving pulse of the community. Wilson et al. (2012) examined the patterns of population change in metropolitan and micropolitan areas and found that metropolitan areas generally grew the fastest between 2000 and 2010. Along with this finding, Wilson et al. (2012) also concluded that downtown areas in the metropolitan counties had some of the fastest growth rates; for example, Chicago increased by 48,000 people within 2 miles of City Hall (the U.S. Census Bureau-designated center of the downtown area). Small towns, however, are experiencing some of the same types Cityscape 71 Cityscape: A Journal of Policy Development and Research • Volume 17, Number 1 • 2015 U.S. Department of Housing and Urban Development • Office of Policy Development and Research Gribb of population increases. The Census Bureau identified San Marcos, Texas, as one of the fastest growing communities in the United States, with a population increase of more than 20 percent between 2010 and 2013 (U.S. Census Bureau, 2014). Greenfield (2012) similarly identified that small towns are growing across the United States. The downtown areas are becoming the hallmark of regrowth and the core of the city (Glaeser, 2012). The trends in redevelopment of downtowns have been an ongoing process for the past 60 years, starting with the urban renewal projects in the early 1960s. Robertson (1999) identified a number of strategies to revitalize downtowns, specifically for small towns. In a study of 57 small-town development strategies, the following 9 strategies were identified by most of the communities: (1) historic preservation, (2) downtown housing, (3) waterfront development along with nightlife and entertainment, (4) new office development, (5) pedestrian improvements, (6) tourism, (7) traffic circulation changes, (8) Main Street approach, and (9) parking facilities and a convention center (Robertson, 1999). Several other studies demonstrated that it is imperative to reinvigorate the downtowns in communities (Faulk, 2006; Filion et al., 2004;
Leinberger, 2005; Rypkema, 2003). The development of the downtown area provides a number of challenges for the local community, depending on the strategies it pursues.
A vibrant downtown is marked by mixed-use activities and a sense of place. These two characteristics are part of the increase in residential activities in downtowns (Birch, 2009; Cook and Bentley, 1986). The higher population densities in the downtown provide a potential market capture for retail, entertainment, and cultural activities (Ferguson, 2005). Wachs (2013: 1162) found that “[y]oung, highly educated professionals move downtown to consciously reject the suburban cul de sacs where they grew up. Millions of senior citizens of means are choosing to retire in central city locations increasingly served by Starbucks, Whole Foods, and Trader Joe’s markets.” In an earlier study by Filion et al. (2004) on the revitalized downtown areas, however, the most successful areas had several elements in common: university campus nearby, seat of government, and historical character. If the community does not have these characteristics, however, the downtown can be a central place for employment and provide housing options for the local citizens. Wachs (2013) believes a number of downtown development activities are enhanced by the importance of transportation connectivity. As important as connectivity is as an element in downtown development, the availability of parking for both customers and residents is even more critical. The American Planning Association (APA) report on off-street parking (Bergman, 1991: ii) states that “…there is tremendous citizen concern about the availability of parking, its effect on the transportation network, and, ultimately, on the quality of life in a community.” With increasing population growth in downtown areas, the impetus to revitalize the downtown, and the concerns for parking and transportation, several factors need to be analyzed.
Objectives and Approach This study examines downtown residential land use and its demand on parking. In the process of analyzing downtown residential parking demand, this research project has several objectives.
1. Locate and inventory all land uses in downtown Laramie, Wyoming.
2. Locate and inventory downtown Laramie’s on- and off-street parking.
72 Urban Problems and Spatial Methods 3-D Residential Land Use and Downtown Parking: An Analysis of Demand Index
3. Create a spatial residential parking demand model based on the land uses within a set distance from each parking area.
4. Identify transportation and parking strategies that promote downtown residential development.
Unlike most land use studies and parking demand analyses, however, this study uses a different approach. First, the land use downtown is inventoried and analyzed using three-dimensional (3-D) spatial referencing. Each building downtown is inventoried floor by floor to record all land uses on each floor and their relative location on the floor. In the past, land use was recorded only for the first floor or a total count of land use was identified for a whole building without any spatial reference. Second, unlike most parking studies, which analyze the demand for parking based on the land use (ITE, 2010), the count of currently available off-street parking spaces, and the count of additional spaces needed to accommodate the new land use, this study assumes that the number of downtown parking spaces is fixed and that the probability of creating new spaces is low to none.
Thus, this research attempts to calculate the parking demand generated by the land uses around each individual parking space and views demand from the parking space perspective, not from the land use perspective. This study is specifically concerned about downtown residential parking, its availability, and demand competition.
Parking demands generally are based on the zoning and the amount of parking required for each land use type within the zone. The Institute of Transportation Engineers (ITE) created a guide (2010) that presents the parking demand for more than 105 different land uses. Most transportation engineers, consultants, and planners use this guide to determine parking demands. The guide, however, was developed from studies of isolated land uses in suburban areas (ITE, 2010).
To represent the full range of land areas, the fourth edition identifies five different area types:
(1) central business district, (2) central city (not including the central business district), (3) suburban centers, (4) suburban, and (5) rural (ITE, 2010). The demand model used in this analysis used the available information only for the central business district or central city uses.
Parking Demands in Downtown Laramie, Wyoming This study is concerned with residential parking demands in the Downtown Commercial (DC) zone district of Laramie, Wyoming. The DC zone district encompasses 25 blocks covering 29.6 hectares (73.1 acres; exhibit 1). Laramie is the third largest city in Wyoming, with an estimated 2013 population of 31,814 (http://quickfacts.census.gov). The home of the University of Wyoming (UW), Laramie has a fluctuating population but one of the most stable economies in a state noted for its boom-bust cycles of energy development. The city is attempting to encourage growth in the downtown area. In the City of Laramie’s Comprehensive Plan (2007), a major goal for the downtown area is— Increase residential population in the Downtown through changes to the current zoning regulations to encourage mixed-use buildings and upper floor rental or condominium units.
(City of Laramie, 2007: chapter 7, page 9) Thus, an emphasis in the plan is to restructure planning policies and governmental regulations to lessen the barriers to downtown residential development.
Note: The inset shows the location of Laramie on a map of Wyoming.