آزمون ماهیت زمانی اختلال اجتماعی از طریق ساختمان های متروکه و فضاهای میان بافتی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|37410||2015||18 صفحه PDF||سفارش دهید||14080 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Social Science Research, Volume 54, November 2015, Pages 177–194
Abstract With the recent housing crisis, studying abandoned buildings has once again become important. However, it has been some time since abandoned buildings were the subject of direct study, leaving scholars with scant knowledge about the characteristics of abandoned buildings, how they change, and their relationship to neighborhood processes. To fill this gap, we employed longitudinal photographic and SSO evaluations of 36 abandoned buildings and their immediate surroundings in Chicago for one year (n = 587). Results demonstrate the presence and severity of social disorder cues vary across time points and the time of day of observation. There is a relationship between abandoned buildings and social disorder, though the relationship is not a trend. Also, social disorder is diminished around extremely decayed buildings. Lastly, we find that our results are driven by the measurement of places ignored by most SSO studies, including alleys and the rear side of buildings.
Introduction Do abandoned buildings serve as contagion points for disorder? While derelict buildings have long been seen as blight on neighborhoods, the recent housing crisis and the long decline of Rust Belt cities, like Cleveland and Detroit, have reanimated this longstanding question in sociological and criminological research. Abandoned buildings2 are considered signs of disorder, evidence of a neglected urban landscape, and a harbinger of the slow decline of neighborhoods (Skogan, 1990, Sampson and Raudenbush, 1999 and Hunter, 1985). Furthermore, languishing abandoned buildings damage property values and investment in communities (Immergluck, 2012 and Immergluck and Smith, 2006), and property investors regularly use signs of disorder, including abandoned buildings, to judge if buying a property is a good investment (Immergluck, 2012). Moreover, abandoned buildings are seen as hot spots and contagions for crime since they provide privacy and little guardianship (Spelman, 1993 and Eck and Spelman, 1987). As such, they can be used for prostitution, drug dealing and using, squatting, and other illegal activities like theft (Eck and Spelman, 1987, Ouellet et al., 1991 and Spelman, 1993). It is surprising then, that given the centrality of abandoned buildings in neighborhood processes and crime, we have little recent scholarly information about them. Abandoned buildings are unlikely to be static, though most studies of abandoned buildings only monitor the change in crime and disorder around buildings (Spelman, 1993 and Eck and Spelman, 1987), not changes to the building. Not only are buildings dynamic, one abandoned building is not like the other: some buildings possess few signs of dereliction, while others possess overwhelming signs of damage. When measured “objectively,”3 abandoned buildings are typically quantified as a binary variable: being present or not. Measuring abandoned buildings in this manner ignores the several dimensions of variability inherent in buildings, such as how severe the structural damage to the building is or whether or not the building is unsecured. Once again, an in-depth analysis of abandoned buildings, what they are like, and how they change over time, is wanted. How do abandoned buildings change? Do they play a role in the generation of additional disorder? Our study aims to address these questions by conducting a longitudinal photographic investigation of abandoned buildings where we not only include the front of buildings but venture into alleyways to document their backs. This enables us to examine changes to the buildings and the disorder occurring both on and around them. Our results will shed light on how disorder cues—particularly physical disorder cues—change both over time and in the severity of their condition. While disorder is generally thought of as temporal in nature (Hipp, 2007), little research documents how disorder changes over time, and there are no studies that examine disorder’s flux across short timelines as does our study. Here, our work is the first to document the timeline surrounding the trajectory of disorder; as such, we will be able to ascertain whether disorder is generated over the course of a week or if it has a longer time line, closer to a year. Furthermore, our study will examine how variability in the condition or severity of a disorder cue is related to later social disorder. This is not only important for disorder studies, but also urban studies and planning research interested in the relationship between disorder, vacancy, crime, and other neighborhood problems. Without comprehending the trajectory of disorder, understanding the relationship between neighborhood processes and disorder, abandoned buildings, and vacant properties is difficult. The implications of this work will assist cities in understanding the problems and dynamic nature of abandoned buildings. Many cities, especially those engaging in community or “hot spot” policing, take great efforts to reduce the number and detrimental effects of abandoned buildings (Braga and Bond, 2008, Skogan, 1990 and Spelman, 1993), citing crime, budgetary drains, and a low likelihood of being returned to productive use as reasons for abatement and demolition (United States Conference of US Mayors, 2006). Only a few studies have examined the effect of abandoned buildings on neighborhood processes (see Spelman, 1993 and Skogan et al., 2004), leaving a large gap between rigorous study of abandonment and the housing crisis of 2007. Our research fills this important information void for cities and their emergency services, like police and fire departments.
نتیجه گیری انگلیسی
Results Before discussing the results from the models, we begin by examining our summary statistics across several time points. Note that all the variables, with the exception of the time of day variables, are standardized. First, while the overall physical state of the building does not see much variability, both the front and back of the building see a diminished physical state in time points 6 and 12, with the back of the building being most severe. Building specific disorder, whether overall, front, or back of the building, sees a general decreasing trend with a slight uptick at time point 18. While the changes in the front of the building are moderate, the changes to building specific disorder at the back of the building are more drastic. Social disorder—the dependent variable in the coming analyses—also has variability over time. While at time point 1 it is below average, by time point 18, it is well above average, with rises and falls across time points. The variability in these summary statistics suggests that (1) disorder cues do vary over time, both in the long view and the short view, and (2) disorder cues, like abandoned buildings, can vary in their severity. Table 3 shows the results from the cross sectional models examining the relationship between the physical state of the building, building specific disorder, and social disorder. Model 1 used the overall physical state of the building and building-specific disorder to predict social disorder, while models 2 and 3 used the physical state of the front and back of the building, as well as the front and back of building-specific disorder to predict social disorder. In Model 1, we see that overall physical state of the building does not significantly affect the social disorder surrounding the building, but building specific disorder does (p < 0.05). Remembering that the outcome and the scales are standardized, every one standard deviation increase in building specific disorder is associated with a 0.02 standard deviation reduction in social disorder. We see similar results in Model 2, which includes only the physical state of the front of the building and building specific disorder scales, however the effect is only significant at p < 0.1. In Model 3, which includes the back of the building scales, building specific disorder is not significant. Here the effect is positive, suggesting the more problematic the physical state of the back of the building, social disorder around that building is also highly problematic. Finally, we ran likelihood-ratio tests to assess if the overall building was significantly different from either the front or back of the building models. The tests were significant, demonstrating that the models are indeed different. Table 3. Fixed effect linear regression models predicting social disorder. 1 2 3 b SE b SE b SE Overall physical state of the building 0.011 (0.017) Overall building specific disorder −0.023⁎ (0.012) Front: physical state of the building 0.003 (0.016) Front of the building specific disorder −0.023+ (0.013) Back: physical state of the building 0.024 (0.013) Back of the building specific disorder −0.016 (0.011) Time 0.006⁎⁎ (0.002) 0.006⁎⁎ (0.002) 0.006⁎⁎ (0.002) Photo taken mid-morning 0.127⁎⁎ (0.046) 0.121⁎⁎ (0.045) 0.127⁎⁎ (0.045) Photo taken early evening 0.194⁎⁎ (0.046) 0.187⁎⁎ (0.046) 0.195⁎⁎ (0.046) Photo taken mid-evening 0.150⁎⁎ (0.049) 0.145⁎⁎ (0.049) 0.149⁎⁎ (0.049) Constant −0.119⁎ (0.049) −0.113⁎ (0.048) −0.121⁎ (0.047) Observations 587 587 587 R-squared 0.944 0.943 0.944 Buildings 36 36 36 Standard errors in parentheses. ⁎ p < 0.05. ⁎⁎ p < 0.01. Table options Next, because of their consistent effects across the three models shown in Table 3, we discuss the overall effect of the control variables. Variables for the time of day the photograph of the building was taken are significant; later times in the day have higher levels of disorder. This is expected given that other studies find that disorder to be more prevalent in the evening (Sampson and Raudenbush, 1999). Also significant is the linear control for time, showing social disorder around buildings is getting worse during the study period, regardless of building. 6.1. Longitudinal models The longitudinal models are presented in Table 4, Table 5 and Table 6 and take on a structure similar to that of Table 3. Beginning with Table 4, net of the controls, lagged social disorder is the most powerful predictor of current social disorder. Here, for every one standard deviation increase in lagged social disorder, we see a 0.11 standard deviation decrease in current social disorder. This finding was consistent across all models. The effect of overall building specific disorder remains significant and negative: for every one standard deviation increase in the overall building specific disorder, social disorder surrounding the building decreases by approximately 0.03 standard deviations. While the size of the coefficient is small, this lends credence to the idea that building specific disorder conveys something salient about the usability of the building and the space around it, especially since the timelines are short, ranging between a week and a month. Next, changes in either the physical state of the building or building specific disorder does not elicit a change in the social disorder surrounding the building. The models for the front of the building (Table 5) show very similar results and will not be discussed. Table 4. Fixed Effects Linear Regression Models Predicting Social Disorder with Changes in the Building. 1 2 3 b SE b SE b SE Lagged social disorder 0.114⁎⁎ (0.043) 0.110⁎ (0.043) 0.112⁎⁎ (0.043) Overall building condition 0.026 (0.022) 0.014 (0.018) 0.027 (0.022) Overall building specific disorder −0.025⁎ (0.012) −0.032⁎ (0.014) −0.033⁎ (0.014) Change in the physical state of the building −0.021 (0.025) −0.026 (0.026) Change in building specific disorder 0.016 (0.018) 0.019 (0.018) Time 0.005⁎ (0.002) 0.005⁎ (0.002) 0.005⁎ (0.002) Photo taken mid-morning 0.107⁎ (0.046) 0.110⁎ (0.046) 0.113⁎ (0.046) Photo taken early evening 0.203⁎⁎ (0.046) 0.206⁎⁎ (0.047) 0.209⁎⁎ (0.047) Photo taken mid-evening 0.134⁎⁎ (0.049) 0.137⁎⁎ (0.049) 0.141⁎⁎ (0.049) Constant −0.114⁎ (0.050) −0.112⁎ (0.050) −0.121⁎ (0.050) Observations 542 542 542 R-squared 0.944 0.948 0.948 Buildings 36 36 36 Standard errors in parentheses. ⁎ p < 0.05. ⁎⁎ p < 0.01. Table options Table 5. Fixed effects linear regression models predicting social disorder with changes in the front of the building. 1 2 3 b SE b SE b SE Lagged social disorder 0.118⁎⁎ (0.043) 0.113⁎⁎ (0.043) 0.115⁎⁎ (0.043) Front: physical state of the building 0.015 (0.020) 0.004 (0.016) 0.014 (0.020) Front: building-specific disorder −0.022 (0.013) −0.029 (0.015) −0.029 (0.015) Front: change physical state of the building −0.020 (0.026) −0.020 (0.026) Front: change in building-specific disorder 0.017 (0.020) 0.017 (0.020) Time 0.005⁎ (0.002) 0.005⁎ (0.002) 0.005⁎ (0.002) Photo taken mid-morning 0.099⁎ (0.045) 0.100⁎ (0.045) 0.102⁎ (0.045) Photo taken early evening 0.196⁎⁎ (0.046) 0.195⁎⁎ (0.046) 0.199⁎⁎ (0.046) Photo taken mid-evening 0.129⁎⁎ (0.048) 0.130⁎⁎ (0.049) 0.133⁎⁎ (0.049) Constant −0.107⁎ (0.050) −0.103⁎ (0.049) −0.111⁎ (0.050) Observations 542 542 542 R-squared 0.948 0.943 0.947 Buildings 36 36 36 Standard errors in parentheses. ⁎ p < 0.05. ⁎⁎ p < 0.01. Table options Table 6. Fixed effects linear regression models predicting social disorder with changes in the back of the building. 1 2 3 b SE b SE b SE Lagged social disorder 0.114⁎⁎ (0.043) 0.116⁎⁎ (0.043) 0.113⁎⁎ (0.043) Back: physical state of the building 0.032 (0.017) 0.024 (0.014) 0.031 (0.017) Back: building-specific disorder −0.019 (0.011) −0.026 (0.014) −0.026 (0.014) Back: change in physical state of the building −0.014 (0.019) −0.014 (0.019) Back: change in building-specific disorder 0.015 (0.017) 0.015 (0.017) Time 0.005⁎ (0.002) 0.005⁎ (0.002) 0.005⁎ (0.002) Photo taken mid-morning 0.107⁎ (0.045) 0.109⁎ (0.045) 0.110⁎ (0.045) Photo taken early evening 0.201⁎⁎ (0.046) 0.205⁎⁎ (0.046) 0.206⁎⁎ (0.046) Photo taken mid-evening 0.130⁎⁎ (0.048) 0.133⁎⁎ (0.048) 0.134⁎⁎ (0.048) Constant −0.110⁎ (0.048) −0.112⁎ (0.048) −0.115⁎ (0.048) Observations 542 542 542 R-squared 0.947 0.948 0.944 Buildings 36 36 36 Standard errors in parentheses. ⁎ p < 0.05. ⁎⁎ p < 0.01. Table options Lastly, the back of the building models in Table 6 show slightly different results. Like the previous models, net of the controls, the lagged social disorder variable is the best predictor of current social disorder around the building; note that the change variables are not significant. However, both the physical state of the back of building and building specific disorder are significant at p < 0.1. As before, building specific disorder has a negative impact on social disorder. Yet, the physical state of the back of building has a positive relationship with social disorder, meaning that as the physical state of the building worsens (reverse coded), social disorder around the building increases. Note that these effects approach significance at p < 0.1. The control variables shown in Table 4, Table 5 and Table 6 behaved similarly to those in Table 3. The variables for the time of day the photograph were significant, as well as the linear control for time.