«Moving to opportunity voluMe 14, nuMber 2 • 2012 U.S. Department of Housing and Urban Development | Office of Policy Development and Research ...»
Detective and specialized units operate citywide or by grouped (that is, multiple districts) geographical distributions. Each district has between 9 and 15 beats, each staffed by one or two police officers 24 hours a day. The beat is both a unit of analysis and a response unit. Workload variations based on the time of day require additional patrol units within a district.
Chicago’s district is equivalent to the New York Police Department’s precinct and the Los Angeles Police Department’s division.
That is, they need to use Geographic Information Systems.
Static geography used for responses and reporting can now be made more dynamic because of near-realtime information about police workloads and community needs. The beat may be passé.
Incorporating geographic data when making resource deployment decisions enables the police to become more responsive to each neighborhood’s particular needs. This approach is a useful component of intelligence-led policing (Ratcliffe, 2008), a broad, strategic approach to making deployment decisions for the provision of public safety.
Still, police continue to struggle with decisions of resource deployment based on need. I would suggest that the police listen to both researchers and citizens when trying to understand what elements define a neighborhood. The differences in definitions of neighborhood are not in conflict, but are rather the same landscape viewed through different lenses.
Acknowledgments This article is based on an article that originally appeared in Geography & Public Safety Bulletin, Vol. II, Issue II, Neighborhoods (Buslik, 2009).
Author Marc S. Buslik is a captain with the Chicago Police Department and an adjunct professor at the University of Illinois at Chicago.
References Buslik, Marc S. 2009. “Not in My Neighborhood: An Essay on Policing Place,” Geography & Public Safety Bulletin 2 (2): 3–6. Also available at http://www.nij.gov/nij/maps/gps-bulletin-v2i2.pdf.
Guerry, Andre-Michel. 1833. Essay on the Moral Statistics of France: A Sociological Report to the French Academy of Science. Translated by Hugh P. Whitt and Victor W. Reinking. 2002. Lewiston, NY: Edwin Mellen Press.
Liberman, Akiva. 2007. Adolescents, Neighborhoods, and Violence: Recent Findings From the Project on Human Development in Chicago Neighborhoods. NIJ Research in Brief, NCJ 217397. Washington, DC: U.S. Department of Justice, National Institute of Justice.
Maltz, Michael D. 1995. “Criminality in Space and Time: Life Course Analysis and the MicroEcology of Crime.” In Crime and Place, edited by John E. Eck and David Weisburd. Monsey, NY:
Willow Tree Press, Criminal Justice Press: 315–348.
Maltz, Michael D., Andrew C. Gordon, and Warren Friedman. 1991. Mapping Crime in Its Community Setting: Event Geography Analysis. With contributions by Marc Buslik, Robert K.
LeBailley, Paul Schnorr, Douglas R. Thomson, and John P. Walsh. New York: Springer-Verlag.
Quetelet, Adolphe J. 1831. Research on the Propensity for Crime at Different Ages. Translated by Sawyer F. Sylvester. 1984. Cincinnati: Anderson Publishing Company.
Ratcliffe, Jerry. 2008. Intelligence-Led Policing. Cullompton, United Kingdom: Willan Publishing.
Sampson, Robert J., Stephen W. Raudenbush, and Felton Earls. 1995. “Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy,” Science 277 (5328): 918–924.
Shaw, Clifford R., and Henry D. McKay. 1942. Juvenile Delinquency and Urban Areas: A Study of Rates of Delinquents in Relation to Differential Characteristics of Local Communities in American Cities.
Chicago: University of Chicago Press.
Thrasher, Frederic. 1927. The Gang. Chicago: University of Chicago Press.
Weisburd, David. 2008. Place-Based Policing: Ideas in American Policing. Washington, DC: Police Foundation.
Additional Reading Cohen, Lawrence E., and Marcus Felson. 1979. “Social Change and Crime Rate Trends: A Routine Activities Approach,” American Sociological Review 44: 588–605.
242 Point of Contention: Defining Neighborhoods Geographic Patterns of Serious Mortgage Delinquency: CrossMSA Comparisons Lariece M. Brown Federal Deposit Insurance Corporation Hui-Chin Chen Melissa T. Narragon Freddie Mac Paul S. Calem Federal Reserve System The views expressed in this article are those of the authors and do not necessarily reflect those of the Federal Deposit Insurance Corporation, the Federal Housing Finance Agency, the Federal Reserve Board, or Freddie Mac.
Abstract This article examines the distribution of impaired mortgages across neighborhoods, defined at the ZIP Code level, in 91 metropolitan areas as of the fourth quarter of 2008, well into the recent U.S. mortgage crisis. We catalogue serious mortgage delinquency patterns by metropolitan area based on features of the geographic distribution, including measures of dispersion across neighborhoods and of spatial autocorrelation. The findings are potentially informative for assessing local and neighborhood consequences of the mortgage crisis and for selecting and implementing strategies to ameliorate the effects of foreclosure.
Introduction The tremendous volume of mortgage delinquencies and foreclosures since 2007 is an ongoing national crisis, but fashioning an appropriate policy or private-sector response requires assessing the local manifestations of the crisis. That the appropriate response depends on the neighborhood distribution of seriously delinquent mortgages in a metropolitan area—the extent to which such mortgages are concentrated in high-foreclosure neighborhoods and whether the latter are sparse or numerous, and are clustered together, dispersed, or isolated—has become increasingly clear.
For example, Goldstein (2010) introduced a data-based tool labeled “Market Value Analysis” that can be used to target public-sector and nonprofit neighborhood stabilization funds.1 The author emphasized that “targeting places where the problem is manageable and the surrounding markets have strength is critical to success” (Goldstein, 2010: 73). An illustrative application to the city of Philadelphia identified neighborhoods where vacancy and foreclosure were geographically confined so that interventions are likely to succeed.
This article surveys and classifies the variety of spatial patterns of serious delinquency observed across U.S. metropolitan areas. The article’s primary objectives are to highlight important differences in the spatial distribution of mortgage delinquency across metropolitan areas and to promote discussion of what public- and private-sector strategies are most suitable in each context. In particular, our typology may facilitate information sharing among cities with similar circumstances.
Secondarily, the article examines some housing market and economic conditions associated with the different spatial patterns. Although overall delinquency rates are highest in cities with large house price declines or high unemployment rates, this examination highlights how most other cities have high-delinquency pockets, mostly because of subprime lending concentrations.
Specifically, this article examines the mortgage delinquency distribution across neighborhoods, defined at the ZIP Code level, within U.S. metropolitan statistical areas (MSAs) as of the fourth
quarter of 2008, well into the mortgage crisis. The results classify metropolitan areas into six groups:
1. Low-to-moderate mean and high spatial autocorrelation: a modest number of high- or moderately high-delinquency neighborhoods that are clustered together.
2. High mean and standard deviation: wide variation across neighborhoods, with most delinquencies occurring in distressed neighborhoods.
3. High positive skewness: mostly multiple high-delinquency neighborhoods, some with extremely high delinquency rates.
4. Low-to-moderate mean, high positive skewness, and steep gradient around the peak delinquency neighborhood: a modest number of neighborhoods distinguished by high delinquency rates, including at least one spatial outlier.
The analytical approach constructs a set of neighborhood indicators, such as foreclosure and vacancy rates, assessed at the census block-group level, and uses them to cluster neighborhoods into categories reflecting dimensionality and degree of distress.
244 Refereed Papers Geographic Patterns of Serious Mortgage Delinquency: Cross-MSA Comparisons
5. Steep gradient around the peak delinquency neighborhood, indicating at least one spatial outlier: in general, isolated problem neighborhoods.
6. All other cities: somewhat more varied, but generally exhibiting moderate mean and low-tomoderate standard deviation of spatial delinquency.
This article contributes to a developing literature analyzing foreclosures and REO (Real Estate Owned) properties from a geographic perspective and deriving implications for neighborhood stabilization strategy.2 Immergluck (2009) classified metropolitan areas based on the level and change in density of REO properties from 2006 through 2008 and compared REO accumulation across
central city and suburban locations. The analysis highlighted three types of metropolitan areas:
(1) areas with low-to-moderate initial REO densities and stable prices, (2) those with initially high REO density and either stable prices or declines in value and increases in REO density from 2006 through 2008, and (3) “boom and bust” areas characterized by steep declines in home values accompanied by rising REO density over this period. The latter category tended to have higher REO concentrations in suburban areas.3 The author emphasized that “understanding the accumulation of REO inventories across and within metropolitan areas is important for formulating policies and informing community development practice regarding how to stabilize communities and neighborhoods that have been affected by surging foreclosures and vacant properties” (Immergluck, 2009: 28).4 Immergluck (2010a) revisited the subject, drilling down to the neighborhood (ZIP Code) level to investigate factors affecting REO accumulation from 2006 through 2008. The analysis indicated that the locations of high-risk lending activity and rapid housing development explain most of the urban-versus-suburban distribution of REO accumulation across metropolitan areas.5 Edmiston (2009) examined factors associated with foreclosure rate differences across census tracts within the 10th Federal Reserve District as of year-end 2008.6 The analysis found that concentrations of foreclosures in lower income areas are explained by concentrations of subprime mortgages.
The analysis in this article proceeds as follows. We first calculate distributional moments of the ZIP Code-level delinquency rates, and several measures of their spatial distribution across ZIP Codes. We next conduct a cluster analysis (using the principal component measures) to determine metropolitan area groupings based on common geographical patterns. Finally, we conduct a principal components regression analysis, exploring the relationship of these distributional moments and spatial measures (reduced to their principal components) to subprime lending patterns and economic factors.
REO properties are those that have been acquired by lenders via foreclosure.
The analysis also indicates that among suburban ZIP Codes, those with long commute times experienced larger REO increases over the November 2006-to-2008 period than those with shorter commute times.
For instance, the paper suggests, as an implication of disproportionate REO shares in ZIP Codes with long commute times, that “it may be unwise to spend scarce resources attempting to redevelop residential patterns that may not be highly sustainable in the context of more conservative mortgage markets or higher long-term energy and transportation costs” (Immergluck, 2009: 28).
Immergluck (2010b) examined both levels of and changes in REO activity from August 2006 through August 2008 across metropolitan areas, particularly in relation to changes in home values and the legal environment affecting foreclosures.
The 10th Federal Reserve District consists of Colorado, Kansas, Nebraska, Oklahoma, Wyoming, and parts of western Missouri and northern New Mexico.
The article is organized correspondingly. The next section describes our data sources. The section on geospatial characterization follows with the calculation of the distributional moments and spatial measures and their principal components. The section following that presents the cluster analysis, emphasizing the implications of the results for developing appropriate policy responses.
The principal component regression analysis precedes the concluding section.
Data Sources We draw data for the study from several sources. We obtain estimates of prime and subprime mortgage delinquency rates as of October 2008 by ZIP Code, using the CoreLogic TrueStandings Servicing® online data analytics tool.7 This online business intelligence platform accesses the prime and subprime mortgage databases of CoreLogic. These databases provide current information on the payment status of active mortgages serviced by the top mortgage-servicing institutions or sold to Fannie Mae or Freddie Mac.8 Historical information for both paid-off and active loans is also available, by origination month, as are the state, county, and ZIP Code location of the financed property. We restrict our attention to first-lien, conventional mortgages. For this article, we define delinquency as 60 or more days past due.9 The CoreLogic databases do not provide a full count of all active mortgage loans in all ZIP Codes, because not all institutions that service mortgages contribute to these databases. Therefore, we adjust the active loan counts from the CoreLogic servicing data based on an estimate of the undercount in each ZIP Code.10 Specifically, we measure the undercount by comparing the number of 2005 and 2006 mortgage originations in the CoreLogic data against the number reported to federal regulatory authorities in Home Mortgage Disclosure Act (HMDA) data. The procedure is discussed in greater detail in the following section.
Information about TrueStandings Servicing® is available at http://www.corelogic.com.