Abstract
Studies of neighborhood change rely on interpolated data to cope with inconsistent boundaries of geographic units over time. The standard approach introduces error by assuming, counterfactually, that all kinds of people are distributed in the same manner within tracts as the whole population. This study evaluates estimates of 2,000 neighborhood characteristics using 2010 boundaries in the Longitudinal Tract Data Base (LTDB) that uses the standard approach, and an alternative trait-based (TB) method that uses additional small area data to account for spatial heterogeneity. Both are compared to the true (but confidential) original census data. For variables that are available from full-count census data at the block level (including race, age, and some housing characteristics), the TB estimates are much better than the LTDB estimates. The same general approach is ineffective, however, when the small area data are subject to sampling variability and published with less spatial granularity.
研究社区变化, 需要采用数据插值来处理地理单元边界因时间变化而导致的不一致。常用方法假设人口普查区内各类人群有相同的分布模式, 从而带来了错误。本文使用纵向普查区数据库(LTDB)的2010年边界, 评估了2,000个社区特征的估计值。LTDB采用了标准方法和替代性的基于特征(TB)方法。TB方法利用额外的小区域数据来解释空间异质性。两种方法都与真实(但保密)原始人口普查数据进行了比较。对于街区层面的完整人口普查数据变量(种族、年龄和某些住房特征), TB估计值远优于LTDB估计值。然而, 当小区域数据受到采样可变性影响或以较低空间颗粒度发布时, 相同的方法无效。
Los estudios sobre transformación barrial dependen de datos interpolados para hacer frente a la inconsistencia en los límites de unidades geográficas a través del tiempo. El enfoque estándar lleva a error al asumir, en contra de los hechos, que todo tipo de gente se distribuye de la misma manera dentro de los tractos que el conjunto de la población. Este estudio evalúa los estimativos de 2.000 características barriales usando los límites del 2010 en la Base Longitudinal de Datos por Tracto (LTDB), que usa el enfoque estándar, y un método alternativo basado en rasgos (TB) que usa datos adicionales de áreas pequeñas para tener en cuenta la heterogeneidad espacial. Ambos se comparan con los datos censales originales verdaderos (aunque confidenciales). Sobre las variables que están disponibles a partir de los datos censales completos a nivel de manzana (incluyendo raza, edad y algunas características de la vivienda), los cálculos TB son mucho mejores que los estimativos de LTDB. No obstante, el mismo planteamiento general no es efectivo, cuando los datos de área pequeñas se sujetan a la variabilidad del muestreo y se publican con menos granulidad espacial.
Disclosure Statement
No potential conflict of interest was reported by the authors.
Additional information
Funding
Notes on contributors
John R. Logan
JOHN R. LOGAN is Professor of Sociology at Brown University, Providence, RI 02910. E-mail: [email protected]. He directs the American Communities Project, which includes the Longitudinal Tract Data Base (LTDB). His recent research includes studies of neighborhood change and spatial inequality in contemporary U.S. cities and in the late nineteenth and twentieth centuries.
Wenquan Zhang
WENQUAN ZHANG is Associate Professor of Sociology at the University of Wisconsin, Whitewater, WI 53190. E-mail: [email protected]. He is conducting research on the changing racial composition of neighborhoods in U.S. cities, including the phenomenon of especially diverse global neighborhoods, and he is using confidential census data to examine movements in and out of these neighborhoods.
Zengwang Xu
ZENGWANG XU is Associate Professor of Geography at the University of Wisconsin, Milwaukee, WI 53201. E-mail: [email protected]. He collaborated in development of the original LTDB. His ongoing projects include studies on neighborhood effects on birth outcome and violence incidents in City of Milwaukee, epidemic diffusion along social contact networks, population migration, and effects of hurricane damage on populations along the Gulf and Atlantic coasts.