Optimized spatial information for 1990, 2000, and 2010 U.S. census microdata.

We report on the successful completion of a project to upgrade the positional accuracy of every response to the 1990, 2000, and 2010 U.S. decennial censuses. The resulting data set, called Optimized Spatial Census Information Linked Across Time (OSCILAT), resides within the restricted-access data warehouse of the Federal Statistical Research Data Center (FSRDC) system where it is available for use with approval from the U.S. Census Bureau. OSCILAT greatly improves the accuracy and completeness of spatial information for older censuses conducted prior to major quality improvements undertaken by the Bureau. Our work enables more precise spatial and longitudinal analysis of census data and supports exact tabulations of census responses for arbitrary spatial units, including tabulating responses from 1990, 2000, and 2010 within 2020 block boundaries for precise measures of change over time for small geographic areas.

Across the Rural-Urban Universe: Two Continuous Indices of Urbanization for U.S. Census Microdata.

Microdata from U.S. decennial censuses and the American Community Survey are a key resource for social science and policy analysis, enabling researchers to investigate relationships among all reported characteristics for individual respondents and their households. To protect privacy, the Census Bureau restricts the detail of geographic information in public use microdata, and this complicates how researchers can investigate and account for variations across levels of urbanization when analyzing microdata. One option is to focus on metropolitan status, which can be determined exactly for most microdata records and approximated for others, but a binary metro/nonmetro classification is still coarse and limited on its own, emphasizing one aspect of rural-urban variation and discounting others. To address these issues, we compute two continuous indices for public use microdata-average tract density and average metro/micro-area population-using population-weighted geometric means. We show how these indices correspond to two key dimensions of urbanization-concentration and size-and we demonstrate their utility through an examination of disparities in poverty throughout the rural-urban universe. Poverty rates vary across settlement types in nonlinear ways: rates are lowest in moderately dense parts of major metro areas, and rates are higher in both low- and high-density areas, as well as in smaller commuting systems. Using the two indices also reveals that correlations between poverty and demographic characteristics vary considerably across settlement types. Both indices are now available for recent census microdata via IPUMS USA (https://usa.ipums.org).

Hybrid Areal Interpolation of Census Counts from 2000 Blocks to 2010 Geographies.

To measure population changes in areas where census unit boundaries do not align across time, a common approach is to interpolate data from one census's units to another's. This article presents a broad assessment of areal interpolation models for estimating counts of 2000 characteristics in 2010 census units throughout the United States. We interpolate from 2000 census block data using 4 types of ancillary data to guide interpolation: 2010 block densities, imperviousness data, road buffers, and water body polygons. We test 8 binary dasymetric (BD) models and 8 target-density weighting (TDW) models, each using a unique combination of the 4 ancillary data types, and derive 2 hybrid models that blend the best-performing BD and TDW models. The most accurate model is a hybrid that generally gives high weight to TDW (allocating 2000 data in proportion to 2010 densities) but gives increasing weight to a BD model (allocating data uniformly within developed land near roads) in proportion to the estimated 2000-2010 rate of change within each block. Although for most 2010 census units, this hybrid model's estimates differ little from the simplest model's estimates, there are still many areas where the estimates differ considerably. Estimates from the final model, along with lower and upper bounds for each estimate, are publicly available for over 1,000 population and housing characteristics at 10 geographic levels via the National Historical Geographic Information System (NHGIS - http://nhgis.org).

Because Muncie's Densities Are Not Manhattan's: Using Geographical Weighting in the EM Algorithm for Areal Interpolation.

Areal interpolation transforms data for a variable of interest from a set of source zones to estimate the same variable's distribution over a set of target zones. One common practice has been to guide interpolation by using ancillary control zones that are related to the variable of interest's spatial distribution. This guidance typically involves using source zone data to estimate the density of the variable of interest within each control zone. This article introduces a novel approach to density estimation, the geographically weighted expectation-maximization (GWEM) algorithm, which combines features of two previously used techniques, the expectation-maximization (EM) algorithm and geographically weighted regression. The EM algorithm provides a framework for incorporating proper constraints on data distributions, and using geographical weighting allows estimated control-zone density ratios to vary spatially. We assess the accuracy of GWEM by applying it with land-use/land-cover ancillary data to population counts from a nationwide sample of 1980 United States census tract pairs. We find that GWEM generally is more accurate in this setting than several previously studied methods. Because target-density weighting (TDW)-using 1970 tract densities to guide interpolation-outperforms GWEM in many cases, we also consider two GWEM-TDW hybrid approaches, and find them to improve estimates substantially.

Bicomponent Trend Maps: A Multivariate Approach to Visualizing Geographic Time Series.

The most straightforward approaches to temporal mapping cannot effectively illustrate all potentially significant aspects of spatio-temporal patterns across many regions and times. This paper introduces an alternative approach, bicomponent trend mapping, which employs a combination of principal component analysis and bivariate choropleth mapping to illustrate two distinct dimensions of long-term trend variations. The approach also employs a bicomponent trend matrix, a graphic that illustrates an array of typical trend types corresponding to different combinations of scores on two principal components. This matrix is useful not only as a legend for bicomponent trend maps but also as a general means of visualizing principal components. To demonstrate and assess the new approach, the paper focuses on the task of illustrating population trends from 1950 to 2000 in census tracts throughout major U.S. urban cores. In a single static display, bicomponent trend mapping is not able to depict as wide a variety of trend properties as some other multivariate mapping approaches, but it can make relationships among trend classes easier to interpret, and it offers some unique flexibility in classification that could be particularly useful in an interactive data exploration environment.