Census 2000 provided the most detailed demographic picture of the United States ever released by the Census Bureau.  Census geographies from states down to Census Blocks carry thousands of demographic data items for each unit of geography.

The Census 2000 view of the United States, as comprehensive as it is, still isn’t sufficient.  It provides a static picture.  As a decision-maker, you really need to know what your market was like in the past, what it looks like today post-Census 2000, and what changes are likely to take place in the future.  Devonshire, through its relationship with Scan/US, Inc., offers this updated information bringing a full picture of the past, present, and future of your markets to you.

Applying the Power of Census 2000 to the Update Process

The 2001–2006 Demographic Updates were based on the Census Bureau’s SF1 100% Survey data released in late Spring 2001.  The 2002–2007 Demographic Updates were the first to incorporate the Census 2000 SF3 Sample data released in Fall 2002.  The all important household income estimates and projections in the 2002–2007 update series were anchored to the detailed 2000 household income data made available for the first time in that Census 2000 SF3 release.  The annual estimates and projections continue to build on this solid foundation annually integrating extensive public and private data that point the way to the future.

The Contribution of the TIGER/Line Files

The term TIGER is the acronym for Topologically Integrated Geographic Encoding and Referencing which is the name for the system and digital database developed at the U.S. Census Bureau to support its mapping needs for the Decennial Census and other Bureau programs.  The TIGER/Line files are the digital database of geographic features, such as roads, railroads, rivers, lakes, legal boundaries, census statistical boundaries, etc. covering the entire United States.

The TIGER system was first used in support of the 1990 decennial census.  Therefore, the TIGER/Line files provide a digital history of United States cartography: roads, political boundaries, and census population geography, as captured by change-encoded linear features from 1990 to the current day.  TIGER will continue to capture this growth.

This means that, for the first time, the 1990 Census geographies with their demographic data could be mapped into the Census 2000 geographies by computer with unerring accuracy.  Even performing this transition at the smallest census geography, 1990 census blocks to 2000 census blocks, can be accomplished accurately by using the date of a street’s development to correctly weight the redistribution. Those 2000 blocks that do not contain street segments built before 1990 had no population in 1990.  The accuracy of assignment of 1990 demographics to the 2000 census geographies at the block level is essential to maintaining data integrity when summing up to the block groups.  This 1990 to 2000 demographic trend is used to inform the models that will estimate and project the demographic change for the 2000 block groups during the first decade of the 21st century.

Tracking Household Change Down to the Street Segment

The TIGER/Line files have also allowed a process to be implemented by which household change can be assigned to the street segment in the block where it occurs.  The process utilizes the TIGER segment address ranges, the United States Postal Service (USPS) carrier route drop counts, the ADVO ZIP+4 deliverable household counts, and the USPS ZIP+4 inventory for the same time period.  The integration of these source files through geocoding results in a derived household count for census blocks for the current time period.  In 2000, these block level derived household counts were generated and benchmarked against the Census 2000 block household counts.  Now, annually, the same process is gone through to generate block derived household counts for the current year.  The change between the 2000 baseline derived household counts and the current derived counts can then be translated into household change at the census block level.  By combining the 1990 to 2000 household change data with the 2000 to current year estimate of household change, an appropriate rate of change for the projection year can be calculated.  These block level estimates and projections then get rolled up into base estimates and projections of population and household change for the parent block groups.

Adding Demographic Detail to the Base Estimates and Projections

To these base estimate and projections of population and households for block groups, detailed demographic profiles are applied that turn the plain vanilla households into fully configured consuming units.  Determining the age, sex and race of the population and how those characteristics vary among households of different types (families and non-families, big and small, rich and poor) is the other essential ingredient of the demographic update process.  A top down, bottom up approach is used in this portion of the update process.

First, an extensive profile of demographic characteristics is developed for each county relying heavily on the information resources of the Census Bureau, Bureau of Labor Statistics and Regional Economic Information System data (REIS) of the Bureau of Economic Analysis.  The county level household income profile is based on a model of structural change in household money income, utilizing the most current “Money Income in the United States” data series and an econometric model of change in aggregate money income, using the income series from REIS.

Second, estimates for each block group’s share of the demographic profile of the county of which it is a part is developed.  A block group’s profile is derived using a series of life cycle models to merge the initial block group population and household estimates with the base-reference, detailed 2000 demographic profiles.  These life cycle models age the population, adjust the characteristics of the base households, estimate the characteristics of new households, and roll the household income distribution forward, consistent with measured and projected infrastructure change.

The third and final step of reconciling county level control demographics with the component block group distributions is performed with the aid of a proprietary matrix balancing routine, which preserves the control marginals while maximally maintaining the underlying block group profiles.  The result is a robust set of block group estimates that preserves the unique character of each block group, but at the same time, reflect the impact of demographic trends and specific county change.

Scan/US, Inc.’s demographic update methodology has evolved and has been continually refined over the last three decades.  The architects of the current update were the owners and founders of Urban Decision Systems, Inc. when in 1978 they published their first small area updates for the United States.  Today, they are committed to continuing to improve their small area demographic update.