CensusCD 1990 Long Form in 2010 Boundaries allows users to access US Census data from 1990 and easily compare it with the 2010 Census data. It is the finest source of Census data from 1990 expressed at all of the 2010 ACS geographies. The CensusCD 1990 Long Form in 2010 Boundaries is based upon the long form (STF 3): questions answered by one in six households in the 1990 census. It includes detailed demographic data about topics such as population, household structure, income, poverty, education level, employment, housing costs, immigration, and other variables. The 1990 Long Form in 2000 Boundaries is an invaluable resource for policy makers, community organizations, and researchers who want to analyze the changes that have occurred in the U.S. from 1990 to 2010.
Why Normalize 1990 Data to 2010 Geographic Boundaries?
Geographic boundaries change from census to census based on many factors. In the example below Williamson County, Texas is shown
broken-out at the census tract level. At a glance, one can observe that there are changes in the census tracts from 1990 to 2010. In fact, Williamson County split tracts into multiple parts, with the overall number of tracts growing from 64 to 89 tracts.
Williamson County, TX
Total Census Tracts (64)
Williamson County, TX
Total Census Tracts (89)
Taking a closer look at Pima County from 1990 to 2000 reveals the modifications in tract boundaries that significantly impact how researchers study and account for changes in a population over time. For example, below is a single 1990 Tract # 020898 that was broken into 7 tracts in the 2010 Census. It is a straight forward 1 to 7 correspondence.
|Williamson County, Texas
1990 Tract Boundary
2010 Census Tract Boundaries
However, the changing tract boundaries from 1990 to 2010 are not always that straight forward. In the next example from Williamson County, there are three tracts, # 020101, 020102 & 020200, which change into 11 new tracts in 2010. The correspondence in this case is 3 to 11. From these two examples in Williamson County, it becomes evident why it is so difficult for researchers to contend with the issue of changing geographic boundaries over time.
1990 Tract ID# 004024 & 004028
2000 Census Tract Boundaries
Normalizing the Data
The actual re-mapping procedure for converting data from 1990 to 2010 boundaries is quite complicated.
Here is a link for those wishing a more technical explanation of this task (click here). The basic procedure
was to use the TIGER/Line 2010 relationship between 2010 blocks with those from 2000 and then from
2000 to those in 1990. This allowed us to identify how block boundaries had changed between censuses.
For example, if a 1990 block split into two for the 2000 census, others tracts merged, while some tracts
both merged and then split. The table below shows the three scenarios (many to one, one to many, and
many to many).
There are over 11 million blocks in the US in 2010, so in fact these are very small geographic areas.
But in order to be more precise, when necessary we broke blocks down. When the blocks split then the
matter of how to split the population becomes a problem. (When they merge you just add the numbers).
In order to determine how to subdivide blocks we looked at the TIGER Street files. The assumption being
that people live on or near streets, so the number of addresses on a street will indicate the approximate
weight to give to that area of the block.
From the blocks or block parts we created a Block Weighting File. These population weights were then
applied to the various other counts to convert them from 1990 to 2000 and then from 2000 to 2010 block
boundaries. Once the data had been calculated at the block level we were then able to sum up the
blocks to the various other geographies. Testing was then done to assure the accuracy and validity of the
weighting method compared with the original numbers from the 1990 Long Form data set.