Census Data

When deciding what demographic data product that you need for your analysis, there are 4 questions:

What years of data do you want to analyze?

We have different products and packages that cover 1960, 1970, 1980, 1990, 2000 and 2010, as well as current year estimates and 5 year projections. CENSUS 2000 LONG FORM (SF3) VETERANS AS A PERCENT OF TOTAL POPULATION

Do you want to compare data from one time to another?

Some of our products and packages cover data from multiple time periods. If you want to compare multiple datasets from different time periods, you must deal with changing boundary definitions. In our Normalized (weighted) datasets we have dealt with this issue by cross-walking all of the data to the same Census year 2000 or soon 2010 geographies. Therefore, you can study 1970, 1980, 1990, 2000 and 2010 data in the same geographic boundaries and easily compare them.

What type of data do you need?

Each product has a complete list of variables available for review on its product web page at www.GeoLytics.com. Below is an overview of the different census datasets:

• Long Form (SF3) - is gathered from approximately 1 out of 6 households. This is the most comprehensive demographic dataset and includes not only demographic variables such as population, race, age, and household composition, but also has specific socio-economic and housing variables such as income, poverty, housing value, mortgage, rent paid, employment, occupation, education, ancestry, language, commute to work, disabilities, veterans/military status, and much more.
• American Community Survey (ACS) – replaces the Long Form starting in 2010. This has most of the same variables as the SF3 but is released annually and covers only a fraction of the households covered in the SF3. It is comprised of 3 data sets: 1year estimates, 3 year estimates and 5 year estimates. To look at the data for the smaller geographies you need to access the 5 year estimtes.
• Short Form (SF1) - is gathered from every household and is the 100% count. It covers population, race (and Hispanic designation), age of children, presence of older relatives, family composition, household size, owning vs. renting, occupancy status, and group quarters. In the 2000 Short Form there are several key indicators that were NOT included in the Long Form data set: country of origin for Asians and Hispanics; specific breakouts for multiracial people; and details about Group Quarters (prison, military barracks, nursing homes, etc).
• Redistricting - this product has a very limited variable set (total population, race, and age breakouts). But, it contains more geographic selections than other products such as school districts and state voting districts. It is predominately used for its geographies.
• Block Products - these products contain Short Form or Redistricting datasets but all of the data is expressed at the BLOCK level only. As a cautionary note, there are 8.2 million blocks in the 2000 Census so this creates very large data files even when looking at small areas.

What geographies do you want to look at the data in?

For example, do you want to look at a whole county, a whole state, compare counties, or look at the whole nation? Your choices vary from Block level geography in our Block Products, to the Block Group, Tract, ZCTA (zip code), and more in our complete census products. To determine what geographies are available for each individual product go to that product's web page and open the geography tab across the top. See the chart below: