The goals of the census are intelligently articulated within Marx’s quote:

“In its best interests, a civilized nation counts and profiles its people and institutions. The abiding mission of the U.S. Census Bureau is to provide, ably and objectively, these basic statistics about the people and the economy of the nation to the Congress, the executive branch, the business and academic communities, and the general public. The success of a census rests not only on collecting data, but also on linking those data to geographic areas.”

For detailed instruction on extracting and applying census data I recommend studying Unlocking the Census with GIS by Peters, Alan and Heather McDonald, 2004. The authors explain census data, census geography and various tools and skill sets needed to generate information from raw data.

The retrieval of census data requires the collecting and merging of disjoint datasets. The geographical data is found in one of eighteen TIGER Files. There are other “canned” sources from a multitude of other sources, some free, some purchased. My project is concerned with “raw” data.

The TIGER files are zipped files that need to be downloaded and extracted for personal use. Shapefiles for downloading of limited geographies are available. But using raw data allows for better control all available geographies that you can use toward your information output. Most GIS software is capable of processing TIGER formats. The TIGER files are series of string texts that the software parses to build topologically correct points, lines, and polygons. There are also free standing software programs that can process the data into shape files for importation into the GIS. ERSI Arcview 9.0 has such capability but I have not tested that process. Instead I used TGR2shp software.

The demographic data from census questionnaires and surveys may be retrieved from the American Fact Finder web page where you can download raw or semi-processed Census Questionnaire Data. Various pre-configured demographic data tables with associated maps are available from the American Fact Finder web page. But again, raw data allows for better control of your information output. American Fact Finder data sets consist of numerous summary census and survey tabulations. The U.S The U.S. Census Bureau's homepage provides numerous links to data dictionaries and information on the various census and surveys. Exploring this web site is well worth the effort. One of the issues facing those of us knew to census data is becoming familiar with the data tables. The census tables from American Fact Finder have cryptic headings. The second row provides discriptions. However, this information will confuse most databases and should be removed upon extraction. Luckily there is a lot of supporting documentation available on the American Fact Finder web site. Below is an example of a table heading with descriptions.

Census Table Headings

P003003	P003004	P003005	P003006	P003007
Total population: Population of one race; White alone	Total population: Population of one race; Black or African American alone	Total population: Population of one race; Population of one race; American Indian and Alaska Native alone	Total population: Population of one race; Asian alone	Total population: Population of one race; Native Hawaiian and Other Pacific Islander alone

Extracting the data is but the first steps in analyzing census data. Once the raw data is downloaded and unzipped the work of joining the geography with the census data begins. The American Fact Finder defines this as “advanced” processing. However, anyone with a basic knowledge of rational databases and data joins can do this quite easily. The main technical issue is that the Geo_IDs for the Census Data and for the TIGER files are usually parsed differently so that some data manipulation must be made to allow for the tables to join in the GIS. For instance the Census Tract might be defined as 0600140210010120 in one data set but 4021 in another. After parsing the data for joining the tables the next issue is data format. Once the Geo and Census tables are joined and the measurement levels (nominal, ordinal, interval/ratio) defined it’s pure GIS analysis. The tools I used for completing this process consisted of basic spreadsheets (MS Excel),databases (MS Access), TRG2shp software, DBFView, and the American Fact Finder. Although I downloaded and studied the Ferrett tool I was able to proceed without using it.

Albany, California census blocks, census block groups, and census tracts. Now that I have the basic geography I can add cartographic elements (street names, etc) although they should be kept at a minimum. One reference map should be sufficient. Now that I have my base maps prepared in ArcView, I am ready to join census data downloaded from American Fact Finder web site. Click here to see the TIGER tables that were parsed for this geography

Much of the economic and housing census data is only available at the census block group and census tract level. This is due to confidentiality and because statistical sampling error is introduced when a disproportionately small population is used for analysis. The typical census tract population is 4,000. Click here to see the TIGER tables that were parsed for this geography

Census blocks, census block groups, and census tracts. Much of the economic and housing census data is only available at the census block group and census tract level. This is due to confidentiality and because statistical sampling error is introduced when a disproportionately small population is used for analysis. The typical census tract population is 4,000. Click here to see the TIGER tables that were parsed for this geography

Here I have pie charts showing the ratio of the racial distribution. Click here to see the census tables.

Here I have pie graphs showing percentage of tenancy defined as owned or rented. Click here to see the census tables.

Median Income at census tract level. Here I have bar graphs showing median income of races for each tract. (Please note that the average income is not 51K as the bar graph might suggest.) Click here to see the census tables.

One of the challenges I faced was importing the tables into the map layer. While ESRI allows for this, the process proved very fragile and so I resulted to hyper linking table data to most of the maps. I found that the best process was to export the tables to MS Excel, format, save as a DBF file, than use DBFView to export to HTML. MS Access is also my preferred method for processing tables. Projecting the data onto a map can be a bit of a challenge. I found it best to experiment with various levels of symbolism to see which communicated the information the best.

I find two of the maps particularly interesting. The Tenancy map displays the ratio of rented housing units versus owned housing units. Less than 1% of the residences living in census tract 4204 own the property. This part of Albany contains a lot of student housing and small high density triplexes. People living in rented communities are generally in more need of social services.

The second map of interest is the Total Housing Units map. The Albany tract with the highest number of units is Tract 4203. This area of Albany is dominated by Albany Hill which is flanked by high rise condominiums on the west and high density apartments on its eastern and southern slopes. Albany hill is still largely undeveloped

I had wanted display geospatial photos of Albany High which architecturally resembles a junior college campus and compare them to El Cerrito High which architecturally resembles a junior state or federal penitentiary. These schools may be located less than 2 miles apart geographically but architecturally they are worlds apart. During my project I learned that El Cerrito High is being demolished in the summer of 2005 and replaced with a state-of-the-art campus-like facility. I am proposing that overall the grades of the students attending the new El Cerrito High will rise and that morale will improve.

Demographic maps supported by census data can give compelling arguments in helping social scientists target critical areas that need additional financial resources. I plan to continue my study of the census and will be using the census for my project in the Spatial Analysis course offered at Los Rios Junior College.

Demers, Michael N., (2005), “Introduction to Digital Geography”, Fundamentals of Geographic Information Systems. 3rd ed, John Wiley Sons, Inc., Hoboken, New Jersey, pg 11

Peters, Alan and Heather McDonald. (2004) Unlocking the Census with GIS., Redlands, California.

Marx, R.W., 1986, “The TIGER System: Automating the Geographic Structure of the United States Census.” Government Publications Review, 13:181-201.

George Farnsworth and James P. Currry, 1970, “The DIME Geocoding System.” Washington: U.S. Bureau of the Census [for sale by the Supt of Docs, U.S. Govt. Print. Off.] Census Use Study Report, No. 4.