The database is a geographical database based on the level of the census block. Census blocks roughly correspond to city blocks in urban areas. There are approximately four hundred thousand census blocks in the State of California. The census block format is ideal for a couple of reasons. First, many types of analysis are best done at the census block level. Second, it is the lowest level building block which can be used in redistricting since it is the lowest level at which population counts are reported.
Data which accords naturally to such a database is of course census data, some of which is released at the census block level. Data which is not released at that level can be statistically allocated back to the census block level and can then be used at higher levels of aggregation, such as the census tract. Other data in the database must be merged to the census block through a process called geocoding. An example of such data is the registered voter file where voters are assigned to precincts, a county-defined area, rather than census blocks. To be useful to an integrated database, registered voters must be merged to the census geography. The software required for this is involved and complex since it must keep track of the relationship between thirteen million registered voters, four hundred thousand census blocks, forty thousand registration precincts and thirty thousand election precincts.
The census-block based database can be displayed geographically. There are numerous programs with such capability extant. Much of the usefulness of this database can be realized through geographic manipulation that can be taught to naive users within several hours. The archival data are, however, valuable in and of themselves, even without the display programs.
| Copyright © 1997-2004 | [ROAD Home] | Questions? Contact the ROAD webmaster. |