The Census Bureau's 1984 Address List Compilation Test

GAO discussed the Census Bureau's 1984 address list compilation test, specifically: (1) the importance of a good address list; (2) past list development techniques and problems; (3) the 1984 test results; and (4) prospects for list development for 1990 and beyond. GAO noted that the Bureau: (1) requires a complete and accurate mailing list because it conducts the census primarily by mailing questionnaires to households; (2) has purchased commercial vendor mailing lists for large urban areas; (3) developed its rural and small urban area mailing lists by sending employees out to locate people; (4) has experienced delays and inaccuracies in developing its lists with these methods; and (5) tested various combinations of list development and update options in 1984. GAO evaluated the results of the tests and found that: (1) for large urban areas, obtaining initial lists from vendors was less expensive on a per-address basis than purchasing the Postal Service's list or using the 1980 list; (2) after the Bureau updated the commercial list, it remained the least expensive and there were no major differences in the lists' accuracy; and (3) for rural areas, the Postal Service list was less expensive than the Bureau's, but it contained a large number of incorrect geographic designations. The Bureau concluded that there was no significant advantage in using the Postal Service as its primary source for creating the list for the 1990 census and planned to use the same methodology that it used in 1980. GAO agreed with this position and endorsed Bureau plans to automate activities previously done manually.