Census of Government Finance, 2007: A Discussion of Response Rate Quality Issues, Data User Concerns, and Plans for Data Improvements

摘要

In the release of the preliminary data for the 2007 Census of Governments: Finance in September 2009, we issued a report summarizing the limitations of the preliminary data. While most of the preliminary data for general purpose governments and schools was of good quality, we cautioned the data user when using the preliminary special district data. The impact of the known errors on the released aggregates was discussed in detail in the paper. We also issued a mitigation strategy designed to help address these known limitations. Both long- and short-term strategies were discussed. In this revision of the paper, we are updating the previous report with the results of the additional work done in October and November of 2009. This report will serve as documentation of the initial limitations and the results of our efforts to mitigate some of the errors prior to the release of the final 2007 data. In the release of the preliminary data for the 2007 Census of Governments: Finance, we cautioned our data users of problems that were to be addressed in subsequent releases of revised data in 2009. Two states had unit response rates around 40 percent (Delaware and New Mexico). The impact of these low response rates was discussed in the initial report. We identified 27 states with other problems at the state by type-of-government level (predominately for special districts).