In the 1982 Government Employment Response Evaluation Study, 318 local governments were sampled to measure the error in the data reported to the Census Bureau by local governments. Using the knowledge gained from this evaluation survey, future censuses and surveys will be improved through redesigning forms and procedures in order to reduce reporting, clerical, and editing errors and to reduce respondent burden. Although this study was designed specifically to study respondent errors in the employment phase of the 1952 Census of Governments, the annual survey of government employment should also benefit from this study since some of the same information is collected and published annually in that survey.