A New World of Social Science Data for Small Areas: Opportunities and Challenges

The American Community Survey is a reality:  1-year period estimates were released for large geographic areas beginning in 2006; the first 3-year period estimates for smaller areas were released in 2009; and the first 5-year period estimates will be released for groups of city blocks, census tracts, and very small governmental areas in 2010.  The ACS data are an improvement over the once-every-10 years decennial census long-form-sample data because they are more frequent and timely; they also have more complete response.  However, because the ACS is a continuous monthly survey, the period estimates pose challenges to users who are accustomed to point-in-time data.  In addition, the estimates have larger sampling errors than the corresponding long-form-sample estimates.  This talk will compare and contrast the ACS and long-form-sample data on a number of technical and substantive dimensions.  While challenging to use, the ACS promises to support richer research on our diverse society and economy than ever before possible for states, cities, and smaller geographic areas.