The American Community Survey’s biggest asset also presents its biggest challenge, according to the summary of a recent workshop that brought together two dozen Census Bureau staff and expert users to evaluate the ACS and discuss how to make it more useful. Specifically, users said the local-level estimates in the ACS are its most valuable feature, but that their biggest concern is the large margins of error associated with these small-area estimates.

The workshop summary is available on the website of the Population Reference Bureau, which co-hosted the event in June with the Annie E. Casey Foundation. In addition to summarizing the workshop, the hosts released a non-scientific survey of participants conducted before the meeting got under way, and links to some of the presentations.

The American Community Survey is a widely used national survey taken by the Census Bureau that replaced the decennial census long form. It provides annually updated population and housing data. For small areas, multiple years of data from the survey need to be combined to produce a large enough sample size to generate reliable estimates. The workshop focused particularly on the estimates that roll together five years of data to produce information about neighborhoods or other areas with populations of less than 20,000. The first set of five-year estimates, for 2005-2009, were released in December. ACS estimates for 2006-2010 are due to be released later this year.

Workshop participants shared their strategies for dealing with the large standard errors in the five-year estimates, and asked the Census Bureau to offer more guidance and tools to help them work with ACS data. Some said the bureau’s recently redesigned data-retrieval tool, American FactFinder, is difficult to use. But they praised the rich variety of available data.

However, another issue that came up at the conference, according to the summary, was the “strain on staff resources” at the Census Bureau because of the large volume of data being released, just as users are asking for more help. Census Bureau officials say are doing a systematic review of the entire ACS program, in part to address these issues. Under discussion is the possibility of modifying or eliminating some data products, which the bureau recognizes would require input from users, according to the workshop summary.