In 2002 and 2003, the Census Bureau, at the request of Congress, conducted research to determine whether the American Community Survey (ACS) could be implemented as a voluntary, rather than a mandatory, survey. A test was designed to provide answers to key questions about the impact, if any, that a change to voluntary methods would have on mail response, survey quality, and costs. This test was conducted between March and June of 2003 and only housing units were included, since there was no group quarters sample in that year. While the test was not a randomized experiment, the Census Bureau (U.S. Census Bureau, 2003) did conclude that:

• A dramatic decrease occurred in mail response when the survey was voluntary. The mail cooperation rate fell by over 20 percentage points, and the final response rate after all three modes was about 5 percentage points lower. The reliability of estimates was adversely impacted by the reduction in the total number of completed interviews.
• The estimated annual cost of implementing the ACS would increase by at least 38 percent if the survey was voluntary and the current number of respondent cases was maintained.
• Perhaps of greatest concern, the use of voluntary methods had a negative impact on traditionally low response areas that will compromise our ability to produce reliable data for these areas and for small population groups such as Blacks, Hispanics, Asians, American Indians, and Alaska Natives.