The Census Bureau conducted a test in summer 2014 of new methods and advanced technologies that could significantly improve the upcoming 2020 Census.

The 2014 Census Test took place in parts of Washington, D.C. and parts of Montgomery County, Maryland from the end of June through September 2014. Census Day was July 1, 2014.

The purpose of the 2014 Census Test was to test the effectiveness and efficiency of operations we are considering for use during the 2020 Census. Within the two sites, we examined different strategies designed to encourage householders to respond via mail, Internet and other options (called “self-response”).  The Census Bureau also tested different strategies to target in-person interviews to more efficiently follow-up with households that don’t self-respond (called “non-response follow-up” or “NRFU” field operations).

The scope of the test included:

• testing of the Internet self-response mode and the contact strategies for Internet pre-registration;
• use of e-mail and automated voice invitations;
• testing of the mobile devices used by field staff to enumerate households that don’t self respond;
• testing of alternative NRFU contact strategies (number of contacts per address, telephone and/or personal visit contact combinations, whether and when to conduct proxy interviews, etc.) for non-responders;
• use of existing government data sources information to identify cases to remove from the non-responding workload when the administrative data are considered sufficient for enumeration; and
• testing of different adaptive design methodologies to manage field enumerator work assignments.
  

Any analysis from this test will be focused on the methods tested.  We did not produce updated population counts or other demographic characteristics from this test.

Expected learnings from this test

We tested alternative treatments for both self-response and non-response follow-up (NRFU) data collection operations. We then compared results (including cost and productivity differences) across alternatives tested for each component (self response and NRFU) to answer research questions and inform preliminary design decisions for the 2020 Census.

We also conducted focus groups in the test locations to examine reactions to the alternate contact, response, and existing government data sources use; any privacy or confidentiality concerns; and how the Census Bureau might address these concerns through micro- or macro-messaging.