Every household will have the option of responding online, by mail, or by phone.
Nearly every household will receive an invitation to participate in the 2020 Census from either a postal worker or a census worker.
95% of households will receive their census invitation in the mail.
Almost 5% of households will receive their census invitation when a census taker drops it off. In these areas, the majority of households may not receive mail at their home’s physical location (like households that use PO boxes or areas recently affected by natural disasters).?
Less than 1% of households will be counted in person by a census taker, instead of being invited to respond on their own. We do this in very remote areas like parts of northern Maine, remote Alaska, and in select American Indian areas that ask to be counted in person.
The 2020 Census has special procedures to count people who don’t live in households, such as students living in university housing or people experiencing homelessness.
When the census starts in 2020, about 80 percent of addresses will receive an invitation letter with instructions on how to respond online or by telephone using a unique ID. The remaining 20 percent of addresses — selected because they are less likely to have or use the internet — will receive both an invitation letter (with a unique ID) and a paper questionnaire with postage-paid return envelope. After three mailed requests to complete the census online or by phone, unresponsive households will receive a paper questionnaire and return instructions on the fourth mailing.
Around May 9, 2020, the Census Bureau will begin NRFU, to count households or determine the status of housing units that did not self-respond. This is the costliest census operation, so the Census Bureau strives to obtain the highest self-response rate possible.
The Census Bureau will use administrative data, primarily from the U.S. Postal Service, local governments, and third-party commercial vendors, to identify and remove vacant housing units from the NRFU universe. The Census Bureau will mail a final postcard to those addresses, inviting someone to respond or contact the Census Bureau if the home is, indeed, occupied.
Enumerators will visit all non-responding households (that is, occupied housing units that haven’t responded) at least once. If no one answers the door or if the “head of household” is unavailable, the enumerator will leave a “Notice of Visit,” a note explaining the attempt and encouraging the occupants to self-respond. Unlike previous decennial censuses, census enumerators will use smart devices to collect data, instead of the traditional pen and paper. Subject to the rules described below, enumerators can visit an unresponsive household up to six times.
The Census Bureau has tested the use of high-quality administrative data, previously collected by other federal government agencies for other purposes or from previous census surveys, to enumerate some households that do not respond to the first in-person visit. The Census Bureau estimates that it could count about five percent of all households (or about 15 percent of unresponsive households) this way, but it has not announced a final plan for using this method at the time this report went to press.
If three in-person attempts to count a household are unsuccessful, enumerators will attempt to conduct an interview with a proxy respondent if they determine the proxy has sufficient knowledge of who lived in the housing unit on April 1, 2020. Proxies can include:
Relatives of the occupants
Landlords or building managers
Real estate agents and new occupants (if the residents moved around the time of Census Day)
Local government employees (clerks, tax collectors and other administrative staff)
Utility workers or postal service employees
Otherwise, enumerators can make up to three additional contacts in person or by telephone (no more than six in total) to complete the census form.
Once all attempts to count a housing unit have been exhausted, the Census Bureau will use federal and local administrative records to fill in missing information. Examples of administrative records include:
IRS documents (1040 Forms)
Medicare and Medicaid records
Social Security Numerical Identification System records
U.S. Post Office files
Supplemental Nutrition Assistance Program (SNAP)
Records from previous Census Bureau surveys
If high-quality administrative data are not available, the Census Bureau will use statistical imputation methods to count households that appear to be occupied.
After three mailed requests to complete the census online or by phone, unresponsive households will receive a paper questionnaire and return instructions on the fourth mailing.