The Census Bureau will test using U.S. Postal Service workers to collect in-person responses in two Southern cities this spring, a small experiment with potentially large consequences for how the next decade’s population counts are conducted. The pilot — rolling out invitations online starting May 1 and fieldwork beginning in June — raises fresh questions about privacy, cost and whether neighbors’ familiarity with mail carriers boosts response rates.
The bureau says it will invite roughly 154,600 residents in Spartanburg, South Carolina, and Huntsville, Alabama, to fill out a test questionnaire online. For households that do not respond digitally, postal carriers will be tapped to follow up in person.
How the pilots will operate differs between the two cities. In Spartanburg, 25 postal employees will stop at homes on their regular routes, identify themselves as postal workers and administer the test questions during their shift. Those carriers will receive their usual USPS pay; the Bureau of Labor Statistics reported an average postal carrier wage of about $28.79 per hour in 2024. Non-postal census workers hired for the Spartanburg test will be paid roughly $17.75 an hour.
New York Times hit with lawsuit from US rights agency: claims white staffer was denied promotion
Horoscopes April 27: what every zodiac sign should expect today
In Huntsville, the plan calls for 25 postal volunteers who will collect answers outside their normal delivery hours — evenings or weekends — and present themselves as Census Bureau staff. Those participants would be paid the same rate as other temporary census hires, roughly $19.75 an hour.
Postal officials argue carriers bring an advantage: long-established daily contact with residents and local knowledge of routes that could improve outreach. The National Association of Letter Carriers framed the assignment as something workers often see as a community service — a chance to help connect neighbors with the census.
But critics question whether the experiment will deliver better results or save money. A previous review by the Government Accountability Office in 2011 concluded that using mail carriers was unlikely to be cost-effective because regular postal pay is substantially higher than temporary census wages. This month, 21 Democratic state attorneys general pressed the Commerce Department for evidence that the new approach would lower costs or raise participation.
Confidentiality is another flashpoint. A similar idea was abandoned during preparations for the 2020 cycle after officials found the Census Bureau’s strict secrecy rules clashed with postal policies that allow certain address information to be shared with other agencies. In response to current concerns, the Census Bureau says participating postal workers will receive the same training as its own census takers and will take the agency’s lifetime oath to protect respondents’ information.
- Locations: Spartanburg, SC and Huntsville, AL
- Sample size: ~154,600 residents invited online beginning May 1
- Fieldwork: Postal carriers begin in June
- Staffing: 25 postal workers in each city
- Pay: Postal average ~ $28.79/hr (2024 BLS); Spartanburg non-postal hires ~$17.75/hr; Huntsville census hires ~$19.75/hr
- Official safeguards: training and lifetime confidentiality oath required
- Main concerns: cost-effectiveness, legal/confidentiality conflicts, potential response bias
Complicating the pilot are recent changes to the 2026 census test announced by the current administration that critics warn could affect how the full 2030 count is shaped. Officials pared back several test sites, narrowed online language options to English from a broader slate that included Spanish and Chinese, and shifted to a questionnaire modeled on the longer American Community Survey — a form that includes a citizenship question — rather than the much shorter traditional census instrument. Advocates say those moves reduce the test’s ability to predict real-world outcomes.
The outcome of these small-scale trials will be watched closely by advocacy groups, state officials and Congressional overseers because the decennial census underpins federal funding formulas and the apportionment of seats in the U.S. House of Representatives. Any change that affects who gets counted or how reliably people respond has tangible budgetary and political consequences.
Next steps: the bureau will collect results from the two pilots, compare response rates, cost metrics and confidentiality performance, and then decide whether to scale the approach. For now, the experiment is a narrow test with outsized stakes — both for how census work is done and for the communities whose services and representation depend on accurate counts.
