Census Participation


by Charles Rynerson

change in census responseIn 2010, the Census Bureau focused outreach in hard- to-count (HTC) census tracts, based on data from the 2000 census. Twelve variables that were correlated with nonresponse rates in 1990 and 2000 were used to derive  the HTC score:

  • Housing: % vacant, % multi-family,% renter, % crowded;
  • Households: % not husband/wife, % linguistically isolated, % no phone, % recent movers, % public assistance;
  • Persons: % below poverty, % unemployed, % not HS graduate

Census workers have to conduct more costly door-to- door visits to count households that did not mail back their form. Changes in the mail response rate may reflect the success of outreach efforts or may have resulted from  demographic shifts that the 10 year old data used to determine the HTC score could not account for.

The census tracts with the lowest mail response rates in the 2000 Census were within N/NE Portland’s Humboldt, King, Boise, and Eliot neighborhoods, Central Portland’s Old Town, and Central Beaverton and Hillsboro. These tracts all had mail response rates below 65%. Statewide response rates were 77% in Oregon and 75% in Washington.

In 2010, the lowest response rates were largely in Multnomah County neighborhoods east of I-205, with the very lowest in North Gresham, at 64% and in Fairview/Wood Village, at 63%. Statewide 2010 response rates were 76% in both Oregon and Washington, nearly identical to the 2000 rates in both states.

Most of the biggest increases were in the Beaverton/Hillsboro area, SW Portland, and Inner N/NE Portland.  For example, a census tract that includes portions of the Humboldt and Piedmont neighborhoods in N/NE Portland increased from 71% in 2000 to 85% in 2010. Most of the biggest drops in mail response occurred within Multnomah County east of I-205, including a census tract in NE Portland’s Argay Terrace neighborhood where mail response dropped from 89% to 72%.  census response 2000-2010