The Decennial Census and Homelessness: Then and Now

written by naehblog
March 31, 2010

This is an excerpt from a longer report by HRI Research Associate Megan Henry. Big hat tip to her!

In honor of the 2010 Census, here’s a few interesting facts about the way the census has included people experiencing homelessness over the years:

  • The first Census occurred in 1790, and consisted of just four questions. Since then, the United States has increased efforts, changed methods, altered principle criteria and time frames, and expanded the survey considerably.
  • In 1850, for the first time, “pauper” was included on the Census questionnaire alongside deaf, dumb, blind, insane, idiotic, and convict. Paupers were considered poor beyond a failure to meet an income threshold, and the word indicates a sort of dependency on government or private assistance to survive.
  • By 1854, when the Census results had come out, the New York Times was contesting the numbers. An articled identified, quite specifically, that the “outdoor paupers” were not included in the U.S. Census, and asserted that the Census’ number of paupers (134,392) was deficient by 50 percent.
  • Between 1870 and 1920, anywhere between 40,000 and 75,000 people were counted living in poorhouses.
  • While some Censes of the 19th century included populations similar to those we consider “homeless,” only Census operations beginning in 1970 identified specific efforts to count the homeless population.

Obviously, the way the Census counts people experiencing homelessness has changed. As we speak, census workers are counting: today, they’ll count those staying outside; yesterday they counted people at soup thkitchens and mobile food vans. They started Monday at emergency shelters.

For the 2000 census, Census Bureau implemented “Service-Based Enumeration,” counting people at emergency shelters, transitional shelters, shelters for unaccompanied youth, hotels, motels, soup kitchens, regularly scheduled mobile food vans, and targeted outdoor locations.

In 2000, 283,898 people were counted as homeless in 14,817 locations. Approximately 50 percent of the locations were shelters. 62 percent of people were counted at shelters, 28 percent at soup kitchens/food vans, and 10 percent outdoors.

There were numerous issues associated with the 2000 count: the limited number of outdoor sites included in the count, shelters that were not open during the day were not included, and shelters that were only open for two days a week were not included.

Since the last census, the Census Bureau has worked with local government and advocacy organizations to create a more comprehensive list of shelters and homeless service providers. Also, a “Be Counted” questionnaire has been developed for people who do not believe they were counted in the census effort. Questionnaire assistance centers (QAC) will be set up in libraries, post offices, community centers, gas stations, etc. and will provide people with assistance in filling out the questionnaire.

What about the census in your community? Do you think it will be accurate?