Wilder Research Center

The Minnesota Homelessness Survey

Wilder Research Center has conducted five statewide surveys of people experiencing homelessness, in 1991, 1994, 1997, 2000, and 2003.

In 2000, more than 500 volunteers received two hours of interview training by satellite television broadcast carried at sixteen sites around the state. The training session provided instructions, tips to reduce interviewer bias, and a role-play exercise. Volunteers completed 2,500 interviews in thirty-six hours at 288 sites around Minnesota, including emergency shelters, battered women's shelters and safe homes, time-limited transitional housing programs, teen shelters and transitional programs, and forty-one non-shelter locations including food sites and encampments. Each respondent received five dollars at the completion of the interview. A census of providers showed that the interviews included 40 percent of all persons living in shelters. At larger shelter sites, survey participants are chosen by random sampling. All survey results are weighted to reflect the entire known sheltered population on the night of the survey.

One important methodological limitation of the statewide survey is the fact that it relies on a convenience sample of non-shelter locations; it does not include a comprehensive sampling of soup kitchens or [Page 597]other feeding sites. However, virtually every shelter, safe house, voucher site, and transitional program is included in the survey and homeless outreach workers assist in the identification of street locations and dropin centers where non-sheltered homeless persons can be found. In 2000, 92 percent of those asked to complete the survey agreed, both in shelters and in non-shelter locations, yielding a large enough sample for analysis of specific populations, such as military veterans, American Indians, African-Americans, and people with mental illness. Other strengths of the survey include an actively involved advisory group whose wide range of perspectives and organizations (advocates, service providers, shelter operators, state agency staff, and nonprofit organizations) enhances the practical value of study results. Volunteer interviewers learn about the issue through face-to-face interaction with people experiencing homelessness. They also lower the cost of this massive data collection project, since interviewers are not compensated for attending training and conducting interviews.

Over the years, the Minnesota Homeless Survey has provided documentation for many grant applications, resulting in tens of millions of dollars to support homeless programs and housing development.

...