Poll on homelessness: How it was done

Two years ago, the Los Angeles Times worked with the Los Angeles Business Council Institute to conduct a groundbreaking poll about attitudes toward homelessness. This year’s survey was designed to see to what extent public opinion has shifted after nearly two years of a deadly pandemic.

As they did in 2019, the LABC Institute hired Hart Research to conduct the survey in cooperation with The Times. The polling firm, based in Washington, D.C., has extensive experience around the country, including in Los Angeles.

The survey questioned 906 Los Angeles County voters from Oct. 27 to Nov. 3, by telephone (602 interviews) and online (304 interviews). The results have an overall margin of error of 3.3 percentage points in either direction. Margins of error for subgroups are larger.

The firm also conducted two focus groups with L.A. County voters, one prior to the poll and one after it.

The Times contracted with the Institute for Governmental Studies at UC Berkeley, which has conducted other surveys for The Times over the last several years, to allow its poll director, Mark DiCamillo, to serve as a consultant to The Times on the survey.

Times reporters and editors suggested questions for the survey and participated in discussions about question wording and survey design. Final decisions about the contents of the survey were made by the LABC Institute.

Respondents for the survey were selected from the California voter file, a public database of registered voters. Interviews were conducted in either English or Spanish, depending on the respondent’s preference. Online respondents were selected from an online panel and screened for their registration status and other demographic characteristics in order to participate in the survey.

The survey conducted extra interviews among Black and Asian American Pacific Islander voters to allow adequate analysis of those groups. In all, 126 Black voters and 135 AAPI voters were interviewed. In the final data, racial groups have been weighted back to reflect their actual proportions of the L.A. County electorate.

The survey set quotas to achieve the proper geographic, demographic and partisan distribution; once the interviews were completed, Hart Research applied minimal weights to geographic region, party registration, age, gender and education in order to ensure that the sample reflects the overall demographics of registered voters in L.A. County.