Words Matter: The Importance of Issue Framing and the Case of Affordable Housing

Abstract

Problem: Does how planners talk about issues have an impact on public support?

Purpose: I examine whether the words used to describe a policy make a difference for the level of acceptance of that policy.

Methods: I conducted a random survey of suburban residents. One-half of the sample were randomly chosen to answer a question about their support for affordable housing; the other half were asked to answer the identical question except that it substituted the words lifecycle housing for the previous phrase.

Results and conclusions: The difference in wording produced a difference in public opinion of more than 30 percentage points, with those whose question referred to lifecycle housing expressing greater support. This difference was significant among White, non-Hispanic respondents; there was no significant difference among other (non-White) respondents. The phrase affordable housing seems to be stigmatized and perhaps associated with race among the residents of this affluent, White suburb. This study examines public support in the context of an opinion survey and these results may not apply to attitudes about actual projects or policies.

Takeaway for practice: Under the right circumstances, the words planners use might expand policy options available to local officials. Proper framing may allow some previously infeasible policies to succeed.

Research support: None

Keywords:

• affordable housing
• issue-framing
• race
• suburbs

Notes

a. Hispanic respondents are included in the non-White category.

¹ . The city used procedures advocated by Dillman (2000) to enhance response rates, sending an initial mailing with questionnaire and self-addressed, stamped, return envelope. They followed up with a reminder postcard and a second questionnaire. The survey over-sampled renters to ensure a sufficient number of responses from them, since they make up a small percentage of the city's residents.

² . The city survey used “Hispanic” as a category in the question asking about race. This is different from the decennial Census, which asks separate questions about race and ethnic origin. Thus the race/ethnicity categories used to report the results of this study should be interpreted as (1) non-Hispanic White, and (2) all others. That is, those identifying themselves as Hispanic are grouped together with non-Whites in the results.

³ . One-third (33.3%) or 445 of the respondents had incomes greater than $100,000.