Assessing the Impact of Grounds for Opposition in Local Multifamily Housing Decision Making

This multi-method, interdisciplinary, research project investigates the prevalence of demographically coded language in opposition to multifamily housing (MFH), and its effects on housing/zoning decisions. Textual data on MFH decision making are taken from case transcripts from federal and state fair housing lawsuits and recently decided cases in Arizona. The study more generally draws upon data that include meeting minutes, legal decisions, news articles, and administrative data.

The research methods include manual textual and thematic analysis of textual data on MFH decision making, machine learning textual and thematic analysis of such data, an online cultural consensus theory-based survey, and qualitative comparative analysis of interview data. Stage 1 of the research collects publicly available textual data on MFH decision making from federal and state fair housing lawsuits and Arizona (AZ) news media and public meetings between 2021 and 2024.

Manual and machine learning (ML) textual and thematic analysis, a meaning decoding test, and descriptive statistics are used to identify the prevalence of demographically coded opposition to MFH and to develop a preliminary code words corpus and decoding guide. Stage 2 deploys in-depth interviews and a cultural consensus theory (CCT) analysis with an estimated 100 MFH stakeholders in a subsample of 20 lawsuits and AZ cases to help verify the corpus with a second application of the “decoding test” and refine the guide.

Manual and ML thematic analysis, qualitative comparative analysis, and ML textual and statistical analysis are then employed to investigate the qualities and dynamics of demographically coded opposition. The research contributions will include a refined theory of coded demographic language in MFH decision making, a corpus and decoding guide to assist future research on use of demographically coded language in housing settings, insights on the benefits and drawbacks of using ML to study demographically coded language, and fair housing agency data collection, training, and public education tools.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.