HMDA Mortgage Data: Geographic Analysis with FIPS Codes

The Home Mortgage Disclosure Act (HMDA) requires most mortgage lenders to report detailed loan application data to federal regulators. The resulting public dataset — published annually by the Consumer Financial Protection Bureau (CFPB) — contains millions of loan records with geographic identifiers including county FIPS code, Census tract GEOID, and MSA code. It's one of the most powerful tools for studying lending patterns, housing market access, and potential fair lending violations.

Each HMDA record includes a 5-digit county FIPS code and an 11-digit Census tract GEOID. For county-level analysis, aggregate on the county FIPS field. For neighborhood-level analysis, use the full tract GEOID. The county FIPS allows direct joining with Census demographic data (ACS), BLS employment data, HUD housing data, and any other FIPS-based dataset — enabling rich contextual analysis of lending patterns against local economic conditions.

A common research pattern: join HMDA denial rates (computed from the loan action field) by county to Census median income and racial composition data, using county FIPS as the key. This reveals whether denial rates correlate with demographic characteristics at the county level — a standard analysis in fair lending research. For metro-level analysis, use the HMDA MSA field and join to MSA data on our MSA pages.

The CFPB's HMDA Data Browser provides a web interface, but serious analysis requires downloading the flat files. The HMDA LAR (Loan Application Register) files are available for download from the CFPB's website and contain the full column set including geography. Identify the county FIPS codes for your study area using the search tool or state-level county listings like Texas or Florida.