U.S. Housing Wealth Recovery and Vulnerability Data, 2007–2021

Harvard Dataverse hosts the replication archive for a study on spatial persistence in U.S. housing wealth from 2007 to 2021. The dataset covers the contiguous United States at the Public Use Microdata Area level across five survey years, including the Great Recession and COVID-19 pandemic. It contains raw input data, pre-computed analysis files, and R scripts to reproduce all tables, figures, and statistical results from the associated paper.

Use Cases

• Calculate a real housing-wealth recovery index based on the described methodology.
• Perform spatial autocorrelation tests using the provided global Moran's I statistic and LISA cluster maps.
• Develop a k-means recovery typology linked to FEMA social-vulnerability and natural-hazard risk scores.
• Conduct instrumental-variables analysis of Great Recession severity and COVID-era recovery using the described Bartik industry-mix shock and HMDA lending share.

Strengths

• Enables full reproduction of every table, figure, and statistical result from the published paper.
• Covers 2,351 Public Use Microdata Areas across five survey years spanning 2007 to 2021.
• Integrates data from multiple public sources including IPUMS ACS, FEMA National Risk Index, FHFA House Price Index, and HMDA.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

Publicly available U.S. government sources and openICPSR.

Collection Method

Aggregated from multiple public datasets for replication of a specific research paper.

Time Range

2007–2021

Freshness

Last updated 2026-06-16 04:50:07; freshness should be verified.

Geography

Contiguous United States