U.S. public data reports and Census analysis
Read comparisons built from compatible Census measures and download their evidence.
-
Housing costs and household income →
A reproducible state comparison of median gross rent, median household income and the share of computable cash-rent households with rent burden. One ACS period, explicit universes, no causal claim.
What makes a report reproducible
A report needs the measures behind its narrative. Its source release, entity identifiers, reference periods, transformations and chart selection are recorded with the download. A later source update does not silently replace the observations in a frozen report.
Cross-topic comparisons join geographic records, not individual households. Two area-level totals cannot show that the same household has both characteristics. Income and rent medians describe different universes, so dividing them does not reproduce household cost burden.
The housing report covers the 50 states and the District of Columbia in one ACS product. County-level context is available in profiles and comparisons; the report does not assign state median values to local households.
Related interactive analyses
The Florida and California renter comparison switches between the 30% and 50% burden thresholds and explains why a ratio of rent and income medians cannot reproduce household burden. The attainment and income analysis separates adult and household universes. The broadband analysis distinguishes household subscriptions from provider-reported availability. Each article supplies its own visible selection and source record.
These analyses complement the report collection with focused questions rather than expanding a report into unsupported topics. Their downloads retain source product, geographic identifiers and measurement periods so a comparison remains interpretable after publication.
Interpretation standards
Reports distinguish observations from possible explanations. Descriptive relationships do not establish what caused a difference or isolate a policy effect. Model-generated outcomes can already use a proposed explanatory variable as an input; those relationships require particular care before being described as independent discoveries.
The initial housing report uses directly observed Census source tables and an explicit rent-burden formula. Other reports will be added when the source joins, rights, uncertainty and editorial interpretation are ready. There is no report builder or automatically generated missing section in this release.