Actually Clean Data ↗

SOURCES, VERSIONS & WHAT CHANGED

2024 county income and poverty: source and preparation record

The first ACD income and poverty reference prepares annual Census estimates for county-level planning and research.

3,144county identifiers
3,143counties with estimates
7measures per county
90%confidence level for published intervals

Source edition and ACD preparation

The estimates describe 2024. Census released the SAIPE county/state files January 27, 2026; their January 7 creation tag is a different date. ACD captured and prepared this edition October 17. This is our first release of these retained inputs, with no prior ACD or historical comparison claimed.

What preparation adds

We check every numerical cell across Census’s text and Excel representations, preserve five-character county FIPS, and use the verified Excel names. The text truncates Lower Connecticut River Valley Planning Region at 45 characters; our label audit records the full official name and its original text. The download supplies both a one-row-per-county reference and a long measure table with original cells.

Coverage and accounting

All 3,196 source rows have an outcome: 3,144 counties retained, 51 state/DC records and one national record excluded from the county tables. The 22,008 long rows preserve all seven measures per county. Kalawao has source-missing estimates, leaving 3,143 usable counties. Connecticut uses nine planning regions; Puerto Rico and other territories are outside this file.

Measures and uncertainty

Poverty counts and rates cover all ages, children under 18, and related children ages 5–17 in families. These populations overlap. Median household income is a separate household measure in 2024 dollars. Both published 90% confidence bounds remain with each estimate; missing values remain distinct from zero. Annual SAIPE estimates are not interchangeable with ACS five-year statistics, and point-estimate differences alone do not establish significance.

Reprocessing and future editions

Original files, lossless staging, processing code and publication identities are retained. Independent acceptance checks every delivered source-derived field and row outcome; offline replay produces identical output. This fixed edition costs $29. Processing improvements on these source inputs are included for buyers; new source editions may be sold separately after entitlement review. No subscription or refresh schedule is included.

Try the county comparison example · Inspect the product and samples · Free original Census data. Independent preparation; no Census endorsement.