When ACS is too coarse
Some tract-level metrics in a report would be too sparse to publish if read directly from ACS — the survey’s confidence intervals get wide at finer geographies. In those cases, PriceDNA falls back to a documented chain: tract → county → metro → state → national. The fallback is visible: if you see the same number across the 1-, 3-, and 5-mile views on a report, the data resolution is broader than the radius, and PriceDNA is keeping the value consistent rather than inventing precision.
census.gov/programs-surveys/acs