Remote work and business AI adoption
This chart tests whether states with more working from home also report more business AI use. The relationship is positive and survives several simple sensitivity checks, but it combines differently timed Census surveys and does not show that remote work causes AI adoption.
What does it show?
States with more remote work also report more business AI use, although the chart cannot show which one causes the other.
Methodology
Matches 52 geographies (50 states, DC, and Puerto Rico) with usable values from two Census products. The x-axis is the ACS 2024 share of workers age 16+ who usually worked from home as their primary commute mode. The y-axis is the latest BTOS state estimate (2026-08-09) of employer businesses using AI in any business function during the prior two weeks. The dashed line and headline statistics are an unweighted OLS fit across geographies. The chart also reports a Spearman rank correlation, a sensitivity test excluding DC and Puerto Rico, and a workforce-weighted fit using each geography's ACS worker count. These are cross-sectional associations between differently timed surveys—not evidence that remote work causes AI adoption. BTOS state estimates have sampling error and suppressed observations are omitted; ACS primary commute mode understates occasional and hybrid remote work.