Data & Accountability • Friday, November 20, 2026

Census Business Survey Shows Firmer Demand Expectations, With Limits

November 19's dashboard marks a statistically significant rise in the future-demand index; the figures are survey indexes, not growth rates.

Tabitha Tally — approved human correspondent in a representative reporting setting

Illustrative correspondent image; not a photograph of the reported event.

By Tabitha Tally • Friday, November 20, 2026

The Census Bureau's November 19 Business Trends and Outlook Survey dashboard shows an improvement in expectations for demand, alongside a mixed picture of current business conditions. Using data as of November 15, the future-demand index reached 51.5, up 1.2 points from the preceding collection period. The dashboard marks that change as statistically significant at the ninety-five-percent confidence level. That is a specific survey finding, rather than a prediction that national sales will increase by 1.2 percent.

Other readings make the picture more complicated. Current performance was 54.9, while the revenue index was 40.8 and the employee index was 47.3. The input-price index stood at 71.2. The dashboard separates conditions reported for the previous two weeks from expectations covering the next six months. Those different time windows mean that current pressure and greater optimism can appear in the same release without representing contradictory measurements of a single period.

An October Federal Reserve research note explains the logic of diffusion indexes using the survey's manufacturing data. Responses indicating increase, no change and decrease receive scores of one hundred, fifty and zero before averaging. For those change measures, fifty is the balance point between the directions. The approach describes how broadly firms report a movement; it does not measure the dollar value of their combined sales or the magnitude of every company's change. An index therefore answers a different question from an official growth-rate estimate.

The Census methodology also shows why the survey can publish so frequently. Approximately 1.2 million businesses are sampled annually and assigned to six panels, with roughly 200,000 businesses invited in each two-week period. A business stays in its selected panel across cycles and receives one invitation per cycle. Invited cases are not equivalent to completed responses. Readers assessing precision should use the published estimates and uncertainty information rather than assume that every sampled company answered every release.

The program tracks nonfarm employer businesses in covered sectors and provides geographic and industry detail that can reveal conditions hidden by a national figure. Its sample and question content have changed over time, which makes documentation important when constructing a trend. The bureau itself warns on the dashboard that question wording has changed. A comparison should keep the measure, geography, reference period and wording consistent before drawing a conclusion about how far an industry or state has moved.

For communities trying to understand business conditions before slower statistics arrive, the release provides a useful pulse. The demand finding deserves attention because the bureau identifies a meaningful change in that index; small movements elsewhere should not automatically receive the same description. The next releases will show whether the signal persists. For now, the defensible reading is limited but informative: expectations improved on one published measure while current revenue, employment and cost indicators continued to tell different parts of the business story.

Sources

Follow the Entire Crew

Follow @FrontPageCrew for daily reporting and newsroom updates.