Event-driven breakdowns of economic releases. What's real, what's noise, what it means.
Released May 14 · Source: U.S. Census Bureau Advance Monthly Retail Trade Survey (CB26-78); values verified against FRED series RSAFS
U.S. retail and food services sales rose 0.5% in April from March, to a seasonally adjusted $757.1 billion. March was revised down to +1.6% from a previously reported +1.7%. February stands at +0.9%. The three-month average — February, March, April — runs at +1.0% per month. Year-over-year, headline sales are up 4.9%.
Two different signals sit inside one release. The Census headline — retail and food services — covers what U.S. consumers spent at retail establishments and at restaurants and bars combined. A separate cut, retail trade only, strips out food services to isolate goods spending; that subset rose 0.5% in April and is running 5.2% year-over-year. For the consumer-spending pulse including restaurants and bars, the headline is the right number, and it shows April at +0.5% after March's revised +1.6%. For the goods-only read, the retail-trade-only number is the right one, and it shows broadly the same monthly pace with a stronger annual figure.
The composition inside the release tells a sharper story than the topline. Nonstore retailers — the bucket that captures e-commerce and direct-to-consumer — are running 11.1% year-over-year. Food services and drinking places are running 2.7% year-over-year. The two ends of the consumer basket are growing at materially different speeds: how households spend has shifted toward online and away from in-restaurant, even when total spending growth is steady.
The +0.5% April print following March's +1.6% reads optically like a deceleration. The three-month average is the correct frame, not the month-to-month bounce. March was the strongest of the three months; April giving some of it back is what monthly mean reversion looks like. The three months read together as a single early-spring stretch averaging roughly 1% per month, not as a sharp inflection.
March's downward revision — from +1.7% to +1.6% — is a small adjustment, but Census flags it explicitly in the release because revisions matter. Advance monthly data is subject to revision in subsequent releases as more establishments report. A single tenth is not a story; the pattern of revisions over time is.
Retail sales are reported in nominal dollars and are not adjusted for inflation. A reader comparing this release directly to real consumer-spending measures will get a misleading picture; some portion of the year-over-year increase reflects price growth rather than volume growth. The release does not separate the two.
The April release shows a consumer that has slowed from March's pace but has not rolled over. Headline growth cooled month over month; the three-month average held near 1%; year-over-year growth ran at 4.9%. The household-spending pulse is softer than March but firmer than the late-2025 stretch when monthly prints clustered near zero.
The composition matters as much as the level. The 11.1% year-over-year pace for nonstore retailers against 2.7% for food services and drinking places is a durable shift in where consumer dollars land, not a monthly quirk. Even when the total grows steadily, the channels underneath it are moving at very different speeds.
The release tells us what U.S. consumers spent at retail and food-service establishments in April. It does not tell us what they spent on services more broadly, which is a larger share of total consumption and is captured elsewhere. It also does not tell us how the spending was financed — whether from income, savings, or credit — which is the question that determines durability.
The headline cooled; the trend did not change.