US Macro

Retail Sales (Year-over-Year)

Retail sales measure the total dollar value of goods sold by US retailers each month — the highest-frequency available read on consumer spending behavior. The year-over-year growth rate is the single best near-term indicator of whether the US consumer is healthy or stressed, and consumer spending makes up roughly 68% of US GDP.

May 14, 2026RSAFSconsumer-spending · us-macro · retail · census-bureau

RETAIL_SALES

—

The most timely real-time read on US consumer behavior. Every month, the Census Bureau collects sales data from 13,000 retailers and publishes the result within 15 days — faster than any other consumer-spending data series. Because consumer spending is roughly 68% of US GDP, the retail sales monthly print is one of the highest-information macro releases available. When retail sales grow strongly, the economy is being pulled along by household demand; when they soften, that pull weakens.

RETAIL_SALES

—

What it measures

Retail sales is the monthly dollar value of sales at US retail establishments and food-services establishments:

The underlying series — FRED RSAFS — is the "Advance Retail Sales: Retail and Food Services" (excluding motor vehicles and parts), published monthly by the US Census Bureau. The release is approximately 15 days after the reference month ends, at 8:30 AM ET as part of Census's "Advance Monthly Sales for Retail and Food Services" report.

Our dashboard displays the year-over-year change (computed by our service layer from the level series). Recent readings have been in the +3% to +6% YoY range, reflecting a healthy but cooling consumer. The 2021 peak was over +25% YoY (against the pandemic-suppressed 2020 base); the 2020 COVID trough briefly hit -8% YoY before fiscal stimulus reversed it.

Why it matters

Two angles.

The GDP-input angle. Consumer spending is approximately 68% of US GDP — the dominant component by a wide margin. Retail sales captures roughly 30% of that (the goods-purchase component); the remaining 38% of GDP-contributing consumer spending is in services (healthcare, education, finance, etc.), which are measured at lower frequencies. So retail sales is the highest-frequency window into the consumer-spending component of GDP. A 1pp surprise in monthly retail sales growth typically translates to roughly 0.05-0.10 percentage points on next-quarter GDP estimates. The retail sales release is therefore directly market-moving and feeds into Fed reaction function expectations.

The early-warning-for-recession angle. When consumer spending softens, it tends to show up in retail sales before it shows up in employment data or GDP. The retail-sales YoY series is therefore a leading indicator within the broader business-cycle dashboard. Specifically: a sustained slowdown from +5% YoY toward +1% YoY (or going negative) is one of the cleanest signals of recession proximity. The 2008 recession was preceded by retail sales YoY turning negative in early 2008; the COVID recession was preceded by a near-instant collapse to -8% YoY in April 2020. Watching retail sales monthly is one of the cheapest forms of recession-risk monitoring.

What moves it, and what it moves

Moves retail sales:

Disposable household income. Wages (from NFP) and transfer payments (from fiscal policy) are the dominant drivers of household spending capacity.
Consumer confidence. When consumers are confident about job security and future income, they spend more discretionary income on retail goods. U-Mich and Conference Board confidence indices are leading indicators.
Inflation. High inflation reduces real purchasing power and shifts spending toward essentials (groceries, gasoline) and away from discretionary categories.
Interest rates. Affect housing-related retail (furniture, appliances, building materials) and durable-goods purchases (autos, electronics) through borrowing costs.
Weather. Severe weather events (hurricanes, polar vortices, heat domes) produce 0.5-2% temporary swings in monthly retail sales.
Fiscal stimulus. Direct payments during COVID produced the largest single-month retail-sales surge in the series' history.

Retail sales moves:

Next-quarter GDP forecasts (analysts rebuild estimates after each monthly print).
Retail-sector equity prices (XRT ETF, individual retailer stocks).
Inflation-related Fed policy expectations (strong retail sales = strong demand = upward inflation pressure).
Consumer-discretionary vs. consumer-staples relative-performance rotation in equities.
Bond yields (a strong retail print typically lifts UST2Y by 5-10 bps).
Currency markets (DXY firms on strong retail prints).

A worked example: the 2020-2022 pandemic-and-recovery cycle

US retail sales (the YoY series) entered 2020 growing at approximately +3% YoY — a steady, modest consumer-spending environment.

The COVID crash:

March 2020: retail sales fell -8.7% month-over-month — the largest single-month decline in the series' history
April 2020: -14.7% MoM — the trough; YoY growth had collapsed to roughly -19%
CARES Act fiscal support ($1,200 stimulus checks, expanded unemployment benefits, PPP loans) began flowing in late April-May 2020

The recovery began as rapidly as the decline:

May 2020: +18% MoM rebound
June-September 2020: another +14% cumulative
By September 2020: YoY retail sales growth was positive again, around +5%
The recovery was driven by the unusual combination of fiscal support (more transfer income to households) + reduced spending opportunities (lockdowns reducing services consumption, with the money flowing into goods purchases instead)

The post-COVID boom:

2021: retail sales grew strongly through most of the year. The peak was +25.3% YoY in April 2021 (a base-effect peak — comparing to the April 2020 trough)
Stimulus from American Rescue Plan in March 2021 supercharged spending
Inventory shortages and supply-chain issues drove price increases that further inflated nominal retail sales numbers
2022: as the base effect faded and Fed tightening took hold, YoY growth decelerated to mid-single-digits
2023-2024: settled into the +3-5% YoY range — strong but normalized

The episode demonstrated that retail sales can swing dramatically based on fiscal policy and consumer confidence, but also that the series has powerful base-effect distortions during sharp cycle turns. For policy and analysis purposes, the 12-month moving average of YoY growth is often more useful than the single monthly print.

The current cycle, and the open question

The debates around the consumer:

Bifurcation across income tiers. Recent data suggests upper-income consumers are spending steadily while lower-income consumers are showing strain (rising delinquencies on credit cards and auto loans, declining real wage growth in lower-paid sectors). The aggregate retail sales number can mask this composition shift.
Goods vs services rebalancing. During COVID, spending shifted dramatically to goods (helping retail sales). Post-COVID, the rebalancing toward services has been gradual but persistent. Retail sales (which is goods-heavy) may understate the services-spending rotation.
Inflation-versus-volume. Nominal retail sales growth of +4% YoY with 3% retail-sector inflation means real volume growth of only +1%. The composition matters for what kind of economic activity is actually happening.
Consumer leverage. Household debt levels (credit cards, auto loans, student loans) have been rising in real terms. If consumer balance sheets are weakening, retail sales growth could decelerate even with no income loss.

What you watch: the monthly Census release (typically 15-17 of each month); the YoY growth rate (the most meaningful single number); the "control group" subset (which strips out volatile components and is what feeds GDP); credit card delinquency data from the NY Fed (a complementary read on consumer financial health); and Conference Board / U-Mich confidence indices as leading indicators.

FAQ

What's included in 'retail sales' and what's excluded?

Included: motor vehicles, gasoline, electronics, furniture, building materials, food and beverages (grocery and restaurants), clothing, sporting goods, books, music, general merchandise (Walmart, Target, etc.), and a 'nonstore retailers' category (which is mostly e-commerce — Amazon dominates this line). The series we track — FRED RSAFS — is the 'Advance Retail Sales: Retail and Food Services, Excluding Motor Vehicles and Parts' which excludes auto sales for clarity (auto sales are volatile and reported separately by Census, and inclusion would distort the underlying retail trend). Excluded entirely: most services purchases (healthcare, education, financial services, travel, professional services, housing rent). This is important context: retail sales captures roughly 30% of consumer spending; the broader 'Personal Consumption Expenditures' (PCE) data — published monthly by BEA — captures the full 68% of GDP that consumer spending represents.

How is retail sales measured?

The Census Bureau's Monthly Retail Trade Survey samples approximately 13,000 retail businesses each month and asks them to report total sales for the month. The Census Bureau publishes an 'advance' estimate roughly 15 days after the reference month ends, then revisions in subsequent releases as more establishments respond. The advance release is the most market-moving — it's typically the first comprehensive read on consumer behavior for the prior month, ahead of slower-frequency data like PCE. The release is at 8:30 AM ET and is part of Census's 'Advance Monthly Sales for Retail and Food Services' release.

What's the difference between nominal and real retail sales?

Nominal retail sales is what the headline FRED RSAFS series publishes — dollar values without adjustment for inflation. Real retail sales adjusts for the inflation of the retail-sector basket. When inflation is meaningful (e.g., 2022's 9.1% CPI peak), nominal retail sales can grow strongly even if real consumption is flat or declining. For policy and economic analysis, real retail sales is the more meaningful metric. In 2022, for example, nominal retail sales grew approximately 8-9% YoY at the peak, but real retail sales (after adjusting for retail-sector inflation) was actually negative — consumers were buying fewer physical units, just paying more for them.

How does e-commerce affect retail sales measurement?

E-commerce is captured in the 'nonstore retailers' category, which has grown from approximately 5% of total retail sales in 2010 to roughly 15-17% in 2024. Amazon alone is estimated to account for over 40% of US e-commerce by sales. The Census Bureau's methodology has been gradually refined to better capture e-commerce specifically — there's a separate 'Quarterly E-Commerce Report' that breaks out online sales from physical-store sales. The retail-sales headline includes both, but the composition shift toward online has implications for monthly seasonality, the impact of weather events (online sales are less weather-sensitive than physical-store sales), and the elasticity of consumer spending to economic conditions.

Why is the year-over-year change (the YoY series) more informative than month-over-month?

Two reasons. (1) Seasonality: retail sales are heavily seasonal (Q4 holiday spending, back-to-school, etc.); MoM changes need substantial seasonal adjustment, which introduces methodological complexity and noise. YoY changes naturally control for seasonality. (2) Base effects: comparing December 2024 to November 2024 isn't very informative — December is always bigger. Comparing December 2024 to December 2023 directly tells you whether holiday spending was stronger or weaker than the prior year, which is the question that actually matters for understanding the consumer. The FRED RSAFS series gives nominal levels; the YoY transform we apply in our service computes (current month / prior-year same month) − 1, which is the conventional way to read the series.