September’s retail sales report looks stronger in dollars than it does as a measure of household comfort. Rising gasoline costs can lift the value of sales while leaving consumers with less money for everything else. Add concerns about expensive borrowing, and the central question becomes whether spending is being supported by income or increasingly stretched household finances.
The evidence calls for caution. It does not, by itself, establish an economy-wide collapse in consumption.
NSA retail sales: the starting point for our analysis
Our analysis focuses on non-seasonally adjusted (NSA) retail sales: estimated dollar spending before adjustments for recurring seasonal, holiday and trading-day patterns. On this basis, our September comparison shows a 3.154% month-on-month decline and a 2.633% year-on-year increase, before adjusting for inflation.
The two comparisons answer different questions. Spending fell from August but remained above its year-earlier level in dollar terms. The monthly decline includes seasonal and calendar effects; the year-on-year increase must also be assessed against the change in prices before drawing conclusions about purchasing power.
The composition of spending matters alongside the total. Sales excluding motor vehicles help assess demand beyond large vehicle purchases, while gasoline-station sales show how much retail expenditure is being directed toward fuel and related purchases. These categories overlap and should not be added together as separate contributions to the total.
Our concern is that higher essential costs can sustain dollar turnover while squeezing the spending capacity available elsewhere. A rising gasoline share would reinforce that concern, but gasoline-station revenue should be distinguished from the quantity of fuel purchased and from gasoline’s share of total household consumption.
The seasonally adjusted headline: a separate comparison
The U.S. Census Bureau’s October 15 release reports September 2026 retail and food-service sales of approximately $745 billion, up 0.8% from August and 6.3% from September 2025. These headline figures are seasonally adjusted but are not adjusted for inflation. August’s monthly increase was revised to 1.3%.
The SA headline and our NSA comparison should be read separately. They may differ because of adjustment treatment, series coverage and data revisions. Like-for-like comparisons require the same retail definition, reference periods and release vintage. Neither seasonal adjustment nor the absence of it removes inflation.
Reader’s guide: NSA, SA, nominal and real
- NSA — non-seasonally adjusted
- Estimates before seasonal, holiday and trading-day adjustments. These retain calendar-related movements as well as changes in economic conditions.
- SA — seasonally adjusted
- Estimates adjusted to help identify movements beyond recurring seasonal and calendar patterns.
- Nominal — current dollars
- Spending measured at the prices prevailing at the time. Both NSA and SA retail-sales figures can be nominal.
- Real — inflation-adjusted
- Spending converted into comparable purchasing-power terms using a price index. Real does not mean unadjusted.
Methodology note: seasonal adjustment and inflation adjustment are separate operations. Either an NSA or an SA series can be expressed in real terms with a suitable, consistently treated deflator. This article prioritizes the NSA perspective while reporting the SA headline separately. See the Census Bureau’s retail definitions.
Energy is the pressure point
BLS’s September CPI report shows overall consumer prices rose 0.6% month over month and 3.7% year over year. Gasoline prices rose 9.1% on a seasonally adjusted monthly basis, accounting for more than half the monthly increase in the overall index.
For households that need a vehicle to work, higher fuel bills are difficult to avoid immediately. Unless income rises or savings absorb the difference, more spending at the pump leaves less room for restaurants, clothing, household goods and other purchases.
Businesses face a related problem. Transportation and distribution costs can pressure margins or feed through to selling prices. The effect depends on contracts, hedging, competition and the ability to pass costs to customers. Energy therefore matters beyond its direct weight in the consumer basket.
From nominal sales to real purchasing power
There are two consistent ways to examine purchasing power: calculate inflation-adjusted growth directly, or convert the entire sales series into constant dollars and then calculate its growth.
Exact real growth (%) = [(1 + nominal growth / 100) ÷ (1 + inflation / 100) − 1] × 100. Subtracting the inflation rate from nominal growth provides an approximation; its accuracy depends on the rates involved, rather than a fixed four-year time limit.
For illustration, combining our NSA nominal year-on-year increase of 2.633% with an all-items CPI deflator of 3.7% gives approximately −1.029% real growth. Simple subtraction gives approximately −1.067%. This illustration uses a broad purchasing-power deflator rather than a retail-specific price index.
For a full series expressed in a common reference period’s dollars, the calculation is: nominal sales in each period × reference-period CPI ÷ that period’s CPI. Calculating growth from this deflated series gives the same result as the exact growth formula when the inputs are consistent. The BLS guide to purchasing power and constant dollars explains this ratio method.
All-items CPI includes services and housing that do not match the retail-sales basket. It is therefore a broad purchasing-power proxy, rather than an official retail-volume measure. Sub-sector analysis also requires care: gasoline revenue deflated by all-items CPI measures something different from fuel quantities estimated using gasoline prices.
Monthly, year-on-year and cumulative year-to-date comparisons should retain separate labels. For real year-to-date analysis, deflate the monthly observations consistently before aggregating and comparing matching periods. A year-to-date comparison with the same months of the previous year is not automatically an annualized growth rate.
The BEA’s personal consumption expenditures data provide the broader framework for goods and services spending, including real consumption. Retail sales alone cannot describe the entire consumer economy.
The Kansas City Fed adds a consumer-credit warning
The Kansas City Fed’s October 15 Economic Bulletin listing introduces research titled Financially Distressed Consumers May Have Turned to High-Cost Credit to Cope with Inflation. Its published summary associates financial distress with inflation-related borrowing and more expensive credit.
The economic concern is straightforward: borrowing may preserve purchases today while committing future income to repayments and interest. If this becomes persistent, spending can remain superficially resilient even as financial flexibility deteriorates.
That is a warning to investigate, rather than evidence that all consumers are distressed or that a banking crisis has begun. Financial difficulty, late payments, charge-offs and non-performing loans are distinct indicators. The scale and direction of each matter.
Automotive demand deserves particular attention
Vehicles combine a large purchase price with financing, insurance, maintenance and fuel costs. This makes the sector a useful place to monitor affordability pressure.
However, a severe automotive crisis cannot be inferred simply by comparing total retail sales with sales excluding vehicles. A stronger diagnosis requires vehicle unit sales, incentives, inventories, lending conditions, delinquencies and manufacturer margins. Weakness in one measure should be tested against the others.
Iran-related risk: follow the economic transmission
For markets, the relevant channels from a prolonged conflict involving Iran are energy supply, shipping reliability, insurance costs, public spending and uncertainty. A sustained disruption could hurt purchasing power and complicate inflation control; de-escalation could ease those pressures.
Defense expenditure also carries an opportunity cost, although economic effects depend on how spending is financed and where it occurs. The useful analytical question is whether security developments increase or reduce the economy’s energy and financing burden.
What would strengthen or weaken the consumer warning?
- Real income: sustained wage and disposable-income gains would help households absorb higher essentials costs.
- Spending breadth: weakness spreading across discretionary categories would be more concerning than an isolated decline.
- Credit quality: rising late payments and losses would make borrowing-led resilience less reassuring.
- Energy: lasting relief would release spending capacity; another sustained rise would tighten the squeeze.
Why this matters for markets
Consumer strain can coexist with strong performance in selected corporations and equity indexes. Investors may focus on future productivity, global revenues or durable earnings while households adjust to higher living costs.
Our view: energy costs and the quality of consumer financing deserve close attention, even when headline spending rises. The important issue is how much genuine purchasing power remains after essential bills and debt service.
Read the companion analysis: Why Markets Keep Recovering Despite Economic Warning Signs. For the inflation background, see our September CPI analysis.