The September fall in US consumer confidence is a warning about household expectations, not yet a measurement of September purchases. The Conference Board reported an index of 81.9, down 6.7 points from August's 88.6. Reuters placed the reading near a 12-and-a-half-year low. The survey's message deserves attention because both views of current conditions and the next six months deteriorated. It should not, however, be translated directly into a fall in retail revenue.
There is a timing problem at the heart of the interpretation. The September survey collected preliminary responses through September 23. The latest broad spending numbers available when the result was published describe August, before most of the period households were asked to assess. Those August data were positive. The financially useful question is whether weaker confidence will eventually reach actual spending, and through which household constraints.
Present conditions and expectations both weakened
The Conference Board's September release shows the Present Situation Index falling 7.9 points to 109.3 and the Expectations Index dropping 5.9 points to 63.6. That combination matters more than a single headline level. Respondents were not merely less optimistic about the future; their assessment of business and labor conditions now also softened. The release says current business conditions were viewed negatively on net for the first time since September 2024.
This is a survey of attitudes and intended behavior. It is not a tally of what households spent, and its components have their own questions and sampling limits. The Conference Board says the preliminary results came from an online sample and lists September 23 as the cutoff. Write-in responses frequently referred to high prices, oil and gas, but those reports of concern do not establish that any single price or policy event caused the index's movement. Political views, income and lived financial experience can all affect responses.
The release also found softer plans for many discretionary services, while planned vacations held up modestly. These are intentions, not orders or receipts. For a business exposed to travel, dining or household goods, the survey can suggest where to look next. It cannot calculate the company's revenue from a confidence index alone.
The employment differential is the transmission point
Perceived job availability gives the survey a potential link to consumption. The share calling jobs plentiful slipped while the share saying they were hard to get rose. Their difference narrowed to 1.7 percentage points, according to the Conference Board. A household worried about replacement income may delay a large purchase even before anyone loses a job. That is an economic mechanism, not a claim that the survey has already measured a fall in employment or that every respondent will cut spending.
The latest official jobs release covers a different month. The Bureau of Labor Statistics reported 162,000 additional payroll jobs in August and an unchanged 4.1% unemployment rate. Those numbers do not invalidate September's worsening perceptions: they precede most of the survey period and measure employment rather than sentiment. They also prevent a simple assertion that the September survey itself proves a current labor-market collapse. The next employment figures, and revisions to the earlier figures, are better tests of whether anxiety and observed conditions are converging.
The same distinction matters for pay. A person can be employed and still feel squeezed by living costs or the prospect of a weaker job market. The Conference Board reported that net expectations for household income remained positive but declined. That combination is compatible with continued spending alongside caution. The article's inference is that the risk may first show in postponed optional purchases, not necessarily in a simultaneous drop across all categories.
August purchases tell a different, earlier story
The Bureau of Economic Analysis reported that August personal consumption expenditure rose 0.9% in current dollars and 0.6% after adjusting for prices. Real disposable personal income was flat on the month. Those are actual expenditure estimates, albeit subject to revision. They show that spending had not already fallen in August, while the flat real-income reading also cautions against assuming the same pace can continue indefinitely.
The Census Bureau's advance retail and food-services release also showed August sales up 1.2% from July. That retail measure is nominal, narrower than BEA's full household consumption measure, and explicitly not adjusted for price changes. It should not be added to BEA's figure or described as a second estimate of real consumption. The two releases instead provide separate checks that August transactions were still moving upward in their respective scopes.
An apparent contradiction between the September survey and August purchases therefore may simply be a calendar difference. It may also reflect the familiar gap between what consumers say and what they do, or a shift toward necessities while some discretionary categories weaken. The present evidence does not decide among those explanations. A retailer's own comparable sales, transaction counts and mix would help, but one broad index cannot provide them.
The gap closes only with later behavior
The strongest case for taking the confidence slump as an early economic signal is that August purchases occurred before September's survey deterioration. If later payrolls, real income and inflation-adjusted consumption weaken together, the survey will look more like a timely warning. If transactions and labor income remain resilient, the confidence fall may describe financial stress without a matching aggregate spending contraction. Both are conditional interpretations, not forecasts.
For investors, the order of evidence is the discipline. September confidence identifies a household risk, especially around job security and optional spending. August BEA and Census data establish that broad purchasing had still grown in the prior month. Subsequent monthly releases and company sales disclosures can test whether the warning travels through household budgets into revenue. Until then, calling the survey either proof of recession or irrelevant noise would overstate what the data show.