Consumer sentiment just dropped 6% in a month — but people's inflation forecast actually eased

The University of Michigan's Surveys of Consumers released its final August 2026 reading on August 28, and the headline number moved the "wrong" way relative to what people said they expect prices to do. The Index of Consumer Sentiment fell to 51.7, down from 55.2 in July — a roughly 6% drop in a single month, and about 11% below where it stood a year earlier. At the same time, the survey's year-ahead inflation-expectation question actually eased, to 4.0% from July's 4.2%. Here's what the survey actually measures, why those two numbers can move in opposite directions in the same release, and why the Federal Reserve pays attention to the expectations piece specifically.

What this survey is — and isn't

The University of Michigan's Surveys of Consumers is a monthly telephone survey that asks a sample of U.S. households how they feel about their own finances, the broader economy, and where they expect prices to head. It is not a government price measurement. That's a meaningful distinction from the inflation numbers this site has covered from CPI and PCE: those are built from actual transaction and pricing data collected by the Bureau of Labor Statistics and Bureau of Economic Analysis. The Michigan survey instead asks people what they expect — a forecast of sentiment and belief, not a measurement of what already happened. Both kinds of numbers are useful, but they answer different questions: CPI and PCE tell you what prices did; this survey tells you what people think prices will do, and how they feel about it.

What the August reading showed

The headline Index of Consumer Sentiment fell to 51.7 in August's final reading, per the survey's own release, down from 55.2 in July — about a 6% monthly decline, and roughly 11% below August 2025's reading of 58.2. The two sub-indexes that make up the headline number both declined: the Current Economic Conditions index fell to 51.9 from 54.8, and the Index of Consumer Expectations fell to 51.5 from 55.4. The survey covered interviews conducted July 28 through August 24, 2026.

The same release carries the survey's inflation-expectations questions, which moved differently. The year-ahead inflation expectation — what respondents think prices will do over the next 12 months — eased to 4.0%, down from 4.2% in July. That's still elevated by recent standards: per the survey's own release, the current reading "substantially exceeds" the 3.4% seen in February 2026, before the Iran conflict began, as well as every reading from 2024. The longer-run expectation, covering the next 5 to 10 years, held at 3.3% for a third straight month, unchanged.

So in the same survey: how people feel about the economy got noticeably worse, while what people expect inflation specifically to do over the next year got slightly better. Both are real findings from the same release, not a data error.

Why sentiment and inflation expectations can diverge

These aren't the same question. Sentiment reflects a household's overall read on its own finances and the broader economy — job security, business conditions, whether now is a good time to make a big purchase. The inflation-expectations questions ask specifically about the expected path of prices. A household can grow more pessimistic about the economy generally — worried about jobs, business conditions, or their own financial position — without necessarily raising their specific price forecast for the next 12 months, and vice versa. The survey's own August release attributes the broader sentiment decline to persistent concern about high prices' effect on household finances and a weaker outlook for business conditions, even as the specific year-ahead price forecast ticked down.

One note on the inflation-expectations figure specifically: some news coverage earlier in August cited a preliminary reading of 4.3% for year-ahead inflation expectations, based on the survey's mid-month partial data. The 4.0% figure here is the final, end-of-month reading, covering the complete survey period through August 24 — a normal revision as more responses come in, not a conflicting or corrected number.

Why the Fed watches this specific number

The Federal Reserve's own inflation target is explicitly defined around the PCE price index — the actual measured data, not a survey. But policymakers also track survey-based inflation expectations, including this one, as a separate signal: whether households and businesses still believe the Fed will bring inflation back to target over time. Economists generally treat expectations as partly self-reinforcing — if people expect higher prices to persist, that belief can show up in wage negotiations and pricing decisions, making the expectation more likely to hold. A reading like August's, where the year-ahead expectation eased even as broader sentiment soured, is one data point the Fed can weigh alongside its own CPI and PCE data. It headed into its September 16, 2026 meeting with that reading in hand — and went on to raise its target range a quarter point to 3.75%-4.00%; see this site's rate-hike explainer for what changed.

None of this tells you what the Fed will actually decide. One survey, in one month, is an input — not a preview of a rate decision.

What this doesn't tell you

This survey doesn't measure what prices actually did — for that, see the CPI and PCE reports this site already covers, both built from real transaction and pricing data rather than expectations. It also doesn't describe any individual household's experience; it's a national survey sample, and sentiment can vary widely by income, region, and other factors the headline number doesn't capture. And a single month's move — in either the sentiment index or the expectations figure — doesn't establish a new trend on its own.

How we sourced this

Every figure above comes directly from the University of Michigan Surveys of Consumers' own published release for August 2026. We're showing our work here for the same reason we publish our scoring methodology and disclose how we make money elsewhere: a number is more useful when you can see exactly where it came from and check it yourself.

The takeaway

Consumer sentiment fell sharply in August 2026 — down about 6% for the month, per the University of Michigan's own final reading — while the same survey's year-ahead inflation expectation actually eased slightly, to 4.0% from 4.2%. Those two findings aren't a contradiction; they're answers to two different questions in the same release. The Fed watches the expectations figure specifically as a signal of whether the public still trusts it to bring inflation back to target — but one month's survey, like one month of any economic data, is an input, not a forecast of what happens next.