October 19, 2026
There’s a particular kind of economic gaslighting that happens when the headlines say business is expanding and your grocery receipt says otherwise.
The latest numbers seem to tell two different stories. The ISM Services index came in at 54.9 in September. Above 50 means expansion, and 54.9 is a solid reading. Meanwhile, the University of Michigan’s consumer survey showed overall sentiment at 48.1, its second-lowest reading on record. Its Expectations Index was even lower, at 46.3.
So which is it: a growing economy or a gloomy one?
Both. That’s the puzzle, and you’re not imagining the gap.
Two numbers, two different questions
The ISM Services index surveys purchasing managers at service businesses. It asks whether things like activity, new orders, employment and prices are improving or worsening compared with the previous month. A reading above 50 means more respondents reported improvement than deterioration. September’s 54.9 says service-sector activity was expanding.
That’s useful information about what businesses are doing now. It is not a vote on whether a household feels comfortable paying the bills.
The University of Michigan survey asks households how they see their own finances and the economic outlook. The headline Consumer Sentiment Index fell to 48.1 in September. The 46.3 figure is the Expectations Index, one component of the broader measure. Both are weak; the distinction matters if we want to describe the numbers accurately. The survey’s September results show a public that is particularly worried about what comes next.
One survey asks, “Are businesses getting busier?” The other asks, “Do people feel okay about their money?” Those questions can have different answers without either survey being broken.
Busy and broke can happen at the same time
A restaurant can have more customers, higher sales and a full reservation book, and still hear every table complain about the bill. The restaurant is measuring activity. The customers are measuring affordability.
That’s the first big distinction: activity measures how much is happening; sentiment reflects what it feels like to afford it.
A business can report rising sales because customers are paying more, buying more, or both. But when a family’s food, rent, insurance and electricity bills climb, the extra dollars going out the door may not feel like prosperity. “We spent more” and “we got more comfortable” are not the same sentence.
Your inflation isn’t necessarily the average
The inflation number everyone hears about is an average across a broad basket of goods and services. But no household buys the average basket in exactly the same proportions.
Some expenses are occasional: a new television, a new car, a big appliance. Others show up every week or every month. Food, gasoline, rent, utilities and insurance can weigh heavily on a household budget. If those essentials are the categories that have become more expensive for you, the official average may not match what your wallet is experiencing.
That doesn’t mean the overall inflation measure is fake. It means an average is a summary, not a personal receipt.
There’s another bit of economic vocabulary that often gets lost: the inflation rate is not the price level. Inflation measures how quickly prices are changing. The price level is what those prices have reached.
If inflation slows, prices are still rising, just more slowly. Prices generally fall only when there is deflation, and broad price declines aren’t what people should count on. As the St. Louis Fed explains, easing inflation does not undo the price increases that already happened.
That’s why the memory of 2019 still matters. A shopper doesn’t compare today’s groceries only with last month’s prices. They remember what the same items used to cost. If the price level stays higher, that comparison keeps stinging even after inflation cools. Lower inflation is welcome news; it is not a time machine.
One survey looks forward; the other mostly looks around
Business activity data tells us what’s happening in the current month. Consumer sentiment is heavily influenced by what people expect will happen to them next.
If you’re worried about your job, your next rent renewal, borrowing costs or another jump in household expenses, a busy restaurant district won’t necessarily calm you down. September sentiment reflected concerns about high prices and interest rates, among other pressures.
And it’s reasonable for a household to care about uncertainty. A person deciding whether to change jobs, replace a car or take on a mortgage isn’t making that decision based on a services-sector index. They’re thinking about their own paycheck and the costs they expect to face.
The K-shaped economy is more than a chart
Some households own homes, stocks or businesses that have gained value. Others depend mostly on wages and have little room between their paycheck and their bills. Those groups can experience the same economy very differently, and they answer the same sentiment survey.
The contrast showed up in earnings data, too. According to the BLS real earnings reports, real average hourly earnings weakened during Q2. The Hamilton Project’s analysis describes two consecutive quarterly declines in inflation-adjusted hourly pay. The first such stretch since 2022. That is not the same as saying every worker got a pay cut. It does mean prices outpaced pay on that measure during the period.
Meanwhile, a strong market can lift the balance sheet of someone who owns financial assets without doing much for someone who has to buy groceries this afternoon. The economy’s gains don’t land evenly.
Growth totals can hide that, too. Real GDP grew at a 1.5% annual rate in Q2, according to the Bureau of Economic Analysis. AI-related investment has become an important part of the growth story. Our working estimate is that growth would run closer to 1% if AI-related investment were stripped out. That is an analytical counterfactual, not an official BEA statistic, and the result depends on what spending gets counted as AI-related.
The larger point is simpler: a headline growth figure can rise even if the typical household doesn’t feel a matching improvement in its finances.
The mood has politics in it, but isn’t only politics
Consumer sentiment has been detached from some hard economic measures for years. Partisan views can influence how people answer questions about the economy, whichever party is in power. That is one reason not to treat a survey as a clean measure of household finances.
But politics doesn’t explain away the price-level problem. People have long memories for the bills that went up and stayed up. When a survey asks whether families feel better off, respondents are not required to forget what their rent, groceries or insurance cost a few years ago.
Which number should you trust?
Both, but for different reasons.
The ISM index tells us whether a large part of the business machine is running. Consumer sentiment tells us how people inside that machine feel about their finances and what they expect next. One is not a rebuttal to the other.
A machine can run well while its passengers are uncomfortable. But if households feel squeezed for long enough, they may delay purchases, avoid taking on debt or cut back on services. That’s why forecasters watch sentiment as a possible early warning, even though it’s a survey of feelings rather than a count of transactions.
A strong activity reading deserves attention. So does a deeply unhappy public. The useful question isn’t which number wins. It’s whether business strength is turning into paychecks and breathing room for more households, or whether the good news is staying mostly on the business side of the table.
Be mindful, be watchful and good luck.
Educational and investment disclaimer: This article is for educational and informational purposes only. It is not investment, financial, tax, or legal advice. Economic conditions and individual circumstances vary. Speak with a qualified professional before making financial or investment decisions.