Retail Sales Fell 0.6% in July; Consumer Sentiment Index Fell to 51.0 in Early August

The Wrong Number Fell First

Retail sales dropped 0.6% in July, the Census Bureau said on August 14, 2026 — the first monthly decline in nine months and the steepest since May 2025^[1][3]. Americans still spent $763.6 billion at stores and restaurants, up 5.0% from a year earlier^[1][3]. That same morning, a second report landed: the University of Michigan's early August reading on how people feel about the economy fell to 51.0, down from 55.2 in July, an 8% drop that erased two months of gains^[2][7].

Here is the part that trips up both sides of the argument. The retail number is about July. The sentiment number is about August. They are not measuring the same month, even though nearly every story about them treats them as one event^[2][7]. That gap matters, because the thing everyone blames for both — the U.S. and Israeli strikes on Iran, which began February 28, and Iran's retaliation against shipping in the Gulf — hit fuel prices hardest in August, not July^[7][9].

So the July spending drop and the August mood drop may not share the same cause at all. Untangling that is the real story here, and it is where the politics comes in.

Two True Numbers That Pull Apart

Start with what nobody disputes. Sales fell 0.6% from June to July^[1][3]. And sales are still 5.0% higher than a year ago, with the May-through-July stretch running 6.3% above the same months in 2025^[1]. A single soft month, sitting inside a year of overall growth, can be read two ways depending on which one you're looking at.

There's a wrinkle worth knowing before either number means much: retail sales are counted in dollars spent, not in the number of things people bought^[1]. When prices go up, the dollar total can rise even if people are actually buying less. So the 5.0% annual gain isn't a clean measure of more shopping — some of it is just higher prices. Strip out the 3.4% inflation rate over the same period, and the real increase in stuff bought is closer to 1.6%^[1][8].

Now look inside July's decline. Car sales fell 1.8% and gas station sales fell 0.9% — those two categories did most of the damage^[10]. Restaurants and bars, the spending people usually cut first when they're worried about money, actually rose 0.5%^[10]. Take out cars and gas entirely, and sales fell just 0.2%^[10]. That looks less like a household in retreat and more like people skipping one big purchase while still going out to eat.

One more asterisk: this is an "advance" estimate, built from a smaller, faster sample than the final numbers Census publishes later, and it carries a margin of error of plus-or-minus 0.4 percentage points^[1]. So the true July drop is probably somewhere between about 0.2% and 1.0%. It's a real decline. Its exact size is still uncertain.

What People Expect Is Not What's Happening

The sentiment survey is a different kind of number. It doesn't measure spending at all — it asks people how they feel about the economy, on a scale where January 1966 equals 100^[2]. And what people said in early August is that they now expect prices to rise 4.3% over the next year, up from 3.4% right before the war started^[2][7][9].

That 4.3% figure is the one both political camps keep pointing to, and it's worth understanding why it carries so much weight. Actual inflation in July ran at 3.4%, and it was ticking down, not up — a tenth of a point better than June^[8]. So the thing that actually got worse in August wasn't prices. It was what people expect prices to do next.

That distinction is not just semantics. Economists watch inflation expectations closely because they can become self-fulfilling: if workers believe prices will keep climbing, they push for bigger raises; if businesses believe the same, they raise prices ahead of their own rising costs. The belief itself can help cause the outcome. That is the exact mechanism the Federal Reserve worries about, and it's why a rate increase — not a cut — was even on the table in a month when spending was falling^[11].

Only 8% of consumers now expect their income to grow faster than prices over the next year^[2]. That means 92% expect to lose ground. And the pain isn't evenly spread: Michigan's data show the sharpest declines among older consumers, lower-income consumers, and people without a college degree^[2][9]. It also cut across party lines — but Republicans posted the steepest drop of any group, with their sentiment now running 19% below where it stood before the strikes began^[9][13].

An Outside Shock, or a Choice That Was Made

This is where the coverage splits, and both sides are working from the same set of facts.

The administration's allies argue this is a shock imported from overseas, not a policy failure at home. Their case rests on three points. First, this isn't a demand collapse — sales are still up 5.0% over last year, so one soft month inside a rising trend is closer to noise than a turning point^[1]. Second, measured inflation actually improved in July, down to 3.4%^[8]. What worsened was how people feel about future prices, and that tracks gas pumps and war headlines more than economic reality. Third, the categories that dragged the number down — cars and gas — are volatile and price-driven, not signs of families going without^[10]. The Daily Caller's headline put it plainly: the war "casts a shadow" over the economy, with the war as the actor and the administration absent from the frame^[5]. The Washington Examiner led with "uncertainty about Iran," a word choice that frames this as a feeling that could resolve on its own, rather than a cost that has to be paid^[9].

Democrats and much of the mainstream press argue the shock was a choice, not something that happened to the country. The U.S. and Israel started the strikes on February 28; the resulting disruption to a region that moves roughly a fifth of the world's oil and gas was a foreseeable consequence, not an accident^[7][9]. Reuters wire copy, syndicated widely, described households as "battered by fallout from President Donald Trump's Iran war" and tied the numbers directly to the November midterms^[3][4]. That possessive — "Trump's Iran war" — assigns ownership of the shock inside what is otherwise a straight news report, which is its own kind of framing choice. CNN's headline centered consumer "frustration," importing an emotion from the sentiment survey into the separate retail story^[4].

A third, quieter reading, favored by some economists, skips the war argument almost entirely: households have been squeezed by years of elevated prices, and July was simply the month the spending finally cracked^[4][9].

Chinese state media, in outlets like Xinhua, took a different approach than either American camp — flat, numbers-heavy prose with almost no adjectives, but a causal chain that runs only one direction: American strikes on Iran, leading to American household pain, with no U.S. official quoted offering a competing view^[7].

The Fed's Bad Timing

All of this lands on the Federal Reserve's desk ahead of its September 15-16 meeting, and it's an unusually awkward moment for the central bank^[3].

Normally, weak spending numbers make a rate increase less likely, because raising rates makes borrowing more expensive — for car loans, credit cards, mortgages, business debt — which cools spending further. Nobody raises rates into a slowdown on purpose. But the Fed isn't just watching spending. It's watching those inflation expectations, because if expectations climb high enough, they can become self-fulfilling before actual prices ever move.

That's the bind: weak retail sales argue for holding rates steady or cutting. Rising inflation expectations argue for raising them, to head off a problem before it starts. The Fed's benchmark rate has sat at 3.50%-3.75% since it held steady in July^[3][11]. After these two reports came out, market-implied odds of a September increase fell to about 30.6%, down from 33.9% the day before and roughly 50% a month earlier^[3]. That's a real shift, but it still leaves a meaningful chance of a hike on the table — in a month when the headline economic story is that spending fell.

What Comes Next

Both numbers in this story carry a warning label their coverage often drops. The July retail figure is an advance estimate that will be revised^[1]. The August sentiment reading is preliminary too — and last month's preliminary reading was revised upward after the fact^[6]. Sentiment and actual spending have diverged before: people have told surveys for years that the economy feels terrible while continuing to shop anyway^[1][2].

What isn't in dispute is the direction of one number: households now expect 4.3% inflation over the next year, up from 3.4% before the strikes began^[2][9]. If oil markets settle down, that expectation could ease on its own. If it doesn't, it becomes the Fed's problem to solve — with a tool that cools prices by also cooling the spending that just came in weak.