Spending Rose $36.3 Billion in July. Adjusted for Prices, $1.3 Billion.

The Bureau of Economic Analysis published July's personal income and outlays report on August 26. Consumer spending went up $36.3 billion, which is the number that made the headlines.

Adjusted for prices, it went up $1.3 billion.

That is the same month, the same release, two lines apart. Households spent measurably more money and came home with almost exactly as much stuff as they had in June. If you sell things for a living and July felt busy in the till and flat in the stockroom, this is the national accounts telling you that you read your own month correctly.

Nominal and real are not two views of one number¶

The distinction doing all the work here is old and boring and gets flattened in almost every write-up of a spending report, so it is worth thirty seconds.

Current-dollar (nominal) spending is what actually left people's accounts. Real spending is that same activity restated at constant prices, so that a change in the number means a change in the quantity of goods and services bought rather than a change in what they cost. When the two diverge, the gap is price.

In July they diverged almost completely. Spending in dollars rose 0.2%. Spending in volume rose less than 0.1%. The dollars moved; the cart did not fill up.

Goods lost $49.9 billion. Services took $86.2 billion.¶

The headline is a net figure, and netting hides the actual event. BEA states that the $36.3 billion increase reflected an $86.2 billion rise in spending on services partly offset by a decrease of $49.9 billion in spending on goods.

Those two numbers reconcile to the headline exactly: 86.2 minus 49.9 is 36.3. There is no residual and no rounding to argue about. In current dollars, goods spending did not merely grow slowly in July. It shrank, by an amount roughly the size of the entire net increase, and then some.

Services86.2$B

Net PCE (the headline)36.3$B

Goods-49.9$B

If your business sells a physical product, that middle bar is the one being reported to you and the bottom bar is the one you actually operate in.

The part the release does not tell you¶

Here is where an honest read has to stop short of the satisfying conclusion.

BEA gives the goods-versus-services split in current dollars only. It publishes a real PCE total and a price index, but this release does not decompose real spending into goods and services, and it does not name a single detail category — no motor vehicles line, no gasoline line, no health care line. The detail exists in the underlying tables; it is not in the release.

That matters because the obvious inference is not safe. "Goods spending fell $49.9 billion, therefore people bought a lot fewer goods" would be true only if goods prices held still, and July is precisely the month when they did not: the same fuel-price move that pulled the CPI down that month runs through the goods side of PCE as well. Gasoline is a good. If goods prices fell, part of that $49.9 billion is cheaper fuel rather than emptier baskets, and the release does not let you say how much.

Your customers got a raise in July and did not spend it¶

The behavioural finding is the one nobody led with, and it is sitting right there in the arithmetic.

Disposable income rose 0.5%. Spending rose 0.2%. The difference did not evaporate — it went into saving, which is why the saving rate rose from 2.6% in June to 3.0% in July. BEA attributes the income gain mainly to private wages and salaries, along with Medicaid and Medicare and personal dividend income.

So: households received more money in July, at a faster rate than prices rose, and responded by holding onto a larger share of it than they had the month before. That is not a story about consumers running out of money. It is a story about consumers having a bit more of it and declining to spend it.

“

Income up 0.5%. Spending up 0.2%. Volume up roughly nothing. The extra money arrived and stayed put.

”

For an operator, those two readings call for opposite responses, which is why the distinction is worth the paragraph. A tapped-out customer is a pricing and payment-terms problem. A cautious customer with money is a persuasion and priority problem — the sale is still affordable, it is just no longer urgent.

This is the third dataset saying the same thing¶

The most useful thing about the July PCE report is not that it is surprising. It is that it is the third independent measurement system to produce the same shape this summer.

Fiserv's card-transaction data showed small business sales up 1.6% year over year in July while transactions fell 1.6% and the average ticket climbed 3.2% — nine consecutive months of fewer customers spending more each. July retail sales, meanwhile, fell 0.6% on a headline dragged down by categories most small operators do not compete in.

Now the national accounts, built from an entirely different collection method than either of those, report dollars up and volume flat. Card processors see it at the terminal. The Census retail survey sees it at the store. BEA sees it in the aggregate. The instruments disagree about many things and agree about this one.

That convergence is what makes the pattern worth planning around rather than watching. A single dataset showing ticket-up-traffic-down could be a sampling artifact. Three, collected in three different ways, is a condition.

What to actually do with this¶

1. 1

   Separate your own revenue growth into price and volume before you celebrate it

   This is the entire lesson of the report applied to one business. If your revenue is up 4% and your unit count is flat, you have had the same July the country had. You cannot see that from a revenue line alone — it requires tracking units, transactions, or hours alongside dollars. The habit of recording the underlying counts, not just the totals is what makes the distinction visible at all.

2. 2

   Check whether your last price increase actually held

   Volume-flat, dollars-up is the signature of a market absorbing price increases. That is genuinely good news for margin and genuinely bad news as a growth strategy, because it only works until it doesn't. Knowing where your price sits relative to your actual cost structure and your customers' alternatives is what tells you whether you have room for another one.

3. 3

   If you sell goods, treat the services shift as a competitor

   The $86.2 billion that moved into services in July is not abstract. It is the same household budget choosing an experience, a subscription, a repair, or a premium over a physical purchase. A goods business is now competing for share of wallet against categories it does not appear to compete with.

4. 4

   Do not read the saving-rate rise as customers being broke

   A 3.0% saving rate is low by most historical standards, and it went up. Households had more disposable income and spent a smaller fraction of it. The constraint in July looks more like willingness than capacity, which is a different sales problem with different fixes.

5. 5

   Wait for the detail tables before drawing category conclusions

   The release names no categories. If your planning depends on knowing whether the goods decline was vehicles, fuel, or general merchandise, that answer lives in BEA's underlying tables and the next release on September 30, not in the August 26 summary.

The till and the stockroom¶

The gap between $36.3 billion and $1.3 billion is not a discrepancy or an error, and neither number is the "real" one. They measure two different things that happened simultaneously in July: money moved, and goods largely did not.

Every dataset available to a small operator right now is describing a version of that. More dollars per customer, not more customers. Higher tickets against flat traffic. A revenue line that grows while the stockroom empties at the same rate it did last year. It is a survivable condition and, for margin, sometimes a pleasant one.

It is also a ceiling. Price-led growth ends when the customer stops absorbing, and nothing in this release says where that point is. What it does say is that in July, households had a little more money than the month before and chose to keep more of it.

Watch the units, not the totals. It's just business.

Sources

1. U.S. Bureau of Economic Analysis — Personal Income and Outlays, July 2026
   +How this was checked

   Released August 26, 2026. Personal income +$115.1 billion (0.4%); disposable personal income +$125.9 billion (0.5%); PCE +$36.3 billion (0.2%), reflecting an $86.2 billion increase in services partly offset by a $49.9 billion decrease in goods. Real PCE +$1.3 billion (less than 0.1% at a monthly rate); real DPI +0.4%. Personal saving $712.0 billion; personal saving rate 3.0%. PCE price index +0.2% monthly and +3.7% year over year; core PCE excluding food and energy +0.2% monthly and +3.3% year over year. Income increase led by private wages and salaries, Medicaid and Medicare, and personal dividend income. April–June updated with revised employment and benefits information. All figures confirmed by direct fetch of the BEA release; the release contains no detail-category breakdown and no real goods/services split.

2. U.S. Bureau of Economic Analysis — Personal Saving Rate
   +How this was checked

   BEA's definition of the personal saving rate as personal saving expressed as a percentage of disposable personal income. Confirms the July 2026 rate of 3.0% against 2.6% in June 2026, and gives September 30, 2026 as the next scheduled release date. Confirmed by direct fetch.