I see this constantly remarked upon, so I wanted to check the data.
Figure 1: Consumption, 2017$ (bold black, left scale), Services Consumption, 2017$ (tan, left scale), Retail Sales in 1999M12$ (light blue, left scale), all in logs 2025M04=0; U.Mich. Survey of Consumers expectations (red, right scale). Source: BEA, Census, BLS, U.Michigan Survey of Consumers, and author’s calculations.
Whether the consumer is surprisingly resilient depends on what one expected consumption to do given determinants (income, expected income, costs, uncertainty, wealth…). Without a full fledged model, I can still observe that real consumption’s trend has changed to flat. Since this series includes durables consumption expenditures — which has been distorted due to front-running tariffs — one can look to services consumption. This variable should be more sensitive to permanent income (in the Hall (1979) sense). Here, the break in trend is apparent as well.
Finally, real retail sales have declined since March. With a June observation, we have a more recent reading on the consumer’s behavior — and it’s not indicating a strong resurgence.
Consumer expectations have improved, and one would expect something of a rebound in consumption in July. However, expectations have not recovered to anything like November levels (77, compared to July preliminary 59).
In the above graph, like a number of similar others I have seen in the past several weeks, the outsized gain in December and pullback in January is doing a lot of work. In other words, the flatness in 2025 may be an artifact of incompletely resolved Holiday seasonality.
The three month moving average of the several real consumption series indicates a slowdown in growth rather than a complete stall. The weekly Redbook data is similar. And as of at least last week, consumers weren’t cutting back on dining out, either.
Of course, that could change with next week’s report. But for now, consumer spending looks intact to me.