Low Vibes Aren’t Keeping America’s Economy Down

According to one prominent news outlet, “Americans still feel worse about the economy than at almost any point in modern history.” The proof? The University of Michigan’s (U-Mich’s) widely watched consumer sentiment poll.^[i] Yet a host of results-oriented US data paints a brighter picture. Which is right? We think the latter. This is your friendly reminder sentiment surveys aren’t useful for assessing economic or market conditions, especially lately. Don’t rely on them to determine your next portfolio move.

First the survey’s results: The U-Mich’s Index of Consumer Sentiment fell to 48.1 in September from August’s 51.7, a four-month low.^[ii] Respondents’ outlooks for both their current and expected personal financial situations weakened while concerns over higher prices rose. They expect inflation to speed 4.6% over the next 12 months, faster than the 3.4% in February, before the Iran war began.^[iii]

Interestingly, while the survey’s questions regarding political affiliation and sentiment have long shown a divide based on who occupies the White House, U-Mich found moods are falling across the political spectrum: “After particularly large declines in sentiment this month, Republican sentiment is now 20% lower than January 2026; Democrats are down 13% over the same period.”^[iv] Granted, the level of Republican sentiment (77.2) is still double Democrats’ (35.6) and far exceeds Independents’ (47.7). But the reading for Republicans is the lowest since President Donald Trump won 2024’s election. 

September’s U-Mich result is also history’s second-lowest, but it, too, isn’t out of line with recent trends—the index produced its four lowest-ever readings within the past six months and has been below its long-term average for the past five years. (Exhibit 1)

Exhibit 1: University of Michigan Index of Consumer Sentiment

Source: St. Louis Federal Reserve, as of 9/25/2026. University of Michigan Index of Consumer Sentiment,  January 1978 – September 2026.

But are things really the worst they have been over the past 50 years? Other prominent datasets suggest that is far from the case. See the latest output data, which were positive for consumers and businesses alike. August retail sales jumped 1.2% m/m, beating expectations.^[v] Lest you think this was simply a function of higher gasoline bills, sales excluding auto fuel also rose 1.2% while “control-group” sales (which exclude volatile categories, including cars, gasoline, building materials and food services) climbed 1.4%.^[vi]

On the heavy industry side, industrial production was flat while the manufacturing subsector slipped -0.3% m/m.^[vii] That said, manufacturing had risen for seven straight months prior to August’s reading, which some economists attributed to businesses’ frontloading orders earlier in the year to avoid potential shortages and higher prices due to the Iran war. Moreover, in last Friday’s durable goods report, “core” capital goods orders (which exclude defense and transportation categories) rose 1.6% m/m, stronger than July’s 0.6% growth and their fourth-straight positive reading.^[viii] Today’s orders are tomorrow’s production, and this subcomponent represents equipment investment (about 39% of GDP’s business investment category)—implying companies are spending.^[ix]

September business surveys, which don’t measure actual output but sketch growth trends, also looked solid. S&P Global’s flash composite purchasing managers’ index (PMI) hit 58.4, a 62-month high, with services delivering a robust 58.7 and manufacturing registering 57.0.^[x] PMIs reveal the breadth of expansion, not magnitude, but the flash reading indicates private sector business activity broadened.

The Atlanta Fed’s “GDPNow,” a “nowcast” that combines incoming data releases to try to predict the eventual GDP release, projects 5.0% annualized Q3 growth, a swift pickup from Q2’s 1.5%.[xi] Now, there has been a sizable chasm between nominal (not adjusted for inflation) and real (i.e., inflation-adjusted) GDP lately: Q2 real GDP grew 1.5% annualized, well behind the 8.0% rate on a nominal basis.^[xii] Considering the aforementioned output data are reported in nominal terms, that would partially explain the disconnect between what inflation-scarred people are feeling and what the numbers are saying. Though these output figures may reflect elevated prices to a degree, inflation-adjusted GDP still grew last quarter and looks likely to grow this quarter, too. The US economy may not be gangbusters, but it is expanding—a stark contrast from the “lowest on record” U-Mich consumer sentiment readings over the past several months.

So what is behind this disconnect between survey findings and the economic data? A simple reality: Feelings don’t correlate with action. Numerous factors influence sentiment polls. See politics, in which partisan sentiment clearly affects how people view the same exact economy. (Exhibit 2)

Exhibit 2: U-Mich Index of Consumer Sentiment Within Political Party

Source: University of Michigan, as of 9/28/2026.

Two, there are well-known issues with surveys. In 2024, the University of Michigan began conducting its survey via the Internet rather than cellphone, due to the difficulty in contacting people by the latter.^[xiii] Some polling experts found people are more negative when surveyed online—which could be skewing the U-Mich index.^[xiv] Anyone who reads the comment section of major media can probably attest to this trollism.

None of this means the U-Mich survey or other sentiment polls are “wrong.” Rather, this is the inherent limitation in recording people’s feelings, which tend to vacillate based on what just happened (e.g., a headline story or recent market volatility). Said another way, people don’t “feel” rising orders of goods meant to last more than three years or whether a business survey finds increased activity from the previous month. But people do feel inflation taking a bite out of their purchasing power or higher prices at the gas station or grocery store. They may feel worried or even scared when headlines warn $100 oil will hurt regular households’ wallets or a certain politician will impose radical-sounding legislation should they win elected office in November.

When it comes to any economic indicator, it is critical for investors to question the headline takeaway. None is all-telling—even ones with established histories. For example, the yield curve’s inversion has a reliable record of signaling recession. But a nearly two-year-long inversion from 2022 – 2024 didn’t lead to a US economic downturn (for more on why, see our commentary from that time). Sentiment surveys can provide a broad sense of where moods are, which is worth monitoring since stocks move most on the gap between expectations and reality. But moods won’t tell you anything about future output, so don’t use surveys as a gauge for how the economy is faring.