The headline numbers appear stable. Inflation, measured by the Consumer Price Index, has nearly vanished — running at just +0.44% year-over-year as of February 2026, the lowest sustained rate since before the COVID-19 pandemic. The federal funds rate has stabilized. The labor market has not collapsed. By the traditional macro scorecards, the crisis that began in 2022 appears largely resolved. And yet the American consumer disagrees, sharply and consistently, across every major sentiment measure the federal government tracks.
The University of Michigan Consumer Sentiment Index hit 56.4 in January 2026 — a three-year low and a 21.3% decline from 71.7 in January 2025. That reading sits just 6.4 points above the June 2022 all-time modern low of 50.0, when inflation was running at 9.1% annually. The paradox is stark: inflation is at near-zero, and consumers feel nearly as bad as they did at the height of the price surge.
This piece examines five federal data series — sourced directly from the Federal Reserve Economic Data (FRED) system — to document the divergence between macro stability indicators and consumer-side stress in early 2026. The data is publicly available and does not require interpretation to identify the gap.
Jan 2026 (was 71.7 in Jan 2025)
Q4 2025, Annualized
Jan 2026 (was 5.5% in Apr 2025)
What Michigan Sentiment Actually Measures
The University of Michigan Consumer Sentiment Index is not a simple "feel good" poll. It is a structured monthly survey dating to 1946, now calibrated to a baseline of 100 in Q1 1966. The survey covers five dimensions: current personal financial conditions, expected personal financial conditions, current buying conditions for large household items, short-term business conditions, and long-term business conditions.
A reading of 56.4 puts the January 2026 figure in deep pessimism territory by historical standards. In 2000, the index regularly printed above 100. In the recovery years of 2017–2019, it averaged around 97. Even during the turbulent months of 2020, it recovered to the mid-70s within a year. The current 56.4 — coming after a year of near-zero inflation and stable employment — is anomalous by any post-1980 benchmark outside of the June 2022 spike.
The trajectory is also telling. Sentiment was 71.7 in January 2025, fell to 64.7 in February, partially recovered to 61.7 in July, then deteriorated sharply in Q4 2025: 53.6 in October, 51.0 in November, 52.9 in December, and 56.4 in January 2026. The back-half of 2025 produced the worst sustained readings since the immediate post-pandemic period.
At 56.4, the January 2026 Michigan Sentiment reading is 43.6 points below the 1966 baseline of 100. It sits 6.4 points above the modern all-time low recorded at the height of 9.1% inflation in June 2022.
— FRED Series UMCSENT · Federal Reserve Bank of St. Louis
The GDP Deceleration: 4.4% to 0.7% in One Quarter
The second data series illuminating the consumer stress picture is real GDP growth (FRED series A191RL1Q225SBEA), released by the Bureau of Economic Analysis. Q4 2025 saw real GDP grow at just 0.7% annualized — down from 4.4% in Q3 2025, a deceleration of 3.7 percentage points in a single quarter.
To put the magnitude in context: Q2 2025 registered 3.8% annualized growth. Q3 2025 registered 4.4%. The Q4 2025 slowdown to 0.7% represents an 84% deceleration from the prior quarter's pace. The real GDP level reached $24,065.956 billion in chained 2017 dollars as of Q4 2025. Year-over-year real GDP growth remained positive at approximately +2.03%, but the quarterly trajectory — a contraction in Q1 2025, recovery in Q2-Q3, then sharp deceleration in Q4 — shows substantial volatility rather than the smooth growth the headline annual figure implies.
This is not a recession by technical definition (two consecutive quarters of negative growth), but it is evidence of an economy losing growth momentum precisely as consumers are losing confidence. The two data series, from different federal agencies and measuring different things, are telling the same story from different angles. You can track real GDP history at the GTP data dashboard GDP tab.
From 4.4% annualized in Q3 2025 to 0.7% in Q4 2025 · Source: FRED A191RL1Q225SBEA (BEA)
The Saving Rate as a Stress Indicator
The Personal Saving Rate (FRED series PSAVERT) provides a third lens on household financial stress. As of January 2026, the saving rate stood at 4.5% — down from 5.1% in January 2025 and from a recent peak of 5.5% in April 2025. The Q4 2025 average hovered around 4.0%, representing the lowest sustained saving period since the economy reopened post-pandemic.
The significance is structural: pre-pandemic, American households typically saved 6–7% of disposable income. The persistent sub-5% saving rate of 2025 indicates that households are drawing down financial buffers to sustain consumption at current levels — spending more than they are comfortably saving, even as they report deep pessimism about personal and national economic conditions.
This pattern — declining savings paired with declining sentiment — is the opposite of what standard economic theory would predict in a low-inflation environment. When prices stabilize, real purchasing power should improve, which typically allows households to both save more and feel better. The data shows neither outcome. You can track inflation and labor indicators at the GTP inflation tab.
The CPI Paradox: Zero Inflation, Maximum Pessimism
The Consumer Price Index for all urban consumers (FRED series CPIAUCSL) recorded a level of 327.460 as of February 2026, representing a year-over-year increase of just +0.44%. That is the lowest annual inflation rate since before the COVID-era price surge, a dramatic reversal from the 9.1% peak of June 2022. On a month-over-month basis, prices rose only 0.27% from January to February 2026.
Standard macroeconomic models predict that inflation disinflation — prices stabilizing after a surge — should improve consumer sentiment as real purchasing power recovers. That transmission mechanism has failed to materialize in the federal data. While the CPI has nearly flatlined, sentiment has continued to deteriorate. The divergence suggests that the inflation of 2021–2023, while over as a rate phenomenon, has left persistent damage to household balance sheets and economic confidence that zero-percent inflation does not automatically repair.
Put plainly: prices stopped rising, but they did not fall back to pre-surge levels. The cumulative price level remains dramatically higher than 2020 baselines even at +0.44% annual growth. Consumers appear to be reacting to price levels, not price rates of change — a distinction that headline inflation figures obscure.
The Sahm Rule: Below Recession Threshold, But Watching
The Sahm Rule Recession Indicator (FRED series SAHMREALTIME) stood at 0.27 percentage points as of February 2026 — below the 0.50 pp threshold that has historically triggered recession signals. The indicator works by comparing the three-month average unemployment rate to its 12-month low; a reading above 0.50 pp has preceded every U.S. recession in the modern era.
The current reading does not signal recession. But the trajectory matters for context. The Sahm indicator peaked at 0.57 pp in August 2024, briefly crossing the recession threshold during the period of rapid Fed rate normalization. It has since receded as the labor market stabilized — but that stabilization has occurred at a softer level than the cycle lows. Unemployment in February 2026 registered 4.4%, up from 4.3% in January and significantly above the 3.4% trough recorded in January 2023.
The labor market is not in crisis, but it has absorbed a meaningful softening. The Sahm decline from 0.57 back toward 0.27 reflects normalization, not renewed strength. Workers who lived through the 2024 scare — briefly the most elevated Sahm reading since the COVID shock — have reason to retain caution about job security even as the headline number improves. You can track labor data at the GTP labor indicators dashboard.
(Threshold: 0.50pp)
Feb 2026 (vs 3.4% in Jan 2023)
Mar 25, 2026 — Just Re-Positive
What Federal Data Can and Cannot Tell Us
These five FRED data series — Michigan Sentiment, Real GDP growth, Personal Saving Rate, CPI, and the Sahm Rule — are the official metrics that federal policymakers, the Federal Reserve, and Congress use to assess economic health. They are all publicly available, updated on documented schedules, and carry no partisan interpretation. The divergence they collectively document — between macro stability indicators and consumer-side stress indicators — is a quantifiable fact embedded in federal databases.
What the data cannot tell us is why the transmission from disinflation to confidence has broken down. That requires interpretation. Possible explanations include cumulative price-level effects (consumers reacting to total cost increases since 2020), labor market insecurity despite low unemployment, debt service burdens from the rate-hike cycle, or a longer-duration erosion of economic trust that does not respond quickly to monthly data releases. The federal data documents the gap; it does not explain it.
What the data can tell us — unambiguously — is that as of the most recent official releases, American consumers are reporting the most pessimistic outlook in three years, in an economy that grew at less than 1% in Q4 2025, with a saving rate at multi-year lows. These are not contested figures. They are the federal government's own numbers, available to anyone at fred.stlouisfed.org.
All data in this article is sourced from the Federal Reserve Bank of St. Louis FRED database, which aggregates official series from the Bureau of Economic Analysis (BEA), Bureau of Labor Statistics (BLS), and University of Michigan. GDP figures represent the BEA's third estimate for Q4 2025, published 2026-03-13. Michigan Sentiment data is current through January 2026 (FRED updated 2026-02-20); February 2026 preliminary data was not yet available in the FRED pull as of publication. CPI and Personal Saving Rate data are current through January–February 2026. Sahm Rule data is current through February 2026. Yield curve spread data is current through March 25, 2026.
All FRED series are cited inline with direct links to specific series pages. Local file paths have not been cited. All sources are approved primary government sources (.gov or Federal Reserve system).