Why Consumer Sentiment Is Crashing — And Why It Matters

Michigan Consumer Sentiment: Current Reading and Trend

The University of Michigan's Index of Consumer Sentiment (UMCSent) is released in two parts: a preliminary reading mid-month and a final reading at month-end. The March 2026 final reading came in at 77.2, down 4.8 points from the February reading of 82.0 and down substantially from the cycle high of 99.1 reached in December 2024.

The decline of 22 points over 15 months represents the sharpest sentiment deterioration since 2021-2022 (the inflation shock period) and is steeper than the 16-point decline seen during the 2019-2020 pandemic recession lead-up. The subcomponents of the sentiment index reveal concentrated weakness in consumer expectations and buying intentions, not just current conditions assessment.

The current conditions component (based on household assessment of current financial situation and economic conditions) stands at 87.3, down 6.2 points from December. The expectations component (6-month-ahead outlook) is at 69.5, down 18.1 points from December. This bifurcation — conditions falling modestly while expectations plummet — is a classic recession-leading pattern. Consumers are saying: "Things are okay now, but I believe they will be much worse in six months."

Gas Prices as the #1 Driver of Sentiment Collapse

The University of Michigan's monthly survey asks respondents to identify the most important factor affecting their personal finances and economic outlook. The March 2026 survey revealed that gasoline prices were cited by 38% of respondents as the primary concern — the highest share since 2008, when oil spiked to $147 and triggered the financial crisis.

Gasoline prices have indeed spiked. Regular unleaded gasoline is currently averaging $3.89 per gallon nationally, up from $2.61 in December 2025 — a 49% increase in just four months. Diesel is up 52%, to $4.27 per gallon. These increases flow directly from the $105 Brent crude price driven by the Israel-Iran conflict and resulting supply destruction, as detailed in our oil price history article.

The psychological impact of gas price increases is disproportionate to the actual household budget impact. At current driving patterns, a 49% gas price increase adds approximately $1,200-$1,500 annually to the typical household's transportation budget. For a household earning $75,000 per year, this represents roughly 1.8% of annual income — material but not catastrophic. Yet survey data shows that consumers perceive gasoline price increases as roughly 3-4x their actual budget impact.

This perception gap matters because consumer spending is driven by perceived wealth and expectations, not just actual cash flow. A consumer who believes gas will hit $5 per gallon and stay there is more likely to cut discretionary spending today in anticipation of future constraints, even if current finances are stable. This precautionary saving behavior is precisely what survey data shows: buying intentions for durable goods (cars, appliances, furniture) have collapsed 15-18% in the March sentiment survey, the largest drop since April 2020.

The Sentiment-to-Spending Transmission Mechanism

Consumer spending comprises roughly 70% of US GDP. Sentiment indices like the Michigan reading are leading indicators of consumer spending because they capture consumer expectations about the future. When sentiment falls, consumers cut discretionary spending in anticipation of lower future income or economic stress.

The transmission mechanism operates with a 2-3 month lag. A sentiment decline in March typically shows up as retail sales weakness in May. This lag reflects the time it takes for survey respondents' changed expectations to translate into actual spending behavior. Historically, the correlation between Michigan sentiment changes and subsequent retail sales changes is approximately 0.72 — strong but not perfect, indicating sentiment captures 50%+ of the variance in spending.

The current March 2026 sentiment decline of 4.8 points is substantial. Using historical elasticity estimates, a 4.8-point decline should translate to approximately 1.8-2.2% decline in retail sales growth over the subsequent 2-3 months, all else equal. Forward-looking retail sales forecasts for April-May 2026 (made before the March sentiment data was fully absorbed) were projecting 2.5-3.0% year-over-year sales growth. The sentiment decline suggests actual April-May sales will likely run at 0.5-1.5% year-over-year growth, a significant deceleration.

Sentiment Declines and Recession: Historical Precedent

The historical record is clear: sharp sentiment declines precede recessions. In the lead-up to the 2001 recession, Michigan sentiment fell from 107 (cycle high in 2000) to 82 by August 2001 — a 25-point decline that began 6-9 months before the NBER-dated recession. The lag reflected the tech bubble unwind and post-9/11 uncertainty.

In the lead-up to the 2008-2009 financial crisis, sentiment fell from 89.6 (October 2007) to 54.9 (February 2009) — a 34.7-point collapse that reflected both the financial shock (declining wealth, falling home prices) and deteriorating labor market (job losses accelerating). The sharpest declines occurred 6 months before the trough, not at the trough.

In the 2020 pandemic recession, sentiment fell from 101.4 (November 2019) to 71.0 (April 2020) — a 30.4-point collapse in just five months, the sharpest decline on record. Recovery was nearly as fast (sentiment back above 80 by September 2020) because stimulus and reopening confidence rebounded expectations quickly.

The current decline of 22 points over 15 months is substantial but not yet at historical recession-onset magnitudes. The 2001, 2008, and 2020 precedents all show declines exceeding 25-30 points before official recession. However, the current trajectory matters: if sentiment falls another 10-15 points over the next 2-3 months (a risk given potential further oil price increases or labor market deterioration), the current cycle would match historical precedent.

Retail Sales and Housing Sentiment Subcomponents

The overall Michigan sentiment reading of 77.2 masks sector-specific weakness. Survey respondents were asked about buying intentions for various categories: automobiles, home furnishings, major appliances, and home purchases (intent to buy in next 6 months). The automobile buying intention index fell 18% in March, the steepest monthly decline since April 2020. Home purchase intentions fell 12%, also a multi-year low.

These declines are driven by the combination of gas prices and higher interest rates. Automobile purchases are elastic to gas price expectations (when gas is perceived as expensive and rising, consumers delay car purchases in favor of more efficient vehicles or hold existing vehicles longer). Home purchase intentions are elastic to mortgage rates — at 7.2%, mortgage rates have priced out entry-level buyers and are pushing existing homeowners to delay downsizing or upgrading. A 1% rise in mortgage rates reduces home purchase demand by approximately 10-15% historically.

Retail sales have held up better than sentiment would suggest (January-February 2026 retail sales were at +3.2% year-over-year), but this strength is likely driven by earlier economic momentum and tax refunds rather than forward-looking demand. The sentiment data suggests this momentum is about to hit a wall.

Housing sentiment is particularly concerning because housing is the asset that drives consumer wealth for median households. A decline in home purchase intentions coupled with falling home price expectations (down 8 points in the March survey from February) creates a wealth destruction narrative that amplifies precautionary saving behavior.

What to Watch: April Retail Sales and Unemployment

The intersection of consumer sentiment data with the labor market deterioration tracked by the Sahm Rule (currently 0.37 and rising, as detailed in our April jobs report analysis) creates a pincer movement on consumer behavior. Falling sentiment suggests consumer are preparing for worse times ahead; rising unemployment validates those concerns and turns pessimistic expectations into actual income loss.

For business leaders and investors, the key thresholds to monitor in April-May 2026 are:

First, April retail sales (due May 15) must show deceleration to below 1.5% year-over-year growth for the sentiment transmission to be confirming recession signal.

Second, unemployment must not creep above 4.3% — the Sahm Rule threshold of 0.50 — in May. Any further Sahm acceleration would lock in the recession signal from sentiment and labor market simultaneously.

Third, Michigan sentiment in April 2026 (preliminary reading due May 15, final reading May 30) must stabilize or begin recovering. Further declines below 75 would indicate that sentiment is entering the "recession begins" territory.

GeoWire's composite recession probability model incorporates the Michigan sentiment reading as one of six inputs. The current reading contributes approximately 20% to overall recession probability. Combined with the Sahm Rule trajectory (30%), oil shock (25%), yield curve un-inversion (12%), credit spreads (8%), and geopolitical risk (5%), the consumer sentiment factor is one of the more material forward indicators in our dashboard model.