Consumption Is Not Confidence: The Income Constraint
Consumer sentiment has been depressed since 2022, yet spending growth remains solid. The explanation is simple: people spend based on income, not feelings.
The Great Disconnect: Sentiment Crashes, Spending Holds
Consumer sentiment (blue) vs spending growth (green) since 2015. Indexed for comparison.
Source: FRED (UMCSENT, PCE). Monthly data.
Since 2022, financial commentary has been obsessed with "the resilient consumer." Despite record-low sentiment readings, Americans keep spending. Pundits attribute this to "strong balance sheets," "excess savings," or simply consumer irrationality.
The explanation is simpler: consumers don't spend based on how they feel. They spend based on how much money they have coming in.
The Core Finding
Income growth explains 14x more of spending variance than sentiment does. The R² for income → spending is 0.085; for sentiment → spending, it's just 0.006. Confidence surveys measure mood, not purchasing power.
I. The Regression Evidence
We tested three potential drivers of consumer spending growth using 25 years of monthly data:
| Driver | Correlation | R² (Explanatory Power) | Interpretation |
|---|---|---|---|
| Nominal Income Growth | 0.292 | 0.085 | Moderate but meaningful |
| Sentiment Level | 0.077 | 0.006 | Nearly zero |
| Sentiment Change | 0.081 | 0.007 | Nearly zero |
N = 309 months (2000-2025). All variables measured as YoY % change except sentiment level.
The numbers don't lie: income growth has an R² of 0.085 for predicting spending, while sentiment has just 0.006. That's a 14x difference in explanatory power.
Yes, 0.085 is a modest R²—plenty of other factors influence spending (interest rates, wealth effects, demographics). But the point isn't that income perfectly predicts spending. The point is that sentiment barely predicts it at all.
II. The Four Quadrants
The most striking evidence comes from examining what happens when sentiment and spending diverge. We divided the data into four categories based on whether each variable was above or below the median:
| Category | Months | Avg Income Growth | Avg Spending Growth | Avg Sentiment |
|---|---|---|---|---|
| Pessimistic but Spending | 44 | +5.33% | +7.71% | 63.0 |
| Optimistic and Spending | 45 | +5.99% | +6.24% | 99.9 |
| Optimistic but Not Spending | 32 | +4.52% | +3.83% | 96.4 |
| Pessimistic and Not Spending | 34 | +2.54% | +1.94% | 64.9 |
The "Pessimistic but Spending" quadrant tells the whole story. With sentiment at 63 (bottom quartile), you'd expect weak consumption. Instead, spending growth was 7.71%—the highest of any category.
Why? Look at income growth: 5.33%. The money was there, so people spent it—regardless of how they answered survey questions about economic conditions.
Conversely, the "Optimistic but Not Spending" group had sentiment at 96.4 (near all-time highs) but spending growth of just 3.83%. The culprit: income growth was lower at 4.52%.
III. The 2022-2025 Case Study
The recent period provides a real-time demonstration. Consumer sentiment collapsed in 2022 to its lowest readings since the 1980s—and has stayed depressed through 2025. Yet spending growth has remained solidly positive.
Annual Averages: Sentiment vs Spending
Sentiment crashed in 2022 but spending stayed strong
2022: Sentiment averaged just 59.0—the lowest annual reading since the early 1980s. Yet spending grew 9.81%. Why? Consumers were spending down pandemic-era savings (the spending-income gap was +6.56 percentage points).
2023-2025: Sentiment remained depressed (59-72 range), but spending growth normalized to 5-6%. With savings depleted, spending aligned more closely with income. But the point remains: at no point did low sentiment prevent consumers from spending.
The Savings Buffer Matters
In 2022, spending exceeded income growth by 6.56 percentage points—consumers drew down savings. By 2023-2025, that gap closed to near zero. The "resilient consumer" was partly a story of excess savings, not defiance of economic gravity. Once savings normalized, spending tracked income again.
IV. Implications
This analysis suggests several practical conclusions:
1. Don't trade sentiment surveys. Consumer confidence readings generate headlines but have minimal predictive power for actual spending. The correlation is near zero.
2. Watch income, not mood. Personal income growth (FRED: PI) is a better indicator of consumer spending trends than any sentiment survey.
3. The "resilient consumer" narrative is misleading. Consumers aren't defying economic logic—they're following it. They spend when they have income. They stop when they don't.
4. Labor market matters most. Since income depends primarily on employment and wages, the labor market is the true leading indicator for consumer spending. Jobs data beats confidence data.
Track the Indicators
V. Conclusion
Consumption is not confidence. The R² tells the story: income explains 14x more spending variance than sentiment. The quadrant analysis confirms it: pessimistic consumers with money spend; optimistic consumers without money don't.
The "resilient consumer" narrative of 2022-2025 wasn't about psychology—it was about arithmetic. Consumers had income (and for a while, excess savings). So they spent. When the savings ran out, spending aligned with income. At no point did mood override math.
For investors, the implication is clear: ignore sentiment surveys. Watch income and employment. That's where the spending signal actually lives.
Explore the Data
Related Insights
Methodology
Data from FRED: Personal Income (PI), Personal Consumption Expenditures (PCE), University of Michigan Consumer Sentiment (UMCSENT). All growth rates calculated year-over-year. Correlation and R² calculated using standard OLS regression. Sample period: January 2000 through September 2025 (309 months). Quadrant analysis divided data at the 25th and 75th percentiles for sentiment and 50th percentile for spending growth.