Consumer Confidence and Sentiment Surveys

Consumer spending drives approximately 70% of US GDP. That single number explains why two monthly surveys—the Conference Board Consumer Confidence Index and the University of Michigan Consumer Sentiment Index—command outsized attention from investors, the Federal Reserve, and financial media. These aren't just opinion polls. They're early-warning systems for shifts in the spending engine that powers the American economy.

TL;DR: Consumer confidence and sentiment surveys measure how optimistic or pessimistic households feel about economic conditions. The two major US surveys (Conference Board and Michigan) use different methodologies and sometimes diverge—understanding both, especially Michigan's inflation expectations, gives you an edge in interpreting macro data before hard numbers confirm the trend.

What Consumer Confidence and Sentiment Surveys Actually Measure

Consumer confidence surveys quantify something inherently fuzzy: how people feel about the economy. That sounds soft, but the behavioral mechanism is concrete. When consumers feel pessimistic, they delay major purchases (cars, appliances, homes), reduce discretionary spending, and increase savings rates. When they feel optimistic, they do the opposite. The surveys attempt to capture these psychological states before they show up in retail sales or GDP prints.

The point is: Consumer attitudes are a leading indicator of consumer behavior. You don't need sentiment data to tell you what already happened—you need it to anticipate what's about to happen.

Two surveys dominate the landscape in the United States:

• Conference Board Consumer Confidence Index (CCI): Published monthly since 1967, surveying approximately 3,000 households. Its questions emphasize labor market conditions (whether jobs are "plentiful" or "hard to get").
• University of Michigan Consumer Sentiment Index (MCSI): Published since 1946 (the longest-running consumer survey in the US), surveying approximately 500 consumers. Its questions emphasize personal financial conditions and inflation expectations.

Both produce a headline number and two sub-indices, but they weight different aspects of consumer psychology. This is why they sometimes tell different stories—and why reading both together matters more than reading either alone.

How the Conference Board Index Works (Structure and Mechanics)

The Conference Board mails questionnaires to a rotating panel of 3,000 households each month, asking five questions:

1. Current business conditions (good, normal, or bad)
2. Current employment conditions (jobs plentiful, not so many, or hard to get)
3. Expected business conditions in six months
4. Expected employment conditions in six months
5. Expected family income in six months

Responses are compiled into three indices:

• Present Situation Index: Derived from the two "current conditions" questions
• Expectations Index: Derived from the three forward-looking questions
• Headline Consumer Confidence Index: A weighted composite of both

The base period is 1985 = 100, meaning a reading of 100 represents average confidence levels from that year. In practice, the index has ranged from a low near 25 (during the 2008-09 financial crisis) to highs above 140 (during strong labor markets).

Interpretation thresholds (these are guidelines, not hard rules):

Why this matters: The Conference Board's labor market questions make its Present Situation Index a useful cross-check against official employment data. The "Jobs Plentiful minus Jobs Hard to Get" spread (often called the labor differential) correlates closely with the unemployment rate—sometimes signaling turns before the Bureau of Labor Statistics data confirms them.

How the Michigan Index Works (and Why the Fed Watches It)

The University of Michigan conducts telephone interviews with approximately 500 consumers each month, asking about:

• Personal financial situation (current versus a year ago, and expected over the next year)
• Economy-wide business conditions (expected over the next year and five years)
• Buying conditions for large household items

This produces three indices:

• Index of Current Economic Conditions (ICC)
• Index of Consumer Expectations (ICE)
• Headline Consumer Sentiment Index

The base period is 1966 Q1 = 100. Because the base year is different from the Conference Board's, you cannot directly compare numerical levels between the two surveys. A Michigan reading of 70 and a Conference Board reading of 70 do not mean the same thing.

The critical differentiator: Michigan includes explicit questions about inflation expectations—both one-year and five-year horizons. The Federal Reserve monitors these readings closely because inflation expectations can become self-fulfilling. If consumers expect prices to rise 4%+ over the next year, they tend to demand higher wages, accelerate purchases (pulling demand forward), and accept higher prices from retailers. This behavior actually produces the inflation they expected.

The point is: When five-year inflation expectations drift above 3%, it signals that long-run inflation anchoring (something the Fed considers essential to its credibility) may be weakening. That's when you should expect more hawkish Fed rhetoric—or action.

Worked Example: Reading an Actual Release

Here's how to interpret a real Conference Board release using October 2024 data:

The numbers:

• Present Situation Index: 138.0
• Expectations Index: 89.1
• Headline Consumer Confidence: 108.7

Step 1: Headline versus consensus. The consensus forecast was 99.5. The actual reading of 108.7 represented a beat of +9.2 points—a large positive surprise. Markets reacted by pricing in stronger consumer spending ahead.

Step 2: Present versus expectations gap. The Present Situation Index at 138.0 sits well above the long-term average, reflecting a tight labor market where jobs remain relatively plentiful. But the Expectations Index at 89.1 is notably weaker—consumers feel good about today but are uncertain about six months from now.

Why this matters: A wide gap between present conditions and expectations (in this case, roughly 49 points) often signals a transition period. Consumers are spending based on current income and employment, but their forward caution suggests they may pull back if conditions deteriorate. This is exactly the kind of signal that helps you contextualize retail sales data in subsequent months.

Step 3: Check the revision. The previous month's headline was revised from 98.7 to 99.2—a revision of +0.5 points. Revisions of 2–3 points are common for Conference Board data, so always note the revised prior reading, not just the initial print. For Michigan preliminary-to-final revisions, swings of 2–5 points occur regularly (the preliminary release surveys only about 60% of the final sample).

A sample Michigan reading for context: In November 2024, Michigan's preliminary one-year inflation expectation came in at 2.6%, down from 5.4% in early 2022. That moderation helped support the Fed's decision to begin easing rates—demonstrating how a single survey component can influence monetary policy trajectory.

When the Two Surveys Diverge (and What It Tells You)

Because the Conference Board emphasizes labor market conditions and Michigan emphasizes financial conditions and inflation, the two surveys sometimes send conflicting signals. These divergences aren't noise—they're information.

The practical lesson: When the surveys diverge, look at what's driving each one. If Conference Board is strong because "jobs plentiful" responses are elevated but Michigan is weak because inflation expectations jumped, you're looking at a labor market that's tight enough to sustain spending in the near term but faces a potential squeeze from rising prices. That's a very different macro picture than "consumers are pessimistic" (which is how headlines might frame a weak Michigan print in isolation).

Release Timing and Market Impact

The point is: Markets react most to the preliminary Michigan reading because it arrives earliest in the data cycle. By the time Conference Board publishes, traders have already positioned around Michigan's signal. The final Michigan revision rarely moves markets unless it differs substantially (more than 3–4 points) from the preliminary.

Both surveys are released at 10:00 AM ET, which means any market impact shows up in the first 15–30 minutes of trading after the release (or in pre-market futures if leaked early, which occasionally happens with Michigan data).

Limitations and Pitfalls (What the Surveys Don't Tell You)

Sentiment data has real predictive value, but it comes with significant caveats that can trip you up if you're not careful.

Political polarization bias. Since approximately 2016, partisan identity has increasingly contaminated survey responses. Democrats and Republicans report dramatically different confidence levels depending on which party holds the White House—regardless of actual economic conditions. The swing can be 30–40 points on the Michigan index based purely on political affiliation. This doesn't mean the data is useless, but it means you need to look at sub-components (like inflation expectations and buying conditions) rather than treating the headline number as a pure economic signal.

The attitude-behavior gap. Consumers routinely report pessimism while continuing to spend. This gap between stated sentiment and actual behavior is well-documented. During periods of elevated inflation in 2022–2023, sentiment readings were at recessionary levels, yet real consumer spending grew. Hard data (retail sales from the Census Bureau, personal consumption expenditures from the Bureau of Economic Analysis) always trumps survey data when the two conflict.

Media amplification effects. Negative news coverage depresses sentiment readings even when underlying conditions (employment, income, wealth) are stable or improving. Survey respondents are influenced by what they've read and watched recently, which creates a negativity bias in the data during periods of intense media coverage.

Small sample size (Michigan). With only ~500 respondents versus Conference Board's ~3,000, the Michigan survey carries more sampling noise. This is partly why preliminary-to-final revisions can be significant.

Common mistakes to avoid:

• Overweighting sentiment in spending forecasts. Sentiment explains direction, not magnitude. A 10-point drop in confidence doesn't map linearly to a specific decline in spending.
• Ignoring the expectations component. The expectations sub-index is more forward-looking than the present situation component—focus your attention there.
• Comparing levels across decades. Structural shifts in how consumers respond to surveys (including the political polarization effect) make cross-decade level comparisons unreliable. Focus on changes and trends, not absolute levels.
• Treating preliminary Michigan as final. Always note that the preliminary number will be revised, and position accordingly.

Checklist for Sentiment Releases

Before the release (preparation)

• Know the consensus forecast for the headline and key sub-indices
• Check recent stock market performance and gasoline prices (the two strongest short-term sentiment drivers)
• Note any major political events that might amplify partisan response bias
• Review the prior month's revision to calibrate expectations

After the release (interpretation)

• Compare headline to consensus—was it a surprise, and by how much?
• Separate the expectations component from the present situation component
• Check Michigan's one-year and five-year inflation expectations (if Michigan release)
• Look for present-versus-expectations divergence within each survey
• Compare Conference Board and Michigan signals if both are available
• Cross-reference against hard data: retail sales (Census Bureau), personal consumption (Bureau of Economic Analysis), and employment (Bureau of Labor Statistics)

Ongoing monitoring

• Track the Conference Board's "Jobs Plentiful minus Jobs Hard to Get" differential monthly—when it narrows by 10+ points from its peak, it often precedes unemployment rate increases within 3–6 months
• Watch Michigan's five-year inflation expectations for any sustained move above 3%—this is the threshold where the Fed's inflation-fighting credibility comes into question

The core principle: Consumer sentiment surveys are thermometers, not thermostats. They measure the temperature of consumer psychology with useful (but imperfect) accuracy. Read them alongside hard spending and employment data, adjust for partisan noise, and focus on directional changes rather than absolute levels. That combination gives you a practical edge in anticipating where the consumer economy—and by extension, 70% of GDP—is heading next.