Home / Wiki / U / University of Michigan Consumer Sentiment Index

University of Michigan Consumer Sentiment Index

Economics Markets and Financial TheoryUpdated 16 Nov 2026

The University of Michigan Consumer Sentiment Index is a monthly indicator of how American consumers feel about their own finances and about the economy, based on the Surveys of Consumers run by the university's Institute for Social Research. It summarises answers to five questions on current and expected personal finances, business conditions and buying conditions into a single index, alongside sub-indices for current conditions and expectations. The survey also asks households what inflation they expect, and these expectations are among its most closely followed results.

The survey dates from 1946, when the economist George Katona began studying how consumers' attitudes affect their spending, and is one of the oldest continuous measures of consumer sentiment. Its main US counterpart is the Conference Board Consumer Confidence Index, which uses different questions and tends to give more weight to the labor market. Expectations questions from the Michigan survey feed, together with questions from the Conference Board's own survey, into one of the components of the Conference Board Leading Economic Index.

01

How the survey works

For decades the survey was conducted by telephone with a sample of around 500 households a month, a mix of new respondents and people re-interviewed from six months earlier. During 2024 it moved to a web-based method, which changed the sample and the way questions are answered, and analysts treat comparisons across the switch with some caution. Results are released twice a month: a preliminary reading around the middle of the month, based on part of the sample, and a final reading near the end of the month.

02

Calculation

For each of the five core questions, a relative score is computed as the percentage of favourable replies minus the percentage of unfavourable replies, plus 100. The five scores are combined as:

ICS=X1+X2+X3+X4+X56.7558+2.0\text{ICS} = \frac{X_1 + X_2 + X_3 + X_4 + X_5}{6.7558} + 2.0
  • X1X_1 and X2X_2 are the relative scores for personal finances compared with a year ago and expected a year ahead.
  • X3X_3 and X4X_4 are the scores for expected business conditions over the next 12 months and the next five years.
  • X5X_5 is the score for whether it is a good time to buy major household items.

The divisor and constant scale the index to equal 100 in the base period, the first quarter of 1966, and correct for changes in sample design. As a worked example, relative scores of 90, 100, 80, 85 and 110 sum to 465, giving 465÷6.7558+2.0≈70.8465 \div 6.7558 + 2.0 \approx 70.8. The current conditions index uses X1X_1 and X5X_5, and the expectations index uses X2X_2, X3X_3 and X4X_4.

03

Inflation expectations

Each month respondents are asked how much they expect prices to change over the next year and over the next five to ten years. The median answers are widely cited as a measure of household inflation expectations. The Federal Reserve watches the longer-term measure in particular, because expectations that drift upward can feed into wage demands and pricing and make inflation harder to control. Surprises in the preliminary reading can move Treasury yields and interest rate futures.

04

How markets and economists use it

Consumer spending, measured as personal consumption expenditures, is the largest component of US GDP, and sentiment surveys are used as an early signal of whether households will spend or save. The index tends to fall ahead of and during a recession and to rise in recoveries. Analysts also use the detailed responses, such as views on buying conditions for homes and cars, to gauge demand in particular sectors.

The link between sentiment and spending is loose in the short run. Research has found that sentiment adds some information to forecasts of consumption beyond income and wealth, but that much of the time spending holds up even when sentiment is weak. Readings are strongly influenced by petrol prices, inflation, stock market moves and news coverage.

05

Criticism and limitations

Like all surveys, the index is subject to sampling error, and the monthly changes are often within the margin of noise. Since the late 2010s, answers have become strongly divided by political affiliation, with supporters of the party in the White House reporting much more favourable views than opponents, which can make shifts after elections hard to interpret. Some economists have also noted periods, such as 2022 to 2024, when sentiment was very low while spending and employment remained strong. The change in survey method in 2024 added a further source of uncertainty in comparing recent readings with history.

Last updated 16 Nov 2026