US consumer sentiment fell in September, and the headline number tells the story before any commentary does: the University of Michigan's Index of Consumer Sentiment dropped to 55.1 from 58.2 in August, a 5.1-point decline of roughly 9%. That reading was the lowest since June and sat just above the all-time low of 50.0 recorded in June 2022, the trough of the inflation spike. The decline ran against a falling unemployment rate and against the fact that gasoline prices had come down from their summer peak. Understanding why requires looking at how the index is built, what its two components measure, and which prices respondents named when they were surveyed.

What the 55.1 Reading Measures and How the Index Is Built

The Index of Consumer Sentiment is a monthly survey run by the University of Michigan's Surveys of Consumers, which has fielded it continuously since 1946. It rests on five questions, not one. Two ask about personal finances β€” current versus a year ago, and expected versus a year ahead. Two ask about business conditions over the next twelve months and over the next five years. The fifth asks whether now is a good time to buy major household goods. Each answer gets scored on a relative scale, and the five scores are averaged into an index benchmarked to 100 against the February 1966 baseline.

That construction matters because it means the index is not a spending forecast. It is a sentiment reading weighted toward expectations. Three of the five questions ask about the future, which is why the index moves sharply on news about prices, interest rates, and employment rather than on the current state of the economy alone. A person can be employed, paid on time, and still report that now is a terrible time to make a large purchase. The index captures exactly that gap.

The September drop of 5.1 points came with a component split that explains most of the headline. The Current Economic Conditions index held up better than the Expectations index. Consumers were telling interviewers that their present situation had not collapsed. What changed was their view of where things were heading. That pattern β€” present conditions steady, expectations sliding β€” is the signature of a sentiment shock tied to price and policy news rather than to layoffs.

Readings near the 55 mark are historically rare. The index has spent most of its history between 80 and 100. It first broke below 60 during the 1980 recession and again in 2008, and the 2022 slide to 50.0 was the lowest print in the survey's seventy-plus-year record. A September reading of 55.1 therefore sits in the bottom decile of the historical distribution. When the survey has printed in this range, it has usually coincided with a period when inflation was high relative to wages or when the outlook for interest rates was sharply uncertain.

Why Prices, Not Payrolls, Dominated the September Survey

The price that consumers named most consistently in the September interviews was groceries. Food-at-home inflation had cooled from its 2022 peak but had not reversed, and households compared their weekly receipts against what they had paid two and three years earlier. A shopper who sees a carton of eggs at a level still well above what it was in 2019 does not experience that as cooling inflation. They experience it as a higher baseline. The index captures that remembered baseline, not the month-over-month change.

Gasoline told a different story and is the survey's clearest inconsistency. National average gasoline prices fell through the late summer and into September, and in past cycles a sustained drop at the pump has lifted sentiment. In September it did not. That is one finding the top-line coverage tends to flatten. When a falling gas price fails to lift the index, it usually means the drag is coming from something the consumer sees more often than the price board β€” rent, insurance, or the grocery bill.

Rent and housing costs carried weight as well. The shelter component of the consumer price index was still running well above its pre-2020 pace, and for renters the monthly payment is the largest single line item in the budget. Any household renewing a lease in the third quarter of that year faced a rent that had reset upward from the prior lease term. That reset does not show up in the sentiment survey as a single question, but it shapes answers to every question about personal finances.

Interest rates completed the picture. Mortgage rates and credit card rates were at levels not seen in more than two decades, and a consumer considering a car loan or a home purchase in September was quoting a monthly payment well above what the same purchase would have cost three years earlier. The fifth survey question β€” is now a good time to buy major household goods β€” captures that directly. High rates make any financed purchase more expensive, and the survey registered it.

The Expectations Index and the Gap Between Present and Future

The Expectations Index broke down more sharply than the headline, and its level is the number to watch if you want to track whether sentiment recovers. Expectations rest on questions about business conditions one year out and five years out, plus the personal-finance outlook. A reading in the low 50s on expectations, paired with current conditions in the low 60s, describes a consumer who says "I am fine now, but I do not like where this is going." That is a fragile configuration because it can flip quickly if the present also deteriorates.

Expectations indexes lead behavior more reliably than current-conditions indexes. Consumers who expect conditions to worsen pull back on discretionary purchases before they lose income. They delay the car, the appliance, and the vacation. The sentiment index cannot tell you the size of that pullback, but a 5-point drop in expectations is the kind of move that has preceded slower discretionary spending in prior cycles. That is a correlation the survey has shown repeatedly, and it is the reason economists weight the expectations component when they read the monthly print.

The one-year inflation expectation is the other number that came with the September release. It is a separate measure from the sentiment index but published in the same report, and it asks consumers how much they expect prices to rise over the next twelve months. When sentiment falls and one-year inflation expectations rise at the same time, the interpretation is straightforward: consumers are not just unhappy about prices, they expect more of the same. That combination has historically been the harder one for policymakers to manage.

There is a limit to how much the five-year expectation tells you. Long-run inflation expectations have been remarkably stable across the past two decades even in months when the one-year measure jumped. That stability is a partial reassurance. It suggests households still believe the current inflation episode is temporary even when they are angry about it. But stability on the five-year question did not prevent the headline index from falling, which tells you the near-term experience of prices dominates how people answer.

How September Compares With the 2022 Trough and the 2008 Collapse

Period Headline reading Primary driver Unemployment rate 30-year mortgage rate (approx.)
June 2022 trough 50.0 CPI inflation at 9.1% year over year 3.6% 5.5–5.8%
September reading 55.1 Sticky grocery and shelter costs Near multi-decade low Near 7.5–8% range
Late 2008 55.3 Financial crisis, mass layoffs Rising toward 10% 5.5–6%

The comparison table is where the September number gets interesting. The all-time low of 50.0 in June 2022 came with headline inflation at 9.1% and gasoline above $5 a gallon in many metros. The September reading of 55.1 came with inflation meaningfully lower than that peak. Yet sentiment was only 5 points above the record low. The gap between the inflation rate and the sentiment reading tells you prices had not come down β€” they had only stopped rising as fast. Households do not grade on the second derivative.

The 2008 comparison cuts the other way. Sentiment fell into the mid-50s during the financial crisis when unemployment was climbing toward 10%. In September, unemployment was near a multi-decade low. Consumers reported sentiment as bad as the 2008 period while the labor market was in far better shape. That divergence is the central puzzle of the September report, and it is why the reading drew more attention than a routine monthly print would.

One detail the comparison surfaces: the mortgage rate. In June 2022 the 30-year fixed averaged in the mid-5% range. By September it was near 7.5–8%, depending on the week and lender. A consumer in 2022 was angry about prices. A consumer in September was angry about prices and faced a financing cost that roughly doubled the monthly payment on the same house compared with three years earlier. That combination is what pushed the "good time to buy" question to a near-record low.

Which Households Drove the Decline and Which Held Steady

Sentiment fell across income groups but not evenly. The steepest declines came among lower- and middle-income households, and the survey's own commentary has repeatedly flagged that split. Households with the least room in their budget feel a grocery price increase as a direct subtraction from something else. Higher-income households absorb it and report more stable sentiment. When the overall index drops 5 points, the distribution underneath is usually wider than the headline suggests.

Age mattered too. Younger consumers, who are more likely to be renting and more likely to be carrying variable-rate debt, reported weaker sentiment than older consumers, many of whom own homes outright and hold fixed-rate mortgages. A homeowner who refinanced at 3% in 2021 is insulated from the rate shock. A renter renewing a lease and a first-time buyer quoting a 7.5% mortgage are not. The same national index covers both, and the aggregate hides that insulation gap.

Partisan composition is the variable the survey handles carefully and the coverage handles badly. The University of Michigan publishes a partisan breakdown, and sentiment among respondents of the party holding the presidency has historically run higher than among the opposition. That gap widened during the 2022–2023 period. A September drop needs to be read against that baseline: if the decline was concentrated in one partisan group, it tells you more about political reaction than about economic behavior. The survey's own analysts have noted this effect explicitly.

Geography is the fourth split. Urban renters in high-cost metros reported the weakest sentiment. Rural homeowners in lower-cost areas reported considerably stronger sentiment. Since the index is a national average, a 5-point move can come from a sharp decline in a large urban sample and no change elsewhere. Anyone using the September reading to forecast national retail sales should look at the geographic composition before drawing a conclusion.

Reading the Margin of Error Before You Treat 55.1 as a Precise Number

The monthly Survey of Consumers typically interviews around 500 households, and the index carries a margin of error of roughly 3 points at the 95% confidence level. That means a 5.1-point drop is meaningful and outside the noise band, but the specific figure of 55.1 is an estimate with a range around it. A two-month comparison of 55.1 versus 55.3 is not a real difference. Anyone quoting the decimal as a precise measurement is overreading the data.

The sampling frame is a telephone sample, and response rates on telephone surveys have fallen for two decades. The Michigan survey weighted its sample to census demographics to compensate, but weighting corrects for known imbalances and cannot correct for the possibility that people who answer surveys differ systematically from people who do not. That limitation is not unique to this index. It applies to every consumer survey, and it is why the sentiment index is best read as a directional signal across months rather than an absolute measure.

Revisions are rare but not unheard of. The preliminary September reading is typically released late in the month, and the final reading follows a couple of weeks later. A preliminary-to-final revision of more than a point has happened when late interviews shifted the sample composition. If you are citing the September figure in a report, cite the month and the preliminary or final status. Quoting 55.1 without saying whether it is preliminary or final is the most common error in secondary coverage.

The survey also runs a separate index of consumer expectations and a separate current-conditions index, both published alongside the headline. The three numbers should be read together. When the headline falls and the current-conditions index holds near 60 while expectations slide toward 50, the story is about the future. When all three fall together, the story is about the present. September's split put it firmly in the first category, and that distinction should govern how the number is used.

Common Misreadings That Distort the September Headline

The first misreading is treating the index as a spending forecast. A sentiment reading does not tell you how much consumers will spend next quarter. It tells you how they feel. The relationship between sentiment and spending is loose and lagged, and it has weakened in recent cycles as households with fixed-rate mortgages insulated themselves from rate changes. Anyone building a retail forecast on 55.1 alone is using one input where three or four are needed.

The second misreading is confusing the sentiment index with the Conference Board's Consumer Confidence Index. The two surveys ask different questions and use different scales. The Conference Board index benchmarks to 1985, and it weights the labor market more heavily than Michigan does. The two indexes can move in opposite directions in the same month. When they do, the divergence itself is informative: Michigan's heavier weighting toward prices explains why it fell sharply in September while the labor-market-driven confidence index held up better.

The third is quoting the number without the reference point. A reading of 55.1 is meaningless without the historical distribution. Against the February 1966 baseline of 100, 55.1 is a poor reading. Against June 2022's 50.0, it is a slight improvement. Both comparisons are accurate. Which one you lead with determines the story you tell, and reputable coverage states both.

The fourth is assuming the index measures the whole economy. It measures sentiment among surveyed households. The stock market can rally while sentiment falls, and it often has. GDP can grow while sentiment sits in the 50s, as it did in 2022 and 2023. The index is one window into household psychology, and September's print told you that window was showing strain even as the headline macro data looked healthy.

Procurement and Policy Consequences for October and Beyond

For a business that sells discretionary goods, a 5-point drop in expectations is a prompt to look at order books rather than a reason to slash inventory. The historical pattern is that sentiment in the mid-50s precedes softer discretionary demand by one to two quarters, weighted toward big-ticket financed purchases. A retailer selling appliances, furniture, or vehicles should expect more price sensitivity and more payment-plan questions in the months that follow. A grocery or utility business is less exposed to the sentiment shift.

For a lender, the signal is about credit demand and credit quality at the same time. Consumers reporting gloomy sentiment and facing higher rates are likely to delay borrowing, which reduces origination volume. They are also more likely to draw down savings and lean on credit cards, which raises delinquency risk in the subprime segment first. September's reading does not tell you the size of either effect. It flags which direction to watch.

For a policymaker, the September print narrows the room to declare victory on inflation. If the public is still reporting sentiment at 55.1 while inflation has fallen from its peak, then the perceived inflation problem persists even as the measured one improves. That gap matters for communication strategy. A central bank that says inflation is under control while households report record-low sentiment risks losing credibility on the very measure it is trying to influence.

For an investor, the sentiment index is a weak timing signal and a useful context signal. It does not tell you when markets will turn. It does tell you that household stress is elevated, which is consistent with a consumer-led slowdown scenario and inconsistent with a boom scenario. The index is best used as one input among several, and the September print was a modest negative for the consumer discretionary picture without being a recession call.

FAQ

Did consumer sentiment fall in September even though unemployment was low?

Yes, and the divergence is the most notable feature of the report. Sentiment at 55.1 coexisted with unemployment near a multi-decade low because the survey weights prices and the future outlook more heavily than current employment status. A household can be employed and still report that now is a bad time to buy a home at a 7.5% mortgage rate. The index captures that gap between income security and purchase affordability.

How does the Michigan index differ from the Conference Board's Consumer Confidence Index?

The Michigan index benchmarks to February 1966 and weights prices heavily; the Conference Board index benchmarks to 1985 and weights the labor market. The two can move in opposite directions in the same month, and when they diverge the labor-market component usually explains why. In September the Conference Board measure held up better because employment was strong while prices remained a sore point.

What was the all-time low for US consumer sentiment?

The record low is 50.0, recorded in June 2022 when headline inflation reached 9.1% year over year and gasoline prices peaked. The September reading of 55.1 sat 5.1 points above that floor. The index has printed below 60 only in three periods: the early 1980s, 2008–2009, and 2022–2023, which tells you how rare a 55 print is against the full 1946–2023 history.

Is a 5-point drop in one month statistically meaningful?

It depends on the margin of error, which is roughly 3 points at the 95% confidence level on a sample of about 500 households. A 5.1-point drop clears that threshold, so the move is meaningful. A one- or two-point movement between adjacent months is not. Any comparison of 55.1 against 55.3 in a different month is inside the noise band and should not be treated as a real change.

What prices did consumers cite most often in the September survey?

Groceries and shelter led the complaints, according to the survey commentary accompanying the release. Gasoline was falling during the period, which is why the sentiment decline caught some analysts off guard. When a falling pump price fails to lift the index, the drag is usually coming from the grocery bill or rent, both of which consumers see more frequently than they see the price board.

Can sentiment fall while consumer spending keeps rising?

Yes, and it did during the 2022–2023 period. Households with fixed-rate mortgages at low rates were insulated from the rate shock, and excess savings built up during the pandemic supported spending even as reported sentiment sat near record lows. Sentiment and spending are correlated loosely and with a lag, so a poor reading is a caution flag, not a spending forecast.

Does the index predict recessions?

No. It has fallen before recessions and failed to fall before others. The 2022–2023 period produced sentiment in the 50s without a recession. The index is a sentiment gauge, not a forecasting model, and it has no designated recession threshold. Treating a sub-60 print as a recession signal has produced false positives in recent cycles.