The Personal Saving Rate Is Low.
Americans Haven’t Run Out of Savings.

On May 28, Axios ran a story headlined The household crunch: Americans are spending faster than they earn it.[1] The anchor statistic came straight from that morning’s government release: the U.S. personal saving rate had fallen to 2.6% in April, down from 3.2% in March and 4.3% in January.[2] The story noted that the rate had been lower only briefly in 2022 and in the mid-2000s, before the financial crisis. Within days, versions of the same story — Americans are burning through their savings — were all over the financial news.

On September 30, the Bureau of Economic Analysis published its annual update of the national accounts, revising personal income and outlays back to January 2021.[3] In the current data, April’s saving rate is 4.4%, not 2.6%. March is 5.1%. January is 5.6%.[4] The latest month, August, comes in at 4.1%.[3]

That is not a story about dishonest reporting. The 2.6% was the official estimate on the day it was published, and it was fair to cite it. It is a story about what a single monthly number can and cannot carry. Even after the revision, households are saving a smaller share of their income than they did before the pandemic. But a smaller share of this month’s income is a very different claim from “the savings are gone.”

The saving rate measures the flow, not the water in the tub

The personal saving rate is defined as personal saving divided by disposable (after-tax) personal income. Personal saving, in turn, is simply income left over after taxes and spending.[5] Nobody counts it directly. It is an accounting residual: the government estimates what Americans earned and what they spent, and the difference is “saving.” That is why it moves so much when either side of the subtraction is revised.

The easiest way to see what it does and doesn’t tell you is to picture a bathtub. Your after-tax income is the water pouring in from the faucet. Your spending is the water going down the drain. Whatever comes in but doesn’t drain out stays in the tub — that is saving — and the saving rate is the share of the incoming water that stays. Your savings account, your 401(k) and your home equity are the water already in the tub. When the saving rate falls from 7% to 4%, more of what comes in is going down the drain, so the tub fills more slowly. It does not mean the tub is empty. The water level only drops when the drain runs faster than the faucet.

Two more features of this particular measure matter:

• It ignores investment gains. BEA’s national accounts exclude capital gains — a rising stock market or a rising home value is a change in the price of something you already own, not new income.[5] A family whose index funds rose 15% last year shows up with zero saving from that gain.
• It is income-weighted. The national rate is total saving divided by total income, so households with the most income count the most. It is not the saving rate of a typical family, and it cannot be read as one.

For context, the rate averaged a little over 7% in 2019, the last full year before COVID.[4] In April 2020 it spiked above 30% — not because Americans suddenly became thrifty, but because stimulus checks arrived while restaurants, travel and stores were closed. Neither number is a normal reading of household behavior.

Four things that all get called “savings”

Most of the confusion in savings headlines comes from treating four different measures as one. Two terms sort them out. A flow is money moving over a period of time — this month’s paycheck, or the part of it you kept. A balance is the amount sitting in an account at a single moment — what your statement shows today. In the bathtub, the flow is the water running in and out; the balance is the water level. Economists call a balance a “stock,” which has nothing to do with the stock market.

A single family can show up in all four rows at once: saving a small share of this year’s paycheck, holding more in a money-market fund than it did in 2019, carrying a healthy retirement balance — and still struggling to cover a surprise car repair.

“Excess savings are gone” — gone compared to what?

The other half of the depletion story comes from pandemic “excess savings.” Economists at the Federal Reserve Bank of San Francisco estimated that Americans built up roughly $2.1 trillion more than they would have saved without the pandemic, peaking in August 2021, and that this surplus was fully drawn down by March 2024.[6]

Read that carefully. “Zero” on that chart means households had returned to the saving path the model projected for a world without COVID. It does not mean they had zero savings. It means the extra was gone.

And the projection itself is the hard part. The San Francisco authors say plainly that their estimates are “highly sensitive to the methodology used and the assumptions made about the pre-pandemic trend.”[6] Researchers at the Boston Fed ran the same exercise with different, equally plausible baselines and found a gap of more than $1.5 trillion between estimates of how much excess saving was left in 2023 — anywhere from about $321 billion to roughly $2 trillion.[7] When the answer swings that much with the choice of baseline, the exact date the surplus “ran out” is a modeling result, not a fact about anyone’s bank account.

A broader San Francisco Fed measure that counts stocks, homes and other assets tells a parallel story: inflation-adjusted excess household wealth peaked around $13 trillion in late 2021 and fell back to its projected trend by late 2022.[8] Again, back to trend — not to nothing.

Where the money actually went

If you look only at checking and savings accounts, the picture can look worse than it really is. The JPMorganChase Institute, which tracks 4.7 million Chase households, found that by May 2025 those bank balances were about 23% above 2019 levels — well short of the roughly 40% growth past trends would predict.[9] Taken alone, that reads like depletion.

But when the Institute added the places families had moved cash — money-market funds, CDs and investment accounts — total cash reserves were growing again, up roughly 3% to 5% year over year in 2025.[9] When savings accounts paid almost nothing and money-market funds paid 4% or more, moving money out of checking was not spending it. It was a sensible decision that checking-account records show as a withdrawal.

Bank of America’s data point in the same direction. In its June 2026 Consumer Checkpoint, median checking and savings balances for a fixed group of customers were still above 2019 levels in every income group through May 2026, even after adjusting for inflation.[10]

Both datasets come with a caveat worth stating: they cover one bank’s customers, not a nationally representative sample. They are directional evidence, not a census. However, they are direct observations of real balances, which is more than the national saving rate can offer.

Every time you see the word “savings” in a headline, ask two questions: is this a flow or a balance — and whose savings are we talking about?

The strain is real. It is also concentrated.

None of this means everyone is fine. The strongest evidence in this whole debate is about who, not how much.

Economists at the Bureau of Labor Statistics and BEA built the first full distribution of personal saving by combining household surveys with administrative data. In 2022, aggregate saving was about 3% of personal income — positive — but saving was negative for the bottom half of the income distribution. The bottom 10% spent more than twice their income — more than $2 out the door for every $1 coming in — while the top 1% spent a small fraction of theirs.[11] A positive national number was made of large positive saving at the top and dissaving below the middle. The authors describe that polarization as “large and persistent,” with only a temporary interruption during COVID.

The Federal Reserve’s latest Survey of Household Economics and Decisionmaking, fielded in October 2025, shows the same split from a different angle:[12]

• Overall: 73% of adults said they were doing at least okay financially — stable from 2024, slightly below the 75% of 2018–2020.
• By income: 91% of adults with income of $100,000 or more were doing at least okay, against 45% of those under $25,000.
• Emergencies: 63% would cover a $400 emergency expense with cash or its equivalent, unchanged from the prior year.
• Rainy-day funds: 55% had three months of expenses set aside — 75% of adults earning $100,000 or more, but 21% of those under $25,000.

That is not a picture of collapse. It is a picture of a country where most households are holding steady and a large minority have very little room. A single national saving rate averages those two groups together and describes neither.

One honest wrinkle: the Boston Fed’s county-level analysis found that pandemic excess savings did not run down faster in low-income counties than in high-income ones, at least through 2022.[7]

The distribution story is strong, but it rests on different data sources that do not all line up neatly — which is one more reason to be wary of anyone who sums it up in a single sentence.

How to read the next savings headline

The saving rate will be reported again next month, and the month after that. Four questions will tell you most of what a headline is worth:

1. Is the number a flow, a balance, or a modeled gap from trend? Only a balance tells you how much money people have saved.
2. What is it divided by, and is it adjusted for inflation? A nominal balance can rise while its buying power falls.
3. Which release is it from? Monthly saving estimates are preliminary. The April figure moved 1.8 percentage points.
4. Is an aggregate being used to describe a typical household? Without data broken out by income, it can’t.

What this means for your family

The national saving rate is interesting. Your own saving rate is useful. It is the one number in this whole debate you control, and it is calculated exactly the same way: what you kept from this month’s take-home pay, divided by that take-home pay. Track it for a few months and you will know more about your family’s position than any headline can tell you.

The bigger lesson is the one the headlines blur: flows and balances are different things, and you need both. A healthy saving rate fills the tub. An emergency fund in a high-yield account is the water you can reach next week. Retirement accounts and home equity are water you cannot easily reach at all. Families get into trouble when they have plenty of the last kind and too little of the second.

And if you have kids, the bathtub is a lesson worth teaching early. A child who understands that “how much I saved this week” and “how much is in my jar” are two different numbers is already reading the news better than most of it is written.

Common Questions

Is the personal saving rate the same as how much I should be saving?

No. The national rate is an average across the whole economy, weighted toward high earners, and it excludes retirement-plan gains and home values. It is not a target for any one household. Your own saving rate — what you keep from each paycheck, including 401(k) contributions you choose to make — is the number to compare against your goals.

Why do saving-rate numbers change after they’re published?

Because saving is calculated as income minus spending, and both are estimated from early, partial data. As tax records, wage data and other sources arrive, BEA revises income and spending — and the small difference between two large numbers can swing a lot. The September 2026 annual update revised figures going back to January 2021.

The Bottom Line

The U.S. personal saving rate is below its pre-pandemic norm, even after this fall’s large upward revisions. The pandemic surplus has faded back to trend by the most widely cited estimates, though when and by how much depends on the model. Many lower-income households are under real financial pressure. All of that is true. What the evidence does not support is the universal version — that Americans as a whole have run out of savings. A tub filling more slowly is not an empty tub, and an average is not a household.