↖︎ Vishal Singh
Federal Reserve · Survey of Household Economics & Decisionmaking

The Peak
and the Plateau

For a decade, Americans felt steadily better about their money. Then 2022 took it back — and the recovery never came.

The national mood, in one line. The share of adults who say they are at least “doing okay” financially rose for years, crested at 78% in 2021 — then fell five points in a single year and has held flat ever since.

Each year the Federal Reserve asks tens of thousands of Americans a deceptively simple question: how are you doing, financially? The answers, gathered since 2013, trace a steady climb out of the Great Recession. By 2019 the country had its best readings on record. Then the pandemic arrived — and, strangely, the numbers got even better.

Stimulus checks, paused student loans, and a year with nowhere to spend pushed savings up and worry down. In 2021, more adults reported emergency savings, more could cover a surprise bill in cash, and fewer felt they were falling behind than at any point the survey had measured. It was the high-water mark.

The tide went out in 2022.

01 — The whole decade, five ways

Everything rose. Everything reversed.

Five well-being measures for all adults. Four point up when life improves; one — feeling worse off — points up when it doesn’t. They move together.

One shared shape. Each measure climbs through the 2010s, crests in 2021, and breaks in 2022. The differences are in how far each fell — and how little has been clawed back since.


02 — The reversal

The year the floor moved

Inflation did what a pandemic couldn’t. The share of adults who felt worse off than a year ago nearly doubled between 2021 and 2022 — from one in five to better than one in three — the sharpest jump in the survey’s history.

A cliff, not a slope. “Worse off” spiked to 35% in 2022. It has eased since, but at 28% in 2025 it remains far above the calm of the late 2010s, when barely one in eight adults felt they were slipping.

By 2025, none of the five measures had returned to its 2021 peak. The plateau is the story.

This is the quiet part. A single bad year can be a shock; what the data shows is a reset. Comfort, savings, and the cushion to absorb a $400 surprise all settled onto a lower shelf and stayed there — three years and counting.


03 — The gaps that didn’t close

A national average hides two countries

Behind every “all adults” line sit groups living very different financial lives. Pick a measure and a way to slice it.

Measure
Compare by

Tip: hover any line for year-by-year values. The faint line is the all-adult average for context.


04 — The persistent floor

Almost everyone has a bank account. Almost.

Some gaps narrow with time. This one won’t budge. Banking is effectively universal — 94% of adults, year after year — yet among adults without a high-school degree, roughly one in four remains outside the banking system, exactly as in 2015. A decade of rising tides never reached the lowest shore.

Universal, except where it isn’t. The headline rate sits flat near the top of the chart. The line for adults with less than a high-school degree never climbs to meet it — the surest sign that an average can be high and a problem can still be stuck.

About the data & a few honest caveats

Source
Federal Reserve Board, Survey of Household Economics and Decisionmaking (SHED), annual releases. Figures are the published group estimates, in percent.
Coverage
“Doing okay” and the $400-expense and worse-off questions run from 2013–2014; emergency savings and bank-account ownership begin in 2015. Lines simply start where each series starts — nothing is back-filled.
Direction
Four measures are framed so that up is good. “Worse off than a year ago” is the exception: there, a rising line means more hardship, and it is shown in clay rather than teal.
Missing cells
The parents / no-children breakdown was not reported for several measures in 2013 and 2014. Those points are left blank rather than guessed, so a few group lines start a year or two later than the rest.
Reading gaps
Group gaps are differences in percentage points between published estimates and are not significance-tested here. Treat small year-to-year wiggles as noise; trust the multi-year shape.