What a Soft Landing Would Actually Mean

A soft landing is an informal label, not an official status: it describes a period in which the Federal Reserve raises interest rates to slow inflation, and inflation stabilizes or falls without a recession, or with only a negligible one. No agency declares it, so it can only be judged after the fact, using recession dating, unemployment and inflation data. If it happens, it means a cooler economy and slower price growth, not lower prices, and not necessarily a return to pre-2022 interest rates.

Key takeaways

Alan Blinder's 2023 review of eleven Fed tightening cycles from 1965 to 2019 rates five as soft or "softish" and calls only 1994–95 a perfect soft landing.
The NBER, which dates US recessions, does not use the two-negative-quarters rule and has announced peaks 4 to 12 months after they occurred.
The Sahm rule read 0.00 for September 2026, against a 0.50 trigger; it crossed 0.50 in July 2024 with no NBER-dated recession following.
As of October 10, 2026, no recession has been dated since 2020, but inflation was 3.4% in August and the Fed raised rates in September, so the 2022 episode has no settled verdict.
Disinflation is not deflation: consumer prices in August 2026 were about 30% above December 2019, and a soft landing would not reverse that.

A term with no official definition

Nothing in law or Fed policy defines a soft landing. The law sets goals: Section 2A of the Federal Reserve Act directs the Fed to promote "maximum employment, stable prices, and moderate long-term interest rates." The Federal Open Market Committee (FOMC), which sets rates, defines stable prices in its Statement on Longer-Run Goals and Monetary Policy Strategy (reaffirmed effective January 27, 2026) as 2% annual inflation in the price index for personal consumption expenditures (PCE). It sets no fixed employment number, judging maximum employment not directly measurable.

A soft landing is the hoped-for path between those goals. In "Landings, Soft and Hard: The Federal Reserve, 1965–2022" (Journal of Economic Perspectives, Winter 2023), Alan Blinder, a former Fed vice chair, describes it as an outcome

"with inflation stabilized or reduced and either no recession or a negligible one."

He treats landings as a continuum, with a looser middle category: a landing is "softish" if real GDP falls by less than 1%, or if no NBER recession begins for at least a year after the tightening cycle ends.

Fed officials use the term just as loosely. In a January 2024 speech, "Heading for a Soft Landing?", Richmond Fed President Tom Barkin described it as inflation returning to normal while the economy stays healthy, and warned of the "wrong airport": inflation settling above 2%.

How Blinder scored eleven tightening cycles

Blinder's Table 1 covers eleven cycles from September 1965 to January 2019, measured with the effective federal funds rate. He grades five as soft or softish (1965–66, 1967–69, 1983–84, 1993–95 and 1999–2000), though for 1965–66 and 1983–84 he asks "but was it a landing?" Three were hard (1972–74, 1977–80, 1980–81) and two hard but not caused by the Fed (2004–06, 2015–19); 1988–89 would probably have landed softly but for the Gulf War. He also gives the opposing reading: setting aside the last two, six of the remaining nine cycles were followed by recessions.

How economists would recognize it

A soft landing is inferred from three kinds of evidence (recession dating, unemployment and inflation), and each comes with a delay or a blind spot.

NBER dating and the two-quarter shortcut

The National Bureau of Economic Research's Business Cycle Dating Committee keeps the standard chronology of US recessions. On its business cycle dating page, it defines a recession as a significant decline in activity spread across the economy and lasting more than a few months, judged by depth, diffusion and duration, with most weight on monthly indicators such as payroll employment and real personal income less transfers. Its dating FAQ explains that two consecutive negative quarters are not its test: most recessions it has dated include them, but the 2001 recession did not.

The reverse happened in 2022. The Bureau of Economic Analysis' advance estimate of July 28, 2022 showed real GDP falling at a 1.6% annual rate in the first quarter and 0.9% in the second. The NBER dated no recession, and the current real GDP series on FRED shows the second quarter revised to growth of about 0.6%.

The committee also waits for data to settle. Dates from its announcements list, with the elapsed months its FAQ gives, show the lag:

Turning point	Announced	Delay (NBER count)
Peak, July 1990	April 25, 1991	9 months (after the March 1991 trough)
Peak, March 2001	November 26, 2001	8 months
Peak, December 2007	December 1, 2008	12 months
Peak, February 2020	June 8, 2020	4 months
Trough, April 2020	July 19, 2021	15 months

The 1990 peak was announced after the recession had ended, so a "no recession" reading for any recent month stays provisional for close to a year.

The Sahm rule and how to compute it

Economist Claudia Sahm's indicator, published on FRED as the Real-time Sahm Rule Recession Indicator, signals the start of a recession when the three-month average of the unemployment rate rises 0.50 percentage point or more above its lowest three-month average of the previous 12 months.

Sahm value (month t) = A(t) − min[ A(t−1), A(t−2), … , A(t−12) ]
A(t) = average unemployment rate of months t, t−1 and t−2
Signal: Sahm value ≥ 0.50 percentage point

The table applies the formula to FRED's unemployment rate series (UNRATE) as of October 10, 2026. October 2025 has no value because household survey data were not collected during the lapse in appropriations; averages that would include it are skipped. The results match FRED's real-time series.

Month	Unemployment rate	3-month average	Prior 12-month low of averages	Sahm value
April 2026	4.3%	4.33%	4.20%	0.13
May 2026	4.3%	4.30%	4.20%	0.10
June 2026	4.2%	4.27%	4.20%	0.07
July 2026	4.1%	4.20%	4.23%	−0.03
August 2026	4.1%	4.13%	4.20%	−0.07
September 2026	4.2%	4.13%	4.13%	0.00

The rule has limits. On FRED's real-time series it reached 0.53 for July 2024 and 0.57 for August 2024, yet the NBER has dated no recession since. Sahm wrote in a July 26, 2024 post that the rule was likely overstating the labor market's weakening because immigration was raising labor supply, and called it a recession indicator, not a forecast. Unemployment can rise because more people are looking for work, not only because employers are cutting jobs.

1994–95 and 2022–23 side by side

The Fed's historical rate-change archive lists seven increases between February 4, 1994 and February 1, 1995, and its current table lists eleven between March 17, 2022 and July 27, 2023. Unemployment and CPI figures are from FRED's UNRATE and CPIAUCNS series for the months of the first and last hikes.

	1994–95	2022–23
Number of increases	7	11
Target before / after	3.00% / 6.00%	0–0.25% / 5.25–5.50%
Total increase	3.00 points (300 basis points)	5.25 points (525 basis points)
Largest single move	0.75 point, once	0.75 point, four times
First to last hike	About 12 months	About 16 months
Unemployment, first / last hike month	6.6% / 5.4%	3.7% / 3.5%
CPI, 12-month change, first / last hike month	2.5% / 2.9%	8.5% / 3.2%
NBER recession afterward	None	None dated as of October 10, 2026

Blinder puts the 1993–95 cycle at 309 basis points because he uses the monthly effective rate from December 1993. He also notes why it was easier: the 1994 tightening was preemptive and did not lower inflation, only headed off a feared rise. In 2022 the Fed started with CPI inflation above 8%.

Where the 2022 episode stands in October 2026

Recession dating: the NBER's announcement list ends with the April 2020 trough, announced July 19, 2021.
Output: real GDP grew at a 2.2% annual rate in the second quarter of 2026 (BEA third estimate, September 30, 2026), and FRED's current series shows no decline since the first quarter of 2022.
Jobs: unemployment was 4.2% in September 2026, with payrolls up 29,000, according to the Employment Situation release of October 2, 2026.
Inflation: CPI rose 3.4% in the 12 months to August 2026, or 2.4% excluding food and energy (BLS, September 11, 2026). BEA puts PCE inflation at 3.4% and core PCE at 3.0% for August 2026, against the 2% goal.
Policy: after cutting 1.75 points between September 2024 and December 2025, the FOMC raised its target range on September 16, 2026 to 3.75–4.00%, describing inflation as still elevated.

So far the recession half of a soft landing has held, but inflation has not settled at 2%: 12-month CPI inflation was 2.4% in January 2026 and 4.2% in May (CPIAUCNS).

What it would mean for households

Prices stay where they are

The St. Louis Fed defines disinflation as a decrease in the rate of inflation, and deflation as a sustained decrease in the price level. A soft landing aims at disinflation.

Illustrative example: by CPIAUCNS, a basket that cost $100 in December 2019 cost about $130.36 in August 2026, 30.4% more. At exactly 2% inflation, it would cost about $133 a year later, not $100. Since 2009, the 12-month CPI change has turned negative only in parts of 2009 and 2015.

Rates may not return to pre-2022 levels

In the FOMC's Summary of Economic Projections of September 16, 2026, the median participant's longer-run federal funds rate is 3.2%. The median in the December 2021 projections was 2.5%. These are projections of a normal rate, not commitments, and they have changed before.

A cooler job market, not mass layoffs

So far in this episode, labor demand has cooled mainly through fewer openings and fewer job switches rather than firings. The JOLTS release of September 29, 2026 put August 2026 openings at 4.3% of jobs, down from 7.5% in March 2022 on FRED's openings series. Quits were 1.9%, against 3.0% in April 2022 (quits series), and layoffs and discharges were 1.0%, against 2.0% in April 2009 (layoffs series). Finding a job gets harder while layoffs stay low.

The national number hides local ones

For August 2026, the state unemployment release lists rates from 2.0% in South Dakota to 5.1% in California, Connecticut and Oregon, and 5.7% in the District of Columbia, against 4.1% nationally. Blinder notes that higher rates hit demand for houses and motor vehicles hardest, so places that depend on those industries can feel a slowdown the national figures smooth over.

What it would mean for investors

Markets trade on expectations, so much of a landing is priced before the data confirm it. A New York Fed staff report by Kenneth Kuttner (February 2000) used fed funds futures to show that bond yields barely respond to anticipated rate changes but respond strongly to surprises. Ben Bernanke and Kenneth Kuttner (2004) found that an unexpected 0.25-point cut was associated with about a 1% rise in broad stock indexes. By the same logic, an expected soft landing need not move prices much when confirmed.

By FINRA's rule of thumb, an existing bond falls or rises by about its duration in percent for each one-point move in rates, in the opposite direction. Stock prices also depend on expected earnings, which a recession usually hits harder than a soft landing would.

The 1994 cycle shows that bonds can lose value during the tightening that precedes a soft landing. The monthly average 10-year Treasury yield rose from 5.75% in January 1994 to 7.96% in November 1994, and was back at 5.71% by December 1995. Illustrative example: at a duration of 7, a 2.21-point rise implies a price drop of about 15.5% before interest income. A hard landing does not guarantee the reverse. The Fed cut its target from 6.5% to 1.75% during 2001, yet the 10-year yield went only from 5.24% in December 2000 to 5.09% in December 2001; between September 2007 and December 2008, as the target fell from 5.25% to 0–0.25%, the yield dropped from 4.52% to 2.42%. After the 1995 landing, the Fed eased by only 0.75 point.

Frequently asked questions

How long after the last hike is a landing judged?

There is no fixed window. Blinder's "softish" test asks whether an NBER recession starts within a year of the last hike, and the NBER has taken up to 12 months to announce a peak, so the real-economy verdict can take about two years. For the cycle that ended on July 27, 2023, that window has passed with no peak dated; the inflation side remains open.

Is a "delayed landing" a soft landing?

Not yet. In his January 2024 speech, Barkin set a soft landing apart from other paths. These included a delayed landing, because the economy "continues to defy expectations"; an economy that could "run out of fuel"; and "unexpected turbulence" from shocks. Until inflation actually comes down without a recession, the soft-landing test is not met.

Why does the Fed target PCE inflation instead of the CPI?

The FOMC's Statement on Longer-Run Goals defines its 2% goal using the PCE price index, which BEA publishes. Blinder notes the Fed has had an official numerical target only since 2012 and that CPI inflation has historically averaged about 0.5 point higher than PCE inflation over long periods. The gap varies: in August 2026 both headline measures were 3.4%, while core CPI (2.4%) was below core PCE (3.0%).

Can a soft landing include a short recession?

Under Blinder's broader "softish" category, yes: a recession with a real GDP decline under 1% can still count. He classes 1967–69 that way. The recession that started in 1970 cut real GDP by only about 0.6% in his account, even though unemployment rose from 3.8% to 6.1% by August 1971.

Does the Sahm rule use revised or first-release data?

SAHMREALTIME uses the unemployment rates available in each month. BLS revises the rate each year at the start of January, so a value recomputed from today's data can differ: July 2024 read 0.53 in real time but 0.50 from the current UNRATE series. Sahm's July 2024 post notes that seasonal revisions had softened earlier readings too.

Checking the signals yourself

Monthly, after the Employment Situation release: note the unemployment rate and the SAHMREALTIME value on FRED, and whether it is near 0.50.
Monthly: compare 12-month CPI with PCE and core PCE; the 2% goal is set in PCE terms.
After each FOMC meeting: read the statement's wording on inflation and jobs; four times a year, check the projections' median longer-run rate.
Quarterly: look at real GDP, remembering that early estimates are revised and two negative quarters are not the NBER's test.
Any time: check the NBER announcements page; a new peak date is the clearest sign a landing was not soft, and it may come months later.
For a portfolio: check the duration of any bond funds and what a one-point rate move would do to them. The site's investment growth and purchasing power calculators show how inflation changes the real value of savings.