The Soft Landing Nobody Believed In

The Bet Everyone Lost

Fast forward to the end of 2022. Inflation had peaked above 9 percent the Federal Reserve had just raised interest rates at the fastest pace in four decades and the consensus among economists banking strategists and most of financial Twitter was that a recession in 2023 was all but inevitable. The logic was simple and historically grounded. Every time the Fed had fought inflation so aggressively unemployment soared and the economy contracted. The phraseTo avoid that result a soft landing he used to say to himself with a satisfied smile

2024 was the year the smile disappeared. Inflation fell toward the Fed's 2 percent target the labor market cooled without collapsing growth approached 3 percent and in September the Fed began cutting rates. It was the most dubious economic outcome of the decade and it happened anyway

How Inflation Fell Without a Crash

The standard model says that inflation falls when demand weakens which usually means job losses. That's not primarily what happened. Inflation fell because the supply side of the economy recovered. Pandemic-era shipping problems cleared up factories caught up with delays energy prices stabilized after the 2022 peak and crucially the American workforce grew partly thanks to a rebound in participation and partly toimmigration.More workers and more goods meant the economy could grow with its demand instead of choking on it

By September 2024 the headline CPI had fallen to 2.4 percent down from 9.1 percent at the June 2022 peak. The Fed never got a clear 2.0 percent figure and the last stretch proved difficult but the direction was unmistakable and was achieved with unemployment close to 4 percent instead of the 6 or 7 percent that many models called for

A Worked Example: What This Disinflation Actually Cost

Economists have a figure to pay the price of killing inflation. It's called the sacrifice ratio and measures how many percentage points of years of excess unemployment were necessary for each percentage point of inflation removed. Calculating it for 2022 to 2024 is the clearest way to see how unusual this episode was

Let's start with the denominator: the disinflation achieved. The headline CPI fell from 9.1 percent in June 2022 to 2.4 percent in September 2024. That's a removal of 6.7 percentage points

Now the numerator excess unemployment. Take full employment at around 4.4 percent which is a common estimate of the rate below which the labor market is heating up. Compare actual unemployment to it each year and add up the gaps

The gaps are negative. Unemployment ran down the benchmark for all disinflation. Divide minus 2.0 by 6.7 and the sacrifice rate will be around minus 0.3

There is assumed to be no negative sacrifice ratio. It says the economy eliminated 6.7 points of inflation while the labor market was warmer than full employment at all times

Now do the same calculation on the episode everyone was comparing it to. Volcker's disinflation took the CPI from about 14.8 percent in the early 1980s to about 3.2 percent in 1983 or 11.6 points lower. Against a full employment benchmark of close to 6 percent for that time unemployment was about 1.6 points higher in 1981 3.7 in 1982 3.6 in 1983.and 1.5 in 1984 for a cumulative excess of about 10.4 points per year. Divide it and the sacrifice ratio is about 0.9

Place the two side by side. Volcker: about 0.9 points of excess unemployment for every point of inflation removed. The 2022 to 2024 episode: about negative 0.3. These are rounded illustrative figures that use annual averages and a questionable benchmark and moving the benchmark by a few tenths changes the figures without changing the sign

That signal is the whole story. That's why the forecasting professionals were wrong and why the models were wrong rather than simply miscalibrated. Every demand-side framework assumes that inflation is reduced by creating slack. This disinflation created nothing. Whatever eliminated inflation was not operating through the channel that the models were observing

The Scare That Tested It

The landing was not without turbulence. In the summer of 2024 unemployment rose to around 4.3 percent and that increase triggered the sahm rule a historically reliable recession indicator that triggers when the unemployment rate rises half a point from its recent low. Markets wobbled sharply in early August. For a few weeks the recession side seemed vindicated

It was a false alarm. The rise in unemployment was mainly due to new workers entering the workforce and taking time to find jobs not to layoffs which remained low throughout the year. The episode became a lesson in how to read indicators mechanically instead of reading the data beneath them. Even the best recession signal can fail when the workforce itself is expanding

The Cuts Begin

On September 18 2024 the Federal Reserve cut rates for the first time since the pandemic emergency and opened with an unusually large half-point move lowering the target range from its high of 5.25 to 5.50 percent. Two-quarter-point cuts followed in November and December ending the year between 4.25 and 4.50 percent. The message was that the fight against inflation had been won convincingly enough tostart releasing the brake

The markets loved it. The S&P 500 gained about 23 percent in 2024 on top of a similar gain in 2023 driven by AI trading covered in our 2023 entry and simple relief from falling rates. Corporate borrowers refinanced business activity thawed and the recession which had been twelve months away for three straight years quietly missed forecasts

Case Study: 1994 and the Only Other Soft Landing

The reason 2024 was so doubted is that there was essentially a precedent and even that is disputed

Between February 1994 and February 1995 Greenspan's Fed doubled the federal funds rate from 3 percent to 6 percent across seven increases including a surprise half-point move and a full point. It was a preemptive adjustment against inflation that had not yet appeared which is the hardest kind to justify politically because what you're preventing never happens and therefore never proves right to you

No recession followed. Growth continued unemployment continued to fall and inflation remained contained. It became the go-to case for the phrase soft landing and much of the reputation Greenspan maintained over the next decade

The part left out of the story is what it cost elsewhere. 1994 was one of the worst years for bonds in modern history severe enough to be remembered as the great bond massacre. Orange County California which had leveraged its mutual fund by betting that rates would stay low lost approximately $1.7 billion and declared bankruptcy in December 1994 the largest municipal bankruptcy in the country at the time. A soft landing forthe economy was a hard landing for anyone in the opposite position

The aftermath matters more. The subsequent easing of the mid-1990s combined with genuine productivity growth fueled the most extreme episode of stock valuations in American history which ended in 2000. It is at least arguable that the soft landing of 1994 did not prevent the reckoning so much as change its direction and 2024 deserves the same question rather than the same applause

Why Almost Everyone Got It Wrong

Three reasons worth internalizing because they will apply to the next cycle as well. First forecasters relied on demand-side models and underestimated the supply recovery which is understandable because supply shocks of this magnitude had not occurred since the 1970s. Second the starting point was unusual: households and businesses had set low rates and maintained unusually strong balance sheets so rate increases were a little slower than expected.history suggested. Third incentives matter predicting recessions is professionally safer than predicting soft landings because being wrong pessimistically is forgiven and being wrong optimistically is not

The honest asterisk is that the landing was softer for asset owners than for everyone else. Cumulative prices were still more than 20 percent above 2019 housing affordability was the worst in four decades and the pain of that price level is a big part of the politics that shaped 2025 and 2026. A soft landing for the economy is not the same as feeling whole

Where Calling It a Soft Landing Is Too Generous

Writing this from 2026 gives me an advantage that the 2024 commentators didn't have and makes the triumphant version harder to defend

The plane took off again. In June 2026 inflation was back above 4 percent. If you draw the line in September 2024 it's a landing. If you draw it two years later it looks more like a pause in a longer inflationary episode and where you draw the line is a choice rather than a finding. The Federal Reserve made three cuts in an economy that would need to reverse those cuts which is obviously not the behavior of an institution that had won

Credit for luck. Disinflation was largely due to supply and the Fed doesn't control supply. Shipping lanes were cleared energy was established and the workforce expanded and none of those were political achievements. A negative sacrifice ratio can mean that the central bank was clever or it can mean that inflation was going to fade anyway and that the increases were largely ceremonial. The data can't tell them apart and the profession has generally chosen the flattering reading

The supply of labor was unique and was reversed. A large proportion of the labor force growth that made demand noninflationary came from immigration and immigration policy changed dramatically afterwards. If the mechanism that produced the soft landing has since been deactivated the episode is less a repeatable pattern than a fortunate conjunction

The failure of the Sahm rule is not reassuring. The standard reading is that the indicator failed and the economy was doing well. The uncomfortable alternative is that our best recession signal stopped working leaving us with one less instrument for next time. Being wrong in a useful direction is still being wrong

My own view is that 2024 was a genuine achievement and less than celebrated and that the 2026 inflation readings should make everyone who declared victory check the tape

How I Actually Judge a Landing

After seeing how they called this one and then partially canceled it this is what I'd watch instead of the headline

First I calculate the sacrifice rate from the example above rather than reading the narrative because it turns a mood into a number. A disinflation achieved without costs to the labor market is a supply story or a measurement problem and in any case it is not the demand story that the models describe

Second I distinguish where disinflation came from before giving anyone credit. Supply-driven disinflation is reversible unlike demand-driven disinflation because ships can get stuck again and oil can skyrocket again. That distinction alone would have flagged the 2026 relapse as a live risk in 2024

Third I wait longer than seems reasonable before calling. Two years after the fact 2024 reads differently than it did at the time and there was no information available in 2024 that would have resolved it

Fourth I try to note when a forecast is professionally safe rather than analytically supported. The consensus on the recession was safe. The universal consensus on the soft landing that formed in 2025 was by then also safe. They both erred in the same direction as the crowd and neither cost anyone anything

That's how I try to think of it. It's a description of the method not a forecast or advice

The Bottom Line

2024 produced the outcome that almost no serious forecaster believed in: inflation near target unemployment near 4 percent growth near 3 percent and a central bank cutting rates to strengthen rather than crisis

The arithmetic shows how unusual it was. Volcker eliminated 11.6 points of inflation at a cost of about 10.4 percentage points years of excess unemployment a sacrifice rate close to 0.9. Disinflation from 2022 to 2024 eliminated 6.7 points while unemployment was below full employment the entire time a ratio close to negative 0.3. That number is not supposed to exist and its existence is the reason whythe models failed rather than simply being wrong

It stands as the strongest counterexample in modern history to the claim that disinflation requires a recession. Remember it in three ways: as a genuine policy and supply success as a warning about consensus and as a reminder that 1994 was also called a soft landing and that the reckoning simply came elsewhere six years later. When all the forecasts point in the same direction the interesting question is what would have to be true for them all to be wrong