Disinflation

Disinflation is a fall in the rate of inflation while that rate remains positive. If prices across the economy rose 9 percent over one year and then 4 percent over the next, that second year was disinflationary, and prices at the end of it were higher than at any point in the first. The word describes the speed at which prices are climbing, not how high they have climbed. Disinflation is not the same thing as deflation, which is an outright fall in the price level and a far more damaging condition; the entry on deflation draws that line in full.

The term is not a commentator's coinage. It appears in the Federal Reserve's own record, used to describe what policy was trying to produce during the tightening cycle that began in 2022. At the July 2023 press conference the Federal Reserve's then-Chair, Jerome Powell, said the Committee had "always expected that the disinflationary process would stem from" both the normalization of pandemic-era supply and demand conditions and restrictive monetary policy.

The single most useful thing to understand about disinflation is arithmetic rather than economics. A falling inflation rate leaves the price level exactly where the earlier inflation put it. Inflation compounds on the level; disinflation operates on the rate. So the year that inflation drops from 9 percent to 3 percent is a year in which the cost of living rises another 3 percent on top of the 9. Nothing in the mechanism gives any of it back, and nothing is supposed to.

That gap between the statistic and the experience is the source of a durable public confusion. A household hears that inflation has come down, checks the grocery bill, finds it unchanged from last month's higher level, and concludes the number is wrong or dishonest. The number is neither. It is answering a different question from the one being asked. "How fast are prices rising?" and "how much do things cost compared with a few years ago?" have different answers, and disinflation only ever moves the first.

The second thing worth knowing is that engineering a disinflation is a policy project with a plausible cost, and the people running it have said so. In July 2023 Powell described the position as one where "we've been able to achieve disinflation so far without any meaningful negative impact on the labor market." Later in the same press conference, having again noted "the beginnings of disinflation without any real costs in the labor market," he added the qualification that matters: "the historical record suggests that there's very likely to be some softening in labor market conditions," and that such softening was "still likely as we—as we go forward with this process." A disinflation that arrives without a rise in unemployment and without a recession is the outcome people mean by a soft landing, and it is not the only way a disinflation can end.

Disinflation can also come from the supply side rather than from policy. If a bottleneck clears, a shipping route reopens or an energy shock unwinds, measured inflation slows without anyone restraining demand. Distinguishing the two matters for what happens next, because a supply-driven slowdown can reverse when the next shock arrives, while a demand-driven one reflects a change in conditions that policy created and can hold.

There is no separate disinflation statistic to look up. Disinflation is something a reader observes in the ordinary inflation data by comparing one period's rate with an earlier one. In the United States the two series that carry those rates are the Consumer Price Index, published monthly by the Bureau of Labor Statistics, and the price index for personal consumption expenditures, published monthly by the Bureau of Economic Analysis. Current readings are on BLS.gov and BEA.gov, and this entry deliberately states none, because the readings change every month while the mechanics do not.

A hypothetical example of the level-versus-rate point. Suppose a basket of goods costs $100 at the start of year one.

• Year one inflation is 9 percent, so the basket ends the year at $109.00.
• Year two inflation is 3 percent, so it ends at $112.27.
• Year three inflation is 2 percent, so it ends at $114.52.

Years two and three are a disinflation, and by any measure a successful one, since the rate fell from 9 percent to 2 percent. The basket now costs 14.5 percent more than it did at the start. Every number in that sequence is doing what it is supposed to do. A household looking at the last line and a policymaker looking at the last rate are both correct, and they are describing different facts.

The practical consequence for a plan is that disinflation changes the growth assumption, not the starting point. Fixed income that was set before the episode, a pension without a cost-of-living escalator or an annuity payment with no adjustment, does not recover its lost purchasing power because the inflation rate later fell. It stops losing ground as quickly. Those are different outcomes and only one of them is worth planning around.

What disinflation does for a household

• Slows the erosion of savings, fixed payments and cash balances, which is the thing that compounds against a long plan.
• Usually accompanies falling short-term interest rates in time, which lowers the cost of variable-rate borrowing.
• Makes planning assumptions more reliable, since a stable rate is easier to project across a thirty-year retirement than a volatile one.

What it does not do, and what it can cost

• It does not reverse the price increases that came before it. The level stays where the earlier inflation left it.
• A fixed income that fell behind during a high-inflation stretch stays behind. Nothing about a lower rate restores it.
• Engineering a disinflation generally involves restraining demand. In 2023 the Federal Reserve's then-Chair said the historical record suggests some softening in labor market conditions is likely to accompany it.
• Disinflation driven by a temporary supply improvement rather than by a change in demand can reverse when the next shock arrives.