Average Inflation Targeting

The average inflation targeting framework, adopted by the Federal Reserve in 2020, permits inflation to run above its long-term target in some years provided it averages that target—typically 2%—over a multi-year horizon. Rather than enforcing a symmetric ceiling, the framework acknowledges that policymakers tolerate temporary overshoots if they are offset by undershoots, helping preserve demand during weak periods.

The shift from symmetric to flexible targets

When the Federal Reserve formally endorsed inflation targeting in the 1990s, the benchmark was a symmetric 2% target—inflation should neither exceed nor fall short significantly. This framework aligned with global central banking practice and reflected the view that any deviation from target, in either direction, imposed welfare costs. However, the 2008–2012 financial crisis and subsequent slow recovery exposed a weakness: a symmetric target meant that if inflation fell below 2% (or threatened to), the Fed would tighten policy to steer it back upward, potentially choking off fragile recovery. In practice, hitting the target was asymmetrical—central banks raised rates aggressively to suppress overshoots but hesitated to loosen enough to permit temporary undershoots.

By 2019, academic and policy discussion began questioning whether this asymmetry was intentional or accidental. The 2020 framework shift made it deliberate. The Fed now permits inflation to run “moderately above 2 percent for some time” after periods of below-target inflation, using the explicit phrase “following periods in which inflation has been running persistently below 2 percent.” This language imports a backward-looking element: past shortfalls justify present tolerance for overshoots.

How averaging works in practice

Average inflation targeting is not a hard numerical guarantee. A 2% average over what time window? The Fed’s statement does not specify, speaking instead of “over time”—typically interpreted as a business cycle (five to ten years, though sometimes longer). The practical effect is that a central bank commits to a medium-term path, not a quarter-by-quarter level. If inflation rises to 2.5% or 3%, the bank will not immediately tighten to force it back; if prior years saw 1%, the 2.5% period brings the longer average closer to target.

This reasoning shifts policy horizon. A central bank watching a single year’s inflation report faces pressure to move rates; a central bank averaging over a decade can absorb annual noise. The cost is credibility risk: if the public believes overshoots are permitted, inflation expectations may drift upward, becoming self-fulfilling. The Fed and other adopters depend on their long-earned reputation—their track record of low inflation—to maintain anchored expectations even as near-term inflation runs hot.

The actual mechanics rely on forward guidance. Rather than announcing a numerical average, the Fed signals that it will remain accommodative (keeping interest rates low or growing the money supply) for longer than traditional policy would suggest, permitting demand and inflation to rise until they have “made up” prior underperformance. This guidance is themselves a form of target; if markets believe the promise, they price in higher future inflation and lower future real returns, encouraging current demand.

Why central banks adopted it

The deepest motivation emerged from the zero lower bound crisis. When nominal interest rates fall to zero, central banks cannot cut further by conventional means, and inflation (and inflation expectations) may fall toward zero or deflation. Once trapped at the lower bound, central banks have limited levers—quantitative easing, forward guidance, or large deficits financed by the central bank. Average inflation targeting is, in part, a commitment device: by pre-announcing tolerance for overshoots, the Fed says “we will let inflation rise and won’t slam on the brakes prematurely.” This encourages firms and workers to raise prices and wages, pulling inflation up even when interest rates are already at zero.

A secondary rationale is asymmetric risk. Deflation, once started, is notoriously hard to reverse; the real burden of debts rises, unemployment tends to persist, and expectations anchor in the deflationary direction. Inflation, even at 3–4%, is unpleasant but does not trigger the same spiral of despair. By tolerating temporary overshoots, policymakers reduce the risk of slipping into a deflationary trap. For many economists, the 2010s experience (persistent below-target inflation in developed economies) vindicated this caution: a symmetric target, they argued, had been too rigid.

Trade-offs and criticisms

Average inflation targeting trades certainty for flexibility. Households and firms can no longer rely on consistent 2% annual inflation; it may be 1% one year and 3% the next. For savers and lenders, this is costly—a 3% year wipes purchasing power faster than expected. For borrowers, it is helpful. Over the cycle, it should net to 2%, but the distribution of gains and losses is uneven.

Critics also worry about anchoring expectations. If the public hears “inflation can run above target for a while,” many may interpret that as permission for higher inflation generally. The Fed’s reputation shields it from this risk—for now. But if inflation remains above target for many years, or if the Fed explicitly backtracks from the framework to allow even higher overshoots (as some feared during 2021–2023), the credibility effect could erode. At that point, inflation expectations would rise, making future disinflation costly.

A third concern is measurement. Inflation is reported with lags and is subject to revisions. If a central bank is averaging over a long horizon, it may not know whether it is on track toward the target until years later. This opacity can invite overconfidence—policymakers may believe they are on the path to average 2% while inflation is drifting higher. The Federal Reserve’s experience in 2021–2023, when inflation surged and the Fed was slower than some critics believed necessary to react, illustrated this risk.

Adoption beyond the Federal Reserve

The European Central Bank adopted a similar “flexibly symmetric” target in 2021, though less explicitly named. Canada’s central bank uses a band (1–3%) and has emphasized flexibility within it. The Bank of Japan, facing deflation for decades, had embraced a 2% target with similar tolerance for overshoots (though less formally). Adoption signals recognition that the pre-2008 inflation-targeting framework, while successful in disinflating from the 1980s, may be poorly suited to an era where the lower bound on interest rates is binding and demand growth is weak.

See also