Integrated Average Inflation Targeting

An integrated average inflation targeting framework allows a central bank to undershoot its inflation target temporarily during downturns and then aim for higher inflation later to bring the multi-year average back to the 2% goal—all while responding to persistent employment shortfalls, not just price misses. It is more flexible and employment-conscious than traditional point-in-time targeting.

Standard average inflation targeting vs. integrated approach

Standard average inflation targeting (AIT) emerged in the 2010s as a response to the zero-lower-bound trap. The Federal Reserve consistently undershot its 2% target during the long post-2008 recovery. A simple AIT rule says: “We’ll target 2% average inflation over, say, the last five years.” If you’ve averaged 1.5% for four years, you permit 2.5% in year five to bring the backward-looking average back to 2%. This acknowledges that a central bank recovering from a deflationary shock may need to run hot for a spell.

Integrated average inflation targeting goes further. It embeds the employment dual mandate—not as an afterthought, but as an integral part of the averaging rule itself. The central bank announces it will tolerate inflation underruns when unemployment is materially above the natural rate, and will make those underruns up later. This is not “ignore inflation when unemployment is high”—it is “temporarily undershoot the 2% target if doing so brings unemployment closer to its natural level.” The integration comes from treating the unemployment gap and the inflation gap as joint arguments in a single policy reaction function.

In practice, the Federal Reserve adopted a version of this in August 2020, labeling it “flexible average inflation targeting.” The integration component meant the Fed would not automatically hike rates merely because inflation had come in above 2% if unemployment remained elevated. The averaging window and the employment clause became inseparable.

Why not just point-in-time inflation targeting?

Before average targeting, central banks aimed to keep inflation at 2% at every moment. A spike in crude oil that pushes inflation to 2.5% triggers an immediate rate hike. This approach has a flaw: in a severe downturn—a financial crisis, a pandemic—inflation often falls sharply anyway because demand collapses. Hiking rates aggressively into a recession to bring inflation back to 2% right now can deepen unemployment and create a deflationary spiral.

The Federal Reserve learned this the hard way after 2008. Despite near-zero rates and massive quantitative easing, inflation stayed below 2% for nearly a decade. Raising rates in 2015 and 2018 only tightened the screws further. By averaging inflation over a longer horizon, a central bank can acknowledge: “Yes, we are temporarily undershooting 2%, but we expect to hit it on a five-year view, so we will tolerate the undershoot and focus on unemployment.”

Integrated average targeting builds on this insight by making employment shortfalls explicit in the policy rule itself.

The employment shortfall clause

Standard AIT focuses on the inflation path. Integrated AIT adds: the central bank will explicitly not raise rates until unemployment is close to its natural rate, even if the averaging path suggests inflation should be higher. This sounds subtle, but it is profound.

Imagine unemployment is 5% and the natural rate is 4%. The Fed could achieve its 2% average inflation target by raising rates now and tolerating higher unemployment going forward. But an integrated framework says: “No—we will run inflation a bit above 2% until unemployment falls to 4%, then allow it to pull back.” The unemployment gap actively shapes the inflation path.

This is sometimes called “price stability in the context of maximum employment,” the Fed’s formal dual mandate. Before 2020, the Fed often treated these as separate objectives, with an informal hierarchy (inflation first, employment second). Integrated average targeting makes them genuinely co-equal.

How make-up inflation works in practice

The logic of integrated AIT is arithmetic: if inflation averages 1.5% for three years while the Fed is fixing unemployment, the central bank must later run inflation above 2% to bring the five-year average to 2%. The “make-up” isn’t a penalty; it’s a forced rebalancing.

Suppose:

• Year 1: inflation 1.0%, unemployment 5.5% (above natural rate of 4%).
• Year 2: inflation 1.2%, unemployment 5.0%.
• Year 3: inflation 1.5%, unemployment 4.2%.
• Three-year average so far: 1.23%.

To hit 2% over five years, the Fed must now run inflation at roughly 3.0% in years 4 and 5. This forward guidance matters: if financial markets and workers believe the Fed will eventually hit 2% over a five-year horizon, inflation expectations stay anchored at 2% even during the undershoot period. Nobody panic-sells long-term bonds; nobody demands wage escalators based on a temporary 1.5% inflation. Credibility of the long-term target preserves stability.

Without the make-up clause, an inflation target is a ceiling, not an average. The Fed could perpetually run low and call it “flexibility.” Credibly committing to make-up inflation—i.e., to higher inflation later—anchors expectations and makes the whole framework self-enforcing.

Criticism and practical constraints

Integrated AIT has critics. One argument: “Forward-looking inflation expectations are already anchored; just keep them anchored and focus on employment.” Why invite higher inflation into the arithmetic? Another: “Real-world unemployment gaps and natural rates are uncertain. You can’t engineer a precise employment-inflation tradeoff.” A third: “Political pressure will erode the make-up commitment. The Fed will undershoot and then forget to raise inflation later.”

All three critiques have merit. Central banks do not have a thermometer for the “natural rate of unemployment” (it shifts with demographics, technology, and policy). Integrating it into the policy rule introduces model risk. And central-bank independence, while strong, is not absolute; a Fed that runs inflation at 3.5% for an extended period faces calls to tighten, regardless of unemployment.

In practice, the Federal Reserve has adopted integrated average inflation targeting rhetorically but has not rigorously committed to an arithmetic make-up. The 2020 framework promised flexibility; when inflation finally overshot in 2021–2022, the Fed raised rates sharply, essentially abandoning the make-up horizon and pivoting back to near-term inflation control. This suggests the integration is more aspirational than mechanical.

Comparison to other monetary frameworks

A gold standard removes discretion entirely; inflation targets the long-run average price level set by gold. A Taylor rule uses a mechanical formula (a + current inflation + 0.5 × inflation gap + 0.5 × output gap). Integrated average inflation targeting is more flexible than both—it allows the central bank to smooth shocks and prioritize employment—but it requires more judgment and is more vulnerable to political pressure.

Some economists prefer a price-level target (aiming for a specific nominal GDP or price level, not a growth rate). This automatically triggers make-up inflation because the target is an end state, not a growth rate. Integrated average targeting tries to approximate this logic without the rigidity.

See also