Don't Throw out the Makeup Policy Baby with the FAIT Bathwater

Don’t FAIT-ally Overcorrect

In my last newsletter, I discussed Athanasios Orphanides' arguments in favor of monetary policy rules that are robust to uncertainty. Orphanides advocates specifically for rules targeting growth rates rather than levels. His reasoning, which is supported by empirical evidence, is that growth-rate targets are less susceptible to real-time measurement errors than level targets, which typically rely on uncertain and frequently revised estimates such as the output gap and the natural interest rate. By adopting a growth-rate approach, Orphanides contends that central banks could reduce vulnerability to data revisions and measurement inaccuracies, resulting in more stable and predictable monetary policy outcomes.

As a longtime champion of NGDP level targeting, I will confess that his arguments gave me pause. They also helped me better appreciate Bennett McCallum’s advocacy for NGDP growth rate targeting. He worried that level targeting inherently demands more discretion in practice because it requires policymakers to rapidly "catch up" whenever nominal GDP deviates from its target growth path. This excess discretion, he argued, could undermine stability and lead to more volatile macroeconomic outcomes.

Some observers, in fact, contend that McCallum’s concerns were borne out with the Fed’s Flexible Average Inflation Targeting (FAIT) framework. FAIT, a form of temporary price level targeting, was introduced in the fall of 2020 and was soon followed by an overheated economy. Some see a causal link between the two and claim it demonstrates the danger of makeup policy as outlined by McCallum.

These concerns about FAIT have not gone unnoticed by monetary policymakers themselves. The Federal Reserve's Federal Open Market Committee (FOMC) is currently conducting its quinquennial review of its monetary policy framework, in which FAIT is being evaluated. From what little we do know about the proceedings, the FOMC appears to be leaning toward abandoning FAIT and returning to the non-makeup policy of FIT (flexible inflation targeting). The first hint of this change emerged in November last year when Fed Chair Jerome Powell participated in a Q&A with Catherine Rampell. She asked him what to expect from the framework review and he said that “the base case should be more like a traditional reaction function where you don’t promise an overshoot. You just target inflation.”

More recently, the January 2025 FOMC minutes revealed that the rest of the committee appears to be leaning in that direction as well:

FOMC “participants assessed that it was important to consider potential revisions to the statement, with particular attention to some of the elements introduced in 2020. Participants highlighted as areas of consideration… the approach of aiming to achieve inflation moderately above 2 percent following periods of persistently below-target inflation.”

When both the Fed Chair and the FOMC minutes indicate a desire to return to a standard FIT approach, it does not bode well for the future of makeup policy.

So, should we give up on level targeting? I say, “Not so fast.” Yes, the Fed’s framework needs some adjustments, but I think it would be a mistake to throw the makeup policy baby out with the FAIT bathwater for three reasons. First, policymakers, like all humans, are subject to cognitive biases that may lead to macroeconomic policy errors without the discipline imposed by level targets. Second, the 2021-2022 experience shows us that makeup policy can actually work when not overdone. Finally, there are workarounds to the McCallum’s worry about level targets being destabilizing due to increased discretionary monetary policy. Rather than abandoning makeup policy altogether, we should focus on refining its implementation. Let's examine each of these points in turn.

Cognitive Biases in Macroeconomic Policy

To make this point, I want us to imagine an economy facing an unprecedented collapse. Nominal income—total current dollar spending across the economy—falls by nearly half. Factories shut down, businesses close, millions lose their jobs, and despair spreads. The general price level plummets nearly 30%, making debt burdens heavier and economic recovery elusive. After years of contraction, policymakers finally act to stimulate a recovery. But just as conditions begin improving, they grow anxious—not about depressed incomes or joblessness, but about inflation. Even with prices still far below pre-crisis levels, they fear a rapid rise and prematurely tighten policy, choking off the recovery and pushing the economy back into recession. Somehow, they miss the price level forest for the inflation trees.

If this scenario sounds absurd, you’re right—except it actually happened. It's precisely what unfolded in the United States during the Great Depression. Between 1929 and 1933, nominal GDP fell by roughly half, and prices plunged about 30%. Yet policymakers in the mid-1930s worried excessively about inflation and tightened macroeconomic policy prematurely, triggering the recession of 1937-1938 and delaying full recovery for years.

Why did policymakers make this mistake? While political pressures and flawed economic doctrines played a role, a deeper cognitive bias likely shaped their decision-making: growth rate bias. This is the tendency to focus on percentage changes rather than absolute levels. Because inflation was rising from a low base, policymakers saw it as a warning sign of overheating rather than as part of a necessary recovery. They reacted to rising inflation rates while failing to recognize that the price level remained far below pre-crisis norms. This misperception led them to tighten policy prematurely, undermining the recovery even though the economy still had a long way to go.

This historical episode underscores why makeup monetary policies—such as nominal GDP level targeting—can help discipline cognitive bias. Had policymakers instead committed clearly to fully restoring nominal income to its pre-crisis trajectory, their anxieties about inflation would have been properly contextualized and kept in check.

Now one might object that major economic crises like the Great Depressions are rare, so growth targets work most of the time. However, Japan in the late 1990s and early 2000s, the Federal Reserve after the Great Recession, and the ECB during the Eurozone Crisis show that these events do happen—and that policymakers continue to fall prey to growth rate bias. Unless we believe these types of crises—and the cognitive biases they trigger—are gone for good, a truly robust monetary policy rule that anchors policy to levels rather than just growth rates makes sense.

Makeup Policy Actually Worked

One of the most remarkable but overlooked lessons from the pandemic period is that makeup policy was tried and worked. Unlike the collapse in nominal income in 2008 that persisted for years and weighed on the economy, the collapse in nominal income in 2020 recovered quickly and returned to its trend path by 2021. It was not a foregone conclusion that the economy would recover this quickly from a $2 trillion-plus shortfall. How did it happen?

David Andolfatto and Fernando Martin show that the success of this makeup policy was contingent on fiscal policy, which acted as a form of insurance, propping up household incomes. Fiscal policy, therefore, in this experience—and likely in general—is an important part of makeup policy when there are big collapses in nominal income. Nonetheless, it is the monetary policy rule that sets public expectations for what is possible for makeup policy and, in turn, fiscal policy. Put differently, the messaging surrounding FAIT—it would tolerate inflation overshoots that corrected for past undershoots—created an environment that enabled aggressive use of fiscal policy.

While makeup policy worked, it was also overused, causing the economy to overshoot a stable growth path for nominal income. This can be seen in the figure below. Andolfatto and Martin argue that the fiscal support could have been half the size and makeup policy still would have closed the hole in the economy. Oopsies!

All of this is to say that makeup policy, or level targeting, has proven effective, but it needs better calibration next time. A clearer level target might have helped manage expectations and fiscal policy more effectively.

How to Address McCallum’s Discretion Worries

Finally, let’s return to McCallum’s concern about level targeting. His worries about the discretionary nature of makeup policy—what he referred to as the “splurge” to quickly close the gap—does not have to be the end of the story. First, If an NGDP level target is highly credible, private-sector expectations would do much of the adjustment themselves, making large discretionary interventions by the central bank less necessary. Under a credible NGDP level target, forward-looking markets and consumers anticipate policy moves, adjusting their spending, pricing, and investment decisions proactively. This, in theory, stabilizes nominal GDP without requiring aggressive policy "catch-ups."

In other words, the central bank's clear commitment to restore the targeted NGDP path acts as powerful forward guidance. It sets market expectations effectively, reducing the magnitude of discretionary policy interventions needed. Thus, a credible level-targeting regime can, in principle, achieve stable outcomes without succumbing to the discretionary pitfalls McCallum highlights.

Second, one could also try a hybrid approach where Athanasios Orphanides’ growth rate target is used in most situations, but still allow for a threshold condition where a level targeting kicks in once nominal income deviates beyond some percent from some trend. For example, if we take τ to be our threshold and D(t) to be an indicator variable, then one formulation would be as follows:

\(\begin{align} \Delta i_t &= \alpha_n E_t \left({ \Delta n_{t{\scriptscriptstyle +3}}} - {\Delta n^*_{t{\scriptscriptstyle +3}}} \right) + \beta_n D_tE_t \left( { n_{t{\scriptscriptstyle +3}}} - {n^*_{t{\scriptscriptstyle +3}}} \right) \\ D_t &= \begin{cases} 1, & \text{if } |E_t \left( { n_{t{\scriptscriptstyle +3}}} - {n^*_{t{\scriptscriptstyle +3}}} \right)| \geq \tau \\[6pt] 0, & \text{otherwise} \end{cases} \end{align} \)

What size should the nominal income gap threshold, τ, be is a tough question. Here is what ChatGPT recommends based on one’s preferences for level interventions.

Conclusion

The likely demise of FAIT should not be mistaken for the failure of makeup policy altogether. Rather, it should be seen as an opportunity to refine and improve its application. The concerns raised by Orphanides and McCallum about discretion and measurement errors are valid, but they do not justify abandoning level targeting outright. Rather, they highlight the need for a more structured, rule-based approach—one that limits discretion while preserving the long-term benefits of a stable nominal anchor.

History has shown us that cognitive biases, particularly growth rate bias, can lead policymakers to premature tightening, exacerbating economic downturns. A credible level target disciplines these biases by anchoring expectations to a well-defined nominal path. Additionally, the pandemic experience demonstrated that makeup policy, when properly implemented, can quickly restore economic stability, though better calibration is needed to avoid excessive overshooting.

One promising path forward is a hybrid framework, where a growth-rate rule operates in normal times, but level targeting is activated when nominal income deviates beyond a reasonable threshold. This approach would balance robustness against uncertainty with the long-term benefits of a level anchor, making monetary policy more resilient to both measurement errors and cognitive pitfalls.

Rather than discarding makeup policy due to its recent missteps, we should refine its use to ensure it remains a valuable tool for stabilizing the economy. The challenge ahead is not whether to use level targeting, but how to design it in a way that enhances credibility, minimizes discretion, and fosters economic stability.

---

Upcoming and Further Reading

• I will be interviewing Paul Blustein and Kate Judge separately next week for the podcast. Please drop question suggestions in the comments. I plan to ask for question suggestions for the following week’s recording(s) on each of my Thursday posts!

• Next Thursday, I will continue our conversation on the Fed’s framework review. I will move beyond the makeup policy discussion and discuss in what ways the FOMC can best deal with negative supply shocks like tariffs. No surprise where I will land on it! Also, at some point I hope to show how I would reword the consensus statement.

• My colleague, Scott Sumner, has written a wonderful paper on the “Princeton School of Macroeconomics” that worked tirelessly on this subject for years. This group included Ben Bernanke, Gauti Eggertson, Paul Krugman, Lars Svensson, and Michael Woodford.

• If you want more discussion on the Fed’s framework review and cannot wait until next week, check out the Mercatus Center’s policy brief series on it.

  Photo Credit: Jose Luis Magana (The AP)

  ---