Literature Review: Structural Shift and Forecast Performances

We review two recent papers. The first argues that the United States has transitioned to a new regime characterized by an inflation rate persistently above the Federal Reserve’s target. The second finds that, since the onset of the pandemic, households have outperformed both professional forecasters and the FOMC in predicting inflation.

Paper #1

Is the U.S. in an Above-Target Inflation Regime?

By  Fernando M. Martin, Federal Reserve Bank of St. Louis.

What the paper does

The paper argues that the US has transitioned to a regime with an inflation rate above target. Specifically, Martin (2025) investigates whether the U.S. economy has transitioned into a persistent inflation regime where inflation remains reliably above the 2% target. He documents the evolution of inflation and explores whether the current elevated inflation is broadly based across consumption categories, whether the distribution of inflation across items remains wide, and whether longer-term inflation expectations have shifted in response.

Main results

The author finds that inflation has indeed remained above the 2% target since March 2021 and that in the post-pandemic era (2023-25) prices are rising at around 2.7% annually for the PCE excluding energy index. He identifies three distinct inflation regimes (shown in Figure 1): 2012-20 (below target, ~1.5 % inflation), 2021-22 (surge, ~5.5 %), and 2023-25 (above target but moderate, ~2.7 %). He further demonstrates that the above-target inflation regime is broad-based: in August 2025 over half of consumption expenditures in the PCE excluding energy were on items experiencing inflation of 3% or more annually, compared to only about 27% of expenditures in that category in January 2012. Additionally, the distribution of inflation across product categories remains more dispersed and shifted to the right in the post-pandemic era than in the pre-pandemic era.

Figure 1. The Price Level and Three Inflation Regimes

A critique of the paper

The paper is descriptive and not structural, so it does not identify causal mechanisms for the regime shifts. In other words, the trend-fitting method is simple and may oversimplify dynamic behaviour. In any case, the simplicity of the analysis is, to us, a strength of the paper because, precisely for this reason, it is very hard to reject the evidence in Figure 1. Indeed, Figure 2 shows the price level from PriceStat for the US (screenshot taken from Cleveland Fed event, here) and includes recent weeks/days not covered by official BLS data. The PriceStat evidence appears in line with what argued by Martin (2025).

Figure 2. PriceStat US Aggregate Price Index

Policy implications

 

In our view, the evidence remains quite clear: it is difficult to reject the hypothesis that inflation in the United States is—and will likely remain—above target going forward. The burden of proof now lies with the data in the months ahead.

Paper #2

Forecasting Inflation During the Pandemic: Who Got It Right?

By François Gourio and Kavitya Sarma (Federal Reserve Bank of Chicago, Chicago Fed Letter No. 513, 2025)

What the paper does

The paper argues that households have outperformed both professional forecasters and the FOMC since the pandemic. Specifically, Gourio et al. (2025) evaluates the accuracy of inflation forecasts made during the pandemic period (2020–24) by three groups—households, professional forecasters, and Federal Open Market Committee (FOMC) participants—and compares their performance with the pre-pandemic years from 2011 to 2019. The authors aim to understand how well different agents anticipated the unexpected surge and persistence of inflation, focusing on core PCE inflation for the FOMC and professional forecasters and on CPI inflation for households.

Main results

The analysis reveals three main findings. First, all groups experienced much larger forecast errors during the pandemic era, with the mean absolute error roughly tripling compared with the pre-pandemic period (Figure 3 reports the mean absolute errors for projections of core PCE, SEP vs professional forecasters). This deterioration in accuracy applied to short-run forecasts (within the current year) as well as medium-term projections one or two years ahead. Second, professional forecasters and FOMC participants produced remarkably similar forecasts. While both groups failed to predict the magnitude and duration of the inflation shock, their errors were virtually indistinguishable, suggesting that their models and information sets were closely aligned. Third, and most surprisingly, households provided more accurate inflation expectations than professionals or policymakers. During the pandemic years, household forecasts systematically outperformed those of experts, even after adjusting for the well-known upward bias in consumers’ inflation expectations. This represents a stark reversal of the pre-2020 pattern, when household forecasts were considerably less reliable.

Figure 3. Mean absolute errors for projections of core PCE inflation: SEP projections versus professional forecasters’ projections

Note: All the forecast errors (mean absolute error) results are in percentage points. Core PCE inflation is year-over-year inflation excluding food and energy components as reported in the initial data release of the Price Index for Personal Consumption Expenditures from the U.S. Bureau of Economic Analysis. SEP stands for the Summary of Economic Projections from the Federal Open Market Committee of the Federal Reserve System; SPD, the Survey of Primary Dealers, conducted by the New York Fed; and SPF, the Survey of Professional Forecasters, conducted by the Philadelphia Fed.

A critique of the paper

The primary limitation of the analysis is a familiar one when working with Survey of Professional Forecasters (SPF) data: the survey is published infrequently, forecasts are submitted at different times, and the panel composition changes over time. Given these factors, it is not surprising that the SPF does not outperform the FOMC.

That said, the observed comovement between SPF forecasts and those of the FOMC suggests that professional forecasters are capable of anticipating the Fed’s outlook—provided they update their forecasts more frequently.

In other words: this is exactly what we aim to do. : )

Contact

Want something more tailored?

We provide tailored consulting on ad-hoc projects.