After five years of high inflation, U.S. inflation expectations remain only moderately elevated, with no immediate signs of becoming unanchored

Overview

The report examines whether more than five consecutive years of above-target U.S. inflation have significantly raised public expectations or made them more prone to becoming unanchored. Drawing on nearly a decade of academic research, household and business surveys, and a model based on individuals' historical inflation experiences, Goldman Sachs concludes that short-term expectations are somewhat elevated, but long-term expectations and their sensitivity to actual inflation have not reached levels warranting immediate concern.

Core views

The report first explains why inflation expectations matter: they pass through to wages, firms' pricing, and consumption and investment decisions. Average estimates from academic research show that for every 1 percentage point change in short-term inflation expectations, consumers' wage expectations and firms' pricing behavior change by roughly 21 basis points and 29 basis points, respectively—meaningful pass-through, but far from one-for-one. Rising inflation expectations are also generally interpreted by households and firms as a deterioration in the economic outlook, rather than simply a signal to bring spending forward, and may therefore reduce consumption, investment, and employment. The second core conclusion is that the public forms expectations primarily based on inflation it has actually experienced, with recent experience, experience over the entire course of adulthood, and formative-years experience all playing important roles. The research review shows that for every 1 percentage point increase in actual inflation, long-term inflation expectations among households and professional forecasters rise by an average of 13 basis points and 9 basis points, respectively. People who have experienced higher lifetime inflation generally believe that new inflationary episodes will last longer, and their long-term expectations are also more responsive to current inflation. For studies that report only changes in short-term expectations, the report assumes a 25% pass-through from short-term to long-term expectations, a ratio estimated using University of Michigan data and consistent with existing research. Third, monetary policy communication has limited direct power to anchor household and business expectations. During periods of low inflation, the public pays little attention to the Federal Reserve: in a 2018 survey, respondents on average believed that the Fed's inflation target was above 8%. High inflation increased attention, and in the same survey conducted in 2025, the estimate fell to slightly above 3%, but the effect of actual high inflation on expectations outweighed this improvement in awareness. The report therefore argues that relying solely on Fed communication and credibility would be insufficient to stabilize household and business expectations if inflation became uncontrolled; the truly critical factor is a sustained decline in actual inflation. Regarding the current situation, market-based expectations and professional forecasters' expectations remain well anchored; the report focuses more on households and firms, whose expectations directly affect economic decisions. More than five years of high inflation were preceded by over a decade of low inflation, and the net effect of the two cycles has brought long-term household expectations broadly back to pre-global-financial-crisis or mid-2000s levels. Short-term household and business expectations are somewhat elevated, particularly after rebounding in response to this year's oil price shock. Different surveys send different signals about long-term expectations. The New York Fed survey indicates that the recent period of high inflation has mainly caused younger people, who had previously known only the low-inflation environment following the global financial crisis, to move their expectations closer to those of older groups that experienced alternating periods of high and low inflation from the 1960s through the 1980s. The University of Michigan's 5-to-10-year inflation expectation measure has reached 3.3%, only 0.1 percentage point below its peak over the 25 years before the pandemic. Goldman Sachs believes this elevated reading partly reflects survey responses becoming more politicized after the 2024 election and an increase in extreme responses following the shift to online surveys. Differences among surveys therefore cannot all be interpreted as genuine deterioration in expectations. To reduce distortions caused by changes in survey methodology, the report uses individual-level University of Michigan microdata from 1978—2026 to adapt an academic model based on inflation memory. The model examines how an individual's average lifetime inflation experience, recent inflation experience, and the similarity between the current environment and past experience affect both the level of 5-to-10-year expectations and their sensitivity to actual inflation, with lifetime experience calculated from age 16. For a 1 percentage point increase in the relevant variables, the coefficients of average lifetime experience, recent experience, and similarity between current and historical experience on the level of long-term expectations are 0.58, 0.34, and 0.25, respectively; their coefficients on the sensitivity of expectations to changes in actual inflation are 0.04, 0.04, and -0.02, respectively. The model supports the conclusion that both lifetime and recent experiences of high inflation raise the level and sensitivity of long-term expectations. The model shows that recent high inflation has indeed significantly increased the level and sensitivity of long-term expectations relative to 2019, with a larger increase among younger respondents than among older respondents. However, after combining the effects of more than a decade of low inflation, over five years of high inflation, and the gradual fading of memories of high inflation from decades ago, both indicators remain below their mid-2000s levels and only slightly above the counterfactual result that assumes inflation remained at 2% throughout 2009—2026. Thus, although the high-inflation period has left a mark, it is not sufficient to demonstrate that long-term expectations have become significantly unanchored. The report also examines the dispersion of expectations, perceived inflation persistence, and public attention to inflation. The dispersion of long-term expectations in the New York Fed survey is moderately elevated, while that in the University of Michigan survey is at a historical high. Dispersion remains elevated even when calculated within political parties, although part of the increase in the Michigan data may stem from changes in survey methodology. Perceived persistence is significantly above the pre-pandemic norm in the Michigan survey but close to normal in the New York Fed survey. Meanwhile, attention—as measured inversely by the share of respondents unable to provide short-term inflation expectations—rose significantly when pandemic-era inflation surged but has broadly returned to pre-pandemic levels over the past few years. Overall, Goldman Sachs judges that short-term expectations are somewhat elevated, long-term expectations are broadly consistent with pre-global-financial-crisis levels, and expectations in aggregate are at most moderately elevated, with no immediate risk of becoming unanchored. The outlook depends on actual inflation: if oil prices stabilize, tariff effects gradually drop out of year-over-year calculations, and inflation returns to target by the end of 2027 as forecast, lower actual inflation and the public's increasing distance from the shocks of recent years should exert downward pressure on inflation expectations in 2027.

Analysis framework

The report first reviews academic research to explain how inflation expectations pass through to wages, pricing, and economic activity, as well as the respective roles of actual inflation experience and monetary policy communication. It then compares indicators for households, firms, professional forecasters, and markets, with particular emphasis on household and business surveys. To address discrepancies between the New York Fed and University of Michigan surveys, the report further uses individual-level historical data to construct a memory model that estimates how recent, lifetime, and similar historical experiences affect the level and sensitivity of expectations. It then uses a counterfactual path of sustained 2% inflation to assess the net effect of the past two inflation cycles. Finally, it cross-validates the findings using the dispersion of expectations, perceived persistence, and public attention.

Methodology notes

• Macroeconomic framework

  Academic research review of inflation-expectation pass-through coefficients

  The report synthesizes estimates from multiple economic studies to measure how actual inflation passes through to long-term expectations and how short-term expectations affect wages and firms' pricing. For studies that report only short-term changes, the report assumes a 25% pass-through from short-term to long-term expectations.

• Macroeconomic framework

  Memory model based on lifetime inflation experience

  The model uses individual-level University of Michigan survey microdata from 1978—2026, constructs lifetime inflation experience beginning at age 16, and incorporates both recent experience and the similarity between the current environment and historical experience to explain the level and anchoring of long-term expectations.

• Macroeconomic framework

  Counterfactual analysis of sustained 2% inflation

  The report compares the model results with a hypothetical path in which inflation remained at 2% throughout 2009—2026 to isolate the net effects of more than a decade of low inflation, over five years of high inflation, and the fading memory of earlier high inflation.

• Macroeconomic framework

  Indicators of expectation dispersion, perceived persistence, and attention

  The report measures expectation dispersion using the standard deviation in the Michigan survey and the interquartile range in the New York Fed survey, measures perceived persistence by regressing individuals' long-term expectations on their short-term expectations, and inversely measures public attention to inflation using the proportion responding “don't know.”

Key data

• Average pass-through from short-term expectations to wage expectations21 basis pointsThe average change in consumers' wage expectations when short-term inflation expectations change by 1 percentage point.
• Average pass-through from short-term expectations to firms' pricing29 basis pointsThe average change in firms' pricing behavior when short-term inflation expectations change by 1 percentage point.
• Average pass-through from actual inflation to households' long-term expectations13 basis pointsThe average increase in households' long-term inflation expectations corresponding to a 1 percentage point increase in actual inflation.
• Average pass-through from actual inflation to professional forecasters' long-term expectations9 basis pointsThe average increase in professional forecasters' long-term expectations corresponding to a 1 percentage point increase in actual inflation.
• Public estimate of the Federal Reserve's inflation targetAbove 8% in 2018; slightly above 3% in 2025Public attention to monetary policy increased during the high-inflation period, but the report believes that actual inflation remained the dominant influence.
• University of Michigan 5-to-10-year inflation expectations3.3%Only 0.1 percentage point below the peak over the 25 years before the pandemic.
• Memory model sample period1978—2026Uses individual-level University of Michigan survey microdata, with lifetime inflation experience calculated from age 16.
• Memory model coefficients for the level of long-term expectations0.58, 0.34, 0.25Respectively correspond to a 1 percentage point increase in average lifetime inflation experience, recent inflation experience, and the similarity between the current environment and past experience.
• Memory model coefficients for expectation sensitivity0.04, 0.04, -0.02Respectively correspond to the effects of average lifetime experience, recent experience, and similarity between current and historical experience on the sensitivity of expectations to changes in actual inflation.
• Counterfactual inflation pathSustained at 2% throughout 2009—2026The current level and sensitivity of long-term expectations are only slightly above this counterfactual result.
• Forecast for inflation to return to targetBy the end of 2027Conditional on stable oil prices and tariff effects dropping out of year-over-year calculations.

Impact & implications

The report's policy implication is that stabilizing inflation expectations cannot rely primarily on Federal Reserve communication, because households and firms typically increase their attention to policy only after high inflation has already occurred, at which point actual inflation exerts a stronger influence on expectations. Sustained reductions in actual inflation are therefore essential to lowering short-term expectations again and preventing the sensitivity of long-term expectations from rising further. Current data do not support a conclusion of immediate unanchoring, but younger groups are more sensitive to recent high inflation, and significant differences among surveys also mean that no single indicator should be relied upon.

Risks

• If above-target inflation persists for too long, household and business inflation expectations could rise further and become more prone to becoming unanchored.

What to watch

• Monitor whether actual inflation can return to target by the end of 2027, because the report argues that sustained declines in actual inflation anchor expectations more effectively than policy communication.
• Monitor whether oil prices stabilize and whether short-term household and business expectations, previously driven higher by the oil price shock, decline.
• Monitor when tariff effects drop out of year-over-year inflation calculations and the resulting extent of inflation cooling.
• Continue comparing long-term expectations, expectation dispersion, and perceived inflation persistence in the New York Fed and University of Michigan surveys.
• Monitor whether the level of expectations among younger groups and their sensitivity to actual inflation continue to rise.
• Monitor whether public attention to inflation remains near normal pre-pandemic levels.