Sticky wage norms and the real wage cost of unexpected inflation

5757 S. University Ave. Chicago, IL 60637 Main: 773.702.5599 bfi.uchicago.eduWORKING PAPER · NO. 2026-108 Sticky Wage Norms and the Real Wage Cost of Unexpected Inflation Erik Hurst, Christina Patterson, Nela Richardson, and Ye Liv Wang AUGUST 2026 Sticky Wage Norms and the Real Wage Cost of Unexpected Inflation∗ Erik Hurst Christina Patterson University of Chicago University of Chicago Nela Richardson Ye Liv Wang ADP Research ADP Research August 12, 2026 Abstract We use a sample of administrative payroll data covering a large and nationally repre- sentative share of U.S. workers to study how wages adjusted during the recent inflation period. Most firms apply a single modal annual nominal wage increase to the majority of their workers, and these firm-level norms changed little during the recent period of unexpected inflation. As a result, nominal wages did not keep pace with prices for a large share of workers who stayed at their firms. Forty-three percent of workers contin- uously employed at the same firm over the four years spanning 2021–2024 experienced a real wage decline, with a mean loss of roughly nine percent among those who fell be- hind. Workers could escape sticky wage norms by changing employers — job-changers’ wages rose nearly one-for-one with inflation — but switching was too infrequent to matter for most. Even accounting for job-changers, 37 percent of all workers saw real wages decline over the 4-year period. Indexing firms’ modal raises one-for-one to inflation would have closed roughly 40 percent of the resulting shortfall relative to pre- pandemic trend. Drawing on cross-country evidence from Belgium, whose wages are automatically indexed to inflation, we show that incomplete wage indexation, rather than inflation itself, helps explain the persistence of depressed consumer sentiment during the 2021–2024 period. ∗Email: erik.hurst@chicagobooth.edu, christina.patterson@chicagobooth.edu, nela.richardson@adp.com, and liv.wang@adp.com. We thank Gianluca Violante, Greg Kaplan, Jonathan Parker and Benjamin Schoefer for very helpful comments. We also thank seminar participants at the 2026 NBER Summer Institute. Per our data use agreement, ADP approved the paper topic ex-ante. Additionally, ADP reviewed the paper prior to the paper’s circulation with the sole focus of making sure the paper did not release information that would compromise the privacy of their clients or would reveal proprietary information about the ADP business model. 1 Introduction After nearly four decades of low and stable inflation, U.S. consumer prices rose sharply and unexpectedly from mid-2021 through late-2023, with the inflation rate peaking at 9 percent in June of 2022. During this period, U.S. consumer sentiment fell to levels not seen since the depths of the Great Recession despite the unemployment rate being at historically low levels.1 Over 70 percent of Americans in 2022 reported that inflation was a “very big problem” (Pew Research Center 2022). In early 2024, when the inflation rate returned to normal levels, over 40 percent of Americans still claimed that “inflation/high cost of living” was the most important financial problem facing their family (Gallup 2024).2 Americans’ unhappiness with inflation has been cited as an important factor in voting patterns during the 2024 presidential election (Steinberg et al. 2025; Ria˜no and Trebbi 2026; Leonhardt 2024). Importantly, even three years after the inflation rate peaked, Americans are still reporting that inflation and affordability are pressing concerns. As of 2025, U.S. consumer sentiment remains depressed, roughly 30 percent of Americans still report that cost of living is their most pressing financial problem, and over 60 percent of Americans still report that inflation is a very big problem. In this paper, we argue that Americans’ enduring discontent with inflation reflects the lasting real wage losses generated during the inflationary episode. Using a sample of admin- istrative payroll data from ADP covering roughly 16 million workers per month, we examine real wage dynamics and firm wage-setting from 2016 through 2025. We show that most firms anchor annual raises to a common wage norm applied broadly across workers, and that these norms adjusted only modestly when inflation surged. Larger off-cycle raises and job changes allowed some workers to keep pace, but these responses were too limited to restore real wages for the typical worker. A temporary inflation shock therefore produced a persistent downward shift in real wages, helping explain why Americans’ dissatisfaction outlasted the inflation episode itself. Figure 1 serves as the launching point for our paper. The figure shows a real wage index for the United States computed using either data from ADP (solid line) or data from the Current Population Survey (CPS) (dashed line). The two lines track each other closely, highlighting that the ADP data is representative of the U.S. labor market with respect to 1The University of Michigan Consumer Sentiment Index bottomed out at a level of 57.4 in the fourth quarter of 2008. In the third quarter of 2022, the Consumer Sentiment Index plummeted to 56.1. For reference, the index only fell to 74.1 in the second quarter of 2020, when employment losses during the pandemic were at their largest. 2For comparison, the share reporting inflation as the most important financial problem facing their family averaged around 8 percent in the Gallup surveys conducted between 2000 and 2021 — never exceeding 18 percent even at the peak of the Great Recession. 1 Figure 1: U.S. Real Wages 2017–2025: ADP and CPS Notes: Figure shows real wage indices using CPS and ADP panel micro data. The CPS wage series comes from the weighted median nominal wage growth for all matched CPS respondents as reported by the Atlanta Fed Wage Growth Tracker. The ADP wage series combines the median nominal wage growth of job-stayers and job-changers. An index is created for both series using the monthly nominal growth rates, normalized to 1 in January 2017, and deflated by the corresponding monthly CPI-U in January 2017 prices. The red dashed line extrapolates the 2017–2019 linear trend in the ADP real wage index through December 2025. See text for additional details. wage dynamics — a point we discuss in greater depth in Section 2.3 Median real wages fell by 4 percent during the inflation period, and did not recover to its 2020 level until late 2024. Moreover, zero real wage growth is not an appropriate benchmark for assessing whether real wages kept up. The ADP and CPS wage indices are defined by using the wage growth of a given worker over time. There is a large literature documenting that the wages of a given worker increase year-over-year due to the accumulation of general or firm-specific human capital; such life cycle wage growth occurs even in a world where there is no aggregate productivity growth. The red dashed line in Figure 1 is the predicted real wage path over the 2020–2025 period, based on extrapolating the 2017–2019 trend in the ADP data forward. During the 2017–2019 pre-period, real wages grew at roughly 2.2 percent per year.4 As of December 2025, the real wage index stands roughly 7 percent below where real wages would have been had the 2017–2019 trend continued. The 2017–2019 trend is a demanding benchmark, since real wage growth in the late 2010s was unusually strong. Measured against 3The CPS data come from the Atlanta Fed Wage Growth Tracker. Both the ADP and CPS series combine data from job-stayers and job-changers and use the panel structure of the data to compute the median nominal wage growth of individuals who are employed 12 months apart. We create a nominal wage index using the growth rates, normalizing the index to 1 in January 2017. We deflate the nominal wage index by the corresponding month’s CPI-U in January 2017 prices. Section 2 provides additional details on how we constructed this figure. 4This is in line with estimates of the annual return to experience. See, for example, Mincer (1974), Topel (1991), Altonji and Williams (2005), and Dustmann and Meghir (2005). 2 average real wage growth of 1.5 percent per year over the longer 2000–2019 period, the shortfall as of December 2025 was still roughly 4 percent. The goal of this paper is to use the richness of the ADP data to shed light on the mechanisms driving the dynamics of real wages during this period. First, we document that job-stayers’ wages did not keep pace with inflation. Median year-over-year real wage growth fell from roughly 1 percent in 2017–2019 to about negative 4 percent during the inflation surge. Although real wage growth returned to its pre-period rate by mid-2023, the earlier losses were never made up, leaving job-stayers’ real wages persistently below trend. The long panel reveals how widespread these cumulative losses were. Among workers continuously employed at the same firm from December 2020 through December 2024, 43 percent ended the period with lower real wages than at its outset; conditional on a decline, the mean loss was 9 percent and the median loss was roughly 7 percent. More broadly, 55 percent averaged less than 1 percent annual real wage growth over this 4-year period. Both the incidence and magnitude of these losses were far greater than among comparable four-year job-stayers before the pandemic. Thus, nominal wage growth failed to keep pace with inflation for a substantial share of workers who remained with the same employer. We next examine the firm-level wage-setting practices behind these losses. Because ADP records base-wage changes for every worker within a firm, we can identify the firm-specific month in which most annual raises occur—the firm’s “on-cycle” month—and distinguish these raises from “off-cycle” adjustments in other months. Off-cycle raises are systematically larger and are consistent with worker-specific events such as promotions or responses to outside offers.