On My Mind: The “Scarlet K”—A misleading consumption divergence story

The US economic outlook has been branded with a new scarlet letter: a K. Countless media reports warn us that US growth has bifurcated, that affluent households now carry the entire economy on their shoulders, and that any wobble in equity markets will bring the whole edifice down.

The "K-Shaped Thesis" in a nutshell goes as follows: US consumption has split into two arms—affluent households spending freely on the back of equity gains, and lower-income households cutting back under the weight of high inflation and stagnant wages. Aggregate growth, the argument runs, is now structurally dependent on a narrow asset-rich slice of the population, and thus highly vulnerable to any correction in stock prices. This narrative, in my view, is quite misleading; it vastly overstates both the shift in consumption trends and the fragility it implies.

Making sense of the data

The first thing to note is that the underlying data are shaky, patchy and late.

The most alarmist conclusions come from a Moody's report,¹ claiming that the top 10% of households (by income) account for nearly half of all consumer spending, following a K-shaped divergence that started around 2020. Moody's reaches that number through a rather circuitous route. The report takes changes in household assets and liabilities in the Federal Reserve's (Fed) quarterly Financial Accounts to estimate personal savings (latest data for fourth quarter 2025). Then it uses data from the Survey of Consumer Finances to map changes in savings and personal disposable income to different income groups. The difference gives consumption by income groups.

There are several difficulties with this approach, not least the fact that the most recent Survey of Consumer Finances dates to 2022. This is a model-based construction layered on top of triennial survey data and held-constant portfolio assumptions. It is not a direct read of what households are buying.

Isn’t there a more direct measure of consumer spending?

Yes—the Bureau of Labor Statistics (BLS) Consumer Expenditure Survey. It puts the top 10%'s spending share at 23% in 2024, broadly stable since 2004. This survey covers only 60% of personal consumption expenditures. To address this issue, the BLS now maps it to personal consumption expenditures data, (only through 2023 so far)² reaching nearly identical results. The Consumption Expenditure Survey is believed to underestimate consumption by higher-income households. But then in all likelihood it has always done so—and shows no change in their consumption share.

The BLS figures do not tell us what's been happening over the past 18 months. However, Bank of America reports on credit and debit card spending by its customers, which does offer additional insight. This limited sample shows a significant divergence between lower, middle, and upper-income customers starting around mid-2025. The New York Fed’s quarterly Economic Heterogeneity Indicators show a moderate divergence in spending between lower, higher, and middle-income households, which emerged in 2023 and has remained stable since.³ In both cases, rather than a K, we see an upward tilted E: Three lines that diverge briefly and then follow broadly parallel upward trends.

On balance, as the Federal Reserve Bank of Minneapolis concludes in a March 2026 review, the available data "do not align to tell a clear, K-shaped story."⁴

The rising tide that lifts all boats

What can we conclude from this patchwork of data? There is evidence of some recent divergence in consumption patterns. It is tempting to link it to rising equity prices and high inflation, but the evidence does not align so neatly on the timing. BLS data peg the consumption share of the top 10% at about one quarter. It may be underestimated, but by 100%? And given that the BLS data show the share held broadly constant for 20 years, it seems rather implausible that it would have broken out so sharply over just the past 18 months.

Concerns about the supposed fragility of the US consumption outlook seem correspondingly exaggerated to me. First, the observed divergence appears modest. Second, wealthy households have, by definition, a healthy cushion—it is unlikely that a stock market correction will cause consumption to crater. A recent analysis by the Federal Reserve Bank of Dallas, which adopts a methodology similar to Moody's, finds divergence in the consumption of low-income, middle-income and high-income households of only a few percentage points since the 1990s. It concludes this makes the economy only slightly more vulnerable to weakening returns on financial assets.⁵

In my view, the “K-shape” narrative is profoundly misleading and diverts attention from a very important fact: In the United States overall, incomes have been rising across the board, and meaningfully so. The “K” moniker suggests that while “richer” households have been earning and spending more, “poorer” households suffered shrinking incomes and consumption. That is not at all what’s been happening, according to the data. Yes, households at the top of the income distribution have increased their share of the pie, but incomes have been rising broadly across the population. This is a crucial distinction, because we need to separate the issue of inequality from that of the economy’s resilience. Inequality is an important problem in itself. But when everyone’s incomes are rising, the economy becomes more resilient even if some households gain more than others.

A recent analysis by Rose and Winship⁶ divides households into five categories, with real income thresholds defined relative to the poverty line: poor, lower middle-class, core middle-class, upper middle-class and rich. It finds that between 1979 and 2024, the share of households that are poor or lower middle-class has declined from 53.8% to 34.5%, whereas the share of households in core or upper middle-class rose from 45.9% to 61.9%. The upper middle-class has seen the biggest rise in size, by over 20 percentage points. The rich have been getting richer, but so has everyone else, and this confers a high degree of resilience to the US economy.

Case in point: The latest New York Federal Reserve survey of consumer expectations shows that since the beginning of the year, spending growth expectations for households making less than US$50,000 a year have outpaced those of higher-income households.