For the past year, you have undoubtedly heard a certain number mentioned in one way or another nearly every week. 45.8% of all consumer spending in the United States is attributed to the top 10% of earners. It sounds conclusive. It has the clarity of a conclusion that has already been debated and decided. It is used by cable hosts. Staff members at the Senate incorporate it into talking points. Wealth managers send it to clients via email with just one line of commentary, as though no commentary is required. However, the number starts to feel more like a snapshot taken under a particular type of light when you press on it.
The problem is that the number depends on how economists treat a concept known as imputed rent in ways that most readers are unaware of. That is the approximate monetary value of the housing services that a homeowner essentially “pays” for by owning as opposed to renting. Although it is a legitimate concept that has been utilized in national accounting for many years, no one actually spends it.
| Topic | The K-Shaped Economy’s measurement problem |
| Core statistic in question | Top 10% of earners responsible for 45.8% of U.S. consumer spending (2024) |
| Source of the figure | Moody’s Analytics |
| Originally reported | Mark Zandi, chief economist, Moody’s Analytics |
| Revised methodology | Imputed rent now treated differently; figure recalculated downward |
| Key counter-data source | Federal Reserve Bank of New York, Liberty Street Economics blog |
| When divergence began | Roughly 2023, after pandemic-era subsidies expired |
| High-income threshold | Households earning above $125,000 annually |
| Lower-tier signals | Rising credit card balances; average now $6,519 per consumer |
| Credit research source | TransUnion Q1 2026 Consumer Pulse |
| Related Fed research | New York Fed Liberty Street Economics |
| Top 1% wealth share | Approximately 32% of total U.S. wealth |
| Why it matters | Policy decisions, recession risk models, election framing all lean on this figure |
The top decile’s contribution skyrockets when it is included in consumer spending and heavily weighted toward homeowners, who tend to be wealthier. If you remove it again or handle it more cautiously, the number decreases by a few points. Not enough to break the K-shaped narrative. However, enough to raise the question of how much of the most widely cited version of that tale was a methodological decision disguised as a discovery.
It’s difficult to ignore how fast the initial framing gained traction. The 45.8% number moved in a manner that most economic statistics never do after Moody’s published its analysis. There is a feeling that it came just when people needed a single figure to describe how they already felt—that the nation had split into two economies operating in parallel. The mood was supported by the data. And once a number does that, it is very hard to go back.
Liberty Street researchers at the New York Fed have been more cautious. Their research reveals a true split that started around 2023, when lower-income households bore the brunt of persistent inflation and pandemic-era support waned. There is no question about that part. The story is evident if you stroll through any suburban strip mall on a Tuesday afternoon: the wine bar two doors down has reservations for the following month, and the dollar store has a line at the register. The same general pattern can be seen from a different perspective in TransUnion’s data on credit card balances, which currently average $6,519 per customer.
However, “real” and “as dramatic as advertised” are not the same thing. Political strategy memos and interest rate forecasts have both been justified by the K-shaped narrative. The consequences spread if the headline figure exaggerates the concentration of spending by even four or five percentage points. Recession models may be calibrated to a phantom if they assume that the economy is particularly vulnerable to a pullback by the wealthy. It’s possible that policymakers who believe that a small percentage of luxury consumers support consumption are operating from a sketch rather than a map.
This does not negate the existence of the divide. Since 2020, homeowners‘ property values have increased by about 50%. Approximately 32% of the country’s wealth is still controlled by the top 1%. These facts are hard. The certainty that one neat percentage has been permitted to represent a much messier reality is what is soft. As you watch this develop, you begin to believe that the most pertinent question is not whether the K-shaped economy exists—which it obviously does—but rather whether the nation has been debating the incorrect version of it for nearly two years.