No, the top 10% of households are not responsible for 50% of personal spending
Update: After failing to reach him through Moody’s, I was able to contact Mark Zandi directly, and he kindly provided me with a document explaining the methodology. I’m still very skeptical, but I plan to attempt a replication when I get the time.
Last Sunday, the Wall Street Journal wrote up a new report by Mark Zandi of Moody’s Analytics with a sensational claim that the top 10% households by income now account for nearly half of all spending:
The top 10% of earners—households making about $250,000 a year or more—are splurging on everything from vacations to designer handbags, buoyed by big gains in stocks, real estate and other assets.
Those consumers now account for 49.7% of all spending, a record in data going back to 1989, according to an analysis by Moody’s Analytics. Three decades ago, they accounted for about 36%.
All this means that economic growth is unusually reliant on rich Americans continuing to shell out. Mark Zandi, chief economist at Moody’s Analytics, estimated that spending by the top 10% alone accounted for almost one-third of gross domestic product.
The story has now metastasized to Bloomberg, with this chart:
I’ve been unable to find the actual report, and have been unsuccessful in my attempts to contact Zandi and Moody’s Analytics to get information on the methodology, so it’s possible that there is some definition "of “personal spending” for which this chart is accurate.
However, under the most straightforward interpretation, the chart is incorrect, and the claim that spending by the top 10% accounts for 33% of GDP is definitely false, and so far off that it should be obvious to anyone familiar with the broad strokes of the data on economic inequality in the United States.
The main source of data we have on the distribution of consumption spending in the United States in the BLS’s Consumer Expenditure Survey, which regularly surveys a sample of Americans on income and expenditures. Of particular relevance here are the annual demographic tables, which, among other things, show the distribution of expenditures among consumer units1 by decile of income. Earlier years are available in PDF, but data for 2023 are currently available only in Excel format.
In 2023, the mean annual expenditure for all households was $77,280, and $180,758 for the top decile. To find the top 10%’s share, we can just divide the latter by ten times the former, yielding 23.4%, less than half the share claimed by Zandi. These numbers include pension and Social Security contributions, which are a larger percentage of expenditures for the top decile than overall; if we exclude these for all deciles, the top 10% share falls to 21.8%. I tried excluding housing and health care, on the theory that Zandi might not count these as personal or consumer spending, but it only increased the top 10%’s share by about 2 percentage points.
Now, these are 2023 data, and the chart above shows a jump around the beginning of 2024, and the 49.7% figure specifically is up to Q3 2024. But the jump appears to be about 4 percentage points, nowhere near enough to bridge the vast gulf between Zandi’s claim and the CEX data.
Okay, but what if the CEX data are just wrong? Maybe they undersample the ultra-rich, and in doing so miss half of the spending of the top 10%.
We can sanity-check this by looking at the top 10%’s share of after-tax income. It is virtually always the case that consumption is distributed considerably more equally than income. High-income households tend to save money to consume at a later time when their incomes are lower (such as in retirement), leading to a compression of consumption inequality relative to income inequality. The top decile’s share of after-tax income should give us a loose upper bound on the top decile’s share of consumption.
The CBO publishes an annual report on the distribution of household income with a three-year lag; the 2021 report was just released last September. Table C-3 shows that the after-tax income share of the top 10% was 34.3% in 2019 and 33.3% in 2020, and 37.1% in 2021. According to the Moody’s chart above, the top 10% share of consumption was around 44-47% in those years. In order to reach that share of spending, the top 10% would have had to spend all of their annual income and dip deep into their savings.
There have been some attempts to take a more rigorous approach to correcting for underreporting in the CEX and include components not measured in the CEX:
A Distributional Approach to U.S. Personal Consumption Expenditures: An Overview: Using a broader measure of consumption expenditures, the BLS found that the share of expenditures for the entire top 20% was about 41% for 2017-2021.
Consumption and Income Inequality in the US Since the 1960s: Meyer and Sullivan found that consumption inequality was consistently and substantially lower than after-tax income inequality.
The claim that spending of the top 10% accounts for a third of GDP is, I believe, obtained by multiplying the supposed 49.7% of spending attributed to the top 10% by the personal consumption share of GDP, currently about 68%, yielding just over one third.
Again, without seeing Zandi’s methodology, I can’t conclusively say that there’s not some sense in which the chart above is accurate. But if this is correct, it’s a huge finding that completely overturns everything we thought we knew about the distribution of personal consumption expenditures in the US, and more explanation than “We used Federal Reserve data” is needed here.
As explained in the linked Wikipedia article, a consumer unit can be: “(1) All members of a particular household who are related by blood, marriage, adoption, or other legal arrangements; (2) a person living alone or sharing a household with others or living as a roomer in a private home or lodging house or in permanent living quarters in a hotel or motel, but who is financially independent; or (3) two or more persons living together who use their incomes to make joint expenditure decisions.”