"Excess" Savings Are Higher Now (Sort Of): More Highlights from the 2023 Comprehensive NIPA Revisions
How changes in pre-pandemic estimates of asset income affect the aggregate numbers. Plus, a look at the August PCE inflation numbers.
The latest comprehensive revisions to national income suggest that American consumers are less financially stressed than previously believed—at least relative to what might have been expected in the absence of the pandemic. While previous estimates implied that the “excess” savings accumulated in 2020-2021 were almost exhausted, the new figures suggest that Americans have retained the majority of this “excess”. At current rates of erosion, this “excess” could persist for almost another three years. Prior estimates implied that it would be gone within months. However, a closer look at the data suggests that the implications for the trajectory of spending, growth, and inflation have not changed.
A Brief Recap of What “Excess” Savings Are
Americans—like their counterparts across the rest of the rich world—saved far more than normal during the first year of the pandemic. The virus limited consumers’ opportunities to (safely) spend money, crushing many businesses’ sales. In response, they cut their wage bills via layoffs, hours reductions, and pay cuts.
Rightly concerned by the prospect of a downward cascade into a financial crisis and depression, governments around the world responded by borrowing and spending tens trillions of dollars to support incomes. The net result was that personal incomes rose even as consumer spending fell far more, causing saving to rise.1 In the U.S., the cumulative impact from February 2020 through March 2021 was worth $2.1 trillion. Much of that surplus went into bank accounts and money market funds, although some also went towards housing2 and debt reduction.
Once mass vaccination allowed the economy to reopen, spending normalized—and then began to rise faster than disposable income as inflation started to bite. The household saving rate had switched from being far above normal to far below. Ostensibly, whatever “excess” had been accumulated earlier was now being used to finance more consumer spending than would otherwise have been possible. Many feared that this was contributing to inflation, although there was also hope—or fear, depending on who you asked—that the imminent disappearance of the “excess” would force consumer spending lower and therefore help restore normalcy.
Regular readers know that I have repeatedly questioned elements of this narrative. In particular, the pre-revision data implied that much of the 2022 decline in the saving rate could be attributed to companies’ (apparent) preference for rewarding shareholders via buybacks rather than dividends. That created large capital gains tax exposures, which count against personal income, while depressing dividend income. And since capital gains are not counted as income in the national accounts,3 the net result was that “household saving” was lower even as “corporate saving” was higher. By contrast, consumer spending has consistently grown in line with aggregate wage income since the reopening in Spring 2021, which suggested that the changes in personal saving rates were not being driven by the vast majority of people who do the vast majority of consumer spending.
How the Revisions Change this Picture
As I explained in my previous note, the 2023 comprehensive revisions included methodological changes that redefined some financial corporate interest payments as profits and that lowered the value of dividends paid to households by REITs. Meanwhile, new tax data from the IRS caused the Bureau of Economic Analysis (BEA), which compiles the national accounts, to raise their estimates of corporate profits and dividend payouts while slashing their estimates of self-emploment and noncorporate business income. Finally, the BEA also concluded that American homeowners were “spending” more in imputed rent on their homes.
The net effect is that capital income earned by households was lower than previously believed before the pandemic and higher than previously believed more recently.
This has ramifications for any estimate of “excess” saving. All of us who have been making these estimates in the past three years have relied on some pre-pandemic baseline. The new data imply that Americans saved about $1.6 trillion less between January 2013 and January 2020 than previously believed ($6.0 trillion vs. $7.6 trillion). Relative to the lower baseline, the post-reopening saving rate does not look as low, which also means that the “excess” is not diminishing nearly as quickly.
However, this is due almost entirely to changing estimates of asset income before the pandemic.