When the Commerce Department releases its preliminary estimate for second-quarter economic output on Thursday morning, the numbers will be historically terrible.
They will also be confusing.
Forecasters expect the report to show that gross domestic product — the broadest measure of goods and services produced in the United States — fell at an annual rate of about 35 percent.
But wait. If you read that last paragraph quickly, you might have come away with the impression that the economy shrank by more than a third in a mere three months. That’s wrong. If forecasters are on target, economic output was about 10 percent lower in the second quarter than in the first — still awful, but not quite as scary-sounding as a 35 percent drop.
Either way, it is expected to be the worst quarter in the 70-plus years that quarterly G.D.P. statistics have been compiled. But here’s why you may see two widely different numbers.
The United States has traditionally reported G.D.P. and some other economic statistics as annual rates. Rather than simply giving the percentage change from one quarter to the next, the government reports how much G.D.P. would grow or shrink if that rate of change were sustained for a full year. (Because growth rates compound on themselves, this calculation is a bit more complicated than simply multiplying by four. The figures are also adjusted for seasonal patterns, so they are more properly described as “seasonally adjusted annual rates.”)
Annual rates make it easier for analysts to compare data collected over different time periods. If you’ve ever estimated how much you’d save over a full year if you kicked your daily latte habit, or worked out how many home runs a favorite player would hit if his current hot streak lasted for a full season, you’ve performed a similar calculation.
But when annual rates are applied to short-term changes, the results can be misleading, as Neil Irwin of The Upshot explained in May. If you received a $500 bonus one month, you wouldn’t think of it as a “$6,000 raise, on an annualized basis,” because you know it’s a one-time windfall, not a long-term change in your income.
Right now, the economy is going through extreme short-term changes. Activity largely halted in much of the country in April, rebounded sharply in May and June, and now looks as if it might be slowing again as surging coronavirus cases force states to slow or reverse reopenings. Those changes are real, and will have a huge effect on family incomes, business profits and state tax revenues. But it doesn’t make much sense to think of them on an annualized basis.
For that reason, in Times coverage of Thursday’s G.D.P. report, we plan to emphasize the simple, nonannualized percentage change from the first quarter to the second. (This is hardly revolutionary — it’s the way most of the rest of the world already reports G.D.P. data.) We’ll still provide the standard annualized figures for people who are used to seeing them that way. And where relevant, we’ll show other calculations, such as the change from a year earlier, or the change since the start of the pandemic.
We plan to follow the same approach for the third quarter, which will almost certainly show a big — and equally misleading — increase in economic activity after the big shutdown in the spring.
