The Washington Post wonders if the US has fallen into a new recession. The motivation for asking is the stall-speed growth in first-quarter GDP, which increased by a slim 0.2%. The fractional gain is quoted in quarter-over-quarter terms, which is the standard reference. By that measure, there’s virtually no growth. But are quarterly comparisons the best way to measure GDP in search of macro danger? It’s a timely question because the annual change in GDP through this year’s first quarter reflects a substantially stronger trend. In fact, GDP’s 3.0% year-over-year advance in Q1 accelerated, rising at the strongest pace in more than a year.
Here’s how GDP changes compare in quarter-over-quarter and year-over-year terms:
Looking at the Q1 data from both perspectives dispenses radically different views. Using the quarterly comparison (blue line), the economy appears to be sinking into a new slump. But the year-over-year data (red line) indicates that growth is picking up.
Why the divergence? The explanation is largely due to 2014’s first-quarter slide. GDP contracted 0.5% in the first three months of 2014 on a quarter-over-quarter basis, which means that the slight increase in GDP for the quarter that just passed looks quite good in annual terms.
Does that mean that the annual comparison for GDP is misleading? Maybe, but before we answer definitively let’s consider the year-over-year change through history as a guide for assessing recession risk. One lesson in the data since 1950: there have been no recessions when GDP is rising at 3.0% in annual terms.
The telltale sign for annual comparisons in the search for heightened recession risk is a sharp deceleration in growth when rates of increase are below 2.5%. By that standard, the threat of recession was low in the first quarter.
That’s also the message in several alternative measures of business cycle risk. Professor Jeremy Piger’s recession-risk index through February estimates the probability of a downturn as virtually nil. The Capital Spectator’s estimate of recession risk was similarly low for the March profile. The three-month average of the Chicago Fed National Activity Index at the end of Q1 is weaker, although the latest reading is still well short of the tipping point that has historically been linked with the start of downturns.
In other words, the decision on whether the US has slipped into a new recession will probably be decided with second-quarter data and well before we see the government’s first Q2 GDP estimate that’s scheduled for release on July 30. Meantime, trying to date the start of a recession based on the quarterly change via a single GDP data point is like trying to evaluate the stock market’s trend by looking at one closing price.
Monitoring the business cycle with GDP numbers in general isn’t a great idea to begin with… quarterly data in particularly can be quite noisy. But if you’re going to look at quarterly comparisons, some perspectives are more valuable than others. For example, crunching the quarterly changes via a probit model, for example, offers a more robust view of the trend. But here too there’s no smoking gun as of Q1, as I discussed last week.
We may ultimately learn that the US economy entered another recession. The Atlanta Fed’s current Q2 nowcast certainly looks weak, with an estimate of growth at just 0.8%. The trend is certainly wobbly these days. But while it feels like a good time to make a dramatic statement about recession risk, the numbers overall still don’t offer much support for assuming the worst. That may change in the weeks ahead, but for now a simple fact endures: you can’t make a recession call based on the latest quarterly change in Q1 GDP.