Following a decline of -1.6% in the first quarter, the advance estimate from the Bureau of Economic Analysis suggests that gross domestic product (GDP) decreased at an annual rate of -0.9% in the second quarter. That revelation has sparked a debate about whether or not the US is actually in recession.
The official arbiter of the recession record in the US is the National Bureau of Economic Research (NBER). If you care to listen, you’ll hear the dusty boffins that form the NBER’s Business Cycle Dating Committee consulting on the ‘depth, diffusion and duration’ of the various data points. Their criteria marks a recession out as a ‘significant decline in economic activity that is spread across the economy and lasts for more than a few months’.
My own definition is not dissimilar. In my mind, recessions are characterised by a sustained decline in output and a similarly sustained incline in joblessness.
There is a more commonly accepted definition though – one that has the benefit of being less subjective – and it simply states that a recession comprises two or more consecutive quarters of shrinking GDP. On that basis, and assuming that the data is not significantly revised, the US was indeed in recession during Q1 and Q2.
Having said that, it’s worth mentioning that the commonly accepted definition is not without its problems. For a start, it pays no heed to the extent of the decline, leading to the absurd implication that a contraction of just -0.1% in each of two consecutive quarters would signify recession but a drop of -10% in one quarter would not. More importantly, it makes no reference to the employment situation.
Let’s delve into the controversy by looking at some numbers…
Before we do, bear in mind that our American cousins have some funny habits, like smiling at strangers, and keeping eggs in the fridge. Another is quoting quarterly GDP results at an ‘annual rate’ rather than as a straightforward change like we do over here. So, while a reported -1.6% annual rate of decline for Q1 sounds dramatic, it amounts to a quarter-on-quarter change of just -0.4%. And the -0.9% annual rate of decline in Q2 translates into a fall of just -0.2%.
As it happens, +/-0.2% is well within the kinds of limits we’ve frequently seen revised entirely away between the initial ‘advance’ estimate and later more complete estimates.
Somewhat counterintuitively, the employment situation has not deteriorated over the period in question, at least at the headline level. Indeed, the Bureau of Labor Statistics reports an increase in employed positions and, on a separate measure, a decrease in the rate of unemployment. ‘Total nonfarm payrolls’ grew by an additional 2.7 million jobs and the ‘household survey’ reveals unemployment fell from 4.0% in January to 3.6% in June.
Thus far, the US economy has lost -0.6% of output with no deterioration in the employment situation. I wonder where that kind of a recession would rank in the list of post-war recessions…
Prior to the current recession, and excluding the ‘unusual’ 2020 pandemic recession, I count 9 which accord with the commonly accepted definition. Most of those recessions lasted just two quarters but there is one comprising four consecutive quarters of contraction and two further recessions lasting as long as three quarters. The average decline in output amounts to -2.1% of GDP, sandwiched between -4.0% on one extreme and -0.5% on the other. And every one of those recessions has seen an increase in the rate of unemployment, ranging between +0.5% and +2.6%. Interestingly, the lower the starting rate of unemployment, the larger the increase during the ensuing recession.
Given that context, the Q1-Q2 recession appears mild. It’s about as mild as one could hope for in fact.
That makes sense to me. At the start of this year, the shape of the yield curve was consistent with a reasonably benign outlook. The yield on the 10-year treasury stood 0.8% higher than that on the 2-year treasury, giving rise to an attractive, upward-sloping yield curve.
Unfortunately, the shape of the yield curve is much less appealing today. The 10-year treasury yield is stooped -0.4% lower than the 2-year rate to form an ‘inversion’ or downward-sloping curve. Deep inversions like the one we see today are very often a warning that something, somewhere is going wrong.
It seems to me that participants in the bond market are increasingly anxious. And not about those two quarters just gone.