Measuring the Speed of Consuming Slack
How fast is slack consumed? How can we measure the speed at which slack is consumed?
We can use capacity utilization and unemployment to set up a simple measure for consuming slack by dividing capacity utilization by the unemployment rate in terms of percentage change year-over-year. (Thanks to Alan Thomson at Terra Firma Financial who created this graph in response to a graph in my previous post.)
When the plot is positive, slack is being consumed on a year-over-year basis. Capacity utilization is rising as unemployment is falling. The height of the plot gives us an estimate of how fast slack is being consumed in terms of both labor and capital. The graph shows there is a general limit around 20% to how fast slack can be consumed.
We can see that since the beginning of 2014, the consumption of slack has accelerated… and reached a level of 25% in June, which in the past signaled a peak. The normal pattern would now be that the consumption of slack will proceed at a slower pace. We shall see.
In theory, as long as the plot stays positive, slack is being consumed. However, the pattern in past business cycles is that towards the end of a business cycle, the economy shows a downward trend in its ability to consume slack. Eventually, there is a recession and slack widens again (plot goes negative).
The speed of consuming slack is important to the Fed… and will determine how and when Janet Yellen raises nominal rates. In the past, the Fed tended to raise its Funds rate when slack was being consumed at a slower pace near the end of a business cycle. Yet Janet Yellen seems to be basing her decision on how fast slack is consumed. The implication is that if slack is consumed slower from here on out (because the plot peaked), the odds of holding the effective Fed rate near the ZLB increase.
Here is another version of the graph using quarterly data. (quarter-to-quarter % change) (link)
2nd quarter 2014 rose quite a bit from 1st quarter at a pace of over 7%. A rise like that has not been common since the early 80’s. One might expect the plot to adjust back down fairly quickly. And this is what we will see. With just one month of capacity utilization data needed to complete the 3rd quarter, the plot in this quarterly graph will certainly decline to below 2%.
If you had used these graphs back in 2007, you would have suspected that a recession was forming months ahead of time. So these graphs of “slack consumption” are good to put in your toolkit. (They are posted and automatically updated at my effective demand blog.)
That doesn’t show recession, that shows the end phase of recovery into the mist between boom and bust. The mid-70’s, 80’s and 90’s all had it. Early 70’s and 00’s didn’t(because they weren’t real recessions).