On the sustainability of low unemployment

JW Mason:

But let’s suppose that today’s unemployment rate of 3.6 percent is sustainable—which it certainly seems to be, given that it is, in fact, being sustained.

Mark Thoma, July 2017 at The Fiscal Times:
“The inevitability of another recession is evident in a graph of the unemployment rate. Notice that, before 1970, it was common for the unemployment rate to reach a low point and then hover around that point for several years. For example, the unemployment rate was around 4 percent for an extended period in both the mid to late 1950s and 1960s. But since 1970 the unemployment rate has behaved differently. Instead of reaching a low point and then leveling off for a period of time, it has tended to “bounce” off the low point and almost immediately begin rising again.”

I put together a graph to show unemployment hovering around low points, or not:

Graph #1

The data begins at a low in the late 1940s. I put a red line thru it to suggest unemployment hovering around that level.

The next four lows each begin after a sharp decline in unemployment. I put the red line where the sharp decline stops. In each case unemployment does seem to hover near that level for at least several months, confirming Mark Thoma's observation.

Regarding the highest of these four red lines: Perhaps I should have stopped that line before the unemployment spike of the 1960 recession. Then at the end of that spike I'd need another flat red line where the blue seems to hover at a higher low, before falling to the last of those red lines. This correction, if I made it, would also support Thoma's view, unemployment hovering around the lows in the years before 1970.

After 1970 it is easy to find "bounce point" lows that separate the fall of unemployment from the rise. I marked them with black arrows (and in one case, a double arrow).

Thoma's right: After 1970, unemployment behaves differently. Looking at this graph, it is easy to doubt JW Mason's assertion that the current low in unemployment is "being sustained".


After 1970, unemployment behaves differently. The spikes on the graph in the early years show unemployment rising smoothly to a peak, and then falling smoothly. The fall is never less than half the size of the rise.

After 1970, the declines are more jiggy than smooth. After the 1975 recession, the smooth fall of the spike is only about one-quarter the size of the rise. The smooth rise and fall associated with the 1982 recession are about equal in size, but that's the last time that occurs. The fall after the 1991 recession appears to be smooth and about the size of the rise; but it isn't smooth: Click the graph to see it bigger, and you'll see what I mean.


Antonio Fatas, 12 March 2019 in The 2020 (US) Recession:
“This post is based on a research note I wrote asking whether low unemployment is sustainable. The answer is a clear no for the US. Low level of unemployment are good predictors of the tail risk event of a recession, a sharp increase in unemployment rates. These dynamics are related to the build up of financial and macroeconomic imbalances. If this pattern is to repeated, and given the current level of unemployment rate, a US recession must be around the corner. For details on the analysis, the research note including additional results is available on my web site: Fatas (2019).”

Antonio Fatas finds the same pattern in unemployment that Mark Thoma finds. I want to say Fatas seems more certain than Thoma that the pattern tells us recession is in the cards; but Thoma does say "the inevitability of another recession is evident" in the graph.

Having looked at the graph myself, I have to say that Fatas and Thoma together offer a perspective that challenges Mason's optimism and underscores the feebleness of his assertion that

"today’s unemployment rate of 3.6 percent ... certainly seems to be [sustainable], given that it is, in fact, being sustained."

But let me take another look. For Mason also says: “Every month that the US records an unemployment rate below 4 percent suggests that these low unemployment rates are indeed sustainable…”

It's not just the 3.6% level (sustained for two months now) to which Mason refers. It's anything less than four percent. This should give us a different graph. Let's see:

Graph #2

Again, all the sustainability seems to occur before 1970.

What stands out here, to my eye, is how seldom unemployment goes below four percent. That, and the scantiness of unemployment below 4% in the current period. I'd need my rose-color glasses to see unemployment "being sustained" at present.

Yeah, I thought there might be something there, but I don't see it.