Slow growth? Since when?

Since 2010?

The economy has been disappointing on the low side consistently for several months now and that makes me and other people worry that maybe this slow growth will linger longer than we now think -- Alan Blinder, 2011

 

 Since 2001?

This slow growth is not some new phenomenon, but rather the way it has been for 15 years and counting. -- Neil Irwin, 2016

 

Since 1974?

[G]rowth in US living standards slowed after 1973 -- Scott Sumner, 2010

For the entire post-WWII period?

On our end, we looked at seven decades of GDP growth data to shed some light on the trend growth potential for the U.S. economy. After stripping out recessions, which can be caused to externalities that may not reflect productive capacity (such as oil shocks or a derivatives-based financial crisis) we calculated the 10-year average GDP growth rate. It shows a clear, steady decline and a close relationship with productivity. -- Raul Elizalde, 2019 

 

 

For a long time, I accepted 1973 as the end of the good years. But making graphs for the blog, I noticed that they sometimes suggested 1966 as the end of the good years. It opened my eyes and made me cautious.

Only recently did I find Raul Elizalde's article at Forbes. Of course I was fascinated. But I couldn't duplicate his graph. By email, I asked him about it. By email, Elizalde explained his method:

I downloaded quarterly real GDP percentage rates from FRED (A191RL1Q225SBEA). Then I eliminated all the quarters with a negative number and the following positive. My goal was to measure GDP growth during “normal” times in order to reflect the long-term growth trend of the U.S. In order to do this I took out all the negative quarters (which are outliers in the sense that the long-term trend is positive, and often due to shocks) and also took out the bounces, defined as the first positive quarter after one or more negative quarters (because bounces are often outsized when the shock disappears and the pent-up demand produce a number larger than the long-term trend). That’s why the lines and mountain in the graph seem to plateau at certain times (because of the missing negative or bounce quarters). The resulting data points are what you see.

It works:

The graph starts at 1957 because the data points are ten-year averages.
Data used for the graph goes back to 1947.

I knew about the "bounces", where recovery from recession brings an unusually big increase in economic growth. I remember Krugman pointing out the bounce of the 1930s after the depression bottomed out. I remember American Thinker pointing out the bounce of 1984 after the double-dip recession bottomed out.

I knew about the bounces, but it never occurred to me to allow for the bounce effect when figuring the trend of long-term growth. 

Impressed by Elizalde's graph, I was.