The stories we tell
By Scott Sumner
The field of economics is a set of stories that we tell to better understand the economy. I thought of this when reading a new post by Matthew Klein:
If I had to pick one chart to tell the story of the U.S. economy since the end of WWII, it would be this:
Indeed Klein views the post-2006 slowdown in RGDP per capita as being so important that his Substack is entitled ‘The Overshoot”, an indication that we need to work on reversing the recent undershoot.
Of course literary theorists know that stories can have many interpretations. What is Kafka’s Metamorphosis actually about?
If I were to tell a story about this a graph, the year 2006 would be of no importance at all. I would begin my story by focusing on productivity, not output per capita. Doesn’t real GDP per capita measure productivity? No, it does not.
So here’s my story. Between 1900 and 1973, America’s engineers and inventors made extraordinary strides that completely transformed a wide range of our industries. Here’s a 1903 airplane above a 1968 airplane:
I won’t bother showing you a 2021 airplane, as it looks similar to a 1968 airplane. Its interior electronics and engines are better, but the percentage gain is tiny compared to the gains during 1903-1968. Great strides were made in many other industries as well, including autos, home appliances, lighting, and infrastructure such as indoor plumbing. Life expectancy soared much higher.
After 1973, America’s engineers and inventors made extraordinary gains in one industry—computers. Productivity growth slowed sharply, with one exception. There was an upward blip in productivity growth during 1995-2004, which might have been real or might reflect a mismeasurement of the impact of PCs on the US economy. It’s a “matter of opinion”. In any case, that brief surge ended in 2004.
So if 1973 was the turning point when productivity growth slowed sharply, why does Klein’s graph make it look like 2006 was the turning point? What’s my “story”?
It turns out that various demographic changes disguised the productivity slowdown for several decades, at least when examined from the perspective of RGDP/person. After 1973, the share of the population that worked rose for several decades, as (non-working) children became a smaller share of our population and as more women entered the labor force. This growth in the labor force roughly offset the decline in output per worker, keeping RGDP/person growing at a relatively steady rate even as growth in RGDP per worker was slowing sharply.
By the mid-1990s, this demographic transition had mostly played out and RGDP/person growth would have slowed sharply if not for the nine year PC boomlet in productivity (or at least measured productivity.) In 2004, that boomlet ended and productivity growth went back to the slow rate of 1973-1995, where it has stayed ever since. Now there were no longer any special factors to disguise the productivity slowdown, and hence RGDP/person slowed sharply. And as boomers began retiring, the demographic dividend started to move in the other direction, a perfect storm of factors slowing RGDP/person.
In my story, there is nothing special about 2006. You had productivity growth slow after 1973 because our tech people were no longer able to radically transform almost all our industries, instead radically transforming only the computer industry. Of course it’s possible that at some point computers become so powerful that robots begin transforming almost all our industries, leading to a renewal of fast productivity growth. But we aren’t there yet. (It’s also possible that AI will destroy life on Earth.)
Every story has policy implications. If you believe something dramatic happened after 2006, you might focus on policies that could prevent a repeat of the Great Recession. I also want to focus on policies that prevent a repeat of the Great Recession, but not because I think the Great Recession had a significant impact on long run growth in RGDP/person. I see no evidence for that claim. The 1995-2004 boomlet in productivity ended even before the Great Recession of 2008. Rather, I want to prevent another Great Recession because big recessions are very bad.