This graph shows US total factor productivity from 1947 until 2020. The last 20 years are the flattest part of the graph.
It’s a time that has seen the rise of cloud computing, mobile apps and agile thinking. None of which appear to have moved the productivity dial.
Until 1970 there is a sharp rise. Then, around the time business computers became commonplace, the productivity rise tapers off. We see a burst in the late 1990s and the start of the 20020s, then it flattens.
Austin Vernon writes at length about software’s negligible impact on productivity. It’s worth a read if you can divert yourself from your own productivity for five minutes.
Software a management technology
He gets into economic concepts. His key point is that software is not a general purpose technology but a management technology.
General purpose technologies drive productivity growth in ways that are easy to understand. That term covers technologies like electricity, railways and steam engines.
Software has more in common with assembly lines.
These have a slower, plodding effect on productivity. In part that’s because they deal with high levels of complexity. In comparison building railway networks is straightforward and easy to repeat.
He says their benefits don’t show up in society at large until the companies using them have a significant market share.
“The current status quo means we don’t get productivity growth until these software-driven companies become behemoths.
Amazon was founded in 1994, almost thirty years ago. In 2020, it was still less than 10 percent of total retail sales 1. Is it any wonder that we haven’t seen robust productivity gains? Amazon is still mapping and digitising processes at prodigious rates.
In other words, Vernon thinks we’re just getting started with software productivity gains. If he is right, then we can expect more productivity as the world’s technology giants grow larger. Of course, that brings another set of problems.
- I think Vernon is referring to US sales here. If you know otherwise please help me correct this footnote. ↩︎