In his important recent essay “The Return of Nature”, Jesse Ausubel of Rockefeller University, highlights a wonderful phenomenon: we humans have been giving land back to nature, since we no longer need it for our purposes. Even though the world’s population continues to increase, we have in all likelihood past the peak for farmland; the number of acres under cultivation globally has been slowly declining, and will continue to do so. Forest loss also appears to have been reversed in recent years, and our use of natural resources such aluminium, copper and timber is decreasing over time.
The article’s subtitle, “How technology liberates the environment”, identifies why this is. Software, sensors, data, autonomous machines and all the other digital tools of the second machine age allow us to use a lot fewer atoms throughout the economy. Precision agriculture enables great crop yields with much less water and fertilizer. Cutting-edge design software envisions buildings that are lighter and more energy efficient than any before. Robot-heavy warehouses can pack goods very tightly, and so be smaller. Autonomous cars, when (not if) they come, will mean fewer vehicles in total and fewer parking garages in cities. Drones will replace delivery trucks. And so on.
The pervasiveness of this process, which Mr. Ausubel labels “de-materialization,” might well be part of the reason that business investment has been so sluggish, even in the US, where profits and overall growth have been relatively robust. Why build a new factory, after all, if a few new computer-controlled machine tools and some scheduling software will allow you to boost output enough from existing ones? And why build a new data centre to run that software when you can just put it all in the cloud?
One strong piece of evidence in support of the de-materialization hypothesis is the fact that while overall business investment has been in the doldrums for a while, U.S. corporate investment in software is near an all-time high. It was at 1.824 percent of gross domestic product in the second quarter of 2015, very near its previous peak of 1.837 percent in Q1 2001, when a combination of Y2K paranoia and dot-com hysteria caused a sharp and unprecedented spike in software spending.
There’s no spike just this time — just a pretty steady increase in the software intensity of the economy. I bet it will continue, and continue to drive de-materialization. If your business is based on selling lots of atoms, this might not be happy news for you.
This blog first appeared in FT.com on Sept. 29, 2015 here.