The Economist article points out:

In 2005 the American Society of Civil Engineers estimated that $1.6 trillion was needed over five years to bring just the existing infrastructure into good repair. This does not account for future needs. By 2020 freight volumes are projected to be 70% greater than in 1998. By 2050 America’s population is expected to reach 420m, 50% more than in 2000. Much of this growth will take place in metropolitan areas, where the infrastructure is already run down.

Recently, I was having a conversation with a friend where he asserted that India would be the largest produce exporter in the world. Because of lack of infrastructure (roads, trains), they can’t even get their produce from the country to their cities before it rots. In the US, our civic infrastructure has given us an economic edge for a good part of the last century. Going forward, we’re being outspent on infrastructure by most European nations and by China.

I think it’s interesting to analogize world economies to a technology stack. It’s great to move further up the stack to more powerful levels of abstraction (high-level languages vs. knowledge workers not involved in manufacturing), but if the bottom part of your stack suffers from bit-rot or someone innovates a cheaper stack, it’s hard to compete.