Is Internet access like electricity? (Is Internet access a general purpose technology?)

In the leadup to the passage of the infrastrucutre bill, a WSJ article cited two studies “showing that expanding high-quality internet access could have significant economic benefits.” I did a little digging into the first study, from the Aspen Economic Strategy Group. It found that universal high-speed internet service would raise labor productivity by 1.1%, using the results of a survey.

Surveys? Eh.

Here’s the dirty little secret of broadband. It’s not the game-changer that many think it is. In the language of economics, it’s not a general-purpose technology or GPT, which would mean it would impact productivity. Electricity is a GPT and it has been shown time and again to impact productivity. Those trends are easy to find. As for broadband, it’s difficult to find a big impact on economic growth.

People don’t just buy Internet access, they typically adopt a bundle of technologies, a smartphone or a computer alongside an Internet connection. A meta-study of economic research, which is an empirical compilation of all studies to determine trends, helps to tease out the various impacts of the bundle. They found that broadband access doesn’t have an impact on economic growth. Rather, its adoption and use of cell phones and computers that drive new growth.

All the evidence suggests that broadband policy should contine to shift focus. The conversation is still dominated by the avilaibility gap and putting money into the ground even there isn’t good data on its size. There is a lot of money in that bill and very little of it is decicated to new pilot programs.

Extra: Austin Vernon has a great post on computers as GPTs.

First published Aug 1, 2022