In my last newsletter I wrote about the possible economic implications of AI, arguing that while they may be big, history suggests they will take longer to materialize than many people expect.
Since life is what it is, several people have come back to me, citing a prediction I made in 1998 that the growth of the Internet would soon slow, and that by 2005 or so it will become clear that the impact of the Internet on the economy was no greater than faxes. I said it actually, in a throwaway piece I wrote for The Red Herring magazine a piece I still don’t remember writing, but I think I was trying to be provocative.
Of course I was wrong about the internet going dead and I admitted it. So it went. Show me an economist who claims they’ve never made a wrong prediction, and I’ll show you someone who is dishonest or unwilling to take intellectual risks.
But how wrong was I, really, about the economic impact of the Internet? Or, since this shouldn’t be about me, have the last few decades generally justified the visionaries who claimed that information technology would change everything? Or have they justified techno-skeptics like economist Robert Gordon, who argued in a 2016 book that late 20th and early 21st century innovations were far less fundamental than those between 1870 and 1940?
Well, by the numbers, the skeptics have won the argument, hands down.
In that latest newsletter, we looked at decadal labor productivity growth rates, which suggested that information technology did indeed drive a surge in economic growth between the mid-1990s and mid-2000s, but which was relatively modest and short-lived. Today I would like to take a slightly different approach.
The Bureau of Labor Statistics produces historical estimates, dating back to 1948, of both labor productivity and total factor productivity, an estimate of the productivity of All inputs, including capital and labor, which is widely used by economists as a measure of technological progress. A truly fundamental technological innovation should cause sustained growth in both of these measures, especially total factor productivity.
So let’s look at those 25-year rates of change in these two measures. I choose 25 years partly because it is roughly a generation and partly because I made my prediction wrong 25 years ago. So here is the growth in labor productivity over the last 25 years following each date on the horizontal axis:
And here is the growth in total factor productivity, measured in the same way:
See the great productivity boom that followed the rise of the Internet? Me niether.
True, there are a couple of excuses you can come up with for the gap between the numbers and the hype. One is to say that the Internet has done really great things for the economy, but that they’ve been offset by negative factors, a deterioration in the work ethic, or the mysterious decline in construction productivity, or something like that.
The other is that official economic growth numbers fail to capture many unseen gains. I get a lot of pleasure from being able to stream live musical performances on YouTube; as far as I know, those benefits are not counted in gross domestic product. Indeed, official economic growth surely understates the true progress of the human condition. But this has been true for a long time.
Before the Internet, official economic data did not directly capture the benefits of, for example, declines in infant mortality due to improved sanitation or the dramatic improvement in air quality after 1970. These invisible gains are greater now compared to the past? It’s doubtful.
Perhaps the key point is that no one is arguing that the Internet has been useless; certainly, it contributed to economic growth. The argument instead is that its advantages were not exceptionally great compared to those of earlier and less glamorous technologies. For example, around 1920, only one in five American households had a washing machine; by 1970 nearly everyone had one or had access to it. Don’t you think it made a big difference? Are you sure it made less difference than widespread broadband access?
Because the fact is, while information movement is important, we still live in a material world: Most of what we consume are physical things or person services, which haven’t been drastically impacted by the internet.
Now, it’s entirely possible that AI or, at any rate, things called AI, whether or not they guarantee that designation is going to be really, really, a big thing. But one thing we should have learned from the history of information technology is that things that seem exceptionally fascinating don’t have to be particularly useful, and vice versa.
Shocked that I still remember it.
#Opinion #internet #cheap #disappointment