I haven't had time to write a review of The Second Machine Age, the new book by +Andrew McAfee
and +Erik Brynjolfsson
, but I just want to comment real quick on something that's stuck in my mind ever since I read the book, but it seems to have been completely skipped over in all other discussion I've seen anywhere about the book, perhaps because it's too "mathy" for most people to discuss. But seems hugely significant.
And that is, that labor income distribution in the industrial age (or First Machine Age as they call it) follows a normal (or Gaussian) distribution -- a bell curve, while in the information age (or Second Machine Age, as they call it), it follows a power law curve.
Actually in the industrial age it follows a log-normal distribution -- I know there are smart people here and someone would correct me if I didn't get that detail correct, but it doesn't matter for the basic point. A log-normal distribution is a bell curve that leans to one side (because there's a logarithm function figuring into its shape).
Anyway, let's use an example. Let's say the average wage in some industry is $10/hour. Nice round number, maybe enough to live on but not in Silicon Valley. With a normal distribution, this would mean most people earn $10, with fewer and fewer people earning more or less the farther you go from $10. With a power-law distribution, however, the majority of people would earn $0.01/hour, with a handful of people earning hundreds of millions or even billions -- such high numbers that the average still comes out to $10. So in both scenarios, the average income is $10, but the distribution is completely different.
What McAfee and Brynjolfsson are saying is that the industrial revolution naturally produces the normal distribution (ok, technically log-normal), and the information revolution (or Second Machine Age) naturally produces the power-law distribution. They say even for jobs where you wouldn't expect it -- service jobs where you'd expect everyone to be paid about the same -- you'll still see it because the effect propagates through the economy, through the networks of monetary exchange.