Paul Krugman is continually taken to task for predicting in 1998 that the Internet would be no more important economically than the fax machine by 2005. Culturally, of course, this new medium has been a watershed event. But he had a point on some level: the Internet–and computers, more broadly–still disappoint from a productivity perspective. Either that or all conventional measurements are insufficient to gauge this new machine. At his Financial Times blog, Andrew McAfee, co-author with Erik Brynjolfsson of 2014’s wonderful The Second Machine Age, wonders about the confusing state of contemporary economics. An excerpt:

The economy’s behaviour is puzzling these days. No matter what you think is going on, there are some facts — important ones — that don’t fit your theory well at all, and/or some important things left unexplained.

For example, if you believe that technological progress is reshaping the economy (as Erik and I do) then you’ve got to explain why productivity growth is so low. As Larry Summers pointed out on the first panel, strong labour productivity growth is the first thing you’d expect to see if tech progress really were taking off and reshaping the economy, disrupting industries, hollowing out the middle class, and so on. So why has it been so weak for the past 10 years? Is it because of mismeasurement? William Baumol’s “Cost Disease” (the idea that all the job growth has come in manual, low-productivity sectors)? Or is it that recent tech progress is in fact economically unimpressive, as Robert Gordon and others believe?

If you believe that tech progress has not been that significant, however, you’ve got to explain why labor’s share of income is declining around the world.•