Thinking a bit more about Baumol’s Cost Disease it’s perhaps useful to think of this as a rectangle with four options rather than just two. You have sectors that feature above-average productivity increases and you have sectors that feature below-average productivity increases. Then you have sectors where people increase consumption over time and sectors where they don’t.
As illustrative examples, I’ve picked smartphones, health care, servants, and timber to fill out the quadrant. In the logging sector, over time machines let us get better and better at cutting down more trees. To an extent this leads to higher real per capita consumption of lumber, but not nearly to an extent that keeps pace with productivity gains. The result is that the share of the population engaged in logging steadily falls over time. In the opposite corner with health care you get the true cost disease. Productivity in this sector is lagging, but an extremely high share of people’s marginal dollar ends up going to health care. Consequently, health care employment skyrockets.
These are the two things you hear the most about. In the bottom left hand corner we wonder “where have all the good jobs gone” as mechanization and outsourcing let us get all the stuff we need with fewer workers. In the top right we wonder “how can we afford this” as lack of productivity gains leads to steady increases in the relative cost of some items.
But the other quadrants are also worth thinking about. With the servants, we might ask where all the bad jobs have gone. The same factors that have made health care less affordable have also made butlers and governesses less affordable. But what we do is get by with fewer butlers. Even in an era of sky-high income inequality, employing people in this way just doesn’t make nearly as much sense for people of means as it used to. Unlike the rich of the gilded age, today’s affluent can spend their money on luxury cars and air travel to vacation homes rather than butlers. Then in the top-left corner you get your growth sectors like smartphones today. Over the past ten years, the productivity of our mobile phone sector has skyrocketed (remember to try to apply the quality adjustment in your head) and as a result people spend way more on mobile phones and related goods (data plans, apps, etc.) than they did ten years ago. A phone as good as the Samsung Galaxy S III would have been prohibitively expensive to make in 2010, to the extent that almost nobody would want to buy one. But as more and more people become able to afford increasingly awesome phones, we have more people than ever working in the app economy.