The link between rational individual behavior and collectively desirable outcomes is extremely tenuous.

For the three decades after WWII, incomes grew at about 3 percent a year for people up and down the income ladder, but since then most income growth has occurred among the top quintile. And among that group, most of the income growth has occurred among the top 5 percent. The pattern repeats itself all the way up. Most of the growth among the top 5 percent has been among the top 1 percent, and most of the growth among that group has been among the top one-tenth of one percent.

Questions like, "Is my suit OK?", or "Is my job performance satisfactory?", are impossible to think about in the absence of a suitable frame of reference. For an interview suit to serve its purpose, it must make you look good relative to other candidates for the job you want. For your job performance to be satisfactory, it must compare favorably with the performance of others who want the same promotion you do. As Charles Darwin saw clearly, much of life is graded on the curve, and conventional economic models completely ignore that fact.

A good school is a relative concept, and the better schools are located in more expensive neighborhoods. But when everyone bids more for a house in a better school district, they succeed only in bidding up the prices of those houses. As before, 50 percent of all children will attend schools in the bottom half of the school quality distribution. As in the familiar stadium metaphor, all stand, hoping to get a better view, only to discover that no one sees better than if all had remained seated.

When the rich build bigger, they shift the frame of reference that shapes the demands of the near rich, who travel in the same social circles. Perhaps it's now the custom in those circles to host your daughter's wedding reception at home rather than in a hotel or country club. So the near rich feel they too need a house with a ballroom. And when they build bigger, they shift the frame of reference for the group just below them, and so on, all the way down.

We're in a classic demand-shortfall recession. There aren't enough jobs because total spending is too low. Consumers won't lead the way because they're busy paying down debt and are fearful they'll lose their jobs, if they haven't already. Businesses, which are currently sitting on mountains of cash, won't spend either, because they already have sufficient capacity to produce more than people are willing to buy.

In short both the things we feel we need and the things available for us to buy depend largely—beyond some point, almost entirely—on the things that others choose to buy.