The pie fallacies

Too many people see the world as zero sum. They think of wealth and prosperity — “the pie” — as fixed, and so assume that if one person gets more, someone else will end up with less.

The world isn’t zero sum. (In 2004, Paul Graham called this sort of thinking the Pie Fallacy.) But way too many of the people who delight in pointing this out then go on to assume, explicitly or implicitly, that because the size of the pie isn’t fixed, the distribution doesn’t matter, or at least isn’t worth focusing on. This is just as absurd as the idea that the pie is fixed. I’ve tried to explain why before, but many pie-growers seem immune to the logic of it.

However, we have empirical evidence, thanks to Raj Chetty and colleagues. Given ballooning inequality over the past 40 years, they asked, essentially, how would the number of Americans who end up better off than their parents be different had the pie been divided as fairly as it was mid-century, vs. if the pie had grown as fast as it had at its modern peak?

Obviously, both matter a lot. But they estimate that how the pie is divided mattered more:

The researchers ran a clever simulation recreating the last several decades with the same G.D.P. growth but without the post-1970 rise in inequality. When they did, the share of 1980 babies who grew up to out-earn their parents jumped to 80 percent, from 50 percent. The rise was considerably smaller (to 62 percent) in the simulation that kept inequality constant but imagined that growth returned to its old, faster path.

The fact that the world isn’t zero sum isn’t a license to ignore the distribution.

At this point, the more ideological of the pie-growers argue that more equitable distribution will mean less growth. That is wrong, generally speaking. But it’s at least less annoying than the outright insistence that, because the pie isn’t fixed, questions of distribution just don’t matter.