Thomas Piketty's Capital in the Twenty-First Century has taken to U.S. by storm. It's been denounced as warmed-over Marxism, praised for its distinctly un-Marxist analytical coherence, denounced as insidious anti-Marxism, praised for its magisterial use of masses of data, denounced for sloppy and tendentious distortions of data, ... A surprisingly small fraction of the people discussing it seem to have actually read it.
I was lucky enough to pick up what seemed to be one of the last three copies for sale in Berkeley on a recent trip, so here's an attempt to distill the main argument briefly. I'll focus on some key economic facts that Piketty takes from data rather than from analysis, since understanding these facts is likely to be the key to understanding where we're headed and how we might change directions.
The first part of the book, told with numbing repetitiveness, does not directly concern inequality at all. It describes changes in the ratio of total capital to annual production. Piketty makes an overwhelming, unchallenged case that this wealth/income ratio has grown a lot in the last few decades, and is approaching values (say ~6) typical before WWI. The period in which some of us grew up, post WWII, was an anomalous result of the shocks of the wars, the depression, and the Russian revolution. These destroyed a lot of capital and constrained the behavior of capitalists.
It's a little hard initially for a reader to see why that growth of capital should be taken as a bad thing. Isn't it better to have more of our work done with enhanced productivity due to accumulated knowledge and machinery? Why wouldn't we want wealth to grow indefinitely?
Piketty's answer shows up next. Throughout the period for which we have records, wealth has always been much more unevenly distributed than income from labor. That's true even now in the U.S., despite our anomalously large wage inequality. Almost everywhere, the bottom half of the population owns almost nothing. Thus unless the fraction of capital in private hands shrinks or the rate of return (ROI) shrinks, a higher fraction of income will come from capital ownership, and this will be much more unequal than income from work.
Will the growing wealth be held privately? The data show that total public wealth (assets-debts) has converged to around zero, less than +/- a year's income, in modern market economies. So unless we somehow understand and change that general fact, the growing wealth will be held privately.
Will the ROI drop much as the wealth to income ratio grows? Marx thought it would. Classical economics predicts that it will, since when capital is plentiful people shouldn't have to promise much ROI to get some. Many economists have been fond of a simplified relation for that drop, in which the returns on capital stay at a fixed fraction of total income even as the wealth/income ratio increases. Piketty shows that in fact ROI tends to stay up near 5%/year regardless of such predictions. He discusses possible reasons briefly and without making any conclusions. The general impression, however, is that the radical inequality in power that goes with radical wealth inequality coupled with large wealth/income ratios allows a variety of exercises of political and social power to counteract the expected drop in ROI.
Thus putting together these empirical ingredients implies that income inequality will grow as the fraction of income coming from capital grows. Europe shows that pattern. In the U.S., extreme inequality in payment for work currently obscures that pattern except in the highest income brackets.
One stage of the argument is not purely empirical. Piketty refers to analytical work showing how a tail of very large fortunes tends to be generated by normal economic processes. As the quantity (r-g), annual ROI minus annual fractional growth of production, gets large, two things happen. Private wealth gains more rapidly on production, and, within realistically simplified models, the distribution of private wealth tends toward a more unequal steady-state. (For the last model-based argument, Piketty refers to some fairly obscure on-line notes. It is not hard to reproduce the result for very simple models, however.)
Piketty make an additional argument, based on common sense bolstered with some nice data, that ROI is largest for the biggest fortunes. That's just what classical theory would predict, since the biggest fortunes can withstand the highest short-term risk and pay for the most sophisticated information and analysis. They can also obtain the most favorable laws, regulations, and extra-legal maneuvers. Of course, this extra factor just worsens the trend toward greater inequality.
Finally, Piketty concludes with some proposals. The key one is to have some modest international wealth tax. This would be a more systematic way of getting at what Teddy Roosevelt intended with our estate tax- preventing the growth of a hereditary aristocracy.
Several important comments have been published on Piketty. I'll comment on a few of them.
Many people have noted that Piketty's remedy, a wealth tax, sounds reasonable and moderate but might be as hard to attain as a revolution. They have an obvious point, but the alternatives aren't so clear either.
DeLong has noted that the "r" used by Piketty is a short-hand for a collection of r's. One gives bare returns, and is important for directly calculating how much income is coming from a given amount of wealth. Another subtracts from that taxes and typical consumption expected for the rich, to get the quantity that enters into the growth of wealth and the form of the tail of the steady-state distribution. I doubt that Piketty would dispute this obvious point, implicit in some of his arguments, but it is true that most of his presentation mushes this up. (See also.)
Giles in the Financial Times has claimed a variety of data- handling errors, of which one would be consequential. Giles claims that wealth inequality hasn't grown recently in Britain the way that Piketty claims. Piketty has responded that in fact he has used more reliable and consistent data sources than Giles, and that his results are backed by independent analyses. Notice, however, that this dispute concerns only a detail on a detail- whether wealth inequality has already started to increase in one country. The central point, that wealth is already extremely unequal and that the importance of wealth is growing relative to wage income, is entirely untouched by this dispute, as others have noted.
I'm a bit worried about the relatively scanty attention Piketty pays to the growth rate (he guesses it will run about 1.5%) and the natural environment. Given our current effects on the climate, on water supplies, and so forth, it's by no means clear that a full accounting of our current growth would give a positive number. In other words, we're creating an environmental deficit, negative wealth, at a growing rate. Piketty is aware of these issues, which inform his choice of a fairly low expected growth rate, but I wonder if our uncertainty of the magnitude of them might swamp all other issues.
Analytical follow-ups to Piketty's work might focus on several key facts that appear empirically in Piketty's work.
Why does the net public wealth stay near zero, except in post-revolutionary societies, which have not been great economic successes?
Why is private wealth always so unequally distributed? (The simple models tend to describe the high-end tail more than the bulk of the population.)
How does ROI stay high even when there's lots of capital available?
What becomes of all this if environmental factors give a net negative growth?
The answers may in part lie in some fairly depressing behavioral economics. I suspect that the arguments of Robert Frank about how competitive arms races drive luxury consumption and suck resources into that zero-sum or negative-sum sector are relevant. It is interesting that Frank's suggested remedies (taxing consumption, not income) would probably worsen the problems Piketty addresses. Some sort of unified approach to these problems is needed.