WE ARE ALL NEOCONOMISTS NOW:

An easy way to remove the impediment to growth is to move toward a consumption tax by allowing the full and immediate deductibility of capital investment.

The argument rests on two points. First, consumption taxes are better for economic growth than are income taxes. Second, allowing full expensing (immediate deductibility) of investment turns the current tax system into a consumption tax.

Consumption taxes are better for economic growth because they create stronger incentives to save and invest than do income taxes. Under an income tax, a person who consumes what he earns immediately is taxed once, specifically on the earnings that he receives in that year. If instead he invests what he earns, the interest on that investment, which is compensation for deferring consumption, is also taxed. This pushes him toward consuming more now and saving less.

The reduced incentive to save that results from taxing returns drives up interest rates and retards investment. Incentives to invest would be improved if the returns were untaxed. By contrast, a consumption tax does not tax the returns to investment. It taxes only once, at the time that actual consumption occurs. Moving to a consumption tax eliminates the tax on returns to investment and improves investment incentives.

Allowing investment expenses to be fully and immediately deductible turns an income tax into a consumption tax, but the logic is subtle. All of an economy's output is used to produce either current consumption or investment goods. If all income, which must equal output, is taxed, then both consumption and investment are taxed. But if we tax only the part of output that is not investment by allowing investment expenditures to be deductible, all that remains is consumption so only consumption is taxed.

So what's the problem?

Quite a few things, but this to start with: There's a persistent tension between the limits of the data he presents and the grandiosity of the conclusions he draws. At times this borders on schizophrenia. In introducing each set of data, he's all caution and modesty, as he should be, because measurement problems arise at every stage. Almost in the next paragraph, he states a conclusion that goes beyond what the data would support even if it were unimpeachable.

This tendency is apparent all through the book, but most marked at the end, when he sums up his findings about "the central contradiction of capitalism":

The inequality r>g [the rate of return on capital is greater than the rate of economic growth] implies that wealth accumulated in the past grows more rapidly than output and wages. This inequality expresses a fundamental logical contradiction. The entrepreneur inevitably tends to become a rentier, more and more dominant over those who own nothing but their labor. Once constituted, capital reproduces itself faster than output increases. The past devours the future. The consequences for the long-term dynamics of the wealth distribution are potentially terrifying ...

Every claim in that dramatic summing up is either unsupported or contradicted by Piketty's own data and analysis. (I'm not counting the unintelligible. The past devours the future?)

As he explains elsewhere, r>g isn't enough by itself to trigger the dynamic he describes. If capital grows faster than the economy, inequality will indeed tend to increase because ownership of capital is concentrated -- though much less so than in the past. But capital will outpace the economy only if owners of capital save a sufficiently large part of the income they derive from it. (Suppose they save none of it: Their wealth won't grow at all.)

You might say this misses the point. Wolf offers this clarification: Piketty "argues that the ratio of capital to income will rise without limit so long as the rate of return is significantly higher than the economy's rate of growth. This, he holds, has normally been the case." That's better: The gap between r and g has to be "significant." The bigger the gap, the more likely it is that saving will build capital faster than output rises -- and Piketty does show that the gap usually has been big.

The trouble is, he also shows that capital-to-output ratios in Britain and France in the 18th and 19th centuries, when r exceeded g by very wide margins, were stable, not rising inexorably. The same was true of the share of national income paid to owners of capital. In Britain, the capitalists' share of income was about the same in 1910 as it had been in 1770, according to Piketty's numbers. In France, it was less in 1900 than it had been in 1820.

What about the 21st century? Perhaps the capitalists' share will rise inexorably in future -- and that's what matters.

Perhaps it will, but Piketty advances reasons to doubt this too. He expects r to be a bit lower and g a bit higher than their respective historical averages. There are many other factors to consider, as he says, but on his own analysis the chances are good that the future gap between return on capital and growth will be smaller than the gap that failed to produce an inexorably rising capital share in the two centuries before 1914.

As I worked through the book, I became preoccupied with another gap: the one between the findings Piketty explains cautiously and statements such as, "The consequences for the long-term dynamics of the wealth distribution are potentially terrifying."

Piketty's terror at rising inequality is an important data point for the reader. It has perhaps influenced his judgment and his tendentious reading of his own evidence. It could also explain why the book has been greeted with such erotic intensity: It meets the need for a work of deep research and scholarly respectability which affirms that inequality, as Cassidy remarked, is "a defining issue of our era."

Maybe. But nobody should think it's the only issue. For Piketty, it is. Aside from its other flaws, "Capital in the 21st Century" invites readers to believe not just that inequality is important but that nothing else matters.

Follow @brothersjudd