Comments Guidelines

All comments are pre-moderated. No spam, slurs, personal attacks, or foul language will be allowed.

Wednesday, September 10, 2014

Understanding Piketty, part 5 (conclusion)

Thomas Piketty's Capital in the Twenty-First Century is the first book to make a data-driven examination of economic inequality. Based on hundreds of years worth of data, it attempts to determine the long-term trends in inequality and the social and political consequences that follow from them.

In this final post, I want to highlight the most important points of the book, including a few I have not yet discussed. Beyond that, I want to consider parts of the book that are perhaps a bit less persuasive.

First of all, the data has been almost unchallenged. The one person who claimed substantial flaws in it, Chris Giles of the Financial Times, is road kill.

Second, three major results emerge from the data bringing dearly-held economists' views into question. A) There is no Kuznets Curve: Developed countries do not keep getting more equal; rather, the data show that they have become less equal since about 1980. B) There is no fixed share for capital and labor income, as assumed by the Cobb-Douglas production function: Capital's share of national income has risen since 1975. C) Franco Modigliani's view that most savings was for retirement, not inheritance, is wrong. Depending on the country, no more than 20% of private wealth is in the form of annuitized wealth that ends at death.

Third, the big theoretical payoff is that some economists' happy stories about how everyone earns their marginal productivity are simply incorrect. These bedtime stories may make the rich feel like their high incomes and wealth are deserved. The fact of the matter, though, is that high incomes are not the result of merit but of bargaining power. The increase of capital mobility since the 1970s is one element in disciplining labor, while the reduction in the top income tax rate gave top corporate executives more incentive to push for large wage increases and exploit the large uncertainty regarding their individual contribution to corporate success.

Fourth and most obviously, r>g* is no historical necessity, but it has held true virtually everywhere for all of human history. As long as it is true, there is a tendency for inequality to worsen.

Moving on to aspects of the book I have not previously covered, one discussion that stood out was Piketty's discussion of the weakness of measures of gross domestic product (p. 92). In particular, he notes that there are no good quality measures for adjusting GDP:
For example, if a private health insurance system costs more than a public system but does not yield truly superior quality (as a comparison of the United States and Europe suggests), then GDP will be artificially overvalued in countries that rely mainly on private insurance.
Parenthetically, it seems to me that any high-cost low-quality system would overstate GDP, whether it's private or public. But the point to remember is that we are talking big bucks here: if the United States were spending merely what the #2 country (Netherlands, in terms of percent of GDP) does, we would be spending almost $1 trillion less, so presumably this means U.S. GDP is overstated by $1 trillion. That's still a lot of money!

As I discussed before, Piketty advocates a global annual tax on wealth as the solution to the problem of inequality. However, he relegates an alternative global tax, on financial transactions, to a single paragraph plus a single footnote. He claims that an FTT would "dry up" "high frequency transactions," and for that reason would not raise much revenue. Of course, this would depend on which transactions are taxed (James Tobin had originally proposed taxing foreign exchange transactions) and what the tax rate is. A balance can be struck between "throwing sand in the wheels," as Tobin described it, and raising revenue. Contra Piketty, I don't think it is something that can be rejected out of hand, and I plan to discuss an FTT more fully in the future.

So what's wrong with the book? Honestly, not much. I mentioned before that I wasn't fully persuaded by Piketty's evidence that bigger fortunes necessarily earn higher rates of return. However, this is not a big issue, especially as the claim does seem fairly plausible.

At times, however, Piketty's political arguments seem almost ad hoc. He attributes (p. 509) the rise of Reaganism and Thatcherism in part to a feeling people had that other countries were catching up to them. He presents no evidence for this claim, which does not strike me as particularly plausible. Similarly, he lectures the leaders of large EU countries (p. 523) for their failure to align taxation among the Member States, rejecting their leaders' point that EU institutions (unanimity is required for changes affecting direct taxation) and other Member States (read: Ireland) can block fiscal coordination indefinitely. But it's true! It's right there in the Treaty! So he's a little too glib about politics for my tastes; but then, I'm a political scientist, so perhaps I'm not the most neutral of sources.

Bottom line: You've already bought the book, so take it off the coffee table and read it! It may take you a few weeks, or a few months, but you'll be glad you did.

* r>g means that the rate of return on investment, r, is greater than an economy's growth rate, g.

Cross-posted at Angry Bear.


  1. Piketty fails to distinguish between Land and Capital, a distinction which the classical economists -- Adam Smith, John Stuart Mill, David Ricardo, Henry George and others -- took very seriously. By positing a 2-factor economy rather than 3-factor (Land, Labor and Capital -- in that order!) he misses out on some very worthwhile texture which explains much of our concentration of wealth, power and income. Piketty implies that land is merely a subset of capital. Naturally flowing from that is seeing "Socialism" as the only alternative to "Capitalism."

    When Land -- agricultural, yes, but more importantly urban and suburban and coastal land, land made valuable by infrastructure and population -- and nonrenewable natural resources -- are recognized as being not "capital" but "Land," a lot of other solutions begin to emerge. See, for example, Joseph Stiglitz's short article in the September Harper's Magazine. See also Walt Rybeck's "Re-Solving the Economic Puzzle," and "The Mason Gaffney Reader: Essays on Solving the Unsolvable." Or go to Henry George's "Progress and Poverty," either in the original unabridged or in Bob Drake's modern abridgment -- both of which are online. All these may suggest a third way by which we can improve our society and economy in a just, efficient, logical way that serves the common good.

  2. “So what's wrong with the book? Honestly, not much. I mentioned before that I wasn't fully persuaded by Piketty's evidence that bigger fortunes necessarily earn higher rates of return. However, this is not a big issue, especially as the claim does seem fairly plausible.”

    I would suggest he is right. Think of how capital was increased from the beginning of the USA till 1930s when the efforts from TR to FDR put in place rules governing capitalism. Then threw the effort of the Federal Reserve these governing principles were eliminated again.

    Capitalism without restriction, start with a company, then a corporation, then part of a trust and at this stage there is little competition between these trust or innovation as these trust control the market.

    Many thank for condensing the book

  3. It seems to me that the premise of the book, r>g, is itself questionable. More questionable is the idea that large fortunes will have higher returns than small ones. (What is more likely is that the owner of a small fortune will spend a higher percentage of his.) The reason I question the latter assumption is that larger fortunes have fewer places to be invested, and that the greatest growth occurs in smaller companies which are too small for the larger investments.

    Now let's look at the r>g assumption. Let's say the GDP increased 2% last year. So in 2012, we produced 100 units of goods and services, and in 2013 we produced 102 units. Most of what we "produce" is consumed. In fact, much of what we put into our Gross Domestic "Product" does not even constitute "productive labor" as defined by Adam Smith. There is no "product" in a haircut. But that $10 you paid is added to our GDP. But if you cut your own hair, the $10 you save is not added to the GDP.

    Now, let's look at an even more egregious example. Let's say there are two unemployed women. Their husbands work and the women stay home and do the dishes, laundry, cooking, etc. One day these women decide to work for each other -- they will each clean the other's house, do the other's laundry, cooking, etc. Each will pay the other $50 every day for these services. They just added $100 per day to the GDP, and nothing actually changed, nothing was produced. The economy "grew".

    So the GDP is itself something of a myth.

    But I digress.

    When, in 2013, we produced 102 units of goods and services, did we invest 100 units of capital to do it? Of course not. So why compare the return on investments with the increase in GDP? It simply makes no sense. Returns come out of the growth of that investment, not out of the total increase in production. A company one year can produce the same amount, or even less, than it did the previous year, and still return a profit. The seed capital for next year's product is not all of this year's product.

    The idea that r>g is a Bad Thing and leads to "infinite wealth concentration" is simply preposterous.

    1. It's well-known that GDP is not perfect, but I think you are mistaken on your main point. Piketty presents the data showing that r>g in Table 12.1. Unless you've got actual data showing otherwise, you are just playing in the "dialogue of the deaf" which you can no longer get away with when there is now substantial data available.

      Moreover, your example of a company generating profits while producing less actually shows capital becoming more concentrated. In any event, there is no point challenging Piketty with made-up examples when there are real data out there. r>g has been true for most of history, and inequality has risen when that was true. r>g was not true for a big part of the 20th century, during which time inequality fell. But it's true again, and inequality has once again increased. Real data is the price of admission to serious discussion of this issue: ask Chris Giles.

      As I noted before, I do agree with you that Piketty is not fully persuasive that r keeps increasing for bigger fortunes. Compare Table 12.2 (returns on university endowments) to Table 12.1: even universities with a minimum endowment of $100 million did better than the 45 richest adults, though the biggest endowments did the best.

    2. I did not dispute that r>g, only that that is a Bad Thing.

      Only if g in that equation is the growth of investments (not GDP), would that inequality result in the increased concentration of wealth.

      Piketty is looking at the wrong growth rate.

  4. If "r>g* is no historical necessity, but it has held true virtually everywhere for all of human history" then we would presumably expect wealth inequality to get gradually worse for virtually all of human history. It hasn't, however - it has always been bad, but it has not progressed in a steady line - so there are probably regular disruptions to capital accumulation throughout history that would be worth investigating. Quite possibly wars are one of them, plagues another.