Yglesias Interviews Piketty
If you’re getting sick of the coverage, don’t worry: After I post my review (coming in a few weeks), I will probably be done with this topic.
In the meantime, Piketty says some interesting things in this interview with Matt Yglesias. For example:
I think what [economists are] doing wrong is that in order to distinguish themselves from other disciplines, in order to look like we all are scientists, they use too much complicated math just for the sake of it.
Math is fine. Math is very cool. But very often, they tend to push for more sophisticated math just to push off other people. It’s an easy way to have the appearance of scientificity. For a real mathematician or physicist, the math will not be terribly impressive but it’s enough to impress those economics departments that are less good at math and those social scientists who are less good at math.
…
When I do sum all these assets and their market value, I do not mean to suggest that this is an adequate summary of everything. I am very much aware that this operation of summing up everything as a single market value of all capital assets and call it K, and this is capital, it’s an incredibly abstract operation.I think it can be useful for some purpose, as long as we have a critical eye and keep a critical eye on what it means. One should not put too much weight on this abstract operation, and on how much I believe in it.
I certainly do not believe in a model with a single form of capital. The one good model where we produce apples and capital is just a physical accumulation of apples is not an adequate model to describe any society at any period in time.
…
Inflation has proved to be very useful to reduce the large stocks of public debts that we had in the 20th century. Now the progressive wealth tax, in a way, is the same thing as inflation, but this is sort of a civilized form of inflation.It’s like inflation, but you can make sure that people with limited wealth would not be hurt, and people with billions would pay more. With inflation you have chaos, in that you don’t actually know who’s going to pay for it.
Very often, not only do you destroy the public debt, but you also destroy the savings accounts of lower and middle class people. I think this is why Europe today, for instance, has a very hard time with inflation.
There is an alternate universe (similar to ours but different in key ways) in which Piketty is a hard-core Misesian.
More on Piketty
Here’s my latest Mises Canada post, in which I show how Brad DeLong actually (though it’s not his intent) concedes that Piketty’s entire case is built on quicksand.
Matt Rognlie (who I think is this PhD econ student at MIT) left what may be the most important comments in a MR blog post this year. Let me quote one crucial part and explain what Rognlie is saying:
Krugman correctly highlights the importance of the elasticity of substitution between capital and labor, but like everyone else (including, apparently, Piketty himself) he misses a subtle but absolutely crucial point.
When economists discuss this elasticity, they generally do so in the context of a gross production function (*not* net of depreciation). In this setting, the elasticity of substitution gives the relationship between the capital-output ratio K/Y and the user cost of capital, which is r+delta, the sum of the relevant real rate of return and the depreciation rate. For instance, if this elasticity is 1.5 and r+delta decreases by a factor of 2, then (moving along the demand curve) K/Y will increase by a factor of 2^(1.5) = 2.8.
Piketty, on the other hand, uses only net concepts, as they are relevant for understanding net income. When he talks about the critical importance of an elasticity of substitution greater than one, he means an elasticity of substitution in the *net* production function. This is a very different concept…
This is no mere quibble. For the US capital stock, the average depreciation rate is a little above delta=5%. Suppose that we take Piketty’s starting point of r=5%. Then r/(r+delta) = 1/2, and the net production function elasticities that matter to Piketty’s argument are only 1/2 of the corresponding elasticities for the gross production function!
…
What does this all mean for the Piketty’s central points – that total capital income rK/Y will increase, and that r-g will grow? His model imposes a constant, exogenous net savings rate ‘s’, which brings him to the “second fundamental law of capitalism”, which is that asymptotically K/Y = s/g. The worry is that as g decreases due to demographics and (possibly) slower per capita growth, this will lead to a very large increase in K/Y. But, of course, this only means an increase in net capital income rK/Y if Piketty’s elasticity of substitution is above 1, or if equivalently the usual elasticity of substitution is above 2. This is already a very high value, and frankly one to be treated with skepticism.
OK it’s past midnight as I type this out, so I reserve the right to change this later. But–especially if you follow the link to my Mises Canada post and see how DeLong paraphrased Rognlie’s concerns–I think this is what’s going on:
With standard estimates of the parameter values you plug into an aggregate production function, you get the result that if the market value of the capital stock keeps increasing relative to annual output (i.e. K/Y or “total wealth” expressed as a multiple of GDP), then the share of annual income going to the capitalists (i.e. rK/Y) goes down, meaning that the share of annual income going to the workers (i.e. wL/Y) goes up. That’s not the outcome Rognlie Piketty wants; he needs inequality (in terms of the percentage of income earned by the “top x%”) to be increasing in order to scare everyone into supporting a global wealth tax.
Now to get that result, Piketty argues why we should opt for the higher end of the range of values on those parameters. But Rognlie is pointing out that Piketty is mixed up: he is making a basic mistake in going from one type of application to another. Once you adjust for depreciation, you realize that Piketty’s needed parameter values are *way* outside the range of plausibility. Hence Rognlie concludes: “Unless I’m missing something, the formal apparatus in Piketty’s book simply is not capable of generating the results he touts…Perhaps there are modifications to the framework that can redeem it, but as it currently stands I’m baffled.”
Climate Change and Captain Kirk
For some time now I’ve been trying to convey just how ludicrous it is that the US government’s anti-carbon policies are based on computer simulations of the global economic/climate system through the year 2300. A recent piece by David Kreutzer and Kevin Dayaratna (an economist and computer programmer, respectively, at the Heritage Foundation) is the best I’ve yet seen:
When you switch on the kitchen light tonight, how will it affect Captain James T. Kirk and the intrepid crew of the Starship Enterprise?
Captain Kirk is a fictional character, of course, but the question — thanks to the EPA — is all too real. The agency calls it the “Social Cost of Carbon.” In the SCC they claim to have an estimate, measured in dollars per ton of CO2, for all the damage that your free-and-easy light-switching today will impose on the world from now until the year 2300, at which point Captain Kirk would be 77. To save Kirk and the rest of future Earth from a panoply of speculative, incremental horrors, the EPA has plans to nudge your carbony little fingers away from that switch and many others.
The two authors then go on to describe their results when they tweak 2 of the 3 computer models that the Obama Administration’s Working Group used to calculate the “social cost of carbon”:
Like the EPA, we ran the DICE model (created by William Nordhaus) and the FUND model (by Richard Tol). But we used more recent peer-reviewed estimates of CO2’s impact on world temperature (here and here), and we ran the models using a 7 percent discount rate (that is, the rate at which we value short-term effects over long-term effects), as stipulated by the OMB and curiously ignored by the EPA, which used 2.5 percent, 3 percent, and 5 percent.
The simple substitution of the discount rate drops the SCC estimate by more than 80 percent in the DICE model and pushes it to zero or even negative for the FUND model. (Here it might be appropriate to observe a moment of sympathetic silence for Richard Tol, a lead and convening IPCC author, who was recently excoriated by his former brethren for publicly stating the painfully obvious: Some impacts of warming are good.)
Updating the estimates of CO2’s impact on temperatures drops the SCC estimates by 40 to 80 percent. When the 7 percent discount rate is used in conjunction with the updated CO2 impacts, the FUND model indicates a better-than-even chance the SCC is negative.
That is, using more up-to-date numbers renders a range of estimates so broad that it is not clear whether CO2 emissions should be restricted or subsidized. Are these changes debatable? Absolutely. Is the debate on these topics over? Absolutely not. We cannot calculate how today’s trip to the grocery will harm humanity in three centuries. Nevertheless, the EPA is forging ahead with regulations employing a SCC of about $40 per ton for the near term. [Bold added.]
Incidentally, I haven’t personally vetted their computer runs, but their results are consistent with my own simulations run on the DICE model when I tweaked it for a paper in The Independent Review. And as I explained in my Senate testimony last summer, the FUND model does indeed show “positive externalities” of global warming for modest warming, which means that the use of a high enough discount rate would render a prima facie case for subsidizing carbon dioxide emissions.
Of course, we aren’t seriously proposing government subsidies on the grounds that CO2 emissions confer “social benefits” that aren’t being fully internalized by the emitters. The point is that even taking the Administration’s case on its own terms, it’s not even clear that CO2 emissions are a bad thing. And to repeat, we’re not inventing “the Heritage Foundation computer model.” Rather, we are running 2 of the 3 computer models that the Obama Administration team used to come up with these figures for federal agencies.
Last point: You might wonder why they (and I) didn’t try peeking under the hood of the third model, the so-called PAGE model. The answer is that it’s proprietary, and the creator (Chris Hope) doesn’t make the code publicly available. (If you email him, he gives conditions under which you can use it, which involves paying him for a pretty expensive training session.) That’s of course his prerogative–and I mean that sincerely, I don’t expect him to volunteer his time or post his proprietary code for the world to download–but it makes it impossible for outside researchers to vet the government’s reported figures when one of the computer models involved is inaccessible. One would think that when federal officials selected the representative computer models to use when estimating the “social cost of carbon,” they would pick models that are publicly available.
It’s Easy to Boo When the Guy’s Not Right There
This happens at conferences when people talk about how “David Friedman pwned Murphy” in the Porcfest debate and I’m standing right behind them.
Some Background on Capital Controversy
I am going to be putting out stuff in the coming month on why Thomas Piketty’s use of a neoclassical aggregate production function is so dubious–Austrians and Post-Keynesians agree–but if you just can’t bear the suspense, here are three links:
==> I walk through the “reswitching” controversy, focusing on Paul Samuelson’s excellent numerical example. Note that he sold this in a very interesting way: It is actually a concession that his side lost in the Cambridge Capital Controversy (that’s what Galbraith was alluding to in his critique of Piketty), but he shakes it off and portrays it as the death of Bohm-Bawerkian analysis. (Strictly speaking, Bohm-Bawerk said nothing incorrect during the CCC, because he had been long dead when it occurred.)
==> I explain (and critique) Sraffa’s approach to explaining the income flowing to capitalists. (Sraffa rejected the marginal productivity approach to income determination when it came to the capitalists.)
==> If you’re super geeky, and want to see how even-handed I am, in this academic paper I devote one section to showing that I think the Austrians did not fully answer Sraffa’s critique of Hayek.
Potpourri
==> Tom Woods interviews Jordan Page, a musician who made a name for himself among liberty fans by opening for a lot of Ron Paul’s appearances.
==> I really liked this Stefan Molyneux video on heroism. (Thanks to a Molyneux fan who didn’t like my video on “Rio 2.”) I don’t think I agree with it, but it’s a fascinating conspiracy theory.
==> Regarding Catalan’s claims about Krugman vs. Rothbard, I’m trying to find any Austrian recognition of the fact that you can get specialization just relying on economies of scale. I ask because I am pretty sure we criticized the statements of Mises and Rothbard on this point back in the early 2000s at the Rothbard Graduate Seminar (when I was a student); Block et al. published the idea here. (I’m not saying the two episodes were related.) Does anyone know if any Austrian alluded to the concept, before Krugman wrote it up formally?
==> Here’s an IER post I did on the Massachusetts gubernatorial race, where the Democrats are all vying for carbon tax street cred.
==> Richard Ebeling on the lower cost of a truly limited government.
Compton and Long Beach Togetha on This One: Piketty Has No Clue About Capital
It’s rare that you will see Peter Klein and James Galbraith agree, but they do. And after getting my own copy of Capital in the 21st Century, I can join the party: Thomas Piketty doesn’t have the foggiest idea what economists are arguing about when they bring up concerns over aggregation in capital theory.
Here’s Klein (his is the second review, under Hunter Lewis):
Piketty understands “capital” as a homogeneous, liquid pool of funds, not a heterogeneous stock of capital assets. This is not merely a terminological issue, as those familiar with the debates on capital theory from the 1930s and 1940s are well aware. Piketty’s approach focuses on the quantity of capital and, more importantly, the rate of return on capital. But these concepts make little sense from the perspective of Austrian capital theory, which emphasizes the complexity, variety, and quality of the economy’s capital structure. There is no way to measure the quantity of capital, nor would such a number be meaningful. The value of heterogeneous capital goods depends on their place in an entrepreneur’s subjective production plan. Production is fraught with uncertainty. Entrepreneurs acquire, deploy, combine, and recombine capital goods in anticipation of profit, but there is no such thing as a “rate of return on invested capital.”
Profits are amounts, not rates. The old notion of capital as a pool of funds that generates a rate of return automatically, just by existing, is incomprehensible from the perspective of modern production theory.
Piketty wants to provide a theory relevant to growth, which requires physical capital as its input. And yet he deploys an empirical measure that is unrelated to productive physical capital and whose dollar value depends, in part, on the return on capital. Where does the rate of return come from? Piketty never says. He merely asserts that the return on capital has usually averaged a certain value, say 5 percent on land in the nineteenth century, and higher in the twentieth.
The basic neoclassical theory holds that the rate of return on capital depends on its (marginal) productivity. In that case, we must be thinking of physical capital—and this (again) appears to be Piketty’s view. But the effort to build a theory of physical capital with a technological rate-of-return collapsed long ago, under a withering challenge from critics based in Cambridge, England in the 1950s and 1960s, notably Joan Robinson, Piero Sraffa, and Luigi Pasinetti.
After quoting from Piketty’s (3-page) discussion of the Cambridge Capital Controversy–in which Piketty somehow manages to conclude that Solow’s model “carried the day”–Galbraith points out, “And Solow’s model did not carry the day. In 1966 Samuelson conceded the Cambridge argument!”
I am really trying not to be a know-it-all about this, lashing out with suppressed jealousy at the guy who is getting worldwide acclaim. But I am pretty sure my jaw literally dropped when I went right to the section on the Cambridge Controversy in Piketty’s book (which I obtained just a few hours ago, and have only spent 30 minutes perusing so far).
Does anyone know, does Piketty’s book elsewhere deal with the problem of aggregating capital? I mean, since he doesn’t even bring it up in three pages devoted to a debate about aggregating capital (really: Wikipedia tells you more in the first two paragraphs than in Piketty’s 3 pages), I would be surprised if he mentioned it elsewhere…but I don’t want to assume. Who knows how these Frenchmen write books?
Hayek > Krugman
I focused on Mises in my Mises Canada post, but I also should highlight this part:
Thus we see the where Finegold is coming from, when he claims that Krugman is clearly a better (or more important) economist than Rothbard, and that Krugman is indeed comparable to Mises and Hayek. It will be easier for me to tackle the latter claim first. Since it’s such a slam-dunk case, I’ll just focus on Mises, but I would personally put Hayek’s 1937and 1945 papers on knowledge as far more important than the entire body of work produced by Paul Krugman during his lifetime (thus far, since we don’t know what he’ll do tomorrow). And no, I don’t just mean, “I hate Krugman’s policy conclusions.” I am talking about contributions to economic science. I would much rather the people getting PhDs in economics this year read and absorb Hayek’s two papers, than understand how economies of scale can explain trade patterns.
Recent Comments