A Taxonomy of Piketty's reviews

Brad DeLong wrote a lengthy review of Piketty's Capital, paraphrasing in the title the famous ISLM (ISLL in Hick's original terminology) paper by John Hicks, which deserves a review by itself. In particular the discussion of Piketty's analytical model, which is, as noted before, incredibly problematic. It is worth noticing that Brad argues somehow that there two kind of reviews, namely: those that are useful and worth reading, in which he includes Tyler Cowen's review (which I've only indirectly alluded to here and here), and the ones that are wrong and are 'distracted by irrelevances,' which include Galbraith and Palley's reviews.*

Brad's idea is not bad, I mean of a taxonomy of the reviews of Piketty, but his take is neither useful to understand the differences, nor constructive for dialogue. So here is my very brief taxonomy.

There are only a few that I included, not because these are the more important necessarily, but because they seem to be representative. I made two analytical distinctions. Mainstream and Heterodox, which is based on the theoretical background, and whether the authors accept the Neoclassical paradigm or not (Murphy and Austrians do, in a confused way). Also, reviews are classified as favorable or critical of the policy prescriptions in Piketty's book. Basically, one distinction is theoretical and the other is policy oriented.

Note that the interesting thing that emerges is that there are NO, at least to my knowledge, Critical Heterodox views, that is, heterodox authors that are against higher taxes and wealth taxes.

The basis for the mainstream critiques (Cowen and Murphy) is that taxes create distortions and lead to lower growth making things worse. After all greed is good, or something like that. This comes from the more radical fringe of the pro-markets are efficient wing of the profession. The conventional defense of Piketty is basically based on New Keynesian (or old in the case of Solow) views, and it suggests that market imperfections are sufficiently large that there is space for redistribution, and tax policy is the main way to go to redress the increase in inequality of the last 30 years.

Finally the Heterodox views (Galbraith, Palley and Lance Taylor), and here I included a non-economist, David Harvey that follows a distinctly Marxist view,** which all suggest that Piketty raises and important point, which has been discussed by heterodox economists forever, that the empirical analysis is also an addition to our understanding of inequality, but that there are theoretical flaws, which both mainstream groups (Piketty would be closer to a New Keynesian, and he is a Socialist, or at least supported Holland in the last election) share, and that this limits, but does NOT disqualify the argument for redistribution.

This illustrates also a point made in this blog for a while. On the one hand, heterodox groups are politically closer to some New Keynesian authors, which, however, remain firmly based on the orthodox notion that markets unimpeded by imperfections produce optimal outcomes. Once, the New Keynesians get rid of some of the limitations of their theoretical framework, the essential being the natural rate (of interest or unemployment), their political argument would be more coherent and stronger.

* The ones that are wrong or distract point to the logical flaws in Piketty's theoretical model, by the way.

** I didn't include Cassidy's good review, simply because as a journalist the mainstream/heterodox distinction doesn't quite apply, but his views are fundamentally favorable. By the way, the Financial Times, also would not fit that dichotomy, but it certainly fits more clearly the political one, and has come clearly against Piketty for his empirical mistakes (here). For a response see Branko Milanovic here (h/t Daniele Tavani).

More on Murphy, and Rowe on the Natural Rate of Interest

My post on Robert Murphy's critique of Piketty generated a few comments, and a good debate (see the comments section here). But there are a few things worth clarifying, and also Robert pointed out a post by Nick Rowe, which is also worth discussing in more detail.

As I noted before there seems to be a confusion among Austrians, which think that their notion of the rate of interest is purely based on intertemporal consumption (savings) preferences, and is not open to the problems of the capital debates (this is as old as Austrian economics, by the way; for more below). It would not be, in their view, equivalent to the natural rate of interest of Wicksell and the Loanable Funds Theory of the rate of interest.

First, let me get back to Robert Murphy's original post, which led to my previous post. Just to remind you his argument was that the non-Austrian mainstream (and Piketty, as a result) confused financial or monetary measures of capital with purely physical ones. It's worth quoting extensively from his post. He says:

"If a firm hires a specific capital good for a unit of time, the payment is the rental price of the capital good. For example, suppose that a warehouse pays $100,000 per year to an independent company that maintains fleets of forklifts. These annual payments are clearly due to the "marginal product" of the forklifts; the warehouse can sell more of its own services to its customers when it has use of the forklifts. 

However, these technological facts tell us nothing about the rate of interest enjoyed by the owners of the forklifts. In order to determine that, we would have to know the market price of the forklifts. For example, if the forklifts that the independent company rents out to the warehouse could be sold on the open market for $1 million, then their owners would enjoy a 10-percent return each year on their invested capital. But if the forklifts could be sold for $2 million, then the $100,000 payments—due to the "marginal product" of the forklifts—would correspond to only a 5-percent interest rate. As this simple example illustrates, knowledge of the marginal product of capital, per se, does not allow us to pin down the rate of interest."

Note that this is a triviality, and by no means contradictory with the conventional neoclassical analysis. It only says that the rate of interest specific to a particular capital good (forklifts) and its price are inversely related. That per se, certainly does NOT mean that "the relationship between the productivity of capital and the interest rate is not [direct]." The point is that, in marginalist economics, the entrepreneur would 'hire' more capital to the point were the additional (marginal) cost would equalize the additional (marginal) revenue that can be obtained from using one more unit of capital, and the latter would depend on how much more output the additional capital (forklift in this case) unit would bring. So according to changes in prices of the capital good, and, as a result, of its rate of return, the capital good would be used if it provides a gain over the interest rate. If there is an advantage in using the capital good (forklift), then more will be used, pushing its price up, and bringing its remuneration down into equilibrium with the rate of interest (the natural one).

The point of the capital debates (go here) is that there is no reason to believe that a certain technology (forklifts) would be more profitable at low rates of interest, while at higher rates of interest firms would switch to manually powered hoists (more labor intensive, arguably), for example, to lift the cargo. It would be even impossible to define clearly that one technology (forklifts) is more capital intensive than another (manual hoists). The point is that there is no relation between intensity (relative scarcity) of the use of a capital good and its remuneration.

For example, in the conventional story if the price of the forklift goes down, more capitalists would be willing to buy it, supposedly substituting other technologies (which are now relatively more expensive for the cheaper one). Yet, the fall in the price of the forklift might reduce the remuneration of the producer of forklifts, even if demand increased, since the increase in the quantity sold might very well be trumped by the effect of a lower price. Also, and more importantly for us, the decrease in the price of forklifts might lead to a reduction in the demand for forklifts. This could be the case, for instance, if the decrease in the price of forklifts and lower remuneration reduces the forklift producers' demand for other goods, which are produced, in turn, using forklifts, leading to a lower demand for forklifts. The changes in the price of the forklift, and its remuneration, are not directly correlated to its relative use (how many forklifts are bought and used in production).

Note that the fact that forklifts are produced by means of forklifts, is central to this perverse effect (the absence of any discussion of re-switching in Robert's discussion of the capital debates is telling). Here it is also worth understanding why Sraffa used the old classical and Marxist terminology of means of production rather than factors of production. A means of production is produced (like the forklift) by using means of production (including forklifts). Robert is actually utilizing the notion of a factor of production, even though he uses emptily the same terminology as Sraffians (means of production), which means that the impact of the production on capital goods (forklifts) on the production of capital goods is actually ignored.

Further, Robert does NOT deny that supply and demand determine prices and by substitution lead to allocation of resources (which makes him, and all Austrians, marginalist).* He seems just to be suggesting that a monetary rate of interest might be at some point different than the rate of remuneration of forklifts.** And it sure can. However, there must be some reason, for an agent not to invest in forklifts if the remuneration is higher than the monetary rate of interest. With free entry, and using Robert's conventional (Austrian) supply and demand logic, the entry should bring prices down, univocally lead to more demand for forklifts and equalize the marginal productivity of the forklifts and remuneration to the natural rate of interest.

Note that this opens up the question of the time preference, the other leg in Loanable Funds Theory of the rate of interest. Assume that you start from a situation in which the rate of return on forklifts is the same as the monetary rate of interest. Now assume that for some reason (Robert would say a change in intertemporal consumer preferences) the monetary rate of interest changed. Then, all of a sudden the demand for forklifts should increase, and the prices of forklifts go up, reducing its remuneration to the new equilibrium. This is when Nick's post comes in handy (again, link here).

Nick shows a very conventional story of the Loanable Funds Theory. On the one hand, we have the conventional Production Possibilities Frontier (PPF, in red), which shows how much more consumption in the future can be obtained by using less resources to produce consumption goods in the present. That is basically the marginal productivity story, in this case with the traditional neoclassical assumption of marginal diminishing returns, since the technology only allows for more consumption tomorrow at a decreasing rate (graph from Nick's post).

On the other hand, you have the indifference curve (in blue) and its slope represents the marginal intertemporal rate of substitution, which gives you how much economic agents are willing to part with consumption today in order to obtain more consumption tomorrow. When the two curves are tangential, and the marginal productivity of capital equals the marginal rate of substitution you are in equilibrium. Two things are important to note here. The intertemporal notion used in this discussion, is not exactly the same as the intertemporal notion of equilibrium used in General Equilibrium models. Not only the notion of capital above is aggregative, but more relevantly, the individual capital goods, when they are considered, would have to obtain a long-term uniform rate of profit. In fact, Bhöm-Bawerk used this notion, which was then lifted by Wicksell and Fisher (cited by Nick).***

In addition, Nick suggests that the marginal productivity of capital is NOT necessary to determine the rate of interest, but the marginal rate of transformation at which we transform less consumption goods today into more tomorrow does. Actually this is an empty distinction, since the rate at which one investment good allows you to produce (transform) more consumption goods in the future is, essentially, its marginal productivity.

This is an old and well-known confusion by Bhöm-Böhm-Bawerk, who wanted to suggest that interest rates were not the remuneration of marginal productivity of capital. His solution revolved about the notion of roundaboutness of productive process, and it does not scape the notion that marginal productivity is still relevant in the Austrian framework, and Wicksell, as well as Fisher (and if I recall correctly even J.B. Clark was too) seemed to be aware of the limitations of Böhm-Bawerk's analysis (a full explanation would require another post).

In sum, this is another case of a mainstream author that does think that markets (supply and demand) determine prices efficiently, producing the correct allocation of resources (in this case capital) confused with his own theory (for a similar lack of understanding of his own neoclassical theory by Noah Smith go here). There is the added perversity of trying to use a critique of his own theory (the capital debates) to show that the theoretically challenged but politically progressive Piketty is wrong (he is, but that idea of wealth taxes is NOT the problem).

The lack of understanding of neoclassical economics by neoclassical economists is, not surprisingly, a result of their defeat in the capital debates, and the fragmentation of teaching thereafter, something that I referred to as the return of vulgar economics. It could be said that the mainstream graduate programs are now basically the production of confused economists by means of confused economists.

* Actually his argument against Piketty's tax is that it would distort prices, and hence the incentives for entrepreneurs to invest, being detrimental to growth. So there is a lot of faith in the powers of supply and demand to allocate resources efficiently.

** One wonders if Robert read the Hayek-Sraffa debate on own rates of interest, in which Hayek committed a similar mistake.

*** Again here there is a terrible confusion in Robert's understanding of the meaning of the intertemporal models, since the latter presume, inconsistently, that the notion of a uniform rate of profit can be abandoned. That's why the intertemporal General Equilibrium models remain short-term models. By the way, as shown in the graph above the determination of the rate of interest (1+r), which is the slope at the tangential point of the PPF and the indifference curve, is the natural rate and is open to the capital debates critique.

Robert Murphy, the Austrian theory of the rate of interest and Piketty's 'Capital'

In the comments to another post it was suggested that I checked Robert Murphy's discussion of the relevance of the capital debates for Austrian economics.* The basis for my comments is Murphy's recent post on the topic here. It seems that the capital debates are somehow connected to a critique of Piketty's views on inequality from an Austrian point of view, but the post here, which was also linked in the comments, is less than clear about that.

The question is why would the capital debates, which basically criticize the main tenets of marginalism, be relevant for a marginalist school of thought like the Austrians. Shouldn't the logical flaws of marginalism affect Austrians too? [The answer is yes, by the way, but we'll get to that].

First of all, Murphy gets the main point of the capital debates wrong. He seems to think that the lack of a natural rate of interest results from the difference between aggregate capital, which must be measured in monetary terms (which he refers to as financial capital), and physical capital (which he, interestingly, refers to in the Sraffian terminology of Marx and the classical political economy authors as produced means of production). Note that it does NOT matter whether capital is in aggregative (monetary) form or if you have an array of physical capital goods, as I explained before, it is still necessary to equate aggregate investment to savings.

In Keynesian economics the equalization of investment to savings is done by the multiplier process and by variations of the level of income leaving space for a monetary story for the rate of interest. In all neoclassical (marginalist) models, including the Austrian, it is the rate of interest that equilibrates investment to full employment savings. That rate of interest is the natural rate of interest.

Murphy seems to think rather confusedly, that the idea of disaggregated capital implies that Austrian (really it would be the case for any model without aggregative measures of capital, like the Arrow-Debreu General Equilibrium too, which is hardly an Austrian model) models do not have a natural rate.** Let me repeat it then, any model with disaggregated means of production (capital goods) still requires for the equilibration of aggregate investment to full employment savings, it requires a measure of the quantity of capital that corresponds to aggregate investment, and that means, by necessity, a rate of interest that equilibrates investment and savings. The point of the capital debates is that there is no direct relationship between the intensity of the use of capital and its remuneration, that is, no guarantee that at lower rates of interest more capital would be used, and full utilization of resources would be produced by the free interplay of market forces (something that Murphy, as an Austrian, believes in).

Worst, Murphy seems to think that the capital debates applies only to the capital market. He says: "the relationship between the productivity of capital and the interest rate is not directly analogous to the relationship between the productivity of labor and the wage rate." As it turns, the point of the capital debates is that if you reduce the wage rate, there is also no guarantee that more labor would be utilized, and there is no necessary relationship between marginal productivity of labor and real wages (and the evidence in favor of that is also flimsy, to say the least). In other words, the capital debates apply to the marginalist labor market too. No 'factor of production' is remunerated according to marginal productivities (Samuelson got that right, all the neoclassical parables are problematic, and it is a bit surprising to find this amount of confusion so long after the capital debates have been resolved).

So its seems that it is not just Piketty that "has no clue about Capital." The fascinating thing about Murphy's critique of Piketty's lack of knowledge about capital, is that for him this suggests that Piketty's  wealth tax would be a bad idea (note that critiques from the left, like Galbraith or Palley are not against a wealth tax, but suggest that inequality must be combated in other ways too, with stronger unions, more regulated capital, full employment policies, etc.). Here Murphy seems to think, like Tyler Cowen, that taxes would punish entrepreneurs and reduce the dynamism of capitalism. Because, you know (wink wink nudge nudge, say no more), without taxes, and other impediments, capital would be more efficiently utilized. So here is an economist that criticizes higher taxes as the solution for inequality, by suggesting that the notion of capital used to defend a wealth tax is flawed, and uses that very same flawed notion of capital (without even getting it) to defend a laissez-faire solution.

* By the way, not the first time I'm asked to comment on Austrians (see here and here). Austrians stand for economics like Libertarians for politics, and they are a militant group, which would be my guess of why there are so many people concerned with Austrian theory. Love for Hayek and Ayn Rand are highly correlated among teenage students, in my experience. And that's not very good company for Hayek.

** The fact that Murphy does think the capital debates are about aggregation is clear when he asks: "Does anyone know, does Piketty’s book elsewhere deal with the problem of aggregating capital?"