Make-Up: Things I Didn't Have Time to Post

Some things I haven't had time to get to:

First, the debate over whether skill-based technical change can explain changes in income inequality in recent decades has heated up again recently. This was discussed at a session at the AEA meetings in San Diego (I was a discussant on one of the other papers in the session), and the following articles do a good job of characterizing the debate:

An Inequality Debate Heats Up - Jared Bernstein

The Smartphone Have-Nots - Adam Davidson

(My own view is that both are right, it's both skill-based technical change and changes in policy.)

Second, I also wanted to get to the debate over "too big to fail":

The Legacy of Timothy Geithner - Simon Johnson

Full WSJ Transcript of the Geithner Interview - WSJ

Update: Meant to include this from Andrew Haldane: Have We Solved Too Big To Fail? 

The Simon Johnson article is critical of Tim Geithner on the too big to fail front, and Geithener provides a response in the WSJ interview when he answers the question near the start of the interview (and follow-ups) from David Wessel: "A lot of people say that by any measure the big banks are bigger than they were before the crisis. Are they still 'too big to fail,' especially if more than one goes at the same time?" (My own view is that banks have exceeded the size needed to exploit efficiency gains, their size gives them too much economic and political power, and large size creates vulnerabilities as big banks become key nodes in the financial network. Small size doesn't prevent crashes, so it's by no means a cure-all for financial problems, but it could still help.)

Third, Fred Moseley emailed a follow-up to something of his I posted earlier, and I've been trying to find time to introduce it properly (Fred chaired the AEA session on inequality discussed above, and invited me as a discussant). Unfortunately, I haven't had time to think about the microeconomics of his comment, or how current theory addresses these points and I don't want to assert things unless I'm sure. If any microeconomists (or macroeconomists) want to take on his criticisms (or reinforce them), that would be great.

However, there is a broader point here I want to discuss. The comment below says (about inequality), "One does not have to use the very dubious marginal productivity theory to explain these important phenomena. Marx’s theory provides a perfectly adequate explanation without the extremely problematic concepts of marginal products of labor and capital." My problem is that the exploitation theory in Marx is based upon the labor theory of value (LTV), and the LTV is wrong. It does not provide a coherent theory of prices. How can we believe the conclusions of a theory with incorrect foundations? For this reason, I believe the theory of exploitation needs to be updated to incorporate modern value theory, and nothing beats the utility theory of value that came out of the fight between the Marxists and the neoclassical economists in the 1800s.

I understand the idea the Marginal Product (MP) theory is an apologetic for the distribution of income within the neoclassical model. But for me, the important question is why laborers have not received the share of income that the MP theory of distribution says they should have received. What went wrong? Why, relative to the MP benchmark, did too much flow to the top, and too little to the working class (and it seems to me you have to go beyond skill-based technical change to answer this question because SBTC seems quite consistent with MP theory)? There may very well be a theory of exploitation here that a Marxist can love, but why not cast it in these terms, i.e. why not expalin why income flows have been distorted? Why not base it upon a MP theory of value rather than the incorrect LTV, and then explain how exploitation -- distortions to income flows toward the top -- works within this framework?

Anyway, I've sat on this comment far too long (apologies for that), so here it is:

This is a brief response to a recent post by Paul Krugman on his blog about “capital-biased technological change” as an explanation of stagnant real wages and the declining wage share of income. This is a follow-up to a previous comment of mine on Economists’ View, which was a response to a previous post on this subject by Krugman.

Krugman still has not clearly defined what he means by “capital-biased technological change”, but he says that he is following Hicks (1932), and Hicks definition is technological change which increases the marginal product of capital (MPK) more than the marginal production of labor (MPL); i.e. ↑MPK > ↑MPL (at a given K/L ratio). Thus, Krugman’s definition of “capital-biased technological change” is in terms of the marginal products of the marginal productivity theory of distribution.

Krugman presents an example of office work with two inputs (capital and labor) and two possible techniques, one capital-intensive and one labor-intensive. However, this example does not include raw materials (and other intermediate inputs), and thus does not address the criticism that raw materials render the concept of the MPL (or the MPK) impossible (a criticism which I made in my earlier post and which goes back to Hobson and Pareto in the early 20th century). Raw materials cannot be held constant (as the concept of marginal product of labor or capital requires) while increasing labor (or capital) and output. In order to produce another car, one must have additional wheels, brakes, etc. And between labor and raw materials, there is only one factor proportion possible (i.e. only one “technique”), so Krugman’s example of two techniques and extrapolation to a “bunch” of techniques (the usual hand-waving to approach a “smooth isoquant”) does not apply.

In order to reassure readers with doubts about marginal productivity theory (“if you’re worried”), Krugman just asserts that “labor and machines are paid their marginal products”. I argue that this conclusion is invalid in cases that include raw materials, since marginal products do not exist in these cases.

I agree with Krugman that technological change is replacing labor, not just low-skilled labor, but also medium-skilled labor, and even increasingly high-skilled labor (as he has been writing about).

But this important phenomenon cannot be analyzed in terms of MPK and MPL because these marginal products do not exist.

There are a number of other major (and probably insoluble) problems with marginal productivity theory that are fairly well known and have been known for a long time: the “aggregation” problem, the “adding-up” problem, the “reswitching” problem, the “multi-causality” problem (which is similar to the “raw materials” problem in that it renders marginal products impossible), the precise definition of the “price of capital”, the lack of adequate theories of the supply of either labor or capital, etc.. But the impossibility of marginal products in production processes that include raw materials is the most obvious insoluble problem.

Marx’s theory predicted in the early days of capitalism that technological change would tend to be labor-saving (in the usual sense of using less labor to produce the same quantity of output, not in Hicks’ and Krugman’s sense of a ratio of marginal products), and this labor-saving technological change would cause increasing unemployment (the “reserve army of the unemployed”) which in turn would put downward pressure on wages and the wage share of income (Capital, Volume 1, Chapter 25). He called this important conclusion “The General Law of Capital Accumulation” (the title of Chapter 25). One does not have to use the very dubious marginal productivity theory to explain these important phenomena. Marx’s theory provides a perfectly adequate explanation without the extremely problematic concepts of marginal products of labor and capital.

To his credit, Krugman acknowledges in another recent post that the current capital-labor dimension of inequality “has echoes of old-fashioned Marxism – which shouldn’t be reason to ignore the facts, but too often it is.” But Krugman still wants to explain these “uncomfortable” facts in terms of marginal products and marginal productivity theory, in spite of the many well-known deficiencies of this theory. I suggest that he should not ignore, not only these facts, but also Marx’s robust theoretical explanation of these facts and the capital-labor dimension of inequality.

Fred Moseley