Jeremie Cohen-Setton Has a Bunch of LInks as He Summarizes the Current State of the "Robots" Debate
The Political Medicaid Expansion Wars Begin…

In Which I Find Cass Sunstein's Implied Author Not Very Appealing...

Cass Sunstein:

Krugman, Krauthammer and Their Implied Authors: Implied authors are the imaginary people whom authors create as they put words on a page. Implied authors have their own personalities…. Wayne Booth initially developed the idea of the implied author, which he followed with a claim that works of literature consist of “company we keep,” with potentially major effects on how readers see the world and perhaps even operate in it….

Nonfiction writers… have implied authors, as well. At his best, Washington Post columnist George Will is sharp, witty and appealingly above the fray, but his implied author can be pretentious, and he sometimes wears his erudition on his sleeve. New York Times columnist Paul Krugman is a Nobel winner and a national treasure, and he knows what he is talking about, but at his worst, his implied author is arrogant and self-absorbed. Charles Krauthammer of the Post may be a great guy, but his implied author is struggling with a serious anger- management problem. Whether or not you agree with him, the New York Times’s David Brooks has a wonderful implied author--humble, open, appealingly tentative….

The concept of the implied author puts a spotlight on a real problem with current political discussions…. As Booth emphasized, the characteristics of implied authors tend to be contagious. In particular, contempt and suspicion, and a fundamental lack of generosity, spread like wildfire…

Paul Krugman:

The Conscience of a Liberal: December 11, 2012, 9:11 am: Human Versus Physical Capital: One more entry in the robots and all that discussion, just to stress how much recent trends require a new storyline. Our discourse on inequality has been dominated for decades by issues of education and talent — and for what were good reasons at the time. There was a big increase in the college premium in the 1980s and to some extent in the 1990s — but since then, not so much…. [T]he people at BLS have sent me an update of their labor share calculations: So the story has totally shifted; if you want to understand what’s happening to income distribution in the 21st century economy, you need to stop talking so much about skills, and start talking much more about profits and who owns the capital.

Mea culpa: I myself didn’t grasp this until recently. But it’s really crucial.

A Brief Note on Mary Matalin: The briefer the better, I’d say. But anyway, yesterday on “This Week” she attacked me for being a “polemicist”; what’s interesting is the context. I was trying to correct a claim she was making that was factually wrong. She threw out a number — $1.7 trillion from capping deductions — as evidence that the G.O.P. claim that it can raise lots of revenue from high-income Americans without raising tax rates is true.

But I know that number, which is discussed here. It’s what you get from capping itemized deductions at $17,000 — which has two problems. First, a significant part of the burden would fall on families well below the 98th percentile, so it would not be equivalent to raising the top rate. Second, it would devastate charitable giving. If you try to fix these things, you get a much smaller number; the White House says around $450 billion.

So Matalin yelled “polemicist” precisely when I was trying to ensure that we had the facts straight. Are you surprised?

Technology and Wages, the Analytics (Wonkish): Obviously I’m getting a lot of reaction to my stuff on robots and all that. (My copy editor, last night: “Thank God, it’s not about the fiscal cliff!”) My sense is, however, that a lot of the reaction, both positive and negative, involves misunderstanding the economic logic, with some readers believing that technological progress can never hurt workers, others believing that rapid productivity growth always hurts workers; neither is true. So here’s an attempt to explain what’s going on in the theory; cognoscenti will recognize it as nothing more than an exposition of J.R, Hicks’s analysis of the whole thing in his 1932 Theory of Wages (pdf).

Start with the notion of an aggregate production function, which relates economy-wide output to economy-wide inputs of capital and labor. Yes, that sort of aggregation does violence to the complexity of reality. So? Furthermore, for current purposes, hold the quantity of capital fixed and show how output varies with the quantity of labor. We expect the relationship to look like the lower curve in this figure (we’ll get to the upper curves in a minute).

Now, in a perfectly competitive economy (don’t worry, we’ll talk about what happens if not in a minute), we would expect the labor force to achieve full employment by accepting whatever real wage is consistent with said full employment. And what is that real wage? It’s the marginal product of labor at that point — which, graphically, is the slope of the aggregate production function where it crosses the vertical blue line.

Now suppose that we have technological progress. This manifests itself — indeed, in this context is basically defined as — an upward shift in the production function. I’ve shown two alternative curves, to make a point. Technology A and technology B are drawn so as to yield exactly the same level of output at full employment — which also says that both would lead to exactly the same rise in measured labor productivity. But they don’t have the same effect on real wages! Technology A is just a proportional upward shift in the original production function — which is “Hicks-neutral” technological change. As a result, the slope of the function where it crosses the blue line rises by that same proportion: real wages rise by the same amount as productivity.

But technology B is different — the gains are bigger at lower levels of employment, which is to say higher ratios of capital to labor (because the amount of capital is held fixed for this exercise). As a result, it is much flatter where it crosses the full employment line — which says that it would lead to much lower real wages than technology BA. In fact, as I’ve drawn it, it leads to lower real wages than under the original technology.

What we’ve just seen, then, is that the effect of technological progress on wages depends on the bias of the progress; if it’s capital-biased, workers won’t share fully in productivity gains, and if it’s strongly enough capital-biased, they can actually be made worse off.

So it’s wrong to assume, as many people on the right seem to, that gains from technology always trickle down to workers; not necessarily. It’s also wrong to assume, as some (but not all) on the left sometimes seem to — e.g., William Greider — that rapid productivity growth is necessarily jobs- or wage-destroying. It all depends.

What’s happening right now is that we are seeing a significant shift of income away from labor at the same time that we’re seeing new technologies that look, on a cursory overview, as if they’re capital-biased. So we could be looking at my technology B story above.

There are, however, other possibilities — including the possibility that the fact that we don’t actually have perfect competition is playing a big role here.

So that’s the story so far. And it’s important stuff.

Technology or Monopoly Power?: More on robots and all that: first, here’s the chart from the BLS (pdf), which focused on nonfarm business. Dean Baker warns me that the trend is a bit slower if you look at net output, because depreciation is a rising share of the total. Still, something major is happening.

Nick Rowe makes a good point, however, which is not so much a critique of the robot story as a general puzzle. Income has been shifting to capital, which would seem to increase the return to investment; but real interest rates are low by historical standards, and were low even before the financial crisis, suggesting that maybe the return to investment is if anything low. You might be able to make some headway here by stressing the different between safe assets and risky investments, but it is a puzzle. Rowe suggests that a third factor, land, may be soaking up the excess returns and being misclassified as part of capital income. Logically, this could be true; I have doubts about whether it can be a major factor empirically, although obviously the thing to do is check it out.

But there’s another possible resolution: monopoly power. Barry Lynn and Philip Longman have argued that we’re seeing a rapid rise in market concentration and market power. The thing about market power is that it could simultaneously raise the average rents to capital and reduce the return on investment as perceived by corporations, which would now take into account the negative effects of capacity growth on their markups. So a rising-monopoly-power story would be one way to resolve the seeming paradox of rapidly rising profits and low real interest rates.

As they say, this calls for more research; but the starting point is to realize that there’s something happening here, what it is ain’t exactly clear, but it’s potentially really important.

Paul Krugman's implied author feels to me like somebody who, as his first and overriding concern, wants to get the facts and arguments right.

Cass Sunstein's implied author feels to me to be sending a very different vibe.

That George Will is the kind of guy who would claim without reasons or evidence or logic that Romney would win Minnesota--that tells me a lot more important about Will than Sunstein's claim that Will's implied author is "retentious… wears his erudition on his sleeve". Sunstein claims that Brooks has "a wonderful implied author--humble, open, appealingly tentative". When I read this, I think that Brooks is the kind of guy who was neither humble nor open nor tentative nor appealing when he claimed--without reasons or evidence or logic--that Nate Silver's election projections were "getting into silly land" and that Silver "think[s he is]… a wizard. That's not possible".

Comments