I love a good fish-in-the-barrel hunting party. Here’s Fredrik deBoer on Tom Friedman’s latest, as deBoer puts it, “gem-encrusted masterpiece“:

Tom Friedman takes aim at a basic mathematical concept when he declares that”average is over.” He then proceeds to demonstrate that he doesn’t understand how averages work.

I’ve been arguing for a while now that “average is over.” It has to be when every boss has cheaper, easier, faster access to software, automation, robots, cheap foreign labor and cheap foreign genius that can produce above-average so easily. Everyone needs to find their unique value-add, their “extra,” and be constantly re-engineering themselves if they want to obtain, or advance in, a decent job that can’t be digitized.

Consider this article published in The New York Times on April 23: “EASTON, N.Y. — Something strange is happening at farms in upstate New York. The cows are milking themselves. Desperate for reliable labor and buoyed by soaring prices, dairy operations across the state are charging into a brave new world of udder care: robotic milkers, which feed and milk cow after cow without the help of a single farmhand.”

Overnight, an average farmhand went from knowing how to milk a cow to having to learn how to program and operate the robotic cow-milker — to keep a job. That takes above-average skills.

Average cannot be over. The average can move, and any individual’s position on the distribution relative to the average can change. But there will always be an average. Friedman is like the living embodiment of that Garrison Keillor joke about the town where all the children are above average. Tom: if everyone learns how to program and operate the robotic cow-milker, that becomes an average skill, and whatever economic advantage accrued to that skill will disappear. If everyone actually had “their ‘extra,’” then it wouldn’t be an extra.  The sentence “Everyone needs to find their unique value-added” might as well be the sentence “I do not know what the word ‘unique’ means.”  You can tell individuals to get ahead and say screw the rest and blather on about unique value-added skill sets. But those everyones just don’t work, Tommie.

If you teach every kid to code, there’s no value in knowing how to code. If a college education becomes a universal public good, it’s great for democracy, for our communal intellectual and aesthetic fulfillment, and for the moral principle of equal opportunity to explore, think, and create. But it means there’s no more economic value in having a college education. Average may be over, Tom, but the three sigma rule sure isn’t, and so what are you gonna do about the 68% of people who are always going to be pretty close to average?

So: in the spirit of my last post, I will proceed to defend the indefensible Tom Friedman.

Here are some possible meanings to “average is over” that are not mathematical nonsense:

  • While it will always be possible to calculate a variety of different “averages” for American wages, the colloquial meaning of “average” assumes that these numbers are similar. And they are decreasingly so. The mean, for example, may be rising, while the median may be falling – and the mode may be getting less and less meaningful as a number at all. All of which would mean that, as an analytical tool for talking about the health of the economy, the “average” wage is becoming less and less useful.
  • America has an above-average per-capita income relative to the rest of the world. If we want to maintain that above-average income, we need to upgrade the skills of our workforce so that our workforce’s skills remain above-average relative to the rest of the world. Otherwise, the relatively-unskilled portion of our workforce will see its income degrade toward the global average. Which would be substantially below-average relative to the historic American experience.
  • Unskilled or semi-skilled labor may be obsolescing on a global basis, as more and more tasks, not only in manufacturing, come to be automated. The “average” person in terms of cognitive ability has, for centuries, relied on readily-transferable but extremely limited skills to provide a limited livelihood, first as farm workers, then as industrial laborers, finally in a variety of low-level white- and pink-collar service jobs. Recent stages in this process have required meaningful cognitive advances simply to stay put and remain productively employable. In the future, nobody will remain productively employable without more specialized skills and the ability to acquire new specialized skills in a rapid timeframe. What will happen to those who don’t make the cut remains unspecified.

I don’t know for sure, of course, because Friedman speaks in buzzwords rather than in language intended to convey actual meaning, but it’s plausible that all three of the above meanings are floating around somewhere in his brain when he says “average is over.” And all three of those meanings are reasonably plausible.

But what they all point to, as deBoer’s piece is intended to point out, is the question of distribution. If “average is over” means that a smaller percentage of Americans have incomes close to the average – the bell curve is flattening, or the center of the curve is moving down as the right-hand tail gets longer and fatter – then that’s an important fact about our economy that demands explanation. And action.

What that action should be depends in part on our understanding of why it’s happening. If the main driver of stagnating median wages and rising inequality is the rise in the average skill level in China and India (the second meaning of “average is over”), then upgrading the average skills of the American workforce is a plausible response.

But if we’re going to ask the question in that form, it’s worth noting that some of the countries that have done a better job than America of moving up the value chain and keeping median wages up approach their workforces very differently than we do. Japan has a quasi-feudal corporate culture dominated by large conglomerates, with strong two-way loyalty and effective lifetime employment. Germany’s economic engine, on the other hand, is mid-sized firms, and the Rhenish model of capitalism gives workers an explicit seat at the table in the management of enterprises – as opposed to managing enterprises exclusively to maximize shareholder value. Finland couples a flexible labor market and relatively easy entrepreneurship with a very generous welfare state and effectively universal unionization. Canada’s boom has been driven substantially by natural resource wealth, but they also have a highly protected and regulated financial sector, and an immigration policy that, while it attracts more immigrants per-capita than the USA does, also focuses on attracting the most-skilled.

In other words, it’s all well and good to say “we have to upgrade our skills to stay ahead,” but we don’t have a political culture that presents a good way of pursuing that goal. We have a hard time talking about unionization, immigration, finance, etc. as relevant factors in determining the actual skill level of the American workforce, or the incentive structure that drives that skill level. We assume, instead, that upgrading skills is a matter of individual initiative, and that the state’s role, if any, is to facilitate that initiative, whether by reducing regulation or providing opportunities for retraining and the like.

And if automation is the driver – as Friedman’s example suggests – then upgrading skills may not be enough. In a utopian future where robots wind up doing most of the work, we simply won’t need that many people to manage the robots, or to design better robots. In that scenario, questions of ownership of the means of production once again become central, just as Marx predicted.

I’m skeptical of the last notion – that we’re simply going to make most humans obsolete – for reasons that deBoer would no doubt agree with (basically, that I think we’re nowhere near true artificial intelligence, and hence nowhere near the creation of robots that could truly replace the most essential human capacities). But the second proposition points in a variety of potential directions, not all of which are dreamt of in Friedman’s philosophy. So let’s take the opportunity to talk about that. Maybe Friedman will eventually even listen, and open up his mind a bit.