It is one of the great explications of economics of modern times: Written in 1958 by libertarian Leonard Read and subsequently performed by Milton Friedman as The Pencil, a couple of minutes of Free to Choose, the ten-part television series he made with his economist wife, Rose Director Friedman, broadcast and published in 1980.
Friedman comes alive as he enumerates the various products required for the manufacture of a simple pencil: the wood (and, of course, the saw that cut down the tree, the steel that made the saw, the iron ore that made the steel and so on), graphite, rubber, paint (“This brass ferrule? I haven’t the slightest idea where it comes from”).
Literally thousands of people cooperated to make this pencil – people who don’t speak the same language, who practice different religions, who might hate one another if they ever met.
This is Friedman as he was experienced by those around him, sparks shooting out of his eyes. The insight itself might as well have been Frederic Bastiat in 1850 explaining the provisioning of Paris, or Adam Smith himself in 1776 writing about the economics of the pin factory.
There is a problem, though. None of these master explicators have so much as word to say about how the pencil comes into being. Nor, for that matter, does most present-day economics, which remains mainly prices and quantities. As Luis Garicano, of the London School of Economics, and Esteban Rossi-Hansberg, of Princeton University, write in a new article for the seventh edition of the Annual Review of Economics:
Mainstream economic models still abstract from modeling the organizational problem that is necessarily embedded in any production process.Typically these jump directly to the formulation of a production function that depends on total quantities of a pre-determined and inflexible set of inputs.
In other words, economics assumes the pencil. Though this approach is often practical, Garicano and Rossi-Hansberg write, it ignores some very important issues, those surrounding not just the companies that make the products that make pencils, and the pencils themselves, but the terms under which all their employees work, and, ultimately, the societies in which they live.
In “Knowledge-based Hierarchies: Using Organizations to Understand the Economy,” Garicano and Rossi-Hansberg lay out in some detail a prospectus for an organization-based view of economics. The approach, they say, promises to shed new light on many of the most pressing problems of the present day: evolution of wage inequality, the growth and productivity of firms, the gains from trade, the possibilities for economic development, from off-shoring and the formulation of international teams, — and, ultimately, the taxation of all that.
The authors note that, at least since Frank Knight described the role of entrepreneurs, in Risk, Uncertainty, and Profit, in 1921, economists have recognized the importance of understanding the organization of work. Nobel laureates Herbert Simon and Kenneth Arrow each tackled the issue of hierarchy. Roy Radner, of Bell Laboratory and New York University, went further than any other in developing a theory of teams, especially, the authors say, in “The Organization of Decentralized Information Processing,” in 1993.
But all the early theorizing, economic though it may have been in its concern for incentives and information, was done in isolation from analysis of the market itself, according to Garicano and Rossi-Hansberg. The first papers had nothing tp say about the effects of one organization on all the others, or about the implications of the fact that people differ greatly in their skills.
That changed in 1978, the authors say. A decade earlier, legal scholar Henry Manne had noted that a better pianist had higher earnings not only because of his skill; his reputation meant that he played in larger halls. The insight led Manne to conjecture that large corporations existed to allocate the production most efficiently of managers, like so many pianists of different levels of ability.
It was Robert Lucas, of the University of Chicago, who took up the task in 1978 of showing precisely how such “superstar” effects might account for the size of firms, with CEOs of different abilities hiring masses of undifferentiated workers – and why scale might be an important aspect of organization. He succeeded, mainly in the latter, generating fresh interest among economists in the work of business historian Alfred Chandler.
It was Sherwin Rosen, of the University of Chicago, with “The Economics of Superstars,” in 1982, who convincingly made the case that the increasing salaries paid to managers had to do with the increase in scale of the operations over which they preside (and, with athletes, singers and others, the size of the audiences for whom they perform). A good manager might increase the productivity of all workers; the competition among firms to hire the best might cause the winners to build more and larger teams; but Rosen didn’t succeed at building hierarchical levels into his model. He died in 2001, at 62, a few months after he organized the meetings of the American Economic Association as president.
Many others took up the work, including Garicano and Rossi-Hansberg. It was the Nineties, not long after a flurry of work on the determinants of economic growth spelled out for the first time in formal terms the special properties of knowledge as an input in production. The work on skills and layers in hierarchies gained traction once knowledge entered the picture.
At the meetings of the American Economic Association this weekend in Boston, a pair of sessions were devoted to going over that old ground, one on the “new growth economics” of the Eighties, another on the “optimal growth” literature of the Sixties. Those hoping for clear outcomes were disappointed.
Chicago’s Lucas; Paul Romer, of New York University; and Philippe Aghion, of Harvard University, talked at cross purposes, sometimes bitterly, while Aghion’s research partner, Peter Howitt, of Brown University, looked on.But Gene Grossman, of Princeton University, who with Elhanan Helpman, of Harvard University, was another contestant in what turned out to be a memorable race, put succinctly in his prepared remarks what he thought had happened:
Up until the mid-1980, studies of growth focused primarily on the accumulation of physical capital. But capital accumulation at a rate faster than the rate of population growth is likely to meet diminishing returns that can drive the marginal product of capital below a threshold in which the incentives for ongoing investment vanish. This observation led Romer (1990), Lucas, (1988), Aghion and Howitt (1992) Grossman and Helpman (1991) and others to focus instead on the accumulation of knowledge, be it embodied in textbooks and firms as “technology” or in people as “human capital.” Knowledge is different from physical capital inasmuch as it is often non-rivalrous; its use by one person or firm in some application does not preclude its simultaneous or subsequent use by others.
My guess is that “Knowledge-based Hierarchies: Using Organizations to Understand the Economy” will mark a watershed in this debate, the point after which arguments about the significance of knowledge will be downhill. “If one worker on his own doesn’t know how to program a robot, a team of ten similar worker will also fail,” write Garicano and Rossi-Hansberg. The only question is whether to make or buy the necessary know-how.
What’s new here is the implication that as inequality at top of the wage distribution grows, inequality at the bottom will diminish less, as the middle class is hollowed out.
[E]xperts, the superstars of the knowledge economy, earn a lot more while less knowledgeable workers become more equal since their knowledge becomes less useful. Moreover, communications technology allows superstars to leverage their expertise by hiring many workers who know little, thereby casting a shadow on the best workers who used to be the ones exclusively working with them. We call this the shadow of superstars.
For a poignant example of the shadow, see last week’s cover story in The Economist, Workers on Tap. The lead editorial rejoices that a young computer programmer in San Francisco can live like a princess, with chauffeurs, maids, chefs, personal shoppers. How? In There’s an App for That, the magazine explains that entrepreneurs are hiring “service pros” to perform nearly every conceivable service – Uber, Handy, SpoonRockert, Instacart are among the startups. These free-lancers earn something like $18 an hour. The most industrious among them, something like twenty percent of the workforce, earn as much as $30,000 a year. The entrepreneurs get rich. The taxi drivers, restaurauteurs, grocers and secretaries who used to enjoy middle class livings are pressed.
Work on the organization-based view of economics is just beginning: beyond lie all the interesting questions of industrial organization, economic development, trade and public finance. Much of the agenda is set out in the volume whose appearance marked the formal beginnings of the field, The Handbook of Organizational Economics (Princeton, 2013), edited by Robert Gibbons, of the Sloan School of Management of the Massachusetts Institute of Technology, and John Roberts, of Stanford University’s Graduate School of Business. Included is a lucid survey of the hierarchies literature by Garicano and Timothy Van Zandt, of INSEAD.
The next great expositor of economics, whoever she or he turns out to be, will give a very different account of the pencil.
Andrew W. Marshall retired last week after 41 years as director of the Defense Department’s Office of Net Assessment, the Pentagon’s internal think-tank. A graduate of the University of Chicago, a veteran of the Cowles Commission and RAND Corp., Marshall was originally appointed by President Richard Nixon, at the behest of Defense Secretary James Schlesinger, and reappointed by every president since. He served fourteen secretaries with little external commotion.
A biography to be published next week, The Last Warrior: Andrew Marshall and the Shaping of Modern American Defense Strategy (2015, Basic Books), by two former aides, Andrew Krepinevich and Barry Watts, is already generating commotion. Expect to hear more about Marshall in the coming year.