
Mike Munger on the Division of Labor 4/2/2007
What this covers
Michael Munger of Duke University joins host Russ Roberts to examine how the division of labor creates wealth. The conversation centers on Adam Smith's insight that specialization—not differences in innate ability—drives productivity gains. Munger and Roberts distinguish between two sources of specialization: Ricardian differences (static advantages based on skill or endowment) and Smithian economies of scale (dynamic gains from dividing tasks into narrower steps, which yield disproportionately large output even among identically skilled workers). They trace how technology and capital investment enable producers to add more specialized steps when markets are large enough to justify them, and they ground these abstractions in concrete examples: pin manufacturing consolidating from hundreds of factories to just two while output soared, and boots falling to roughly 5 percent of their 1800 cost in terms of hours a worker must labor to buy them.
The discussion balances two sides of specialization's advance. On one side lies dramatic material gain: cheaper goods, longer lives, and more choices. On the other lies the erosion of the personal and local—the shift from knowing your shoemaker to buying from strangers via algorithms and machines, and the job displacement that accompanies automation. Munger argues this trade-off is not consciously chosen but emerges inexorably from competition and the pursuit of efficiency. Roberts emphasizes that while specific jobs vanish, productivity gains make goods cheap enough that freed resources flow into new industries and opportunities, a benefit that remains "unseen" compared to the visible pain of displaced workers. The pair also challenge protectionist framing, noting that China—not America—shed the most manufacturing jobs in the 1990s, and that impersonal exchange with strangers, oddly, produces more wealth than restricting trade to those you know or trust.
Munger and Roberts argue that the division of labor—not innate ability differences—is the deep, often underappreciated source of wealth creation, and that as it advances it inexorably makes exchange more impersonal while making goods dramatically cheaper, a trade-off that overall raises human welfare.
- Smithian specialization yields economies of scale (more than proportional output) even without Ricardian ability differences
- Specialization requires a large enough market ('extent of the market') to justify capital investment and division of tasks
- Productivity gains reduce employment in a given field but create wealth and new opportunities elsewhere; job loss is 'seen' while gains are 'unseen'
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Skills and tools develop dynamically through repeated production: a person who incidentally makes a few pairs of shoes learns tricks, designs specialized tools, and builds a reputation, eventually becoming a dedicated shoemaker whose techniques can be taught to others—so tools are 'congealed ideas' easily passed on while skills can take long to learn.
“Tools are sort of a congealed idea that I can easily hand on to someone else. But learning skills might just take a long time.”
It is paradoxical and miraculous that, despite the modern buyer knowing essentially nothing about how boots are made—far less than the 1800 buyer—boots are far easier and cheaper to obtain; our knowledge has become so specialized that ignorance of production coexists with abundance.
“you don't know anything about boots. What a miraculous thing that... You know less about boots as the buyer of a boot than the guy in 1800.”
The advance of the division of labor makes exchange more impersonal (a Walmart purchase made by strangers, untouched by human hands, transshipped by machines) and trades away the emotional community bond of knowing your maker in return for dramatically cheaper goods—an outcome no one consciously decided but which emerged from the inexorable progress of specialization.
“I have purchased a pair of shoes that was made in another country... by a bunch of strangers... never touched except by machines. And this pleases me no end because mostly I wanted shoes.”
Restricting your trade to only people you know, only people you like, or only those who meet a moral/religious/political standard makes you very poor, because you can then trade with only a few people and the informational cost of even discovering whether others are 'good enough' is enormous; Smith's insight is that trading with strangers and keeping them strangers—judging only the goods, not the person—produces wealth.
“if you only trade with people you know, or if you only trade with people you like... you will be very poor because it means you will only trade with a few people.”
When output is large enough, you can exploit economies of scale—adding capital that would not be worth adding for small output—and Smith's genius was recognizing that dividing a process into specialized tasks yields more than proportionally greater output (e.g. 16, 20, or 100 times) even without any Ricardian differences between workers; this near-free-lunch is the surprising true source of wealth that materialist/Marxist labor-value views miss.
“if I have four times as much resources, I get way more than four times as much output, and you don't need Ricardian specialization to take that happen”
By 'extent of the market' Smith did not mean a place to buy goods, but the zone or area within which one can exchange things, and crucially the creation of a market means producing things explicitly for trade rather than for one's own use—which requires an infrastructure of expectations (security, sound currency, fair value) that barter alone does not need.
“What he meant by markets was that if we change our production decisions... I have to anticipate... that I'm going to be able to exchange that for something else.”
There are two distinct sources of specialization: Ricardian specialization (driven by differences between people in ability, training, or endowments such as land) which is essentially static, and Smithian specialization (economies of scale from dividing tasks) which is dynamic and arises even among identically-skilled people—a distinction drawn out in a paper by Buchanan and Yoon.
“this insight comes from a paper by Buchanan and Yoon, where they make a distinction between Ricardian and Smithian sources of specialization.”
It is asking too much of the human mind to anticipate in advance that dividing a task into four specialized steps yields more than four times the output; historically people first exploited obvious Ricardian differences ('you're good at this'), and the super-proportional economies-of-scale effect emerged unexpectedly, which is why it took Smith's genius to recognize its importance.
“I think it's asking too much of the human mind to say, 'I recognize that if I divide this into four different tasks and have each of them specialize, then I'll get more than four times as much output.'”
In every business there is a relentless attempt to strip labor out of the process and make it less personal; though sometimes sad and hard on those involved, the longer perspective shows this transforms the world by creating wealth, and wealth is good because it lets us live longer, richer lives—'richer' meaning more choices and quality of life (health care, time with family), not merely money.
“in every business, there's a relentless attempt to strip labor out of the process... specialization creates wealth and wealth is good.”
English pin manufacturing consolidated dramatically: hundreds of pin factories around 1776 fell to ~11 in Gloucester plus ~20 in Birmingham by 1820, to 12 by 1940, and to just 2 by 1960—yet total pins produced rose tenfold between 1940 and 1960 while employment fell to about 1% of its former level, illustrating productivity-driven concentration.
“By 1940, there were 12 and by 1960, there were two pin making factories. But the total number of pins produced had gone up by 10 times just between 1940 and 1960.”
Measured by how long the average person must work to buy them, boots cost about 5% of their 1800 price—roughly a twentyfold decline in real cost—because almost no human labor now touches them; in 1800 a pair took days or weeks of work to afford, today a few hours.
“they're about 5% of the price of a pair of boots in 1800. Measured by how long a person the average person has to work to to buy them.”
The vegetarian who happens to own fertile pastures will tend to become a cattle farmer once trade is possible, because the opportunity to trade leads people to specialize in what they are relatively good at and what their resources suit—even something they personally dislike—since doing so increases their command over the things they actually want.
“the vegetarian who happens to own fertile pastures, we would expect to become a cattle farmer.”
When two farmers each specialize and then swap (wool for cotton, beef for chicken), the total amount of stuff produced is unchanged—so it is zero-sum in output—but both are made better off, so it is not zero-sum in happiness or welfare.
“it's zero-sum in the sense of of output. We've rearranged the output, but in terms of happiness, it's not zero-sum. Both of us have gained from the trade.”
Specialization without the opportunity to trade actually makes you poorer, because if you specialize and the potential market is small, you produce far more of one thing than can be sold, which is wasteful.
“without that opportunity to trade, that specialization actually makes us poorer”
Aristotle's distinction between value in use and value in exchange survives in our minds (e.g. a homemade gift vs a store-bought one), and markets tend to drive the 'value in use' framing out of us, replacing it with value in exchange—what something costs to obtain from someone who made it.
“Aristotle's idea that value in use is superior to value in exchange is something that market drive the markets drive out of us.”
Technology and specialization reduce the number of job opportunities in a particular field (and may cause brief unemployment if change is rapid), but because the goods become so much cheaper, people have more resources to spend on other things, and new employment opportunities are created elsewhere—so the apparent shrinking of jobs is not like musical chairs losing a seat.
“what happens is that because the pins are so much cheaper, people have more resources to spend on other things and new employment opportunities get created in all those other areas.”
Many lost jobs (toll-booth takers, buggy-whip makers) are jobs no one of the next generation aspires to anyway; the Connecticut Turnpike found nearly a third of toll receipts went just to pay the collectors, so eliminating them via automation is efficient even if it displaces current workers who may deserve some transitional help.
“They did a study and found that nearly a third of all the toll receipts were going to pay the toll booth takers.”
The optimal number of steps in dividing a production process (e.g. Smith's roughly 18 steps for pin-making) is not fixed but depends on transportation costs and how large the market is, since more steps mean more employees and qualitatively more output—so a larger physical market justifies more steps.
“the appropriate number of steps in the pin making process actually depends on transportation costs and how many other people you can sell your pin to”
The nation that lost the most manufacturing jobs between 1990 and 2000 was China—the same country protectionists claim America is 'shipping jobs to'—because manufacturing employment there fell as technology and productivity replaced artisanal labor, demonstrating that productivity, not trade per se, drives manufacturing job loss everywhere.
“What nation in the world lost the most manufacturing jobs between 1990 and 2000? ... The answer is China.”
Unemployment in a sector is caused by increases in productivity (capital and automation dramatically raising output per worker), not by shipping jobs overseas; a few people in an automated factory can make 100,000 pairs of shoes, and the real payoff is that society gets more and different stuff (iPods, electric light) rather than merely shifting jobs to equipment-making.
“unemployment is not caused by shipping our jobs overseas. Unemployment is caused by increases in productivity.”
Division of labor was, for Adam Smith, the principal source of the creation of the wealth of the world: competition drives people to find cheaper ways to produce, and the biggest such way is dividing a process into more steps so people become good at each step, design new tools, and produce far more.
“division of labor for Smith was in some ways the source of the creation of the wealth of the world”
Amazon's algorithmic, impersonal recommendation system (based on what you and similar buyers purchased) delivers advice of far higher quality than even the smartest independent bookseller, because no individual can know what Amazon's entire readership knows; thus the move to big-box and online retail trades the human touch for both vastly more selection and better advice, leading people to buy more and like their purchases more.
“I don't care how smart he is. He doesn't know what Amazon's readership knows.”
Most skill is acquired, not innate: Smith recognized intrinsic talents exist but considered them not very important, because the skills we have we develop by doing tasks repeatedly, and switching between many tasks (as in self-sufficiency) is costly in time and lost proficiency.
“Smith recognized that intrinsic skills and talents exist, but they're not that important. A lot of the skills and talents that we have we develop by doing things over and over again.”
Keeping production 'in the family' or 'in the village' does not make you richer; unless you are genuinely very good at making a good (and almost no one is at, e.g., shoes), self-provision or local-only provision is deceptively impoverishing.
“we could make more money, couldn't we, if we kept that production in the family or in the village? The answer is no. We couldn't. It's deceptive”
The Williamsburg artisans (shoemaker, blacksmith, cooper, printer, wigmaker, saddler) illustrate that pre-industrial specialization existed but generated little wealth: the shoemaker cannot even supply enough shoes for the modern reenactors, saddles were luxury goods because most people couldn't afford a horse, and people were poor fundamentally because they lacked capital.
“he cannot supply enough shoes to fulfill the demand for the actors of Williamsburg today”
Personalized, local exchange should not be over-romanticized: a personal maker like 'John' could be a jerk who deliberately makes your shoes poorly out of personal animus, leaving you no recourse—so impersonal exchange has the virtue of insulating transactions from interpersonal conflict.
“John could be a jerk... and he deliberately makes your shoes really poorly, and you just have to suck it up and wear them anyway.”
The information that a knowledgeable independent retailer provides (e.g. hardware-store advice) is a real service worth paying for, and it is a moral obligation to buy from them rather than take the advice for free and purchase at a cheaper store—but as detached, packaged information it can be provided more cheaply, which is what big-box and online stores do.
“when I get the advice to shop there as well, rather than take the advice for free and then go use it at the cheaper store. I think that it should... be packaged together.”
If you find the impersonal Walmart-style economy unpleasant, you can voluntarily 'time travel into the past' by joining the Amish, living primitively in the Amazon, or starting a kibbutz-like commune that trades only locally—but you will live very poorly, and most people choose not to.
“if you find the Walmart experience sufficiently unpleasant you can live very poorly but very communally if you choose to. Most people don't choose to.”
Variety among tasks is intrinsically pleasant, which partly explains a (misplaced) romance about primitive self-sufficient life, in contrast to the Charlie Chaplin 'Modern Times' nightmare of doing one repetitive motion endlessly.
“Variety is pleasant. It's nice to move among different tasks.”
The demand for shoes differs by sex, with women on average demanding more pairs than men.
“that the demand for shoes differs by sex. I think that is a true statement.”