My Account

Warning: You are using a browser that does not support angularJS. Some site functionality will not be available to you. Please consider updating to a newer version.

Comparative Advantage: An Idea Whose Time Has Passed

Michael Munger

Many economists will tell you that the most important principle in economics is comparative advantage — the idea that it is expensive to grow oranges in Alaska or to flood rice paddies in Saudi Arabia, so Alaska and Saudi Arabia should import oranges and rice, respectively, and base local production on the advantages of local conditions. We got this idea from classical economist David Ricardo, who famously observed in 1821 that England and Portugal would both be wealthier if Portugal exported its wine and imported England’s textiles, and vice versa.

Comparative advantage is not a separate concept at all. It is simply an explanation of the implications of the division of labor and opportunity cost. 

Ricardo’s principle even demonstrated the advantages of trading with those who are less productive at everything. For example, my wife is an attorney. She is also a fast and accurate typist. Yet she hires a secretary who is considerably slower at typing. Secretaries get paid less than attorneys, so if my wife specializes in law and the secretary specializes in typing, my wife can earn more for her firm and a secretary gets a job. Both end up better off. That’s true even though my wife is better at both jobs: comparative advantage means trade helps everyone.

Division of Labor Trumps Comparative Advantage

The problem is that fixed comparative advantage — derived from weather, culture, and location — is vanishing in the modern world. Ricardo’s classical formulation leaves no space for human creativity, no role for division of labor, and no room for innovation to affect the dynamics of cost.

So economists have it wrong, as my friend Russ Roberts argued in 2010. The most important principle in economics is opportunity cost. Here’s proof: you can define opportunity cost without resorting to comparative advantage. But you can’t possibly define comparative advantage without invoking opportunity cost.

The notion of comparative advantage is empirically misleading, because it sounds deterministic. There are few situations where fixed factors make the relative opportunity costs of different actions immutable. Instead, cost and productivity differences are endogenous, the consequence of human ingenuity and the division of labor. Today’s cost advantage for one country may disappear if another country finds a better, cheaper way to produce the product. And the way to specialize is to exploit the division of labor.

Sock City

What nation lost the most manufacturing jobs from 1990 to 2000? China. That may seem surprising, given the media stereotype of how we “ship US jobs overseas,” but it’s true (PDF). In 1990, Chinese manufacturing meant large sheds filled with hundreds of people working with sewing machines and other small tools. The scale was huge, with at least 100 million manufacturing workers. But productivity was terrible.

In the late 1990s, China began to automate, taking advantage of division of labor. A thousand women with sewing machines in a barn turned into 25 women running enormous machines in a factory, with a gigantic increase in productivity. A fraction of the workers produced 10 times as much output, increasing productivity a hundredfold and forcing 97.5 percent of the workers out. But those workers found new jobs, as China used the division of labor more and more effectively. The country’s advantage was not climate, not soil quality, but human action consciously designing production processes that were cheaper and faster.

Nowhere did productivity rise faster than in the city of Datang in Guangdong Province. Part of Datang is actually called “Sock City,” because more than a third of all socks sold in the entire world (yes, the world) are manufactured there. Datang boasts more than 8,000 hosiery makers ringing the city center, and they produced more than 11 billion pairs of socks in 2012. The socks you’ll find at Walmart — or even at Neiman Marcus or another more upscale store — were likely made in or around Datang.

The concentration of manufacturing at Sock City means this: there is a well-developed labor market for exotic sock-making specialties. The occupations that are well known in Datang don’t exist elsewhere, because no other location has been able to take such full advantage of the division of labor. What limits the division of labor in Datang? Only the extent of the market, just as Adam Smith said in The Wealth of Nations. And remember that Datang is producing at a rate of nearly two pairs of socks per year, for every human on the planet. Datang’s market is Earth.

It wasn’t always that way. Datang does not have any comparative advantage, at least not in the way Ricardo meant. The climate is not especially favorable, the city is not near an essential natural-resource base, and sock making is not part of traditional culture. Datang’s dominance is new and is overtaking historical frontrunners like Fort Payne, Alabama, the self-proclaimed Sock Capital of the World.

Fort Payne “began making stockings in 1907 and once boasted of producing 1 of every 8 pairs worn on the planet,” writes Don Lee in “The New Foreign Aid,” published April 10, 2005, in the Los Angeles Times. However, he explains,

China’s advantages in the global marketplace are moving well beyond cheap equipment, material and labor. The country also exploits something called clustering in a way that the United States just can’t match.… Industrial clusters are like one-stop production centers, achieving economies of scale and driving innovation by geographically bunching suppliers, manufacturers and contractors.…

Meanwhile, American producers, pummeled by imports from China and elsewhere, saw their share of the US hosiery market fall from 69% in 2000 to 44% in 2003, according to the latest industry data.

Comparative advantage is fixed and exogenous. Opportunity cost is mutable, the product of innovation. Datang’s Sock City itself may soon lose its dominance.

Who “should” produce socks? Comparative advantage here is no guide; the situation is more like comparing two street porters who appear to be quite similar. One of the street porters figures out ways to make socks much more cheaply. Over time, the advantage in opportunity cost grows because of the improvements in dexterity, tool use, and design of new production processes. Human ingenuity created an opportunity for nations to specialize in activities where their opportunity costs were lower. Specialization and trade are what produce prosperity, and opportunity costs guide the choice of what each country should specialize in. My comparative advantage today may be your comparative advantage next year. But all the street porters started out the same.

Focus on Opportunity Cost

Economists routinely act as if three related key concepts — division of labor, comparative advantage, and opportunity cost — are distinct.

They are not. Comparative advantage is not a separate concept at all. It is simply an explanation of the implications of the division of labor (the engine that drives prosperity) and opportunity cost (the concept that guides the choice of which activities a person, or a nation, should specialize in).

Admittedly, it was a significant intellectual achievement to show that the weaker trading partner benefits from trade, even if the stronger partner is better at everything. But those fixed differences have largely disappeared in many markets. The question of what should be produced, and where, is now answered by dynamic processes of market signals and price movements, driven by human ingenuity and creativity. The cost savings resulting from successfully dividing labor and automating production processes dwarf the considerations that made comparative advantage a useful concept in economics.

Let’s downgrade comparative advantage from our list of key concepts in economics, and recognize that the human mind is the mainspring of a market economy.

Related Articles


{{}} - {{relArticle.pub_date | date : 'MMMM dd, yyyy'}}