Implementing market-balanced trade

"Friedrich List does not refute Adam Smith. He generalizes Adam Smith."

Mercantilism is hard. Tariffs are one possible element of a successful trade policy. While tariffs worked well in East Asia in the second half of the 20th century, they worked badly in Latin America at the same time. Tariffs do not guarantee success.

In this paywalled post, I’ll explain the generalized trade framework known as import certificates, a Warren Buffett proposal which flamed out in the Senate in 2006. This is mainstream Gray Mirror—a rare take. The idea is not mine or Buffett’s, but much older.

The key win of market-balanced trade is that the policymaker sets the level of balance, and the market implements the level of balance. There are no arbitrary tariff numbers. (China’s currency-manipulation scheme does something similar in a different way.) So MBT has a clean quality which has been lacking in the new administration’s measures.

If you cannot afford to peek behind the paywall, just read National System of Political Economy, by Friedrich List. While List died in poverty, his ideas were embraced both by Wilhelmine Germany and postwar East Asia—where anyone can see the result, both in statistics and with their own two eyes.

It is erroneous to say that List refutes Adam Smith. List himself takes great pains to deny it. Given Adam Smith’s assumptions, List tells you, he is completely right. But his assumptions amount to denying the existence of nations. Smith is not studying political economy, but a special case of political economy: cosmopolitical economy, economics for a planet made up of a single nation.

Einstein did not refute Newton—he generalized Newton. Friedrich List does not refute Adam Smith. He generalizes Adam Smith. Newton’s equations work perfectly well at low speeds. At high speed, we need Einstein. Smith’s theories work perfectly well in a politically unified world. In a politically disunited world, we need List.

This is not to say that List’s ideas always work—they are tools to be used properly. What is important to understand is not the tools, but the framework behind them.

The fundamental idea of mercantilist trade policy is that the nation is a firm. The goal of this firm is to maximize the value of its assets—to improve the land and its people, and augment its treasury.

These are very different economic metrics than we use in these rotting United States. The statistics we use—“growth,” “inflation,” “unemployment,” etc—are artifacts of the axial age of 20th-century economics, the 1930s to the 1950s. Like many elements of our political language, these concepts did not exist before the New Deal. Yet our ancestors had trade, finance, and political economy—and seem to have done well at them, at least judging by the buildings they left us.

What does it even mean, for instance, to improve the people of a country? Well—has China not improved her people—along with the buildings and factories on her land? Certainly, but… how do you measure that? Regulatory changes post 1865 have severely complicated this econometric problem. But it is a problem of metrics, not of reality. Not everything real is measurable—yet the immeasurable remains no less real.

We can say that, in general, human beings are improved by work; and the best work to improve them is work at the limit of their human capacity for skill. This effect is not in any way measured by “growth,” and measured very imperfectly by “unemployment.” The value of buildings and factories can be measured by price. But even here, markets finds it hard to measure the beauty of a building. The immeasurable is everywhere, for those with eyes to see.

Once we start treating nations as firms, we have a huge variety of tools from the world of capitalism to diagnose and improve their ills. Corporate assets can also be hard to value. Cash flow, however, is impossible to fake.

If we equate the “value of the land and its people” with the common good, we see quickly that a trade policy which turns a profit (like China’s) is probably going to correlate better with the common good than a trade policy which makes a loss. There’s a reason we associate trade deficits with decaying countries—the same reason why we associate chronic losses with decaying companies.

A profit? Welcome to the most fundamental principle of mercantilist trade policy: for a mercantilist, the balance of trade is the P/L. A trade surplus is a profit. A trade deficit is a loss. President Trump, of course, seems to understand this instinctively. It’s quite possible that he understands trade better than all economics professors. Unfortunately, this may not mean he understands it all that well—the fundamental paradox of the second Trump administration, in all its greatness and retardation.

A mercantilist model

Let’s understand trade and import certificates by building a general trade model whose knobs can be tweaked either for the status quo, or for fully balanced trade.

We’ll start by market-neutral transformations of the status quo—rewriting existing economic activities while maintaining existing incentives, so that supply, demand and price do not change.

In the new system, there is zero trade between private entities in either nation. All trade and financial flows are intergovernmental.

You can’t buy a bottle of wine from France. You can only ask your government to buy you a bottle of wine from France. Imagine that this process is so frictionless that it is no more difficult than shipping the wine is now. You place the order with the External Revenue Service, which buys the wine from the government of France, which buys the wine from the vintner. It’s exactly the same as the current flow.

When you import, the government buys something from another government, then sells it to you. When you export, you make something, then sell it to the government, which sells it to another government.

This is all privately directed activity. The government does not employ any wine-tasters. You might even order a single bottle of Bordeaux from the website of a vineyard you like, and have it go through this whole process automatically for you, with no less friction than in shipping now.

However, a little bit of friction can solve one trade problem: tariff fraud. A tariff is charged on the declared value of goods, which is just a number somebody made up. Lowballing declared value is an old game in the import-export world.

However, if the External Revenue Service owns the goods on the way in, even for a moment, it has a choice. It could sell the wine to you, who ordered it. Or it could sell the wine, perhaps at a standard 15% markup, to some sharky bandit company which makes a living off finding underpriced goods—the tariff equivalent of a Harberger tax. Thus, the market itself detects fraud.

The empty wine bottle has not augmented the value of the land and/or people of the United States. So drinking Bordeaux instead of Cabernet is a net loss to the United States. If you see why, you may have had your first Trumpian mercantilist thought. (Alas, drinking Cabernet instead of Bordeaux is a net loss to any person of taste—but this purely hedonic cost does not appear on any balance sheet, and nor should it.)

At the level of the nation as firm, deciding whether the people of the US should drink Cabernet or Bordeaux when they order “the red, please,” is a buy-or-build decision. If you “build” your wine, you lose whatever other services the winemakers would have been providing—but you don’t send any money to France. If you “buy” it, you get to use these human assets for other things (they could be medical billing analysts, for instance), but you send a bunch of money to France.

Balancing mechanisms

Now, let’s introduce market-balanced-trade mechanisms. On the blockchain!

When a US firm sells $1 of goods to the government to export, it receives two things: $1 token, and 1 $ERC token—an export receipt. When a US firm buys $1 of imported goods from the government, it has to send $1, and 0 $ERC. We are still in the status quo. All economic activity will continue as it continues today.

Now, turn the knob so that an import requires 1 ERC token. Voila: balanced trade! The export receipt has become an import certificate.

Now, every imported dollar of goods must be matched by an exported dollar of goods. What will the price of ERC tokens be? The market will set it—at whatever level is needed to create fully balanced trade. No prediction of elasticity is necessary, nor is there any tariff formula.

Imagine a wildly asymmetric trade deficit, where exports are $1T and imports $4T. We impose this regimen. If exports stay the same, now we can only import $1T of goods (at the foreign price), because buyers will have to compete for 1T import certificates. Exports will not stay the same, however, because exporters are being subsidized by importers—since they get to sell their ERC tokens.

There are adjustable knobs all over this system. If you want half-balanced trade, you can charge only 0.5 ERC for an import certificate. If you want a tariff rather than an export subsidy, give the ERC tokens to the government, not the exporter. Or half. Etc.

This system is multilateral: it balances trade across all partners and all goods. If you want bilaterally balanced trade, use a different token for every trading partner. Trade can also be segregated into types—for instance, a traditional role of “less-developed” or colonial countries is to import manufactured goods, and export raw materials.

If the US doesn’t want to be Argentina with nukes and the global reserve currency, it can have separate tokens for manufactured goods and raw materials, so that we have to be separately self-sufficient in both. This again will happen instantly—if the switch is flicked instantly. With plenty of disruption, of course.

Flicking these kinds of switches instantly is actually quite crude. In general, because the structure of tariffs is intended to shift economic incentives, and thus economic structures, around trade, sudden change creates disruption—and Stonks Go Down.

If the knob is turned smoothly—phasing in restrictions over months and/or years—it matches the possible pace of structural change. No one can build a factory overnight. The incentive patterns should shift predictably over a long period of time.

Is a tariff working? A tariff/export subsidy/import certificate scheme works in two ways. In one, it is a consumption tax. In another, it creates effective structural change.

A tax is just a tax. An import duty of any kind on cacao or bananas or other tropical crops is just a tax. It is not going to make these crops grow in the US. The only thing it will do is to make people eat less chocolate.

Perhaps the worst way for managed trade to turn out is in the middle—when local companies can build cars, but only shitty cars. Latin America is peculiarly afflicted. The best way is to build an industry so strong that it can abandon trade protections. Again: mercantilism is a tool for success, not a recipe for success. There is no recipe.

The goal of mercantilism

But what is the goal of a tariff regime? The goal is simple—we already covered it.

At the level of the nation as firm, trade policy works in two ways. First, the nation turns a profit, which is good for its international position—and, more important, an exercise that helps keep the whole nation in efficient, profitable shape. Second, the nation preserves and enhances its human capital, its physical capital, and its intellectual capital, by demanding labor from its citizens. Value of the assets.

China sends huge quantities of products to America and in return receives pieces of green paper. The pieces of green paper are not the primary benefit to China. In fact, if America disappeared into the sea, the Chinese government might rationally choose to simulate America—producing the same amount of stuff, dumping it into the Pacific, and paying the factories with the same pieces of paper. Why? Because its assets are the Chinese people, and working hard technology is good for the Chinese people.

Just as exercise is good for a person, the exercise of production is good for people, for teams, and for societies. Especially in the age of AI, the primary function of the state will be the conservation and production of labor demand. When humans do not need to work, they need to be made to work—on tasks which fit their skills and aptitudes. Contrary to 20th-century dogma, skills and aptitudes are not infinitely adjustable.

Nike has 450,000 shoe-making workers in Vietnam. Suppose we set a narrow balance of trade on shoes—to import a dollar of shoes, you have to export a dollar of shoes. Given today’s structural realities, this is essentially just banning shoe imports. But unlike bananas, shoes actually can be made in America by Americans.

The result will be an enormous domestic demand for American shoes, which means a similar demand for American shoemakers—skilled, semiskilled, and/or unskilled. We might even need a million of them. Are there a million Americans who would be, in our daring parlance, improved, by jobs—not just jobs but careers, because good trade policy is structural, and has the lifespan of human beings and human organizations—as shoemakers?

And what happens when robots make better, cheaper shoes? Ultimately, in the 21st century, the state has no choice but to turn the productive economy into a videogame—a simulation. In a videogame, there are no natural monsters. All danger, all death, all difficulty, is artificial.

If artificial barriers to trade seem strange, they are only the first weird harbinger of a future of artificial difficulty—restricting technology to produce high-quality labor demand. Artificial difficulty is more useful in old technology—restoring the earliest professions that the Industrial Revolution destroyed. Such restrictions should be chosen by the quality of the labor demand they produce—for example, switching household goods from industrial production to artisanal production. That many people love to be artisans in their spare time helps us see what would fulfill them—certainly, no one ever wanted to be a medical billing analyst in their spare time.

For example, one form of trade barrier with interesting structural properties is a barrier to shipping. Suppose there were no financial controls, and anyone could bring goods from the Old World to the New World—but only on motorless sailing ships. This would recreate another type of American—the sailor. Who wouldn’t rather be an American sailor than an American medical billing analyst?

By 20th century standards, we are doing enormous damage to productivity and thus GDP. True—but by 20th century standards, AI is creating enormous amounts of productivity. Through economic restrictions that create artificial difficulty, we spend this productivity on making a better, more human life for everyone.