Category Archives: Economics

Tariffs Won’t Achieve America’s Goals

Over at Morning Consult, Iain Murray and I have an op-ed explaining why tariffs are ill-suited to achieving the Trump administration’s economic and foreign policy goals:

We often talk about trade between nations as if it is done through the medium of government. But “Mexico” does not trade with “America.” The actual trade going on is between individuals, to the tune of billions of transactions a year. That trade is by definition win-win. Individuals only give something up if they expect something better in return.

That means the term “trade deficit” is incredibly misleading. When Americans buy $592 billion worth of goods from the European Union, and the EU buys $500 billion worth of goods from the United States, both sides are better off. Americans have gotten $592 billion worth of value from trade with the EU, while sending only $500 billion of goods across the Atlantic. The “deficit” is actually a surplus of value. Both sides win.

When policymakers think of aggregates instead of individuals, confusion reigns. In this case, President Trump thinks hurting the economy will help it. As economist Joan Robinson famously said, if another country dumps rocks into its harbors, the proper response is not to dump rocks in your own harbors.

There are better ways. Taking the remaining rocks out of America’s harbors will only help the U.S. economy, and give it more resources to achieve its other policy goals.

Read the whole thing here.


Trump Trade Tariffs Hit Ford Motor Co. Hard

This is press statement on unintended, but not unforeseeable consequences of tariffs. Originally posted at

Ford Motor Company just announced layoffs in the midst of a reported $1 billion in tariff-related losses. Auto sales are down due to trade tariffs that President Trump imposed on metal and other car-related materials. Competitive Enterprise Institute trade policy expert Ryan Young says this sort of economic calamity is not surprising.

“President Trump’s push towards government-managed trade is starting to show its effects. New trade barriers have already cost Ford a billion dollars, and the company, already in the midst of a reorganization, is laying off employees.

“Trump’s new taxes on foreign goods are intended to stimulate American manufacturing, but they are having the opposite effect, just as economists across the political spectrum predicted. U.S. manufacturing output was already near a record high before the new tariffs, and did not need any help, especially of this counterproductive variety.

“President Trump should repeal his new tariffs, and Congress should pass legislation preventing him from causing further damage to the economy.”

Ryan Young coauthored a recent report explaining all the economic harms caused by trade tariffs and other barriers. See:

Analysis: Common Myths and Facts about Trade

Report: Traders of the Lost Ark

William Nordhaus, Paul Romer Win 2018 Economics Nobel Prize

Both of this year’s economics Nobel laureates have been on the short list for some time. Both are deserving, as David Henderson writes in The Wall Street Journal. Paul Romer is best known for his work on the nature of economic growth, and William Nordhaus for his environmental economics, as well as coauthoring, with fellow laureate Paul Samuelson, the 20th century’s most widely used economics textbook (a copy of the 18th edition sits on my shelf). In fact, my only surprise at hearing that Nordhaus won was that I mistakenly thought he already had.

Nominees are not always so deserving; the prize committee often uses the award to make a political statement. Paul Krugman, for example, the 2008 laureate, produced excellent scholarship earlier in his career on international trade. But he later ditched economics for political activism, to the point of famously switchingsides on policy issues based on which party proposed them. His strong partisanship almost certainly played a role in his prize—and its timing, announced in October of an election year.

More notoriously, the committee more or less had to give F.A. Hayek a prize due to the quality, quantity, breadth, and influence of his scholarship. But it felt the need to balance Hayek’s unfashionable pro-market views by co-awarding the 1974 prize to the anti-market Gunnar Myrdal, who publicly favored using eugenicsto improve the welfare state. Myrdal’s work led to more than 60,000 people being sterilized in his native Sweden, 90 percent of them women.

None of this should tar Nordhaus and Romer’s prize—I am not the only spectator who was surprised by the lack of political message their selections send. Their awards are merit-based, which speaks well both of their work and of the committee’s.

CEI scholars and fellow travelers will find much to disagree with in Nordhaus’ work—but in some ways he also echoes Julian Simon. The biggest criticism is that Nordhaus was an early advocate of a carbon tax. One sees its theoretical appeal to economists—it’s an attempt to create a missing market. But there are serious practical problems.

Nordhaus recently estimated that an ideal tax would be about $30 per ton of carbon emissions. But this sits inside a range of as low as $6 per ton to as high as $93 per ton. This 15-fold difference is a sophisticated way of saying “I have no idea.” Future estimates will also have to account for declining emissions in richer, and still-growing, countries without a carbon tax. And all this ignores public choice concerns. The politicians who would enact a carbon tax do not have strong incentives to act in the public interest, to put it mildly. Moreover, as Henderson notes, “the Nordhaus model shows that the cost of a policy to limit the temperature increase to only 2.7 degrees Fahrenheit by 2100 would have been $37.03 trillion—$16.4 trillion more than the cost of doing nothing after accounting for the damage done by carbon emissions.” So carbon tax opportunity costs are substantial, too.

But Nordhaus’ most famous article is brilliant, almost literally—it’s about light. It also captures the spirit of Julian Simon, as well as the larger reason why economics is worth understanding in the first place. In “Do Real-Output and Real-Wage Measures Capture Reality? The History of Lighting Suggests Not,” Nordhaus shows that macroeconomic statistics don’t capture the full spectrum of improvement in human well-being over history. His chosen example is artificial light.

Open fires were the dominant light source for most of human history. These often take a half hour for even an expert to build and require time-consuming wood foraging and chopping. The indoor air pollution wood and dung fires cause is still a leading cause of death in areas without electricity. Candles and fat- and gas-burning lamps came along later, and were a massive improvement. Electric light bulbs were even better. Even accounting for the fossil fuels they consume, these have a far lighter environmental footprint, are safer, and provide brighter, more consistent light with very low maintenance. Since Nordhaus’ 1996 article, the improvement has continued with LED technology that provides customizable color temperature and brightness, with less power consumption and longer bulb life. The process will almost certainly continue into the future.

Today’s better, safer light sources are also cheaper—much cheaper. Nordhaus calculates the first-generation compact fluorescent bulbs from 1992 were nearly 30,000 times cheaper, per lumen generated than the open fires our ancestors used. That’s not 30,000 percent, that’s 30,000-fold. And CFLs turned out to be a dud technology! Talk about first world problems. Current LED technology is even better. And this has improved the lives of more than just humans. Whales are grateful for the extinction of whale-oil lamps rather than themselves. Forests are grateful that campfires are today mostly an occasional novelty for wealthy city-dwellers, rather than the dominant global light source.

Nordhaus has often fallen into the blackboard economics trap that fells many a talented economist. But his article on the price of light teaches an important lesson that well deserves a Nobel. Official statistics understate human betterment. There is a vast tapestry of non-statistical improvements. Some of them can be quantified, such as the price of light over time. Others cannot, such as the incredible ease with which people can find music, movies, and books today, or the quality difference between a 1988 Camry and a 2018 Camry, accounting for inflation. There is a vast literature on this subject, and Nordhaus prominently influenced it.

Remember, human betterment is the core reason to study economics in the first place. People can’t become better off if they don’t know what causes the wealth of nations. GDP and other statistics are useful. But economists who focus too closely on them not only understate the amount of human betterment that sound economic policy enables, but they can misdirect their time and talents into areas that could possibly cause harm, not betterment.

Economist Don Boudreaux points to another Nordhaus article that makes a similar point about the importance of innovation and the entrepreneurial spirit that makes it possible: “nearly all—about 98 percent—of the benefits of technological innovation are captured, not by the entrepreneurs and businesses who introduce them, but by the general public.” The article, 2004’s “Schumpeterian Profits in the American Economy: Theory and Measurement,” invokes another name familiar in Competitive Enterprise Institute circles. It’s easy to see the billions of dollars that the Mark Zuckerbergs and Jeff Bezoses earn for themselves. By Nordhaus’ estimate, they create about 49 times more value than that for complete strangers such as you and me.

Paul Romer’s work, at its best, follows a similar theme. He is best known for his work on the nature of economic growth. It ties directly into what many economists, myself included, consider the most important graph of all time, pictured below.

People lived on the equivalent of three dollars a day for most of human history, from Mesopatamia through the Middle Ages and the Renaissance, up until about 1800 or so. Now, in the last one tenth of one percent of our species’ history, the average person in rich countries suddenly has fifty times that to live on. More than half of the world’s population, poorest countries included, now qualify as middle class. The proportion of people living in absolute poverty, defined as $1.90 per day or less, is now below ten percent for the first time in history. The absolute number of people in absolute poverty is declining, even as population increases. How the heck is this happening? Romer has helped improve our understanding of this important question.

Unlike cable news economists, Romer has an attention span longer than a fiscal quarter. This enables him to look at what the real engines of economic growth are, and what made today’s mass prosperity possible—and how it can continue to lift people out of poverty in the future.

Romer is not the first to say that technology, innovation, and change are the most important long-term factors driving growth. But many previous theorists, most prominently Robert Solow, thought economies tended to work towards a stable equilibrium state, until some outside technological shock, such as the electric lightbulb or the automobile, changes where that equilibrium point lies. The economy then lurches towards that new point, where it would remain mostly still until another shock comes along.

Romer argues that technological change is not an outside shock, it is built into the human condition. People innovate all the time, and always have—just look at the evolution of stone tools before anything like a modern economy ever existed. The proper concern for policymakers then, is enacting public policies that do not stifle the natural human impulse to change and innovate. Long-run institutional structures matter; whatever the Dow Jones Industrial Average did last quarter does not.

Romer prefers a government with a handsier approach than we do at CEI. But on those fundamentals, we very much agree. Property rights, freedom to trade, rule of law, limited corruption, and the usual Washington Consensus policies tend to work quite well. Countries with governments that more or less follow them are far more prosperous and free than those that don’t, as this year’s just-released 2018 edition of the Economic Freedom of the World report shows.

Romer’s shifting of technology in blackboard models from an exogenous variable to an endogenous variable seems like some ivory tower dispute of little consequence. But it is really about human nature, and human betterment. Despite our inborn tendency to innovate and exchange with each other, poverty is still our default state. A quick glance at the graph above makes this obvious enough. It is wealth and growth that need to be understood. One implication of Romer’s theory is that innovation, while impossible to predict, is very possible to suppress.

Regardless of one’s political leanings, the goal is nearly always the same: freedom and prosperity for as many people as possible. Nordhaus and Romer’s work, in their separate ways, tie into that common theme, and their Nobel reflects that. Congratulations to them both.

New NAFTA Could Have Been Much Worse

The new USMC (United States-Mexico-Canada) trade agreement isn’t very different from the old NAFTA (North American Free Trade Agreement), and that’s a good thing. Given the Trump administration’s emphasis on government-managed trade, it could have been much worse. Now President Trump can claim a political victory and hopefully turn his attention to non-trade issues, while actual trade policy remains mostly unchanged.

The agreement now goes to Congress. Passage is likely, though a possible party change in next month’s election could complicate matters. Incoming Mexican president Andrés Manuel López Obrador is the other political variable still in play. He has similar views to President Trump on foreign trade. But as a show of political good faith to current President Enrique Peña Nieto, he has indicated he will likely sign the agreement despite his misgivings. López Obrador’s populist reputation does not make this a guarantee, however. But Canada’s joining the agreement means the USMC Agreement will likely clear all political hurdles.

The 1,812-page agreement leaves intact the mostly tariff-free relationship between the U.S., Canada, and Mexico. It even has a few improvements, such as a slight liberalization of Canada’s dairy policy. U.S. agriculture policy will remain heavily subsidized and insulated from competition, however. Among the downsides are new wage and country-of-origin rules that will make cars more expensive for consumers and potentially disrupt carefully designed supply chains.

The negotiations also missed some significant reform opportunities. Oddly enough, this is actually a good thing. The current U.S. administration and the incoming Mexican administration both favor government-managed trade, so the timing was bad to open trade negotiations in the first place. While this is a sad commentary on contemporary politics, negotiators scored a victory by not letting their bosses make things significantly worse.

The most obvious missed opportunity is that recent U.S. tariffs against Canadian and Mexican steel and aluminum will remain in place, allegedly for national security reasons. As these new tariffs make themselves known throughout the supply chain, they will further increase car prices, plus other items such as houses and manufacturing equipment. There are other avenues for reforming these tariffs and the retaliations they caused, though they may have to wait until the next administration.

Hyper-complicated country-of-origin rules in the USMC Agreement run for 234 pages, and are similar to the ones in the old NAFTA. These are guides for figuring out how to apply tariffs to, say, a good designed in one country and built in a second country, from parts made in still other countries that are not even members of NAFTA/USMCA. This complexity could be avoided by simply doing away with tariffs altogether, but that was never going to happen in the current political environment. Again, the economy is better off simply for the new USMC Agreement not making things significantly worse.

Also troubling is a general NAFTA/USMCA ethos under which some countries determine other countries’ regulatory policies for them. This is generally due to trade-unrelated policies in trade agreements, mostly on labor, environmental, and intellectual property issues. Right now, it is the U.S. formulating Mexico’s regulatory policies, so there is little uproar about it, at least in the U.S. (indeed, this may have been included to increase support from labor union interests). But if the tables ever turn, this will quickly become a hot-button issue that could sink the agreement.

In short, NAFTA has a new name, but it’s still NAFTA. Pundits will just have to remember that USMC now means “United States, Mexico, and Canada” in addition to “United States Marine Corps.” There are a few marginal changes here and there in the USMC Agreement, some good and some bad. But a major bullet has been dodged between America and two of its largest trading partners. That the Trump administration is calling it a victory means that a major economic loss has been avoided for the time being. It would have been better to leave well enough alone, but under the circumstances, this may be about the best possible outcome.

Is There Such a Thing as Too Much Trade?

A common argument for free trade is that fewer trade barriers mean more trade. That argument is mostly true—there are a lot of deals people want to make but don’t, because trade barriers make them too expensive to be worthwhile. But there is more to the story.

The thinking goes like this: there is a point when enough is enough. After a certain amount of wheeling and dealing, people obtain the best possible mix of goods and services they can under the circumstances. Then they stop trading. The transaction costs of further action exceed the benefits. In a weird, non-Panglossian way, this would be the best of all possible worlds, because people could do no better.

Where exactly is that magical point? Nobody knows. But preventing people from trying to find out doesn’t do any good, and almost certainly does harm. And that’s the argument against trade barriers, even in a world with no trade (free trade in a trade-free world?). Trade barriers prevent that discovery process from getting people as close as they can get to that equilibrium point.

A skeptic could argue, correctly, that that equilibrium point will never be reached. It’s just a blackboard argument that doesn’t resemble a real-life economy.

Skeptics have another good point: that unknown, unachievable optimum is a moving target. And tariffs or not, the frictions of commerce are already easing up all the time anyway—online shopping, innovations in shipping technology, and peer-to-peer sharing services such as Uber and Airbnb are lowering transaction costs every year, making it easier than ever for people to trade. These innovations are constantly moving the equilibrium goalposts from one unknown point to another unknown point, neither of which can ever be reached anyway.

The transaction cost point in particular is especially important in today’s economy, as Michael Munger points out in his book “Tomorrow 3.0” (see my review here). But this is not an argument in favor of trade barriers. This is an example of what Harold Demsetz called the Nirvana Fallacy—real-world markets can’t achieve idealized blackboard perfection, therefore government intervention is necessary, or least indifferent. Real-world free trade will never achieve idealized perfection, therefore trade barriers matter less than free trade purists think.

This line of thinking does no one any good. As the famous Green Bay Packers coach Vince Lombardi pointed out, perfection is impossible. But in constantly chasing after it, you can achieve excellence. No realistic free trader argues that a tariff-free world would be perfectly efficient. But it would be better than the current trade regime, and is worth chasing. In doing so, we can achieve excellence as best we can in this non-blackboard world.

All this is a long-winded way of answering the question this post’s title asks. Is there such a thing as too much trade? Yes. And people should be free to try and find out where that point is.

For more, read the new CEI paper “Traders of the Lost Ark: Rediscovering a Moral and Economic Case for Free Trade” here.

A New Front in the Trade War: Overseas Private Investment

Tariffs get most of the press in today’s trade debate, and for good reason. Tariff rates under Trump have roughly doubled in less than two years, and more might be on the way. But tariffs are not the only way government tries to manage trade. Another is in finance. In particular, an agency called the Overseas Private Investment Corporation, or OPIC, might be doubling in size shortly, to a total portfolio size of $60 billion.

Unfortunately, on September 26 the House passed legislation to increase and rename OPIC. The BUILD Act, a bill that would do just that, was folded into must-pass 1,205-page legislation to reauthorize the Federal Aviation Administration. The House passed it on September 26, with the Senate to follow suit, likely before the end of the month, calendar permitting. Absent an effort to remove the BUILD Act provisions from the Senate version or during reconciliation, President Trump’s trade war will open a new front, hurting both the economy and America’s foreign policy interests to the tune of about $480 per U.S. household.

OPIC is basically an economic development agency. This sounds benign; it is not. In a 2015 paper, I pointed out that OPIC costs roughly $719,000 per job it supports. By my calculations from OPIC’s 2017 data, it has become more efficient, with 2017’s new projects costing $523,076 per job supported. Even with this improvement, OPIC-supported (they refuse to use the term “created”) jobs still cost nearly ten times per capita U.S. GDP. OPIC’s total portfolio size is currently a little more than $23 billion, with $3.8 billion of that coming in 2017.

OPIC operates by providing loan guarantees and other financial products for business projects in other countries, often to the benefit of U.S. businesses with the right political connections. As part of its reorganization plan, the Trump administration is proposing to fold a few similar agencies into OPIC, double its size, and change its name. A July 17 press release about the BUILD Act states:

The Administration supports H.R. 5105/S. 2463, the “Better Utilization of Investments Leading to Development (BUILD) Act of 2018,” which would consolidate the Overseas Private Investment Corporation (OPIC) and other development-finance programs into a reformed United States International Development Finance Corporation (USDFC).  This action will catalyze market-based, private-sector development and economic growth in less-developed countries and advance the foreign policy interests of the United States.

There are a lot of reasons the Trump administration views OPIC and the proposed USDFC as useful to its trade and foreign policy agendas. One of them is China. As OPIC states in a September 25 press release, after the BUILD Act was folded into the FAA reauthorization bill:

At a time when China is investing heavily in emerging markets under a state-directed model, the U.S. model offers an alternative for advancing development in a manner that is financially sound, adheres to high standards and avoids debt traps.

There is a lot to unpack from these statements. First, government-directed investment is not market-based development, as the administration alleges. OPIC is a government agency that transfers money from taxpayers to private businesses. OPIC makes the same mistake in its statement. As an alternative to one state-directed model, they offer another state-directed model.

Nor is this financially sound, as OPIC alleges. OPIC claims a profit on its investments, covering its own costs and reducing the federal deficit by $262 million last year. This is based on unusual accounting practices that are rarely used in the private sector, as my paper points out. OPIC may very well show a loss under the same accounting standards the private sector uses. But let’s be charitable and grant OPIC the $262 million in 2017 deficit reduction for the sake of argument. Its financial soundness claims are still weak, even with this artificial strengthening.

The reason is the old economist’s question, “as compared to what?” In other words, what are the other possible uses for the resources OPIC uses? What if OPIC’s portfolio had instead been invested by private investors? They have their own money at stake; OPIC does not. Private investment is meritocratic; OPIC admits its investments are politically driven. In many years, as much as 40 percent of its investments are earmarked for green investments, for example. With OPIC’s politicized incentive structure, its investment decisions are likely far less profitable than likely alternative options. OPIC does less with more, on purpose. Private investors take the opposite approach, also on purpose. The solution to this problem is not to double OPIC’s size.

The other big issue here is China. China has its own versions of OPIC and similar investment agencies, which also make politically-directed investments in strategic countries. It is in America’s interest to make similar investments to build alliances and relationships in the region, the thinking goes. Foreign policy concerns outweigh the financial cost.

That’s a subjective value judgment, but let’s grant it here for the sake of argument. The trouble is that the administration’s foreign policy goals are unclear. President Trump is ever-concerned about trade deficits, despite endless “Groundhog Day” meetings where his aides explain again and again that trade deficits don’t hurt the economy because people get something in return.

If successful—and remember, that’s a bigger “if” than in non-state-run investment models—OPIC’s foreign investments would likely increase net U.S. imports. This would add to America’s trade deficit. While this is neither good nor bad for America’s economic health, it contradicts President Trump’s long-held stance on trade deficits. It is unclear if he understands this, or if other concerns for him outweigh a likely increase in the trade deficit.

President Trump is also concerned about the Chinese government’s unfair trade practices, and rightly so. I explained in an earlier post why tariffs will not encourage China to liberalize its trade policies. The same argument applies here to government-directed foreign investments. Basically, we’re making strategically located foreign investments because China is, too.

The trouble is that China is unlikely to respond to increased U.S. foreign investment in its neighborhood by making completely unrelated reforms to its IP abuses, technology transfer, and government ownership policies. If it responds at all, it will likely be the same way it responds to higher tariffs—retaliation with more of the same. This will almost certainly not increase economic or political freedom in China, or improve its bad-faith trading behavior.

Given the corruption opportunities state-directed investments create, ramping up such activity by both the U.S. and China could actually make East Asia less stable, not more U.S.-friendly. Of all the ways to advance America’s foreign policy interests, doubling OPIC’s size and renaming it the USDFC is surely not on the list. This would likely be the case even if the administration had clear policy objectives.

OPIC is bad enough as it is. Doubling its size and rebranding it with a different name will not improve its policy outcomes, it will make them worse. Some of the investments will turn out to be boondoggles, wasting perfectly good capital that could have created more value in less politicized hands. This will dampen future economic growth. A bigger OPIC could well increase the trade deficit. And it almost certainly will not induce China to make needed trade reforms. It could also increase corruption in China’s surrounding countries, weakening potential U.S. allies. In short: bad idea. The BUILD Act provisions should be removed from the FAA reauthorization bill. OPIC should be closed, not doubled.

Tariffs and Opportunity Costs

Today’s unsubtle trade debate largely ignores a subtle, but vitally important concept: opportunity costs. Direct harms from tariffs are easy enough to point out. Steel and aluminum tariffs mean new buildings and cars cost more now, for example. But opportunity costs are more abstract. The economist Frederic Bastiat famously called them “the unseen.”

For example, NAFTA renegotiations could potentially impose a 2.5 percent tariff on cars that don’t satisfy certain new requirements. An affected $20,000 car would then cost an extra $500. Where before a family could have a new car and $500 worth of other goods of their choosing, the administration prefers that they have only the car. And not a different or better car, mind you. The same car, just more expensive.

What value could the family have gotten from that lost $500? That is the tariff’s opportunity cost. Maybe it’s some new clothing and shoes for kids who just won’t stop growing. Maybe it’s replacing that beat-up dinette or recliner. Maybe it’s a couple of nice date nights for the parents, and some spending money for the neighbor kid who babysits. Maybe it’s a gift to charity. Not even the world’s most talented statisticians can suss out exactly what those lost opportunities might be.

The Trump administration’s new tariffs are a multi-billion dollar tax increase. Think of the sheer number of those little everyday decisions never made because of the money people spend on those tariffs instead of other goods. They add up to a big impact, even if they will remain forever unseen.

Another example. Many tariffs are part of a larger policy goal of preserving manufacturing jobs. A lot of people don’t know this, but service-sector jobs tend to pay more than manufacturing jobs. If you don’t believe me, the data are clear, and they are here. Manufacturing workers make about a dollar per hour less than the average worker. The difference adds up to about $2,000 per year for a full-time worker.

Using tariffs to artificially prevent manufacturing workers from migrating to higher-paying jobs hurts working class families. Mortgage payments are harder to keep up with than they need to be, and maybe that family vacation won’t happen after all. Those are opportunity costs that tariff supporters need to account for, and mostly don’t.

Relatedly, people who think America’s manufacturing sector is in decline should be heartened to know that manufacturing output was already at or near an all-time high before any of the new tariffs took effect. Moreover, this was accomplished with fewer workers than in years past. So the country gets staggering manufacturing output and a couple million peoples’ time and talent freed up for other, additional pursuits. Not one or the other, both. People today are creating and buying more with less than before, which is a good thing. Public policies that accomplish the opposite, such as tariffs, should be rejected by dynamists and nostalgists alike. It is better to create opportunities than prevent them.

Back to that $2,000 pay difference. Compensation is tied to productivity; this is why teenagers almost always make less than mid-career workers. When tariffs artificially keep workers in less productive jobs, as they are currently doing with the steel and aluminum industries, people’s time and talents aren’t being put to their most valuable uses, creating opportunity costs. One of those workers in a legacy job could have an idea in her head for a new product or her own business that could pay more and create more value for people than her current job does. She deserves the chance to give it a go. Maybe an existing startup is having trouble finding enough workers because tariffs and other policies induce people to stay put at the steel mill. That is consumers’ loss, and it doesn’t need to be.

If there were fewer tariffs and other industrial policies, the pay difference between manufacturing and other sectors would be smaller. Think of it like this: that $2,000 pay gap encourages workers to leave manufacturing jobs. As they do so, each additional service job pays progressively less until average pay is equal to manufacturing. At that point, there is no point in leaving manufacturing, since the money is the same.

At the same time, manufacturing employers would increase their pay to prevent workers from leaving, up until it matched service sector pay. The two sector’s pay would meet somewhere in the middle of where they are now. Economists call this converging tendency the law of one price. Tariffs get in the way of that process, creating still more opportunity costs. That wage difference is a sign that some people could be doing better, but something is costing them the opportunity—such as tariffs.

Tariffs don’t account for all of the manufacturing-service pay gap, mind you. Subsidies, tax breaks, and regulations can have similar effects. Training for new skills isn’t free, and even simple inertia matters, too. And those last two aren’t necessarily bad things. For some people, learning a different skillset is more trouble than it’s worth. Some people simply enjoy their current job more than they would an extra dollar an hour. They are not wrong. They have valid personal preferences that should be respected. But they should at least have the choice. Tariffs and other policies substitute the President’s preferences for workers’ preferences, and account for a chunk of the manufacturing-service sector pay gap, and all the lost opportunities that represents.

These types of opportunity cost arguments are impossible to explain in soundbite form. They will never appear in a meme, or a cable news shoutfest. There are no data for opportunity costs, and they don’t make for compelling news stories. I think it was P.J. O’Rourke who once pointed out the difficulty of holding a press conference in front of a building that was never built, that doesn’t house a business that was never founded, while not flanked by workers who were never hired, and are not making a product that was never invented. Even so, opportunity costs matter. They affect real people, and must be part of the trade debate.

For more arguments that deserve a place in the trade debate, read Iain Murray’s and my “Traders of the Lost Ark” study here.