Campaigning for the presidency, Donald Trump made the North America Free Trade Agreement the centerpiece of his attack on globalization. NAFTA, he said, was “the single worst deal ever approved.” Dismantling it quickly became key to his plan to Make America Great Again.
To be fair, Trump was far from the first to condemn the deal, which has long been used as a scapegoat for all that is wrong with the U.S. economy. For instance, during the 2008 presidential campaign, the agreement was regularly assailed by the two major Democratic Party aspirants. Barack Obama claimed it helped businesses at the expense of American workers, and hence was responsible for jacking up unemployment. Hillary Clinton foreswore any new multilateral trade agreements and vowed that she would toughen what she said was insufficiently pro-American enforcement of NAFTA.
The fact that it was taken over the finish line by the administration led by her husband, who lobbied hard for its adoption, likely made it harder for Clinton to pander to protectionist voters and politicians, during either that campaign or the one she ran in 2016. Indeed, on December 8, 1993, after several years of three-nation parleying and political wrangling in Washington, President Bill Clinton had signed NAFTA into law. It came into force on January 1, 1994, with the objective of gradually eliminating most tariffs on products traded between the United States, Mexico, and Canada by January 2008. Notably, it was the first major U.S. trade deal with a poor country since the trade liberalization that started immediately after World War II.
The debate that preceded the adoption of NAFTA, no less than the one occurring today, was passionate. In his book Pop Internationalism (1996), economist Paul Krugman said, “Not since the Smoot-Hawley tariff has trade legislation produced such a bitter polarization.” He added: “The intensity of this debate cannot be understood in terms of the real content or likely consequences of the agreement, nor is the debate’s outcome likely to turn on any serious examination of the evidence.”
A Deal that Was Oversold by Everyone
The anti-NAFTA crowd back then argued, using a term coined by 1992 presidential candidate Ross Perot, that the agreement would create a “giant sucking sound” of jobs going south—roughly 5.9 million, Perot estimated—thanks to unscrupulous U.S. manufacturers taking advantage of the cheap labor of Mexicans. This opposition was rooted in a populism that persists today and is expressed vividly by our current President. Among the desires of Perot’s populists then, and Trump’s now, is to stop the ongoing transformation of our country into a service economy. It is a movement whose champions are oblivious to the fact that this transformation is driven far more by technological innovation than by trade. Their consistent and quixotic belief is that this transformation can be stopped by erecting trade barriers against imports from low-wage countries.
During the initial NAFTA debate, populists insisted that preserving the small (4 percent) tariffs that the U.S. imposed on Mexican manufacturing imports—along with keeping somewhat higher tariffs on a few agricultural products and a handful of quantitative restrictions—would somehow stop the percentage decline in U.S. manufacturing jobs—a trend that had started not in the 1990s, but in the 1950s. They didn’t realize that if low wages in Mexico had held such great appeal for U.S. industry, that giant sucking sound would have been heard decades before NAFTA. With pre-NAFTA tariff rates already being quite low, U.S. firms would have already moved to Mexico.
On the other side, the agreement’s defenders claimed that a $6.2 trillion unified North American market and a booming post-NAFTA economy in Mexico would boost U.S. exports to our southern neighbor. That export boost, we were told, would create hundreds of thousands of additional jobs here in the United States. President Clinton declared that NAFTA would create 200,000 American jobs in its first two years and a million jobs in its first five years. These would be high-paying jobs that would do wonders for American “competitiveness.”
The truth is that the threatened pains and promised benefits from the agreement were both exaggerated.
Reality Is Much More Boring
Study after study, including ones from the U.S. International Trade Commission and the International Monetary Fund, showed that while NAFTA resulted in a significant increase in total trade, a surge in cross-border investment, and the creation of a North American supply chain, it also resulted in small but significant national gains, with intense (but very small overall) competitive disruptions for certain U.S. industries. There is also some evidence that it helped make the Mexican economy a bit more stable. Finally, NAFTA unquestionably consolidated the position of Canada and Mexico as the United States’ largest trading partners. In 2016, U.S. companies exported $262 billion to Mexico and more than $320 billion to Canada alone.
It was to be expected. Canada and the United States were already close trading partners. As my colleague at the Mercatus Center, Daniel Griswold, noted recently: “The U.S. and Canadian auto sectors have been integrated since the U.S.-Canadian Automotive Products Agreement of 1965. The two nations implemented the U.S.-Canada Free Trade Agreement in 1988, which became NAFTA when Mexico joined in 1994.”
In addition, prior to NAFTA’s adoption, tariffs between the three countries, and between each of them and countries not party to the agreement, were already quite low thanks to eight rounds of multilateral trade negotiation under General Agreement on Tariffs and Trade (GATT). This is not to say that it wasn’t worth lowering them further, or eliminating them altogether. It does mean we shouldn’t be surprised that NAFTA’s final impact (positive or negative) on the American economy was modest.
Just prior to NAFTA, the U.S. market alone was $5.5 trillion. The Mexican market would only add $0.2 trillion (3.6 percent of the U.S. GDP) and the Canadian market would add $0.5 trillion (9 percent of the U.S. GDP). That’s not a major change, especially considering that U.S. exporters already had significant access to the two markets before the agreement was adopted.
Not surprising, either, is that NAFTA didn’t have much impact on total employment in the three countries. As economists know, trade—like all competition, whether domestic or international—causes particular jobs to disappear, particular companies to go out of business, and particular sectors of the economy to shrink. But over the long run trade doesn’t affect the total number of jobs or the overall rate of employment or unemployment.
Job Losses from Imports Are Compensated for in Other Ways
There are a few reasons for this conclusion. First, workers, capital, and other resources will shift within the domestic economy, replacing the jobs that have been eliminated by import competition with new jobs elsewhere in the country.
This shift in resource-allocation happens in many ways. One channel is that foreigners who are enriched with U.S. dollars earned by selling imports to us will send those dollars back to the United States as demand for U.S. exports. Higher exports induced by higher imports require prompt U.S. producers to ramp up their production—and, hence, employment—to meet this new demand.
Another way in which dollars return to the United States is when foreigners invest their dollars in U.S. treasury bills and other dollar-denominated financial instruments, as well as investing directly in U.S.-based businesses. When a foreign firm invests its dollars in America, our economy grows, as does employment in those enterprises in which the investments occur. Think about a Japanese automaker building and operating a plant in the United States. That plant employs U.S. workers.
Finally, consumers benefit from the lower prices on imported and import-competing consumer goods. These consumer savings become extra cash. Some of this extra cash is saved and invested in America’s economy—hence there are more jobs where these investments occur. Some of this extra cash is spent on domestically produced goods and services, too, which creates employment gains in these sectors.
But also overlooked is the fact that manufacturers are consumers as well. In fact, our biggest U.S. importers are also our largest exporters. These imports are raw materials, intermediate inputs, and capital machinery whose low prices help U.S. manufacturers to produce at lower cost. Lower production costs, in turn, mean higher profits, increased appeal to investors, and ultimately more employment in those sectors that benefit from these low-priced, imported inputs. The bottom line is that focusing merely on the particular jobs that are lost because of imports ignores the offsetting job-creation made possible by trade.
Aggregate employment is also unaffected by international trade because of other counterbalancing forces, such as monetary policy and adjustments in foreign exchange rates. In his 2009 book Mad About Trade, Dan Griswold wrote:
If a surge in imports did cause widespread layoffs in certain sectors, the resulting increase in unemployment would push the federal reserve to tilt toward a looser monetary policy and lower interest rates to stimulate the overall economy. Increased imports would also have the effect of pumping more dollars into international markets, causing the dollar to depreciate in foreign currency markets. A weaker dollar in turn would make U.S. exports more attractive, stimulating employment in export sectors while dampening demand for imports, offsetting initial job losses.
What Freer Trade Actually Accomplished
The byproducts of shifts in resources induced by freer trade are threefold.
First, jobs destroyed by imports are replaced by new and better jobs in relatively more productive industries, resulting in increased overall economic efficiency of the national economy as a whole. Second, there is better exploitation of economies of scale thanks to larger and freer markets. And third, there is greater competition between producers, leading to reduced inefficiencies and more innovation. These byproducts combine to make clear that, while NAFTA had no effect on overall employment, it did indeed make the North American labor force—including that in the United States—more productive.
But what about the argument we hear today, used against China, that when firms and workers in high-wage countries compete with firms and workers in low-wage countries, wages in high-wage countries must fall? Those who make this argument fail to understand why wages in countries like the United States are higher than in countries such as Mexico. Workers in high-wage countries are more productive, on average, than are workers in low-wage countries. (This greater productivity is caused by the greater amounts of capital—including human capital and infrastructure—in operation within high-wage countries.)
Yes, the typical American worker was paid multiple times what the typical Mexican worker was paid. But because the typical American worker also produced multiple times more than did the typical Mexican worker, Mexican workers weren’t an especially attractive bargain. Mexican workers’ average pay was much lower than that of Americans in order to compensate employers for Mexican workers’ much lower productivity.
The significance of this becomes clear if we highlight the productivity differential rather than the pay differential. How worried would Americans have been in 1993 if someone had been warning them that, under a prospective free trade agreement with Mexico and Canada, high-productivity American workers would have to compete with low-productivity Mexican workers?
Not surprisingly, there’s no evidence—none—that NAFTA depressed Americans’ wages. The inflation-adjusted average hourly earnings of production and nonsupervisory workers in the United States rose rather steadily in NAFTA’s first 20 years in force. These wages—the ones that were thought to be most at risk under the agreement—were about 13 percent higher, in real terms, in 2014 than they were back in 1994. (For an interesting review of the literature on this issue, see a 2014 paper published by the Peterson Institute on “NAFTA at 20: Misleading Charges and Positive Achievements.”)
While NAFTA was falsely accused of causing many economic ills (for example, depressing total U.S. employment and wages), it truly did play a significant role in reshaping North American economic relations. The result was not only impressive further economic integration, but also the creation of strong, large, and very efficient North American supply chains.
For instance, in the 1990s, many U.S. companies and foreign companies operating in the United States shifted their sourcing of raw materials and intermediary goods (such as auto parts) from Asia to Mexico. Some academics have argued that the peso crisis of 1994-1995 prompted this move as Mexican labor markets became relatively cheaper, which if added to the fact that a move to Mexico lowered both transportation costs and supply-chain risk, made Mexican workers an attractive bargain. This ensured that more of domestic companies’ production remained in the United States, as we saw with the automotive industry. It also enhanced the advantages that foreign companies enjoyed by transitioning their production from their home countries to North America, which thereby increased their reliance on this new North American supply chain. Given that America has numerous free-trade agreements with multiple countries around the world, these foreign companies also found it advantageous to export from the United States as opposed to exporting from their own country.
Did It Need Updating? Yes, But . . .
After 25 years, the pact did need to be modernized and fine-tuned, mostly as a result of the impressive technological advancements and the explosion of digital trade that we’ve witnessed since the mid-1990s. However, the biggest reform NAFTA needed was more liberalization, such as the removal of Canada’s dairy tariffs. The United States also kept a few small protections of its dairy and peanut markets. A revised NAFTA also should have included incentives for Mexico to privatize its state-owned enterprises.
There was a broad consensus on this, which is why much of the modernization that was sensible for NAFTA—such as the liberalization of Canada’s dairy market—was included in the Trans-Pacific Partnership, the prospective trade agreement among the three NAFTA nations plus nine other Pacific Rim countries. Unfortunately, once inaugurated, President Trump promptly withdrew from the TPP—a move that postponed many of the worthwhile upgrades to NAFTA. This action fits perfectly with the rest of the administration’s behavior on trade, and it reflects its desire for a radical reversal of U.S. trade policy. As a result, and instead of pursuing more liberalization, Trump labelled NAFTA “a disastrous deal” and threatened to withdraw from it unless Canada and Mexico agreed to his demands.
The end result, the United States-Mexico-Canada Agreement (USMCA)—the framework for which was announced on October 1 in Washington—is on balance a step backward from what an upgrade of NAFTA should have looked like. The new trilateral pact, which has not yet been submitted to Congress for approval, leaves in place many of the existing protectionist policies that are part of the original NAFTA, like U.S. sugar subsidies and American and Canadian dairy protection of domestic markets. And what liberalization there is in the USMCA is minor, especially when compared to what was already agreed to in the TPP.
The Cato Institute’s Scott Lincicome calculates that Canada agreed to open its dairy market by 0.34 percent more in the USMCA than in the TPP. Moreover, Canada did agree on October 1 to open its wine market a little further than it did in the TPP. According to the Office of the U.S. Trade Representative, this wine agreement is in exchange for the United States’ providing “new access to Canada for dairy, peanuts, processed peanut products, and a limited amount of sugar and sugar containing products.”
The truth of the matter is that, for all the complaints from Mr. Trump about tariffs, there wasn’t much room for progress on that front between Canada, Mexico, and the United States. According to data from the World Trade Organization, under NAFTA, all U.S. exports to Mexico face tariffs of zero percent. In addition, all non-agricultural U.S. exports to Canada enter that country duty-free. And despite all the talk about that pesky 270 percent Canadian tariff on U.S. dairy products, 97 percent of U.S. agricultural exports to Canada are duty-free.
Another good change is the increase in the de minimis thresholds, meaning that the value thresholds for importing goods without having to pay duties have been raised in Mexico and Canada. These higher thresholds will especially benefit small businesses, as they tend to conduct a larger share of transactions online. In addition, as Dan Ikenson of the Cato Institute notes, “the USMCA sets out reasonable rules in its Digital Trade chapter,” such as a prohibition on governments imposing localization requirements or any particular data architectures that reduce the efficacy of digital services.
Not that Ikenson is a fan of the USMCA over all. In fact, he considers it a disaster compared to the TPP. “The only certainty,” he writes, “is that the USMCA is better than a U.S. withdrawal from NAFTA without a replacement agreement.” That’s because, in spite of a few improvements, the overwhelming majority of the deal is more protectionist than the TPP, and possibly a wash or a slightly negative deal compared to the old NAFTA.
We can be relieved that the Trump administration set aside its least reasonable demands, which included: a minimum of 50 percent mandatory U.S. content on automobiles entering the United States to benefit from the new NAFTA duty-free treatment; a ban on student visas for Chinese nationals; and an every-five-year sunset clause. It’s nevertheless true that if the United States hadn’t dropped these poison pills, we likely wouldn’t have had this new deal—a deal filled with lots of destructive provisions.
In particular, the USMCA leaves in place the steel and aluminum tariffs. As a result, retaliatory tariffs from Canada and Mexico are also remaining in place. The disastrous effects of the Trump import tax have been felt throughout the metal-consuming industries in the United States and have led to over 34,000 requests for exemption.
The auto section of the deal is another protectionist provision. For automobiles to enter the United States duty-free from Canada or Mexico, at least 75 percent of their content must originate in the United States, Mexico, or Canada. That’s up from the current 62.5 percent. In addition, the deal ties duty-free treatment of automobiles to the requirement that at least 40 percent of each vehicle be produced by workers who are paid at least $16 per hour. These requirements will increase the cost that domestic or foreign automakers producing in the United States must incur to manufacture automobiles. In turn, the prices that Americans pay for automobiles will also rise.
The USMCA is also the perfect illustration of this administration’s utter lack of basic trade understanding. According to the President’s logic, if you make it more expensive to import parts into the United States from Mexico and Canada, U.S. automakers will have to source more of their parts here. This sourcing of more parts in America, he believes, will create jobs. Not likely. For one thing, the President refuses to acknowledge either that the decline of manufacturing jobs long predates NAFTA or that U.S. manufacturing output is nevertheless near an all-time high. A look at the decline in manufacturing employment shows no signs of having accelerated or slowed as a result of NAFTA. Therefore, this new deal will not change this trend.
Raising the cost of producing cars in the United States is hardly a sure-fire recipe for more jobs. Some auto producers might change their supply chains to conform to the new trade deal, but others might decide instead to pay the current 2.5 percent tariffs on imported parts. Either way, the resulting higher production costs—and higher prices in showrooms—will cause fewer automobiles to be produced in the United States.
Finally, the President fails to understand that the future of the U.S. auto industry lies in exporting. But of course raising the cost of producing cars in the United States makes it more difficult for companies producing cars within the United States to export them. This reduction in the U.S. automobile industry’s competitiveness will produce the effect he doesn’t want: It will increase the offshoring of some sectors of the auto industry, many of which will turn to Asia.
The Replacement’s Better Than Nothing—But Still Not That Great
The biggest political issue with trade is that the workers who lose their jobs as a result of competition are visible, while the millions whose lives are made better off, the innovation, and the new and more productive jobs resulting from it, are mostly unseen. It is easy to blame trade for eliminating jobs, when labor-saving innovations that result in higher wages are mostly the culprit. Behind this impulse is the persistent, and absolutely misguided, belief that trade with other countries is a zero-sum game—where only one country is the winner, and the winner takes all the benefits at the expense of its trading partners.
For all of these reasons, the last quarter-century has seen NAFTA serve as the scapegoat of the protectionists—who obsess over the discrete harms while ignoring the much larger gains—and the demagogues like President Trump who are all too willing to assert ignorant facts about trade and make promises to naïve Americans that will never materialize.
The end result: a more protectionist deal, the USMCA, that was agreed to by the three countries at the cost of a fractured relationship with our closest trading partners. It’s a deal that will facilitate offshoring. It will make it harder to produce cars here, and harder to export them to foreign car-buyers. That’s hardly a win, but it has the merit of putting an end to the uncertainty brought on by the constant threats of an eminent NAFTA withdrawal.
That is, of course, if this deal wins final approval from Congress. With the Democrats now taking control of the House of Representatives, the deal is once again in jeopardy. First, the Democrats who have expressed skepticism about the deal could block it, which might well infuriate President Trump, who may then decide to go back to his original plan and withdraw from NAFTA. This would be terrible for everyone, as tariffs in the three countries would go back up to levels not seen for almost 25 years.
The second scenario is slightly better but not good, either. In order to get his deal through and still claim a win, Mr. Trump may be willing to give in to the Democrats on their demands for more stringent labor and environmental protections. This will increase the cost of doing business in the United States and heighten the incentive to offshore production even further. Not a great prospect.
 The text of NAFTA was negotiated by the Bush 41 administration. President Clinton lobbied Congress to get the implementing language passed. He did negotiate labor and environmental side agreements that were not part of the NAFTA text.
The American economy lost its competitiveness in part because of a trade-policy shift inaugurated by NAFTA.
We need to keep the true justifications for free trade in mind as we assess NAFTA.
Who could quarrel with a more stable Mexico?
The tiresomely familiar “decline” trope—it used to come mostly from the political Left, but now comes from many on the Right—is as unsubstantiated as ever.