The North American Free Trade Agreement (NAFTA) is a pact eliminating most trade barriers between the U.S., Canada and Mexico that went into effect on January 1, 1994. Some of its provisions were implemented immediately; others were staggered over the following 15 years. Today, just over a month into its 24th year, NAFTA appears to be on its last legs. U.S. President Donald Trump repeatedly attacked it during his campaign, and two days after his inauguration, he promised to begin renegotiating the agreement, saying he expected "a very good result for Mexico, for the United States, for everybody involved."
Failing that, he has previously said he would pull out of the bloc – a relatively simple process, according to article 2205 of the treaty: "A Party may withdraw from this Agreement six months after it provides written notice of withdrawal to the other Parties. If a Party withdraws, the Agreement shall remain in force for the remaining Parties." Experts disagree about whether Trump would need Congress' approval to abandon the agreement.
Why do Trump and many of his supporters see NAFTA as "the worst trade deal maybe ever," when others see its main shortcoming as a lack of ambition – and the solution as yet more regional integration? What was promised? What was delivered? Who are NAFTA's winners, and who are its losers?
|Skip to section|
|1. United States||2. Mexico|
|3. Canada||4. China, Tech and the Crisis|
What Did NAFTA Accomplish?
NAFTA's immediate aim was to increase cross-border commerce in North America, and in that respect it undoubtedly succeeded. By lowering or eliminating tariffs and reducing some nontariff barriers, such as Mexican local-content requirements, NAFTA spurred a surge in trade and investment. Most of the increase came from U.S.-Mexico trade, which totaled $481.5 billion in 2015, and U.S.-Canada trade, which totaled $518.2 billion. Trade between Mexico and Canada, though by far the fastest-growing channel between 1993 and 2015, totaled just $34.3 billion.
|Trade volumes (million USD)|
|Channel||2015||1993||Nominal increase||Real increase*|
|*Adjusted for inflation using BLS core CPI; source: Mexican Embassy in Canada|
It is probably safe to give NAFTA at least part of the credit for doubling real trade among its signatories. Unfortunately that's where the easy assessments of the deal's effects end.
From 1993 to 2015, the U.S.'s real per-capita gross domestic product (GDP) grew 39.3% to $51,638 (2010 USD). Canada's per-capita GDP grew 40.3% to $50,001, and Mexico's grew 24.1% to $9,511. In other words, Mexico's output per person has grown more slowly than that of Canada or the U.S., despite the fact that it was barely a fifth of its neighbors' to begin with. Normally one would expect an emerging market economy's growth to outpace that of developed economies.
Can We Know?
Does that mean that Canada and the U.S. are NAFTA's winners, and Mexico is its loser? Perhaps, but if so, why did Trump debut his campaign in June 2015 with, "When do we beat Mexico at the border? They're laughing at us, at our stupidity. And now they are beating us economically"?
Because, in a way, Mexico does beat the U.S. at the border. Prior to NAFTA, the trade balance in goods between the two countries was modestly in favor of the U.S. Today Mexico sells close to $60 billion more to the U.S. than it buys from its northern neighbor. NAFTA is an enormous and enormously complicated deal; looking at economic growth can lead to one conclusion, while looking at the balance of trade leads to another.
Jobs (skip to top)
When Bill Clinton signed the bill authorizing NAFTA in 1993, he said the trade deal "means jobs. American jobs, and good-paying American jobs." His independent opponent in the 1992 election, Ross Perot, had warned that the flight of jobs across the southern border would produce a "giant sucking sound."
At 4.8% in January, the unemployment rate is lower than it was at the end of 1993 (6.5%). It fell steadily from 1994 to 2001, and while it picked up following the tech bubble's burst, it did not reach its pre-NAFTA level again until October 2008. The fallout from the financial crisis kept it above 6.5% until March 2014.
Finding a direct link between NAFTA and overall employment trends is difficult. The partially union-funded Economic Policy Institute estimated in 2014 that 851,700 net jobs had been displaced by the U.S.'s trade deficit with Mexico, which amounted to 0.6% of the U.S. labor force at the end of 2013. In a 2015 report, the Congressional Research Service (CRS) said that NAFTA "did not cause the huge job losses feared by the critics." On the other hand, it allowed that "in some sectors, trade-related effects could have been more significant, especially in those industries that were more exposed to the removal of tariff and non-tariff trade barriers, such as the textile, apparel, automotive, and agriculture industries."
NAFTA's implementation has coincided with a 30% drop in manufacturing employment, from 17.7 million jobs at the end of 1993 to 12.3 million at the end of 2016.
Whether NAFTA is directly responsible for this decline is difficult to say, however. The automotive industry is usually considered to be one of the hardest-hit by the agreement; yet although the U.S. vehicle market was immediately opened up to Mexican competition, employment in the sector grew for years after NAFTA's introduction, peaking at nearly 1.3 million in October 2000. Jobs began to slip away at that point, and losses grew steeper with the financial crisis. At its low in June 2009, American auto manufacturing employed just 623,000 people. While that figure has since risen to 948,000, it remains 27% below its pre-NAFTA level.
Anecdotal evidence supports the idea that these jobs went to Mexico. Wages in Mexico are a fraction of what they are in the U.S. All major American car makers now have factories south of the border, and prior to Trump's twitter campaign against offshoring, a few were openly planning to ship more jobs abroad. Yet while the job losses are tough to deny, they may be less severe than in a hypothetical NAFTA-less world.
The CRS notes that "many economists and other observers have credited NAFTA with helping U.S. manufacturing industries, especially the U.S. auto industry, become more globally competitive through the development of supply chains." Carmakers did not move their entire operations to Mexico; they now straddle the border. A 2011 working paper by the Hong Kong Institute for Monetary Research estimates that a U.S. import from Mexico contains 40% U.S. content. For Canada the corresponding figure is 25%. Meanwhile it is 4% for China and 2% for Japan.
While thousands of U.S. auto workers undoubtedly lost their jobs as a result of NAFTA, they might have fared worse without it. By integrating supply chains across North America, keeping a significant share of production in the U.S. became an option for car makers. Otherwise they may have been unable to compete with Asian rivals, causing even more jobs to depart. "Without the ability to move lower-wage jobs to Mexico we would have lost the whole industry," UC San Diego economist Gordon Hanson told the New York Times in March 2016. On the other hand, it may be impossible to know what would have happened in a hypothetical scenario.
Garment manufacturing is another industry that was particularly hard-hit by offshoring. Total employment in the sector has declined by nearly 85% since NAFTA was signed, but according to the Commerce Department, Mexico was only the sixth largest source of textile imports from January to November 2016 ($4.1 billion), behind China ($35.9 b), Vietnam ($10.5 b), India ($6.7 b), Bangladesh ($5.1 b) and Indonesia ($4.6 b). Not only are none of these other countries members of NAFTA – none has a free trade agreement with the U.S.
An important point that often gets lost in assessments of NAFTA's impacts is its effects on prices. The Consumer Price Index (CPI), a measure of inflation based on a basket of goods and services, rose by 65.6% from December 1993 to December 2016, according to the Bureau of Labor Statistics (BLS). During the same period, however, apparel prices fell 7.5%. Still, the decline in garment prices is no easier to pin directly on NAFTA than the decline in garment manufacturing.
Because people with lower incomes spend a larger portion of their earnings on clothes and other goods that are cheaper to import than to produce domestically, they would probably suffer the most from a turn towards protectionism – just as many of them did from trade liberalization. According to a 2015 study by Pablo Fajgelbaum
and Amit K. Khandelwal, the average real income loss from completely shutting off trade would be 4% for the highest-earning 10% of the U.S. population, but 69% for the poorest 10%.
Part of the justification for NAFTA was that it would reduce illegal immigration from Mexico to the U.S. The number of Mexican immigrants – of any legal status – living in the U.S. nearly doubled from 1980 to 1990, when it reached an unprecedented 4.3 million. Boosters argued that uniting the U.S. and Mexican markets would lead to gradual convergence in wages and living standards, reducing Mexicans' motive to cross the Rio Grande. Mexico's president at the time, Carlos Salinas de Gortiari, said the country would "export goods, not people."
Instead the number of Mexican immigrants more than doubled – again – from 1990 to 2000, when it approached 9.2 million. According to Pew, the flow has reversed, at least temporarily: 140,000 more Mexicans left the U.S. than entered it from 2009 to 2014, likely due to the effects of the financial crisis. One reason NAFTA did not cause the expected reduction in immigration was the peso crisis of 1994-1995, which sent the Mexican economy into recession. Another is that reducing Mexican corn tariffs did not prompt Mexican corn farmers to plant other, more lucrative crops; it prompted them to give up farming. A third is that the Mexican government did not follow through with promised infrastructure investments, which largely confined the pact's effects on manufacturing to the north of the country.
Trade Balance and Volume
Critics of NAFTA commonly focus on the U.S.'s trade balance with Mexico. While the U.S. enjoys a slight advantage in services trade, exporting $30.8 billion in 2015 while importing $21.6 billion, its overall trade balance with the country is negative due to a yawning $58.8 billion 2016 deficit in merchandise trade. That compares to a surplus of $1.7 billion in 1993 (in 1993 USD, the 2016 deficit was $36.1 billion).
But while Mexico is "beating us economically" in a mercantile sense, imports were not solely responsible for the 264% real growth in merchandise trade from 1993 to 2016. Real exports to Mexico more than tripled during that period, growing by 213%; imports outpaced them, however, at 317%.
The U.S.'s balance in services trade with Canada is positive: it imported $30.2 billion in 2015 and exported $57.3 billion. Its merchandise trade balance is negative – the U.S. imported $9.1 billion more in goods from Canada than it exported in 2016 – but the surplus in services trade eclipses the deficit in merchandise trade. The U.S.'s total trade surplus with Canada was $11.9 billion in 2015.
Real goods exports to Canada grew by 50% from 1993 to 2016; real goods imports grew by 41%. It would appear that NAFTA improved the U.S.'s trade position vis-à-vis Canada. In fact the two countries had already had a free trade agreement in place since 1988, but the pattern holds: the U.S.'s merchandise trade deficit with Canada was even steeper in 1987 than it was in 1993.
If NAFTA had any net effect on the overall economy, it was barely perceptible. A 2003 report by the Congressional Budget Office concluded that the deal "increased annual U.S. GDP, but by a very small amount – probably no more than a few billion dollars, or a few hundredths of a percent." The CRS cited that report in 2015, suggesting it hadn't come to a different conclusion.
NAFTA displays the classic free-trade quandary: diffuse benefits with concentrated costs. While the economy as a whole may have seen a slight boost, certain sectors and communities experienced profound disruption. A town in the Southeast loses hundreds of jobs when a textile mill closes, but hundreds of thousands of people find their clothes marginally cheaper. Depending on how you quantify it, the overall economic gain is probably greater, but barely perceptible at the individual level; the overall economic loss is small in the grand scheme of things, but devastating for those it affects directly.
For optimists in Mexico in 1994, NAFTA seemed to be full of promise. The deal was in a fact an extension of the 1988 Canada-U.S. Free Trade Agreement, and it was the first to link an emerging market economy to developed ones. The country had recently undergone tough reforms, beginning a transition from the kind of economic policies one-party states pursue to free-market orthodoxy. NAFTA supporters argued that tying the economy in with those of its richer northern neighbors would lock in those reforms and boost economic growth, eventually leading to convergence in living standards between the three economies.
Almost immediately, a currency crisis struck. Between the fourth quarter of 1994 and the second quarter of 1995, local-currency GDP shrank by 9.5%. Despite President Salinas' prediction that the country would begin exporting "goods, not people," emigration to the U.S. accelerated. In addition to the recession, the removal of corn tariffs contributed to the exodus: according to a 2014 report by the left-leaning Center for Economic and Policy Research (CEPR), family farm employment fell by 58%, from 8.4 million in 1991 to 3.5 million in 2007. Due to growth in other agricultural sectors, the net loss was 1.9 million jobs.
CEPR argues that Mexico could have achieved per-capita output on par with Portugal's if its 1960-1980 growth rate had held. Instead it clocked the 18th-worst rate of 20 Latin American countries, growing at an average of just 0.9% per year from 1994 to 2013. The country's poverty rate was almost unchanged from 1994 to 2012.
NAFTA does appear to have locked in some of Mexico's economic reforms: the country has not nationalized industries or run up massive fiscal deficits since the 1994-1995 recession. But changes to the old economic models were not accompanied by political changes – at least not immediately. Jorge Castañeda, who served as Mexico's foreign minister during Vicente Fox Quesada's administration, argued in a December 2013 article in Foreign Affairs that NAFTA provided "life support" to the Institutional Revolutionary Party (PRI), which had been in power without interruption since 1929. Fox, a member of the National Action Party, broke PRI's streak upon becoming president in 2000.
Mexico's experience with NAFTA was not all bad, however. The country became a car manufacturing hub, with General Motors Co. (GM), Fiat Chrysler Automobiles N.V. (FCAU), Nissan Motor Co., Volkswagen AG, Ford Motor Co. (F), Honda Motor Co. (HMC), Toyota Motor Co. (TM) and dozens of others operating in the country – not to mention hundreds of parts manufacturers. These and other industries owe their growth in part to the more than four-fold real increase in U.S. foreign direct investment (FDI) in Mexico since 1993. On the other hand, FDI in Mexico from all sources (the U.S. is usually the largest contributor) lags behind other Latin American economies as a share of GDP, according to Castañeda.
Led by the auto industry, the largest export category, Mexican manufacturers maintain a $58.8 billion trade surplus in goods with the U.S.; prior to NAFTA there was a deficit. They have also contributed to the growth of a small, educated middle class: Mexico had around 9 engineering graduates per 10,000 people in 2015, compared to 7 in the U.S.
Finally, the increase in Mexican imports from the U.S. has driven consumer goods prices down, contributing to broader prosperity: "if Mexico has become a middle-class society, as many now argue," Castañeda wrote in 2013, "it is largely due to this transformation." Yet he concludes that NAFTA "has delivered on practically none of its economic promises." He advocates a more comprehensive deal, with provisions for energy, migration, security and education – "more NAFTA, not less." That seems unlikely today.
Canada experienced a more modest increase in trade with the U.S. than Mexico did as a result of NAFTA, at an inflation-adjusted 63.5% (Canada-Mexico trade remains negligible). Unlike Mexico, it does not enjoy a trade surplus with the U.S.; while it sells more goods to the U.S. than it buys, a sizable services trade deficit with its southern neighbor brings the overall balance to -$11.9 billion in 2015.
Canada did enjoy a 243% real increase in FDI from the U.S. between 1993 and 2013, and real GDP per head grew faster – just barely – than its neighbor's from 1993 to 2015, though it remains about 3.2% lower.
As with the U.S. and Mexico, NAFTA did not deliver on its Canadian boosters' most extravagant promises; nor did it bring about its opponents' worst fears. The Canadian auto industry has complained that low Mexican wages have siphoned jobs out of the country: when General Motors cut 625 jobs at an Ontario plant to move them to Mexico in January, Unifor, the country's largest private-sector union, blamed NAFTA. Jim Stanford, an economist working for the union, told CBC News in 2013 that NAFTA had sparked a "manufacturing catastrophe in the country."
Supporters sometimes cite oil exports as evidence that NAFTA has helped Canada: according to MIT's Observatory of Economic Complexity, the U.S. imported $37.8 billion worth of crude oil in 1993, with 18.4% of it coming from Saudi Arabia and 13.2% of it coming from Canada. In 2015 Canada sold the U.S. $49.8 billion, or 41% of its total crude imports. In real terms, Canada's sales of oil to the U.S. grew 527% over that period, and it has been its neighbor's largest supplier since 2006.
U.S. crude oil imports, 1993: $37.8 billion current USD
U.S. crude oil imports, 2015: $120 billion current USD
On the other hand, Canada has long sold the U.S. 99% or more of its total oil exports: it did so even before the two countries stuck a free-trade agreement in 1988. In other words, NAFTA does not appear to have done much to open the U.S. market to Canadian crude. It was already wide open; Canadians just produced more.
Overall, NAFTA was neither devastating nor transformational for Canada's economy. Opponents of the 1988 free trade agreement had warned that Canada would become a glorified 51st state. While that didn't happen, Canada didn't close the productivity gap with the U.S. either: Canada's GDP per hours worked was 74% of the U.S.'s in 2012, according to the OECD.
China, Tech and the Crisis
An honest assessment of NAFTA is difficult because it is impossible to hold every other variable constant and look at the deal's effects in a vacuum. China's rapid ascent to the become the world's number-one exporter of goods and its second-largest economy happened while NAFTA's provisions were going into effect. The U.S. bought just 5.8% of its imports from China in 1993, according to MIT; in 2015, 21% of imports came from the country.
Hanson, David Autor and David Dorn argued in a 2013 paper that the surge in import competition from 1990 to 2007 "explains one-quarter of the contemporaneous aggregate decline in US manufacturing employment." While they acknowledged that Mexico and other countries "may also matter for US labor-market outcomes," their focus was unquestionably China; the country did – controversially – join the World Trade Organization in 2001, but it is not a party to NAFTA. Meanwhile Japan saw its share of U.S. imports decline from 19% to 6% from 1993 to 2015. Japan is not a party to NAFTA either.
U.S. imports by origin, 1993: $542 billion current USD
U.S. imports by origin, 2015: $2.16 trillion current USD
NAFTA is often blamed for things that could not be its fault: in 1999 the Christian Science Monitor wrote of an Arkansas town that it "would collapse, some said, like so many NAFTA ghost towns that lost needle-trade and manufacturing jobs to places such as Sri Lanka or Honduras." Sri Lanka and Honduras are not parties to the agreement.
Yet there is something to this conflation of NAFTA with globalization writ large. The deal "initiated a new generation of trade agreements in the Western Hemisphere and other parts of the world," the CRS writes, so that "NAFTA" has understandably – if not correctly – become shorthand for 20 years of broad diplomatic, political and commercial consensus that free trade is generally a good thing.
Isolating NAFTA's effects is also difficult due to rapid technological change: the supercomputers of the 1990s boasted a fraction of the processing power of today's smartphones, and the internet was not yet fully commercialized when NAFTA was signed. Real U.S. manufacturing output rose 57.7% from 1993 to 2016, even as employment in the sector plummeted; both trends are largely due to automation. The CRS quotes Hanson, who puts technology second behind China in terms of employment impacts since 2000. NAFTA, he says, is "far less important."
Finally, three discrete events have had major impacts on the North American economy, none of which can be traced to NAFTA. The tech bubble's bust put a dent in growth. The September 11 attacks led to a crackdown on border crossings, particularly between the U.S. and Mexico, but also between the U.S. and Canada: Michael Wilson, Canada's minister of international trade from 1991 to 1993, wrote in a 2013 Foreign Affairs article that same-day crossings from the U.S. to Canada fell nearly 70% from 2000 to 2012 to a four-decade low.
Finally, the 2008 financial crisis had a profound impact on the global economy, making it difficult to pinpoint one trade deal's effect. Outside of particular industries, where the effect is still not entirely clear-cut, NAFTA had little obvious impact – good or bad – on North American economies. That it is now in danger of being scrapped probably has little to do with its own merits or flaws, and much more to do with automation, China's rise and the political fallout from September 11 and the 2008 financial crisis.