The Truth about Trade

Imports have inarguably affected U.S. manufacturing companies and workers — no serious free-trader argues otherwise. But criticism of trade and its impact on American workers has acquired a sharper edge in the Age of Trump, bolstered in part by a recent study from labor economists David Autor, David Dorn, and Gordon Hanson that found that the recent surge in Chinese imports to the United States has inflicted pronounced harms on the wages and labor-force participation of U.S. workers in local markets (e.g., mill towns) that face direct competition with those imports. Trump fans and longtime trade skeptics on both the left and the right have seized on this study as the final “proof” that free trade — in particular, trade with China — has been a disaster for the United States and its workers, and that a heavy dose of protectionism, through Trump’s tariffs or export subsidies, could produce a manufacturing renaissance in America.

Reality, however, begs to differ.

First, even assuming Autor et al. are entirely correct about the harms of Chinese imports — a conclusion about which George Mason University economist Scott Sumner has raised legitimate questions — there remains no evidence that imports are the primary driver of U.S. manufacturing-job losses, or that the U.S. manufacturing sector is actually in decline. In fact, American manufacturers began slowly and steadily shedding workers as a share of the U.S. work force in the late 1940s and in sheer numerical terms in 1979 — long before the North American Free Trade Agreement existed or Chinese imports were more than a rounding error in U.S. GDP.

By contrast, the United States has gained about 54 million jobs since 1980, 30-plus million of which came after the creation of NAFTA and the World Trade Organization in the mid 1990s. Meanwhile, it is a myth that the United States “doesn’t make anything anymore” or that trade agreements have caused a “giant sucking sound” as investment and jobs go elsewhere. Our manufacturers continue to set production and export records, and the United States is the world’s second-largest manufacturer (17.2 percent of total global output) and third-largest exporter. America also remains the world’s top destination for foreign direct investment ($384 billion in 2015 alone) — more than double second-place Hong Kong and almost triple third-place China. Much of this investment went to U.S. manufacturing assets, as shiny new BMW, Toyota, and other foreign-owned plants across the American South attest.

For these and other reasons, it is widely accepted that U.S. manufacturing “decline” has been limited to employment, and that these losses were primarily caused by productivity gains, not trade. Indeed, even the most pessimistic academic studies on imports and manufacturing jobs have found only a limited connection between the two. Autor, Dorn, and Hanson found in 2013, for example, that “import competition explains [only] one-quarter of the contemporaneous aggregate decline in US manufacturing employment” between 1990 and 2007. Other studies have been even more sanguine. For example, a recent Ball State study attributed almost 90 percent of all U.S. manufacturing-job losses since 2000 to productivity gains. “Had we kept 2000-levels of productivity and applied them to 2010-levels of production,” the authors write, “we would have required 20.9 million manufacturing workers. Instead, we employed only 12.1 million.” Thus, it is simply wrong to blame import competition for the disappearance of American manufacturing jobs or the supposed destruction of U.S. industrial capacity.

Second, despite its harms to some manufacturing interests, free trade also has generated broadbased benefits for U.S. consumers, businesses, and workers. In The Payoff to America from Global Integration, economists with the Peterson Institute found that past global-trade liberalization through the WTO and other efforts generated between $2,800 and $5,000 in additional income for the average American and between $7,100 and $12,900 for the average household. The consumer gains from trade disproportionally accrue to America’s poor and middle class. A 2015 study by Pablo Fajgelbaum and Amit Khandelwal finds that these groups, because they concentrate spending in more-traded sectors such as food and clothing, enjoy almost 90 percent of the consumer benefits of trade. These benefits are even more concentrated for Chinese imports, since poor and middle-class American consumers are more likely than their richer counterparts to shop at “big box” stores such as Target and Walmart that carry a lot of made-in-China goods.

American businesses, of course, also benefit. More than half of all imports (including those from China) are inputs and capital goods consumed by other American manufacturers to make globally competitive products. Raising these firms’ costs via tariffs would mean fewer employees, if not outright bankruptcy — a particularly bad outcome given that downstream industries (e.g., steelmakers) typically employ far more workers than their upstream counterparts (e.g., steel users). Non-manufacturers benefit, too — whether they be retailers such as the Gap, transportation and logistics companies such as FedEx, or multinational firms such as Apple, which assembles iPhones in China but generates most of their final sale price through marketing, design, engineering, and even manufacturing done in the United States. (Chinese manufacturers themselves earn only a few dollars from an iPhone’s assembly.) U.S. exporters such as Caterpillar and Boeing also gain from trade, and many foreign markets wouldn’t be open without reciprocal trade agreements such as NAFTA. According to the Business Roundtable, in 2014, U.S. free-trade-agreement (FTA) partners purchased 13 times more goods per capita from the United States than non-FTA countries did.

Third, free trade and protectionism raise serious moral issues. Protective tariffs force American families and businesses to subsidize — through hidden, regressive taxes — the small share of U.S. manufacturers and workers (and the tiny portion of the total economy and work force) that compete directly with the imports at issue. For this reason, labor unions such as the United Steelworkers expend considerable financial and in-kind resources lobbying the federal government to insulate them from foreign competition, at a huge cost to American consumers. When the Steelworkers convinced President Obama to impose 35 percent tariffs on Chinese tires in 2009, the result was, even under the best assumptions, a few unionized jobs saved at a cost to U.S. consumers of $900,000 per job — precisely the type of crony-capitalist boondoggle that, in any form other than that of a hidden tax targeting a foreign “adversary,” would engender hostile political opposition from the right.

Finally, even if it were morally and economically advantageous for the United States to embrace protectionism, it’s almost certainly impossible for it do so. U.S. manufacturers have evolved over decades to become integral links in a breathtakingly complex global value chain — whereby producers across continents cooperate to produce a single product based on their respective comparative advantages — that could not be severed without crippling both them and the global economy. According to the WTO, for example, almost 40 percent of all U.S. exports are involved in global value chains; almost 31 percent of exports from China, Canada, or Mexico contain U.S. inputs; and almost 34 percent of U.S. exports contain inputs from these same three countries. Perhaps the automobile industry, more than any other, makes this point clear: The National Highway Traffic Safety Administration estimates that the American “Big Three” automakers produce only five of the top 20 most “domestic” cars (defined by their total share of U.S. and Canadian auto parts) sold in the United States in 2016. Killing the “virus” of global trade integration would surely kill the host, too.

Given these realities, even the authors of the protectionists’ favorite new study reject protectionism and instead point to problems in the U.S. labor market that keep it from adjusting adequately to the effects of trade or other disruptions. In releasing a previous study of trade with China, David Autor in 2012 stated, “I’m not anti-trade, but it is important to realize that there are reasons why people worry about this issue.” He elaborated: “We do not have a good set of policies at present for helping workers adjust to trade or, for that matter, to any kind of technological change.” His co-author Gordon Hanson told the New York Times earlier this year:

U.S. labor-market data attest to this problem — one that exacerbates job dislocations arising from trade, technological innovation, or any other disruptive but ultimately beneficial phenomenon. Total non-farm job openings, for example, are at their highest point on record (including well over a million unfilled jobs in “blue collar” fields such as manufacturing, construction, and transportation) and continue to outpace hirings. Workers have recently appeared more willing to quit their jobs and seek others, but the civilian labor-force-participation rate has hovered near its lowest point (62.5 percent) since the late 1970s — a problem caused in part by the fact that workers have become less likely to move to areas with better employment opportunities, choosing instead to remain in places hit hard by the Great Recession and to drop out of the labor force entirely.

More-complex measures of labor dynamism corroborate the aforementioned numbers: The Goldman Sachs Labor Market Dynamism Tracker, which synthesizes various labor reports, shows that, after remaining positive through the 1980s and ’90s, U.S. labor dynamism — the natural, beneficial replacement of old jobs with new ones, owing in part to the willingness of workers to seek new jobs and their ability to obtain them — dove into negative territory in 2001 and has remained there ever since. A recent study by Steven Davis and John Haltiwanger found that the “U.S. economy experienced large, broad-based declines in labor market fluidity in recent decades,” and that this reduction in fluidity had “harmful consequences for productivity, real wages and employment.”

Some of the troubling decline in U.S. labor dynamism is a matter of an aging work force disproportionately composed of Baby Boomers — older workers are far less likely than their younger counterparts to change jobs. However, three types of distinct government policy failures have amplified the problem. First, state and federal policies prevent Americans from saving enough wealth to cope with unexpected financial calamities or to enable them to take professional risks. Over 60 percent of Americans have less than $1,000 in their savings accounts, and economists on the right and the left agree that our current tax and entitlement policies discourage private savings — unlike, say, those of Canada, which has a highly successful and popular system of tax-free savings accounts.

Meanwhile, the costs of health care, child care, and education — all highly subsidized, protected, and regulated — have risen far faster than the rate of inflation, and there is little doubt that government intervention has played a role in this trend. The New York Fed in 2015, for example, found a strong link between federal student aid and the skyrocketing cost of tuition. Such cost increases disproportionately harm poor and middle-class Americans and force them to spend more of their stagnating wages on these essential services — money that could have gone into a savings account. Indeed, the rampant inflation in these sectors stands in stark contrast to the declining prices of other goods, such as clothing, toys, and electronics, that are less subsidized, more open to competition (foreign and domestic), and less subject to onerous government regulation.

Second, government policy actively discourages Americans from finding work in burgeoning fields. Perhaps the most brazen example of such policies is the federal tax code’s business deduction for work-related education, which permits a worker to deduct education and training expenses from his taxable income, but only if they relate to his current job. Thus, a textilefactory worker can get a tax benefit for new training on the latest garment machine, but he cannot get the same benefit for night classes to become a certified IT specialist. Such a system discourages workers in dying fields from preparing themselves for a new career.

An assortment of other government policies also undermines a worker’s ability or willingness to change jobs. In their aforementioned study on collapsing U.S. labor dynamism, Davis and Haltiwanger identified five specific contributors: employment-protection laws (which protect employees from being fired because of certain actions or immutable characteristics) that “suppress labor market flows, sometimes to a powerful extent”; laws that erode the employmentat-will doctrine (which permits employers to fire employees without cause); occupational-licensing laws and other labor-supply restrictions; minimum-wage laws; and the tax code’s preference for employer-provided health insurance. At the same time, the United States has witnessed a distressing collapse in business dynamism — the creation and destruction of firms — which has had the consequence of entrenching workers in large, existing firms while reducing job openings in new and innovative ones. According to one recent study, a big cause of the recent collapse of business dynamism is the federal government’s response to the Great Recession, which involved “defensive policies to protect large firms and existing employment, rather than proactive policies to encourage entrepreneurship and new venture/​job creation.” None of this is good for people looking for a job or considering a career change.

Finally, current government policy has failed to help displaced workers when disaster strikes, and has very likely made things worse. Most notably, the Trade Adjustment Assistance (TAA) program, intended to subsidize U.S. workers affected by import competition, is a notorious failure: Not only are TAA’s costs too high and its eligibility criteria too loose, but multiple studies commissioned by the Labor Department have found that TAA participants are worse off, as measured by future wages and benefits, than similarly situated jobless individuals outside the program. (TAA also breeds the misconception that trade is somehow different from, and worse than, other forms of beneficial economic disruption, such as automation.)

Other federal job-training programs are similarly inefficacious. A 2011 Government Accountability Office study, for example, found that the federal government had 47 different, often overlapping job-training programs spanning nine federal agencies at a cost of $18 billion per year. Only five had been subject to any sort of impact analysis since 2004; thus, “little is known about the effectiveness of [the] employment and training programs” identified. A 2014 reform of this system, the Workforce Innovation and Opportunity Act, eliminated 15 programs (while maintaining the rest, despite their long history of subpar results) but failed to impose any sort of rigorous multi-site evaluation and accountability system. Without these simple reforms, or other, more radical ones, there is no way to ensure that the “reformed” federal job programs won’t continue their long record of failing American workers and taxpayers.

Unfortunately, the private sector has not succeeded where our problematic government jobtraining system has failed, and government policy may actually deter it from attempting to do so. Private-sector job-training programs, for example, seem to be disappearing: The Labor Department estimates that “formal programs that combine on-the-job learning with mentorships and classroom education fell 40% in the U.S. between 2003 and 2013,” and the 2015 Economic Report of the President found substantial declines in the percentages of American workers receiving employer-paid training (19.4 percent to 11.2 percent) or on-the-job training (13.1 percent to 8.4 percent) between 1996 and 2008. There are legitimate concerns that such programs have simply been sloughed off in favor of ineffective government programs. Tax and regulatory costs might also play a discouraging role: According to one analysis, a $14-per-hour worker has a true cost to his employer of almost $20 per hour because of federal and state taxes plus an array of mandated and voluntary benefits and job training. As labor costs continue to rise, companies are more often looking for skilled workers whom they don’t have to train.

Federal unemployment benefits also have the potential to discourage workers from searching for and obtaining a job. Most troubling is the current Social Security Disability Insurance (SSDI) system, which, because of its generous benefits, lax eligibility criteria, and lack of rigorous enforcement, has become, according to the Manhattan Institute’s Scott Winship, “a permanent dole for a rising number of adults with limited earning potential who clearly are physically able to work.” The numbers bear this out: Between 1990 and 2014, the percentage of working-age individuals who receive SSDI benefits more than doubled, from 2.3 percent to 5.1 percent. Basic unemployment insurance also raises concerns: Four economists examined the effects of North Carolina’s 2013 cuts to unemployment benefits and found that previous benefit extensions had had a significant negative effect on the state’s employment level, number of job openings, and labor-force-participation rate — harms that “dominate any potential stimulative effect that some ascribe to such policies.” The same economists found similar discouraging results at the national level.

Free trade — with China or any other country — has demonstrable benefits for American families and businesses. To the extent he denies this, Donald Trump is entirely wrong. However, the economic anxiety propelling Trump reflects very real problems in America’s labor market — problems caused not by Chinese imports or any other type of creative destruction but by multiple government-policy failures and a resulting collapse of labor dynamism. The solutions to these problems are complex and deserving of substantial debate. But the analysis I have presented should provide some clues. Most simply, U.S. workers should receive the same tax benefit for job training unrelated to their current job as they do now for training related to it. SSDI and unemployment-insurance eligibility requirements should be tightened and redesigned to ensure that able-bodied adults are looking for, and accepting, available work. Occupational-licensing reform should be a priority, particularly at the state level. Federal job-training programs should be consolidated, if not eliminated outright — perhaps through a simple voucher for dislocated workers to use at accredited community colleges or vocational schools, or a single block grant to states for local experimentation with programs that support, instead of crowd out, private-sector training initiatives such as apprenticeships.

More broadly, tax-free savings accounts, similar to those in Canada, also should be explored, as should ways to increase the portability of health care and other benefits currently tied to people’s jobs. (Eliminating the tax preference for employer-provided health insurance would be the most obvious solution.) Finally, the federal government should more seriously consider, and attempt to rectify, the inflationary harms caused by its subsidization and overregulation of basic essential services such as health care and higher education — opening them to global competition would be a great place to start.

None of these ideas is a silver bullet, but the problems they would seek to address, and the palpable economic anxiety of Americans, clearly show that reform is needed. Protectionism not only would ensure that these problems aren’t fixed but would actually make things far worse.