A Mediterranean Shipping Co. (MSC) ship loaded with freight containers is seen in this aerial photograph approaching the Port of Long Beach in Long Beach, California, U.S., on Friday, Feb. 1, 2013. (Tim Rue/Bloomberg)

Is free trade bad — or good — for American workers?

This presidential election gives us two opposing views. According to Republican presidential nominee Donald Trump, “we are killing ourselves with trade pacts that are no good for us and no good for our workers” — and so he proposes higher tariffs to boost employment and wages. Democratic presidential nominee Hillary Clinton, by contrast, argues that free trade leads to “better growth,” which means workers’ wages will go up in the U.S. industries that compete around the globe. (Yes, she came out against the Trans-Pacific Partnership during the primary, pushed left by Sen. Bernie Sanders, but she continues to praise the benefits of free trade for U.S. workers.)

Trump promises that protectionism will benefit workers in America’s import-competing industries. Clinton promises that free trade will benefits workers in America’s export-oriented industries.

They’re both wrong.

No trade policy will automatically increase U.S. workers’ wages. Trade policies can increase the profitability of certain firms or industries, but whether those profits are shared with workers depends on labor’s bargaining power.

The economic models that political scientists use are based on a very unrealistic assumption

Both candidates’ views of trade are rooted in the neoclassical economic models that dominate the field of political science. For instance, political scientists often analyze international trade with help from the Ricardo-Viner model, which predicts that a trade policy that benefits an industry will automatically increases wages for that industry’s workers. According to this theory, free trade increases wages for workers in export-oriented industries, while trade protectionism increases wages for workers in import-competing industries. But these predictions depend on the unrealistic assumption of full employment.

Here’s how its logic works. Say that a protectionist trade policy helps a particular U.S. industry — for example, steel — by imposing a tariff that increases the domestic price of steel. Because steel now sells for more, steelworkers can produce a greater value of steel each day. If steelworkers originally produced $100 of steel a day, a 10 percent price increase enables them to produce $110 of steel a day. As a result of such increased productivity (technically, an increase in the marginal revenue product of labor), it becomes profitable for steel firms to hire more workers. But since no one is looking for jobs in an economy with full employment, the expanding industry has to raise their wages to attract workers.

The same logic holds for free trade and wages. Free trade grants U.S. exporters — say, the high tech software industry — access to markets abroad, which increases demand for their products. The ensuing price increase enables their workers to produce a higher value of software each day, which makes it profitable for the firm to hire more workers, and leads to automatic wage increases.

Some recent research has offered what’s been called “’new’ new trade theory.” This theory recognizes that only a small number of American companies are competitive enough to export products. But its predictions about wages are based on the same logic. In this view, free trade permits these firms to sell more products abroad, which increases both demand and the product’s price. The higher price increases the value of goods produced by the firms’ workers, which leads to an increase in hiring. Since this theory also assumes full employment, increased hiring leads to increased wages.

But we are not living in a full employment economy. And in our economy, wages only increase if workers are powerful enough to bargain for raises.

Workers don’t gain anything from new trade policies unless they have real bargaining power

Even when companies make more money, they do not automatically pass that money on to their workers. As I argue in my new book, workers only share in the profits produced by trade policy reform if they can rely on basic labor rights, such as the rights to organize, bargain collectively, and strike.

In an article published last year, I examine the relationship between wage growth and productivity growth in 28 manufacturing industries, across 117 countries, from 1986 to 2002. The analysis relies on labor rights data that measures how often a country restricts their workers’ rights to act collectively. The results demonstrate that when labor rights are weakly protected, the trade models systematically exaggerate how much workers will gain from trade policies that benefit their industries. As you can see in the figure below, wages rise along with productivity growth when labor rights are well protected. Wages don’t go up at all if labor rights are regularly violated.

The blue dots represent industries in countries that strongly protect workers’ rights to organize, such as Sweden and France. The red ones are from countries that regularly deny workers’ rights, such as Colombia and Malaysia.

In the past, workers wouldn’t support the trade policies their employers wanted until they won labor rights and profit-sharing agreements

Although political scientists frequently rely on these neoclassical trade models, U.S. workers know — and have known for more than a century — that they do not automatically share in the profits generated by trade policies that benefit their industries. Rather than support trade policies that increase profits without increasing wages, U.S. labor unions have often opposed trade policies that benefit their own industries.

And yet, no matter how often capital and labor have disagreed about trade over the past 150 years, conventional political science theories and analyses regularly overlook such class conflict.

As I explain in a recent article, during the late 19th century workers were only willing to support the high import tariffs that their employers wanted after they’d established “profit-sharing institutions.” These institutions formally recognized a series of things that workers sought: labor unions, industry-wide wage contracts, and explicit agreement that wages would rise along with profits.

For example, the steelworkers’ union negotiated a contract that pegged wages directly to the price of steel. When the tariff increased the price of steel, wages rose automatically. As long as such agreements lasted, the union favored high steel tariffs. However, when employers discarded such agreements — as Andrew Carnegie did when he broke the steelworkers’ union at the infamous Homestead steel strike of 1892 — workers quickly stopped supporting high tariffs.

My book shows that similar divisions shaped U.S. trade politics during the Great Depression and World War II, as well. Globally competitive companies such as Ford Motor Company and U.S. Steel consistently lobbied the government in favor of free trade. At the time, these companies’ workers were struggling to unionize — and they openly rejected the idea that free trade would improve their wages or lives. However, once labor unions finally established profit-sharing institutions in the early 1940s — following the passage of the Wagner Act — those same workers began supporting free trade, alongside their employers.

How will new trade policies affect American workers?

U.S. labor unions are at their weakest point in more than a century, leaving workers without the bargaining power necessary for profit-sharing. No matter how trade policy changes after January — whether it becomes more protectionist or more pro-trade — workers aren’t likely to come away with higher wages.

Adam Dean is an assistant professor in the department of political science at Middlebury College. His book “From Conflict to Coalition: Profit-Sharing Institutions and the Political Economy of Trade” was just published by Cambridge University Press.