Tonight’s State of the Union address should contain a healthy amount of talk about trade. But other than a showy pullout from the Trans-Pacific Partnership days into President Donald Trump’s first term, the talk has produced little action. And would-be supporters of a fairer trade agenda have grown restless. “We’re terribly disappointed and hugely frustrated,” Leo Gerard, the president of United Steelworkers, told CNN. “There’s been no action that has done anything to protect and defend American jobs. ... In some cases, we’re worse off now than we were then.”

But protecting American jobs means more than setting tariffs or quotas. In fact, in the very steel industry where Gerard wants to see action, a deeper problem has emerged: a duopoly in domestic steel production that, if empowered as the two major suppliers to the U.S. market, would have an enormous impact on what Americans build and eventually what they buy. The situation demonstrates that it’s not enough to devise industrial policy based just on foreign competition; you have to look at the domestic companies making the goods, too.

Take last week’s imposition of tariffs on the foreign import of solar panels and washing machines, Trump’s first major protectionist step. The reviews were mixed—not to mention complicated.

Paul Krugman wrote that the solar-panel tariffs will destroy more jobs than they create, because the bulk of domestic employment in solar panels is in installation, and if the tariffs create higher prices, fewer people will opt to install. Interestingly, Al Gore half-heartedly defended the action because the tariffs are temporary and not as large as they could have been. But that’s the problem: Because they are temporary and modest, they’re highly unlikely to lead to long-term domestic investments in solar manufacturing in the United States.

The washing-machine tariffs have also, weirdly, caused discontent at the U.S.-based factories of top foreign manufacturers, which are claiming that lower profits from imports generally will cause them to reduce investment in U.S. factories. As a result, LG has announced price increases on their washing machines, and Samsung has warned that their expansion efforts would be harmed by the move. But the whole point of the tariff is to move factories onshore, which both LG and Samsung are already doing. The tariffs should accelerate that process if the foreign companies want cheaper access to the U.S. market.

These preliminary spats are just a prelude to the much more momentous upcoming decision on steel. On January 11, Commerce Secretary Wilbur Ross finally delivered a “Section 232” reportoriginally expected back in Juneabout the threat that steel imports pose to national security. Since steel imports increased 17.5 percent in 2017, it’s widely believed that the report—though not yet public—will signal that dumping cheap steel into the United States could impede government needs in the event of a war, for example. Trump has 90 days to issue a response to the investigation, which could include unlimited tariffs or even restrictions on steel imports.

Reports indicate that administration officials are at odds over what to do. Trade hawks point to tariffs on Chinese steel imposed through international-trade cases that led to a 5 percent drop in imports. (Though China has since routed its steel through other countries.) Meanwhile, doves argue that broad-based restrictions would increase prices, disrupt global markets, and trigger retaliations that could lead to a trade war. Labor positions the decision as the last chance to “save American steel.”

But nobody is looking at which steel companies might actually benefit from the change. A little-noticed report in an industry trade paper earlier this month points to a surprising answer: mainly just two.

The report, from American Metal Market magazine, highlights Commercial Metals Company’s (CMC) purchase of four mills from a competitor named Gerdau. Once the deal closes, CMC will solidify a 50 percent share of the market for rebar, the reinforcing beams used in bridges, buildings, roadways, and nearly every structure built in America.

Seth Rosenfeld, a market analyst with Jeffries, saw the acquisition as “further aiding pricing discipline in an already consolidated market.” That term, “pricing discipline,” has a very specific meaning. It refers to the ability to gouge purchasers for higher prices, in this case for rebar. If the deal clears, CMC and one other company, Nucor, would control close to 85 percent of the domestic market. Nucor would thus get a “free-rider” benefit from higher prices in a consolidated market, Rosenfeld said, because CMC is doing all the consolidation work for them.

So, if Trump moves to restrict imports of steel rebar, which ran at about 1.5 million tons last year, he will be handing over the market to two companies that would have a serious incentive to jack up prices, given the lack of competition. CMC stock has been running wild since the first of the year, up 20 percent in a month. This anticipation of inflated prices can be seen across all steel markets, not just rebar: A research note from Zacks states that domestic steelmakers would have “more pricing power” if Trump triggers Section 232 restrictions.

What does this mean for you? In addition to construction, steel is used in automobiles, rail lines, aircraft, energy infrastructure, packaging, and appliances. Go into your kitchen and you’ll find a dozen stainless-steel products without looking too hard. Imagine the prices of all of that shooting up because a restrictionist policy meant to save U.S. jobs funnels power into the hands of a tiny group of steel CEOs. Imagine the impact on all the companies that make those products, and their workers.

The United States does have a domestic steel-market problem. Annual production has fallen 23 percent in the past decade and a half, and the entire domestic industry has struggled to survive. But the survival strategy to this point has been to merge with one another, on the idea that bigger is better. If the Trump administration just snaps on tariffs and gives these behemoths free rein, the benefits will be hindered by the costs. The nation would need antitrust enforcement to limit concentration in the steel industry, or else the United States will just walk from one bad situation into another. The free market will not sort this out. Only the kind of antitrust enforcement that served this country well for a century will create true competition in steel to the benefit of suppliers, manufacturers, and consumers.

The broader point is that no one policy can save U.S. manufacturing or bolster middle-skill, high-quality jobs. You have to combine them for the best effect. The American steelworker is a dying breed, and illegal trade practices must be stopped so they don’t go extinct. But a policy that just creates domestic steel monopolies does nobody any good.