In an insightful paper about the relationship between globalization and populism, economist Dani Rodrik proposes the following thought experiment:

“Suppose Harry and John own two firms that compete with each other. Ask yourself how you feel about each of the scenarios below. In each of them, Harry outcompetes John, resulting in John and his employees losing their jobs. Should they be blocked or allowed to run their course?

(1) Harry works hard, saves and invests a lot, and comes up with new techniques and products, while John lags behind.

(2) Harry finds a cheaper (or higher quality) supplier in Germany.

(3) Harry outsources to a supplier in Bangladesh, which employs workers in 12-hour-a-day shifts and under hazardous conditions.

(4) Harry brings Bangladeshi workers to the US under temporary contracts, and puts them to work under conditions that violate domestic labor, environmental, and safety laws.”

Rodrik points out that each scenario generates winners and losers, but Harry gains more than John loses, so if all we cared about was growth, we’d be fine will all of them. “But most audiences react very differently to them.”

No one objects to the first scenario, and few object to the second. But the next two, by invoking outsourcing to a low-wage, low-standards economy, thereby violating domestic norms (3) and laws (4), introduce a sense of unfairness. In those last two scenarios, John and his staff are losing not because they’re slacking off but because Howard has found a way to tilt the playing field his way.

Most political and economic elites have long been okay with No. 3. Outsourcing of jobs is the price we must pay, they argue, for the increased flow of goods, services and capital and the efficiencies they deliver. Yes, multinational firms will engage in global wage arbitrage, and that will hurt U.S. workers’ paychecks. But it will also lower the prices they face at Walmart and Best Buy. (It should be noted that lower prices did not prevent inflation-adjusted wage stagnation for many workers hit by these dynamics.)

Thus, to most in the policy realm, outsourcing doesn’t constitute unfair trade. To them, it’s practices like those engaged in by China: requiring technology transfers by foreign firms that want access to China’s huge market, or the heavy subsidy of exporters and chosen industries.

Thus, there is a gulf between what most people and what elites view as most unfair about trade. For working people, unfair trade is throwing them or their neighbors into competition with those who can sharply undercut their pay and living standards. For elites, unfair trade is anything that diminishes the competitive advantages and profitability of U.S. multinationals.

The political system has long supported the companies’ definition of unfairness, not the workers’. Alongside tariff reduction, trade agreements turned to protecting investors, intellectual property and patents, and making sure big Pharma and big finance could safely penetrate foreign markets.

Candidate Donald Trump recognized a huge opening here to, at least rhetorically, side with the workers against the global forces that are “ripping them off” thanks to “horrible trade deals.” To this day, even as he enacts tariffs (which invite retaliatory tariffs) that hurt them, his constituents remain loyal. Apparently, if people feel you’ve got their back, they’ll put up with a lot.

But will tariffs address either side’s version of unfair trade? Will they help workers hurt by outsourcing or companies hurt by China’s subsidies and intellectual property theft?

It’s hard to say who’ll win a war when you’re in the middle of it, but I’m skeptical on both fronts.

Though Trump hasn’t said what winning looks like (he just assures us it’s easy), I suspect it’s a lower trade deficit and import substitution through expanded domestic capacity. But his agenda is likely to boost the value of the dollar and worsen the trade deficit, and, even considering his new $12 billion payoff to farmers, I doubt he’s willing to sustain this war to the point where producers build domestic plants to replace foreign ones, doubts that are amplified by the deep integration of global supply chains into American production. As I’ve recently written, tariffs can’t unscramble the globalization omelet.

It’s possible that trade war leads China to reduce some tariffs, but I don’t see it changing its trade practices, as they are at the core of its growth model, a.k.a. “Made in China 2025.” The Chinese are too deeply committed to capturing market share in artificial intelligence, robotics, advanced computing, next-generation vehicles and high tech in general to change their plans for U.S. soybeans (especially when they can get them from Brazil).

But if the tariffs won’t work, what will?

I’ve written extensively about policies that circulate some of the benefits of trade into helping those hurt by it, but here I’d like to talk about the China problem.

I say we fight fire with fire, meaning not IP theft, of course, but aggressive investments in future technologies. Somehow, our fetish over “private markets” and disdain for government has wiped our collective memory of the role that government-supported R&D has played in virtually every important technological breakthrough, including  railroads, aviation and space travel, advanced computing, the Internet, GPS and much more.

In fact, federal R&D spending as a share of GDP has been flat or falling for more than a decade, which is not surprising if you’ve been following the budget cuts in precisely these areas (at least, outside of defense spending).

That’s misguided policy. Instead of applying an outdated tool — tariffs — to unscramble the globalization omelet and move backward, we should be looking around the next corner and investing in people, research and technology that will give us the economic edge we’re rapidly losing.