The White House, nine days after having said the China trade war was “on hold,” yesterday flipped the switch back to “on.” As of now, the Trump administration again plans to move forward with imposing 25 percent tariffs on $50 billion of imports from China by the middle of June.

To try to ascertain the position of President Trump on trade at any point in time is like trying to ascertain the position of a subatomic particle: It is better viewed as having a probabilistic distribution than as having a true position that could be discovered through sufficiently acute observation and analysis. Even under normal circumstances, it could be hard to keep track of a half dozen or more different China trade issues running simultaneously. Some of the actions that are at stake have a basis in economic or political or legal logic, while some do not.

{mosads}We can begin by noting that when the United States goes against international institutions and deals, for example, by imposing tariffs that violate World Trade Organization or other international agreements, aside from immediate negative economic effects, we give other countries good reason to mistrust the value of our word and to engage in similar behavior themselves. The calculus is that much worse because the United States had the overwhelmingly dominant role in setting up such institutions as the WTO in the first place. We did it because such international agreements and institutions are in our national interest.

The least logical of the recurring Trump obsessions is the demand that China reduce the U.S. bilateral trade deficit. The Chinese government has no policy levers that could reduce the bilateral balance by the mooted $100 billion or $200 billion. If China wanted to give Trump a superficial “win,” it could reduce the bilateral trade deficit some by buying more American natural gas or by routing smartphone exports through a third country like South Korea.

But the economics would be illusory. There would be virtually no effect on the overall U.S. trade deficit: We would otherwise sell the natural gas to some other country, and in reality those smartphones already get about 95 percent of their value added from South Korea, the United States and other countries anyway. Focusing on the bilateral balance is a waste of time. Regardless what happens with China, the overall U.S. trade deficit will rise this year as a result of the Republican tax cuts, enacted at a time when the economy is already producing at the limit of its capacity.

Similarly unmoored from both economic and legal logic are the tariffs that Trump has imposed on imports of steel and aluminum, and threatens to impose on automobiles and other goods. With respect to legal criteria, the national security justification that he has invoked is flimsy in the extreme and will likely inspire other countries to retaliate. With respect to economic criteria, the beneficiaries in the steel industry, for example, are vastly outnumbered by those who will be hurt because they use steel as inputs, they consume the final product, or they produce export goods that will be hit by the repercussions, such as farmers.

Much better grounded are the U.S. penalties legally imposed on the Chinese company ZTE for violating international sanctions against Iran and North Korea. But this is precisely the issue on which Trump seems most clearly to have let China off the hook. In between the two extremes are some grievances that the United States could reasonably pursue in negotiations with China if done intelligently.

These include intellectual property rights and a Chinese practice of requiring foreign firms that want to set up in China to share technology. But to pursue such areas effectively, the United States should do it in cooperation with allies like Europe and under international agreements like the WTO and the Trans Pacific Partnership.

Trump has been doing the opposite. His approach to China has been so inconsistent that observers have had to stretch to find explanations. Suggested explanations include allegations of quid pro quos for Trump family businesses, divisions between “nationalists” and “business” factions on his staff, his tendency to be persuaded by whomever he last talked to, a desire to preempt Democratic protectionists, and an instinct to do the opposite of whatever President Obama did.

It is easier to say what the Trump approach is not. It is not the deliberate or competent application of a coherent strategy informed by a clear set of American policy objectives, an understanding of the incentives of the other side, or an appreciation of the implications that impulsive tweets can have for the global system over the long term or even for the next move in the tactical game in the short term.

The rise of China as a great power over the last three decades has had little to do with the terms of any international trade agreements that were made or could have been made. The bilateral trade issues currently under discussion could all be viewed as low stakes. This is especially true when compared to some larger national security issues.

We need help from China with a newly nuclear North Korea more than we need it to give in to any of our on again, off again trade demands, even those that might make some sense. Think how bad it would be if President Trump applied to Kim Jong Un the same pattern of erratic alternation back and forth between inflammatory bluster and inexplicable concession that he has applied to China. Oh, wait…

Jeffrey Frankel is the James Harpel professor of capital formation and growth at Harvard University. He served on the White House Council of Economic Advisers under President Reagan and President Clinton.