You argue that the United States, not just China, has become a major driver of the new economic statecraft. How did Washington move from being the chief architect of a rules-based neoliberal order to increasingly relying on coercive trade and investment tools?

The most immediate answer is the Trump administration beginning in 2017, when President Trump decided to go after China and a number of other countries that he believed were engaging in unfair trade. Many people point to that moment, but it is important to place it in historical context. The United States has often engaged in coercive economic statecraft. Under the Reagan administration, for example, we restricted cars from Japan because many people were concerned that Japan was going to become the next superpower before its economy faltered at the end of the 1980s.

So, this is not the first time the United States has used coercive economic tools, and other countries use them too. What is different about the Trump administration is that the instruments it used have not been subject to review by the World Trade Organization. When the United States used measures like Super 301 against Japan, those policies were eventually set aside because these kinds of unilateral actions should not be used. What President Trump has done is essentially say that we do not need to worry about the rules-based system. That is a significant change.

At the same time, this trend goes beyond Trump. The Biden administration also showed much less commitment to a multilateral, or even a mini-lateral, trading order. It did not rejoin the Trans-Pacific Partnership, which became the Comprehensive and Progressive Agreement for Trans-Pacific Partnership. It also did not pursue a trade agreement with Europe on the model of the Transatlantic Trade and Investment Partnership that the Obama administration had negotiated. The Biden administration, too, was quite protectionist.

This is a broader trend in American politics after China joined the WTO in 2001. There has been growing anxiety about the rise of Chinese imports and China’s industrial policy, and that helps explain what the United States has been doing.

When countries use export controls, sanctions, tariffs, investment screening, and industrial policy at the same time, how should we distinguish between legitimate economic security concerns and simple protectionism?

In a democracy, it is very easy for firms to lobby for protectionism. You have the steel industry lobbying for protection, the textile industry lobbying for protection, and everyone claims they are important for national security. When I teach in the MBA program, I tell students that the two best ways to get protection are to scream “national security” or “jobs.” And if that does not work, scream “jobs and national security.”

Those arguments appeal to politicians who worry that their constituencies will be upset or that they will seem weak on national security. So in a democracy, it is very difficult to develop clear criteria that distinguish industries genuinely vital to national security from industries that are simply seeking protection because they do not want to compete.

Some of the work I have been doing with collaborators has focused on whether there is a way to define in advance which industries are truly important for national security and which are simply using that language to avoid market competition. That is an important task, because otherwise you end up with a lot of protectionism. You end up protecting sunset industries instead of sunrise industries. This is not just a problem for the United States. It is a recurring problem in many countries where lobbying pressures are strong.

For middle powers caught between Washington and Beijing, what is the real strategic dilemma today? Are they being pushed to choose sides, or are they finding ways to hedge through mini-lateral and sector-specific arrangements?

It is difficult for middle powers, but they do have options. They can pursue classic strategies of strengthening themselves and developing their own industries. If you look at South Korea and Taiwan, they have built advanced high-tech sectors. Taiwan and Korea are both major semiconductor producers, and Korea has also developed other high-technology industries, including automobiles and electric vehicles.

In the international realm, the picture varies. Taiwan faces a particularly acute dilemma because it is not recognized by China, and many other countries do not recognize it as a sovereign state. But countries such as South Korea, Malaysia, and Thailand are all trying to maneuver.

The United States made a strategic mistake by abandoning the effort to promote the Trans-Pacific Partnership, which included many Southeast Asian countries. The idea behind the TPP was, in some sense, to contain China through an institutional arrangement. Meanwhile, China has pursued the Regional Comprehensive Economic Partnership, or RCEP, which includes the ten ASEAN countries plus Australia, New Zealand, China, Japan, and Korea. RCEP is now the largest free trade agreement in the world.

We can now see the development of competing institutional arrangements. China is deeply embedded in RCEP, while the United States is no longer pursuing an institutional strategy. Instead, it has shifted toward unilateral and bilateral approaches and pays far less attention to mini-lateral or multilateral ones. That leaves middle powers trying to hedge and adapt in a more fragmented environment.

You suggest that the global order may not collapse so much as fragment. What would a fragmented economic order actually look like in practice for trade, supply chains, and international institutions over the next decade?

At least in theory, the World Trade Organization was supposed to cover all of global trade. We had a neoliberal trading order in which all countries, including China, the United States, and Russia, were members. But as we move into a more bipolar world, with a China-led sphere and a U.S.-led sphere, that model is becoming harder to sustain.

For a long time, businesses and business schools promoted global supply chains built around maximum efficiency. That often meant locating production in China or, in some cases, Mexico. The assumption was that you should source from the lowest-cost supplier. But crises such as the Japanese earthquake and the COVID-19 pandemic exposed how vulnerable those supply chains were.

As a result, many countries, especially the United States, began talking about reshoring and friend-shoring. Reshoring means bringing production back to the United States. Friend-shoring means moving production to countries such as Mexico or other trusted partners. But even that strategy has become less clear, because the United States has also gone after Canada, Mexico, and Europe. So it is not always obvious anymore who counts as a “friend.”

Chinese supply chains continue to move into places like Vietnam, and American companies have also gone into Vietnam and India. But with uncertainty around tariff policy, companies are unsure whether they will be able to manufacture abroad and still access the U.S. market. That uncertainty makes long-term planning difficult.

Therefore, we are going to see a fragmentation of supply chains. We may end up with something like a North American supply chain that includes the United States, Canada, Mexico, and perhaps parts of Latin America, alongside an Asian supply chain built around China, ASEAN, and other participating countries. Some of that is already happening in Asia because of RCEP.

In North America, however, the United States has still not decided how it wants to treat Canada and Mexico. That matters especially in sectors like automobiles, where supply chains are highly integrated across all three countries because of NAFTA and then the USMCA, which Trump himself negotiated. Now that agreement is coming up for renewal, and even auto executives are unhappy with efforts to restrict imports from Canada. So, fragmentation does not necessarily mean collapse, but it does mean a far less coherent and predictable system.

If the WTO and other postwar institutions are no longer fully equipped to manage trade in the current environment, where do you see the most meaningful adaptation happening? Is it inside existing institutions, through regional blocs, or through looser coalitions built around critical sectors such as semiconductors, green energy, and AI?

Part of the answer depends on how you think about international agreements. You can pursue a unilateral strategy, where you only rely on domestic policy. You can pursue a bilateral strategy between two countries. You can have a mini-lateral strategy involving a group of countries, like the TPP or RCEP. And then you can have a global or multilateral strategy such as the WTO.

But there is another dimension too. Agreements can be global and cover many products, or they can be sector-specific. There can be sectoral agreements on textiles, for example, or broader agreements covering nearly all products. There is also a combination of these forms in regional frameworks, as with the USMCA or the European Union.

What that means is that there are many possible configurations. The earlier hope was that bilateral and mini-lateral agreements would eventually become nested within the WTO and reinforce it. Jagdish Bhagwati famously described this as the distinction between “building blocks” and “stumbling blocks.” But that is no longer clear.

Instead, we may be heading toward a world of overlapping bilateral, regional, sectoral, and mini-lateral agreements that are not fully consistent with one another. The United States has negotiated bilaterally with many countries. China has done the same. The European Union represents a broad regional arrangement. APEC is a looser forum that is not even really a trade agreement.

The likely result is greater fragmentation and less attention to the WTO. We are moving away from a unified global system and toward a patchwork of agreements that may or may not fit together. That makes life much harder for firms and governments alike, because it becomes increasingly difficult to know which rules apply in which context.

To what extent are the current challenges to maintaining an open global trading order a function of the current U.S. administration, as opposed to structural challenges that extend beyond the Trump administration, such as differentiated labor costs, capital mobility, or political polarization in advanced industrial societies?

The first point to keep in mind is that the neoliberal order the United States created after World War II served the country very well overall, except in a few respects. We did not do enough to promote labor mobility, and some industries that came under pressure, such as steel, textiles, and later automobiles, were not always able to adjust effectively to global competition.

The United States has not done well in developing policies to help workers adjust. President Trump was able to tap into very real insecurities among people losing jobs and build a coalition against the global neoliberal order. This backlash had been building for a long time. Trump did not invent it, but he seized on it and ran with it.

One of the striking political shifts is that Republicans, traditionally the party of free trade, has become the party of protectionism, while Democrats, who were often more protectionist over the past half century, are less so now. It is a remarkable reversal.

What many neoliberal economists underestimate is how difficult adjustment really is. There has long been wishful thinking that workers will simply find new jobs, get more education, and transform their skill sets. But getting a steel worker to become an AI worker is not trivial. It requires entirely different training. And Americans are going to college less than before.

This is where AI becomes especially important and AI may ultimately pose an even more fundamental challenge than trade. A great many white-collar jobs are going to be threatened. We still have not developed a serious plan for how to handle that. Some people talk about universal basic income, but the broader issue is whether new jobs will actually appear and whether displaced workers will have the skills to do them.

We can already see this problem. Samsung and TSMC, for example, are investing in Arizona and Texas, but they say they cannot find workers with the necessary skills. We need much stronger efforts at both the state and federal levels to train people for new industries. In Europe, this is often described as active labor market policy: helping people maintain healthcare, subsidizing them during transition, and supporting them as they move into new jobs. In the United States, we have more of a welfare model for those who cannot find work, but not a well-developed transition system.

That matters even more now because AI is disrupting the labor market in ways we are only beginning to understand. Many entry-level jobs in law, consulting, and software may shrink. Even at Berkeley, computer science enrollments are declining because students can already see that some software work can be done by AI systems like Claude.

So we need to think about the domestic economy, the international economy, and technological change all at the same time. The government’s job is not just to provide security in the narrow sense. It is also to make sure people are healthy, economically secure, and able to adapt. If we fail to manage these transitions, we will see more backlash, more lobbying for protection, and more political instability. That is the deeper structural problem.