Trump 2.0: Semiconductor Industry Policy Dynamics — DSET Interviews Jimmy Goodrich

Introduction

Jimmy Goodrich is a senior advisor for technology analysis at the RAND Corporation, senior associate at the Center for Strategic and International Studies, and nonresident fellow at the University of California Institute on Global Conflict and Cooperation. With nearly two decades of experience in the technology sector, Goodrich has established himself as a leading expert on the intersection of technology, geopolitics, and national security. As the former Vice President for Global Policy at the Semiconductor Industry Association, he led global policy and supply chain initiatives as well as efforts to successfully secure $52 billion in funding for the CHIPS and Science Act. Prior to his role at the SIA, Goodrich also directed China policy at the Information Technology Industry Council and worked in China’s technology sector for seven years. Ahead of Trump’s inauguration, DSET had the opportunity to speak with Goodrich, whose extensive experience and insights continue to shape critical conversations on Taiwan-US economic security cooperation.

Trump 2.0: U.S. Tech Policy Toward China

DSET: How do you think the Biden and Trump administrations will differ in their approaches to China tech policy? What do Trump’s cabinet picks suggest about trade and export control policies?

Jimmy Goodrich: There is bipartisan consensus amongst D.C. policymakers that China is one of the most important strategic challenges facing the United States and its allies. Toward the end of the Obama administration, we saw increasing concern about China’s semiconductor developments. Their Secretary of Commerce gave a speech in the last month of their administration noting the rise of China’s semiconductor industry. 

The Trump administration then took a number of actions, including strengthening CFIUS to block acquisitions, implementing export controls, and inviting TSMC to build a fab in the U.S. This sparked the idea for the CHIPS Act. The Biden administration then took the baton, rallying Congress to pass the CHIPS Act and overseeing its implementation. They also expanded Trump-era export controls with major rule packages in October 2022, October 2023, and most recently in December 2024.

We don’t know what the final cabinet will look like. However, the nominees include known China hawks like Marco Rubio for Secretary of State and Mike Waltz for National Security Advisor. Even the nominees with finance backgrounds have expressed strong stances on China, particularly on tariffs, in op-eds and podcasts. There is a broad consensus on being tough on China for legitimate reasons: concerns over Taiwan Strait stability, the South China Sea, human rights, unfair market access, industrial subsidies, and dumping practices. Over the past eight years, these issues awakened the U.S. public and political establishment to the economic and technological challenges posed by China.

This stance is also popular with voters. Polling by Pew and others shows incredibly low favorability ratings for China in the U.S. This is an issue where politicians enjoy broad bipartisan support. For many of the officials shaping the U.S. China tech policy, this is not just political—they genuinely believe in their mission.

So far, President Trump’s cabinet picks reflect his preference for hearing from different personalities with different views. He likes having different ideas put forward and debated to see who can win based on the merits of their argument. 

Waltz has worked on a think tank effort looking at China’s technology in AI. He seems to support strong export controls but simultaneously believe that the U.S. should not burden domestic companies with obtrusive regulation like the Biden administration’s executive order on AI and utilization of the Defense Production Act to require companies to report algorithms and models to the government. This is an area where Republicans and Democrats disagree. Whereas the Biden administration was considering the long-term risks of AI safety, Republicans believe this places too much burden on industry. 

The disagreement is not about the fact that there are risks to AI. We have heard from Waltz that AI, if misused by the PRC, could be a national security threat to the U.S. However, Republicans less frequently talk about AI safety issues emanating from domestic companies.  

The Future Prospects of the CHIPS Act

DSET: Do you expect Trump, with a Republican Congress, to push for a second CHIPS Act? 

Jimmy Goodrich: There is strong bipartisan support for enhancing U.S. domestic semiconductor production. There is also strong support for participation by foreign invested enterprises. Let’s not forget that the Trump administration invited TSMC to build a fab in Arizona. Congress and the administration said that this should not only be about American companies. And frankly, the U.S. has no choice but to work with Taiwan, given its leading role in cutting-edge chipmaking. In fact, this started with supporting Taiwanese companies in the U.S. Moreover, Intel’s funding is now being reduced. TSMC may be the biggest breadwinner at the end of the day. 

But on both sides of the aisle, politicians and experts want to see more investment in the U.S. TSMC’s investments in the U.S. are fantastic, but they want more advanced technology to be produced at a larger scale in Arizona. That may take time. TSMC has already committed to several expansions, but can more be done, and how? 

When the tax credit for the CHIPS and Science Act expires in a few years, I am hoping that Republicans will support its renewal because they generally support tax policy incentives. But we need to see how current projects pan out to know whether there will be a CHIPS 2.0.

At the end of the day, the supply chain is still global. The U.S. cannot build everything it needs on its own shores, and it shouldn’t. We need to rely on our allies, whether that is Taiwan, Japan, Korea, Europe, Singapore, India, Vietnam, or Malaysia. Their role is not decreasing in this supply chain.

DSET: How do you think Trump’s “America First” agenda will play into whether a second CHIPS Act or other policies favor American companies?

Jimmy Goodrich: My understanding of “America First” does not mean that the U.S. doesn’t work with allies. For instance, the Trump administration worked with the Netherlands to coordinate export controls on extreme ultraviolet lithography. 

The difference between the Biden and Trump administrations is that the Biden administration has allies first and America as equal, whereas the Trump administration always has America at the core of their interests. When interests align, they are willing to work together. When they don’t, the Trump administration is willing to use U.S. leverage more forcefully than the Biden administration. In practice, this means that the U.S. government could threaten reduced military assistance or tariffs because they will not treat you as an equal if you are not willing to help them. I scratch your back, and you scratch mine.

Trump’s Tariff Policies

DSET: Do you think tariffs could be implemented against Taiwan to attract not only TSMC but also Taiwanese firms in advanced packaging, materials, and server assembly supply chains to the U.S.? 

Jimmy Goodrich: Obviously, tariffs are the Trump administration’s favorite tool. There is no doubt that we will see increased usage of tariffs. That being said, there is more bipartisan consensus around the usage of tariffs than you’d expect. The Biden administration did not fundamentally roll back the tariffs on China. They also imposed or increased tariffs on strategic items such as electric vehicles and semiconductors from the PRC.

Do tariffs work? The semiconductor industry is really complex. Design, front-end, back-end, and integration often occur in different countries. A simple tariff on Taiwan to force more production in the U.S. wouldn’t necessarily work. There may not be as many chips flowing into the U.S. from Taiwan as you’d think. For instance, many of them might be assembled in Malaysia then integrated into a product in Mexico that is ultimately imported to the U.S. in an AI server. 

The best approach to encourage onshore production is to cultivate talent and have an attractive investment requirement, low cost energy, tax and fiscal incentives, and favorable regulations––not to threaten tariffs. 

The Trump administration has talked extensively about how they want to reinvigorate the American economy and boost investment in manufacturing. If America can further lower energy costs, reduce regulation, and have good fiscal tax policy, then I think they could achieve that objective.

DSET: Looking at potential policy actions, could Taiwanese firms collaborate with Silicon Valley to prevent such tariffs, possibly forging alliances with influential figures like Elon Musk? Alternatively, what strategies would you recommend for Taiwan to mitigate these risks?

Jimmy Goodrich: Taiwanese and American companies are already tied at the hip. Whether it’s xAI, Dell, Nvidia, or Intel, everyone is dependent on each other. Taiwanese companies are dependent on U.S. IP, software, equipment, and materials. U.S. companies are dependent on Taiwanese companies for contract manufacturing. And Taiwan itself is dependent on many of the finished products. This is a uniquely intermeshed technology ecosystem. Just look at TSMC and its investment in Arizona. Many of TSMC’s Taiwanese suppliers who do engineering, materials handling, or environmental engineering are coming into Arizona and investing as well. These are great success stories.

Challenges in Export Controls

DSET: What do you see as the primary gaps in U.S. unilateral export controls?

Jimmy Goodrich: Successive administrations have issued hundreds of pages of regulation on both unilateral and multilateral controls to expand the scope of technology subject to control: EUV then DUV lithography, AI chips, HBM memory, or advanced node production facilities for NAND, DRAM, and logic. However, it is still narrow compared to the overall scope of U.S.-China trade in semiconductors. 

A big deficit lies in oversight and implementation. Huawei managed to gain access to TSMC, and Chinese AI companies have either smuggled in tens, if not hundreds, of thousands of GPUs or used large data centers outside of China. Furthermore, public reports from Bloomberg and SemiAnalysis show Huawei has built up a large network of production facilities. A recent SemiAnalysis piece showed how somebody can connect two fabs as a way to possibly avoid export controls 

The most surprising thing has been the US government’s slow response. China is a large country with huge resources. Like any country, they are agile in responding to US export control restrictions. It should be assumed that any country will have a counter-strategy to any U.S. action, but the U.S. response to that has been pretty slow or inadequate. The Biden administration said that this third round of controls has addressed the circumvention, but the jury is still out on whether or not it will. It’s a big question. Does the U.S. government have enough resources to do what SemiAnalysis called the “whack-a-mole game”?

There are increasing calls from Republicans to strengthen implementation and oversight of the regulations. The House Foreign Affairs Committee and the Select Committee on China have issued statements about this over the last few months. This will be a space to watch.

DSET: When evaluating whether the scope of export controls should include legacy chips, what considerations should be factored into the discussion?

Jimmy Goodrich: Export controls, China’s domestic market, and China’s pre-existing stated objective to build a self-sufficient fortress economy based on dual circulation have all led China to rapidly expand the pace of its capacity addition in front-end semiconductor manufacturing, primarily for 200-300mm legacy logic or foundational semiconductors, which are essentially 20-180nm chips. These go into electric vehicles, solar panels, IoT devices, etc. Even an advanced server will have foundational semiconductors fabricated on larger feature sizes that do things like regulate the power and temperature of the server. 

China is building the largest number of foundational semiconductor fabs, with over 40 planned or under construction. In steel, aluminum, shipbuilding, LED displays, electric vehicles, and batteries, we have seen big Chinese boosts in capacity not necessarily aligned with market demand that create pressure on incumbents outside of China, who don’t have access to the same state capital or subsidies and just cannot compete. This is what market analysts and some in Washington are concerned about. 

It is already impacting Chinese companies. An article in Caixin business magazine this year said that even SMIC is facing pressure from domestic Chinese startup foundries that are undercutting them on pricing. The question is if or when will it impact companies outside of China. Right now, we are seeing the Chinese economy become more domestic-oriented. More domestic capacity is being filled by domestic companies. But at some point, will they export? That has been the story of every other industry. 

Furthermore, this could lead to dependence. If China does dump products in overseas markets, then Ford Motor or Toyota could become dependent on the Chinese market for these foundational chips. There are arguments for and against whether it is happening. But it has happened in many other sectors, so the concern is warranted. 

The solutions to this challenge are complicated. The first difficulty is that China’s self-sufficiency in mature node chips is much stronger than it is in the advanced chips. They have an increasingly competitive semiconductor equipment industry and a materials industry that can provide most of the technology needed down to 90nm. That still means China has foreign dependencies for 65-28nm mature nodes, but China is making progress. The second difficulty is that there is a larger set of countries that can make equipment and materials that feed into the foundational chip market, making allied coordination more complex. Japan, the Netherlands, Taiwan, South Korea, Israel, Singapore, Malaysia, Germany, Austria, and Italy all make equipment feeding into that foundational chip market. 

Beyond export controls, we can look at trade remedies like countervailing duties, tariffs, or border restrictions. There is a system largely based on the WTO that can address the dumping of products with low-priced Chinese chips, although the WTO is not operating how people hoped it would. 

Right now, we are in a pretty upward boom cycle for some parts of the semiconductor industry, particularly advanced logic and memory. Legacy is a little bit more flat. We are also seeing declines over the smartphone market and other areas. If we saw a big drop in demand and bust of the AI market, for example, that could certainly really hurt both of these sectors. That’s out of the control of politicians.

DSET: How might the incoming administration impact progress toward successful multilateral export control regimes?

Jimmy Goodrich: It is too early to tell. What will the overarching framework of the Trump administration’s approach to China be? How will China approach the Trump administration? Are they going to proactively go on a charm offensive? Will there be another Mar-a-Lago summit with the chocolate cake? Or will it be a “wait and see” where neither side wants to make the first move. Will Trump go all-in on tariffs as he said he would? And what will China’s counter-strategy be?  

With six to eight years to think about how to respond, China is more prepared than in Trump 1.0. They have doubled down on dual circulation and a fortress economy. They have developed a big toolkit of regulations, including counter-sanctions, sanctions-blocking rules, restrictions on rare earth and materials, and their own Unreliable Entity List. They have already sanctioned Micron and threatened an investigation into Intel. Just this week, government trade associations put out a coordinated message saying “Don’t trust American chips” so they could strengthen that restriction. We are all focused on Trump, but don’t forget that China has a say in everything too.

Impact of Latest Investment Review Measures

DSET: The U.S. recently issued a Final Rule to implement outbound direct investment (ODI) screening. What is your perspective on the purpose or anticipated impact of this new program?

Jimmy Goodrich: The US is playing catch-up to Taiwan, which has had an outbound investment screening regime in semiconductors for quite some time. The Biden administration adopted similar rules in the executive order and CHIPS Act guardrails. Recipients of funds from the CHIPS Act office cannot do a lot of expansion in China. Of course, there is the outbound investment legislation as well. 

Foreign investment into China has dropped off in the semiconductor space, although Foxconn has maintained some investments and TSMC, SK Hynix, Samsung, Intel, and a few other companies still have their fabs there. Interestingly, European companies are quietly expanding their presence. STMicroelectronics, a French and Italian analog chip maker, built a joint venture factory in Chongqing and announced a new partnership with Huali. Just today, NXP said they are open to expanding production cooperation in China. This is not a one-way street with everyone exiting. Others are still going into the market.

Advancement of US-Taiwan Strategic Export Control Partnership

DSET: In practice, U.S. extraterritorial controls prevent major Taiwanese semiconductor firms from conducting business with Huawei, reducing the incentive for the Taiwanese government to strengthen its domestic export controls. What are your insights regarding this gap between the U.S. and Taiwan in strategic approaches to technology export controls?

Jimmy Goodrich: The U.S. has the most robust and aggressive export control regime with regards to the PRC. If Taiwan is falling short in any area, many other countries are too. That being said, Taiwan is the hub for global semiconductor production as well as home to the contract outsourced chip manufacturing business model and lots of really important talent. What Taiwan does matters significantly to the security of the global semiconductor ecosystem. 

Having worked with government, business, and academia between the U.S. and Taiwan on this exact question for over a decade, I think Taiwan has a stronger focus than the U.S. in some ways but weaker focus in others. Taiwan is very focused on preserving the security of its own industrial base–perhaps more so than the U.S. On the flipside, Taiwan is less concerned than the U.S. about the export of its technology to the PRC in ways that could be misused.

Many years ago, Taiwan established explicit rules about what Taiwanese companies could and could not do in China. The preeminent goal was preserving Taiwan’s global leadership. They did not want TSMC, UMC, or Powerchip to offshore critical capability while building factories in China. The N minus two then N minus one rules required companies to build at least one fab in Taiwan for every fab built in China. Fabs in China had to be a certain number of technology generations behind those in Taiwan, and any investment in China had to be approved. 

In addition, Taiwan has a very robust regime preventing the theft of intellectual property and, more importantly, talent by China. Taiwan strengthened its economic security and trade secrets protection laws, and its Ministry of Justice has aggressively pursued violations. Taiwan is one of the only countries to completely prohibit all investment by Chinese chip design firms. They have also made headhunting on behalf of Chinese companies illegal. These are all things that the U.S. can learn from Taiwan. South Korea also does very well in this area.  

Surprisingly however, Taiwan’s government currently has a laissez-faire approach toward high technology dual-use trade with China. Unlike the U.S., Taiwan has not created an entity list, maintained an end user list, or established a military end use rule requiring Taiwanese companies to determine whether the items they sell could end up in the Chinese military. The perspective in Taiwan has been to look the other way, or “睜一隻眼閉一隻眼 (turn a blind eye).”

The recent example of Bitmain and TSMC is just the tip of the iceberg. Several years ago, it was reported that Taiwanese companies were producing for a PLA-owned company. The Washington Post also investigated the export of machine tools from Taiwanese companies to China and Russia. Taiwan lacks strong enforcement or even in some cases regulation of dual use technologies that could be used for military systems in China. While Taiwan observes the Wassenar Arrangement, many now perceive that as insufficient on its own without stronger end-use and technology-based controls outside the scope of these arrangements. 

Ironically, Taiwan has the most to lose from that weakness. Taiwan is directly staring down the military threat from China more so than any other country. Many policy analysts in D.C. are perplexed that the Taiwanese government and society haven’t paid attention to this. In the Trump administration, a senior White House official told the Financial Times that Taiwan was selling China the chips in the missiles that are pointed back at them.