Editor's note:
One of the most complex issues in contemporary international relations is the status of the self-governing entity of Taiwan and its government in Taipei, formally called the Republic of China (ROC). Is Taiwan a sovereign state? Who gets to decide? What is the legal relationship between the ROC and the People’s Republic of China (PRC)? How much international support for Taiwan is grounded in law and how much in policy? Does international law recognize Taiwan’s right of self-defense or the right of its friends to come to its aid?
During 2024-2026, the U.S.-Asia Law Institute (USALI) hosted a series of speakers to address these and related questions from different perspectives. In this November 6, 2025 talk, Pasha Hsieh, the Lee Kong Chian Professor of Law at the Singapore Management University Yong Pung How School of Law, explains how states that lack formal diplomatic relations with Taiwan nonetheless enter into trade and bilateral investment agreements with it, blurring the line between recognition and non-recognition.
The following excerpts have been edited lightly for clarity and brevity. A full recording of the program can be found here. The discussion was moderated by Katherine Wilhelm, executive director of the U.S.-Asia Law Institute.
***
By Pasha Hsieh
My research focuses on trade agreements and recognition policy. These two topics may sound quite different, but they can be closely intertwined. I’m a lawyer by training, and my Taiwanese background makes me quite interested in issues of recognition and non-recognition in international law. At some point, I came across works written by political philosophers and international relations scholars. They discuss recognition theory, but often from very different angles. They frequently refer to what legal scholars have written but not the other way around. My goal is to bridge the gap between law and political science on the issue of recognition theory. A few years ago, I received two separate messages from my friends who work closely with European and Asian governments. Interestingly, they both asked me the same question, whether trade agreements with Taiwan would constitute implied recognition. That was the first time I realized that this topic carries real practical significance for foreign governments. My talk today focuses on my forthcoming article on the US-Taiwan trade agreement, which is part of my broader book project titled Beyond Non-Recognition.
Let me first discuss how international lawyers and IR scholars understand recognition. First, recognition, statehood, and sovereignty are actually three distinct concepts. Recognition is a unilateral political act with legal consequences such as entitlement to sovereign immunity. The recognition of states concerns whether an entity is recognized as an international person, whereas the recognition of governments concerns which government represent the states. International law textbooks will usually classify Taiwan or the Republic of China under [issues of] government recognition. But today, what really matters in legal practice is state recognition, whether Taiwan can be treated as if it were a state.
Recognition can be explicitly conferred or implied. Scholarly work and judicial cases suggest that concluding a treaty can amount to implied recognition. An example is an FCN Treaty – a treaty of friendship, commerce, and navigation. Now a controversial question is whether a modern trade agreement can be considered as a treaty in this sense.
Recognition in international law involves the debate between constitutive theory and declaratory theory. Influenced by positivism, the constitutive theory holds that recognition [by other states] is essential to statehood. In contrast, the declaratory theory asserts that recognition only serves as a declaratory acknowledgement of that fact. For most international law, the debate is largely settled. The Montevideo Convention [on the Rights and Duties of States, signed in 1933 by the United States and 18 Latin American and Caribbean countries] clearly says that the political existence of a state is independent of recognition. That’s the declaratory theory.
What is sovereignty? It includes both internal and external sovereignty. External sovereignty, also known as international legal sovereignty, relates to a state's independence and its recognition by other states. Statehood and sovereignty overlap in the element of independence, but diverge when it comes to recognition.
International law largely treats recognition and non-recognition as a binary exercise, either black or white.
International law largely treats recognition and non-recognition as a binary exercise, either black or white. When we talk about non-recognition, it is often associated with the duty not to recognize illegally annexed territories. A typical case is the Stimson Doctrine regarding Manchukuo, Japan's puppet state during World War II. However, the International Court of Justice created a very narrow exception in the [1971] Advisory Opinion on Namibia. The court said that certain public law acts, such as birth and marriage registration certificates, could be recognized despite the non-recognition of the illegal regime.
Now, let me explain the evolution of recognition in political science. Hegel first developed recognition theory in the context of inter-subjective human behavior. According to him, the existence of the self depends on the recognition of the other. A classic example is that the existence of a master depends on the recognition of his or her slave. From around 2010, a new trend emerged. International relations scholars have applied Hegel’s recognition theory to interpret interstate relations. Traditionally, neo-realists argued that states are primarily motivated by the pursuit of security and material power. Modern IR scholars argue that this is not a whole story. Influenced by the constructivism, they claim that struggles for recognition of identity, such as prestige as a great power, also motivates state behavior.
Another group of international relations scholars goes further. They argue that, in addition to the psychological element of identity, states also struggle for status, which demonstrates dignity based on normative standards. International law is particularly relevant here, as it provides normative benchmarks for status recognition. Overall, international relations scholars believe that recognition is still constitutive and that it is not binary. There are diverse and gradual forms of recognition, even within the space of non-recognition. However, they admit that these forms are difficult to measure.
Overall, international relations scholars believe that recognition is still constitutive and that it is not binary. They are diverse and gradual forms of recognition, even within the space of non-recognition.
I try to address this methodological challenge by using trade agreements as a lens. My overall argument is that collective forms of legal recognition under trade agreements can blur the line between implied recognition and non-recognition. The extent of legal recognition, in fact, far exceeds the Namibia exception that I talked about.
After discussing the theories, let me look at the empirical information on US trade agreements. Traditionally, there are three types of agreements: free trade agreements (FTAs), bilateral investment treaties (BITs), and trade and investment framework agreements (TIFAs), which provide a framework for dialogue with the aim of eventually reaching a BIT. Taiwan currently has nine FTA-style agreements, including ECFA (Economic Cooperation Framework Agreement) with China, as well as more than thirty bilateral investment treaties.
In the US, these traditional approaches are no longer the favored method. Both the Biden and Trump administrations have relied on non-traditional agreements, such as the IPEF (Indo-Pacific Economic Framework for Prosperity, initiated in 2022) or reciprocal trade agreements. So far, the US has only concluded three full-fledged FTAs in the Indo-Pacific, with Singapore, Australia, and Korea. …
President Biden viewed the IPEF as a key pillar of his Indo-Pacific strategy. The IPEF is not an FTA, as it contains no tariff or service commitments. It's organized around a four-pillar structure, with each pillar intended to result in a separate agreement [on trade, supply chains, clean energy, and a fair economy]. … Some ASEAN (Association of Southeast Asian Nations) countries were quite concerned about Taiwan's participation in the IPEF. As a result, Biden signed a separate deal called the US-Taiwan Initiative on 21st Century Trade, which is the focus of my article. It is like a mini-IPEF. Both sides signed the first agreement under the initiative in 2023, and Congress passed the Implementation Act for this agreement.
… Let's now zoom in on US-Taiwan agreements. Toward the end of World War II, the United States signed an FCN Treaty with the ROC government. This treaty represents the very first generation of US trade agreements. The FCN Treaty remains valid in both the United States and Taiwan under the Taiwan Relations Act, which Congress enacted following President Carter's de-recognition of the ROC. From the 1990s, bilateral interactions became more normalized. With the signing of the US-Taiwan Trade and Investment Framework Agreement in 1994 and the Customs Mutual Assistance Agreement in 2001, the two sides also established additional dialogue mechanisms. All of this initiative culminated in the US-Taiwan Initiative First Agreement in 2023.
US congressional members have regularly written letters to the Office of the US Trade Representative and the president, urging the signing of an FTA with Taiwan.
What are the policies that prompted the US-Taiwan trade agreement? Of course, it relates to Chinese President Xi Jinping's escalating approach to Taiwan, including gray-zone military actions and economic coercion. China has also tried to squeeze Taiwan's international space by persuading ten countries to switch recognition to Beijing since 2016. We should note that the termination of diplomatic relations does not automatically lead to the termination of FTAs. Panama is an example. It severed ties with Taiwan in 2017, but the FTA continued to remain operational. …
US congressional members have regularly written letters to the Office of the US Trade Representative and the president, urging the signing of an FTA with Taiwan. … To understand the US-Taiwan Initiative, we first need to determine the legal nature of the trade agreement, as only a treaty can trigger implied recognition. Under WTO law, the answer is quite straightforward. The US-Taiwan agreement is neither an FTA nor an interim agreement, as it does not meet the “substantially all the trade” requirement [under the General Agreements on Tariffs and Trade, Art. XXIV:8(b)]. But that doesn't fully answer the question. How about the Vienna Convention on the Law of Treaties? The Vienna Convention adopts a broad view by defining a treaty as an international agreement concluded between states. The convention also envisions that the agreement may be concluded by a subject of international law other than a state. The ICJ has also broadly interpreted treaties, saying that all instruments such as minutes of meetings or MOUs can be regarded as treaties so long as they have a binding character. From its substance, the US-Taiwan Initiative First Agreement is far more detailed than the FCN Treaty that I alluded to. And of course, for implied recognition to arise, there must be the government's unequivocal intent.
The ICJ has broadly interpreted treaties, saying that all instruments such as minutes of meetings or MOUs can be regarded as treaties so long as they have binding character. And of course, for implied recognition to arise, there must be the government's unequivocal intent.
In 1929, the [US] Court of Appeals for the Ninth Circuit dealt with an interesting case. It’s actually about an insurance policy that was signed by the Beiyang government in Beijing. But Chiang Kai-shek’s government in Nanjing demanded payment. So, there was a two-ROC problem. The court held that, due to various factors including the signing of a treaty, implied recognition was accorded to Chiang's ROC government.
In the 2000s, there were two interesting cases about Singapore Airlines. Both cases involved Taiwan's sovereign immunity, and the Singapore Court of Appeal said that the bilateral [trade] agreement [between Singapore and Taiwan] only indicated cooperation, and Singapore did not have the intent to recognize Taiwan. This was quite different from the Canadian court's decision [which found that Taiwan fulfilled the criteria for statehood and therefore recognized Taiwan’s sovereign immunity]. …
What does the US-Taiwan Initiative actually recognize? From an international perspective, it recognizes key statehood criteria. First, it recognizes the defined territory of Taiwan. The scope of ROC territory was actually controversial when Taiwan and the United States negotiated their mutual defense treaty, as the treaty only included Taiwan and Penghu. It did not extend to Kinmen and Matzu, two small islands near China. The US-Taiwan Initiative avoids the problem by simply stating that the territory is what Taiwan's representative represents. The US-Taiwan Initiative also recognizes the government. It includes references to government agencies in charge of defense and foreign affairs, as well as the president.
Since 1979, about 370 agreements have been signed through two de facto embassies, technically NGOs: the American Institute in Taiwan, or AIT, and TECRO, the Taipei Economic and Cultural Representative Office in Washington, D.C. What is unique in the US-Taiwan Initiative is the use of designated representatives, allowing these de facto embassies to delegate their authorities to government agencies, including the USTR and Taiwan's Office of Negotiations. In other words, the agreement facilitates direct government-to-government talks, and the committees established under the agreement also institutionalized this dialogue.
The agreement facilitates direct [US-Taiwan] government-to-government talks, and the committees established under the agreement also institutionalized this dialogue.
Finally, the independence requirement is also addressed. This is often the most challenging statehood criterion for Taiwan. After the agreement, the PRC warned the US not to conclude any agreement with sovereign implication or official nature. Sovereignty manifests independence. So even Beijing agrees that the agreement reinforces Taiwan's capacity to enter into relations. And these legal recognitions also re-enforce Taiwan's identity and status, and they further provide empirical support for the international relations argument that recognition can take gradual forms.
The US-Taiwan Initiative First Agreement covers five areas, and it is expected that the second agreement will cover the remaining areas, including agriculture, environmental, and labor issues. Many of these issues are also comparable to those under Trump's reciprocal trade agreements. It is generally agreed that, although there are no tariff commitments, the agreement helps reduce trade costs by facilitating customs clearance. There are additional benefits and cooperation on top of the existing customs agreement and the WTO Trade Facilitation Agreement. Examples include fewer formalities for express shipments and the recognition of electronic invoices and data. …
This research is about the intersection of recognition theories in international law and international relations using the recent US-Taiwan trade agreement as a case study. I hope it sheds light on the ongoing development of global regionalism and the Indo-Pacific strategy. I welcome your comments and questions.
***
Q&A
Wilhelm: Are you adopting the view that statehood requires recognition of other states, or just saying that this is how IR scholarship views the situation? And as a corollary, are you arguing that Taiwan’s trade and investment agreements constitute the necessary implied recognition?
Hsieh: In terms of the constitutive theory and declarative theory, we know that it's no longer a debate in international law. I was a member of the International Law Association Committee on Recognition/Non-recognition in International Law, and although we provided some examples about the continued pragmatic significance of the constitutive theory, we international lawyers generally agree that the prevailing theory is the declaratory theory. So I was quite shocked when I read IR literature saying that, actually, international lawyers got this wrong. For state action, they still find that the concept of recognition is important because the struggle for identity and status will actually motivate state action. So, from their perspective, the constitutive theory is still useful.
Back to the trade agreements. As I said, Taiwan currently has nine FTA-type agreements. Taiwan also has more than thirty bilateral investment agreements, as well as double taxation agreements. I think it would be an exaggeration to say that all of these agreements would trigger implied recognition. But it would be reasonable for me to assert that the collective agreements will actually blur the line between non-recognition and implied recognition.
I think it would be an exaggeration to say that all of these agreements would trigger implied recognition. But it would be reasonable for me to assert that the collective agreements will actually blur the line between non-recognition and implied recognition.
A very interesting issue about implied recognition is that implied recognition usually depends on the intent of the government. However, it is for the government to interpret its intent. So, it depends on what the government says or how the court interprets the government action. In practice, a government can sign extensive agreements with Taiwan. So long as it says it upholds the one-China policy, there would be no implied recognition. I think this is the approach of many countries when they engage Taiwan in many international agreements.
Usually for FTAs and bilateral investment agreements, they will not include a one-China policy in the agreement, but of course they will have an interesting arrangement for some sensitive terms. For example, how do you deal with investor-state disputes if you don't use the word “state” in the agreement. So many countries in their agreements will use “authorities” instead of “state.” Another interesting issue … in the context of Taiwan, [governments] often use “BIA,” where “A” means agreement or arrangement instead of a treaty. This has sort of toned-down sovereign implication.
Wilhelm: How does enforcement work without formal diplomatic relations?
Hsieh: All the [US-Taiwan] agreements were actually concluded not by government agencies per se, but by de facto embassies, representative offices. They are technically NGOs. Can those NGO agreements bind the states? It's actually a public law issue raised by some scholars. But so far, it only stayed at the academic discussion level. It's safe to say that all the countries that have concluded all these agreements with Taiwan have not challenged the binding effect of these agreements. I think it's more precise to say that they treat those agreements as if they were binding.
Wilhelm: Is Taiwan pursuing trade and investment agreements as a deliberate strategy to expand Taiwan's international recognition and to strengthen its status and legitimacy in the world? Or are they pursuing them simply because of the pragmatic need to have agreements to facilitate trade?
Hsieh: The so-called domino theory helped explain why Taiwan launched its FTA strategy in the 2000s, as Taiwan was quite concerned that the trade diversion effect of other FTA would undermine Taiwan's exports. [Former President] Tsai Ing-wen’s New Southbound Policy emphasized concluding more investment agreements with Southeast Asian countries to provide better investment protection for Taiwanese companies. The overall objective is to protect Taiwan's trade and investment interests, and not necessarily pursue a different form of recognition. But of course, those agreements naturally come with different forms of legal recognition, which is also important to Taiwan.
Wilhelm: In IR theory, is there any discussion of the practical value of implied recognition? In other words, is it any use?
Hsieh: Implied recognition is actually an international [law] term; it's not an IR term. Implied recognition and explicit recognition are actually the same in terms of legal effect. Why can it be important? Because once recognition occurs, regardless which type of recognition, the entity, such as Taiwan, will be entitled to certain benefits such as sovereign immunity, which is quite important, as well as diplomatic privileges and immunities, which are quite significant for Taiwan’s de facto embassies. Let me give an example: should Taiwanese diplomats pay taxes in the host countries? And can Taiwan use the diplomatic bag? These are the very practical issues that Taiwan constantly faces, and implied recognition could help overcome these issues. Of course, all these issues have been negotiated bilaterally so far. But implied recognition would be an important threshold for these rights.
Audience: What has been the PRC attitude toward trade agreements with Taiwan that do not imply recognition?
Hsieh: The interesting thing about the PRC position is that there's no fixed position. For [former Taiwan President] Ma Ying-jeou’s government, it's okay to sign agreements with New Zealand and Singapore. I recently read a memoir by former Taiwan Foreign Minister Chien Foo. He talked about the [2013 economic partnership] agreement between Singapore and Taiwan, and actually, Singapore got sort of an implicit endorsement from [PRC Foreign Minister] Wang Yi. … This is what happened during the Ma administration. But for the Tsai Ing-wen administration and the Lai Ching-te administration, I think the answer is quite clear. Any agreement with Taiwan will not be tolerated.
Wilhelm: The US-Taiwan Initiative First Agreement took effect in January 2025 right ahead of the change of US administrations. Has there been any change in how it is implemented over the transition from Biden to Trump?
Hsieh: After both sides concluded the First Agreement in 2023, they started discussing the second agreement. The most recent meeting took place in 2024. Now Trump came in and Trump had a very different strategy, an America First strategy. Trump is now negotiating so-called reciprocal trade agreements with various countries, including Taiwan. There's no plan for Trump to abandon the US-Taiwan Initiative. In fact, Trump cannot abandon it [because Congress implemented it in law.] But there is no update on the second agreement. But we should also be optimistic because I recently read the text of US reciprocal trade agreements with Cambodia and Malaysia. If you compare them to the US-Taiwan Initiative or IPEF, there were actually quite a lot of overlapping issues, such as digital trade, labor, and environmental issues. So Taiwan will continue to discuss those issues with the United States, just under a different agreement. The priority is to negotiate a reciprocal trade agreement. [Update: The US and Taiwan signed an Agreement on Reciprocal Trade on February 12, 2026. As of May 2026, Taiwan’s Legislative Yuan has not yet voted on it.]
Audience: It seems that as long as economic benefits can be maintained, the Americans will remain a protector of Taiwan. That is, the relationship is transactional. Does Taiwan need to remain relevant and helpful to US interests?
Hsieh: Of course, Taiwan wants to be relevant. Many observers would pay attention to semiconductors. TSMC has announced a $165 billion investment in the United States. … Investment commitments by Taiwan will be important to the Trump administration, but we also need to realize that it will take a long time. It only takes nineteen months to build a semiconductor fabrication plant, commonly known as a fab, in Taiwan, but it takes more than thirty months in the US. And the new H-1B visa policy will also increase the labor costs. And cultural differences are another challenge. TSMC is actually facing a class action lawsuit brought by almost thirty US employees alleging different forms of discrimination, including the use of Chinese in the workplace. So there are a lot of practical challenges. There’s a gap between what happens on the ground and investment commitments.
Overall, of course, Taiwan needs to be relevant to the US, but I would say that it's not just for a trade perspective. Politically, Taiwan also is quite important to the US, right? … In the article, I say that the fall of Singapore to Japan during World War II signaled the end of the British Century. If Taiwan were to fall to communist China, US allies in the region would no longer trust US commitments and would either become part of the Sinosphere or develop their own deterrent capabilities, such as nuclear weapons. The US will no longer be recognized as a resident power in the Indo-Pacific. So naturally, it will be the end of the American Century. This outcome will not be acceptable to any US leadership.
Wilhelm: Can a country – even one as economically successful as Taiwan – survive over time if it does not have formal diplomatic recognition?
Hsieh: Taiwan’s case can demonstrate that recognition is critical. But here, recognition is not necessarily defined as diplomatic recognition. This is the key point in my research. What I try to prove is that there are actually different forms of legal recognition under various agreements, and all of these forms are important to Taiwan for pragmatic purposes.
***
Pasha Hsieh is the Lee Kong Chian Professor of Law at the Singapore Management University Yong Pung How School of Law.
Suggested citation:
Pasha Hsieh, “What Do Trade Agreements Say About Taiwan?” USALI Talking Points, May 9, 2026, https://usali.org/publications/talking-points-what-do-trade-agreements-say-about-taiwan.