Yan Xuetong: assessing a decade of predictions to inform future forecasts

Tsinghua's IR scholar acknowledged overestimation of China's growth, oversight of Asia-Pacific dynamics, Russia-North Korea ties, and the Israel-Hamas war despite an 82.3% accuracy rate.

In 2013, Professor Yan Xuetong, Director of the Institute of International Relations at Tsinghua University, published the book "Inertia of History: China and the World by 2023 历史的惯性：未来十年的中国与世界". Critics at the time described it as a work predicting the trends in the international landscape over the next decade: Could China become a superpower within the next ten years? Could East Asia replace Europe as the center of the world? Would the European Union, Russia, and Japan lose their luster? Would the BRICS countries become history? Should China abandon its non-alignment policy? What kind of diplomacy should China adopt? Among China's neighboring countries, who are friends and who are foes? With what kind of ideology should China lead the world?

A decade later, on January 13, 2024, a seminar titled "Predicting International Relations - Revisiting 'Inertia of History'" was convened at Tsinghua University to evaluate the precision of Yan's forecasts. Esteemed scholars contributed their assessments, and at the end of the seminar, Professor Yan offered a comprehensive review of his initial predictions. Despite achieving an 82.3% accuracy rate in his predictions, Yan highlighted some of his key missteps:

• Overestimation of China's Growth: Optimistic projections, fueled by national pride, led to an underestimation of how quickly the "new normal" would proliferate across different domains, resulting in an overly optimistic view of China's growth rate in various sectors.

• Strategic Realignments in Asia-Pacific: The "Inertia of History" overlooked the shift among several of China's neighbors from a dual-reliance strategy—economic growth with China and security with the US—to strengthening ties more exclusively with the US.

• Geopolitical Developments: The "Inertia of History" notably underestimated certain geopolitical shifts in the latter half of the forecast period, particularly in 2022 and 2023. Notable errors include the emergence of Russia as North Korea's most important strategic partner and the Israel-Hamas war which challenged the principle of human rights over sovereignty.

Addressing these inaccuracies, Yan proposed five guidelines to refine the art of forecasting in international relations. A key cautionary note within these guidelines is the uncertain nature of the power disparity between China and the US. While China and the US are projected to further widen their comprehensive national power gap with other major powers, the potential for an increasing power disparity between China and the US cannot be dismissed, wrote Yan.

This article was written by Professor Yan Xuetong and originally published by Phoenix New Media.

---

Verification of Predictions in "Inertia of History" and Suggestions Thereof

People lacking a basic understanding of predictive research might mistakenly think that the field of international relations can only conduct hindsight analysis and is incapable of making useful predictions. This misunderstanding arises from a lack of awareness that a crucial part of predictive research involves verifying predictions. Unverified conjectures in international relations are devoid of significance, just as it would be meaningless to conduct a scientific experiment without analyzing its outcomes or to issue a weather forecast without subsequently checking the actual weather.

Predictive research in international relations is underpinned by three foundational pillars: theoretical guidance, scientific methodologies, and empirical knowledge. These three elements are essential for scientific prediction, and the absence of any one of them disqualifies it as a scientific forecast. Acknowledging the mistakes in the predictions made in "Inertia of History" can reinforce the scientific ethos of pursuing truth from facts. Evaluating the causes of these inaccuracies can help avoid repeating the same mistakes, while the analysis of correct predictions can refine and enhance the future capacity for predictive research.

According to preliminary statistics, "Inertia of History" made a total of 298 predictions, of which two could not be judged right or wrong due to the lack of data for 2023. The prediction that China and the US would rank the top two countries that have the highest number of internet uses in 2023 was only half correct, as India surpassed the United States to become the country with the second-largest internet user base in 2022. Therefore, out of the 296 predictions that were verified, 243.5 were correct, accounting for 82.3%; 52.5 were incorrect, accounting for 17.7%.

Achieving an accuracy rate above the 65% benchmark confirms that the book's methodologies are scientifically robust, with their substantial reference value further affirmed by an accuracy exceeding 75%. Moreover, as the accuracy rate approaches 85%, these methodologies are nearing the practical value standard. In comparison, current weather forecast accuracy surpasses 95% for a 24-hour period, approximates 90% for 48 hours, and hovers around 70% for 96 hours. Weather prediction science should serve as a model for international relations predictive research. While perfect accuracy is unattainable, the ongoing improvement of prediction accuracy and precision is essential.

The analysis of prediction errors yields several recommendations for forecasting international relations over the coming decade:

I. Predictions should account for the growing influence of populism on international relations, considering the historical trend towards deglobalization and its implications. Many prediction inaccuracies in the book originated from an underestimation of populism's ascent and its effect on major nations' shift towards deglobalization strategies. Presently, policymakers in the United States and Europe are moving away from the liberal international order, with the United States emerging as a leading advocate for deglobalization.

Deglobalization is unfolding across both economic and political domains. The surge in populism, which began less than a decade ago, is poised to continue its upward trajectory over the next ten years. The strategic dynamic between China and the United States is predominantly characterized by competition, with collaborative efforts focused on averting the escalation of conflicts into warfare. The anticipation of heightened mutual caution among major powers for so-called "economic security" and "technological superiority" suggests an increasing trend of deglobalization-related conflicts and confrontations. There will likely be a reinforcement of trade protectionism and decoupling policies, along with more frequent violations of human rights norms. Populist leaders encounter increased regime instability, and their foreign policies are often designed to consolidate domestic power. This results in foreign policy approaches that can swing widely and frequently in response to domestic policy shifts.

II. Predictions should incorporate both systemic and domestic-level variables. Trends should be forecasted using systemic factors, while processes and events should be analyzed based on the influence of decision-makers' power. Systemic factors, characterized by strong inertia, play a critical role in shaping the development trend of the international landscape. "Inertia of History" achieved greater accuracy in predicting the comparative strength of major powers and the direction of international landscape changes, primarily due to the slow pace of changes in power dynamics. Given the recent formation of the bipolar competition pattern between China and the US, this dynamic is expected to remain stable over the next decade, serving as a constant in predictions.

Prediction errors in "Inertia of History" often arose from underestimating the impact of decision-makers' power. The power wielded by these individuals can significantly alter the course of international relations through the creation of pivotal events, potentially accelerating or reversing historical trends. Instances like Russia's annexation of Crimea in 2014, the US's trade war with China in 2017, Germany's stringent policy towards China in 2021, the Philippines escalating the Ren'ai Reef dispute in 2023, and Israel's mass killing of civilians in Gaza in 2023, underscore the influence of decision-makers' efforts to maintain their ruling power. Predictions for the next ten years should account for the increased likelihood of using force abroad to sustain personal power, meaning that armed conflicts may become more frequent than in the previous decade.

III. Minimize the impact of subjective biases on predictions. In the current geopolitical framework, where nation-states are the dominant entities, forecasters often fall into the trap of overestimating their own country's development prospects due to national biases. This tendency towards optimistic projections, fueled by national pride, led to an underestimation of how quickly the "new normal" would proliferate across different domains, resulting in an overly optimistic view of China's growth rate in various sectors in "Inertia of History". Some prediction errors related to overestimating strength growth involved overestimating the impact of the widened power gap on bilateral strategic relations. For instance, contrary to expectations, neighboring countries like Japan, South Korea, and the Philippines did not seek to proactively enhance bilateral ties in response to China's growing strength; instead, they pursued less cooperative policies.

In forecasting the next decade, predictions should be anchored in realism theory, steering away from liberalism and constructivism, prioritizing the resurgence of power politics, and paying more attention to the inclinations of confrontational tactics. The coming decade is expected to witness an uptick in international disputes and a tightening of constraints on global cooperation, both of which will complicate efforts to spur global economic growth. Major powers are highly likely to grapple with low growth, no growth, or negative growth. While China and the US are projected to further widen their comprehensive national power gap with other major powers, the potential for an increasing power disparity between China and the US cannot be dismissed.

IV. Emphasize international relations predictions that deviate from mainstream opinions. The accuracy of a prediction is linked to the likelihood of the trend or event it forecasts. Since mainstream societal beliefs often mirror the most probable trends or events, predictions conforming to these expectations, even if accurate, offer minimal social and academic contribution. The merit of publishing such forecasts is subject to debate. Conversely, predictions that diverge from mainstream views, namely those concerning unlikely trends and events, hold greater social and academic value.

The prediction in "Inertia of History" that China would widen the gap in strength with other major powers from 2013 to 2023 exemplifies a prediction aligned with probable trends. However, it overlooked the less likely scenario of countries increasing their strategic distance from China amid US-China competition. Several of China's neighbors moved away from their dual-reliance strategy on China for economic growth and the US for security, opting instead to enhance their economic and security ties with the US, thereby tilting the strategic balance in the Asia-Pacific region in favor of the US. For the upcoming decade, predictions concerning such unlikely developments are of greater societal importance. These include: Amidst rising global military conflicts, which regions can foster security cooperation? Amidst the deglobalization trend, what areas of global governance can make progress? In an era of sluggish global economic performance, which nations can achieve growth against the odds? And in the context of widespread populism, which countries will push forward with democratization?

V. Segment the forecast period into an initial five years and a subsequent five years. The inertia of events weakens over time, which means predictions for the near term tend to be more accurate than those for the longer term. The inaccurate international relations predictions in the "Inertia of History" are primarily noted in the latter half of 2013-2023, particularly in 2022 and 2023. Notable errors include failure to predict China losing its status as America's largest trading partner, the US becoming India's top trading partner, India rising to become the world's second-largest internet user, escalated maritime disputes between China and the Philippines, Russia's emergence as North Korea's most important strategic partner, and the Israel-Hamas war challenging the principle of human rights over sovereignty.

Given the principle that inertia's impact fades with time, forecasting the next decade of international relations should be divided into two segments: the initial and latter five years. Early predictions should leverage inertia, while those for the latter period should increasingly consider potential pivotal changes. This bifurcated approach aligns more accurately with the dynamic nature of international relations, where the potential for significant shifts grows with the extension of the timeline.

In China, the community of researchers focusing on predictive studies of international relations is relatively small, with methodologies and research levels yet to reach a more advanced stage. To foster the evolution of this field, it is crucial to cultivate a scientific ethos that embraces predictive inaccuracies, to adopt sophisticated forecasting techniques, and to continually refine predictive theories. I hope this critique of predictions in "Inertia of History" will serve to promote the development of predictive research in international relations.