The Deglobalization Signal
Supply chain disruptions, geopolitical fragmentation, and industrial policy have reversed decades of globalization. Understanding what is actually changing — and what isn't — requires looking past the political rhetoric.
The globalization story of the late twentieth century had a clean narrative: trade liberalization, the fall of the Soviet Union, and China's integration into the WTO had created an interconnected world economy where goods, capital, and services flowed across borders with unprecedented freedom, generating growth and reducing poverty on a scale without historical precedent. The narrative was substantially correct and substantially incomplete.
It was correct about the aggregate outcomes: global poverty rates have declined dramatically since 1990, from approximately 36 percent to under 10 percent of world population; global GDP has grown at rates that would have seemed implausible to forecasters in 1970. It was incomplete about the distributional outcomes: the gains from globalization accrued disproportionately to the wealthy in advanced economies and to the emerging middle classes in China and other manufacturing-intensive developing countries, while the costs fell disproportionately on manufacturing workers in advanced economies who competed with low-wage imports.
The political consequences of that incompleteness are now visible in the trade policy of every major advanced economy: the US CHIPS Act and Inflation Reduction Act, the EU's Critical Raw Materials Act and Green Deal industrial policy, Japan's economic security legislation, all represent a shift toward state-directed industrial policy and supply chain localization that would have been ideologically unthinkable in the Washington Consensus era.
What is actually changing
The deglobalization narrative is partly real and partly rhetorical. Understanding which is which requires distinguishing between different types of global economic integration.
Trade in manufactured goods — the most politically visible form of globalization — has genuinely flattened after decades of rapid growth. The ratio of goods trade to global GDP peaked around 2008 and has been roughly flat since, with significant declines during the 2008 financial crisis, the 2020 pandemic, and the 2022 geopolitical disruptions. The secular growth trend that characterized the 1990s and 2000s has ended.
The reasons are multiple. China's manufacturing wages have risen substantially — the labor cost arbitrage that drove globalization's first era has partly exhausted itself as Chinese workers have moved into the middle class. Automation has reduced the labor cost component of manufacturing, making near-shoring and on-shoring more competitive relative to offshore production. Geopolitical risk — the possibility that conflict, sanctions, or regulatory changes will disrupt supply chains — has been repriced upward after the Russia sanctions and the COVID supply chain disruptions.
Trade in services — less visible but growing — has continued to expand and is less affected by the political backlash against goods trade. Digital services, financial services, and professional services are increasingly traded globally in ways that are difficult to measure but significant in value.
Capital flows remain deeply globalized, though their composition has shifted: foreign direct investment has declined as a share of global GDP, while portfolio investment and cross-border lending have remained substantial.
The CHIPS Act and industrial policy
The most consequential US industrial policy initiative in decades is the CHIPS and Science Act of 2022, which appropriated $52 billion for domestic semiconductor manufacturing incentives and $200 billion for science research. The strategic rationale was explicit: semiconductor manufacturing, concentrated in Taiwan and South Korea for critical leading-edge fabrication, represents an unacceptable supply chain vulnerability in a world where China might attempt to seize Taiwan and where semiconductor shortages have disrupted automobile and consumer electronics manufacturing.
The program has had early successes: TSMC announced a $65 billion investment in Arizona fabs; Samsung announced major Texas expansion; Intel announced $20 billion in Ohio construction. These announcements represent genuine commitments to domestic manufacturing capacity that would not have occurred without the incentive structure.
The program's success in actually manufacturing competitive chips in the United States at scale remains to be demonstrated. Semiconductor fabrication requires not just capital investment but a supply chain ecosystem — specialized chemical suppliers, equipment manufacturers, trained engineers, water infrastructure — that takes years to decades to develop. Taiwan's semiconductor cluster did not emerge from a single incentive program; it emerged from forty years of coordinated industrial policy and ecosystem development.
Metaculus forecasts a 58 percent probability that US domestic semiconductor manufacturing capacity for leading-edge chips (3nm and below) will be at least 15 percent of global capacity by 2030, up from less than 2 percent today. The investment is real; whether it translates into sustained competitive manufacturing capacity is the open question.
The China decoupling paradox
The most politically discussed form of deglobalization — the decoupling of US and Chinese economic relationships — is simultaneously more advanced and less complete than the rhetoric suggests.
Bilateral US-China goods trade declined after the Trump tariffs of 2018-19 and has not recovered to pre-tariff levels. But the decline in direct US-China trade has been partially offset by the rise of third-country intermediation: Chinese manufacturers have increased production in Vietnam, Mexico, Thailand, and other countries from which goods are exported to the United States without triggering China-specific tariffs. This is not genuine decoupling; it is trade routing.
The more significant decoupling is in capital flows: US venture and private equity investment in Chinese technology companies has declined sharply since 2021, driven by both regulatory restrictions (US outbound investment rules) and commercial risk aversion. Chinese access to US semiconductor technology, cloud services, and advanced software has been progressively restricted.
This partial decoupling is simultaneously too slow (from a US national security perspective that seeks to limit China's access to advanced technology) and too fast (from a global economic perspective that values the efficiency gains from integrated supply chains). Kalshi was trading a contract on whether US-China bilateral goods trade will fall below $500 billion annually — roughly half of current levels — before 2030 at 21 percent. The gradual nature of decoupling suggests the lower probability is appropriate, but the direction is not in doubt.
Marcus Webb is a contributing writer at The Auguro covering fiscal policy, markets, and the political economy of American government.