Burning bridges or bridging divides: Geopolitics and collaboration in research and innovation
This paper analyzes the association between geopolitical alignment and international research and innovation collaborations using data spanning 2000–2023. Using ideal point distance measures derived from UN General Assembly voting patterns, we find that increases in geopolitical distance are associated with reductions in international collaborations. A one-unit increase in geopolitical distance (roughly the distance between a median country pair) over three years is associated with a 3.1% reduction in collaborations in scientific publications (0.8% per standard deviation). We also find that this relationship is asymmetric: worsening geopolitical relations are associated with reductions in collaborations in scientific publications that are roughly three times larger than the gains associated with equivalent improvements, suggesting that international research partnerships are easier to damage than to build. At the technology level, critical technologies exhibit approximately three times the expected level of international collaboration compared to other fields (conditional on country-pair effects), yet are also more sensitive to shifts in geopolitical alignment, raising concerns about technology sovereignty. At the same time, despite ongoing geopolitical tensions, overall collaboration patterns have not undergone substantial changes in recent years, suggesting resilience in international research and innovation networks.
AI and Trade: Why Europe Cannot Afford to Lag on Adoption
Artificial intelligence (AI), and generative AI in particular, is poised to transform productivity across a broad range of activities, with the strongest effects concentrated in knowledge-intensive services such as finance, professional services, and ICT. Its economic impact will nevertheless depend on how quickly countries adopt and integrate it into their economies. Evidence points to substantial cross-country differences in adoption, particularly within Europe. Yet AI is not only a domestic transformation; it is also a productivity shock transmitted through international trade. Productivity gains abroad lower import prices and reshape competitiveness across countries and sectors. Our analysis shows that these forces interact: countries can benefit from foreign AI progress through cheaper imports, but without sufficient domestic adoption, they risk losing competitiveness in AI-exposed sectors. The global diffusion of AI therefore makes domestic adoption capacity and openness to trade complementary determinants of future growth.
AI Meets Trade: Global Linkages and the Cross-country Distribution of the Gains from AI
This paper provides estimates of expected per capita real income gains from AI over the next decade in OECD and G20 economies. It relies on a multi-country, multi-sector general equilibrium model to incorporate the role of international trade and considers different scenarios regarding AI adoption paths and AI capabilities. In our central scenario, AI-driven productivity gains vary widely across countries and are expected to raise per capita real income growth by 0.1–0.95 percentage points annually. The model’s dynamics reveal a key insight: while countries lagging in AI adoption can gain from cheaper, AI-intensive imports generated by global diffusion, maintaining competitiveness – especially in highly AI-exposed sectors – ultimately requires strong domestic AI adoption. As a further channel, the paper quantifies the welfare effects generated by international knowledge spillovers boosting AI adoption of countries with relatively lower AI adoption.
Mapping Economic Opportunities in Global Clean Energy Supply Chains
The energy transition offers countries that can manufacture clean energy technologies substantial opportunities for sustainable economic growth. This paper provides a framework for context-aware industrial policy by applying economic complexity theory to a newly constructed dataset of twelve key clean energy supply chains (CESCs). We find that CESCs are diverse but highly interdependent; they are also growing faster and are more concentrated than other industries. CESCs exhibit substantial entry, exit and competitive churn, and countries are more likely to enter CESC industries that are related to their existing productive capabilities. We also explore changing global competitiveness and country positioning in these industries, and draw out implications of these patterns for industrial policymakers.
Innovation Capabilities Outlook 2026
Knowledge is expanding globally, yet most countries struggle to harness this growth effectively. Global innovation remains strikingly concentrated: a small number of leading economies account for the vast majority of scientific publications, patents, trademarks, and advanced exports, whereas most contribute less than 1 percent to any innovation dimension. Success does not require a big push in all fields, but instead lies in strategically diversifying into complex skills while at the same time maintaining intensity in high-value areas – a balancing act that only the most sophisticated innovation ecosystems have mastered.
Mapping the global innovation landscape
The Innovation Capabilities Outlook (ICO) 2026 analyzes 2,508 innovation capabilities across four dimensions – science, technology, entrepreneurship, and production – using comprehensive datasets spanning 2001–2023. The analysis reveals that innovation emergence depends critically on connections between these four dimensions, with the most sophisticated capabilities emerging only in highly diversified ecosystems able to support complex, interdependent knowledge networks.
A tale of two innovation worlds
Global innovation output has expanded dramatically, yet this growth remains highly uneven and concentrated in no more than 30 percent of the world’s economies. Asian economies – led by China, India and Viet Nam – have mastered sophisticated capability development strategies, consistently achieving both smart diversification (gaining breadth and complexity simultaneously) and smart capability management (intensifying focus on high-value skills while protecting them with complementary knowledge). In contrast, many established and emerging economies struggle with this dual challenge: 46 percent of ecosystems have not meaningfully diversified, and complexity gains remain elusive for 70 percent of economies.
Strategic opportunities
The ICO 2026 identifies substantial untapped potential – only 10 percent of economies fulfill their technological potential. Ecosystems collectively underperform by 339,000 technological innovations annually. Regional patterns reveal distinct strategic pathways: Europe possesses strong foundations, but struggles with technological translation; Asia shows balanced capabilities, but faces entrepreneurial commercialization challenges; and Africa should focus on foundational capability building while gradually targeting more complex activities.
Policy implications
Innovation policy cannot rely on one-size-fits-all approaches. Success requires tailoring strategies to regional development levels, existing capability portfolios, and institutional contexts. Countries that align innovation investments with these evidence-based insights can break traditional development constraints and accelerate a transition toward knowledge-based competitiveness. The systematic nature both of diversification constraints and untapped potential suggests that targeted, level-appropriate interventions yield the highest probability of success.
The Innovation Capabilities Outlook 2026 was developed through a partnership between WIPO and Harvard University’s Growth Lab (HGL), under the general direction of Daren Tang (Director General) and Marco Alemán (Assistant Director General). The report was supervised by Carsten Fink (Chief Economist) and Ricardo Hausmann (founder and Director of HGL), prepared by a team led by Julio Raffo (Head of Innovation Economy Section, WIPO) and Muhammed A. Yildirim (Director of Academic Research, HGL). The team included Christian Chacua, Matte Hartog, Shreyas Gadgin Matha, and Federico Moscatelli.
Tackling Discrepancies in Trade Data: The Harvard Growth Lab International Trade Datasets
Bilateral trade data informs foreign and domestic policy decisions, serves as a growth indicator, determines tariffs, and is the basis for financial and investment decisions for corporations. Accurate trade data translates into better decision-making. However, the raw bilateral trade data reported by UN Comtrade suffer from two structural problems: reporting differences between country partners and countries reporting in different product classification systems, which require product-level harmonization to compare data across countries. In this paper, we address these challenges by combining a mirroring technique and a data-driven concordance method. Mirroring reconciles importer and exporter differences by imputing country reliability scores and applying a weighted country-pair average to calculate the estimated trade value. We harmonize product classifications across vintages by calculating conversion weights that reflect a product’s market share. The resulting publicly available datasets mitigate issues in raw trade statistics, reducing reporting inconsistencies while maintaining product-level granularity across six decades.
Global Networks, Monetary Policy and Trade
We develop a novel framework to study the interaction between monetary policy and trade. Our New Keynesian open economy model incorporates international production networks, sectoral heterogeneity in price rigidities, and trade distortions. We decompose the general equilibrium response to trade shocks into distinct channels that account for demand shifts, policy effects, exchange rate adjustments, expectations, price stickiness, and input–output linkages. Tariffs act simultaneously as demand and supply shocks, leading to endogenous fragmentation through changes in trade and production network linkages. We show that the net impact of tariffs on domestic inflation, output, employment, and the dollar depends on the endogenous monetary policy response in both the tariff-imposing and tariff-exposed countries, within a global general equilibrium framework. Our quantitative exercise replicates the observed effects of the 2018 tariffs on the U.S. economy and predicts a 1.6 pp decline in U.S. output, a 0.8 pp rise in inflation, and a 4.8% appreciation of the dollar in response to a retaliatory trade war linked to tariffs announced on “Liberation Day.” Tariff threats, even in the absence of actual implementation, are self-defeating— leading to a 4.1% appreciation of the dollar, 0.6% deflation, and a 0.7 pp decline in output, as agents re-optimize in anticipation of future distortions. Dollar appreciates less or even can depreciate under retaliation, tariff threats, and increased global uncertainty.
Tackling Discrepancies in Trade Data: The Harvard Growth Lab International Trade Datasets
Bilateral trade data informs foreign and domestic policy decisions, serves as a growth indicator, determines tariffs, and is the basis for financial and investment decisions for corporations. Accurate trade data translates into better decision-making. However, the raw bilateral trade data reported by UN Comtrade suffer from two structural problems: reporting differences between country partners and countries reporting in different product classification systems, which require product-level harmonization to compare data across countries. In this paper, we address these challenges by combining a mirroring technique and a data-driven concordance method. Mirroring reconciles importer and exporter differences by imputing country reliability scores and applying a weighted country-pair average to calculate the estimated trade value. We harmonize product classifications across vintages by calculating conversion weights that reflect a product’s market share. The resulting publicly available datasets mitigate issues in raw trade statistics, reducing reporting inconsistencies while maintaining product-level granularity across six decades.
De Facto Openness to Immigration
Various factors influence why some countries are more open to immigration than others. Policy is only one of them. We design country-specifc measures of openness to immigration that aim to capture de facto levels of openness to immigration, complementing existing de jure measures of immigration, based on enacted immigration laws and policy measures. We estimate these for 148 countries and three years (2000, 2010, and 2020). For a subset of countries, we also distinguish between openness towards tertiary-educated migrants and less than tertiary-educated migrants. Using the measures, we show that most places in the World today are closed to immigration, and a few regions are very open. The World became more open in the first decade of the millennium, an opening mainly driven by the Western World and the Gulf countries. Moreover, we show that other factors equal, countries that increased their openness to immigration, reduced their old-age dependency ratios, and experienced slower real wage growth, arguably a sign of relaxing labor and skill shortages.
Explore the country rankings in our interactive visualization website and learn more about the project, Leveraging the Global Talent Pool to Jumpstart Prosperity in Emerging Economies.
On the Design of Effective Sanctions: The Case of Bans on Exports to Russia
We build on Baqaee and Farhi (2019, 2021) and derive a theoretically-grounded criterion that allows targeting bans on exports to a sanctioned country at the level of ∼5000 6-digit HS products. The criterion implies that the costs to the sanctioned country are highly convex in the market share of the sanctioning parties. Hence, there are large benefits from coordinating export bans among a broad coalition of countries. Applying our results to Russia reveals that sanctions imposed by the EU and the US in response to Russia’s invasion of Ukraine are not systematically related to our arguments once we condition on Russia’s total imports of a product from participating countries. We discuss drivers of these differences, and then provide a quantitative evaluation of the export bans to show that (i) they are very effective with the welfare loss typically ∼100 times larger for Russia than for the sanctioners; (ii) improved coordination of the sanctions and targeting sanctions based on our criterion allows to increase the costs to Russia by about 80% with little to no extra cost to the sanctioners; and (iii) there is scope for increasing the cost to Russia further by expanding the set of sanctioned products.