Comparative global outlook findings

Japan

Fragmented interdependence, not decoupling. MRI’s earlier multipolar forecast is explicitly stress-tested against NRI’s later reassessment of China: China grew 5.2% in 2023, but the IMF forecast cited by Kiuchi falls to 3.3% by 2029 and the US-China nominal-GDP share gap is projected to widen to 7.3 percentage points. METI/JIIA therefore emphasise strategic autonomy and indispensability inside open networks rather than autarky (Kiuchi, 2024a; METI, 2025a, 2026; Sasae, 2026).

India

Capacity-constrained interdependence. ORF and CSEP treat trade reconfiguration, China+1 strategies and Global South agency as a rebalance of globalisation, not its end. Cross-border finance, data, technology and trade remain essential, but access is increasingly filtered by security, standards, infrastructure and trusted partnerships (Mukherjee & Jaishankar, 2024; Stirling & Karacsony, 2026b).

Singapore

Managed interdependence is the base case. ISEAS finds sector-specific GVC reconfiguration rather than broad regionalisation; WTO evidence used in the report shows inter-bloc goods trade growing about 4% more slowly than intra-bloc trade after the Ukraine war, without a general collapse in cross-bloc trade (Doarest & Wihardja, 2024; WTO, 2024).

UK/Western

Fragmentation without autarky. McKinsey/WEF evidence pushes economic security and strategic technology into the centre of competition, while WTO evidence shows adaptation rather than trade collapse. The Western lens therefore assumes a still-connected economy with higher friction, duplicated capacity and more policy risk (McKinsey Global Institute, 2022; World Economic Forum, 2026).

Japan

Economic-security literature treats safety, reliability, resilience and robustness as preconditions of economic activity. MRI’s circularity work also shows that resilience can create value by reducing import exposure and creating recoverable domestic stocks; an illustrative model estimates about JPY1tn annual domestic value from wider circulation of steel, plastics, textiles and cement (MRI, 2024; Takayama, 2026).

India

CDRI makes the mechanism explicit: average service-disruption losses are estimated at about 7.4 times direct infrastructure damage, so continuity and recovery capacity can have economic value far beyond repair costs. India-based work therefore treats redundancy, diversified supply and lifecycle resilience as productive capacity where service failure is systemically costly (CDRI, 2025; Roy & Swaisgood, 2023).

Singapore

CSF’s “more shocks, more connections, more speed” framing makes resilience an operating model rather than a contingency reserve. RSIS cites evidence that 82% of surveyed firms were affected by tariffs and 43% planned supply-chain changes, indicating that dual sourcing, flexibility and option preservation are becoming operational practices (CSF, n.d.; Future Issues and Technology Cluster, 2025).

UK/Western

Western models increasingly monetise resilience through avoided losses, insurance and capital allocation. PwC projects US$151.1tn of cumulative infrastructure spending through 2050, while Deloitte models large annual infrastructure losses from natural hazards, reinforcing the economic case for continuity and adaptation rather than pure lowest-cost optimisation (Deloitte, 2025; PwC, 2026).

Japan

The Japanese lens treats scarcity as a chain: reliable power, grid access, water/cooling, critical materials, skills, trusted data, planning and consent. Critical-mineral refining concentration illustrates the point: the leading refining country averaged about 72% across key minerals excluding rare earths in 2025 (IEA, 2026; MRI, 2022). Circularity and recovery rights are therefore forms of supply creation.

India

India-based institutions define scarcity as “connected capacity and trusted capability”: grid, ports, water, cooling, compute, critical minerals, skills, insurance, consent and delivery institutions. CEEW finds four countries supplied more than 70% of solar exports and more than 80% of wind-generator exports over the period examined, demonstrating concentration beyond raw materials (CEEW, 2023a, 2023b).

Singapore

Singapore’s lens is especially attentive to hidden bottlenecks. RSIS uses tritium as an extreme example: a global stock of roughly 20 kg against an estimated requirement of around 55 kg per year for a 1 GW fusion reactor. The lesson is methodological: a small input can constrain a very large system (Future Issues and Technology Cluster, 2025).

UK/Western

The Western lens links scarcity to investable throughput: critical-mineral refining, transformers, grid queues, cooling, water, labour and planning. UNEP projects global resource extraction could rise roughly 60% from 2020 levels by 2060 without a change in trajectory, while IEA sees a potential copper supply gap of roughly 25% against 2035 primary needs (IEA, 2026; UNEP-IRP, 2024).

Japan

Japanese foresight links digitalisation to distributed energy systems and current policy now coordinates electricity, telecoms and data-centre location through “watt-bit” planning. Global data-centre electricity demand is projected at about 945 TWh by 2030, close to 3% of global electricity use, making digital growth physically locational (IEA, 2025; METI, 2025c; MRI, 2020).

India

India-based work treats clean energy as an economic and strategic input only when generation is joined to grid, ports, storage, cooling, data centres, industrial demand, housing and skills. The core issue is conversion of energy potential into productive clusters, not generation capacity alone (Sarma & Chandola, 2025; Stirling & Karacsony, 2026c).

Singapore

Singapore-based institutions stress regional connectivity, network coordination and the physical dependencies of digital economies. Energy security is not self-sufficiency: it is diversification plus reliable networks, standards and cross-border coordination (CSF, 2022; ISEAS-Yusof Ishak Institute, 2025).

UK/Western

The Western lens is most explicit about scale. IEA central scenarios show electricity demand about 40% higher by 2035; roughly US$1tn/yr is going into electricity supply versus about US$400bn into grids, and IEA has highlighted ~80m km of grids needing addition/replacement by 2040 (IEA, 2023, 2025b).

Japan

Japanese evidence integrates adaptation with disaster resilience, consumption systems and long-duration viability. IGES’s 1.5-degree pathway gives modelled lifestyle-footprint benchmarks of 2.5 tCO2e/person in 2030, 1.4 in 2040 and 0.7 in 2050; the important mechanism is that housing, mobility and food systems jointly determine whether places remain viable (IGES et al., 2019; Ministry of the Environment, Japan, 2021).

India

India-based work is strongest on service continuity and multi-hazard infrastructure. CDRI and T20 analysis treat adaptation across the asset lifecycle and emphasise that the social and economic cost of service interruption can dwarf direct physical damage (CDRI, 2025; Roy & Swaisgood, 2023).

Singapore

Singapore institutions increasingly connect physical risk and natural capital to finance. SGFIN and CSF frame adaptation as a question of asset usability, funding, natural-system capacity and legitimacy. UNEP validation used in the report estimates developing-country adaptation needs at US$310-365bn per year by 2035 versus US$26bn of international public adaptation finance in 2023 (SGFIN, 2026; UNEP, 2025).

UK/Western

Western evidence converts physical risk into loss, insurance and capital metrics. Deloitte models average annual infrastructure hazard losses around US$460bn by 2050; Swiss Re reports 2025 natural-catastrophe economic losses of about US$220bn, with roughly 49% insured (Deloitte, 2025; Swiss Re Institute, 2026).

Japan

IOG argues that frontier AI becomes power only when capital, talent, compute, cloud, chips, universities, firms, data and governance operate as an ecosystem. Japanese policy adds robotics, human augmentation and iterative “living document” governance, while power and water constraints make AI a physical-system issue (METI & MIC, 2026; Shiono et al., 2025).

India

India-based analysis decomposes AI into sovereign capability, infrastructure, compute, chips, data centres, economics, labour, data and connectivity. The implication is uneven diffusion: productivity depends on whether firms and workers can access the stack, while concentration in compute, chips and data centres creates strategic dependency (Sarma & Chandola, 2025; Ray, 2021).

Singapore

Singapore emphasises anticipatory governance and trust. RSIS treats AI advantage as an ecosystem of standards, talent, infrastructure and public legitimacy, while CSF frames governance as adaptive because technology cycles can outrun fixed rules (CSF, 2024; Future Issues and Technology Cluster, 2025).

UK/Western

Western evidence quantifies both labour and infrastructure. ILO estimates one in four workers is in an occupation with some GenAI exposure, but only 3.3% of global employment is in the highest-exposure category; IEA projects data-centre electricity near 945 TWh by 2030 (IEA, 2025; ILO, 2025).

Japan

Japan-based analysis links labour supply directly to national economic power: Kiuchi cites a five-year Chinese employee-growth outlook of about -0.56% per year and estimates US immigration added about 0.33 percentage points to 2023 real-GDP growth. NISTEP/IGES add participation, trust, loneliness and meaning as transition constraints (Kiuchi, 2024a; Mao et al., 2019; NISTEP, 2026).

India

India-based institutions focus on the distribution of skills and opportunity: technology and urbanisation generate broad prosperity only if labour-market institutions, education, migration, health and social protection diffuse capability beyond leading firms and cities. WEF validation indicates 59 of every 100 workers may require training by 2030 (Kumar et al., 2026; WEF, 2025).

Singapore

Singapore’s lens highlights ageing and social atomisation. WHO validation reports loneliness affecting about 15.8% of people globally, with around 871,000 associated deaths annually; CSF treats social connection and trust as conditions for adoption, labour participation and institutional resilience (CSF, 2024; WHO, 2025).

UK/Western

Western institutions emphasise ageing, participation and human capital. BCG finds all 38 OECD countries ageing by 2050 and 33 with at least 21% of the population aged 65+. UN demographic projections reinforce the shift in dependency structures (BCG, 2025; UN DESA, 2024).

Japan

NISTEP’s 2050 exercise combines 46 experts/practitioners, 14 workshops and 12 scenarios, showing that uncertainty management itself requires institutional capacity. Japanese sources repeatedly pair capability with limits: strategic autonomy with rules, technology with ELSI, and coordination with participation (NISTEP, 2026; Hosoya & Kundnani, 2024).

India

India-based sources stress implementation conversion: capital and policy ambition do not create outcomes without institutions able to sequence infrastructure, contract, coordinate, learn and retain legitimacy. The key bottleneck is often delivery architecture rather than finance alone (Mukherjee & Jaishankar, 2024; CDRI, 2025).

Singapore

Singapore provides the strongest explicit foresight-operating-model lens. CSF’s Driving Forces 2040 uses 17 driving-force cards and five wildcards; its handbook stresses signposts, stress tests, options and organisational embedding. Foresight is valuable only if it changes decisions (CSF, 2022, n.d.).

UK/Western

Western sources emphasise operating-model redesign and anticipatory governance. Deloitte synthesises more than 200 government transformation cases; OECD/WEF surveyed 167 foresight practitioners in 55 countries, showing that institutional learning and governance are themselves being reshaped by AI (Deloitte, 2026; OECD & WEF, 2025).