Greening in the Wrong Places: Geography, Policy Distortions, and the Hidden Costs of Misallocated Green Investment

Policy Center for the New South, PP-13/26

Jorge Arbache & Otaviano Canuto

1. Decarbonization as a Problem of Space, Prices, and Coordination

Global decarbonization is typically framed as a technological race, a financing challenge, and/or a problem of carbon pricing. There is a focus on the green premia still paid to replace carbon-emitting technologies with clean alternatives, although such premia have declined because of technological learning and falling clean-energy generation costs. Another focus is that existing stocks of investments in renewable energy have been shown to be insufficient to serve as a full alternative. The implications of a shift from fossil fuels to critical minerals upstream of production chains are also highlighted. As a result, the spotlight is generally on the possible toll to be paid ‘on the road to decarbonization’ (‘greenflation’) (Canuto, 2021a; 2021b).

These perspectives are important but incomplete. They understate a more fundamental transformation: decarbonization is also reorganizing the geography of global production and changing relative prices (Arbache and Esteves, 2023; Arbache, 2025a; 2026a). As energy systems change, the location of production is becoming increasingly important for both economic efficiency and emissions reduction.

This transformation is particularly visible in hard-to-abate and energy-intensive sectors including hydrogen, aluminum, steel, fertilizers, and sustainable aviation fuels, for which production costs depend heavily on geographically distributed factors—renewable resources, electricity costs, water, mineral endowments, and logistics infrastructure. As these factors grow in importance, geography is re-emerging as a central determinant of comparative advantage (Arbache, 2023; 2025a; 2026b).

This paper builds on a broader strand of recent work and develops five core arguments (Arbache, 2025d):

1. Structural changes in relative prices are reshaping comparative advantage under decarbonization;
2. The physical characteristics of renewable energy systems are reinforcing the economic importance of geography;
3. Industrial policy interventions distort effective prices and redirect investment toward structurally inefficient locations;
4. Demand uncertainty and project bankability are additional major constraints on efficient capital allocation; and
5. These distortions generate broader consequences for technological learning, global value chains, development, and the future geography of industrialization.

The paper also argues that decarbonization is producing a ‘rematerialization’ of the global economy, contrary to earlier predictions that geography would become less important in a digital world (Arbache, 2026b; Yang & Canuto, 2024). The central argument: decarbonization is not simply a technological transition; it is simultaneously a spatial, industrial, trade, and development transition (Arbache, 2025a).

The notion of ‘dematerialization’ of the economy came from the increasing share of intangible products in GDP, and the rising weight of intangible assets in economic growth and the composition of wealth of nations (Canuto and Daoulas, 2019). The digital revolution of course has reinforced this notion, bringing with it implications in terms of changes in the role of land (Yang and Canuto, 2024).

The technological changes that have made Globalization 2.0 possible since the 1990s led to a view that geographical constraints were becoming less relevant. Containerization in transport, and information and communication technologies, led to the breakdown and geographical scattering of value chains (Canuto, 2021c).

Thomas Friedman’s bestseller The World Is Flat (Friedman, 2007) highlighted the strong forces pushing the world toward a single economic platform. Technology-fueled globalization in the provision of services, and the widespread organization of production processes as global value chains (GVCs) were part of his narrative. As Friedman acknowledged, the world was not entirely there yet. Barriers remained in terms of logistics performance, trade facilitation bottlenecks, and international physical connectivity (Canuto, 2013). But the idea of decreasing geographical constraints. However, the road to decarbonization and the corresponding transformation of the energy system has reinstated the weight of geography.

2. Energy Density and the Structural Return of Geography

The reemergence of geography as a central economic variable is rooted in a physical transformation of the energy system (Arbache, 2026c,d). Fossil fuels—characterized by high energy density and well-developed logistics—allowed energy production and industrial activity to be geographically separated. Renewable energy systems operate under fundamentally different conditions: wind and solar resources are spatially uneven, electricity transmission involves high capital expenditure and physical constraints, and hydrogen as an energy carrier introduces significant costs through compression, liquefaction, and conversion into derivatives.

These physical constraints translate directly into economic outcomes. This logic defines “powershoring” (Arbache, 2022; Arbache and Esteves, 2023). Powershoring is not simply a policy strategy; it is an economically efficient response to the physics of the energy transition. By relocating production to energy-rich regions, it reduces system costs, avoids transmission losses, and accelerates the deployment of low-carbon industrial capacity. However, this adjustment mechanism depends critically on the functioning of relative prices: if prices are distorted, adjustment is delayed or blocked.

3. Rematerialization and the Return of Geography

Influential strands of late twentieth-century thought predicted that digitalization and intangible assets would reduce the importance of geography. Decarbonization is producing the opposite effect. The clean energy transition is extraordinarily material-intensive, requiring vast quantities of copper, aluminum, steel, lithium, nickel, rare earths, transmission infrastructure, land, and water. In other words, while reducing dependence on fossil-fuel extraction, the transition is increasing dependence on geographically specific renewable resources, industrial minerals, and natural capital.

In the context of GVCs, the relevant question is not whether entire industries should relocate, but whether specific energy-intensive upstream stages should move to lower-cost locations, while downstream activities remain near consumer markets. This is precisely the logic of powershoring: rather than reshoring entire industries, advanced economies could anchor downstream value-added activities domestically while allowing energy-intensive upstream stages to relocate toward regions where clean energy is structurally cheaper. This would reconcile industrial policy, resilience, and allocative efficiency, not by eliminating specialization, but by aligning it with the new geography of clean energy costs (Arbache, 2025c; 2026e).

4. From Comparative Advantage to Distorted Competitiveness

In a decarbonizing world, the standard Heckscher-Ohlin framework becomes insufficient, as it is based on relative endowments of facrors of production. Clean energy and natural capital must be incorporated explicitly as production factors. Countries with abundant and low-cost renewable energy, combined with complementary natural resources, should specialize in energy-intensive low-carbon goods—a prediction that follows directly from standard microeconomic reasoning. What has changed is the structure of costs.

A critical distinction emerges between two concepts: structural competitiveness, which reflects underlying production fundamentals including energy costs and natural resource endowments, and effective competitiveness, which reflects the prices firms actually face after accounting for subsidies, trade barriers, certification systems, and regulatory asymmetries. In many sectors, policy interventions (like non-tariff barriers and subsidies) are now substantial enough to materially alter competitiveness rankings across countries (Arbache, 2026a).

The cost structure in hard-to-abate sectors makes geography decisive. In green hydrogen, electricity represents 60% to 80% of production costs via electrolysis (IEA, 2023). In aluminum, electricity accounts for 30% to 40% of smelting costs. Given that renewable electricity costs vary across countries by factors of two to four in harmonized estimates (OECD/NEA, 2020), the resulting cost differentials are sufficient to determine industrial location. Table 1 illustrates this dispersion.

Table 1: Harmonized Renewable Electricity Costs Across Selected Countries (LCOE, USD/MWh, 7% discount rate)

Country Group	Country	Solar PV ($/MWh)	Onshore wind ($/MWh)
Renewable-resource abundant	India	35.5	35.8
Renewable-resource abundant	Brazil	46.0	33.6
Renewable-resource abundant	Australia	37.1	43.0
Reference large market	United States	43.7	61.3
Structurally higher cost	Belgium	90.2	67.2
Structurally higher cost	Japan	172.1	140.2
Structurally higher cost	Korea	96.6	113.3

Source: Authors’ compilation from IEA/OECD-NEA, Projected Costs of Generating Electricity 2020.

5. Distorted Prices and the Architecture of Green Industrial Policy

The present phase of the energy transition is characterized by unprecedented policy intervention. We distinguish between two broad categories: corrective policies that address market failures, internalize externalities, and build institutional infrastructure, and distortive policies that permanently offset structural disadvantages through sustained fiscal transfers, local-content mandates, or trade barriers that override comparative advantage (Arbache, 2025c; 2026e).

The cumulative effect of current interventions is to redirect investment toward structurally inefficient locations. In hydrogen, the U.S. Inflation Reduction Act Section 45V Production Tax Credit was approved to provide up to $3 per kilo—a wedge that can approach or exceed the entire structural cost advantage of renewable-rich exporters. While the original Inflation Reduction Act (IRA) of 2022 established this credit to last through 2032, a new law—the One Big Beautiful Act of 2025 (Public Law 119-21)—has significantly altered its timeline. Under the 2025 legislation, the 45V credit is being phased out and will be eliminated for any facility that begins construction after December 31, 2027. However, it remains an example of our point about the effect of interventions.

In Europe, policy initiatives the Hydrogen Bank, ReFuelEU, and the carbon border adjustment mechanism (CBAM) combine subsidies, requirements, and regulatory barriers that reshape effective competitiveness. Similar interventions are being increasingly seen in Canada, Australia, Japan, China, and the Middle East. What is emerging is not a coherent global market for decarbonization, but a fragmented system in which governments simultaneously subsidize domestic production, restrict external competition, and selectively create demand under nationally defined regulatory frameworks.

6. Structural Costs, Policy Wedges, and the Logic of Location Decisions

Understanding why green investment flows to suboptimal locations requires a distinction to be made between two layers of cost that firms face when deciding where to produce.

The first layer is structural cost: the underlying cost of producing a low-carbon good based on fundamental economic geography. In most hard-to-abate sectors, the dominant component is the local price of clean energy at the point of industrial use. This can account for the greatest part of total production costs in industries such as green hydrogen and aluminum smelting. The remaining components are variable production inputs (labor, water, logistics, raw materials) and capital-related costs. Because renewable energy prices differ so dramatically across geographies, this structural cost layer already establishes a strong presumption on where production should locate: near the cheapest, cleanest, and most abundant energy sources.

The second layer is effective cost: what firms actually pay after accounting for three categories of policy intervention. First, domestic subsidies and tax credits reduce the costs faced by producers in countries where such support is offered. Second, tariffs and trade barriers raise the effective cost for accessing destination markets of goods produced elsewhere. Third, regulatory compliance costs and certification frictions, such as the need to prove that hydrogen was produced using a specific grid-accounting methodology, or that sustainable aviation fuels meet particular traceability standards, add costs that differ by origin country and destination market.

Firms invest where effective costs are lowest. When these two cost layers align—when the places with the lowest structural costs also face the lowest effective costs—investment flows efficiently toward renewable-rich locations, and decarbonization advances faster and more affordably. When they diverge because subsidies make structurally expensive production appear competitive, or because trade and regulatory barriers raise the effective costs facing structurally efficient producers, investment is redirected away from where it should most economically go.

The main implication is that the geography of decarbonization investment is not primarily determined by physics or economics, but by the size of policy wedges: the larger the gap between structural and effective costs, the further the resulting investment geography deviates from the efficient optimum.

7. Demand Uncertainty and the Bankability Constraint

Cost distortions alone do not fully explain the slow reallocation of green investment. Even when a producer in a renewable-rich country has a genuine structural cost advantage, and even when that advantage survives the distortions described above, projects frequently fail to attract financing for a different reason: investors are not confident that the output will find buyers.

This is the bankability problem. A project only becomes financeable when investors expect, with reasonable confidence, that future revenues will cover operating costs and generate a return on the capital invested. In practice, this requires not just competitive costs but a credible, accessible market for the output. In the current landscape, that credibility is frequently absent.

The sources of demand uncertainty in clean energy markets are multiple and reinforcing. For example, certification requirements for green hydrogen vary across importing regions, making it unclear whether a project built to one set of standards will be eligible to sell in other markets. Long-term offtake agreements—typically required by banks before financing large projects—are difficult to secure when buyers themselves are uncertain about future regulatory requirements. Fragmented standards create situations in which the same physical product may qualify as ‘green’ in one jurisdiction but not another, depending on criteria that are still being negotiated.

The result is a gap between projects that are economically viable in principle and those that are financeable in practice. A structurally efficient project with competitive production costs remains unbankable if the likelihood of securing market access at prices sufficient to cover costs and service debt is too low—regardless of how good the fundamentals are.

This bankability constraint and the price-distortion problem described earlier interact and reinforce each other. When subsidies favor domestic production in importing regions, they simultaneously raise the effective cost for external producers and reduce the likelihood that those producers will be eligible for the procurement frameworks through which offtake agreements are structured. Conversely, when institutions are weak and markets are fragmented, the relative attractiveness of subsidized domestic production increases even further. The true cost of current policy fragmentation is likely more than the sum of its individual parts: price distortions and institutional gaps compound each other, making the geography of decarbonization less efficient, less inclusive, and more costly than either mechanism in isolation would suggest.

8. Misallocation: From Micro Distortions to Aggregate Losses

When green investment is directed systematically toward structurally higher-cost locations, the aggregate economic consequences are significant. The welfare cost of misallocation equals the difference between what it actually costs to produce a given volume of clean goods under the current distorted geography, and what it would cost to produce the same volume under an efficient geography, in which production is located where structural costs are lowest. Multiplied across the full scale of projected clean energy markets, this gap translates into substantial and unnecessary economic losses every year.

Table 2 provides illustrative estimates for three important sectors under mid-2030s market scale assumptions.

Table 2: Illustrative Aggregate Cost Implications of Misallocation (projected mid-2030s)

Sector	Projected output value ($ billions)	Cost penalty (%)	Additional annual cost ($ billions)
Green hydrogen	300	15%	45
Sustainable aviation fuels	150	20%	30
Low-carbon aluminum	250	10%	25
Total (illustrative)	700	—	~100+

Source: Authors’ calibration based on IEA WEO 2023, IRENA 2022, ICAO 2050 Net Zero roadmap and sectoral assessments. Estimates are illustrative of direction and scale, not forecasts.

Even these substantial static estimates are likely to understate the true cost of misallocation, because they omit dynamic inefficiencies that accumulate over time. When deployment is systematically concentrated in higher-cost locations, the global cost curve for clean technologies declines more slowly because scale economies and learning-by-doing effects, which drive down costs over time, are spread across a smaller and more expensive base. Fiscal costs compound through subsidy competition: as countries outbid each other to attract investment, regardless of structural advantage, the cumulative public expenditure required to sustain a given level of deployment rises. Distortions propagate through value chains, as higher upstream costs in green hydrogen or aluminum cascade into the costs of fertilizers, transport, construction materials, and renewable infrastructure itself. Meanwhile, climate outcomes worsen, as higher costs slow the speed of deployment and delay the emissions reductions the transition is meant to achieve (Arbache, 2025d).

9. Development Implications

For many developing economies, low-carbon industrialization may represent one of the largest industrial opportunities in decades. Compared to previous waves of industrialization, where latecomers faced high technological barriers and sharply shrinking opportunities in labor-intensive manufacturing, several energy-intensive, low-carbon sectors retain a degree of geographical openness that creates meaningful entry possibilities for resource-rich economies. Brazil has structural advantages in sustainable aviation fuels, green ammonia, low-carbon aluminum, and green chemicals. In Chile, it is green hydrogen. For Namibia, Morocco, and Egypt, it is hydrogen derivatives, while in Indonesia it is battery value chains.

If current policy distortions continue to suppress investment in structurally efficient locations, developing economies may remain concentrated in lower-value segments of global trade, while higher-value activities remain concentrated elsewhere. This would replicate familiar patterns of uneven industrial development (Arbache and Esteves, 2023; Arbache, 2025a). The irony is that the current transition creates unusually favorable conditions for breaking this pattern. Realizing that opportunity requires institutional frameworks that enable efficient cross-border specialization, rather than systematically suppressing it.

10. Conclusion

The global decarbonization challenge is frequently misunderstood when treated primarily as a technological problem. The transition is equally a problem of economic geography, relative prices, institutional coordination, and global production reallocation. We have identified two major distortions: the growing divergence between structural costs and effective prices created by industrial policy interventions, and demand uncertainty that leaves structurally competitive projects unbankable. Together, these distortions slow technological learning, increase fiscal burdens, delay emissions reductions, reinforce industrial concentration, and suppress development opportunities in renewable-rich emerging economies.

The concept of powershoring captures the efficient response to these structural changes: as the energy intensity of industrial production interacts with the spatial fixity of renewable resources, the efficient organization of global production increasingly requires upstream activities to locate on the basis of energy geography, rather than market proximity. Whether this reorganization occurs at scale will depend on whether global policy frameworks enable it, or continue to suppress it through sustained price distortions and institutional fragmentation.

The central policy challenge is therefore not whether governments should intervene, but whether intervention helps align climate goals, economic efficiency, resilience, and broader development opportunities. The future of decarbonization may depend less on whether the world can invent enough clean technologies, and more on whether it can enable optimum deployment of these technologies in line with economics, geography, and physics.

References

Arbache, J. (2022). Powershoring. Development Bank of Latin America and the Caribbean (CAF). https://www.caf.com/en/blog/powershoring/

Arbache, J. (2023). Back to Geography. Development Bank of Latin America and the Caribbean (CAF). https://www.caf.com/en/blog/back-to-geography/

Arbache, J. (2025a). The Role of Trade in Global Energy Transition. CEBRI / iCS. https://cebri.org/en/doc/400/the-role-of-trade-in-the-global-energy-transition

Arbache, J. (2025c). Efficiency or Resilience? A Paradigm Shift in Global Value Chains. LinkedIn.

Arbache, J. (2025d). The Geography of Decarbonization: Distorted Relative Prices and the Misallocation of Global Green Investment. Texto de Discussão do Setor Elétrico Nº 156, updated May 2026, GESEL, UFRJ.

Arbache, J. (2025d). Trade Is the World’s Untapped Climate Accelerator. Forum on Trade, Environment & the SDGs (TESS), Geneva Graduate Institute.

Arbache, J. (2026a). Geography, Relative Prices, and the Reorganization of Global Production. LinkedIn.

Arbache, J. (2026b). From the Weightless Economy to the Return of Geography. LinkedIn.

Arbache, J. (2026c). Geography, Relative Prices, and the Reorganization of Global Production. LinkedIn.

Arbache, J. (2026d). Powershoring, Energy Density, and the Trade Economics of Decarbonization. LinkedIn.

Arbache, J. (2026e). Why Industrial Decarbonization Should Be Designed Around Value Chains, Not Plants. LinkedIn.

Arbache, J., & Esteves, L. A. (2023). Resilience with efficiency: How powershoring can contribute to the decarbonization and economic development of Latin America and the Caribbean. CAF. https://scioteca.caf.com/handle/123456789/2393

Arrow, K. J. (1962). The economic implications of learning by doing. Review of Economic Studies, 29(3), 155–173.

Baldwin, R. (2016). The Great Convergence: Information Technology and the New Globalization. Harvard University Press.

BloombergNEF. (2024). Hydrogen Levelized Cost Update 2024.

Canuto, O. (2013). Trade: The World Is Not Flat Yet. World Bank.

Canuto, O. & Daoulas, C. (2019). Natural Wealth and Economic Growth: The Case of Sub-Saharan Africa. Policy Center for the New South, Policy Paper PP-19/12.

Canuto, O. (2021a). The Road to Decarbonization. Policy Center for the New South.

Canuto, O. (2021b). Decarbonization and “Greenflation.” Policy Center for the New South, Policy Brief PB-51/21.

Canuto, O. (2021c). Climbing a High Ladder – Development in the Global Economy. Policy Center for the New South.

Energy Transitions Commission. (2023). Material and Resource Requirements for the Energy Transition.

European Commission. (2024a,b). European Hydrogen Bank; ReFuelEU Aviation Regulation. Brussels.

European Union. (2023). Regulation (EU) 2023/956 establishing a carbon border adjustment mechanism.

Friedman, T. (2007). The World Is Flat. Picador.

Gereffi, G., Humphrey, J., & Sturgeon, T. (2005). The governance of global value chains. Review of International Political Economy, 12(1), 78–104.

H2Global Foundation. (2024). H2Global Mechanism: Double Auction Model for Hydrogen Market Creation.

Hydrogen Council. (2024). Hydrogen Insights 2024.

ICAO. (2024). Net Zero 2050 Roadmap. Montreal.

IEA. (2019). The Future of Hydrogen. Paris. | IEA. (2020). Iron and Steel Technology Roadmap. Paris. | IEA. (2021). Aluminium Tracking Report. Paris.

IEA. (2023). World Energy Outlook 2023. Paris. | IEA. (2024). Global Hydrogen Review 2024. Paris.

IMF. (2023). Geoeconomic Fragmentation and the Future of Multilateralism. Washington, DC.

IRENA. (2022a). Green Hydrogen Cost Reduction. Abu Dhabi.

IRENA. (2022b). World Energy Transitions Outlook 2022.

McKinsey & Company. (2024). Sustainable Aviation Fuels: Brazil’s Competitive Potential.

OECD Nuclear Energy Agency. (2020). Projected Costs of Generating Electricity 2020. Paris: OECD/NEA.

OECD. (2023, 2024). Industrial Policy for Net Zero Transition; The Rising Role of Industrial Policy in Climate Transition. Paris.

Prebisch, R. (1950). The Economic Development of Latin America and Its Principal Problems. United Nations ECLA.

Quah, D. (1999). The weightless economy in economic development. CEPR Discussion Paper, LSE.

U.S. IRS. (2024). Section 45V Clean Hydrogen Production Tax Credit Guidance. | U.S. DOE. (2024). Clean Hydrogen Production Tax Credit Guidance.

UNCTAD. (2023). Trade and Development Report 2023. Geneva.

Way, R., Ives, M., Mealy, P., & Farmer, J. D. (2022). Empirically grounded technology forecasts and the energy transition. Joule, 6(9), 2057–2082.

World Bank. (2020). World Development Report 2020: Trading for Development in the Age of Global Value Chains. | World Bank. (2023). Scaling Hydrogen Finance in Emerging Markets.

WTO and World Bank. (2023). Trade Policy Tools for Climate Action. Geneva/Washington.

Wright, T. P. (1936). Factors affecting the cost of airplanes. Journal of the Aeronautical Sciences, 3(4), 122–128.

Yang, P. & Canuto, O. (2024). Capital, Labor, and Land in the Digital Transition. Policy Center for the New South, Policy Paper PP-17/24.

Jorge Arbache, Associate professor, University of Brasilia and Dom Cabral Foundation (FDC), and Senior Fellow, Climate and Society Institute (iCS). He is a former deputy minister and chief economist at Brazil’s Ministry of Planning, vice president for the private sector at the Development Bank of Latin America and the Caribbean, board member at BNDES, and senior economist at the World Bank.

Otaviano Canuto, a former vice president and executive director of the World Bank, executive director of the International Monetary Fund, vice president of the Inter-American Development Bank, and deputy minister of finance for international affairs of Brazil, is currently a senior fellow at the Policy Center for the New South and a non-resident senior fellow at the Brookings Institution.