Why the Major Energy Transition Deals in Southeast Asia Are Not Taking Off
In November 2021, in Glasgow, the G7 governments and the European Union presented what they described as a new architecture of climate finance: the Just Energy Transition Partnerships. The idea was ambitious in its design. Rather than channelling funds toward scattered projects, donors committed tens of billions of dollars to specific economies where coal was deeply embedded in their electricity matrix, in exchange for concrete decarbonisation plans, regulatory reforms, and protection for workers in the sector. South Africa was the pilot. Indonesia and Vietnam came next.
Four years later, the balance sheet is uncomfortable. Funds have not flowed at the promised pace, large-scale renewable energy projects have still not reached financial close, and in March 2026 the United States government formally withdrew its participation in these partnerships, removing with it more than 3 billion dollars in commitments linked to Vietnam and Indonesia. What once appeared to be a paradigm shift in climate finance now faces an even more uncomfortable question: whether the problem was always the design, or whether the ambition of the instrument simply exceeded the political capacity of the countries that were meant to receive it.
The Model That Ignored the Political Economy of Coal
The architecture of the energy transition partnerships rested on a reasonable but incomplete premise: that recipient governments, when presented with a sufficiently large financing package, would have both the incentives and the capacity to carry out deep structural reforms. The problem is that those reforms do not happen in a vacuum. They happen within political systems where state-owned electricity companies are power actors, where coal sustains employment and fiscal revenues in specific regions, and where the price of electricity is not merely a market variable but an instrument of social cohesion.
In Indonesia, the state-owned company PLN supplies electricity to 98 percent of households in the country. The investment plan approved in 2023 estimated total needs of 97 billion dollars through 2030, of which the partnership covered only a fraction. The remainder depended on the private sector entering massively, attracted by reforms that PLN and the government never fully adopted. Among those reforms were the elimination of domestic price caps on coal, the increase of electricity tariffs for consumers, and the granting of sovereign guarantees for priority projects. Each of those measures carried a real political cost. None could be executed by technical decree.
The result is revealing. In early 2026, only 2.9 billion dollars had been approved under the framework of the Indonesian partnership, and of that figure, 1.8 billion corresponded to a loan from the Japanese agency JICA for the Jakarta metro, which did not form part of the programme's original objective. The bulk of the remainder consisted of loans for institutional capacity-building of the government, not for concrete renewable generation projects. The Cirata floating solar project, one of the few large-scale works that moved forward, was built outside the partnership framework.
Vietnam presents a different but equally revealing pattern. The country did experience a renewable boom between 2019 and 2023: wind and solar rose from 0.4 percent to 14 percent of electricity generated. But that growth happened before the partnership existed, driven by a fixed incentive tariff established in 2017, and its costs were absorbed primarily by the state-owned company EVN. Between 2022 and 2023, EVN accumulated losses equivalent to 1.97 billion dollars, pressured by rising payments to private producers and the obligation to purchase energy at above-market tariffs. By 2023, EVN's share of national installed capacity had fallen from 61 to 37 percent in just seven years. This experience, far from serving as a model, functions as a warning: both for EVN, which is hardly disposed to repeat it, and for Indonesia, which observes it with scepticism.
When the Promise and the Structure Do Not Align
The withdrawal of the United States in March 2026 did not create the problem, but it made it visible. The announcement by the Treasury Secretary exposed that a significant proportion of the commitments within these partnerships were not grants or concessional loans, but credits at market rates. Vietnam was the only one of the three countries that insisted from the outset that at least 7.5 billion dollars should arrive at rates lower than those of the open market. Neither Indonesia nor South Africa imposed that condition with the same firmness. The result was a design in which recipient countries assumed the risk of external debt while being asked to sacrifice tariff revenues and control over their own strategic companies.
This is not a failure of political will on the part of the recipients. It is a miscalibration of the instrument against the real economy. In South Africa, Eskom, the state electricity company, was running operating losses of more than 1.1 billion dollars when the agreement was signed in 2021. By 2025 it managed to record its first profit in eight years, equivalent to 1 billion dollars, but only after the government absorbed a substantial portion of its debt and consumers endured tariff increases that raised the company's revenues by 67 percent between 2021 and 2025. The costs of the adjustment were paid, in different proportions, by the state and by households. The international private sector watched, assessed the residual risk, and did not enter at the expected scale.
What the analysis of these three economies makes clear is a structural gap between two logics that the instrument never resolved. The donor's logic assumes that the main obstacle is financing, and that if capital is provided in sufficient quantities, markets and institutions will reform themselves to take advantage of it. The recipient's logic starts from the premise that the main asset being managed is the stability of the electricity system, and that any reform that raises prices, transfers risk to the state, or erodes control over the state-owned company is politically costly and potentially destabilising. These are not irrational logics. They are incompatible on the terms in which the agreement was framed.
A joint report by Bain and Standard Chartered warned that more than 35 percent of green investments announced in Southeast Asia could fail to materialise if bottlenecks in electricity grids, regulatory stability, and project execution capacity are not resolved. That percentage is nearly double what is observed in other comparable regions. The figure quantifies precisely the gap between announcement and conversion into real assets.
What Partial Failure Teaches About Climate Finance at Scale
To call these partnerships an outright failure would be inaccurate. In all three countries, debates about electricity sector reform accelerated, some regulatory frameworks advanced, and in cases such as South Africa the restructuring of Eskom, however slow, is taking place. The instrument generated institutional pressure that would not otherwise have existed. But the distance between what was promised and what was delivered is large enough to draw lessons that transcend the specific case.
The first is that climate finance at scale cannot be designed while ignoring the governance structure of the energy sector in the recipient country. State-owned electricity companies are not obstacles to be circumvented through external conditions: they are central actors whose behaviour determines whether projects reach construction or remain as plans. Designing a transition programme that requires PLN or EVN to cede market power, take on debt in foreign currency, and accept dollar-denominated contracts, without first resolving the financial equation of those companies, is to build on an assumption that the data does not support.
The second lesson is that combining market-rate loans with demands for deep structural reform creates a risk asymmetry that recipient countries have rational reasons to resist. If the projects work, the return flows to the international private sector. If the projects or the reforms generate tariff or political instability, the state absorbs the cost. Vietnam experienced that asymmetry first-hand between 2019 and 2023, and the institutional memory of that episode conditions its willingness to repeat it under any label whatsoever.
The third lesson, perhaps the most uncomfortable for the architects of these instruments, is that the social justice component never received financing proportional to its rhetorical importance. In the case of South Africa, barely 50 million dollars of the 8.5 billion dollars in the original package were directed toward economic diversification, worker retraining, and social inclusion. The coal-dependent communities that give the instrument its name and its political legitimacy received a marginal fraction of the resources. When the promise of justice is more narrative than budgetary, the domestic political actors who might defend the programme before their own constituencies have no concrete arguments to offer.
The American withdrawal accelerated the outcome but did not cause it. What these partnerships face is a design tension that no change of administration can resolve: an instrument conceived to reorganise national energy markets without a sufficiently robust theory of how the political economy of those markets actually functions. The next versions of this type of programme, if they come, will need to start from there.
