Inside the finance ministry of a low-income country, debt sustainability has long ceased to be a matter of simply adding up what is owed abroad. Domestic bonds must be refinanced, money preserved for roads or power plants, higher interest rates absorbed, and sometimes the damage from a drought or a flood repaired. With every new loan, a development question moves a little closer to becoming a solvency question.

Thousands of miles away, in Washington, that boundary is partly drawn by a model.

On September 21, 2026, the International Monetary Fund and the World Bank approved an overhaul of their joint Debt Sustainability Framework for Low-Income Countries. The instrument is little known outside finance ministries, financial institutions and sovereign debt markets. Yet it carries considerable weight. It is used to identify a country's vulnerabilities, assess its debt-carrying capacity and inform financing decisions by both institutions. Its analyses are also used by governments and creditors to assess sovereign risk.

The reform is due to become operational in the second half of 2027. It represents the most significant change to the framework since its previous major overhaul in 2017.

What is changing, however, is not merely the model.

It is the world the model is trying to measure.

A framework built for a different kind of debt

When the current framework was designed, debt in the world's poorest economies could still largely be understood through a relatively familiar relationship: governments with shallow financial markets borrowed primarily from other governments, multilateral institutions or external creditors, often on concessional terms.

That architecture has not disappeared. But it is no longer enough to describe what now sits on sovereign balance sheets.

Sources of financing have multiplied. Some governments gained access to international capital markets. Others borrowed more heavily from their own banks, pension funds and institutional investors. Commercial financing became more important, while state-owned enterprises, guarantees and various contingent liabilities made the state's true exposure increasingly difficult to measure.

The IMF and the World Bank themselves acknowledge this transformation. Their review comes in an environment marked by elevated debt vulnerabilities, a more diverse creditor base and greater reliance on both domestic and external financing on commercial terms.

Debt has therefore not simply grown.

Its geography has changed.

And that transformation alters the nature of a sovereign debt crisis.

When the creditor is at home

A restructuring of external debt can impose losses on international banks, funds or creditor governments. Restructuring domestic debt produces a different mechanism: the institutions holding government securities may be the same ones safeguarding household deposits, financing businesses and managing part of the country's savings.

Reducing the burden on the state can then weaken the institutions financing the economy.

This is one reason domestic debt occupies a much larger place in the new framework. The IMF and the World Bank intend to assess more systematically the risks it creates for both sovereign sustainability and the domestic financial system.

The distinction matters.

A government may appear relatively insulated from an external shock because much of its debt is denominated in its own currency and held at home. But that apparent protection may simply have moved the risk. When domestic banks accumulate sovereign bonds, public finances and bank balance sheets become increasingly intertwined. A crisis of the state damages the banks; weaker banks, in turn, reduce the economy's capacity to finance the state.

The boundary between a sovereign crisis and a financial crisis becomes much harder to identify.

The new framework is an attempt to see this intermediate territory more clearly.

Measuring what the state really owes

Another difficulty arises before anyone can decide whether debt is sustainable: determining exactly what counts as debt.

A sovereign balance sheet does not necessarily end with bonds issued by the Treasury. A state-owned enterprise can borrow on the assumption that the government will intervene if it fails. A government can guarantee financing without immediately spending anything. Some liabilities can remain outside the fiscal perimeter for years before suddenly returning to the public accounts.

The revised framework will therefore broaden scrutiny of state-owned enterprises, guarantees and insufficiently documented liabilities.

The change sounds technical. It goes to the heart of sovereign-risk measurement.

Public debt can appear manageable until obligations that were not clearly visible become actual claims on the state. Recent debt crises have repeatedly demonstrated that the gap between reported debt and genuine exposure can become especially consequential precisely when a country has the least room to absorb it.

The reform therefore introduces an indicator designed to assess confidence in the debt data being used. It also strengthens realism checks and stress scenarios intended to test official fiscal trajectories against what an economy might actually be able to sustain.

Yet the attempt to measure debt more accurately leads to an even harder question: is measuring existing debt enough when countries face radically different future needs?

Everything that remains to be built

For many low-income economies, the fiscal equation contains a feature that debt ratios capture imperfectly: much of the capital already available to developed economies has yet to be built.

Power grids, roads, ports, irrigation systems, hospitals, schools and digital infrastructure all require financing even as refinancing costs rise. Climate adaptation is adding another layer of investment needs.

The contradiction is profound.

Restricting borrowing can protect a country from a future financial crisis. But restricting investment too severely can also weaken the growth that would allow the country to sustain its debt in the first place.

The revised framework will therefore introduce a longer-term analysis intended, among other things, to better assess how development and climate-adaptation needs affect public debt. This does not mean automatically treating debt incurred for infrastructure as sustainable. A useless power plant remains a liability; a road that fails to generate the expected economic gains still has to be paid for.

The question becomes more subtle: how much can a country invest today without undermining its ability to finance what still needs to be built tomorrow?

That question gradually shifts the meaning of sustainability itself.

The boundary is not a number

Since its creation in 2005, the IMF-World Bank framework has attempted to solve a difficult problem: turning uncertain economic trajectories into categories clear enough to guide financial decisions.

The system combines debt-carrying capacity, debt indicators, thresholds, projections, stress scenarios and staff judgment. The resulting analyses can classify a country according to its risk of debt distress or conclude that its debt has become unsustainable.

Those words have consequences.

They influence how the IMF and the World Bank lend. They can affect the conditions under which a government is permitted or encouraged to assume new debt. They are read by other creditors deciding whether to continue financing a country. And when conditions deteriorate far enough, the sustainability assessment becomes a central part of discussions over a possible restructuring.

The reform now seeks to distinguish more precisely between different forms of vulnerability and to make assessments more responsive to the characteristics of individual economies. The harmonized discount rate used by the two institutions will remain at 5%.

But as the measurement becomes more precise, another question emerges: who can verify the boundary it draws?

The paradox of precision

Some parameters of the new framework will not initially be public.

The probability thresholds used to generate the mechanical signal of debt unsustainability are, at least initially, expected to remain confidential, along with elements of that signal in country analyses. A majority of IMF Executive Directors accepted this approach during the transition to the new framework.

There are technical arguments for preventing a mechanical indicator from being interpreted as a verdict in its own right. A debt sustainability analysis is not an algorithm capable of predicting a crisis with precision. Projections depend on assumptions about growth, interest rates, government revenue, exchange rates and events that remain inherently uncertain.

But the decision creates a tension.

The institutions want to produce a more granular assessment, better adapted to the characteristics of individual economies, while temporarily keeping some of the parameters contributing to that assessment outside the public domain. Governments, creditors, researchers and civil society may be able to observe the result without necessarily being able to reconstruct the entire quantitative path that produced it.

Greater sophistication can improve diagnosis.

It can also make that diagnosis harder to reproduce.

This question will acquire its full significance when the new framework takes effect in 2027. Only then will it become possible to see which countries change category, how domestic debt alters assessments, and whether incorporating development needs genuinely creates greater differentiation among economies previously assessed through more uniform thresholds.

Because behind the models lies a much more tangible decision.

A low-income country sometimes has to borrow because it already carries too much debt. It sometimes has to borrow because it still lacks enough roads, electricity, water systems, housing or protection against a changing climate. Between those two realities lies a boundary economists call sustainability.

From 2027, the IMF and the World Bank will draw it differently.

And for dozens of countries, moving that line a few inches on a model could move billions in the real world.

Main sources

International Monetary Fund — 2026 Review of IMF and World Bank Debt Sustainability Framework for Low Income Countries; consultation documents and technical documentation related to the reform.

World Bank — 2026 Review of IMF and WBG Debt Sustainability Framework for Low Income Countries; Debt Sustainability Framework and Debt Sustainability Analysis documentation.

International Monetary Fund — Macroeconomic Developments and Prospects in Low-Income Countries — 2026.

International Monetary Fund — Managing Director's statement following the G20 Finance Ministers and Central Bank Governors meeting, September 1, 2026.

Reuters — coverage of the IMF and World Bank boards' approval of the reform, September 21, 2026.