What happens to a country's creditworthiness when some of its debts are secret?
Undisclosed Debt Sustainability
Laura Alfaro, Fabio Kanczuk (2023) · AEA Papers and Proceedings 112
Article ↗ · Replication data ↗ (openICPSR)
The archive requires an account we do not hold, so the files were not opened. What follows rests on the published abstract and on this platform's own observations.
What the authors found
Alfaro and Kanczuk address the rise of non-Paris Club lending, notably from China, to low- and middle-income countries. Unlike traditional sovereign debt these arrangements are not public, so other creditors cannot see their size. Adapting a standard sovereign default model to incomplete information, the authors find undisclosed debt substantially reduces the sustainability of traditional creditors' claims. Disclosure would deliver significant welfare gains to recipient countries while reducing how much they can sustainably borrow.
What our data adds
External debt is now loaded, and the distribution is stark. Mozambique owes 350.6 per cent of its gross national income abroad, Lebanon 331.6, Mongolia 181.9 and Senegal 150.7. The ratio that decides whether any of that is affordable is debt service against exports, because exports are what earns the foreign currency a foreign debt must be repaid in: El Salvador spends 96.2 per cent of its export earnings servicing debt, Haiti 63.2 and Egypt 49.2. A country in that position is running to stand still, and it is doing so on debts that are disclosed. The paper's subject is the debts that are not.
Each country at its own most recent year.
Exports earn the foreign currency a foreign debt must be repaid in, which is why this is the affordability ratio.
| Figure | Value | Indicator |
|---|---|---|
| Mozambique, external debt as a share of GNI | 350.6% | DT.DOD.DECT.GN.ZS |
| Lebanon | 331.6% | DT.DOD.DECT.GN.ZS |
| El Salvador, debt service as a share of exports | 96.2% | DT.TDS.DECT.EX.ZS |
| Haiti | 63.2% | DT.TDS.DECT.EX.ZS |
Every figure here is computed from observations in this database and can be charted, downloaded or queried.
What this cannot show
The series comes from the World Bank's Debtor Reporting System, which rests on what borrowing countries report, so it is precisely the thing Alfaro and Kanczuk argue is incomplete. Undisclosed debt is by definition absent from it, and no dataset can fix that — work in this area proceeds by assembling loan-level records from contracts and press reports, which is a document shape rather than a series. Nor does this series carry the creditor composition the paper turns on: it says how much is owed, not to whom. So the figures above set the scale of the problem and cannot address the paper's actual claim.
The data behind it
External debt stocks, total (DOD, current US$)
DT.DOD.DECT.CD
WBExternal debt stocks (% of GNI)
DT.DOD.DECT.GN.ZS
WBTotal debt service (% of exports of goods, services and primary income)
DT.TDS.DECT.EX.ZS