The research, published in the open-access journal Heliyon by Jerry Ogutu Sumba, Kennedy Nyabuto Ocharo and Paul Joshua Mugambi, is notable for how it frames the question. Instead of asking simply whether debt is good or bad for growth, it asks where the switch happens, whether the answer depends on a country’s income level, and whether the quality of a nation’s institutions changes how much debt an economy can safely swallow. Economists have debated these questions for decades, with studies variously reporting positive, negative or no relationship between borrowing and growth, largely because linear models obscure the tipping points that matter in the real world.

The historical backdrop makes the findings urgent. Sub-Saharan Africa’s external debt surged in the 1980s, peaking at about 58 percent of GDP in 1994 before the Heavily Indebted Poor Countries Initiative and successive Paris Club relief agreements wrote off large arrears and restructured obligations. The region then enjoyed a lull: external debt fell to a low of 21.85 percent of GDP in 2008, a stretch marked by relatively stable growth. That discipline did not last. Spurred by heavy borrowing for infrastructure and military spending, the median public debt-to-GDP ratio in the region jumped from 28.8 percent in 2012 to 59.1 percent in 2022, and by 2020 the regional external debt ratio had climbed back to 47.16 percent just as GDP growth sank to its lowest point, minus 2.01 percent, a decline the authors note was compounded by the COVID-19 pandemic but consistent with debt’s persistent negative influence.

To pin down the turning point rather than assume it, the researchers used a Dynamic Panel Threshold Model applied to an unbalanced panel of 44 countries from 2005 to 2022, deliberately starting after the debt relief era so the sample reflects a homogeneous, post-restructuring phase. The method, developed by Seo and Shin, solves a problem that has hampered earlier threshold studies: it allows both the threshold variable and the other regressors to be endogenous, meaning it does not pretend that debt and growth move independently of each other. Previous static approaches assumed strict exogeneity, while standard generalized method of moments estimation, though capable of handling endogeneity, forced a linearity that hid exactly the non-linear behavior the researchers sought.

The results confirm a regime-switching relationship. In the full sample, external debt carries a small positive coefficient of 0.09 below the estimated threshold of 47.23 percent of GDP, then flips to a significantly negative coefficient of minus 0.17 above it. Splitting the sample, the threshold sits at 42.64 percent for the 20 low-income countries and 49.81 percent for the 24 middle-income countries. Intriguingly, the 95 percent confidence intervals for these two groups overlap, which means the apparent income-based difference is not statistically distinguishable. That finding contradicts a common assumption, embedded in many debt sustainability frameworks, that middle-income countries can inherently carry proportionally heavier external debt because of stronger fiscal capacity.

Diagnostic tests support the robustness of the estimate. Arellano-Bond tests show first-order but no second-order serial correlation, the Hansen J statistic is insignificant, and excluding Liberia, whose extreme 2005 debt stock of nearly 498 percent of GDP towers over the sample mean of 42.43 percent, moves the threshold only marginally to 46.91 percent. The study also documents a consistently negative effect of domestic debt on growth across all groupings, though milder in middle-income countries, plausibly reflecting deeper and more liquid local financial markets. A strong negative pairwise correlation between external debt and GDP growth, minus 0.58, underscores the central tension in the data.

The most consequential findings concern institutions. Drawing on public choice theory, which holds that self-interested political elites determine whether borrowed funds are invested productively or siphoned into private utilities, the researchers measured institutional quality as the mean of six Worldwide Governance Indicators, spanning rule of law, corruption control, regulatory quality, government effectiveness, voice and accountability, and political stability. When institutional quality was modeled as a threshold variable, debt exerted a significant negative effect on growth in countries below an index level of roughly 0.77 and a positive effect above it. In other words, the so-called debt trap is not purely a function of how much a country borrows; it is also a consequence of how well its institutions function.

Stronger institutions do more than soften the blow of existing debt; they actively raise the ceiling. When the external debt and institutional quality interaction was added to the model, the effective threshold for the full sample shifted upward from 47.23 percent to 53.17 percent of GDP. Middle-income countries gained the most, with their threshold rising by 6.42 percentage points and the strongest interaction coefficient, while low-income countries saw a smaller shift of 3.86 points that was not statistically significant, reflecting persistently weak governance, pervasive corruption and limited state capacity to deploy borrowed funds well. The picture that emerges is of governance, not income, as the central, policy-actionable determinant of debt absorption capacity.

The practical implications are considerable. The authors recommend that sub-Saharan governments keep external debt below the estimated 47 percent benchmark and, crucially, treat institutional strengthening as an integral component of debt management rather than a separate governance aspiration, through sustained commitment to rule of law, transparency in public financial management, anti-corruption efforts and judicial independence. For multilateral lenders such as the IMF and World Bank, the results offer an empirical basis for linking financing terms to measurable governance indicators instead of relying solely on aggregate income-based ceilings. The authors also flag limitations: the model identifies statistical turning points rather than causal transmission channels, and it assumes a shared threshold within each sample, so the estimates should be read as regional benchmarks rather than country-specific limits, a gap future research with heterogeneous threshold models could close.

Subject of Research: The threshold effect of external debt on economic growth in sub-Saharan African countries and how income level and institutional quality condition debt sustainability

Article Title: External debt and economic growth among sub-Saharan African countries, a threshold analysis

Article References: Sumba, J. O., Ocharo, K. N., & Mugambi, P. J. (2026). External debt and economic growth among sub-Saharan African countries, a threshold analysis. Heliyon, 12(15), Article e45539. https://doi.org/10.1016/j.heliyon.2026.e45539

Image Credits: AI Generated

DOI: 10.1016/j.heliyon.2026.e45539

Keywords: external debt, economic growth, sub-Saharan Africa, debt threshold, debt overhang, institutional quality, Dynamic Panel Threshold Model, GDP, IMF, World Bank, public debt sustainability, governance