Domestic Debt, External Debt, and Inflation in Kenya: An ARDL-ECM Approach

1. INTRODUCTION

Public debt worldwide has been on a generally rising path for some twenty years now, pushed up by widening budget deficits, bigger outlays on development and social welfare, stimulus packages, deeper financial integration across borders, and repeated rounds of crisis-driven borrowing. The 2007–08 financial crisis, the Eurozone sovereign-debt troubles of 2009–12, and the COVID-19 pandemic each left sovereign balance sheets more stretched than before; by 2020, global public debt had reached 99% of GDP, the sharpest one-year jump since the Second World War. Advanced economies still carry the largest debt loads in absolute terms for example the G20 averaged 85.9% of GDP in 2021, well above Africa's 67.8% of GDP yet it is developing countries that face the sharper sustainability problem, given their thinner fiscal buffers, higher borrowing costs, and exposure to shocks originating elsewhere. Adewuyi et al. (2026). In Kenya's case, the debt-to-GDP ratio bottomed out at 32.2% in 2009 before climbing to 67.3% in 2022, pushing past every recommended ceiling on the books to 60% from the IMF, 64% from the World Bank, 65% from the African Union, and 50% from the East African Community. Behind that climb sit large outlays tied to devolution and development projects, a bloated wage bill, chronic budget shortfalls, and continued recourse to both concessional and commercial lenders abroad. Domestic and external sources now make up roughly equal shares of the total, a pattern that emerged deliberately after 2009 as MTDS cycles tilted toward domestic borrowing to limit currency exposure. Even so, the ceilings keep getting breached and then reset higher, and Kenya has been flagged by outside observers as a country at risk of a debt crisis. One lever that advanced-economy debt-management research has explored, but which developing-country research has largely skipped over, is inflation itself specifically, inflation that runs ahead of what markets had priced in, which can chip away at the real value of fixed-rate debt. Germany's hyperinflation of the 1920s, the inflationary stretch in the United States after the Second World War, and Latin America's hyperinflations of the 1980s and 1990s all lined up with sharp falls in real public-debt burdens. Whether anything comparable holds for a lower-middle-income, sub-Saharan African economy like Kenya and if so, how large the effect might be has not yet been tested. That is the question this paper takes up. The matching objectives are to estimate the relationship between inflation and the two debt categories, and to simulate how a 2% shock would alter each over five- and ten-year horizons. Three things make this research important. First, the existing evidence base is built almost entirely on advanced economies, so Kenya an emerging-market, lower-middle-income economy in sub-Saharan Africa fills a gap that simply has no developing-country comparator yet. Second, the National Treasury and the Central Bank of Kenya (CBK) stand to gain a directly usable input as they design future MTDS cycles. Third, by treating domestic and external debt as separate objects of analysis rather than lumping them into one aggregate, the paper adds a layer of detail that much of the existing literature skips over.

1.2 Research Questions and Objectives

Two questions frame the analysis:

• How does inflation affect Kenya's domestic and external public debt, and
• What would a 2-percentage-point inflation shock do to each?

2. LITERATURE REVIEW

2.1 Theoretical Framework: Shock Inflation

Keynes (1923) was among the first to describe how surprise inflation could be used to whittle down a government's fixed nominal obligations, pointing to several European governments that had, after the First World War, partly financed their debts by printing money in some cases cutting the real value of that debt by half or more. The channel runs through the Fisher relationship, r ≈ i − π: when inflation (π) surprises markets on the upside, the real interest rate (r) paid on existing fixed-rate, non-indexed debt falls temporarily, easing the real cost of servicing it. Three conditions make the effect bigger: debt with longer maturities, since rates cannot adjust quickly; debt denominated in domestic currency and not indexed to inflation; and inflation that catches markets off guard rather than inflation that has already been built into nominal rates. A second, related route is financial repression (McKinnon, 1974), which pairs capped or administered interest rates, a captive base of domestic investors, and capital controls with sustained inflation to push real interest rates persistently negative. This route can, in principle, achieve more than inflation surprises alone, but it depends on policy tools such as; interest-rate ceilings, compulsory holdings of government paper, and capital controls that sit outside the model used here. For that reason, financial repression is not adopted as the main analytical lens in this paper, though its complementary role is touched on later.

2.2 Empirical Evidence from Advanced Economies

Looking across 28 countries from 1900 to 2009, Reinhart and Sbrancia (2011) found that roughly a quarter of the years they classified as debt-liquidation years lined up with surprise-inflation episodes; the most extreme instances for example Germany in the 1920s, Argentina and Brazil in the 1990s saw public debt wiped out entirely, along with the collapse of domestic debt markets. A follow-up study covering ten advanced and emerging economies from 1945 to 1980 (Reinhart & Sbrancia, 2015) found liquidation conditions present in 92% of years, with the United States and United Kingdom each saving 2–3% of GDP annually in interest costs cumulatively on the order of 30–40% of GDP across a decade. Working with intertemporal-budget and Laffer-curve constraints on United States data spanning 1940–2009, Aizenman and Marion (2011) showed that even a modest 6% inflation shock could cut the debt ratio by about 20% within four years, with most of that effect falling on longer-maturity, non-indexed debt held abroad in domestic currency. Equiza- Goñi (2016) found a 1% permanent inflation shock cut Euro-Area debt by 4.2%, more than double the 1.7% figure estimated for the United States is a gap the author attributes to the Euro Area's longer debt maturities. Akitoby, Binder and Komatsuzaki (2017) simulated G7 net debt ratios for 2012–2017 and found that pushing inflation to 6% could lower net debt by 11–14.5% within five years, but only for medium- and long-term, non-indexed debt denominated in domestic currency. By contrast, Hilscher, Raviv and Reis (2022), using a forward-looking probabilistic model of United States debt, argued that meaningful debt debasement via inflation was unlikely there precisely because so much US debt is short-term and privately held serving as a reminder that maturity structure, not inflation alone, drives the result. Fukunaga, Komatsuzaki and Matsuoka (2022) simulated 19 advanced economies and found that a persistent rise to 6% inflation cut short-term debt by roughly 10% and long-term debt by roughly 10.1% over a decade, with the effect amplified wherever financial repression accompanied the inflation shock.

2.3 The Gap for Developing Countries

Nearly all of the empirical work on inflation-driven debt liquidation comes from the United States, the Euro Area, or other G7 economies; none of it, before now, has looked at a developing sub-Saharan African economy. Kenya-specific debt research has instead gone in the opposite direction, cataloguing the costs of debt accumulation and its drag on private investment (Kamundia, 2015; King’wara, 2014), its link to inflation (Ndemange, 2018), its association with currency depreciation (Arisa, 2020), and its bearing on income inequality (Obiero, 2022) poses a question without asking whether inflation could itself be turned into a debt-management instrument. That is the gap this paper tries to close.

3. METHODOLOGY

3.1 Theoretical Model

Combining the government's budget-constraint identity with the Fisher equation yields the standard debt-dynamics equation applied here: bₜ = −sₜ + bₜ₋₁ × (1 + iₜ) / [(1 + g) (1 + πₜ)] Here bₜ denotes the debt-to-GDP ratio, sₜ the primary surplus as a share of GDP, iₜ the implicit nominal interest rate on debt, gₜ real GDP growth, and πₜ the inflation rate. Holding everything else fixed, the identity shows higher inflation, pulling the debt ratio down which is the theoretical motivation for the empirical specification that follows.

3.2 Empirical Model Specification

Two structurally identical models were estimated, distinguishing domestic (d) from external (e) debt: bₜᵈ = β₀ + β₁sₜ + β₂bₜ₋₁ᵈ + β₃iₜᵈ + β₄gₜ + β₅πₜ + eₜ (Objective 1: domestic debt) bₜᵉ = β₀ + β₁sₜ + β₂bₜ₋₁ᵉ + β₃iₜᵉ + β₄gₜ + β₅πₜ + eₜ (Objective 2: external debt) The dataset spans 1983–2022 (40 annual observations) and was compiled from the Kenya National Bureau of Statistics (KNBS) and the Central Bank of Kenya (CBK). Implicit interest rates on each debt category were obtained by dividing each year's nominal interest payments by the previous year's outstanding stock, following the approach set out by Ryan and Maana (2014).

3.3 Estimation Strategy

Stationarity was checked with both the Augmented Dickey-Fuller (ADF) and Kwiatkowski–Phillips–Schmidt–Shin (KPSS) tests. Since the series showed a mix of I(0) and I(1) properties, cointegration was tested using the Pesaran, Shin and Smith (2001) ARDL bounds approach, which confirmed a levels relationship for both models at the 5% level. In the domestic- debt case, the long-run error-correction term came out positive and statistically insignificant as an ‘explosive’ correction term — so the short-run ARDL (4,4,1,3,3,4) specification was used instead. The external-debt model, on the other hand, produced a negative and significant error-correction term and was therefore kept in error-correction form as ARDL-ECM (1,0,3,4,3,2). Both specifications were then put through the standard battery of diagnostics: Breusch-Godfrey for autocorrelation, pairwise correlations for multicollinearity, Shapiro-Wilk and Jarque-Bera for normality, White's test for heteroskedasticity, Ramsey RESET for functional-form misspecification, and CUSUM-of-squares for parameter stability.

3.4 Forecasting and Shock-Inflation Simulation

Baseline forecasts out to 2027 and 2032 were built using the local-projection method. A persistent 2-percentage-point shock (Δₜ) was then layered onto the 2022 inflation rate, following the simulation logic set out by Akitoby et al. (2017), and the resulting ‘treated’ path was compared against the untreated baseline to gauge whether a liquidation effect shows up and by how much.

4. RESULTS AND DISCUSSION

4.1 Descriptive Statistics

Across the 1983–2022 window, domestic debt averaged 18.4% of GDP (ranging from 2.5% to 33.6%), and external debt averaged 32.8% (15.7% to 81.6%). The primary balance never moved out of deficit, averaging ~10.7% of GDP across the period. Domestic borrowing carried a strikingly high implicit interest rate of 25.8% on average, spiking to 116.8% in 1998 against just 3.7% on external debt, a gap that reflects Kenya's continued access to concessional terms abroad. Inflation itself averaged 11.1% a year, swinging between a low of 1.6% and a high of 46.0% in 1993. Figure 1. Composition of Public Debt in Kenya (share of domestic vs. external debt), 1983–2022. Source: authors computation from CBK data.

4.2 Inflation and Debt Trends

Inflation in Kenya has swung considerably over time averaging 17% through the 1990s, with a 46% spike in 1993 right after the 1992 elections before settling into single digits during the 2010s and 2020s. Since the Central Bank of Kenya adopted its 5±2.5% flexible target in 2005, that band has nonetheless been breached more than once amid economic and political shocks (Figure 2).

4.3 Stationarity, Cointegration and Diagnostics

The KPSS tests pointed to stationarity across all variables, while the ADF tests flagged domestic debt, external debt, and the implicit external interest rate as non-stationary at levels, but stationary once first differenced. ARDL bounds testing confirmed a cointegrating relationship in both models — F = 23.805 for domestic debt and F = 15.937 for external debt, both significant at 5%. Neither model showed problems on the standard diagnostic checks: no autocorrelation (Breusch- Godfrey), no heteroskedasticity (White's test, p = 0.420 and 0.422), no functional-form misspecification (Ramsey RESET, p = 0.157 and 0.286), and stable parameters (CUSUM of squares). Fit was strong in both cases, with R² of 99.3% for the domestic model and 99.1% for the external model.

4.4 The Effect of Inflation on Domestic Public Debt

In the short-run ARDL (4,4,1,3,3,4) results, every coefficient on current and lagged inflation carried the expected negative sign, but only the first and third lags reached statistical significance at 5%. A one-point rise in inflation one and three years earlier was associated with domestic debt falling by 0.13 and 0.46 percentage points respectively, and the Wald test confirms these lags are jointly significant. The pattern points to a liquidation effect that kicks in with a delay rather than immediate liquidation plausibly because Kenya's domestic debt sits mostly in shorter-maturity, domestically held instruments, matching the broader finding that shorter maturities are less responsive to inflation surprises (Shittu, et. al., 2026). The data also point to a non-linear link between the implicit domestic interest rate and debt: the squared interest- rate term is significant at several lags, implying that interest-rate effects on debt accumulation do not move in a straight line.

4.5 The Effect of Inflation on External Public Debt

The ARDL-ECM (1,0,3,4,3,2) model returned a negative, statistically significant error-correction term (−0.517, p < 0.001), confirming the system moves back toward long-run equilibrium over time. In that long run, a one-point rise in inflation tracks with a 2.35-point fall in the external debt ratio — a much bigger effect than what shows up for domestic debt, and one that fits the broader literature's finding that longer-maturity debt, which describes Kenya's external obligations to bilateral and multilateral concessional lenders, erodes more readily under inflation (Okwoche and Nikolaidou, 2024).). Short-run dynamics tell a different story, though: inflation and its lags are positively linked to external debt in the near term, meaning the liquidation effect on external debt is something that builds over years rather than showing up quickly. The long-run relationship turns at roughly 22.6% inflation, beyond which the liquidating effect starts to fade. The implicit external interest rate also carries a significant positive long-run link to debt where a one-point drop in that rate corresponds to a 3.49-point fall in external debt which fits the Fisher-effect channel through which inflation lowers real borrowing costs.

4.6 Forecast and Shock-Inflation Simulation Results

Figures 4 and 5 present schematic versions of the local-projection forecasts, built to match the magnitudes reported in the underlying analysis.

5. CONCLUSIONS AND POLICY IMPLICATIONS

This paper set out to test, for the first time in a developing-country setting, whether surprise inflation could be used to liquidate Kenya's public debt, looking at domestic and external debt separately over 1983–2022. Three conclusions stand out. First, inflation barely touches Kenya's domestic debt, which fits with that debt being short-maturity and held mostly at home. Second, external debt reacts far more strongly to inflation shocks, but only after a substantial delay: a 2% shock initially made the external position slightly worse over five years before what looks like a large reduction emerges after ten. Third, none of this would work as a quick fix by using inflation deliberately as a debt tool would require sustaining it over a long stretch rather than firing off a single shock, which matches what the wider literature says about the importance of persistence (Hilscher et al., 2022; Krause & Moyen, 2016). A few policy points follow from this. Within the existing 5±2.5% inflation target, the National Treasury and the Central Bank of Kenya might factor the debt-management side-effects of inflation into future MTDS revisions, rather than treating inflation purely as a cost to be squeezed out. Since the liquidation effect found here sits mostly with longer-maturity, externally held debt, any deliberate use of the channel would also need to go hand in hand with lengthening the average maturity of domestic debt and coordinating fiscal and monetary policy closely enough that a fiscally driven inflation shock is not simply cancelled out by monetary tightening. Given how small the estimated benefit is next to inflation's well- known costs such as eroded purchasing power, currency depreciation, and the risk of unrest, as seen in the protests over Kenya's 2024 Finance Bill. This paper stops short of recommending inflation as a frontline debt-management tool for Kenya. It is better thought of as a secondary factor worth keeping in view when fiscal and monetary policy are being coordinated. Two limitations are worth flagging: there is no comparable developing-country evidence to benchmark these results against, and the analysis is constrained by the data limits that come with a 40-year annual series. Later work could test larger shock sizes, break debt down by maturity in more detail, and look at how shock inflation interacts with financial- repression tools which administered interest rates, mandatory domestic-investor holdings, and the like which sat outside the present model but which the literature suggests would strengthen any liquidation effect.

REFERENCES

1. Adewuyi, A. A., Shittu, S. A., & Agyemang, J. K. (2026). The impact of debt overhang on private domestic investment in Nigeria. Redeemer's University Journal of Management and Social Sciences. https://runjmss.com/index.php/runojs/article/view/150
2. Aimola, A. U., & Monkam, N. (2025). Causal links between public debt and inflation in sub-Saharan African countries. The Journal of Economic Asymmetries, 32. https://doi.org/10.1016/j.jeca.2025.e00428
3. Aizenman, J., & Marion, N. (2011). Using inflation to erode the US public debt. Journal of Macroeconomics, 33(4), 524– 541. https://doi.org/10.1016/j.jmacro.2011.09.001
4. Akitoby, B., Binder, A., & Komatsuzaki, T. (2017). Inflation and public debt reversals in the G7 countries. Journal of Banking and Financial Economics, 1(7), 5–27. https://doi.org/10.7172/2353-6845.jbfe.2017.1.2
5. Arisa, G. M. (2020). Impact of external debt on inflation and exchange rate in Kenya [Unpublished master's thesis]. Kenyatta University. http://ir-library.ku.ac.ke/handle/123456789/21495 Cytonn Investments. (2025, July 20). Review of Kenya's public debt 2025. https://cytonn.com/topicals/review-of- kenyas-1
6. Does external debt drive inflation in Sudan? Evidence from symmetric and asymmetric ARDL approaches. (2023). Journal of Business and Socio-economic Development. https://doi.org/10.1108/jbsed-03-2023-0023
7. Equiza-Goñi, J. (2016). Government debt maturity and debt dynamics in euro area countries. Journal of Macroeconomics, 49, 292–311. https://doi.org/10.1016/j.jmacro.2016.01.005
8. Fukunaga, I., Komatsuzaki, T., & Matsuoka, H. (2019). Inflation and public debt reversals in advanced economies (IMF Working Paper No. 2019/297). International Monetary Fund. https://doi.org/10.5089/9781513521596.001
9. Hilscher, J., Raviv, A., & Reis, R. (2022). Inflating away the public debt? An empirical assessment. The Review of Financial Studies, 35(3), 1553–1595. https://doi.org/10.1093/rfs/hhab018
10. International Journal of Economics and Finance. (2025). Fiscal deficits and inflation in Mozambique: An ARDL approach, 2017–2023. International Journal of Economics and Finance, 17(5). https://ccsenet.org/journal/index.php/ijef/article/download/0/0/51555/56026
11. Kamundia, S. W. (2015). The effects of public debt on private investments and economic growth in Kenya. EKP, 13(3), 1–78.
12. Keynes, J. M. (1923). A tract on monetary reform. Macmillan. https://doi.org/10.4337/9781788118569.00031
13. King'wara, R. (2014). The impact of public debt on private investment in Kenya. Developing Country Studies, 4(22), 88– 97.
14. Krause, M. U., & Moyen, S. (2016). Public debt and changing inflation targets. American Economic Journal: Macroeconomics, 8(4), 142–176. https://doi.org/10.1257/mac.20140288
15. McKinnon, R. I. (1974). Money and capital in economic development. International Journal, 29(4), 649. https://doi.org/10.2307/40201473
16. National Treasury and Economic Planning, Republic of Kenya. (2025). Annual public debt management report, FY 2024/2025. https://www.treasury.go.ke/wp-content/uploads/2025/11/Annual-Public-Debt- Report-2024-2025.pdf
17. Nazamuddin, B. S., Wahyuni, S. S., Fakhruddin, F., & Fitriyani, F. (2024). The nexus between foreign exchange and external debt in Indonesia: Evidence from linear and nonlinear ARDL approaches. Journal of the Asia Pacific Economy, 29(2), 810–836. https://doi.org/10.1080/13547860.2022.2054153
18. Obiero, W. L. (2022). Do public and internal debt cause income inequality? Evidence from Kenya. Journal of Economics, Finance and Administrative Science, 27(53), 124–138. https://doi.org/10.1108/JEFAS-05-2021-0049
19. Okwoche, P., & Nikolaidou, E. (2024). Determinants of external, domestic, and total public debt in Nigeria: The role of conflict, arms imports, and military expenditure. Defence and Peace Economics, 35(2), 227–242. https://doi.org/10.1080/10242694.2022.2145709
20. Pratap, A., & Mishra, B. R. (2024). External debt and economic growth: Does the role of governance matter in emerging countries? SAGE Open, 14(2). https://doi.org/10.1177/21582440241243254
21. Reinhart, C. M., & Sbrancia, M. B. (2011). The liquidation of government debt (NBER Working Paper No. 16893). National Bureau of Economic Research. https://doi.org/10.3386/w16893
22. Reinhart, C. M., & Sbrancia, M. B. (2015). The liquidation of government debt. Economic Policy, 30(82), 291–333. https://doi.org/10.1093/epolic/eiv003
23. Ryan, T., & Maana, I. (2014). An assessment of Kenya's public debt dynamics and sustainability. Central Bank of Kenya.
24. Shittu, S. A., Adewale, A. A., & Agyemang, J. K. (2026). Public debt and inflation dynamics: A time-series analysis of a developing economy. Journal of Business and Social Review in Emerging Economies, 12(1), 15–28. https://doi.org/10.26710/jbsee.v12i1.3670
25. World Bank. (2025). International debt report 2025: Analysis of external debt stock and debt flows as of end-2024. World Bank Group. https://thedocs.worldbank.org/en/doc/6e72b0ded996306fa01f5db7a0c38b19-0050052021/related/IDR2025- Ch1.pdf