Просмотр новости
Найдите то, что Вас интересует
Спутники
РН
SpaceX
Роскосмос
NASA
СпутникиРНSpaceXРоскосмосNASABackground The COVID-19 pandemic found many governments around the world, and South Africa, in particular, unprepared financially; hence, they had to borrow. This paper aims to assess the sustainability of government debt in South Africa from 1997Q4 to 2023Q4. Method Data was collected from the South Africa Reserve Bank. The empirical investigation is conducted using an Autoregressive Distributed Lag (ARDL) framework. Results Bound cointegration technique outcome showed that there is cointegration between government expenditure and government revenue, meaning that there is fiscal sustainability in South Africa during the period of study. Further empirical appraisal using the ARDL technique showed that the relationship between domestic debt and primary balance is positive and significant. This means that debt is becoming unsustainable in South Africa. This study recommends that the government’s domestic debt needs to be reduced. Originality To assess the sustainability of debt, using cointegration and the ADRL techniques. Also, the debt variable was examined by looking at both the domestic and foreign debt.
On a global basis, financing national budgets requires sustainable financing policies worldwide to boost economic growth. When tax revenues fall below government spending estimates, they usually have no choice but to raise taxes and borrow domestically or internationally (Owusu-Nantwi & Erickson, 2016). Relying on credit as an alternative means for governments to avoid tax burdens leads to public debt (Ogunmuyiwa, 2011). These efforts have led to the accumulation of sovereign debt in many countries, resulting in the economic recession and debt crisis experienced by many developed and developing countries in the early 2000s (Donayre & Taivan, 2017). As a result, academic and political debates have renewed on the causal link between public debt and economic growth (Gómez-Puig & Sosvilla-Rivero, 2017). Studies show that countries are unlikely to run a budget surplus in which public debt is inevitable (Adom, 2016). Nevertheless, accumulating public debt at unfeasible levels can slow economic growth (Adom, 2016).
Global debt levels, which were already on an upward trajectory, experienced sharp acceleration due to the COVID-19 pandemic (Kose et al.,2020). To finance essential health interventions due to COVID-19 and mitigate the severe economic fallout, governments worldwide have been compelled to implement substantial fiscal stimulus packages, leading to widened financial deficits and increased borrowing from both domestic and international sources (Guo et al., 2025). The COVID-19 pandemic found many governments around the world, particularly South Africa, unprepared financially; hence, they had to borrow.
South Africa has experienced an increase in public debt over the years, from 63.5% of GDP in 2019/20 to 81.8 of GDP in 2020/21. This has stimulated interest in research on the future of South Africa’s debt. High levels of public debt beyond the country’s capacity work against economic growth and development of the economy. Not being able to sustain the level of debt will result in severe economic suffering.
When debt is not sustainable, it has an unfavorable effect on the country’s budget deficit. The national budget is negatively affected by the high cost of service debt. Consequently, the government finds it difficult to expand public services and investments. There are numerous consequences for the economy when debt is not sustainable; hence, there is a need to assess fiscal sustainability in South Africa.
After the 2007–2008 Global Financial Crisis (GFC), it was particularly challenging for the South African economy to restructure its public debt, which was accumulated by countercyclical fiscal policies. The same primary end occurred after the COVID-19 pandemic; consequently, South Africa sustained a deceleration in the extent of its economic growth. In 2017, public debt approached the sustainable maximum level of 50% of GDP. Because of the potential risk of endangering the stability of the government budget, the expense of debt payment continues to be a concern, which affects the expansion of public service and investment.
South Africa experienced an increase in public debt over the years. Heights were reached during the Covid-19 period from 63.5% of GDP in 2019/20 to 81.8 of GDP in 2020/21 ( South African Reserve Bank, 2021). This has stimulated concern among different role players in the economy about the future of South Africa’s debt. High levels of public debt beyond the country’s capacity work against the growth and development of the economy (Özmen and Mutascu, (2024). Not being able to sustain the level of debt will result in severe economic suffering.
The high level of public debt issuance and weak demand for debt from abroad have increased the government’s reliance on the domestic financial sector for financing (National Treasury, 2022). Consequently, dependence on the domestic financial sector for funding has mostly increased, whereby the government has drawn down its cash deposits, which are held by the SARB, and increased short-term borrowing through loans obtained internationally. In South Africa, the fiscal deficit is already high and continues to increase as a result of managing the impact of the COVID-19 pandemic by increasing spending on certain sectors of the economy, leading to an increase in public debt.
When debt is not sustainable, it has an unfavorable effect on the country’s budget deficit. The national budget is negatively affected by the high cost of service debt. Consequently, the government finds it difficult to expand public services and investments. Continuous rating downgrades translate to unaffordable debt costs, deteriorating asset values (such as retirement, other savings, and property) and reduction in disposable income for many”. There are numerous consequences to the economy when debt is not sustainable. Hence, there is a need to assess debt in South Africa by answering the research question: Is debt sustainable in South Africa?. This study contributes significantly to the literature in several ways. First, it stands out as one of the main studies that examines the sustainability of public debt within South African economies while accounting for the Covid19 Era. Additionally, this study employs a robust methodological approach that aids in appraising the sustainability of public debt in the short and long run. Finally, the study makes a comparative appraisal of two measures of debt sustainability that are relatively lacking in the extant literature.
According to the IMF (2002), debt sustainability is a condition whereby the borrower can pay off debts without making excessive future corrections to the income and spending balance. On the other hand, Wyplosz (2005) states that debt sustainability aims to measure the point at which the debt of a country is high and reaches a point where it can no longer be serviced. In addition, sustainability is linked to solvency, that is, a country’s ability to service debt in the long run (Naraidoo & Raputsoane, 2015). Redda (2020) and Abdulnasser (2002) postulate that debt sustainability refers to whether the government can maintain its borrowing pattern or alter the policy settings in the long run to satisfy its budget constraints. The crucial factor of debt sustainability is a country’s capability to service its debt in the long run or its solvency (Naraidoo & Raputsoane, 2015). In a sense, there should be limited or no difficulty for the country to meet its long-term financial obligations to its creditors.
Ramu (2021) assessed public debt sustainability in India at Karnataka Sub-National level from 1991 to 2018. The study revealed that Karnataka has managed to maintain the sustainability of its public debt over the past decades. The government’s growing fiscal discipline has contributed to the sustainability of public debt in the short and long run. This study supports the findings by Balbir, Atri, and Prakash (2017), whereby the study covered 20 Indian states using panel data from 1980–81 to 2015–16. The study shows that fiscal policies used in those selected states have stabilized the increase in debt and managed to sustain the existing debt in the long run.
Chandia et al. (2019) compared the sustainability of public debt and the external debt burden of India and Pakistan for 1971–2017. This study used two equations to analyze the debt dynamics equation for overall public debt sustainability and the debt dynamics equation for external debt sustainability. The study found that the primary budget deficit and current account deficit contributed to the growth of public and external debt for the two countries. Therefore, the study concluded that in Pakistan and India, public debt and external debt are sustainable but in a weak form. On the other hand, Campos, Marín and Badilla (2020), assessed the sustainability of the public debt of Costa Rica. This study used two techniques: the fiscal policy reaction function (FRF) and the calculation of the debt-stabilizing primary fiscal balance using the government’s intertemporal budget constraint. Using annual data from 1974 to 2018, the study showed that, in the short and long run, the public debt of Costa Rica has been unsustainable. Therefore, the study concludes that servicing debt in that country is unsustainable.
Campeanu, Stoian, and Roman (2006) examine fiscal sustainability in Romania based on the reaction function. The study used quarterly data from 1991 to 2005 using ordinary least squares. The study concluded that Romania’s fiscal policy was characterized by weak public debt sustainability. For European economies, Curtaşu (2011) examined fiscal sustainability as an aspect that remains a debatable topic for most economists. Employing estimation techniques such as unit-root tests, cointegration tests, and fiscal reaction function tests, the study made use of annual data spanning 1970 to 2012.
Swaziland, Nxumalo and Hlope (2018) assessed the fiscal sustainability of the country by employing the Hakkio and Rush technique along with the Trehan and Walsh technique. The study concluded that Swaziland was not on a sustainable fiscal path as its public expenditure was greater than its revenue, which continuously increased its budget deficits. The authors support the concluding remarks of Curtaşu (2011) by stipulating that solvency risks should be minimized by implementing fiscal policies to achieve sustainability.
Beqiraj, Forte, and Fedeli (2018) investigated public debt sustainability in OECD countries from the period 1991 to 2015 using panel data. The study shows that there is a long-run relationship between debt and structural primary balance. These results support the view that the long-term discretionary government response to raising the debt-GDP ratio is negative. It is postulated that governments do not consider the long-term effects that counteract the increases in debt, and hence, the intertemporal budget constraint is not satisfied.
Another group of countries studied by Mahmood, Arby, and Sherazi (2014) compared debt sustainability in SAARC countries (India, Sri Lanka, Pakistan, and Bangladesh) using the threshold debt ratios for assessing debt sustainability in those countries. The study results showed that the four countries have been experiencing fluctuations in the unsustainability of debt, mainly due to imbalances in fiscal and current accounts. Nega (2021) studied the effects of debt sustainability and economic growth in low-income sub-Saharan African countries. Employing the random and effect method for panel data spanning 2000 to 2017, the study found that high debt has a negative effect on economic growth. In a sense, instead of committing government revenue to more productive investments, it is used to service existing debt. The study concluded that low-income countries in SSA face unsustainable public debt.
Ganyaupfu (2014) evaluated the fiscal sustainability of South Africa for the period 1990–2013 using the VECM technique. The study results concluded that South African fiscal policy during the study period showed a sustainable path. However, new evidence by Redda (2020) revealed different conclusions by studying the sustainability of public debt and budget deficits in South Africa from 2000 to 2018. Employing the VECM, this study postulates that public debt in South Africa is unsustainable. The study elaborated that this was mainly due to insufficient tax revenue due to the narrow budget deficit. The study further indicates that South Africa needs to increase its tax base to service its debt, which can be achieved through job creation in the country, which will increase economic growth.
The above-mentioned literature provides different methods for analyzing the sustainability of public debt. However, the literature shows insufficient empirical studies focusing on the extent of debt sustainability in South Africa. Therefore, the purpose of this study is to provide new evidence of sustainability concerning South Africa’s public debt and further raise awareness of the potential loss of confidence by investors in South Africa, which may lead to continuous credit rating downgrades if debt is unsustainable.
Fiscal policy, primarily through government spending and taxation, serves as a key instrument for influencing aggregate demand. As Blanchard (2011) notes, the fiscal balance, or the current budget position, is essentially the difference between a government’s revenue and its expenditure.:
where Bt is the balance (the difference between spending and income) at time t. Gt is total spending and Tt is tax revenue (Mah et al., 2013). However, the public debt equation takes the following form.
The total accumulated debt is represented by Dt, and is derived from.
Dt=(1+r)Dt−1+Gt−Tt.
where r denotes the interest rate of debt. Public debt measures the total accumulated debt and all other related debt service costs (Mah et al., 2013; Mah, 2012). Bagni (2004) disagrees with the IBC and states that it is just an imposed constraint by creditors on the debtors.
Critiquing traditional IBC-based sustainability analysis, Bohn (2005) puts forward Model-Based Sustainability (MBS) as an alternative that accommodates uncertainty. This model assumes that optimizing creditors prevent long-term negative government debt and that financial markets are complete (Mah, 2012; Mosikari and Mah, 2024). Accordingly, the MBS criterion can be expressed as:
Bt=∑n=0∞Et(Ut,nPBt+n).
where. Ut,n is the pricing kernel for contingent claims, and PBt+n is the difference between revenue and expenditure, excluding interest expenditure. The MBS criterion differs from the IBC in its future surpluses, which depend on the distribution of primary surpluses across the state’s nature. This model was adopted in the present study.
The aim of this study is to assess the sustainability of government debt in South Africa using ARDL. According to Curtaşu (2011) and Kaur et al. (2018), fiscal sustainability occurs when cointegration is present between government revenue and government expenditure. Bohn (1998) and Ganyaupfu (2014) state that if there is a significant negative relationship between the primary balance-to-GDP ratio and debt-to-GDP ratio, then the fiscal policy is sustainable, or debt is sustainable.
The government expenditure and government revenue approach to sustainability tests for cointegration between government expenditure and government revenue (Kaur et al. 2018; Hakkio and Rush (1991). In this approach, if cointegration exists between government expenditure and government revenue, then the fiscal position or debt of that country is sustainable in the long run.
GRt=β0+β1GEt+β1DDEBTt+β1FDEBTt+et
Where GR stands for national government revenue as a percentage of GDP, GE is national government expenditure as a percentage of GDP, DDEBT denotes total domestic debt as a percentage of GDP, FDEBT stands for total foreign debt as a percentage of GDP, and t is the time period. The second approach is that of the budget deficit-to-GDP and debt-to-GDP ratios, as mentioned by Bohn (1998) and Ganyaupfu (2014). It states that if a negative relationship exists between the budget deficit-to-GDP ratio and debt-to-GDP ratio, then the fiscal policy or debt is sustainable.
PBt=β0+β1DDEBTt+β2FDEBTt+β3GE+β4GR+β4GDPGAPt+et
Where GDPGAP stand for GDP gap.
Quarterly data from 1997Q4 to 2023Q4 were collected from the South African Reserve Bank for the following variables: national government revenue as a percentage of GDP, national government expenditure as a percentage of GDP (GE), total loan debt of national government: total foreign debt as a percentage of GDP (FDEBT), total loan debt of national government: total domestic debt as a percentage of GDP (DDEBT), Primary Balance (PB), and GDPGAP. The percentage difference between actual GDP, Potential GDP, and potential GDP is generated using the Hodrick-Prescott filter.
This study uses the ARDL model to examine the relationships among the variables for cointegration and long-run and short-run relationships. The ARDL technique was used when the variables were stationary at I(0) and I(1). ARDL begins by determining the order of integration.
The Perron (PP) unit root test was used to test for stationarity. The unit-root test reveals the effects of shocks on the variables over time. Because the augmented Dickey–Fuller test has heterogeneity in the distribution of the disturbance term, this study used the PP test to avoid such issues in the unit root test (Asteriou & Hall, 2011). The Phillips-Perron tests (PP) correct the t-statistics of the coefficient in the autoregressive regression model (Asteriou & Hall, 2011). This is done as well at the intercept, intercept and trend, intercept, and none. The next step was the selection of an appropriate lag.
After the order of integration is determined, the next step is the bound cointegration. The ARDL bounds cointegration test determines the long-run relationship between variables. According to Pesaran et al. (1997), the ARDL Bounds test is advantageous because it can be applied regardless of whether the regressor is I(0) or I(1), thus eliminating the need for prior testing connected to standard cointegration analysis. Pesaran and Smith (2001) argued that the ARDL bounds cointegration test uses the F-statistic to determine whether cointegration exists. When the F-statistic falls within the critical values of I(0) and I(1), the results become inconclusive. However, when the F-statistic value is above the upper bound of I(1), cointegration occurs. However, when the F-statistic value is below the lower bound of I(0), there is no cointegration.
The next step is the error-correction model. According to Yunus et al. (2014), the Error Correction Model (ECM) determines how the change in a variable in the current term is related to the gap between its expression in the previous term. ECM simply provides the speed of adjustment from disequilibrium.
This was followed by the diagnostic and stability tests. According to Hendry (1980), diagnostic testing assists in determining the weaknesses and strengths of the model. This study will perform the following diagnostic tests: normality test, Autocorrelation LM test, heteroscedasticity test, and Ramsey Reset test. Mantalos (2010) suggested using the normality test to check whether the residuals were normally distributed. Serial correlation expresses the relationship between a variable at hand and a lagged rendition of the variable over different periods. Heteroscedasticity is defined as a circumstance in which the variance of the error term in a regression model changes significantly changing.
To assess fiscal sustainability in South Africa, we first present descriptive statistics, followed by the unit root test, the bound test for cointegration, the long-term and short-term results, and finally, the different diagnostic test outcomes.
Table 1 presents the descriptive statistics of the variables. The descriptive outcome demonstrates that the primary balance has a mean value of 0.050476, with a standard deviation of 3.162585, which is above the mean. The maximum and minimum values of the primary balance were − 10.90000 and 6.500000, respectively. This shows great dispersion in the primary balance during the study period. Furthermore, mean values of 37.36667, 4.476190, −3.173333, and − 0.000174 for domestic debt, with corresponding standard deviations of 12.79344, 1.702742, 3.493325, and 0.084005, are reported for domestic debt, foreign debt, deficit, and government expenditure gap, respectively. This generally points to the fact that during the study period, there were great variations in the different variables, either above or below their means.
The Phillip Perron unit root was employed to examine the order of integration of the different variables in this study. This will guide the adoption of the appropriate estimation techniques. The unit root test results presented in Table 2 reveal that the null hypothesis that the primary balance has a unit root is rejected at 1% for all three variants (intercept, intercept, and trend, without intercept and trend) of the Phillips Perron test results. This shows that the primary balance is stationary at level and, as such, follows an I(0) process. Further, the null hypothesis of domestic and foreign debt-containing unit roots is not rejected at level; however, at the first difference, the null hypothesis is rejected. This demonstrates that the variables were stationary at the first difference. Finally, the null hypothesis of government deficit and government expenditure gap containing unit roots is rejected at level; hence, these variables are stationary at level. Amidst the mixed level of stationarity, this study uses the times ARDL estimation technique that requires the existence of a mixed order of stationarity of variables, that is, I(0) and I(1).
According to Curtaşu (2011), Kaur et al. (2018), and Hakkio and Rush (1991), when cointegration is present between government revenue and government expenditure, the fiscal position or debt of that country is sustainable in the long run.
4.3.1 Correlation
Table 3 presents the results of the pairwise correlation matrices among the variables. The correlation matrices reveal a negative correlation between domestic debt and primary balance, foreign debt and primary balance, and equal government expenditure gap and primary balance. A positive correlation was established between government deficit and primary balance. This provides an a priori relationship that can exist between the independent variables and primary balance.
4.3.2 Lag length selection results
The lag-length results are listed in Table 4. The lag length of one is the chosen lag for national government revenue as a percentage of GDP and national government expenditure as a percentage of the GDP dataset. A lag of four was chosen for the estimation to assess the sustainability of debt.
4.3.3 Cointegration test
After the unit root test and the selection of the lag length, we examine the data for cointegration to establish the existence of the long-run relationship. In this regard, the bound test for cointegration is used, given that the ARDL approach is appropriate for the estimation technique. The result in Table 5 of the bound test reports an F-statistic of 6.05156, which lies above the upper limit of the 1% critical bound coefficient of 4.66. This implies that there is cointegration among the variables of the model, and hence, a long-run relationship.
In this approach, since cointegration exists between government expenditure and government revenue, the fiscal position or debt in South Africa is sustainable in the long run (Kaur et al., 2018; Hakkio & Rush, 1991).
4.3.4 ARDL results
Table 6 presents the results of the ARDL model. It is worth noting that because this study is interested in fiscal sustainability, we interpret the long-run results. Appendix 1 presents the short-run outcomes.
The long-term ARDL results show a positive relationship between government revenue and government expenditure in South Africa. This result is significant at the 1% level. The focus is on government expenditure and revenue variables. Both increase and decrease simultaneously. Debt sustainability assessment is about having a clear understanding of what might shape future debt dynamics; having a good grasp of fiscal policy behavior is key. Consequently, governments may either default or undertake extraordinary fiscal adjustments. With the interpretation of the different results, we proceeded to a diagnostic test to confirm if the estimated model was good.
4.3.5 The diagnostic test results
To examine the estimated ARDL model for stability, we used the Ramsey-Reset test. Diagnostic tests, such as autocorrelation, heteroscedasticity, and normal tests, were performed. The test results are listed in Table 7. The diagnostic test results show that, at the 5% significance level, the estimated model does not suffer from autocorrelation of residuals. There are no heteroskedasticity issues; the estimated model is normally distributed, and the model is stable. This indicates that the estimated model is efficient and suitable for statistical inference.
Bohn (1998) and Ganyaupfu (2014) state that if there is a significant negative relationship between the primary balance-to-GDP ratio and debt-to-GDP ratio, then the fiscal policy is sustainable, or debt is sustainable.
4.4.1 Correlation
Table 8 presents the results of the pairwise correlation matrices among the variables. The correlation matrices reveal a negative correlation between domestic debt and primary balance, foreign debt and primary balance, and equal government expenditure gap and primary balance. A positive correlation was established between government deficit and primary balance. This provides an a priori relationship that can exist between the independent variables and primary balance.
4.4.2 Lag length selection results
The lag-length results are listed in Table 4. The lag length of one is the chosen lag for national government revenue as a percentage of GDP and national government expenditure as a percentage of the GDP dataset. A lag of one is chosen for the estimation to assess the sustainability of debt.
4.4.3 Cointegration test
After the unit root test and the selection of the lag length, we examine the data for cointegration to establish the existence of the long-run relationship. The result in Table 9 of the bound test, which tests cointegration for the model under consideration, reports an F-statistic of 5.636747, which lies above the upper limit of the 1% critical bound coefficient of 4.15. This implies that there is cointegration among the variables of the model, and hence, a long-run relationship.
4.4.4 ARDL Long results
The results of the ARDL model are presented in Table 10.
The long-term ARDL results show that there is a positive and statistically significant relationship between primary balance and domestic debt, while there is a negative and insignificant relationship between primary balance and foreign debt. Domestic debt is positive and significant, indicating that it is not sustainable in South Africa. Foreign debt is sustainable but its relationship is insignificant. Because the relationship is insignificant, it means that the relationship cannot be considered. This result shows that foreign debt does not enhance fiscal sustainability. As mentioned by Bohn (1998) and Ganyaupfu (2014), if there is a negative relationship between the primary balance-to-GDP ratio and debt-to-GDP ratio, then fiscal policy or debt is sustainable.
The relationship between the primary balance and government expenditure is negative, while it is positive for government revenue. The relationship between the primary balance and the GDP gap is negative. All the results were statistically significant.
According to Debrun, Ostry, Willems, and Wyplosz (2019), if a country’s primary balance is unable to sustain the high interest payable to service debt, public debt is at the risk of being explosive. Consequently, governments may either default or undertake extraordinary fiscal adjustments. With the interpretation of the different results, we proceeded to a diagnostic test to confirm if the estimated model was good.
4.4.5 ARDL short results
The ARDL short-term results in Table 11 show that lags 1 to 3 of the primary balance, domestic debt, and lags 1 and 2, government expenditure, and government revenue lags 1 to 3 are all statistically significant in the short run. The GDP gap is insignificant, but lags 1 and 2 are statistically significant.
4.4.6 The stability test results
We further examined the estimated ARDL model for other diagnostic tests, such as autocorrelation, heteroscedasticity, and normal tests, while using Ramsey Reset for the stability test. The diagnostic test results show that, at a 5% significance level, the estimated model does not suffer from autocorrelation of residuals. The test results are presented in Table 12. There are no heteroskedasticity issues, and equally, the estimated model is normally distributed. This indicates that the estimated model is efficient and suitable for statistical inference.
The ADRL technique was used to assess debt sustainability. Quarterly data were employed for the South African economy from 1997Q4 to 2023Q4. The results show that there is a cointegration between government revenue and government expenditure, confirming fiscal sustainability in South Africa. Furthermore, empirical appraisal using the ARDL technique shows that domestic debt and primary balance are positive and statistically significant. Thus, domestic debt is becoming unsustainable in South Africa. Foreign debt is negative, which means it is sustainable but insignificant.
Based on this outcome, this study recommends that government domestic debt be reduced. This could be achieved through the implementation of initiatives to reduce government spending and boost income, such as cutting subsidies, revamping tax structures, and improving tax collection efficiency. Second, there is a need to strive for a sustainable domestic debt-to-GDP ratio through constant budgetary constraints.
The author(s) declared that no grants were involved in supporting this work.
© 2026 Mah G. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current Reviewer Status:
AWAITING PEER REVIEW
AWAITING PEER REVIEW
?
Key to Reviewer Statuses VIEW HIDE
ApprovedThe paper is scientifically sound in its current form and only minor, if any, improvements are suggested
Approved with reservations A number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.
Not approvedFundamental flaws in the paper seriously undermine the findings and conclusions