This study examines how economic stability, particularly inflation thresholds, affects export growth rates among 12 COMESA countries. We gathered annual panel data from 2000 to 2022, focusing on variables that are grounded in economic theories and empirical studies on international trade and macroeconomic stability, for instance [27,48–51]. The selection of variables was guided by their relevance to the research question and their role in explaining export performance under varying inflationary conditions, as well as their availability across COMESA countries. Missing data were handled using linear interpolation methods to ensure continuity without introducing bias. The number of interpolated gaps totaled nearly a dozen from over 275 observations.

Key macroeconomic factors influencing export growth include GDP, trade openness, and inflation, with export growth serving as the dependent variable. Inflation is treated as the threshold variable to capture its nonlinear effects on export growth. The choice of inflation as the threshold variable in analyzing the role of economic stability in boosting exports in COMESA, particularly within a Panel Threshold Regression (PTR) model, is theoretically sound and empirically supported by various economic principles [52,53]. Inflation, as a key indicator of macroeconomic stability, directly impacts several channels relevant to export performance.

The nexus between export growth and key macroeconomic variables such as GDP, inflation, and trade openness represent a crucial focus in the field of international economics. A number of studies have examined these relationships across various countries and timeframes [23,28–32,49,54,55]. These investigations provide valuable insights into how these macroeconomic factors interact and influence export performance globally. The data for these variables were sourced from reliable and reputable databases such as the World Bank’s World Development Indicators (WDI) and Our World in Data (OWID). These comprehensive and standardized datasets are widely recognized and facilitate cross-country comparisons over time, as supported by various studies utilizing similar datasets [56]. Details of the variables and sources used are provided in Table 1.

In this study, annual export growth rates serve as the dependent variable, reflecting the ability of COMESA countries to expand their trade activities over time. This metric directly measures the effectiveness of these nations in international trade. To evaluate the impact of economic factors, we include GDP as an indicator of economic size and overall productivity, which is essential for a country’s capacity to produce exportable goods. GDP provides insight into domestic production capabilities, crucial for understanding supply-side influences on exports. Trade openness is assessed through the ratio of total trade (exports plus imports) to GDP, illustrating a country’s integration into global markets. This variable reflects policy initiatives aimed at enhancing global economic participation, which can attract foreign investment and boost export performance. Inflation rates are included as both an explanatory and threshold variable to explore their nonlinear effects on export growth. Inflation plays a dual role by influencing competitiveness and economic stability. By treating inflation as a threshold variable, the study seeks to identify specific levels at which inflation begins to either hinder or enhance export performance. Table 2 lists the 12 COMESA member countries included in this study. These countries were chosen solely due to data availability.

The remaining COMESA member countries were excluded from the study due to a lack of available data (refer to S1 Table in the Supporting information section for detailed information).

The methodological framework for this investigation, as depicted in Fig 1, is systematically organized around a series of pivotal steps aimed at examining the repercussions of inflation on export expansion through the implementation of threshold regression analysis. First, the study defined/selected inflation (Infl) as the threshold variable, and the dataset is divided into two distinct regimes—low inflation and high inflation—based on a predetermined threshold value. Next, a grid search is conducted to determine the optimal threshold by evaluating various values of Infl and minimizing the residual sum of squares (RSS). Grid search is a recognized approach for assessing the robustness of threshold parameters [57]. Subsequent to this, the investigation of nonlinear relationships is conducted to ascertain the presence of threshold effects; this necessitates the evaluation of whether nonlinear models yield a better fit relative to linear models, employing techniques such as polynomial regression, particularly quadratic regression, to elucidate potential complexities [58,59]. The analysis subsequently advances to a Panel Data Threshold Regression (PTR), as delineated by [27], which is employed in this study to assess the interactions among inflation, GDP, and trade openness to understand their combined effects on export growth. Finally, the findings are represented visually by graphing export growth against inflation, incorporating fitted lines for both the low and high inflation regimes, thus facilitating a more lucid interpretation of the influence of inflation on export dynamics across varying economic contexts. This thorough approach guarantees a rigorous analysis of the threshold effects of inflation on exports within the COMESA region.

The research framework of this study is to investigate whether inflation acts as a threshold variable that moderates the relationship between export growth and key economic factors such as GDP and trade openness in COMESA countries. The Panel Threshold Regression (PTR) model, introduced by Hansen (1999), extends the linear panel data regression model to account for nonlinear threshold effects. Threshold effects analysis is an econometric technique that identifies critical values (thresholds) of an independent variable, at which the relationships between other variables change significantly [1,48,60]. The PTR model is an analytical tool used to investigate nonlinear relationships between variables, particularly when these relationships change as a result of another variable exceeding a specific threshold [61,62]. A nonlinear relationship is characterized by the phenomenon wherein the influence of one variable on another does not exhibit a consistent change; rather, the nature of this relationship may fluctuate across varying levels of the independent variable. For example, within the framework of export growth, the impact of GDP on export levels may escalate at an increasing rate after surpassing a certain GDP threshold, suggesting that as national economies experience growth, their ability to augment exports may become more evident, albeit not uniformly across differing GDP magnitudes.

The PTR approach is frequently applied in economic and financial research to identify regime-switching behaviors and structural breaks within data [63,64]. The PTR model is particularly well-suited for addressing heterogeneity; it enables the interaction between independent and dependent variables to differ across various regimes established by the threshold variable [65]. Therefore, to identify nonlinear relationships and estimate threshold effects of inflation on export growth, our study employed a PTR model. This nonlinearity arises because the relationship between the dependent variable and independent variables change depending on whether the threshold variable crosses a specific value.

It is essential to introduce the equation of linear relationships before transitioning to nonlinear relationships and estimating threshold effects in our study. To ensure consistency and comparability across countries, all variables were transformed into natural logarithms. This transformation helps to stabilize variance, reduce skewness, and interpret coefficients as elasticities in regression models. For this study, using Export Growth (EG) as the dependent variable and economic growth (GDP) and Trade Openness (TO) as independent variables, the linear panel data regression model can be expressed as:

Where, Ln stands for the natural logarithm, EGit refers to export growth for country i at time t, GDPit represents economic growth for country i at time t, and TOit implies trade openness for country i at time t. Here α is the intercept term, while β₁ and γ₁ are coefficients representing the effect of each independent variable on export growth, and εit is the error term capturing unobserved factors affecting export growth. This equation assumes a linear relationship between export growth and its determinants across all countries and time periods in the dataset, with no threshold effect or nonlinearity in the relationship between these variables.

Unlike linear models where coefficients remain constant across all observations, PTR models allow for different coefficients in different regimes. The PTR model is widely used in economic studies to capture regime changes or structural breaks caused by specific threshold levels. Taking inflation (LnInfl) as the threshold variable, we can write the PTR model explicitly as:

Here, Inflit denotes inflation (a threshold variable) for country i at time t, (c > 0: Threshold value that splits the data into two regimes. This equation shows that when inflation (LnInfl) is below or equal to a certain threshold level (c), one set of coefficients applies (α₁, β₁, γ₁). When inflation exceeds this threshold level, another set of coefficients applies (α₂, β₂, γ₂).

The PTR model allows for nonlinear relationships by introducing threshold effects. The decision to use PTR instead of linear regression ones in this study is based on the nature of the economic phenomena being analyzed. Export growth, inflation, GDP, and trade openness often exhibit complex interactions that cannot be adequately captured by simple linear models. Employing PTR allows the study to: (i) Capture threshold effects where key variables like inflation influence export growth differently across distinct regimes. (ii) Provide more accurate empirical evidence consistent with economic theory and real-world observations. (iii) Improve model fit and predictive power by accounting for heterogeneity among member states. (iv) Offer nuanced policy recommendations tailored to specific conditions within COMESA countries.

In economic studies, relationships between variables are often assumed to be linear for simplicity. However, testing for nonlinear relationship is crucial because it allows researchers to capture the complex interactions among variables and better understand the true nature of the relationships. For instance, in this study, export growth may not respond uniformly to changes in inflation, GDP, or trade openness. Threshold effects, where the impact of one variable changes depending on the level of another variable, are a common form of nonlinearity.

By conducting nonlinear relationship testing, such as fitting quadratic regression models or incorporating interaction terms with threshold variables (inflation regimes), we can identify whether these nonlinearities exist and how they influence the dependent variable. This approach ensures that policy recommendations derived from the analysis are more accurate and reflective of real-world dynamics. Without testing for nonlinearity, important insights could be missed, leading to biased or oversimplified conclusions and potentially ineffective or counterproductive policy measures. The equations for nonlinear relationship testing used in the study can be expressed as:

Where, the additional term (LnInfl)2 captures the nonlinearity in the relationship between inflation and export growth.

The interaction model incorporates a dummy variable for inflation regimes (LowRegime = 1, HighRegime = 0) and interaction terms between explanatory variables and the regime. The equation is:

Where. LowRegime represents a dummy variable for inflation regimes, β₀: Intercept, β₁: Effect of GDP on export growth, β₂: Effect of being in a low-inflation regime, β₃: Effect of trade openness on export growth, β₄: Direct effect of inflation, though not statistically significant, β₅: Interaction effect between GDP and low-inflation regime, and β₆: Interaction effect between trade openness and low-inflation regime.

Furthermore, to identify whether the effect of one independent variable on the dependent variable changes depending on the level of another independent variable and to improve the explanatory power of the model by accounting for complex interdependencies among variables that would otherwise be overlooked in a simple additive model, this study employed interaction terms in a regression model. Testing interaction terms in a regression model is crucial for understanding how two or more variables jointly influence the dependent variable. In this case, testing the interaction terms between inflation and other variables (GDP and trade openness) provides insights into how inflation moderates or amplifies their effects on export growth. It provides a deeper understanding of how multiple factors work together to influence export growth, enabling more effective policymaking and economic planning. The regression model with interaction terms can be expressed as follows:

Where, β₀,...,β₅: Coefficients estimated by the regression, (LnInfl⋅LnGDP) = Infl_GDP: Interaction term between inflation and GDP, and (LnInfl⋅LnTO) = Infl_TO: Interaction term between inflation and trade openness.

This study investigates the impact of economic stability on export growth within the COMESA region, specifically examining threshold effects using inflation as the threshold variable. By splitting the data into low inflation and high inflation regimes based on the median inflation value, we gain insights into the economic dynamics across these two regimes from 2000 to 2022. The summary statistics, presented in the Table 3, outline key metrics including export growth (LnEG), GDP (LnGDP), trade openness (LnTO), and inflation (LnInfl) for both regimes.

The result provides insights into the economic dynamics of COMESA countries under low- and high-inflation conditions. In terms of export growth, both inflation regimes presented almost identical mean values, around 4.625 to 4.626. However, the low inflation regime exhibited a wider range, with a maximum export growth of 4.985 compared to 5.215 in the high inflation group. This suggests that economies with lower inflation may experience more stable trading conditions, potentially leading to higher export growth and resilience against economic shocks [36]. Elevated inflation rates can result in the depreciation of a currency, potentially providing a short-term increase in exports. However, this scenario often generates economic uncertainty, which can deter investment [39]. On the other hand, low and stable inflation is typically viewed as beneficial for the economy, fostering an environment conducive to stability, investment, and growth. Nonetheless, maintaining inflation at excessively low levels is also critical, as it can pose its own set of economic challenges and risks [66].

Economic growth showed notable differences between the two regimes, particularly in the lower bounds. The minimum LnGDP growth was 1.696 in the low inflation group, contrasting sharply with the high inflation group’s minimum of 1.097. This indicates that countries in low inflation environments tend to have a stronger baseline for economic growth, while those facing higher inflation may encounter challenges that hinder robust economic performance. Notably, the mean LnGDP growth values were slightly higher in the high inflation regime, at approximately 3.126, compared to 3.163 in the low inflation context. This indicates the possibility that high inflation regions can still generate significant growth, albeit with greater risk.

When examining trade openness, the data revealed that the average openness was significantly higher in the low inflation regime at 4.173 versus 3.845 in the high inflation regime. This implies that countries maintaining low inflation may enjoy more favorable conditions for trade. Furthermore, although both regimes exhibited high maximum values for trade openness, the lower bound in the high inflation grouping (0.693) suggests that some countries are considerably less open to trade due to the economic pressures stemming from inflation. The inflation variable clearly distinguished the two groups, with the low inflation regime reporting a mean of 2.225, while the high inflation regime averaged 2.986. The broader range of inflation values in the high inflation context, where the minimum was 2.565 and the maximum reached 5.491, highlights the economic fragility encountered by these countries.

The analysis indicates that countries experiencing high inflation should emphasize policies aimed at stabilizing inflation, given its negative correlation with GDP growth and trade openness [36,66]. High inflation can erode competitiveness by increasing production costs and reducing profit margins for exporters. Conversely, maintaining lower inflation appears beneficial for sustaining higher export growth and economic resilience [66]. Therefore, further investigation into sector-specific performances, alongside comprehensive regression analyses, could yield valuable insights, particularly regarding the relationship between inflation and economic outcomes. This approach may assist policymakers in implementing strategies that promote economic stability and growth, tailored to the unique challenges posed by varying inflation levels.

The average export value in the COMESA region is approximately 4.63, which aligns with its median, suggesting a fairly balanced distribution. When analyzing individual countries export performance, significant variability exists, with export values ranging from a minimum of 0.49 to a maximum of 5.22 (see Fig 2). Most countries show relatively stable export growth with only minor fluctuations, although there was a notable decline around 2011, likely due to economic or external factors impacting exports during that time. Kenya stands out with the most consistent export growth among the analyzed countries, marked by a low standard deviation of 0.072. This stability highlights the crucial role of trade openness and inflation rates in shaping export performance. Political stability and governance, human capital, trade agreements and policies, and infrastructure development create a resilient and adaptive export environment that enables consistent growth.

Economic relationships are frequently nonlinear, meaning that the impact of one variable on another can change once a certain level or “threshold” is crossed [67,68]. The threshold effect analysis seeks to reveal nonlinear patterns that traditional linear regression approaches might miss, thereby providing a more nuanced understanding of the relationship between inflation and export growth. This analysis is a crucial econometric tool aimed at examining how the relationship between variables shifts when a specific threshold value of another variable is crossed [27]. In the realm of export growth in COMESA, scholars implies that the influence of trade openness and economic growth on export performance exhibit variability contingent upon heterogynous levels of inflation [69]. A low, stable inflation environment might foster export growth by providing a predictable economic climate, encouraging investment, and reducing transaction costs [70,71]. Nevertheless, if inflation surpasses an optimal threshold, it may jeopardize economic stability, escalate production expenditures, erode purchasing power, disrupt supply chain, and subsequently impede export performance and competitiveness [70,72–74]. By determining an optimal threshold value for inflation, this study seeks to offer actionable insights into how economic stability shapes export performance among COMESA member countries [70,71]. Table 4 summarizes the results that reflects the key variables and their coefficients under different regimes.

The results of this analysis reveal an optimal threshold value for inflation, identified at 3.662%, and a corresponding minimum Residual Sum of Squares (RSS) of 28.552. These parameters are essential for evaluating the model’s goodness of fit. Moreover, our findings align closely with those of [55], who conducted an extensive study on inflation thresholds in the African context. Ndoricimpa’s research suggests that an inflation rate exceeding 5.43% negatively impacts economic growth, while maintaining inflation rates below this threshold does not exhibit detrimental effects. This correlation emphasizes the significance of identifying accurate inflation thresholds to inform economic policy and promote sustainable growth.

The comprehensive investigation conducted within the parameters of this study has ascertained that the optimal threshold inflation rate for COMESA during the extensive timeframe spanning from the year 2000–2022 has been identified as precisely 3.662%, a figure that holds significant implications for economic policy within the region. This particular inflation rate corresponds remarkably closely with the lower boundary of the inflation targets that are conventionally established in both emerging and developing economies, which typically fluctuate within a range of approximately 3% to 7%, thereby indicating a broader context for economic performance [75–77]. Furthermore, this identified inflation rate also occupies a position at the upper limit of the more conservative inflation range of 2% to 3% that is commonly targeted by central banking authorities in various developed economies, which collectively suggests that this figure represents a judicious and realistically attainable goal for the maintenance of economic stability throughout the COMESA region [75–77].

Fig 3 illustrates how RSS changes across different potential threshold values of inflation (grid search) [57], with a clear minimum point at a threshold value of 3.662, indicating a structural break in the relationship between export growth and its predictors, suggesting that splitting the data into two regimes, low inflation (LnInfl≤3.662) and high inflation (LnInfl>3.662), at this point provides the best model fit by reducing prediction errors.

Generally, the threshold effects analysis reveals that inflation serves as a significant determinant in understanding the interplay between GDP, trade openness, and export growth in the COMESA region. The divergent effects in low- and high-inflation regimes underline the importance of maintaining economic stability to foster an environment where both GDP growth and trade openness can effectively contribute to boosting export performance.

Scholars agree that testing for nonlinear relationships is imperative, given several economic theories posit that the effects of policies or economic conditions are not invariably consistent [69,78]. One of the key objectives of this analysis is to comprehend how sensitivity to economic conditions varies across different inflation levels. Through the integration of nonlinear specifications, this study seeks to enhance economic models, thereby providing a more robust framework that accurately encapsulates the complexities of these interrelations.

In our exploration of the relationship between export growth and its predictors (GDP, trade openness, and inflation), we sought to determine whether there is a nonlinear connection based on panel data during the study period (see Table 5). This inquiry highlights that the interactions among these factors may change under varying economic circumstances, particularly influenced by inflation rates. First, we compared the fit of a linear regression model against a quadratic regression model to see if including polynomial regression for inflation capture potential nonlinearities in the relationship and provided a better explanation of export growth [58].

The results from the ANOVA indicated that the quadratic model (Model 2) significantly outperform (Model 1), the traditional linear model (lm), obtaining an F-statistic of 0.011 with a p-value of 0.013. These results suggest that there is curvature in the relationship between inflation and export growth, supporting the hypothesis of nonlinearity [58]. The presence of nonlinear relationships suggests that the effects of GDP and trade openness on export growth might not be uniform; for example, a certain level of inflation could lead to disproportionately larger increases in exports as economic growth rises.

By utilizing a PTR approach, we divided the data based on inflation levels, specifically focusing on a threshold defined by the median inflation rate. This division allowed us to analyze the impacts of GDP and trade openness in two distinct inflation regimes: low inflation and high inflation. The results of the PTR analysis for the low- and high-inflation regime, presented in Table 6, reveal significant relationships between export growth, GDP, and trade openness.

In the low-inflation regime, which includes observations where inflation is below the median threshold, our findings revealed that LnGDP has a strong positive effect on export growth, with a coefficient of 0.348 and a highly significant p-value (1.802e-08). This suggests that for every 1% increase in LnGDP, export growth experiences an approximate increase of 0.348%. Scholars like [79] provides evidence that supports our findings. He conducted a study on over 40 African countries from 1960 to 2018, using various econometric methods to investigate the relationship between export growth and GDP growth. The study found a positive, bidirectional relationship between exports and economic growth. This significant relationship underscores the importance of economic growth in fostering exports during periods of low inflation [79,80]. However, while trade openness also has a positive coefficient of 0.091, it is not statistically significant (p = 0.159). This indicates that although greater trade openness could contribute to higher exports, its effect is less pronounced in a low-inflation context. Furthermore, the insignificance of trade openness may reflect persistent structural challenges faced by COMESA economies, such as non-tariff barriers, inadequate infrastructure for cross-border trade, lack of logistics networks, regulatory bottlenecks, limited diversification of traded goods, and insufficient regional integration frameworks within COMESA. These issues may persist even in the presence of open trade policies, potentially leading to increased transaction costs, currency depreciation, and diminished competitiveness in global markets driven by price instability.

When we examined the high-inflation regime, where inflation exceeds the median threshold, the dynamics shifted substantially. Here, LnGDP remained significant, with an even larger positive coefficient of 0.779 (p = 2.538e-07), indicating that the impact of LnGDP on export growth is even stronger in times of high inflation. Specifically, a 1% increase in LnGDP corresponds to an impressive 0.779% increase in export growth. On the other hand, trade openness in this regime displayed a negative coefficient of −0.127, which is statistically significant (p = 0.046). This result is particularly striking as it suggests that during high-inflation periods, increased trade openness may actually hinder export growth. Our findings contradict the results of [39,81]. The results also highlight that the relationship between trade openness and export growth is not always positive. Some studies suggest the export-led growth hypothesis does not hold for all countries [11,82]. Other studies also confirm that the impact of trade openness can vary based on a country’s level of development and openness [36,83]. This unexpected finding could point to potential structural inefficiencies within COMESA economies that manifest during inflationary episodes. In terms of model performance, the F-statistic of both regimes and their associated p-value of < 0.001 confirm that the model as a whole is statistically significant, confirming their overall reliability.

To further illustrate how inflation interacts with export growth across different regimes, Fig 4 depicts the relationship between export growth and inflation, showcasing fitted lines from both models. By visually representing these dynamics, we can better grasp the nuanced effects of inflation on export growth in the context of COMESA economies.

Overall, our findings highlight that in the low inflation regime, maintaining economic stability through controlled inflation fosters an environment where economic growth significantly enhances export performance within COMESA countries during the study period. However, trade openness does not show a direct and significant impact on exports under these conditions, likely due to structural challenges or inefficiencies in policy implementation. In contrast, during high inflation periods, while economic expansion continues to support exports, increased trade openness may expose economies to risks that negatively affect export performance. To address these issues, policymakers should prioritize macroeconomic stability by implementing effective monetary and fiscal policies to control inflation. Additionally, investments in productivity-enhancing measures such as infrastructure development and technological advancements are essential to reduce production costs and improve competitiveness in global markets.

In this section we investigated how inflation interacts with other key economic factors, namely GDP and trade openness, in affecting export growth within the COMESA region from 2000 to 2022. We included interaction terms in our regression model to capture these relationships more accurately, see Table 7. Our findings revealed several important insights. First, the model’s intercept, estimated at 4.794, indicates a strong baseline level of export growth when all other factors are held constant. This is statistically significant, reflecting meaningful export levels in the absence of other influences. When looking specifically at GDP, we found a coefficient of −0.203, suggesting a negative relationship with export growth; however, this result was not statistically significant, indicating that LnGDP alone might not consistently affect export performance in our model. In terms of trade openness, we observed a positive coefficient of 0.117, but, again, this was not statistically significant, indicating that while trade openness may positively influence exports, its effect is not strong enough to be conclusive. On the other hand, inflation had a notable negative effect with a coefficient of −0.722, which was statistically significant (p = 0.032). This suggests that rising inflation rates tend to decrease export growth, pointing to the detrimental impact of inflation on economic stability and competitiveness.

A particularly interesting finding came from the interaction term between inflation and GDP (LnInfl_GDP), which showed a significant positive coefficient of 0.284 (p = 0.009). This result suggests that as GDP increases, the negative impact of inflation on export growth is decreased. Essentially, economic growth can help counterbalance some of the adverse effects of inflation on exports, highlighting the importance of fostering growth to stabilize trade outcomes. In the case of the interaction between inflation and trade openness (LnInfl_TO), the coefficient was −0.042. While this result was not statistically significant, it indicates a potential trend where the positive effects of trade openness on export performance might be diminished in high-inflation environments. This points to a complex interaction where higher LnGDP can potentially counterbalance the adverse effects of inflation on exports, evidenced by a statistically significant positive interaction term between inflation and LnGDP, while the interaction with trade openness remains less conclusive [2,84].

Economic growth has the potential to alleviate some of the detrimental effects of inflation on exports by enhancing productivity and bolstering international competitiveness [85]. Inflation can diminish a nation’s export competitiveness by elevating production costs and the prices of goods destined for international markets [86]. Nonetheless, economic growth can effectively counteract these adverse effects through several mechanisms.

First, economic growth typically facilitates advancements in technology and infrastructure, which can reduce production costs [87]. This increase in efficiency can mitigate the inflationary pressures on export prices, thereby rendering them more competitive on the global stage. Second, economic growth can lead to currency appreciation [88]. A stronger domestic currency enhances the attractiveness of exports for foreign purchasers by improving their purchasing power [89]. This dynamic can contribute to the maintenance or potential increase in export volumes even amidst inflationary challenges. Finally, economic growth can invigorate foreign demand for a country’s goods and services [90]. As economies worldwide expand, the demand for imports tends to rise, presenting exporting nations with greater opportunities to boost their sales [49].

The overall model fit is strong, as indicated by an F-statistic of 13.72 with a highly significant p-value (<6e-12), suggesting that the model explains a substantial portion of variation in export growth across COMESA countries during the study period. Generally, these results contribute to a nuanced understanding of how economic stability, particularly through inflation and GDP dynamics, influences export growth in the COMESA region. They emphasize the need for strategies focused on stabilizing inflation to create a more favorable environment for exports. Additionally, policymakers should prioritize boosting GDP, as economic growth can help mitigate some of the challenges posed by inflation.