1 Introduction

More recently, climate change has emerged as a persistent global issue with far-reaching effects including food insecurity especially in emerging economies. The Southern African region has been among the most impacted region by climate change particularly the agricultural value chain. In this regard, climate change has become one of the most pressing issues affecting attainment of sustainable agriculture and food security [1]. One area where it is particularly concerning is in emerging economies, where agriculture plays a significant role in their economy and food security [2]. Climate change has led to unpredictable weather patterns, droughts, floods, and extreme temperatures that have affected crop yields and livestock production [3]. According to Mahofa [4], climate change impacts on agricultural trade and food security are significant in emerging economies. The changing climatic patterns have led to substantial declines in crop yields, affecting availability and affordability of food thereby resulting in increased food prices in most emerging economies [5]. Addressing climate change has become crucial for ensuring sustainable agricultural trade and food security in emerging economies [5]. Hence, there have been collaborative efforts among governments and private sector players to climate change [4].

The impacts of climate change on agricultural trade and food security have also been largely felt in the Southern African region [6]. The region has been vulnerable to recurring climatic shocks such that the Intergovernmental Panel on Climate Change (IPCC) labelled it as a climate change “hotspot” [7]. According to the IPCC [8], by the year 2050, average temperatures in Southern Africa are expected to increase by about 1.6 ℃ and average precipitation is expected to decline by 10%. This trend has been a regional issue even as it has also been noted that if no actions are made, most Southern African countries will be net-importers of food by 2050 [8]. This climate change trend will cause significant decline in agricultural productivity and trade as changes in agricultural productivity affect agricultural trade through changes in agricultural comparative advantage [9, 10].

In Southern Africa, climate change has caused significant decline in agricultural production. According to FAO [5], due to global warming, cereal production in Southern Africa has significantly declined for the past decade and is projected to further decline by over 20% by 2030. This decline in agricultural production in Southern Africa has been linked to food insecurity and increased food import bills threatening the fiscus of most nations in the region [11]. For the past two decades, Southern Africa has witnessed spiralling food import bills which increased from US$35 billion to US$43 billion between 2019 and 2022 [5]. According to the African Development Bank’s [12] projections, food imports for Southern Africa will reach US$90 billion by 2025. In support, the International Monetary Fund (IMF) [13] reported that food prices in Southern Africa surged by an average of 23.9% between 2020 and 2022 which is the highest since the 2007/8 global financial crisis. This increasing food import bill for Southern Africa has attracted mounting attention and has been considered a worrisome trend for a region that was once an agricultural export powerhouse but now food import dependant [14]. This has contributed to agricultural trade deficits facing the region over the past decade [5].

Climate change has been cited as one of the driving factors for agricultural trade deficits and food insecurities [13]. To revise these trends, Southern African countries need to fight climate change so as to improve agricultural productivity and boost agricultural exports [14]. Previous empirical studies have cited negative effects of climate change on agricultural trade flows [10, 15,16,17]. However, compared to the available literature, there are few empirical studies on climate change impacts on agricultural trade and food security. Precisely, there is lack of empirical evidence in Southern Africa which is considered the climate change “hotspot” and net food importer. The findings of previous studies done in other regions may not therefore be applicable to Southern Africa’s context as Mahofa [4] argue that climate changes vary by location such that climate change impacts on agricultural productivity and trade also vary. Hence, there are significant gaps in existing literature which this research intents to fill.

This study reports the impacts of climate change on agricultural trade in Southern Africa. Understanding climate change impacts on agricultural trade and food security in Southern African region through empirical analysis is of importance for several reasons. One of the major reasons is that the results will aid in effective policy-making towards addressing climate change and improved agricultural trade flows. According to Nhundu et al. [18], employing agricultural policies without empirical understanding result in inappropriate policies. More so, agricultural trade flows are important in promoting and strengthening resilience of agricultural supply chains from climate change shocks [19, 20]. Hence, there is really need for this research to understand the climate change impacts on agricultural trade and food security in the context of Southern Africa. In doing so, this research will make significant contributions to theory, policy and practice. In line with the research aim, the following hypotheses were developed:

H1: Climate change has significant negative impacts on agricultural trade.

H2: Climate change has significant negative impacts on food security.

2 Material and methods

The research is a secondary based using panel data for the period 2012 to 2021. The time period was selected based on availability of data. Secondary annual panel data for the Southern African countries were collected and the sample was based on data availability. Of the 15 Southern African countries, 12 countries (Angola, Eswatini, Botswana, Lesotho, Mauritius, Mozambique, Malawi, Namibia, Tanzania, Zambia, South Africa and Zimbabwe) were included in the sample whilst three countries (Comoros, Democratic Republic of Congo and Seychelles) were excluded due to data unavailability.

Data was obtained from FAOSTAT and World Bank databases (Table 1). Data collected was analysed using the dynamic panel regression analysis making use of the Generalised Method of Moments (GMM) technique. The dynamic panel GMM regression model was estimated to address endogeneity and homogeneity which lead to unbiased estimates according to Hsiao [21]. The other motivation for employing the GMM technique is that Ullah, Akhtar and Zaefarian [22] cited that the most popular econometric approach for estimating dynamic panel regression models is the GMM technique that banks on lagged variables as instruments. The research aimed to estimate the following dynamic panel econometric models developed basing on specifications of Tekce and Deniz [15], Mahrous [23] and Affoh et al. [24]. The logarithmic function was employed so that the estimates can be interpreted as elasticities and to reduce problems of non-normality [21].

$$ln{AGTRADE}_{it}={\beta }_{0}+{\beta }_{1}ln{AGTRADE}_{it-1}+ {\beta }_{2}{lnTEMP}_{it-1}+ {\beta }_{3}{lnPREC}_{ti-1}+ {\beta }_{4}{AGDP}_{it}+ {\beta }_{5}{FI}_{it}{+ \beta }_{6}{lnPG}_{it}+ {\uplambda }_{t}+{\upmu }_{t}$$
(1)
$$ln{FPI}_{it}={\alpha }_{0}+ {\mathrm{\alpha }}_{1}ln{FPI}_{it-1}+ {\mathrm{\alpha }}_{2}{lnAGTRADE}_{it}+ {\mathrm{\alpha }}_{3}{lnTEMP}_{ti-1}+ {\mathrm{\alpha }}_{4}{lnPREC}_{ti-1}+{\mathrm{\alpha }}_{5}{FI}_{ti}+ {\mathrm{\alpha }}_{6}{lnPG}_{it}+ {\uplambda }_{t}+{\upmu }_{t}$$
(2)

where; ln- natural logarithm; AGTRADE- agricultural trade balance as a percent of gross domestic product (GDP); \({\beta }_{0}\) and \({\mathrm{\alpha }}_{0}\)—regression coefficients; \(({\beta }_{1}-{\beta }_{6})\) and \(({\alpha }_{1}-{\alpha }_{6})\)—regression-coefficients; TEMP annual mean temperatures (°C) as proxy for climate change; PREC- annual mean precipitation as proxy for climate change, FI food inflation measured by food consumer price index (CPI), AGDP agricultural GDP growth rate (%); FPI food production index as a proxy for food security, PG population growth rate (%); \({\upmu }_{t}-\) error term,\({\uplambda }_{t}\)country fixed effects and \({\updelta }_{t}-\) time fixed effects. The stated models are estimated using Eviews version 13 software. Inclusion of food inflation, population growth and agricultural GDP as control variables was justified following the recommendations of Adesete et al. [25] and Affoh et al. [24].

Table 1 Description of variables
Full size table

The panel GMM regression model was estimated to address endogeneity and homogeneity which lead to unbiased estimates [21]. Pre-and post-estimation tests such as unit root, Hausman, normality, multicollinearity and autocorrelation were carried out. Panel unit root tests were carried out using the Levin, Lin and Chu (LLC) test. Normality of residuals was tested using the Jarque–Bera (JB) test. More so, the pair-wise correlation matrix was employed to check for the problem of multicollinearity. Autocorrelation was checked using the Durbin-Watson (DW) statistic. Lastly, the Hausman test was done to check for the most appropriate panel model between the fixed-effects (FE) and random effects (RE) models.

3 Results and discussion

This section presents and discusses empirical results of the study. Table 2 presents the summary of the descriptive statistics for the variables as per each country over the ten-year period (2012–2021).

Table 2 Descriptive statistics (2012–2021)
Full size table

From Table 2, AGTRADE had mean statistics for the 12 countries ranging from 0.83 to 7.26 showing that for the period 2012 to 2021, agriculture trade balance in the Southern African region averaged between approximately 0.8% and 7.3% of GDP. More so, for TEMP, the mean statistics ranged from of 12.26 to 24.53 showing that on average mean annual temperatures in the Southern African region for the 10 years averaged between 12.26 and 24.33 ℃. The mean statistics for PREC also imply that mean annual precipitation in Southern Africa was approximately between 0.01 and 3061.67 mm. The mean statistics for AGDP show that on average, the value of agricultural GDP has been between − 21.18% and 5.86% of national GDP in Southern Africa. Additionally, the mean of 15.02 for food inflation (FI) indicate food inflation in the Southern African region averaged between − 3.34% and 601.02% during the same period. Lastly, the mean statistics for PG indicate the on average, the population in Southern African countries averaged approximately between 0.01% and 3.76% on an annual basis. Table 3 shows the results of the panel unit root tests.

Table 3 Panel unit root test results
Full size table

The results presented in Table 3 show that three variables (AGDP, FI and PG) were found to contain no unit root tests at level whilst the remaining four variables (AGTRADE, TEMP, PREC and FPI) were found to unit root tests at level. However, the series for the four variables became stationary after first differencing. In this regard, robust and unbiased estimates were obtained. Table 4 presents the collinearity matrix.

Table 4 Collinearity matrix
Full size table

From Table 4, the correlations for paired independent variables were significantly less than 0.8 implying that there were no serious problems of collinearity among the independent variables. According to Gujarati [26], the rule of thumb is for pair-wise correlations between two regressors is 0.8 where correlation coefficients in excess of 0.8 imply high multicollinearity. More so, baseline panel GMM regression estimations were done. From these, the Hausman test was undertaken, and the results are presented in Table 5.

Table 5 Hausman test results
Full size table

From Table 5, for model (1), a Hausman statistic of 0.001 was estimated with a p-value of 1.00 (P > 0.05) indicating that the RE GMM model was the most appropriate. On the other hand, for model (2), a Hausman statistic of 337.32 with a p-value of 0.000 (P < 0.001) implying that FE GMM model was the most appropriate. Besides, the JB test was undertaken to test for normality of residuals and the results are presented in Table 6.

Table 6 Jarque–Bera normality test
Full size table

The results shown in Table 6 show that for model (1) (RE model) and model (2) (FE model), the Jarque–Bera statistics were 1.30 (P = 0.52 > 0.05) and 0.89 (P = 0.64 > 0.05). These show that the residuals followed a normal distribution. More so, the Kurtosis and Skewness statistics are close to the values three and zero respectively showing normal distribution [21]. The results of the robust panel GMM regressions are presented in Table 7. For each of the two models, three models (pooled ordinal least square, RE GMM and FE GMM models) were estimated. However, basing on the results of the Hausman tests, results for the RE model and FE model were interpreted for models (1) and (2) respectively. As indicated in Table 7, model (1) and model (2) had coefficients of determination (R-squared) of 0.86 and 0.52 respectively. These results show goodness of fit. More so, models (1) and (2) estimated DW statistics of 2.31 and 1.66 respectively (Table 7). These DW statistics fell in the range 1.5 to 2.5 showing that the models did not suffer from autocorrelation [21].

Table 7 Results of robust panel GMM regression model) (Models 1–2)
Full size table

As shown in Table 7, the RE GMM model (1) estimated the impacts of climate change on agricultural trade whilst the FE GMM model (2) estimated the impacts of climate change on food security in the Southern African region. For the RE model (1), the one-period lagged variables for the climate change variables (TEMP and PREC) were found to have statistically significant negative coefficients. For TEMP, the coefficient of − 0.25 (P ≤ 0.01) show that a percent increase in mean temperatures in 1 year can result in about a 0.25% decline in agricultural trade flows in the succeeding season. This is because, increase in temperatures may significantly reduce agricultural productivity leading to low output supplied to the market. On the other hand, for PREC, the coefficient of − 0.06 (P < 0.01) show that changes in precipitation negatively impacts agricultural trade. This suggests that a percentage change in the amount of precipitation can cause agricultural trade to fall be approximately 0.06%. These results show that climate change has significant impacts on agricultural trade in the Southern African region. Basing on these results, the hypothesis that climate change has significant negative impacts on agricultural trade is not being rejected but upheld. In other words, the results have proven that climate change is threating agricultural trade flows in the region such that the region may continue to be a net-food importer. The results corroborate those of Tekce and Deniz [15], Khan et al. [17], Adesete et al. [25], Fusco [27], Baptista et al. [11], Affoh et al. [24], Brenton et al. [2] and Dumortier et al. [1] who in their respective studies also found negative impacts of climate change on agricultural trade.

Besides the negative climate change effects, other factors negatively impacting agricultural trade in Southern Africa have been found to include food prices proxied by food inflation (FI) (β = − 0.001; P < 0.001) and PG (β = − 0.04; P < 0.001). These have further implications in food security. Similar findings were also reported by Mahrous [23], Wiebe et al. [3], Smith and Glauber [20], Adesete et al. [25] and Affoh et al. [24].

More so, from the FE GMM model (2), temperature (TEMP) was found to have insignificant effects on food security in the Southern African region (α = 0.09; P > 0.05). This could be a pointer of adoption of adaptation strategies to temperature changes in the Southern African region. On the other side, changes in precipitation (PREC) have significant positive effects on food security (α = 0.17; P < 0.05) using model (2). These results show that a percentage increase in precipitation may cause an increase in agricultural trade flows by approximately 0.17%. The results show inconclusive effects of climate change on food security but suggests the possibility of upholding the research hypothesis that climate change negatively impacts food security. The findings of this study align towards that of Mahrous [23] and Brenton et al. [2] who found significant negative impacts of climate change on food security.

4 Conclusion

The results of the random effects GMM model revealed significant negative climate change impacts on agricultural trade. However, the fixed effects GMM model provided inconclusive results regarding the impacts of climate change on Southern Africa’s food security. Besides, the results have significant implications for policy, theory and practice. The negative climate change impacts on agricultural trade may aid policymakers in formulating effective policies to promote agricultural trade such as the Southern Africa Development Committee (SADC) Regional Agricultural Policy and trade facilitation policies. Also, the study will inform policies to mitigate climate change in the Southern African region such as the SADC Climate Change Policy. Furthermore, the study will inform policies and strategies for promoting food security including the SADC Food and Nutrition Security Strategy (2015–2025). Precisely, the findings of the study will provide insights to policymakers towards formulating policies to ensure sustainable food security through climate change mitigation and adaption and agricultural trade facilitation.

The research has also made significant contributions to the existing climate change literature. Basing on the results, there is need for concerted efforts towards climate change adaptation and mitigation for sustainable agriculture. However, the research was not exhaustive. The use of only precipitation and temperature as indicators could be considered as a limitation to this study other important indicators such as carbon dioxide emissions were excluded. Hence, the subject may be further explored by future researchers using different methodologies and indicators. Another limitation is that the study was carried out in a panel of Southern African countries with different climatic conditions such that the findings are unlikely to be consistent across the countries. Hence, such a research could be done limited to country-specific settings for each of the Southern African countries.