Mingzheng ZHAO, Fangfang ZHENG, Xiaomeng MA

College of Economics and Management, Henan Agricultural University, Zhengzhou 450046, China

Abstract The dependability of world major soybean exporters was measured and its influencing factors were analyzed through constructing the export output elasticity indicator. It was found that since 2010, among the world major soybean exporters, the dependability of Brazil, India, Russia and Ukraine increased, the dependability of Argentina and Paraguay dropped, while that of Canada, the USA and Uruguay was basically stable. Among the influencing factors of the dependability, the increase in international soybean prices and abundance of cultivated land resources are helpful for increasing the dependability, while the increase of population and improvement of economic development will reduce the dependability. China should reduce its dependence on countries with low export dependability, and actively develop production cooperation with import source countries to promote their dependability.

Key words Soybean trade, Dependability, Resource endowment, Economic development

1 Research background and present situation

In the "14Five-Year Plan" period, the soybean import will play a more prominent role in supporting China’s food security. This period is a critical transition period from achieving the goal of building a moderately prosperous society in an all-round way to building a modern and powerful socialist country in an all-round way. With the increasing food demand from economic development, China has become a world major food importer. Since 2015, China’s annual food imports have remained above 100 million t. In 2020, it imported 142.621 million t of grain, of which 10 032.7 million t was soybean, making China the world largest soybean importer. As predicted by the Chinese Academy of Social Sciences, by the end of the "14Five-Year Plan" period, China’s food demand will likely have a gap of about 130 million t, of which the soybean gap is about 100 million t. Thus, a large-scale soybean import will last for a long time.

Under such circumstance, the risks associated with China’s existing soybean import pattern are increasing with the expansion of the import scale. (i) Import sources are highly concentrated. In 2019, more than 90% of China’s soybean imports depended on the top three exporters. Such high concentration increases the risk of domestic and foreign food price linkage, and also poses certain political risks. (ii) The uncertainty of the global soybean market environment is increasing. The emerging anti-globalization thoughts and the outbreak of the COVID-19 have exacerbated the turbulence of the international trading environment. Due to Sino-US trade frictions, China reduced its soybean imports from the USA. This has won China the initiative in trade disputes, but it increased the pressure on domestic soybean supply and increased the risk of domestic food and oil prices and inflation. The rapid spread of COVID-19 in the world has caused abnormal fluctuations in the global food market including soybeans, leading to the danger of interruption in the international food logistics and supply chain. In the context of further expansion of the scale of soybean imports, China urgently needs to optimize the layout of soybean import sources, improve the reliability of soybean import sources, and reduce food import risks, so as to provide a solid guarantee for food security during the "14Five-Year Plan" period.

The No.1 Central Document for 2021 stressed optimizing the layout of trade of agricultural products and implementing the strategy of diversification of agricultural products imports. At the time of optimizing the layout of soybean import sources, it is not enough to simply increase the number of import source countries. It is also necessary to consider the country’s soybean export dependability, as a key decisive factor of China’s soybean import. The dependability of soybean exporters determines whether they can export soybeans to China steadily and continuously. If China takes soybean exporters with low dependability as the main source of soybean imports, once their soybean export volume to China declines or fluctuates, it will exert a huge impact on the stability of China’s soybean supply. Therefore, China should select those exporters with high dependability, take precautions, and predict the trend of changes in the dependability, and give early warning of import risks.

There have been extensive and in-depth studies about characteristics of global soybean trade and export dependability of agricultural products. These provide some helpful references for our study.

(i) Characteristics of global soybean trade. On the one hand, most studies on soybean exporters are concentrated in three countries: the USA, Brazil and Argentina. Reis

et

al.

analyzed the main soybean trading partners of the USA, Brazil and Argentina and found that due to the constraints of commodity prices and logistics costs, countries prefer to import soybeans rather than import soybean oil or soybean meal. Muhammad, Xia Pei, and Sun Jiangming studied the impact of price risks (price changes) in the USA, Brazil, and Argentina on the export competition in China’s soybean market. They found that Argentina is the only country in the Chinese market that is affected by its own price risk. For every 1% increase in price risk, China’s imports will decline by 1.11%; for every 1% increase in Brazilian prices, China’s imports from Argentina will increase by 0.51%; for every 1% increase in Argentina’s price risk, China’s imports from the USA will increase by 0.79%. Besides, these three countries have different export demand elasticity for soybean exports to China. The USA had the largest export elasticity, followed by Argentina, and Brazil had the smallest export elasticity. Lee

et

al.

studied the reasons for the decline in Argentina’s soybean exports. From 2002 to 2015, Argentina implemented an export license system for soybean exports, which restricted the soybean export volume; compared with soybeans, Argentina’s soybean meal and soybean oil had lower export tax rates, which encouraged its export of processed soybean products. On the other hand, studies on soybean importers are mainly concentrated in China, because China imports the most soybeans in the world. In the study on the characteristics of China’s soybean import trade, Lin Dayan

et

al.

found that the land endowment, labor endowment, capital endowment, economies of scale, transportation conditions, the fragmentation of China’s import market, market demand, and third-party market effects of import source countries are the main factor affecting China’s soybean import pattern. In addition, the time difference of soybean harvest between the northern and southern hemispheres also helps China import soybeans from southern hemisphere countries, reducing the fluctuation of soybean imports.(ii) Dependability of agricultural product export. Using the import dependence evaluation model, Liu Linqi analyzed the import dependence risk of China’s main food varieties and found that due to the inequality of agricultural imports and exports, China’s soybean import dependence risks mainly came from the USA and Brazil. Zhao Mingzheng constructed an export output elasticity indicator, evaluated the dependability of four food varieties in the international market, and found that the international market for maize had the lowest dependability, and the international market for soybeans had a higher dependability. Using this indicator, Zhang Zhexi

et

al.

and Li Huiyan

et

al.

analyzed the dependability of China’s maize import source countries, and Li Fangmin

et

al.

analyzed the dependability of China’s pork import source countries. Liu Hongduo and Chen Xiaoshan analyzed the factors affecting trade stability between China and the countries along the Belt and Road, and found that countries signed regional trade agreements and RMB local currency swap agreements, increased the degree of trade liberalization of countries along the Belt and Road, and reduced the country risks of countries along the Belt and Road, and both are helpful for improving the trade stability. Wang Shuai analyzed the risks of inland transportation and maritime transportation of food, and put forward policy recommendations such as diversifying import source countries, increasing import sources in the Black Sea region, and reducing dependence on the USA and Brazil. In the opinion of He

et

al.

, if China reduces imports of soybeans from the USA, American soybeans will be replaced by South American soybeans and local soybeans, and because South America has a comparative advantage over the USA, it will ultimately reduce the environmental cost of global soybean production. Hu Bingbing held that in the context of Sino-US trade friction, China’s agricultural trade needs to explore new development ideas. At present, China’s agricultural product trade only exerts part of its static comparative advantages, but not plays its dynamic intergenerational advantages. Due to the intergenerational differences in development between China and the countries along the Belt and Road, intergenerational advantages can be taken to bring new division of labor and improve trade welfare.

Previous studies have reached valuable conclusions and suggestions on the dependability of soybean international trade and agricultural product trade, but some points need to be improved. (i) Most studies focus on the USA, Brazil and Argentina, and not adequate attention has been paid to other major soybean exporters. In addition to the above three countries, the largest exporters of soybeans also include Canada, Paraguay, Uruguay, and Ukraine. If China wants to reduce the concentration of soybean trade imports, it should not only focus on the USA, Brazil, and Argentina, but also focus on other soybean exporters. (ii) Lack of in-depth analysis of dependability of soybean exports. As stated above, in the process of optimizing the layout of import sources, it is required to pay close attention to the changes in the export dependability of import source countries. Otherwise, even if the source of soybean imports is scattered, it may affect the stability of China’s soybean supply due to the low dependability of the new import source country. In view of these, we would improve the existing study from the following two aspects. (i) we extended the scope of the study to the nine countries with the largest soybean exports in recent years, including Argentina, Brazil, Canada, India, Paraguay, Russia, Ukraine, the USA and Uruguay, to grasp the overall export situation of soybean exporters. (ii) By constructing indicators and econometric models, we studied changing trend and influencing factors of the export dependability of soybean exporters, so as to provide information support for grasping the changing rules of soybean export dependability and optimizing the source of China’s soybean import.

2 Analysis framework, model setting and data description

2.1 Analysis framework

2.1.1

Evaluation indicator of soybean export dependability. With the continuous decline of transportation costs in international trade and the reduction of international trade barriers, the international soybean trade has become increasingly active, and the soybean export volume of world major soybean exporters has increased year by year. However, merely based on the increase in soybean exports, it is improper to believe that the dependability of soybean trade is increased. For soybean exporters, the increased supply of soybeans can be used for domestic demand as well as for external exports, and the part of trade is excess soybeans that cannot be consumed domestically. If the increase in domestic demand for soybeans in a country is greater than the increase in supply (the sum of output and imports in the current year), then it will be difficult to maintain the soybean exports in the future. In order to find out the true trend of soybean trade, it is necessary to make clear the proportion of increased supply used for domestic consumption and foreign exports. If large portion of the increase is used for domestic consumption, it can be considered that the dependability of country’s soybean exports is not high; on the contrary, if larger portion of the increase is used for foreign exports, it can be considered that the dependability of country’s soybean exports is high. In this study, we use the concept of "export output elasticity" to measure the dependability of the world’s major soybean exporters. Export output elasticity refers to the ratio of the increase in soybean exports to the increase in supply compared with the base period. It can be calculated using the following formula:

Export output elasticity = (Current year export volume/Base period export volume)/(Current year output/Base period output) = Current year export output ratio/Base period export output ratio

(1)

Export output elasticity is used to measure whether the increased supply of soybeans is used for domestic consumption or for export compared with the base period. If the export output elasticity is greater than 1, it means that more soybeans are used for export in the part of the increase in soybean supply; compared with the base period, the export dependability is increased; if the export output elasticity is less than 1, it means that more soybeans are used for domestic consumption, and the export dependability is decreased.

2.1.2

Analysis framework of factors influencing the dependability of soybean exports. A country’s willingness to export soybeans and its ability to export soybeans are jointly affected by domestic and international factors. International factors mainly refer to the international soybean price level and these factors constitute the "pulling force" of soybean exports and affect the willingness of soybean exports. When the international soybean price is high, it will attract soybean exporters’ willingness to increase exports, which will help increase the dependability of soybean exports; conversely, it will reduce the dependability of soybean exports. Domestic factors constitute the "pushing force" of soybean exports and determine the ability of soybean exports. Soybean exports are essentially the remaining part of soybean supply minus domestic consumption, so the export ability of a soybean exporter depends on its soybean supply and demand. If the domestic soybean demand is stable and the supply increases, there will be more soybeans for export, and the soybean export dependability will be higher; if the supply of soybeans is stable and the domestic demand increases faster, the soybeans available for export will decrease, then dependability of soybean exports will decline. From the perspective of demand, population and economic development level are important influencing factors. In terms of supply, soybean supply consists of production and imports. For soybean exporters, their soybean imports are very small and can be ignored. Therefore, the main factor determining soybean supply is soybean production. Production is determined by cultivated land endowment and unit yield, and the latter is closely related to the economic development level. As mentioned earlier, the main factors influencing soybean exports include international soybean prices, economic development level, population size, and cultivated land endowment (Fig.1).

Fig.1 Analysis framework of factors influencing the dependability of soybean exports

(i) International soybean price. International soybean prices affect the willingness of soybean exports. When the international soybean prices rise, a soybean exporter tends to increase soybean exports, which is conducive to increasing soybean exports. When other factors remain unchanged, the dependability of the country’s soybean export will be increased. (ii) Population size. The population size is the basic factor that determines the demand for food, and soybean is no exception. The larger the population, the more soybeans are consumed. When other factors remain constant, an increase in population will reduce the soybean export volume and reduce the dependability of soybean exports. (iii) Per capita demand. When the total population is constant, the higher the per capita soybean demand, the higher the total soybean demand. When other factors remain unchanged, it will lead to a decrease in soybean exports, ultimately leading to a decline in the dependability of soybean exports. The change in per capita soybean demand is mainly affected by the per capita income level. When the per capita income increases, the residents’ dietary structure will shift from plant-based food to animal-based food, and the feed grain demand will increase, which will increase the soybean demand. (iv) Unit yield of soybean. When soybean planting area is constant, the unit yield of soybean will increase, which will help increase soybean supply. When other factors remain unchanged, soybean exports increase, which leads to an increase in the dependability of soybean exports. The increase in unit yield of soybean depends on the national economic development. Only when the development of the national economy drives the popularization of modern agricultural production factors such as farmland water conservancy, road transportation, and improved varieties of agricultural machinery, can the unit yield of crops including soybeans increase steadily. (v) Cultivated land endowment. Under the premise of meeting certain agricultural production conditions, the more abundant cultivated land resources and the more cultivated land available for soybean production, the stronger the soybean production capacity. When other factors remain unchanged, the cultivated land endowment is helpful to increase soybean exports and increase the dependability of soybean exports. (vi) Economic development level. As an essential domestic factor, economic development has a more complex impact on the dependability of soybean exports. It influences the dependability of soybean exports from two aspects: per capita demand and unit yield of soybean. On the one hand, economic growth will change the diet of residents, leading to an increase in the consumption of meat, eggs, milk and other livestock products in the diet, which in turn will lead to an increase in the consumption of feed grains. Soybean is one of the most important feed grains, so domestic per capita demand for soybeans will increase with economic growth. When other factors remain unchanged, it is not conducive to the increase in the dependability of soybean exports. On the other hand, the increase of unit yield of soybean depends on the development of technology research and development, agricultural technology extension system and related farmland infrastructure, which are closely related to the level of economic development. Therefore, the higher the level of economic development, the higher the soybean production efficiency of countries with comparative advantages in soybean production, and the stronger the soybean production capacity, which will help increase the supply of soybeans and increase the dependability of soybean exports. Therefore, the ultimate impact of economic growth on the dependability of soybean exports depends on the superimposed effect of these two influencing paths. If economic growth has a greater impact on soybean demand, it will reduce the dependability of soybean exports; if it has a greater impact on soybean supply, it will increase the dependability of soybean exports.

2.2 Model setting of dependability influencing factors

According to the above theoretical analysis, we constructed the following panel model formula:

(2)

where subscripts

i

and

t

represent the country and year respectively; ln(

es

) is the logarithm of the export output ratio and is used to describe the changes in the dependability of soybean exports. According to formula (2), the dependability of soybean exports is essentially the comparison between the ratio of export output in the current year and the ratio of export output in the base period. If the ratio of export output increases, it means that the dependability of exports increases. Therefore, the export output ratio can be used to characterize the dependability of soybean exports. ln(

sp

) is the logarithm of international soybean prices; ln(

pgdp

) is the logarithm of per capita GDP. In order to eliminate the influence of inflation and make countries comparable, it is measured according to the per capita GDP data provided by the World Bank in 2010 constant prices in USD. In order to further explore the influence path of per capita GDP on the dependability, the different power terms of ln(

pgdp

) are added. The superscript

j

represents the power of ln(_

pgdp

_),

j

=1,…,

N

; ln(

pgdp

) is the logarithm of the population size, and the population size is measured by the total population. ln(

pland

) is the logarithm of the per capita soybean planting area and is used to describe the cultivated land resource endowment of a country. Soybean is a typical land-intensive crop. The more cultivated land resources are in the country, the higher the per capita planting area of soybeans. ln(

pland

)×ln(

pgdp

) is the interaction term of the logarithm of the per capita soybean planting area and the logarithm of the per capita GDP. It is used to measure the impact of economic development on farmland infrastructure and unit yield of soybean, and then affect the possibility of dependability of soybean exports.

Z

is a control variable and it mainly includes population structure, urbanization level, and the proportion of agricultural GDP in total GDP. The population structure and urbanization level are measured by the proportion of the population over 60 years old and the proportion of the urban population, respectively.

α

is the constant term,

β

-

β

are the regression coefficient of each explanatory variable and are used to measure the influence of each variable on the dependability of soybean exports, and,

ε

is the stochastic error.

2.3 Variable measurement and data source description

In this study, we used the panel data from the top nine soybean exporters in 1990-2019 (Argentina, Brazil, Canada, India, Paraguay, Russia, Ukraine, the USA, and Uruguay) as the research sample. In 2020, soybean exports from these 9 countries accounted for 99.6% of the total global soybean exports. The descriptive statistical results of each variable indicator are listed in Table 1.

3 Empirical results and tests

We first used the export output elasticity indicator to analyze the dependability of soybean exporters, and then carried out the regression analysis on the factors determining the dependability of soybean exporters, and finally conducted a robustness test.

3.1 Calculation of the dependability of soybean exporters

Table 2 shows the soybean export output elasticity of soybean exporters in 2010-2020 based on the 3-year average data of 2009-2011. When the export output elasticity is greater than 1, it means that the dependability is increased, and when it is less than 1, it means that the dependability is decreased. From Table 2, it can be seen that since 2010, the dependability of soybean exports in Argentina and Paraguay declined. Especially, the dependability of Argentina was not stable, and the fluctuation range was large. The countries with great increase of soybean export dependability were Brazil, India, Russia and Ukraine. India and Russia had larger export output elasticity values, mainly due to the lower export volume of soybeans in the base period; the soybean export dependability of Canada, the USA, and Uruguay were basically stable with little fluctuation. The soybean export output elasticity of the USA remained at around 1.1 in most years, which means that its soybean export dependability was basically stable. However, it suddenly dropped to 0.89 in 2018, which may be caused by the Sino-US trade friction during that year.

Table 1 Descriptive statistics

Table 2 Soybean export output elasticity of soybean exporters

To further analyze the reasons for the changes in soybean export output elasticity of world major soybean exporters, Table 3 shows the average annual growth rate of soybean yield, export volume, domestic consumption of the above nine countries in 2010-2020, and the proportion of various types of consumption in total soybean consumption in 2020. Soybean is an important food resource for any country. It will not be exported before there is excess. Therefore, changes in export volume are closely related to changes in yield and domestic consumption. Which of the faster the growth rate of yield or the growth rate of domestic consumption determines the trend of changes in the dependability of soybean exports. In the two countries with decline of the dependability of soybean export, the decline in soybean yield in Argentina and the increase in domestic consumption are the main reasons for the decline the dependability of soybean export. In 2010-2020, Argentina’s soybean yield dropped by an average annual rate of 0.21%, and domestic consumption increased by an average annual rate of 1.45%. Faster growth rate of Paraguay’s domestic soybean consumption than the yield growth rate is the main reason for the decline in its dependability of soybean export. The average annual growth rate of domestic soybean consumption in Paraguay was 8.89%, and the average annual growth rate of production was 3.7%. In the soybean consumption of Argentina and Paraguay, crushed consumption accounted for a relatively high proportion and grew rapidly, which is a main driving factor for the increase in domestic soybean consumption growth rate. In the four countries with increased dependability soybean exports, the growth rate of yield was higher than the growth rate of domestic consumption. This is the fundamental reason for the increase in dependability of soybean exports. The average annual growth rates of soybean yield in Brazil, India, Russia, and Ukraine were 5.85%, 0.36%, 14.27%, and 6.32%, respectively, and the average annual growth rates of domestic consumption were 2.18%, 0.25%, 9.88%, and 4.05%, respectively. Among the three countries with relative stable dependability of soybean exports, the average annual growth rates of yield in Canada, the USA, and Uruguay were 3.63%, 2.19%, and 1.72%, respectively, and the average annual growth rates of domestic consumption were 5.29%, 2.73% and 2.92%, respectively. Considering that the yield base of soybean exporters was greater than the domestic consumption base, the yield growth rate was slightly lower than the domestic consumption growth rate, which means that the growth of soybean yield and domestic consumption was more balanced, so the dependability of soybean exports was relatively stable. In summary, in order to strengthen the stability and sustainability of soybean imports, China should strengthen cooperation with countries with high dependability of soybean exports and reduce cooperation with countries with low dependability. From the perspective of soybean export output elasticity, China should take the initiative to increase the soybean imports from Brazil, India, Russia and Ukraine, maintain the soybean imports from the USA, Canada and Uruguay, and not rely too much on Argentina and Paraguay. In addition, India, Russia, and Ukraine are relatively close to China and are countries along the Belt and Road. Importing soybeans from these countries will not only help reduce the pressure on soybean transportation by sea, but also strengthen the cooperation between China and the countries along the Belt and Road.

3.2 Regression analysis on factors influencing the dependability of soybean exports

3.2.1

Model estimation results. Through comparing the estimation results of mixed regression model, fixed effects model and random effects model on panel data, we found that the fixed effects model is the most effective. Through testing the heteroscedasticity and autocorrelation estimated by the fixed effects model, we found that the sample data has significant heteroscedasticity and autocorrelation. We used a comprehensive feasible generalized least squares method (FGLS) to correct heteroscedasticity and autocorrelation. The basic regression results are shown in the first column of Table 4. To comprehensively examine the functional relationship between the economic development level and the dependability of soybean exports, we introduced the interaction term of ln(

pgdp

) and ln(

pland

) into formula (2), and the regression results are shown in the second column of Table 4. From the regression results that after adding the interaction term, the interaction terms of the logarithm of the per capita soybean planting area and the logarithm of the per capita GDP are no longer significant. The path effect of the dependability is not significant. Taking the four major food crops in the world as an example, in 1991-2019, the average annual growth rate of global rice, wheat, maize, and soybean yields were 52.00, 38.15, 75.46, and 28.45 kg/ha, respectively. Among the four food crops, the increase in unit yield has the weakest contribution to the increase in soybean yield. Thus, soybean is a typical land-intensive agricultural product, and it is more likely to increase yield through area expansion. Therefore, in the follow-up analysis, we did not add this interaction item. Then, we introduced the second and third terms of ln(

pgdp

) into formula (2) in turn, and the regression results are shown in the third and fourth columns of Table 4. From the regression results, the sub-items of the logarithm of per capita GDP are significant in the regression. According to the regression results in the third column, the first term coefficient of the logarithm of per capita GDP is -6.412, and the second term coefficient is 0.277, which means that an increase in per capita GDP will lead to a decrease and then an increase in the dependability of soybean exports. According to the derivation rule of the quadratic equation of one unknown, when the first derivative is equal to zero at the inflection point, the logarithm of per capita GDP is 11.574 (6.412÷(2×0.277)), and the value of per capita GDP is 106 298.6 (e) USD. This means that as long as the per capita GDP is less than 100 000 USD, the increase in the per capita GDP will lead to a decline in the dependability of soybean exports. The per capita GDP of most countries is less than 100 000 USD, which shows that the increase in per capita GDP of most countries in a long period of time will lead to a decline in the dependability of soybean exports. Therefore, analysis with the model in the first column can meet the needs of analysis. For the consideration of model simplicity, we only kept the first term of per capita GDP, and conducted analysis based on the first column.

Table 3 Soybean yield, export volume and consumption growth rate and proportion of soybean exporters (%)

Table 4 Model estimation results

(i) In terms of the international soybean price, the coefficient is positive and significant at the 1% level, indicating that the increase in international soybean prices is conducive to attracting soybean exporters to increase soybean exports. When other factors remain unchanged, it will increase the export supply ratio and increase the dependability of soybean exports. (ii) For the economic development level, the coefficient of per capita GDP is negative and significant at the level of 1%, indicating that the increase in per capita GDP leads to an increase in domestic per capita soybean demand. When other factors remain unchanged, the export volume of soybeans will decrease and the dependability of soybean exports will decline. (iii) In terms of population, the coefficient is negative and significant at the level of 1%, indicating that the increase in population leads to an increase in domestic soybean demand. When other conditions remain unchanged, the soybean export volume will decrease, the ratio of soybean export output will decrease, and the dependability of soybean exports will decline. (iv) As regards the per capita soybean planting area, the coefficient is positive and significant at the level of 1%, indicating that the increase in per capita soybean planting area will help increase soybean yield. When other factors remain unchanged, the export volume of soybeans will increase, which will help increase the dependability of soybean exports. The increase in per capita planting area will help increase the dependability of soybean exports, which is a manifestation of the characteristics of soybean as a land-intensive agricultural product. Therefore, countries with abundant cultivated land resource endowments have more advantages in soybean production.

3.2.2

Robustness test. In this study, we performed the robustness test by replacing variables, transforming the model estimation method, and shortening the sample time span to ensure the reliability of the above results. The results are shown in Table 5. The first column of Table 5 is the regression result after replacing the original independent variable with the lagging one-period data of all independent variables. Regression with the lag period data of the explanatory variables can solve the endogeneity problem to a certain extent. Compared with the second column, the coefficient signs and significance levels of the international soybean price, per capita GDP, population, and per capita soybean planting area in the regression results of the third column are basically the same as those in the first column of Table 4. The second column is to use the fixed effects regression method to regress the data. From the results, the sign directions of the coefficients of the main explanatory variables are basically consistent with the second column of Table 4.

Table 5 Robustness test of the model

The third and fourth columns of Table 5 are regressions after reducing the sample size. America is currently the most important continent for soybean exports. Therefore, the data of American countries are separately extracted from the sample for regression. The results are shown in the third column of Table 5. From the regression results, except for the sign of the coefficient of per capita soybean planting area, which is negative, the signs of the coefficients of other main explanatory variables are consistent with the first column of Table 4. Among them, the coefficients that pass the significance test are the international soybean price and the population, indicating that these two factors have a significant impact on the dependability of American soybean exporters. The fourth column of Table 5 is a regression performed by shortening the sample time span to 2000-2019. Compared with the first column of Table 4, after shortening the sample time span, the main explanatory variables are still significant and the sign direction remains unchanged, which means that shortening the time span does not significantly change the research results. On the whole, the regression results are robust.

4 Conclusions and policy recommendations

4.1 Conclusions

(i) The dependability of soybean exporters is dynamically changing. Compared with 2010, the dependability of of soybean exports in Brazil, India, Russia and Ukraine is increasing, Canada, the USA and Uruguay basically remain stable, while Argentina and Paraguay are declining significantly. (ii) The dependability of soybean exports is affected jointly by international and domestic factors. Among them, the increase in international soybean prices and the expansion of per capita soybean planting area will help increase the dependability of soybean exports; the increase in the domestic population and the improvement of economic development level will reduce the dependability of soybean exports. (iii) Soybean is a typical land-intensive agricultural product, it mainly relies on the expansion of planting area to increase yield. Economic development drives the increase of unit yield and does not have a significant effect on the increase of soybean yield.

4.2 Policy recommendations

(i) It is recommended to increase the soybean imports from countries with high dependability of soybean exports, and reduce the dependence on countries with low export dependability and gradually adjust the current situation of high concentration of soybean imports through properly arranging the country of origin of soybean imports. In 2019, China’s soybean imports reached 88 585 900 t, of which soybeans from Argentina, Brazil, Canada, the USA and Uruguay accounted for 9.92%, 65.11%, 2.56%, 19.21% and 2.33%, respectively. Among these five countries, Brazil had a higher dependability of soybean exports, Canada, the USA and Uruguay had a relatively stable dependability of soybean exports, and the dependability of soybean exports in Argentina has continued to decline. China should reduce imports of soybeans from Argentina and reduce its dependence on Argentina. (ii) China should increase the soybeans imports from Russia and Ukraine. Ukraine and Russia exported 732 600 and 21 600 t of soybeans to China in 2019, which are relatively small. These two countries have huge soybean production potential, and their dependability of soybean export is increasing year by year. Increasing soybean imports from these two countries will help to improve the current situation of China’s soybean imports being too dependent on the American countries. In addition, these two countries are along the Belt and Road. Increasing soybean imports from them will help enrich the connotation of the Belt and Road Initiative strategic partnership. (iii) China should actively develop cooperation with soybean exporters in the development of soybean production potential. The increase in per capita soybean planting area is helpful for increasing the dependability of soybean exports. Therefore, China should actively cooperate with soybean exporters in land resource development, agricultural infrastructure construction, agricultural product storage and logistics, agricultural materials and agricultural machinery investment, and agricultural product purchases,

etc.

, assist them in improving the soybean production capacity and improving the dependability of soybean exports, so as to increase the stability of soybean import.