A Study on the Trade Efficiency and Potential of China’s Agricultural Products Export to Association of South East Asian Nations Countries: Empirical Analysis Based on Segmented Products

Abstract

This study examines the current state of China’s agricultural exports to ASEAN countries using a segmented export structure analysis via a stochastic frontier gravity model, based on panel data from 2007 to 2020. The results indicate that: (1) China’s primary agricultural exports to ASEAN countries include plant products, food and beverages, and tobacco, with animal products mainly exported to Thailand, plant products mainly exported to Vietnam, and animal and plant fats, food, beverages, and tobacco mainly exported to Malaysia. (2) The economic scale and population size of China and ASEAN countries have differing impacts on various markets, while distance significantly negatively affects the exports of animal products, plant products, food, beverages, and tobacco. Additionally, ASEAN countries’ per capita carbon emissions positively influence the exports of these product categories. (3) The liner shipping connectivity index is significantly negatively correlated with trade inefficiency. The influences of financial freedom, trade freedom, investment freedom, and government expenditure on trade inefficiency vary across ASEAN countries. (4) China’s export efficiency for animal products, plant products, food, beverages, and tobacco has increased rapidly to Thailand and Vietnam, with Malaysia and Singapore showing high export efficiency, while that to Cambodia lags. (5) China exhibits significant export potential to Thailand, Indonesia, and Vietnam, with substantial expansion opportunities in Indonesia. Moreover, China’s export potential and opportunities in Cambodia are steadily increasing.

1. Introduction

With the accelerated advancement of global economic integration, cooperation between China and ASEAN countries in the agricultural sector has become increasingly close in recent years. China–ASEAN bilateral trade, especially agricultural trade, is essential to China’s foreign trade. ASEAN countries have been China’s largest trade partners for four consecutive years, with the average annual growth rate of bilateral trade reaching 8.8% between 2013 and 2023 [1]. In the context of optimizing trade rules [2] and enhancing cross-regional trade facilitation, China–ASEAN agricultural trade has achieved significant growth. By 2021, bilateral agricultural trade had basically achieved “zero tariffs”, making the countries one another’s main export markets for agricultural products [3]. The average annual growth rate of bilateral agricultural trade has even been as high as 11% [4].

However, in the broader context of the global low-carbon economic transition, ASEAN countries are actively promoting the development of a green and low-carbon economy through industrial upgrading and transformation. Green consumption preferences have become an important engine for reshaping the pattern of agricultural trade. As a major exporter of agricultural products, whether China’s agricultural export trade structure will be affected by the differing low-carbon economic goals of ASEAN countries is an urgent question to be clarified.

The methodology of regional trade studies mainly consists of ex ante assessments based on general equilibrium models and ex-post assessments based on gravity models. Confined by the sensitivity of the general equilibrium model [3], at present, relevant studies are mostly carried out in terms of ex-post assessment, such as forecasting the potential of China–ASEAN bilateral trade in agricultural products, evaluating trade efficiency, and exploring the influencing factors. Trade potential prediction takes trade competitiveness and complementarity as the main evaluation tools, and some scholars have estimated that China–ASEAN bilateral agricultural trade potential is the main evaluation tool from the perspective of trade geography [5]. From the perspective of trade geography, some scholars conclude that China–ASEAN agricultural trade is mainly complementary, and the bilateral trade potential is large. There are also scholars who have a perspective on commodity trade structure [6]. From the perspective of commodity trade structure, some scholars find that the China–ASEAN trade cooperation space is expanding in low-technology products, while in medium- and high-technology products, there is a state of competition and integration. Studies on the trade efficiency of agricultural products show that there are obvious differences in country-specific efficiency in China–ASEAN bilateral trade, and trade efficiency is negatively correlated with trade volume [7]. As above, trade efficiency is negatively correlated with trade volume. The influencing factors are “soft factors”, such as trade and investment freedom, government expenditure and integrity, as well as liner shipping connectivity [8]. Different forms of “hard power”, such as liner transportation connectivity and communication infrastructure, all play an obvious role [9]. Therefore, it is important to emphasize the bilateral trade between China and ASEAN countries. Therefore, it has become a consensus among scholars to emphasize the geographical orientation and commodity structure of China–ASEAN bilateral trade [10]. Therefore, it has become a consensus among scholars to emphasize the geographical orientation and commodity structure in China–ASEAN bilateral trade. From the perspective of research methodology, in addition to the traditional gravity model, the propensity matching score method (PSM) has been adopted [11]; the propensity matching score method (PSM) and double difference model (DID) have obvious advantages in capturing disturbances that do not change over time [12] while the synthetic control method (SCM) and survival analysis have also been used in studies that tap into the variability of trade effects across member countries [13].

To summarize, although rich in research on China’s agricultural trade efficiency and trade potential using stochastic frontier gravity modeling, most of the studies are on total agricultural trade, and there is relatively little literature on segmented agricultural export trade. Given the serious dependence on plant products and animal and vegetable oils and fats in China’s agricultural trade deficit [14], it is more relevant to subdivide China–ASEAN agricultural export trade into four HS codes. In addition, a few scholars have introduced carbon emissions into the model to conclude that there are significant positive effects and regional differences in China’s foreign merchandise import and export trade [15,16]. However, is there also a positive effect on China’s agricultural exports? Whether a similar pattern exists for China’s agricultural exports needs to be further verified. This paper utilizes the Uncomtrade database of China–ASEAN agricultural export trade data from 2007 to 2020, incorporates per capita carbon emissions into the stochastic frontier gravity model, and draws a picture of the spatio-temporal evolution and product differences in China–ASEAN agricultural export trade with a view to providing suggestions for the formulation of a more targeted agricultural export trade policy.

In view of this, this paper aims to conduct an in-depth analysis of the current status of China’s agricultural exports to ASEAN countries, particularly focusing on the export situation of different segmented products (such as animal products, plant products, animal and vegetable oils and fats, food, beverages, and tobacco). It analyzes the spatiotemporal evolution factors that affect trade efficiency and product differentiation and explores the inducing effects of a low-carbon economy on China–ASEAN agricultural export trade. Using the Stochastic Frontier Gravity Model and based on panel data from 2007 to 2020, this paper empirically analyzes the efficiency and potential of China’s agricultural exports to ASEAN countries, aiming to provide references for formulating more precise agricultural export trade policies.

The structure of this paper is organized as follows: the second part presents the theoretical analysis and research hypotheses, proposing the research hypotheses of this paper based on the Preference Similarity Theory, the Traction Growth Theory, and the Fallacy of Composition Theory. The third part discusses the research design, introducing the data sources, variable selection, and model construction. The fourth part is the empirical analysis, including model applicability testing, analysis of the results of the Stochastic Frontier Gravity Model, analysis of the results of the trade inefficiency model, and analysis of export efficiency and potential. The fifth part presents the conclusions and discussions, summarizing the research findings, proposing policy recommendations, and indicating future research directions.

2. Theoretical Analysis and Research Hypotheses

2.1. Preference Similarity Theory

The preference similarity theory proposed by the Swedish economist Linder explains for the first time the phenomenon of intra-industry trade from a demand perspective. The export power of agricultural trade comes from the increase in domestic demand, which promotes the expansion of market capacity and thus realizes economies of scale. Preference similarity, in turn, depends on factors such as income and culture [17]. Cultural factors can be subdivided into living customs characterized by public boundaries and Chinese cultural identity characterized by a common language. Based on this, the following research hypotheses are proposed:

H1: 

An increase in China’s total agricultural output will promote the growth of China’s agricultural exports to ASEAN countries.

H2: 

A common border between the two countries will facilitate the growth of China’s agricultural exports to ASEAN countries.

H3: 

A common language between the two countries has a positive impact on the growth of China’s agricultural exports to ASEAN countries.

2.2. Traction Growth Theory

Lewis’s traction growth theory, which dates back to the 1950s, states that a nation’s degree of economic development determines how much it imports. More robust import demand is a direct result of faster economic development, which impacts export volume for the exporting nation. Consequently, the traction of agricultural output in the importing nation determines the level of agricultural export trade [18]. As a result, the growth of agricultural production in the importing nation determines the growth of agricultural export commerce. The importing country’s level of economic openness concurrently influences the pulling impact. Openness in banking and investment policies is substitutive, whereas openness in shipping and trade policies supports commerce. Based on this, the following research hypotheses are proposed:

H4: 

An increase in the agricultural output value of ASEAN countries will promote the growth of China’s agricultural exports to ASEAN countries.

H5: 

The degree of economic policy openness (shipping, finance, trade, investment) of ASEAN countries has a mixed impact on China–ASEAN agricultural exports.

2.3. Synthetic Fallacy Theory

The synthetic fallacy argument often explains why the East Asian export-oriented economic model cannot be replicated indefinitely. Due to the limited market capacity of the importing country, once the importing country has an oversupply of a specific type of product, it will inevitably adopt trade protection policies to restrict imports [19]. Therefore, the commodity demand of the importing country is the starting point and foothold for the formulation of import trade policy, and the demand is divided into three categories: segmented demand, sophisticated and picky demand, and forward-looking demand. In the case of China’s agricultural exports to ASEAN countries, segmented demand is represented by the exports of the four HS-coded commodities under agricultural products. In comparison, forward-looking demand is reflected in the extent to which the exports of the four HS-coded commodities are responsive to the per capita carbon emissions of the ASEAN countries under the dual-carbon perspective. Based on this, the following research hypotheses are proposed:

H6: 

China’s agricultural exports to ASEAN countries are characterized by a markedly differentiated segmented commodity structure;

H7: 

There is heterogeneity in the carbon emissions per capita level across ASEAN countries, which influences China’s segmented agricultural exports to ASEAN countries.

Based on this, the theoretical analysis framework was constructed, as shown in Figure 1.

3. Research Design

3.1. Data Source and Variable Selection

Due to serious missing data from Brunei and Laos, for the accuracy of the data source, this study excludes Brunei and Laos and finally selected eight countries as samples, namely Indonesia, Malaysia, the Philippines, Thailand, Singapore, Cambodia, Myanmar, and Vietnam. Due to the missing sample data both before 2007 and after 2020, this study chooses the panel data from 2007–2020. The variable descriptions and data sources are shown in

Table 1. Description of variables and data sources.

Explanatory Variable	Y	China’s Agricultural Exports to Country j in Period t	Predictor	UNComtrade Database
Explanatory variable	ACGDP	China’s gross agricultural output in period t	+	World Bank WDI database
	AGDP	Gross agricultural output in country j in period t	+	World Bank WDI database
	DIST	Distance between the two capitals	-	French CEPII database
	BOR	Whether the two countries share a common border	+	French CEPII database
	LANG	Do the two countries share a common language?	+	French CEPII database
	CARBON	Per capita carbon emissions in country j	+	World Bank WDI database
Covariate	μ	Trade inefficiencies
	SHP	Liner shipping connectivity index for country j in year t	-	World Bank WDI database
	FIN	Financial freedom index for country j in year t	+	American Heritage Foundation
	TRA	Trade freedom index for country j in year t	-	American Heritage Foundation
	INV	Freedom of investment index for country j in year t	+	American Heritage Foundation
	GOV	Level of government expenditure in country j in year t	+	American Heritage Foundation

.

3.2. Selection of Variables

Explained variable: The variables explained in this paper are China’s exports to the eight ASEAN countries from 2007 to 2020 and the segmented agricultural product exports under the HS category 4 code.

Explanatory variable: Referring to the definition in the mainstream gravity model [20], the total agricultural output value, distance, cultural similarity of the import and export countries, and the carbon emission level of the importing country are taken as explanatory variables, among which the total agricultural output value of the two countries and the carbon emission level of the importing country are the core explanatory variables of this paper.

Concomitant variable: Considering the bridge effect of the economic opening policy of the importing country on trade, this paper analyzes the non-efficiency items from the perspectives of shipping connectivity, finance, trade, investment, and government expenditure.

3.3. Modeling

3.3.1. Stochastic Frontier Gravity Model

Referring to the existing stochastic frontier gravity models [21], the specific form is set as follows:

$$Y_{ijt}=f\left(X_{ijt},\beta \right)exp\left(v_{ijt}\right)exp\left(-u_{ijt}\right), u_{ijt}\ge 0$$

(1)

Take the logarithm of the formula and obtain the formula:

$$ln{Y}_{ijt}=lnf\left(X_{ijt},\beta \right)+v_{ijt}-u_{ijt }$$

(2)

In the formula, i represents the exporting country and j represents the importing country. Y_ijt represents the actual trade volume of exports from country _i to country _j, X_ijt represents the factors affecting the trade volume (including economic size, common language, geographical distance, etc.), and β represents the parameters to be estimated. v_ijt − u_ijt is a compound error term, where v_ijt is a random error term, usually set to obey a normal distribution. u_ijt is a non-efficiency term, which represents the artificially caused trade resistance factors, and is set to obey the semi-normal distribution or the broken tail normal distribution. According to whether the non-efficiency term u_ijt changes with time, there are sometimes invariant models and time-varying models.

$${Y^{*}}_{ijt}=f\left(X_{ijt},\beta \right)exp\left(v_{ijt}\right)$$

(3)

$$T{E}_{ijt}=Y_{ijt}/ {Y^{*}}_{ijt}=exp\left(-u_{ijt}\right)$$

(4)

When u_ijt = 0, Formula (3) can be obtained from Formula (1), where Y^*_ijt represents the maximum export value that can be achieved without trade resistance, that is, trade potential. In Formula (4), TE_ijt represents trade efficiency, which is the ratio of actual export value to trade potential. The value of TE_ijt is [0,1]. When the trade inefficiency u_ijt = 0, TE_ijt = 1, the actual export value equals the trade potential value, and the trade relationship between the two countries is the optimal situation. When u_ijt > 0, 0 < TE_ijt < 1, At this time, there is trade resistance, and the actual export value is less than the potential value of trade.

3.3.2. Model Construction

The stochastic frontier gravity model in this paper includes economic size and geographic distance, which are depicted in Figure 2. Drawing on previous studies that grouped natural factors such as the economic size of the two countries, their geographic distance, and whether or not they are landlocked into the model [22], and analyzes human factors such as infrastructure and institutional environment in the trade inefficiency term model [23]. Meanwhile, under the dual-carbon perspective, therefore, this paper introduces the environmental variables of carbon emissions into the stochastic frontier analysis framework. As a result, the stochastic frontier gravity model of China’s agricultural exports to ASEAN countries constructed in this paper is logarithmically treated as follows:

$$ln{Y}_{ijt}={\beta }_0+{\beta }_1\mathrm{l}\mathrm{n}\mathrm{A}CGD{P}_{it}+{\beta }_{2}lnAGD{P}_{jt}+{\beta }_{3}lnDIS{T}_{ij}+{\beta }_{4}BO{R}_{ij}+{\beta }_{5}LAN{G}_{ij}+{\beta }_{6}CARBO{N}_{ij}+V_{ij}-U_{ij}$$

(5)

Combined with the previous research methods, the trade inefficiency model constructed in this paper is as follows:

$$u_{ijt}={\alpha }_0+{\alpha }_{1}SH{P}_{jt}+{\alpha }_{2}FI{N}_{jt}+{\alpha }_{3}TR{A}_{jt}+{\alpha }_{4}IN{V}_{jt}+{\alpha }_{5}GO{V}_{jt}+{\epsilon }_{ijt}$$

(6)

4. Empirical Analysis

4.1. Model Applicability Tests

Based on the above theoretical model and panel data, the regression analysis of the model was performed using the software frontier.4.1. To ensure the applicability of the model, it is necessary to use the likelihood ratio test (LR test) to test the time-variability of trade inefficiencies and trade efficiencies successively before regression analysis. As can be seen from

Table 2. Model 1 applicability test.

Original Hypothesis	Constrained Model	Unconstrained Model	LR Statistic	5% Threshold	Test Conclusion
Absence of trade inefficiencies	−93.369	−3.108	180.522	5.991	rejection
Trade inefficiency inaction	−3.108	4.086	14.388	3.841	rejection

,

Table 3. Model 2 applicability test.

Original Hypothesis	Constrained Model	Unconstrained Model	LR Statistic	5% Threshold	Test Conclusion
Absence of trade inefficiencies	−178.324	−150.550	55.548	5.991	rejection
Trade inefficiency inaction	−150.550	−143.146	14.808	3.841	rejection

and

Table 4. Model 3 applicability test.

Original Hypothesis	Constrained Model	Unconstrained Model	LR Statistic	5% Threshold	Test Conclusion
Absence of trade inefficiencies	105.976	59.126	93.700	5.991	rejection
Trade inefficiency inaction	−59.126	54.338	9.576	3.841	rejection

the original hypothesis of China’s total agricultural exports to ASEAN countries and the export of animal and plant products in subdivided products are rejected at the significance level of 5%.

Table 5. Model 4 applicability test.

Original Hypothesis	Constrained Model	Unconstrained Model	LR Statistic	5% Threshold	Test Conclusion
Absence of trade inefficiencies	246.980	239.843	14.274	5.991	rejection
Trade inefficiency inaction	239.843	238.503	2.68	3.841	no refusal

displays that the export of animal and vegetable oils to ASEAN countries cannot reject the null hypothesis that trade inefficiency does not change, which indicates that trade inefficiency does not change with time. For ASEAN food, beverage, and tobacco exports, as can be seen from

Table 6. Model 5 applicability test.

Original Hypothesis	Constrained Model	Unconstrained Model	LR Statistic	5% Threshold	Test Conclusion
Absence of trade inefficiencies	97.629	11.357	217.972	5.991	rejection
Trade inefficiency inaction	11.357	13.630	0.546	3.841	rejection

, the null hypothesis was rejected at the 5% significance level. Therefore, there are trade inefficiencies and trade efficiency changes over time.

4.2. Analysis of Model Results

4.2.1. Analysis of Stochastic Frontier Gravity Model Results

In this paper, by using the software frontier.4.1, a stochastic frontier analysis is carried out on the panel data of China’s exported sub-products to ASEAN member states. The estimated results of the stochastic frontier gravity model are shown in

Table 7. Stochastic frontier gravity model estimation results.

Variant	Model 1		Model 2		Model 3		Model 4		Model 5
	Ratio	t-Value	Ratio	t-Value	Ratio	t-Value	Ratio	t-Value	Ratio	t-Value
constant term (math.)	−17.874 ***	−6.961	−33.438 ***	−6.702	−14.783 ***	−3.684	13.013 ***	3.283	−14.764 ***	−8.041
lnACGDP	1.253 ***	13.071	1.491 ***	8.225	1.005 ***	7.745	0.125	0.463	1.170 ***	19.364
lnAGDP	0.227 ***	2.863	0.245 ***	3.053	0.313 ***	4.028	0.104	0.525	0.140 ***	3.322
lnDIST	−0.072	−1.045	0.501 **	2.033	−0.344	−0.293	−0.394	−0.806	−0.038	−0.656
BOR	−0.017	−0.128	−0.843 **	−2.506	0.425 **	2.445	0.812	0.814	−0.629 ***	−5.644
LANG	−0.301	−0.615	−1.235 *	−1.839	−1.452 ***	−2.636	−2.414 *	−1.736	−0.711 ***	−2.709
lnCARBON	0.820 ***	11.115	0.764 ***	2.766	1.357 ***	13.965	3.025 ***	6.703	0.543 ***	8.367
σ2	5.960	0.683	3.359 ***	2.947	4.552 **	2.076	12.971 **	2.420	3.212 *	1.913
γ	0.993 ***	97.154	0.823 ***	9.887	0.973 ***	65.778	0.709 ***	4.886	0.990 ***	145.111
μ	−4.866	−0.543	−3.324 *	−1.948	−4.209 **	−2.401	−6.065	−1.255	−3.565 **	−2.303
η	−0.030 ***	−4.327	0.079 ***	4.859	−0.050 ***	−3.302			−0.012 **	−2.457
log-likelihood	4.086		−143.146		−54.338		−239.843		13.630
LR test	194.911		70.357		103.275		14.273		222.519

“*”, “**”, and “***” indicate significance at 10%, 5%, and 1% levels, respectively.

, among which the output results of the time-varying model are adopted except for the time-invariant model for animal and vegetable oils.

In the time-varying model, η all pass the 5% significance level test, indicating that the trade inefficiency term changes with time. The coefficient for animal products is positive, indicating a decrease in trade inefficiencies over time, while the coefficient for total agricultural exports, plant products, food and beverages, and tobacco is negative, indicating a slight increase in trade inefficiencies over time.

Model 1 indicates that the gross agricultural output value (ACGDP) of China, the gross agricultural output value (AGDP) of ASEAN member countries, and the per capita carbon emissions (CARBON) of ASEAN member countries are the main factors affecting the value of China’s agricultural exports to the eight ASEAN countries. These factors have a significant positive impact on China’s exports of agricultural products, which is consistent with the expected sign. In contrast, there is no discernible difference in the two countries’ distance (DIST), shared language (LANG), or border (BOR).

Model 2 indicates that, for the first type of animal products, China’s total agricultural output value (ACGDP), ASEAN member countries’ total agricultural output value (AGDP), and ASEAN member countries’ per capita carbon emissions (CARBON) have a significant positive impact on China’s animal product exports, which is consistent with the expected symbols. The distance (DIST) between the two countries has a significant positive impact on China’s export of this kind of product, which is inconsistent with the expected symbol. Maybe it is because the export of this kind of product is also affected by other factors, such as trade facilitation to shorten the trade time further and cost, increasing foreign investment in China’s agriculture, etc. Therefore, it may be biased to consider the impact of distance on the export of animal products separately. Whether there is a common border (BOR) or not and a common language (LANG) has a significant negative impact on the export of such products in China, which is inconsistent with the expected symbols. Maybe it is because, although common borders and languages serve as carriers to promote cross-cultural communication between the two sides of the trade, their use situations may differ in different cultural environments. We should fully consider the different cultural backgrounds, ways of thinking, values, and customs of different countries.

Model 3 indicates that, for the second type of plant products, China’s total agricultural output value (ACGDP), ASEAN member countries’ total agricultural output value (AGDP), whether there is a common border (BOR), and per capita carbon emissions (CARBON) of ASEAN member countries have a significant positive impact on the export of Chinese plant products, which is consistent with the expected symbols. Whether there is a common language (LANG) significantly negatively impacts the export of Chinese plant products, which is inconsistent with the expected symbol. Maybe it is because language as a communication tool has a profound cultural background, and sometimes, the same word will express different meanings, which will cause disputes and misunderstandings. The distance between the two countries has no significant effect on the export of such products.

Model 4 shows that for the third type of animal and vegetable oils, per capita carbon emissions (CARBON) of ASEAN member countries have a significant positive impact on China’s animal and vegetable oil exports, consistent with the expected symbol. Whether there is a common language (LANG) has a significant negative impact on the export of such products in China, which is inconsistent with the expected symbol, which may also be due to the different use of language in different cultural environments, and the cultural background of each country should be fully considered. The total agricultural output value of China (ACGDP), the total agricultural output value of ASEAN member states (AGDP), the distance (DIST) between the two countries, and whether there is a common border (BOR) has no significant influence on the export of such products.

Model 5 indicates how China’s gross agricultural output value (ACGDP), the gross agricultural output value (AGDP) of ASEAN member nations, and the per capita carbon emissions of ASEAN member countries relate to the fourth category of food, beverage, and tobacco. China’s exports of this class of goods are significantly boosted by CARBON, which is in line with the predicted sign. Regardless of whether a common language exists, the language (LANG) significantly reduces China’s exports of this class of goods, which defies the predicted trend. The distance (DIST) between the two nations and the existence of a shared border (BOR) has no appreciable impact on China’s exports in this category.

4.2.2. Analysis of the Results of the Trade Inefficiency Model

To analyze the impact of human factors on trade inefficiency, this paper adopts the “one-step method” to conduct a regression analysis, and the results are shown in

Table 8. Estimated results of the trade inefficiency model.

Variant	Model 1		Model 2		Model 3		Model 4		Model 5
	Ratio	t-Value	Ratio	t-Value	Ratio	t-Value	Ratio	t-Value	Ratio	t-Value
TRA	−0.021	−1.055	−0.160 ***	−5.563	−0.035 **	−2.080	−0.163	−1.632	−0.032	−1.275
SHP	−0.073 ***	−10.361	−0.104 ***	−7.049	−0.079 ***	−9.410	−0.256 ***	−3.597	−0.091 ***	−9.717
FIN	−0.006	−0.555	0.048	1.015	−0.012	−0.696	−0.214 **	−2.414	0.082 ***	3.800
INV	0.022 **	2.210	0.054	1.504	0.028 **	2.030	0.246 ***	3.232	0.042 ***	2.638
GOV	0.015	0.847	0.096 ***	12.265	0.059 ***	3.454	0.164 **	2.273	−0.080 *	−1.768

“*”, “**”, and “***” indicate significance at 10%, 5%, and 1% levels, respectively.

. (1) Trade Freedom TRA: The effect of Trade Freedom TRA on total agricultural exports, animal and vegetable oils and fats, food and beverages, and tobacco exports is not significant, while it has a significant and negative effect on animal and plant exports, which is in line with the expectation, suggesting that the higher the degree of trade openness of ASEAN member countries, the more it helps to improve the trade efficiency of these three types of products. (2) Liner Shipping Connectivity Index SHP: Liner Shipping Connectivity Index SHP passed the significance level test and is negative, consistent with expectations, indicating that ASEAN member countries have better shipping facilities and more robust shipping capacity has a significant positive impact on improving trade efficiency, which contributes to the export of China’s agricultural products. (3) Financial Freedom FIN: Financial Freedom FIN has no significant effect on total agricultural exports, animal products, and plant products, and has a significant and positive effect on food and beverages and tobacco, indicating that the more open the financial sector of ASEAN member countries, the more limited the efficiency of China’s exports of food and beverages and tobacco will be, and the effect on animal and plant oils and fats is negative, indicating that it helps to improve the efficiency of the trade in animal and plant oils and fats exports. (4) Investment Freedom INV: The impact of investment freedom INV on China’s export of animal products is insignificant. In contrast, the impact on total agricultural exports, plant products, animal and vegetable fats and oils, and food, beverages, and tobacco exports is significant and positive, which is in line with expectations, suggesting that the more liberalized investment environment of ASEAN member countries will instead limit the efficiency of the trade of these three types of products. (5) Government Expenditure Level GOV: The government expenditure level of ASEAN member countries has no significant effect on the total export of agricultural products. It has a significant and negative effect on China’s export of food beverages and tobacco, indicating that its government expenditure level helps to improve the efficiency of the export trade of food beverages and tobacco. In contrast, it has a significant and positive effect on China’s export of animal products, plant products, and animal and plant fats and oils, which is in line with expectations, suggesting that the higher level of government expenditure of ASEAN countries, the more they can fully utilize the domestic resource endowment and geographic advantages of member countries to meet their own needs and therefore will reduce imports of such agricultural products to China, thus reducing trade efficiency.

4.2.3. Export Efficiency and Potential Analysis

(1)

Trade efficiency analysis

Table 9. Efficiency of China’s segmented agricultural exports to ASEAN member countries.

Nations	Total Agricultural Exports			Animal Products			Plant Products
	2007	2014	2020	2007	2014	2020	2007	2014	2020
Myanmar (or Burma)	0.19	0.32	0.32	0.93	0.10	0.11	0.11	0.48	0.40
Cambodia	0.09	0.10	0.07	0.00	0.10	0.03	0.03	0.09	0.03
Indonesia	0.85	0.59	0.52	0.13	0.15	0.06	0.06	0.61	0.69
Malaysia	0.98	0.95	0.97	0.56	0.91	0.22	0.22	0.90	0.95
The Philippines	0.90	0.80	0.90	0.75	0.67	0.49	0.49	0.50	0.89
Singapore	0.95	0.99	0.92	0.95	0.96	0.24	0.24	0.94	0.97
Thailand	0.39	0.78	0.90	0.38	0.71	0.52	0.52	0.79	0.83
Vietnam	0.44	0.78	0.78	0.70	0.67	0.56	0.56	0.89	0.87
Nations	Animal and Vegetable Fats and Oils					Food and Beverages and Tobacco
	2007		2014	2020		2007	2014		2020
Myanmar (or Burma)	0.53		0.06	0.43		0.89	0.85		0.82
Cambodia	0.13		0.03	0.05		0.22	0.11		0.10
Indonesia	0.37		0.77	0.70		0.72	0.83		0.51
Malaysia	0.69		0.68	0.87		0.95	0.89		0.96
The Philippines	0.47		0.51	0.62		0.87	0.92		0.79
Singapore	0.59		0.73	0.87		0.95	0.94		0.86
Thailand	0.44		0.63	0.66		0.46	0.47		0.75
Vietnam	0.65		0.57	0.50		0.68	0.95		0.83

displays the trade efficiency of China’s segmented agricultural exports to ASEAN countries, which is further measured based on the results of the one-step technique of the stochastic frontier gravity model. Better export efficiency is indicated by a larger value for export efficiency, which takes values between 0 and 1.

From the perspective of the total export of agricultural products, the efficiency of China’s agricultural export trade with ASEAN members is extremely unbalanced. China’s export efficiency to Malaysia and Singapore ranks first and has remained above 0.9, which is due to the strong trade complementarity between China and Singapore [24]. The efficiency of China’s exports to the Philippines first declined and then rebounded, while the efficiency of exports to Thailand and Myanmar steadily increased. On the contrary, China’s agricultural export efficiency to Indonesia continues to decline, and its trade efficiency to Myanmar and Cambodia is always low, which indicates that China has not effectively developed its agricultural export market [25]. China should take advantage of its resource endowment and use advanced agricultural technology to enhance the added value of agricultural products, so as to further improve its trade efficiency with these countries.

From the animal product market perspective, the trade efficiency of China’s exports of such agricultural products to Singapore and Myanmar declined rapidly. It rose sharply for Malaysia and Thailand before falling. While the trade efficiency of Cambodia and Indonesia has been at a low level, in 2020, China’s trade efficiency of animal product exports to the Philippines, Thailand, and Vietnam is the highest, which indicates that China has good trade conditions with these countries [24]. In the plant products market, China’s export efficiency of plant products to Malaysia, Singapore, Indonesia, the Philippines, and Thailand has been growing, with Malaysia and Singapore reaching more than 0.9, followed by Thailand and the Philippines. China’s trade efficiency of such products to Vietnam rose first and then fell, while the trade efficiency of plant products to Cambodia has always been low. In the animal and vegetable oil market, the export efficiency of animal and vegetable oil from China to Thailand, the Philippines, and Singapore is increasing. The efficiency of China’s exports of such agricultural products to Indonesia declined after growth, while in contrast with Malaysia, exports of such products declined slightly and then recovered, which indicates that China can rely on the free trade area to strengthen bilateral trade cooperation to enhance the comparative advantage of agricultural products. In the food and beverage and tobacco market, the trade efficiency of China’s export of such agricultural products to Malaysia and Singapore has always been at a high level, and the export efficiency of Thailand has proliferated. The efficiency of China’s food and beverages and tobacco exports to the Philippines and Myanmar was high. In contrast, the efficiency of such agricultural exports to Indonesia and Vietnam increased and then declined.

In terms of countries, China’s export efficiency to Myanmar is relatively high in the food, beverage, and tobacco market, and to Indonesia in the plant products, animal and vegetable oils, food and beverages, and tobacco market, with export efficiency of 0.5–0.7. Due to their geographical advantages and economic structure characteristics, Malaysia and Singapore have strong complementarity with China’s agricultural products trade [24], so they have high export efficiency in these four product markets. In contrast, China’s export efficiency of these four agricultural products to the Philippines has decreased compared with Malaysia and Singapore, but it remains high. China’s export efficiency of animal products, plant products, food, beverages, and tobacco to Thailand and Vietnam has increased rapidly, indicating that its export of agricultural products has reached a particular scale. However, it can continue to increase the added value of agricultural products to achieve further growth. In contrast, China’s export efficiency of these four products to Cambodia is always low [25], indicating that it can combine its market demand. We will strengthen cooperation among countries and promote unimpeded trade in agricultural products.

(2)

Analysis of trade potential

The greatest value of China’s agricultural export commerce with ASEAN countries, free from any obstacles, is the export potential measured in this article. This study uses the formula “export potential = actual export value/export efficiency” to calculate China’s export potential to ASEAN countries in terms of segmented agricultural products between 2007 and 2020.

Table 10. China’s export potential and expandable space for segmented agricultural products to ASEAN member countries (millions of dollars).

Nations	Export Potential and Scope for Expansion	Total Agricultural Exports			Animal Products			Plant Products
		2007	2014	2020	2007	2014	2020	2007	2014	2020
Myanmar (or Burma)	Export potential	360.84	1264.96	2290.81	14.21	112.31	261.64	134.95	568.54	1111.88
	Expandable space	293.63	854.98	1547.40	0.94	101.13	232.79	120.07	295.98	672.10
Cambodia	Export potential	211.74	763.09	1665.57	24.37	206.45	589.90	38.72	221.10	542.55
	Expandable space	191.85	688.42	1554.29	24.37	185.88	570.28	37.43	200.92	526.74
Indonesia	Export potential	1024.60	3061.10	4354.81	145.55	1014.24	1857.85	10,858.40	1361.77	1952.58
	Expandable space	158.45	1253.30	2106.61	127.01	862.02	1749.13	10,222.99	525.24	603.29
Malaysia	Export potential	1058.34	2831.87	3571.98	128.42	819.93	1269.02	2649.08	1274.88	1636.64
	Expandable space	23.26	138.89	105.69	56.00	76.58	989.89	2066.39	124.32	88.12
The Philippines	Export potential	598.20	1781.48	2531.74	81.17	558.53	1011.74	585.32	692.95	1008.26
	Expandable space	60.73	352.26	258.30	20.70	181.68	516.77	298.97	343.19	106.68
Singapore	Export potential	355.60	961.07	1134.56	33.11	198.12	291.50	494.65	286.46	400.29
	Expandable space	17.89	11.04	90.86	1.80	7.25	221.79	376.36	16.30	11.87
Thailand	Export potential	1279.67	3561.03	4700.96	179.12	1145.53	1926.57	433.57	1672.71	2136.31
	Expandable space	783.02	795.32	478.91	111.29	327.07	915.53	206.04	354.25	352.87
Vietnam	Export potential	1024.71	3737.67	7022.05	42.30	363.39	893.56	584.37	2364.06	4704.62
	Expandable space	573.55	816.15	1579.03	12.60	118.97	396.25	259.14	255.80	599.09
Nations	Export Potential and Scope for Expansion	Animal and Vegetable Fats and Oils				Food and Beverages and Tobacco
		2007		2014	2020	2007		2014	2020
Myanmar (or Burma)	Export potential	0.69		1.03	3.84	43.29		148.82	332.88
	Expandable space	0.33		0.97	2.18	4.61		22.64	59.77
Cambodia	Export potential	0.50		1.09	4.28	83.18		303.43	785.47
	Expandable space	0.43		1.07	4.06	64.65		269.54	709.84
Indonesia	Export potential	4.20		16.83	16.97	291.76		965.83	1518.84
	Expandable space	2.64		3.92	5.07	81.13		159.69	740.53
Malaysia	Export potential	22.74		24.71	151.73	381.91		881.82	1575.31
	Expandable space	6.99		8.02	20.48	17.70		99.44	67.91
The Philippines	Export potential	1.96		4.36	6.87	219.03		760.31	1108.15
	Expandable space	1.04		2.13	2.62	29.28		59.93	235.50
Singapore	Export potential	20.41		27.60	107.95	186.04		497.40	573.66
	Expandable space	8.28		7.44	14.46	10.06		28.55	81.57
Thailand	Export potential	11.59		19.97	25.20	426.90		1311.20	1885.38
	Expandable space	6.47		7.45	8.58	230.74		694.92	474.43
Vietnam	Export potential	9.42		10.53	20.67	132.15		594.76	1002.14
	Expandable space	3.34		4.55	10.37	41.99		31.91	172.27

presents the results of a calculation of the potential for development of segmented agricultural exports (trade potential–actual trade value) for the years 2007, 2014, and 2020. This will help to clarify with which partner countries China has more room for development in segmented agricultural exports.

With the rapid development of economic globalization, the export potential of four types of agricultural products of China and ASEAN member states from 2007 to 2020 grew, and the space for expansion also expanded. In the animal products market, China’s export potential to Thailand and Indonesia was expected to be the largest in 2020, and the actual trade volume to Indonesia was expected to be low, so the space for expansion was expected to be the largest. The second is China’s export potential for animal products from Malaysia and the Philippines. In the plant products market, the countries with high export potential in 2020 were Vietnam, Thailand, Indonesia, Malaysia, and Myanmar. Among them, China was found to have the most room for expansion in Myanmar. In the animal and vegetable oil market, China’s export potential to Malaysia and Singapore was the highest in 2020. The country with the most room for expansion was Malaysia. In the food, beverage, and tobacco market, the countries with the highest export potential in 2020 were Thailand, Malaysia, Indonesia, the Philippines, and Vietnam. The countries with the most room for expansion were Indonesia and Cambodia. In general, among ASEAN member states, China’s export potential to Vietnam, Thailand, and Indonesia is relatively large [9], while its expansion space to Indonesia is relatively large, which indicates that China can increase its export of agricultural products and release the trade potential of agricultural exports [8]. In addition, the actual trade volume of China’s exports to Thailand continues to increase. The growth rate is faster than the growth of the export potential to Thailand, so the expansion space has declined. In contrast, the export potential and the expansion space for Cambodia continue to rise, indicating that the growth rate of the actual trade volume of China’s export of agricultural products to Cambodia is lower than the growth of the export potential. Cambodia’s contribution to China’s agricultural exports is limited, so China should increase the export of agricultural products to this country and pay more attention to exports to other countries [9].

5. Conclusions and Discussion

Based on panel data of China’s agricultural export trade to eight ASEAN countries from 2007 to 2020, this study empirically examines the spatiotemporal evolution and product differences of China’s agricultural exports to ASEAN countries by measuring the efficiency and potential of China’s agricultural exports. The study employs a stochastic frontier gravity model to analyze the core factors affecting China’s agricultural exports. The results reveal that China’s agricultural exports to ASEAN countries mainly include plant products, food and beverages, and tobacco, with animal products mainly exported to Thailand, plant products mainly exported to Vietnam, and animal and vegetable oils, fats, food and beverages, and tobacco mainly exported to Malaysia.

Further analysis shows that China’s agricultural gross domestic product (ACGDP) significantly impacts agricultural exports to ASEAN countries, confirming Hypothesis 1. A shared border (BOR) has varying effects on different product categories, generally positively impacting some categories while having insignificant or adverse effects on others, partially confirming Hypothesis 2. Conversely, the presence of a common language (LANG) significantly negatively impacts agricultural exports, contrary to conventional expectations. The possible reason for this is that using a common language sometimes translates into effective marketing and communication strategies. Differences in marketing channels, promotional methods, and packaging may hinder the acceptance of agricultural products in the target market, reflecting the complex role of cultural factors in agricultural trade and thus confirming Hypothesis 3. ASEAN countries’ agricultural gross domestic product (GDP) also positively impacts China’s agricultural exports, indicating that the increase in ASEAN countries’ agricultural production drives the demand for Chinese agricultural products, verifying Hypothesis 4. The indicators of economic policy openness (such as the Shipping Connectivity Index (SHP), Financial Freedom (FIN), Trade Freedom (TRA), Investment Freedom (INV), and Government Expenditure (GOV)) have varying impacts, both positive and negative. It is worth noting that while geographical distance typically increases transportation costs and transit time, the results are the opposite. The positive impact of economic scale and market demand on agricultural exports may outweigh the negative impact of distance. Large economies with high demand levels can offset the costs associated with distance and cultural differences, supporting Hypothesis 5.

Furthermore, China’s agricultural exports to ASEAN countries exhibit significant differences across different product categories (such as animal products, plant products, animal and vegetable oils, food and beverages, and tobacco). This shows a distinct segmented commodity structure, verifying Hypothesis 6. ASEAN countries’ per capita carbon emissions (CARBON) have a significant positive impact on China’s exports of different categories of agricultural products, indicating that ASEAN countries’ transition to a low-carbon economy has changed demand preferences and affected the structure of China’s agricultural exports, thus confirming Hypothesis 7.

This study makes marginal contributions in several aspects. Firstly, it introduces ASEAN countries’ per capita carbon emissions as an explanatory variable into the stochastic frontier gravity model, enriching the literature by analyzing the impact of the low-carbon economic transition on China’s agricultural exports. Secondly, segmented product analysis reveals differences in export efficiency and potential across various agricultural product categories, providing more targeted policy recommendations. Thirdly, incorporating low-carbon economic factors offers a new perspective on understanding changes in the agricultural trade pattern against the backdrop of global climate change. Lastly, panel data spanning a long time horizon provide empirical evidence for the efficiency and potential of China’s agricultural exports to ASEAN, facilitating deeper cooperation between China and ASEAN in the agricultural sector.

Despite our achievements, this study also has limitations. Due to data unavailability, Brunei and Laos were excluded from the sample, potentially affecting the representativeness of the findings. Future research should obtain more comprehensive data to improve sample coverage. Additionally, this study focuses primarily on static analyses of export efficiency and potential; dynamic mechanisms, such as changes in trade network structures and market demand fluctuations, could be further explored to gain a more holistic understanding of agricultural export dynamics.

6. Policy Recommendations

First, investment in agriculture should be increased, and agricultural production technology and efficiency should be enhanced. To further boost China’s agricultural exports to ASEAN countries, the government should intensify agricultural Investment, mainly supporting the advancement of agricultural production technology and efficiency. Strengthening China’s gross agricultural output will effectively facilitate agricultural export growth, enabling China to secure a larger share of the international market.

Second, leverage geographical advantages and strengthen cross-border trade cooperation with ASEAN countries sharing common borders:

Given the positive influence of shared borders on specific product categories, the government should actively promote cross-border trade cooperation with ASEAN nations that share borders. By harnessing geographical advantages and reducing logistics costs, China can further augment its agricultural exports to these countries, enhancing its competitiveness in the ASEAN market.

Third, conduct in-depth research on cultural differences in ASEAN countries and formulate precise marketing strategies:

In response to the negative impact of shared language on agricultural exports, the government and enterprises should delve deeper into the cultural disparities in ASEAN nations. By crafting targeted marketing strategies, China can effectively circumvent cultural barriers that hinder exports, thereby elevating the acceptance of Chinese agricultural products in the ASEAN market and its expanding market share.

Fourth, strengthen agricultural cooperation and exchanges with ASEAN countries and keep abreast of agricultural development trends:

The growth of ASEAN countries’ gross agricultural output significantly influences China’s agricultural exports. Consequently, the government should foster closer agricultural cooperation and exchange with ASEAN nations, staying informed about ASEAN’s agricultural development trends. This will facilitate adjusting and optimizing China’s agricultural export structure to cater to ASEAN market demands, ultimately enhancing export performance.

Fifth, promote economic policy liberalization in ASEAN countries and reduce trade barriers. Given the mixed effects of economic policy openness in ASEAN nations, the government should actively engage in economic cooperation with ASEAN countries, driving the liberalization of trade, finance, shipping, and other sectors forward. By signing free trade agreements, establishing economic and trade cooperation mechanisms, and other means, China can reduce trade barriers and uncertainties, fostering a more conducive environment for agricultural exports.

Sixth, optimize the export commodity structure based on ASEAN market demands. Acknowledging the distinctive segmented commodity structure characteristics of China’s agricultural exports to ASEAN countries, the government and enterprises should tailor their export commodity mix to ASEAN countries’ specific needs. By augmenting the market share of competitive segmented commodities, China can further consolidate and expand its position in the ASEAN market, enhancing its overall competitiveness.

Seventh, address ASEAN countries’ low-carbon environmental demands and boost exports of low-carbon agricultural products. Amidst the low-carbon economic transformation, ASEAN countries’ varying per capita carbon emissions exert heterogeneous influences on China’s agricultural exports. Consequently, the government and enterprises should heed ASEAN countries’ evolving low-carbon environmental needs, particularly by increasing exports of plant products, animal and vegetable fats and oils, and other low-carbon agricultural products. By satisfying ASEAN countries’ low-carbon demands, China can further enhance the competitiveness of its agricultural products in the ASEAN market, fostering sustainable growth in agricultural exports.