China–U.S. Trade Friction and China’s Agricultural Machinery Imports: Mechanism and Empirical Evidence

Abstract

Based on the monthly panel data of China’s imports of agricultural machinery products from 2016–2022, this paper uses a multi-period double-difference model to assess the impact of China’s imposition of counter-tariffs on China’s imports of agricultural machinery in the context of U.S.–China trade friction. It is found that China’s implementation of counter-tariffs significantly reduces China’s imports of agricultural machinery products from the U.S. and significantly increases imports from 16 other countries, but the trade diversion effect is lower than the trade suppression effect. Mechanism analysis finds that China–U.S. trade friction affects the technological innovation capacity of agricultural machinery enterprises and the degree of uncertainty in the Chinese economy, which in turn affects China’s agricultural machinery imports. Heterogeneity analysis finds that China–U.S. trade friction has a more significant inhibitory effect on seeding and planting and fertilizing machinery, drainage, irrigation, and water lifting machinery, and other machinery imported from the U.S., and a more significant diversionary effect on agricultural primary processing machinery and harvesting machinery imported from 16 other countries. The imposition of countervailing tariffs mainly affected imports of complete machinery products rather than machinery spare parts.

1. Introduction

The No.1 Document of the central government in 2024 pointed out that “it is necessary to strengthen the support of agricultural science and technology”. Agricultural machinery and equipment is an important material foundation for strengthening the support of agricultural science and technology and promoting the comprehensive revitalization of the countryside, and trade in agricultural machinery plays a crucial role in promoting the high-quality development of the agricultural machinery industry. At present, China is striving to promote the construction of an open global economy and is in a critical period of high-quality development and high-level opening to the outside world. However, in recent years, the U.S. government has practiced trade protectionism and unilateralism, and has imposed economic sanctions on China, which not only undermines the economic order, but also further increases the uncertainty faced by China’s foreign trade policy, which brings challenges to China’s economic development. In August 2017, the U.S. launched a Section 301 investigation into China, and has implemented three rounds and five tariff hikes on China since March 2018, involving amounts of USD 50 billion, USD 200 billion, and USD 300 billion, respectively, with a wide range of levies and high tax rates [1]. In response, China implemented tariff countermeasures and imposed tariffs at different rates on imports from the U.S. as a policy countermeasure, involving USD 50 billion, USD 60 billion, and USD 75 billion, respectively. The trade friction between China and the U.S. is highly uncertain and will continue to affect the global economic and trade development patterns for some time to come [2].

In the trade friction, the levy list issued by the U.S. government includes a large number of agricultural machinery products, of which key products such as tractors and harvesting machinery have been subject to tariff rates ranging from 20–25%, and China has therefore implemented tariff countermeasures on agricultural machinery products imported from the United States. According to the statistics of the China Agricultural Machinery Industry Yearbook, the United States is an important source country for China’s imports of agricultural machinery products, and imports of agricultural machinery products from the United States accounted for about 15% of all imports in the past years. The United States is the world’s strongest industrialized country, and China is the world’s largest importer of agricultural machinery. The two countries have a far-reaching impact on the world agricultural machinery trade market. Whether the United States imposed tariffs on China, or China’s implementation of tariff countermeasures against the United States, to a certain extent, may change the structure of imports of agricultural machinery products and market patterns. Therefore, it is necessary to study in depth, in the context of U.S.–China trade friction, China’s agricultural machinery products import structure and market changes, as well as the question of whether the imposition of tariffs on imports of agricultural machinery is effective,, to ensure China’s stable and safe development of the agricultural machinery industry.

At present, there is rich literature on the study of trade friction between China and the United States in the academic community. In terms of theoretical research, Bown et al. (2007) proposed various trade effects that may exist after the occurrence of trade friction, including trade disruption effect, trade recession effect, trade deflection effect and trade diversion effect, by constructing a three-country model [3]. Lv Jianxing et al. (2023) studied the export deflection effect generated by trade friction [4]. Through empirical research, some scholars believe that the trade friction between China and the U.S. mainly suppresses industrial trade through the imposition of tariffs on each other, and the tariffs imposed by the U.S. on China’s exports increase the cost of China’s exports to the U.S. and cut down on the comparative advantage of China’s products exported to the U.S. [5,6]. After China implemented tariff countermeasures, the import tariffs on different products were passed on to varying degrees to related intermediate and final products within the U.S., with U.S. consumers and importers carrying the main tariff costs [7,8]. Many scholars have also focused on the impact of trade frictions on industrial trade [9,10]. On the one hand, the CGE model [11] and the GTAP model [12,13,14] have been used to conduct ex ante simulation analysis of trade friction scenarios and to predict and project the possible impacts of trade frictions on industries, which not only involves the macroeconomics and various sectors of the two countries (i.e., trade frictions have a significant negative impact on the economies of the two countries), but also how trade frictions will affect individual industries, including automotive parts and components [15] and trade in agricultural products [16,17]. As trade frictions intensify and economic uncertainty grows, most industries will be hit by more pronounced negative shocks, with Chinese exports in a variety of sectors declining, leading to a further rise in welfare losses [18,19]. On the other hand, some scholars have also assessed the impact of trade frictions on industrial trade through ex post analysis, arguing that after the U.S. imposes tariffs on China, China’s counter-tariffs reduce the volume of imports from the U.S. as well as the average price of imports, which can significantly inhibit China’s imports of commodities from the U.S., thus forming an effective counterattack against the U.S. tariff increase [20,21].

In addition, the research on agricultural machinery trade mostly focuses on the potential measurement [22] and competitiveness assessment [23] with various countries and lacks the results of the trade friction between China and the United States on China’s agricultural machinery trade. Lian Xiaolu et al. (2007) argued that since China’s accession to the WTO, China’s import and export of agricultural machinery products has shown continuous growth, and the trade market and product structure are relatively stable [24]. Zhang Meng et al. (2015) argued that economies of scale, market size, per capita income level, degree of mechanization, price advantage, and output value of agricultural machinery products can promote trade in the agricultural machinery industry, whereas the degree of differentiation of agricultural machinery products, geographic distance, and the area of arable land per labor in the importing country can hinder intra-industry trade in agricultural machinery [22]. According to Zhang Meng et al. (2016), in recent years, the trade scale of China’s agricultural machinery products has been growing rapidly, and its competitiveness has been improving, which has achieved the optimization of the trade structure to a certain extent; however, the relative export competitiveness of China’s agricultural machinery products has shown a continuous decline [23]. Some scholars have also studied the trade potential of agricultural machinery products between China and RCEP member countries [25] and Belt and Road countries [26]. With the continuous development of the agricultural machinery industry, the export competitiveness of China’s agricultural machinery products has been improving, but the trade of some products still faces structural challenges [27].

In summary, the existing literature focuses mostly on the impact of China–U.S. trade friction on macroeconomics and other industries, but there are few studies on agricultural machinery trade using factual data, and there is almost a gap in the research on the impact of China’s implementation of counter-tariffs on China’s imports of agricultural machinery in the context of trade friction, as well as on exploring how China–U.S. trade friction has altered China’s import pattern of agricultural machinery products. This paper will enrich this part of the research by selecting the monthly panel data of China’s agricultural machinery imports from 2016–2022 as the initial sample and applying the multi-period double-difference method to test the effect of China’s implementation of counter-tariffs against the background of trade friction between China and the United States and to conduct the product heterogeneity analysis, so as to measure the specific impacts of the U.S.–China trade friction on China’s imported agricultural machinery product structure and market. Compared with the established literature, the marginal contribution of this paper may be reflected in several aspects. In terms of research perspective, it may enrich and expand the existing literature on the trade effects of U.S.–China trade friction on industries, focusing on the policy effects of U.S.–China trade friction on China’s imports of different agricultural machinery products. In terms of research content, the ex post policy effects of implementing tariff countermeasures are assessed using the double-difference method, and the possible impact mechanisms as well as the heterogeneity of different products are analyzed, which will be of reference significance for the risk and response to changes in China’s agricultural machinery trade brought about by the Sino–U.S. trade friction in the future, as well as for the issue of how to further adjust the structure of China’s agricultural machinery products, the import market, and the practice of trade policy.

2. Theoretical Mechanisms

The trade friction between China and the United States represents a strategic decision for China and the United States to compete for future economic dominance. As one of the most important macro events in the international economic field in recent years, it has caused huge losses to China’s welfare and will profoundly affect the pattern of the global economic and trade development in the next decade and even longer [2]. On the one hand, the direct result of the tariff increase is to raise the trade cost of China’s agricultural machinery products imported from the U.S., which leads to an increase in the price of imports from the U.S., weakening the competitiveness of U.S. products in China’s market and reducing imports from the U.S. Under open economic conditions, after the imposition of tariffs on agricultural machinery products imported from the United States, the import price and transaction costs of agricultural machinery products rise, causing absolute and relative price changes [28]. Absolute price increases cause a trade inhibition effect. The United States directly reduces imports of agricultural machinery products, and the income effect will lead to a decline in real import revenue in China, resulting in a reduction in the volume of imported products, including different types of agricultural machinery products. There are differences in the income elasticity of imports of different types of agricultural machinery products. Reducing the volume of imports to varying degrees further leads to changes in the structure of the types of agricultural machinery products imports, which also causes a trade inhibition effect. At the same time, a considerable part of the products imported from the United States is high-end intelligent agricultural equipment. Therefore, when the U.S. trade friction on the international supply chain of enterprises increases uncertainty, the higher prices of domestic agricultural enterprises from the U.S. exacerbate the market risk of the domestic agricultural enterprises. Agricultural machinery enterprises will increase their R&D efforts, expand their R&D investment in high-end intelligent “necklace” technologies of agricultural machinery, and reduce their dependence on some high-end agricultural machinery products imported from the U.S. [5].

On the other hand, as a result of the implementation of the tariff policy, the direction of imports of some of our agricultural machinery products has changed, leading to the triggering of indirect effects. At the same time, due to the increase in tariffs, the competitiveness of the United States in our market has been weakened, which may have an impact on the quality of the products imported or lead to an increase in our imports from other countries to compensate for the losses resulting from the cessation of imports from the United States. As a practical matter, it becomes difficult to adjust product quality in the short term, as firms often face difficulties in financing, purchasing high-end equipment, and acquiring advanced technology [29]. As a result, domestic firms facing counter-tariffs may choose to adjust not the production of their products, but their import strategy for their products. Such adjustments include changing the import choices and import sources of products, rather than improving product quality in a timely manner. Specifically, after China imposes tariffs on agricultural machinery products imported from the U.S., the increase in relative prices leads to a reduction in imports from the U.S., and the reduction is shifted to imports from other countries, generating a trade diversion effect. Among other things, the trade diversion effect is widely differentiated at both the country level and the product mix level. Shifting to other major agricultural machinery countries, due to the level of tariff preferences, geographical distance and terrain environment, agricultural output value, and other factors [25], not only leads to differences in the level of economic development, but also, the production and export of advantageous agricultural machinery products are different; thus, there are various differences in trade with China. This shift in agricultural machinery product imports from the United States to other countries will change China’s agricultural machinery products import market structure [30]. However, the drastic uncertainty brought by the U.S.–China trade friction to the global economic development, especially China’s economic development, may also affect the agricultural machinery enterprises, leading consumers to reduce their investment and consumption, and thus reduce the demand for agricultural machinery products imported from other countries in order to reduce the risk [9].

This paper interprets the trade remedies in the study of Carter et al. (2018) as counter-tariffs and mathematically verifies the trade suppression and diversion effects after the imposition of tariffs [31]. The specific derivation process is as follows:

Suppose there are n countries in the world market. Each country has only one agricultural machinery manufacturer, manufacturer i, producing one agricultural machinery product; each manufacturer has relatively independent shares in the domestic and international markets; and all are in a Cournot equilibrium. The Cournot equilibrium is a special case of Nash equilibrium in which each firm cannot increase its profit by changing its own output given the output of its competitors. This equilibrium state indicates that, without external intervention, the vendors have reached a self-implementing stable state.

In this case, it is assumed that manufacturer 1 operates only in the domestic market, while manufacturers 2 and 3 operate in both the domestic and international markets, and manufacturer 4 does not currently trade with the country 1 market. The aggregate supply in the domestic market is $Q_i={\sum }_{i=1}^n{q}_{ji}+q_i$, the total supply in the foreign market is $Q_j={\sum }_{i=1}^n{q}_{ji}+q_j$, and the total output in each country is $q_j^{*}={\sum }_{i=1}^n{q}_{ji}+q_j$. Consumers generally believe that domestically manufactured products and imported products are perfectly substitutable. The technological level among competitors is comparable, and the demand function for each farm machinery manufacturer is $p=Q_i+Y_i$, where $Y_i$ is a possible factor affecting the demand for imports. $c=(q_i^{*},W_i)$ is the cost function of the agricultural machinery manufacturers, and $W_i$ is the cost factor. Assume that marginal costs are strictly convex, i.e., both the first- and second-order derivatives are greater than 0. Assume that the cost of trade between the two countries is only related to tariffs, that the exporters bear the full cost of tariffs ${\tau }_{ji}$, and that the exporter bears the full cost of the tariff. The profit function of each manufacturer is:

$${\Pi }_i=p\left(Q_i,Y_i\right)q_i+p\left(Q_j,Y_j\right)q_{ji}-{\tau }_{ji}{q}_{ji}-c(q_i^{*},W_i)$$

(1)

According to the profit function, profit is the difference between total revenue and total cost. Total revenue includes both domestic and foreign markets, and total cost consists of tariff cost and production cost. According to the principle of profit maximization, each manufacturer chooses the output $q_i$. Assuming that country 1 imposes a tariff on country 2 at this time, the first-order conditions for each country’s market are

$${\displaystyle \frac{\partial {\Pi }_i}{\partial q_i}}=p\left(Q_i,Y_i\right)+p^{\prime }\left(Q_j,Y_j\right)q_i-c^{\prime }\left(q_i^{*},W_i\right)=0$$

(2)

The first-order condition for countries 2, 3, and 4 to export to country 1 to maximize profits is

$${\displaystyle \frac{\partial {\Pi }_{i1}}{\partial q_{i1}}}=p\left(Q_1,Y_1\right)+p^{\prime }\left(Q_1,Y_1\right)q_{i1}-c^{\prime }\left(q_i^{*},W_i\right)=0$$

(3)

The optimal response function of each manufacturer in the domestic market is usually expressed as:

$$q_i=R_i[p\left(Q_i,Y_i\right),c(q_i^{*},W_i\left)\right]$$

(4)

The general form of the response function for foreign markets is:

$$q_{ij}=R_{ij}[p\left(Q_i,Y_i\right),{\tau }_{ji},c(q_i^{*},W_i\left)\right]$$

(5)

The Cournot equilibrium solution for the respective vendor in each country can be obtained by solving the above equation. Since the cost function is strictly convex, vendors 2, 3, and 4 will adjust their sales in the domestic and foreign markets until marginal revenues in the domestic and foreign markets are equal. Under strict assumptions, when the tariff increase is implemented, the vendors must readjust their market share allocations to each other. Among other things, before the imposition of the tariff, vendor 4 did not export to country 1 because the cost of exporting was too high, but after the imposition of the tariff by country 1 on country 2, the price of country 1’s products increases until the price increases to the point where vendor 4 can earn a positive profit from exporting, and then vendor 4 enters the market to export to country 1 and competes with the rest of the country, i.e., there is a trade-creation effect. Under the combined effect of these effects, the policy effect of China–U.S. trade friction on China’s imports of agricultural machinery products is studied.

Based on the above analysis, this paper presents:

Hypothesis 1 (H1).

Under the trade friction between China and the United States, China’s implementation of counter-tariffs will significantly inhibit China’s imports of agricultural machinery products from the United States. There is a trade inhibition effect. a significant increase in imports of agricultural machinery products from other countries to make up for the reduction in imports from the United States. There is a trade diversion effect, but there are differences in imports of different products from different countries.

Hypothesis 2 (H2).

China–U.S. trade frictions may affect China’s agricultural machinery imports through two mechanisms: the technological innovation capacity of Chinese agricultural machinery firms and China’s economic uncertainty.

3. Typical Facts

In recent years, China has been expanding its opening to the outside world, and China’s agricultural machinery trade has realized rapid development. The high demand for agricultural machinery has led to a rapid expansion of the scale of imports, which has pushed China to become the world’s largest agricultural machinery manufacturer as well as the world’s largest importer. The United States has been the third-largest importer of agricultural machinery products in China, and the space for economic and trade cooperation between China and the United States has been expanding. China’s imports of agricultural machinery products from the United States have continued to rise, from 5.04 trillion U.S. dollars in 2003 to 20.65 trillion U.S. dollars in 2017, an increase of up to 309.72%. Considering that “COVID-19” has affected the international trade of a variety of products to a certain extent, this paper mainly analyzes the trend of change until 2020.

Since the trade friction occurred, with the United States of America continuing to impose tariff sanctions on China’s agricultural machinery products and China imposing counter-tariffs on imports from the United States. agricultural machinery product imports from the United States declined from 20.65 trillion U.S. dollars in 2017 to 19.52 trillion U.S. dollars in 2019, a decline of 5.47%. During the same period, China’s imports of agricultural machinery products from the world fell by 4.8%, indicating that China and the United States imposing these tariffs, in the overall scale of imports, has had a limited impact. At the same time, imports from other major agricultural machinery countries have risen significantly. China imported 26.43 trillion U.S. dollars’ worth of agricultural machinery from Germany in 2019, an increase of 6.43% compared with 2017, and imported 32.92 trillion U.S. dollars’ worth of agricultural machinery from Japan, an increase of 3.85% compared with 2017. The details are shown in Figure 1. Figure 1 shows the major market distribution of China’s agricultural machinery imports before and after the trade friction. After 2020, the growth trend of China’s imports of agricultural machinery products from the U.S. as well as from other countries did not show any significant change due to the impact of the new crown pneumonia epidemic. As a result, China is not only an important importer of agricultural machinery in the world, but also an important trading partner of several large agricultural machinery exporting countries such as Germany, the United States, and Japan.

China’s main imports of agricultural machinery include harvesting machinery, irrigation, drainage and well-drilling machinery (including all kinds of pumps and their parts), internal combustion engines, and all kinds of agricultural machinery parts. Since the occurrence of trade friction, and the trade diversion effect due to the existence of different products, the agricultural machinery products import structure and market have changed. Several types of product changes are worth paying attention to. First of all, trade friction has had less impact on harvesting machinery less impact. For China’s imports of harvesting machinery, Germany, the United States, and Japan are the three major countries of origin, accounting for about 70% of China’s total imports of harvesting machinery. Since the occurrence of trade friction, imports from the U.S. of harvesting machinery such as combine harvesters and their parts, other unlisted harvesters, and other products decreased by up to 20% in 2018–2019. Imports from other countries increased by 19.8% over the same period, of which imports from Germany increased significantly, with a significant transfer effect. However, more than 90% of cotton pickers in harvesting machinery rely on imports from the United States, which increased from USD 1.90 trillion in 2018 to USD 2.11 trillion in 2019, an increase of 11.16%. Possible reasons for the increase, rather than decrease, in imports of such products from the U.S. is that the cotton picking machine import trade market is single, the product is inelastic, less affected by price, more dependent on the U.S. export market, and more import tariffs are borne by domestic agricultural machinery enterprises. Secondly, the trade friction significantly affected the main two categories of machinery, respectively, irrigation, water supply, and well drilling machinery and seeding, planting, and fertilizer machinery. Among them, the main import source countries of drainage, irrigation, and well-drilling machinery are South Korea, Japan, Germany and the United States. Sowing, planting, and fertilization machinery import source countries are mainly Germany, Italy, Japan, and the United States. Since the trade friction, China’s imports of agricultural machinery products partially shifted to these countries; China’s imports of these two types of agricultural machinery products from these countries increased by more than 30%. Although the United States is one of the main sources of imports of these machines, the import source countries are rich and diverse, and China can seek alternatives in other countries. In addition, the impact on internal combustion engines and their parts is very obvious. Their main source countries are the U.S., Japan, and the U.K. In 2018, China’s imported internal combustion engines and their parts trade amounted to USD 72.58 trillion, while the imported trade from the U.S. amounted to USD 21.12 trillion, accounting for 29.10% of the total imports of internal combustion engines. And in 2019, China’s trade in imported internal combustion engines and their parts amounted to USD 67.48 trillion, a decrease of up to 7.56% year-on-year, while the imported trade from the U.S. amounted to 18.37 trillion dollars, down as much as 14.97% year-on-year, and the year-on-year increase in imports from other countries was 7.41%. The internal combustion engine produced by the United States is an important seller that cannot be replaced by China’s imports of internal combustion engines in the coming period. Despite the trade diversion effect, it could not offset the decrease from the United States.

Along with this development process, the trade friction between China and the United States has had a significant impact on the changes in the structure and market of agricultural machinery imports. At present, agricultural machinery and equipment is an important material and equipment basis for China to realize food security, hiding food in the ground and food in technology. The agricultural machinery trade has also become an important support to guarantee the development of the agricultural machinery industry under the established resource constraints. Therefore, it is necessary to further study in depth the far-reaching impact of trade friction on agricultural machinery imports and even the development of the entire agricultural machinery industry.

4. Model Setting and Data Processing

4.1. Modeling

The double-difference method (DID), as one of the methods of policy effect assessment, has been widely used by scholars in recent years [32,33]. Quasi-natural experiment is an experimental condition formed in a natural state due to some external shock or policy change, which can be used to test economic theory or evaluate the effect of policy. In this kind of experiment, researchers use actual events as experimental conditions to analyze the effects of these events by comparing the reactions of different groups or regions. In particular, China–U.S. trade frictions involve a wide range of products, covering almost the vast majority of China–U.S. trade. This wide coverage means that it is difficult for the power of a certain industry or a specific enterprise to interfere with the policy-making of both sides, thus making the tariff conflict an exogenous policy impact, which is very suitable for studying the impact on China’s agricultural machinery import trade [20]. Referring to the study of Zhang Zhengyu et al. (2024), this paper takes the U.S.–China trade friction as a quasi-natural experiment and adopts the double-difference model to assess the policy effect of China’s implementation of taxing countermeasures on China’s imports of agricultural machinery in the context of the U.S.–China trade friction [34]. The basic form of the model is as follows.

$${lnimort}_{i,t}={\alpha }_0+{\beta }_{0}DID+{Control}_{i,t}+{\mu }_i+{\phi }_t+{\epsilon }_{it}$$

(6)

Among them, $i$ represents the HS eight-digit code imported agricultural machinery products, and $t$ represents the month. ${lnimort}_{i,t}$ is a general term representing the explanatory variables and denoting the logarithmic value of China’s monthly imports of agricultural machinery products from different countries. $DID$ is the ${Treat}_i\times {Post}_{i,t}$ an interaction term, a core explanatory variable, used to estimate the impact of the policy; ${Treat}_i$ is defined as the group dummy variable for China’s implementation of counter-tariff increases. ${Post}_{i,t}$ represents the time dummy variable for China’s implementation of the counter-tariff increase. The coefficient ${\beta }_0$ is the policy effect generated by China’s implementation of tariff countermeasures on U.S. agricultural machinery products and incorporates the trade creation effect into the trade diversion effect. ${Control}_{i,t}$ is a collection of control variables. ${\mu }_i$ represents individual fixed effects, and ${\phi }_t$ represents month fixed effects, each of which is used to control for unobservable factors that vary with product and time. ${\epsilon }_{it}$ is a random error term, which may be region or time dependent. The double-difference model mitigates omitted variable bias by controlling for time and product fixed effects. In addition, all indicators except the dummy variables take logarithmic form to reduce heteroskedasticity.

4.2. Selection of Variables

4.2.1. Explanatory Variable

The explanatory variable is the logarithmic value of China’s monthly imports of agricultural machinery products with HS eight-digit codes from 2016 to 2022 [25], and the selected importing countries are the United States, Germany, Japan, Italy, the Netherlands, France, the United Kingdom, the Czech Republic, Sweden, Austria, Mexico, Hungary, Belgium, Turkey, India, Brazil, and South Korea. These countries are Amercia and the other top 16 importers of agricultural machinery products in China in 2022, accounting for about 80% of China’s import market share of agricultural machinery products in that year, which is more fully representative.

4.2.2. Core Explanatory Variables

${Treat}_i\times {Post}_{i,t}$ is the cross term of the policy and time dummy variables; i.e., it is the core explanatory variable, and its coefficient ${\beta }_0$ is the policy effect that this paper focuses on. The sample time span used in this paper is the monthly data on imports of agricultural machinery products from January 2016 to December 2022 for the analysis of the implementation of tariff countermeasures. It should be noted in particular that China’s implementation of tariff countermeasures on relevant agricultural machinery products is embodied in the list of tariff countermeasure products published by the Tariff Commission of the State Council, with a total of 100 HS eight-digit-coded agricultural machinery products involved in the levy. The tariff countermeasures were implemented on 35 agricultural machinery products in September 2018, and the implementation of tariff countermeasures on the remaining products is scheduled for June 2019, Cf. Wang et al. (2022) [35] and Amiti et al. (2020) [36] studies, the time of implementing tariff countermeasures in batches is determined as September 2018 and June 2019. After data processing, this paper involves 63 agricultural machinery products imported by HS eight-digit code, and the experimental group includes 24 items, of which 10 agricultural machinery products are levied with tariffs in September 2018, and 14 agricultural machinery products are levied with tariffs in June 2019. The control group includes 39 items.

4.2.3. Control Variable

With reference to the “natural trading partner” hypothesis and existing literature, the selected control variables include exchange rate, agricultural output share, per capita GDP, and population [37,38,39]. The exchange rate data in this article represent the prices of domestic currencies and are calculated based on the monthly average exchange rate data provided by the State Administration of Foreign Exchange of the People’s Republic of China. Exchange rate changes will directly affect the export and import of a country, affect the production and operation of enterprises, affect international investment and cross-border mergers and acquisitions, and affect the structure and pattern of international trade. If the dollar depreciates, China’s export volume of agricultural machinery products increases, and imports decrease. On the contrary, the export volume of China’s agricultural machinery products decreases, and the import volume increases. The proportion of agricultural output represents the strength of a country’s agricultural economy. The influence of the proportion of agricultural output on the agricultural machinery trade is mainly reflected in the change of total factor productivity, technological progress, and the degree of openness of the agricultural machinery products market. If the proportion is higher, it means that the agricultural economic strength is stronger, and the demand for agricultural machinery is higher. GDP per capita and population are annual figures that measure the level of economic growth of a country. Regarding per capita GDP, the stronger the economic strength of the country, the higher the degree of agricultural mechanization may be, and the higher the demand for agricultural machinery. Population represents labor and also affects agricultural machinery.

In addition, in rare cases where data is missing in the control variable, linear interpolation is used to supplement the data.

4.3. Data Description

The time span of this paper’s study is January 2016–December 2022, involving monthly data on China’s customs HS eight-digit code imports of agricultural machinery products, including China’s trade in agricultural machinery products imported from the United States and 16 other countries. The exchange rate data are from the State Administration of Foreign Exchange of China, and the data of other control variables are from the State Administration of Foreign Exchange of China and the World Bank. Through correlation analysis of the data, the absolute value of the correlation coefficient between the variables is low, and no multicollinearity problem is found.

Table 1. Description of variables and data sources.

Variable Type	Variable Symbol	Variable Description	Data Sources
explanatory variable	lnimport1	China’s imports of agricultural machinery products from the United States	Customs statistics
	lnimport2	China’s imports of agricultural machinery products from 16 countries
explanatory variable	Treat	group dummy variable	State Council Tariff Commission (PRC)
	Post	time dummy variable
	Treat × Post	policy effect
	reer	Real effective exchange rate of the renminbi	State Administration of Foreign Exchange (SAFE) of China
	pgdp	GDP per capita	World Bank
	pop	demographic	World Bank
	agr	Share of agricultural output in importing countries	World Bank

illustrates the specific data.

Table 2. Descriptive statistics of variables.

Variant	Sample Size	Average Value	(Statistics) Standard Deviation	Minimum Value	Maximum Values	Upper Quartile
lnimport1	4754	12.53	2.19	2.48	18.36	12.70
lnimport2	5197	14.15	14.15	1.39	18.95	14.40
lnreer	9951	6.51	0.03	6.45	6.57	6.51
lnagr	9951	2.00	0.04	1.95	2.09	2.03
lnpop	9951	19.61	0.01	19.58	19.61	19.61
lnpgdp	9951	9.24	0.16	9.00	9.45	9.22

shows descriptive statistics of variables.

5. Empirical Analysis and Testing

5.1. Analysis of Baseline Regression Results

The results of the benchmark regression are shown in

Table 3. Trade effects of imports of agricultural machinery products after China’s implementation of countervailing tariffs.

Variant	(1)	(2)	(3)	(4)
	Lnimport1	Lnimport1	Lnimport2	Lnimport2
DID	−1.792 *** (−0.212)	−1.790 *** (0.215)	0.371 *** (0.084)	0.370 *** (0.084)
control variable	uncontrolled	containment	uncontrolled	containment
constant term (math.)	12.19 *** (0.183)	9.805 *** (1.886)	5.682 *** (0.153)	6.036 *** (1.146)
sample size	4754	4754	5192	5192
R2	0.160	0.169	0.069	0.069
product fixed effect	containment	containment	containment	containment
time fixed effect	containment	containment	containment	containment

Note: *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space. Same table below.

, and the regression results are all clustered using the HS eight-digit code product-level clustering standard errors. China’s imports of agricultural machinery products from the U.S. are shown in columns (1) and (2), and China’s imports of agricultural machinery products from the other 16 countries are shown in columns (3) and (4). Columns (1) and (3) include only the core explanatory variables and no control variables. The core estimated coefficients for both imports from the U.S. and imports from the other 16 countries are both significant at the significance level of 1%, which suggests that the tariff countermeasures implemented by China not only significantly reduced China’s imports of agricultural machinery products from the U.S., but also significantly increased China’s imports of agricultural machinery products from the other 16 countries. The effect of tariff counteraction policy implemented by China against the United States in terms of imported agricultural machinery products is significant, which is manifested in an obvious trade suppression effect and trade diversion effect. In order to eliminate the interference of unconsidered variables on the results, the corresponding data are obtained in columns (2) and (4) of the table after the introduction of control variables. The core estimated coefficients of imports from the U.S. and imports from the other 16 countries, although changed, are still significantly negative at the 1% level, which indicates that the regression results are relatively robust. By comparing the coefficients, it can be seen that after the implementation of the policy, although China significantly increased imports of agricultural machinery products from the other 16 countries, this did not offset the decline in imports of agricultural machinery products from the U.S. The increase in imports of agricultural machinery products from the other 16 countries is smaller than the decline in imports of agricultural machinery products from the U.S. That is to say, the positive trade diversion effect is smaller than the negative trade inhibition effect, and the total trade effect is negative. Hypothesis H1 is verified.

5.2. Parallel Trend Hypothesis Testing

An important basic assumption of the $DID$ model is the parallel trend assumption; i.e., imports of agricultural machinery products without tariffs and with tariffs should have the same trend before the imposition of countervailing tariffs. In this paper, the following model is developed using the event study method proposed by Jacobson et al. (2005) [40]:

$${lnimort}_{i,t}=\alpha +{\sum }_{t=-5}^{t=5}{\delta }_t{D}_{it}+\beta {Control}_{i,t}+{\mu }_i+{\phi }_t+{\epsilon }_{it}$$

(7)

where $D_{it}$ is a set of dummy variables, and the meaning of the remaining variable symbols is consistent with the baseline model. In this paper, the data of the five months before the policy implementation are combined into period −5, while the data of the five years after the policy implementation are combined into period 5, and the data of the period before the policy comes into effect are excluded as the reference group, in order to avoid the problem of full covariance.

Figure 2 reports the results of the parallel trend test of China’s imports of agricultural machinery products from the U.S. Its 95% confidence interval before the implementation of China’s countermeasure tariffs includes a value of 0, indicating that the coefficient is not significantly different from 0 at the 5% statistical level, which means that the treatment and control groups are more consistent with the trend of changes in the size of the trade before the policy shocks, and there is no significant difference. After the imposition of countermeasure tariffs, the coefficient shows a significant downward trend and is significantly negative, indicating that China’s tariff countermeasures significantly reduced China’s imports from the U.S. of agricultural machinery products. The research sample passes the hypothesis test of parallel trends.

Figure 3 reports the results of the parallel trend test of China’s imports of agricultural machinery products from the other 16 countries. Before China implemented counter-tariffs, the coefficient was close to 0, indicating that there was no significant difference between the treatment group and the control group. After China’s implementation of counter-tariffs, the coefficient obviously showed an upward trend, and the coefficient was significantly positive in each period, which indicates that China’s implementation of tariff counteraction on U.S. imports of agricultural machinery products significantly increased China’s imports of agricultural machinery products from the other 16 countries. The fluctuation may have occurred because of the instability of the global agricultural machinery trade market after the trade friction. The sample data on China’s imports of agricultural machinery from 16 other countries passed the parallel trend hypothesis test.

5.3. Placebo Test

In order to verify that the above empirical results are valid and not the result of random selection, and to avoid the influence of unobserved factors on the results, this part first conducts a placebo test at the time level according to Yan Bing et al. (2021), i.e., replacing the time of the policy shock and adjusting the sample data interval [41]. Setting the time of the policy shock at a period before the actual policy shock time and deleting the samples in the year when the policy time actually occurs, as well as in the months after it, is a way to examine whether there is still a trade effect. The sample time is set from January 2017 to June 2019 by advancing the implementation of tariff countermeasures by one year.

Table 4. Placebo test results.

Variant	(1)	(2)
	Lnimport1	Lnimport2
DID	−0.112 (0.196)	0.336 (0.279)
constant term (math.)	15.970 *** (1.512)	4.491 *** (0.657)
sample size	1752	1890
R2	0.025	0.014
product fixed effect	containment	containment
time fixed effect	containment	containment

Note: *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

shows the regression results after replacing the policy time. Columns (1) and (2) of

Table 5. Robustness test results.

Variant	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Replacement of Explanatory Variable Measures		PSM-DID		Policy Uniqueness Test		Endogeneity Problem
	Lnimport1%	Lnimport2%	Lnimport1	Lnimport2	Lnimport1	Lnimport2	Lnimport1	Lnimport2
DID	−0.445 ** (0.181)	0.26 *** (0.071)	−1.772 *** (0.224)	0.370 *** (0.084)	−0.748 *** (0.189)	0.320 *** (0.070)	−1.783 *** (0.213)	0.367 *** (0.080)
constant term (math.)	−2.804 (1.950)	4.58 *** (0.63)	12.64 *** (2.143)	6.036 *** (1.146)	12.162 *** (2.076)	5.080 *** (1.175)	9.800 *** (1.881)	6.032 *** (1.142)
sample size	4754	5170	3650	5292	2702	3024	4737	5175
R2	0.039	0.030	0.233	0.069	0.061	0.028	0.168	0.067
product fixed effect	containment	containment	containment	containment	containment	containment	containment	containment
time fixed effect	containment	containment	containment	containment	containment	containment	containment	containment

Note: ** p < 0.05, *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

show that the estimated coefficients of the core variables of China’s imports of agricultural machinery products from the U.S. and imports from the other 16 countries after the replacement of the policy time point do not reach the significant level, indicating that the original regression results are robust.

Second, the placebo test is conducted by randomly selecting the experimental group. An unrepeated random sampling of agricultural machinery products is conducted to categorize the selected products into a pseudo-treatment group and to create a placebo-tested trade friction dummy variable. To exclude the interference of other factors, 500-times sampling is conducted. Figure 4 and Figure 5 report the estimated coefficients and densities, and the corresponding p-value distributions, for the 500-times pseudo-treatment group for agricultural machinery products imported from the U.S. and those imported from 16 other countries. The results show that the means of the two regression coefficients are close to 0, most of the p-values are greater than 0.1, and most of the estimated coefficients do not pass the test of significance, thus ruling out the possibility of pseudo-regression and indicating that the test results are robust.

5.4. Robustness Tests

This section uses four methods to verify the robustness of the results. Table 5 shows the results of the four robustness tests. (1) Change the measurement of the explanatory variables. Replace the original explanatory variable, China’s imports of agricultural machinery products from the U.S. and from the other 16 countries, with the percentage of China’s imports of agricultural machinery products from the U.S. and the other 16 countries of all China’s imports of agricultural machinery products, and then re-test [42]. The principle of this method is to replace an indicator that has the same properties as the explained variable under study. After taking the percentage as the explained variable, the percentage makes the original explained variable relatively smaller, and the result is still significant, but the value is smaller, which indicates that the original regression result is significant.

(2) PSM-DID test. In order to avoid the endogeneity of the importing countries selected under trade friction (i.e., other major agricultural machinery importing countries are countries that also have a higher level of economic development in the world), the PSM-DID method is used to conduct a robust-type test of the policy effects of tariff countermeasures. Specifically, a logit model is built, controlling for other variables, and a least-nearest-neighbor matching method with 1:1 no-putback sampling is used for matching at the country level. PSM-DID can remove samples that are not in the common support domain through sample rematching, potential confounding factors can be more effectively controlled, and the trend of treatment group and control group before intervention is more consistent, thus improving the accuracy of estimation. Therefore, we used PSM-DID in the relevant inspection of the Sino–U.S. trade friction on the import of Chinese agricultural machinery. When the PSM-DID method was used, logit regression was carried out on the control variables through the dummy variables that were subject to trade friction, and the propensity score was obtained. Specifically, logit regression models are often used to estimate the relationship between a binary variable and one or more explanatory variables. The goal of PSM is to make the two groups as similar as possible in all observable features by matching individuals from the treatment group and the control group. This can reduce the bias caused by sample self-selection and improve the accuracy of estimation. The role of logit regression is to provide a suitable statistical model that predicts the probability of an individual receiving processing based on a set of covariables, and thus calculates the propensity score. After matching, most of the observations are distributed within the common trend range of the propensity score, indicating a relatively balanced distribution between the treatment and control groups [43].

(3) Policy uniqueness test. The outbreak of new crown epidemic in 2020 led to different degrees of negative impact on trade in different regions. In order to exclude the possible interference caused by the row-control epidemic in the same period, the data after January 2020 are excluded and retested [44]. According to the results in

Table 6. Regression results of testing mechanism of technological innovation capacity of agricultural machinery enterprises.

Variant	(1)	(2)	(3)
	Lnimport1	Lnimport1	Lnimport1
DID	−1.790 *** (0.215)	−1.797 (0.224)	−1.804 ** (0.310)
DID-RD		−0.002 * (0.001)
DID-lnRD			−0.023 * (0.015)
control variable	containment	containment	containment
constant term (math.)	9.805 *** (1.886)	9.806 (1.992)	9.832 (1.992)
sample size	4754	4754	4754
R2	0.169	0.171	0.171
product fixed effect	containment	containment	containment
time fixed effect	containment	containment	containment

Note: * p < 0.1, ** p < 0.05, *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

, the results of all four robustness tests support that the U.S.–China trade friction significantly inhibits China’s imports of agricultural machinery products.

(4) Endogeneity problem. In (2), we briefly discussed the endogeneity problem. Further, consider the endogenous problems caused by reverse causality. It is found that there may be reverse causality between Sino–U.S. trade friction and China’s agricultural machinery import volume, and the endogenous problems caused by it make the estimated results biased and inconsistent. In order to strengthen the robustness of the conclusion, we adopted a one-stage lag for the core explanatory variables to weaken the influence of reverse causality on the regression results. In the estimation results with one-period lag (columns (7) and (8) of Table 6), the impact of Sino–U.S. trade friction on China’s agricultural machinery imports to the U.S. is significantly negative at 1% level, and the impact from other 16 countries is also significantly negative at 1% level, which is consistent with the baseline regression as a whole.

6. Further Analysis

6.1. Mechanism Analysis

The results of the previous study show that the U.S.–China trade friction significantly inhibits China’s imports of agricultural machinery products from the U.S. and also significantly increases imports from 16 other countries. However, the path of this effect has not yet been tested. Based on the theoretical analysis in the previous paper, this paper further explores the impact path of the mechanism in terms of the import price of Chinese agricultural machinery products and China’s economic uncertainty. This paper examines the mechanism of U.S.–China trade friction on China’s imports of agricultural machinery, and builds a mechanism testing model based on Equation (1):

$${lnimort}_{i,t}={\alpha }_0+{\beta }_{0}DID+{\beta }_{1}DID·R+{Control}_{i,t}+{\mu }_i+{\phi }_t+{\epsilon }_{it}$$

(8)

where variable $R$ is the mechanism variable, $DID$ is the interaction term that measures the effect of the influence mechanism, and the other variables have the same meaning as in Equation (1).

6.1.1. Technological Innovation Capacity of Agricultural Machinery Enterprises

Through the research and development of new products and technologies, enterprises can continuously improve their technology innovation ability, enhance their competitiveness in the market, and lay the foundation for long-term development. Considering that China-U.S. trade friction will increase uncertainty in the international supply chain, including import prices of U.S. agricultural machinery, exacerbating the market risk of domestic agricultural machinery enterprises, agricultural machinery enterprises will increase R&D efforts to expand the agricultural machinery “neckline” technology R&D investment, to reduce the dependence on imports of high-end agricultural machinery products from the U.S. and promote the domestic great cycle. However, the enterprise technology innovation, R&D capabilities, mastery of core technology, and ability to withstand risks from the United States to reduce imports of agricultural machinery products, due to the U.S. trade friction impact, are smaller.

Using company R&D expenditures (RD) can measure the technological innovation capacity of agricultural machinery companies and serve as a mechanism variable to measure China’s imports of agricultural machinery from the U.S. [32] The interaction term of RD and DID measures the mechanism effect of China’s imports of agricultural machinery. The baseline regression of model (3) is carried out, and the results are shown in Table 6. Columns (2) and (3) indicate that the addition of the mechanism variable increases the significance of the original DID variable, and the cross-multiplier term of the mechanism variable is significant and negative. This indicates that the U.S.–China trade friction does affect China’s imports of agricultural machinery from the U.S. Through stronger technological innovation ability, mastery of core technology, and stronger ability to withstand risk, agricultural machinery enterprises will more significantly reduce imports of agricultural machinery products from the U.S.

6.1.2. Economic Uncertainty

The trade friction between the United States and China is a destabilizing factor for the global economy, adding uncertainty to global economic development. The unjustified U.S. sanctions against China pose a huge potential risk to the global supply chain of Chinese enterprises and also adversely affect China’s import and export trade, adding to the uncertainty of China’s economic development. Agricultural machinery enterprises and consumers may reduce their investment and consumption, thereby lowering the demand for imported agricultural machinery products.

Referring to the Economic Uncertainty Index (EPU) studied by Baker, Bloom, and Davis (Baker, Bloom, and Davis, 2016) to measure the degree of economic uncertainty in China, it is used as a mechanism variable to measure China’s agricultural machinery imports from 16 other countries [45]. The interaction term of the economic uncertainty variable and the DID measures the mechanism effect of China’s agricultural machinery imports in the face of the trade friction between China and the U.S. A benchmark regression of model (3) is conducted, and the results are shown in

Table 7. Regression results of the economic uncertainty mechanism test.

Variant	(1)	(2)	(3)
	Lnimport2	Lnimport2	Lnimport2
DID	0.370 *** (0.084)	0.348 * (0.074)	0.304 ** (0.053)
DID-EPU		−0.001 *** (0.000)
DID-lnEPU			−0.015 *** (0.002)
control variable	containment	containment	containment
constant term (math.)	6.036 *** (1.146)	6.010 *** (1.336)	6.013 *** (1.336)
sample size	5192	5192	5192
R2	0.069	0.070	0.072
product fixed effect	containment	containment	containment
time fixed effect	containment	containment	containment

Note: * p < 0.1, ** p < 0.05, *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

. Columns (2) and (3) indicate that the cross-multiplier term of the mechanism variable is significant and negative at the 1% level after adding the mechanism variable. This indicates that the trade friction between China and the U.S. increases China’s economic uncertainty, and the greater the economic uncertainty, the greater the significantly positive degree of China’s agricultural machinery imports from the other 16 countries. This indicates that the U.S.–China trade friction increases China’s economic uncertainty, which in turn reduces the degree of increase in imports of agricultural machinery products from other 16 countries to cope with the risk of trade friction. The mechanism test verifies hypothesis H2.

6.2. Heterogeneity Analysis

6.2.1. Based on Different Classifications of Agricultural Machinery Products Imported from the United States of Heterogeneity Analysis

Referring to the study of Zhang Meng et al. (2016) on the definition of agricultural machinery products, we distinguish the impact of trade friction on the import of different agricultural machinery products [25].

Table 8. Distinguish different classifications of agricultural machinery products imported from the United States analysis.

Variant	Lnimport1
	(1) Drainage, Irrigation and Water Lifting Implements and Their Machinery	(2) Sowing and Planting and Fertilizing Machinery	(3) Agricultural Products Primary Processing Machinery	(4) Harvesting Machinery	(5) Other Machinery
DID	−1.680 *** (0.311)	−2.366 * (0.711)	−3.013 *** (0.470)	−0.360 (0.410)	−0.404 (0.322)
control variable	containment	containment	containment	containment	containment
constant term (math.)	13.04 *** (2.772)	11.39 * (5.022)	3.658 (7.655)	14.11 (1.19)	7.172 (4.910)
sample size	1776	587	583	168	709
R2	0.320	0.288	0.344	0.169	0.164
product fixed effect	containment	containment	containment	containment	containment
time fixed effect	containment	containment	containment	containment	containment

Note: * p < 0.1, *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

shows the regression results of the impact of China–U.S. trade friction on imports of different categorized agricultural machinery products from the United States. Columns (1)–(5) are drainage, irrigation, and water-lifting implements and their machinery; seeding, planting, and fertilization machinery; agricultural primary processing machinery,;harvesting machinery; and other machinery. According to Table 8, the decrease in imports from the U.S. is mainly reflected in irrigation and drainage implements and their machinery, planting and fertilization machinery, and agricultural preliminary processing machinery. The trade-suppressing effect on harvesting machinery and other machinery is not obvious. The explanation is: (1) The imposition of tariffs on irrigation and drainage equipment and machinery and agricultural products processing machinery and the sum of the types of taxes accounted for more than 50% of all taxed products, so such machinery trade inhibition effect is significant. (2) Harvesting machinery and other machinery include many kinds of agricultural machinery products, with rich import source countries, which can be imported from other countries as substitutes, and China imposes fewer tariffs on harvesting machinery and other machinery, so the inhibitory effect on these two kinds of machinery products is not significant.

6.2.2. Heterogeneity Analysis of Imports of Agricultural Machinery Products from 16 Other Countries Based on Different Classifications

Further, this paper differentiates the impact of trade friction on imports of different agricultural machinery products from other countries.

Table 9. Distinguish different classifications of agricultural machinery products from other 16 countries import analysis.

Variant	Lnimport2
	(1) Drainage, Irrigation and Water Lifting Implements and Their Machinery	(2) Sowing and Planting and Fertilizing Machinery	(3) Agricultural Products Primary Processing Machinery	(4) Harvesting Machinery	(5) Other Machinery
DID	0.374 *** (0.098)	0.673 (0.317)	0.346 *** (0.094)	1.283 ** (0.257)	0.0948 (0.384)
control variable	containment	containment	containment	containment	containment
constant term (math.)	6.514 *** (0.869)	7.862 ** (1.938)	4.759 ** (1.140)	8.734 (26.302)	2.304 (3.046)
sample size	18	756	672	252	756
R2	0.234	0.118	0.183	0.131	0.138
product fixed effect	containment	containment	containment	containment	containment
time fixed effect	containment	containment	containment	containment	containment

Note: ** p < 0.05, *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

shows the regression results of the impact of China–U.S. trade friction on the imports of different categorized agricultural machinery products from 16 other countries. According to Table 9, the increase in imports from the other 16 countries is mainly manifested in the imports of irrigation, drainage, and water lifting implements and their machinery, agricultural primary processing machinery, and harvesting machinery, while the transfer effect on seeding, planting, and fertilizing machinery and other machinery is not significant. The explanation is: (1) Irrigation machinery, water lifting equipment, and processing and harvesting machinery have rich import source countries, and the main importing countries are concentrated in Europe. These products from the United States imports are part of the significant transfer to other countries, so imports from the other 16 countries increased significantly. (2) Sowing, planting, and fertilization machinery is the main source of imports from the United States. Such products have a strong dependence on the United States, so in a short period of time, it is more difficult to seek other countries to replace the market. (3) In recent years, the increase in demand from foreign markets for imports of drainage, irrigation, and water lifting equipment and machinery from the other 16 countries, to reduce imports from the U.S., did not significantly increase the situation. The test again verified hypothesis H1.

6.2.3. Distinguish the Heterogeneity Analysis of Complete Machinery and Machinery Spare Parts

This paper further divides agricultural machinery products into complete machinery products and machinery spare parts.

Table 10. Distinguishing between complete machinery and machinery spare parts analysis.

Variant	Lnimport1		Lnimport2
	(1) Complete Machinery	(2) Mechanical Spare Parts	(1) Complete Machinery	(2) Mechanical Spare Parts
DID	−1.993 *** (0.233)	−0.410 (−1.303)	0.322 ** (0.096)	0.634 (0.420)
control variable	containment	containment	containment	containment
constant term (math.)	9.282 *** (2.301)	12.41 (3.197)	7.257 *** (1.110)	2.167 (3.269)
sample size	3550	1204	4032	1260
R2	0.201	0.122	0.069	0.133
product fixed effect	containment	containment	containment	containment
time fixed effect	containment	containment	containment	containment

Note: ** p < 0.05, *** p < 0.01, robust standard errors in parentheses; control variable regression results are retained for space.

reports the impact of the U.S.–China trade friction on complete machinery products and machinery spare parts. The coefficient of imports of complete machinery products from the U.S. is significantly negative, and the coefficient of machinery spare parts is insignificant, i.e., the inhibitory effect of imports of complete agricultural machinery products from the U.S. is obvious, indicating that the import demand elasticity of complete machinery products is large, and China’s implementation of counter-tariffs in this regard can effectively reduce the U.S. agricultural machinery enterprises to export agricultural machinery products to China. The core coefficient of imports of complete machinery products from the other 16 countries is significantly positive, and imports of complete machinery products from the other 16 countries increases significantly. That is, after China’s implementation of tariff countermeasures, the trade diversion effect of China’s imports of agricultural machinery products is significant and mainly reflected in the import diversion of complete agricultural machinery products. Possible reasons are: On the one hand, Europe and the United States of America’s large agricultural machinery production of high-end products has multi-functional, composite operations, large-scale, refined, high-efficiency, cost-saving and other advantages. While China produces single function agricultural machinery products, the import demand for large-scale complete agricultural machinery and equipment is high. On the other hand, China is rich in labor force and is a big country in the manufacture and export of machinery spare parts, which have a smaller import demand. The above reasons make the demand for machinery spare parts less sensitive than the demand for complete mechanical products, so that the impact of trade friction on the two produces more significant differences. This also reflects the country’s urgent need for technological innovation, research, and development of large-scale composite agricultural machinery.

7. Conclusions and Implications

This paper examines the impact of China–U.S. trade friction on China’s imports of agricultural machinery products from the perspective of trade effect by using a multi-period double-difference model and selecting the monthly import data of Chinese agricultural machinery products with HS eight-digit codes from 2016 to 2022. The results of the study show that: (1) The trade friction between China and the United States significantly reduces the trade volume of China’s agricultural machinery products imported from the United States, and the trade suppression effect significantly increases the trade volume of agricultural machinery imports from 16 other countries; there is a trade transfer effect, and imports are mostly transferred to Japan, South Korea, and other countries. The conclusions of the results are still robust after various types of tests. (2) A mechanism test of the agricultural machinery enterprise technology innovation ability, and China’s economic uncertainty as a channel to influence China’s agricultural machinery imports to be analyzed, found that the role of the mechanism of agricultural machinery enterprise technology innovation ability and economic uncertainty on the degree of China’s imports of agricultural machinery is significant, and the negative effect. (3) Further, after China’s implementation of tariff countermeasures, seeding planting and fertilization machinery, harvesting machinery, and other agricultural machinery imports, there is a more significant trade inhibition effect, while the impact on the drainage, irrigation, and water lifting equipment and its machinery, and agricultural product processing machinery imports is not significant. (4) In addition, the U.S.–China trade friction mainly affects the imports of complete machinery products; the impact on machinery spare parts is not obvious.

This paper’s conclusion has certain enlightenments for coping with Sino–U.S. trade friction and adjusting China’s agricultural machinery imports and market structure. (1) Build a more complete import diversification system for agricultural machinery products. Sino–U.S. trade frictions are likely to redefine the global trade pattern of agricultural machinery products and bring development opportunities for the diversification of China’s agricultural machinery products imports. In this Sino–U.S. economic and trade friction, the key to China’s successful response to the trade disputes provoked by the United States lies in the continuous promotion of the diversification strategy of agricultural imports. Therefore, China should continue to adjust the import of agricultural machinery products on the basis of considering the trade interests of the existing major sources of agricultural machinery products and further rely on the “Belt and Road” initiative, regional trade agreements, special trade arrangements, etc., to build a stable and reliable external grain supply system. (2) Use the leading role of the government in the development of agricultural machinery trade, increase investment in scientific research, strengthen technological innovation, support the deep integration of production, research and use, accelerate the process of technology research and development, speed up the shortcomings of agricultural mechanization, break the monopoly barriers of foreign enterprises on high-end technology, and reduce the dependence on high-end agricultural machinery imports, in order to resist the risk of external shocks. (3) The agricultural machinery industry can consider establishing a long-term trade friction monitoring mechanism. Establish a perfect system for information collection, information standardization, information storage, information integration and information sharing. Through data collection, personnel regularly grasp the relevant trade information of agriculture-related websites and conduct audits and comparative analyses, in order to obtain timely and effective trade-friction early-warning information. For early warning analysis: Conduct an in-depth analysis of the collected information, identify potential trade friction risk points, and propose corresponding countermeasures. Establish an expert consultation mechanism, invite industry experts to study and predict the situation of trade friction, and provide professional consultation and advice for enterprises. Formulate specific coping plans and measures for different types of trade frictions, such as adjusting export strategies and strengthening communication and coordination with the international market. (4) Promote the development of the domestic market for agricultural machinery. Agricultural machinery enterprises should make full use of national agricultural machinery purchase subsidies and other policies to stimulate market demand and reduce the cost to farmers of buying agricultural machinery. Strengthen communication and coordination with local government departments and strive for more policy support and resource investment, so as to increase marketing efforts in the domestic market and enhance brand awareness and influence. Expand sales channels and networks, strengthen cooperation with dealers, cooperatives and other institutions, and improve market coverage.

Industrial development and the building of a strong trading nation are complementary and mutually reinforcing. In the face of the complex and changing international environment, it is also important to emphasize the role of trade in agricultural machinery products in international negotiations. As China continues to expand its opening to the outside world, trade in agricultural machinery has become an important bargaining chip in trade negotiations between countries. China and the United States are the world’s leading economic powers. If the United States is still pursuing trade protectionism and tariff sanctions against China, the implementation of tariff countermeasures on agricultural machinery products can be used as a means of response to play a significant policy effect.

There are also some limitations to this paper. The analysis of the overall import volume and trade structure of China’s agricultural machinery products is limited to Sino–U.S. trade frictions, and the analysis of intermediate and final products of agricultural machinery is lacking. The trade flow of products is not limited to the trade volume of final products, so it is a future research direction to explore the impact of Sino–U.S. trade friction on specific products and whether it will indirectly affect other industries.