Research on the Policy Effects and Impact Mechanisms of the Belt and Road Initiative on China’s Forest Products Trade

Abstract

The Belt and Road Initiative, as an important measure for China in terms of opening up and participating in international economic and trade cooperation, has become a new driving force for the sustainable development of China’s forest products trade. This paper takes the Belt and Road Initiative as a policy event and evaluates its policy effects on the development of China’s forest products trade from the causal level through the difference-in-differences model (DID), explores the policy effect in detail from the perspectives of product heterogeneity and regional heterogeneity, and clarifies the specific impact mechanism. The main results are as follows: (1) there is a significant policy promotion effect of the Belt and Road Initiative on the growth of the bilateral trade scale of forest products between China and the countries along the route. (2) In terms of product structure, the policy promotion effect of the Belt and Road Initiative is mainly manifested in processed wood products. (3) In terms of regional distribution, the policy promotion effects of the Belt and Road Initiative are mainly concentrated in Europe, Africa, and Asia. (4) The “logistics performance, political partnership with China, and Internet penetration” of trading countries play a significant positive mediating role in the policy effects of the Belt and Road Initiative. Therefore, in view of the significant role of the Belt and Road Initiative in promoting the development of bilateral forest products trade, China should promote more countries to participate in the joint construction of the Belt and Road and tap new momentum for the development of the forest products trade by focusing on key countries, priority areas, and key products.

1. Introduction

As an important part of China’s foreign economic trade, forest products trade has many effects, such as driving the economic growth of forest areas, promoting the income of forest farmers, ensuring domestic timber security and realizing trade in foreign exchange, etc. It inherits the goal of leapfrog development of modern forestry in the 21st century in China and is highly integrated with the macro layout of high-quality development [1,2,3]. Although China’s forest products industry has developed rapidly and has become one of the world’s largest forest products trading countries, the internal and external challenges it faces are becoming increasingly severe. From the perspective of import, with the increase of global attention to forest resources, countries around the world have successively introduced more rigorous policies of felling and export for timber one after another, which has exacerbated the long-standing contradiction between supply and demand for timber in China [4,5]. From the perspective of exports, many factors, such as the rise of international trade protection awareness, have led to the continued deterioration of the global business environment, making China’s forest products encounter “anti-dumping, anti-subsidy,” and green trade barriers in the export [6]. In addition, with the global increase in demand for wood products that can be proven to be of legal origin, Chinese forestry enterprises are required to undergo forest certification when exporting forest products, the direct consequence of which is higher trade costs, shrinking profits, and declining competitiveness [7,8,9].

The proposal of the Belt and Road Initiative has pointed out a new direction for the development of China’s forest products trade. First, after more than eight years of unremitting efforts, the Belt and Road Initiative has yielded fruitful results. The level of connectivity between China and countries along the route has been qualitatively improved, which provides an excellent business environment for bilateral and multilateral trade in forest wood products [10,11]. Secondly, there are many countries along the Belt and Road with different forest resources, capital and technology endowments, and the supply and demand of these countries for forest products are also different. Therefore, it shows that there is a huge space and potential for mutual cooperation between these countries [12,13].

As a “Chinese solution” to actively open up and cooperate with the outside world and promote global prosperity and development, the policy effects of the Belt and Road Initiative have attracted widespread attention from scholars in various fields, involving macroeconomic development, green and sustainable development, national soft power enhancement and so on. Most studies have shown that the Belt and Road Initiative can achieve the improvement of China’s export efficiency and potential, the development of the tourism industry, and the growth of residential income; that is, a direct promotion effect on China’s macroeconomic development level [14,15,16,17]. Some scholars have proposed that the Belt and Road Initiative has significantly improved the green total factor productivity and reduced corporate pollution in domestic cities along the route and promoted China’s green development of Outward Foreign Direct Investment (OFDI) to developing countries [18,19]. In addition, some scholars have analyzed the investment in countries along the Belt and Road to significantly enhance China’s soft power [20].

A small amount of literature has explored the mechanism of the policy effects of the Belt and Road Initiative. On the one hand, the Chinese government has always occupied the main position of resource allocation in the Belt and Road Initiative. Some studies have incorporated government behavior and policy factors into the impact mechanism of the Belt and Road Initiative and empirically analyze that government policies have played an important role in driving enterprises to reduce environmental pollution, develop trade channels, and many other aspects [21,22,23]. On the other hand, since sustainable development is the core goal of the Belt and Road Initiative, the promotion effect of the initiative on macroeconomic development has been widely studied. Some scholars have suggested that the level of infrastructure, optimal allocation of resources, industrial upgrading and technological progress play a vital role in the process of the Belt and Road Initiative affecting the sustainable economic development of participating countries [24,25].

The existing empirical methods for policy effects can be divided into indirect research methods and direct research methods [26,27]. Among them, the indirect research method is based on the gravity model and stochastic frontier gravity model to indirectly evaluate the policy impact of the Belt and Road Initiative based on the empirical analysis of trade influencing factors [28,29]. The direct research method has emerged in recent years, and with the maturity of the difference-in-differences model (DID), scholars have mostly used this method to causally analyze the policy effects of the Belt and Road Initiative on direct investment and international trade between China and countries along the route [15,25,30]. Due to the exogenous nature of policy shocks and the restricted nature of the target population for policy use, the sample affected by the policy and the sample not affected by the policy will change as a result of the policy effect, and the difference-in-differences model (DID) can control the systematic differences between the experimental group and the control group by comparing the differences before and after the experiment, so as to test the implementation effect of a particular policy [31].

In summary, the existing literature provides a rich research perspective and a mature research methodology for this paper. However, there are the following research shortcomings: first, the policy promotion effects of the Belt and Road Initiative on international trade between China and countries along the route have been confirmed by many scholars, but whether it is applicable to the field of forest products trade remains to be proven; secondly, the research on policy effect analysis mostly appears in recent years, mostly at the macro level, and few scholars have conducted a comparative analysis of the heterogeneity of different products and different regions from the perspective of policy effect analysis; thirdly, few scholars have paid attention to the specific intermediaries of the policy impact of the initiative on trade, that is, the emphasis on the impact mechanism is slightly insufficient. Therefore, based on the analysis of the policy effects of the Belt and Road Initiative on the trade of forest products between China and the countries along the route, this paper delves into product heterogeneity and regional heterogeneity and emphatically analyzes the specific impact mechanisms of the initiative. This paper mainly discusses the following three core issues: First of all, is the Belt and Road Initiative conducive to improving the scale of China’s forest products import and export trade? Secondly, what is the heterogeneity of the impact of the Belt and Road Initiative on the scale of China’s forest products import and export trade at the product and spatial levels? Finally, what is the impact mechanism of the Belt and Road Initiative on the scale of China’s forest products import and export?

Compared with existing studies, the possible innovations of this paper are mainly focused on the following three points: first, the research perspective level. Exploring the specific impact of the Belt and Road Initiative on the forest products trade between China and the countries along the route from the perspective of intuitive policy effect analysis is an important complement to the content of the policy analysis of the Belt and Road Initiative and the high-quality development of forest products trade. Secondly, the research content level. This paper fully considers the product heterogeneity and regional heterogeneity effects of the Belt and Road Initiative on the development of China’s forest products trade and introduces the influence mechanism analysis, which provides a realistic basis for decision-makers to formulate trade strategies for different products in different regions, so as to maximize trade and economic well-being. Thirdly, the research method level. Based on the traditional gravity model, the difference-in-differences model (DID) is constructed to evaluate the policy effect of the Belt and Road Initiative, which realizes the application of the direct policy analysis method in the empirical framework of international trade and makes up for the defect that the gravity model cannot analyze the policy effect from the causal level.

2. Research Design

2.1. Model Setting

The gravity model relates bilateral trade flows to the scale of economic development and trade costs of trading partners and describes this relationship as a function of the “gravitational equation.” Since its proposal, it has been rapidly popularized in international trade research and has gradually become a classical method for research in the field of international trade because of its wide application space. In this paper, we consider the advantages of the gravity model and difference-in-differences model (DID) and construct a difference-in-differences (DID) model based on the traditional gravity model to causally analyze the impact of the Belt and Road Initiative on China’s forest products trade. First, the gravitational model of forest products trade is constructed in the traditional sense, as shown in Equation (1).

$$\mathrm{l}\mathrm{n}({\mathrm{y}}_{\mathrm{i}\mathrm{t}})={\mathsf{\beta }}_0+{\displaystyle \sum _{\mathrm{m}}^{}{\mathsf{\beta }}_{\mathrm{m}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{m}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{n}}^{}{\mathsf{\beta }}_{\mathrm{n}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{n}\mathrm{i}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{s}}^{}{\mathsf{\beta }}_{\mathrm{s}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{s}\mathrm{i}})}+{\mathrm{e}}_{\mathrm{i}\mathrm{t}}$$

(1)

where i is the country that trades forest products with China, t is the year of trade, y_it is the trade flow of forest products between China and the trading countries, and this paper intends to use the import value (Imp), export value (Exp) and total trade value (Trd) for the analysis. The independent variables in the gravity model can be divided into three categories according to their subscripts: one (x_t) is the time-varying variables of China, the other (x_it) is the time-varying variables of trading countries, and the third (x_i) is the non-time-varying constants of China and trading countries. β is the marginal impact of each covariate on trade flows. e_it is the unobservable factors affecting trade flows, which are assumed to obey a normal independent homogeneous distribution, and Equation (1) defines a simple OLS model. On the basis of Equation (1), the policy variables of the Belt and Road Initiative are introduced, thus rewriting Equation (1) as (2).

$$\mathrm{l}\mathrm{n}({\mathrm{y}}_{\mathrm{i}\mathrm{t}})={\mathsf{\beta }}_0+\theta {\mathrm{D}}_{\mathrm{it}}+{\displaystyle \sum _{\mathrm{m}}^{}{\mathsf{\beta }}_{\mathrm{m}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{m}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{n}}^{}{\mathsf{\beta }}_{\mathrm{n}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{n}\mathrm{i}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{s}}^{}{\mathsf{\beta }}_{\mathrm{s}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{s}\mathrm{i}})}+{\mu }_{\mathrm{i}}+{\nu }_{\mathrm{t}}+{\epsilon }_{\mathrm{it}}$$

(2)

where D_it is the policy dummy variable for whether trading country i joins the Belt and Road Initiative in year t. The error term consists of three components, which are the country-fixed effect μ_i, time-fixed effect ν_t, and random error ε_it. Thus, Equation (2) defines a standard multi-temporal DID model, where D_it is equal to the interaction term of the experimental group dummy variable (whether or not to join the initiative country) and the initiative time dummy variable, and θ is the main parameter of interest in this paper to capture the average treatment effect of joining the Belt and Road Initiative on forest products trade flows.

In order to determine the relative impact of the initiative on the trade of forest products in international trade, this paper also estimates Equation (3). Where z_it is the total import value (Timp), export value (Texp), or total trade value (Ttrd) occurring between China and the trading countries.

$$\mathrm{l}\mathrm{n}({\mathrm{y}}_{\mathrm{i}\mathrm{t}}/{\mathrm{z}}_{\mathrm{it}})={\mathrm{b}}_0+\theta {\mathrm{D}}_{\mathrm{it}}+{\displaystyle \sum _{\mathrm{m}}^{}{\mathrm{b}}_{\mathrm{m}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{m}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{n}}^{}{\mathrm{b}}_{\mathrm{n}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{n}\mathrm{i}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{s}}^{}{\mathrm{b}}_{\mathrm{s}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{s}\mathrm{i}})}+{\mu }_{\mathrm{i}}+{\nu }_{\mathrm{t}}+{\epsilon }_{\mathrm{it}}$$

(3)

The key assumption of the DID model is the parallel trend assumption. In this paper, we refer to the event study approach in the form of a model setup to observe the changes in trade trends before and after joining the initiative separately during the window period. Equation (4) defines the above model. The subscripts of the policy dummy variable D are transformed from it to ij, where j is used to indicate the state of the agreement countries before or after the event of joining the Belt and Road Initiative in year j. By setting this state variable year by year, the pre-event trend changes and post-event dynamic trends can be observed through α_j.

$$\mathrm{l}\mathrm{n}({\mathrm{y}}_{\mathrm{i}\mathrm{t}})={\mathsf{\beta }}_0+{\displaystyle \sum _{\mathrm{j}=-5}^{J=3}{\alpha }_{\mathrm{j}}}{\mathrm{D}}_{\mathrm{ij}}+{\displaystyle \sum _{\mathrm{m}}^{}{\mathsf{\beta }}_{\mathrm{m}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{m}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{n}}^{}{\mathsf{\beta }}_{\mathrm{n}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{n}\mathrm{i}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{s}}^{}{\mathsf{\beta }}_{\mathrm{s}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{s}\mathrm{i}})}+{\mu }_{\mathrm{i}}+{\nu }_{\mathrm{t}}+{\epsilon }_{\mathrm{it}}$$

(4)

Finally, a typical mediating effect model is constructed to analyze the impact mechanism of the Belt and Road Initiative on China’s forest products trade. Three regression equations are constructed to verify whether the specific variables play a mediating mechanism. Among them, the base equation is Equation (2), and θ represents the effect of the initiative on forest products trade. The other two regression equations are shown in Equations (5) and (6), which represent the mediating variable M_it, and the mediating variables chosen in this paper include political relations, logistics performance, etc. Similar to Equation (2), a in Equation (5) can determine the effect of the Belt and Road Initiative on the mediating variable M_it. By introducing the mediating variables in (2), we obtain Equation (6), where b is the effect of the mediating variables on trade values and θ′ is the effect of the Belt and Road Initiative on trade after controlling for the mediating variables. In the linear model, θ = θ′ + ab can be confirmed. Among them, θ′ can be interpreted as the “direct effect” of the policy, and ab can be interpreted as the “indirect effect” of the policy through the mediating variable M_it.

$${\mathrm{M}}_{\mathrm{it}}={\mathsf{\beta }}_0+a{\mathrm{D}}_{\mathrm{it}}+{\displaystyle \sum _{\mathrm{m}}^{}{\mathsf{\beta }}_{\mathrm{m}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{m}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{n}}^{}{\mathsf{\beta }}_{\mathrm{n}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{n}\mathrm{i}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{s}}^{}{\mathsf{\beta }}_{\mathrm{s}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{s}\mathrm{i}})}+{\mu }_{\mathrm{i}}+{\nu }_{\mathrm{t}}+{\epsilon }_{\mathrm{it}}$$

(5)

$$\mathrm{l}\mathrm{n}({\mathrm{y}}_{\mathrm{i}\mathrm{t}})={\mathsf{\beta }}_0+{\theta }^{’}{\mathrm{D}}_{\mathrm{i}\mathrm{t}}{+\mathrm{bM}}_{\mathrm{it}}+{\displaystyle \sum _{\mathrm{m}}^{}{\mathsf{\beta }}_{\mathrm{m}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{m}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{n}}^{}{\mathsf{\beta }}_{\mathrm{n}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{n}\mathrm{i}\mathrm{t}})}+{\displaystyle \sum _{\mathrm{s}}^{}{\mathsf{\beta }}_{\mathrm{s}}\mathrm{l}\mathrm{n}({\mathrm{x}}_{\mathrm{s}\mathrm{i}})}+{\mu }_{\mathrm{i}}+{\nu }_{\mathrm{t}}+{\epsilon }_{\mathrm{it}}$$

(6)

2.2. Variable Selection and Interpretation

The dependent variables used in the empirical analysis of this paper are import value, export value, and the total trade value of forest products between China and major countries in the world. The forest products defined in this paper refer to wood forest products in general, specifically logs, sawn timber, wood pulp, wood-based panels, wood products, wood furniture and paper products. Due to the limitation of space, these kinds of product codes are not described in this paper. The data on trade values comes from the trade database UN Comtrade. We identified the major countries with which China trades forest products based on total import value, total export value, and total trade value for each of the seven product categories mentioned above. Thus defining the sample frame used in the data analysis session of this paper, which differs in the three dependent variable models. The range of years chosen for the sample in this paper is 11 years cumulatively from 2009 to 2019.

In this paper, the sample frames are selected based on the top 50 countries in terms of trade values of forest products with China in 2019 (

Table 1. Sample frame of China’s forest products trading countries.

CNTY	Tot	Imp	Exp	Time	CNTY	Tot	Imp	Exp	Time	CNTY	Tot	Imp	Exp	Time
USA	1	1	1		ARE	26		16	2018	OMN			45	2018
CAN	2	3	7		URY	27	14		2018	BGD			36	2016
RUS	3	2	24	2017	ESP	28	43	21		PRK			49
JPN	4	11	2		BEL	29	41	23		BLR		46		2014
BRA	5	4	30		PNG	30	15		2018	PER			40	2019
IDN	6	5	19	2018	POL	31	34	27	2019	COL			46
GBR	7	17	3		MEX	32		22		KEN			50
AUS	8	13	4		CZE	33	20		2015	JOR			48
DEU	9	9	9		LAO	34	22		2016	ECU		39		2018
NZL	10	6	31	2017	IRQ	35		25	2017	LTU		49		2017
VNM	11	19	6	2017	NGA	36		26	2018	IRL		47	44
KOR	12	23	5	2015	SLB	37	24		2019	GRC			47	2018
THA	13	10	17	2017	ISR	38		28		GNQ		33		2019
CHL	14	7	34	2018	TUR	39		29	2015	COG		30		2018
MYS	15	28	8	2017	KHM	40		33	2016	PRT		42		2018
SGP	16	18	11	2017	IRN	41		32	2017	QAT			43	2014
FRA	17	21	10		ROU	42	32		2017	SLE		40		2018
FIN	18	8			MOZ	43	29		2018	KWT			42	2018
SWE	19	12	37		AUT	44	26		2018	CHE		48
NLD	20	35	14		GAB	45	27		2018	LVA		45		2017
ITA	21	16	20	2019	CMR	46	31		2018	ARG		38
IND	22	37	15		PAK	47		35	2017	GHA		44		2018
PHL	23		13	2018	UKR	48	36		2015	EGY			38	2016
SUA	24		12	2016	MMR	49		39
ZAF	25	25	18	2018	DNK	50	50	41

(1) the ranking in the table is calculated according to the total trade value, import value and export value in 2019; (2) the sample of Hong Kong is removed from the ranking calculation; (3) a blank year indicates that the country has not joined the Belt and Road Initiative.

). Table 1 describes the sample frame countries of three dependent variables statistically, and the three sample frames are the balanced panel data of “50 countries × 11 years”. According to statistics, the total trade value of the top 50 countries accounted for 89.75% of China’s total trade value of forest products. China’s total import value of forest products from the top 50 countries accounted for 96.88% of the total import value. The total export value of China’s forest products from the top 50 countries accounts for 89.07% of the total export value. Therefore, it can be seen that the sample countries selected in this paper are highly representative.

The core independent variable in this paper is the policy dummy variable (D) of countries joining the Belt and Road Initiative; the main variables in the traditional gravity model are all covariates of this paper. Specifically, the log of GDP of trading countries (lngdp), the log of GDP of China (lngdp), the log of population size of trading countries (lnpop), the log of population size of China (lncpop), and the log of geographical distance between China and trading countries (lndis), all of which are obtained from the World Bank statistical database. In addition, this paper will also analyze the mechanism of the impact of the Belt and Road Initiative on trade scale, and the selected mediating mechanism variables include three categories: first, the state of logistics infrastructure (lp); secondly, political partnership (pt); thirdly, Internet penetration (internet). Among them, the state of logistics infrastructure is measured through the Logistics Performance Index, which is derived from the World Bank statistical database. A political partnership is measured by summing two variables, which are partnership and the number of important bilateral documents. Internet penetration is also an important indicator used to reflect a country’s level of economic development and information infrastructure development, which is obtained from the annual panel data provided by the World Bank database.

3. Regression Analysis of the Model

3.1. Analysis of the Effect of Belt and Road Initiative on Forest Products Trade

First, estimate by Equation (2) (

Table 2. Effect of joining the Belt and Road Initiative on the trade of forest products.

Variable	Import (Lnimp)		Export (Lnexp)		Total (Lntrd)
	(1)	(2)	(3)	(4)	(5)	(6)
D	0.138	0.457 ***	0.172 *	0.284 ***	−0.123	0.209 ***
	(0.64)	(2.88)	(1.84)	(3.66)	(−0.93)	(3.25)
lngdp	0.387 ***	0.218	0.813 ***	0.991 ***	0.507 ***	0.533 ***
	(8.42)	(0.60)	(32.10)	(6.10)	(17.29)	(3.45)
lnpop	−0.008	−1.396	−0.118 ***	0.309	−0.130 ***	1.009
	(−0.13)	(−0.85)	(−5.13)	(0.37)	(−3.56)	(0.83)
lncgdp	1.424		1.144 ***		0.981 *
	(1.63)		(2.94)		(1.79)
lncpop	−3.627		−7.155		0.347
	(−0.24)		(−1.05)		(0.04)
lndis	0.044		−0.524 ***		−0.100
	(0.39)		(−11.54)		(−1.43)
Country fixed effect	Uncontrolled	Controlled	Uncontrolled	Controlled	Uncontrolled	Controlled
Time fixed effect	Uncontrolled	Controlled	Uncontrolled	Controlled	Uncontrolled	Controlled
constant term	76.124	40.232	157.054	5.745	0.996	−1.782
	(0.24)	(1.48)	(1.13)	(0.40)	(0.01)	(−0.09)
Observation	550	550	550	550	550	550
Adjusted R2	0.246	0.883	0.718	0.959	0.469	0.951
F-Statistics	30.892	17.977	233.397	17.509	81.799	12.226
p-Value	0.000	0.000	0.000	0.000	0.000	0.000

(1) sample frames of importing countries, exporting countries and trading countries are from Table 1; (2) the t-values corresponding to the estimated coefficients are indicated in parentheses and are calculated by the robust standard errors of clustering at the country level; (3) * and *** indicate the statistical significance levels of 10%, and 1%, respectively; (4) the adjusted R² reported in the fixed effects model in the table includes the contribution of fixed effects.

). As shown in Table 2, all models show a good overall fit. Comparing the R² of different models, it can be found that the explanatory effect of the models can be substantially improved after adding fixed effect parameters to the models; in comparison, the fitting effect of the forest products export model is significantly better than that of the import model. This indicates that the gravity model explains China’s forest products export trade more strongly than the import trade. After testing, some of the estimated coefficients of the covariates in the model are no longer statistically significant because there is some covariance among the independent variables. Among them, the covariate that matches the theoretical expectation and performs robustly is the GDP of the trading country with positive and statistically significant coefficients at the 1% level in all five models, followed by China’s GDP, while the geographical distance is only statistically significant in the export trade model, indicating that the main factors affecting China’s forest products trade lie in the macroeconomic development of the trading countries and China.

As shown in Table 2, when fixed effects are not introduced in the model, the estimated coefficients of the policy dummy variable D can’t be used to reflect the policy effects. Therefore, the impact of joining the Belt and Road Initiative is only examined in the fixed effect model. The estimated coefficients of the policy variable D are significantly positive in all three models, where the import value, export value and total trade value of forest products of China and the major trading partners joining the Belt and Road Initiative increase by 45.7%, 28.4% and 20.9% on average, respectively. The above findings suggest that trading partners joining the Belt and Road Initiative can effectively boost forest products trade with China, and the boosting effect on China’s import value is stronger than that on China’s export value.

3.2. Product Heterogeneity Analysis

In this paper, Equation (2) is estimated separately for the import and export trade of seven types of forest products (

Table 3. Effect of joining the Belt and Road Initiative on the trade of forest products.

Variable	Logs	Sawn	Pulp	Panels	Products	Furniture	Paper
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Import (lnimp)
D	−0.046	0.488 ***	0.024	0.942 ***	0.565 **	0.102	0.567 ***
	(−0.17)	(3.36)	(0.06)	(4.44)	(2.50)	(0.57)	(2.59)
Observation	513	535	345	486	486	433	487
Adjusted R2	0.791	0.913	0.814	0.827	0.829	0.921	0.930
Main Variable R2	0.054	0.096	0.011	0.277	0.219	0.177	0.754
F-Statistics	10.560	14.942	1.983	5.098	7.118	5.366	3.046
p-Value	0.000	0.000	0.022	0.000	0.000	0.000	0.000
Export (lnexp)
D	0.400	0.086	0.566 **	0.277 ***	0.340 ***	0.007	0.304 ***
	(0.24)	(0.55)	(2.07)	(2.87)	(5.31)	(0.08)	(8.06)
Observation	33	546	369	550	550	550	550
Adjusted R2	0.583	0.871	0.779	0.856	0.949	0.916	0.968
Main Variable R2	0.115	0.527	0.168	0.373	0.742	0.596	0.586
F-Statistics	0.620	6.836	4.041	20.903	51.766	11.342	145.975
p-Value	0.793	0.000	0.000	0.000	0.000	0.000	0.000
Total(lntrd)
D	−1.610	0.282 **	−0.110	0.322 ***	0.174 ***	−0.023	0.347 ***
	(−0.71)	(2.36)	(−0.48)	(3.66)	(2.63)	(−0.26)	(7.84)
Observation	28	463	277	480	503	471	508
Adjusted R2	0.504	0.913	0.917	0.926	0.960	0.934	0.976
Main Variable R2	0.256	0.073	0.115	0.561	0.794	0.549	0.730
F-Statistics	0.418	3.638	1.370	20.003	45.062	7.999	91.781
p-Value	0.917	0.000	0.175	0.000	0.000	0.000	0.000

(1) country fixed effect and year fixed effect are controlled in all models; (2) due to the limitation of space, the estimated values of constant terms are omitted in the table; (3) the adjusted R² reported in the table includes the contribution of fixed effect, and the main variable R² refers to the contribution of three variables in the model: D, lngdp and lnpop; (4) the values in brackets indicate the t-value corresponding to the estimation coefficient, which is obtained by the cluster robust standard error calculation at the country level; (5) ** and *** represent statistical significance levels of 5%, and 1% respectively.

). For comparison purposes, the sample frames of import, export and total trade value are no longer screened separately based on the trade size of each type of product.

As shown in Table 3, except for the logs and wood pulp models, all five forest product categories performed well in the overall fit of the three models. The regularity can be more clearly summarized by looking at the contribution of the three main variables to the model R². Overall, four categories of processed products, such as wood-based panels, wood products, wood furniture and paper products, are explained by three main variables to a greater extent than three kinds of raw materials, such as logs, sawn timber and wood pulp. That is to say, the R² of the raw material forest product model mainly comes from the contribution of the two-way fixed effect reflecting the national heterogeneity and time heterogeneity, and the fixed effect itself can be understood as an idiosyncratic factor that cannot be effectively explained; on the contrary, the trade of processed products is explained by gravity model to a higher degree.

Similar conclusions can be obtained by comparing the estimated coefficients of the policy dummy variable D. First, according to the estimation results of the import value model, it can be seen that the accession of trading countries to the Belt and Road Initiative significantly increases China’s imports of sawn timber, wood-based panels, wood products, and paper products by 48.8%, 94.2%, 56.5%, and 56.7%, respectively, while there is no significant change in the import value of logs, wood pulp, and wood furniture. Secondly, according to the export value model, the accession of trading countries to the initiative significantly increased China’s exports of wood pulp, wood-based panels, wood products, and paper products by 56.6%, 27.7%, 34.0%, and 30.4%, respectively, while there is no significant change in the exports of logs, sawn timber, and wood furniture. In addition, according to the trade value model, the accession of trading countries to the initiative significantly increased China’s trade value of sawn timber, wood-based panels, wood products, and paper products by 28.2%, 32.2%, 17.4%, and 34.4%, respectively, while there is no significant change in the trade value of logs, wood pulp, and wood furniture.

Therefore, it can be argued that the gravitational model explains China’s forest products trade mainly at the level of trade in processed products, while the explanation of raw materials trade is relatively weak; similarly, the promotion effect of trade countries joining the Belt and Road Initiative on China’s forest products trade is also mainly reflected in the impact on trade in processed products. The reason for this is due to the characteristics of different forest products. As mentioned above, wood raw material products are highly valued by the world’s major forest products trading countries due to their strategic resource properties, resulting in relatively small supply and demand elasticity, so they are not easily affected by economic fluctuations and policy shocks. This is not the case for processed wood products, whose commodity attributes are more prominent. The Belt and Road Initiative has fully enhanced the level and capacity of linkage development between China and the countries along the route, bringing the potential and effectiveness of the domestic and international markets into full play, thus realizing the growth of the scale of processed wood products between China and the countries along the route.

3.3. Regional Heterogeneity Analysis

In this paper, we introduce the interaction term of the policy dummy variable and the continent dummy variable based on Equation (2) and then perform the regression. The estimation results are reported in

Table 4. Effect of introducing trade size effects of regional heterogeneity.

Product	Import (Lnimp)					Export (Lnexp)					Total (Lntrd)
	Asia	America/Oceania	Europe	Africa	R2	Asia	America/Oceania	Europe	Africa	R2	Asia	America/Oceania	Europe	Africa	R2
All	0.029	0.241	0.785 ***	0.497 **	0.886	0.289 ***	0.335 ***	0.195 *	0.316 ***	0.959	0.124 **	0.159	0.438 ***	0.120 *	0.952
	(0.18)	(1.21)	(5.41)	(2.52)		(5.81)	(2.86)	(1.77)	(2.95)		(2.11)	(1.54)	(5.22)	(1.87)
Logs	−1.129 ***	0.895 *	−0.065	0.731	0.796	0.400				0.583	−1.610				0.504
	(−2.59)	(1.69)	(−0.16)	(1.39)		(0.24)					(−0.71)
Sawn Timber	−0.217	0.115	0.898 ***	0.841 ***	0.916	0.137	0.488	−0.489	0.114	0.871	−0.344 **	0.491 **	0.837 ***	0.984 ***	0.921
	(−0.94)	(0.41)	(4.36)	(3.03)		(0.78)	(1.18)	(−1.26)	(0.30)		(−2.28)	(2.06)	(4.55)	(3.37)
Wood Pulp	−0.049	−0.566	0.524	−0.728	0.813	0.606 **	−0.702	0.569	0.470	0.777	−0.048	−0.067	−0.321	−0.142	0.916
	(−0.08)	(−0.71)	(0.83)	(−0.51)		(1.99)	(−0.46)	(0.86)	(0.74)		(−0.18)	(−0.06)	(−0.68)	(−0.21)
Wood-based Panels	0.198	0.446	1.261 ***	1.594 ***	0.830	0.276 **	0.659 **	−0.021	0.336	0.856	0.215 **	0.139	0.401 ***	0.785 ***	0.927
	(0.59)	(0.89)	(4.26)	(3.59)		(2.54)	(2.57)	(−0.09)	(1.43)		(1.98)	(0.57)	(2.76)	(3.51)
Wood Products	0.438	0.054	0.980 ***	0.071	0.829	0.394 ***	0.209	0.250	0.174	0.949	0.314 ***	−0.147	0.151	−0.262	0.961
	(1.24)	(0.11)	(2.99)	(0.14)		(5.48)	(1.23)	(1.57)	(1.12)		(3.92)	(−0.91)	(1.35)	(−1.57)
Wood Furniture	0.378	1.196 **	−0.301	−0.376	0.922	−0.111	0.310	0.273	0.263	0.916	−0.190 *	0.381	0.088	0.625 **	0.935
	(1.45)	(2.57)	(−1.25)	(−0.55)		(−1.14)	(1.35)	(1.27)	(1.26)		(−1.77)	(1.38)	(0.59)	(2.21)
Paper Products	0.984 ***	−1.186 ***	0.765 **	1.017 *	0.933	0.373 ***	0.214 **	0.234 **	−0.015	0.969	0.409 ***	0.132	0.368 ***	0.060	0.977
	(2.89)	(−2.75)	(2.52)	(1.93)		(8.93)	(2.17)	(2.53)	(−0.16)		(7.63)	(1.35)	(4.98)	(0.48)

(1) lngdp, lnpop, country fixed effect and year fixed effect are controlled in all models; (2) due to the limitation of space, the estimated values of lngdp, lnpop and constant terms are omitted in the table; (3) adjusted R² reported in the table includes the contribution of fixed effect; (4) values in brackets indicate t-values corresponding to estimated coefficients, calculated through robust standard errors of clustering at country level; (5) *, **, and *** indicate statistical significance levels of 10%, 5%, and 1%, respectively.

, where the dependent variable in the first column is the import and export trade value of seven types of forest products combined, and the dependent variables in the last seven columns are the trade value of each type of forest product individually.

First, observe the estimation results of the forest products trade model for the first column summation. The estimation results of the import value model can be interpreted that the accession of trading countries to the Belt and Road Initiative significantly increases China’s forest products imports from European and African trading partner countries by 78.5% and 49.7%, respectively, while there is no significant change in imports from Asia, the Americas, and Oceania. The estimation results of the export value model indicate that the accession of trading countries to the Initiative significantly increased China’s forest products exports to Asia, America or Oceania, Europe, and Africa by 28.9%, 33.5%, 19.5%, and 31.6%, respectively. The estimation results of the total trade value model indicate that the accession of trading countries to the Initiative significantly increased China’s trade value of forest products with Asia, Europe, and Africa by 12.4%, 43.8%, and 12.0%, respectively. The above results confirm that there is very significant regional heterogeneity in the treatment effect of the Initiative on forest products trade. The impact of the Initiative on forest product imports is mainly concentrated in European and African countries, while the regional heterogeneity of the impact effect on forest product exports is relatively weak, with significant increases in exports to different regional trading partners.

Then, the impact of the latter seven columns on the trade of various forest products is observed. In the import value model, the impact of the Belt and Road Initiative on China’s imports of logs, sawn timber, wood-based panels and wood products is mainly reflected in European and African countries, and the impact on the trade of wood furniture is mainly reflected in American or Oceanian countries, while the impact on the trade of paper products shows a significant structural adjustment, with imports from Asian, European and African countries rising and imports from American or Oceanian countries falling. The estimation results of the export value model show that the impact of the Belt and Road Initiative on China’s exports of wood pulp and wood products is mainly reflected in Asian countries, the impact on the exports of wood-based panels in the Americas Oceania and Asian countries, and the impact on the exports of paper products in Asian and European countries. In the total trade value model, the impact of the Belt and Road Initiative on the trade of forest products between China and Asian countries shows an increase in paper products, wood products and wood-based panels and a decrease in sawn timber and wood furniture; the impact on China and European countries shows an increase in the trade of sawn timber, wood-based panels and paper products, and the impact on China and African countries shows an increase in the trade of sawn timber, wood-based panels and wood furniture.

Through the above comparison, it can be found that under the impact effect of the Belt and Road Initiative on China’s forest products trade in terms of overall promotion, there are very significant regional differences and structural product differences at the same time. In terms of total trade, the highest increase is in China’s trade with European forest products; in terms of individual products, the increase in China’s trade in African-specific forest products is also very obvious, while China’s trade with Asian countries shows a structural adjustment with some products increasing and some products decreasing, in contrast, China’s trade with American and Oceanian countries is the least affected. This paper speculates that the specific reasons are as follows: First, Europe and Africa as the world’s most and least developed economic regions, respectively, and their forest industry development structure is quite different from China, thus contributing to the bilateral trade has a very high complementarity. China has exported low-end processed wood products to European and African countries for a long time, at the same time importing high-end wood products from Europe and Africa’s raw wood materials. The promotion of the Belt and Road Initiative has led to the continuous optimization of the regional trade environment and the further enhancement of bilateral and multilateral trade complementarity; secondly, Asian countries are mostly emerging forest products trading countries whose development structure of forest industry is similar to China’s, and the bilateral has long maintained a state of both cooperation and competition; thirdly, the Belt and Road Initiative has always taken Asia, Europe and Africa as the key promotion areas in the process of promotion, and the Americas and Oceania have not been effectively developed. This is also a key area for the future of the Belt and Road Initiative

4. Analysis of the Mediating Effects of Policy Effects

According to the previous section, there are multiple possible mechanisms for the impact of the Belt and Road Initiative on international trade. This paper verifies whether the mediation effect is established from three aspects: the state of logistics infrastructure, political partnership and Internet penetration (

Table 5. Mediating effect test of the effect of the Belt and Road Initiative on China’s forest products trade.

Variable	Import (Lnimp)					Export (Lnexp)					Total (Lntrd)
	Lny (1)	Lp (2)	Pt (3)	Internet (4)	Lny (5)	Lny (1)	Lp (2)	Pt (3)	Internet (4)	Lny (5)	Lny (1)	Lp (2)	Pt (3)	Internet (4)	Lny (5)
D	0.457 ***	0.073 ***	0.205 *	5.891 ***	0.312 ***	0.284 ***	0.119 ***	0.249	6.449 ***	0.210 ***	0.209 ***	0.091 ***	0.221	7.120 ***	0.121 ***
	(4.51)	(2.62)	(1.96)	(6.45)	(3.62)	(6.45)	(5.05)	(1.37)	(5.62)	(5.61)	(4.33)	(3.94)	(1.60)	(6.00)	(3.02)
lngdp	0.218	0.149 ***	−0.709 *	2.310	0.229	0.991 ***	−0.003	−0.297	−1.000	0.993 ***	0.533 ***	0.127 ***	−0.932 **	0.727	0.490 ***
	(1.13)	(2.82)	(−1.93)	(1.33)	(1.18)	(11.27)	(−0.07)	(−0.82)	(−0.44)	(11.35)	(5.20)	(2.60)	(−2.39)	(0.29)	(4.78)
lnpop	−1.396 *	−0.391 **	4.812 ***	−5.369	−1.494 **	0.309	0.241 *	4.200 ***	57.258 ***	0.075	1.009 **	−0.356 *	6.111 ***	28.549 ***	1.008 **
	(−1.95)	(−1.99)	(3.53)	(−0.83)	(−2.07)	(1.14)	(1.66)	(3.76)	(8.11)	(0.26)	(2.42)	(−1.80)	(3.86)	(2.79)	(2.40)
lp					0.288					−0.047					0.349 ***
					(1.18)					(−0.56)					(3.70)
pt					0.201 ***					0.085					0.029
					(2.86)					(0.88)					(0.36)
Internet					0.014 ***					0.009 ***					0.007 *
					(2.78)					(2.88)					(1.92)
C	40.232 ***	8.527 ***	−69.864 ***	117.870	41.660 ***	5.745	−0.942	−64.700 ***	−935.497 ***	9.743 **	−1.782	8.226 **	−92.199 ***	−453.750 ***	−2.672
	(3.38)	(2.61)	(−3.08)	(1.10)	(3.46)	(1.26)	(−0.38)	(−3.43)	(−7.85)	(2.01)	(−0.26)	(2.48)	(−3.48)	(−2.65)	(−0.38)
R2	0.883	0.961	0.696	0.970	0.904	0.959	0.960	0.702	0.945	0.969	0.951	0.971	0.705	0.951	0.959
F	20.218	4.351	2.413	96.354	17.352	67.236	7.554	3.839	99.861	55.896
p-Value	0.000	0.000	0.004	0.000	0.000	0.000	0.000	0.000	0.000	0.000

(1) lny in the table denotes lnimp, lnexp and lntrd in the samples of importing country, exporting country and trading country, respectively; (2) country fixed effect and year fixed effect are controlled in all models, and the adjusted R² reported in the table includes the contribution of fixed effect; (3) values in brackets denote t-values corresponding to estimated coefficients, calculated by robust standard errors of clustering at the country level; (4) *, **, and *** indicate statistical significance levels of 10%, 5%, and 1%, respectively.

). For comparison purposes, the estimated results in Table 2 reflecting the treatment effect of the Belt and Road Initiative on forest products trade are replicated in column 1 of Table 5. Columns 2, 3, and 4 report the model estimation results with the three mediating variables as dependent variables according to Equation (5), and column 5 reports the model estimation results with the introduction of mediating variables in Equation (6).

The effects of the Belt and Road Initiative on the mediating variables are first observed in columns 2, 3, and 4, that is, the estimated results of the coefficient a in Equation (5). In the sample of importing countries, trading countries’ joining the Belt and Road Initiative has a significant boosting effect on the state of logistics infrastructure, political partnership with China, and Internet penetration. In the sample of exporting countries and the sample of trading countries, the trading countries’ joining the Belt and Road Initiative has a significant boosting effect only on the state of logistics infrastructure and Internet penetration, and the effect on political partnership is not confirmed. In terms of estimated coefficients, joining the Belt and Road Initiative improved the logistics performance of China’s forest products trading partners by an average of 0.07–0.12 points (total score of 5); increased the domestic Internet penetration rate of China’s forest products trading partners by 5.89 to 7.12 percentage points; and increased the political partnership between China and its forest products importing countries by an average of 0.21 units.

Then observe the estimation results of coefficient b and coefficient θ′ in column 5. In the sample of importing countries, coefficient b of the effect of the state of logistics infrastructure on the import value of forest products is not significant, while the coefficient b of the effect of political partnership and Internet penetration on the import value of forest products is both statistically significant at the 1% level. Controlling for other variables, each one-unit increase in political partnerships or each one-percentage-point increase in Internet penetration increases China’s forest product imports from trading countries by 20.1% and 1.4%, respectively. After controlling for mediating variables, the treatment effect of the Belt and Road policy dummy variable D on forest product imports in the model decreases from 45.7% to 31.2%. In other words, the direct effect of the Belt and Road Initiative on forest product import trade is 31.2%, and the indirect effect through the three mediating variables is 14.5%, with the direct and indirect effects accounting for 68.3% and 31.7% of the total effect, respectively. Calculating the indirect effects of the three mediating variables shows that the indirect effect brought about by Internet penetration has the highest proportion, reaching 18.0% of the total effect; the indirect effect of political partnership has 9.0%.

In the sample of exporting countries, the coefficients of the effects of the state of logistics infrastructure and political partnership on the export value of forest products are not significant, and only the coefficient b of the effect of Internet penetration on the export value of forest products is statistically significant at the 1% level. For every 1 percentage point increase in Internet penetration in trading countries, China’s forest products exports increase by 0.9% accordingly. After controlling for the mediating variables, the treatment effect of the Belt and Road policy dummy variable D on the export value of forest products in the model decreases from 28.4% to 21.0%. That is, the direct effect of the Belt and Road Initiative on forest product export trade is 21.0%, and the indirect effect through the three mediating variables is 7.4%, with the direct and indirect effects accounting for 74.1% and 25.9% of the total effect, respectively. Among them, the indirect effect brought by Internet penetration has the highest proportion, reaching 20.4% of the total effect.

In the total trade sample, the effect of political partnerships only on the trade value of forest products is not significant, and the effect of the state of logistics infrastructure and Internet penetration on the trade value of forest products are both statistically significant at the 1% level. After controlling for the mediating variables, the treatment effect of the Belt and Road policy dummy variable D in the model decreases from 20.9% to 12.1%. That is, the direct effect of the Belt and Road Initiative on the trade value of forest products is 12.1%, and the indirect effect through the three mediating variables is 8.8%, with the direct and indirect effects accounting for 57.9% and 42.1% of the total effect, respectively. Among them, the indirect effect brought by Internet penetration has the highest proportion, reaching 20.4% of the total effect; the indirect effect brought by logistics performance is 15.2%.

In summary, the three mediating variables selected in this paper do play a certain degree of mediating effect in the promotion effect of the Belt and Road Initiative on China’s forest products trade. In contrast, the mediating effect of logistics performance is not robust enough, mainly because the coefficient b is not significant enough, and the mediating effect of political cooperation is only confirmed in the sample of importing countries but not in the other two samples, mainly because the coefficient a and the coefficient b are not significant.

The reason for the empirical results of the mechanism analysis is determined by the distinctive characteristics of the Belt and Road Initiative. First of all, the significance of the results of logistics infrastructure status variables is mainly reflected in the overall trade level. Most of the countries along the Belt and Road are developing countries with relatively backward economic levels and infrastructure construction. Infrastructure is an important prerequisite for the long-term stable growth of a country’s product production and trade scale. China has invested a lot in infrastructure in these countries, which has greatly improved the level of facility connectivity in the regions along the route and promoted the development of overall bilateral trade to a certain extent. The direct benefits of infrastructure are not obvious, and the income cycle is often very long, so the impact on the sub-trade level, such as import and export, still needs time to be verified. Bilateral political relations are significantly positive only when China imports, indicating that China’s policy communication’ with countries along the route has made a strong contribution to the import of China’s forest products and the safety of domestic timber supply. In the future, policy communication should be the guide to promote China’s forest products to enter the markets of more countries along the route. The variable of Internet popularity is significantly positive at the overall import and export levels. In the process of promoting the Belt and Road Initiative, China not only overflows its mature network and other infrastructure construction processes to underdeveloped areas but also actively shares its own development experience with other co-founding countries. In today’s world, the speed of technological upgrading is accelerating, and emerging technologies such as the digital economy are gradually leading the fourth wave of the industrial revolution. The rapid development of Internet infrastructure makes enterprises more motivated and capable of continuously improving production processes and increasing the scale of production and trade.

5. Discussion

As an important platform for China’s international economic and trade cooperation, the Belt and Road Initiative is a successful practice for China to “bring in” and “go out,” and the high-quality development of international trade is an important part of China’s “high-quality development” strategy. The high-quality development of international trade is an important part of China’s “high-quality development” strategy. This paper explores the economic impact of the Belt and Road Initiative from the perspective of upgrading the scale of China’s forest products trade and deeply explores the policy effects of the Belt and Road Initiative on the economic and trade development of China and the countries along the route, opening up a new perspective for studying the upgrading of China’s international trade and evaluating the effects of the Belt and Road Initiative. Compared with existing studies, this paper focuses on the analysis of the following scientific issues:

First, the product heterogeneity analysis of policy effects. Different types of wood forest products differ from the perspective of the production process, factor inputs, and uses. Wood forest products can be divided into raw materials and processed products according to their processing degree and can be divided into resource-intensive products, capital-technology-intensive products and labor-intensive products according to the input of factors in the production process. There are obvious differences in the import and export structure of different types of forest products in China. Then, it is entirely possible that there are structural differences in the import and export trade of seven different types of forest products in China as a result of trade partners joining the Belt and Road Initiative. The empirical analysis shows that the policy promotion effect of the Belt and Road Initiative on the development of China’s forest products trade is mainly manifested at the product level for wood processed products, and the effect on wood raw materials is very weak.

Secondly, the regional heterogeneity differences of the policy effects. There are two initial purposes for examining regional heterogeneity. On the one hand, there are obvious geographical differences in the distribution of forestry resources and the layout of forest products processing industries in the world, and on the other hand, the geographical direction of China’s promotion of the Belt and Road Initiative is gradually expanding from Europe and Asia to Africa, America and Oceania. From this starting point, it can be argued that the impact of China’s wood forest products trade by policy shocks may also vary across regions. The empirical results confirm that there is significant regional heterogeneity in the policy effects of the Belt and Road Initiative, and the policy effects are mainly concentrated in Europe, Africa and Asia.

Thirdly, the mechanistic mechanism by which the policy effect works. The impact of the Belt and Road Initiative on the import and export of forest products in China may vary due to the differences in the state of logistics infrastructure, Internet penetration and bilateral political relations of the trading countries. This paper uses these conditions as moderating variables for the first time and introduces their interaction terms with policy variables in the model. This proves that China’s policy orientation of “discussing, sharing and building together” and achieving mutual benefit and a win-win situation with countries along the route is the only way to promote the Belt and Road Initiative and achieve the long-term development of Chinese forest products.

The research in this paper still has the following shortcomings and directions for further research: first, since the Belt and Road Initiative was proposed in late 2013 and formally implemented in 2014 for less than 9 years, this paper actually assesses a short-term impact effect of the Belt and Road Initiative on the development of China’s forest products trade. In fact, it may take longer to present the full picture of the Belt and Road policy, so an assessment of the long-term policy effects on this topic is a way forward. Secondly, there are many countries along the Belt and Road, with different levels of economic development and national income, so this paper does not analyze in detail the impact of the Belt and Road Initiative on countries with different levels of economic development, which could be a future This could be a complementary study in the future.

6. Conclusions

Taking the Belt and Road Initiative as a policy event, this paper makes a causal evaluation of the impact of the participation of trading partners in the Belt and Road Initiative on China’s forest products trade, focusing on the product heterogeneity and regional heterogeneity of the impact. And the mechanism of the occurrence of the impact effect was analyzed by using the logistics performance of trading countries, political cooperation relationship with China and Internet penetration as mediating variables. The following main conclusions were obtained.

First, it is confirmed that the method of constructing a difference-in-differences model based on the gravity model is applicable to the study of China’s forest products trade and the effects of the Belt and Road Initiative. The results show that, on the one hand, the participation of trading partners in the Belt and Road Initiative has played a significant role in promoting the growth of bilateral forest products trade. This confirms that the Belt and Road Initiative meets the development needs of China and the countries along the route and is an important link to achieve mutual benefits and win-win results; on the other hand, the promotion effect of the Initiative on China’s imports is stronger than that on China’s exports.

Secondly, there are obvious product heterogeneity and regional heterogeneity effects on the effect of trade partners’ joining the Belt and Road Initiative on China’s forest products trade. From the product dimension, the gravity model explains China’s forest products trade mainly in terms of trade in processed products. From the regional dimension, the total trade of forest products between China and Europe and Africa has the relatively highest increase, and the trade of forest products between China and Asian countries shows a structural adjustment, with some products increasing while some products decreasing, compared to the trade of forest products between China and American Oceanian countries, which is least affected by the initiative.

Thirdly, the state of the logistics infrastructure of trading countries, political cooperation with China, and Internet penetration play significant positive mediating roles in the promotion effect of the Belt and Road Initiative on China’s forest products trade. Among them, Internet penetration plays the highest proportion of the mediating role. In contrast, the mediating effect of the state of logistics infrastructure is confirmed only in the total trade sample frame, while the mediating effect of political partnership is confirmed only in the importing country sample frame.

Based on the above findings, this paper can get the following enlightenment: first, in view of the significant role of the Belt and Road Initiative in promoting the development of bilateral forest products trade, China should promote more countries, especially developing countries with more trade potential, to participate in the joint construction of the Belt and Road, while the existing Belt and Road co-construction countries can improve the cooperation mechanism and jointly explore new dynamics of international trade development, so as to obtain more fruitful new dividends; secondly, in view of the different impacts of the Belt and Road Initiative on different forest products in different countries along the route, China should adjust measures according to the specific time, region and country, focus on key countries, priority areas and key products to make key breakthroughs. Finally, based on the test results of the impact mechanism, China should continue to take “policy communication” as a guide and enhance its bilateral political relations with countries along the route, adhere to the “facility connectivity” as the foundation, give consideration to the construction of “hard connectivity” based on logistics and transportation and “soft connectivity” based on network communication, effectively improve the level of regional trade facilitation along the route, and create a good business environment for the development of bilateral forest products trade.