Measurement and Validation of Market Power in China’s Log Import Trade—Empirical Analysis Based on PTM Model and AIDS Model

Abstract

China is the world’s largest importer of logs, possessing the scale to exert significant influence in the international market. This paper uses a fixed-effect variable coefficient Pricing-to-Market panel model to measure China’s market power in log import trade. It also utilizes the Almost Ideal Demand System model from an elasticity perspective to explore the market behavior characteristics of various source countries in China’s log import trade, thereby validating the mechanism of market power. The results indicate that: ① China’s main trading partners can be categorized into four groups according to their market power in the log import trade. Specifically, China holds superlative market power in log imports from Indonesia, Malaysia, and Myanmar; holds strong market power in log imports from Russia, the Democratic Republic of the Congo, and Mozambique; holds weak market power in log imports from Papua New Guinea, Equatorial Guinea, France, Germany, Australia, and New Zealand; and holds no market power in Japan, Cameroon, and the United States. ② As China’s expenditure on log imports increases, there is a tendency to purchase high-quality precious wood and a greater concern for the legality of market transactions. Consequently, China is anticipated to augment its imports from source countries with no or weak market power. ③ The simple price elasticity of log imports from each source country is negative. Source countries with stronger market power tend to increase prices to achieve higher total revenue, while those with weaker market power are more inclined to lower prices to achieve the same. ④ Log products from various source countries are complementary in the Chinese market, indicating that China’s substantial demand for logs relies on the simultaneous supply from multiple countries and diverse wood types. Based on the existence or absence of market power in China’s log import trade, this paper provides targeted insights into enhancing international market power and reducing trade losses.

1. Introduction

China is the world’s largest importer of logs (Gao et al., 2024) [1]. According to the latest statistics from the UN Comtrade database, China’s log import volume (HS code: 4403) has increased from 16.86 million m³ in 2001 to 34.14 million m³ in 2023, with an annual growth rate of 3.26%. China’s share in global log import trade has also grown from 14.33% in 2001 to 36.39% in 2023 (since the trade volume of Hong Kong, Macao, and Taiwan is counted separately, the term “China” in this paper refers only to mainland China). For the year 2023, the top five countries by share in global log import trade are China (36.39%), Austria (8.12%), Sweden (7.78%), India (5.71%), and Germany (4.77%) [2]. This stark difference highlights the substantial disparity between China’s share and that of the other top four countries, indicating that China holds a dominant position in the international log import market. Given this dominance, it is crucial to examine whether China possesses significant market power in any of its source countries and whether its bargaining power is proportionate to its import volume. Additionally, it is important to assess whether China can leverage its market power to effectively manage international price fluctuations and mitigate the negative impacts of global market dynamics. These factors directly influence China’s economic benefits and risks in the international log trade arena.

Currently, scholars have conducted extensive research on how to measure international market power, with two mainstream approaches. The first is indicator-based measurement, primarily adopted by the Structuralist school, which uses various econometric techniques to reflect market concentration. This generally includes market structure-based indicators, such as the international market concentration index and the Herfindahl-Hirschman Index (Utton et al., 1983) [3]; market performance-based indicators, such as profit margin, Bain Index, and Lerner Index (Bain, 1951; Demsetz, 1973) [4,5]; and market behavior-based indicators, such as core technology ownership rate (Berger et al., 1998) [6]. However, because there is a lack of a theoretical foundation linking market share to market power, a high market share does not necessarily indicate the presence of market power. Moreover, in practical applications, marginal costs are difficult to quantify, and relying solely on this method can easily lead to erroneous conclusions. As a result, the frequency of using this approach is relatively low. The second is model-based measurement, mainly adopted by New Empirical Industrial Organization (NEIO) theory, which measures market power by deriving economic models. These include the Price-Cost Margin (PCM) model (Loecker, 2011) [7], the Residual Demand Elasticity (RDE) model (Goldberg et al., 1999; Engel, 2006; Hellerstein, 2008; Nakajima, 2012) [8,9,10,11], the Two-country Partial Equilibrium Model (SMR model) (Song et al., 2009; Yamaura, 2011; Lv et al., 2023) [12,13,14], and the Pricing-to-Market (PTM) model based on Exchange Rate Pass-Through (Krugman, 1986; Dizgah et al., 2019) [15,16]. Among these, the PCM model is based on Solow residual theory and relies on statistical parameters, providing relatively accurate results. However, the model’s application requires strict assumptions, such as constant returns to scale and Hicks-neutral technological progress, and demands high-quality data, making it difficult to implement in empirical studies (Kim et al., 2017) [17]. In comparison, the RDE and PTM models have a solid theoretical foundation and are easier to apply empirically, making them widely adopted in practical research (Corsetti et al., 2022) [18]. Building on the RDE model, Song et al. (2009) [12] introduced the inverse residual demand function, inverse residual supply function, and supply-demand equilibrium conditions, thereby constructing the SMR model. This model allows measurement from both the buyer’s and seller’s perspectives simultaneously and is considered one of the most scientifically robust models for calculating international market power (Zhu et al., 2019; Wang et al., 2023) [19,20]. However, the SMR model focuses on a limited number of major trading countries, narrowing its scope of analysis. In contrast, the PTM model comprehensively considers the impact of market structure, consumer utility, and other factors, making it suitable for examining the market power of importers across all trading countries, which more closely aligns with the research objectives of this study (Fitzgerald et al., 2014; Chen et al., 2023) [21,22].

Research on China’s international market power by scholars has predominantly focused on sectors such as agricultural products (Song et al., 2009 [12]; Dai et al., 2021 [23]; Chen et al., 2022 [24]; Zheng et al., 2022 [25]; Lv et al., 2023 [14]; Yan et al., 2023 [26]), mineral resources (Zhu et al., 2019 [19]; Zhu et al., 2019 [27]; Marz et al., 2023 [28]), electricity markets (Cheng et al., 2023; Ji et al., 2023; Lin et al., 2023) [29,30,31], medical product markets (Wu et al., 2022) [32], and food and tobacco markets (Dai et al., 2018) [33]. In contrast, research on the timber trade is relatively limited. Given this gap, this study uses the PTM panel model to measure China’s market power in log imports.

The presence of market power in China’s log imports is influenced not only by trade dynamics between China and the source countries but also by the competitive relationships among these source countries (Zhu, 2023) [34]. Thus, after assessing the market power of China’s log imports, it is essential to further investigate the market behavior characteristics of these source countries. Research on differentiated import demand, where the focus is on the demand for products from different source countries, frequently utilizes models such as the Rotterdam model, the Almost Ideal Demand System (AIDS) model, and the General model. Among these, the AIDS model is particularly advantageous due to its flexible functional specification, which allows for the simultaneous examination of consumer behavior and market characteristics, making it highly relevant for studies in international trade (Thompson, 2013; Michael, 2020; Tereza et al., 2022) [35,36,37]. Hence, this study uses the AIDS model to analyze the market behavior of the source countries for China’s log imports, from an elasticity perspective, to validate the mechanisms of market power. Based on the existence or non-existence of market forces, this study draws on the development experience of countries with mature market forces and proposes policy recommendations for China’s timber import trade to mitigate trade losses.

Compared to existing literature, this study may offer the following innovations: First, it uniquely combines market power analysis with elasticity analysis, validating the mechanisms of market power from an elasticity perspective. Second, it calculates and classifies the market power of all log import source countries, addressing the limitation of previous research, which has been restricted to individual markets and wooden furniture products.

2. Materials and Methods

2.1. Theoretical Analysis

Industrial organization theory defines international market power as the comprehensive control capability of a nation’s industry in the international market (Farrell et al., 1985) [38]. Essentially, it is the outcome of global industries competing with each other in terms of productivity and technological levels under conditions of an open economy. This study primarily focuses on two questions, which will be discussed in the following paragraphs.

2.1.1. Whether Market Power Exists

Early research often conflated market power with monopoly power, although the two concepts are fundamentally different. Both relate to a firm’s ability to price above marginal cost, but they diverge significantly in nature. Unlike monopoly power, market power typically does not lead to profits that exceed the average competitive level, and it is generally short-lived, unable to be sustained in the long term. Consequently, firms with market power should not be the primary targets of antitrust enforcement. Moreover, subsequent studies have shown that a firm possessing monopoly power does not necessarily have market power, and conversely, a firm with market power does not always hold monopoly power. This further underscores that market power is not equivalent to monopoly power (Bannock, 2005) [39]. From a microeconomic perspective, the mechanism by which international market power operates is illustrated in Figure 1. In this diagram, D, MC, MR, and AR represent the demand curve (downward-sloping), marginal cost curve, marginal revenue curve, and average revenue curve (which coincides with the demand curve), respectively. According to the profit-maximization condition in an imperfectly competitive market (MR = MC), the firm’s equilibrium output and price should occur at point E, where MR intersects MC, corresponding to Q^m and MC^m. However, since the D curve lies above the MR curve, the firm can set its transaction price based on the demand curve at P^m. When P^m − MC^m > 0, the firm has pricing power, indicating the presence of market power. The greater the value of P^m − MC^m, the stronger the firm’s market power, which is visually depicted by the vertical distance between P^m and MC^m in the figure (Young, 2000) [40]. This relationship is intricately linked to the price elasticity of demand (ε), with the price-cost markup expressed as (P^m − MC^m)/P^m = −1/ε. The left-hand side of the equation represents the price-cost markup.

2.1.2. Market Behavior Characteristics of Import Source Countries

When analyzing the market behavior characteristics of the source countries, several key aspects necessitate consideration: the sequence in which source countries benefit as China’s log import expenditures increase, changes in the structure of China’s log import market, the sensitivity of log exports from each source country to price changes, and whether products from different source countries act as substitutes or complements. This analysis will provide insights into the mechanisms of market power from an elasticity perspective, thereby enhancing the understanding of the estimated results pertaining to market power.

The AIDS model, owing to its advantages of superiority and practicality, as discussed in the introduction, has been widely utilized in empirical research concerning consumption issues. This model, rooted in simultaneous equations, thoroughly accounts for consumers’ market behavior and bears distinct economic significance. For an exhaustive explanation of the model’s specific construction process, kindly consult Section 2.3.2.

2.2. Data Source

This study selected countries that accounted for more than 1% of China’s total log imports from January 2001 to December 2023 as the research sample. The countries included are Australia, Canada, Cameroon, the Democratic Republic of the Congo (DRC), Equatorial Guinea, France, Gabon, Germany, Indonesia, Japan, Laos, Malaysia, Mozambique, Myanmar, New Zealand, Papua New Guinea, Russia, the Solomon Islands, and the United States. Specifically: ① The log prices from each source country are calculated based on trade value per trade volume. The data on trade value and trade volume for logs imported from these countries are sourced from the National Research Network statistical database [41]. ② Nominal exchange rate data for these countries are obtained from the Economy Prediction System (EPS) database [42]. ③ The real exchange rate is calculated as the nominal exchange rate multiplied by China’s monthly Consumer Price Index (CPI) divided by the source country’s monthly CPI. Due to the lack of a specific log price index, and because previous studies have shown that results derived from CPI calculations are less accurate than those from nominal exchange rates (Yumkella et al., 1994; Wang et al., 2017) [43,44], this method is used. As a robustness check, the real exchange rate calculation results are also presented, with CPI data for the countries sourced from the BVD-EIU Countrydata database [45].

2.3. Methods

2.3.1. PTM Model

Krugman (1986) introduced PTM theory. PTM induced by exchange rates occurs when a change in bilateral exchange rates between an exporter and several importers alters the price ratio paid by the importers [15]. Building upon this theoretical framework, Goldberg et al. (1999) devised a representative export pricing strategy model in accordance with the profit maximization principle [8]. Consider a scenario where an exporting nation supplies goods to N destination countries, with the demand in each country defined as follows:

$$q_{it}=f\left(p_{it}{e}_{it}\right)z_{it}\hspace{1em}\forall i=1\dots N,\forall t=1\dots T$$

(1)

In Equation (1), q_it represents the demand from destination country i in period t; p_it denotes the export price set by the exporter for i, expressed in the exporter’s domestic currency in period t; e_it signifies the bilateral exchange rate between the exporter and destination country i in period t, expressed as the number of units of the destination country’s currency exchanged for one unit of the exporter’s currency; z_it is a random variable accounting for shifts in the demand curve. Additionally, the production costs incurred by exporting countries are given by:

$$C_t=C\left({\displaystyle \sum _{i=1}^N{q}_{it}}\right){\delta }_t\hspace{1em}\forall i=1\dots N,\forall t=1\dots T$$

(2)

In Equation (2), C_t represents the aggregate costs associated with all destination countries i, expressed in the currency of the exporting country; and δ_t is a random variable that induces changes in the cost function, such as variations in the price of an input product in period t. Consequently, the profit maximization of the exporting country can be articulated as follows:

$$\max \pi ={\displaystyle \sum _{i=1}^N\left[p_{it}f\left(p_{it}{e}_{it}\right)z_{it}\right]}-C\left\{{\displaystyle \sum _{i=1}^N\left[f\left(p_{it}{e}_{it}\right)z_{it}\right]}\right\}{\delta }_t$$

(3)

In Equation (3), taking the derivative with respect to p_it and expressing it in terms of elasticity, the first-order condition is obtained as follows:

$$p_{it}=c_t\left({\displaystyle \frac{{\epsilon }_t^i}{{\epsilon }_t^i-1}}\right)\hspace{1em}\forall i=1\dots N,\forall t=1\dots T$$

(4)

In Formula (4), c_t represents the common marginal cost faced by the exporting country in period t, and ${\epsilon }_t^i$ denotes the absolute value of the price elasticity of demand faced by the exporting country in destination country i in period t. Therefore, the price expressed in the exporter’s currency should be equal to the price-cost markup, with the markup depending on the demand elasticity faced by the exporter in destination market i (Varma et al., 2016) [46]. If the demand elasticity in destination country i is not constant, then changes in the exchange rate between the exporting country and the destination country will affect the transaction price by influencing either the marginal cost or the demand elasticity. Specifically, changes in marginal cost will impact other destination countries, while changes in demand elasticity will only affect the destination country where the exchange rate change occurs (Dawson et al., 2017) [47]. Taking the natural logarithm of Formula (4) and performing a total differential yields the final form of the export PTM model as follows:

$$\ln p_{it}=\alpha +{\beta }_i\ln e_{it}+{\lambda }_i+{\theta }_t+{\mu }_{it}\hspace{1em}\forall i=1\dots N,\forall t=1\dots T$$

(5)

In Formula (5), p_it represents the price in the exporter’s currency for exports from the exporting country to country i in period t; β_i is the parameter to be estimated, reflecting the elasticity of export prices to exchange rate fluctuations and serving as a measure of the exchange rate pass-through effect in destination country i; λ_i captures country-specific effects; θ_t represents the time effect as a monthly dummy variable to control for seasonal influences, thus eliminating the need for seasonal adjustment of the price and exchange rate series; and μ_it is the error term.

Building on the export pricing strategy model developed by Goldberg et al. (1999) [8], Manitra et al. (2001) [48] introduced the import PTM model. The specific form of the PTM model is as follows:

$$\ln r_{jt}=\phi +{\varphi }_j\ln e_{jt}+{\lambda }_j+{\theta }_t+{\mu }_{jt}\hspace{1em}\forall j=1\dots N,\forall t=1\dots T$$

(6)

In Equation (6), r_jt represents the log price in yuan per cubic meter for imports from country j to China at time t; e_jt represents the bilateral exchange rate between China and source country j, expressed as the amount of RMB (the lawful currency of China) per unit of j’s currency; φ_j is the parameter to be estimated, reflecting the elasticity of log import prices with respect to exchange rate fluctuations and measures the exchange rate pass-through effect for country j; λ_j represents country-specific effects; θ_t accounts for time effects as a monthly dummy variable to control for seasonal variations, thus eliminating the need for seasonal adjustment of the price and exchange rate series; and μ_jt is the error term. Some scholars have attempted to extend the PTM model by including additional control variables such as GDP. However, results indicate that the original PTM model by Goldberg et al. (1999) [8] provides more reliable outcomes. This is primarily because, from a multilateral trade perspective, the PTM model’s inclusion of individual and time effects already accounts for unobservable heterogeneity. Adding additional control variables may reduce degrees of freedom and impact the accuracy of the model’s results (Griffith et al., 2001 [49]; Pall et al., 2013 [50]; Wang et al., 2017 [44]).

For ease of presentation, we adopt the definitions proposed by Goldberg et al. (2008) [51] as follows: Exchange Rate Pass-Through Effect (ERPT*) reflects the impact of exchange rate fluctuations on the price of logs denominated in foreign currency, representing the market power of log-exporting countries. Pricing-to-Market Ability (PTM*) reflects the effect of exchange rate fluctuations on the price of logs denominated in RMB (φ_j), representing the market power of China in log imports. It is noted that PTM* = ERPT* + 1. Assuming constant costs, when foreign currency appreciates, the research results can be categorized into the following seven scenarios based on the values of ERPT* and PTM* (see

Table 1. The relationship between China’s log import demand elasticity, ERPT, PTM, and market power.

εjt	ERPT*	The Import Price in Foreign Currency	PTM* (φj)	The Import Price in Local Currency	Exchange Rate Pass-Through and Market Power (China’s Imports)
Perfectly inelastic (εjt = 0)	>0	Increase	>1	Increase > exchange rate fluctuation	Reverse pass-through of exchange rates, no market power
Perfectly inelastic (εjt = 0)	=0	Remain unchanged	=1	Increase = exchange rate fluctuation	Complete non-pass-through of exchange rates, no market power
Lack of elasticity (0 < εjt < 1)	(−½, 0)	0 < decrease < ½ of the exchange rate fluctuation	(½, 1)	½ < increase < exchange rate fluctuation	Incomplete pass-through of exchange rates, relatively weak market power
Unit elasticity (εjt = 1)	=−½	Decrease = ½ of the exchange rate fluctuation	=½	Increase = ½ of exchange rate fluctuation	Incomplete pass-through of exchange rates, equal market power
Highly elastic (εjt > 1)	(−1, −½)	½ < decrease < exchange rate fluctuation	(0, ½)	0 < increase < ½ of exchange rate fluctuation	Incomplete pass-through of exchange rates, relatively strong market power
Infinite elasticity (εjt → ∞)	=−1	decrease = exchange rate fluctuation	=0	Remain unchanged	Complete pass-through of exchange rates, strong market power
Infinite elasticity (εjt → ∞)	<−1	decrease > exchange rate fluctuation	<0	Decrease	Excessive pass-through of exchange rates, dominant market power

Note: Based on PTM theory; ε^j_t represents the absolute value of the price elasticity of demand faced by country j in China; “domestic currency” refers to RMB; “foreign currency” refers to the currency of country j; exchange rate fluctuations are exemplified by the appreciation of the foreign currency; analysis of exchange rate pass-through is based on the interpretation of ERPT* values; market power refers to China’s import market power in the source country, and market power is relative—strong import market power for China in the source country indicates weaker market power for the source country’s exports.

for details):

① ERPT* > 0, PTM* > 1: Reverse Pass-Through Effect. In this case, the demand faced by the source country in the Chinese market is perfectly inelastic. The entire cost increase resulting from the appreciation of the foreign currency is borne by China’s imports, allowing the source country to raise the cost mark-up and amplify the effects of exchange rate fluctuations. This indicates that the source country has exceptionally strong market power in the Chinese market, while China’s import market power is negligible.

② ERPT* = 0, PTM* = 1: Complete Non-Pass-Through Effect. Similar to the previous case, the demand from the source country in the Chinese market is perfectly inelastic. When the foreign currency appreciates by 1%, the price in foreign currency remains unchanged, while the price in RMB increases by the same percentage of 1%. This situation indicates that China’s import market power is negligible.

③ −½ < ERPT* < 0, ½ < PTM* < 1: Incomplete Pass-Through Effect. In this case, the demand faced by the source country in the Chinese market is relatively inelastic and not sensitive to price changes. When the foreign currency appreciates, the source country absorbs a small portion of the exchange rate fluctuation cost and slightly reduces its prices, while passing most of the cost increase to China, resulting in a significant rise in the price denominated in RMB. This indicates that the source country has substantial market power in the Chinese market, whereas China’s market power is relatively weak.

④ ERPT* = −½, PTM* = ½: Incomplete Pass-Through Effect. Here, the demand elasticity faced by the source country in the Chinese market is equal to 1, meaning that both the source country and China share the exchange rate fluctuation impact equally. This situation reflects balanced market power between the source country and China, with both parties possessing equivalent market influence.

⑤ −1 < ERPT* < −½, 0 < PTM* < ½: Incomplete Pass-Through Effect. In this case, the demand faced by the source country in the Chinese market is elastic, with an elasticity greater than 1. The source country bears a significant portion of the exchange rate fluctuation cost and can only pass a small fraction of the cost increase to the Chinese market to maintain its market share. This indicates that China has considerable market power in this source country’s market.

⑥ ERPT* = −1, PTM* = 0: Complete Pass-Through Effect. In this case, the entire burden of exchange rate fluctuations is borne by the source country, with no change in the price expressed in RMB. This indicates that China possesses significant market power in the market for this source country.

⑦ ERPT* < −1, PTM* < 0: Excessive Pass-Through Effect. Here, the source country faces infinite demand elasticity in the Chinese market, meaning that even minor price changes are highly sensitive. When the foreign currency appreciates, the price in foreign currency decreases by more than the appreciation rate, and the price in RMB also declines. This reflects that China has exceptionally strong market power in the market for this source country.

2.3.2. AIDS Model

To gain further insights into the market behavior characteristics of source countries, it is imperative to use the AIDS model, a widely recognized framework for analyzing differentiated demand (Thompson, 2013 [35]; Michael, 2020 [36]; Tereza et al., 2022 [37]; Zhu, 2023 [34]). The AIDS model, initially developed by Deaton et al. (1980) [52], is based on the concept of minimizing consumer expenditure, given the prices and levels of utility (Woo et al., 2022) [53]. This model is valued for its flexibility in functional form, facilitating straightforward estimation and allowing for the derivation of critical parameters, such as consumer expenditure elasticity, simple price elasticity, and cross-price elasticity. Furthermore, the AIDS model enables a comprehensive analysis of consumer behavior and market characteristics, thereby offering substantial economic insights. As a result, it is extensively utilized in the field of international trade research (Pascoe et al., 2023; Bronnmann et al., 2024) [54,55]. In this study, the specific formulation of the AIDS model is as follows:

$$w_{it}={\alpha }_i+{\beta }_i\ln \left({\displaystyle \frac{E_t}{P_t^{*}}}\right)+{\displaystyle \sum _{j=1}^n{\gamma }_{ij}\ln p_{jt}+u_{it}}$$

(7)

$$\ln p_t^{*}={\displaystyle \sum _{i=1}^n{w}_{it}\ln (p_{it})}$$

(8)

In Equation (7), w_it represents the proportion of the value of timber imported from country i to China in year t relative to the total value of China’s timber imports in year t; E_t denotes the total value of China’s timber imports in year t; P_t* is the timber import price index, calculated according to Equation (8), and can also be referred to as the overall timber import price level; p is the unit price of timber imported from country i in year t; α_i, β_i, and r_ij, are parameters to be estimated; u_it is the random error term. The model must satisfy three constraint conditions: additivity, homogeneity, and symmetry, as specified in Equations (9)–(11) (Forgenie et al., 2023; Squires et al., 2023) [56,57].

$${\displaystyle \sum _{i=1}^n{\alpha }_i=1}, {\displaystyle \sum _{i=1}^n{\beta }_i=0}$$

(9)

$${\displaystyle \sum _{i=1}^n{r}_{ij}={\displaystyle \sum _{j=1}^n{r}_{ij}=0}}$$

(10)

$$r_{ij}=r_{ji}$$

(11)

Based on the estimated parameters, the expenditure elasticity of China’s timber imports (η_i) can be further calculated as follows:

$${\eta }_i=1+{\displaystyle \frac{{\beta }_i}{{\overline{w}}_i}}$$

(12)

Marshall’s uncompensated price elasticity is given by:

$${\epsilon }_{ij}^m=-{\delta }_{ij}+{\displaystyle \frac{r_{ij}}{{\overline{w}}_i}}-{\beta }_i{\displaystyle \frac{{\overline{w}}_j}{{\overline{w}}_i}}$$

(13)

In Equations (12) and (13), ${\overline{w}}_i$ represents the mean proportion of the value of timber imported from country i to China’s total timber imports; δ_ij is the Kronecker delta, where δ_ij = 1 when i and j are equal, indicating simple price elasticity as shown in Equation (14), and δ_ij = 0 when i and j are not equal, indicating cross-price elasticity as shown in Equation (15) (Hsu et al., 2023) [58].

$${\epsilon }_{ii}=-1+{\displaystyle \frac{r_{ii}}{{\overline{w}}_i}}-{\beta }_i$$

(14)

$${\epsilon }_{ij}={\displaystyle \frac{r_{ij}}{{\overline{w}}_i}}-{\beta }_i{\displaystyle \frac{{\overline{w}}_j}{{\overline{w}}_i}}$$

(15)

3. Results

3.1. Analysis of Market Power Measurement

3.1.1. Diagnostic Tests for the PTM Model

To ensure the scientific validity of the regression results, this study conducted a series of diagnostic tests on the PTM model. These tests include stationarity tests, cointegration tests, Hausman tests, two-way fixed effects tests, varying coefficient model tests, Wald heteroskedasticity tests, Wooldridge autocorrelation tests, and multicollinearity tests. Ultimately, we determined that a varying coefficient PTM model with both individual and time fixed effects should be constructed. The model was estimated using Feasible Generalized Least Squares (FGLS) regression analysis, iterating until convergence to address issues like heteroskedasticity and ensure the most reliable estimation results. The specific diagnostic process is as follows:

(1)

Stationarity Test:

To avoid the “spurious regression” phenomenon, this study used the Fisher-type test and Levin–Lin–Chu (LLC) test to assess the stationarity of the panel data. The Fisher-type test is suitable for heterogeneous roots, while the LLC test is appropriate for homogeneous roots. Only when both tests simultaneously reject the null hypothesis (H0) can it be concluded that the panel data does not contain either heterogeneous or homogeneous unit roots, meaning the panel data is stationary. Let rlp denote the price of China’s timber imports and elp denote the bilateral exchange rate between China and its timber import source countries. The test results, as shown in

Table 2. Stability test of each sequence.

Variables Name	Fisher-Type				Levin–Lin–Chu	Consequence
	p Value	Z Value	L* Value	Pm Value	t Value
lnrlp	494.94 ***	−18.03 ***	−31.37 ***	52.41 ***	−6.60 ***	Stable
lnelp	−78.91 ***	−4.67 ***	−4.55 ***	4.69 ***	−1.65 ***	Stable

Note: Calculated through Stata 15 software; *** represent significance level of 1%.

, indicate that all sequences are stationary and constitute long panel data.

(2)

Cointegration Test

The Pedroni Residual test and Kao Residual test were used to determine the cointegration relationship among the variables. The Pedroni test was used for assessing cointegration in heterogeneous panels, while the Kao test was applied for homogeneous panels. Cointegration is indicated only when both tests simultaneously reject the null hypothesis (H0). The results, as presented in

Table 3. Cointegration test of various variables.

Variables Name	Pedroni Residual			Kao Residual	Consequence
	rho Value	PP Value	ADF Value	t Value
Lnrlp and lnelp	−57.67 ***	−25.97 ***	−13.57 ***	−2.85 ***	There exists a cointegration relationship

Note: Calculated through Stata 15 software; *** represent significance level of 1%.

, reveal a long-term cointegration relationship between lnrlp and lnelp.

(3)

Hausman Test

The Hausman test was used to determine whether to use a fixed effects model or a random effects model. The fixed effects model assumes that the unobserved effects, which do not vary over time, are correlated with the disturbance term, whereas the random effects model assumes no such correlation. The null hypothesis (H0) of the Hausman test is that although estimates from the random effects model and fixed effects model converge, the random effects model is preferred. If H0 is accepted, the random effects model is deemed more appropriate; if H0 is rejected, the fixed effects model is preferred. For the Chinese timber import model, the Hausman test yielded a χ² statistic of 3.55, which significantly rejects the null hypothesis at the 5% level, indicating that the fixed effects model is more suitable.

(4)

Double Fixed Effects Test

After confirming the use of a fixed effects model, further testing was conducted to determine whether both individual and time effects should be fixed. The individual fixed effect assumes no correlation between the disturbance term and the explanatory variables across periods, addressing the issue of omitted variables that vary by individual but not by time. Similarly, the time fixed effect assumes consistent time effects across periods. The F-statistics for individual and time fixed effects were 233.76 and 1.12, respectively, which significantly reject the null hypothesis at the 1% and 10% levels. Therefore, a double fixed effects model with both individual and time effects is appropriate.

(5)

Varying Coefficient Model Test

The varying coefficient model, also known as the “variable parameter model”, involves constructing regression equations for each individual where the slopes differ. Essentially, the slopes are treated as random variables. The null hypothesis (H0) for this model is β₁ = … = β_n. If this hypothesis is rejected, it indicates that a varying coefficient model is more appropriate, as it better captures economic behavior in practice. The test results show a χ² statistic of 3.3 × 10⁵, which significantly rejects the null hypothesis at the 1% level, thus supporting the use of a varying coefficient model.

(6)

Wald Heteroscedasticity Test

The Wald test was used to examine heteroscedasticity in the panel model. The null hypothesis (H0) is that the variances of the disturbance terms across different individuals are equal. The calculated χ² statistic is 7435.87, which significantly rejects the null hypothesis at the 1% level, indicating the presence of heteroscedasticity.

(7)

Wooldridge Autocorrelation Test

The Wooldridge test was used to assess autocorrelation in the panel model. The computed F-statistic is 17.32, which significantly rejects the null hypothesis at the 1% level, indicating the presence of autocorrelation.

(8)

Multicollinearity Test

Multicollinearity issues were rarely encountered in practical data. To assess multicollinearity, the variance inflation factor (VIF) was calculated. The VIF value obtained is 1, indicating that multicollinearity is not a concern.

Based on tests 1–8, in PTM model estimation, Feasible Generalized Least Squares (FGLS) should be used to ensure the reliability of the estimation results.

3.1.2. Analysis of PTM Model Empirical Results

Using the nominal exchange rate estimates as the primary basis for analysis, with actual exchange rate estimates used for robustness checks, the results are shown in

Table 4. Estimation results of market power for imported logs.

Import Source	Nominal Exchange Rate			Real Exchange Rate
	ERPT*	PTM* (φj)	Market Power	ERPT*	PTM* (φj)	Market Power
Indonesia	−2.94	−1.94 ***	Superlative	−2.51	−1.51 ***	Superlative
Malaysia	−2.40	−1.40 ***	Superlative	−2.41	−1.41 ***	Superlative
Myanmar	−1.20	−0.20 ***	Superlative	−1.21	−0.21 ***	Superlative
Russia	−0.95	0.05 **	Strong	−0.94	0.06 ***	Strong
The DRC	−0.88	0.12 ***	Strong	−0.88	0.12 ***	Strong
Mozambique	−0.57	0.43 ***	Strong	−0.51	0.49 ***	Strong
Papua New Guinea	−0.43	0.57 ***	Weak	−0.34	0.66 ***	Weak
Equatorial Guinea	−0.19	0.81 ***	Weak	−0.09	0.91 ***	Weak
France	−0.18	0.82 ***	Weak	−0.30	0.70 ***	Weak
Germany	−0.08	0.92 ***	Weak	−0.09	0.91 ***	Weak
Australia	−0.07	0.93 ***	Weak	−0.14	0.86 ***	Weak
New Zealand	−0.01	0.99 ***	Weak	−0.21	0.79 ***	Weak
Japan	0.19	1.19 ***	No	0.14	1.14 ***	No
Cameroon	0.47	1.47 ***	No	0.38	1.38 ***	No
The United States	3.14	4.14 ***	No	2.46	3.46 ***	No
Laos	−2.19	−1.19	Not significant	−2.87	−1.87	Not significant
The Solomon Islands	−1.36	−0.36	Not significant	−1.25	−0.25	Not significant
Gabon	−1.21	−0.21	Not significant	−1.47	−0.47	Not significant
Canada	−1.05	−0.05	Not significant	−1.23	−0.23	Not significant
R2 = 0.79				R2 = 0.78
Log Likelihood = −788.09				Log Likelihood = −914.77
F-statistic = 57.54				F-statistic = 53.47
Prob (F-statistic) = 0.00				Prob (F-statistic) = 0.00

Note: Calculated through Stata 15 software; ***, ** represent significance levels of 1%, and 5%, respectively.

. Although the specific values of φ_j vary, the final conclusions regarding market power are largely consistent. This indicates that inflation has a minimal impact on market power in China’s timber import trade and does not bring about substantial changes, reflecting the robustness of the constructed PTM model. Table 4 reveals significant differences in market power among various timber source countries. Based on the magnitude of market power, China’s timber import sources can be classified into the following four categories:

(1)

Category 1: China Holds Superlative Market Power

This category includes Indonesia, Malaysia, and Myanmar. When the currency of these three countries appreciates by 1%, the timber prices in their currencies decline by 2.94%, 2.40%, and 1.20%, respectively, which is greater than the currency appreciation. Timber prices in RMB decrease by 1.94%, 1.40%, and 0.20%, respectively, indicating that China has superlative market power in importing timber from these markets, with the cost of currency appreciation entirely borne by the source countries.

These countries are Southeast Asian lower-middle- and upper-middle-income nations with abundant forest resources. China primarily imports hardwoods such as Merbau, Belian, and Teak from these countries for use in constructing docks, bridges, and flooring. Their lack of market power in China’s timber import trade is due to several reasons. First, poor forest management and lack of long-term planning have led to degraded self-renewal and growth capacity of forest resources, making sustained high-intensity export logging challenging. To protect their resources, these countries have implemented timber export bans, providing opportunities for countries like the DRC, Germany, New Zealand, Canada, and Uruguay to increase their market share in China’s timber import trade. Second, as per the latest data from the World Development Indicators database, by 2021, Indonesia, Malaysia, and Myanmar had forest areas of 915,300, 190,600, and 282,500 square kilometers, respectively. Southeast Asian timber is generally coarser, lighter in oil content, and of lower quality compared to South American timber. High-quality timbers such as Uruguayan Rosewood and South American Padauk are rarer and more resistant to decay and insects. Additionally, Southeast Asian timber prices are significantly higher than those from African countries, and the diversity of tree species is less. Consequently, the Southeast Asian timber export market is increasingly impacted, with Indonesia, Malaysia, and Myanmar’s share of China’s total timber imports declining from 10.09%, 9.01%, and 3.42% in 2001 to 0.01%, 0.09%, and 0.01% in 2023. Third, the proximity of these countries to China and ineffective implementation of timber export bans contribute to the persistent issue of illegal logging and smuggling, reducing their international competitiveness (Obidzinski et al., 2007) [59].

(2)

Category 2: China Holds Strong Market Power

This category includes Russia, the DRC, and Mozambique. When the currency of these countries appreciates by 1%, timber prices in their currencies decrease by 0.95%, 0.88%, and 0.57%, respectively, which is greater than half of the currency appreciation. Conversely, timber prices in RMB increase by 0.05%, 0.12%, and 0.43%, respectively, which is less than half of the currency appreciation. This indicates that China holds significant market power in importing timber from these markets, with most of the cost of currency appreciation borne by the source countries and only a small portion passed on to China.

Among these countries, Russia is classified as a high-income country, while the DRC and Mozambique are categorized as lower-middle-income and low-income countries, respectively. These countries are rich in forest resources. China imports coniferous woods such as Siberian Larch and Scots Pine from Russia, as well as hardwoods like Ash, for use in manufacturing shipbuilding materials, packaging crates, flooring, roofing, and furniture, capitalizing on the strength and decay resistance of these woods. From the DRC and Mozambique, China imports hardwoods like Sapele, Wenge, and African Blackwood for use in high-quality flooring, carved crafts, interior decoration, and premium furniture, leveraging their gloss, oil content, and bending properties. As of 2021, Russia, the DRC, and Mozambique had forest areas of 8,153,100, 219,300, and 365,000 square kilometers, respectively. Despite these large areas of high-quality timber being in demand by China, the market power of these countries in China’s timber imports is relatively weak. This is primarily due to the following reasons: First, China is the largest importer of timber from these countries, with China accounting for 69.18% and 58.24% of Russia and the DRC’s timber exports in 2021, and 98.40% of Mozambique’s timber exports in 2019. However, Russia, the DRC, and Mozambique only account for 7.13%, 0.85%, and 1.57% of China’s timber imports, respectively, indicating that the Chinese market is more important to these countries than vice versa. Second, these countries’ forest industries are underdeveloped, with inadequate funding leading to challenges in modernizing forestry machinery. This results in a lack of technological and equipment support for timber extraction and processing, with exports primarily consisting of raw timber and minimally processed lumber, resulting in low added value (Simeone, 2013) [60]. Third, the rapidly changing timber harvesting policies in Russia and the poor socio-economic conditions, including security and disease issues in the DRC and Mozambique, contribute to a challenging investment environment and higher risks. Additionally, inadequate forest infrastructure, undeveloped logging road networks, and difficult extraction conditions further deter foreign investment (Li et al., 2019) [61]. Fourth, forest management in Russia, the DRC, and Mozambique is relatively chaotic, with a lack of accurate and detailed data on forest production and management. There are significant volumes of mature and over-mature timber that need to be harvested promptly to avoid pest infestations and decay. Moreover, illegal logging and timber smuggling are prevalent, with precious timber often being exported at lower prices by misclassifying timber grades. Fifth, global warming and local practices such as slash-and-burn agriculture increase the likelihood of forest fires, severely impacting the regeneration and growth of both natural and artificial forests. In Russia, warming temperatures reduce the number of cold-adapted conifer species and shorten the logging period, making timber more susceptible to pests and diseases. These factors collectively diminish the international competitiveness of timber from these countries.

(3)

Category 3: China Holds Weak Market Power

This category includes Papua New Guinea, Equatorial Guinea, France, Germany, Australia, and New Zealand. When the currencies of these countries appreciate by 1%, the log prices in their currencies decrease by 0.43%, 0.19%, 0.18%, 0.08%, 0.07%, and 0.01%, respectively, all less than half of the currency appreciation rate. In RMB terms, the log prices increase by 0.57%, 0.81%, 0.82%, 0.92%, 0.93%, and 0.99%, respectively, all more than half of the appreciation rate. This indicates that China has weak market power in these markets, with the majority of the cost increase due to currency appreciation being passed on to China, while only a small portion is absorbed by the source countries.

These countries are a mix of middle-income (Papua New Guinea), high-income (Australia and New Zealand), and upper-middle-income (Equatorial Guinea) nations in Oceania, Africa, and Europe. They all have abundant forest resources. As of 2021, the forest areas are as follows: Papua New Guinea (358,200 square kilometers), Equatorial Guinea (24,400 square kilometers), France (173,400 square kilometers), Germany (114,200 square kilometers), Australia (1,340,100 square kilometers), and New Zealand (99,100 square kilometers). The strong market position of these countries in China’s log import trade is due to several factors. Firstly, for Papua New Guinea, which is similar to Malaysia in its timber offerings, China imports timber species such as satinwood, ironwood, and Merbau for products including veneer, plywood, and high-end furniture. In 2023, Papua New Guinea accounted for 7.56% of China’s total log imports, making it the fourth-largest source. The region’s forest resources are predominantly owned by clans and tribes (95%), with only 5% government-owned. The timber industry is largely monopolized by foreign investors from Australia, Japan, and Malaysia. Before timber export, a series of procedures including logging permit approval, species pricing approval, and export quota approval must be completed, which enhances the bargaining power. Secondly, Australia and New Zealand primarily supply China with radiata pine and other softwoods used in manufacturing engineered wood products, furniture, paper products, and railway sleepers. In 2023, New Zealand was the largest source of logs for China, contributing 34.61% of the total imports, while Australia was the fifth-largest source in 2019, but its share has gradually decreased due to restrictive policies. Both Australia and New Zealand are advanced in forestry, emphasizing sustainable forest management and conservation. They implement robust policies to combat forest fires and pest infestations, invest in seedling propagation, forest management, and wood processing technologies, and actively establish forestry certification systems while encouraging foreign investment. Consequently, the market for legal timber sources has diminished due to increased efforts against illegal logging, leading to a reduced supply of legal timber. Given China’s high demand for radiata pine, its broad applications, and the difficulty in finding substitutes, its bargaining power remains strong. Thirdly, with regard to Equatorial Guinea, China primarily imports high-value woods such as okoumé, padouk, okan, and ebony for producing high-end furniture, fine woodworking, moldings, musical instruments, and collectibles. In 2023, Equatorial Guinea accounted for 0.50% of China’s total timber imports, making it the 26th largest source of timber for China. The country’s strong bargaining power in China’s timber imports is largely due to the high quality and economic value of its timber resources, compounded by the gradual restrictions on timber exports, which have led to increased scarcity of timber. Lastly, France and Germany export hardwoods such as oak, beech, walnut, and ash, and softwoods including fir, spruce, and pine to China for premium furniture, interior woodwork, plywood, and packaging. In 2023, France and Germany accounted for 3.20% and 9.03% of China’s total timber imports, making them the 6th and 3rd largest sources of timber for China, respectively. Both countries are leaders in woodworking and forestry machinery, possessing advanced technology in timber cultivation and processing. Their forest management practices adhere strictly to Forest Stewardship Council (FSC) certification standards, with detailed, standardized management systems for afforestation, thinning, harvesting, and road construction in forest areas. They also focus on biodiversity conservation and maintaining forest ecological balance, while promoting the development of professional forestry management skills and continually improving infrastructure such as main roads, secondary roads, and access routes for machinery and personnel. These practices contribute to a steady increase in timber yield and enhance their international competitiveness.

(4)

Fourth Category: China Holds No Market Power

This category includes Japan, Cameroon, and the United States. When the currencies of these countries appreciate by 1%, the price of logs expressed in their currencies increases by 0.19%, 0.47%, and 3.14%, respectively, and the price in RMB increases by 1.19%, 1.47%, and 4.14%, respectively. This indicates that China lacks market power in these markets, with the entire cost increase due to currency appreciation being absorbed by China. Additionally, the estimated coefficients for Laos, the Solomon Islands, Gabon, and Canada are not significant and are therefore not analyzed separately, indicating no market power.

In this category of countries, Japan is an Asian high-income nation, Cameroon is an African lower-middle-income nation, and the United States is a North American high-income nation. As of 2021, Japan, Cameroon, and the United States have forest areas of 249,400, 202,800, and 3,098,000 square kilometers, respectively. Their bargaining power in China’s log import trade is notably strong, primarily due to the following reasons: China primarily imports precious timber species from Cameroon, such as rosewood, ebony, and ironwood, which are used in the production of high-end furniture, fine woodworking, musical instruments, and crafts. In 2023, Cameroon accounted for 1.94% of China’s log imports, making it the 12th largest source of imports. Similar to Equatorial Guinea, the valuable and irreplaceable timber varieties from Cameroon, combined with the government’s efforts to protect local forest resources and boost domestic employment by raising log export tariffs or even banning raw log exports in favor of processed products, have significantly reduced the supply of logs. Secondly, in the case of Japan and the United States, China primarily imports artificial forest wood such as Japanese cedar, Hinoki cypress, and larch, as well as broadleaf timber including oak, hickory, and beech from these countries. These materials are used for manufacturing bridges, furniture, paper products, steamers, and plywood. In 2023, Japan and the United States accounted for 2.64% and 10.92% of China’s total wood imports, respectively, ranking as the 9th and 2nd largest sources of imports. Similar to France and Germany, both countries follow an intensive, large-scale, and efficient development model in the wood processing industry. They continuously leverage forestry technology in forest management, promote integrated logging and reforestation, advance rapid-growing tree species improvement and pest control technologies, develop intelligent and high-efficiency forestry machinery, and enhance information and communication technology infrastructure in the supply chain. This facilitates real-time transmission of information related to forest production, management, harvesting, and distribution, while providing financial support during negotiations with foreign enterprises. Particularly in the United States, forest resource ownership is clearly delineated, with the principle of “who owns, who manages, who profits” being applied, minimizing property disputes and fully mobilizing the initiative of private landowners. This clarity allows for long-term planning and substantial investment in infrastructure. Moreover, the distinct roles of public and private forests, where one focuses on social and ecological benefits and the other on economic benefits, have significantly enhanced the international competitiveness of the United States timber in China’s import trade.

3.2. Analysis of Market Behavior Characteristics of Source Countries

In this section, we used Seemingly Unrelated Regression (SUR) to estimate the AIDS model, aiming to elucidate the market behavior characteristics of source countries. The SUR method is widely used in the estimation of multiple equation systems, as it effectively addresses estimation biases in simultaneous equations, significantly improves estimation efficiency, and enhances the robustness of the estimation results. Based on Equations (12) and (13), we calculated the expenditure elasticity, simple price elasticity, and cross-price elasticity of China’s log imports, as shown in

Table 5. Estimation results of expenditure elasticity and price elasticity.

	Expenditure Elasticity	①	②	③	④	⑤	⑥	⑦	⑧	⑨	⑩	⑪	⑫	⑬	⑭	⑮	⑯	⑰	⑱	⑲
①	1.47	−0.73	−0.03	0.18	−0.37	0.06	0.04	−0.09	−0.09	−0.19	−0.14	0.16	0.01	−0.06	0.29	−0.30	−0.16	−0.24	0.21	−0.08
②	1.09	−0.01	−0.26	−0.01	−0.20	−0.58	0.05	−0.05	0.21	0.05	−0.78	0.11	−0.07	−0.01	−0.17	0.25	−0.45	−0.17	0.26	−0.30
③	0.04	0.19	−0.01	−1.25	−0.02	−0.12	0.47	0.03	−0.00	0.37	0.26	−0.05	0.04	−0.04	−0.57	−0.05	−0.37	0.06	0.06	−0.02
④	1.21	−0.55	−0.42	−0.04	−0.67	−0.14	0.60	−0.30	−0.16	−0.16	−0.06	0.04	0.00	0.10	0.04	0.24	−1.84	0.32	1.63	−0.03
⑤	1.46	0.08	−0.22	−0.12	−0.09	−0.29	−0.21	−0.03	0.04	0.17	0.39	0.44	−0.02	−0.11	0.07	0.13	−0.16	−0.18	−0.39	−0.54
⑥	0.99	0.05	0.07	0.07	0.40	−0.21	−1.37	0.30	0.19	−0.03	−0.12	−0.14	−0.23	0.26	0.09	0.54	−1.01	0.51	−0.62	−0.01
⑦	1.19	−0.09	−0.08	0.03	−0.22	−0.03	0.33	−0.79	0.09	0.09	0.87	−0.10	0.02	−0.07	0.12	−0.08	−0.59	−0.17	−0.30	−0.48
⑧	1.85	−0.20	0.30	−0.02	−0.25	0.01	0.46	0.18	−0.38	−0.12	0.31	−0.07	0.08	−0.32	0.14	−0.54	−2.71	−0.29	1.05	−0.26
⑨	1.03	−1.01	0.08	0.57	−0.01	0.16	−0.00	0.08	−0.31	−0.19	0.16	0.05	0.04	−0.00	−0.16	−0.00	−0.04	−0.13	−0.03	0.45
⑩	1.58	−0.14	−1.12	0.15	−0.04	0.03	−0.14	0.66	0.13	0.18	−0.67	0.13	0.08	−0.29	0.66	−0.31	−1.31	−1.41	−0.10	0.21
⑪	−0.92	0.61	0.63	−0.12	−0.06	1.01	0.52	−0.25	−0.07	0.22	0.51	−1.24	0.43	0.71	−0.37	0.13	0.31	−0.36	−0.81	−0.27
⑫	1.85	0.03	−0.73	0.27	−0.23	0.03	−1.57	0.09	0.23	0.11	0.51	0.83	−0.87	−0.89	0.16	−0.17	2.62	−0.62	−0.25	−0.65
⑬	1.65	−0.45	−0.18	−0.35	0.42	−0.26	1.69	−0.44	−0.86	−0.23	−1.96	1.36	−0.49	−0.42	−0.23	0.36	−0.78	0.65	−0.96	−0.50
⑭	0.68	0.31	−0.13	−0.32	0.04	0.08	0.06	0.13	0.16	−0.26	0.74	−0.09	0.21	−0.10	−1.67	0.07	0.72	0.04	−0.84	0.14
⑮	0.98	−0.41	0.29	−0.07	0.22	0.21	0.76	−0.10	−0.29	−0.05	−0.42	0.04	−0.03	0.08	0.14	−2.07	0.10	0.23	0.38	−0.13
⑯	1.46	−0.41	−0.29	−0.07	−0.22	−0.21	−0.76	−0.10	−0.29	−0.15	−0.42	0.04	−0.03	0.08	0.14	−0.07	−0.44	0.23	0.38	0.09
⑰	1.14	−0.07	−0.09	0.02	0.08	−0.06	0.19	−0.05	−0.03	−0.05	−0.50	−0.05	−0.03	0.05	0.01	0.06	−0.26	−0.56	−0.23	0.11
⑱	0.64	0.22	0.36	0.01	0.10	−0.01	−0.05	−0.09	0.05	−0.23	−0.33	−0.11	−0.05	−0.01	−0.27	0.03	0.13	−0.03	−1.47	0.29
⑲	1.55	−0.03	−0.17	−0.02	−0.01	−0.22	−0.02	−0.16	−0.03	0.19	−0.08	−0.05	−0.02	−0.02	0.05	−0.04	−0.69	0.09	0.92	−0.23
Breusch−Pagan test of independence: χ2(171) = 895.36, p = 0.00

Note: ①–⑲ respectively represent Australia, Solomon Islands, Myanmar, Cameroon, Canada, the DRC, Equatorial Guinea, France, Gabon, Germany, Indonesia, Japan, Laos, Malaysia, Mozambique, New Zealand, Papua New Guinea, Russia, and the United States.

. The Breusch-Pagan test’s χ² statistic indicates that the null hypothesis of no contemporaneous correlation among the disturbance terms across equations is significantly rejected at the 1% level, suggesting that the regression results of the model are reliable. The diagonals refer to simple price elasticity, while the rest (except the first column) to cross-price elasticities and the direction of influence for the cross-price elasticity.

3.2.1. Analysis of Expenditure Elasticity

Expenditure elasticity reflects the extent to which changes in China’s spending on log imports affect the import volumes from various source countries, under the condition that other factors remain constant. It also indicates the order of benefit for China’s log import source countries and the changes in the structure of China’s log import market. As shown in Table 5, except for Indonesia (−0.92), the expenditure elasticities for all source countries are positive, indicating a positive correlation between import volumes from source countries and China’s expenditure on log imports. This suggests that China’s substantial demand for logs is challenging to satisfy with the supply from any single country.

Specifically, China’s expenditure elasticity is greater than 1 for logs imported from France (1.85), Japan (1.85), Laos (1.65), Germany (1.58), the United States (1.55), Australia (1.47), New Zealand (1.46), Canada (1.46), Cameroon (1.21), Equatorial Guinea (1.19), Papua New Guinea (1.14), the Solomon Islands (1.09), and Gabon (1.03). This indicates that when China’s expenditure on log imports increases by 1%, the import volume from these countries increases by more than 1%, reflecting elasticity. Combined with Table 4, China holds weaker or non-existent market power in log trade with these 13 countries. However, China continues to increase its log imports from these countries, indicating that Chinese log importers have a higher overall evaluation of the log products from these countries. It also suggests that as China’s log expenditure increases, consumers tend to prefer purchasing higher-quality precious wood and place greater importance on the standardization and legality of market transactions. Conversely, China’s expenditure elasticity is less than 1 for logs imported from the DRC (0.99), Mozambique (0.98), Malaysia (0.68), Russia (0.64), and Myanmar (0.04). This means that when China’s expenditure on log imports increases by 1%, the import volume from these countries increases by less than 1%, indicating inelasticity. For instance, Russia has long been one of China’s primary sources of log imports. The stable trade relationship, geographical proximity, strong demand for logs in China, and Russia’s vast timber production have all reduced China’s sensitivity to the price of imported Russian logs. A notable exception is Indonesia, where China’s expenditure elasticity for log imports is less than 1. This reflects two factors: First, Indonesia’s severe forest degradation has led to a short-term shortage of available timber; second, Indonesian logs are considered lower-grade products for China. This reaffirms that as Chinese expenditure increases, consumers are more inclined to seek out higher-quality wood markets, consistent with economic principles.

Additionally, as China’s expenditure on log imports increases, the country that benefits the most is France, while the one that benefits the least is Indonesia, which has even lost part of its share in China’s log import market. Notably, Indonesia is also the country where China holds the weakest market power in log import trade. This further illustrates that when China’s log import expenditure increases, price is not the only factor considered by importers.

3.2.2. Simple Price Elasticity

Simple price elasticity measures the sensitivity of the quantity of logs imported by China from each source country in response to changes in the import price. It also helps determine how changes in import prices (whether increases or decreases) affect total revenue. As shown in Table 5, the simple price elasticity for all import source countries is negative, indicating that the demand for logs imported by China from these countries is for ordinary goods—meaning that the demand for log imports moves inversely with changes in log prices.

The simple price elasticities for logs imported by China from Japan (−0.87), Equatorial Guinea (−0.79), Australia (−0.73), Germany (−0.67), Cameroon (−0.67), Papua New Guinea (−0.56), New Zealand (−0.44), Laos (−0.42), France (−0.38), Canada (−0.29), the Solomon Islands (−0.26), the United States (−0.23), and Gabon (−0.19) all have absolute values less than 1. This indicates that the demand for logs from these countries in the Chinese market is inelastic, meaning these countries possess strong market power in China’s log import trade and can increase their total revenue by raising prices. In contrast, the simple price elasticities for logs imported by China from Mozambique (−2.07), Malaysia (−1.67), Russia (−1.47), the DRC (−1.37), Myanmar (−1.25), and Indonesia (−1.24) all have absolute values greater than 1, indicating that the demand for logs from these countries in the Chinese market is elastic. These countries thus have weaker market power in China’s log import trade and are more likely to increase total revenue by lowering prices. This conclusion is consistent with the findings in Table 4, verifying the accuracy of the PTM conclusions.

3.2.3. Cross-Price Elasticity

Cross-price elasticity can be used to measure the relationship between the logs imported from different source countries in the Chinese market. If the cross-price elasticity is greater than 0, it indicates a substitution relationship between the logs from the two countries; if it is less than 0, it indicates a complementary relationship. As shown in Table 5, the cross-price elasticities for China’s log imports exhibit both positive and negative values, with negative values being more common. This indicates that in the Chinese log import market, there are both competitive and complementary relationships between the major source countries, though the predominant relationship is complementary. This discovery confirms the estimation of expenditure elasticity, indicating that as China’s demand for logs steadily increases, the market relies on a comprehensive supply from different regions and tree species to meet this demand. Therefore, cross-price elasticity may be negative.

Specifically, the overall cross-price elasticities for New Zealand (−6.49), the United States (−1.98), Australia (−1.88), the Solomon Islands (−1.74), Papua New Guinea (−1.47), Germany (−1.06), France (−1.03), Laos (−0.64), Cameroon (−0.46), Canada (−0.25), Equatorial Guinea (−0.19), Gabon (−0.09), and Japan (−0.06) in the Chinese market are less than 0, indicating that the wood from these countries primarily exhibits a complementary relationship with that from other source countries, facing relatively less market competition. In contrast, Malaysia (0.14), Myanmar (0.16), Mozambique (0.25), Russia (0.36), the DRC (2.42), and Indonesia (2.54) have positive overall cross-price elasticities, indicating that their wood mainly exhibits a competitive relationship with that from other source countries, making them more easily substitutable and facing greater market competition. For example, Indonesian wood can be easily replaced by wood from Canada (1.01), Laos (0.71), the Solomon Islands (0.63), Australia (0.61), the DRC (0.52), Germany (0.51), Japan (0.43), New Zealand (0.31), Gabon (0.22), and Mozambique (0.13). Moreover, the impact of price changes in Australian wood on the volume of Chinese imports of Indonesian wood (0.61) is greater than the impact of price changes in Indonesian wood on the volume of Chinese imports of Australian wood (0.16), indicating that Australian wood has greater market power in China compared to Indonesian wood, consistent with the market power classification in Table 4.

4. Discussion

4.1. Enhancing Import Market Power

Research indicates that China has strong market power with suppliers from Indonesia, Malaysia, and Myanmar; moderate market power with Russia, the DRC, and Mozambique; weak market power with Papua New Guinea, Equatorial Guinea, France, Germany, Australia, and New Zealand; and no market power with Japan, Cameroon, and the United States.

Analysis reveals that the probability of having significant market power increases if the following conditions are met: higher local income levels; better quality of forest management; valuable and irreplaceable timber species; lower timber prices for the same quality; well-regulated markets with complete export approval procedures; substantial market share (Li et al., 2008; Karthikeyan et al., 2013) [62,63]; high levels of forestry mechanization and modernization (Silva et al., 2019) [64]; well-developed local infrastructure and smooth information exchange platforms; stable foreign investment policies; emphasis on forest protection and commitment to legal timber trade (Sheng et al., 2019) [65]; clear forest resource ownership without property disputes; ample professional and skilled forestry management personnel; and high levels of pricing coordination (Cheng et al., 2023 [29]; Zou et al., 2024 [66]).

Thus, while a large market share is one condition for achieving market power, it does not necessarily guarantee it. In other words, “market power” is not equivalent to “monopoly power”, confirming the view that “using market concentration to measure market power lacks theoretical foundation”. This provides a deeper and richer understanding of “market power”. Therefore, to enhance China’s import market power, efforts should address the factors mentioned above, and adjustments to the import market structure can be made to increase market share in countries with significant market power.

This is pivotal for fostering the stable growth of China’s log import trade and safeguarding the security of its domestic wood resource supply. Given China’s heavy reliance on log imports, acquiring market power commensurate with its trade volume would effectively mitigate the adverse effects of the large country effect, bolster its capacity to navigate international market risks, and consequently minimize trade losses. Concurrently, augmenting market power serves to stabilize domestic market prices, ensure the security of domestic wood resources, and further propel the sustainable development of the domestic forestry industry.

China’s growing log imports have received extensive attention around the world, with many observers concerned that the country’s increasing demands could have a huge impact on global timber markets and forest resource conservation and is criticized as a menace to the world’s forest sustainability by the international community (Zhang et al., 2017; Barbu et al., 2021) [67,68]. Yet, most of these studies are descriptive without in-depth quantitative analysis or consideration of the complex interrelationships among different countries and economic sectors. Hence, there is a lack of robust evidence to support the idea that China’s growing log imports have a substantial adverse effect on forest conservation at a global level (Zhang et al., 2007) [69]. In fact, a significant portion of the logs imported by China serves as intermediate products for the production of final goods intended for re-export. As such, China’s imports displace wood processing production in other countries. Because of the displacement effect, the actual impact of China’s imports on world forest production and the strain on world forest resources is less pronounced than the import figures might indicate (Zhang et al., 2007) [69]. Of course, there’s no denying that China’s increased imports could exert additional pressure on the already vulnerable forest resources in some countries, highlighting the importance of Chinese companies enhancing their social responsibility in overseas investments and the Chinese government formulating relevant policies (Dong et al., 2018) [70]. Furthermore, China should actively utilize its timber import control measures to contribute to international efforts in combating illegal logging and deforestation (Wang et al., 2023) [71].

In addition, the ascending market power of China poses certain risks to exporting countries, such as the inability of previously dominant market players to further elevate prices. On the one hand, this could lead to reduced profitability for original exporting nations within the Chinese market, compelling them to seek new export destinations, consequently diminishing their supply of raw wood to China and jeopardizing the security of China’s raw wood supply chain. On the other hand, a balanced market dynamic, where buyers and sellers are relatively equal in strength, is more conducive to maintaining market stability, influenced as it is by both supply and demand forces. Looking ahead, China’s wood industry must remain poised to adapt to market shifts, including monitoring trends related to deforestation-free initiatives and global and regional economic conditions. This necessitates China to emphasize enhancing domestic wood supplies to decrease reliance on imported wood (Zhang et al., 2023) [72] and improve the efficiency of forest management, which has been demonstrated as another crucial aspect in mitigating dependency on foreign forest resources (Zhang et al., 2021) [73].

4.2. Diversified Import Strategy

From the perspective of elasticity analysis, the results indicate that the wood products from various source countries in China’s timber import trade are ordinary goods, with no extreme cases of infinite or zero elasticity. This suggests that the Chinese timber import market is relatively rational and regulated, which is related to the immense demand for timber in China, and multiple countries with various qualities of wood are needed to meet this demand. Therefore, it is essential to continue implementing a diversified import strategy (Niquidet et al., 2013; Cheng et al., 2015) [74,75]. In fact, the issue of high concentration in China’s timber import trade market has improved and is moving towards diversification.

If the market power of the importing source country is strong, its expenditure elasticity in the Chinese market is greater than 1, and its simple price elasticity is less than 1. This means that when China increases its expenditure on timber, consumers are more inclined to purchase timber from such markets, indicating a preference for higher-quality, more regulated, and more legitimate timber markets, where price is not the only consideration. At the same time, the source country can achieve higher total revenue by increasing prices, which is consistent with economic theory (Sun et al., 2017) [76]. Conversely, if the market power of the importing source country is weak, its expenditure elasticity in the Chinese market is less than 1, and its simple price elasticity is greater than 1. Such countries can only achieve higher total revenue by lowering prices. For instance, China’s plywood export trade has gradually lost its low-price advantage, and further price reductions could lead to greater trade losses. Therefore, improving international market power for timber imports or plywood exports is crucial to avoid trade losses and even achieve trade gains.

Additionally, within the domestic market, China should enhance the construction of state reserve forest bases, increase domestic timber supply capacity, and actively explore new alternative materials to address the issue of forest resource scarcity (Guan et al., 2023) [77].

5. Conclusions

This study employs panel data from January 2001 to December 2023 and constructs a fixed-effects varying coefficient PTM panel model to measure market power in China’s log import trade. Using the AIDS model from an elasticity perspective, it explores the market behavior characteristics of different log-supplying countries in China’s log import market and validates the mechanisms behind market power. The results reveal:

(1)

China’s main trading partners can be categorized into four groups according to their market power in the log import trade. First, China holds superlative market power in log imports from Indonesia, Malaysia, and Myanmar. When the currency of these three countries appreciates by 1%, the timber prices in their currencies decline by 2.94%, 2.40%, and 1.20%, respectively, which is greater than the currency appreciation. Timber prices in RMB decrease by 1.94%, 1.40%, and 0.20%, respectively. Second, China holds strong market power in log imports from Russia, the DRC, and Mozambique. When the currency of these countries appreciates by 1%, timber prices in their currencies decrease by 0.95%, 0.88%, and 0.57%, respectively, which is greater than half of the currency appreciation. Conversely, timber prices in RMB increase by 0.05%, 0.12%, and 0.43%, respectively. Third, China holds weak market power in log imports from Papua New Guinea, Equatorial Guinea, France, Germany, Australia, and New Zealand. When the currencies of these countries appreciate by 1%, the log prices in their currencies decrease by 0.43%, 0.19%, 0.18%, 0.08%, 0.07%, and 0.01%, respectively, all less than half of the currency appreciation rate. In RMB terms, the log prices increase by 0.57%, 0.81%, 0.82%, 0.92%, 0.93%, and 0.99%, respectively. Finally, China holds no market power in Japan, Cameroon, and the United States. When the currencies of these countries appreciate by 1%, the price of logs expressed in their currencies increases by 0.19%, 0.47%, and 3.14% respectively, and the price in RMB increases by 1.19%, 1.47%, and 4.14% respectively.

(2)

As China increases its log import expenditure, it tends to purchase higher-quality and more valuable timber, while also placing greater emphasis on market transaction legality. Consequently, China is anticipated to augment its imports from source countries with no or weak market power. As illustrated in Table 5, with the exception of Indonesia, which has an expenditure elasticity of −0.92, all import source countries exhibit positive expenditure elasticities. Furthermore, China’s expenditure elasticities surpass 1 in several countries, including France (1.85), Japan (1.85), Laos (1.65), Germany (1.58), the United States (1.55), Australia (1.47), New Zealand (1.46), Canada (1.46), Cameroon (1.21), Equatorial Guinea (1.19), Papua New Guinea (1.14), the Solomon Islands (1.09), and Gabon (1.03), indicating a high degree of elasticity.

(3)

The simple price elasticity of logs from all source countries is negative. Specifically, the simple price elasticities for logs imported by China from Japan (−0.87), Equatorial Guinea (−0.79), Australia (−0.73), Germany (−0.67), Cameroon (−0.67), Papua New Guinea (−0.56), New Zealand (−0.44), Laos (−0.42), France (−0.38), Canada (−0.29), the Solomon Islands (−0.26), the United States (−0.23), and Gabon (−0.19) all have absolute values less than 1. In contrast, the simple price elasticities for logs imported by China from Mozambique (−2.07), Malaysia (−1.67), Russia (−1.47), the DRC (−1.37), Myanmar (−1.25), and Indonesia (−1.24) all have absolute values greater than 1. Countries with stronger market power can achieve higher total revenues by raising prices, whereas countries with weaker market power are more inclined to achieve higher total revenues by lowering prices.

(4)

Logs from different source countries are complementary in the Chinese market, indicating that China’s substantial log demand relies on simultaneous supplies from multiple countries and various types of timber. Specifically, the overall cross-price elasticities for New Zealand (−6.49), the United States (−1.98), Australia (−1.88), the Solomon Islands (−1.74), Papua New Guinea (−1.47), Germany (−1.06), France (−1.03), Laos (−0.64), Cameroon (−0.46), Canada (−0.25), Equatorial Guinea (−0.19), Gabon (−0.09), and Japan (−0.06) in the Chinese market are less than 0. Based on these findings, this study provides targeted recommendations for enhancing international market power and reducing trade losses.

Due to space limitations, this paper presents only the analysis results of China’s log import market power and does not cover market power analyses of chemical wood pulp imports or plywood exports. A comparative analysis of different product trade market characteristics will be presented in subsequent research and publications. As mentioned in the introduction, while the PTM model effectively analyzes China’s market power across various import source countries, it falls short in comparison to the SMR model, which has the capability to assess the situation from both buyer and seller perspectives concurrently. Regarding sample selection, although this study has focused on major source countries for China’s log imports, ensuring high representativeness, it does not encompass all import sources, thereby imposing certain constraints.