Price Competition or Quality Competition? Evidence of Forest Products in Top Exporting Countries

Abstract

Price competition and quality competition are the main ways to increase and maintain international competitiveness in the world. However, competitive strategies can vary significantly from product to product. With the utility function and trade gravity model as the research framework in this paper, we use the top 10 countries in terms of the trade volume of forest products from 2012 to 2021 as samples to systematically explore the differential impacts of price and quality on the international competitiveness of forest products. The results of the panel data model that show the regression coefficient of forest product price in terms of international competitiveness are significantly negative, while the regression coefficient of forest product quality is significantly positive, but the absolute value of the regression coefficient of product quality is higher than that of forest product price. Thus, the price and quality of forest products are key factors affecting international competitiveness in general, with the quality of forest products having a higher impact on international competitiveness than the price. However, a further analysis of different forest product categories and countries revealed significant differences in the significance and magnitude of price and quality impacts on international competitiveness. The quality of forest products contributes more to international competitiveness in Brazil, Canada, Germany, Russia, Sweden and the United States. Conversely, the price of forest products in China, Finland, Italy and Poland contributes more to international competitiveness. Therefore, an objective choice of price, quality or a quality:price ratio strategy, taking into account the industry and characteristics of forest products in each country, can contribute to the sustainable improvement of the international competitiveness of forest products.

1. Introduction

Trade barriers have decreased significantly since the beginning of the 21st century due to rapid developments in packaging and shipping logistics. The global export of forest products has reached USD 410.8 billion in 2021, an increase of USD 33.9 billion from 2012 (source: UN Comtrade database). During this period, developing countries such as China, Brazil and Russia have gradually eroded the market share of high-income countries and become major trading countries for forest products due to their cheap labor and production costs and location benefits. However, some scholars have found that the size of trade in forest products in developing countries is not commensurate with its position of quality. China and other developing countries export forest products of low quality at low prices that are still at the low end of the product value chain [1].

In fact, both product price and quality are important ways for a country to increase or maintain international competitiveness. Whether a country adopts price-based or quality-based competition depends on the characteristics of its own industry for forest products as well as the market environment of the export destination country [2]. Therefore, a systematic study of the impact of the price and quality of forest products on international competitiveness is important to answer the question of which advantages top exporters rely on to enhance international competitiveness. This research is also of practical importance for all exporters of forest products to understand the current state and development of competition in the world and adapt their international competitive strategies to improve their competitiveness.

In recent years, with the development of firm heterogeneous theory, many scholars have progressively focused on export product competition patterns; that is, whether export products competitiveness depends on product price competition or product quality [2]. Melitz (2003) proposed the theory of heterogeneous firms, arguing that firms compete with each other in terms of price and the most efficient firms can achieve increased competitiveness in distant markets at low prices [3]. Subsequently, Melitz and Ottaviana (2008), Bernard et al. (2011) and other scholars further extended and confirmed the theory of firm heterogeneity [4,5,6]. On this basis, Panayiotis and Bardaka (2010) and Bernardina Algieri (2014) analyzed the impact of price and non-price factors on export performance [7,8].

However, Melitz (2003) considered productivity heterogeneity and ignored product quality heterogeneity in firm heterogeneity theory [3]. Baldwin and Harrigan (2011) proposed a quality-modified heterogeneous firm model, which argued that firms can achieve increased market competitiveness through both product quality and price, and higher-quality firms are the most competitive with heterogeneous quality increasing with the firms’ heterogeneous cost [9]. Amit Khandelwal (2006) utilized both unit value and quantity information to infer quality; depending on price, imports with higher market shares are assigned higher quality [10]. Khandelwal’s research has provided fresh ideas for measuring product quality and facilitated research related to product quality. Based on the theory of product quality heterogeneity, Chen Lili (2014) systematically studied the international competition patterns of products from the world’s top 50 exporting countries and found that quality-based competition is not always better than price-based competition [2].

Export prices are an important aspect of heterogeneous enterprise theory. Existing research has found that the main factors affecting export prices are productivity, product quality, distance to the export destination country, exchange rate and tariffs [6,9]. There are three measures of export prices in empirical analysis. The first one is the Tornqvist index of export prices, which is calculated using the price weight method to obtain the deflator of export prices [11]. The second measure is the relative unit price index, which is a relative value obtained by dividing the unit price of an export product by the global average unit price of that product [12]. The third method is the unit value measure, which directly divides the value by the quantity, and is more widely used [13].

Determining the scope of forest products has been the focus of many studies. The classification of forest products has been explored separately by different organizations and scholars. In the Yearbook of Forest Products of the food and agriculture organization of the United Nations (FAO), forest products comprise roundwood, wood charcoal and chips, sawn wood and veneer sheets, wood-based panels, wood pulp and recovered paper, wood paper and paperboard. Cross-references to the central product classification (CPC), Customs Codes and Standard International Trade Classification (SITC), are also available in the FAO Statistical Yearbook for the subcategories of forest products. The Chinese government classifies forest products as roundwood, sawn wood, wood-based panels, wood pulp, wood based-paper and paperboard, wood products and wood furniture in its Forestry and Grassland Statistics Almanac. Based on the classification given by the FAO and the Chinese government, Shi divided forest products into eight categories, including as logs, wood furniture, etc., and explained in detail the corresponding HS and SITC codes for the eight categories of forest products [14]. Forest products were divided based on the classification of wood products into two sectors (wood and paper) and three processing levels: raw wood, semi-finished and finished items [1].

Meanwhile, scholars at home and abroad have used indicative indicators to study the comparative advantages of forest products in major exporting countries including the United States, China, Brazil, Canada, Russian Sweden, Finland and Vietnam, and some studies have roughly discussed the impact of product price and product quality on international competitiveness [15,16,17]. Comparing the international competitiveness of the European furniture industry, Grzegorzewska (2019) noted that the competitiveness of the Polish furniture industry is mainly due to the lower prices of its products [18]. Slobodan Cvetanović (2019) found that the success of the wood industry lies in a newly active comparative advantage in selected Southeast Europe countries, since competitive advantage based on prices and costs obviously disappears [19]. Su Haiying et al. (2020) found that China’s exported forest products are in low-value-added, but its participation in global value chains is gradually increasing [1]. Jiang Bo and Dai Yongwu (2023) analyzed the competitiveness of forest products in major countries and found that improved quality enhances the international competitiveness of major forest product [20]. In addition, factors affecting international competitiveness include the green certification of products, level of technology, and consumer preferences in export markets [14]. Additionally, the impact of recycling renewable resources and used products on the international competitiveness of forest products is gradually increasing with the growing environmental awareness of the public.

In summary, we find that due to the difficulty in measuring the quality of forest products, few scholars have studied the relationship between product quality and international competitiveness. Similarly, fewer scholars have focused on the impact of quality and price on the international competitiveness of forest products. Therefore, we systematically analyze the impact of the price and quality of forest products on international competitiveness in terms of the product dimension in this paper, using the top ten countries for forest product trade as a study sample. We find that the price and quality of forest products play a significant role in international competitiveness, but their effects vary significantly across products and country samples.

The marginal contribution of this paper to previous results comprises the following two points. First, the impact of forest product prices and quality was analyzed based on a broader sample of countries and sub-divisions of forest products. Most of the current literature focuses on the comparative advantages of forest products in individual countries, and the empirical analysis of the factors affecting international competitiveness is minor, especially the lack of research on the impact of forest product prices and product quality on competitiveness. In this paper, we empirically analyze the differential impact of price and quality on the international competitiveness of forest products across different categories and countries in the product dimension, enriching empirical studies on the international competitiveness of forest products. Second, unlike most studies that explain the sources of international competitiveness only in terms of a single dimension of price or quality advantage, this paper provides a relatively consistent theoretical explanation for the sources of international competitiveness of forest products in major countries by conducting empirical analyses in the framework of a price and quality model.

2. Materials and Methods

2.1. Preferences and Market Structure

Based on Meltiz’s (2003) theory of firm heterogeneity [3], Baldwin and Harrigan (2011) and Fan Haichao (2020) have introduced product quality into the study of international trade and constructed a model of the heterogeneous quality of products [9,21]. On the basis of the above scholars, we introduced product quality heterogeneity into the utility function and proposed the theoretical model in this paper. We assume that the utility of a consumer depends on both the quantity and the quality of the product consumed. Thus, the agent’s utility function for the export destination country c, is shown in Equation (1):

$$U_c={\left\{{{\displaystyle {\int }_{\omega ={\mathsf{\Omega }}_c}\left[q_{ic}{(\omega )}^{\frac{\eta }{\sigma }}{x}_{ic}(\omega )\right]}}^{\frac{\sigma -1}{\sigma }}d\omega \right\}}^{{}^{\frac{\sigma -1}{\sigma }}}$$

(1)

In Equation (1), i is the country of export, c is the export destination country, and $i,c\in \left\{1,\hspace{0.33em}\cdots ,\hspace{0.33em}n\right\}$. $\omega$ is a consumer product, $x_{ic}(\omega )$ is the quantity of each variety consumed, and $q_{ic}(\omega )$ is the quality of each variety consumed. The set of products $\mathsf{\Omega }$, that can be purchased by consumers of c, is not the same for different countries. $\eta >0$ is the elasticity of consumer demand with respect to product quality, and is used to measure the quality difference between products. $\sigma >1$ is the elasticity of substitution between varieties.

With the objective of optimizing the utility function ($U_c$), we compute Equation (1) to obtain the $\omega$’s quantity demand function, $x_{ic}(\omega )$ [18], shown in Equation (2):

$$x_{ic}(\omega )={\left[q_{ic}(\omega )\right]}^{\eta }\frac{{[p_{ic}(\omega )]}^{-\sigma }}{P_c^{1-\sigma }}{Y}_j$$

(2)

In Equation (2), $p_{ij}(\omega )$ is the price of $\omega$, $Y_j$ is the total consumption expenditure in the export destination country, $j$, and $P_j={\left\{{{\displaystyle {\int }_{\omega \in {\mathsf{\Omega }}_j}\left[p_{ij}(\omega /q_{ij}(\omega ))\right]}}^{1-\sigma }d\omega \right\}}^{\frac{1}{1-\sigma }}$ is a quality-adjusted price-added index. Equation (2) shows that both product price and product quality are key factors affecting the quantity demand for a product. When the price of the product is given, the higher the quality of the product, the higher the demand for the quantity. When the quality of the product is given, the lower the price of the product, the higher the demand for the quantity. Therefore, in view of the market demand in export destination countries, the key to the competitiveness of products from exporting countries lies in product quality and product price. This is consistent with the conclusion obtained by solving for corporate profit maximization through the framework of corporate heterogeneity theory.

Product competitiveness is usually defined as the ability of an exported product to expand in the international market, corresponding to the amount of demand from a demand perspective [22]. It can be inferred that, just like the quantity demanded, the international competitiveness of a product is affected by the price and quality of the product. The lower the price and the better the quality of the exported products, the stronger the international competition.

2.2. Model Assumption

The trade gravity model is an influential method for analyzing international trade flows [23] which argues that bilateral trade flows are positively correlated with the size of two economies and negatively correlated with the spatial distance between two countries. Subsequently, scholars have introduced exchange rate, resource endowment and other factors into the trade gravity model [24], which is used to study the influencing factors of trade flows, and these research results provide a reference for the study in this paper. With reference to the basic formalism of the trading gravity model and the objectives of this paper, we set the model in the following form.

$$\begin{array}{c}\ln i{c}_{ijt}=\alpha +{\beta }_1\ln qualit{y}_{ijt}+{\beta }_2\ln pric{e}_{ijt}+\eta Z_{ijt}+u_{ij}\\ Z_{ijt}=[\ln lparno, \ln exrate, \ln aforest, \ln dis, \ln dft, \ln gdp, \ln fgdp]\end{array}$$

(3)

In Equation (3), $i$ is the country, $j$ is the main forest product, $t$ is the specific year, and $i{c}_{ijt}$ is the index of international competitiveness of exported forest products, while $qualit{y}_{ijt}$ is the index of quality of exported forest products, and $pric{e}_{ijt}$ is the weighted price of exported forest products. $Z_{ijt}$ is a subset of optional variables that have previously been identified as potentially important explanatory variables.

2.2.1. Measurement of International Competitiveness

There are a number of indicators to measure international competitiveness, with market share (MS, Equation (4)), revealed comparative advantage (RCA, Equation (5)), trade competitiveness (TC, Equation (6)) and relative trade advantage (RTA, Equation (7)) being the most commonly used [25,26,27]. MS focuses on the absolute size of exported products, RCA focuses only on the relative performance of exportation [17], compared to the average level of that product in the world, and RTA and TC favor countries with export surpluses. To address the single-dimensional nature of individual indicators, some scholars have employed objective weights to compute comprehensive international competitiveness indicators with multiple indicators [28]. Therefore, we also use the entropy weighting method to calculate the weights of the RCA, MS, RTA and TC indicators and then obtain the index of the international competitiveness of forest products, so as to comprehensively assess the current state of the international competitiveness of forest products.

$$m{s}_{ijt}=e{x}_{ijt}/e{x}_{wjt}$$

(4)

$$rc{a}_{ijt}=(e{x}_{ijt}/e{x}_{it})/(e{x}_{wjt}/e{x}_{wt})$$

(5)

$$t{c}_{ijt}=(e{x}_{ijt}-i{m}_{ijt})/(e{x}_{ijt}+i{m}_{ijt})$$

(6)

$$rt{a}_{ijt}=(\frac{e{x}_{ijt}/e{x}_{it}}{e{x}_{wjt}/e{x}_{wt}})-(\frac{i{m}_{ijt}/i{m}_{it}}{i{m}_{wjt}/i{m}_{wt}})$$

(7)

In Equations (4)–(7), $e{x}_{ijt}$ is the amount of exports of forest products from country $i$, $e{x}_{it}$ is the amount of exports from country $i$, $e{x}_{wjt}$ is the amount of global exports of forest products, and $e{x}_{wt}$ is the amount of global exports. $im$ is the amount of imports, and the meanings of the specific symbols are similar to those of $ex$.

2.2.2. Measurement of Product Quality

The quality of exported products is usually measured indirectly using the export unit value method and the nested logit method [29]. In the unit-value approach, the average price of a product is used as a proxy indicator, which makes it difficult to exclude the effects of production cost differences, exchange rate changes and other factors, hence more scholars use the nested logit method model to measure product quality. Khandelwal assumed that the market share of export products depends on product price and product quality in the nested logit method; conditional on price, imports with higher market shares are assigned higher quality [10]. Therefore, the nested logit model is set as in Equation (8):

$$\ln S_{icjt}-\ln S_{icj0}=\sigma \ln pric{e}_{icjt}+\theta \ln n{s}_{icjht}-\phi \ln gd{p}_{it}$$

(8)

In Equation (8), $S_{icjt}$ is the market share of forest products, $j$, in year $t$ from country $i$ to destination country c, $pric{e}_{icjt}$ is the price of forest products, j, $h$ is a HS-6 product of species $j$, $n{s}_{jhtg}$ is the market share of $h$ in $j$, and $gd{p}_{it}$ is the gross domestic product per capita of country $i$. The assumptions of the nested-logit model are consistent with the reality of consumers’ pursuit of cost-effective products, so we also use this model to estimate the quality of forest products.

First, we estimate a panel data model of Equation (8) using bilateral trade data with six-digit HS codes to obtain regression coefficients for the variables ($\stackrel{⌢}{\sigma }$, $\stackrel{⌢}{\theta }$ and $\stackrel{⌢}{\phi }$). Then, based on Equation (9), we obtain the quality of forest products at each 6-digit HS code for different bilateral trades. Further processing quality in Equation (10), the relative quality of forest products is calculated. Finally, we computed the relative quality of all 6-digit HS code forest products under the same category weighted by trade value to obtain the weighted quality of the eight categories of forest products in each country.

$$cqualit{y}_{icjt}=\ln S_{icjt}-\stackrel{⌢}{\sigma }\ln pric{e}_{icjt}-\stackrel{⌢}{\theta }\ln n{s}_{icjht}+\stackrel{⌢}{\phi }\ln gd{p}_{it}$$

(9)

$$qualit{y}_{icjt}=\frac{cqualit{y}_{icjt}-\min (cqualit{y}_{icjt})}{\max (cqualit{y}_{icjt})-\min (cqualit{y}_{icjt})}$$

(10)

2.2.3. Measurement of Product Price

Product price is the key explanatory variable, which we measure using forest product-weighted prices (shown in Equation (11)). We obtain the average price of each product via the HS code and then calculate the weighted average price of subcategory forest products weighted by their export scale. $P_{ijt}$ is the price index of forest products, $Tv$ is the amount of forest products exported, $Q$ is the quantity of forest products exported, and the symbols i, j, h and t agree with the above.

$$P_{ijt}={\displaystyle \sum _{j=1}^m{w}_{ijht}{p}_{ijht}}={\displaystyle \sum _{j=1}^m\frac{T{v}_{ijht}}{T{v}_{ijt}}\frac{T{v}_{ijht}}{Q_{ijht}}}$$

(11)

2.2.4. Measurement of Other Factors

The set of control variables, Z, consists of seven variables: the number of long-term trading partners ($\ln lparno$), exchange rate ($\ln exrate$), per capita GDP of the export destination country ($\ln fgdp$), distance to the export destination country ($\ln dis$), forest area per capita ($\ln aforest$), and external dependence of forest products ($\ln dft$), which have been argued by other scholars to be factors affecting international competitiveness [24,28]. $\ln lparno$ is the number of countries that have exported consistently for more than three years, and the formulae for other selected indicators are as follows:

$$fgd{p}_{ijt}={\displaystyle \sum _{c=1}^C{w}_{ijct}gd{p}_{ct}={\displaystyle \sum _{c=1}^C\frac{e{x}_{ijct}}{e{x}_{ijt}}}}gd{p}_{ct}$$

(12)

$$di{s}_{ijt}={\displaystyle \sum _{c=1}^C{w}_{ijct}di{s}_{ic}={\displaystyle \sum _{c=1}^C\frac{e{x}_{ijct}}{e{x}_{ijt}}}}di{s}_{ic}$$

(13)

$$df{t}_{ijt}=\frac{e{x}_{ijt}+i{m}_{ijt}}{{\mathrm{GDP}}_{it}}$$

(14)

3. Samples and Dates

3.1. Samples

Our study focuses on wood-based forest products, and, unless otherwise stated, the forest products in this paper are wood-based forest products. The scientific classification of product categories is a prerequisite for our study of forest products. In the introductory chapter we stated the approach taken by some institutions and scholars to the classification of forest products. We used Shi’s approach of Shi (2015) as it is more explicit, operational, and consistent with the purpose of this research paper [14,30]. All forest products with a four- or six-digit HS code were aggregated into eight groups, which are wood furniture, wood products, wood chips, roundwood, sawn wood, wood-based paper products, wood-based panels and wood pulp. Meanwhile, the corresponding customs codes for each category of forest products are listed in

Table 1. The forest products classified into eight groups at the HS code level.

Types	HS Code	Types	HS Code
wood chips	4401, 4402, 4404, 4405	roundwood	4403
sawn wood	4406, 4407	wood-based panels	4408–4412
wood products	4413–4421	wood pulp	4701–4705
wood-based paper products	4801–4813	wood furniture	940161, 940169, 940330–940360

.

Sample selection was also a focus of the study. We measured the average annual export value of forest products for all countries in the world from 2012 to 2021 and selected the top 10 as the sample countries for this paper. These are Brazil, Canada, China, Finland, Germany, Italy, Poland, Russia, Sweden and the United States, which together account for more than half of the world’s export market for forest products (source: UN Comtrade database).

The UN Comtrade database provides national trade data for forest products exported between 2012 and 2021. Prices, quality and long-term partners for forest products are available through calculations. The World Bank database provides GDP per capita, forest area and exchange rates. We obtained geographic distances between countries from the CEPII website.

3.2. Stationarity Test of Data

Table 2. Descriptive statistics.

Variables	Observations	Mean	Sd	Min	Max
lnic	800	0.338	1.006	−4.414	2.000
lnquality	800	−0.844	0.150	−2.147	−0.999
lnprice	800	−1.259	1.313	−4.241	4.183
lnlparno	800	4.221	0.818	0.693	5.278
lnexrate	800	0.936	1.301	−0.336	4.150
lnaforest	800	4.622	1.572	2.620	6.908
lndis	800	8.023	0.725	6.621	9.741
lndft	800	−6.743	1.327	−3.290	−12.737
lngdp	800	10.10	0.758	8.699	11.00
lnfgdp	800	9.596	0.513	7.101	10.46

lists the descriptive statistical values of the main variables. In this paper, we use stata16 for data processing and model estimation when not otherwise specified.

Table 3. The results of the panel unit root test.

Variables	Year	Observations	IPS	LLC
lnic	10	800	−5.322 ***	−4.500 ***
lnquality	10	800	−5.290 ***	−15.193 ***
lnprice	10	800	−14.096 ***	−7.4502 ***
lnlparno	10	800	−5.955 ***	−24.111 ***
lnexrate	10	800	−1.361 *	−8.819 ***
lnaforest	10	800	−3.356 ***	−6.876 ***
lndis	10	800	−4.092 ***	−19.147 ***
lndft	10	800	−16.442 ***	−55.671 ***
lngdp	10	800	−7.283 ***	−33.308 ***
lnfgdp	10	800	−6.001 ***	−18.807 ***

*** p < 0.01, and * p < 0.1.

shows the results of the panel unit root test for the main variables, and we can see that the IPS (Im–Pesaran–Shin) and LLC (Levin–Lin–Chu) test results are in agreement for all panel series. For the lnexrate, IPS and LLS reject the null hypothesis that all panels contain unit roots, with significance at 10% and 1%, respectively. For all other variables, IPS and LLS reject the null hypothesis of significance at 1% and show that the panels of these variables are stationary.

3.3. Model Test

The sample data in this paper are panel data including three dimensions: country, product type and year. To successfully perform the estimation of 3D panel data, we use the group method to reduce 3D panel data into 2D panel data by grouping countries and product types. The advantage of downscaling is that it makes it easy to estimate the coefficients of the variables using conventional panel estimation methods, while simultaneously addressing the effects of heteroskedasticity, cross-sectional correlation, and serial correlation [31].

Before model estimation, we examined the product price and quality correlations. Using Pearson correlation analysis, we found that the correlation coefficient between the price and the quality of forest products is −0.089, which is not significant. Also, for panel model (lnquality = c + alnprice) estimation, the coefficient of the lnprice is not significant. These findings above suggest that the price and quality of forest products are linearly uncorrelated.

The model’s formalism and estimation methods are also the focus of this paper. The F-test shows that the fixed-effects model is better than the mixed-effects model, and the Hausman test shows that the fixed-effects model is better than the random-effects model. The modified Wald test indicates inter-group heterogeneity in the full-sample panel data, and the Wooldridge test indicates intra-group autocorrelation in the full-sample panel data. To solve the above problems, we use the feasible generalized least squares method (FGLS) in this paper, which improves the estimation efficiency and the accuracy of statistical inference [32].

4. Results

4.1. Robustness Analysis

Table 4. The estimation results of different models.

Model Class	RE	FE-WRG	FE-FGLS	FE-FGLS	FE-FGLS
Object	ALL	ALL	All	Except BRA and CAN	Except SWE and USA
lnquality	0.150 ***	0.149 ***	0.182 ***	0.184 ***	0.216 ***
	(2.600)	(2.641)	(4.898)	(6.657)	(5.936)
lnprice	−0.172 **	−0.214 **	−0.114 **	−0.107 ***	−0.100 ***
	(−1.972)	(−2.145)	(−2.327)	(−4.543)	(−3.268)
lnlparno	0.358 ***	0.277 *	0.477 ***	0.659 ***	0.493 ***
	(2.681)	(1.971)	(6.479)	(8.589)	(7.502)
lnexrate	−0.181	−0.247	−0.007	−0.074	−0.161
	(−1.592)	(−1.203)	(0.050)	(−0.541)	(−0.609)
lnaforest	0.320 ***	0.929	0.313 ***	0.527 ***	0.319 ***
	(3.519)	(1.496)	(2.843)	(3.352)	(4.834)
lndis	0.152	0.138	0.170 **	0.174 *	0.196 ***
	(1.301)	(1.066)	(2.057)	(1.909)	(2.743)
lndft	2.275 ***	2.603 ***	1.616 ***	1.613 ***	2.033 ***
	(2.602)	(2.840)	(3.718)	(3.041)	(5.263)
lngdp	−0.342 **	−0.486 **	−0.025	−0.334	−0.235
	(−2.459)	(−2.337)	(−0.139)	(−1.381)	(−0.782)
lnfgdp	2.147 **	2.474 **	1.622 ***	1.481 ***	1.973 ***
	(2.370)	(2.638)	(3.672)	(2.672)	(5.175)
Constant	−11.325 **	−14.005 **	−13.13 ***	−10.305 ***	−12.658 ***
	(−2.051)	(−2.540)	(−4.547)	(−3.078)	(−3.538)
i.year	Yes	Yes	Yes	Yes	Yes
Observations	800	800	800	640	640

Robust t-statistics (z-statistics) in parentheses; *** p < 0.01, ** p < 0.05, and * p < 0.1; WRG: within regression estimator.

lists the estimation results for different models, the first column of which is the full-sample random-effects model, while the others are fixed-effects models with different estimation methods or samples. In the model’s RE, FE-WRG (within the regression estimator) and FE-FGLS, the estimated coefficients and significance levels for the key variables, forest product quality and price are consistent, and the estimated coefficients for the control variables are in the same direction of influence with slight differences in significance levels. The relative consistency and stability of the estimation results across different models justifies the model settings and choices.

Robustness analysis is a key aspect of model testing. In this paper, international competitiveness is a composite index, and it is difficult to reflect the full picture of the international competitiveness of forest products by choosing other single indicators. We can use the unit value as a proxy indicator of the quality of forest products, but that is highly correlated with the price of forest products. Therefore, there is no suitable alternative indicator for the key variables, hence we choose to remove some of the country samples for robustness analysis in this paper. We again test the model on the panel data by removing the first two and last two country samples based on the alphabetical ordering of the sample countries. The results of the estimation are shown in columns 5 and 6 in Table 4. Comparing the full-sample estimation results with the partial-sample results, the coefficients for forest product quality and price are essentially the same in size and direction with improved significance, which indicates that the fixed-effects FGLS results are accurate and robust.

4.2. Analysis of Overall Results

We can find the following conclusions from Table 4. First, the quality of forest products in each country has a significant positive impact on the international competitiveness of forest products. In all models, the regression coefficient for product quality was greater than 0.1 with 1% significance, thus indicating that product quality improvement is conducive to the increased international competitiveness of forest products, which is consistent with the conclusion of the heterogeneous firm model for product quality in the paper by Baldwin and Harrigan. Second, the price of forest products in each country has a significant negative impact on the international competitiveness of forest products. In all models, the regression coefficient for the product price is negative and passes the 5% significance test, indicating that a decrease in the product price favors an increase in the international competitiveness of forest products. The negative effect of price is consistent with the implicit assumption of Khandelwal (2010)’s product quality model that international competitive market share for forest products is negatively related to product price [10]. The above empirical analysis confirms that the characteristics of forest product trade are consistent with the theory of enterprise heterogeneity, and that both product price and quality are key factors affecting the international competitiveness of forest products. Any improvement in the quality of forest products or a reduction in the price of products will boost the international competitiveness of forest products. However, the optimal competitive strategy with which to promote international competitiveness is to increase the product quality:price ratio of forest products.

However, there are significant differences in the magnitude and significance of the regression coefficients for forest product quality and price. In the FE-FGLS model, the estimated coefficient of forest product quality was 0.182 (p < 0.01) and the estimated coefficient of product price was −0.114 (p < 0.05). This suggests that a 1% increase in the quality of forest products contributes more to an improvement in international competitiveness than a 1% reduction in the price of their products. In summary, the quality of forest products plays a more obvious role in improving the competitiveness of forest products than does the price, which is consistent with the author’s viewpoint in another paper, that is, the current improvement in the international competitiveness of forest products is mainly a result of the improvement in product quality [20].

4.3. Results for Different Product Categories

To further identify the differential impact of forest product price and quality on international competitiveness, we estimate the regression coefficients of the variables in this paper for different forest product categories, and the estimation results are shown in

Table 5. The estimation results of different product categories via FGLS.

	Round Wood	Wood Chips	Sawn Wood	Wood-Based Panels	Wood Products	Wood Pulp	Wood-Based Paper Products	Wood Furniture
lnquality	0.338 ***	0.438 ***	0.060 **	0.269 ***	0.304 ***	0.182 ***	0.431 ***	0.330 ***
	(5.766)	(15.485)	(1.974)	(4.702)	(5.927)	(5.148)	(8.453)	(6.124)
lnprice	0.051	−0.179 ***	−0.513 ***	−0.011	0.050 **	−0.063	−0.470 ***	0.188 ***
	(1.232)	(−5.822)	(−9.392)	(−0.370)	(2.038)	(−0.875)	(6.933)	(3.050)
Control	YES	YES	YES	YES	YES	YES	YES	YES
Constant	−2.971	6.381 *	3.417	−5.893	−3.194	−18.923 **	−12.246 **	20.788 ***
	(−0.962)	(1.864)	(1.020)	(−1.299)	(−0.759)	(−2.571)	(−2.078)	(3.706)
i.Year	YES	YES	YES	YES	YES	YES	YES	YES
Observations	100	100	100	100	100	100	100	100

z-statistics in parentheses; *** p < 0.01, ** p < 0.05, and * p < 0.1.

. Since the Wald and Wooldridge tests indicate the presence of inter-group heterogeneity and intra-group autocorrelation in the panel data for categorical forest products, we use FGLS to estimate the panel model for the eight categories of sub-divided forest products.

The results show that there are significant differences in the impact of the price and quality of different types of forest products on international competitiveness. Forest product quality across all categories had a significant positive effect on international competitiveness, which is consistent with the findings from the overall sample. However, there are differences in the magnitude of the quality coefficients for different forest products, reflecting the differential impact of product quality on international competitiveness.

Prices of wood chips, sawn wood and wood-based paper products had a significant negative impact on international competitiveness, which is consistent with the overall findings. Prices of logs, wood-based manufactured boards and wood pulp have had a non-significant impact on international competitiveness. This is mainly attributed to the weakening of the role of price due to significant improvements in the quality of these types of forest products, e.g., international competition for wood pulp only increased by 12% from 2012 to 2021, but the quality of its products increased by 29.33%. Puzzlingly, the price of wood products and wood furniture has had a significant positive impact on international competitiveness. We believe that wood products and wood furniture products are high-value-added products compared to other forest products, with relatively higher product quality, a higher degree of differentiation, lower consumer price sensitivity, and easier transportation over further distances to obtain higher premiums, which result in a significant positive impact on prices [33].

4.4. Results for Different Countries

Similarly, we use FGLS to estimate the impact of the price and quality of forest products on international competitiveness in different countries, as shown in

Table 6. The estimation results of different countries via FGLS.

Country	lnquality	lnprice	Control	Constant	i.Year	Observations
BRA	0.132 *** (−3.79)	−0.023 (−1.360)	YES	0	Yes	80
CAN	0.247 *** (−3.676)	0.041 (−1.307)	YES	−64.683 *** (−7.535)	Yes	80
CHN	0.073 *** (−3.733)	−0.205 ** (−2.554)	YES	110.713 *** (−9.365)	Yes	80
DEU	0.321 *** (−4.391)	−0.145 *** (−3.522)	YES	−16.683 *** (−3.720)	Yes	80
FIN	0.11 (−1.261)	−0.081 * (−1.778)	YES	0	Yes	80
ITA	0.226 *** (−3.987)	−0.227 *** (−5.455)	YES	−14.941 *** (−3.645)	Yes	80
POL	0.011 (−0.248)	−0.228 *** (−10.326)	YES	0	Yes	80
RUS	0.131 * (−1.833)	−0.111 ** (−2.518)	YES	0	Yes	80
SWE	0.919 *** (−10.548)	0.311 *** (15.374)	YES	−42.585 *** (−12.927)	Yes	80
USA	0.250 *** (−3.055)	0.051 *** (−2.686)	YES	−10.845 *** (−6.424)	Yes	80

z-statistics in parentheses; *** p < 0.01, ** p < 0.05, and * p < 0.1.

. Consistent with the overall sample results, the quality of forest products in most countries has a significantly positive effect on the international competitiveness of forest products, while the price of forest products has a significantly negative effect.

More specifically, Finland has significant competitiveness in high-value-added forest products [17], but the effect of product quality on international competitiveness is not significant. Poland’s product quality does not play a significant role in the international competitiveness of forest products, but the price of its products is significantly negative with a large value, which is essentially in line with Grzegorzewska’s study that the competitiveness of Poland’s furniture industry mainly stems from the lower price of its products [18]. The insignificant impact of forest product prices on international competitiveness in Canada may be related to the overall superior quality of its forest products, resulting in a weaker role for prices. U.S. prices show a positive and small effect on the international competitiveness of forest products, which is related to a high share of transportation costs in the price and a high premium in product quality, as U.S. forest products are exported over the longest distances among the sample countries [33]. The significantly positive impact of forest product prices on international competitiveness, with a relatively high coefficient, is an unusual result. We believe that it may be the brand, technology and other factors that contribute to the relatively high price:performance ratio of their products.

5. Discussion

Some studies have shown that rich countries have a better preference for product quality [34]. In other words, the quality of exports from rich countries is better than that from developing countries, but this does not work for trade in forest products. Comparing the quality of forest products in the observed countries, the average value of quality for forest products in developing countries, including Brazil, China and Russia, is slightly higher than that in developed countries, with Brazil ranked third and Russia fifth among the countries investigated. The situation is similar for the forest products category, where in the case of wood pulp, Brazil and China rank second and third among the countries observed in terms of the mean value of the product quality index. The degree of affluence of exporting countries is indeed conducive to the quality of forest product products, but resource endowment, technology level, and production efficiency are also important factors affecting the quality of forest products [29].

Currently, more scholars are advocating that developing countries should upgrade the quality of their products and replace price competition with quality competition to enhance their international competitiveness [2]. In practice, Chinese and Brazilian forest products are following this strategy to improve the quality and the added value of their products [1,35]. In this paper, we find that improving the quality of forest products is definitely conducive to international competitiveness, but do not conclude that quality competition is the only way for developing countries to capture international markets. Germany, Italy and Poland can still rely on price competition to enhance the international competitiveness of forest products, and price competition remains a sharp weapon for these countries to capture the international market for forest products with. Therefore, countries should consider their own circumstances and choose price competition or quality competition, or improve the cost-effectiveness of forest products.

Finally, we explored the impact of product quality and price on the international competitiveness of forest products and found some interesting conclusions, but our study still has some limitations. Firstly, our empirical analysis is profoundly dependent on international trade data, and the results of our study are strongly affected by the accuracy of the data. Secondly, the impact of price and product quality on international competitiveness may be non-linear [36], which will be the direction of our future research. Third, the export prices of forest products may have fluctuated significantly during the COVID-19 pandemic due to logistical embargoes and market controls, but we do not discuss this in this paper.

6. Conclusions

In the framework of the utility function and trade gravity model, we investigate the impact of the price and quality of forest products on international competitiveness, focusing on the top 10 countries in terms of the global trade volume of forest products.

The estimation results for the full sample show that the regression coefficient for forest product price has a significantly negative effect on international competitiveness, while the regression coefficient for the quality has a significantly positive effect, but the absolute value of the regression coefficient for product quality is higher than that for forest product price. The results indicate that improving the quality or lowering the price of forest products will significantly improve the international competitiveness of forest products, but quality is more effective than price in improving international competitiveness.

However, the direction, magnitude and significance of the impact of forest product price and quality on the international competitiveness of forest products varies significantly across forest product categories and sample countries. The estimation results for different types of forest products show that the quality of forest products contributes more to international competitiveness in round wood, wood chips, wood-based panels, wood products and wood furniture, while the price of forest products contributes more to international competitiveness in sawn wood, wood-based paper products. Similarly, the quality of forest products contributes more to international competitiveness in Brazil, Canada, Germany, Russia, Sweden and the United States. Conversely, the price of forest products in China, Finland, Italy and Poland contributes more to international competitiveness. This fully reflects the complexity and variability of the factors affecting the international competitiveness of forest products, so we need to conduct a specific analysis based on the specific situation of the international competitiveness of forest products in each country.

At the same time, we find that the quality preferences of rich countries are not fully applicable to trade in forest products. The view of some scholars that developing countries must enhance their international competitiveness with an improved quality of forest products is biased and may even hinder the enhancement of the international competitiveness of forest products in developing countries.

Therefore, we suggest that all countries should have a clear understanding of the level of international competitiveness of their forest products and its main influencing factors, and select targeted paths to enhance the international competitiveness of forest products, such as improving production efficiency (lowering prices), product quality, and quality-price ratio. The optimal international competitive strategy in China, Poland and Finland is a price competition strategy, and the optimal strategy of Brazil, Canada, Germany, Sweden and the United States is a quality competition strategy, and there is not much of a difference in the effect of Italy and Russia choosing a price competition or quality competition strategy. Considering the types of forest products, it is preferable for countries with round wood, wood chips, wood-based panels, wood products and wood furniture as their major export products to choose a quality competition strategy, and for countries with other forest products as their major export products to choose a price competition strategy.