Can the Reform of “Streamline Administration and Delegate Power, Improve Regulation, and Optimize Services” Help Drive Export Diversification of Wood-Processing Enterprises?

Abstract

The Chinese government has often been suspected of excessively intervening in the market. The reform of “Streamline Administration and Delegate Power, Improve Regulation, and Optimize Services” (RSDO) is a key initiative by the Chinese government to help enterprises improve their international competitive advantage through institutional reforms. Few studies have empirically tested the effects of RSDO on enterprises’ export diversification. Using the data of 2141 wood-processing enterprises (WPEs) in China in 2014, this study applies zero-truncated negative binomial regression and the Tobit model to evaluate the effects of prefecture-level RSDO on enterprises’ export product and market diversification. The results show that the RSDO can enhance WPEs’ export product diversification without significantly impacting export market diversification. Regarding specific indicators, the reforms of “Streamline Administration and Delegate Power” and “Improve Regulation” significantly positively affect export product and market diversification. In contrast, the optimization of government services shows no significant impact. Heterogeneity tests show that the RSDO promotes export diversification in wooden products and furniture manufacturing enterprises, smaller enterprises, domestic enterprises, and those with weaker operational capabilities. The impact mechanism shows that the RSDO aids sample enterprises in enhancing export product diversification by lowering operational costs and supports export market diversification by encouraging technological innovation.

1. Introduction

On 2 April 2025, the United States announced the imposition of a 10-percent “minimum baseline tariff” on all imports and an “individualized reciprocal higher tariff” on the countries and regions with which the United States “has the largest trade deficits”. Subsequently, the Trump administration imposed a minimum tariff rate of 145% on Chinese goods imported to the United States. This event once again highlights potential disruption in export-oriented economies due to an overly concentrated export market. Export market diversification is a crucial strategy to deal with tariff barriers and even other trade protectionism [1]. And export product diversification can help a country or enterprise to mitigate international market risks, counter external demand shocks, and achieve stable export growth [2]. However, the uniqueness of the Chinese economy provides a crucial background for research. According to data released by the General Administration of Customs of China in 2024, however, export value to the five developed countries or regions (the United States, European Union, Japan, South Korea, Australia, the United Kingdom, and Canada) accounted for 42.93% of China’s total export value. In a government-led economy, local governments exert significant influence over resource allocation, with export diversification emerging as a core objective of structural reform. Thus, China needs to enhance its export diversification further, similar to the challenges many countries face worldwide. Currently, research on the factors influencing export diversification remains limited, particularly regarding the impact mechanisms across different development stages, enterprise sizes, and industry characteristics, which have yet to be fully explored. Therefore, it is crucial to identify the key factors affecting export diversification, which can help find policy breakthroughs that enable enterprises to enhance their levels of export diversification.

Institutional factors could be one of the significant influences on the diversification of enterprises’ export markets or products. RSDO offers a valuable experimental opportunity. Traditional international trade theories construct models and analyze theories based on the assumption that “institutions are fixed and there are no transaction costs” [3]. This led scholars to overlook the impacts of institutional factors on regional or enterprise export trade. However, the “trade driven by institutions” theory has attracted academic attention [4]. For example, enterprises must frequently interact with local governments and relevant functional departments, preparing various documents and undergoing approval and regulatory procedures. If there are institutional and systemic barriers, friction costs will arise, equivalent to enterprises being subjected to “hidden taxes”. This will reduce the enterprises’ cost advantages, thereby restricting the development and diversification of international markets. By streamlining approval processes and optimizing supervision, RSDO directly reduces institutional transaction costs for enterprises. This aligns with Williamson’s transaction cost theory and offers a valuable experimental opportunity for research. Therefore, analyzing the impact of RSDO on export diversification from the perspective of institutional change can provide micro-level evidence for the theory of “institutionally driven trade.

The RSDO began in 2013, and the Chinese government views the RSDO as a ‘preemptive move’ and a ‘strategic initiative’, designed to deepen political and economic reforms, transform government functions, optimize the business environment, and lower institutional costs for enterprises. By the end of July 2014, the State Council had abolished and delegated 468 administrative approval items. A total of 4350 approval items were canceled, delegated, or simplified by provincial governments across the country [5]. However, there are regional differences in the breadth, depth, and speed of the RSDO. Data from the ‘National Private Enterprise Survey Database in 2014’ show that Jiangsu Province accounted for 7 of the top 20 prefecture-level cities in the RSDO evaluation, far outperforming other provinces. Does the regional institutional environment difference caused by the differing progress of the RSDO affect enterprise export diversification? Is the effect positive or negative? What are the underlying mechanisms? A scientific and systematic empirical analysis based on objective data is needed to determine this, to provide reliable decision-making references for government policies.

Some scholars have evaluated the impact of institutional factors, such as institution quality [6], business environment [7], and trade facilitation measures [8,9], on export product or market diversification based on cross-country, cross-province, national, and industry-level data. However, most existing studies rely on national or provincial data, thereby overlooking the heterogeneity of institutional reforms among Chinese municipal administrative units and their influence on export diversification for small and micro-enterprises. In addition, some research based on micro-enterprise data from China [2,10,11] mainly evaluates the export diversification promotion effect of a single policy (such as the foreign trade registration system) based on data from Chinese enterprises before 2007, which lacks timeliness. Therefore, it is imperative to systematically analyze the mechanisms through which RSDO affects export diversification, considering differences in municipal-level reforms and micro-enterprise data.

The wood-processing industry in China has long been subject to complex and often burdensome administrative procedures, including regulations on logging, wood processing, and export licensing. These regulatory challenges have contributed to high transaction costs and limited the flexibility of enterprises to respond to changing international market demands. Given these challenges, the RSDO reforms, which aim to simplify administrative processes, reduce regulatory burdens, and promote market liberalization, are of particular relevance to this sector. By focusing on wood-processing enterprises (WPEs), this study seeks to examine how these reforms can alleviate the regulatory obstacles faced by the industry and foster greater export diversification. Furthermore, as a major sector in China’s manufacturing industry, WPEs provide a clear example of how institutional reforms can enhance operational efficiency, stimulate product innovation, and open up new market opportunities in the global economy. The Chinese wood-processing industry faces the reality of over-concentrated export markets and an export product structure that still needs optimization. The average number of export products of 2141 wood-processing enterprises in this study is only 6.32. Additionally, most enterprises’ export product diversification index (from 0 to ∞) is below 0.2. It is significantly influenced by institutional factors, such as environmental regulations, safety production supervision, logging permits, wood processing and operation permits, and certification of available timber resources [12,13]. Moreover, the central government actively promotes the RSDO in the forestry sector, such as approving twenty-seven administrative powers, including the filing and approving of pilot enterprises for processing imported logs and exporting sawn timber. To this end, this study takes wood-processing enterprises as the research subject, using data from the “National Private Enterprise Survey Database in 2014” to quantify and measure the prefecture-level cities’ RSDO. It then evaluates the effects of the reform on the enterprise export market and product diversification, enriching the micro-level evidence of how institutional factors influence export diversification.

The potential academic contributions of this study include three aspects: First, this study evaluates the impact of RSDO reform in prefecture-level cities on enterprise export products and market diversification using zero-truncated negative binomial regression and Tobit models. Previous studies mainly focus on specific aspects of the RSDO, such as streamlining administration and the delegation of powers (e.g., foreign trade “registration system”). Studies that comprehensively evaluate the trade effects of the RSDO are scarce. This study tentatively evaluates the effects of the RSDO and its subcomponents on WPEs’ export product and market diversification. It contributes to theoretical research on the RSDO’s economic effects. Second, unlike the existing literature, this study conducts heterogeneity tests on enterprises across different industry types, sizes, ownership structures, and asset operation capabilities. This study shifts the research perspective on the trade effects of institutional quality or business environment from the national, provincial, and industry levels to the prefecture-level city and micro-enterprise levels. The observed data on the effectiveness of the RSDO at the prefecture-level city level are obtained, which fully consider the institutional (business) environment heterogeneity between prefecture-level cities within a province and its impact on enterprise export behavior. This allows for a more accurate assessment of the trade effects of institutional or business environments. Third, this study employs the Bootstrap method to test whether the RSDO reform influences enterprise export behavior through cost reduction and technological innovation promotion. It examines whether the RSDO indirectly affects the export diversification of sample enterprises through two pathways: lowering operating costs and promoting technological innovation. It systematically identifies the mechanisms through which the RSDO exerts its influence. We aim to uncover more valuable decision-making insights.

2. Literature Review and Critique

The literature on measuring export diversification levels and identifying their key influencing factors is relatively abundant. This study focuses solely on the research literature examining the effects of institutional factors closely related to the scope of the RSDO. To provide a comprehensive review, this study includes the literature on the impact of institutional factors on the expansion of exports (i.e., increased export products), export market size, and related topics.

2.1. The Impact of Institutional Factors on International Trade

Kamuganga [14] found that export costs, export time, operating costs, export procedures, weak export support institutions, and institutional quality significantly negatively impacted the product diversification index, number of export products, export market diversification index, and the number of export markets. An important marginal contribution of this study is the increased consideration of export markets. Dennis and Shepherd [15] and Persson [8] argued that reductions in export costs (transportation costs of a 20-foot container) and export time (the average number of days required to complete customs clearance for export products) could effectively promote an increase in the number of export products in developing countries. Gul et al. [16] reached the same conclusion. From an industrial perspective, Liu and Li [17] announced that increasing an industry’s contract intensity and enforcement rate improved the diversification of export products in China. Bian and Qiang [10] shifted the focus to the micro-enterprise level, using the change in foreign trade operation from the “approval system” to the “registration system” on 1 July 2004 as an experimental opportunity, and they found that the “registration system” significantly increased the export product variety of export enterprises by boosting domestic export market supply, providing empirical evidence at the micro level in the Chinese context. Zhang et al. [11] found that being located in export processing zones did not help enterprises explore new export markets but significantly increased the variety of new export products. However, both studies are based on empirical research of very specific and singular policies, which offer limited decision-making reference value for current policy designers. Hu and He [2] argued that regions with lower marketization levels and higher government intervention suppressed export enterprises’ expansion into new export markets.

2.2. The Impact of Regulations on International Competitiveness

Most current studies rely on cross-country and cross-year data, covering developing countries, transitional economies, and developed nations, with conclusions that offer useful reference points. Cabral and Veiga [18] found that improvements in government accountability, rule of law, political stability, administrative efficiency, and corruption control could significantly enhance a country’s export diversification level. Elhiraika and Mbate’s [19] research further confirmed the positive effect of administrative efficiency. Balavac and Pugh verified the promoting effect of lower domestic market entry costs on enhancing a country’s export product diversification level and increasing export products. Giri et al. [20] found that the higher the governance quality (corruption control, rule of law, and government institutional quality), the higher the level of export product diversification in a country. Azam [21] argued that optimizing industrial policies in Pakistan, including increased export subsidies and tax rebates, significantly promoted export product diversification. Rémy and Ayivodji [7] asserted that increasing trade openness (i.e., simplifying import and export approval procedures) could help to enhance a country’s export product diversification level. Yin et al. [22] showed that business environment improvements promote the export variety of a country. Xiong and Wu [23] found that countries with better contract systems had more export product types in contract-intensive industries. The above literature only focused on the impacts of institutional factors on export diversification at the product level. Since the Chinese government emphasizes market diversification or decentralization, the decision-making reference value of the above literature is limited.

2.3. The Impact of Trade Facilitation Policies on Export Diversification

Relevant institutional arrangements for trade facilitation directly impact enterprise exports’ fixed and variable costs. This impact on export diversification is a key area of concern in academic research. Some scholars have constructed trade facilitation measurement variables and tested their impact on a country’s export diversification. Wilson et al. [24] declared that trade facilitation promoted increased export products by reducing fixed and variable costs. For 2007-17 and 92 countries, Kurul [25] suggested a significant positive relationship between border efficiency and variety in products and markets, and the availability and quality of infrastructure enhance variety in markets. The trade facilitation data in the “World Bank Doing Business data” are the most commonly used. Beverelli et al. [26] used the trade facilitation index compiled by the OECD and found that trade facilitation promoted the growth of a country’s export product and market numbers. Using the same trade facilitation data, Fontagné et al. [9] shifted the research perspective from the national and industry levels to the enterprise level. They found that optimizing trade facilitation measures significantly increased the number of export products for small-scale European export enterprises. Surprisingly, there is still a lack of substantial research studying the relationship between institutional factors and export diversification in the Chinese context. Measuring trade facilitation at the provincial level fails to capture institutional differences between prefecture-level cities and counties, such as varying administrative approval efficiency and customs clearance procedures within and outside the Free Trade Zone in Fujian Province. This makes it difficult to accurately assess the trade effects of the RSDO or other reforms. Lv and Huang [27] claimed that trade facilitation significantly increased the number of exported products in China. Ouyang and Park [28] suggested that implementing digital trade facilitation significantly improved the extensive margin of exports. Given the imbalanced market-oriented reforms in China, which result in varying institutional environments across regions, Duan and Liu [29] calculated a trade facilitation index for 31 provinces (autonomous regions and municipalities). They found that trade facilitation significantly increased the number of export markets and products for enterprises. Launched in 2013, the RSDO reform in China focuses on streamlining administrative procedures, delegating powers, and optimizing services to reduce institutional costs for enterprises. By July 2014, the State Council had abolished or delegated 468 administrative approval items, and provincial governments had canceled, delegated, or simplified a total of 4350 approval items [5,23]. The reform aimed to alleviate regulatory burdens, thereby enabling enterprises to diversify their export markets and products. However, regional differences in the implementation of the RSDO have led to varying impacts on export diversification across different prefecture-level cities [10,29]. Brazil has implemented various programs aimed at diversifying its export base, including the simplification of export procedures and the provision of incentives for small- and medium-sized enterprises (SMEs) to enter international markets. While these programs have led to some diversification, Brazil continues to face challenges related to infrastructure and regulatory complexities that hinder the full potential of export diversification. India’s “Make in India” initiative and the EU’s Single Market reforms similarly aimed at reducing regulatory barriers but faced hurdles such as infrastructure limitations and differing national regulations, respectively. These experiences provide valuable insights into the potential and challenges of export diversification in the context of institutional reforms.

In summary, there is little research that systematically examines the impact effect of the RSDO on enterprise export diversification. The contribution of the RSDO in advancing China’s export diversification strategy remains unclear. Research that examined the effects of institutional factors, such as trade facilitation, export time, and intellectual property protection systems, on export diversification from national, provincial, and industry perspectives mainly focused on export product diversification, with more emphasis on the number of export products rather than the degree of export product dispersion or diversification. There is less research on export market diversification or dispersion. In addition, considering that county-level and municipal-level governments and relevant departments directly impact enterprises within their jurisdiction, previous studies evaluating provincial institutional quality’s impact on enterprise export diversification may misestimate the trade effects of institutional quality. Given this, this study introduces new data sources to examine the status of the RSDO at the prefecture city level. It evaluates its impacts on WPEs’ export product and market diversification, which makes a significant marginal contribution to theoretical and policy reference value.

3. Theoretical Analysis and Research Methodology

3.1. Theoretical Analysis

In March 2013, the Chinese government launched a commitment to reduce at least one-third of administrative approval items across all departments of the State Council within five years, marking the beginning of the “Streamline Administration and Delegate Power” (RSDO) reform. This initiative aimed to enhance the efficiency and effectiveness of governance while fostering market-driven economic activities. As part of this broader effort, the Chinese government later introduced the “Improve Regulation” and “Optimize Services” reforms in 2015, which together form a comprehensive institutional framework. The core of the “Streamline Administration and Delegate Power” initiative lies in reducing administrative approvals and transferring non-essential government functions to the market, thereby alleviating constraints on businesses and allowing the market to take a more decisive role in resource allocation. This approach, consistent with “Institutional Economics Theory”, aims to reduce transaction costs and increase economic efficiency, ultimately stimulating the vitality of market entities and fostering innovation. The “Improve Regulation” reform focuses on enhancing regulatory oversight during and after the business process, shifting from a model of “strict entry and loose regulation” to “loose entry and strict regulation”. This shift seeks to foster fair competition, mitigate market distortions, and ensure that deregulation does not lead to chaotic market behavior. This aspect of the reform aligns with “Dynamic Capabilities Theory”, which emphasizes how institutional improvements can strengthen firms’ capacities to innovate, adapt, and effectively compete in a newly liberated market environment. The “Optimize Services” component is designed to improve government service delivery by streamlining procedures, increasing service awareness, and introducing innovative methods of interaction with businesses. By ensuring that public services remain efficient and accessible even as administrative powers are reduced, this reform aims to avoid a “service vacuum” and ensure continued support for enterprise activities. This service optimization directly supports businesses’ ability to engage in more agile market behavior and expansion, which is consistent with “New Trade Theory”. The theory highlights how reducing operational costs and improving market access through enhanced governmental services can enable firms to diversify their products and markets, positioning them more effectively in global trade.

Together, these reforms aim to create a more business-friendly environment by reducing institutional and operational barriers, allowing firms to leverage market forces more effectively. By lowering administrative burdens, strengthening regulatory frameworks, and improving service delivery, the RSDO reforms enhance the overall competitiveness of enterprises, fostering export diversification and contributing to broader economic growth. This integrated approach not only aligns with theoretical frameworks but also provides a realistic and actionable pathway for promoting sustainable economic development in China.

3.1.1. Analysis of the Direct Impact of the RSDO

The RSDO covers a wide range. Thus, almost all departments potentially serve as reform implementers. In policy recommendations for promoting export diversification, governments should place greater emphasis on improving product quality and fostering innovation. The success of RSDO reforms lies not only in the growth in the number of products but also in innovation-driven export diversification. Governments should further streamline administrative processes, enhance intellectual property protection, and reduce innovation costs to provide stronger support for enterprise innovation. Particularly for small- and medium-sized enterprises (SMEs), governments can help reduce barriers to entering new markets, assisting them in improving product innovation and competitiveness, thereby driving higher-quality export diversification, for example, reducing and simplifying export-related approvals, improving customs clearance efficiency, providing international market information, and promoting overseas markets through streamlining administration and optimizing services. This can greatly reduce the difficulty for wood-processing enterprises to start an export business for the first time and the difficulty of entering new markets. This may directly promote wood-processing enterprises entering export markets and developing new markets, thus achieving higher export market diversification and expansion levels. Improvements in regulatory quality, judicial efficiency, intellectual property protection, and market dispute resolution have reduced the space for counterfeit and substandard products while increasing the cost of violations. WPEs are stimulated to place more emphasis and invest more resources in improving product quality and developing new products. This could promote an increase in the number of wood-processing enterprises exporting products. In summary, hypothesis H1 is proposed:

H1: 

Local RSDO has a direct positive effect on wood-processing enterprises’ export diversification.

3.1.2. Analysis of the Impact Mechanism of the RSDO

(1) Reducing operating costs. When analyzing export diversification, although the number of export products is a fundamental measure, we must also consider product quality and innovation. Product quality and innovation are critical drivers of export diversification, as they not only determine a product’s competitiveness in international markets but also directly impact whether enterprises can enter new markets and maintain their market share. To achieve long-term export diversification, enterprises must continuously focus on improving product quality and fostering innovation, developing unique and high-value-added products. Through innovation, enterprises can not only increase the variety of export products but also enhance market share and enter higher-end markets. Classical and new trade theories suggest that country A enterprises with lower export prices for a particular product can enter the export market and trade with other countries. Additionally, if a new country demands the product internationally and its domestic price is higher than imports from country A, enterprises in country A can enter new export markets. Low prices stem from low costs, which are driven by higher production efficiency (due to technology level) or more abundant and cheaper factor endowments [30]. This study’s average cost-to-income ratio (operating costs and period expenses to sales revenue) of 2141 WPEs is 84.01%. Even for enterprises below the 25th percentile, the average exceeds 80%, highlighting the need to help wood-processing enterprises reduce operating costs. Regional RSDO may help WPEs establish cost advantages through various channels, thereby enhancing the level of export product and market diversification for enterprises. The ‘Streamline Administration and Delegate Power’ reform covers all operational processes, including business establishment, material procurement, production, domestic sales, export, and transportation. For example, in February 2013, the Second Plenary Session of the 18th Central Committee proposed reforms to the business registration system. Reforms such as the ‘subscribed capital registration system’ and the ‘cancellation of annual business inspections’ were later implemented. The establishment of the “power list” and “negative list” systems ensures that “what is not authorized by law cannot be done” and “what is not prohibited by law can be done” [31]. Administrative service fees and government funds are managed through a catalog list system, canceling, suspending, and reducing over 1100 central and provincial government administrative service fees (including those in the import and export processes), resulting in a total reduction in market entities’ burdens by over CNY 3 trillion [32]. The mentioned reforms help enterprises reduce institutional costs [33], improve production and operating efficiency, and accelerate the production and supply of new products. The above improvements can offset the fixed and variable costs of export enterprises entering new export markets and reduce the sunk costs incurred after the failure to develop new markets. This may encourage enterprises to attempt to open new export markets. Additionally, these reforms may help enterprises open new export markets and increase product variety in existing markets, based on price advantages and product uniqueness [34]. Thus, hypothesis H2 is proposed:

H2: 

The RSDO can enhance the export diversification level of enterprises by reducing operating costs.

(2) Promoting technological innovation. RSDO reforms, by simplifying administrative processes and lowering market entry barriers, have reduced operational costs for enterprises, creating favorable conditions for innovation and product quality improvement. Enterprises can allocate the cost savings to research and development and technological innovation, thereby enhancing product quality and competitiveness. This not only helps enterprises expand product variety in existing markets but also promotes their entry into new international markets. By providing a more equitable and competitive market environment, RSDO reforms encourage enterprises to drive export product diversification through innovation. Posner [35] proposed the technology gap theory, which argues that the technology gap is one of the sources of comparative advantage and export benefits in international trade. Helpman and Krugman [36] argued that innovation and supplying differentiated products are the basic conditions for enterprises to establish market power, expand sales scale, and obtain increasing returns to scale. The “trade driven by institutions” theory suggests that improvements in institutional quality can promote trade through at least three pathways: promoting technological innovation, human capital development and accumulation, and increasing returns to scale. Chinese wood-processing enterprises undergo a “cost curse” (i.e., costs rise uncontrollably after falling to an optimal point) as their cost advantage diminishes. They must accumulate technology to develop innovative products, achieving ‘what others don’t have, I possess’. Alternatively, they can produce differentiated products at the same cost, improve products to achieve ‘what others have, I excel at’, or offer superior products more cheaply. The RSDO can create a favorable external environment for technological innovation and quality improvement. First, the RSDO significantly lowers market entry barriers, entry time, and costs for new enterprises. This encourages more competitors in an industry, fostering market competition and stimulating enterprises to develop new products or explore new markets for survival and growth [37]. Second, the RSDO provides a fair and competitive market environment, compressing the space for homogeneous or imitation products, which encourages the development of diversified products [38]. Forcing enterprises to meet market demand by expanding the production of differentiated products enhances the expansion of export markets and increases the variety of export products for enterprises. Third, the RSDO ensures enterprises’ safety, stability, legality, fairness, freedom, convenience, and confidence, promoting the free flow of capital, technology, markets, and talent. This will enable enterprises to focus on long-term development to unlock existing capabilities and accumulate new ones, increase substantial innovations, and then build a foundation for productivity improvements and product diversification through technological diversification. In summary, hypothesis H3 is proposed:

H3: 

The RSDO can enhance the export diversification level of enterprises by promoting technological innovation. The mechanism of the impact of the RSDO on the export diversification of wood-processing enterprises is shown in Figure 1.

3.2. Research Methodology

(1) Zero-truncated negative binomial regression. The number of WPE export products and markets is used as measurement variables for export diversification. Both are typical non-negative discrete variables suitable for count models, such as Poisson regression, negative binomial model, zero-truncated Poisson model, and zero-truncated negative binomial regression. Among them, the Poisson regression model has an important assumption: the conditional mean of the dependent variable is equal to its conditional variance. In this study, the number of export products and export markets of 2141 sample enterprises varies greatly. Their variances (116.46, 159.27) are much larger than the means (6.32, 11.55), making Poisson regression unsuitable. The negative binomial model does not require the expected value of the variable to equal the variance, which solves the problem present in Poisson regression [39]. Moreover, export product data types and market quantities are “non-negative truncated integer data (zero-truncated data)”. The likelihood function must be adjusted for consistent estimation, whether it is the Poisson or negative binomial model [40]. In summary, the zero-truncated negative binomial regression is introduced to identify the effects of the RSDO on the number of export products and export markets of WPEs. The initial Poisson regression model is designed:

$$E\left(y_i\left|x_i,\beta \right.\right)=Var\left(y_i\left|x_i,\beta \right.\right)={\lambda }_i=\exp (\beta x_i)$$

(1)

In the equation, y_i is the dependent variable, representing the number of export products or export markets for the i-th wood-processing enterprise. Since the condition of “the mean equals the variance for the number of export products and the number of export markets” is not met, the Poisson model is transformed into a negative binomial distribution model. At this point, an unobserved effect, denoted as ${\epsilon }_i$, is introduced into the conditional mean, which can affect the number of export products and markets for the sample enterprises. At this point, an unobserved effect, denoted as a, is introduced into the conditional mean, which can affect the number of export products and markets for the sample enterprises. The Poisson model is extended to:

$$E\left(y_i\left|x_i,\beta \right.\right)=\exp (\beta x_i+{\epsilon }_i)$$

(2)

Taking the natural logarithm of both sides of Equation (2) results in the negative binomial distribution model:

$$Ln\left[E\left(y_i\left|x_i,\beta \right.\right)\right]=\alpha +\eta RSD{O}_i+\beta x_i+{\epsilon }_i$$

(3)

In Equation (3), x_i is the vector of control variables for the i-th enterprise, and $\delta$ is the vector of estimated coefficients for the control variables; η represents the effect of the RSDO on the number of export products or export markets, controlling for other explanatory variables; α is the constant term; ε_i is the random error term. The model estimation results based on zero-truncated negative binomial regression will produce an overdispersion parameter alpha; if the model rejects the null hypothesis of alpha = 0, it indicates that the negative binomial distribution model is superior to the Poisson regression model. Conversely, the opposite is true. The model estimation results will also generate the log-likelihood statistic and its p-value. If the p-value is less than 0.05, the zero-truncated negative binomial regression is superior to the standard negative binomial regression.

(2) Tobit model. Two of the measurement variables for export diversification, the export product diversification index and the export market diversification index, are continuous variables between 0 and 1. They are classified as “non-negative truncated data”; some values are 0, not following a normal distribution, making it impossible to directly use OLS regression. To address this, the Tobit model is introduced, which is known as a censored or truncated regression model and is one of the limited dependent variable regression models. Based on the left-censoring characteristic of the data, the Tobit model can set the rule that “values of the dependent variable equal to or below the threshold of 0 will be censored”. The standard Tobit model is specified as:

$$\begin{array}{l}{diversity}_i^{\ast }=\alpha +{\beta }_{1}RSD{O}_i+\delta x_i+{\epsilon }_i\\ \left\{\begin{array}{l}{diversity}_i={diversity}_i^{\ast },If,{diversity}_i^{\ast }>0\\ {diversity}_i=0,\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\hspace{0.25em}\mathrm{If}, {diversity}_i\le 0\end{array}\right.\end{array}$$

(4)

In the equation, $diversit{y}_i^{\ast }$ represents the latent variable of the export product (or market) diversification index for the i-th enterprise; ${diversity}_i$ represents the observed value of the export product (or market) diversification index for the i-th enterprise; β₁ is the effect of the RSDO on the enterprise’s export product (or market) diversification index; x_i is the vector of control variables for the i-th enterprise; δ is the vector of estimated coefficients for the control variables; α is the constant term; ε_i is the random error term.

(3) The model’s endogeneity issue in Equations (3) and (4) must be considered and tested. Using “whether the party secretary of the prefecture-level city has been replaced (replacement = 1; no replacement = 0)” as an instrumental variable for the RSDO, the rationale is that the short tenure and promotion tournaments lead to a rational official’s optimal response of “immediately upon taking a new position, pushing harder and striving for quick results” [41]. In the first half of 2013, various levels of government in China launched a wave of “streamlining administration, delegating powers, and improving services” reform. This reform became a key opportunity for new party secretaries to establish their achievements. It naturally became one of the main tasks for newly appointed party secretaries and local government officials to meet the demand for ‘short-term results’ and fully promote the RSDO. Therefore, the change in party secretaries in 2013 is significantly related to the local RSDO in 2014 but has no direct connection with the export behavior of wood-processing enterprises in 2014, thus meeting the homogeneity requirement for the instrumental variable. First, the two-step method of endogeneity Hausman test proposed by Wooldridge is selected to test the endogeneity problems of Equation (3) [41]. In the first step, regression is carried out with the RSDO as the dependent variable and instrumental variable as well as other existing control variables as independent variables to obtain the residual term. In the second step, the residual term is incorporated into Equation (3). The results show that regression coefficients of the residual term are, respectively, −1.044 (p-value being 0.721) and −0.679 (0.669) for the equation with the export product diversification index and export market diversification index as the dependent variable, indicating that the RSDO can be regarded as an exogenous variable. Second, using the “ivtobit” code in Stata 14.0 and the Wald test (null hypothesis: all explanatory variables are exogenous), the IV Tobit estimation for Equation (4) is conducted. The results show that the Wald statistics are 0.32 and 0.28, with p-values of 0.5699 and 0.5974, respectively. So, we cannot reject the null hypothesis of “α = 0” for homogeneity, meaning there is no endogeneity issue in Equation (4) fitting, and the RSDO is not an endogenous variable. Therefore, subsequent empirical analyses will use the Tobit model to estimate and discuss results.

(4) The indirect effect testing method is the Bootstrap method. The testing method for indirect effects is similar to mediation, but there are still differences. Mediation effects assume that the explanatory variable X significantly influences dependent Y. However, indirect effects do not require this assumption because many indirect paths may exist [42]. The sum of indirect effects from different paths may be close to 0, leading to no significant effect of X on Y. An initial exploration using the export market diversification index (Y) shows that the RSDO (X) has no significant effect on Y, which cannot meet the requirements for mediation effect testing. Therefore, this study tests the indirect effect of the RSDO on the export behavior of wood-processing enterprises, identifying the impacting pathways of the reform. Currently, the most commonly used methods include stepwise testing, the Sobel test, and the Bootstrap method. Drawing on the research of Zhong et al. [43], the Bootstrap method is chosen to test whether the RSDO affects the export behavior of wood-processing enterprises indirectly by promoting technological innovation and reducing operating costs.

4. Data Sources and Descriptive Statistical Analysis

4.1. Variable Design

(1) Dependent variable. When measuring enterprise export diversification, two common methods are typically used: one is to count the number of export products for enterprises based on the internationally recognized Harmonized System (HS) 8-digit product code; the other one is to count the number of export markets for enterprises based on the export destination country (market) codes. The second method involves indirect measures such as the Herfindahl–Hirschman Index (HHI), Theil’s Entropy Index, and other similar indicators [44]. To ensure the robustness and reliability of the econometric analysis results, this study uses the two measurement methods mentioned above, including enterprise export diversification measurement variables, such as the number of export products, export product diversification index, number of export markets, and export market diversification index. Among them, the HHI is used to calculate the export market diversification index at the enterprise level, with the formula as follows:

$$deversity=1-{\displaystyle \sum _{i=1}^n{s}_i^2}$$

(5)

The observed values are distributed between 0 and 1. In the equation, s_i represents the proportion of a specific product’s export value to a specific destination country in the total export value of the enterprise. Furthermore, the export product diversification index is calculated using Theil’s Entropy Index. The specific steps are as follows: An export enterprise can export n types of HS 4-digit products, and these n products belong to j categories of 2-digit product codes (n ≥ j). The entropy index (pER) between 4-digit products within 2-digit product categories is defined as related diversification, and the entropy index (pEU) between 2-digit products is defined as unrelated diversification. The enterprise product diversification is the sum of related and unrelated diversification [45]. The specific formula is as follows:

$$pER={\displaystyle \sum _{i\in j}\frac{p_i}{p_j}}\ln {\displaystyle \frac{p_j}{p_i}}pEU={\displaystyle \sum _{j=1}^j{p}_j\ln \frac{1}{p_j}}$$

(6)

In the equation, i represents the product type at the 4-digit level; j represents the product type at the 2-digit level; and p represents the enterprise’s export value. This study’s export product diversification index is standardized, so its observed values are distributed between 0 and 1. The larger the values of these four export diversification measurement indicators, the higher the enterprise’s export diversification level.

Explanatory variable: RSDO. The “National Private Enterprise Survey Database in 2014” include comprehensive information on the RSDO. First, the question “What are the main factors you think contribute to the improvement of the business environment?” (multiple choice) includes but is not limited to two options: “reduced administrative approvals” and “enhanced service awareness of government departments”. For these two options, enterprises that select them are assigned a value of 1. Otherwise, zero is assigned. Second, the question “What are the main problems in current market regulation?” (multiple choice) includes overlapping regulatory functions, unclear departmental responsibilities, selective law enforcement, insufficient punishment for violations, excessive penalties, and others (please specify). Enterprises that do not select any options are assigned a value of 6, while those selecting one to six options are assigned values of 5, 4, 3, 2, 1, and 0, respectively. Third, the average of the sum of scores for the three items for all companies in each postal area is calculated and used to measure the status of RSDO for the postal area. After matching with the “China Industrial Enterprises Database”, the measure of the RSDO at the prefecture-level city is indirectly obtained based on geographic location information. The larger the observed value of this variable, the better the effectiveness of the RSDO. Mediating variables: To capture the indirect effects of the RSDO on export diversification through two pathways—reducing operating costs and promoting technological innovation—interaction terms are included between the RSDO and the two mediating variables. Operating costs are measured by the ratio of sales costs to sales revenue, multiplied by 100%. Technological innovation is measured by total factor productivity calculated using the Solow Residual Method. The total output value of the enterprise is the dependent variable, while the number of employees and the total value of fixed assets are the independent variables. The residual term from the estimated model is the Solow residual, which serves as the measure of technological innovation.

Control variables: Based on enterprise heterogeneity theory, dynamic capabilities theory, and research by Lin et al. [46], control variables at the enterprise level include rent-seeking behavior, subsidy income, size, years of establishment, management expense ratio, debt ratio, comprehensive tax rate, financial expense ratio, and whether the enterprise is state-owned, foreign-funded, or invested in by Hong Kong, Macau, or Taiwan. Excess management expenses are used to measure rent-seeking behavior in enterprises [47]. The specific calculation is as follows: the management expense ratio (management expenses/sales revenue × 100%) is the dependent variable, while main business income, debt ratio, main business profit margin, and number of employees are independent variables. The residual term from the estimated model is the measure of rent-seeking behavior. Additionally, control variables at the prefecture city level are included, such as the GDP size of the prefecture-level city, per capita GDP, distance to the nearest coastal port, telecom business income, and the number of internet users [48,49]. The specific variable explanations are provided in

Table 1. Interpretation of variables and results of descriptive statistics.

Variant	Explanation	Average Value	Standard Deviation	Minimum Value	Maximum Values
Number of export products	Obtained according to the method described in Section 4.1	6.32	6.17	1.00	393.00
Export product diversification index	Obtained according to the method described in Section 4.1	0.34	0.39	0.00	0.99
Number of export markets	Obtained according to the method described in Section 4.1	11.55	2.02	1.00	104.00
Export market diversification index	Obtained according to the method described in Section 4.1	0.57	0.38	0.00	0.99
RSDO	Obtained according to the method described in Section 4.1	3.05	2.16	0.279	6.63
Share of unusual expenditures	(Government contribution + public relations and hospitality + fees and charges)/sales revenue × 100%	4.84	0.31	3.78	6.18
Technological innovation	Calculations based on the Solow residual value method	2.49	2.59	0.002	12.66
Business cost	Cost of sales/revenue from sales × 100%	86.08	7.82	69.85	100.49
Rent-seeking behavior	Obtained according to the method described in the previous section	−0.04	0.60	−5.75	14.97
Subsidized income	The logarithmic of the current year’s government subsidy income	0.48	1.91	0.00	12.95
Scale	The logarithmic value of total assets	10.54	1.24	6.22	17.63
Founding Years	2013-Factory start-up time	9.01	4.96	1.00	53.00
Management cost ratio	Administrative expenses/sales revenue × 100%	0.58	0.27	−0.94	2.45
Gearing ratio	Total liabilities/total assets × 100%	48.15	4.12	0.31	111.05
Composite tax rate	Total tax revenue/sales revenue × 100%	1.13	1.19	-0.89	6.21
Financial cost ratio	Finance costs/sales revenue × 100%	13.38	7.13	3.10	19.87
Whether it is a state-owned enterprise	State-owned enterprises = 1; otherwise = 0	0.01	0.00	0.00	1.00
Whether it is a foreign company	Foreign-invested enterprises = 1; otherwise = 0	0.07	0.39	0.00	1.00
Whether it is a Hong Kong, Macao, or Taiwan enterprise	Hong Kong, Macao and Taiwan investment = 1; otherwise = 0	0.09	0.41	0.00	1.00
Prefecture-level city GDP size	The logarithmic value of GDP of prefecture-level cities	17.41	0.84	14.73	19.19
GDP per capita of prefecture-level cities	The logarithmic value of per capita GDP of prefecture-level cities	11.06	0.73	9.04	13.06
Distance to the nearest coastal port	The logarithmic value of the distance between the prefecture-level city government location and the nearest coastal port	4.43	1.33	0.57	7.15
Revenue from telecommunications operations	The logarithmic value of telecommunication revenues at the prefecture-level	13.35	0.88	9.71	15.60
Number of Internet users	Internet users per 100 population (households)	5.14	0.862	2.30	6.64

.

4.2. Data Sources

The survey involves 6144 enterprises, covering all 31 provincial-level administrative units (excluding Hong Kong, Macau, and Taiwan), 236 postal areas, and 538 counties. This county encompasses counties with high, medium, and low levels of economic development, effectively ensuring the representativeness of the data. To improve accuracy, 81 postal areas with fewer than 10 enterprises in the survey sample were excluded, leaving 155 postal areas. Data on corporate public relations expenses, government fees paid, and government allocations due are also sourced from these data. Second, microdata such as the financial status of WPEs are obtained from the “China Industrial Enterprises Database in 2014”. The enterprise sample in this database includes all state-owned and non-state-owned enterprises with main business income exceeding CNY 20 million. Data processing is performed following the approach of Lin et al. [46]. The industrial enterprise data are then matched with the data on the RSDO in 155 postal areas based on the postal code rules (the first two digits represent the province, and the third digit represents the postal area). Using the first three digits of the postal code, the industrial enterprise data match the reform effectiveness data of 155 postal areas. Duplicate prefecture-level cities within the same postal area are removed, leaving 148 prefecture-level cities and 11,353 WPEs. Export transaction records are selected, excluding records from trading intermediaries, enterprises with export quantities less than 1, export values under 50 USD, or those with only one export destination. In addition, the records with missing or abnormal values for variables such as enterprise names are removed. Then, the data are matched based on the enterprise name with the 11,353 WPEs, and samples with duplicate company names are removed. We obtain a dataset of 2143 companies with export businesses, distributed across 23 provinces and 103 prefecture-level cities. Fourth, prefecture-level cities’ total and per capita GDP data come from the “China City Statistical Yearbook (2015)”. The longitude and latitude of ports and the government locations of prefecture-level cities are based on Google Maps, and port data come from the Ministry of Transport’s “National Coastal Port Layout Plan in 2006”.

Despite the limitations of the 2014 dataset, such as its restriction to a single year, which may limit the observation of the long-term effects of RSDO, its representativeness, timeliness, and comprehensiveness render it a suitable choice for evaluating the impact of RSDO on export diversification. First, the “National Private Enterprise Survey Database in 2014” provide the most comprehensive and closest original data at the prefecture city level for the RSDO content, effectively addressing the issue of missing data for the RSDO evaluation. Compared to the “World Bank Doing Business data”, it helps to bring the research focus from the national and provincial levels down to the prefecture city level. Second, the effectiveness of the RSDO had entered its visible stage, and the surveyed enterprises could perceive and evaluate it. After the RSDO was launched in 2013, the central and local governments quickly advanced it. By the end of July 2014, the State Council had already canceled or delegated 468 administrative approval items. The National Private Enterprise Survey Database in 2014 began in June of 2013, and the surveyed enterprises could perceive and provide an objective evaluation of the RSDO. The reliability of the RSDO evaluation data ensured the reliability of empirical analysis results in this study. Other scholars have used these data in related studies on the RSDO [50]. Third, it meets the timeliness requirements for policy effectiveness evaluation. Policy implementation and effectiveness evaluations are generally conducted one year after implementation. Although the “National Private Enterprise Survey Database in 2014” baseline is 2013, the respondents may be evaluated based on their perceptions since 2014. By this time, differences in the progress and effectiveness of the RSDO between prefecture-level cities have emerged. It is now possible to quantitatively assess whether there is a significant difference in the level of export diversification in regions with varying effectiveness.

4.3. Descriptive Statistical Analysis

The descriptive statistical analysis in Table 1 shows that there are differences in the maximum and minimum values of most variables, including the RSDO, technological innovation, operating costs, and rent-seeking behavior, indicating that the distribution of observations among enterprises exhibits a dispersed characteristic. This may affect the export product and market diversification of WPEs, making the inclusion of these variables both feasible and necessary. The average number of export markets for the sample enterprises is greater than the number of export products, and the export market diversification index is higher than the export product diversification index. For that, the level of WPEs’ export market diversification is higher than that of product diversification.

Figure 2 shows the scatter plots between each dependent variable and the RSDO. The fitting value curves that slope upwards to the right in the four graphs indicate that in regions with a better “RSDO”, the level of WPEs’ export diversification is higher. It can be preliminarily concluded that the RSDO has a positive effect, which still needs to be scientifically tested through econometric analysis.

WPEs’ export product diversification level was still very low, with the number of export products concentrated below 10, accounting for 83.62% of the total sample. And the overall sample mean was only 6.32. The export product diversification index was mostly below 0.2, indicating the limited expansion of export products. Second, the number of export markets was concentrated below 20, accounting for 82.19% of the total samples, with an overall mean of 11.55. However, the export market diversification index was mostly above 0.6, indicating that the distribution of enterprise export value across markets was more diversified. The WPEs’ export market diversification was higher than export product diversification.

The RSDO status was processed for 23 provinces and 103 prefecture-level cities, ranging from 0 to 8. A larger value indicates a higher evaluation of the local RSDO by enterprises in the jurisdiction. The evaluation results of the RSDO follow a normal distribution, with values concentrated between 4.7 and 5.3. Based on the actual evaluation results from enterprises, the top 20 cities are as follows: Yuncheng (6.18), Huai’an (5.73), Jinhua (5.56), Ganzhou (5.54), Suqian (5.53), Hefei (5.48), Fushun (5.46), Xuzhou (5.45), Shangqiu (5.44), Jiangmen (5.41), Guilin (5.40), Baoshan (5.40), Chuzhou (5.39), Wuxi (5.34), Ji’an (5.31), Yangzhou (5.30), Yancheng (5.28), Lianyungang (5.27), Chengdu (5.23), Shantou (5.22). Among them, Jiangsu Province accounted for seven cities (58% of its 12 prefecture-level city samples), and Guangdong, Zhejiang, Anhui, and Jiangxi provinces each accounted for two cities. Due to space limitations, the evaluation results of the RSDO for all prefecture-level city samples cannot be presented. However, it can be observed that Jiangsu Province has the best results in terms of the effectiveness of the RSDO.

5. Analysis of Empirical Results

5.1. Baseline Regression Analysis

From the second and fourth columns of

Table 2. Benchmark fitting results.

	Export Product Diversification		Export Market Diversification
	Number of Products	Diversity Index	Number of Markets	Diversity Index
RSDO	0.250 **	0.013 **	0.176	0.008
	(0.13)	(0.007)	(0.12)	(0.02)
Technological innovations	0.038 **	0.003 **	0.042 ***	0.012 ***
	(0.02)	(0.00)	(0.02)	(0.00)
Business cost	−0.038 **	−0.002 *	−0.027 **	−0.006 *
	(0.02)	(0.00)	(0.01)	(0.00)
Rent-seeking behavior	−0.319 ***	−0.013 **	−0.224 ***	−0.048 ***
	(0.07)	(0.01)	(0.06)	(0.02)
Government subsidies	0.026	0.002	0.081 ***	0.009 **
	(0.02)	(0.00)	(0.02)	(0.00)
Size of the enterprise	0.058	−0.001	0.102 ***	0.006
	(0.06)	(0.00)	(0.04)	(0.01)
Founding Years	0.017	0.002 ***	0.005	0.002
	(0.01)	(0.00)	(0.01)	(0.00)
Gearing ratio	−0.864 ***	−0.041 **	−0.232	−0.096 *
	(0.27)	(0.02)	(0.24)	(0.06)
Asset turnover ratio	0.321 ***	0.014 ***	0.205 ***	0.049 ***
	(0.07)	(0.01)	(0.06)	(0.02)
Tax burdens	−0.064	−0.005 *	−0.044	−0.005
	(0.07)	(0.00)	(0.03)	(0.01)
Financial cost ratio	−0.023 **	−0.001 **	−0.013 *	−0.001
	(0.01)	0.00	(0.01)	(0.00)
Whether it is a state-owned enterprise	−18.497 ***	−0.712	−2.906	−0.345
	(0.37)	(.)	(2.92)	(0.31)
Whether it is a foreign-invested enterprise	0.340 *	0.021 ***	−0.309 ***	−0.056 ***
	(0.18)	(0.01)	(0.10)	(0.02)
Whether it is a Hong Kong, Macao or, Taiwan investment enterprise	0.158	0.001	−0.121	−0.043 **
	(0.11)	(0.01)	(0.09)	(0.02)
Geographic location	0.567 ***	0.021 **	0.956 ***	0.074 ***
	(0.167)	(0.009)	(0.254)	(0.026)
Industry categories	−1.101 ***	−0.045 ***	−0.377 ***	−0.108 ***
	(0.107)	(0.006)	(0.095)	(0.018)
Prefecture-level city GDP size	−0.06	−0.035 ***	−0.181 *	−0.126 ***
	(0.32)	(0.01)	(0.11)	(0.03)
GDP per capita in prefecture-level cities	0.239	0.013 **	0.147 **	0.030 *
	(0.19)	(0.01)	(0.07)	(0.02)
Distance to the nearest coastal port	−0.009	−0.003	−0.04	−0.009
	(0.08)	(0.00)	(0.03)	(0.01)
Revenue from telecommunication services in prefecture-level cities	0.012	0.006	0.134	0.024
	(0.09)	(0.01)	(0.15)	(0.04)
Number of Internet users in prefecture-level cities	0.000	0.000 ***	0.001 *	0.000 ***
	(0.00)	0.00	0.00	0.00
Constant term	−5.059 *	0.075	−0.496	0.415
	(2.59)	(0.15)	(1.33)	(0.42)
Superdistribution adjoint parameter alpha = 0	2.838 ***(0.835)		9.204 *** (5.187)
Log-likelihood statistics	−5553.248 ***(0.000)		−6964.179 *** (0.000)
Sigma		0.106 *** (0.00)		0.323 *** (0.05)

Note: *, **, *** represent significance at the confidence levels of 90%, 95%, and 99%, respectively, with standard errors in parentheses.

, the overdispersion parameter values (alpha) for the export product type count equation and the export market count equation are 1.038 and 2.034, respectively. Both are significant at the 95% confidence level, meaning the null hypothesis of ‘alpha = 0’ (i.e., rejecting Poisson regression) is rejected. There is indeed overdispersion in the export product type count for the sample enterprises, and negative binomial regression should be used. The log-likelihood values for the two equations are −5639.028 and −6974.729, whose p-values are both significant, indicating that the model is overall significant. We reject the standard negative binomial regression. In conclusion, applying a zero-inflated negative binomial regression model is reliable.

Economically, the results suggest that RSDO enhances innovation opportunities for businesses by optimizing the business environment through measures, such as streamlining and accelerating approval processes, ensuring fair regulation, and improving public services. For example, canceling production license management for engineered wood products allows enterprises to respond quickly to market demands and profit from them [51]. The positive effects of the RSDO create a favorable business environment, which helps enterprises establish optimistic expectations and encourages them to increase product development efforts rather than waiting and observing [52]. The above conclusion is generally consistent with the literature on how the institutional environment [6,53], business environment [9,29], and trade facilitation [7,8,40] impact the increase in export products at the national or enterprise level. Optimizing the business environment can lower both fixed and variable costs for enterprises, thereby facilitating their entry into new markets and the development of new products.

Second, the RSDO does not significantly impact the WPEs’ export market diversification. As seen in the fourth and fifth columns of Table 2, the estimated coefficients for the RSDO on the number of export markets and the export market diversification index are still not significant at the 90% confidence level. The expansion of new markets and the promotion of export market diversification by WPEs are unrelated to the effectiveness of regional RSDO. The fundamental motivation for enterprises to open new markets is that the marginal revenue from selling products in new markets is greater than zero. From the above empirical conclusion, the RSDO (including but not limited to simplified trade approvals, optimized customs services, etc.) has failed to help WPEs improve the marginal returns from market development. The underlying reason may be that it fails to significantly reduce the fixed costs of new market development and the variable costs of operations. This is inconsistent with the research literature on how intellectual property protection systems [28], trade facilitation [26], and business environments [14,29] affect the growth of export market numbers at the national or enterprise level. A possible reason is that the processing trade still accounted for a significant portion of China’s wood-processing industry during the sample period. Due to processing trade contract restrictions and control by overseas parent companies, the export market choices of processing trade enterprises may have become rigid, constraining the increase in export market numbers and the improvement in export diversification [54]. Whether the enterprise is foreign-invested or has Hong Kong, Macau, or Taiwan investment hurts export market diversification, which is evidence.

5.2. Subdivided Indicator Tests

The ‘Streamline Administration and Delegate Power’ reform has consistently enhanced the quality of export products and market diversification for WPEs, as shown in

Table 3. Sub-indicator examination of the RSDO—export product diversification.

	Number of Export Products			Export Product Diversification Index
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Simplification of administrative powers	1.309 ***			0.066 ***
	(0.473)			(0.023)
Improvement of regulation		0.057 **			0.003
		(0.026)			(0.002)
Optimization of services			−0.392			−0.020
			(0.286)			(0.021)
Control variables	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
N	2141	2141	2141	2141	2141	2141

Note: **, *** represent significance at the confidence levels of 95% and 99%, respectively, with standard errors in parentheses.

and

Table 4. Sub-indicator examination of the RSDO—export market diversification.

	Number of Export Markets			Export Market Diversification Index
	Model 7	Model 8	Model 9	Model 10	Model 11	Model 12
Simplification of administrative powers	0.276 ***			0.314 ***
	(0.067)			(0.067)
Improvement of regulation		0.011 *			0.013 **
		(0.006)			(0.006)
Optimization of services			−0.025			−0.017
			(0.062)			(0.062)
Control variables	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
N	2141	2141	2141	2141	2141	2141

Note: *, **, *** represent significance at the confidence levels of 90%, 95%, and 99%, respectively, with standard errors in parentheses.

. This finding aligns with the micro-effect research on trade approval procedures, export time, and similar variables [7,8,14,39]. After streamlining administrative approval, government intervention in the operations of wood-processing enterprises decreases. This improves the efficiency of enterprises’ responses to customer demand in new markets, promoting the extensive marginal expansion of both product and market diversification.

Second, the improvement in regulation has a significant positive effect on the export product and market quantity and the export market diversification index for wood-processing enterprises. This finding aligns with the macro-level research of Yin and Gao [55] and Chen and Wang [56]. The improvement in regulation is mainly reflected in fully promoting fair regulation, maintaining a market order of fair competition, reducing the costs enterprises incur from administrative regulation, and creating greater cost-saving opportunities for enterprises’ product development and expansion into new international markets.

Third, optimizing government services does not significantly impact WPEs’ export product and market diversification. This contrasts with the empirical conclusions of scholars, like Cabral et al. [18], Elhiraika and Mbate [19], and Giri et al. [20], who attribute regional and industry export diversification to administrative efficiency and government institution quality. A possible reason is that the effects of government service optimization have not yet manifested during the sample period, or the speed of service optimization by various levels of government and relevant departments has been slower than the simplification of administrative powers and regulation improvement.

5.3. Heterogeneity Test

Considering that the impacts of the RSDO may vary across different types of enterprises, heterogeneity tests are conducted based on four dimensions: industry type, enterprise size, enterprise nature, and asset operation capability. Industry categories include wood products and furniture manufacturing enterprises and paper product manufacturing enterprises. Enterprise nature is divided into foreign-invested enterprises (including foreign enterprises in Hong Kong, Macau, and Taiwan-invested enterprises) and domestic enterprises. Sample enterprises are classified as large or small scale based on the 50th percentile of total assets. Enterprises are also divided into those with strong and weak operational capabilities based on the 50th percentile of asset turnover. The test results are detailed in

Table 5. Results of heterogeneity test.

	Business Classification	Export Product Diversification		Export Market Diversification
		Number of Products	Diversity Index	Number of Markets	Diversity Index
Industry category	Wood products and wood furniture manufacturers (N = 1644)	0.271 ***	0.021 ***	0.327 ***	0.033
	Paper and paper products manufacturers (N = 497)	−0.186	−0.029	−0.868	−0.119 **
Enterprise size	Larger enterprises (N = 1045)	0.137	0.011	−0.146	−0.010
	Smaller enterprises (N = 1096)	0.331 ***	0.014	0.349 ***	0.029
Nature of business	Foreign enterprises (N = 862)	0.006	−0.004	−0.057	−0.013
	Domestic enterprises (1279)	0.375 ***	0.023 ***	0.305 ***	0.018
Asset operating capacity	High operational capacity (N = 635)	−0.013	0.003	−0.639	−0.125 ***
	Weak operational capacity (N = 1506)	0.282 ***	0.015 **	0.273	0.036 *

Note: *, **, *** represent significance at the confidence levels of 90%, 95%, and 99%, respectively. Control variables for each equation have been accounted for.

. First, the regional RSDO significantly impacts the export market and product diversification of wood products and furniture manufacturing enterprises. The RSDO has no significant impact on the export diversification of paper and paper product manufacturing; it even has a significant negative impact on the export market diversification. In addition, it has no significant impact on the export diversification of large-scale enterprises but has a significant positive impact on small-scale enterprises. Similarly, paper and paper products manufacturing and large-scale enterprises in their counties are major taxpayers. These enterprises are more likely to receive government ‘privileges’ in approvals, public services, and market regulation, which results in a minimal impact from the RSDO on their export diversification. However, for small-scale wood products and furniture manufacturing enterprises that do not receive government ‘preferential treatment,’ the RSDO helps provide a fair market position and competitive standing. This reduces operating costs and supports the marginal expansion of their products and markets.

Second, the regional RSDO has no significant impact on the export diversification of foreign-invested enterprises but has a significant positive effect on domestic enterprises’ export product and market diversification. This result is because foreign-invested enterprises benefit from numerous “government privileges” due to government investment promotion policies. For example, the average comprehensive tax rate (1.05%) and financial expense ratio (3.23%) of foreign-invested enterprises are lower than those of domestic enterprises (2.20%, 10.25%). They are less affected by regional administrative efficiency and market regulation. Furthermore, foreign-invested enterprises view mainland China as a “production base, supplying the global market”, with more stable exports and are less influenced by regional administrative efficiency and market regulation. Domestic enterprises are crucial in achieving China’s strategic goals of export diversification. The RSDO can help them improve their export diversification levels, again demonstrating the practical importance of advancing the RSDO for China’s export diversification strategy.

The regional RSDO has no significant impact on the export diversification of enterprises with strong asset management capabilities and even has a significant negative impact on the export market diversification index. However, it has a significant positive effect on the export product and market diversification of enterprises with weaker asset management capabilities. The main reason is that enterprises with stronger asset management capabilities have advantages in technological innovation, product innovation, management innovation, and business model innovation, as well as higher asset utilization efficiency, all of which help these enterprises outperform those with weaker asset management capabilities in areas such as export product structure expansion and optimization and international market development.

5.4. Indirect Effects Test

By examining the significance and direction of the interaction terms between the RSDO and technological innovation and operating costs, the indirect effects of regional RSDO can be assessed.

First, Models 13 to 18 in

Table 6. Examination of the indirect effects of the RSDO—export product diversification.

	Number of Export Products			Export Product Diversification Index
	Model 13	Model 14	Model 15	Model 16	Model 17	Model 18
RSDO	0.287 **	0.346 *	0.251 *	0.017 **	0.082 *	0.016 **
	(0.133)	(0.182)	(0.129)	(0.007)	(0.045)	(0.007)
RSDO × technological innovation	0.028			0.001
	(0.034)			(0.001)
RSDO × Operating costs		0.003 ***			0.001 *
		(0.001)			(0.001)
Control variable	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled

Note: *, **, *** represent significance at the confidence levels of 90%, 95%, and 99%, respectively, with standard errors in parentheses.

show that at the 90% confidence level, the RSDO’s coefficients are significantly positive, indicating that the RSDO has a significant direct positive effect. Models 14 and 17 show that the interaction term between the RSDO and operating costs is significant at the 90% confidence level. Similarly, as operating costs significantly negatively impact the export product type count and export product diversification index of sample enterprises (as seen in Table 2), improvements in regional RSDO help WPEs reduce operating costs. This lowers the cost ‘threshold’ for new products to enter export markets or gain a price advantage in international markets. However, the interaction terms between the RSDO and technological innovation did not reach a significant level, indicating that the RSDO did not promote WPEs’ export product diversification strategy through technological innovation.

Second, Models 19-24 in

Table 7. Examination of the indirect effects of the RSDO—export market diversification.

	Number of Export Markets			Export Market Diversification Index
	Model 19	Model 20	Model 21	Model 22	Model 23	Model 24
RSDO	0.271 **	−0.379	0.179	0.026	−0.140	0.008
	(0.137)	(0.731)	(0.122)	(0.022)	(0.150)	(0.020)
RSDO × technological innovation	0.063 *			0.013 *
	(0.035)			(0.008)
RSDO × Operating costs		−0.006			−0.002
		(0.008)			(0.002)
Control variable	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled

Note: *, ** represent significance at the confidence levels of 90% and 95%, respectively, with standard errors in parentheses.

show that, after adding the interaction terms, the impact coefficients of the RSDO variable are mostly insignificant, indicating that the direct effects of regional RSDO are not apparent. From Models 19 and 22, it can be seen that the interaction term between the RSDO and technological innovation has coefficients of 0.063 and 0.013 for the number of export markets and the export market diversification index, which are significant at the 90% confidence level. The premise that technological innovation has a significant positive impact (as seen in Table 2) indicates that regional RSDO will promote WPEs’ technological innovation, thereby enhancing their export market diversification. However, the interaction term between the RSDO and operating costs was insignificant, indicating that the RSDO did not promote WPEs’ export market diversification by lowering operating costs.

6. Research Findings, Discussion, and Policy Implications

6.1. Conclusions

This study is based on cross-sectional data from 2143 wood-processing enterprises in 2014. It introduces zero-truncated negative binomial regression and Tobit models to assess the effects of regional RSDO on enterprise export market and product diversification. The results show that regional RSDO has a significant positive effect on export product diversification but no significant effect on export market diversification. Second, regarding the sub-indicators of the RSDO, the reforms of “Streamline Administration and Delegate Power” and “Improve Regulation” have significant positive effects on export product and market diversification. In contrast, the optimization of government services shows no significant effect. Specifically, we hypothesize that the export market diversification of wood-processing enterprises may be more constrained by external factors (e.g., international demand fluctuations, trade barriers, and industry-specific restrictions) than by regional administrative reforms. In this sector, many enterprises, especially those involved in processing trade, may have limited control over their export market choices due to contractual obligations with foreign parent companies or the nature of their export agreements. Additionally, we argue that the impact of RSDO on export market diversification may take longer to manifest, as firms often require a period of time to build relationships and establish a market presence in new regions. The non-significant effect could be an indication that market diversification in this industry is not immediately influenced by administrative changes alone, and that other market-driven and industry-specific factors are more dominant in the short term. Third, the RSDO has a significantly stronger promoting effect on the export diversification level of wood products and furniture manufacturing enterprises, small-scale enterprises, domestic enterprises, and enterprises with weaker asset management capabilities. Fourth, the mechanism tests show that regional RSDO improves WPEs’ export product diversification by reducing operating costs and promoting export market diversification through technological innovation.

6.2. Policy Implications

First, local governments should explore the potential space for the RSDO. Following elaboration and requirements for “creating a market-oriented, law-based, and internationalized business environment”, governments at all levels should focus on the “bottlenecks” and “pain points” in the reform and focus on removing barriers that restrict the vitality of market entities. Governments should help enterprises reduce fixed and variable export costs, boosting their motivation and confidence to develop new markets and products. It has been observed that enterprises in Jiangsu Province give the highest overall evaluation of the RSDO in their prefecture-level cities, and further investigation and analysis of Jiangsu Province’s reform experience should be strengthened, so as to summarize the practices suitable for nationwide promotion. By reducing enterprises’ export-related costs, this reduction stimulates their incentive to pursue new markets and innovate products.

Second, more attention should be given to consulting domestic enterprises’ demands during the RSDO process. Governments should reduce the “policy advantage gap” between domestic enterprises and foreign-invested enterprises, such as revoking the policy of super-national treatment for foreign-invested enterprises, which can create a fairer competitive environment for domestic enterprises. Governments should provide a favorable institutional environment for domestic enterprises to explore new international markets and develop products that meet international market requirements. Similarly, more attention should be given to the needs of enterprises with strong operational capabilities. This can help peer enterprises enhance their export diversification capabilities through spillover effects.

Third, the revised “Forests Law of the People’s Republic of China” took effect on 1 July 2020. It abolished some systems, such as the wood transportation permit, inspection code, and wood production plan, and simplified the difficulty of wood-harvesting approval to some extent. However, the difficulty in approving industrial land for wood-processing enterprises still exists. Based on the “Notice on Deepening the Forest Harvesting ‘Streamline Administration and Delegate Power, Improve Regulation, and Optimize Services”, further reform is needed to deepen reforms and tackle the “tough issues”. Governments should reduce wood production and harvesting costs and ensure the security of China’s wood raw material supply. Institutional barriers should be eliminated for forest–paper integration enterprises, forest–board integration enterprises, and wood-processing enterprises with forest land. RSDO has significantly enhanced the diversification of export markets for enterprises through technological innovation. Local governments are advised to bolster enterprise technological innovation via policy support, including R&D subsidies and incentives for technology introduction. Additionally, fostering cooperation among enterprises, universities, and research institutions to establish an integrated industry–academia–research system can further strengthen the international competitiveness of enterprises.

This study is subject to limitations regarding data timeliness; the 2014 data utilized may not adequately capture the recent effectiveness of the RSDO reform. Future research is advised to incorporate more recent data to assess the long-term impact. Additionally, this study’s focus on the wood products industry restricts the generalizability of the findings. Future research should extend to other manufacturing and service industries to validate the broader applicability of RSDO. The absence of detailed enterprise-level cost and technological innovation data may also hinder in-depth mechanism analysis. Therefore, future studies should leverage more granular enterprise survey data to examine the impact of RSDO on corporate behavior. Potential future research directions include multi-industry comparative analysis, long-term effect analysis, policy interaction effect research, and international comparative studies. These approaches will enhance the understanding of RSDO’s influence on export diversification and offer more targeted policy recommendations.