Evaluation of operational efficiency in China's pharmaceutical industry and analysis of environmental impacts

Sun, Jiaqiang; Rosli, Anita Binti; Daud, Adrian; Yan, Xia Article Evaluation of operational efficiency in China's pharmaceutical industry and analysis of environmental impacts Economies Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Sun, Jiaqiang; Rosli, Anita Binti; Daud, Adrian; Yan, Xia (2025) : Evaluation of operational efficiency in China's pharmaceutical industry and analysis of environmental impacts, Economies, ISSN 2227-7099, MDPI, Basel, Vol. 13, Iss. 4, pp. 1-35, https://doi.org/10.3390/economies13040090 This Version is available at: https://hdl.handle.net/10419/329370 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Academic Editor: Angeliki N. Menegaki Received: 10 February 2025 Revised: 17 March 2025 Accepted: 20 March 2025 Published: 27 March 2025 Citation: Sun, J., Rosli, A. B., Daud, A., & Yan, X. (2025). Evaluation of Operational Efficiency in China’s Pharmaceutical Industry and Analysis of Environmental Impacts. Economies, 13(4), 90. https://doi.org/10.3390/ economies13040090 Copyright: © 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/ licenses/by/4.0/). Article Evaluation of Operational Efficiency in China’s Pharmaceutical Industry and Analysis of Environmental Impacts Jiaqiang Sun 1,2 , Anita Binti Rosli 1,* , Adrian Daud 1,3 and Xia Yan 1 1Department of Social Science & Management, Faculty of Humanities, Management & Science, Universiti Putra Malaysia, Bintulu Campus, Nyabau Road, Bintulu 97008, Sarawak, Malaysia; [email protected] (J.S.); [email protected] (A.D.); [email protected] (X.Y.) 2Strategic Research Center of Chengdu Maidison Pharmaceutical Technology Co., Ltd., No. 733, East Section, Hubin Road, Xinglong Subdistrict, Tianfu New Area, Chengdu 610000, China 3Institute of Ecosystem Science Borneo, Universiti Putra Malaysia, Bintulu Sarawak Campus, Nyabau Road, Bintulu 97008, Sarawak, Malaysia *Correspondence: anitar[email protected] Abstract: The pharmaceutical industry is a cornerstone of national economies and plays a critical role in public health. However, China’s pharmaceutical industry faces significant challenges, including regional disparities in development. The existing research on operational efficiency evaluation primarily focuses on financial or innovation metrics, lacking a comprehensive approach. Moreover, studies on the environmental impact on operational efficiency often rely on a limited set of indicators, failing to offer a holistic understanding of how environmental factors influence efficiency. This study aims to address these gaps by comprehensively evaluating operational efficiency and analyzing the impact of broader environmental factors on efficiency. To achieve these objectives, the study employs a ThreeStage Data Envelopment Analysis method combined with Principal Component Analysis to evaluate the operational efficiency of the pharmaceutical industry across 31 provinces in China, considering both financial and innovation dimensions.The findings reveal that overall efficiency has improved annually, with regional disparities gradually narrowing. Specifically, innovation capability and innovation environment have a positive impact on operational efficiency, while living standards and openness exhibit a negative correlation. Additionally, the current environmental conditions in the northwestern region are found to be conducive to the development of the pharmaceutical industry. This study is the first to integrate three-stage data envelopment analysis with principal component analysis, constructing a comprehensive framework for analyzing the relationship between environmental factors and operational efficiency. The results provide empirical evidence for policymakers aiming to enhance the efficiency of the pharmaceutical industry. Keywords: efficiency research; environmental impact; efficiency distribution; regional disparities 1. Introduction Since the Policy of China’s Reform and Opening-Up policy, the pharmaceutical industry (PI) in China has experienced rapid development. Particularly since 2000, the total output value of the pharmaceutical industry has increased by 13 times, with the compound annual growth rate of 12.85% (National Bureau of Statistics of China,2023). Currently, China has become one of the significant components of the supply chain in the pharmaceutical industry. However, despite the remarkable achievements in recent years, Economies 2025,13, 90 https://doi.org/10.3390/economies13040090 Economies 2025,13, 90 2 of 35 the pharmaceutical industry of China still faces severe challenges in terms of innovation, scale efficiency, low profit, and sustainable development (Bhardwaj,2024;Borja Reis & Pinto,2022). Firstly, China’s PI started late and invested less in research and development (R&D), resulting in a lower share of the international pharmaceutical market and the international competitiveness of the pharmaceutical industry (H. Guo & Shi,2021;Jia,2022;G. Wang & Zhang,2023). In addition, China’s regional economic development is not balanced. The efficiency of the PI in various regions varies significantly; whether in the innovation capacity or the overall level of operation, there is an uneven phenomenon. At the same time, because pharmaceutical companies, as part of a highly polluting industry, along with the overall development of China’s economy, some developed regions (such as Beijing, Shanghai, etc.) are gradually shifting from the production of raw materials and chemical drugs to the development of biopharmaceuticals and other fields. Many pharmaceutical companies are, therefore, relocating to more environmentally favorable regions in search of further development (Bai et al.,2022;Dou & Han,2019;Lai et al.,2020;T. Liu et al., 2020). It is important to note that the daily operations of an industry or enterprise are often constrained by regional political, economic, cultural, and technological conditions. Therefore, when selecting a development area, industries typically prioritize regions with comparative advantages. At the same time, the formation of industrial clusters has a significant impact on the operational efficiency of local industries. This regional selection and environmental adaptation not only determine the competitiveness of enterprises but also profoundly influence the overall development pattern of the industry (P. Chen et al., 2025;Imran et al.,2024a,2017;W. Wang et al.,2024;M. Zhang et al.,2024). To address these challenges mentioned above, a series of policies have been proposed at the national level to promote the high-quality development of the pharmaceutical industry. For instance, the 14th Five-Year Plan for the Development of the Pharmaceutical Industry emphasizes the need to enhance resource utilization efficiency, reduce pollution emissions, and achieve a green and low-carbon transformation. Simultaneously, local governments are exploring differentiated strategies aligned with regional development objectives, aiming to strengthen policy support and environmental regulation to foster the growth of the pharmaceutical industry. However, the implementation of these policies faces several challenges. Notably, there is a lack of systematic quantitative research on the specific impact of environmental constraints on industrial efficiency. Additionally, policy formulation often lacks a sufficiently targeted and evidence-based foundation. Addressing how to comprehensively improve the efficiency of the regional pharmaceutical industry under policy guidance has become a critical and urgent issue requiring resolution (Y. Liu et al.,2022;Xu et al.,2022). Despite the importance of these endeavors, there are still significant research gaps on how to assess the effectiveness of the implementation of these policies, and in particular, how these policies specifically affect the development of the pharmaceutical industry. Most of the current evaluations of firm efficiency focus on a single dimension such as finance or innovation, neglecting the fact that the pharmaceutical industry is a combination of both dimensions. In addition, the existing studies mainly limit the environmental impact to a single indicator, adopting one indicator to represent a certain evaluation dimension and lacking a systematic study of the overall environmental impact on efficiency. To fill the current research gaps described above, this study evaluated PI efficiency from a comprehensive perspective and reveals the influence of comprehensive environmental factors on efficiency. To achieve the research objectives, this study focused on the pharmaceutical industry across 31 provincial-level administrative units in China. It employed a three-stage data envelopment Economies 2025,13, 90 3 of 35 analysis (DEA) method combined with principal component analysis (PCA) to comprehensively evaluate the operational efficiency of China’s pharmaceutical industry in both both financial and innovation dimensions. PCA is used to extract key components from a large number of environmental factors, which are then used as environmental variables to reveal the impact of the comprehensive operational environment on the efficiency of the pharmaceutical industry. The innovations of this study are reflected in the following aspects: First, the existing research on the impact of environmental factors on efficiency typically relies on single indicators to substitute for specific dimensions in regression analysis. In contrast, this study employs PCA to extract principal components from a large number of environmental indicators, thereby enhancing the scientific rigor and feasibility of the conclusions. Second, the integration of PCA effectively eliminates multicollinearity issues, improving the robustness and explanatory power of the regression model. Finally, in the evaluation of pharmaceutical-industry operational efficiency, previous studies have predominantly focused on a single dimension, such as financial or innovation indicators. This study adopted a multidimensional, comprehensive evaluation approach, providing a more holistic and precise assessment of pharmaceutical-industry efficiency. This study has theoretical and practical significance. By integrating multiple evaluation dimensions and employing the combination of the three-stage DEA and PCA, this research provides a more comprehensive assessment of the characteristics of the pharmaceutical industry and its influencing factors, addressing the limitations of traditional single-dimensional analysis. Furthermore, the study conducted an in-depth analysis of the drivers of efficiency changes, revealing the effects of factors such as pure technical efficiency and economies of scale on performance and thereby offering a refined theoretical framework for efficiency analysis. From an economic perspective, the research further investigates regional influencing factors, exploring how economic elements such as the local economy and local innovation influence the efficiency of the PI. These analyses not only enhance the understanding of efficiency dynamics in the pharmaceutical sector but also provide a theoretical foundation for formulating targeted industry policies and optimizing resource allocation, underscoring the study’s academic and economic significance. This study is organized into six sections. Section 1outlines the research background, identifies the research problem, and defines the study objectives. Section 2conducts a comprehensive literature review, summarizing relevant theoretical frameworks and efficiency measurement methodologies, justifying the adoption of the three-stage approach, and highlighting gaps in the existing literature and the conceptual framework of this study. Section 3details the research methodology, encompassing data sources, theoretical models, and the selection of variables. Section 4presents the empirical results, beginning with data validation, followed by the first-stage result, principal component extraction and regression analysis (second stage), and DEA-based efficiency measurement (third stage). Section 5discusses the distribution of efficiency, its temporal evolution over a decade, regional efficiency disparities, the influence of environmental factors on efficiency across regions, and a comparative analysis between the findings of this study and prior research. Finally, Section 6concludes the study by summarizing the main findings, the implication for policy-makers, and the limitations of this study, proposing directions for future research. 2. Literature Review 2.1. The Basic Theory and General Measurement Methods of Efficiency Value maximization theory: This theory posits that an industry or enterprises should balance financial performance with the consideration of various stakeholders’ interests and the equilibrium between long-term and short-term goals. Specifically, a comprehensive evaluation of the industry or companies should integrate both financial outcomes and inno- Economies 2025,13, 90 4 of 35 vation, reflecting the balance of short-term and long-term interests. Therefore, this theory emphasizes the necessity of evaluating the industry using the financial and innovation dimensions (Colm,1960;Friedman,2007;Grossman & Stiglitz,1977). Regional cluster economic theory: Michael Porter’s regional cluster economic theory suggests that the competitiveness of an industry is influenced by multiple factors such as technology, government policies, and local resources. Although environmental regulations may pose challenges, they can foster technological innovation, enhancing productivity and competitiveness. Furthermore, regional clusters facilitate cooperation, knowledge sharing, and technological advancement among firms, thereby strengthening the competitive advantage of enterprises and promoting overall regional economic growth. Therefore, the study of the impact of regional environments on industry efficiency is of significant theoretical and practical importance (Martin & Sunley,2003;Porter,1990b,1998). Resource-based view (RBV) theory: The resource-based view (RBV) theory asserts that the competitive advantage of firms and industries stems from their unique resources and capabilities, particularly those that possess value, rarity, inimitability, and non-substitutability (VRIN). Effective resource allocation, especially the integration and dynamic adjustment of resources, is crucial for maintaining competitive advantage. This theory provides a theoretical foundation for analyzing local industry resources and their impact on competitive advantage (J. Barney,1991;J. B. Barney & Arikan,2005;Ge et al.,2024;Kaur & Singh,2024; Teece et al.,1997;M. Zhang et al.,2024). Efficiency, as a concept, was first articulated by the Italian economist Pareto, who framed it as an optimal state of resource allocation, commonly called “Pareto optimality”. In this state, no reallocation of resources can increase the benefit of one party without reducing the benefit of others. Pareto optimality fundamentally reveals the economic meaning of efficiency, which is to achieve optimal economic output through the rational distribution of resources. Building on this theory, Farrell (1957) further developed the theory of efficiency by decomposing it into technical efficiency and allocative efficiency. Farrell’s work laid a solid foundation for modern efficiency evaluation theory and provided a theoretical basis for the development of efficiency measurement methods. Efficiency measurement methods are mainly divided into two categories: parametric and non-parametric methods. Among these, stochastic frontier analysis (SFA), as a representative of parametric methods, is commonly used, while data envelopment analysis (DEA) is a typical non-parametric method (Greene,2008;Lampe & Hilgers,2015;Resti,2000). 2.2. Efficiency Measurement Methods: A Comparison of SFA and DEA SFA, a classical parametric method, was proposed by Aigner et al. (1977). The basic principle of SFA is to decompose the efficiency of decision-making units (DMUs) into two parts: random error and an inefficiency term. The random error reflects the impact of uncontrollable factors such as the external environment, such as weather, policy changes, or market fluctuations; and the inefficiency term reflects the loss of efficiency caused by the internal management or technical level of the firm (Aigner et al.,1977;Li & Fan,2009). SFA evaluates the technical efficiency and the total efficiency by constructing a production frontier function combined with maximum likelihood estimation techniques. The advantage of this method is that it can distinguish between the impact of external uncontrollable factors and internal management factors on efficiency, thus making the assessment results closer to the actual situation. Due to its wide applicability, SFA has been widely used in many fields such as agriculture, industry and environmental research (Kumbhakar & Tsionas,2011;Lee & Jeon, 2023;Ðoki´c et al.,2022;Singh et al.,2020;J. Wang et al.,2020;R. Wang & Duan,2023). However, as a parametric method, SFA requires assumptions about the functional form of the production function and the distribution of inefficiency and random error. If Economies 2025,13, 90 5 of 35 the assumptions deviate from the actual situation, it may lead to biased efficiency estimates. Moreover, the application of SFA involves certain limitations when dealing with inputs and outputs that have different units and dimensions (Ahmed & Melesse,2018;Kalirajan & Shand,1994;Madaleno & Moutinho,2023;Moulay Ali et al.,2024). DEA is a method that constructs an best-practice frontier based on the DMUs and then compares other DMUs with this to determine their relative efficiency. This method offers several advantages in efficiency evaluation. First, it does not require predefined assumptions about the functional form of the production function, thus avoiding the model bias that may result from incorrect function specification (Charnes et al.,1978). Second, DEA can handle multi-input, multi-output efficiency issues simultaneously without the need to weight these variables, which gives it an advantage when dealing with multidimensional data (Banker et al.,1984). Moreover, DEA can identify sources of inefficiency, such as scale inefficiency or insufficient technical efficiency, providing decision-makers with directions for improvement (Cooper et al.,2007). In terms of sample size requirements, DEA is well suited to small sample data, making it particularly useful for industry studies or case analyses (Zhu,2009). Additionally, DEA can provide improvement paths for non-DEA efficient DMUs by analyzing reference sets and clarifying the directions for adjusting inputs or outputs (Tone,2001). Finally, the flexibility of DEA models is considerable, with the ability to incorporate various extensions, such as three-stage DEA, weight-restricted models, and indirect economic efficiency models (Camanho et al.,2024;Mergoni et al., 2024). With the development of DEA research, it has found wide application in efficiency testing across various industries (André et al.,2024;C. Guo et al.,2024;Y. Guo et al.,2024; Rashid et al.,2024;Sun et al.,2024). However, DEA also faces certain limitations. First, it cannot separate the influence of external environmental factors and random noise on efficiency, which may lead to efficiency estimates being affected by external conditions or data fluctuations. Second, DEA is unable to further analyze the specific sources of inefficiency for non-DEA efficient DMUs, such as whether it is due to management issues, environmental constraints, or random factors (Coelli et al.,2005;Dyson et al.,2001;Hjalmarsson et al.,1996;Joe & Wu,1996;Ma,2010). 2.3. The Method of Three-Stage DEA and Advantage To address the limitations of traditional data envelopment analysis (DEA) in handling efficiency measurement errors caused by environmental factors and the constraints of stochastic frontier analysis (SFA), which requires assumptions about error distribution and struggles with multiple inputs and outputs, Fried et al. (2002) proposed the threestage DEA method. This approach integrates DEA and SFA to improve the accuracy of efficiency measurement. In the first stage, a conventional DEA model is used to compute initial efficiency scores and identify slack variables. In the second stage, SFA is applied to regress the slack variables against environmental factors, allowing for the separation of environmental influences and stochastic disturbances on efficiency. Finally, in the third stage, after eliminating the effects of environmental and random factors, the adjusted input data are used in a final DEA calculation to obtain a more precise measure of technical efficiency (Charnes et al.,1978;Fried et al.,2002;Luo,2012). Compared to traditional DEA methods, the three-stage DEA approach offers significant advantages. It retains the flexibility of DEA, such as not requiring a predefined production function, while also isolating the effects of external environmental factors and stochastic noise, thereby enhancing the accuracy of efficiency evaluation (Avkiran & Rowlands,2008;Fried et al.,2002;Na et al.,2019;Z. Wang et al.,2017). Consequently, this method has been widely applied in various fields, including energy management, environmental governance, public services, the pharmaceutical industry, and innovation Economies 2025,13, 90 6 of 35 efficiency assessment, providing a more reliable tool for efficiency measurement (G. Chen & Chen,2024;Guanglan & Zhening,2024;Shi et al.,2025;Song & Ma,2024;Z. Wang et al., 2024;Wei & Zhao,2024;L. Zhang & Cui,2024). 2.4. The Current Research on Efficiency in the PI Currently, there is limited research on the efficiency of the pharmaceutical industry (PI), and existing studies tend to focus on a single dimension, such as innovation efficiency or financial efficiency. 2.4.1. Innovation Efficiency Research Innovation is one of the key characteristics of the pharmaceutical industry. However, research on innovation efficiency is still very limited. In studies of listed pharmaceutical companies or the pharmaceutical industry, Xiong and Meng (2019) used DEA analysis to find that, in China’s PI, the efficiency of biopharmaceuticals is the highest, with the main inefficiency stemming from pure technical inefficiency. The efficiency of translating innovation into revenue is also low, primarily due to excessive research and development investment. Moreover, in China, innovation efficiency is not only low but also unevenly distributed (Hao & Ruan,2022;Lai et al.,2020;Qiu et al.,2023). International research has also been conducted on pharmaceutical innovation. SFA and multiple-frontier analysis were applied to evaluate the 705 pharmaceutical companies’ efficiency in America, revealing that different open innovation approaches had distinct impacts on performance (Shin et al.,2018). In terms of the efficiency of large global enterprises, it was found that even globally renowned companies face efficiency challenges (Gascón et al.,2017;Schuhmacher et al.,2023,2021). 2.4.2. Pharmaceutical Companies’ Financial Efficiency Evaluation Financial efficiency evaluation is a common method for assessing companies. In 2013, a method combining data envelopment analysis (DEA) and stochastic frontier analysis (SFA) was applied, and the results revealed that increased technical knowledge reserves significantly improved company revenue, with similar conclusions drawn using different DEA models (Cai & Sun,2013). Xia et al. (2022) applied the BCC-DEA to evaluate the operational efficiency of public pharmaceutical firms in China. Their analysis indicated a general decline in overall financial efficiency, with the exception of the biopharmaceutical sector, which exhibited growth. In contrast, the chemical pharmaceuticals and traditional Chinese medicine sectors experienced a reduction in efficiency. Similarly, Lin et al. (2021) utilized a two-stage network DEA approach combined with Malmquist indices to assess the impact of government subsidies. Their findings suggested that such subsidies had no significant effect on financial efficiency. Further research by Yang (2024) applied a threestage DEA model and Malmquist indices, revealing that the overall efficiency of Chinese pharmaceutical companies remains relatively low, with notable year-to-year variability. Regarding the selection of indicators, domestic researchers typically rely on operating income and profits as output metrics due to the accessibility of these data. In contrast, international scholars often incorporate a wider array of indicators, including human capital efficiency, structural capital efficiency, intellectual capital efficiency, earnings per share, dividends per share, and returns on equity (Hamad & Tarnoczi,2021;Riaz et al.,2023). 2.5. Research Gap Up to now, the research on the efficiency of pharmaceutical companies and the pharmaceutical industry has been limited both in terms of the quantity of studies and the depth of their integration. Economies 2025,13, 90 7 of 35 (1) Current evaluations of the operational efficiency of the PI are based solely on financial or innovation dimensions, lacking comprehensive research. In the innovation dimension, scholars typically use indicators such as sales of new products and the number of patent applications (Hao & Ruan,2022;Lai et al.,2020; Qiu et al.,2023). Other scholars also evaluate the operational efficiency use the new molecular entity (NME) and impact factors of the publication (Gascón et al.,2017; Schuhmacher et al.,2023,2021). The financial evaluation dimension includes a broader range of indicators, with operation revenue and operation profit being the most widely used (Gascón et al.,2017;Lin et al.,2021;Yang,2024). Other financial indicators, such as asset turnover, returns on equity (ROE), and earnings per share (EPS), are also employed (Hamad & Tarnoczi,2021;Riaz et al.,2023;Xia et al.,2022). To date, the only study that has combined both financial and innovation dimensions in evaluating the operational efficiency of pharmaceutical companies is Gascón et al. (2017). No other research has conducted a combination evaluation of both dimensions. However, the financial indicators primarily reflect a company’s short-term profitability; innovation serves as a measure of its long-term growth potential. Therefore, a separate evaluation of the financial and innovation performance overlooks the balance between short-term and long-term interest. (2) The current research on the impact of environmental factors on efficiency is limited. Although previous studies have employed three-stage DEA to analyze the impact of the environment on efficiency, the use of a single indicator to represent a dimension lacks a comprehensive understanding of the environment. Qiu et al. (2023) used the number of employees to represent company size, the number of employees with a bachelor’s degree or higher to represent employee quality, and returns on equity (ROE) and the ratio of total liabilities to total assets (LEV) as environmental indicators. Yang (2024) used government subsidies, per capita GDP, and the years that a company has been established to measure and explain the environmental influence on efficiency. Sun et al. (2024), construct environmental indicators using per capita disposable income to represent wealth levels, the working-age population to represent the labor supply, and local GDP to represent the local economic level. Although these studies explore the impact of environmental factors on efficiency from different perspectives, they remain inadequate. Environmental factors are complex, and using a single indicator for regression analysis cannot fully capture the overall environmental impact. Moreover, when too many environmental indicators are included in regression, potential multicollinearity issues may arise, which could compromise the accuracy of regression results (Baird & Bieber,2016;Haitovsky, 1969;Shrestha,2020). In conclusion, there are notable research gaps in the current research regarding the comprehensive efficiency evaluation of the PI and the in-depth exploration of how the environmental factors impact efficiency. Addressing these gaps will not only contribute to the theoretical framework but also provide a more scientific implication for policy-makers and companies in the PI, improving their operation efficiency. 2.6. The Conceptual Framework of This Study Based on the research gap, the study had two primary objectives: first, to accurately measure the operational efficiency of the PI in both the financial and innovation dimensions and, second, to reveal the impact of comprehensive environmental factors on operational efficiency. To achieve these research objectives, the study employed a combined method of three-stage DEA and PCA. The process can be divided into three stages. (1) Stage 1: initial efficiency measurement. Economies 2025,13, 90 8 of 35 In this stage, DEA was used to calculate the efficiency values of 31 provincial-level regions in China. The main objective was to measure the current efficiency distribution and calculate the sales of input variables. (2) Stage 2: environmental factor analysis and adjustment. At the beginning of this stage, PCA was employed to extract principle components from multiple potential environmental variables. The environmental variable values were calculated based on the variance contribution of each component. The extracted environmental variables were used as independent variables (IV), and the slacks of the inputs from Stage 1 were used as the dependent variable (DV) in the SFA regression. This stage aimed to reveal how the comprehensive environment influences efficiency and then eliminate the impact of environmental factors and random disturbances on the input variables by adjusting the input indicators. (3) Stage 3: efficiency adjustment and recalculation. Adjusted inputs were obtained in Stage 2, and the original output was employed in this stage to recalculate the final and accurate efficiency using the DEA model. This stage aimed to measure the accurate efficiency of PI in China while removing the effects of environmental factors. The conceptual framework of this study was primarily based on the theoretical frameworks of Zhao et al. (2019) and L. Zhang and Cui (2024), which form the analytical model for this research, as well as the PCA-DEA approach proposed by Stevi´c et al. (2022). Figure 1 illustrates the conceptual framework of this study. Figure 1. The conceptual framework of this study. Economies 2025,13, 90 15 of 35 encompass aspects such as economic foundations, consumption levels, foreign investment, local fiscal revenue, investment in technology and education, and the composition of the labor force. (1) Step 1: Method suitability verification. The suitability of the data for principal component analysis (PCA) was verified using the Kaiser–Meyer–Olkin (KMO) measure, and cumulative variance was explained. When the KMO value is greater than 0.6, it indicates strong correlations between the variables, making the data suitable for PCA. If the KMO value is below 0.6, it suggests weak correlations between the variables, making the data unsuitable for PCA. Additionally, when the cumulative variance explained is 70% or higher, it indicates that the extracted principal components effectively explain the majority of the variation in the original data, making the data suitable for further analysis. Conversely, if the cumulative variance explained is below 70%, it suggests that the extracted components do not adequately capture the main information in the data (Hair et al.,2010). Table 4 shows the KMO measure and Bartlett’s test results, while Table 5presents the analysis of cumulative variance before and after rotation. As shown in Table 4, the KMO value is 0.902, which exceeds the acceptable threshold of 0.60, indicating that these environmental factors are well suited for the principal PCA method. Furthermore, Bartlett’s test of sphericity results ( χ2=15737, d f =253, p<0.001 ) demonstrate significant correlations between the variables, further validating the appropriateness of conducting PCA. As shown in Table 5, after PCA was applied, only four main principle components were extracted from the 23 potential environmental factors. These principal components collectively already explain 89.35% of the total variance, indicating a very high level of explanatory power in the dataset. Table 4. The result of KMO and Bartlett’s test. KMO measure of sampling adequacy 0.902 Bartlett’s test of sphericity approximate Chi-Square 15,737 Degrees of freedom 253 Significance 0.000 Table 5. Eigenvalues and variance explained before and after rotation. Component Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Total % Variance Cum. % Total % of Variance Cum. % Total % of Variance Cum. % 1 15.18 66.00 66.00 15.18 66.00 66.00 11.82 51.37 51.37 2 2.95 12.83 78.83 2.95 12.83 78.83 4.19 18.21 69.58 3 1.40 6.07 84.90 1.40 6.07 84.90 2.37 10.29 79.87 4 1.02 4.43 89.33 1.02 4.43 89.33 2.18 9.46 89.33 5 0.58 2.56 90.81 - - - - - - ... - - - - - - - - - 23 0.03 0.02 100.00 - - - - - - (2) Step 2: Rotated component. In PCA, the initially extracted components may be complex, with variable loadings that are difficult to clearly distinguish, which hinders the interpretation of the results. Therefore, rotating the components helps simplify the structure of variable loadings, allowing each variable to concentrate on a few main components. This increases the interpretability of the principal components, clarifies their actual meaning, and enhances the explanatory power regarding the study subject. Table 6presents the results of the rotated component loading matrix. Economies 2025,13, 90 16 of 35 Based on the results of the rotated component matrix in Table 6, four principal components were extracted. To better interpret the environmental factors reflected by each principal component, only variables with loadings greater than 0.5 were retained. For the purpose of the subsequent efficiency analysis, the extracted factors were classified and named according to the characteristics of the principal components. Table 6. Rotated component loadings Component Principle Component 1 Principle Component 2 Principle Component 3 Principle Component 4 Water Pollution Equivalent - - 0.769 - Air Pollution Equivalent - - 0.807 - Full-time R&D Hours 0.927 - - - Annual Number of R&D Projects 0.916 - - - Annual R&D Investment 0.899 - - - New Product Projects 0.95 - - - New Product Investment 0.933 - - - Annual New Product Sales 0.922 - - - Authorized Inventions 0.76 0.516 - Authorized Utility Models 0.898 - - - Authorized Designs 0.885 - - - Per Capita Disposable Income - 0.905 - - Per Capita Consumption Level - 0.908 - - Number of Foreign-Funded Enterprises Registered 0.723 - - - Foreign Investment Amount - - - 0.895 Registered Capital of Foreign Investment - - 0.964 Higher Education Enrollment 0.727 - - - Local General Budget Revenue 0.777 - - - Government Support for Education 0.809 - - - Government Support for Science and Technology 0.809 - - - Government Support for Environmental Protection 0.563 - - - Regional GDP 0.832 - - - Per Capita GDP - 0.880 - - (3) Step 3: Name component. Principal Component 1: economic and technological foundation level (Z1). The first principal component includes variables such as regional economic levels, government fiscal revenue, and government investments in technology, education, innovation, and technological outputs. Therefore, this component reflects the region’s economic development and innovation capacity, and it is named “Economic and technological foundation level”. Principal Component 2: residents’ living standards (Z2). The second principal component comprises variables like per capita GDP, resident income, and consumption levels, which reflect the living quality and economic status of the region’s residents. Thus, this component is named “Residents’ Living Standards”. Principal Component 3: local pollution levels (Z3). The third principal component primarily consists of waste emissions and pollution levels in the natural environment. The higher the pollutant emissions, the higher the local pollution level, which is why this component is named “Local Pollution Level”. Principal Component 4: openness to the foreign market (Z4). The fourth principal component is mainly composed of variables related to foreign investment, reflecting the region’s level of openness and its ability to attract foreign capital. Hence, this component is named “Openness to the foreign market”. Economies 2025,13, 90 17 of 35 4.3.2. The Result of the SFA Regression The objective of this stage was to reveal how environmental variables affect efficiency. To achieve this objective, the principal components extracted from the PCA in the previous stage were used as the IV, while the input slacks obtained in the first stage were used as the DV, and SFA regression was then conducted using Frontier 4.1 to examine the impact of environmental variables on efficiency. The regression results are presented in Table 7. Table 7. Results of SFA on the impact of environmental variables on input slacks. Independent Variable Dependent Variable Slack of Total Asset Slack of Employees Number Slack of R&D Investment Constant Term β0−17973.69 *** −2638.33 *** −32.16 *** t-ratio −14.03 −23.19 −5.09 Economic and technological foundation (Z1) β1−6568.12 *** 218.76 4.51 * t-ratio −2.68 0.70 1.77 Residents’ living standards (Z2) β216,586.74 *** 2117.79 *** 40.48 *** t-ratio 25.28 5.75 10.56 Local pollution levels (Z3)β37163.00 *** 2942.94 ** −42.76 ** t-ratio 39.44 2.01 −2.41 Openness to the foreign market (Z4) β445,187.43*** −4467.38 *** 4.66 t-ratio 9.31 −4.63 0.10 σ23,560,829,700 57,626,093.00 33,054.70 γ0.999 0.999 0.999 LR test of the one-sided error 204.38 204.72 310.85 Note: *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively. Only coefficients with a significance level of 5% or lower ( p≤ 0.05) are included in subsequent calculations, while those above 5% (p>0.05) are excluded. As shown in Table 7, the likelihood ratio (LR) test values for the three input slack variables are 204.38, 204.72, and 310.85, respectively, all significantly exceeding the critical value of the mixed chi-square distribution at the 1% significance level (10.83). This confirms the suitability of the model and the validity of the assumptions. At the same time, the γ values are close to 1, indicating that the input slack is primarily caused by managerial inefficiency, rather than random noise. This suggests that there is significant potential for management optimization to improve operational efficiency. By optimizing management and adjusting strategies, input slack can be effectively reduced, further enhancing the operational efficiency of the pharmaceutical industry. From Table 7, it can be seen that external environmental factors have a complex impact on the operational efficiency of the pharmaceutical industry. For example, the local economic and technological innovation capabilities can reduce asset allocation through innovation, thereby reducing the enterprise’s input. On the other hand, the standard of living increases input from various aspects, indicating that, in regions with higher living standards, input in the pharmaceutical industry increases, which is not conducive to its development. Regions with a higher environmental capacity tend to increase investment in labor and assets, but they can reduce R&D investment. In contrast, regions with higher openness to the outside world increase investment in assets while reducing the number of employees, indicating that foreign investment upgrades asset allocation and reduces labor input. A detailed analysis of these impacts is presented in Section 5. Another research objective of this stage was to exclude the influence of environmental factors and random noise on efficiency to provide a more equitable environment and place Economies 2025,13, 90 18 of 35 all DMUs at the same level of randomness, ultimately calculating the adjusted input values. The specific calculation process was as follows. After Frontier was run, the environmental values were calculated using the coefficients in Table 7, and the principal component shares were obtained through PCA. Subsequently, Formulas (4) – (9) were applied to calculate managerial inefficiency, and random errors were estimated using Formula (10) . Finally, adjusted input values were obtained through Formula (11) . This series of calculations eliminated the interference of environmental impacts and random errors, ensuring that the efficiency values reflect only the managerial issues within the enterprises themselves. 4.4. The Result of Stage 3 In Stage 3, the BCC-DEA model was employed, with the adjusted inputs obtained in Stage 2 and initial outputs. All the results are presented in Table 2. As shown in Table 2, the overall average technical efficiency values in the first and third stages do not differ significantly, with the most notable difference observed in pure technical efficiency. Since this study uses cross-sectional data, the efficiency values obtained represent the relative efficiency between the DMUs. In the third stage, Tianjin remains the region with the highest pharmaceutical-industry efficiency, while Liaoning and Jiangxi have shown significant improvements. This suggests that the efficiency of these two regions was underestimated in the previous stage and indicates that the environmental conditions in these regions are not conducive to the development of the pharmaceutical industry. The regions with the lowest technical efficiency are the same as in the first stage, indicating that the development of the pharmaceutical industry in these regions remains suboptimal. An analysis of efficiency trends reveals notable advancements in areas like Heilongjiang, Beijing, and Sichuan. This implies that the initial assessments may have undervalued the efficiency levels in these regions. Moreover, it underscores the challenging environmental circumstances that are less conducive to the growth of the pharmaceutical sector in these locales. In contrast, the regions that experienced significant declines in efficiency include Tibet, Ningxia, Fujian, and Qinghai. Moreover, all provinces in the northwest have seen a decline in efficiency, suggesting that the external environment is currently more favorable for the development of the pharmaceutical industry in other regions. As shown in the table, the average efficiency has shown a consistent upward trend over the years, increasing from 0.642 in 2013 to 0.767 and indicating a continuous improvement in efficiency levels. Additionally, the probability distribution of efficiency for each year also shows some improvement. The Gaussian distribution graph based on the efficiency distribution across 31 provinces each year is shown in Figure 2. As shown in Figure 2, there was a noticeable change in the overall efficiency distribution. In 2013, the efficiency values were concentrated in the lower range, with a relatively broad distribution, indicating low efficiency and significant disparities. Over the years, the curve gradually shifted to the right, reflecting an increase in efficiency levels, and the distribution became more concentrated, showing higher efficiency values. By 2022, the efficiency level had significantly improved, with a steeper distribution, indicating reduced efficiency disparities and a more balanced overall efficiency. This change reflects the gradual improvement in the efficiency of the pharmaceutical industry and the narrowing of efficiency gaps across different regions. Meanwhile, the coefficient of variation (CV, calculated as average efficiency/SD) decreased from 0.352 to 0.191, indicating a reduction in the disparity of efficiency levels. The decline in the coefficient of variation suggests that the efficiency levels across different regions are becoming more balanced, which may reflect an increase in the fairness of resource allocation or faster efficiency improvements in underdeveloped regions. Economies 2025,13, 90 19 of 35 Figure 2. Gaussian kernel density distribution of TE. This section applies a three-stage DEA method combined with PCA to measure the operational efficiency of the pharmaceutical industry across 31 provinces in China in the holistic dimension and to reveal the impact of comprehensive external factors on operational efficiency. However, the causes of these differences are multifaceted, and the current section will provide a detailed discussion of the factors contributing to these disparities. 5. Discussion In the empirical analysis of the previous chapter, it was found that the overall efficiency of China’s pharmaceutical industry is gradually increasing, with the gap between regions narrowing. However, different regions exhibit distinct trends. Additionally, the study revealed the specific impact of environmental factors on efficiency. The following sections will further explore the underlying causes of these phenomena. 5.1. Discussion on the Efficiency Distribution of the Pharmaceutical Industry Across China’s Regions The results presented in the results section reveal that there are differences in efficiency across regions in both the first and third stages. To further explore the overall differences in the operational efficiency of the PI across regions and the underlying causes, this study decomposed TE into PTE and SE. By using the average values of TE and SE as thresholds, the efficiency of each region was classified into four quadrants: high PTE–high SE, high PTE–low SE, low PTE–high SE, and low PTE–low SE. Figure 3illustrates the relative positions of the regions based on their PTE and SE performance. (1) Efficiency analysis of pharmaceutical enterprises in first-tier regions. Pharmaceutical enterprises in first-tier regions, including Tianjin, Liaoning, and Jiangxi, exhibit high PTE and SE. These regions demonstrate strong resource allocation capabilities and significant economies of scale, leading to higher overall efficiency. While the economic strength of these regions is not the highest, they effectively leverage abundant human resources, low labor costs, and strong government support to optimize resource utilization. Through the synergistic effects of technology, Economies 2025,13, 90 20 of 35 resources, and policies, these regions have developed their unique high-efficiency models. Additionally, the well-established industrial support foundation in these regions promotes inter-industry coordination, which further enhances both technical and scale efficiencies. Figure 3. Average pte and se distribution of PI across Chinese provinces. (2) Efficiency analysis of pharmaceutical enterprises in second-tier regions. Pharmaceutical enterprises in second-tier regions also show high PTE but relatively low SE. These regions include Jiangsu, Shandong, Guangdong, and Ningxia. Except for Ningxia, the other provinces (Jiangsu, Shandong, Guangdong) are among the economically strongest in China. According to the SFA analysis, the high efficiency in Ningxia is primarily due to favorable environmental factors such as strong national support, a large environmental capacity, and the relatively low income and living standards of the local population, which together create a high level of resource allocation. However, enterprises in Jiangsu, Shandong, and Guangdong, despite their strong performance in technical innovation and resource allocation, efficiently utilize technology and management methods to achieve high technical efficiency. These regions’ enterprises possess advanced production technologies and robust R&D capabilities. However, despite their strengths in technological efficiency, they have struggled to achieve economies of scale during expansion, resulting in a persistent situation of decreasing returns to scale (DRS) over the past decade. This phenomenon can be attributed to factors such as market saturation, with enterprises finding it difficult to match increased production capacity with sufficient market demand, inefficient resource allocation, despite strong technological performance, with issues in the distribution of resources such as funds, talent, and equipment, and management bottlenecks, through which the original management system is unable to adapt to the complexity brought about via scale expansion, thereby limiting the improvement of scale efficiency. Economies 2025,13, 90 21 of 35 (3) Efficiency analysis of pharmaceutical enterprises in third-tier regions. Pharmaceutical enterprises in third-tier regions show high SE but relatively low PTE. These regions include Beijing, Shanghai, Chongqing, Hunan, Henan, and Yunnan. The potential causes for the low PTE in these regions vary. For instance, in Chongqing, Beijing, and Shanghai, although these regions possess strong economic and technological foundations, they are hindered by strict environmental and safety policies. High environmental protection requirements in places like Beijing and Shanghai have restricted investment in technological upgrades and innovation, thereby limiting improvements in technical efficiency. Additionally, high production factor costs, such as labor, land, and energy prices, have become significant obstacles to enhancing technological performance. Despite these constraints, enterprises in these regions can achieve economies of scale through market size advantages and industrial agglomeration effects, which allow them to lower unit costs through increased production. In contrast, enterprises in regions like Hunan and Yunnan, despite facing challenges in technical efficiency, have performed well in scale efficiency. These regions’ enterprises have relatively limited investments in technological innovation and R&D, which results in stagnation in technological progress and affects their production efficiency. However, due to lower production factor costs and government policy support, such as tax reductions and industrial subsidies, enterprises in these regions can achieve economies of scale. The local government’s policies further enhance the efficiency of enterprises when expanding production to meet the growing market demand, thereby boosting scale efficiency. (4) Efficiency analysis of pharmaceutical enterprises in fourth-tier regions. Pharmaceutical enterprises in fourth-tier regions, such as Gansu and Shanxi, generally exhibit low TE and PTE. These regions’ enterprises mainly rely on traditional technologies and relatively extensive management models, lacking advanced technical support and process optimization. As a result of insufficient technological innovation, the production efficiency in these regions is low. Moreover, the geographical locations and economic foundations of Gansu and Shanxi make them less attractive for highly skilled talent and technological capital, preventing enterprises from absorbing external advanced technologies and management experience. The severe shortage and outflow of local talent have further constrained technological improvements. Additionally, the weak economic foundation and limited government support for technological innovation and research restrict the ability of enterprises to invest in necessary upgrades, leading to a lack of improvement in technical and pure technical efficiency. 5.2. Discussion on the Changes in Regional Efficiency and Their Causes Over the past decade, the overall efficiency of the PI has shown an upward trend. However, it is important to note that, in some regions, efficiency has declined year by year. Therefore, this study divides the 31 regions of China into seven large areas and decomposes the TE of these areas into PTE and SE. PTE reflects the efficiency improvements achieved through technological innovation or resource allocation optimization in a region, excluding the impact of economies of scale, while SE measures whether a region can effectively achieve economies of scale as production scale increases. Figure 4a–c show the changes in pharmaceutical-industry efficiency in the seven large regions of China over the decade. Economies 2025,13, 90 22 of 35 (a) Trends in TE of the PI in different regions of China. (b) Trends in PTE of the PI in different regions of China. (c) Trends in SE of the PI in different regions of China. Figure 4. Decomposition of the operation efficiency in Stage 3. As shown in Figure 4a, the overall efficiency of China’s PI has steadily increased over the past decade, with most regions experiencing relatively stable changes and minimal fluctuations. This indicates that the overall development trend of the PI is stable, and efficiency is gradually becoming balanced. However, certain regions have shown persistent increases or decreases in efficiency. This suggests that, although the overall trend is positive, some areas still face different challenges and opportunities, necessitating a more in-depth analysis of the underlying causes. The operation efficiency of Central China (including Henan, Hubei, and Hunan) experienced a noticeable decline from 2013 to 2022. A decomposition of TE reveals that the primary reason for this decline is low PTE. The possible causes include inefficient resource allocation and insufficient innovation output in these regions, which have constrained the improvement of PTE. Specifically, the lack of innovation capacity in the central region (including Henan, Hubei, and Hunan) has resulted in poor performance in technological advancement and production efficiency optimization, thereby limiting the enhancement of PTE. These three provinces lag behind in technological innovation, potentially due to insufficient government emphasis on pharmaceutical research and development (R&D), Economies 2025,13, 90 23 of 35 limited investment in R&D, and a continued reliance on traditional production methods. The lack of introduction and application of new technologies has further constrained the improvement of production efficiency. Moreover, the pharmaceutical industry in the central region remains predominantly focused on manufacturing, leading to slow progress in technological upgrading and industrial transformation. The issue of brain drain is also prominent in these provinces, with a shortage of high-end technical and innovative talent further restricting the development of technological innovation. Inadequate policy support and financial investment from the government have weakened the motivation of enterprises to engage in R&D and technological transformation, making it difficult to establish an effective innovation ecosystem. Therefore, the central region faces multiple challenges in improving operational efficiency and technical efficiency. It is imperative to address these issues by increasing R&D investment, optimizing resource allocation, promoting industrial upgrading, and attracting high-end talent to improve the current situation. In contrast, the TE in Southern and Northwestern China has significantly improved from 2013 to 2022. Given the economic characteristics and scale differences between these two regions, further decomposition of the TE into PTE and SE reveals different underlying causes for the improvement. As is shown in Figure 4b, the increase in TE in Southern China is primarily due to improvements in PTE, while the efficiency improvement in Northwestern China is mainly attributed to SE. This suggests that the driving forces behind growth in these regions are distinct, and targeted strategies are needed to promote their continued development. The annual increase in TE in Southern China is primarily due to improvements in pure technical efficiency. Southern China, being one of the most economically developed regions in the country, has seen substantial support from both the government and enterprises for high-tech industries, particularly in sectors such as pharmaceuticals, electronics, and advanced manufacturing, with research and development investments increasing annually. Additionally, the southern region benefits from well-developed infrastructure and a strategic geographical location, attracting a concentration of high-end technical talent and innovative enterprises and thereby further optimizing resource allocation efficiency. Improvements in management practices and production organization methods have also contributed to higher production efficiency. These factors collectively underpin the region’s outstanding performance in PTE, driving the overall improvement in technical efficiency. Moving forward, if the southern region continues to prioritize technological innovation and talent acquisition while further optimizing industrial structure and resource allocation, its technical efficiency is expected to achieve even greater progress. The improvement in TE in Northwestern China is mainly driven by SE. The likely cause is that PI in this region has expanded production scale, achieving economies of scale, which, in turn, has enhanced TE. An analysis of returns to scale (RTS) reveals that, over the past decade, in the total 50 samples from the five provinces in Northwestern China, only one year exhibited decreasing returns to scale (DRS), while the remaining years were under increasing returns to scale (IRS). This indicates that increasing scale investments in the region can further improve efficiency. With the transfer of the PI to Northwestern China, SE in this region has improved, leading to an increase in TE. This industrial relocation is primarily attributed to national policy support, relatively lenient environmental regulations, and abundant environmental capacity. In recent years, the Chinese government has implemented strategies such as the “Western Development” and the “Belt and Road” initiatives, providing tax incentives, fiscal subsidies, and infrastructure support to the northwestern region, thereby attracting a significant number of pharmaceutical enterprises. Simultaneously, the relatively relaxed environmental standards in the northwest have reduced operational costs for enterprises, while its vast land, sparse population, and abundant natural resources Economies 2025,13, 90 24 of 35 have provided favorable conditions for large-scale production. These factors collectively contributed to the improvement of scale efficiency and, through economies of scale, promoted the enhancement of technical efficiency, injecting new momentum into regional economic development. In the future, if the northwestern region can further strengthen infrastructure construction, optimize the business environment, and enhance technological innovation capabilities, the development potential of its pharmaceutical industry will become even more significant. 5.3. The Discussion of Regional Disparities in China’s Pharmaceutical Industry To further discuss the regional disparities of PI across different regions in China, this study employed the coefficient of variation (CV) as an indicator. A larger CV indicates regional disparities between regions, suggesting an uneven distribution of resource allocation and operational efficiency. In contrast, a smaller CV reflects more balanced operational efficiency, implying relatively rational resource distribution. Figure 5presents the CV trend of TE in PI across 31 regions from 2013 to 2022, giving deeper insights into developmental disparities between regions. As shown in Figure 5, both in Stage 1 and Stage 3, despite a slight increase in the coefficient of variation (CV) between 2017 and 2019, the overall trend shows a significant decline. This suggests that the regional disparities in China’s pharmaceutical industry are gradually narrowing. Figure 5. Trend of CV changes in China’s PI from 2013 to 2022. The potential reasons for this include government macro-control measures and the changes in scale efficiency brought about via industrial transfer. Firstly, the government has implemented stringent safety, environmental, and tax policies in developed regions, which have placed significant pressure on the pharmaceutical industry, particularly on pollutionintensive sectors such as active pharmaceutical ingredients (APIs) and pharmaceutical intermediates. This has prompted these industries to gradually shift to underdeveloped regions, contributing to a more balanced regional distribution. Additionally, pharmaceutical companies tend to choose regions with fewer regulations, tax incentives, and larger environmental capacities for their development. However, the increased disparity between 2017 and 2019 could be attributed to major policy adjustments, such as centralized procurement and a consistency evaluation of the generic drug in PI. Industries such as APIs and intermediates, which could not participate in centralized procurement and consistency evaluation, experienced a decline in overall output, thereby increasing regional disparities. Economies 2025,13, 90 31 of 35 Data Availability Statement: Data in this study is available in https://doi.org/10.5281/zenodo.147 55469. Acknowledgments: The authors thank the editorial team and anonymous reviewers for their insightful comments on the article. Conflicts of Interest: Author Jiaqiang Sun was employed by the company Chengdu Maidison Pharmaceutical Technology Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Note 1 Note: This study’s data sources do not include Hong Kong, Macau, or Taiwan, as their statistical data are processed separately in official Chinese reports. References Ahmed, M. H., & Melesse, K. A. (2018). Impact of off-farm activities on technical efficiency: Evidence from maize producers of eastern Ethiopia. Agricultural and Food Economics,6(1), 1–15. [CrossRef] Aigner, D., Lovell, C. K., & Schmidt, P. (1977). Formulation and estimation of stochastic frontier production function models. Journal of Econometrics,6(1), 21–37. [CrossRef] André, F. J., Buendía, A., & Santos-Arteaga, F. J. (2024). Efficient water use and reusing processes across Spanish regions: A circular data envelopment analysis with undesirable inputs. Journal of Cleaner Production,434, 139929. [CrossRef] Avkiran, N. K., & Rowlands, T. (2008). How to better identify the true managerial performance: State of the art using DEA. Omega, 36(2), 317–324. [CrossRef] Bai, H., Irfan, M., & Hao, Y. (2022). How does industrial transfer affect environmental quality? Evidence from China. Journal of Asian Economics,82, 101530. [CrossRef] Baird, G. L., & Bieber, S. L. (2016). The Goldilocks dilemma: Impacts of multicollinearity—A comparison of simple linear regression, multiple regression, and ordered variable regression models. Journal of Modern Applied Statistical Methods,15(1), 18. [CrossRef] Banker, R. D., Charnes, A., & Cooper, W. W. (1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis. Management Science,30(9), 1078–1092. [CrossRef] Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management,17(1), 99–120. [CrossRef] Barney, J. B., & Arikan, A. M. (2005). The resource-based view: Origins and implications. In The Blackwell handbook of strategic management (pp. 123–182). Blackwell Publishers Ltd. Bhardwaj, R. (2024). Economics of the pharmaceutical and medical device industry: Supply chain, trade and innovation. Taylor & Francis. [CrossRef] Borja Reis, C. F. d., & Pinto, J. P. G. (2022). Center–periphery relationships of pharmaceutical value chains: A critical analysis based on goods and knowledge trade flows. Review of Political Economy,34(1), 124–145. [CrossRef] Cai, L., & Sun, Y. (2013). Research on the efficiency of biopharmaceutical enterprises based on DEA and SFA. Science Technology and Management Research,33(2), 24. Camanho, A. S., Silva, M. C., Piran, F. S., & Lacerda, D. P. (2024). A literature review of economic efficiency assessments using Data Envelopment Analysis. European Journal of Operational Research,315(1), 1–18. [CrossRef] Charnes, A., Cooper, W. W., & Rhodes, E. (1978). Measuring the efficiency of decision making units. European Journal of Operational Research,2(6), 429–444. [CrossRef] Chaturvedi, P., Shukla, P., Giri, B. S., Chowdhary, P., Chandra, R., Gupta, P., & Pandey, A. (2021). Prevalence and hazardous impact of pharmaceutical and personal care products and antibiotics in environment: A review on emerging contaminants. Environmental Research,194, 110664. [CrossRef] [PubMed] Chen, G., & Chen, F. (2024). Changes in technological innovation efficiency and influencing factors of listed textile and apparel companies research—Based on three-stage DEA with Tobit modeling. PLoS ONE,19(8), e0307820. [CrossRef] Chen, P., Ding, X., Chen, M., Song, H., & Imran, M. (2025). The impact of resource spatial mismatch on the configuration analysis of agricultural green total factor productivity. Agriculture,15(1), 23. [CrossRef] Coelli, T. J., Rao, D. S. P., O’donnell, C. J., & Battese, G. E. (2005). An introduction to efficiency and productivity analysis. Springer Science & Business Media. [CrossRef] Colm, G. (1960). The theory of public finance: A study in public economy. JSTOR. Cooper, W. W., Seiford, L. M., & Tone, K. (2007). Data envelopment analysis: A comprehensive text with models, applications, references and dea-solver software (Vol. 2). Springer. [CrossRef] Economies 2025,13, 90 32 of 35 Deore, A. B., Dhumane, J. R., Wagh, R., & Sonawane, R. (2019). The stages of drug discovery and development process. Asian Journal of Pharmaceutical Research and Development,7(6), 62–67. [CrossRef] Dou, J., & Han, X. (2019). How does the industry mobility affect pollution industry transfer in China: Empirical test on Pollution Haven Hypothesis and Porter Hypothesis. Journal of Cleaner Production,217, 105–115. [CrossRef] Dyson, R. G., Allen, R., Camanho, A. S., Podinovski, V. V., Sarrico, C. S., & Shale, E. A. (2001). Pitfalls and protocols in DEA. European Journal of Operational Research,132(2), 245–259. [CrossRef] Farrell, M. J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society Series A: Statistics in Society,120(3), 253–281. Feng, M., & Li, X. (2020). Evaluating the efficiency of industrial environmental regulation in China: A three-stage data envelopment analysis approach. Journal of Cleaner Production,242, 118535. [CrossRef] Fried, H. O., Lovell, C. K., Schmidt, S. S., & Yaisawarng, S. (2002). Accounting for environmental effects and statistical noise in data envelopment analysis. Journal of Productivity Analysis,17, 157–174. [CrossRef] Friedman, M. (2007). The social responsibility of business is to increase its profits. In Corporate ethics and corporate governance (pp. 173–178). Springer. Gascón, F., Lozano, J., Ponte, B., & de la Fuente, D. (2017). Measuring the efficiency of large pharmaceutical companies: An industry analysis. The European Journal of Health Economics,18, 587–608. [CrossRef] Ge, X., Imran, M., & Ali, K. (2024). Natural resource-driven prosperity: Unveiling the catalysts of sustainable economic development in the United States. In Natural resources forum. Blackwell Publishing Ltd. Greenacre, M., Groenen, P. J., Hastie, T., d’Enza, A. I., Markos, A., & Tuzhilina, E. (2022). Principal component analysis. Nature Reviews Methods Primers,2(1), 100. [CrossRef] Greene, W. H. (2008). The econometric approach to efficiency analysis. The Measurement of Productive Efficiency and Productivity Growth, 1(1), 92–250. Grossman, S. J., & Stiglitz, J. E. (1977). On value maximization and alternative objectives of the firm. The Journal of Finance,32(2), 389–402. Guanglan, Z., & Zhening, Z. (2024). Evaluation and influencing factors of regional environmental efficiency in China based on three-stage DEA model. SAGE Open,14(4), 21582440241280825. [CrossRef] Guo, C., Song, Q., Yu, M.-M., & Zhang, J. (2024). A digital economy development index based on an improved hierarchical data envelopment analysis approach. European Journal of Operational Research,316(3), 1146–1157. [CrossRef] Guo, H., & Shi, K. (2021). Evaluation of the international competitiveness of China’s pharmaceutical industry and high-quality development strategies during the 14th Five-Year Plan. Journal of Beijing University of Technology (Social Sciences Edition),21(3), 65–79. [CrossRef] Guo, Y., Yu, M.-M., & See, K. F. (2024). Developing a sustainable development goals index for OECD countries: An effectiveness-based hierarchical data envelopment analysis. Environmental Science & Policy,160, 103836. [CrossRef] Hair, J., Black, W., Babin, B., & Anderson, R. (2010). Multivariate data analysis (7th ed.). Prentice Hall. Haitovsky, Y. (1969). Multicollinearity in regression analysis: Comment. The Review of Economics and Statistics,51, 486–489. Hamad, M., & Tarnoczi, T. (2021). Efficiency analysis of companies operating in the pharmaceutical industry in the Visegrad countries. Intellectual Economics,15(2), 131–155. [CrossRef] Hao, B., & Ruan, X. (2022). Research on the technological innovation efficiency of Chinese listed biopharmaceutical enterprises based on the two-stage DEA model. China Pharmacy,12, 7–12. [CrossRef] Hjalmarsson, L., Kumbhakar, S. C., & Heshmati, A. (1996). DEA, DFA and SFA: A comparison. Journal of Productivity Analysis,7, 303–327. [CrossRef] Imran, M., Alam, M. S., Jijian, Z., Ozturk, I., Wahab, S., & Do˘gan, M. (2024a). From resource curse to green growth: Exploring the role of energy utilization and natural resource abundance in economic development. In Natural resources forum. Blackwell Publishing Ltd. [CrossRef] Imran, M., & Jijian, Z. (2023). Exploring the relationship between financial inclusion and natural resource utilization in QUAD economies. Environmental Science and Pollution Research,30(58), 122958–122971. [CrossRef] [PubMed] Imran, M., Khan, M. K., Alam, S., Wahab, S., Tufail, M., & Jijian, Z. (2024b). The implications of the ecological footprint and renewable energy usage on the financial stability of South Asian countries. Financial Innovation,10(1), 102. [CrossRef] Imran, M., Khan, M. K., Wahab, S., Ahmed, B., & Jijian, Z. (2025a). The paradox of resource-richness: Unraveling the effects on financial markets in natural resource abundant economies. Financial Innovation,11(1), 63. [CrossRef] Imran, M., Sattar, A., & Alam, M. S. (2024c). Heterogeneous analysis of free trade agreement between Pakistan and China: A policy guideline for CPEC. Journal of Economic and Administrative Sciences,40(1), 76–94. [CrossRef] Imran, M., Tufail, M., Mo, C., Wahab, S., Khan, M. K., Hoo, W. C., & Ling, Z. (2025b). From resources to resilience: Understanding the impact of standard of living and energy consumption on natural resource rent in Asia. Energy Strategy Reviews,57, 101590. [CrossRef] Economies 2025,13, 90 33 of 35 Imran, M., Zhang, G., & An, H. (2017). Impact of market access and comparative advantage on regional distribution of manufacturing sector. China Finance and Economic Review,5, 1–14. [CrossRef] Jia, F. (2022). Research on the international competitiveness of China’s pharmaceutical industry from the perspective of the global value chain [Unpublished Master’s thesis, Kunming University of Science and Technology]. Joe, Z., & Wu, G. (1996). Dea theory, methods and applications. Science Press. Kalirajan, K. P., & Shand, R. (1994). Stochastic frontier production functions and technical efficiency measurements: A review. Palgrave Macmillan UK. [CrossRef] Kaur, N., & Singh, B. (2024). Dynamics of resource allocation–firm performance relationship: Panel autoregressive distributed lag approach. Managerial and Decision Economics,45(6), 3663–3676. Krugman, P. (1991). Increasing returns and economic geography. Journal of Political Economy,99(3), 483–499. [CrossRef] Kumbhakar, S. C., & Tsionas, E. G. (2011). Some recent developments in efficiency measurement in stochastic frontier models. Journal of Probability and Statistics,2011(1), 603512. [CrossRef] Lai, H., Shi, H., & Zhou, Y. (2020). Regional technology gap and innovation efficiency trap in Chinese pharmaceutical manufacturing industry. PLoS ONE,15(5), e0233093. [CrossRef] [PubMed] Lampe, H. W., & Hilgers, D. (2015). Trajectories of efficiency measurement: A bibliometric analysis of DEA and SFA. European Journal of Operational Research,240(1), 1–21. [CrossRef] Lee, G. H., & Jeon, H. W. (2023). A stochastic frontier analysis (SFA)-based method for detecting changes in manufacturing energy efficiency by sector and time. In IFIP international conference on advances in production management systems (pp. 29–42). Springer. Li, S., & Fan, C. (2009). A review and comparison of SFA and DEA methods. Statistics and Decision,7(283), 25–28. Lin, T.-X., Wu, Z.-h., Ji, X.-x., & Yang, J.-j. (2021). Research on the operating efficiency of Chinese listed pharmaceutical companies based on two-stage network DEA and Malmquist. Mathematical Problems in Engineering,2021(1), 1475781. [CrossRef] Liu, T., Pan, S., Hou, H., & Xu, H. (2020). Analyzing the environmental and economic impact of industrial transfer based on an improved CGE model: Taking the Beijing–Tianjin–Hebei region as an example. Environmental Impact Assessment Review,83, 106386. [CrossRef] Liu, Y., Zhang, N., Xie, C., Jiang, Y., Qin, Y., Zhou, L., Fan, Y., Ren, L., Yin, C., Yang, H., Xie, W., Zhai, Q., Li, G., Chen, H., & Chen, X. (2022). Evolution of drug regulations and regulatory innovation for anticancer drugs in China. Acta Pharmaceutica Sinica B,12(12), 4365–4377. [CrossRef] Luo, D. (2012). A note on estimating managerial inefficiency of three-stage DEA model. Statistical Research,29(4), 104–107. Ma, Z. X. (2010). Data envelopment analysis models and methods. Science Press. Madabushi, R., Seo, P., Zhao, L., Tegenge, M., & Zhu, H. (2022). Role of model-informed drug development approaches in the lifecycle of drug development and regulatory decision-making. Pharmaceutical Research,39(8), 1669–1680. [CrossRef] Madaleno, M., & Moutinho, V. (2023). Stochastic frontier analysis: A review and synthesis. Advanced Mathematical Methods for Economic Efficiency Analysis,692, 55–78. [CrossRef] Martin, R., & Sunley, P. (2003). Deconstructing clusters: Chaotic concept or policy panacea? Journal of Economic Geography,3(1), 5–35. [CrossRef] Mergoni, A., Emrouznejad, A., & De Witte, K. (2024). Fifty years of data envelopment analysis. European Journal of Operational Research. [CrossRef] Mohs, R. C., & Greig, N. H. (2017). Drug discovery and development: Role of basic biological research. Alzh Dement-Trci,3(4), 651–657. [CrossRef] Moulay Ali, H., Guellil, M. S., Mokhtari, F., & Tsabet, A. (2024). The effect of subsidies on tecnical efficiency of Algerian agricultural sector: Using stochastic frontier model (SFA). Discover Sustainability,5(1), 98. [CrossRef] Na, H., Du, T., Sun, W., He, J., Sun, J., Yuan, Y., & Qiu, Z. (2019). Review of evaluation methodologies and influencing factors for energy efficiency of the iron and steel industry. International Journal of Energy Research,43(11), 5659–5677. [CrossRef] National Bureau of Statistics of China. (2023). China statistical yearbook 2023. China Statistics Press. Available online: https:// www.stats.gov.cn (accessed on 27 October 2024). Okeke, E. S., Ezeorba, T. P. C., Okoye, C. O., Chen, Y., Mao, G., Feng, W., & Wu, X. (2022). Environmental and health impact of unrecovered API from pharmaceutical manufacturing wastes: A review of contemporary treatment, recycling and management strategies. Sustainable Chemistry and Pharmacy,30, 100865. [CrossRef] Ðoki´c, D., Novakovi´c, T., Teki´c, D., Matkovski, B., Zeki´c, S., & Mili´c, D. (2022). Technical efficiency of agriculture in the European Union and Western Balkans: SFA method. Agriculture,12(12), 1992. [CrossRef] Porter, M. E. (1990a). New global strategies for competitive advantage. Planning Review,18(3), 4–14. [CrossRef] Porter, M. E. (1990b). The competitive advantage of nations. Macmillan Press LTD. Porter, M. E. (1998). Clusters and the new economics of competition (Vol. 76, No. 6). Harvard Business Review Boston. Qin, Y., Zhang, P., Deng, X., & Liao, G. (2023). Innovation efficiency evaluation of industrial technology research institute based on three-stage DEA. Expert Systems with Applications,224, 120004. [CrossRef] Economies 2025,13, 90 34 of 35 Qiu, L., Yu, R., Hu, F., Zhou, H., & Hu, H. (2023). How can China’s medical manufacturing listed firms improve their technological innovation efficiency? An analysis based on a three-stage DEA model and corporate governance configurations. Technological Forecasting and Social Change,194, 122684. [CrossRef] Rashid, A. A., See, K. F., & Yu, M.-M. (2024). Measuring airline efficiency using a dynamic network data envelopment analysis in the presence of innovation capital. Technological Forecasting and Social Change,206, 123457. [CrossRef] Resti, A. (2000). Efficiency measurement for multi-product industries: A comparison of classic and recent techniques based on simulated data. European Journal of Operational Research,121(3), 559–578. [CrossRef] Riaz, M., Kazmi, S. M. A., Iqbal, M. S., & Hussain, A. (2023). Efficiency analysis of pharmaceutical companies in Pakistan: A case study of ten famous companies. Pakistan Journal of Humanities and Social Sciences,11(3), 3823–3830. [CrossRef] Schuhmacher, A., Hinder, M., von Stegmann und Stein, A., Hartl, D., & Gassmann, O. (2023). Analysis of pharma R&D productivity—A new perspective needed. Drug Discovery Today,28(10), 103726. [CrossRef] Schuhmacher, A., Wilisch, L., Kuss, M., Kandelbauer, A., Hinder, M., & Gassmann, O. (2021). R&D efficiency of leading pharmaceutical companies—A 20-year analysis. Drug Discovery Today,26(8), 1784–1789. [CrossRef] Shi, Y., Charles, V., & Zhu, J. (2025). Bank financial sustainability evaluation: Data envelopment analysis with random forest and Shapley additive explanations. European Journal of Operational Research,321(2), 614–630. [CrossRef] Shin, K., Lee, D., Shin, K., & Kim, E. (2018). Measuring the efficiency of US pharmaceutical companies based on open innovation types. Journal of Open Innovation: Technology, Market, and Complexity,4(3), 34. [CrossRef] Shrestha, N. (2020). Detecting multicollinearity in regression analysis. American Journal of Applied Mathematics and Statistics,8(2), 39–42. [CrossRef] Singh, P. K., Bhatt, R. K., & Kiran, R. (2020). Investigating allocative efficiency, cost efficiency, technical efficiency using stochastic frontier analysis (SFA) of groundnut farm households: Evidence from a case in India. Custos e Agronegocio on Line,16(4), 105–129. Solanki, N., Verma, J., Kumar, S., Seema, Mundlia, J., Saini, R., & Saini, S. (2024). Impact of contract research organizations on pharmaceutical industries: A review. Applied Drug Research, Clinical Trials and Regulatory Affairs, e26673371332371. [CrossRef] Song, H., & Ma, Y. (2024). Measurement and spatial and temporal characterization of logistics efficiency–Based on a three-stage DEA model. Heliyon,10(19), e38455. [CrossRef] [PubMed] Stevi´c, Ž., Miški´c, S., Vojinovi´c, D., Huskanovi´c, E., Stankovi´c, M., & Pamuˇcar, D. (2022). Development of a model for evaluating the efficiency of transport companies: PCA–DEA–MCDM model. Axioms,11(3), 140. [CrossRef] Sun, J., Rosli, A. B., & Daud, A. (2024). Operational efficiency of pharmaceutical companies in China: Based on three-stage DEA with undesirable outputs. Sustainability,17(1), 207. [CrossRef] Tang, Y., & Imran, M. (2024). Trade-offs and synergies: Examining the impact of natural resource rents and energy efficiency on financial development in the RCEP context. In Natural resources forum. Blackwell Publishing Ltd. [CrossRef] Tang, Y., Yin, M., Yang, W., Li, H., Zhong, Y., Mo, L., Liang, Y., Ma, X., & Sun, X. (2019). Emerging pollutants in water environment: Occurrence, monitoring, fate, and risk assessment. Water Environment Research,91(10), 984–991. [CrossRef] Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509–533. Tone, K. (2001). A slacks-based measure of efficiency in data envelopment analysis. European Journal of Operational Research,130(3), 498–509. [CrossRef] Wang, G., & Zhang, S. (2023). Evaluation of the international competitiveness and country dependence of China’s pharmaceutical manufacturing industry based on the global value chain. Science Technology and Industry,23(13), 16–22. Wang, J., Han, D., & Wang, Y. (2020). Empirical research on innovation efficiency in China based on SFA model. In Iop conference series: Earth and environmental science (Vol. 474, p. 072055). IOP. Wang, R., & Duan, Y. (2023). Dynamic comparison on the technical efficiency between China’s EPEs and PEs: Two-dimensional measurement based on SFA. Journal of Cleaner Production,406, 136986. [CrossRef] Wang, W., Imran, M., Ali, K., & Sattar, A. (2024). Green policies and financial development in G7 economies: An in-depth analysis of environmental regulations and green economic growth. In Natural resources forum. Blackwell Publishing Ltd. [CrossRef] Wang, Z., Li, Y., Wang, K., & Huang, Z. (2017). Environment-adjusted operational performance evaluation of solar photovoltaic power plants: A three stage efficiency analysis. Renewable and Sustainable Energy Reviews,76, 1153–1162. [CrossRef] Wang, Z., Zhou, Y., Wang, T., & Zhao, N. (2024). Efficiency of construction waste and carbon reduction in the construction industry: Based on improved three stage SBM-DEA model in China. Engineering, Construction and Architectural Management. [CrossRef] Wei, X., & Zhao, R. (2024). Evaluation and spatial convergence of carbon emission reduction efficiency in China’s power industry: Based on a three-stage DEA model with game cross-efficiency. Science of the Total Environment,906, 167851. [CrossRef] [PubMed] Wu, B., Li, J., Yao, Z., Li, X., Wang, W., Wu, Z., & Zhou, Q. (2023). Characteristics and reduction assessment of GHG emissions from crop residue open burning in China under the targets of carbon peak and carbon neutrality. Science of the Total Environment,905, 167235. [CrossRef] Xia, F., Cui, Y. Y., Liu, J. P., & He, Y. F. (2022). Analysis of operational efficiency of listed pharmaceutical companies in China from 2013 to 2019. Chinese Pharmaceutical Journal,31(11), 1–5. Economies 2025,13, 90 35 of 35 Xiong, A., & Meng, G. (2019). Research on the technological innovation efficiency of pharmaceutical enterprises based on DEA method: A case study of the top 15 listed companies in Shenzhen and Shanghai. Chinese Journal of New Drugs, 6. [CrossRef] Xu, T., Kang, C., & Zhang, H. (2022). China’s efforts towards carbon neutrality: Does energy-saving and emission-reduction policy mitigate carbon emissions? Journal of Environmental Management,316, 115286. [CrossRef] Yan, L. (2012). Empirical study on the regional innovation efficiency evaluation index system in China. Management World, 174–175. [CrossRef] Yang, R. (2024). Operational efficiency analysis of listed biomedical companies in China: Three-stage DEA and Malmquist index. Advances in Economics and Management Research,10(1), 264. Zhang, L., & Cui, J. (2024). Research on total factor energy efficiency in western China based on the three-stage DEA-Tobit model. PLoS ONE,19(4), e0294329. [CrossRef] Zhang, M., Imran, M., & Juanatas, R. A. (2024). Innovate, conserve, grow: A comprehensive analysis of technological innovation, energy utilization, and carbon emission in BRICS. In Natural resources forum. Blackwell Publishing Ltd. Zhang, Y., Wang, Y., Zhang, J., Liu, J., Ruan, J., Jin, X., Liu, D., Lu, Z., & Xu, Z. (2024). Research on waste gas treatment technology and comprehensive environmental performance evaluation for collaborative management of pollution and carbon in China’s pharmaceutical industry based on life cycle assessment (LCA). Science of The Total Environment,919, 170555. [CrossRef] Zhao, H., Guo, S., & Zhao, H. (2019). Provincial energy efficiency of China quantified by three-stage data envelopment analysis. Energy,166, 96–107. [CrossRef] Zhong, S., Liang, S., Zhong, Y., Zheng, Y., & Wang, F. (2022). Measure on innovation efficiency of China’s pharmaceutical manufacturing industry. Frontiers in Public Health,10, 1024997. [CrossRef] [PubMed] Zhu, J. (2009). Quantitative models for performance evaluation and benchmarking: Data envelopment analysis with spreadsheets (Vol. 2). Springer. Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.