Measurement of R&D activity level, analysis of regional differences and spatial convergence of the foreign funded enterprises in China

At the Central Economic Work Conference in December 2023, General Secretary Xi Jinping delivered an important speech emphasizing the need to leverage China’s advantage of a super-large market. The speech highlighted the importance of retaining high-quality existing foreign investments, as well as attracting more high-quality foreign investments to continuously promote high-level openness to the outside world. Foreign R&D investment in China serves as an essential component of high-quality foreign investments, playing a crucial role in the nation’s economic growth (Zhou and Dahal, 2024), technological advancement (Zhao et al., 2024) and the establishment of an open and innovative ecosystem (Jiang et al., 2022). Its significance is increasingly recognized in the context of global knowledge sourcing and sustainable development. In the past decade, the competition to attract high-quality foreign investments has intensified globally, becoming a significant national strategy for many developed countries and emerging economies. For China, attracting and retaining foreign R&D not only is vital for accessing global innovation networks and advanced technologies but also serves as a key driver for fostering indigenous innovation (Farooq, 2023). Understanding the dynamics and impacts of foreign R&D is thus critical for formulating effective policies to enhance China’s global innovation competitiveness (Froese et al., 2019; Simon and Cao, 2023). This article summarizes and analyzes the levels of R&D activities of foreign funded enterprises across different regions from 2011 to 2021. The aim is to serve as a decision-making reference for attracting more foreign R&D centers to establish a presence in China, accelerating the leveraging of China’s advantage of a super-large market, and promoting high-quality economic development in China.

In recent years, scholars have extensively studied the motivations, location selection and economic impacts of the R&D activities of foreign funded enterprises in China. Early research primarily focused on the driving factors and the reasons for the location selection of foreign funded enterprises (Junsong et al., 2017; Castellani and Lavoratori, 2019). With the improvement of China’s openness, the research perspective has gradually expanded to the two-way effect of the R&D activities of foreign funded enterprises on local innovation (Liu et al., 2025), including technological spillover (Han and Feng, 2023), collaborative innovation (Huang et al., 2023) and the potential risk of path dependence (Zhao et al., 2021).

Building upon this foundation, many scholars have also studied the positive effects of the R&D activities of foreign funded enterprises in China. They analyzed the impact pathways, modes, and mechanisms of the R&D activities of foreign funded enterprises on China’s independent innovation system (Añón Higón and Díez-Minguela, 2021; Jiang et al., 2023; Wang and Choi, 2023). Some scholars have pointed out that foreign R&D centers are massively withdrawing from China, leading to a decrease in foreign R&D funding. The factors contributing to the withdrawal of foreign R&D centers include rising R&D costs in China and insufficient protection of intellectual property rights (Luo et al., 2021). At the same time, some scholars have found that due to the intense political power and networks in Chinese firms, qualified foreign institutional investors are less motivated to enhance innovation activities (Ain et al., 2022).

With the vigorous development of the modern economy, gradually reducing barriers that hinder the cross-regional flow of R&D factors has become an important issue. This not only facilitates the rational flow of R&D elements but also helps promote the formation of a regionally balanced innovation pattern and a unified national market, a consensus that many scholars are currently studying (Yang and An, 2020; Zhou et al., 2024).

However, the existing literature has limitations in the following two aspects:

First, the one-sidedness of measurement methods. Most existing studies rely on single-dimensional indicators to measure the R&D activities of foreign funded enterprises, such as R&D investment amount (Veliyath and Sambharya, 2011) or patent output (Tse et al., 2021). Although these methods are operable, they ignore the multi-dimensionality of R&D activities (such as personnel allocation and technological dependence). There is especially a lack of a measurement framework combined with panel data to quantify the regional evolution trend of the R&D activities of foreign funded enterprises. Second, the superficial analysis of regional differences. Although some studies have noticed the agglomeration characteristics of the R&D activities of foreign funded enterprises, their analysis has not systematically distinguished the contribution of cross-regional differences and within-region differences (Li et al., 2021).

The innovation of this study is reflected in the following aspects: First, a multi-dimensional measurement system is constructed to measure the R&D activities of foreign funded enterprises, and the Delphi-entropy weight combination weighting method is adopted to overcome the one-sidedness of a single indicator. Second, the convergence mechanism of regional differences is analyzed. Through the decomposition of the Dagum Gini coefficient and the spatial convergence model, the contribution degrees of inter-regional differences and intra-regional differences are quantified. It will provide an effective analytical framework for understanding the logic of foreign R&D behavior in China and offer decision support and a scientific basis for systematically examining the R&D activities of foreign funded enterprises in China. Based on the research objectives outlined above, we propose the following hypotheses:

The subject of this study is the R&D activities of foreign funded enterprises (hereinafter referred to as foreign R&D activities), which include not only foreign funded enterprises but also enterprises with funds from Hong Kong, Macau and Taiwan. It should be noted that for the purpose of this study, there is no specific threshold of investment amount or percentage of ownership required for an enterprise to be classified as a foreign-funded enterprise. Any enterprise that meets the defined criteria of foreign or Hong Kong, Macau and Taiwan funding sources, regardless of the scale of investment or shareholding proportion, is categorized accordingly. Following China’s official statistical standards, all enterprises registered under the categories listed in Table 1 are included. This classification derives from the National Bureau of Statistics. Foreign R&D activities include the R&D activities conducted by foreign funded enterprises and enterprises with funds from Hong Kong, Macau and Taiwan in mainland China.

Currently, although a large amount of research has been conducted in many Chinese and foreign literature on the investment motives, location choices, organizational structures, and other aspects of foreign R&D activities, there is a lack of clear measurement indicators for the actual level of foreign R&D activities and spatial differences. The following examines the level and geographical distribution pattern of foreign R&D activities in provinces in China from three aspects: total scale, per capita level and dependency degree.

The total scale mainly includes the full-time equivalent of R&D personnel in foreign funded enterprises, R&D expenditure of foreign funded enterprises, and the number of invention patents granted to foreign funded enterprises. The full-time equivalent of R&D personnel and R&D expenditure of foreign funded enterprises represent the input of foreign R&D activities in China (Zhu and Xu, 2022), while the number of patents can better reflect the output of foreign R&D activities in China (Gao and Sammartino, 2024).

To further reflect the spatial differences of foreign R&D activities in China, measurements are conducted on the per capita level indicators, including the full-time equivalent of R&D personnel per capita in foreign funded enterprises, R&D expenditure per capita in foreign funded enterprises, and the number of invention patents per capita in foreign funded enterprises to reflect the per capita level of foreign R&D activities. Measurements are also conducted on dependency indicators, including personnel dependency (the percentage of full-time equivalent R&D personnel in foreign funded enterprises to total full-time equivalent R&D personnel in all enterprises), expenditure dependency (the percentage of R&D expenditure in foreign funded enterprises to total R&D expenditure in all enterprises) and patent dependency (the percentage of invention patents in foreign funded enterprises to total invention patents in all enterprises), to reflect the overall dependency of local enterprises’ R&D level on foreign R&D activities and the extent to which foreign R&D activities in a region affect the local area. Based on these measurements, a measurement indicator system for foreign R&D activities in China is obtained (Table 2).

To make the weighting more realistic and reliable, the Delphi method is used for weighting at the criterion level, while the entropy method is used for weighting at the indicator level (Table 2), resulting in comprehensive weights. The Delphi method is a structured expert consultation process to achieve consensus through iterative rounds of feedback. We invited 13 experts, consisting of 3 professors, 6 associate professors, 2 lecturers and 2 researchers, to assign weights to the first-level indicators. After two rounds of surveys, the weights of the first-level indicators were determined based on the results collected in the last round. To measure the level of foreign R&D activities in China, the specific formula is as follows:

where p‘ij represents the standardized evaluation indicator data, and y represents the level of foreign R&D activities in China.

The three primary indicators (total scale, per capita level and dependency degree) were chosen to holistically capture the multi-dimensional nature of foreign R&D activities. Total scale reflects absolute inputs and outputs, aligning with conventional metrics in innovation studies. Per capita level adjusts these metrics by population, addressing efficiency and resource allocation disparities across provinces, critical for regional comparisons. Dependency degree evaluates the autonomy of foreign-funded enterprises’ R&D activities, a novel addition that quantifies vulnerabilities and external dependence (an underexplored dimension in prior literature). Together, these indicators balance scale, efficiency and self-sufficiency, offering a comprehensive framework to analyze spatial disparities and policy impacts.

Based on the levels of foreign R&D activities in China’s 30 provinces from 2011 to 2021, the Dagum Gini coefficient method was used to analyze regional disparities. This method can better measure the evolution trend of regional inequality and identify the sources of differences. The Dagum Gini coefficient can be decomposed into intra-regional differences, inter-regional differences and super-variable density.

The super-variable density reflects the cross-overlapping phenomenon of various regions, indicating the relative gap situation. Intuitively, reducing foreign R&D activity levels in regions with higher averages while increasing those in regions with lower averages could theoretically decrease the overall Gini coefficient by narrowing inter-regional disparities. However, when subsamples exhibit overlapping distributions (i.e. some high-activity provinces in low-level regions exceed some low-activity provinces in high-level regions), simultaneously elevating foreign R&D activity level in high-activity provinces within low-level regions and reducing it in low-activity provinces within high-level regions may paradoxically increase intra-regional Gini coefficients, diminish inter-regional differences and exacerbate inequality within overlapping segments. This counterintuitive outcome could ultimately raise the aggregate Gini coefficient. The portion of Gini coefficient attributable to such inter-group overlaps is termed super-variable density, which equals zero when all subgroups maintain completely non-overlapping distributions.

The specific formula is as follows:

G represents the overall Gini coefficient, k represents the number of regions, n represents the number of provinces, Y represents the foreign R&D activity level, r and h represent different regions, p and q represent different provinces:

where G_w G_nb and G_t represent the contributions of intra-regional differences, inter-regional differences and inter-regional super-variable density, respectively. P and Q are the proportions of the number of provinces in different regions and the proportion of foreign R&D activity level in each region to the total foreign R&D activity level, respectively. D_rh represents the relative impact of foreign R&D activity level in each region. d_rh represents the total influence of region r and region h, which is the mathematical expectation of the sum of all sample values satisfying $|yrp−yhq|>0$ in regions r and h. p_rh is the first-order moment of the super-variable, which is the mathematical expectation of the sum of all sample values satisfying $|yhq−yrp|>0$ in regions r and h.

To test the possible spatial aggregation and spillover effects on foreign R&D activity levels in China, we apply the Moran Index to explore the spatial correlation of foreign R&D activity levels in each province. The calculation formula is as follows:

where y_i and y_j represent the foreign R&D activity levels of the i-th and j-th provinces, W_ij represents the spatial weight, calculated using an adjacent matrix.

To further explore the evolution trends of foreign R&D activity levels in different regions of the eastern, central and western parts of China, convergence analysis is conducted. The concept of convergence was initially proposed by Barro (1992). It is divided into absolute convergence and conditional convergence. Since then, scholars have continuously enriched convergence-related research, categorizing convergence into absolute convergence and conditional convergence based on the existence of restrictions in each region. Convergence refers to the continuous narrowing of the gap in foreign R&D activity levels among different regions over time, measured using the coefficient of variation.

Absolute convergence refers to the situation where, under the same development structure, regions with lower foreign R&D activity levels experience higher growth rates compared to regions with higher levels, ultimately leading to convergence in overall foreign R&D activity levels in China. With the increasing mobility of economic factors among regions, the spatial dependency between regions is also continuously strengthening. Therefore, when analyzing the convergence of foreign R&D activity levels in China, it is important to fully consider the spatial dependency between regions. This study only analyzes the absolute convergence of foreign R&D activity levels, and the model for absolute convergence based on spatial dependency is as follows:

where Y_i,t and Y_{i,t + 1} denote the foreign R&D activity level on province i in period t and t + 1 respectively, β is the regression coefficient, ρ is the spatial regression coefficient, α is the constant term and ε_it is the error term. The focus parameter is β. If β > 0, it indicates that the regional foreign R&D activity level has the characteristic of dispersion; if β = 0, it indicates that the regional foreign investment R&D activity level is relatively balanced; and if β < 0, it indicates that there is the convergence characteristic of the regional foreign R&D activity level.

The regional foreign R&D activity level converges at a rate $θ=−1Tln(1+β)$⁠.

Panel data from 2011 to 2021 for 30 provincial-level administrative units in China (referred to as provinces, excluding the Tibet Autonomous Region and the Hong Kong, Macao and Taiwan regions) were selected as the research sample. According to national statistical standards, the 30 provinces were divided into three major regions: East, Central and West.

The Eastern region includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan. The Central region includes Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan. The Western region includes Inner Mongolia, Guangxi, Chongqing, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang.

The analysis focuses on measuring foreign R&D activity levels in China. The data sources include the corresponding year’s “China Science and Technology Statistics Yearbook”, “Industrial Enterprise R&D Activity Statistics Yearbook” as well as statistical yearbooks from each province.

The measurement results of the foreign R&D activity levels in 30 provinces of China from 2011 to 2021 are shown in Table 3. The larger the measurement result is, the higher the foreign R&D activity level in that province. To intuitively show the evolution trend of the foreign R&D activity levels in 30 provinces of China, ArcGIS software was used to select the data of the foreign R&D activity level in 2011, 2015, 2018 and 2021 to draw the evolution of the foreign R&D activity levels in China, as shown in Figure 1. According to Jenks’ Natural Breaks method, the provinces are classified. For the convenience of description and analysis, according to the classification from low to high, they are respectively defined as provinces with low levels, provinces with medium-low levels, provinces with medium levels, provinces with medium-high levels and provinces with high levels.

Based on Table 3 and Figure 1, it can be concluded that during the research period, the overall distribution of the foreign R&D activity levels in China shows a characteristic of continuous deepening from the northwestern inland areas to the southeastern coastal areas. The gap between different provinces gradually narrows during the development process. From 2011 to 2021, except for Beijing, the levels in other regions have improved. Although Beijing has experienced a slight decline, it has always maintained a high level.

In 2011, the foreign R&D activity levels in the eastern coastal areas were already relatively high. Regions at a high level included Guangdong, Jiangsu, Shanghai, Beijing, etc. indicating that these areas were hotspots for R&D activities. In contrast, the foreign R&D activity levels in the central and western regions and the northeastern region were relatively low, and most provinces were at a medium-low level. In 2015, the foreign R&D activity levels in the eastern coastal areas continued to maintain a high level and increased. The colors of some provinces in the central and western regions began to deepen, indicating that the foreign R&D levels gradually improved. The foreign R&D activity levels in the three northeastern provinces changed little. In 2018, the eastern coastal areas continued to maintain their leading position. The levels of foreign R&D activities in the central and western regions kept rising. Some provinces such as Chongqing and Hubei had darker colors, indicating that foreign R&D activities developed rapidly in these regions. The levels of foreign R&D activities in the three northeastern provinces still grew slowly, with a relatively large gap. In 2021, the levels in the eastern coastal areas continued to grow steadily, and Anhui joined the ranks of high-level provinces. The levels of foreign R&D activities in some provinces of the central and western regions also tended to grow steadily. Although the three northeastern provinces had improvements, there was still a significant gap compared with the eastern regions, and the overall growth was relatively slow. The above research results and related analyses support H1. There are regional differences in the level of foreign R&D activities in China, with the eastern coastal region leading the way.

According to the Dagum Gini coefficient decomposition method, the Gini coefficients of foreign R&D activities in China from 2011 to 2021 were respectively calculated for the eastern, central, and western regions, and the results are shown in Table 4.

The evolution trends of the overall and intra-regional Gini coefficients of foreign R&D activities in China are shown in Figure 2. From 2011 to 2021, the overall Gini coefficient of foreign R&D activities in China showed a downward trend as a whole, dropping from 0.1949 in 2011 to 0.1329 in 2021. The reason for this phenomenon may be that, against the backdrop of the market economic system and the advocacy of independent innovation by enterprises, the imbalance in the level of foreign R&D activity has gradually eased. The intra-regional differences in the level of foreign R&D activity in the eastern and central regions fluctuated around 0.1 and remained basically stable, indicating that the internal differences within the eastern and central regions are small. By 2021, the eastern region had the smallest internal differences among the three major regions. The intra-regional differences in the level of foreign R&D activity within the western region were relatively large, but it showed a downward trend as a whole, decreasing from 0.2124 in 2011 to 0.1427 in 2021, with a reduction rate of 32.82%, and the internal differences within the region significantly narrowed.

As shown in Figure 3, although the inter-regional differences in foreign R&D activities show a downward trend, there still exists a significant regional differentiation issue, especially with the east-west divide consistently at a high level, while the difference between the east and central regions is the smallest. This indicates that the leading advantage of foreign R&D activities in the eastern region of China remains significant, with the differences in foreign R&D activities between the east-central and central-western regions gradually narrowing.

Figure 4 shows that in 2021, the contribution rate of inter-regional difference G_nb reached approximately 62.33%. This far exceeded the 25.65% contribution rate of intra-regional difference G_w and the 12.01% contribution rate of super variable density G_t. This indicates that the differentiation in the level of foreign R&D activity among the eastern, central and western regions is relatively severe and is an important factor affecting the development of the level of foreign R&D activity in China. Therefore, promoting the continuous reduction of the differences in the level of foreign R&D activity among regions and narrowing the inter-regional differences between the eastern and western regions as soon as possible are the keys to the advancement of the level of foreign R&D activity in China. The data supports H2, that inter-regional differences (especially the gap between the eastern and the western) are the main source of the overall imbalance in foreign R&D activities.

Judging from the overall results of the contribution rates, the impact of the inter-regional imbalance G_nb on the level of foreign R&D activity in China is decreasing, which is consistent with the national policy of advocating coordinated regional development. The contribution rate of the intra-regional difference G_w shows a slow upward trend but remains stable overall, indicating that the intra-regional imbalance has a small impact on the level of foreign R&D activity in China. The contribution rate of the super variable density G_t has been increasing, but there was a slight decline from 2020 to 2021, indicating that in regions with a relatively high overall level of foreign R&D activity, there are some provinces whose levels are lower than those of some provinces in regions with a relatively low overall level. This is the inducement for the overall imbalance in the level of foreign R&D activity in China.

Figure 5 presents the coefficient of variation of foreign R&D activities nationwide and in eastern, central and western regions. Overall, this coefficient has slightly fluctuated and decreased, with an average reduction of 3.63%, indicating a narrowing gap in foreign R&D activities. Regionally, in the eastern area, the coefficient has slightly changed and decreased during the sample period, showing a slow-convergence trend over time. The central region saw a 1.23% increase, but still displays a divergence trend despite the small growth. The western region shows a significant and fast-decreasing trend, indicating stable convergence in foreign R&D activities.

In summary, the coefficient of variation shows convergence in the nation, eastern and western regions, while the central region’s convergence is not obvious.

Using the MATLAB software to obtain the global Moran’s I, the results are shown in Table 5. The results indicate that from 2011 to 2021, the global Moran’s I ranged between 0.4450 and 0.6361 and passed the significance test, suggesting a significant positive spatial correlation in foreign R&D activities across the eastern, central and western regions of the country.

After the Hausman test, the fixed-effects model was used for empirical analysis. Based on the LM test and robust LM test results, the national model uses the fixed-effects model with spatial error; the eastern region uses the fixed-effects model with OLS; the central region uses the fixed-effects model with spatial lag; the western region uses the fixed-effects model with spatial Durbin. Table 6 shows the absolute β-convergence analysis results.

It can be seen that the β coefficients for the entire country and the eastern, central and western regions are all negative and significant at the 1% level, indicating the presence of absolute convergence characteristics in the levels of foreign R&D activities across the entire country and the eastern, central and western regions. The imbalance will gradually decrease. The convergence rates are 5.75%, 8.21%, 8.38% and 10.16% respectively, showing a trend of “western region > central region > eastern region > national level”, with the convergence rate of the central region being similar to that of the eastern level. These findings support H3. Foreign R&D activities show convergence characteristics throughout the country and in the eastern, central and western regions.

The narrowing inter-regional disparities could be partially explained by the investment incentive policies implemented by western local governments(Xu and Jaeyeon, 2022). These measures, including tax concessions and infrastructure subsidies documented in prior studies (Wang et al., 2023; Zhou and Liu, 2023), may have not only redirected foreign investment westward but also potentially facilitated beta-convergence through mechanisms requiring verification – possibly by reducing intra-regional investment barriers and creating secondary innovation hubs that distribute FDI more evenly across western provinces. Second, the Belt and Road Initiative’s spatial economic effects (Hu et al., 2022) appear to have amplified this trend by structurally altering foreign enterprises’ location calculus through improved transnational connectivity. Meanwhile, central regions’ performance aligns with New Economic Geography theories regarding technology spillovers – their proximity to eastern innovation centers enables absorption of coastal R&D externalities and intermediate technology diffusion. The eastern regions’ sustained dominance corresponds with agglomeration economy principles (Zhao et al., 2022), where the Yangtze River Delta and Pearl River Delta metropolises leverage first-mover advantages in institutional maturity, human capital accumulation, and industrial completeness to maintain FDI attractiveness.

This article explores the theme of “the level of foreign R&D activity in China” and arrives at the following research conclusions.

Through measuring and analyzing the level of foreign R&D activity in different regions of China, and examining regional differences and convergence, it can be observed that: in terms of measuring the level of foreign R&D activity in China, the active areas of foreign R&D activity are still concentrated in the eastern coastal regions, while the level of foreign R&D activity in inland regions remains low, showing a basic trend of gradual decrease from the coastal areas to the inland regions and from the east to the west, which is highly consistent with the economic development levels in various regions of China. The eastern region’s entrenched dominance necessitates strategies to channel high-value spillovers inland via innovation corridors (Hu et al., 2022) while curbing excessive polarization (Zhao et al., 2022). Concurrently, central provinces should capitalize on their coastal adjacency by developing specialized human capital to absorb diffused mid-tier technologies, avoiding direct competition with eastern high-tech clusters.

In terms of regional differences, the uneven distribution of R&D activity among foreign enterprises in various provinces across the country is gradually easing. Intra-regional differences are relatively small in the eastern and central regions, while the western region shows larger internal differences that are gradually narrowing. Inter-regional disparities particularly the east-west divide remain the primary contributor to overall imbalance, demanding targeted policy interventions. Policies should sustain infrastructure investments and FDI incentives in western region to accelerate spatial redistribution of foreign R&D activities (Xu and Jaeyeon, 2022). Simultaneously, inter-regional coordination mechanisms should be institutionalized to mitigate the largest eastern-western gap identified in contribution rate analysis.

In terms of convergence characteristics, there are different levels of α and β convergence at the national level and within the eastern, central and western regions. The coefficient of variation indicates α-convergence in the country as a whole, as well as in the eastern and western regions, while the central region shows less evident α-convergence. There is significant absolute β-convergence at the national level and within the eastern, central and western regions. The western region exhibits the fastest convergence rate, suggesting proactive measures in western region could accelerate rebalancing through sustained infrastructure and incentive programs.

These findings underscore the essential role of spatially coordinated governance in capitalizing on convergence trends and mitigating inter-regional gaps to optimize foreign R&D’s contribution to China ’ s innovation landscape.

Our findings offer insights into the evolution of foreign R&D activities in China, informing national strategies for innovation-driven development. The persistent dominance of coastal regions underscores how established hubs leverage agglomeration economies to retain global R&D investments.

Compared to prior studies relying on single metrics, our research more comprehensively captures the complexity of foreign R&D engagement. The Dagum decomposition reveals that inter-regional disparities account for over 60% of overall inequality, with the eastern-western divide as the main challenge. Such detailed analysis is essential for targeted regional policymaking.

The rapid convergence in western provinces showcases the significant impact of regional coordination policies. Tax incentives and infrastructure improvements have successfully redirected foreign R&D resources inland, reducing the eastern-western gap. This validates the government’s efforts to rebalance technological development through place-based interventions. Additionally, by exploring the relationship between regional characteristics and foreign R&D activity levels, our study enriches our understanding of the regional development of foreign R&D in the Chinese context.

The research method of this paper has certain limitations. First, reliance on official statistics may introduce regional data biases, as informal R&D activities or small-scale foreign funded enterprises might be underrepresented, particularly in less-developed provinces. Second, omitted variables, such as local infrastructure quality, educational attainment or firm-level strategic decisions, could influence regional disparities but were excluded due to data constraints. While the Delphi-entropy method mitigates subjectivity in weighting, the exclusive selection of academic experts inherently limits the integration of policy and practitioner perspectives, potentially overlooking practical implementation challenges. Future research could integrate multi-source data and actively engage policymakers in methodology design to enhance robustness.

This study’s provincial-level analysis establishes a foundation for deeper inquiry into the evolving dynamics of foreign R&D activities. Future studies could address the regional data biases arising from reliance on official statistics by integrating multi-source data. For instance, combining official data with data from enterprise surveys or industry reports could better represent informal R&D activities and small-scale foreign funded enterprises, especially in less-developed provinces. Additionally, future research could actively engage policymakers in the methodology design. Their involvement could enhance the integration of policy and practitioner perspectives, mitigating the limitations caused by the exclusive selection of academic experts.