Sage Journals: Discover world-class research

Abstract

Digital transformation has accelerated corporate innovation and strengthened competitive advantage, yet its impact on one of the most critical innovation inputs—corporate R&D expenditure—remains underexplored. This study addresses this gap by employing machine learning-based text analysis on 27,163 observations of Chinese A-share listed firms from 2012 to 2021. By integrating the perspectives of agency theory and dynamic capabilities, I find a positive relationship between digital transformation and corporate R&D expenditure. Additionally, five key dimensions of digital transformation—artificial intelligence, blockchain, cloud computing, big data, and digital technology applications—are shown to significantly increase R&D expenditure. The study further provides evidence that digital transformation enhances the effectiveness of R&D expenditure in producing innovative outcomes. These findings not only advance the theoretical frameworks of agency theory and dynamic capabilities but also offer practical insights for firms seeking to optimize digital strategies to maximize innovation outcomes. While focused on Chinese firms, the results have broader implications for other emerging markets where digital transformation is evolving rapidly.

JEL Classification: G34, O31, O32, O33.

Keywords

corporate R&D expenditures digital transformation innovation performance machine learning text analysis

Introduction

Technological advances in artificial intelligence, blockchain, cloud computing, big data, and digital technology applications have catalyzed a paradigm shift toward digital transformation on a global scale, fundamentally altering industrial structures and competitive dynamics (Guo et al., 2023; Peng & Tao, 2022; Tian et al., 2022; Tu & He, 2023). For firms today, digital transformation has become essential for maintaining competitiveness, reshaping traditional business models, and enabling new approaches to value creation (Loonam et al., 2018; Zhuo & Chen, 2023). The economic implications of digital transformation—through its stimulation of product innovation, improvement in operational efficiency, and enhancement of financial performance—are now widely acknowledged, and this trend has attracted growing interest from economists and policymakers alike (Hess et al., 2020; Petrů et al., 2020; Reuschl et al., 2022; Verhoef et al., 2021).

Within this digital transformation landscape, firms increasingly leverage their digital capabilities to foster innovative products and business models, fundamentally shifting how they allocate resources, especially research and development (R&D). Innovation remains a critical driver of economic growth, with R&D expenditure as a primary input that fuels sustainable innovation (Kogan et al., 2017; Radonić et al., 2021). Yet, despite the essential role of R&D in driving growth, firms often face challenges in securing sufficient funding, limiting their ability to sustain long-term innovation efforts (Du et al., 2022; Foray et al., 2012). Given the transformative shifts introduced by digital technologies across production and governance structures, examining the specific impact of digital transformation on corporate R&D expenditures presents a crucial, though relatively underexplored, research area with significant implications for both corporate strategy and economic policy.

While existing studies emphasize the transformative potential of digital technologies, few have disaggregated digital transformation into its distinct technological components—namely artificial intelligence, blockchain, cloud computing, big data, and digital technology application—and examined their unique effects on corporate R&D expenditure. This disaggregation is vital for firms, as different digital capabilities inform strategic choices and guide the allocation of resources, ultimately determining the direction, and effectiveness of their digital transformation efforts (Zhuo & Chen, 2023). Additionally, R&D expenditure is widely recognized as a critical determinant of innovation performance, which is of paramount importance to both sustained economic growth and competitive advantage. As digital transformation accelerates the adoption of new technologies globally, a clearer understanding of whether, and how, these digital investments facilitate the conversion of R&D efforts into measurable innovation outcomes holds considerable relevance, particularly in emerging economies where digital adoption and innovation are pivotal to economic resilience and growth.

Based on the aforementioned research context, this study seeks to address these gaps by exploring the relationship between digital transformation and corporate R&D expenditure, with a focus on China—a major player in the digital economy, ranking second globally on the digital economic scale. According to the “White Paper on China’s Digital Economy Development (2021)” by the China Academy of Information and Communications Technology, China’s digital economy reached 39.2 trillion yuan (roughly 5.4 trillion USD), accounting for 38.6% of its GDP (Tian et al., 2022). Investigating Chinese firms provides valuable insights into how digital transformation might drive innovation outcomes, especially within emerging markets undergoing rapid digital advancement. Specifically, this study aims to answer three core questions: (1) Does digital transformation increase corporate R&D expenditure? (2) If so, which dimensions of digital transformation play a key role in driving this enhancement? (3) Can digital transformation moderate the relationship between corporate R&D expenditure and innovation performance, potentially enhancing innovation outcomes?

This study makes three key contributions. First, I construct a comprehensive digital transformation index by applying machine learning-based text mining to annual reports from listed firms. I employ the unsupervised learning algorithm word2vec, which generates vector representations of words based on their contextual relationships across large text corpora. This technique is particularly suited for capturing nuanced insights, as it identifies word associations and semantic similarities, thereby enabling a refined measure of digital transformation across firms. By calculating word vector similarities within our dataset, I move beyond general metrics to gain firm-specific insights into digital transformation activities, providing an in-depth perspective on its impact.

Second, I disaggregate digital transformation into five primary dimensions—artificial intelligence, blockchain, cloud computing, big data, and digital technology applications—and investigate each dimension’s influence on R&D expenditure. By opening the “black box” of digital transformation, this approach deepens our understanding of how specific digital strategies affect economic outcomes, shedding light on the mechanisms by which distinct digital tools and technologies drive corporate innovation investments.

Third, I examine whether digital transformation enhances the translation of R&D investments into innovation performance, thereby informing a broader economic perspective on digital transformation’s role in fostering productive innovation. By focusing on this moderating effect, I contribute to the broader discourse on how digital initiatives can amplify R&D’s impact on innovation, a topic of critical interest to economists, and policymakers striving to support innovation-driven growth, particularly in digitally evolving markets.

The paper is organized as follows: Section 1 reviews relevant literature on digital transformation and its influence on corporate innovation and R&D, establishing the research hypotheses. Section 2 describes the research design, including data sources, variable definitions, and empirical models. Section 3 presents the empirical findings, robustness checks, and extended analyses. Finally, Section 4 concludes by summarizing the study’s key insights and discussing implications for future research and policy.

Literature Review and Hypothesis Development

Literature Review

In terms of the literature on digital transformation, many recent studies have examined digital transformation, establishing a general consensus on its definition. Specifically, digital transformation refers to an organization’s strategic adoption and utilization of digital technologies to innovate and reshape its products, services, and operational processes. Vial (2019) describes digital transformation as an adaptive response to environmental changes, achieved by integrating digital tools like mobile computing, artificial intelligence, and cloud computing to overhaul value creation. Likewise, Hess et al. (2020) define digital transformation as incorporating advanced technologies to foster novel product lines, innovative value creation methods, and restructured organizational frameworks. Gong and Ribiere (2021) expand on this by framing digital transformation as a significant shift driven by technology, leading to the generation of advanced value. Additionally, Verhoef et al. (2021) outline digital transformation as a three-stage process: digitization of analog information, digital adaptation of business processes, and strategic transformation of the business model.

Most research investigating the economic impact of digital transformation emphasizes its role in enhancing production and operational aspects within organizations. One central viewpoint suggests that digital technologies enable firms to transcend internal resource constraints, allowing access to external resources for strategic growth in new markets and product development (Çera et al., 2020; Chan et al., 2019; Jum’a & Alkhodary, 2024; Rakovic et al., 2024). Some scholars argue that businesses can improve production efficiency by leveraging digital transformation. This enhancement is achievable through the collection and utilization of consumer data (Earley, 2014), adaptation of organizational structures (Kretschmer & Khashabi, 2020), and adjustments in value creation methods (Rachinger et al., 2019). Furthermore, corporate digital change extends organizational reach by facilitating connectivity and collaboration within supply chains (Raj et al., 2024; Reuschl et al., 2022; Zheng & Chen, 2024), while concurrently improving operational efficiency, value generation, and innovation through ecosystems and platforms (Chan et al., 2019; Gong & Ribiere, 2021; Rachinger et al., 2019). Recent studies have also shown how digital platforms contribute to optimizing innovation ecosystems, driving sustainable corporate growth (e.g., Pu, 2025; Secundo et al., 2024).

Previous studies highlight that sustained R&D expenditures enhance product, service, and technological optimization within the innovation process, contributing to corporate productivity (e.g., Doraszelski & Jaumandreu, 2013; Parisi et al., 2006). Moreover, R&D spending allows firms to secure long-term competitive advantages, transforming R&D practices and strategies into profitable technological innovations that yield substantial returns and increased market valuation (Falk, 2012; Polívka & Dvořáková, 2023; Stacho et al., 2023; Toader et al., 2023). For instance, J. Zhang and Liang (2012) examined the Chinese new energy sector, finding that firm-level R&D expenditures positively impact eco-innovation performance. Their research also highlighted the importance of external resources, including ownership structures and government subsidies, in enhancing R&D-driven performance. At the corporate governance level, Rodrigues et al. (2020) investigated the influence of governance mechanisms on corporate investment, particularly in R&D. Their findings suggest that ownership concentration alleviates agency issues linked to R&D investments, underscoring the importance of governance and external support in maximizing R&D’s benefits for innovation outcomes. Recent literature similarly notes the significance of external financing and collaborative innovation efforts in sustaining R&D-driven growth, particularly in high-tech industries (Leung & Sharma, 2021; Li et al., 2024).

In exploring the relationship between digital transformation and corporate innovation activities, initial studies primarily focused on the role of the Internet. The rapid adoption of the Internet has been instrumental in accelerating the diffusion of knowledge and information, thus fostering corporate innovation (Peng & Tao, 2022; Vial, 2019). Furthermore, Ferreira et al. (2019) analyzed how digital transformation influences innovation, particularly in service and process innovation, shedding light on its positive impact on specific corporate innovation dimensions. More recent research has expanded this view, showing that digital tools like AI and cloud computing are now core to enhancing innovation by enabling data-driven decision-making, automation, and adaptive business processes (Bag et al., 2021; Duan et al., 2019).

In summary, the existing literature on digital transformation primarily focuses on characterizing its features and exploring its influence on business models, organizational structures, and management strategies. However, there remains limited research on the direct effects of digital transformation on corporate R&D expenditures, signifying a need for deeper investigation and analysis in this area.

Hypothesis Development

The Relationship Between Digital Transformation and Corporate R&D Expenditure

Digital transformation has emerged as a critical factor in enhancing corporate R&D capabilities, as it significantly reshapes the way firms manage information, resources, and strategic decision-making. From the perspective of agency theory, one of the core advantages of digital transformation is its ability to reduce information asymmetry between managers and shareholders, thereby addressing long-standing agency conflicts. In typical corporate structures, managers are responsible for operational oversight, including the management of innovation activities and financial disclosures (Iqbal et al., 2022). However, shareholders often face challenges in accessing complete and accurate insights into these operations, resulting in a potential misalignment of interests and decision-making priorities (O’Connor & Rafferty, 2012). Moreover, the inherently risky and long-term nature of R&D investments may encourage managers to prioritize short-term financial performance over strategic, long-term innovation, potentially compromising shareholder interests (Y. Chen & Huang, 2023; Yuan & Wen, 2018).

Digital technologies, such as big data, cloud computing, and blockchain, have transformed corporate information landscapes by enhancing transparency and reducing the scope for managerial discretion in financial reporting (Cong & He, 2019). For instance, these technologies enable firms to automate the processing and analysis of financial data, significantly reducing the risk of information distortion and ensuring more accurate and timely R&D disclosures (Zhuo & Chen, 2023). This increased transparency can effectively minimize the information gaps between managers and shareholders, thereby reducing the risk of misinformation and enhancing overall governance (Ferreira et al., 2019; Iqbal et al., 2022; Peng & Tao, 2022).

Additionally, digital transformation facilitates the integration and synchronization of dispersed resources, creating a more cohesive and efficient innovation ecosystem. Automated financial systems can streamline resource allocation, reduce manual errors, and improve decision-making accuracy, which collectively strengthen the financial sustainability of R&D activities (Llopis-Albert et al., 2021; Urbinati et al., 2020). This approach not only reduces the likelihood of managerial opportunism but also aligns short-term operational goals with long-term innovation objectives, promoting a more stable and transparent financial environment. Based on these considerations, I propose the following hypothesis:

H1: There is a positive relationship between digital transformation and corporate R&D expenditures.

The Relationship Between the Sub-dimensions of Digital Transformation and Corporate R&D Expenditure

Dynamic capabilities, as derived from the resource-based view, were introduced by Teece et al. (1997), emphasizing the ability of firms to build, integrate, and reconfigure internal and external competences to address rapidly changing environments. According to Teece (2007), dynamic capabilities primarily involve three functions: sensing opportunities and threats, seizing them, and reconfiguring resources. These capabilities are critical for fostering innovation and R&D investments, as they enable firms to respond swiftly to technological shifts, market demand changes, and competitive pressures (Chakrabarty & Wang, 2012). The five dimensions of digital transformation—artificial intelligence, blockchain, cloud computing, big data, and digital technology applications—serve as key drivers of firms’ dynamic capabilities, thereby enhancing R&D expenditure through their respective attributes.

Artificial Intelligence (AI): AI enhances a firm’s sensing capability by analyzing vast amounts of data, automating decision-making, and forecasting market trends (Bag et al., 2021). Research shows that the application of AI enables firms to identify innovation opportunities more rapidly and improve the efficiency of R&D investments (Agrawal et al., 2019). By processing data and predicting market movements more intelligently, AI helps reduce information asymmetry and decision-making uncertainty, prompting increased investment in R&D activities (Duan et al., 2019). AI not only accelerates product innovation cycles but also optimizes the allocation of R&D resources, driving firms to invest more in R&D to maintain a competitive edge.

Blockchain (BC): BC technology enhances transparency and trust, particularly in supply chain management and data security (Heo et al., 2024). This increased trust facilitates the seizing of resources, reducing innovation risks. Studies suggest that BC technology mitigates collaboration barriers between firms, fostering inter-organizational R&D collaboration, which in turn increases R&D expenditure (Queiroz & Fosso Wamba, 2019). By creating a secure environment for innovation ecosystems, BC bolsters firms’ confidence in new technologies and R&D.

Cloud Computing (CC): The flexibility and economies of scale provided by CC allow firms to share and utilize computational resources on demand, reducing the fixed costs of R&D (Park et al., 2023). CC enhances a firm’s resource reconfiguration capacity, enabling more efficient resource allocation during innovation processes and allowing firms to respond more effectively to rapid technological changes (Yu et al., 2023). Its use also allows firms to scale R&D projects more flexibly, alleviating financial constraints and promoting sustainable R&D investments (Lăzăroiu et al., 2023).

Big Data (BD): BD technology provides firms with more accurate market insights and supports innovation decision-making through advanced data analysis (Cadden et al., 2023). It significantly enhances firms’ sensing and forecasting abilities, allowing them to better identify R&D directions and innovation opportunities (N. Chen et al., 2012). Moreover, research indicates that BD can optimize R&D processes, reducing inefficiencies and encouraging increased investment in high-potential R&D projects (Xie et al., 2021).

Digital Technology Applications (DTA): Digital technologies span from optimizing production processes to implementing intelligent management systems, enhancing operational efficiency and freeing up resources for R&D expenditure (W. Chen & Srinivasan, 2024). The widespread adoption of digital technologies enables firms to better integrate existing resources, respond more swiftly to market changes, and enhance innovation capabilities, thus promoting R&D input (Urbinati et al., 2020). Furthermore, prior literature suggests that these technologies improve the visibility and controllability of R&D projects, further encouraging firms to increase their R&D investment (e.g., Chandrasekaran et al., 2016).

Based on the discussion above, the five dimensions of digital transformation significantly enhance a firm’s dynamic capabilities and strengthen its competitiveness in the rapidly evolving R&D landscape. These dimensions likely encourage increased R&D expenditure by improving the firm’s ability to sense, seize, and reconfigure resources. Therefore, I posit:

H2: The five dimensions of digital transformation (artificial intelligence, blockchain, cloud computing, big data, and digital technology applications) positively and significantly influence corporate R&D expenditures.

The Role of Digital Transformation on the Conversion of Corporate R&D Expenditure Into Innovation Outcomes

The Agency Theory Perspective

According to agency theory (Jensen & Meckling, 1976), there is often a misalignment between the interests of managers and shareholders, particularly when it comes to investments in R&D (X. Wang, 2024). Managers may be reluctant to engage in high-risk, long-term R&D projects due to the uncertainty involved, potentially resulting in inefficiencies and underutilization of R&D resources (Muhammad et al., 2024).

Digital transformation, through technologies such as AI, big data analytics, and blockchain, improves managerial oversight, transparency, and accountability (Li et al., 2024). These technologies reduce agency costs by providing better monitoring mechanisms and clearer data on R&D performance, which aligns managerial actions more closely with shareholder interests. As a result, the positive effects of R&D expenditure are amplified, leading to greater innovation outcomes. For instance, AI and data analytics can significantly improve project selection by using predictive modeling to assess the potential success of R&D projects (Sharma & Chanda, 2017). This reduces uncertainty and ensures that capital is allocated toward projects with higher expected returns, thereby increasing the efficiency of R&D expenditures (Lăzăroiu et al., 2023). Blockchain technology enhances trust in multi-party R&D collaborations, enabling better resource utilization and ultimately contributing to higher innovation performance (Queiroz & Fosso Wamba, 2019).

The Dynamic Capabilities Perspective

Dynamic capabilities emphasize the ability of firms to integrate, build, and reconfigure internal and external competencies to address rapidly changing environments (Teece, 2007; Teece et al., 1997). Digital transformation enhances a firm’s dynamic capabilities by enabling more agile responses to market shifts and technological changes (Agrawal et al., 2019; Chakrabarty & Wang, 2012). Specifically, digital tools like cloud computing, AI, and big data improve a firm’s ability to sense new innovation opportunities, seize them through strategic decision-making, and reconfigure resources to maximize R&D efficiency (Bag et al., 2021; Duan et al., 2019; Heo et al., 2024).

By utilizing digital technologies, firms can better identify emerging market trends and shifts in consumer demand, which improves the alignment between R&D efforts and external opportunities. For example, cloud computing provides the flexibility to scale R&D operations according to evolving needs, ensuring that resources are deployed efficiently and in real time (H. Chen et al., 2015). This heightened agility in resource allocation allows firms to extract greater value from R&D expenditure, translating it more effectively into innovation performance (Liu et al., 2018).

Moreover, digital transformation supports the reconfiguration of R&D processes to optimize knowledge flows, reduce redundancies, and improve collaboration both within and across organizations (Camarinha-Matos et al., 2019; Warner & Wäger, 2019), all of which contribute to enhanced innovation outcomes. This dynamic adaptability ensures that R&D expenditures are not only maintained but also strategically enhanced, leading to superior innovation performance.

In summary, the above discussion leads me to hypothesize that:

H3: Digital transformation positively moderates the relationship between corporate R&D expenditure and innovation performance, making R&D expenditures more efficient in generating innovation outcomes.

Data and Methodology

Data and Sample

This study is based on data from Chinese A-share listed companies from 2012 to 2021. To ensure the sample’s relevance and quality, a purposive sampling technique was employed, focusing on companies that meet specific criteria aligned with the research objectives. The data used in the analysis includes R&D expenditure and patent application data sourced from the CNRDS database, digital transformation information extracted from corporate annual reports, and additional financial and firm-specific data obtained from the CSMAR database. The data preprocessing involved the following steps: (1) excluding financial firms, given their distinct governance structures and financial performance characteristics compared to non-financial firms in China; (2) removing companies under Special Treatment (ST) status due to consecutive losses over 2 years, as per Chinese securities regulations, to eliminate the potential impact of abnormal financial conditions and possible delisting risks; (3) excluding firms with missing data, to enhance the robustness and usability of the dataset. After data preprocessing, a total of 27,163 firm-year observations were retained for the analysis.

Variables

Dependent Variable

The dependent variable of our research is R&D expenditure (RD), with a natural logarithm of the R&D expenditure disclosed by firms used as the measure of RD. Drawing on relevant literature in the R&D field (e.g., Hu et al., 2020; Iqbal et al., 2022; Mao et al., 2020; Yuan & Wen, 2018), I employ four additional metrics for robustness checks: the ratio of current year R&D expenditure to current year total assets (RD1), the ratio of current year R&D expenditure to current year total revenue (RD2), the ratio of current year R&D expenditure to previous year total assets (RD3), and the ratio of current year R&D expenditure to previous year total revenue (RD4).

Independent Variable

The independent variable in this study is digital transformation (DT), which captures the extent to which firms adopt and integrate digital technologies. DT reflects a company’s strategic focus on leveraging digital tools to enhance operations and innovation capabilities, often articulated through corporate disclosures such as annual reports (Kindermann et al., 2021). Given the strategic importance of these documents, text mining methods were applied to systematically assess firms’ digitalization levels.

Following the approach of Tu and He (2023) and Pu and Zulkafli (2024a), this study utilized Python to gather annual reports from the selected sample firms. The digital transformation measure was constructed based on five core dimensions: Artificial Intelligence (AI), Blockchain (BC), Cloud Computing (CC), Big Data (BD), and Digital Technology Application (DTA). The construction of this variable involved several key steps: (1) seed word identification. Initially, a set of foundational keywords, including “artificial intelligence,”“cloud computing,”“blockchain,”“big data,” and “digital technology application,” was identified. These terms represent the core technologies associated with digital transformation; (2) text preprocessing. Using Python’s Jieba library for Chinese word segmentation and Re module for regular expression matching, all extracted textual content underwent a rigorous preprocessing phase. This included the removal of stop words, punctuation, and irrelevant characters to ensure a clean, analyzable text corpus; (3) word embedding model. The preprocessed text was then analyzed using the word2vec model, a widely adopted machine learning algorithm for generating word vectors. This model effectively captures the contextual relationships between words, enabling the identification of semantically similar terms beyond the initial seed words; (4) seed dictionary construction and frequency calculation. The trained word2vec model produced a seed dictionary encompassing the five digital transformation dimensions. This dictionary was subsequently employed to identify and count the occurrences of digital transformation-related keywords within the annual reports; (5) frequency normalization. Given the typical right-skewed distribution of keyword frequencies in corporate disclosures, the resulting counts were transformed using the natural logarithm function, specifically LN (keyword occurrence frequency + 1). This approach mitigates the impact of extreme outliers and allows for a more balanced representation of firms’ digital transformation intensity.

Control Variables

This study controlled for several factors closely linked to R&D expenditure, drawing on prior research (e.g., Choi et al., 2011; Hou et al., 2017; Hu et al., 2020; Yuan & Wen, 2018; R. Zhang et al., 2022). Firm Size (FS) is measured as the logarithm of total assets, as larger firms typically have more financial resources and a greater capacity for R&D investment. Firm Age (FA), defined as the natural logarithm of the years since establishment plus one, often correlates with accumulated experience and stability, which may impact R&D spending. Financial Leverage (LEV), calculated as the ratio of total debts to total assets, indicates debt reliance, where higher leverage may limit funds available for R&D due to debt repayment obligations. Return on Assets (ROA), the ratio of net income to total assets, reflects profitability, with more profitable firms potentially having greater resources to allocate toward R&D. Sales Growth (SG), the percentage change in operating income from the previous year, indicates revenue expansion, as firms with rapid sales growth may increase R&D investments to sustain their competitive advantage. Board Size (BS), defined as the logarithm of the total number of directors, influences corporate governance and strategic decisions, including R&D commitments. Ownership Concentration (OC), or the percentage of shares held by the largest shareholder, acts as a proxy for concentrated ownership, potentially impacting strategic decisions that either limit, or foster R&D. Finally, Institutional Ownership (IO), represented by the ratio of shares held by institutional investors to total shares, reflects investor influence, as institutional investors often prefer long-term value creation, including R&D investment.

Empirical Specification and Data Analysis Techniques

Following prior studies (e.g., Iqbal et al., 2022; Oliver & Gujarati, 1993; Yuan & Wen, 2018; R. Zhang et al., 2022), I employ a fixed effects model (FEM) to examine our hypotheses concerning the impact of digital transformation on corporate R&D expenditure (H1 and H2). Among the three primary approaches for panel data estimation—Pooled Ordinary Least Squares (POLS), Fixed Effects Model (FEM), and Random Effects Model (REM)—the FEM was selected based on diagnostic tests including the F-test, Breusch-Pagan Lagrangian Multiplier (BPLM) test, and Hausman test. First, the F-test results indicated a significant p-value, rejecting the null hypothesis and favoring FEM over POLS. Second, the BPLM test supported the use of REM over POLS with a significant result. Finally, the Hausman test results were significant, establishing FEM as superior to REM. Thus, FEM is chosen as the optimal model to control for unobserved. The empirical model specification is as follows:

\begin{matrix} R D_{i, t} = α_{0} + α_{1} D T_{i, t} + α_{2} Control s_{i, t} + α_{3} Yea r_{i, t} \\ + α_{4} Industr y_{i, t} + ε \end{matrix}

(1)

where $α_{0}$ denotes the intercept, $α_{1} - α_{4}$ are the coefficients to be estimated; ϵ is the error term; i denotes the cross-sectional dimension for firms; and t denotes the time series dimension. The variables in this model are as follows:

RD_i,t : Corporate R&D expenditure, serving as the dependent variable.

DT_i,t : Digital transformation, the main independent variable of interest, which captures the multidimensional nature of digital transformation through five sub-dimensions—Artificial Intelligence (AI), Blockchain (BC), Cloud Computing (CC), Big Data (BD), and Digital Technology Application (DTA). Each sub-dimension represents a specific aspect of digital transformation and collectively provides a comprehensive measure of firms’ digital advancement.

Controls_i,t : Control variables include Firm Size, Firm Age, Leverage, Return on Assets, Sales Growth, Board Size, Ownership Concentration, and Institutional Ownership.

Year_i,t and Industry_i,t : Fixed effects that capture temporal and sectoral variations in the DT-RD relationship.

To address heteroscedasticity and potential autocorrelation issues, I use firm-level clustered robust standard errors (Cameron & Miller, 2015). I also perform a series of diagnostic tests to confirm the model’s suitability. First, I check for multicollinearity using Variance Inflation Factors (VIF) to ensure no high intercorrelation among variables distorts our estimates. The model framework and selected control variables align with established literature (e.g., Choi et al., 2011; Hou et al., 2017; Hu et al., 2020; Yuan & Wen, 2018; R. Zhang et al., 2022), which highlights that those specific firm characteristics are significant factors influencing corporate R&D activities. Additionally, all continuous variables are winsorized at the 1st and 99th percentiles to reduce the influence of outliers.

Empirical Results and Discussion

Descriptive Statistics

The descriptive statistics for the main variables in our study are presented in Table 1, including the number of observations, mean, standard deviation, minimum, and maximum values for each variable.

Table 1.

Descriptive Statistics.

Variables	Acronym	N	Mean	Std	Min	Max
Corporate R&D Expenditure	RD	27,163	15.643	6.125	0.000	21.669
Corporate R&D Expenditure Indicator 1	RD1	27,163	0.020	0.020	0.000	0.101
Corporate R&D Expenditure Indicator 2	RD2	27,163	0.041	0.047	0.000	0.256
Corporate R&D Expenditure Indicator 3	RD3	27,163	0.024	0.025	0.000	0.126
Corporate R&D Expenditure Indicator 4	RD4	27,163	0.047	0.056	0.000	0.304
Digital Transformation	DT	27,163	1.415	1.388	0.000	5.056
Digital Transformation (Binary Indicator)	DT_Dummy	27,163	0.663	0.473	0.000	1.000
Artificial Intelligence	AI	27,163	0.348	0.731	0.000	3.367
Blockchain	BC	27,163	0.015	0.101	0.000	0.693
Cloud Computing	CC	27,163	0.543	0.954	0.000	3.970
Big Data	BD	27,163	0.514	0.882	0.000	3.801
Digital Technology Application	DTA	27,163	0.944	1.101	0.000	4.078
Innovation Performance Indicator 1	IP1	27,163	2.642	1.734	0.000	6.690
Innovation Performance Indicator 2	IP2	27,163	1.863	1.532	0.000	5.974
Firm Size	FS	27,163	22.267	1.293	19.814	26.153
Firm Age	FA	27,163	2.921	0.319	1.609	3.497
Financial Leverage	LEV	27,163	0.422	0.203	0.050	0.893
Return on Assets	ROA	27,163	0.041	0.063	−0.239	0.222
Sales Growth	SG	27,163	0.170	0.390	−0.544	2.445
Board Size	BS	27,163	2.122	0.198	1.609	2.708
Ownership Concentration	OC	27,163	34.206	14.810	8.630	74.180
Institutional Ownership	IO	27,163	44.383	25.029	0.321	94.529

Source. Own.

Corporate R&D Expenditure (RD) has a mean of 15.643 and a standard deviation of 6.125, with values ranging from 0 to 21.669, which shows substantial variation in R&D investment levels across firms. RD1, RD2, RD3, and RD4 are indicators of specific R&D investment components used in robustness tests, reflecting alternative measurements of R&D allocation. Their descriptive statistics also indicate significant variability, which is consistent with the diversity of corporate R&D expenditure strategies documented in previous literature (e.g., Pu & Zulkafli, 2024b).

Digital Transformation (DT) displays a mean of 1.415 and a standard deviation of 1.388, with values ranging from 0 to 5.056. This spread implies that while digital transformation adoption is advancing among Chinese firms, the majority are at relatively early stages. The binary indicator, DT_Dummy, has a mean of 0.663, showing that approximately 66.3% of the firms have initiated digital transformation to some extent. Additionally, digital transformation is captured through five specific dimensions: Artificial Intelligence (AI), Blockchain (BC), Cloud Computing (CC), Big Data (BD), and Digital Technology Application (DTA). AI has a mean of 0.348, BC a mean of 0.015, CC a mean of 0.543, BD a mean of 0.514, and DTA a mean of 0.944, each showing distinct distributions across firms. These dimensions reflect the multi-faceted nature of digital transformation and highlight substantial variability in firms’ engagement across different digital technologies.

Innovation Performance Indicators are captured through two measures, IP1 and IP2. Following Yuan and Wen (2018), IP1 represents the natural logarithm of 1 plus the total number of the company’s patents (including invention patents, design patents, and utility model patents), with a mean of 2.642. IP2 represents the natural logarithm of 1 plus the company’s invention patents specifically, with a mean of 1.863. These indicators enable us to capture different aspects of firms’ innovation performance.

The control variables are as follows: Firm Size (FS), with a mean of 22.267, reflects the typical scale of firms in our sample, while Firm Age (FA), with a mean of 2.921, represents a balanced sample in terms of the firms’ operational maturity. Financial Leverage (LEV) has an average of 0.422, showing moderate debt levels across firms. Return on Assets (ROA), a measure of profitability, ranges from −0.239 to 0.222, suggesting that while some firms operate at a loss, others are profitable, with variability in performance across the sample. Sales Growth (SG) also shows positive and negative values, indicating the presence of both expanding and contracting firms in the sample. Board Size (BS) has a mean of 2.122, indicating that most firms have relatively standard board sizes. Ownership Concentration (OC) and Institutional Ownership (IO) have means of 34.206 and 44.383, respectively, indicating a moderately high concentration of ownership and institutional investor presence in the sample.

Overall, the level of descriptive analysis provides a foundation for understanding the range and distribution of key variables, aligning well with established literature (e.g., Yuan & Wen, 2018; R. Zhang et al., 2022).

Correlation and Variance Inflation Factor Analysis

The Pearson correlation matrix of the major variables is shown in Table 2. Generally, the correlation coefficients between independent and control variables almost are less than 0.50, and this study further conducts the multicollinearity diagnostic test among the continuous variables. Each of the control variables shows a low VIF from the test (less than 2), which indicates no serious multicollinearity issue in our model.

Table 2.

Pearson Correlation and Variance Inflation Factor Analysis.

Variables	RD	DT	FS	FA	LEV	ROA	SG	BS	OC	IO	VIF
RD	1.000										-
DT	0.176	1.000									1.04
FS	0.033	0.026	1.000								1.78
FA	−0.134	0.019	0.170	1.000							1.06
LEV	−0.138	−0.071	0.518	0.164	1.000						1.71
ROA	0.090	0.020	−0.002	−0.080	−0.358	1.000					1.35
SG	0.039	0.043	0.037	−0.043	0.021	0.260	1.000				1.10
BS	−0.036	−0.079	0.273	0.054	0.153	−0.003	−0.024	1.000			1.12
OC	−0.068	−0.116	0.197	−0.089	0.060	0.126	−0.010	0.026	1.000		1.39
IO	−0.107	−0.092	0.449	0.059	0.211	0.102	0.028	0.233	0.499	1.000	1.66

Source. Own.

Univariate Analysis

In the pursuit of univariate analysis, this study introduced a dummy variable, denoted as Dummy (DT). The variable assumes a value of 1 when the annual report of a firm incorporates keywords or their associated synonyms indicative of digital transformation through machine learning; conversely, it assumes a value of 0 otherwise.

Table 3 presents the results of the T-test regarding the differences between firms with DT and firms without DT in our sample firms. The mean of RD is 16.191 for firms having DT and 14.566 for firms without DT, and the differences are both statistically significant at the 1% level. This means that firms with digital transformation have higher R&D expenditure than firms without digital transformation. Additionally, the positively sloped linear regression line in Figure 1 indicates a positive correlation between DT and RD. These findings imply that firms undergoing digital transformation exhibit higher levels of R&D expenditure compared to those that do not engage in digital transformation

Table 3.

Univariate Analysis.

	Dummy (DT) = 1		Dummy (DT) = 0		Differences
Variable	N	Mean	N	Mean	T-value
RD	18,006	16.191	9,157	14.566	1.625***

Source. Own.

***

Denotes statistical significance at the 1% level.

Figure 1.

The linear fit of DT and RD.

Multivariate Results

In this section, I present the baseline regression results examining the relationship between DT and RD to test Hypothesis 1 (H1). All regression models incorporate both year and industry fixed effects to control for industry-specific heterogeneity and the influence of temporal shocks or policies. The FE estimation employs corrected standard errors clustered at the firm level for robust statistical inference.

Column (1) of Table 4 details the regression outcomes testing the impact of DT on RD. The coefficient for DT is significantly positive at the 1% level (α₁ = .618), indicating a robust positive effect of DT on RD without control variables. As I progress to Columns (2) and (3), which sequentially incorporate control variables, I observe that the positive relationship between DT and RD persists and remains significant at the 1% level, even with the addition of these controls. This consistency underscores that digital transformation significantly boosts R&D expenditure, independent of other variables.

Table 4.

Main Results.

	RD	RD	RD
Variables	(1)	(2)	(3)
DT	0.618*** (12.33)	0.414*** (8.52)	0.412*** (8.50)
FS		1.235*** (18.00)	1.288*** (17.58)
FA		−1.732*** (−8.20)	−1.702*** (−8.11)
LEV		−1.279*** (−3.08)	−1.279*** (−3.06)
ROA		1.239* (1.73)	1.420* (1.87)
SG			0.045 (0.48)
BS			0.368 (1.22)
OC			0.005 (1.10)
IO			−0.010*** (−3.98)
Year FE	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes
N	27,163	27,163	27,163
Adj. R²	.458	.510	.511

Source. Own.

, *** indicate statistical significance at the 10% and 1% levels, respectively; robust standard errors are clustered at the firm, and the related t-statistics are reported in brackets.

The results also provide further insights into the control variables. FS, with a significant positive coefficient at the 1% level, indicates that larger firms tend to invest more in RD, likely due to their broader resource base and capacity for innovation. ROA also shows a positive association with RD, implying that firms with higher profitability are more inclined to allocate funds for R&D, possibly to sustain long-term growth. On the other hand, FA, LEV, and IO exhibit significantly negative coefficients, suggesting that older firms, those with higher debt obligations, and those with larger institutional ownership might face constraints or strategic orientations that limit R&D spending. These results align with previous studies (e.g., Hou et al., 2017; Pu & Zulkafli, 2024b; Sunder et al., 2017) that highlight these characteristics as possible inhibitors to R&D investment. It’s notable that other factors such as SG, BS, and OC did not show significant effects on R&D expenditure, suggesting that their influence may not be as critical for RD decisions within this sample.

Overall, the result confirms a positive relationship between the digital transformation indicators obtained through machine learning and R&D expenditure, emphasizing the significance of digital technologies in driving innovation. This finding aligns with the work of Lukić (2017), who argue that organizations leveraging digital technologies are better positioned to enhance their innovation capabilities, thereby fostering a more dynamic R&D environment. While the measure of RD in this study differs from that used by Ferreira and Teixeira (2019), who employed new product development as a metric of innovation, the conclusions drawn are similar. Notably, Kane et al. (2015) highlight that firms investing in digital transformation initiatives tend to experience a greater alignment between their R&D strategies and market demands, resulting in improved innovation outcomes. This evidence further substantiates the role of digital transformation as a driving force in enhancing innovation inputs by breaking organizational boundaries and bridging information gaps.

Robustness Test

So far, the estimation reveals a positive relationship between digital transformation and corporate R&D expenditure, this section performs a variety of additional tests to check the robustness of the baseline results.

(1) Alternative dependent variables. I utilize R1, R2, R3, and R4 as dependent variables to re-examine our primary baseline model (1). The regression results are presented in columns (1) to (4) of Table 5. The estimated coefficients of the key variables exhibit similar magnitudes and consistent signs compared to the previous analysis, further confirming the robustness of our baseline conclusions.

(2) Alternative independent variable. To ensure the robustness of our results, I replaced the continuous variable DT with a binary variable, DT_Dummy, which takes the value of 1 if the firm has undergone digital transformation and 0 otherwise. The regression analysis in column (5) of Table 5 shows that the coefficients for DT_Dummy have the same signs and consistent significance as those of DT.

(3) Subsample regression test. The COVID-19 pandemic led to the rapid adoption of remote work, compelling companies to quickly implement digital tools to support their employees’ R&D activities (Bachmann & Frutos-Bencze, 2022). Additionally, high-tech firms inherently possess advanced technological capabilities, making it easier for them to implement and leverage emerging digital technologies (Sergei et al., 2023). Considering the unique nature of the sample period and the specific characteristics of high-tech firms, I accordingly excluded the samples from 2020 and 2021 as well as the high-tech firm samples. The DT results in columns (6) and (7) of Table 5 support the previous regression results.

(4) Adjusting the robust standard errors from firm level to industry level. To address potential industry-level correlations that may be overlooked by firm-level adjustments, I further transitioned robust standard errors from the firm level to the industry level to mitigate heteroscedasticity and autocorrelation issues. The estimates in column (8) of Table 5 show consistent magnitude and significance, confirming the robustness of our findings.

(5) Extra fixed effects. To further control for geographical variations and potential local-specific factors, I introduced province and city-level fixed effects based on the location of company registration. The findings in column (9) of Table 5 remain consistent with our previous conclusions, reinforcing the robustness of the results.

Table 5.

Robustness Test.

	RD1	RD2	RD3	RD4	RD	RD	RD	RD	RD
Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
DT	0.002*** (10.36)	0.003*** (7.61)	0.002*** (10.48)	0.004*** (7.43)		0.413*** (7.25)	0.629*** (7.74)	0.412*** (3.40)	0.411*** (8.67)
DT_Dummy					0.698*** (6.36)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Extra FE	No	No	No	No	No	No	No	No	Yes
N	27,163	27,163	27,163	27,163	27,163	19,919	15,640	27,163	27,149
Adj. R²	.454	.470	.459	.481	.508	.529	.501	.511	.554

Source. Own.

***

indicate statistical significance at the 1% level; robust standard errors are clustered at the firm or industry level, and the related t-statistics are reported in brackets.

Endogenous Treatment

The potential endogeneity issues in the DT-RD relationship can lead to a bias in the OLS estimates. Therefore, I employed the propensity score matching (PSM) method and generalized method of moments (GMM) estimation techniques to address sample selection biases, omitted variable concerns, and reverse causality issues effectively.

On the one hand, employing the PSM method, I strive to ensure comparability between firms with DT (treatment group) and those without DT (control group) across other dimensions. Specifically, I use firm-level control variables as covariates and establish 1:1 and 1:3 nearest neighbor matches (see Figure 2a and 2b). The first and second columns of Table 6 report the corresponding re-estimations of the model using different matching ratios. The results indicate that the impact of DT on RD remains consistent irrespective of the matching ratio used.

Figure 2.

(a) Balance of treatment and control (PSM 1:1) and (b) balance of treatment and control (PSM 1:3).

Table 6.

Endogenous Treatment.

	PSM (1:1)	PSM (1:3)	DIF-GMM	SYS-GMM
	RD	RD	RD	RD
Variables	(1)	(2)	(3)	(4)
L.RD			0.801*** (24.54)	0.865*** (19.20)
DT	0.392*** (8.11)	0.391*** (8.10)	0.132*** (2.68)	0.689** (2.48)
Controls	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes
Adj. R²	.514	.513
No. of firms			3,441	3,223
No. of IVs			178	104
p-value of AR1			0.000	0.000
p-value of AR2			0.454	0.860
p-value of Hansen			0.233	0.236
N	27,069	27,069	22,910	19,323

Source. Own.

***

, ** indicate statistical significance at the 10% and 1% levels, respectively; robust standard errors are clustered at the firm or industry level, and the related t-statistics are reported in brackets.

On the other hand, while controlling for a range of factors influencing RD, alongside industry and time fixed effects, I acknowledge potential issues such as omitted variable bias and endogeneity arising from bidirectional causality. GMM is particularly suitable for mitigating these issues (Bond, 2002). Therefore, I employ two-step Difference GMM (DIF-GMM) and two-step System GMM (SYS-GMM) estimation methods to address endogeneity concerns affecting the relationship between DT and RD. Columns (3) and (4) of Table 6 display the outcomes from the dynamic regression of DT on RD. The coefficients associated with DT exhibit statistical significance at the 1% or 5% level following rigorous Arellano-Bond and Hansen overidentification tests, underscoring the robustness of our regression findings.

Placebo Test

Following previous studies (e.g., Z. Chen & Jiang, 2024), I estimated the impact of randomly generated DT on RD using a placebo approach to further address concerns regarding unobserved variables. I created a counterfactual set by randomly selecting numbers and running model (1) 500 and 1,000 times. Figures 3a and 3b visually depict the estimated values and p-values of the impact of random DT on RD. Most coefficients significantly deviate from the actual coefficient in the baseline regression (0.412) and lack statistical significance. This result suggests that the randomly generated numbers are ineffective in influencing RD. Consequently, our conclusions do not appear to be driven by unobserved factors.

Figure 3.

(a) Coefficients of DT randomly assigned on RD (500) and (b) coefficients of DT randomly assigned on RD (1,000).

Extended Analysis

Dimensions of Digital Transformation

The empirical results previously presented indicate that DT enhances RD. An intriguing question arises regarding the dimensions driving this effect.

As discussed in the hypothesis development, numerous studies highlight the transformative potential of artificial intelligence (AI), blockchain (BC), cloud computing (CC), big data (BD), and digital technology applications (DTA) in reshaping organizational operations and fostering innovation (e.g., Guo et al., 2023; Simsek et al., 2019; Srinivasan & Venkatraman, 2018). These perspectives generally support the view that advanced digital technologies provide unprecedented opportunities for firms to optimize innovation processes, enhance R&D decision-making capabilities, promote R&D sustainability, and mitigate agency problems. However, there is currently no clear answer as to which dimension of digital transformation can navigate the complex technological landscape and drive firms to increase RD.

To explore the relationship between the different dimensions of DT and RD, I decompose the indicators, based on digital transformation keywords, into five distinct dimensions: AI, BC, CC, BD, and DTA. Columns (1) to (5) of Table 7 present the results for AI, BC, CC, BD, and DTA, respectively. The overall empirical findings indicate that the five dimensions of digital transformation—AI, BC, CC, BD, and DTA—significantly promote R&D expenditures. These findings are broadly consistent with Wang et al. (2022), who argue that digital technologies not only provide essential tools and infrastructure to support innovative activities but also create a favorable paradigm for continuous improvement and exploration of new R&D support.

Table 7.

Dimensions of digital transformation and corporate R&D expenditures.

	Artificial intelligence	Blockchain	Cloud computing	Big data	Digital technology application
	RD	RD	RD	RD	RD
Variables	(1)	(2)	(3)	(4)	(5)
AI	0.387*** (6.90)
BC		0.679*** (2.71)
CC			0.507*** (9.25)
BD				0.485*** (8.11)
DTA					0.289*** (5.22)
Controls	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes	Yes
N	27,163	27,163	27,163	27,163	27,163
Adj. R²	.508	.506	.510	.509	.508

Source. Own.

***

indicate statistical significance at the 1% level; robust standard errors are clustered at the firm or industry level, and the related t-statistics are reported in brackets.

Moderating Effect of Digital Transformation on Innovation Consequences

Our previous research results indicate that digital transformation promotes R&D across multiple dimensions, providing insights into the dynamics of this relationship. Evidence suggests that continuous R&D investments support firms in developing new technologies and products, fostering technological progress, and thereby improving innovation performance (Leung & Sharma, 2021). In other words, stable and increased R&D not only helps maintain a firm’s competitive advantage in rapidly changing markets but also ensures sustainable innovation returns in the long term.

The moderating role of DT is introduced here for several reasons. First, DT has the potential to accelerate knowledge absorption, cross-departmental communication, and process optimization, all of which enhance the effectiveness of R&D expenditures in producing innovation outputs (Z. Chen & Jiang, 2024; Kretschmer & Khashabi, 2020; Sergei et al., 2023). DT can also facilitate real-time data analytics, which helps firms swiftly respond to market demands, streamline R&D processes, and refine product development efforts (Llopis-Albert et al., 2021; Peng & Tao, 2022; Reuschl et al., 2022). Moreover, DT may play a key role in reducing uncertainty in the R&D-to-innovation pipeline by offering tools for more accurate project monitoring, risk assessment, and resource allocation, thus improving R&D’s impact on innovation performance (Sergei et al., 2023; Tang et al., 2023; J. Zhang & Liang, 2012).

Given these factors, DT can be hypothesized to enhance the R&D-innovation performance link by optimizing the way R&D investments are translated into valuable innovation outputs. DT’s moderating role offers a deeper understanding of how firms can leverage digital advancements to maximize the returns on their R&D investments. Therefore, I employ the following models to analyze these economic effects.

I P_{i, t} = β_{0} + β_{1} R D_{i, t} + β_{2} Control s_{i, t} + β_{3} Yea r_{i, t} + β_{4} Industr y_{i, t} + ε

(2)

\begin{matrix} I P_{i, t} = γ_{0} + γ_{1} R D_{i, t} + γ_{2} D T_{i, t} + γ_{3} D T_{i, t} * R D_{i, t} \\ + γ_{4} Control s_{i, t} + γ_{5} Yea r_{i, t} + γ_{6} Industr y_{i, t} + ε \end{matrix}

(3)

Here, model (2) examines the direct effect of R&D expenditures (RD) on innovation performance (IP), while model (3) tests the moderating effect of digital transformation (DT) on the relationship between RD and IP. Following Yuan and Wen (2018), I measure IP in two ways. The first indicator, IP1, is the natural logarithm of 1 plus the total number of the company’s patents, including invention patents, design patents, and utility model patents. The second indicator, IP2, is the natural logarithm of 1 plus the company’s invention patents. β₁ represents the impact of RD on IP; γ₃ denotes the moderating effect of DT on the relationship between RD and IP.

The results in columns (1) and (3) of Table 8 show that RD is significantly positively correlated with innovation performance (IP1 and IP2), indicating that R&D expenditures can enhance corporate innovation performance. The moderating effect of DT is shown in columns (2) and (4), indicating that digital transformation positively and significantly strengthens the positive relationship between RD and IP. Additionally, I executed a marginal effects plot and a kernel density plot of the moderating variable in one graph, providing additional information to analyze the moderating effect of DT on innovation consequences. As seen in Figure 4a and 4b, as DT increases, the marginal effect of RD on innovation performance (IP1 and IP2) gradually increases and is statistically significant. This further reveals the potential of DT in optimizing innovation processes, enhancing R&D investment efficiency, and facilitating the conversion of R&D expenditures into innovation outcomes.

Table 8.

Moderating effect of digital transformation on the relationship between R&D expenditures and innovation performance.

	IP1	IP1	IP2	IP2
Variables	(1)	(2)	(3)	(4)
RD*DT		0.009*** (4.83)		0.013*** (7.92)
RD	0.096*** (29.21)	0.085*** (23.88)	0.069*** (24.58)	0.054*** (17.30)
DT		−0.029 (−0.96)		−0.098*** (−3.61)
Control variables	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Industry FE	Yes	Yes	Yes	Yes
N	27,163	27,163	27,163	27,163
Adj. R²	0.532	0.538	0.465	0.474

Source. Own.

***

indicate statistical significance at the 1% level; robust standard errors are clustered at the firm or industry level, and the related t-statistics are reported in brackets.

Figure 4.

(a) Effects of DT on the relationship between RD and IP1 and (b) effects of DT on the relationship between RD and IP2.

Discussion

Our findings demonstrate a positive relationship between corporate R&D expenditure and digital transformation, highlighting how digital capabilities have become central to innovation strategies. As the digital revolution progresses, the integration of technologies such as artificial intelligence, cloud computing, and big data continues to reshape traditional business models. Digital transformation’s multidimensional nature allows it to reformulate economic production and service processes, altering organizational structures, operational efficiency, and value creation paradigms (Guo et al., 2023; Peng & Tao, 2022; Reuschl et al., 2022). Our results underscore the importance of digital transformation as a primary driver of R&D expenditures, in alignment with Vial (2019), who suggests that digital technologies elicit adaptive strategic responses within firms, strengthening digital infrastructures’ dynamic capabilities, and, consequently, positively impacting innovation outcomes.

In terms of potential mechanisms, our findings indicate that digital transformation plays a dual role in fostering corporate R&D expenditures. From an agency theory perspective, the integration of digital technologies reduces information asymmetry between managers and shareholders, thereby lowering agency costs and enhancing managerial confidence in R&D investments—a result consistent with Ferreira et al. (2019), Gong and Ribiere (2021), and Rachinger et al. (2019). This reduced information gap smooths decision-making channels, encouraging proactive investments in innovation. From the lens of dynamic capabilities perspective, originally posited by Teece et al. (1997), digital transformation fosters the core capabilities needed to sense, seize, and reconfigure resources in response to rapid changes. Specifically, the five dimensions of digital transformation—artificial intelligence, blockchain, cloud computing, big data, and digital technology applications—serve as key drivers for these capabilities. Each dimension uniquely enhances firms’ abilities to respond flexibly to new opportunities, allocate resources efficiently, and scale R&D efforts in dynamic environments, aligning with prior research (e.g., Chakrabarty & Wang, 2012; Teece, 2007).

Furthermore, our analysis provides new insights into the moderating effects of digital transformation. Specifically, it appears that digital technologies reduce time inefficiencies in R&D processes, which significantly boosts innovation output. This finding supports Srinivasan and Venkatraman (2018), who emphasized that digital platforms streamline R&D operations, resulting in faster innovation cycles. Our results also extend the existing literature by demonstrating that digital transformation’s influence on innovation is multifaceted, operating through both direct R&D expenditure increases and efficiency improvements in the innovation process.

Conclusions

This paper investigates the influence of digital transformation on corporate R&D expenditure, providing novel insights into how digital transformation drives innovation. The findings reveal that digital transformation significantly increases corporate R&D expenditure levels. To address endogeneity concerns, I utilized propensity score matching and generalized method of moments estimation. Additionally, I conducted a placebo test using two counterfactual datasets constructed with the ordinary least squares method, confirming that our results are robust across various econometric treatments and additional robustness checks.

Building on these findings, I further analyzed the five dimensions of digital transformation by constructing independent variables with the word2vec deep learning module. This analysis demonstrated that all five dimensions are significantly associated with changes in R&D expenditure. Given the pivotal role of R&D in corporate innovation processes, I also explored the moderating role of digital transformation in the relationship between R&D expenditure and innovation performance. These results highlight digital transformation’s potential to optimize the innovation process, enhance R&D investment efficiency, and improve the conversion of R&D expenditure into innovative outcomes.

From a theoretical perspective, the study contributes the understanding of digital transformation’s impact on R&D expenditure by applying both agency theory and the dynamic capabilities framework. The agency theory lens elucidates how digital transformation aligns managerial incentives with long-term innovation goals, reducing agency problems in corporate R&D investment. This helps to clarify the underlying economic and strategic mechanisms at play. Additionally, the dynamic capabilities framework underscores how digital transformation empowers firms to adapt resources swiftly and respond to external pressures, fostering continuous innovation. Integrating these two perspectives offers a comprehensive view of the strategic and adaptive mechanisms through which digital transformation influences R&D, advancing our understanding of its role in driving sustained innovation.

This research also provides critical policy and managerial implications. For policymakers, the finding that digital transformation enhances corporate R&D expenditure suggests that priority should be given to advancing digital infrastructure, strengthening digital property rights protections, and providing subsidies for digital technologies. Such measures can encourage firms to leverage digital tools to boost R&D efficiency and innovation outcomes. Additionally, implementing differentiated policies that cater to the specific digitalization needs of various industries can comprehensively elevate the digital innovation capacity of the economy. For economies beyond China undergoing similar digital transformations, it is essential to support companies in developing digital skills and ensuring access to digital resources. Encouraging the use of artificial intelligence, big data, cloud computing, and blockchain can further improve R&D efficiency. For managers, this research highlights that strategically integrating digital transformation into R&D processes can enhance innovation efficiency and sustain long-term innovation. Digital integration empowers firms to translate R&D investment into tangible innovation, strengthening their competitive position in the global market.

However, this study has certain limitations. First, the sample is limited to Chinese listed companies, suggesting that future research could benefit from comparative studies in other emerging and developed economies. Additionally, I tested only the moderating effect of digital transformation on the relationship between R&D expenditure and innovation performance. Future research could delve into the interplay between these factors and other relevant variables, such as regional digitalization levels, R&D subsidies, and innovation sustainability, to yield deeper insights.

Footnotes

Acknowledgements

The initial inspiration for this research stemmed from my academic exchanges with the late Dr. Yuanba Li at Universiti Sains Malaysia. Although I completed this work independently over the past year, his unwavering support, insightful guidance, and generous encouragement during our time together continue to serve as a source of strength and motivation. Despite the 3,300 kilometers that eventually separated us, his intellectual spirit and kindness remained ever close. I am deeply grateful for the many moments of academic exchange and mutual support we shared at USM. On many quiet nights of writing and reflection, I found myself missing those days when we encouraged one another through our academic journeys. This work is dedicated to his memory.

ORCID iD

Tingqian Pu

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Agrawal

Gans

Goldfarb

(2019). The economics of artificial intelligence: An agenda. NBER Books. https://ideas.repec.org/b/nbr/nberbk/agra-1.html

Bachmann

Frutos-Bencze

(2022). R&D and innovation efforts during the COVID-19 pandemic: The role of universities. Journal of Innovation & Knowledge, 7(4), 100238. https://doi.org/10.1016/j.jik.2022.100238

Bag

Gupta

Kumar

Sivarajah

(2021). An integrated artificial intelligence framework for knowledge creation and B2B marketing rational decision making for improving firm performance. Industrial Marketing Management, 92, 178–189. https://doi.org/10.1016/j.indmarman.2020.12.001

Bond

S. R.

(2002). Dynamic panel data models: A guide to micro data methods and practice. Portuguese Economic Journal, 1(2), 141–162. https://doi.org/10.1007/s10258-002-0009-9

Cadden

Weerawardena

Cao

Duan

McIvor

(2023). Examining the role of big data and marketing analytics in SMEs innovation and competitive advantage: A knowledge integration perspective. Journal of Business Research, 168, 114225. https://doi.org/10.1016/j.jbusres.2023.114225

Camarinha-Matos

L. M.

Fornasiero

Ramezani

Ferrada

(2019). Collaborative networks: A pillar of digital transformation. Applied Sciences, 9(24), 5431. https://doi.org/10.3390/app9245431

Cameron

A. C.

Miller

D. L.

(2015). A practitioner’s guide to cluster-robust inference. Journal of Human Resources, 50(2), 317–372. https://doi.org/10.3368/jhr.50.2.317

Çera

Phan

Q. P. T.

Androniceanu

Çera

(2020). Financial capability and technology implications for online shopping. E&M Ekonomie a Management, 23(2), 156–172. https://doi.org/10.15240/tul/001/2020-2-011

Chakrabarty

Wang

(2012). The long-term sustenance of sustainability practices in MNCs: A dynamic capabilities perspective of the role of R&D and Internationalization. Journal of Business Ethics, 110(2), 205–217. https://doi.org/10.1007/s10551-012-1422-3

10.

Chan

C. M. L.

Teoh

S. Y.

Yeow

Pan

(2019). Agility in responding to disruptive digital innovation: Case study of an SME. Information Systems Journal, 29(2), 436–455. https://doi.org/10.1111/isj.12215

11.

Chandrasekaran

Linderman

Sting

F. J.

Benner

M. J.

(2016). Managing R&D project shifts in high-tech organizations: A multi-method study. Production and Operations Management, 25(3), 390–416. https://doi.org/10.1111/poms.12410

12.

Chen

Zhu

Guo

Zhu

Qin

(2015). Towards energy-efficient scheduling for real-time tasks under uncertain cloud computing environment. Journal of Systems and Software, 99, 20–35. https://doi.org/10.1016/j.jss.2014.08.065

13.

Chen

Chiang

Storey

(2012). Business intelligence and analytics: From big data to Big impact. MIS Quarterly, 36(4), 1165. https://doi.org/10.2307/41703503

14.

Chen

Srinivasan

(2024). Going digital: Implications for firm value and performance. Review of Accounting Studies, 29, 1619–1665. https://doi.org/10.1007/s11142-023-09753-0

15.

Chen

Huang

(2023). CEO reputation and corporate risk-taking: Managerial competence or managerial defence? Emerging Markets Finance and Trade, 59, 4028–4053. https://doi.org/10.1080/1540496x.2022.2164691

16.

Chen

Jiang

(2024). Digitalization and corporate investment efficiency: Evidence from China. Journal of International Financial Markets Institutions and Money, 91, 101915. https://doi.org/10.1016/j.intfin.2023.101915

17.

Choi

S. B.

Lee

S. H.

Williams

(2011). Ownership and firm innovation in a transition economy: Evidence from China. Research Policy, 40(3), 441–452. https://doi.org/10.1016/j.respol.2011.01.004

18.

Cong

L. W.

(2019). Blockchain disruption and smart contracts. Review of Financial Studies, 32(5), 1754–1797. https://doi.org/10.1093/rfs/hhz007

19.

Doraszelski

Jaumandreu

(2013). R&D and productivity: Estimating endogenous productivity. Review of Economic Studies, 80(4), 1338–1383. https://doi.org/10.1093/restud/rdt011

20.

Duan

Edwards

J. S.

Dwivedi

Y. K.

(2019). Artificial intelligence for decision making in the era of Big Data – Evolution, challenges and research agenda. International Journal of Information Management, 48, 63–71. https://doi.org/10.1016/j.ijinfomgt.2019.01.021

21.

Kim

P. H.

Fourné

S. P. L.

Wang

(2022). In times of plenty: Slack resources, R&D investment, and entrepreneurial firms in challenging institutional environments. Journal of Business Research, 145, 360–376. https://doi.org/10.1016/j.jbusres.2022.03.004

22.

Earley

(2014). The digital transformation: Staying competitive. IT Professional, 16(2), 58–60. https://doi.org/10.1109/mitp.2014.24

23.

Falk

(2012). Quantile estimates of the impact of R&D intensity on firm performance. Small Business Economics, 39(1), 19–37. https://doi.org/10.1007/s11187-010-9290-7

24.

Ferreira

J. J.

Teixeira

A. A.

(2019). Open innovation and knowledge for fostering business ecosystems. Journal of Innovation & Knowledge, 4(4), 253–255.

25.

Ferreira

J. J. M.

Fernandes

C. I.

Ferreira

F. A. F.

(2019). To be or not to be digital, that is the question: Firm innovation and performance. Journal of Business Research, 101, 583–590. https://doi.org/10.1016/j.jbusres.2018.11.013

26.

Foray

Mowery

D. C.

Nelson

R. R.

(2012). Public R&D and social challenges: What lessons from mission R&D programs? Research Policy, 41(10), 1697–1702. https://doi.org/10.1016/j.respol.2012.07.011

27.

Gong

Ribiere

(2021). Developing a unified definition of digital transformation. Technovation, 102, 102217. https://doi.org/10.1016/j.technovation.2020.102217

28.

Guo

Song

Dou

Wang

(2023). Can digital transformation of the enterprise break the monopoly? Personal and Ubiquitous Computing, 27(4), 1629–1642. https://doi.org/10.1007/s00779-022-01666-0

29.

Heo

J. W.

Ramachandran

G. S.

Dorri

Jurdak

(2024). Blockchain Data Storage Optimisations: A Comprehensive Survey. ACM Computing Surveys, 56(7), 1–27. https://doi.org/10.1145/3645104

30.

Hess

T. M.

Matt

Benlian

Wiesböck

(2020). Options for formulating a digital transformation strategy. Routledge eBooks. 151–173. https://doi.org/10.4324/9780429286797-7

31.

Hou

Yuan

(2017). Corporate innovation and political connections in Chinese listed firms. Pacific-Basin Finance Journal, 46, 158–176. https://doi.org/10.1016/j.pacfin.2017.09.004

32.

Huang

Chen

(2020). Environmental regulation, innovation quality and firms’ competitivity–Quasi-natural experiment based on China’s carbon emissions trading pilot. Ekonomska Istrazivanja-economic Research, 33(1), 3307–3333. https://doi.org/10.1080/1331677x.2020.1771745

33.

Iqbal

J. F.

Fareed

Wan

(2022). Financial leverage and corporate innovation in Chinese public-listed firms. European Journal of Innovation Management, 25(1), 299–323. https://doi.org/10.1108/ejim-04-2020-0161

34.

Jensen

M. C.

Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360. https://doi.org/10.1016/0304-405x(76)90026-x

35.

Jum’a

Alkhodary

(2024). Navigating the Industry 4.0 frontier: Unveiling perceived risk and cost moderators in technology adoption. E&M Ekonomie a Management, 27(3), 53–71. https://doi.org/10.15240/tul/001/2024-5-006

36.

Kane

G. C.

Palmer

Phillips

A. N.

Kiron

(2015). Is your business ready for a digital future. MIT Sloan Management Review, 56(4), 37–44. http://ilp.mit.edu/media/news_articles/smr/2015/56415.pdf

37.

Kindermann

Beutel

Garcia de Lomana

Strese

Bendig

Brettel

(2021). Digital orientation: Conceptualization and operationalization of a new strategic orientation. European Management Journal, 39(5), 645–657. https://doi.org/10.1016/j.emj.2020.10.009

38.

Kogan

Papanikolaou

Seru

Stoffman

(2017). Technological innovation, resource allocation, and growth. Quarterly Journal of Economics, 132(2), 665–712. https://doi.org/10.1093/qje/qjw040

39.

Kretschmer

Khashabi

(2020). Digital transformation and organization design: An integrated approach. California Management Review, 62(4), 86–104. https://doi.org/10.1177/0008125620940296

40.

Lăzăroiu

Bogdan

Geamănu

Hurloiu

Luminiţa

Ştefănescu

(2023). Artificial intelligence algorithms and cloud computing technologies in blockchain-based fintech management. Oeconomia Copernicana, 14(3), 707–730. https://doi.org/10.24136/oc.2023.021

41.

Leung

T. Y.

Sharma

(2021). Differences in the impact of R&D intensity and R&D internationalization on firm performance – mediating role of innovation performance. Journal of Business Research, 131, 81–91. https://doi.org/10.1016/j.jbusres.2021.03.060

42.

Liu

Chan

F. T. S.

Yang

Niu

(2018). Understanding the effect of cloud computing on organizational agility: An empirical examination. International Journal of Information Management, 43, 98–111. https://doi.org/10.1016/j.ijinfomgt.2018.07.010

43.

Xie

Chen

(2024). Corporate digital transformation, governance shifts and executive pay-performance sensitivity. International Review of Financial Analysis, 92, 103060. https://doi.org/10.1016/j.irfa.2023.103060

44.

Llopis-Albert

Rubio

Valero

(2021). Impact of digital transformation on the automotive industry. Technological Forecasting and Social Change, 162, 120343. https://doi.org/10.1016/j.techfore.2020.120343

45.

Loonam

Eaves

Kumar

Parry

(2018). Towards digital transformation: Lessons learned from traditional organizations. Strategic Change, 27(2), 101–109. https://doi.org/10.1002/jsc.2185

46.

Lukić

(2017). Transformation of coordination mechanisms in companies due to new technologies: Evidence from empirical research. Economic Themes, 55(3), 421–436. https://doi.org/10.1515/ethemes-2017-0023

47.

Mao

Wang

Zhao

(2020). Does participating in the standards-setting process promote innovation? Evidence from China. China Economic Review, 63, 101532. https://doi.org/10.1016/j.chieco.2020.101532

48.

Muhammad

Migliori

Consorti

(2024). Corporate governance and R&D investment: Does firm size matter? Technology Analysis and Strategic Management, 36, 518–532. https://doi.org/10.1080/09537325.2022.2042508

49.

Oliver

F. R.

Gujarati

(1993). Essentials of econometrics. Journal of the Royal Statistical Society Series A (General), 156(2), 322. https://doi.org/10.2307/2982744

50.

O’Connor

Rafferty

(2012). Corporate governance and innovation. Journal of Financial and Quantitative Analysis, 47(2), 397–413. https://doi.org/10.1017/s002210901200004x

51.

Parisi

M. L.

Schiantarelli

Sembenelli

(2006). Productivity, innovation and R&D: Micro evidence for Italy. European Economic Review, 50(8), 2037–2061. https://doi.org/10.1016/j.euroecorev.2005.08.002

52.

Park

Han

Lee

(2023). Green Cloud? An empirical analysis of cloud computing and energy efficiency. Management Science, 69(3), 1639–1664. https://doi.org/10.1287/mnsc.2022.4442

53.

Peng

Tao

(2022). Can digital transformation promote enterprise performance?—From the perspective of public policy and innovation. Journal of Innovation & Knowledge, 7(3), 100198. https://doi.org/10.1016/j.jik.2022.100198

54.

Petrů

Kramoliš

Stuchlik

(2020). Marketing tools in the era of digitization and their use in practice by family and other businesses. E&M Ekonomie a Management, 23(1), 199–214. https://doi.org/10.15240/tul/001/2020-1-014

55.

Polívka

Dvořáková

(2023). The importance of Industry 4.0 technologies when selecting an ERP system – An empirical study. E&M Ekonomie a Management, 26(3), 51–69. https://doi.org/10.15240/tul/001/2023-3-004

56.

(2025). The role of digitalization in enhancing green innovation efficiency: Examination of multiple sub-dimensions, transformation speed, and heterogeneity. SAGE Open, 15(2). https://doi.org/10.1177/21582440251335734.

57.

Zulkafli

A. H.

(2024a). How does digital transformation affect innovation quality? E&M Economics and Management, 27(4), 16–32. https://doi.org/10.15240/tul/001/2024-5-021

58.

Zulkafli

A. H.

(2024b). Managerial ownership and corporate innovation: Evidence of patenting activity from Chinese listed manufacturing firms. Cogent Business & Management, 11(1), https://doi.org/10.1080/23311975.2023.2289202

59.

Queiroz

M. M.

Fosso Wamba

(2019). Blockchain adoption challenges in supply chain: An empirical investigation of the main drivers in India and the USA. International Journal of Information Management, 46, 70–82. https://doi.org/10.1016/j.ijinfomgt.2018.11.021

60.

Rachinger

Rauter

Müller

Vorraber

Schirgi

(2019). Digitalization and its influence on business model innovation. Journal of Manufacturing Technology Management, 30(8), 1143–1160. https://doi.org/10.1108/jmtm-01-2018-0020

61.

Radonić

Milosavljević

Knežević

(2021). Intangible Assets as financial performance drivers of IT industry: Evidence from an emerging market. E&M Ekonomie a Management, 24(2), 119–135. https://doi.org/10.15240/tul/001/2021-2-008

62.

Raj

Kumar

Ranjana

Anirvinna

(2024). Existing and forthcoming obstacles in adopting technological advances in vulnerable supply chains. E&M Ekonomie a Management, 27(3), 88–10. https://doi.org/10.15240/tul/001/2024-3-006

63.

Rakovic

Maric

Milutinovic

L. D.

Vukovic

Bjekic

(2024). The role of leadership in managing digital transformation: A systematic literature review. E&M Ekonomie a Management, 27(2), 87–107. https://doi.org/10.15240/tul/001/2024-2-006

64.

Reuschl

A. J.

Deist

M. K.

Maalaoui

(2022). Digital transformation during a pandemic: Stretching the organizational elasticity. Journal of Business Research, 144, 1320–1332. https://doi.org/10.1016/j.jbusres.2022.01.088

65.

Rodrigues

Samagaio

Felício

(2020). Corporate governance and R&D investment by European listed companies. Journal of Business Research, 115, 289–295. https://doi.org/10.1016/j.jbusres.2019.11.070

66.

Sergei

Arkady

Natalya

Pathak

R. D.

Samson

Husain

Sushil

(2023). Digital transformation enablers in high-tech and low-tech companies: A comparative analysis. Australian Journal of Management, 48(4), 801–843. https://doi.org/10.1177/03128962231157102

67.

Secundo

Spilotro

Gast

Corvello

(2024). The transformative power of artificial intelligence within innovation ecosystems: A review and a conceptual framework. Review of Managerial Science. Advance online publication. https://doi.org/10.1007/s11846-024-00828-z.

68.

Sharma

S. K.

Chanda

(2017). Developing a Bayesian belief network model for prediction of R&D project success. Journal of Management Analytics, 4(3), 321–344. https://doi.org/10.1080/23270012.2017.1304291

69.

Simsek

Vaara

Paruchuri

Nadkarni

Shaw

J. D.

(2019). New ways of seeing big data. Academy of Management Journal, 62(4), 971–978. https://doi.org/10.5465/amj.2019.4004

70.

Srinivasan

Venkatraman

(2018). Entrepreneurship in digital platforms: A network-centric view. Strategic Entrepreneurship Journal, 12(1), 54–71. https://doi.org/10.1002/sej.1272

71.

Stacho

Stachová

Lukáč

Kupec

Petrů

(2023). Management Reflections on Innovations in Digitalization, with an emphasis on degree of work autonomy. E&M Ekonomie a Management, 26(1), 78–93. https://doi.org/10.15240/tul/001/2023-1-005

72.

Sunder

S. V.

Zhang

(2017). Pilot CEOs and corporate innovation. Journal of Financial Economics, 123(1), 209–224. https://doi.org/10.1016/j.jfineco.2016.11.002

73.

Tang

Liu

(2023). Effect of digital transformation on enterprises’ green innovation: Empirical evidence from listed companies in China. Energy Economics, 128, 107135. https://doi.org/10.1016/j.eneco.2023.107135

74.

Teece

D. J.

(2007). Explicating dynamic capabilities: The nature and microfoundations of (sustainable) enterprise performance. Strategic Management Journal, 28(13), 1319–1350. https://doi.org/10.1002/smj.640

75.

Teece

D. J.

Pisano

Shuen

(1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509–533. https://doi.org/10.1002/(sici)1097-0266(199708)18:73.0.co;2-z

76.

Tian

Cheng

(2022). Does digital transformation matter for corporate risk-taking? Finance Research Letters, 49, 103107. https://doi.org/10.1016/j.frl.2022.103107

77.

Toader

Paraschiv

Dinu

Manea

Mihai

(2023). The effects of private sector companies’ research and development investments on the adoption of cloud computing services in the European Union. E&M Ekonomie a Management, 26(2), 189–202. https://doi.org/10.15240/tul/001/2023-2-012

78.

(2023). Can Digital Transformation facilitate firms’ M&A: Empirical Discovery based on machine learning. Emerging Markets Finance and Trade, 59(1), 113–128. https://doi.org/10.1080/1540496x.2022.2093105

79.

Urbinati

Chiaroni

Chiesa

Frattini

(2020). The role of digital technologies in open innovation processes: An exploratory multiple case study analysis. R and D Management, 50(1), 136–160. https://doi.org/10.1111/radm.12313

80.

Verhoef

P. C.

Broekhuizen

Bart

Bhattacharya

Dong

J. Q.

Fabian

N. E.

Haenlein

(2021). Digital transformation: A multidisciplinary reflection and research agenda. Journal of Business Research, 122, 889–901. https://doi.org/10.1016/j.jbusres.2019.09.022

81.

Vial

(2019). Understanding digital transformation: A review and a research agenda. Journal of Strategic Information Systems, 28(2), 118–144. https://doi.org/10.1016/j.jsis.2019.01.003

82.

Wang

Zhou

Huang

(2022). Green credit policy, government behavior and green innovation quality of enterprises. Journal of Cleaner Production, 331, 129834. https://doi.org/10.1016/j.jclepro.2021.129834

83.

Wang

(2024). Too much incentive to innovate? CEO stock option exercise and myopic R&D management. Journal of Product Innovation Management, 41, 1141–1164. https://doi.org/10.1111/jpim.12731

84.

Warner

K. S. R.

Wäger

(2019). Building dynamic capabilities for digital transformation: An ongoing process of strategic renewal. Long Range Planning, 52(3), 326–349. https://doi.org/10.1016/j.lrp.2018.12.001

85.

Xie

Wang

Miao

(2021). Big data and emerging market firms’ innovation in an open economy: The diversification strategy perspective. Technological Forecasting and Social Change, 173, 121091. https://doi.org/10.1016/j.techfore.2021.121091

86.

Yuan

Wen

(2018). Managerial foreign experience and corporate innovation. Journal of Corporate Finance, 48, 752–770. https://doi.org/10.1016/j.jcorpfin.2017.12.015

87.

Zhou

Wang

Sang

Huang

(2023). Carbon emission reduction analysis for cloud computing industry: Can carbon emissions trading and technology innovation help? Energy Economics, 125, 106804. https://doi.org/10.1016/j.eneco.2023.106804

88.

Zhang

Liang

X. J.

(2012). Promoting green ICT in China: A framework based on innovation system approaches. Telecommunications Policy, 36(10–11), 997–1013. https://doi.org/10.1016/j.telpol.2012.09.001

89.

Zhang

Xiong

Deng

(2022). Political connection heterogeneity and corporate innovation. Journal of Innovation & Knowledge, 7(3), 100224. https://doi.org/10.1016/j.jik.2022.100224

90.

Zheng

Chen

(2024). Optimization of inventory cost control for SMEs in supply chain transformation: A case study and discussion. E&M Ekonomie a Management, 27(1), 87–107. https://doi.org/10.15240/tul/001/2024-5-002

91.

Zhuo

Chen

(2023). Can digital transformation overcome the enterprise innovation dilemma: Effect, mechanism and effective boundary. Technological Forecasting and Social Change, 190, 122378. https://doi.org/10.1016/j.techfore.2023.122378

How Digital Transformation Shapes Corporate R&D Expenditure: An Exploration of Multidimensional Perspectives and Innovation Consequences

Abstract

Keywords

Introduction

Literature Review and Hypothesis Development

Literature Review

Hypothesis Development

The Relationship Between Digital Transformation and Corporate R&D Expenditure

The Relationship Between the Sub-dimensions of Digital Transformation and Corporate R&D Expenditure

The Role of Digital Transformation on the Conversion of Corporate R&D Expenditure Into Innovation Outcomes

The Agency Theory Perspective

The Dynamic Capabilities Perspective

Data and Methodology

Data and Sample

Variables

Dependent Variable

Independent Variable

Control Variables

Empirical Specification and Data Analysis Techniques

Empirical Results and Discussion

Descriptive Statistics

Correlation and Variance Inflation Factor Analysis

Univariate Analysis

Multivariate Results

Robustness Test

Endogenous Treatment

Placebo Test

Extended Analysis

Dimensions of Digital Transformation

Moderating Effect of Digital Transformation on Innovation Consequences

Discussion

Conclusions

Footnotes

Acknowledgements

ORCID iD

Funding

Declaration of Conflicting Interests

Data Availability Statement

References