The Driving Forces of Digital Transformation: Navigating Peer Effects in Industrial and Regional Ecosystems

Abstract

Understanding the systemic synergy of peer effects on digital transformation is essential for overcoming development bottlenecks and stimulating digital vitality across industrial and regional ecosystems. Utilizing data from Chinese A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2010 to 2024, this study empirically investigates the impact of peer effects on corporate digital transformation and its underlying influencing factors from a systems perspective. The findings reveal significant industry and regional peer effects in corporate digital transformation, indicating that firms’ decision-making is interdependent within broader ecosystems. A greater distance between a focal firm’s prior digital transformation level and that of its peers is associated with a higher level of enthusiasm for such transformation. Similarly, the more a focal firm’s prior performance falls below that of its peers, the stronger its impetus for digital transformation becomes. Furthermore, the influence of the transformation distance on digital transformation enthusiasm exhibits a non-linear threshold effect, which varies with the performance gap. Finally, further analysis indicates that peer effects exert a multiplier effect and that industry-level peer effects in digital transformation significantly enhance firm performance. These conclusions contribute to a deeper understanding of the systemic mechanisms and pathways of corporate digital transformation and offer both theoretical and empirical support for fostering resilient digital economic ecosystems across industries and regions.

1. Introduction

In the digital economy era, digital transformation has become imperative for business operations [1]. The “Overall Layout Plan for Digital China Construction,” ref. [2] issued in 2024 by the Central Committee of the Communist Party of China and the State Council, explicitly aims for China’s level of digital development to be among the world’s highest by 2035. However, enterprises’ current digital transformation efforts are hampered by uneven progress and a widening gap across industries and regions, presenting significant challenges to achieving these goals [3]. From an enterprise decision-making perspective, digital transformation, a long-term strategic initiative, faces high uncertainty in areas like risk assessment, effectiveness measurement, and process evaluation [4]. To mitigate risks associated with strategy implementation and overcome transformation obstacles, enterprises often benchmark against the behaviors and decisions of their peers when formulating digital transformation strategies [5]. Specifically, focal firms (the specific entity whose decisions are the primary subject of our analysis) imitate and learn from the transformation paths of peer firms [6], thereby accelerating their own digitalization process through successful cases [7] and enhancing their competitive advantage [5]. This widespread benchmarking practice highlights a key insight: the actions of peer firms constitute a fundamental decision-making context for managers, reflecting a form of distributed and ecosystem-level leadership. Rather than relying solely on internal strategic formulation, firms are increasingly influenced by pioneering actors within their industrial and regional ecosystems. These leading enterprises, often unintentionally, set reference points and establish norms that guide the strategic choices of other firms, thereby exerting a form of emergent leadership at the system level. This phenomenon raises an important research question: Can the “peer effect”—defined as the phenomenon whereby a firm’s behavior or decisions are influenced by the actions and characteristics of other firms in its reference group [8]—serve as an underlying mechanism within broader digital ecosystems that facilitates digital transformation and helps alleviate systemic imbalances in digital development across industries and regions?

Current research on enterprise digital transformation is still emerging, primarily focusing on the impact of digitalization on the firms themselves. Studies have demonstrated that digital transformation can significantly foster innovation [9,10], enhance total factor productivity [11], and improve ESG performance [12]. These findings collectively confirm the positive impact of digital transformation on firms themselves. However, the systemic external effects generated during the digitalization process have received insufficient attention and discussion. Owing to factors such as incomplete information, competitive pressure, and social norms and recognition, firms are often influenced by their peers when making digital transformation decisions [7]. Nevertheless, scholarly research on the peer effects of digital transformation remains limited and requires further development. Existing scholarly work has confirmed the presence of peer effects in corporate digital transformation [13,14]. Zhang and Du examined the moderating role of Top Management Team (TMT) characteristics in this process [15]. Furthermore, regarding the firm-level impacts of digital transformation peer effects, studies have also found that such effects not only significantly enhance the efficiency of corporate investment decisions [16] but also improve firms’ innovation performance, carbon neutrality performance, and environmental performance [6]. However, the mechanisms and contingent factors underlying peer-effects-driven digital transformation require further exploration. First, existing research predominantly examines the impact of peers’ digital transformation levels on focal firms’ transformation from an absolute perspective, overlooking the relative incentive effect arising from the transformation distance, a key dynamic in system interactions. Second, related studies have not integrated the performance feedback mechanism and particularly lack analysis of the non-linear interaction between the transformation distance and performance discrepancies. Finally, discussion of regional-level peer effects in digital transformation is notably lacking, and there is a severe shortage of multi-level systems studies that consider both industry and geographical dimensions.

To address these research gaps, this study utilizes data from Chinese A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2010 to 2024. Aligned with the “digital collaboration” goals outlined in the “14th Five-Year Plan for Digital Economy Development,” ref. [17] it investigates the effects and influencing factors of the peer effect on enterprise digital transformation through a systems lens. By constructing a dynamic analytical framework grounded in systems thinking and incorporating two key constructs derived from social comparison theory: the transformation distance (the absolute disparity in digital transformation levels between a firm and its peers) and performance feedback (signaled by the performance gap, i.e., the difference between a firm’s performance and the peer average), this paper focuses on examining the existence of dual-dimensional (industry and region) peer effects and the factors influencing how these effects drive digitalization. Specifically, it investigates the incentive effect of the transformation distance on transformation motivation, the non-linear threshold effect induced by performance feedback, and the multiplier effect (a positive feedback loop) of peer effects in digital collaboration.

The marginal contributions of this paper are threefold. First, it moves beyond the absolute perspective to adopt a relative viewpoint, thereby introducing and validating the concept of “transformation distance” as a novel driver of digital transformation. This helps enrich the research on digital transformation mechanisms based on peer effects. Second, in a departure from the existing literature, it integrates performance feedback theory with threshold effect analysis, providing the first empirical verification of the moderating role of firm performance discrepancies on the transformation distance effect and uncovering the associated threshold rule. Third, it empirically tests the collaborative enhancement mechanism through which peer effects elevate the digitalization level of firms, offering a more comprehensive systemic understanding than prior single-level studies.

2. Theoretical Analysis and Hypothesis Construction

2.1. The Peer Effects of Enterprise Digital Transformation

In today’s interconnected systems, strategic decisions are rarely made in isolation. The digital transformation of a firm is embedded in an open, complex, multi-level ecosystem. While subject to various influences, the impact from peer firms is paramount, serving as the most direct and relatable driver in the face of high uncertainty. Peer firms provide a critical, observable reference point. Their successful practices offer concrete “social proof,” which more effectively lowers decision-making uncertainty and perceived risk than abstract policy or vendor information. Thus, strategic choices are profoundly shaped by peer behaviors and leadership signals—the phenomenon of peer effects [8]. To mitigate risks and gain competitive advantage, firms imitate these successful market behaviors [18], leading to strategic convergence through the core mechanisms of competitive learning and institutional isomorphism [5,15]. Within complex organizational systems, competitive learning is a dynamic process through which firms observe, analyze, and judiciously imitate the successful practices of competitors. This process enables them to enhance their competitiveness while simultaneously seeking innovation and differentiation to adapt to market changes. Institutional isomorphism emphasizes behavioral consistency among firms as they respond to external environmental pressures, creating emergent patterns of collective behavior across the system. To meet regulatory requirements, industry standards, and social expectations, firms often adopt similar organizational structures and behavioral patterns [15]. This conformity not only facilitates integration into a broader business ecosystem but also effectively mitigates the risks associated with deviating from established norms. This tendency is particularly evident in the complex and uncertain context of digital transformation, where firms are more likely to observe the practices of their peers—particularly in areas like information technology application and business process optimization—to reduce operational uncertainty and risk [7]. This imitative behavior not only enables firms to assimilate and apply new technologies rapidly but also helps secure their market competitiveness.

Firms within the same industry typically offer similar products or services and operate under comparable market environments and technical conditions [19]. When pioneering firms within an industry achieve success through digital transformation, they establish a strategic leadership position and set a de facto standard. Others often emulate these leading practices not only to maintain competitiveness but also to follow the strategic direction validated by these market leaders. Moreover, digital transformation is often characterized by long cycles and high market information uncertainty. In this context, gathering high-quality information for decision-making is inherently time-consuming and labor-intensive. The digital transformation initiatives of industry peers can provide valuable decision-making benchmarks for firms contemplating such a transition, thereby effectively reducing associated risks and costs [20].

Furthermore, firms within the same region are exposed to similar regional economies, cultures, local policies, and regulations [5,19] and are consequently subject to analogous economic cycles. Geographical proximity serves as a natural conduit that facilitates knowledge spillover [21], lowers the cost of information acquisition, and accelerates information exchange among firms [22]. This proximity fosters both competition and cooperation among firms, thereby facilitating the sharing of knowledge and best practices. When firms in the same region embark on digital transformation, it creates a demonstration effect for focal firms, compelling them to adopt similar strategies to maintain their competitive standing. Moreover, national policies such as the “14th Five-Year Plan for Digital Economy Development” [17] and the “Overall Layout Plan for Digital China Construction” [2] underscore the strategic importance of digital transformation. Local governments often respond by creating a favorable external environment for the digital economy through incentives, supportive policies, and institutional frameworks. Successful digital transformation by leading enterprises can create regional competitive advantages. Their progress sets the pace for the entire region, strengthening the local digital culture and initiating a virtuous cycle that draws broader participation into the regional innovation ecosystem. Based on the above analysis, we propose the following core hypotheses:

H1a: 

An industry peer effect exists in digital transformation; specifically, the level of digital transformation among firms within the same industry positively influences that of the focal firm.

H1b: 

A regional peer effect exists in digital transformation; specifically, the level of digital transformation among firms within the same region positively influences that of the focal firm.

2.2. The Impact of Prior Transformation Distance on Digital Transformation

Social decision-making theory suggests that individuals operating within a social context are influenced by relevant others and adjust their behaviors based on group actions to enhance decision-making quality and efficiency [23]. As boundedly rational entities, firms engage in comparisons with similar firms to assess their competitiveness and strategic direction [24]. In the context of digital transformation, this comparative process encompasses not only current digitalization levels but also anticipations of and responses to future trends. Firms compare the prior digital transformation status of their peers and adjust their current strategies accordingly, potentially giving rise to either a catch-up effect or a lead-maintenance effect.

Specifically, when a focal firm’s digital transformation level lags behind that of its peers, a significant gap from the previous period generates competitive pressure. This pressure is then translated into motivation, driving the firm to accelerate its transformation process to catch up with or even surpass its competitors [5]. Conversely, when a focal firm’s transformation level leads that of its peers, a substantial historical gap provides an opportunity for continuous innovation and system optimization. This prompts increased investment to reduce costs, improve efficiency, consolidate market position, and enhance customer satisfaction. From the perspectives of learning curves and cost-effectiveness, a significant transformation distance implies that firms have greater opportunities to acquire valuable external information and experience. Followers can learn from the successes of leaders, reducing trial-and-error costs and improving transformation efficiency. Conversely, leaders are driven to continuously innovate and optimize, enhancing system operational efficiency, achieving value addition, and maintaining their market leadership position.

Furthermore, from a behavioral economics perspective, a significant transformation distance can trigger psychological effects among decision-makers, including social identity, overconfidence, and loss aversion. These psychological factors can lead to a higher risk preference in transformation decisions, manifesting as an overestimation of potential benefits and an underestimation of associated risks. Additionally, firms generally seek to maintain a positive reputation and social identity. Consequently, a substantial gap between a focal firm and its peers typically triggers actions aimed at maintaining or enhancing its status. The concept of reference point dependence further emphasizes that firms’ decision-making is profoundly shaped by their chosen benchmarks [24]. In digital transformation, the digitalization level of peers serves as a critical reference point, profoundly influencing a focal firm’s perceptions and actions, whether aimed at avoiding losses or maintaining gains. If the focal firm’s transformation level falls below this benchmark, loss aversion motivates active promotion of digital transformation to avoid falling behind. Conversely, if the level exceeds the benchmark, a desire to maintain gains drives continued investment to preserve and expand advantages, and consolidate leadership. Therefore, irrespective of whether a focal firm’s prior digital transformation level was higher or lower than that of its peers, a substantial historical gap exerts a positive influence on its current-period enthusiasm for digital transformation. Based on this reasoning, we propose Hypothesis H2:

H2: 

A positive relationship exists between the prior period’s transformation distance and the focal firm’s current-period enthusiasm for digital transformation. Specifically, the greater the gap between the focal firm’s prior digital transformation level and that of its peers, the higher its enthusiasm for transformation.

2.3. The Impact of Prior Performance Gap on Digital Transformation

Performance feedback theory posits that firms adjust their strategies and behaviors in response to performance. When a focal firm’s prior performance falls below the average of its peer group, this negative performance feedback often triggers introspection into the causes of the performance gap, stimulating more proactive digital transformation initiatives aimed at performance improvement [25,26]. Firms typically use the performance level of their peer group as a reference point. When a firm’s performance falls below this benchmark, loss aversion psychology tends to emerge, driving a search for change to mitigate further losses. Both risk-avoidance and risk-seeking behaviors play significant roles in this process.

Specifically, underperforming firms may exhibit a greater propensity for risk-taking. They perceive that the potential losses from maintaining a conservative strategy at their current performance level are substantial. Consequently, they demonstrate a greater willingness to explore innovative paths, pursue potential opportunities, and seek improvement measures [27]—often adopting high-risk strategic initiatives to overcome performance predicaments [28]. Digital transformation presents a viable opportunity for such firms to achieve rapid performance improvement [26]. In contrast, well-performing firms, already in an advantageous position with stable revenues and strong market performance, are often reluctant to disrupt the status quo. These firms tend to be more risk-averse [29] and consequently adopt a more cautious approach toward digital transformation.

Furthermore, a firm’s risk preference is influenced not only by its current performance but also by the reference point chosen for comparison with its peer group. For underperforming firms, the high performance of the peer group represents an aspirational benchmark. This disparity may stimulate the pursuit of riskier strategies [27,30] in hopes of achieving significant performance improvement. Consequently, they may actively pursue high-risk digital transformation initiatives. The opposite dynamic holds for top-performing firms. Based on this reasoning, we propose Hypothesis H3:

H3: 

A negative performance gap in the prior period (focal firm performance below peers) strengthens the current-period digital transformation tendency of the focal firm, whereas a positive performance gap (focal firm performance above peers) weakens it.

2.4. Threshold Effect

A significantly negative performance gap (i.e., focal firm performance substantially below peers) typically indicates a relatively weak market position and significant competitive pressure [31]. The pressure stemming from this performance lag stimulates a strong catch-up motivation. In this context, a larger transformation distance significantly heightens the firm’s enthusiasm for digital transformation. The substantial performance gap amplifies the perceived urgency for organizational change, prompting decision-makers to aggressively adopt peers’ digital practices to rapidly narrow the gap. Under these conditions, the influence of the prior transformation distance on digital transformation enthusiasm is most pronounced.

As the negative performance gap narrows (i.e., the gap persists but is diminishing), the competitive pressure to catch up eases relatively, and management’s tendency for risk aversion in digital transformation begins to emerge. Although still at a disadvantage, firms may believe that gradual, incremental performance improvements can be achieved without rushing into large-scale digital transformation. Consequently, the influence of the prior transformation distance on transformation enthusiasm weakens. Decision-makers may adopt peers’ digital practices in a more gradual, asymptotic manner, leading to a nonlinear relationship between the prior transformation distance and transformation enthusiasm. When the performance gap becomes positive, some leading firms may exhibit organizational inertia [32,33], believing current performance is sufficient and large-scale transformation is unnecessary to consolidate their advantages. These firms are more inclined to maintain their current technological trajectory, thereby reducing the incentive effect of the transformation distance on digital transformation.

Based on this analysis, we propose the following hypothesis to explore the threshold effect of the prior transformation distance on digital transformation enthusiasm under varying prior performance gaps and to verify the nonlinear relationship:

H4: 

The influence of the prior transformation distance on digital transformation enthusiasm exhibits a threshold effect that depends on the magnitude and sign (positive/negative) of the prior performance gap.

Based on the theoretical reasoning, we develop a conceptual framework (Figure 1) that integrates all the hypothesized relationships. This model not only summarizes the direct peer effects (H1a and H1b) but also illustrates the roles of transformation distance (H2), performance gap (H3), and their interaction through the threshold effect (H4).

3. Research Design

3.1. Sample Selection and Data Sources

Data were sourced from Chinese A-share listed companies spanning the period from 2010 to 2024. The data were obtained from the China Stock Market & Accounting Research (CSMAR) database and subsequently filtered and processed based on the following criteria: (1) Missing data points were supplemented using information compiled from the annual reports of the respective listed companies; (2) Firms under special treatment (ST), those in the financial industry, and those with missing key variable data were excluded; (3) To mitigate the influence of outliers, all continuous variables were winsorized at the 1st and 99th percentiles. This process resulted in a final sample of 49,490 firm-year observations.

3.2. Variable Definitions

3.2.1. Dependent Variable

Digital Transformation (DT). Following Chang et al. [16], we measure digital transformation by constructing a feature word list based on their framework of five core technologies: artificial intelligence, blockchain, cloud computing, big data, and digital technology applications. The frequency of these keywords appearing in a firm’s annual report is used to measure its level of digital transformation. To address the right-skewed distribution of the raw word counts, we add 1 to the total frequency for each firm and then take the natural logarithm.

3.2.2. Independent Variables

The independent variables include: industry peer digital transformation (IDT), regional peer digital transformation (ADT), prior-period transformation distance (JL), and prior-period performance gap (CZ). Following Yang et al. [34] and Grennan [35], we define industry peers as other firms within the same industry as the focal firm, and regional peers as other firms headquartered in the same province (the regional level is defined as the provincial level). Accordingly, IDT (ADT) is measured as the average digital transformation level of all industry (regional) peers in the current fiscal year (t), excluding the focal firm itself. Furthermore, building on the work of Ba et al. [24], JL is defined as the absolute value of the difference in digital transformation between the focal firm and the average of its peers in the preceding fiscal year (t − 1). Similarly, CZ is calculated as the focal firm’s performance in year t − 1 minus the average prior-period performance of its peers in year t − 1. Firm performance is measured by Tobin’s Q, which captures the firm’s market value relative to its asset replacement cost. This market-based metric is chosen for its forward-looking nature and ability to reflect dynamic market expectations, aligning well with the long-term strategic character of digital transformation initiatives.

3.2.3. Control Variables

Consistent with prior literature [36,37], we include a set of control variables: asset-liability ratio (ALR), equity balance ratio (Balance), proportion of independent directors (Indep), CEO duality (Com), total compensation of the top three executives (Pay), ownership concentration (OC), and revenue growth rate (Growth). Furthermore, industry, year, and province fixed effects are included to control for time-invariant unobserved heterogeneity.

Table 1. The definitions of variables.

Variable Type	Variable Name	Symbol	Definition
Dependent Variable	Digital Transformation	DT	Natural logarithm of (word frequency of digital transformation in annual report + 1)
Independent Variables	Industry Peer Digital Transformation	IDT	Average digital transformation level of other firms in the same industry as the focal firm
	Regional Peer Digital Transformation	ADT	Average digital transformation level of other firms in the same region as the focal firm
	prior-period transformation distance	IJL	Absolute difference in digital transformation between focal firm and industry peers in previous period
		AJL	Absolute difference in digital transformation between focal firm and regional peers in previous period
	prior-period performance gap	ICZ	Focal firm’s previous period performance minus average performance of industry peers
		ACZ	Focal firm’s previous period performance minus average performance of regional peers
Control Variables	Asset-Liability Ratio	ALR	Total liabilities divided by total assets
	Equity Balance Ratio	Balance	Shareholding ratio of 2nd–5th largest shareholders divided by shareholding ratio of largest shareholder
	Proportion of Independent Directors	Indep	Number of independent directors divided by total number of directors
	CEO duality	Com	Dummy variable indicating whether chairman and CEO are the same person (1 = yes, 0 = no)
	Total Compensation of Top Three Executives	Pay	Total annual compensation of three highest-paid executives
	Ownership Concentration	OC	Shareholding percentage of largest shareholder
	Revenue Growth Rate	Growth	(Current year operating revenue—previous year operating revenue)/previous year operating revenue

presents the specific definitions of all variables used in this study.

3.3. Model Construction

To test hypotheses H1a and H1b, we estimate the following baseline regression models, presented as Equations (1) and (2): Equation (1) includes year and province fixed effects, whereas Equation (2) includes year and industry fixed effects.

DT = α₀ + α₁IDT_i,t + α₂Controls_i,t + year + province + ε_i,t

(1)

DT = α₀ + α₁ADT_i,t + α₂Controls_i,t + year + industry + ε_i,t

(2)

To test hypothesis H2, a regression model as shown in Equations (3) and (4) was constructed.

DT = β₀ + β₁IJL_i,t + β₂Controls_i,t + year + province + ε_i,t

(3)

DT = β₀ + β₁AJL_i,t + β₂Controls_i,t + year + industry + ε_i,t

(4)

To test hypothesis H3, a regression model as shown in Equations (5) and (6) was constructed.

DT = γ₀ + γ₁ICZ_i,t + γ₂Controls_i,t + year + province + ε_i,t

(5)

DT = γ₀ + γ₁ACZ_i,t + γ₂Controls_i,t + year + industry + ε_i,t

(6)

The following Equations (7) and (8) represent the threshold model. The purpose is to test Hypothesis H4. Mit is the threshold variable for firm i in year t, γ_n is the nth threshold value, and I(∙) is the indicator function. If the expression within the parentheses is true, it takes the value 1; otherwise, it takes the value 0.

DT = θ₀ + θ₁₁IJL_i,t × I(M_i,t ≤ γ₁) + θ₁₂IJL_i,t × I(γ₁ < M_i,t ≤ γ₂) + ⋯⋯ + θ_1nIJL_i,t × I(γ_n−1 < M_i,t ≤ γ_n) + θ_1(n+1)IJL_i,t × I(M_i,t > γ_n) + θ₂Controls_i,t + ε_it

(7)

DT = θ₀ + θ₁₁AJL_i,t × I(M_i,t ≤ γ₁) + θ₁₂AJL_i,t × I(γ₁ < M_i,t ≤ γ₂) + ⋯⋯ + θ_1nAJL_i,t × I(γ_n−1 < M_i,t ≤ γ_n) + θ_1(n+1) AJL_i,t × I(M_i,t > γ_n) + θ₂Controls_i,t + ε_it

(8)

Among them, α₀, β₀, γ₀, θ₀ are intercept terms, Controls are control variables, α₁, α₂, β₁, β₂, γ₁, γ₂, θ₁₁ to θ_1(n+1), and θ₂ are the coefficients of each variable. The subscript i represents the company, and the subscript t represents the time. ε is the error term.

4. Empirical Analysis

4.1. Industry and Regional Peer Effects

Table 2. Test results of industry and regional peer effects in digital transformation.

Variable	(1) DT	(2) DT	(3) DT	(4) DT
IDT	0.9335 *** (155.8589)	0.9332 *** (154.6431)
ADT			0.3470 *** (21.1221)	0.3145 *** (19.0310)
Controls	NO	YES	NO	YES
_cons	0.1042 *** (4.7570)	−0.0430 (−1.0017)	0.0285 (0.5064)	−0.0899 (−1.3048)
N	49,490	49,490	49,490	49,490
R2	0.5161	0.5209	0.4864	0.4914
year	YES	YES	YES	YES
province	YES	YES	NO	NO
industry	NO	NO	YES	YES

Note: T-statistics are reported in parentheses. Asterisks *** denote statistical significance at the 1% level.

reports the estimation results for industry and regional peer effects on digital transformation. Column (1) controls for province and year fixed effects, Column (3) controls for industry and year fixed effects, while Columns (2) and (4) further incorporate the full set of control variables. The coefficients on IDT in columns (1) and (2) are 0.9335 and 0.9332, respectively, both statistically significant at the 1% level. This suggests that a higher level of digital transformation among industry peers is associated with a higher level of digital transformation in the focal firm, supporting the existence of an industry peer effect and validating Hypothesis H1a. Similarly, the coefficients on ADT in columns (3) and (4) are 0.3470 and 0.3145, respectively, and are both significant at the 1% level. This result indicates a positive regional peer effect, thereby confirming Hypothesis H1b.

4.2. Analysis of the Impact of Transformation Distance and Performance Gap on Digital Transformation

Table 3. The Impact of Prior-Period Transformation Distance on Digital Transformation.

Variable	(1) DT	(2) DT	(3) DT	(4) DT
IJL	0.4683 *** (44.3067)	0.4554 *** (43.1412)
AJL			0.3460 *** (40.0025)	0.3468 *** (40.2355)
Controls	NO	YES	NO	YES
_cons	0.4662 *** (17.8168)	0.4124 *** (7.6297)	0.0755 (1.2668)	−0.1453 ** (−2.0041)
N	44,208	44,208	44,208	44,208
R2	0.2897	0.3013	0.5075	0.5138
year	YES	YES	YES	YES
province	YES	YES	NO	NO
industry	NO	NO	YES	YES

Note: T-statistics are reported in parentheses. Asterisks ** and *** denote statistical significance at the 5% and 1% levels, respectively.

presents the results examining the impact of the prior-period transformation distance on current-period digital transformation. As shown in columns (1) to (4), the coefficients on the industry peer transformation distance are 0.4683 and 0.4554, and those on the regional peer transformation distance are 0.3460 and 0.3468, in specifications without and with control variables, respectively. All coefficients are statistically significant at the 1% level. These positive coefficients imply that a larger transformation distance in the prior period stimulates greater digital transformation efforts by the focal firm in the current period. Specifically, a narrower absolute gap between the focal firm’s digital transformation level and that of its peers corresponds to a weaker propensity for digital transformation, whereas a wider absolute gap is associated with a stronger transformation impetus. These findings provide robust empirical support for Hypothesis H2.

Table 4. The Impact of Prior-Period Performance Gap on Digital Transformation.

Variable	(1) DT	(2) DT	(3) DT	(4) DT
ICZ	−0.0254 *** (−5.1973)	−0.0387 *** (−7.8352)
ACZ			−0.0285 *** (−7.2943)	−0.0264 *** (−6.6300)
Controls	NO	YES	NO	YES
_cons	0.6719 *** (23.4795)	0.6389 *** (11.3435)	0.1791 ** (2.3890)	0.0030 (0.0346)
N	44,144	44,144	44,144	44,144
R2	0.2471	0.2619	0.4811	0.4871
year	YES	YES	YES	YES
province	YES	YES	NO	NO
industry	NO	NO	YES	YES

Note: T-statistics are reported in parentheses. Asterisks ** and *** denote statistical significance at the 5% and 1% levels, respectively.

examines the influence of the prior-period performance gap on current-period digital transformation. The results reveal significantly negative coefficients (at the 1% level) for both the industry and regional performance gaps across all specifications. This suggests that a larger prior-period performance gap inhibits the focal firm’s digital transformation in the current period. Specifically, focal firms that outperformed their peers in the prior period show a weaker propensity for digital transformation, while those that underperformed demonstrate a stronger motivation to catch up through transformation. Thus, Hypothesis H3 is supported.

4.3. Endogeneity Test

4.3.1. Lagged One-Period Core Independent Variable

To mitigate potential reverse causality between peer and focal firm digital transformation, we lag all core independent variables by one period and re-estimate Equations (1) and (2). The results are presented in

Table 5. Regression Results of Core Explanatory Variables with One-Period Lag.

Variable	(1) DT	(2) DT
L.IDT	0.8156 *** (125.2778)
L.ADT		0.1947 *** (11.9593)
Controls	YES	YES
_cons	0.0517 (1.0869)	−0.0292 (−0.4035)
N	44,208	44,208
R2	0.4756	0.4885
year	YES	YES
province	YES	NO
industry	NO	YES

Note: T-statistics are reported in parentheses. Asterisks *** denote statistical significance at the 1% level.

. As shown in columns (1) and (2), the coefficients on lagged IDT and ADT remain positive and statistically significant at the 1% level. This suggests that our main findings regarding industry and regional peer effects are robust to controlling for reverse causality.

4.3.2. Propensity Score Matching

We further employ a propensity score matching (PSM) approach to address endogeneity concerns stemming from sample self-selection. Following Li et al. [38], we use one-to-one nearest neighbor matching without replacement. The matching covariates include the debt-to-asset ratio, equity balance ratio, the proportion of independent directors, CEO duality, the total compensation of the top three executives, ownership concentration, and revenue growth rate. Firms are classified into treatment and control groups based on whether their values of the core independent variables are above or below the sample mean. The matched samples are then used for regression analysis. The results are reported in

Table 6. Regression Results Based on Propensity Score Matching (PSM).

Variable	(1) DT	(2) DT	(3) DT	(4) DT	(5) DT	(6) DT
IDT	0.9285 *** (115.9056)
ADT		0.2908 *** (11.9177)
IJL			0.4712 *** (33.6778)
AJL				0.3159 *** (25.7655)
ICZ					−0.0390 *** (−5.9910)
ACZ						−0.0294 *** (−5.2782)
Controls	YES	YES	YES	YES	YES	YES
_cons	0.0391 (0.6194)	−0.0665 (−0.7059)	0.2800 *** (3.7438)	−0.0819 (−0.8694)	0.6330 *** (7.6781)	0.0280 (0.2768)
N	25,816	24,229	24,057	23,898	21,665	20,807
R2	0.5088	0.4801	0.2994	0.5008	0.2598	0.4959
year	YES	YES	YES	YES	YES	YES
province	YES	NO	YES	NO	YES	NO
industry	NO	YES	NO	YES	NO	YES

Note: T-statistics are reported in parentheses. Asterisks *** denote statistical significance at the 1% level.

. Columns (1) to (6) in Table 6 correspond to the re-estimation of our main models (peer effects, transformation distance, and performance gap) using the PSM-matched sample. The results obtained using PSM are consistent with our baseline findings, providing further support for our hypotheses.

4.4. Robustness Test

To ensure the robustness of our findings, we conduct several additional tests.

4.4.1. Controlling for Additional Variables

Drawing on corporate governance and resource dependence theories [39,40], board size can influence a firm’s strategic decisions and resource allocation, potentially affecting its digital transformation initiatives. To mitigate potential omitted variable bias, we include board size as an additional control variable. The results, presented in

Table 7. Robustness Check: Addressing Potential Omitted Variable Bias.

Variable	(1) DT	(2) DT	(3) DT	(4) DT	(5) DT	(6) DT
IDT	0.9329 *** (154.5588)
ADT		0.3137 *** (18.9765)
IJL			0.4551 *** (43.1193)
AJL				0.3469 *** (40.2376)
ICZ					−0.0388 *** (−7.8443)
ACZ						−0.0264 *** (−6.3671)
Controls	YES	YES	YES	YES	YES	YES
_cons	−0.0046 (−0.0959)	−0.0462 (−0.6316)	0.5355 *** (8.8238)	−0.0785 (−1.0213)	0.7732 *** (12.2073)	0.0706 (0.9185)
N	49,490	49,490	44,208	44,208	44,144	44,144
R2	0.5209	0.4915	0.3017	0.5138	0.2622	0.4872
year	YES	YES	YES	YES	YES	YES
province	YES	NO	YES	NO	YES	NO
industry	NO	YES	NO	YES	NO	YES

Note: T-statistics are reported in parentheses. Asterisks *** denote statistical significance at the 1% level.

, are qualitatively unchanged from our baseline estimates. This reinforces our main conclusions and underscores the robustness of our results.

4.4.2. Altering the Sample Period

The COVID-19 pandemic, which emerged in late 2019, represents a significant exogenous shock that likely affected firm performance. Furthermore, the year 2015 witnessed notable policy turbulence and stock market fluctuations in China, which may also constitute a confounding event. To ensure that our findings are not driven by these exogenous shocks, we conduct robustness checks by altering the sample period. Specifically, we first exclude data from 2020 onwards to isolate the pandemic’s impact. Second, we exclude the data from 2015 to address potential confounding from that year’s unique events. The results, reported in

Table 8. Robustness Check with Alternative Sample Selection Criteria.

Variable	(1) DT	(2) DT	(3) DT	(4) DT	(5) DT	(6) DT
IDT	0.9249 *** (104.3591)
ADT		0.2238 *** (9.5126)
IJL			0.7181 *** (46.9137)
AJL				0.5502 *** (43.9554)
ICZ					−0.0283 *** (−3.8701)
ACZ						−0.0145 ** (−2.3833)
Controls	YES	YES	YES	YES	YES	YES
_cons	−0.0832 (−1.3901)	−0.0330 (−0.3710)	0.2131 *** (2.8276)	−0.1580 * (−1.6887)	0.6073 *** (7.5285)	0.0399 (0.4123)
N	23,687	23,687	20,348	20,348	20,314	20,348
R2	0.4720	0.4742	0.2835	0.5354	0.1816	0.4705
year	YES	YES	YES	YES	YES	YES
province	YES	NO	YES	NO	YES	NO
industry	NO	YES	NO	YES	NO	YES

Note: T-statistics are reported in parentheses. Asterisks *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively.

, remain consistent with our baseline findings across all model specifications, further confirming the robustness of our conclusions.

4.5. Threshold Effect Analysis

Prior to estimating the threshold regression models, we first test for the existence of threshold effects. The test results are presented in

Table 9. Threshold Effect Regression Results.

Threshold Variable	Test Type	F-Value	p-Value	Threshold Estimate	Confidence Interval
ICZ	Single Threshold	13.07	0.023	−0.4794	(−0.5282, −0.4657)
	Double Threshold	The double and triple threshold effects are statistically insignificant for the industry performance gap.
	Triple Threshold
ACZ		No significant threshold effect is found for the regional performance gap.
Variable		Coefficient Estimate		t-statistic
ICZ ≤ −0.4794		0.4401 ***		15.5083
ICZ > −0.4794		0.3572 ***		14.6721

Note: T-statistics are reported in parentheses. Asterisks *** denote statistical significance at the 1% level.

. Using the performance gap (CZ) as the threshold variable, we find a statistically significant single threshold effect for industry peers, but no significant double or triple threshold effects. For regional peers, however, we find no evidence of any significant threshold effects.

From Table 9, we observe that when the industry performance gap is less than or equal to the threshold value of −0.4794, the estimated coefficient is 0.4401. When the performance gap exceeds this threshold, the coefficient decreases to 0.3572. As the performance gap varies, the marginal effect of the prior-period transformation distance decreases at the threshold, which aligns with theoretical expectations. The absence of a threshold effect for regional peers may occur because firms within the same region operate under similar external institutional environments, where digital transformation is driven more by policy mandates than by performance competition. Such external intervention could weaken the moderating role of the performance gap on the transformation distance effect, thereby stabilizing its impact. Furthermore, industry leaders might benefit from structural competitive advantages (e.g., technological patents) that create inertia, whereas regional leaders may rely on locational advantages (e.g., port resources) rather than digital capabilities to maintain their status; thus, their leading position may not significantly dampen their digital transformation impetus. These findings provide empirical support for Hypothesis H4.

4.6. Further Analysis

4.6.1. Test of the Multiplier Effect of the Peer Effects

Peer effects exhibit a multiplier effect. A small initial shock may lead to significant changes in group behavior [41]. In the context of digital transformation, the successful transformation of pioneering enterprises can rapidly disseminate, motivating others to follow suit and triggering a wave of digital adoption. The peer effect in digital transformation first exerts a positive influence at the firm level by accelerating the pace of adoption. As more firms accelerate their transformation, this effect accumulates at the industry or regional level, thereby fostering coordinated growth in the overall digital landscape [7].

This multiplier effect operates through several mechanisms. First, it reduces the cost of transformation for subsequent firms [20], as they can learn from the experiences of pioneers and avoid repeating past mistakes. Second, the peer effect fosters inter-firm cooperation and innovation, providing a platform for problem-solving during transformation, which accelerates technological progress and the adoption of new applications. Moreover, competitive pressure within the group is a key factor driving firms to accelerate their digitalization, prompting them to innovate and seek more effective digital solutions. Simultaneously, extensive cooperation exists among interconnected firms [5]. During transformation, firms often collaborate with diverse partners, including other firms and research institutions, to achieve effective resource integration and leverage complementary advantages. Such collaboration not only helps to address technical challenges but also facilitates the emergence and development of new business models. Finally, as more firms within an industry or region engage in digital transformation, governments tend to increase their attention and support by formulating relevant policies, laws, and regulations to guide and promote these efforts. Government support, in turn, further stimulates firms’ enthusiasm for transformation, creating a virtuous cycle that promotes the continuous advancement of the digital level across the entire industry or region.

To test the multiplier characteristics of digital transformation peer effects, we introduce the time variable, its interaction term with industry-peer digital transformation, and its interaction term with regional-peer digital transformation into the model for empirical examination. A positive coefficient on the interaction term would indicate that the time trend positively moderates the peer effect. This suggests that as time progresses, the influence of peer effects expands, driving more firms to adopt digital transformation. This would demonstrate that initial transformation behaviors are gradually amplified through inter-firm interaction and imitation mechanisms, promoting coordinated change within a broader group. The results are presented in

Table 10. Multiplier Effect Test of Peer Influence.

Variable	(1) DT	(2) DT
IDT	0.9325 *** (153.1428)
ADT		0.3094 *** (18.8033)
Year	−0.0053 ** (−2.2395)	−0.0166 *** (−3.4064)
IDT∙Year	0.0021 (1.1698)
ADT∙Year		0.0166 *** (3.8851)
Controls	YES	YES
_cons	10.6359 ** (2.2181)	33.0792 *** (3.3709)
N	49,490	49,490
R2	0.5209	0.4916
year	YES	YES
province	YES	NO
industry	NO	YES

Note: T-statistics are reported in parentheses. Asterisks ** and *** denote statistical significance at the 5% and 1% levels, respectively.

. The coefficient for the interaction between the time variable and industry peer digital transformation is not statistically significant. In contrast, the coefficient for the interaction between the time variable and regional peer digital transformation is 0.0166 and significant at the 1% level.

The lack of significance for the industry peer interaction term may stem from several factors. Within the same industry, intense competition fosters a strong sense of self-protection among firms, making them reluctant to share their proprietary technologies and accumulated knowledge. Even with numerous successful digital transformation cases within an industry, significant heterogeneity in firm size, technical capabilities, and market positioning often makes it difficult for these cases to be effectively replicated by others. Theoretically, peer firms’ digital transformation practices could mutually influence and promote each other. However, in the absence of effective information channels and knowledge-sharing mechanisms within the industry, successful experiences struggle to be disseminated rapidly, learned and adopted by other firms.

In contrast, firms within the same geographical region often possess a natural advantage in information exchange. They engage in high-frequency face-to-face communication, interaction, and in-depth cooperation, which facilitates the rapid circulation and widespread sharing of knowledge and technology within the region. This injects strong momentum into the overall digital transformation process, significantly accelerating its pace. Furthermore, firms within the same region are more likely to be influenced by similar policy environments, market conditions, and cultural contexts. Collectively, these factors create a favorable environment for digital transformation to be promoted in a more coordinated manner.

4.6.2. Economic Consequences of the Peer Effects in Digital Transformation

Digital transformation itself has a significant performance-enhancing effect [42]. The imitating firm’s adaptation, absorption, and improvement of peers’ digital practices primarily occur through three mechanisms: decision optimization [7], risk mitigation [5], and legitimacy acquisition [43]. This process reduces transformation costs and risks, enhances operational efficiency and market recognition, and ultimately leads to gradual improvements in firm performance. The stronger the peer effect, the deeper the imitation learning, and the more significant the performance improvement resulting from this “experience transfer.” As digital transformation is a long-term strategy, its effectiveness is often better reflected in market expectations than in short-term financial metrics. This study employs Tobin’s Q as a performance metric to examine the impact of the digital transformation peer effect on firm performance, as it more accurately captures its role in shaping long-term value and growth potential. The results are presented in

Table 11. Economic Consequences of Peer Effects.

Variable	(1) TobinQ	(2) TobinQ
IDT	0.1473 *** (18.6264)
ADT		−0.0030 (−0.1562)
Controls	YES	YES
_cons	3.2457 *** (52.2900)	3.4356 *** (30.9014)
N	49,490	49,490
R2	0.1359	0.1752
year	YES	YES
province	YES	NO
industry	NO	YES

Note: T-statistics are reported in parentheses. Asterisks *** denote statistical significance at the 1% level.

.

As shown in Table 11, the impact of the industry peer effect on firm performance is positive and significant, whereas the impact of the regional peer effect is not significant. A potential explanation is that firms within the same industry operate under similar market environments, customer demands, technical standards, and norms [21], making it easier for them to identify common ground and relevant benchmarks during digital transformation. The successful experiences and models developed by industry leaders can be more readily learned and imitated by other firms within the same industry. Due to the high consistency in product usage and customer demands among firms within the same industry, imitating peer strategies is more effective for driving performance improvement.

Firms within the same region, however, may exhibit significant heterogeneity in operational models, market channels, customer bases, and other factors, making it difficult to form a unified digital transformation model or strategy. Consequently, even if a regional peer effect exists, its overall effect on enhancing firm performance is not significant. Moreover, policies and measures supporting digital transformation vary across regional governments. Some regions may lack a comprehensive policy support system and guidance mechanism, leaving firms without the necessary guidance and assistance during their transformation. This policy deficiency may further constrain the ability of the regional peer effect to enhance firm performance.

5. Conclusions, Recommendations and Limitations

5.1. Conclusions

Drawing on data from listed companies on the Shanghai and Shenzhen stock exchanges between 2010 and 2024, this study investigates the peer effects in digital transformation, examining how these effects influence firms’ decisions to adopt digital technologies and identifying key factors that moderate this relationship. The main findings are as follows: (1) Digital transformation exhibits both industry and regional peer effects. (2) The transformation distance between the focal firm and its peers is a significant factor motivating digital transformation. Specifically, a larger prior-period transformation distance positively influences the focal firm’s current-period digital transformation efforts. That is, the greater the absolute gap between the focal firm’s digitalization level and that of its peers in the previous period, the stronger its impetus for digital transformation. (3) The performance gap in the prior period exerts a negative influence on the focal firm’s digital transformation in the current period. Specifically, the further the focal firm’s prior performance falls below that of its peers, the stronger its impetus for digital transformation becomes, and conversely, the less motivation it demonstrates when outperforming peers. (4) A threshold effect exists in the influence of the prior-period transformation distance on digital transformation motivation as the prior performance gap varies. Specifically, the marginal effect of the prior-period transformation distance decreases at the threshold within the same industry. (5) Digital transformation peer effects demonstrate multiplier characteristics. (6) The industry peer effect in digital transformation positively influence firm performance.

5.2. Recommendations

Based on our systemic analysis of peer effects, this study proposes recommendations for multi-level stakeholders to navigate and steer the digital transformation ecosystem effectively:

5.2.1. Recommendations for Central Authorities

The central government should focus on top-level design and macro-level guidance to create a conducive environment for nationwide digital transformation.

Develop Differentiated Policy Frameworks: Formulate industrial and regional policies that account for digitalization disparities. The identified peer effects and threshold mechanisms can enhance the precision of policy evaluation and resource allocation, enabling the construction of a nationally coordinated yet differentiated transformation echelon.

Cultivate National Digital Leaders: Stimulate ecosystem-wide transformation by strategically nurturing firms with advanced digital capabilities through national-level financial subsidies and tax incentives. Supporting these “system catalysts” encourages them to widen the transformation distance, thereby raising the reference point and stimulating broader digitalization efforts.

Promote Cross-Regional Collaboration: Invest in or endorse national-level digital transformation platforms that facilitate information sharing and collaboration across regional and industrial boundaries. This helps amplify the positive multiplier effects of peer influence and mitigates regional fragmentation.

Bridge the Digital Infrastructure Gap: Prioritize strategic investment in foundational digital infrastructure (e.g., broadband networks, computing centers) in underdeveloped regions. This is imperative to narrow the digital divide and ensure the ecosystem’s inclusive and balanced evolution, preventing systemic imbalances.

5.2.2. Recommendations for Regional Governments

Local governments should concentrate on creating a favorable local environment, implementing targeted support, and adapting central policies to regional contexts.

Implement Targeted Subsidy Mechanisms: Enhance support for lower-performing local firms through direct subsidies based on digital transformation investment percentages or tax deductions for digital equipment purchases. This helps firms facing competitive pressures to accelerate their digitalization and leverage the incentive effects of the transformation distance.

Establish Local Benchmarking Systems: Identify and promote successfully transformed local firms as benchmarks. Organizing case studies, site visits, and experience-sharing sessions can create a localized environment conducive to emulation and catch-up, directly harnessing the power of peer effects.

Launch Confidence-Building Programs: For chronically underperforming firms, establish dedicated support funds offering low-interest loans and entrepreneurial counseling. Such measures can help overcome confidence deficits, bolster willingness to invest, and enable firms to utilize the transformation distance effect to overcome strategic thresholds.

Provide Tiered Firm Support: Offer high-end resource opportunities (e.g., international exchanges, cutting-edge R&D collaboration) to advanced local firms. For less advanced firms, the focus should be on basic training, technical support, and digital literacy programs to facilitate rapid improvement.

5.2.3. Recommendations for Company Leaders

Firm managers must adopt a proactive and adaptive strategic posture, leveraging insights from peer dynamics for internal decision-making.

Benchmark and Adapt Strategically: Actively monitor the digitalization initiatives of industry and regional peers. Managers should systematically assess their firm’s transformation distance and performance gap relative to these peers to dynamically adjust strategies and resource allocation, avoiding strategic myopia.

Integrate Digital Transformation into Core Strategy: Recognize digital transformation’s role in driving long-term competitiveness, not as a short-term trend. It should be embedded into core strategic planning, with clear pathways emphasizing coordinated changes in technology, organizational structure, and management models.

Forge Strategic Symbiotic Partnerships: Leverage resource complementarity by collaborating with partners such as technology startups, research institutions, or peer firms. Through resource sharing and co-innovation, firms can access critical capabilities and mutually enhance their digital transformation capacity.

Manage Risks of Strategic Interdependence: While learning from peers, remain mindful of the risks associated with herd behavior. Firms must optimize internal resource allocation to ensure transformation initiatives are aligned with their unique context, maximizing effectiveness and preserving competitive differentiation.

5.3. Limitations

Our model specification incorporates a one-year lag to establish temporal precedence, a necessary step for causal identification. However, the transmission of peer influence is likely a process rather than a discrete event. The “true” time-lag might vary across firms and contexts, influenced by factors such as organizational agility and the complexity of the technology being adopted. Therefore, while our findings robustly confirm the existence of peer effects, future research could delve deeper into the “black box” of this process, exploring how the strength and speed of peer influence evolve over time and what firm-level factors moderate this temporal pathway.

Regarding the generalizability of our findings, the empirical evidence of this study is drawn exclusively from Chinese A-share listed companies. While this sample allows for a robust test of our hypotheses within an important context, it necessarily limits the direct applicability of our conclusions. The extent to which these findings generalize to unlisted or small and medium-sized enterprises (SMEs), which often operate under different resource constraints and competitive pressures, remains an open question. Furthermore, the unique institutional environment of China, including its distinct policy landscape and market characteristics, suggests that the strength and operation of these peer effects in other countries or economic regions warrant further investigation. Future research could test the boundary conditions of our theoretical framework by replicating this study in different national contexts.

Our study primarily examines peer effects at an aggregate level. Although we control for industry fixed effects, we do not explicitly investigate the heterogeneity of peer effects across different industries. It is plausible that the mechanisms we identify are more pronounced in technology-intensive sectors or industries with rapid innovation cycles compared to more traditional sectors. Exploring how industry-specific characteristics, such as competitive dynamics, regulatory intensity, and technological opportunity, moderate the peer effect process represents a promising direction for future research.

While our use of keyword frequency to measure digital transformation follows established practice and enables large-sample analysis, this approach may better capture strategic discourse than implementation depth. Future studies could employ surveys or case studies to provide more nuanced measurement.