Chen et al.

Digital Economy and

Digital Economy and Sustainable Development (2023) 1:18
https://doi.org/10.1007/s44265-023-00018-x Sustainable Development

RESEARCH Open Access

Digital transformation and firm

performance: a case study on China’s listed
companies in 2009–2020
Chuanglian Chen1, Yi Zhang1 and Shudan Wang1*

Abstract
Based on a panel fixed effect model, our paper explores the impact of digital transformation on the operation effi-
ciency of Chinese A-share listed enterprises during 2009–2020. It conducts related robustness tests, mechanism analy-
sis, and heterogeneity analysis to provide a comprehensive understanding of the intrinsic relationship between digi-
talization and firm performance. The empirical findings demonstrate that digital transformation has a significantly
positive effect on firm performance. More specifically, rising digital transformation of enterprises leads to an improve-
ment in their operation efficiency. The research further reveals that digital transformation mainly promotes the effi-
ciency of capital-labor allocation, which in turn enhances firm performance. Overall, our study adds to the existing
literature by offering empirical evidence on the relationship between digital transformation and enterprise operation
efficiency, providing valuable insights for practitioners and policymakers.
Keywords Digital transformation, Firm performance, Capital allocation efficiency, Labor allocation efficiency

1 Introduction industrial revolution, featuring with digital technologies

As technology continues to advance as digital transfor- and big data, which has had a profound impact on busi-
mation of enterprises has emerged as a critical area that ness operation (Lasi et al. 2014). Previous research has
has had and will continue to have a profound effect on shown that digital transformation is a key factor in boost-
how business may operate and the way how people may ing firm performance (Bakhshi et al. 2014), reducing pro-
behave. The G20 Digital Economy Development and duction costs (Loebbecke and Picot 2015), enhancing
Cooperation Initiative document, released at the G20 innovation capabilities (Yoo et al. 2012), and optimizing
Hangzhou Summit in 2016, identifies three phases of organizational management structures (Einav and Levin
the digital economy: digitalization of information, digi- 2014) to drive high-quality and efficient corporate devel-
tal transformation, and digitalization of business. Among opment. In a rapidly changing global competitive land-
these, digital transformation involves utilizing digital scape and a volatile economic environment, companies
technologies to aggregate, analyze, utilize, and exchange must actively embrace change and proactively promote
data, in order to make informed decisions and develop digital transformation to stay ahead. Therefore, examin-
innovative business models that drive value creation and ing the impact of digital transformation on the operation
enhance business performance (Verhoef et al. 2021). This efficiency of companies, from the perspective of their
phenomenon is described by Yao (2023) as the fourth digital transformation, is essential for companies to bet-
ter adapt to the new market and future challenges.
*Correspondence: When examining the factors that exert influence
Shudan Wang upon firm performance, the predominant focus within
[email protected]
1
the realm of enterprise industry literature resides in
School of Economics, Jinan University, Guangzhou, China
the analysis of commercial banks’ performance (Berger

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/.
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 2 of 16

et al. 2009; Fungáčová et al. 2013). Concurrently, these companies spanning the years 2009 to 2020. This explo-
influential determinants can be categorized into exter- ration is grounded in the application of a panel fixed-
nal and internal factors. Some scholars have diligently effect model, supplemented by a comprehensive analysis
scrutinized the ramifications of external factors on firm of the associated mechanisms and potential sources of
performance, encompassing aspects such as strategic heterogeneity. To this end, we incorporate the interaction
investment allure (Berger et al. 2009), regional corruption of capital allocation efficiency, labor allocation efficiency,
dynamics (Hanousek et al. 2019), the pervasive impact and technological progress bias in our investigation.
of the COVID-19 pandemic (Khatib and Nour 2021), Our findings unveil a significant and affirmative role
and various other contributory elements. Furthermore, played by enterprise Digital Transformation in bolstering
certain academics have engaged in discourse surround- operating efficiency. This enhancement is chiefly facili-
ing the consequences posed by internal control factors tated by the concurrent optimization of capital and labor
on firm performance. For instance, the purview extends allocation efficiencies. Notably, our inquiry uncovers a
to encompass intracompany risk management protocols pronounced inclination for firms to direct their tech-
(Gordon Lawrence et al. 2009), interrelations amidst nological advancements towards labor-centric avenues
internal executive counterparts (Michiels et al. 2013), when pursuing digital transformation for the purpose of
and the defining characteristics of governing boards augmenting firm performance. These discernments bear
(Pucheta-Martínez and Gallego-Álvarez 2020). profound implications for comprehending the intricate
Of particular note, the extent of digital transformation mechanisms that underlie the relationship between Digi-
undertaken by enterprises, as an intrinsic factor exerting tal Transformation and business efficiency. By contribut-
influence upon firm performance, has engendered a cor- ing to the augmentation of the existing body of relevant
pus of pertinent literature. Bloom et al. (2012) and Bessen literature, our findings furnish valuable insights and serve
and Righi (2019), leveraging data gleaned from an expan- as a foundational underpinning for the formulation and
sive cohort comprising over 4,000 companies across execution of informed policy decisions.
Europe and Asia, alongside 4,262 entities cataloged A distinguishing facet of our study lies in its innovative
within the Compustat database in the United States, dur- conceptualization of digital intelligent transformation
ing the chronological span from 1990 to 2012, respec- as an overarching technological augmentation of both
tively, have empirically established that the evolution of capital and labor elements. This novel approach posits
digital technology has propelled improvements in firm that Digital Transformation intrinsically influences the
performance. Nevertheless, Kim et al. (2011) posited the efficiency of both capital and labor, thereby necessitat-
view that the impact of digital technology on firm per- ing a comprehensive examination of their intertwined
formance lacks a statistically significant positive impetus. dynamics. Additionally, we break new ground by devis-
Certain scholars, exemplified by Hajli et al. (2015), pos- ing a quantitative measure to gauge the extent of enter-
tulate that the dividends reaped from digital transforma- prise digital transformation, employing Python-based
tion may be selectively garnered by particular enterprises web crawling techniques and text analysis methodolo-
or industries. Therefore, a note of circumspection is war- gies. This pioneering methodology presents research-
ranted as Ferreira and Fernandes (2017) underscored the ers with a potent and precise instrument for assessing
inconclusiveness characterizing the polarity of the influ- and appraising the extent of Digital Transformation
ence wielded by digital transformation upon corporate within enterprises, thereby offering a valuable reference
performance. In light of this, prudence dictates refraining point for subsequent research endeavors and practical
from prematurely drawing definitive conclusions. applications.
Through a comprehensive review of the existing
research literature, it becomes apparent that there exists 2 Research hypothesis
a lack of consensus among academic perspectives regard- Internal and external information asymmetry commonly
ing the influence of Digital Transformation on enterprise plague companies. The former refers to the inadequate
operational efficiency. Consequently, a more profound exchange of information among various departments,
exploration of the intricate interplay between these impeding the smooth flow of information and conse-
two phenomena becomes imperative. Furthermore, the quently affecting the overall performance of the enter-
mechanisms through which Digital Transformation prise. The latter denotes the unequal distribution of
exerts its impact on business efficiency have received information between firms and external stakeholders,
limited scholarly attention. In light of these gaps, the pre- thereby increasing uncertainty and risk in transaction pro-
sent study aims to delve into the intricate relationship cesses. Digital transformation, on the one hand, enhances
between corporate Digital Transformation and corporate internal management efficiency (Loonam et al. 2018), and
operating efficiency, utilizing a sample of Chinese listed on the other hand, helps firms become more flexible and
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 3 of 16

responsive to external market changes (Mikalef and Pateli to form performance of the banking industry in China.
2017), facilitating rational planning of growth strategies to However, few studies have deliberately focused on digi-
adapt to consumer demand fluctuations (Kaur and Sood tal transformation and its impact on firm operation effi-
2015). In general, digital transformation has the potential ciency of the Chinese listed firms during 2009–20 when
to improve communication efficiency within and between many enterprises had undergone prodigious digital trans-
firms, reduce information asymmetry, increase informa- formation, providing an unique opportunity for us to tes-
tion delivery transparency and efficiency, thereby lower- tify whether digital transformation can have a significant
ing operational costs (Lindstedt and Nauri 2010; Mikalef influence on firm performance in the Chinese context. In
et al. 2018), search costs and transaction costs (Goldfarb particular, we aim to testify another hypothesis on how
and Tucker 2019). Moreover, Bessen and Righi (2019) and and why digital transformation affects firm performance
other scholars have noted that digital transformation can as put forward by hypothesis 2 below.
increase firms’ productivity. They suggest that through
Hypothesis 2: Digital transformation can improve
digital transformation, firms can manage their costs effec-
the resource and labor allocation efficiency of firms,
tively and generate higher returns, which can ultimately
thereby increasing operational efficiency.
improve their operation efficiency. Based on these find-
ings, we propose hypothesis 1.
3 Research design
Hypothesis 1: Digital transformation significantly 3.1 Data source
improves firm performance. The financial data of the listed companies examined in
our study were sourced from two databases: the China
Firm performance can be enhanced through two
Research Data Service Platform (CNRDS) and the
mechanisms of in their digital transformation progress.
WIND database. Digital transformation measurement
In accordance with the Enterprise Innovation Driver
data are collected from the annual reports of companies
Theory, digital transformation inherently embodies
listed on the Shanghai and Shenzhen Stock Exchanges.
innovation. This paradigm ushers in novel technolo-
Our study uses a sample of Chinese A-share listed
gies, tools, and methodologies, thereby ushering in a
companies during 2009–20, which comprises 12,276
transformative shift in the operational landscape and
firm-year observations in 12 consecutive years after
business model of enterprises. Innovation, as a catalyst,
eliminating missing, duplicated, and invalid samples. To
empowers enterprises to unearth novel resource alloca-
ensure the robustness of the data, a 1% upper and lower
tion strategies and labor organization approaches that
tailing process is employed to filter out extreme values
seamlessly align with evolving market demand dynamics.
that could skew the results.
This innovative stance stands poised to effectively cata-
lyze the efficiency of enterprise resource allocation and
3.2 Model construction
labor deployment, a notion that is substantiated within
To examine the impact of digital transformation on firm
the existing body of literature. Loebbecke and Picot
performance, our study constructs a model that includes
(2015) contend that the use of operational digital tech-
several variables. The dependent variable is total factor
nology enables firms to make better investment choices
productivity (tfp), which is commonly used to measure
at a lower cost, thereby directing resources towards effi-
firm performance. The degree of digital transformation
cient projects and improving the efficiency of capital
is represented by the standardized digital transformation
allocation. Secondly, according to Goldfarb and Tucker
word frequency (ai). A set of control variables (controls)
(2019), digital technology development leads to increased
is included in the model. The subscript it represents the
market transparency, thereby reducing search costs and
data of the i-th variable in year t. Individual fixed effects
tracking costs for firms, which in turn improves resource
or industry fixed effects are denoted by ui , time fixed
allocation efficiency. Additionally, the digital transforma-
effects are represented by µt , and the random error term
tion of enterprises can facilitate the replacement of sim-
is εit . The hypothesis that digital transformation of enter-
ple labor with automated systems (Babina et al. 2021),
prises can improve operational efficiency is tested by
reduce the demand for low-end labor, and remove tem-
assessing the significance of γ1, the coefficient on ai. A
poral and spatial constraints and barriers to labor fac-
significantly positive value of γ1 indicates that hypothesis
tor mobility (Acemoglu and Restrepo 2018), ultimately
1 is supported.
leading to improved labor allocation efficiency. Yang
et al. (2023) suggests that digital transformation not only tfpit = γ0 + γ1 aiit + controlsit + µt + ui + εit (1)
enhances firm productivity but also reduces the income
gap between the high- and low-skilled workers. Chen Additionally, the present study aims to investigate
et al. (2023) posits that digital technologies are conducive the underlying mechanisms through which digital
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 4 of 16

transformation affects firm performance. To this end, lnYit = α + β1 lnKit + β2 lnLit + β3 lnMit + εit (3)
we introduce the interaction terms ( aiit ∗ mechanismit )
in Model 1 to examine the mediating roles of capital Figure 1 illustrates the annual average trend of total
allocation efficiency, labor allocation efficiency, and factor productivity for enterprises. Taking an overall per-
technological progress bias. Specifically, we denote the spective, the average annual TFP of enterprises ranges
capital allocation efficiency as mpk , labor allocation between 3.80 and 3.95, with a maximum value of 3.95 and
efficiency as mpl , and technological progress bias as d . a minimum value of 3.82. There is little variation in the
For ease of interpretation, we de-mean the interaction average TFP across different years. However, upon closer
terms of mpk , mpl , and d . examination of Fig. 1, several noteworthy points emerge.

tfpit = γ0 + γ1 aiit + γ2 aiit ∗ mechanismit + controlsit + µt + ui + εit (2)

3.3 Variable settings Following the financial crisis in 2008, the Chinese

3.3.1 Explained variables (tfp) government implemented a massive expansionary fis-
In our study, we utilize the lp method to gauge the cal policy stimulus program to bolster effective mar-
total factor productivity of enterprises. Semi paramet- ket demand and stabilize the economy. Increased
ric methods such as lp estimation can effectively solve market demand enabled firms to generate revenue
endogeneity and sample selection problems. To achieve from product sales and invest in business expansion to
this, we formulate a model as follows: Yit represents the leverage economies of scale, thereby improving their
main business income, Kit the net investment in fixed performance. Notably, enterprise performance peaked
assets, Lit the number of employees, Mit intermediate in 2011, followed by a subsequent decline that reached
inputs, which is replaced by the data of cash and labor its lowest point in 2015 at 3.82. The year 2015 witnessed
purchased by the enterprise, and εit is a random error significant volatility in China’s capital markets, with the
term. The application of the lp method necessitates the Shanghai Stock Exchange Index experiencing a sharp
utilization of intermediate inputs as proxies. The coeffi- decline and multiple stocks being subject to trading sus-
cients are derived through the regression of Eq. 3. These pensions and circuit breakers. Despite the government’s
coefficients are subsequently integrated into Eq. 3 to efforts to stabilize the market, these events inevitably
compute the projected values. By subsequently juxta- impacted enterprise operations. However, post-2015,
posing these projected values with the actual output enterprise performance gradually rebounded and
values, residual values are calculated, effectively yield- reached a recent peak of 3.93 in 2018. In 2019, the out-
ing the measure of total factor productivity. In relation break of the novel coronavirus (COVID-19) further dis-
to the quantification of total factor productivity, we rupted enterprise operations. External factors, such as
have employed the "prodest" command within Stata to reduced market demand and logistical challenges, cou-
facilitate this calculation. Given the twelve-year dura- pled with internal difficulties related to resuming work
tion of the data, we apply the price index deflator to the and production, contributed to a decline in business
main business income, net investment in fixed assets, efficiency and, subsequently, enterprise performance.
and intermediate inputs prior to calculating total factor Figure 1 also indicates a notable decrease in the average
productivity. A higher tfp value is indicative of better annual TFP of enterprises in 2020, falling to 3.56. This
firm performance. decline can be attributed to the far-reaching effects of

Fig. 1 Annual average total factor productivity trend

Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 5 of 16

the COVID-19 pandemic, which continued to exert its the total number of keywords (236). This method pro-
influence on the global economy, including the enter- vides researchers with an effective tool to more compre-
prise sector in China. hensively and accurately measure and evaluate the degree
Although the average value of total factor productivity of Digital transformation of enterprises. As depicted in
does not show significant variation between years, a closer Fig. 2, there is an observable upward trend in the level of
look reveals the impact of external shocks on firm per- digital transformation across enterprises.
formance. the financial crisis of 2008 stimulated govern- Based on Fig. 2, it is evident that the degree of digi-
ment intervention, leading to improved firm performance. tal transformation in enterprises has generally been
However, fluctuations in subsequent years were influenced increasing since 2009, reaching its peak in 2020, with
by stock market events and the devastating impact of the a range of 0.041 to 0.161. Looking more closely, there
COVID-19 pandemic in 2019. These findings contribute to is a steady rise in the degree of digital transformation
a better understanding of the dynamics of firm performance from 2009 to 2013, with a value of 0.050 in 2013. Subse-
and its broader implications for the Chinese economy. quently, the pace of digital transformation accelerated,
reaching 0.058 in 2014 and further increasing to 0.062 in
3.3.2 Core explanatory variables (ai) 2015. From 2015 onwards, there was explosive growth
Considering the limitations of traditional financial indi- in the degree of digital transformation. To assess the
cators to represent the level of digital transformation, year-on-year growth rates, we can calculate the growth
and the advancements in text mining technology, our rates for various periods. The growth rate of digital
study employs text analysis to measure the digital level transformation from 2014 to 2015 was 0.07, from 2015
of enterprises. Firstly, the annual reports of all listed to 2016 it was 0.37, from 2016 to 2017 it was 0.24, from
enterprises in the sample period from the two Chinese 2017 to 2018 it was 0.21, from 2018 to 2019 it was 0.15,
Stock Exchanges are crawled using the Python software and from 2019 to 2020 it was 0.08. Notably, the period
crawler technology. English annual reports are elimi- of 2015–2016 witnessed the fastest growth in the degree
nated. Secondly, the obtained annual reports are matched of digital transformation, followed by 2016–2017.These
with existing filtered enterprise samples, and unsuccess- findings from Fig. 2 provide valuable insights into the
ful matches are eliminated. Then, 236 keywords related increasing trend of digital transformation in enterprises
to the digital transformation of enterprises are identi- over time. The period from 2015 to 2016 stands out as
fied through an extensive review of literature and policy a period of significant growth, reflecting the increasing
reports, and consultations with experts. These keywords, importance of digital transformation for enterprises.
such as "face recognition, big data, digital intelligence, The subsequent years also demonstrate notable growth
financial technology," are used to build a thesaurus in the rates, highlighting the ongoing adoption and integration
Python software "jieba" package. Finally, machine learn- of digital technologies in firm performance.
ing text analysis function is utilized to mine and analyze
the matched enterprise annual reports, and the standard- 3.3.3 Control variables (controls)
ized word frequencies of the keywords are used as indica- Our study employs several control variables, including
tors of the digitalization level of enterprises. The higher the total asset turnover ratio (turnover ), which is derived
the value of the indicators, the higher the digitalization from the net sales revenue divided by total assets. The
level. Standardized word frequency refers to dividing the growth rate of revenue ( salesr ) is also considered to eval-
keyword word frequency counted by each enterprise by uate the company’s growth ability, which is calculated

Fig. 2 Annual average digital transformation trend

Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 6 of 16

as the difference between current revenue and previous system, and µit represents the technical inefficiency
revenue divided by previous revenue. The stability of the error term.

1 1 1
lnYit = α + β1 lnKit + β2 lnLit + β3 Tt + β4 (lnKit )2 + β5 (lnLit )2 + β6 T 2 (4)
2 2 2
+β7 lnKit lnLit + β8 Tt lnKit + β9 Tt lnLit + εit − µit

company’s asset structure is assessed by fixratio, which is By utilizing the estimation outcomes from Model 4,
calculated as the ratio of fixed assets to total assets. The the relevant capital allocation efficiency, labor alloca-
return on total assets (roa), indicating a company’s prof- tion efficiency, and technological progress bias index
itability, is computed as the ratio of net income to total (Diamond 1965) of the firm can be derived as follows:
assets. The shareholding ratio of the largest shareholder εKit refers to the capital output elasticity, while εLit rep-
(own), expressed as a percentage, is used as a proxy for resents the labor output elasticity. If capital allocation
the concentration of equity in the company. Additionally, efficiency is high, it signifies the effective utilization of
the age of the enterprise (age) is also taken into account, resources in production, resulting in a higher output-to-
which is computed as ln(year of observation—year of input ratio. Similarly, elevated labor allocation efficiency
establishment + 1). implies optimal utilization of the workforce, leading to
comparatively higher output levels. The technological
3.3.4 Mechanism Variables (mechanism) progress bias index, denoted by d, implies that d > 0
Our study focuses on three primary mechanism vari- suggests a biased technological progress towards capi-
ables, namely, capital allocation efficiency ( mpk ), labor tal, while d < 0 implies that technological progress is
allocation efficiency ( mpl ), and technological progress biased towards labor. In contrast, d = 0 indicates a neu-
bias ( d ). To measure these variables, a stochastic fron- tral technical progress bias.
tier model in the form of trans-log production func-
tion is utilized. The model is constructed as follows: Yit mpkit = εKit ∗ Yit /Kit = (β1 + β4 lnKit + β7 lnLit + β8 Tt ) Yit /Kit

represents the output of the enterprise, which is meas- (5)

ured by the main business income of the enterprise. mplit = εLit ∗ Yit /Kit = (β2 + β5 lnKit + β7 lnKit + β9 Tt ) Yit /Kit
Kit denotes the capital input of the enterprise, which (6)
is measured by the net investment in fixed assets of β8 β9
the enterprise, while Lit denotes the labor input of the d= − (7)
εK εL
enterprise, which is measured by the number of active
employees of the enterprise. Tit represents technical
progress, which is the annual time minus the initial 3.3.5 Descriptive statistics
time of the sample span. εit denotes the stochastic error Table 1 presents the descriptive statistics for the vari-
term that reflects the inefficiency of the stochastic ables analyzed in our study. The total factor productivity

Table 1 Descriptive statistics of main variables

Variable Obs Mean Std. Dev Min Max

tfp 11,724 3.8905 0.7158 1.8329 5.7426

ai 11,448 0.0807 0.1374 0.0000 0.9000
turnover 12,274 0.6873 0.5443 0.0384 3.0980
salesr 12,273 10.1908 33.3621 -64.4708 175.0089
fixratio 12,253 0.2420 0.1854 0.0003 0.9393
roa 12,254 2.7619 6.2275 -27.9500 19.7300
age 12,276 3.0054 0.2615 1.7918 4.0073
own 12,276 0.3487 0.1513 0.0821 0.7496
mpk 12,144 0.4465 0.3521 0.0221 2.0029
mpl 12,144 0.4809 0.6806 0.0297 4.4341
d 12,144 -0.0568 0.0057 -0.0718 -0.0465
In Table 1, the variable tfp is Total factor productivity, ai is the degree of Digital transformation of the enterprise, turnover is the turnover rate of total assets, salesr is
the growth rate of operating revenue, fixratio is the ratio of fixed assets, roa is the return rate of total assets, own is the shareholding ratio of the largest shareholder,
age is the age of the enterprise
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 7 of 16

values range from a minimum of 1.8329 to a maximum the technical progress bias index, we observe a maximum
of 5.7426, with a mean of 3.8905 and a standard devia- value of -0.0465, a minimum value of -0.0718, a standard
tion of 0.7158. These results indicate that the total fac- deviation of 0.0057, and a mean value of -0.0568. These
tor productivity of each enterprise differs only slightly. findings suggest that the technical progress of the sample
The digital transformation of enterprises ranges from enterprises tends to be biased towards labor.
0.9000 to 0.0000, with a smaller standard deviation due
to the data being standardized. Meanwhile, in addition to 4 Empirical results and analysis
the explanatory variable of total factor productivity and 4.1 Benchmark return
the explanatory variable of the degree of digital trans- We conduct a Hausman test and obtained a p-value of
formation in enterprises, Table 1 provides descriptive less than 0.01, which leads to the rejection of the origi-
statistics for the control variables, which offer an intui- nal hypothesis. Subsequently, we adopt a fixed-effects
tive and comprehensive understanding of Chinese listed model. Table 2 displays the results of the baseline regres-
enterprises. sion, where the first column presents the outcomes of the
Examining the total asset turnover ratio, we observe regression with only the core explanatory variables based
values ranging from 0.0384 to 3.0980, with a standard on individual and time fixed effects. The results of the
deviation of 0.5443. This indicates that the total asset first column reveal a positive correlation between digital
turnover ratio of enterprises exhibits relatively low vari- transformation and the operational efficiency of enter-
ability, with a mean value of 0.6873. Notably, the con- prises, but the relationship is statistically insignificant.
trol variables that exhibit significant differences across On the other hand, the second column presents the out-
enterprises are the operating income growth rate and comes after the addition of control variables to the first
the return on total assets. The maximum value of the column. The findings show a significant and positive cor-
operating income growth rate is 178.0089, the minimum relation between the degree of digital transformation and
value is -64.4708, and the standard deviation is 33.3621. the business efficiency of the firm. In economic terms,
These figures suggest substantial variation in the oper- a 10% increase in the degree of digital transformation
ating income growth rate among enterprises, with some
even experiencing negative growth. Similarly, the return
on total assets shows considerable diversity, with a maxi-
Table 2 Baseline regression results
mum value of 19.7300, a minimum value of -27.9500, and
a standard deviation of 6.2275. The maximum value of (1) (2) (3) (4)
the fixed assets ratio is 0.9393, while the minimum value Variables tfp tfp tfp tfp
is 0.0003. Since the fixed assets ratio is obtained by divid- ai 0.1088 0.1881a 0.5759a 0.3684a
ing fixed assets by total assets, its values range between 0 (0.0777) (0.0400) (0.1624) (0.0764)
and 1. The standard deviation of the fixed assets ratio is a
turnover 0.9826 1.0485a
0.1854, with a mean value of 0.2420, indicating a relatively
(0.0281) (0.0731)
lower proportion of fixed assets in the sample. Moving
salesr 0.0015a 0.0011a
on to the enterprise age variable, we observe a maximum
(0.0001) (0.0003)
value of 4.0073, a minimum value of 1.7918, and a mean
fixratio 0.2820a 0.2576a
value of 3.0054, suggesting that the sample consists of
(0.0563) (0.0504)
relatively older enterprises. Analyzing the shareholding
roa 0.0029a 0.0073a
ratio of the first largest shareholder, we find a maximum
(0.0009) (0.0015)
value of 0.7496, a minimum value of 0.0821, a standard
age 0.0545 -0.0201
deviation of 0.1513, and a mean value of 0.3487. Further-
(0.1255) (0.0486)
more, considering the mechanism variable of enterprise
own 0.0636 0.3562a
capital allocation efficiency, we note a maximum value
(0.0640) (0.0569)
of 2.0029, a minimum value of 0.0221, a standard devia-
_cons 3.8334a 2.8335a 3.8475a 2.9605a
tion of 0.3521, and a mean value of 0.4465. Similarly, for
(0.0134) (0.3426) (0.0132) (0.1599)
the enterprise labor allocation efficiency, the maximum
r2 0.0174 0.6150 0.3034 0.7860
value is 4.4321. Comparing the capital allocation effi-
N 10,960 10,960 10,960 10,960
ciency and labor allocation efficiency of enterprises, it is
year Yes Yes Yes Yes
evident that labor allocation efficiency exhibits greater
firm Yes Yes No No
fluctuations than capital allocation efficiency, with the
industry No No Yes Yes
mean value of labor allocation efficiency slightly higher
than that of capital allocation efficiency. Lastly, regarding Standard error in parentheses, taking clustering robust standard errors.
a
indicates significant at the statistical levels of 1%
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 8 of 16

results in a 1.881% increase in the business efficiency of to test the overidentification restriction, with a value of
the firm, while controlling for individual and time fixed 71.922 and a p-value of 0.0000, which indicates that the
effects and other control variables. Additionally, the original hypothesis of insufficient instrumental vari-
results of the second column demonstrate a significant able identification is rejected. Overall, these tests provide
and positive correlation between the total asset turno- strong evidence that the instrumental variables employed
ver ratio and the growth rate of operating income of the in our study are a valid substitute for the original core
enterprise with the operating efficiency of the enterprise. explanatory variables. The results in the first column of
This association arises from the fact that the higher these Table 3 demonstrate that digital transformation has a sig-
variables are, the quicker the total asset turnover, and the nificant positive impact on firms’ operational efficiency.
better the operating condition of the enterprise, reflect-
ing a stronger operating capacity and growth ability. The 4.2.2 Lagged variable regression
coefficient of fixed assets ratio also shows a significantly In order to further verify the robustness of our findings,
positive association at the 1% level, where a higher fixed we replace the dependent variable with a first-order lag of
assets ratio alleviates financing constraints and enhances total factor productivity in the second column of Table 3,
total factor productivity, as well as providing operational and controlled for individual and time fixed effects using
flexibility to improve operational efficiency. Moreover, 2sls. The results of the regression analysis reveal that the
the return on total assets of an enterprise exhibits a sig- coefficient of digital transformation is positive at the 10%
nificantly positive correlation with the operating effi- level, thus supporting our initial findings.
ciency of the enterprise. The profitability of the enterprise
is reflected in the return on total assets, and the better 4.2.3 Substitution of core explanatory variables
the profitability, the higher the operating efficiency of the In our study, the primary explanatory variable is the
enterprise. However, firm age and firm equity concen- standardized word frequency. To ensure the robustness
tration are positively but not significantly related to firm of the results, the logarithm of the total word frequency
operation efficiency. Apart from individual and time fixed of keywords is employed as a substitute. The regression
effects, industry and time fixed effect regressions are also results of the third column in Table 3 demonstrate that
conducted, and the results are presented in the third and the coefficient of digital transformation is significant at
fourth columns of Table 2. The coefficient of digital trans- the 1% level. Specifically, the coefficient of digital trans-
formation is 0.5759 at 1% level of significance without the formation is estimated to be 0.0226, indicating that
inclusion of control variables for industry and time fixed digital transformation of enterprises has a positive and
effects, and 0.3684 at 1% level of significance with the significant impact on their operation efficiency.
inclusion of control variables, which also confirms that
digital transformation is conducive to improving business
efficiency, thus proving hypothesis 1.
Table 3 Endogeneity and robustness test results
(1) (2) (3) (4)
4.2 Robustness tests
4.2.1 Endogeneity test Variables ai tfp L.tfp tfp tfp
To address the potential endogeneity issue, our study a b a
ai 0.2179 0.1238 0.0226 0.1325a
employs the two-stage least squares (2SLS) method with
(0.0555) (0.0736) (0.0050) (0.0406)
instrumental variables for testing endogeneity. Spe-
L1.ai 0.6728a
cifically, the first-order and second-order lags of digi-
(0.0332)
tal transformation are selected as instruments for the
L2.ai 0.0417
digital transformation of firms. The left side of the first
(0.0386)
column shows the results of the first stage regression,
F-stat 142.10 104.84
and the right side shows the results of the second stage
r2 0.5546 0.6160 0.4408 0.6179 0.6060
regression. The weak instrument problem is first exam-
N 8470 8470 8470 9944 9254
ined by testing the Wald F statistic, which yields a value
control Yes Yes Yes Yes Yes
of 3017.095 for each instrument, well exceeding the criti-
firm&year Yes Yes Yes Yes Yes
cal value of 10, thereby indicating that weak instruments
are not a concern. Additionally, the Hansen J test is per- The first column of Table 3 shows the results of the endogeneity test. The second
column shows the results of 2sls regressions on the lagged dependent variable,
formed to test the validity of the instrument selection. the third category shows the results of regressions with replacement of the core
The resulting p-value of 0.1962, greater than the conven- explanatory variable ols, and the fourth column shows the results of subsample
regressions
tional significance level of 0.05, confirms that the instru-
Standard error in parentheses, taking clustering robust standard errors. a, and b
ments are valid. Finally, the LM statistic is computed indicate significant at the statistical levels of 1%, and 10%, respectively
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 9 of 16

4.2.4 Subsample regression hence firm performance. The second column of Table 4

The present study employs empirical data spanning shows that the coefficient of digital transformation is still
2009–2020, a period that includes the emergence of the significantly positive after the inclusion of the interac-
COVID-19 pandemic in 2019, which has the potential tion term, and the coefficient of the interaction term is
to affect the study’s empirical outcomes. To address this also significantly positive at 0.2357. It can be shown that
potential issue, a subsample covering the years 2009– when labor allocation efficiency is average, a 10% rise in
2018 is selected for the regression analysis. As shown in digital transformation will raise total factor productivity
the fourth column of Table 3, the regression results indi- by 1.967%. It can also be shown that enterprises improve
cate that the coefficient of digital transformation remains their labor allocation efficiency by increasing the degree
significantly positive at the 1% level, which reaffirms the of digitization, and thus improve their performance. The
robustness of the study’s findings. higher the degree of digitalization, the higher the labor
allocation efficiency, and therefore the better firm perfor-
4.2.5 Mechanism test mance. The third column in Table 4 adds the interaction
To ensure the robustness of the results, our mechanism term between digital transformation and the technologi-
testing and heterogeneity analysis are both empirically cal progress bias index. It can be seen that the coefficient
tested based on two-stage least squares regression. The of Digital transformation is also positive at the signifi-
first column in Table 4 adds the interaction term between cance level of 1%, and the coefficient of interaction item is
digital transformation and firms’ capital allocation effi- -15.8174 at the significance level of 10%. From the above,
ciency to the 2sls regression. After adding the interaction we can see that the mean value of technological progress
term, the coefficient of digital transformation is still sig- bias is negative, so this indicates that Digital transforma-
nificantly positive at the 1% level, and the coefficient of tion can improve firm performance when technological
the interaction term is 0.4452, which is also significantly progress is labor oriented and less than or equal to its
positive at the 10% level. On the one hand, it can be mean value. Looking at it from another angle, when tech-
shown that the degree of digital transformation of a firm nological advancements are biased towards capital, digi-
can significantly improve the operation efficiency when tal transformation can have an adverse impact on firm
the capital allocation efficiency is average, and the higher performance.
the capital allocation efficiency, the greater the effect of
digital transformation on firm performance. In other 4.3 Heterogeneity analysis
words, capital allocation efficiency reinforces the effect 4.3.1 Region‑based heterogeneity analysis
of digital transformation on firm performance. On the Given the pronounced variations in economic develop-
other hand, since the coefficient of the interaction term ment levels across different regions within China, our
is also significantly positive, it also indicates that digital study employs a meticulous heterogeneity analysis. This
transformation improves capital allocation efficiency, and analysis stratifies the nation’s provinces into two distinct
categories: the eastern region and the central-western
region. The primary objective of this segmentation is to
Table 4 Mechanism test results ascertain whether the mechanisms previously identified
operate differentially across these disparate locales. Prior
(1) (2) (3)
to the inclusion of mechanism variables, we embark on
Variables tfp tfp tfp
a sub-sample Two-Stage Least Squares (2SLS) regres-
ai 0.2324a 0.1967a 0.2992a sion. The initial column of Table 5 presents the out-
(0.0603) (0.0563) (0.0804) comes derived from this sub-sample 2SLS regression for
ak 0.4452 c
each delineated region. The outcomes, as showcased in
(0.2488) Table 5, underscore several noteworthy findings. Within
al 0.2357b the second column of the regression outcomes for the
(0.1162) eastern region, the coefficient corresponding to digital
ad -15.8174c transformation stands at 0.2719, manifesting statistical
(9.4219) significance at the 1% level. Nevertheless, the coefficient
r2 0.6179 0.6194 0.6158 pertaining to its interaction term, amounting to 0.3586,
N 8470 8470 8470 fails to attain statistical significance. This suggests an
controls Yes Yes Yes absence of the capital allocation efficiency mechanism
firm&year Yes Yes Yes within the context of the eastern region. Shifting focus
Standard error in parentheses, taking clustering robust standard errors. a, b, and c
to the third column of the regression results for the east-
indicate significant at the statistical levels of 1%, 5%, and 10%, respectively ern region, the coefficient affiliated with the interaction
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 10 of 16

Table 5 Regression results of region-based heterogeneity

East Mid-west

(1) (2) (3) (4) (1) (2) (3) (4)

Variables tfp tfp tfp tfp tfp tfp tfp tfp

ai 0.2708a 0.2719a 0.2383a 0.3513a 0.1321 0.2440b 0.2112b 0.2018

(0.0683) (0.0708) (0.0693) (0.0990) (0.0957) (0.1134) (0.0919) (0.1310)
ak 0.3586 1.3323a
(0.2526) (0.5031)
al 0.1767c 0.8934a
(0.1028) (0.1635)
ad -16.2597 -13.6976
(11.2478) (15.8117)
r2 0.6226 0.6244 0.6255 0.6222 0.6115 0.6165 0.6232 0.6114
N 4968 4968 4968 4968 3502 3502 3502 3502
controls Yes Yes Yes Yes Yes Yes Yes Yes
firm&year Yes Yes Yes Yes Yes Yes Yes Yes
a b c
Standard error in parentheses, taking clustering robust standard errors. , , and indicate significant at the statistical levels of 1%, 5%, and 10%, respectively

term between digital transformation and labor alloca- efficiencies, thereby engendering a more profound ame-
tion efficiency emerges as 0.1767, achieving statistical lioration in business efficiency within the central-western
significance at the 10% level. This observation postulates regions. This notable alignment with the findings of Yang
that digital transformation bears the capacity to pro- et al. (2023) study concerning digital transformation’s
pel labor allocation efficiency, consequently fostering an implications for worker wages among divergent regions
enhancement in enterprise performance. In the realm underscores the cogency of our observations. This pro-
of the central-western region, the 2SLS regression out- nounced difference can be attributed to the central-west-
comes initially indicate an insignificance associated with ern regions’ relatively lower economic development level,
the coefficient linked to digital transformation. However, coupled with the prevailing inadequate emphasis on digi-
upon the inclusion of the interaction term, a distinctive tal transformation. This, in conjunction with the compar-
pattern becomes evident. Specifically, the coefficient atively suboptimal capital and labor allocation efficiency
associated with the interaction between digital trans- of enterprises in the central-western regions vis-à-vis
formation and capital allocation efficiency registers at a their eastern counterparts, contributes to this phenom-
statistically significant value of 1.3323, signifying the 1% enon. Hence, a compelling imperative emerges: an eleva-
level of significance. Correspondingly, the coefficient tion of the digital transformation quotient within the
associated with the interaction between digital transfor- central-western regions holds the potential to mitigate
mation and labor allocation efficiency stands at 0.8934, information asymmetry among enterprises. This strategic
also attaining statistical significance at the 1% level. This augmentation can stimulate the optimization of resource
juxtaposition of outcomes underscores a pivotal infer- allocation, thereby catalyzing an efficacious enhancement
ence: the digital transformation of enterprises does not in capital and labor allocation efficiencies. This collective
exert a direct promotion of enterprise performance. augmentation, in turn, confers a heightened impact upon
Instead, its impact is channeled through the elevation operational efficiency within firms, thus substantiating
of both labor and capital allocation efficiencies, sub- the pronounced significance of digital transformation in
sequently bestowing a salient enhancement upon firm the realm of enterprise dynamics.
performance. A discerning comparison of the regression
results between the eastern and central-western regions 4.3.2 Industry‑based heterogeneity analysis
unveils a notable pattern. Specifically, the coefficients Our study categorizes firms into manufacturing and non-
pertaining to interaction terms, as observed within the manufacturing industries. The coefficients of the first col-
first two columns, exhibit comparatively higher values umn of the sub-sample benchmark regression results of
within the central-western regions. This disparity indi- digital transformation in Table 6 are still significantly pos-
cates an intensified role played by digital transforma- itive. Based on the regression results of the second and
tion in the promotion of capital and labor allocation third columns of Table 6 for the manufacturing industry,
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 11 of 16

Table 6 Regression results of industry-based heterogeneity

Manufacturing Non-Manufacturing

(1) (2) (3) (4) (1) (2) (3) (4)

Variables tfp tfp tfp tfp tfp tfp tfp tfp

ai 0.1662b 0.2697a 0.1717b 0.3070a 0.2730a 0.2689a 0.2293b 0.2749c

(0.0674) (0.0838) (0.0679) (0.0938) (0.0962) (0.0948) (0.0951) (0.1524)
ak 1.3642a 0.1793
(0.3970) (0.1934)
al 0.4974b 0.1887c
(0.2088) (0.1106)
ad -25.6366b -4.3538
(11.7182) (16.2737)
r2 0.6562 0.6653 0.6627 0.6569 0.5866 0.5871 0.5903 0.5866
N 4876 4876 4876 4876 3594 3594 3594 3594
controls Yes Yes Yes Yes Yes Yes Yes Yes
firm&year Yes Yes Yes Yes Yes Yes Yes Yes
a b c
Standard error in parentheses, taking clustering robust standard errors. , , and indicate significant at the statistical levels of 1%, 5%, and 10%, respectively

both interaction terms are significant. Specifically, the efficiency and labor allocation efficiency mechanisms
interaction term coefficient between digital transforma- for manufacturing firms. However, only the labor alloca-
tion and capital allocation efficiency is 1.3642 at the 1% tion efficiency mechanism exists for non-manufacturing
significance level, and the interaction term coefficient firms. In the non-manufacturing sector, firms are pri-
between digital transformation and labor allocation effi- marily engaged in service industries, such as finance,
ciency is 0.4974 at the 5% significance level. These results retail, hospitality, and consulting. Unlike manufacturing
suggest that manufacturing firms can enhance product enterprises, non-manufacturing enterprises rely more
quality and maintain market competitiveness by reduc- on human resources and information technology to pro-
ing costs through digital transformation, thus promot- vide services. Therefore, digital transformation plays a
ing capital and labor allocation efficiency, and ultimately crucial role in enabling non-manufacturing enterprises
improving business efficiency. In addition, the interaction to achieve long-distance transmission and globaliza-
coefficient between digital transformation and enterprise tion of services. Through the application of digital tech-
technological progress bias is -25.6366 at a significance nology, enterprises can offer online services, engage in
level of 5%, indicating that when technological progress remote collaboration, and facilitate virtual communica-
is biased towards labor, it is beneficial for digital transfor- tion, thereby delivering high-quality services that pro-
mation to promote enterprise performance. On the other mote business development and enhance performance.
hand, only the third column of the regression results is Moreover, by implementing advanced digital technolo-
significant for non-manufacturing firms, where the inter- gies and automation systems, companies can enhance
action term coefficient between digital transformation the efficiency and accuracy of their production and ser-
and labor allocation efficiency is 0.1887 at the 10% sig- vice processes. This, in turn, helps them reduce costs and
nificance level. increase production capacity, leading to improved per-
The result indicates that the mechanism of capital allo- formance. These are conducive to the digital transforma-
cation efficiency is not significant for non-manufactur- tion of non-manufacturing companies to promote labor
ing firms, and digital transformation improves business allocation and thus improve firm performance.
efficiency by promoting labor allocation efficiency. By
comparing the results of the two industries, we find that
the coefficient of the 2sls regression for the non-manu- 4.3.3 The nature of corporate equity‑based heterogeneity
facturing industry is larger than that for the manufactur- analysis
ing industry. This indicates that the non-manufacturing Our study concludes by conducting regressions based
industry has a greater potential to improve firm per- on the ownership nature of the firms. From Table 7, the
formance through digital transformation. The mecha- first column of regression results indicates that digital
nism test reveals the existence of both capital allocation transformation can promote the operational efficiency
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 12 of 16

Table 7 Regression results of the nature of corporate equity-based heterogeneity

State Enterprise Non-State Enterprise

(1) (2) (3) (4) (1) (2) (3) (4)

Variables tfp tfp tfp tfp tfp tfp tfp tfp

ai 0.2604a 0.2407b 0.2228b 0.4233a 0.0771 0.0855 0.0625 0.0811

(0.0954) (0.0982) (0.0911) (0.1455) (0.0668) (0.0667) (0.0663) (0.0748)
ak 0.5439 0.3412
(0.5156) (0.2644)
al 0.1539 0.2324b
(0.1882) (0.1132)
ad -28.8068c -1.9924
(16.1050) (9.3645)
r2 0.6462 0.6501 0.6483 0.6459 0.6065 0.6076 0.6097 0.6065
N 2071 2071 2071 2071 5106 5106 5106 5106
controls Yes Yes Yes Yes Yes Yes Yes Yes
firm & year Yes Yes Yes Yes Yes Yes Yes Yes
a b c
Standard error in parentheses, taking clustering robust standard errors. , , and indicate significant at the statistical levels of 1%, 5%, and 10%, respectively

of state-owned enterprises (SOEs), with interaction coef- enterprises are characterized by their small size and high
ficients of 0.5439 and 0.1539 for columns 2 and 3, respec- staff turnover, which results in reduced operational effi-
tively, that are not significant, indicating the absence of ciency and increased operating costs. In comparison to
corresponding mechanisms in SOEs. The coefficient of large enterprises, small businesses often face limitations
digital transformation and technological progress bias is in resources and lack economies of scale. Their smaller
-28.8068 at the 10% level, indicating that but when the production scale makes it challenging to achieve pro-
technological progress bias of SOE firms is labor, digital curement and production cost advantages similar to
transformation is able to boost firm performance. The those of larger counterparts. Moreover, high staff turno-
2sls regression result of digital transformation coefficient ver poses another challenge for non-SOE enterprises.
for non-SOEs is 0.0771, but it is not significant. After Intense competition and rapidly evolving industries con-
adding interaction coefficients, it can be observed that tribute to employees frequently changing jobs, resulting
the interaction coefficient of digital transformation and in the loss of skills and experience and impacting the
labor allocation efficiency is 0.2324 at a 5% level of signif- operational efficiency and business continuity of these
icance. This indicates that digital transformation cannot companies. The constant turnover necessitates additional
directly impact non-SOE enterprises, and they need to investments in recruitment, training, and integrating
promote labor allocation efficiency through digital trans- new employees, thereby increasing labor costs. However,
formation to improve firm performance. digital transformation presents non-SOE enterprises with
On the one hand, due to the large scale and complex an opportunity to address these challenges and enhance
organizational structures of state-owned enterprises their business efficiency. Specially, through digital trans-
(SOEs), the process of adapting and implementing digi- formation, non-SOE enterprises have the opportunity to
tal transformation may take longer. Additionally, the revamp their organizational structures to maximize effi-
resource allocation of SOEs can be influenced by politi- ciency, improve capital allocation efficiency, labor allo-
cal factors and stakeholders, which may result in capital cation efficiency, and reduce costs. This transformative
and labor allocation not being fully based on efficiency process can ultimately enhance firm performance.
principles. On the other hand, during the digital trans-
formation process, SOEs may prioritize improving opera- 4.3.4 Size‑based heterogeneity analysis
tional efficiency and optimizing business processes rather Furthermore, our analysis extends to the examination
than solely focusing on the efficiency of capital and labor of heterogeneity based on firm size, the outcomes of
allocation. Therefore, although the efficiency of capital which are presented below. To delineate, we gauge enter-
and labor allocation in SOEs may not show significant prise size relative to the mean size of the sampled com-
changes, digital transformation can still lead to improve- panies: those surpassing the mean are classified as large
ments in operational efficiency. In contrast, non-SOE enterprises, while those falling short constitute small
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 13 of 16

enterprises. The outcomes for large enterprises manifest embrace change and proactively push forward digital
that digital transformation does not directly engender transformation to better adapt to new markets and future
enhanced business performance. Evidently portrayed in challenges. Our study examines the impact of enterprise
the third column of the table, the interaction coefficient digital transformation on operational efficiency based on
between digital transformation and labor allocation effi- panel data in 2009–2020 for Chinese A-share listed com-
ciency for large enterprises is 0.2366, significant at the panies over a continuous 12-year period, using a panel
5% threshold. This observation suggests that substan- fixed-effects model, and conducts relevant robustness
tial enterprises can enhance labor allocation efficiency tests, mechanism analysis, and heterogeneity analysis.
through digital transformation, thereby fostering enter- Empirical results show that enterprise digital transfor-
prise performance. Contrastingly, digital transforma- mation has a significantly positive impact on firm per-
tion wields a more potent influence on small enterprises, formance measured by total factor productivity. In other
with mechanisms involving capital allocation efficiency words, rising digital transformation is conducive to
and labor allocation efficiency coming to the fore. This improving firm performance the China’s listed firms. The
suggests that digital transformation can augment the Mechanism analysis reveals that enterprise digital trans-
resource allocation and labor efficiency of smaller enter- formation mainly promotes capital and labor allocation
prises, ultimately ameliorating their performance. This efficiency to improve their performance. In addition, we
discrepancy may be attributed to the smaller scale of find that firms’ technological advances are biased toward
these entities and the heightened personnel turnover, labor when it comes to favoring digital transformation
culminating in inefficiencies and elevated operational for firm performance.
costs. However, through digital transformation, smaller From a regional perspective, the sub-regional bench-
enterprises can reimagine their corporate organizational mark regressions indicate that firms in both the Eastern
framework, optimizing efficiency, curtailing costs, and and Midwestern regions are capable of enhancing their
elevating operational efficacy. Importantly, this analysis performance through digital transformation. There is a
mirrors our exploration of both SOEs and non-SOEs, mechanism role of labor allocation efficiency in the east-
thus substantiating the dependability of our findings ern region, while there is a mechanism role of resource
(Table 8). allocation efficiency and labor allocation efficiency in
the central and western regions, i.e., the enterprises in
5 Conclusions and policy implications the central and western regions are able to promote the
In the rapidly evolving technological landscape of the enterprise resource allocation efficiency and labor allo-
world, with continuous changes in the global competi- cation efficiency through digital transformation, which
tive scenario and increasing uncertainty in the economic in turn promotes the performance of the enterprises.
environment, it is imperative for enterprises to actively However, a comparison between the two regions reveals

Table 8 Regression results of size-based heterogeneity analysis

Large Small

(1) (2) (3) (4) (1) (2) (3) (4)

Variables tfp tfp tfp tfp tfp tfp tfp tfp

ai 0.1392 0.1608c 0.1113 0.1602 0.1777c 0.2248b 0.3185a 0.3027c

(0.0895) (0.0924) (0.0886) (0.1070) (0.0908) (0.1000) (0.1078) (0.1574)
ak 0.3138 1.4591b
(0.2717) (0.6143)
al 0.2366b 0.9374a
(0.1206) (0.2005)
ad -9.1833 -17.6084
(10.5107) (16.4757)
r2 0.5512 0.5520 0.5562 0.5510 0.6463 0.6532 0.6544 0.6457
N 3555 3555 3555 3555 3999 3999 3999 3999
controls Yes Yes Yes Yes Yes Yes Yes Yes
firm & year Yes Yes Yes Yes Yes Yes Yes Yes
Standard error in parentheses, taking clustering robust standard errors. a, b, and c indicate significant at the statistical levels of 1%, 5%, and 10%, respectively
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 14 of 16

that digital transformation plays a more substantial role adverse impact of digital transformation on businesses in
in promoting capital allocation efficiency and labor allo- the Mechanisms Test, we have not extensively explored
cation efficiency, leading to improved firm performance, this aspect. However, to pave the way for future explo-
in the Midwest compared to the East. From an industry rations, it would be judicious to consider avenues that
perspective, the coefficient of digital transformation is delve into the realms pertaining to the potential adverse
significantly positive for both manufacturing and non- implications of digital transformation on firm perfor-
manufacturing firms, as demonstrated by the industry mance. This would offer a more comprehensive and
2sls regression. This implies that both types of firms can well-rounded perspective on the multifaceted interplay
enhance their performance through digital transforma- between digital transformation and corporate dynamics.
tion. Regarding the mechanism involved, there is a cor- Our study delves deeply into the ramifications of digi-
responding mechanism role for manufacturing firms. tal transformation on both resource allocation efficiency
Through digital transformation, manufacturing firms are and labor allocation efficiency within enterprises. By
able to improve their labor allocation efficiency and capi- uncovering the intricate interplay among these distinct
tal allocation efficiency, consequently promoting their components, our research offers invaluable insights that
overall performance. At the same time, the technological shed light on real-world business operations and the for-
progress of manufacturing companies is biased towards mulation of strategic initiatives. The findings of our study
labor, which is conducive to digital transformation for suggest that the government should not only encourage
firm performance. Conversely, the mechanism of capital companies to enhance their digital transformation efforts,
allocation efficiency for non-manufacturing enterprises is but also promote the widespread adoption of digital tech-
not significant. Digital transformation, in this case, pri- nologies to provide the necessary technical support and
marily enhances the operating efficiency of enterprises resources for successful implementation. It is essential to
by promoting labor allocation efficiency. From the per- support companies in strengthening their talent devel-
spective of the nature of equity, the benchmark regres- opment and management capabilities to ensure the suc-
sion coefficient of digital transformation for state-owned cess of digital transformation initiatives. Specifically, the
enterprises is significantly positive. However, there is no government needs to prioritize the digital transforma-
mechanism effect observed for capital allocation effi- tion of enterprises in the central and western regions. The
ciency and labor allocation efficiency. The technologi- government should enhance the digital infrastructure in
cal progress of SOE firms is biased towards labor, which these areas, including network coverage and data center
facilitates digital transformation for business perfor- construction. This will provide stable and high-speed digi-
mance. In contrast, the coefficient of the 2sls regression tal environments, fostering the application and innovation
for digital transformation in non-state-owned enterprises of digital technologies. Furthermore, implementing corre-
is positive but statistically insignificant. Nevertheless, a sponding policies and regulations to create a more flexible
mechanism role is present for non-state-owned enter- and favorable business environment for enterprises in the
prises, suggesting that digital transformation does not central and western regions is crucial. For instance, sim-
directly affect them. Instead, it needs to promote the plifying administrative approval procedures and reducing
efficiency of capital and labor allocation through digi- related tax burdens can foster a conducive operating envi-
tal transformation to improve firm performance. Het- ronment. The government should also intensify efforts
erogeneity analysis based on firm size reveals that digital in intellectual property protection to enhance innova-
transformation exerts a more pronounced influence on tion incentives and competitiveness among enterprises
small-scale firms. It facilitates the enhancement of their in these regions. Additionally, the government can facili-
resource allocation efficiency and labor allocation effi- tate collaboration and cooperative innovation between
ciency, consequently leading to performance improve- non-manufacturing enterprises and universities, research
ment. Conversely, for large-scale enterprises, the role institutions, and technology parks. This joint effort will
of digital transformation is primarily centered around foster technology exchange and cooperation, promoting
resource allocation efficiency. digital transformation projects. All of these contribute to
The limitations inherent in our paper stem from the improving the performance of the company.
potential presence of a minor reverse causality between Additionally, companies need to comprehensively con-
digital transformation and total factor productivity, not- sider their current situation and future development, for-
withstanding our meticulous incorporation of pertinent mulate scientifically reasonable digital transformation
control variables. Moreover, it’s important to note that plans and strategies, strengthen research and application
our current research predominantly centers on elucidat- of digital technologies, and establish a digital talent pool.
ing the affirmative influence of digital transformation on Firstly, enterprises need to comprehensively consider
firm performance. While we have briefly discussed the their current situation and future development, including
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 15 of 16

assessing internal resources and capabilities, researching Acemoglu, D., and P. Restrepo. 2018. The race between man and machine: impli-
cations of technology for growth, factor shares, and employment. American
market demand and competitive environment, as well as Economic Review 108 (6): 1488–1542.
evaluating the prospects of digital technology applications. Allen N. Berger, Iftekhar Hasan, Mingming Zhou. 2009. Bank ownership and
It is also important to objectively assess the efficiency of efficiency in China: What will happen in the world’s largest nation?. Journal
of Banking and Finance 33 (1): 113-130.
resource allocation and labor allocation within the organi- Babina T., Fedyk A., He A. X., and Hodson J. 2021. Artificial intelligence, firm
zation. This ensures that the digital transformation plan growth,and industry concentration [R]. Working Paper.
aligns with the actual situation of the enterprise, making Bakhshi H., Bravo-Biosca A., and Mateos-Garcia J. 2014. The analytical firm: estimat-
ing the effect of data and online analytics on firm performance. Nesta Working
it feasible and sustainable. Secondly, enterprises need to Paper. 1–3.
establish clear and specific digital transformation goals and Bessen J., and Righi C. 2019. Shocking technology: what happens when firms make
indicators to measure the effectiveness and outcomes of the large IT investments [R]. SSRN Working Paper.
Bloom, N., S. Raffaella, and V.R. John. 2012. Americans do it better: US multina-
transformation. These goals should be specific, quantifiable, tionals and the productivity miracle. American Economic Review 102 (1):
and aligned with the core business and strategic objectives 167–201.
of the enterprise. When formulating digital transformation Chen S., Li S., Liu Q., Wang C. 2023. Fertilizing the soil: FinTech development and
corporate digital transformation. Digital Economy and Sustainable Develop-
plans and strategies, enterprises should consider digital ment 1: article 1.
technologies and tools that are suitable for their own opera- Diamond, P.A. 1965. Disembodied technical change in a two-sector model. The
tions and integrate them closely with their core business. Review of Economic Studies 32 (2): 161–168.
Einav L., and Levin, J. 2014. Economics in the age of big data. Science
Strengthening research and application of digital tech- 346(6210):1243089.
nologies may involve areas such as artificial intelligence, Ferreira, J., and C. Fernandes. 2017. Resources and capabilities effects on firm
big data analytics, and the Internet of Things. Collabora- performance: What are they? Journal of Knowledge Management 21 (5):
1202–1217.
tion with research institutions and technology vendors Fungáčová Zuzan, Pierre Pessarossi, Laurent Weill. 2013. Is bank competition
can facilitate active technological research and develop- detrimental to efficiency? Evidence from China. China Economic Review 27:
ment to meet business needs and maintain competitive 121-134.
Goldfarb, A., and C. Tucker. 2019. Digital economics. Journal of Economic Literature
advantages. Lastly, enterprises need to establish a digital 57 (1): 3–43.
talent pool to drive the effective implementation of digital Gordon Lawrence A, Loeb Martin P, Tseng Chih-Yang. 2009. Enterprise risk
transformation. This includes recruiting and developing management and firm performance: a contingency perspective. Journal of
Accounting and Public Policy 28 (4): 301–327.
talents with digital skills and knowledge, fostering digital Hajli, M., J.M. Sims, and V. Ibragimov. 2015. Information Technology Productivity
awareness and capabilities among internal employees, and Paradox in the 21St Century. International Journal of Productivity and Perfor-
establishing effective mechanisms for team collaboration mance Management 64 (4): 457–478.
Hanousek, J., A. Shamshur, and J. Tresl. 2019. Firm efficiency, Foreign Ownership
and knowledge sharing. These efforts will facilitate the suc- and CEO Gender in Corrupt Environments. Journal of Corporate Finance. 59:
cessful implementation of digital transformation, leading to 344–360.
improved firm performance and sustainable development. Kaur J., and Sood L. 2015. Comparison between various types of adder topologies.
Khatib, S., and A.I. Nour. 2021. The impact of corporate governance on firm per-
formance during The COVID-19 pandemic: Evidence from malaysia. Journal
of Asian Finance Economics and Business. 8 (2): 0943–0952.
Author contributions
Kim, J., Y.H. Li, and L.D. Zhang. 2011. Corporate tax avoidance and stock price crash
All authors jointly completed this research study, and there is no confict of inter-
risk: Firm-level analysis. Journal of Financial Economics. 100 (3): 639–662.
est between the authors. The authors read and approved the final manuscript.
Lasi H., Fettke P., Kemper H. G., Feld T., and Hoffmann M. 2014. Industry 4.0. Busi-
ness and Information Systems Engineering 6: 239–242.
Funding
Levinsohn, James, and Amil Petrin. 2003. Estimating production functions using
This paper is financially supported by the National Natural Science Foundation
inputs to control for unobservables. The Review of Economic Studies 70 (02):
of China (72071094).
317–341.
Chen Chuanglian is grateful for the funding support provided by the National
Lindstedt, C., and D. Naurin. 2010. Transparency is not enough: making transpar-
Natural Science Foundation of China (72071094).
ency effective in reducing corruption. International Political Science Review
31 (3): 301–311.
Declarations Loebbecke, C., and A. Picot. 2015. Reflections on societal and business model
transformation arising from digitization and big data analytics: a research
Competing interests agenda. Journal of Strategic Information Systems 24 (3): 149–157.
All authors of this article have no objections to the ranking of the authors and Loonam, J., S. Eaves, V. Kumar, and G. Parry. 2018. Towards digital transformation:
there are no conflicts of interest. Lessons learned from traditional organizations. Strategic Change 27 (2):
101–109.
Michiels Anneleen, Wim Voordeckers, Nadine Lybaert, Tensie Steijvers. 2013.
Received: 1 June 2023 Revised: 14 August 2023 Accepted: 28 August 2023 CEO Compensation in private family firms:Pay-for-performance and the
moderating role of ownership and management. Family Business Review 26
(2): 140-160.
Mikalef, P., and A. Pateli. 2017. information technology-enabled dynamic capabili-
ties and their indirect effect on competitive performance: Findings from
References PLS-SEM and FSQCA. Journal of Business Research. 1: 1–16.
Abdul Razak Abdul Hadi, Hafezali Iqbal Hussain, et al. 2018. Bank’s performance Mikalef, P., I.O. Pappas, J. Krogstie, and M.N. Giannakos. 2018. Big data analytics
and its determinants–evidence from Middle East, Indian Sub-continent and capabilities: A systematic literature review and research agenda. Information
African Banks. Polish Journal of Management Studies 17 (07): 17–26. Systems and e-Business Management 16: 547–578.
Chen et al. Digital Economy and Sustainable Development (2023) 1:18 Page 16 of 16

Pucheta-Martínez, M.C., Gallego-Álvarez, I. 2020. Do board characteristics drive

firm performance? An international perspective. Review of Managerial Sci-
ence, 14(6):1251–1297.
Verhoef, P.C., T. Broekhuizen, Y. Bart, et al. 2021. Digital transformation: a multidis-
ciplinary reflection and research agenda. Journal of Business Research. 122
(1): 889–901.
Yang P, Riepe J, Moser K, et al. 2019. Women directors, firm performance, and
firm risk: a causal perspective. The Leadership Quarterly 30 (5): 101297.
Yang, G., Yao, S., Dong, X. 2023. Digital economy and wage gap between high-
and low-skilled workers. Digital Economy and Sustainable Development 1:
article (7).
Yao, S. 2023. Editorial introduction, digital economy and sustainable develop-
ment. 1: article (2).
Yoo, Y., J.B. Richard, and Jr., Kalle L., and Ann M.,. 2012. Organizing for innovation in
the digitized world. Organization Science 23 (5): 1398–1408.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.