Can state ownership help facilitate the digital transformation of private sector enterprises? Evidence from China

Descriptive statistics of research variables

Table 2 reports the descriptive statistical results of the research variables.

Table 2 shows that the mean value and standard deviation of DT are 1.374 and 2.358, respectively, which indicates that the digital transformation degree of the sample firms is generally not high, and there are still big differences in the transformation capabilities of different firms. The mean value of State1 is 0.526, indicating that more than half of the sample firms have state-owned shareholders. The mean value of State2 is 0.159, which shows that with the continuous advancement of “mixed reform,” state-owned shareholders have become an important force that cannot be ignored among PSEs. The mean values of ROA and Lev are 0.067 and 0.485, respectively, indicating that the profitability of the sample firms is generally good, and liabilities are also controlled in a reasonable range. The mean value of SCD is 0.193, accounting for nearly 20%, which shows that in recent years, under the background that the state has vigorously promoted the innovation and development of enterprises, more talent with a technical background in science and engineering has served as directors in PSEs. The mean value of TOP1 is 0.284, which indicates that the largest shareholder of the firm has a strong influence.

Regression results

The impact of state ownership on the digital transformation of PSEs

Table 3 reports the regression results for Hypothesis 1.

Columns (1) and (3) in Table 3 only test the variables State1 and State2 by regression. The results show that the coefficients of State1 and State2 are significantly positive at least at the level of 5%. After adding a series of control variables in columns (2) and (4), the coefficients of State1 and State2 are significantly positive at least at the level of 10%, indicating that PSEs with state ownership can effectively help their digital transformation, verifying Hypothesis 1. Among the control variables, the coefficient of SCD is significantly positive at least at the level of 10%, which indicates that the higher the proportion of directors with a technical background in science and engineering, the more favorable it is for the implementation of the digital transformation strategy. The coefficients of the other controlled variables do not show statistical significance in the regression.

The impact of state ownership with different characteristics on the digital transformation of PSEs

At present, the state ownership holders in PSEs can be divided roughly into the following two types: (1) The State-owned Assets Supervision and Administration Commission (SASAC) or other organizations under the jurisdiction of governments at all levels; and (2) SOEs. Specifically, category (1) state ownership holders usually have a strong official color. The behavior of such state-owned shareholders reflects their specific will to follow government policy. Under the assumption that the state actively wants to help the digital transformation of PSEs, such state ownership in PSEs can be considered an important way for the state to support its digital transformation strategy; as such, they are regarded as strategic investors. In category (2), state ownership holders are mainly SOEs that are looking for financial performance; their official color has faded. Such state ownership in PSEs is based more on economic considerations; thus, they are regarded as financial investors. If a PSE has both of these types of state owners, given the background that the state has encouraged PSEs to pursue high-quality development in recent years, SOEs need to actively respond to the call of the state and actively cooperate with the government’s development policy; thus, this kind of mixed state ownership is also regarded as functioning as a strategic investor. Table 4 reports the regression results for Hypothesis 2.

Columns (1) and (2) in Table 4 show that among the strategic investors, the coefficients of State2 are all significantly positive at the level of 1%. Columns (3) and (4) show that among the financial investors, the coefficient of State2 does not pass the statistical significance test. The results indicate that as a strategic investor, state ownership plays a significant role in helping the digital transformation of PSEs. Correspondingly, state ownership as a financial investor does not help their digital transformation. This verifies Hypothesis 2. State ownership in PSEs that functions as a strategic investor is based on China’s current development policy. By giving play to their roles in resource provision and governance, governments at all levels offer effective support to PSEs in implementing digital transformation. However, SOE state owners in PSEs as financial investors, based on their own economic needs, pay more attention to the short-term financial performance of these PSEs. Thus, they pay scant attention to major strategies involving the long-term development of enterprises, such as digital transformation.

The impact of state ownership on the digital transformation of PSEs in different business environments

For the assessment of the local business environment in which the firm is situated, we referred to the 2018 “Urban Business Environment Index” constructed by Li (2019) as an important basis for our grouping. Specifically, the samples with an index score higher than or equal to the average value in the city where the firm is located are regarded as cities with good business environments, while those with an index score lower than the average are regarded as cities with poor business environments. In addition, the regression results for Hypothesis 2 confirm that state ownership that functions as a financial investor does not play a significant role in helping the digital transformation of the PSEs. To better analyze the impact of state ownership on the digital transformation of PSEs, in Table 5 and the following empirical examination, we delete the sample where state ownership functions as a financial investor. Table 5 reports the regression results for Hypothesis 3.

We can see in columns (1) and (3) in Table 5 that although the coefficients of State1 and State2 are shown as positive in the regression, they are not statistically significant. Columns (2) and (4) show that the coefficients of State1 and State2 are both significantly positive at the level of 1%. This indicates that state ownership plays a more obvious role in helping the digital transformation of PSEs in cities with better business environments, thus verifying Hypothesis 3. In fact, in cities with poor business environments, local governments generally lack sufficient technical and human resources, and the financial state of these local governments is relatively constricted. Under such circumstances, it is difficult to provide substantial support for PSEs to implement digital transformation strategies through state ownership participation. In cities with good business environments, research institutes and universities are often concentrated there. This can provide the necessary manpower and technical support for enterprises to implement a digital transformation, with local governments often having rich financial resources. Amid the backdrop of the nation’s robust digital economy advancement, state ownership participation can be an effective means to bolster the digital transformation of private enterprises, consequently fostering the high-quality development of local industries.

Further analysis

The impact of state ownership on the digital transformation of PSEs under uncertain economic policies

Economic uncertainty pertains to the challenges faced by economic entities in accurately forecasting the timing, nature, and manner of potential adjustments in the current economic policies of the government (Gu et al. 2021). Due to the relatively less robust standing of PSEs than of SOEs in China, PSEs tend to be more responsive to changes in economic policies. Economic policy uncertainty often exerts a more pronounced influence on their strategic decision-making processes. When PSEs are faced with a high policy uncertainty, evaluating policy directions becomes more difficult, thereby raising investment risk and making their development prospects murky (Ma and Hao, 2022). When faced with a lack of investment confidence, enterprises often opt for conservative management strategies, which may involve postponing initial investment plans or scaling back investment endeavors (Cui et al. 2021). For PSEs undergoing digital transformation, the reduction or even interruption of financial subsidies and financing preferences caused by economic policy uncertainty will increase their business risks, thus adversely affecting the implementation of their digital transformation strategy. In that case, the private controlling shareholders, whose business goal is to maximize profits, will be reluctant to invest too much in a digital transformation strategy—namely, a large investment with high risk—based on possible business risks in the future. To investigate whether state ownership remains a substantial factor in aiding the digital transformation of PSEs amid economic policy uncertainty, we introduced the EPU variable. Table 6 reports the corresponding regression results.

We can see in column (1) in Table 6 that the coefficient of EPU is significantly negative at the 1% level, indicating that economic policy uncertainty is not conducive to the implementation of the digital transformation strategy of PSEs. After adding the variables of state ownership into the model, columns (2) and (4) show that the coefficient of EPU is significantly negative at least at the 5% level, and the coefficients of State1 and State2 are significantly positive at the 1% level. To explore whether economic policy uncertainty will affect the digital transformation of PSEs with state ownership, we added interaction terms EPU×State1 and EPU×State2 in columns (3) and (5), respectively. The results show that the coefficients of State1 and State2 are still significantly positive at the 1% level, while those of EPU are significantly negative at the 10% level, and the coefficients of EPU×State1 and EPU×State2 are significantly positive at least at the 5% level, indicating that the risks caused by economic policy uncertainty are not conducive to the implementation of PSE digital transformation. However, state ownership can avoid the adverse impact of policy risks on the digital transformation of PSEs to a certain extent, and the degree of risk avoidance can reach 84.9% (0.186/0.219) and 80.2% (0.227/0.283), respectively.

State ownership in PSEs can provide solid support for enterprises in implementing their digital transformation strategies. On the one hand, by virtue of information superiority, state ownership can eliminate the long-standing information asymmetry problem of PSEs, help enterprises access relevant information on macro-policy adjustments in time, enhance the accuracy of future risk assessment, and reduce the impact of policy uncertainty on business operations, thus ensuring the smooth implementation of digital transformation. On the other hand, the resource advantages brought by state ownership in PSEs are also particularly important to stabilize their digital transformation strategy and related investments. Under the impact of economic policy uncertainty, to avoid business risks, enterprises will take the initiative to cut or even give up some investments. As a high-risk characteristic, an investment related to digital transformation will easily become the first choice for enterprises to cut. As an important national strategy, the government departments behind the state-owned shareholders will make every effort to continuously provide enterprises with various resources, such as funds and equipment, for digital transformation to prevent PSEs from reducing their related investment and strengthen their digital construction.

The impact of state ownership on the digital transformation of PSEs with different industry characteristics

Digital transformation is not a simple technological change, but also the overall transformation of enterprise production and management mode and values. The implementation of this strategy usually has the characteristics of long-term uncertainty (Vial, 2019), thus profoundly affecting the strategic choice of digital transformation among PSEs with different industry characteristics. From the perspective of the enterprise’s internal environment, enterprises engaged in traditional industries compared with those in emerging industries have a better foundation for implementing digital transformation. By establishing a big data processing center and integrating digital resources with technical data, the transmission efficiency of information within enterprises can be effectively improved, and the operation of enterprises can be monitored in real-time, which can significantly reduce operating costs, thus ensuring sound development of the enterprise (Jiang et al. 2022).

Viewed from an external enterprise environment standpoint, the primary objective of a digital transformation strategy for emerging businesses is to judiciously leverage digital technology to elevate their market competitiveness and refine their industrial framework (Zhang et al. 2023). The application of artificial intelligence, big data, and other information technologies by emerging enterprises can effectively reduce information costs and achieve coordinated development through the exchange of information across enterprises (Li et al. 2018). Additionally, emerging enterprises are actively committed to digital construction, which can obtain the latest market information, effectively alleviate long-standing information asymmetry between enterprises and the market, and quickly adjust their own development strategies to enhance their market competitiveness.

According to Huang et al. (2016) industry standard groupings that enterprises belong to, the industries of the sample firms are divided into emerging industries and traditional industries, to examine the impact of state ownership on the digital transformation of PSEs with different industry characteristics. Table 7 reports the corresponding regression results.

Columns (1) and (3) in Table 7 show that although the coefficients of State1 and State2 are positive, they are not statistically significant. The results of columns (2) and (4) show that the coefficients of State1 and State2 are all significantly positive at the 1% level, which indicates that state ownership in PSEs in emerging industries helps their digital transformation. Compared with PSEs engaged in traditional industries, digital transformation is more important for enterprises in emerging industries, and state ownership in such enterprises also plays a more significant role in helping their digital transformation. On the one hand, new and high-tech-driven enterprises need digital technology to establish their competitive advantage in the industry and gain more from digital transformation. By implementing a digital transformation strategy, emerging enterprises will have a stronger incentive to improve their capacity building, thus enhancing the supporting effect of state ownership in such enterprises. On the other hand, China is undergoing a phase of economic transformation and progress, and it urgently needs more emerging technology enterprises that can create high-added value. Diverse policies are also favoring these enterprises to drive the high-quality development of the national economy. According to the policy of “mixed reform” issued by the state recently, we can see that state ownership should invest in non-public enterprises with great development potential and support extensive, new technology. Under the guidance of relevant policies, state ownership in PSEs engaged in emerging industries will better show the “supporting hand” of the government, thus “escorting” the implementation of a digital transformation strategy for such enterprises.

The impact of state ownership on the innovation performances of PSEs

The previous regression results confirm that state ownership in PSEs has a significant positive impact on their digital transformation. Nevertheless, can state ownership enhance the innovation performance of PSEs while simultaneously facilitating their digital transformation? To illustrate the positive consequences of state ownership in enabling the digital transformation of enterprises, we referred to the research methods of Hao et al. (2020) and built the following simultaneous equation model for regression analysis:

In the above model, we first divide PSEs into “Group with state ownership” and “Group without state ownership” according to the principle of whether they have state ownership and then use the three-stage least square (3SLS) estimation for regression analysis. If the coefficient of the variable DT is significantly positive in the regression, it means that with the continuous advancement of the digital transformation of the PSEs, their innovation performance effectively improves. Moreover, to reflect the mediating effect of digital transformation of PSEs between state ownership and innovation performance, the innovation performance of PSEs is measured by the number of invention patents and the sum of design patents and utility model patents in each year. In addition, referring to the research of Li et al. (2020), we also use the variable innovation performance by one lag period (Patent_i,j,t+1) for the regression. Table 8 reports the corresponding regression results.

Columns (1) and (3) in Table 8 show that the coefficients of variable DT are all significantly positive at the 1% level, indicating that the continuous advancement of the digital transformation of PSEs can improve their innovation performance. In the group of PSEs with state ownership, the coefficient of DT is 0.451, which is significantly higher than that of DT in the group of PSEs without state ownership of 0.036; the difference test shows that it is highly significant, indicating that PSEs with state ownership has effectively improved their innovation output because of their faster digital transformation. In addition, columns (2) and (4) show that the coefficient of the variable Patent is significantly positive at the 1% level, which indicates that although there is some endogeneity between digital transformation and innovation performance, the 3SLS estimation in this study can largely avoid this problem. In sum, state ownership can accelerate the digital transformation of PSEs, thus improving their innovation performance.

Endogenous test

The analysis shows that state ownership in PSEs significantly supports their digital transformation. However, it is necessary to consider potential endogeneity issues. Some PSEs may implement digital transformation strategies more effectively, which could attract government support through state ownership. This could result in reverse causality in the regression results. To address this concern and account for omitted variable bias, we use the instrumental variable (IV) method, as suggested by Yao et al. (2018).

In this study, we use the total market value of SOEs in the cities where the sample firms are located and its natural logarithm as the instrumental variable (SV). This variable reflects the achievements of market-oriented reforms and is positively related to state ownership. At the same time, it does not directly influence the digital transformation of PSEs, making it an appropriate instrumental variable.

We apply the two-stage least squares (2SLS) method to handle potential endogeneity (Batrancea, 2022, Batrancea et al. 2023). We assume that the SV is an instrumental variable with a positive correlation with state-owned equity because the total market value of a city’s SOEs can better reflect the achievements made by the local SOEs through market-oriented reform. If the city’s SOEs have a larger total market value, they have a stronger influence in the local area and can directly participate in local PSEs. Moreover, there is no direct relationship between this variable and the digital transformation of PSEs; thus, it is completely suitable as an instrumental variable in this study. Table 9 reports the results of the weak instrumental variable and endogenous test in the first stage of the instrumental variable.

Table 9 shows the results of the weak instrumental variable and endogenous tests in the first stage. The Cragg-Donald value is 18.795, which is greater than the Stock-Yogo bias critical value of 9.834 (15%). This means the hypothesis of a weak instrumental variable is rejected. The Durbin-Wu-Hausman test gives a chi-square value of 5.463 (p = 0.029), showing that state ownership variables are endogenous. This confirms the need for instrumental variables in the analysis.

Table 10 presents the 2SLS regression results using instrumental variable. Columns (1) and (2) show the first stage of the analysis, where the coefficients of SV are significantly positive at the 1% level. Columns (3) and (4) show the second stage, where the coefficients of State1 and State2 are also significantly positive at the 1% level. These results are consistent with the earlier findings. This confirms that the previous empirical results are robust and unaffected by endogeneity.