Do State Ownership Imprints Affect Innovation in Family Firms? The Estimation and Inference of a Panel Model with a Time Trend

Abstract

In the realm of China’s bustling economy, a fascinating dynamic exists between family businesses and state-owned enterprises, characterized by a symbiotic relationship and collaboration. Many family firms today originate from state-owned enterprises. Hence, their earlier ownership institution can influence privatized family firms. However, the impact of this privatization on long-term strategic orientation is not clear. This study endeavors to shed light on this complex matter through meticulous examination, employing estimation and inference techniques through the use of a panel model with a discernible time trend. Combining both imprinting theory and institutional logic theory, this study finds that state ownership imprints intervene in the strategic outcomes of family firms: Privatized family firms induce a lower level of innovation as compared to non-privatized family firms. This research also finds that intergenerational succession weakens this effect while the proportion of state-owned shares strengthens it. Robustness tests, utilizing the PSM method, have been conducted to validate the credibility and reliability of the findings obtained through this study. The findings of this research serve as a testament to the ever-evolving dynamics and interconnectedness prevalent within the intricate tapestry of China’s economic landscape.

1. Introduction

In the course of China’s economic reform, the institutional environment, as well as the nature and structure of organizations, have undergone tremendous change. In the 1990s, a large number of state-owned enterprises (SOEs) and collective-owned enterprises (COEs) were privatized into firms controlled by individuals and families [1,2,3]. This privatization transaction emphasized efficiency and cost control, and it was expected to enhance business vitality and competitiveness. However, in many cases, business privatization has failed in the past. For instance, the privatization of SOEs in Latin American countries, the Philippines, and other economies led to the long-term “middle-income trap” in the 1980s. In fact, these countries lost the opportunity to realize industrialization [4].

Prior research on family business innovation offers diverse perspectives. According to several studies on the willingness of family business members, these members focus on non-economic factors and have the willingness to be innovative with conservatism [5,6,7]. However, according to the research on the resources of family businesses, some recent studies have emphasized the unique resources and advantages of family firms in innovation, such as the deployment of valuable, firm-specific tacit knowledge and superior access to feedback from a trusted network [8,9,10,11]. Yet, few studies discuss whether historical ownership imprints, such as the previous management style or goal orientation in a pre-privatization historical context, can affect the long-term strategy of the family business [12]. In fact, the change in ownership is an important source of innovation [13]. Munari & Oriani (2005) [14] argued that privatized family firms would be influenced by the previous ownership institution and governance model, and this impact may be manifested for a long time. Although some scholars have found that privatization has a positive effect on innovation for SOEs, few studies discuss the effect of ownership privatization on family firms [13]. Similar to the SOEs’ privatization process, or because of this privatization process, many family firms originate from SOEs. What kind of impact would the management philosophy and practice of state ownership have on family firms? This study further expands the understanding and research context of family business innovation by analyzing the impact of ownership privatization. This research answers the following two questions. First, do privatized family firms and privately owned family firms—which have a different history-informed origin—lead to differences in innovation? Second, what is the boundary mechanism of the impact of ownership privatization on innovation?

To answer these questions, this study selected a dataset of Chinese-listed family firms, including characteristics such as the demographics of family members, family ties, governance structure, and financial characteristics of family firms. By using these data, this study can identify the different sources of listed family firms and compare the differences in innovation between privatized and non-privatized family firms. Finally, this research used 9576 observations of Chinese-listed family firms from 2007 to 2020 and estimated using the fixed-effect model for panel regression.

The findings of this study lead to two theoretical contributions. First, it contributes to an in-depth understanding of the innovative behavior of family firms by identifying the heterogeneity of family business innovation. This study helps to improve the understanding of imprinting theory by examining how the institutional logic of external environments affects the strategic behavior of family businesses during the privatization transaction period. By studying the effect of state ownership imprints on family firms’ innovation, this research provides a novel and valuable research perspective for the strategic decisions of family business innovation from both historical and institutional perspectives.

Second, this study furthers the understanding of imprinting theory by examining the boundary mechanisms of state ownership imprints on the innovation strategies of family firms. It focuses on the moderating effects of family intergenerational succession and the proportion of stated-owned shares, further exploring the mechanisms that influence the maintenance and attenuation of state ownership imprints of family firms. This study proposes that the organization is a wrestling arena for members with different types of institutional logic, such as family interest and state ownership tendentiousness. The wrestling can determine whether the imprint is maintained or fades away. The study is a response to the contemporary theoretical call for more understanding of the imprint’s decay and the institutional complexity of family firms [15].

2. Theoretical Aspect and Hypothesis Development

2.1. Theoretical Aspect

Institutional logic refers to the “socially constructed, historical pattern of material practices, assumptions, values, beliefs, and rules” that prescribe and give meaning to social action [16,17]. Institutional logic provides organizational members with principles that guide daily action. The logics correspond, at least partly, to historical imprints of past environments [18]. Conflict arises when historical imprints are challenged by new institutional logic.

Marquis and Tilcsik (2013) [18] define imprinting “as a process whereby, during a brief period of susceptibility, a focal entity develops characteristics that reflect prominent features of the environment, and these characteristics continue to persist despite significant environmental changes in subsequent periods”. According to imprinting theory, characteristics of entity objects are compatible with the salient features of the external environment during sensitive periods. Even if the external environment changes afterward, these characteristics continue to influence the behavior of the focal entity and do not disappear quickly [15].

Prior studies have found that long-term corporate strategy decisions can be influenced by various sources of imprinting, including prior career imprinting, founders’ novelty imprinting, and politicians’ ideological imprinting. However, these studies focus on the social–historical level [19] or the individual–historical level imprints of political leaders or entrepreneurs [20,21]. Few studies have concentrated on organizational-level investigations that analyze the effect of the institutional ownership (initial controlling shareholder) imprint on privatized family business strategy.

According to Marquis and Tilcsik (2013) [18], the imprinting effect consists of three key components: the sensitive period, the formation of the imprint, and the persistence of the imprinting effect. The sensitive period is the period when the entity object is most vulnerable to the external environment. The sensitive period for the formation of the organizational imprint of SOEs in China can be divided into two periods. The first is the period of founding since 1949 [22]. In this period, the business concept, organizational structure, and labor relations of the enterprise were shaped by communist ideology. The second is the public–private partnership transformation [23] since 1978. Especially after 1992, the Chinese government emphasized that SOEs should establish modern operations, such as a shareholding system, separation of the government and enterprises, and other modern corporate governance practices, thus promoting a major wave of privatization of Chinese SOEs [24].

The formation of imprinting consists of three elements: the imprinter, the imprinted object, and the imprinting process [25]. The imprinters are the environmental factors that influence and shape the characteristics of the imprinted object during the sensitive period. The imprinted object is the entity that is influenced by environmental factors [18]. The imprinting process refers to the occurrence of imprinting during the sensitive period [15]. Most SOEs in China were founded by the government in the 1950s [26], thus forming a communist ideology imprint during their sensitive period. As the imprinter, the government established a “unitary” production organization.

The unitary organization has two distinctive features in terms of labor–capital relations: on the one hand, the SOEs provide lifelong employment (“iron rice bowls”), resulting in low labor mobility [27]. On the other hand, by internalizing social services into the unit (so-called “social management”), SOEs provide employees with a set of universal services for old age, sickness, and death [28]. In this implicit contract, managers tend to be less likely to engage in risky, innovative activities; instead, they focus on government policy goals and maintain employment stability [29].

The third component of the imprinting effect is the persistence of the imprint. Once established, two central forces keep imprints in place [25]. First, internal (e.g., organizational governance) and external (e.g., economic institution) structural forces create inertia or path dependence, which can lead to a “lock-in” pattern of imprints and limit the organizations’ ability to change core characteristics [30]. Second, the institutionalized practices and procedures become taken-for-granted norms [31,32]. Thus, state ownership, which has been implemented for 30 years, will become an institutional logical imprint, and it can continue to impact strategy [33], structure [34], decision-making rules [35], cultural values and identity [36], and the development and deployment of knowledge and capabilities [37]. However, the boundary conditions of the imprints are not clear. For instance, little attention has been paid to whether the previous imprint will be weakened in the course of the firm’s ongoing operations.

2.2. Hypothesis Development

2.2.1. Privatized Family Firms and Firm Innovation

There are a few studies focusing on the impact of privatization on business innovation [38,39]. For example, Zulfiqar et al. (2021) [39] recently analyzed the effect of the birth mode of family firms on R&D investment. When individuals or families have control over the SOEs through privatization, the new owners will follow the market logic, focus on efficiency, and maximize shareholder interests. However, as the impact of the state ownership imprints is sustained, privatized family firms are still influenced by political support and state interest [1]. From the motivation perspective, non-privatized family firms and privatized family firms are very different.

It was reported that SOEs had relatively low production and operational efficiency and relatively weak innovation levels during the 1970s–1990s. There are several reasons for this finding. First, SOEs are owned by the state (all people) and become agents without principals [40]. Moreover, the supervision of SOE managers is relatively weak. Second, the main task of SOEs is to fulfill the production plans issued by the government and secure employment. Raw materials and products are distributed by the state, while the marketing function of enterprises is missing or weakened. Their competitive motivation is much lower than those of their market-oriented counterparts [41]. Third, SOE managers are used to the state-ownership ideology. On the one hand, the “iron rice bowl” (lifelong employment) imprint with expectations of stability and security may discourage radical innovation [42]. On the other hand, managers with more political connections tend to be bureaucrats rather than entrepreneurs [43]. These managers lack the motivation to strive for innovation because, in the absence of positive profit-sharing incentives, they cannot benefit from successful innovation. They may also lack the appropriate competencies or skills to manage the firm effectively [13]. Some SOE managers plan to serve in government in the future [44] and focus more on short-term benefits for this purpose, which in turn reduces investments in long-term innovative strategy.

Studies of institutional inertia emphasize the stickiness of norms and practices embedded in organizations [45]. In other words, cognitive factors (including political influence and the state-ownership ideology) formed by managers and employees during sensitive periods can persist [18]. In the case of privatized family firms, for instance, managers and employees retain the mindset of lifelong employment and expectations of stability. Moreover, although the organization is no longer under government control and regulation, the inherited political connections induce managers to consider the government’s preferences. In addition, the supplier relationships of privatized family firms are difficult to change. The business objectives have changed for privatized family firms, but the organizational culture, decision-making process, and behavior of managers and employees are still influenced by the inertia of the previous public ownership. These factors result in low levels of firm innovation.

In contrast, for non-privatized family firms, the profit motivation is rooted in the start-up stage. Their goal setting is to ensure profit maximization and value creation for the owners. Managers of non-privatized firms respond to intense market competition by introducing new products, services, business approaches, and strategies. Under family control, there is consistency between improving firm innovation and maximizing the interests of family shareholders, especially in the face of severe market competition [46]. This is because family executives have ownership and control of the firm, giving them specific power structures and incentives to supervise organizational decision-making processes [47]. Non-privatized family firms have additional advantages in terms of innovation efficiency through access to family resources and the use of their supervisory role [48]. Thus, non-privatized family firms can enhance the efficiency of innovation output and are relatively weakly influenced by political goals. These differences between privatized family firms and non-privatized family firms reflect the fact that the privatized family firms operate within a broader national normative order, while the non-privatized family firms are subject to a broader market normative order. Based on the above arguments, the following hypothesis is proposed in this study.

Hypothesis 1: 

Compared to non-privatized family firms, privatized family firms have lower levels of innovation.

2.2.2. The Weakening Effect of the Imprint from Intergenerational Succession

Based on the institutional logic theory, there could be multiple competing institutional logic in an organization, and actors of the organization tend to choose the logic they prefer [49]. When the imprinted institutional logic conflicts with other institutional logic, the influence of the ideological imprint begins to weaken [15].

Family business owners do not always follow the family-first logic [50]. In privatized family firms, many family business owners were once managers of SOEs, and they were influenced by the state ownership imprint. However, when families start to implement intergenerational succession arrangements, these owners tend to involve their successors in corporate management decisions [51]. The successors, who do not have the state ownership imprint, tend to follow the logic of maximizing the family’s interests [52]. In order to successfully take over the leadership, the successor increasingly exerts an influence in decision-making.

With a change in the original board structure and executive team of the family firm, intergenerational succession will lead to disruption of power and differences in the allocation of resources [53]. Family business successors will not only implement strategic changes [54], but they will also increase their investment in innovation in order to build their own competence and authority [55]. At the same time, based on the altruism of “parental love”, family business owners usually create “secret reserve” resources for their successors in order to stabilize their future performance and build competence authority [56]. This support further promotes innovation. Therefore, the influence of the state ownership imprint on the privatized business is weakened when the owners pay more attention to the influence of the successor. Based on the above arguments, the following hypothesis is proposed in this study.

Hypothesis 2: 

Family intergenerational succession weakens the negative relationship between privatized family firms and firm innovation.

2.2.3. The Reinforcement Effect of the Imprint from State Ownership

The state-owned shares of family firms reflect the indirect influence of the state-owned capital or government. It can weaken the family owners’ full authority in decision-making. State-owned capital represents an important political connection [29]. Although state-owned capital is a “commitment signal” from the government for family firms, it also emphasizes the increase in additional policy burdens, such as redundant staff, children’s education, and investment in some areas that the government needs but cannot return in the short term. Family firms with state-owned shares need to take the government’s objectives into account when making decisions [57]. In this situation, managers will focus less on formulating and implementing innovation strategies actively [58].

Non-privatized family firms, in comparison, are more willing to balance economic interests with social goals to adapt to the dynamics of market competition and changing rules, which in turn leads to higher engagement in innovative activities. Moreover, both the altruism and family supervision mechanisms increase the efficiency of non-privatized family firms in innovation [9]. Therefore, non-privatized family firms are free to pursue their cost-efficiency goals. As a result, the following hypothesis is proposed in this study.

Hypothesis 3: 

A high proportion of state-owned shares strengthens the negative relationship between privatized family firms and firm innovation.

3. Data and Methodology

3.1. Sample and Data

This study takes A-share listed firms from 2007 to 2020 as the research sample from the China Stock Market and Accounting Research (CSMAR) database. The year 2007 was chosen as the starting point of the research because China’s new accounting standards were implemented among listed companies on 1 January 2007; this step was intended to achieve convergence with the International Financial Report Standard. Firms that met the following conditions were considered to be family firms. First, the ultimate control right could be traced back to a natural person or a family. Second, a natural person or family had ultimate control over the listed firm. Third, the ultimate controller was the largest shareholder of the listed firm. The criteria for determining the ultimate control of a family firm are as follows: (1) the natural person or family has 20% or more of the voting rights through direct or indirect ownership of the listed firm; (2) on the premise that the natural person or family is the chairman or general manager of the listed firm, 10% is used as the critical control ratio; (3) if neither of these two conditions is met but the critical holding ratio is above 10%, the family or natural person is the first largest shareholder; at the same time, there is no second-largest shareholder above the 10% critical holding ratio [59].

The independent variable in this study is the privatized family firm. The initial property rights data of listed family firms are mainly based on the firm development history on the official websites and CSMAR’s family firm database. Basic information regarding the actual controller was taken from the Prospectus disclosed on the Juchao Information Website (http://www.cninfo.com.cn/) and then combined with information from the China Family Firm database. The dependent variable is firm innovation, which is obtained from the CSMAR corporate innovation database; other data were taken from CSMAR and China statistical yearbooks. Finally, the research excludes the sample of financial and special treatment firms, resulting in 9576 observations.

3.2. Measurements

3.2.1. Dependent Variable

Firm innovation (Innovation). Following Wang et al. (2021) [60], in order to obtain a holistic view of each firm’s innovative activities, this study constructed three measures: the total number of patents (Inno), the number of invention patents (exploratory innovation, Explor), and the number of design patents and utility patents (exploitative innovation, Exploi). The patent information was obtained from both CSMAR and the State Intellectual Property Office (SIPO). Furthermore, to ensure the reliability of the research results, this study followed Luo et al. (2022) [61] and took the number of authorized patents as the dependent variable (firm innovation) for the robustness test.

3.2.2. Independent Variable

Privatized family firms (PFFs). The privatized family firm [62,63] is mainly defined according to the nature of the initial ownership of family firms. A value of 0 is assigned if a family firm is initially founded by the controlling family, and a value of 1 is assigned if the business is privatized from an SOE or COE [64].

3.2.3. Moderating Variables

Intergenerational succession (Succession) reflects the degree of involvement of family descendants in business management and the family’s intention to inherit and control the business in the long term [65,66]. For the measurement of cross-generational involvement in family firms, the second generation of the immediate family is assigned a value of 1. In contrast, the absence of an immediate family member is assigned a value of 0.

Proportion of state-owned shares (State_ratio). Following Jia et al. (2018) [67], the variable is measured as the proportion of state-owned shares in the total shares in the equity structure disclosed in the annual reports of listed family firms.

3.2.4. Control Variables

In order to accurately analyze the relationship between privatized family firms and firm innovation, according to previous research, the following variables were controlled at both the firm level and the manager level (

Table 1. Control Variables.

Name	Definition
Firm age	A firm’s age is represented by “Firm age” and is defined as the natural logarithm of the number of years that the firm has been established.
Firm size	A firm’s size is represented by “Firm size” and is defined as the logarithm of the company’s general assets.
Board	Board size is represented by “Board” and is defined as the logarithm of the total number of directors on the board.
Inde	Independent directors are represented by “Inde”, which is defined as the number of independent directors divided by the total number of directors.
Duality	A CEO with both positions is represented by “Duality”, which is set to 1 if the CEO is also the chairman; otherwise, it is set to 0.
Holder1	Shareholding concentration is represented by “Holder1” and is defined as the ratio of the number of shares held by the largest shareholder to the total number of shares.
MaleRatio	The proportion of male executives is represented by “MaleRatio” and is defined as the number of male members of the top management team.
MngmFB	Management financial background is represented by “MngmFB” and is defined as the number of members of the top management team with a financial background.
Insinv	Institutional investor holdings are represented by “Insinv”, and it is defined as the ratio of the number of shares held by institutional investors to the total number of shares.
Lev	Financial leverage is represented by “Lev”, defined as total liability divided by total assets.
Growth	The operating revenue growth rate is represented by “Growth”, defined as the ratio of year-end operating income growth to the operating revenue of the firm.
Cash	Firm cash holding levels are represented by “Cash” and are defined as the ratio of the firm’s free cash flow to total assets at the end of the period.
ROA	Profitability is represented by “ROA” and is used as a proxy variable of the firm’s performance; it is defined as net profit divided by total assets.
Tangibility	The firm’s fixed-asset ratio is represented by “Tangibility” and is defined as the ratio of the firm’s total fixed assets to the firm’s total assets.
Year	dummy variables
Industry	dummy variables

): firm age, firm size, board size, ratio of independent directors, CEO with both positions, shareholding concentration, proportion of male executives, finance background of the management, institutional investor holdings, financial leverage, operating revenue growth rate, firm cash holding levels, profitability, firm fixed-asset ratio, year, and industry (Min et al., 2016).

3.3. Model

To test Hypothesis 1, this study proposes the following regression model:

$${\mathit{Innovation}}_{i,t+1}={\alpha }_0+{\alpha }_1\times {\mathit{PFFs}}_{i,t}+{\alpha }_k\sum {\mathit{Control}}_{i,t}+\mathit{Ind}+\mathit{Year}+{\epsilon }_{i,t}$$

(1)

Innovation requires a long-term investment and has a lagging effect. The innovation lags by one year, which reduces the possibility of reverse causality [64].

Further, for Hypothesis 2 and Hypothesis 3, this study proposes regression models to test the moderating effect. The following regression models were established by referring to the model settings of Luo et al. (2022) [61]:

$$\begin{gathered} {\mathit{Innovation}}_{i,t+1}={\beta }_0+{\beta }_1{\mathit{PFFs}}_{i,t}+{\beta }_2\times {\mathit{Succession}}_{i,t}+{\beta }_3{\mathit{PFFs}}_{i,t}\times {\mathit{Succession}}_{i,t} \\ {+\beta }_k\sum {\mathit{Control}}_{i,t}+\mathit{Ind}+\mathit{Year}+{\epsilon }_{i,t} \end{gathered}$$

(2)

$$\begin{gathered} {\mathit{Innovation}}_{i,t+1}={\beta }_0+{\beta }_1{\mathit{PFFs}}_{i,t}+{\beta }_2\times {\mathit{\text{State\_ratio}}}_{i,t}+{\beta }_3{\mathit{PFFs}}_{i,t}\times {\mathit{\text{State\_ratio}}}_{i,t} \\ {+\beta }_k\sum {\mathit{Control}}_{i,t}+\mathit{Ind}+\mathit{Year}+{\epsilon }_{i,t} \end{gathered}$$

(3)

The subscript i represents the i-listed firm, and t represents the year. Referring to the classical literature [52], a fixed-effect model was selected for regression, controlling for year and industry effects separately. $\sum \mathit{Control}$ denotes the summation of the model’s control variables, including the corporate-governance-level variables and financial variables, and ${\epsilon }_{i,t}$ denotes the random error term.

4. Results

4.1. Descriptive Statistics

Table 2. Correlations, Means and Standard Deviations.

Variables	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20
1. Inno	1.263	1.482	1
2. Explor	0.778	1.096	0.861	1
3. Exploi	0.944	1.329	0.919	0.659	1
4. PFFs	0.112	0.316	−0.160	−0.137	−0.135	1
5. Succession	0.558	0.497	0.004	−0.023	0.015	0.014	1
6. State_ratio	0.007	0.034	0.010	0.016	0.008	0.025	−0.018	1
7. Firm size	21.678	1.022	0.033	0.073	0.039	0.062	0.062	−0.019	1
8. Firm age	2.812	0.354	−0.246	−0.181	−0.221	0.224	0.037	−0.062	0.213	1
9. Board	2.198	0.165	0.086	0.077	0.069	−0.039	0.113	0.035	0.131	−0.054	1
10. Duality	0.388	0.487	0.018	0.018	0.016	−0.101	−0.133	0.009	−0.110	−0.028	−0.120	1
11. Inde	0.378	0.052	−0.044	−0.046	−0.031	0.013	−0.103	−0.028	−0.058	0.029	−0.631	0.126	1
12. Holder1	0.376	0.143	0.100	0.065	0.100	−0.112	0.045	−0.024	−0.071	−0.112	−0.065	0.067	0.083	1
13. MaleRatio	0.790	0.117	0.129	0.119	0.110	0.032	0.020	0.020	0.067	−0.088	0.138	−0.044	−0.088	−0.066	1
14.MngmFB	0.657	0.475	−0.016	−0.020	−0.002	−0.010	−0.047	0.008	0.067	0.004	0.048	0.018	−0.016	−0.017	0.009	1
15. InsInv	0.354	0.256	0.014	0.016	0.023	0.184	0.082	0.083	0.245	0.007	0.090	−0.083	−0.078	0.189	0.050	0.062	1
16. Lev	0.374	0.199	−0.073	−0.051	−0.039	0.237	−0.013	0.001	0.370	0.139	0.019	−0.092	0.005	−0.148	0.095	0.050	0.108	1
17. Growth	0.270	0.470	0.074	0.054	0.065	−0.089	−0.070	0.138	−0.055	−0.215	0.003	0.066	0	0.091	0.012	0.017	0.025	−0.159	1
18. Cash	1.103	1.886	0.033	0.016	0.012	−0.094	−0.033	0.041	−0.267	−0.181	0.004	0.070	−0.014	0.137	−0.061	−0.015	−0.049	−0.534	0.317	1
19. ROA	0.043	0.075	0.109	0.104	0.080	−0.066	0.033	0.015	−0.003	−0.092	0.072	0.008	−0.030	0.206	−0.001	−0.026	0.105	−0.356	0.249	0.172	1
20. Tangibility	0.201	0.128	0.016	0.012	0.006	0.065	0.122	−0.01	0.068	0.019	0.070	−0.037	−0.042	0.001	0.077	−0.056	0.083	0.102	−0.218	−0.209	−0.064	1

Notes: N = 9576.

provides the descriptive statistics and correlation coefficients of all variables. The mean value of the total number of patents is 1.263, and the standard deviation is 1.482, indicating that most firms in this study engaged in innovation activities during the sample period. However, the innovation levels of these firms are significantly different. The mean and the standard deviation values of PFFs are 0.112 and 0.316, respectively, indicating that the number of privatized family firms is low. The mean value of intergenerational succession is 0.558, and the standard deviation is 0.497. The mean value of State_ratio is 0.007, and the standard deviation is 0.034. The correlation coefficient of PFFs and firm innovation is −0.160, and the result shows a significant negative correlation at the level of 1%, suggesting that privatized family firms have an inhibiting effect on innovation. Thus, Hypothesis 1 is primarily verified.

4.2. Multiple Regression Analysis

4.2.1. Examining the Relationship between Privatized Family Firms and Firm Innovation

Table 3. Fixed-Effect Regression for the Firm Innovation.

Variables	Model 1	Model 2	Model 3
	Inno	Explor	Exploi
PFFs	−0.625 ***	−0.403 ***	−0.480 ***
	(−14.416)	(−12.501)	(−12.365)
Firm size	0.203 ***	0.166 ***	0.162 ***
	(13.370)	(13.872)	(11.211)
Firm age	−0.388 ***	−0.226 ***	−0.343 ***
	(−8.678)	(−6.488)	(−7.787)
Board	0.341 ***	0.173 **	0.241 ***
	(3.430)	(2.125)	(2.583)
Duality	0.087 ***	0.080 ***	0.069 ***
	(3.402)	(3.910)	(2.814)
Inde	0.273	−0.044	0.356
	(0.883)	(−0.174)	(1.213)
Holder1	0.377 ***	0.116	0.434 ***
	(4.068)	(1.546)	(4.948)
MaleRatio	0.460 ***	0.454 ***	0.228 **
	(4.368)	(5.576)	(2.284)
MngmFB	−0.004	−0.018	0.036
	(−0.141)	(−0.856)	(1.464)
InsInv	−0.142 ***	−0.088 **	−0.092 *
	(−2.728)	(−2.089)	(−1.831)
Lev	−0.188 **	−0.068	−0.052
	(−2.259)	(−1.057)	(−0.685)
Growth	−0.051 *	−0.023	−0.050*
	(−1.770)	(−1.006)	(−1.853)
Cash	−0.027 ***	−0.010	−0.034 ***
	(−3.288)	(−1.629)	(−4.413)
ROA	0.865 ***	0.848 ***	0.460 ***
	(5.149)	(6.456)	(3.083)
Tangibility	−0.439 ***	−0.347 ***	−0.499 ***
	(−4.183)	(−4.165)	(−5.144)
Constant	−3.572 ***	−3.097 ***	−2.748 ***
	(−8.014)	(−8.543)	(−6.464)
Industry	Yes	Yes	Yes
Year	Yes	Yes	Yes
N	9576	9576	9576
Adj−R2	0.3843	0.2812	0.2942

Notes: *** correlation is significant at 1%, ** correlation is significant at 5%, * correlation is significant at 10%. t-statistics in parentheses.

(Model 1) provides the regression results of the privatized family firms on firm innovation. The coefficient of PFFs and firm innovation is −0.625, and it is significant at the level of 1%, suggesting that privatized family firms have a significant negative impact on firm innovation. In other words, the imprinting of state ownership can incline privatized family firms toward a lower level of firm innovation. In addition, the results of the control variables suggest that the coefficients of Size and Duality are both significant and positive, indicating that larger firms where the CEO’s position has more power generally have higher innovation levels. The coefficient of Holder1 exhibits both significant and positive results, indicating that a greater number of shares being held by the largest shareholder is conducive to corporate innovation activities. Lev has a significant negative correlation with Inno at the 1% level, suggesting that when firms can fully utilize financial leverage for financing, their investment in innovation will increase. Cash has a significant negative correlation with Inno at the 1% level, indicating that a higher cash-asset ratio does not necessarily contribute to a higher innovation level, and an excessively high cash-asset ratio will restrict the innovation of firms. Therefore, firms should make full and reasonable use of liquid assets and improve the capital utilization rate. Moreover, considering both the exploratory (Model 2) and exploitative (Model 3) innovation types, the relationship between privatized family firms and innovation is still significantly negative (coef. = −0.403, p < 0.01; coef. = −0.480, p < 0.01). Therefore, these empirical results are consistent with the results of existing studies, which validates Hypothesis 1.

4.2.2. The Moderating Effect of Intergenerational Succession and the Proportion of State-Owned Shares

Table 4. The Moderating Effect of Intergenerational Succession and the Proportion of State-owned Shares.

Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
	Inno	Explor	Exploi	Inno	Explor	Exploi
PFFs	−0.635 ***	−0.411 ***	−0.488 ***	−0.697 ***	−0.422 ***	−0.542 ***
	(−14.646)	(−12.758)	(−12.588)	(−15.749)	(−13.049)	(−13.705)
Succession	−0.094 ***	−0.122 ***	−0.044 *
	(−3.814)	(−6.140)	(−1.849)
PFFs×Succession	0.181 ***	0.125 **	0.157 ***
	(2.677)	(2.520)	(2.642)
State_ratio				−0.020	0.383	−0.158
				(−0.055)	(1.328)	(−0.455)
PFFs×State_ratio				−0.822 ***	−0.214	−0.707 ***
				(−4.108)	(−1.481)	(−4.100)
Firm size	0.206 ***	0.170 ***	0.164 ***	0.201 ***	0.166 ***	0.161 ***
	(13.600)	(14.192)	(11.342)	(13.301)	(13.867)	(11.134)
Firm age	−0.374 ***	−0.211 ***	−0.334 ***	−0.384 ***	−0.225 ***	−0.340 ***
	(−8.335)	(−6.074)	(−7.550)	(−8.601)	(−6.443)	(−7.724)
Board	0.354 ***	0.191 **	0.246 ***	0.337 ***	0.171 **	0.238 **
	(3.567)	(2.356)	(2.634)	(3.400)	(2.101)	(2.554)
Duality	0.076 ***	0.067 ***	0.063 **	0.085 ***	0.079 ***	0.067 ***
	(2.956)	(3.239)	(2.574)	(3.307)	(3.863)	(2.735)
Inde	0.250	−0.081	0.351	0.271	−0.042	0.354
	(0.811)	(−0.325)	(1.198)	(0.878)	(−0.168)	(1.205)
Holder1	0.388 ***	0.133 *	0.438 ***	0.363 ***	0.118	0.421 ***
	(4.188)	(1.763)	(4.990)	(3.910)	(1.563)	(4.776)
MaleRatio	0.453 ***	0.447 ***	0.224 **	0.472 ***	0.457 ***	0.238 **
	(4.308)	(5.492)	(2.246)	(4.481)	(5.618)	(2.385)
MngmFB	−0.008	−0.023	0.033	−0.004	−0.018	0.036
	(−0.322)	(−1.119)	(1.362)	(−0.142)	(−0.870)	(1.469)
InsInv	−0.137 ***	−0.081 *	−0.091 *	−0.129 **	−0.089 **	−0.080
	(−2.644)	(−1.939)	(−1.798)	(−2.466)	(−2.099)	(−1.571)
Lev	−0.208 **	−0.091	−0.064	−0.169 **	−0.062	−0.037
	(−2.512)	(−1.427)	(−0.841)	(−2.039)	(−0.964)	(−0.485)
Growth	−0.055*	−0.029	−0.051 *	−0.051*	−0.027	−0.048 *
	(−1.912)	(−1.277)	(−1.899)	(−1.771)	(−1.160)	(−1.801)
Cash	−0.028 ***	−0.011*	−0.035 ***	−0.026 ***	−0.010	−0.034 ***
	(−3.409)	(−1.813)	(−4.475)	(−3.251)	(−1.613)	(−4.382)
ROA	0.854 ***	0.842 ***	0.450 ***	0.864 ***	0.853 ***	0.458 ***
	(5.101)	(6.433)	(3.020)	(5.141)	(6.493)	(3.067)
Tangibility	−0.416 ***	−0.316 ***	−0.489 ***	−0.445 ***	−0.348 ***	−0.504 ***
	(−3.942)	(−3.781)	(−5.014)	(−4.248)	(−4.186)	(−5.205)
Constant	−3.747 ***	−3.261 ***	−2.861 ***	−3.655 ***	−3.157 ***	−2.809 ***
	(−8.397)	(−8.987)	(−6.709)	(−8.217)	(−8.706)	(−6.610)
Industry/Year	Yes	Yes	Yes	Yes	Yes	Yes
N	9576	9576	9576	9576	9576	9576
Adj−R2	0.385	0.284	0.295	0.385	0.281	0.295

Notes: *** correlation is significant at 1%, ** correlation is significant at 5%, * correlation is significant at 10%. t-statistics in parentheses.

reports the regression results of the moderating effect of intergenerational succession and the proportion of state-owned shares. Model 1 introduces the first-order moderation term of intergenerational succession (PFFs×Succession) to test whether intergenerational succession has a significant moderation effect. The regression results show that the coefficients of the moderating terms of PFFs and Succession are significantly positive, with significant results at the 1% level (coef. = 0.181, p < 0.01), verifying that intergenerational succession weakens the negative impact of privatized family firms on firm innovation. In fact, this moderating effect remains robust when considering different types of patent variables, thus providing strong support for Hypothesis 2. Similarly, the results of Model 4 show that the coefficient of PFFs×State_ratio is significantly negative (coef. = −0.822, p < 0.01), indicating that the proportion of state-owned shares strengthens the negative impact of privatized family firms on firm innovation, which supports Hypothesis 3. It is worth noting that different patent types lead to different results. The results of Model 6 indicate that the coefficient of PFFs×State_ratio is −0.707, which is significant at the 1% level, but the results of Model 5 indicate that the coefficient of the interaction term PFFs×State_ratio is −0.214, which is not significant. These results suggest that a higher proportion of state-owned shares has no significant moderating effect on the relationship between the privatization of family firms and exploratory innovation, while it has a significant effect on exploitative innovation, further supporting Hypothesis 3.

4.3. Robustness Tests

4.3.1. Robustness Tests for Endogeneity Problems

In this study, the Heckman two-stage regression model was used to test the robustness of the endogeneity problem due to the sample selection bias [15]. In the first stage of regression analysis, the natural logarithm of the provincial GDP per capita in the year 2000 (GDP2000) was selected as an instrumental variable. This is because the year 2000 was around the peak period of the privatization of SOEs, and the government’s policy to “invigorate large enterprises while relaxing control over small ones” (zhuada fangxiao) became the guiding strategy for SOE reform.

Table 5. Robustness Tests for Endogeneity Problems: Heckman Two-stage Regression Model Result.

Variables	Model 1	Model 2	Model 3	Model 4
	First-Stage	Second-Stage
	PFFs	Inno	Explor	Exploi
PFFs		−0.884 **	−0.887 ***	−0.266
		(−2.302)	(−2.860)	(−0.717)
GDP 2000	−0.091 ***
	(−15.329)
IMR		−1.216 ***	−0.722 ***	−1.058 ***
		(−17.727)	(−13.008)	(−15.949)
Firm size	−0.038 ***	0.142 ***	0.143 ***	0.120 ***
	(−11.260)	(6.597)	(8.182)	(5.767)
Firm age	0.195 ***	−0.080	0.059	−0.196 **
	(22.088)	(−0.948)	(0.872)	(−2.406)
Board	−0.124 ***	0.126	−0.029	0.114
	(−5.252)	(0.958)	(−0.272)	(0.898)
Duality	−0.043 ***	0.047	0.058 *	0.044
	(−6.917)	(1.259)	(1.911)	(1.205)
Inde	0.069	0.695 *	0.186	0.743 **
	(0.932)	(1.790)	(0.591)	(1.980)
Holder1	−0.244 ***	−0.024	−0.254 **	0.245 *
	(−10.974)	(−0.165)	(−2.169)	(1.751)
MaleRatio	0.048 *	0.307 **	0.451 ***	0.024
	(1.829)	(1.964)	(3.569)	(0.161)
MngmFB	−0.014 **	0.026	−0.007	0.072 **
	(−2.221)	(0.745)	(−0.267)	(2.168)
InsInv	0.245 ***	0.214 **	0.177**	0.152
	(19.757)	(2.090)	(2.137)	(1.538)
Lev	0.446 ***	0.687 ***	0.549 ***	0.604 ***
	(22.332)	(3.616)	(3.573)	(3.290)
Growth	−0.023 ***	−0.095 ***	−0.064 **	−0.074 **
	(−3.335)	(−2.705)	(−2.247)	(−2.201)
Cash	0.016 ***	0.006	0.012	−0.008
	(8.246)	(0.594)	(1.463)	(−0.756)
ROA	0.239 ***	1.452 ***	1.558 ***	0.553 *
	(5.015)	(4.462)	(5.924)	(1.759)
Tangibility	0.052 **	−0.309 **	−0.201 *	−0.471 ***
	(2.096)	(−2.178)	(−1.752)	(−3.433)
Constant	1.286 ***	−0.884**	−0.887 ***	−0.266
	(11.315)	(−2.302)	(−2.860)	(−0.717)
Industry	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	7117	7117	7117	7117
Adj-R2		0.209	0.145	0.154

Notes: *** correlation is significant at 1%, ** correlation is significant at 5%, * correlation is significant at 10%. t-statistics in parentheses. The inverse mills ratio is abbreviated as IMR.

shows the regression results of the Heckman two-stage regression model. The first-stage regression results indicate that the coefficient of the variable GDP2000 is −0.091, which is significant at the 1% level, whereas there should not be an influential relationship between this variable and the level of firm innovation. The results of the second-stage regression indicate that the firm innovation level is still reduced after the privatization of family firms, and this result supports Hypothesis 1.

4.3.2. Robustness Tests for the Propensity Score Matching (PSM) Method

This study first conducted a Probit regression model on the factors that may affect the manner of obtaining initial property rights in family firms. This research identified non-privatized family firms with similar characteristics to privatized family firms as control samples according to the 1:1 matching principle. It compared the effects of differences in initial property rights between the two groups of samples on different pairs (i.e., privatized and non-privatized) of firms’ innovation levels. After referring to existing studies, this research selected several relevant matching variables at the firm level and the macro level as independent variables of the regression model, mainly including control variables in the main hypothesis, such as firm size, profitability, growth, and firm operating years.

Table 6. Robustness Test: Propensity Score Matching (PSM) Method Regression Results Analysis.

Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
	Inno	Explor	Exploi	Inno	Explor	Exploi
PFFs	−0.643 ***	−0.424 ***	−0.512 ***	−0.679 ***	−0.462 ***	−0.520 ***
	(−10.618)	(−9.288)	(−9.119)	(−10.949)	(−9.957)	(−9.160)
Succession	−0.108	−0.076	−0.115
	(−1.206)	(−1.112)	(−1.326)
PFFs×Succession	0.239 **	0.103	0.261 **
	(1.966)	(1.134)	(2.294)
State_ratio				−1.470 **	−0.589	−1.466 **
				(−2.238)	(−1.181)	(−2.530)
PFFs×State_ratio				−0.524 **	−0.457 **	−0.254
				(−2.157)	(−2.544)	(−1.255)
Firm size	0.281 ***	0.199 ***	0.248 ***	0.275 ***	0.194 ***	0.245 ***
	(9.793)	(9.320)	(9.172)	(9.601)	(9.127)	(9.037)
Firm age	−0.758 ***	−0.396 ***	−0.679 ***	−0.785 ***	−0.406 ***	−0.713 ***
	(−4.820)	(−3.476)	(−4.677)	(−4.987)	(−3.571)	(−4.880)
Board	0.640 ***	0.443 ***	0.445 **	0.636 ***	0.438 ***	0.442 **
	(3.208)	(2.846)	(2.433)	(3.190)	(2.814)	(2.414)
Duality	0.009	0.040	0.016	0.038	0.068	0.030
	(0.134)	(0.806)	(0.255)	(0.544)	(1.306)	(0.467)
Inde	1.867 ***	1.190 **	0.959	1.887 ***	1.221 **	0.985
	(2.734)	(2.172)	(1.540)	(2.762)	(2.239)	(1.569)
Holder1	−0.207	−0.490 **	0.083	−0.273	−0.528 **	0.023
	(−0.792)	(−2.372)	(0.343)	(−1.032)	(−2.541)	(0.092)
MaleRatio	0.544 **	0.543 ***	0.157	0.530 **	0.545 ***	0.134
	(2.134)	(2.872)	(0.672)	(2.079)	(2.889)	(0.571)
MngmFB	−0.133 **	−0.141 ***	−0.062	−0.125 *	−0.136 ***	−0.055
	(−2.011)	(−2.815)	(−1.020)	(−1.887)	(−2.724)	(−0.903)
InsInv	0.043	0.262 *	−0.156	0.131	0.311**	−0.085
	(0.248)	(1.905)	(−0.978)	(0.736)	(2.208)	(−0.517)
Lev	−0.631 ***	−0.457 ***	−0.330 ***	−0.596 ***	−0.432 ***	−0.305 **
	(−4.644)	(−4.803)	(−2.709)	(−4.383)	(−4.556)	(−2.493)
Growth	−0.139 **	−0.080 *	−0.132 ***	−0.126 **	−0.071	−0.122 **
	(−2.361)	(−1.782)	(−2.677)	(−2.137)	(−1.569)	(−2.471)
Cash	−0.042 **	−0.013	−0.041 **	−0.040 **	−0.013	−0.040 **
	(−2.107)	(−0.846)	(−2.392)	(−2.083)	(−0.874)	(−2.314)
ROA	0.249	0.327	0.157	0.225	0.309	0.137
	(0.854)	(1.560)	(0.608)	(0.770)	(1.475)	(0.530)
Tangibility	−0.110	−0.024	−0.223	−0.110	−0.034	−0.217
	(−0.520)	(−0.152)	(−1.151)	(−0.524)	(−0.216)	(−1.126)
Constant	−5.023 ***	−4.048 ***	−3.830 ***	−4.853 ***	−3.970 ***	−3.658 ***
	(−5.212)	(−5.495)	(−4.297)	(−5.065)	(−5.413)	(−4.121)
Industry/Year	Yes	Yes	Yes	Yes	Yes	Yes
N	1876	1876	1876	1876	1876	1876
Adj-R2	0.303	0.248	0.241	0.305	0.249	0.241

Notes: *** correlation is significant at 1%. ** correlation is significant at 5%. * correlation is significant at 10%. t-statistics in parentheses.

shows the results of the regression analysis after matching the PSM scores. It can be seen from the table that the coefficient of PFFs in column (1) is −0.643, which is significant at the 1% level, while the coefficient of PFFs×Succession is 0.239, which is significant at the 5% level. The coefficient of PFFs×State_ratio is −0.524, which is significant at the 5% level. These results are highly consistent with the above hypotheses.

4.3.3. Change in the Measurement Method of Firm Innovation

This study used the logarithm of the sum of the number of authorized patents plus 1 to remeasure firm innovation [59]. The logarithm of the sum of the authorized invention patents plus 1 was used to measure exploratory innovation, and the logarithm of the sum of the authorized design and utility patents plus 1 was used to measure exploitative innovation. The regression results are shown in

Table 7. Robustness Test: Substitute Variable for Firm Innovation.

Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
	Au_Inno	Au_Explor	Au_Exploi	Au_Inno	Au_Explor	Au_Exploi
PFFs	−0.697 ***	−0.363 ***	−0.567 ***	−0.760 ***	−0.350 ***	−0.635 ***
	(−16.363)	(−14.346)	(−13.735)	(−16.919)	(−13.592)	(−14.464)
Succession	−0.088 ***	−0.126 ***	−0.053 *
	(−3.246)	(−7.077)	(−1.912)
PFFs×Succession	0.187 ***	0.178 ***	0.132 **
	(2.835)	(4.644)	(2.080)
State_ratio				0.086	0.062	0.005
				(0.229)	(0.276)	(0.013)
PFFs×State_ratio				−0.824 ***	0.030	−0.845 ***
				(−3.962)	(0.275)	(−4.194)
Firm size	0.154 ***	0.143 ***	0.117 ***	0.149 ***	0.138 ***	0.113 ***
	(9.203)	(12.703)	(6.905)	(8.902)	(12.292)	(6.692)
Firm age	−0.488 ***	−0.173 ***	−0.461 ***	−0.498 ***	−0.190 ***	−0.467 ***
	(−10.445)	(−5.958)	(−9.598)	(−10.711)	(−6.511)	(−9.758)
Board	0.414 ***	0.212 ***	0.331 ***	0.398 ***	0.194 ***	0.320 ***
	(3.903)	(3.023)	(3.056)	(3.764)	(2.759)	(2.965)
Duality	−0.029	−0.035 *	−0.009	−0.021	−0.020	−0.005
	(−1.013)	(−1.912)	(−0.308)	(−0.735)	(−1.101)	(−0.175)
Inde	0.164	−0.050	0.302	0.183	−0.014	0.311
	(0.491)	(−0.225)	(0.893)	(0.546)	(−0.064)	(0.918)
Holder1	0.519 ***	−0.058	0.642 ***	0.496 ***	−0.073	0.622 ***
	(5.127)	(−0.878)	(6.230)	(4.895)	(−1.099)	(6.027)
MaleRatio	0.484 ***	0.261 ***	0.418 ***	0.503 ***	0.269 ***	0.435 ***
	(4.094)	(3.495)	(3.477)	(4.248)	(3.586)	(3.612)
MngmFB	−0.031	−0.026	−0.004	−0.027	−0.020	−0.001
	(−1.106)	(−1.403)	(−0.141)	(−0.951)	(−1.089)	(−0.046)
InsInv	−0.123 **	−0.028	−0.065	−0.116 **	−0.036	−0.054
	(−2.148)	(−0.747)	(−1.100)	(−2.000)	(−0.949)	(−0.915)
Lev	0.057	−0.009	0.195 **	0.095	0.016	0.226 **
	(0.643)	(−0.162)	(2.201)	(1.069)	(0.287)	(2.541)
Growth	−0.027	−0.026	−0.012	−0.024	−0.021	−0.011
	(−0.923)	(−1.388)	(−0.428)	(−0.831)	(−1.101)	(−0.369)
Cash	−0.034 ***	−0.004	−0.040 ***	−0.033 ***	−0.003	−0.039 ***
	(−4.444)	(−0.764)	(−4.952)	(−4.277)	(−0.522)	(−4.843)
ROA	1.323 ***	0.603 ***	1.229 ***	1.335 ***	0.613 ***	1.237 ***
	(6.504)	(4.613)	(6.186)	(6.570)	(4.686)	(6.237)
Tangibility	−0.545 ***	−0.295 ***	−0.571 ***	−0.574 ***	−0.327 ***	−0.591 ***
	(−4.796)	(−4.090)	(−5.054)	(−5.070)	(−4.521)	(−5.257)
Constant	−3.159 ***	−3.069 ***	−2.451 ***	−3.075 ***	−2.944 ***	−2.401 ***
	(−6.667)	(−9.427)	(−5.085)	(−6.516)	(−9.050)	(−5.001)
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
N	9571	9571	9571	9571	9571	9571
Adj−R2	0.203	0.132	0.160	0.202	0.127	0.160

Notes: *** correlation is significant at 1%, ** correlation is significant at 5%, * correlation is significant at 10%. t-statistics in parentheses.

, from which it can be seen that there is no significant difference between the regression results and those in previous sections, indicating that the findings of this study are robust.

4.4. Further Research: Analysis of Privatization Time Trends

According to the imprinting theory, with the development of privatization, the impact of the state ownership imprint on the innovation level of family firms will be weakened. To verify the temporal trend of the state ownership imprint of privatized family firms, this study examined the impact of the length of time from the year of privatization (PFFs_time) of family firms on the level of innovation. The results in

Table 8. Further Research: Analysis of Privatization Time Trends.

Variables	Model 1	Model 2	Model 3
	Inno	Explor	Exploi
PFFs_time	0.028 ***	0.022 ***	0.017 *
	(2.702)	(2.895)	(1.857)
Firm size	0.266 ***	0.148 ***	0.244 ***
	(7.919)	(6.424)	(7.826)
Firm age	−0.917 ***	−0.476 ***	−0.850 ***
	(−4.424)	(−3.141)	(−4.639)
Board	0.186	0.094	−0.009
	(0.894)	(0.610)	(−0.050)
Duality	−0.013	−0.017	0.013
	(−0.178)	(−0.319)	(0.197)
Inde	2.344 ***	1.737 ***	0.647
	(2.907)	(2.697)	(0.923)
Holder1	−0.893 ***	−0.945 ***	−0.474 *
	(−2.830)	(−3.787)	(−1.757)
MaleRatio	0.434	0.331	0.067
	(1.511)	(1.591)	(0.266)
MngmFB	−0.141 *	−0.158 ***	−0.082
	(−1.816)	(−2.750)	(−1.186)
InsInv	−0.003	0.353 **	−0.299
	(−0.013)	(2.111)	(−1.613)
Lev	−0.557 ***	−0.416 ***	−0.337 ***
	(−3.845)	(−4.351)	(−2.589)
Growth	−0.171 ***	−0.089 **	−0.152 ***
	(−3.170)	(−2.132)	(−3.562)
Cash	−0.035 *	−0.031 **	−0.012
	(−1.762)	(−2.015)	(−0.774)
ROA	−0.183	−0.023	−0.268
	(−0.551)	(−0.102)	(−0.896)
Tangibility	0.297	0.347 *	0.015
	(1.134)	(1.859)	(0.062)
Constant	−3.654 ***	−2.333 ***	−2.226 **
	(−3.172)	(−2.710)	(−2.214)
Industry	Yes	Yes	Yes
Year	Yes	Yes	Yes
N	1131	1131	1131
Adj−R2	0.221	0.177	0.178

Notes: *** correlation is significant at 1%, ** correlation is significant at 5%, * correlation is significant at 10%. t-statistics in parentheses.

show that there is a significant positive relationship between the time period of privatization and the innovation of family firms (coef. = 0.028, p < 0.01; coef. = 0.022, p < 0.01; coef. = 0.017, p < 0.1). Therefore, the results indicate that the state ownership imprints of privatized family firms fade as the time period of privatization increases, and there is a gradual improvement in the innovation capacity of these firms.

5. Discussion

5.1. Conclusions

This research delves into the enduring impacts of state ownership on the strategic decisions made by family-owned businesses. In rapidly progressing societies, notably China, the study has discovered that family firms that have undergone privatization exhibit a diminished capacity for innovation compared to those that remain non-privatized. The data suggests that state ownership leaves a lasting institutional mark on these firms, and its effects persist over time. This observation aligns with the conclusions drawn by Ruxi et al. in 2022. Furthermore, the insights from this study offer an explanatory framework for understanding the middle-income stagnation experienced by Latin American nations, the Philippines, and several other economies during the 1980s.

Furthermore, several researchers have contended that “endurance doesn’t equate to eternity” [18], suggesting that imprints might not be indefinite. However, the dynamics of how these imprints transform remain largely unexplored. Studies have shown that shares owned by the state intensify the detrimental impact of privatization on the innovation of family businesses. Conversely, the act of passing down the business through generations tends to diminish this effect. The influence of imprinting can persist, provided there are external circumstances and internal motivations that support its sustenance. However, the inherent values and priorities of a family often act as a counterbalance. Specifically, when a family places its own benefits at the forefront during generational handovers, it tends to counteract the enduring effects of state ownership.

5.2. Theoretical Contributions and Practical Implications

5.2.1. Theoretical Contributions

This study delves into the innovation research of family businesses by contrasting the innovation behaviors of privatized versus non-privatized family firms. It highlights the imprinting effect of pre-privatization state ownership and offers a fresh perspective on the strategic behavior of family firms in historical contexts. The work builds on prior research, introducing an organizational-level imprint from the institutional logic of state ownership. The study emphasizes how privatized family firms, influenced by traditional state ownership, differ from non-privatized ones in management philosophies, with the former often aligning with broad social goals and political loyalty.

While most past research focused on the emergence of organizational imprints, this work examines the dynamics of their maintenance, particularly when the imprinter and imprinted entities evolve. It underscores the role of contextual factors in this dynamic, especially for privatized family firms. Challenges like the governance model and policies can hinder innovation. Yet, family successors may find alternative means for profit and innovation, bypassing state-owned capital. This research enriches the understanding of the interplay between intergenerational succession, state-owned shares, and imprint maintenance in family firms—a facet previously under-explored, thereby bolstering the imprinting theory.

5.2.2. Managerial Implications

The findings of this study provide the following suggestions for family business owners and managers. First, family firms, especially privatized ones, should foster innovation by reforming corporate governance and creating a tolerant culture towards mistakes. Leveraging the family’s unique resources, like long-term trust relationships with various stakeholders, can accelerate innovation by converting inputs to outputs and minimizing inefficiencies.

Second, it’s crucial for family firms to have a thorough succession strategy. Potential successors should undergo comprehensive training, starting at the grassroots level and gradually ascending the ranks. A continuous evaluation process should determine the most suitable successor. Given market uncertainties, relying solely on family members might restrict innovation. To ensure sustained innovation, firms might consider hiring experienced professional managers, necessitating the creation of a decentralized and accountable organizational environment.

Third, the government should actively support family firms during privatization by accelerating the transformation of public governance. Family businesses and state-owned enterprises (SOEs) should receive equitable rights and incentives. Reducing policy burdens and limiting governmental interference in family firm operations while ensuring a level playing field is vital.

5.3. Limitations and Future Research

There were some limitations to this study, which need to be addressed in future studies. First, because the state-ownership system may lead to preferential treatment through factors such as government subsidies, preferential taxation, loans, endorsement, and network [68], this imprint can have a positive impact on privatized family firms. Therefore, future research can further distinguish the different effects of the state-ownership imprint on family-first logic. Second, as a creative process, innovation is closely related to industrial dynamics (such as sunrise and sunset industries) [69]. Privatization can also bring about further changes in the supply chain. Since SOEs in China belong to different types of industries and many of them are upstream suppliers, future research on the privatization of family businesses can focus on the heterogeneity of different industry types. The role of supply-chain SOEs in the relationship between the privatization of family firms and innovation is also worth analyzing. Considering the ubiquitous supply chain in the world, its investigation is crucial to understanding its role in the strategic behavior of family businesses [70]. Third, this study examined innovation performance from the perspective of innovation output. Future research may consider other measures of innovation, such as the research quotient [71], differentiation (vertical or horizontal) of new products, and the introduction of new products [72] in different domains.